SCMD 2017

STORMWATER CONVEYANCE
MODELING AND DESIGN

First Edition
Authors
Haestad Methods
S. Rocky Durrans
Managing Editor
Kristen Dietrich
Editors
David Klotz, Adam Strafaci, and Colleen Totz
Contributing Authors
Muneef Ahmad, Thomas E. Barnard,
Peder Hjorth, and Robert Pitt
Exton, Pennsylvania USA

STORMWATER CONVEYANCE MODELING AND DESIGN
First Edition
Copyright © 2007 by Bentley Institute Press
Bentley Systems, Incorporated.
685 Stockton Drive
Exton, Pennsylvania 19341
www.bentley.com
Copyright © 2003 by Haestad Press

Haestad Methods, Incorporated
27 Siemon Company Drive, Suite 200W
Watertown, Connecticut 06795
All rights reserved. Printed in the United States of America. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior
written permission of the publisher.
Graphic image reprinted courtesy of ESRI and is used herein with permission. Copyright © ESRI. All rights reserved. Graphic
image reprinted courtesy of Jim McKibben, CH2MHill, Inc. and ESRI and is used herein with permission.
Indexer: Beaver Wood Associates
Proofreading: Beaver Wood Associates
Special thanks to The New Yorker magazine for the cartoons throughout the book. © The New Yorker Collection from
cartoonbank.com. All rights reserved.
Page 5 – (2000) Liza Donnelly Page 324 – (1988) Warren Miller
Page 12 – (1986) Al Ross Page 365 – (1991) Mort Gerberg
Page 58 – (1990) J.B. Handelsman Page 376 – (2002) Lee Lorenz
Page 102 – (1963) Robert Day Page 400 – (2003) P.C. Vey
Page 131 – (1988) Peter Steiner Page 434 – (1987) Lee Lorenz
Page 140 – (2002) Robert Mankoff Page 481 – (1992) Henry Martin
Page 145 – (1979) William Steig Page 501 – (1988) J.B. Handelsman
Page 198 – (1988) Jack Ziegler Page 525 – (1988) Arnie Levin
Page 209 – (1975) Dean Victor Page 567 – (1996) Robert Day
Page 214 – (2003) Sidney Harris Page 581 – (2000) Jack Ziegler
Page 237 – (1993) Mick Stevens Page 596 – (1970) Vahan Shirvanian
Page 254 – (1990) Danny Shanahan Page 599 – (1970) James Stevenson
Page 280 – (2003) Mary Lawton Page 613 – (1971) Lee Lorenz
Page 288 – (1995) Peter Steiner Page 632 – (2000) David Sipress
Page 306 – (2003) Ted Goff
StormCAD, PondPack, CulvertMaster and FlowMaster are registered trademarks of Bentley Systems, Inc. AutoCAD is a registered
trademark of Autodesk, Inc. ArcMap and ArcView are registered trademarks of ESRI, Inc. All other trademarks, brands, company or
product names not owned by Bentley Systems or its subsidiaries are the property of their respective owners, who may or may not be
affiliated with, connected to, or sponsored by Haestad Methods or its subsidiaries.
Haestad Methods is a registered trade name of Bentley Systems, Inc.
Library of Congress Control Number: 2007925244

ISBN: 978-1-934493-00-7
Bentley Systems, Inc. Phone (US): 800-225-2613

685 Stockton Drive Phone (International): 203-805-1100
Exton, PA 19341, USA Internet: www.bentley.com/books
“Theory can leave questions
unanswered, but practice has
to come up with something.”
-Mason Cooley
To all of my parents: Lola de, Dennis,
Richard, Margaret, Harold, and
Charlotte; for your sacrifices,
tolerance, support, and
encouragement.
-S. Rocky Durrans
“This book is a unique blending of at least six books that I
currently use. . . . I wish it had been available when I took the
Professional Engineer’s Exam!”
Shane Parson, Ph.D., P.E.
Anderson & Associates, Inc.
“Haestad Press has once again published a book that is one of

the best I’ve seen. It will surely become a ‘first to reach for’
when performing stormwater conveyance design.”
Robert M. Steele
LPA Group
“Stormwater Conveyance and Design was written by persons

who have ‘been there and done that’ and who work closely
with design professionals. As such, there is much knowledge
and practical information in the book. While reviewing the
book, I’ve learned something new from every chapter—infor-
mation that will help me do my job better..”
Michael Meadows, Ph.D., P.E.
University of South Carolina
Acknowledgments
Stormwater Conveyance Modeling and Design is the culmination of the efforts of

many individuals, the most important of which are the thousands of stormwater engi-
neering professionals that have dedicated their careers to managing stormwater
and protecting the environment. Bentley Systems has been serving stormwater
engineers around the world for over 20 years and our discussions and meetings with
the commu-nity has provided the inspiration to write this book.
Bentley Institute Press is grateful to all of the talented people who contributed to the
creation of this book, in particular, S. Rocky Durrans, the primary author. Several
others con-tributed content in specific areas, including Robert Pitt (regulatory
practices, stormwater quality management, and U.S. stormwater regulations),
Muneef Ahmad (Canadian stormwater regulations), and Peder Hjorth
(international stormwater policies and practices). Tom Barnard made significant
contributions in organizing the content from these multiple sources and adding
additional material. Mr. Barnard also authored many of the exercises. Shane Parson
contributed material for a number of the sidebar topics found throughout the book,
and Bob Steele provided practical modeling and design examples.
Many engineers at Bentley Systems assisted with editing, reviewing, and adding con-
tent, working problems, and checking for accuracy, making this project a
company-wide effort. Their input played an enormous role in shaping the text into
its current form. Peter Martin led the effort to create and assemble the graphics used
throughout the book. Chris Dahlgren, Caleb Brownell, Cal Hurd, John Slate, and
several others contributed to the illustration efforts. Special thanks go to the
organizations that pro-vided us with reprint permissions for their graphics.
Others involved in the production process for the book include Lissa Jennings (pub-
lishing logistics and proofreading assistance), Rick Brainard and Jim O'Brien (cover
design), Ben Ewing (web page design), Jeanne Moody (index and proofreading), John
Shirley (formatting and page layout), Wes Cogswell (accompanying CD-ROM), and
Amanda Irwin and David Sutley (example problem unit conversions).
Several engineering professionals reviewed the book and provided us with insightful
feedback and suggestions for improvement. Huge thanks go to our peer reviewers:
Roger Kilgore, G.V. Loganathan, Mike Meadows, Shane Parson, and David Wall.
Other core members of the Bentley Institute Press team who put in many hours on
this project and provided a huge amount of support for this effort are David Klotz,
Adam Strafaci, Colleen Totz, and Tom Walski.
Special acknowledgments go to Niclas Ingemarsson, executive vice president, for
his management guidance, and Keith Hodsden, general manager of sales and
technical support, for his assistance in organizing staff resources. Finally, thank
you to John Haestad, whose vision it was to create this book, for his unwavering
support of this project.
Kristen Dietrich
Managing Editor
Authors and Contributing Authors
Stormwater Conveyance Modeling and Design represents a collaborative effort that

combines the experiences of ten contributors and peer reviewers and the
engineers and software developers at Bentley Systems. The authors and
contributing authors are:
Authors
Bentley Systems
Exton, Pennsylvania
S. Rocky Durrans
University of Alabama
Contributing Authors
Muneef Ahmad
GM Sernas & Associates
Thomas E. Barnard
Bentley Systems
Peder Hjorth
Lund University
Robert Pitt
University of Alabama
Bentley Systems
The Bently Systems Engineering Staff is an extremely diverse group of profession-
als from six continents with experience ranging from software development and
engineering consulting, to public works and academia. This broad cross
section of expertise contributes to the development of the most comprehensive
software and educational materials in the civil engineering industry. In addition
to the specific authors credited in this section, many at Bentley Systems
contributed to the success of this book.
S. Rocky Durrans, Ph.D., P.E.
S. Rocky Durrans, Ph.D., P.E., is a professor of Civil and Environmental Engineering
at the University of Alabama, where he is responsible for the hydrologic and hydrau-
lic engineering programs.
Prior to his academic career, Dr. Durrans spent many years in the consulting arena
where he gained extensive experience in the design of stormwater and major urban
drainage systems. He is regarded as an authority on flood and rainfall analysis and has
many publications to his credit.
Muneef Ahmad, P. Eng.

Muneef Ahmad, P. Eng., is a water resources engineer with the Sernas Group, Inc., a
civil engineering consulting firm based in Southern Ontario, Canada. The firm pro-
vides a broad range of services throughout Southern Ontario, including municipal,
water resources, land use planning, transportation planning, class environmental
assessments, and power and utility design. Project-related experience for Mr. Ahmad
includes work for public and private sector clients in hydrologic modeling, hydraulic
design, erosion remediation works, and stormwater management facility design.
Thomas E. Barnard, Ph.D., P.E.

Thomas E. Barnard, Ph.D., P.E., is a senior engineer with Haestad Methods. He has
more than 20 years of experience in environmental engineering working as a
consultant, researcher, educator, and author. He holds a B.S. in civil engineering from
the University of Vermont, an M.S. in environmental engineering from Utah State
University, and a Ph.D. in environmental engineering from Cornell University. His
expertise includes water and wastewater treatment, hazardous waste management,
surface water hydrology, and water quality monitoring systems. Dr. Barnard’s
experience includes hydrologic and water quality modeling of a delta lake in Egypt.
He is a registered professional engineer in Pennsylvania.
Peder Hjorth, Ph.D.

Dr. Peder Hjorth is a full-time faculty member in the Department of Water Resources
Engineering at the University of Lund, where he is responsible for teaching and con-
ducting research in hydrology, hydraulics, and water resources engineering. In his
more than 30 years at Lund University, he and his students have studied erosion prob-
lems, urban drainage, and integrated water resources management. He has also served
some 15 years as Course Director of an international training program on water
resources management in developing countries. In the early 1990s, Dr. Hjorth was a
faculty member at the Asian Institute of Technology in Bangkok, where he and his
students primarily studied problems related to hydropower, irrigation, and urban
water management. He received his M.S. degree in civil engineering from Chalmers
University of Technology in Gothenburg, and his Ph.D. in water resources engineer-
ing from Lund Institute of Technology/Lund University in Lund, Sweden. He has
served as chairman of the International Association for Hydraulic Research (IAHR)
Technical Section on Water Resources Management.
Robert Pitt, Ph.D., P.E.
Robert Pitt, Ph.D., P.E., is the Cudworth Professor of Urban Water Systems and
Director of Environmental Engineering Programs at the University of Alabama. He
currently teaches classes in urban water resources, stressing the integration of hydrol-
ogy and water quality issues. Prior to coming to Tuscaloosa two years ago, he taught
at the University of Alabama at Birmingham for 14 years and previously was a senior
engineer with industry and government for 16 years. Dr. Pitt obtained his Ph.D. from
the University of Wisconsin, Madison, his M.S. in civil engineering from California
State University, San Jose, and his B.S. from Humboldt State University in Arcata,
California. He has been active in stormwater research for most of his professional
career, having conducted research for the EPA, NSF, Environment Canada, and many
state and local agencies, and some industries. He has published numerous reports and
papers, including eight books. He is a Registered Engineer in Wisconsin and is a Dip-
lomate of the American Academy of Environmental Engineers, and a member of sev-
eral professional societies. In 2002, he received a Distinguished Service Award from
the University of Wisconsin recognizing his service to the engineering profession.
Foreword
Every day, stormwater management decisions are made concerning topics such as
land use, site planning, impervious cover, and infrastructure. These decisions, made
by developers, engineers, property owners, and government officials, are often made
on a micro scale even though cumulatively, they have macro-scale effects-not the least
of which is the cost of constructing and maintaining stormwater facilities. Stormwater
management hydraulic models are essential tools for designing and analyzing these
facilities, and through their appropriate application, one can ensure adequate flood
protection without wasting financial resources through over-design or poorly con-
ceived designs.
The common paradigm for stormwater management has always been the efficient
removal of stormwater. Design focused on curb and gutter, ditches, and storm drain
systems. Efficient, in this context, has meant the rapid removal of stormwater from
the point where it occurs. This way of thinking assumes that stormwater is something
to be disposed of rather than used as a resource. Over the years, however, we have
realized that such efficiency frequently causes downstream flooding. This realization
inaugurated a period during which regional detention ponds were considered the cost-
effective means of attenuating flood peaks from upstream development. Because
regional ponds tended to be large facilities, the role of municipal government
increased through study of regional stormwater management needs, identification of
sites, and coordination of financing through public and private sources.
We then realized that such large facilities, frequently located on live drainageways,
impaired the natural stream function. Furthermore, it became apparent that these
regional ponds did little to enhance water quality and may have exacerbated channel
degradation problems. As a result, design practices have evolved to consider infiltra-
tion, storage, sediment transport, and water quality issues. Adapting regional facilities
to meet water quality objectives, through techniques such as extended detention and
wetland retention systems, were among the responses to the newly identified prob-
lems. Regional systems also began to give way to dispersed on-site infiltration and
retention systems that are designed to not only reduce peaks of flood events but also
to mimic the runoff characteristics for the full range hydrological events. The goal
became to capture and treat water on-site, before releasing it downstream.
These changes in thinking have required those involved in stormwater management to
adapt existing modeling tools and to develop new ones to meet the new demands. The
suite of stormwater management facilities that we now have to choose from has
grown. We can expect that as our experience grows with our current stormwater man-
agement philosophy, we will continue to adapt as we recognize new limitations of
standard practices. It may be expected that this will primarily come in the form of
land use and development management strategies rather than new forms of con-
structed facilities.
As the undesired effects of past stormwater management practices are being more
fully realized, stormwater management regulations are also evolving. Rather than
focusing on the infrequent events for flood protection, emphasis is increasingly
placed on facility design that mimics pre-development channel forming discharges as
well as environmental goals such as fish passage. Designing for a channel forming
discharge requires an understanding of the amount of work that a stream can do in
transforming its vertical profile and horizontal pattern. This, in turn, requires mastery
of sediment transport characteristics as well as flow frequency relationships. With
respect to fish passage, we are concerned with defining discharges, depths, and veloc-
ities of lower flow events. Some of these new demands on designers require an under-
standing of continuous simulation modeling in addition to event-based design.
Modeling tools are also evolving to meet today's regulatory environment.
In spite of the changes in stormwater management practices and regulations, and the
evolution of modeling tools, stormwater modeling, analysis, and design are still based
on the same fundamentals. They still require tools that allow for the representation of
precipitation, conveyance, storage, and treatment of stormwater. And, they still
require tools that allow for the geographic variation in rainfall and runoff patterns as
well as jurisdictional variation in policies and regulations.
In 1889, Emil Kuichling introduced the rational method to the United States in his
paper "The Relation Between the Rainfall and the Discharge of Sewers in Populous
Districts." In the intervening century, we have kept what is good and discarded what-
ever did not meet our demands, and we have continued to adapt. Whatever the spe-
cific stormwater requirements and techniques, we still require analysts and designers
who understand the fundamental theory behind the tools and that can appropriately
apply stormwater management models. This book, by Haestad Press, is a valuable
contribution to keeping the practice of stormwater modeling alive and vigorous.
Roger T. Kilgore, P.E.

Kilgore Consulting
Table of Contents
Preface xi
About the Software xv
Chapter 1 Introduction to Stormwater Conveyance

Modeling and Design 1
1.1 Need for Stormwater Conveyance Systems 2
1.2 History of Stormwater Management Systems 5
Ancient Practices............................................................................. 5
Post-Roman Era to the 1700s .......................................................... 6
The Nineteenth Century .................................................................. 7
The Twentieth Century.................................................................... 9
Summary and Outlook .................................................................. 11
1.3 Overview of Stormwater Conveyance System Design 11
Design Objectives ......................................................................... 12
Resources ...................................................................................... 13
Constraints..................................................................................... 13
What Is Best? ................................................................................ 14
1.4 Chapter Summary 14
Chapter 2 System Components, Models,

and the Design Process 17
2.1 Hydrologic Components 18
Precipitation .................................................................................. 18
Runoff............................................................................................ 18
ii Table of Contents
2.2 Hydraulic Components 19

Open Channels...............................................................................19
Culverts..........................................................................................22
Storm Sewer Systems ....................................................................25
Stormwater Detention and Retention ............................................33
Summary of Hydraulic Models for Various Components of
Stormwater Management Systems.............................................34
2.3 The Design Process 36
Master Planning .............................................................................36
Preliminary Design Development .................................................37
Detailed Design .............................................................................38
Construction Drawings, Specifications, and Contract
Documents .................................................................................38
2.4 Integration of Stormwater Conveyance Modeling with CAD
Software and GIS 39
CAD Integration ............................................................................39
GIS Integration ..............................................................................41
2.5 Sources of Model Data 44
Proposed Systems ..........................................................................45
Existing Systems............................................................................46
Chapter 3 Fundamental Laws and Units 51

3.1 Conservation of Mass 51
3.2 Conservation of Energy 55
3.3 Conservation of Momentum 58
3.4 Common Units of Measurement 61
Chapter 4 Modeling Rainfall 67

4.1 The Hydrologic Cycle 68
4.2 Precipitation Types 70
Frontal Precipitation ......................................................................70
Cyclonic Precipitation ...................................................................70
Convective Precipitation................................................................71
Orographic Precipitation................................................................71
iii
4.3 Basic Rainfall Characteristics 72

Depth and Intensity ....................................................................... 73
Temporal Distributions ................................................................. 73
Spatial Distributions...................................................................... 74
Recurrence Interval ....................................................................... 74
4.4 Obtaining Rainfall Data 76
Rainfall Data Collection Methods................................................. 76
Rainfall Data Sources.................................................................... 79
4.5 Types of Rainfall Data 87
Intensity-Duration-Frequency Curves........................................... 87
Temporal Distributions and Hyetographs for Design Storms....... 91
4.6 Rainfall Requirements for Modeling Runoff 101
Chapter 5 Modeling Runoff 107

5.1 Rainfall Abstractions 108
Interception.................................................................................. 108
Depression Storage...................................................................... 109
Infiltration.................................................................................... 110
5.2 Determination of Effective Precipitation (Runoff) 116
Runoff Coefficient ...................................................................... 118
NRCS (SCS) Curve Number Method ......................................... 119
Creating Direct Runoff (Effective Rainfall) Hyetographs .......... 132
5.3 Measures of Basin Response Time 134
Basin Response Time Estimation Methods................................. 135
Time of Concentration and Basin Lag ........................................ 139
5.4 Peak Flow Estimation 140
Rational Method.......................................................................... 140
NRCS (SCS) Peak Flow Estimation ........................................... 144
USGS Regression-Based Formulas............................................. 148
5.5 Hydrograph Estimation 150
Overview of Direct Runoff Hydrograph Estimation
Using a Unit Hydrograph ........................................................ 152
Unit Hydrographs........................................................................ 152
Discrete Convolution................................................................... 158
Kinematic Wave Model .............................................................. 162
5.6 Base Flow 165
5.7 Snowmelt 166
Runoff Potential .......................................................................... 167
Snowmelt Models........................................................................ 167
iv Table of Contents
5.8 Channel Routing 169

Muskingum Routing ....................................................................170
5.9 Reservoir Routing 173
5.10 Summary of Methods for Computing Flows 173
Chapter 6 Flow in Closed Conduits 185

6.1 The Energy Equation 186
The Velocity Distribution Coefficient .........................................187
Graphical Representation of the Energy Equation ......................188
6.2 Resistance to Flow in Closed Conduits 192
The Reynolds Number.................................................................192
The Darcy-Weisbach Equation....................................................192
The Manning Equation ................................................................198
The Chézy Equation ....................................................................200
The Hazen-Williams Equation ....................................................201
6.3 Minor Losses at Bends and Appurtenances 202
Classical Method .........................................................................202
6.4 Momentum and Forces in Pipe Flow 208
Evaluating Forces on Hydraulic Structures .................................210
Chapter 7 Flow in Open Channels 217

7.1 Classification and Properties of Open Channels 218
Geometric Properties of Open Channels .....................................219
7.2 Energy and Constitutive Relationships 223
Pressure Variations in Open-Channel Flow ................................225
Constitutive Relationships for Open-Channel Flow....................225
Minor Losses in Open-Channel Flow..........................................230
7.3 Uniform Flow and Normal Depth 231
Uniform Flow in Simple Channels ..............................................233
Uniform Flow in a Channel with a Gradually Closing Top ........235
Uniform Flow in a Channel with Composite Roughness ............238
Uniform Flow in Compound Cross-Sections ..............................240
7.4 Specific Energy and Critical Flow 244
Specific Energy............................................................................244
Critical Depth and Critical Flow .................................................246
v
Compound Channel Sections ...................................................... 248

Flow Control ............................................................................... 252
Flow Measurement...................................................................... 252
Analysis of Flow in Channel Transitions.................................... 252
7.5 Momentum in Open-Channel Flow 255
Specific Force.............................................................................. 258
The Hydraulic Jump .................................................................... 260
Forces on Submerged Structures................................................. 263
7.6 Weirs 265
Weir Characteristics .................................................................... 265
Sharp-Crested Weirs ................................................................... 266
Broad-Crested Weirs ................................................................... 268
7.7 Orifices 270
7.8 Gradually Varied Flow 271
Water Surface Profile Equation................................................... 272
Classification of Flow Profiles.................................................... 273
Profile Calculations ..................................................................... 275
7.9 Spatially Varied Flow 277
Spatially Varied Flow with an Increasing Discharge.................. 278
Spatially Varied Flow with a Decreasing Discharge .................. 279
Chapter 8 Design of Open Channels 287

8.1 Design Considerations 288
Velocity of Flow.......................................................................... 288
Froude Number ........................................................................... 289
Side Slopes .................................................................................. 289
Longitudinal Slope ...................................................................... 289
Freeboard..................................................................................... 289
Lining Material............................................................................ 290
Cost and Maintenance ................................................................. 290
Safety........................................................................................... 290
Aesthetics .................................................................................... 290
Right-of-Way Requirements ....................................................... 291
Dry-Weather Flows ..................................................................... 291
8.2 Best Hydraulic Section 291
8.3 Design of Lined Channels 294
8.4 Design of Erodible Channels 297
Method of Permissible Velocity.................................................. 297
Method of Tractive Force............................................................ 303
vi Table of Contents
8.5 Design of Grassed Channels 308

Procedure for Design of a Grassed Channel................................312
8.6 Riprap-Lined Channels 315
8.7 Modifying Existing Channels 315
Chapter 9 Culvert Design 323

9.1 Culvert Design Flows 325
9.2 Culvert Types and Materials 326
9.3 Culvert End Treatments 327
9.4 Alignments of Culverts 331
Horizontal Alignment ..................................................................331
Vertical Alignment ......................................................................332
Skewed Culverts ..........................................................................333
9.5 Culvert Hydraulics 334
Flow Controls ..............................................................................335
Graphical Analysis of Culvert Performance................................340
Flow Profile Analysis of Culvert Performance ...........................348
Outlet Velocity ............................................................................354
9.6 Roadway Overtopping 355
9.7 Performance Curves 359
9.8 Erosion Protection 360
9.9 Environmental Considerations 362
Chapter 10 Gutter Flow and Inlet Design 369

10.1 Roadway Gutters 370
Flow in Uniform Gutter Sections ................................................371
Flow in Composite Gutter Sections.............................................373
Other Gutter Sections ..................................................................377
10.2 Inlets 377
Hydraulic Performance Concepts ................................................378
Grate Inlets ..................................................................................380
Curb-Opening Inlets ....................................................................392
Combination Inlets ......................................................................397
Slotted-Drain Inlets......................................................................398
vii
10.3 Inlet Locations and Design Flows 400

Chapter 11 Storm Sewer Pipe System and

Outlet Design 407
11.1 Storm Sewer Types, Components, and Materials 409
Storm Sewer Pipes ...................................................................... 409
Storm Sewer Appurtenances ....................................................... 416
11.2 Alignment of Storm Sewers 421
Horizontal Alignment.................................................................. 421
Vertical Alignment...................................................................... 421
Sewer Slopes ............................................................................... 422
11.3 Initial Pipe System Design 422
Multiple Design-Storm Return Periods....................................... 425
Directly and Indirectly Connected Areas .................................... 426
11.4 Minor Losses in Junction Structures 434
Energy-Loss Method .................................................................. 434
Composite Energy-Loss Method................................................. 440
11.5 Final Design of the Pipe System 444
Surcharged Pipes ......................................................................... 444
Open-Channel Flow .................................................................... 452
Structural Loads and Stresses in Pipes........................................ 460
Flow Routing in Storm Sewers ................................................... 460
11.6 Using Computer Models for Storm Sewer Design and Analysis 461
11.7 Outfall Design and Energy Dissipation 464
Chapter 12 Stormwater Detention 477

12.1 Overview of Stormwater Detention 478
The Advantages of Implementing Detention Ponds ................... 478
Stormwater Detention versus Retention...................................... 479
Additional References on Stormwater Detention........................ 479
12.2 Types of Stormwater Detention Facilities 479
Dry versus Wet Ponds ................................................................. 480
Aboveground versus Underground Detention............................. 481
On-Line versus Off-Line Detention ............................................ 482
Regional versus On-Site Detention ............................................. 482
Independent versus Interconnected Ponds .................................. 484
viii Table of Contents
12.3 Stormwater Detention Modeling Concepts 485

12.4 Detention Design Considerations 487
Pre-Development versus Post-Development Criteria..................488
Downstream Systems ..................................................................488
Mandated Release Rates ..............................................................488
Design Storms..............................................................................488
Peak Flow Timing Issues.............................................................489
Bypassing Flows from Off-Site ...................................................490
Volume Mitigation Issues............................................................491
12.5 Components of Detention Facilities 493
Dam Embankments......................................................................493
Freeboard .....................................................................................493
Pond Bottoms and Side Slopes ....................................................493
Outlet Structures ..........................................................................494
Overflow Spillways .....................................................................496
Outfall Sheet Flow Conditions ....................................................496
Diversion Structures and Conveyances .......................................497
12.6 Routing Data: Storage and Hydraulic Relationships 497
Stage versus Surface Area ...........................................................497
Stage versus Storage Volume ......................................................499
Stage versus Discharge ................................................................505
12.7 Storage Routing Calculations 509
Storage-Indication Method ..........................................................511
Pond Routing by Numerical Integration......................................520
12.8 Iterative Pond Design 525
Chapter 13 Stormwater Pumping 533

13.1 Need for Stormwater Pumping 534
13.2 Pump Types and Installation Methods 536
Pump Types .................................................................................536
Pump Installations .......................................................................540
13.3 Pump Performance and Selection 545
Pump Performance Curves ..........................................................545
System Curves .............................................................................548
Operating Point and Pump Selection...........................................552
Cavitation and Limits on Pump Location....................................553
Installations with Multiple Pumps ...............................................555
13.4 Piping System Requirements 556
Pipe Requirements .......................................................................556
Valves and Fittings ......................................................................558
ix
13.5 Pump Controls 559

13.6 Storage Requirements 560
13.7 Pumping Station Design 563
Hydrology.................................................................................... 563
Development of a Trial Design ................................................... 563
Trial Design Evaluation .............................................................. 564
13.8 Miscellaneous Design Considerations 566
Architectural Considerations....................................................... 566
Power and Control Requirements ............................................... 567
Trash Racks and Grit Chambers.................................................. 567
Monitoring Devices..................................................................... 569
Safety and Ventilation Requirements.......................................... 569
Doorways, Hatches, and Monorails ............................................ 569
Chapter 14 Regulatory and Environmental Issues 575

14.1 International Overview of Regulatory Guidelines and Policies 575
Milestones in International Water Resources Policies................ 576
Agenda 21 and Water Management ............................................ 576
Additional International Guidelines ............................................ 577
Drainage Management in Developing Countries ........................ 580
Regulation of International Watercourses................................... 582
Water and Environment in the European Union ......................... 583
Canadian Stormwater Regulations .............................................. 589
14.2 Regulatory Requirements in the United States 592
Federal Laws and Regulations .................................................... 593
State Laws and Regulations ........................................................ 599
Local Regulations........................................................................ 601
Chapter 15 Stormwater Quality Management 607

15.1 Environmental Effects of Stormwater Runoff 608
Pollutants in Urban Stormwater .................................................. 609
Effects on Streams....................................................................... 612
Receiving-Water Considerations................................................. 613
15.2 Hydrology for Stormwater Quality Control 614
x Table of Contents
15.3 Stormwater Quality Best Management Practices 617

Structural BMPs ..........................................................................618
Effectiveness of Structural BMPs................................................626
Nonstructural BMPs ....................................................................628
15.4 Approaches to Applying Best Management Practices 630
Implementation of Controls in Newly Developing Areas ...........631
Stormwater Controls Suitable for Retrofitting in Existing Areas633
Construction Project Best Management Practices.......................634
Appendix A Physical Properties of Water 641
Appendix B Unit Conversions 643
Appendix C Unsteady Stormwater Flow 645

C.1 Basics of Unsteady Flow Analysis 645
C.2 Hydrodynamic Equations 647
C.3 Approximations of the Hydrodynamic Equations 648
Kinematic Wave Routing ............................................................648
Muskingum Routing ....................................................................648
Muskingum-Cunge Routing ........................................................649
Convex Routing ...........................................................................649
Weighted Translation Routing.....................................................650
Translation Routing .....................................................................650
C.4 Routing method Complications 650
Appendix D References for Common Stormwater

Conveyance System Component Models
and Software 653
Bibliography 655
Preface
Design and analysis of an urban stormwater management system is a fascinating and

complex blend of heuristics and practicality, and of complex hydrologic and
hydraulic analyses associated with runoff-producing storm events and the resulting
mix of over-land, open-channel, and confined flows in a drainage basin and its
stormwater con-veyance features. Designers and analysts must keep in mind not
only the engineering goals of efficient and cost-effective solutions, but also a host of
other goals and constraints pertaining to legal and institutional requirements, public
safety, aesthetics, and main-tainability.
Designers and managers of urban stormwater conveyance systems are

increasingly being required to consider innovative approaches that minimize
runoff production through practices, such as on-site infiltration, and that improve
the water quality of the runoff before discharging it to downstream properties or
receiving waters. The great-est benefits of these new practices tend to be associated
with low- to moderate- intensity storms that occur quite frequently, as they
contribute the bulk of the runoff and pollutant load from an urban drainage
basin on an annual or long-term basis. Although these new practices and
mandates will undoubtedly improve water quality in urban areas, stormwater
management systems will continue to be required to safely and efficiently convey the
large flows resulting from high-intensity storm events. The general public may
demand and expect good water quality in an area, but it is highly unlikely that they
will be willing to sacrifice their safety and well-being to accomplish those goals.
Thus, the need for efficient and cost-effective stormwater conveyance facilities is
not being supplanted by new environmental regulations—it is instead being
augmented by them and making the challenge of design increasingly exciting.
This text includes elements of design pertaining to stormwater quality,

mainly through its discussions of environmental regulations, best management
practices, and an introduction to detention pond design, but its main focus is on
the more traditional aspects of hydrologic prediction and hydraulic design and
analysis of conveyance facilities. In writing the text, an attempt has been made to
produce a volume that inte-grates materials and methods not usually found in a single
reference source. Thus, it is hoped that not only will the experienced design
professional and/or manager find it useful as a desk reference, but also that
students and academics will find it useful for courses dealing with stormwater
management system design.
xii Preface
Chapter Overview
Chapter 1 of this book discusses the need for stormwater management and the general
design process. It also presents a history of the development of wet-weather flow
management technologies.
Chapter 2 discusses the various components of stormwater management systems and
introduces the various types of models and tools, including GIS and CAD, that are
used in their design and analysis. It also summarizes the main elements of the design
process, from master planning through final design and construction contract develop-
ment. The experienced designer will find this chapter informative, although it may be
most valuable to the inexperienced individual who wants to learn about stormwater
management systems and their design and analysis methodologies.
Nearly all hydrologic and hydraulic methods used for design and analysis of storm-
water management systems are applications of the fundamental physical laws of con-
servation of mass, energy, and momentum. Chapter 3 reviews these laws as a basis for
further developments and applications of them in later chapters.
Design of any stormwater system component, such as a culvert or pipe, must be pre-
ceded by estimating the design discharge (flow rate) that the component is to convey.
Hydrology is the science by which such estimates are developed, and, in that context,
involves estimating the runoff rate from a drainage basin for a chosen duration and
frequency of rainfall. Chapter 4 presents sources and methods for modeling design
rainfall events and is followed by methods for conversion of rainfall into runoff in
Chapter 5.
The flow of stormwater through various conveyance system components occurs gen-
erally as confined (pipe) flows and as open-channel flows. Hydraulic methods of anal-
ysis of each of these fundamental types of flow are presented in Chapters 6 and 7,
respectively, which build on the fundamental laws presented in Chapter 3.
With the fundamental tools of hydrologic and hydraulic analysis established in earlier
chapters, Chapters 8 through 12 delve into detailed design procedures for various
components of stormwater conveyance systems. Chapter 8 addresses design of open
channels, and Chapter 9 addresses culvert design. Chapters 10 and 11 cover design of
the most common conveyance system elements, namely roadway gutters, stormwater
inlets, and pipe systems and their outfalls. Chapter 12 addresses design of stormwater
detention facilities, which are often mandated for attenuation of peak runoff rates
from developed lands or for water quality management.
It is occasionally true that stormwater requires pumping from low-lying areas, such
as highway underpasses, and from areas protected by levees. Chapter 13
provides an introduction to the design of stormwater pumping facilities.
Chapter 14 provides a broad overview of the laws and regulations under which
storm-water management systems must be developed and includes both U.S. and
international perspectives. The presentation is necessarily broad, as each
political jurisdiction has its own specific requirements. Readers must acknowledge
this and must take it upon themselves to become familiar with the requirements in
their locale.
xiii
Stormwater management professionals are increasingly being required to address

issues of runoff quality and receiving water impacts and to design system compo-
nents that can control the pollutants in small, frequently occurring rainfall events.
Chapter 15 presents examples of best management practices (BMPs) that are com-
monly applied in the Denver, Colorado (USA), metropolitan region for stormwater
management. Similar methods are often used in many other parts of the nation and
world.
Conventions
Because of the diversity of sources from which material has been gathered, and
because of the differences in mathematical notation from one source to another,
choices have been made regarding notation. Remaining faithful to the notation used in
original references would have given rise to the same notation being used to represent
different things in various places throughout the book. An example of this is the vari-
able Q, used in most hydraulic literature to represent the volumetric discharge of a
flow, but used in NRCS (SCS) hydrologic methods to represent the effective precipi-
tation depth during a storm. Because of the confusion that could be caused, we opted
instead to use a single set of notation throughout, even though this sometimes results
in notation that differs from that found in the original reference sources.
Continuing Education and Problem Sets

Also included in this book are more than 100 hydraulics and modeling problems to
give students and professionals the opportunity to apply the material covered in each
chapter. Some of these problems have short answers, and others require more
thought and may have more than one solution. The accompanying CD in the back of
the book contains academic versions of Bentley's StormCAD, PondPack, Cul-
vertMaster, and FlowMaster software (see “About the Software” on page xv), which
can be used to solve many of the problems. However, we have endeavored to
make this book a valuable resource to all modelers, including those who may be
using other software packages, so these data files are merely a convenience, not a
necessity.
Bentley also publishes a solutions guide that is available to instructors and
professionals.
About the Software
The CD in the back of this book contains academic versions of Bentley's

StormCAD, PondPack, CulvertMaster, and FlowMaster software. The following
provides brief summaries of the software functionality. For detailed information
about how to apply it to solve stormwater problems, see the help systems and
tutorial files included on the CD. The software packages included with this textbook
are fully functional but are not intended for professional use (see license
agreement in the back of this book).
STORMCAD
StormCAD is a powerful, easy-to-use program that helps civil engineers design and
analyze storm sewer systems. StormCAD has a CAD-like interface but does not
require the use of third-party software in order to run. StormCAD provides intuitive
access to the tools you need to model complex hydraulic situations.
PONDPACK
PondPack is a feature-rich stormwater program that analyzes a tremendous range of
situations, from simple sites to complex networked watersheds. The program analyzes
pre- and post-developed watershed conditions and automatically sizes ponds. It also
computes outlet rating curves with tailwater effects, accounts for pond infiltration,
calculates pond detention times, analyzes channels, and models interconnected ponds
with tidal outfalls.
CULVERTMASTER
CulvertMaster makes culvert analysis and design much easier. Use the Quick
Culvert Calculator to solve for any unknown variable, such as headwater depth,
discharge, or culvert size. Evaluate existing culvert systems and develop
improvement strategies using the Culvert Analyzer, or compare and contrast
numerous design options with the Culvert Designer. From simple, single culvert
sections to complex, multi-barrel culverts with roadway overtopping, CulvertMaster
handles them all with ease.
FLOWMASTER
FlowMaster is an efficient and powerful program for the design and analysis of pipes,
ditches, open channels, weirs, orifices, gutters, and inlets. FlowMaster’s “Hydraulic
xvi About the Software
Toolbox” can solve or rate any unknown variable using the Manning, Hazen-
Williams, Kutter, Darcy-Weisbach, and Colebrook-White formulas. FlowMaster’s
inlet computations strictly comply with the latest FHWA Hydraulic Circular Number
22 and AASHTO inlet computation guidelines. Create multiple design trials and com-
pare the results in customizable tabular reports, performance curves, and rating tables.
C H A P T E R
1
Introduction to Stormwater
Conveyance Modeling and Design
Stormwater is water from rain, snowmelt, or melting ice that flows across the land
surface. Conveyance systems for dealing with stormwater are everywhere—in
backyards, streets, parking lots, and parks. They frequently affect people but typically
go unnoticed unless they fail. The consequences of failure range from nuisance
flooding of yards, basements, and roadway travel lanes, to temporary road or bridge
closure and minor property damage, to widespread destruction and even loss of life.
Another consideration when dealing with stormwater that has received increased
attention in recent years is stormwater quality. In urban areas, stormwater often comes
into contact with pollutants such as oil from roadways and parking lots, which it then
carries into surface waters such as lakes and streams. Sediment loads from construc-
tion sites can also be problematic, as can runoff from agricultural lands exposed to
fertilizers, pesticides, and animal waste.
This book presents information necessary to appropriately analyze and design storm-
water conveyance systems, including the following topics:
• Basic components of stormwater systems and conveyance modeling
• Creation of project design storms
• Determination of the runoff rates to the system that will result from these
storms
• Conveyance structure design alternatives
• Hydraulic analyses to determine which alternatives meet design criteria
• Practices for improving stormwater quality
Other factors to be considered, such as cost and available budget and the upstream and
downstream impacts of a project, are also discussed.
2 Introduction to Stormwater Conveyance Modeling and Design Chapter 1
This chapter begins the book by citing statistics and providing background on the fac-
tors driving the need for stormwater conveyance systems. Section 1.2 gives a brief
history of stormwater management, and Section 1.3 provides an overview of issues
that must be addressed in system design.
1.1 NEED FOR STORMWATER CONVEYANCE SYSTEMS

The need to manage stormwater runoff to protect vulnerable areas against destructive
flooding is an age-old problem. Flooding poses a threat to developed areas, agricul-
tural crops and livestock, human life, and highways and other transportation systems.
Flooding happens across all spatial scales—from individual properties and neighbor-
hoods along small creeks, to entire regions bordering major rivers.
Of course, the largest part of the losses is due to major flooding along large streams
and rivers. Therefore, proper designs of storm sewers and culverts, which are the
focus of this book, will only partly cure the growing cost of flooding in terms of mon-
etary damage and loss of life. (Chapter 2 provides an overview of the types of storm-
water facilities presented in this text.) Nevertheless, careful stormwater conveyance
designs that consider not only the service of an immediate local area, but also how
this area merges into and interacts with regional-scale stormwater and flood protec-
tion plans and facilities (sometimes called a “major system” of drainage conveyance
facilities), can alleviate many of the damages associated with frequent storm events.
The magnitude of the flooding problem can be appreciated by citing a few statistics.
Figure 1.1 illustrates the trend in annual flood losses in the United States over the
period from 1903 to 1999, where all losses are expressed in 1997 dollars. The general
upward trend is significant in view of the fact that the vertical axis is logarithmic.
Over the period from 1992 to 1996, it has been estimated that losses due to natural
hazards averaged $54 billion per year, or over $1 billion per week (National Science
and Technology Council, 1997). Flooding alone accounted for about 26 percent of
that total, or roughly $270 million per week.
It is staggering to realize that these huge losses have occurred despite the fact that for
most of the twentieth century, the United States has been largely spared the expense
of a catastrophic natural disaster in a highly populated and developed area. For exam-
ple, a great earthquake of magnitude 8 or more on the Richter scale has not struck a
major metropolitan area since the one in San Francisco, California, in 1906. A Class 4
or 5 hurricane has not struck a major urban area directly since 1926 in Miami, Florida
(van der Vink et al., 1998). However, even though hurricanes have largely spared
major population centers in recent years, they do pose threats every year and have
caused substantial property damage, mainly to the Gulf and Atlantic coasts.
Sobering also is the cost of natural disasters in terms of loss of life. Figure 1.2 shows
estimates published by the U.S. National Climatic Data Center and cited by the
National Research Council (1991). The estimates indicate that during the period from
1941 to 1980, U.S. population-adjusted death rates due to flooding have generally
increased, but those due to lightning and tornadoes have dropped dramatically. Popu-
Section 1.1 Need for Stormwater Conveyance Systems 3
Figure 1.1
Annual flood losses in
the United States from
1903–1999 (as
reported by the U.S.
National Weather
Service, 2000)
lation-adjusted death rates due to tropical cyclones (hurricanes) have remained fairly
constant over that time period. As of 1980, the population-adjusted death rate due to
flooding alone was about as great as that due to lightning, tornadoes, and tropical
cyclones combined. Although the greater share of the loss of life from floods indi-
cated in Figure 1.2 is due to major flooding, a substantial portion is due to flooding of
drainageways in urbanized areas. Since such drainageways are often engineered as
part of an urban stormwater management and conveyance system, they fall within the
scope of this text.
Many factors contribute to the upward trends in flood damages. However, the escala-
tion is due largely to the increased occupancy and use of floodprone lands and coastal
areas. Despite good intentions, public financing and implementation of so-called
“flood control” projects are partly to blame for this state of affairs. The public,
including politicians and decision-makers, are accustomed to hearing the term “flood
control” without fully understanding the limitations of such projects. Often, one con-
sequence of these projects is a false sense of security, so that when the inevitable large
flood does occur, damages are greater than they would have been otherwise.
History has shown that floods cannot be controlled completely. Technical experts
have never intended the term “flood control” to be interpreted in an absolute sense,
but the phrase is often misleading to the public. To avoid public misconceptions about
the level of protection provided by flood control projects, terminology such as “excess
water management,” which is less suggestive of absolute control, should be
considered.
Figure 1.2
Population-adjusted
death rates in the U.S.
due to natural
disasters from 1941 to
1980
The monetary damage and the loss of life illustrated in Figures 1.1 and 1.2, respec-
tively, are aggregates representing totals for the entire United States. Similar informa-
tion may be obtained for other parts of the world. The Natural Hazards Center at the
University of Colorado, Boulder is a clearinghouse for data and other information on
natural hazards worldwide, and provides Internet links to many useful sources. The
Web site for the Hazard Reduction and Recovery Center at Texas A&M University
also provides useful research tools.
Damage and loss of life will continue to be among the most important factors driving
stormwater management. These factors are also the most tangible indicators of the
success (or failure) of the stormwater management process. Other considerations that
drive the need for stormwater management are environmental and water quality con-
cerns, aesthetics, and legal and regulatory considerations. The complexity of the driv-
ing factors and the consequent need for evaluation of the trade-offs among competing
alternatives make stormwater management an exciting and challenging process.
Section 1.2 History of Stormwater Management Systems 5
1.2 HISTORY OF STORMWATER MANAGEMENT

SYSTEMS
As noted in Section 1.1, the need for dealing with stormwater runoff is an age-old
problem. A number of accounts of historical developments in wet-weather flow man-
agement have been published over the years; the summary by Burian et al. (1999) is
one of the most recent and serves as a model for the briefer presentation here. Addi-
tional historic information relevant to regulatory requirements is presented in
Chapter 14.
Ancient Practices
Several ancient civilizations constructed successful surface-water drainage systems.
Some civilizations incorporated the removal of sanitary wastes into the surface runoff
system to form a combined system of sewage. For example, around 3000 B.C., dwell-
ers of the Indus civilization in the city of Mohenjo-Daro (now in Pakistan) con-
structed combined sewers consisting of a simple sanitary sewer system with drains to
remove stormwater from streets. The masonry work and clever design of this system
demonstrate that, in some instances, more care was taken with the sewage system than
with construction of some of the buildings. Other ancient sewer systems were con-
structed by the Mesopotamian Empire in Assyria and Babylonia (circa 2500 B.C.), and
by the Minoans on the island of Crete (3000–1000 B.C.). Yet others were constructed
in parts of the city of Jerusalem (circa 1000 B.C.); and in central Italy by the Etruscans
around 600 B.C. Ruins of some major cities in China indicate that partially under-
ground systems were constructed around 200 A.D. Other examples of ancient sewage-
system builders include the Greeks in Athens, the Macedonians and Greeks under the
rule of Alexander the Great, and the Persians.
Of all the societies of western Asia and Europe, from antiquity until the nineteenth
century, only the Romans built a carefully planned road system with properly drained
surfaces. Specific drainage facilities used by the Romans included occasional curbs
and gutters to direct surface runoff into open channels alongside roadways. They also
graded roadway surfaces to enhance drainage to the channels. This channel-based
sewer system collected not only stormwater runoff from roadway surfaces but also
sanitary and household wastes. During periods of dry weather when flows were insuf-
ficient to flush the channels, the wastes accumulated and caused unsanitary conditions
to develop. Therefore, the channels were covered, evolving into combined sewer sys-
tems. These covered drains ultimately developed into the Roman cloacae, or under-
ground sewers (see Figure 1.3). These sewers were a source of civic pride and
symbolized the advanced stature of the Roman civilization.
Figure 1.3
Outlet of the Cloaca
Maxima on the Tiber
River, Rome (Blake,
1947)
Post-Roman Era to the 1700s

From the time of the Roman Empire through the 1700s, European sewage manage-
ment strategies experienced little advancement and even regressed in terms of sanita-
tion. Along with the decreasing populations of cities after the fall of the Roman
Empire came the abandonment of municipal services such as street lighting, running
water, and sewer systems.
During the Dark Ages, Western society was indifferent to such issues. Most citizens
ignored matters of hygiene and cleanliness, for example. As the Middle Ages pro-
gressed, drainage and sanitary services were developed in response to aggravating
conditions and disease outbreaks. Unfortunately, these systems were disjointed and
poorly planned. Paris and London exemplify European cities with piecemeal systems
at that time.
Resurgence in the development of planned sewage systems from the haphazard sys-
tems of the Middle Ages occurred between the fourteenth and nineteenth centuries.
The sewers constructed in Europe during this time, such as those in London and Paris,
were simply open ditches. Besides being conveyances for stormwater, these channels
became receptacles for trash and other household and sanitary wastes, the accumula-
tion of which caused overflows. As a remedy, the channels were covered to create
combined sewers, essentially replicating the practice of the Romans some 1500 years
earlier.
In Paris, the first planned covered sewer dates to 1370, when Hugues Aubriot con-
structed the Fosse de St. Opportune. This “beltway sewer” discharged into the Seine
River and acted as a collector for the sewers on one side of the Seine. However, for the
most part, sewer systems before the 1700s lacked proper design and were conducted
in a piecemeal fashion. Maintenance and proper operation were virtually neglected.
The systems were poorly functioning and subject to repeated blockages, resulting in
nuisance flooding.
The Nineteenth Century

At the end of the 1700s, the outlook was improving for wet-weather flow manage-
ment in Europe. Society held a belief in progress that became increasingly linked to
technology throughout the 1800s. Until the 1820s, European sewers were constructed
of cut stone or brick with rectangular or roughly rounded bases that contributed to
deposition problems. Substitution of millstone and cement mortar stone made con-
struction of curved and smooth sewer floors easier, thus improving both the self-
cleansing properties and hydraulic efficiency. A variety of new pipe shapes were
developed, including egg-shaped, oval, and v-notch cross sections. Additional
improvements in the early 1800s consisted of increased attention to sewer mainte-
nance and the adoption of minimum slope and velocity criteria to provide adequate
flushing during dry-weather periods.
Comprehensive planning was the next major step in sewage system design. In 1843,
William Lindley, an Englishman living in Hamburg, Germany, was commissioned to
plan and design a system following a major fire (see Figure 1.4). London followed
Hamburg’s successful example by commissioning Joseph Bazalgette to plan and
design its system. Construction of the Main Drainage of London was completed in
1865. In 1858, E. S. Chesbrough designed the first comprehensive wet-weather flow
management system in the United States for the City of Chicago, Illinois. About the
same time, J. W. Adams designed a comprehensive sewer system for Brooklyn, New
York. The designs implemented in the United States often made use of empirical data
obtained from European practice. Because rainfall intensities and storm durations are
highly location dependent, as are factors such as time of concentration and soil type,
the topographic and climatological differences between the United States and Europe
ultimately contributed to design deficiencies.
Figure 1.4
Map of Hamburg,
Germany sewer
system from 1857
(Chesbrough, 1858)
In the middle to late 1800s, rapid urbanization in the United States led to increases in
water consumption and consequent overwhelming of privy-vault and cesspool sys-
tems for wastewater disposal. After much debate, city councils, engineers, and health
groups generally agreed that centralized sewage systems were more cost-effective and
provided greater benefits than did other options.
The question arose, however, as to which type of sewage system should be con-
structed—separate or combined. Initially, the combined sewer scheme was widely
implemented for several reasons, including the fact that there was no European prece-
dent for successful separate systems. Further, a belief existed that combined systems
were cheaper to construct.
There were opponents to this practice, however. For example, Colonel George E.
Waring designed relatively successful separate sewer systems in Lenox, Massachu-
setts, and Memphis, Tennessee. Around the same time (about 1880), at the request of
the U.S. National Board of Health, Rudolph Hering visited Europe to investigate
practices there. Hering’s report recommended combined systems when serving exten-
sive and densely developed areas, and separate systems for areas in which rainwater
did not require removal underground. By the end of the 1800s, engineers had
embraced Hering’s ideas, and combined sewers were recommended for most urban-
ized areas. This philosophy did not change until the 1930s and 1940s when more
wastewater treatment was required, because the contribution of stormwater volume
translated into higher treatment costs. Consequently, separate sewers then became the
more commonly recommended option.
Regardless of the type of sewage (separate or combined), control and treatment of
wastewater discharges were limited. It was commonly assumed that cities could
safely dispose of their wastes into adjacent waterways, with “dilution” being the
“solution to pollution.” By the middle to late 1800s, epidemiological research by John
Snow and William Budd, which showed water’s ability to transmit infectious disease,
and germ and bacterial research by Louis Pasteur and Robert Koch, were beginning to
challenge the effectiveness of dilution strategies. Despite the fact that this fundamen-
tal scientific research demonstrated a connection between polluted waters and dis-
ease, wastewater treatment was not widely practiced at this time. Debate centered on
whether it was better to treat wastewater before discharging it to a stream or for down-
stream users of the water to treat it before distribution as potable water. Most engi-
neers subscribed to the latter approach, effectively ignoring the effects on the habitats
and recreational uses afforded by receiving waters.
In the mid-1800s, estimation of surface stormwater runoff, which was needed to
design appropriately sized conveyance structures, was based mainly on empirical
results. Among the tools used was Roe’s table, a simple look-up table indicating the
pipe size required to convey runoff for a given drainage area and slope based on Roe’s
observations of London sewers (Metcalf and Eddy, 1928). Other methods were based
on empirical and, frequently, ill-founded formulas developed by Adams, McMath,
Talbot, and others. The empirical nature of these methods and their development for
site-specific conditions often resulted in very poor estimates and designs when they
were applied under conditions different from those for which they had been devel-
oped. For example, Talbot’s formula predicts the required cross-sectional area of a
drainage pipe as follows (Merritt, 1976):
(1.1)
A = C 4M 3/ 2
where A = required cross-sectional area of drainage pipe (ft2)

C = coefficient ranging from 0.2 for “flat country not affected by accumu-
lated snow or severe floods” to 1.0 for “steep, rocky ground with abrupt
slopes”
M = the drainage basin area (ac)
Talbot’s formula was based on a rainfall intensity of 4 in/hr, but could be adjusted for
other intensities by proportion. The formula does not consider pipe slope or rough-
ness, factors that are now known to greatly influence required conveyance pipe
diameters.
In the second half of the nineteenth century, hydrologic and hydraulic methods were
enhanced considerably through the research of such notable figures as Antoine Chézy
and Robert Manning, who developed uniform flow formulas, and T. J. Mulvaney and
Emil Kuichling, who developed what is known today as the rational method for
stormwater runoff rate estimation. In 1899, Arthur Talbot performed some of the ini-
tial work on rainfall estimation, producing curves that eventually evolved into
present-day intensity-duration-frequency curves.
The Twentieth Century

By the 1920s, the growing accumulation of rainfall records enabled the use of design
storms in which the time variation of rainfall intensity during a storm event was
explicitly acknowledged and accounted for. Additional advancements during the
twentieth century consisted of the development of unit hydrograph theory by L. K.

Sherman in 1932, and studies of rainfall abstractions and infiltration by W. H. Green
and G. A. Ampt in 1911 and by Robert Horton in 1933. By the middle to latter part of
the century, the U.S. Soil Conservation Service (now the Natural Resource Conserva-
tion Service) had developed simple but effective methods for runoff estimation for
both rural and urban areas. Methods for statistical and probabilistic analysis of floods,
droughts, and other hydrologic, hydraulic, and water quality variables also advanced
rapidly during the twentieth century.
With the development of the digital computer came rapid advancements in technical
tools and methods for wet-weather flow management. Included among these
advances were developments of specialized hydrologic and hydraulic computer pro-
grams (see Figure 1.5), as well as more general computational and data processing
technologies such as spreadsheets, databases, pre- and postprocessors, and geographic
information systems. The computational power of computers enabled engineers to
efficiently optimize their designs through the use of advanced mathematical optimiza-
tion and statistical techniques. Whereas just a few decades ago it was uncommon to
find even a single computer in most design engineering offices, a computer is now an
indispensable tool on every engineer’s desk.
Figure 1.5
Computer programs
for designing storm
sewers became
available in the latter
part of the twentieth
century
Perhaps some of the most significant developments that have occurred in the past few
decades have been related to the recognition of the impacts of stormwater runoff on
receiving waters. By the early 1970s, the U.S. Congress had passed the Water Pollu-
tion Control Act, and increasingly strict water quality regulations have since been
developed. To mitigate wet-weather flow impacts on receiving waters and habitats,
methods of control and treatment of urban stormwater runoff are required. Among
these methods are physical, chemical, and biological treatment processes and storage
and treatment combinations. It should be noted, however, that water quality is not the
only wet-weather factor contributing to habitat degradation downstream of urban
areas. Other factors, including channelization, debris and cover removal, and channel
straightening, must be addressed more adequately in the future.
Section 1.3 Overview of Stormwater Conveyance System Design 11
Summary and Outlook

Most of the tools and methods currently used for prediction of runoff discharges and
volumes and for appropriately sizing required conveyance facilities have been devel-
oped within the last 200 years. That ancient civilizations were able to effectively deal
with stormwater runoff indicates that advanced knowledge and sophisticated tools are
not necessary for relatively simple conveyance systems. This is not the case in mod-
ern society, however, where competing infrastructure needs vie for resources and mul-
tiple goals require evaluation of several designs. Computers and automated control
technologies enable modern designers and system operators to optimize and maintain
complex systems and make the design and operation much more efficient than they
would be otherwise.
What the future will bring is always uncertain. A report by the U.S. EPA’s Urban
Watershed Management Branch (Field et al., 1996) describes five areas in which
research is needed. They are
• Characterization and problem assessment
• Watershed management
• Impacts and control of toxic substances
• Control technologies
• Infrastructure improvement
The third of these areas, which deals with pollutants, has received the least attention
in the past. However, it is beginning to receive considerable attention in the research
community and can be expected to become more of an issue in practice in the coming
years. A growing research area that is absent from this list is sustainable development,
which is concerned with preserving water quality for human use and the environmen-
tal preservation into the future.
1.3 OVERVIEW OF STORMWATER CONVEYANCE

SYSTEM DESIGN
Design of a stormwater conveyance system is a trial-and-error process. The “art” of
design involves defining the objectives and determining, from among a number of
feasible options, the one that best meets those objectives. The design engineer opti-
mizes the use of resources (materials, construction labor, operation and maintenance
costs, and so on) to meet the objectives while satisfying certain constraints (such as
project budgets, right-of-way limitations, and design storm criteria).
Design is different from analysis. Design is more complex than analysis. In a design,
the engineer is not limited to the evaluation of any single combination of inputs in
seeking to meet the desired objectives. An analysis, on the other hand, is simply a
technical evaluation of a possible choice to determine whether it meets the objectives
while satisfying the constraints. The design process may be viewed as consisting of
many analyses, each of which determines whether the design is feasible (that is,
whether it satisfies the design constraints) and whether it meets the project objectives.
Finally, from among all the feasible designs, the “best” one is selected.
With this view of the design process in mind, several questions need to be addressed:
• What is (are) the objective(s) of the design?
• What resources need to be allocated?
• What are the constraints that need to be satisfied?
• How does one define the “best” feasible design? What makes one choice bet-
ter than another?
Only when these questions are explicitly acknowledged and addressed can the design
be considered acceptable. The following subsections are intended to clarify these
questions. However, the issues surrounding these questions are specific to each
design.
Design Objectives
The design objectives are the project goals that the engineer attempts to meet as much
as possible without violating the design constraints. There may be more than one
objective associated with the design of a stormwater conveyance system. The most
obvious objective is probably adequate conveyance of stormwater runoff. However, in
the face of specific regulations regarding return periods for which design flows must
be computed, stormwater conveyance capacity is more likely to be viewed as a con-
straint than as an objective.
It is usually not possible to develop a design that is optimal for all objectives. In other
words, if a particular design alternative is optimal for one objective, then it will usu-
ally be less than optimal for the other objectives. Therefore, some weighting or trade-
Section 1.3 Overview of Stormwater Conveyance System Design 13
off analysis to quantify the relative importance of the objectives is necessary. For fur-
ther reading on multiobjective planning, the reader is referred to Loucks, Stedinger,
and Haith (1981); Goodman (1984); and Grigg (1996).
Many possible objectives may be adopted for a particular project. Historically, mini-
mization of the total life-cycle cost of the project has been an attractive one, probably
due in part to the relative ease with which it can be quantified. For cases in which the
engineer is faced with developing a design for which the capital cost may not exceed
an available budget, cost may not be a design objective, but instead a design con-
straint.
Other possible objectives are
• Minimizing flood damages or traffic hazards during a storm event of a mag-
nitude greater than that for which the primary, or minor, storm sewer system
is designed
• Minimizing aesthetic, physical, chemical, and biological impacts to existing
receiving waters
• Maximizing potential recreational and aesthetic benefits provided by wet
detention facilities
The engineer should seek to define the design objectives with input from the entity
funding the project, other relevant regulating and/or operation and maintenance agen-
cies, the general public, and special-interest groups.
Resources
In conveyance system design, resources allocated to a project (or project costs) con-
sist of raw materials and labor for construction. Operation and maintenance resources
(equipment, manpower, and so forth) required over the project lifetime should also be
included. For instances in which natural channels, wetlands, or other areas where eco-
systems are disturbed, the loss of habitat and the associated biotic communities
should also be considered as “resources” allocated to a project.
Other types of resources that should be considered are those related to acquisitions of
rights-of-way and/or temporary or permanent easements. In some cases, it is neces-
sary to purchase or condemn residential or other properties so that they can be demol-
ished to make way for the planned project improvements. In these instances,
resources allocated to the project should include purchase and relocation costs, but
these costs may be offset in part by less frequent flooding in the project locale.
Constraints
As already noted, a limited budget within which a project must be implemented is one
type of constraint. Additional constraints may be geometric in nature, such as limited
pipe burial depths, existing underground utilities that cannot be relocated, and limited
rights-of-way. Other constraints consist of federal, state, and local regulations within
which a project must be developed, as well as laws regarding water rights and flood
protection.
As a general rule, constraints can be of two types. One type of constraint specifies a
requirement that cannot be violated at any cost. The second type of constraint can be
violated, but the cost associated with its violation may be high. For example, if a
design constraint were to say that an existing underground gas transmission line can-
not be moved, the reality is that the line could be moved, but the cost could be prohib-
itive.
What Is Best?
After the engineer identifies a number of feasible designs, the task turns to the issue of
selection of the best design. But how should one define what is best? What makes one
design better than another? The answer depends on many factors, but the degree to
which each design meets the stated project objective(s) is certainly one basis for com-
parison. For example, if cost minimization were the only objective, then the lowest-
cost alternative would be chosen. The design engineer does not necessarily make the
final selection of the design. Often, the engineer’s client or a regulatory body is the
decision-maker. In such instances, the role of the designer is to provide the decision-
maker with options and quantifications of costs, benefits, and impacts. The designer
essentially acts as an advisor on complex technical issues.
Factors that might be considered in judging the suitability of the various choices are
issues such as safety and aesthetics. Other measures might relate to equity in the dis-
tribution of project benefits. For example, if a citywide stormwater drainage improve-
ment project were to make improvements only in new or politically powerful districts,
while overlooking needs in poor and disadvantaged areas, matters of fairness and
equitability would not be satisfied.
The design process as outlined by this section may seemingly imply that the storm
sewer and culvert design process can be viewed as a mathematical optimization prob-
lem (Loucks, Stedinger, and Haith, 1981). That approach, however, is usually not for-
mally applied in design practice because the degree of quantification necessary to do
so is usually not possible. The practitioner must instead rely heavily on experience
and knowledge of the problem at hand to identify suitable alternatives for detailed
evaluation. Nevertheless, the general idea of optimizing the use of resources to
achieve adopted objectives within certain constraints does succinctly characterize the
overall design process. After the preferred design is identified through this heuristic
process, one can prepare detailed construction plans and specifications.
1.4 CHAPTER SUMMARY

The need to protect against flooding is driven primarily by the loss of life and prop-
erty that flooding causes. Most of the losses can be associated with major flooding of
streams and rivers, and the stormwater conveyance systems described in this text have
a limited effect on events of this magnitude. However, these primary systems signifi-
cantly reduce the damages caused by smaller but more frequent events.
In modern society, storm drainage design is complicated by competing infrastructure
needs and goals that require careful balancing of resources and evaluation of multiple
References 15
possible designs. Computers have greatly simplified this process. Considerations such
as water quality and the need to deal with pollutants are receiving more attention in
recent years.
The design of stormwater conveyance systems is an iterative process that involves
defining project objectives and determining the design alternative that best meets
them, but does not exceed available resources or violate design constraints. The
design process can be viewed as a series of analyses, or technical evaluations of indi-
vidual design alternatives. The “best” design is then selected from among these alter-
natives.
REFERENCES
Blake, M. E. 1947. Ancient Roman Construction in Italy from the Prehistoric Period to Augustus. Washing-
ton, D.C.: Carnegie Institution of Washington.
Burian, S. J., S. J. Nix, S. R. Durrans, R. E. Pitt, C-Y. Fan, and R. Field. 1999. “Historical Development of
Wet-Weather Flow Management,” Journal of Water Resources Planning and Management 125, no. 1:
3–13.
Chesbrough, C. S. 1858. “Chicago sewerage report of the results of examinations made in relation to
sewerage in several European cities in the winter of 1856–57.” Chicago, Illinois.
Field, R., M. Borst, M. Stinson, C-Y. Fan, J. Perdek, D. Sullivan, and T. O’Connor. 1996. Risk Management
Research Plan for Wet Weather Flows. Edison, N. J.: U.S. Environmental Protection Agency.
Goodman, A. S. 1984. Principles of Water Resources Planning. Englewood Cliffs, N. J.: Prentice Hall.
Grigg, N. S. 1996. Water Resources Management: Principles, Regulations, and Cases. New York:
McGraw-Hill.
Loucks, D. P., J. R. Stedinger, and D. A. Haith. 1981. Water Resource Systems Planning and Analysis,
Englewood Cliffs, N. J.: Prentice Hall.
Merritt, F. S., ed. 1976. Standard Handbook for Civil Engineers, 2d ed. New York: McGraw-Hill.
Metcalf, L., and H. P. Eddy. 1928. Design of Sewers. Vol. 1 of American Sewerage Practice. New York:
McGraw-Hill.
National Science and Technology Council. 1997. Committee on Environment and Natural Resources: Fact
Sheet, Natural Disaster Reduction Research Initiative.
National Research Council. 1991. Opportunities in the Hydrologic Sciences. Washington, D.C.: Committee
on Opportunities in the Hydrologic Sciences, Water Science and Technology Board, National Academy
Press.
National Weather Service. 2000. “Flood Losses: Compilation of Flood Loss Statistics.” National Weather
Service, Office of Hydrologic Development. www.nws.noaa.gov/oh/hic/flood_stats/
Flood_loss_time_series.htm.
van der Vink, G., R. M. Allen, J. Chapin, M. Crooks, W. Fraley, J. Krantz, A. M. Lavigne, A. LeCuyer, E.
K. MacColl, W. J. Morgan, B. Ries, E. Robinson, K. Rodriquez, M. Smith, and K. Sponberg. 1998.
“Why the United States is Becoming More Vulnerable to Natural Disasters,” EOS, Transactions of the
American Geophysical Union 79, no. 44: 533–537.
Workers construct
a large, triple-
barrel box culvert
(top). Low flows
are temporarily
diverted around the
disturbed area
through bypass
piping (bottom).
C H A P T E R
2
System Components, Models,
and the Design Process
Stormwater is managed with a network of structures designed to collect, convey,

detain, treat, and discharge runoff. Examples of these structures are storm sewer con-
duits, culverts, drainage ditches, and detention ponds. Generally, these structures act
in a passive mode, relying on the force of gravity to move water. Occasionally, how-
ever, stormwater is actively managed with devices such as pumps, gates, or valves.
In this book, models are defined as representations of the physical behavior of hydro-
logic and/or hydraulic systems. They may be physical, electric-analog, or mathemati-
cal in nature. Mathematical models are based on the laws of physics and typically
apply the principles of conservation of mass, energy, or momentum. Computer mod-
els (such as those on the CD-ROM accompanying this book) are collections of
automated mathematical models.
Models are used to predict how stormwater structures will perform. The basic
elements that must be considered when modeling stormwater management systems
are hydrology (specifically, watershed response to storms), hydraulic capacities of the
system structures, and upstream and downstream impacts (such as the effect of tailwa-
ter elevation on storm sewer system hydraulics or the effect of a storm sewer system
on downstream discharges).
Engineers use models in the design process to simulate the performance of a
stormwater collection and conveyance system under a range of flows and conditions.
The tradeoffs between size and performance can then be evaluated. Hydraulic
performance in terms of resulting flows, headwater elevations, exit velocities, and so
forth is the primary focus of this text. Other factors that must be considered in
selecting system components are construction cost, structural loads, maintenance
requirements, aesthetics, seepage flow (from or into the groundwater), and biological
habitat. The final design is based on compliance with regulations, standards, and,
ultimately, the professional judgment of the engineer.
18 System Components, Models, and the Design Process Chapter 2
Sections 2.1 and 2.2 of this chapter describe stormwater conveyance system compo-
nents and modeling approaches for the purpose of assisting those new or inexperi-
enced in the field of stormwater management in getting a brief overview of basic
concepts, common terminology, and analysis methods. They also serve as a “road
map” for the book by directing the reader where to find further information on various
topics. The stages of the design process, including master planning, preliminary
design, and detailed design, are presented in Section 2.3. Section 2.4 provides a brief
discussion of the integration of computerized stormwater models with CAD and GIS.
The chapter concludes with an overview of model input data sources in Section 2.5.
2.1 Hydrologic Components

Hydrology is the study of the properties, distribution, and effects of water as it cycles
through the earth’s surface, subsurface, and atmosphere. This book is concerned with
hydrology in terms of loading stormwater management systems. Coverage of the
topic is therefore limited to quantifying precipitation and runoff for that purpose.
This section briefly describes the modeling of rainfall and runoff. These topics are
covered in detail in Chapters 4 and 5.
Precipitation
To design or evaluate a storm sewer system, the engineer must determine the recur-
rence frequency of the storm event(s) that the system should be designed to handle.
These events, also called design storms, will typically be dictated by local practice or
regulations. In other instances, it may be up to the engineer to determine what storm
events should be used given the type of project, its impact on the surrounding area,
and the consequences of system failure.
Rainfall intensity varies with time over the course of a storm event. Design storms are
frequently computed by applying a total storm rainfall depth to a synthetic rainfall
distribution (see Section 4.5, page 91). Rainfall distributions may also be developed
from actual precipitation records for an area. Evaluation of multiple storm events for a
single design is often desirable, especially if stormwater will be detained or pumped.
Once the design storms are selected and developed, rainfall data can be combined
with drainage basin characteristics to determine runoff volumes and discharges.
Runoff
Runoff or rainfall excess is the portion of the rainfall that is not lost to interception,
evapotranspiration, or infiltration (see Section 5.1, page 108). It “runs off” over the
ground surface and into streams, ditches, and other stormwater management struc-
tures. Thus, runoff is the hydrologic component of primary interest in stormwater
conveyance modeling.
The methods and models available for computing runoff are numerous. They range
from simple peak discharge calculations for sizing pipes to more complex hydrograph
Section 2.2 Hydraulic Components 19
methods that account for total runoff volume and the change in discharge over time.
The rational method is an example of the former; the unit hydrograph method is an
example of the latter. Methods for computing runoff from snowmelt, which may be a
consideration in some cases, are also available.
Spreadsheets can assist in runoff hydrograph calculation, and many computer pro-
grams for automating the calculations are also available. Computer programs such as
HEC-HMS and PondPack (Haestad, 2003a) offer several models from which the user
can choose.
2.2 Hydraulic Components

Many types of structures are used in stormwater management systems. A conveyance
structure may be as simple as an earthen ditch or as complex as an extensive system of
inlets and subsurface piping that delivers water to a storage facility from which it is
pumped to a treatment plant or receiving stream.
The subsections that follow describe some of the basic properties of open channels,
culverts, storm sewer systems, and detention/retention facilities. Brief descriptions of
commonly used materials and applicable mathematical models are included, as well
as examples of computer software available for model calculations. Table 2.1, which
is located at the end of this section, summarizes the various stormwater system com-
ponents, methods used in modeling these components, and typical model input
requirements. It also provides text references for obtaining more detailed information.
Open Channels
Open channels include natural stream channels, as well as natural swales (depres-
sions) along which water runs following rainfall events. In the constructed or urban-
ized environment, open channels include street gutters, drainage ditches, pipes, lined
or unlined stormwater collection channels, and natural or modified stream channels.
Figure 2.1 shows an example of flow in an open channel with a trapezoidal cross
section.
Flow in a pipe is classified as open-channel flow if there is a water surface at atmo-
spheric pressure (that is, a free water surface). For the purposes of this book, the term
open-channel flow refers to any flow having a free water surface, regardless of
whether it occurs in a closed conduit such as a culvert or storm sewer, aboveground
channel, or other structure. Open-channel flow theory is presented in detail in Chapter
7. However, when the text refers simply to an open channel in the context of physical
characteristics or design approaches, as in Chapter 8, the term is meant to describe
nonpipe conveyances only.
Figure 2.1
Open-channel flow
Prismatic versus Nonprismatic Channels. Open channels exist in a wide

variety of cross-sectional shapes. For analytical purposes, channels may be classified
as prismatic or nonprismatic. In a prismatic channel the variables that describe the
channel geometry—such as its width, side slopes, and longitudinal slope—remain
constant along the length of the channel. In a nonprismatic channel one or more of
these geometric variables changes along the channel length. Natural stream channels,
whose widths and other channel properties are variable, are examples of nonprismatic
open channels. Treatment of open-channel flow in this text deals primarily with pris-
matic, man-made channels. Nevertheless, certain aspects associated with the hydrau-
lics of flow in nonprismatic channels will be addressed where appropriate.
Materials. Open channels can also be classified on the basis of whether they are
lined or unlined. The bottom and sides of an unlined channel consist of natural geo-
logic materials such as earth, gravel, or rock. In a lined channel, the channel bottom
and sides are covered with an erosion-resistant material such as concrete, asphalt,
riprap, or vegetation.
As shown in Chapter 7, the channel’s lining material (or lack thereof) affects hydrau-
lic performance characteristics such as flow depth and velocity. In turn, these same
characteristics affect the erosive forces of the flow, thereby influencing the lining
selection. The selection of the channel lining material is therefore a significant part of
the hydraulic design process.
Model Representation. To model open-channel flow, the engineer must know
the flow condition. With steady flow, flow characteristics such as discharge and veloc-
ity are constant over time at a given point in the channel. When these characteristics
do vary over time as a point, the condition is unsteady.
Steady flow can be further classified as either uniform flow or varied flow. Uniform
flow is a condition in which the depth, velocity, cross-sectional area, and discharge are
constant along the channel length. As a practical matter, this can happen only in a
prismatic channel and only if the flow is steady. In uniform flow, the depth of flow is
called normal depth.
Two models that can be used to describe uniform flow are the Manning equation and
the Chézy equation. In their simplest forms, either of these equations may be used to
calculate velocity or discharge as a function of channel slope, roughness and geome-
try. In design, these models are used in a variety of ways. For example, for a given
channel geometry, slope, roughness, and depth, the discharge may be calculated.
Alternatively, if the channel geometry, material roughness, discharge and maximum
allowable depth are specified, the required slope can be calculated. An example of a
computer model used for uniform flow calculations is FlowMaster. Modeling of uni-
form flow is covered in more detail in Section 7.3.
Varied flow is a condition in which the depth and velocity of flow change along the
length of the channel. If the depth and velocity change only gradually, the flow is said
to be gradually varied flow (GVF). If the depth and velocity change quickly over a
short distance, as in a hydraulic jump, the flow is said to be a rapidly varied flow
(RVF). Gradually varied flow occurs in all nonprismatic channels and in prismatic
channels under the influence of a flow control other than normal depth. Because the
curvatures of flow streamlines are small in a gradually varied flow, the pressure forces
can be assumed to be hydrostatic. Constitutive relationships developed for uniform
flow, such as the Manning equation, can be assumed to be valid for computing friction
losses in gradually varied flow.
Because of the variable channel geometry and the presence of flow controls, the data
required for gradually varied flow analysis are considerably more complex than for
uniform flow analysis. The input includes channel cross-section geometry, reach
lengths, channel roughness, and the water surface elevation at the control section.
HEC-RAS is an example of a computer program frequently used in gradually varied
flow profile calculations. For more information on gradually varied flow calculations,
see Section 7.8.
Channel routing is a procedure by which the outflow at the downstream end of a

channel reach is computed from inflow data and channel characteristics. It accounts
for the effects of channel storage and travel time on discharge rates; thus, it is a type
of unsteady flow modeling. Channel routing is applicable to both open channels and
sewer systems when channel storage and/or travel time affect flow rate to a significant
degree.
Channel routing may be accomplished by using either hydrologic or hydraulic routing
methods. Hydrologic routing is computationally simpler and less data-intensive, but
usually is also less accurate. Hydrologic routing algorithms are based on the physical
principle of conservation of mass and on an assumed storage relationship. Hydraulic
methods are more physically based and employ conservation of mass and conserva-
tion of momentum. Their use requires considerable data on channel geometry and
roughness variables and on initial and boundary conditions of the flow itself. The
Muskingum, Modified Puls, and Convex routing methods are examples of hydrologic
routing models. (See Section 5.8 and Appendix C.) Hydraulic routing involves a solu-
tion of the Saint-Venant equations, and may consist of kinematic, diffusion, or
dynamic wave routing. Such methods are not well-suited to manual calculations but
are coded into software packages such as HEC-RAS.
Culverts
A culvert is a relatively short underground water conveyance conduit. The primary
purpose of a culvert, like that of a bridge, is to provide a means whereby the water in
a stream or other open channel can pass through an obstruction such as a highway or
railway embankment. Figure 2.2 shows a small culvert passing under a driveway.
Materials. Culverts come in a variety of shapes, sizes, and materials. As shown in
Figure 2.3, cross-sectional shapes associated with culverts include circular, box, ellip-
tical (horizontal or vertical orientation), and arched.
Common materials used for culvert design are reinforced precast or cast-in-place con-
crete and corrugated steel. Other materials include corrugated aluminum, polyethyl-
ene, and polyvinyl chloride (PVC). Discussions of these materials, including available
shapes and sizes and sources of further information, are presented in Section 9.2.
The inlet and/or outlet ends of a culvert often receive special treatment. A culvert end
may simply project from the embankment, or it may be mitered to conform to the
embankment slope. It may also be fitted with a special flared end section or with a
headwall and wingwalls. Section 9.3 discusses various types of end treatments and
their hydraulic performance characteristics, as well as special inlet configurations that
can be used when allowable headwater depths are limited.
Model Representation. A complete theoretical analysis of the hydraulics of a
particular culvert installation is complex. Flow conditions vary from one culvert to the
next, and they also vary over time. The barrel of the culvert may flow full or partially
full, depending on upstream or downstream conditions, barrel characteristics, and
inlet geometry. A commonly used method of analysis of culvert hydraulics was devel-
Figure 2.2
A corrugated metal
pipe culvert under a
driveway
oped by the Federal Highway Administration and published as “Hydraulic Design

Series No. 5: Hydraulic Design of Highway Culverts,” often referred to as HDS-5
(Norman, Houghtalen, and Johnston, 2001). Modeling software such as CulvertMas-
ter follows the calculation methods set forth in HDS-5.
The following list describes the principal terms and concepts in culvert hydraulics
More detailed information on culvert hydraulic analysis and HDS-5 methodologies is
provided in Section 9.5 and Section 9.6.
• Headwater depth is the depth (relative to the culvert’s upstream invert eleva-
tion) of flow just upstream of the culvert entrance. Any increase in energy
required to push an increased discharge through a culvert translates into a
greater headwater depth.
• Tailwater depth is the depth of water just downstream of the culvert outlet.
The tailwater surface elevation may be dictated by downstream channel
characteristics, by obstructions, or by a receiving-water elevation. An exit
condition in which the tailwater depth is significantly less than the depth of
flow in the culvert, and thus does not affect the upstream hydraulics, is called
a free outfall.
Figure 2.3
Common cross-
sectional shapes for
culverts
• The flow condition in a culvert can be characterized as either full (pressure)

flow or partially full flow with a free water surface. Both free-surface and
pressure flow can occur simultaneously at different locations within the
same culvert. Uniform flow and gradually varied flow, as discussed earlier
for open channels, apply to free-surface flow in culverts as well.
• The flow in a culvert may be controlled by conditions at either the inlet or
the outlet of the culvert, so the type of control is classified as either inlet
control or outlet control. In the case of inlet control, the hydraulic capacity of
the culvert entrance limits the amount of water conveyed for a given headwa-
ter depth; therefore, hydraulic characteristics downstream of the inlet control
section do not affect the culvert capacity. In an outlet control condition, the
culvert barrel capacity or downstream tailwater elevation limits the amount
of flow that can be conveyed by the culvert for a particular headwater eleva-
tion. The control section for outlet control is located at the barrel exit or fur-
ther downstream.
• The flow velocity at the exit of a culvert is typically higher than that of the
stream channel into which the culvert discharges. High outlet velocity can
cause streambed scour and bank erosion in the vicinity of the culvert outlet.
Minor problems occasionally can be avoided by increasing the barrel rough-
ness, but structures such energy dissipators and/or outlet protection such as
riprap are often necessary.
Storm Sewer Systems

A storm sewer system is a network of primarily subsurface structures for the collec-
tion and conveyance of stormwater runoff. A typical storm sewer system consists of
inlets through which stormwater enters from the land surface, pipes that convey this
water, junction and manhole structures that serve as connection points for pipes and
provide access to the subsurface system, and outlets or outfalls where stormwater is
discharged to treatment facilities, to detention/retention structures, or into a larger
conveyance such as an open channel.
Components and Materials. Storm sewer systems almost always include
gutters, inlets, manholes, pipes, and outlet structures. Some systems may include stor-
age and/or pumping facilities and other structures such as weirs. The following sub-
sections briefly describe each of these components.
Although roadway gutters are not always considered a part of a stormwater manage-
ment system, they are in fact an integral part of the system and are responsible for
conveying runoff that collects in streets to inlets, where it enters the subsurface con-
veyance system. A gutter may be formed by a curb along the edge of a street, or, for
roads without curbs, it may consist of a shallow swale along the road edge. The vari-
ous gutter types are presented in Section 10.1.
Figure 2.4 is a photo of a concrete curb-and-gutter section. The width of flow in the
gutter is called the spread. Depending on the roadway classification and posted speed
limits, allowable spread widths may extend beyond the roadway shoulder width and
into the travel lane.
Curbs are typically composed of either precast, poured-in-place, extruded, or asphalt
concrete. In some cases, a depressed section of gutter may be integrated with the con-
crete curb, and the remainder of the gutter section is composed of the roadway pave-
ment material.
A stormwater inlet is a structure for intercepting stormwater on the ground surface or
in a roadway gutter and conveying it to the subsurface storm sewer piping system.
Common inlet types are grate inlets (Figure 2.5), curb inlets (Figure 2.6), and combi-
nation inlets (Figure 2.7).
For each of these three types, inlets can be further described as either continuous-
grade inlets or sag inlets. A sag inlet is constructed in a low point where water drains
to the inlet from all directions. With a continuous-grade inlet (also referred to as an
inlet on grade), surface water may flow away from the inlet opening in one or more
directions.
Figure 2.4
Concrete curb and
gutter
Below the ground surface, inlets consist of box-like structures that support the inlet
opening (for example, the grate) and connect to the system piping. The inlet box is
typically constructed of precast reinforced concrete. Other materials include poured-
in-place concrete, corrugated metal, brick, and block. A catch basin is a special type
of inlet box designed to trap sediment and debris.
Although not considered inlets per se, runoff from streams and channels can enter
storm sewer systems through culvert-type entrances such as headwalls. Additional
specialized inlet types are discussed in Section 10.2.
Figure 2.5
A grate inlet in a
triangular gutter
section
Figure 2.6
A large curb inlet
Figure 2.7
A combination inlet
Manholes, sometimes called access holes, are usually installed in sewers at locations
where pipes join, change direction, or change size. They provide access to the system
for inspection and maintenance. Manholes for relatively small-diameter sewers
generally consist of a concrete or masonry base, which ideally has been formed to aid
in reducing the energy loss associated with water traveling from the inlet pipe(s) to
the outlet pipe. Figure 2.8 shows a precast concrete manhole structure prior to
installation. Junction structures, like manholes, are used where two or more pipes join
one another. However, junction structures tend to be associated with larger or more
complex installations and are often custom-made.
For storm sewer pipes, circular pipe is the most typical cross-sectional shape, but, as
with culverts, a variety of cross-sectional shapes are available (see Figure 2.3). Ellipti-
cal and arch-shaped cross sections may be used for instances in which vertical or hor-
izontal clearances are limited. Rectangular concrete box sections are also available.
Concrete pipe (shown in Figure 2.9) is available in both nonreinforced and reinforced
varieties and in a range of wall thicknesses and strength classes and is probably the
most commonly used pipe material for storm sewer construction.
Figure 2.8
A precast manhole
structure
Corrugated metal pipe (or CMP) is also used frequently in storm sewers (see Figure
2.10). Like concrete pipe, CMP is manufactured in a number of cross-sectional shapes
and sizes and in strict accordance with standard specifications. Corrugated steel prod-
ucts include circular and arch pipes, structural plate pipe-arches, and structural plate
arches, with the last requiring bolted assembly in the field. Various coatings and steel
thicknesses are available.
Although concrete and corrugated steel pipes are by far the most common in storm
sewer and culvert applications, other materials, such as corrugated aluminum, poly-
ethylene, and polyvinyl chloride (PVC), are used as well. PVC pipes are shown in
Figure 2.11.
Figure 2.9
Concrete pipes
Figure 2.10
Corrugated metal pipe
(CMP)
Storm sewers may outfall to the ground surface, into a natural or man-made body of
water, or into detention or treatment facilities. Exit velocities at storm sewer outfalls
are often large enough to cause erosion problems and potential undermining of the
outfall pipe. Outlet structures dissipate the excess energy of the flow and prevent
undermining. An outlet structure may be as simple as a concrete headwall with a
riprap and filter blanket downstream of the outfall point, or it may consist of a more
complex energy dissipator or stilling basin.
Figure 2.11
PVC pipes
In most cases, stormwater management systems are designed as gravity systems.

Pumping of stormwater is generally undesirable because it significantly increases the
cost of stormwater conveyance but cannot be avoided in some instances. For storm-
water pumping purposes, the most common types of pumps are axial-flow pumps,
radial-flow pumps, and mixed-flow pumps. Both wet-well and dry-well installations
are found in stormwater pumping applications. Chapter 13 provides information on
the design of stormwater pumping systems.
Model Representation. Many components are involved in modeling a storm
sewer system, and various levels of model complexity exist for their analysis. The
hydraulic design or analysis of a storm sewer can typically be broken down into two
major parts: surface hydraulics (that is, gutter flow and inlet capacity) and subsurface
hydraulics (pipe capacity and energy losses). Other possible components are pump
stations and storage facilities.
Levels of analysis that may be performed on storm sewers include:
• Use of a steady-state model that ensures the system can handle peak flows.
• Use of an extended-period simulation (EPS) analysis that applies hydrologic
routing and that may also simulate inter-storm time periods.
• Use of a dynamic model that solves the full Saint-Venant equations.
A steady-state analysis is typically sufficient for design applications that do not

involve system flooding or storage facilities and for which the peak flows from vari-
ous subbasins arrive at the outlet at roughly the same time. This type of analysis is the
primary focus of this text. The paragraphs that follow describe basic steady-state
model requirements for surface systems, subsurface systems, and pumping facilities.
The surface portion of a storm sewer system consists of gutters and other open chan-
nels, and inlet structures. Gutter hydraulics are typically modeled as uniform flow in
an open channel. Specialized forms of the Manning equation have been developed to
facilitate this calculation and are given in Section 10.1. The data required are the dis-
charge, the geometry of the street (longitudinal and transverse slopes) the height of
the curb, and the roughness of the pavement.
Inlets in sag locations operate as weirs for shallow flows and as orifices for greater
flow depths. The forms of the weir and orifice equations used for computing the
capacities of curb-opening and grate inlets are presented in Section 10.2 of this text.
For grated inlets, information on the grate perimeter and open area is required. For
curb inlet calculations, the dimensions of the curb opening and the local depression
are necessary.
Calculations for inlets on grade are somewhat more complex, as all of the flow that
comes toward the inlet may not be intercepted. A portion of the flow in the gutter,
termed bypass flow, may flow over or around the inlet and continue downgradient.
The calculation of intercepted flow and bypass flow is performed using empirical
methods that take into account the velocity of the flow in the gutter relative to the inlet
length, the amount of inlet depression, and the inlet’s splash-over velocity (for grate
inlets, the critical velocity at which a portion of the flow begins to splash over the
inlet). Section 10.2 provides the equations describing the capacity of inlets on grade.
Inlet bypass flows form a surface network for stormwater runoff similar to the way
sewer piping forms a subsurface network. Manual calculations for inlet capture and
bypass can be quite tedious if a project has a large number of inlets on grade. Some
computer programs used for computing inlet and gutter hydraulic characteristics are
simple equation-solvers that look at each structure individually. An example of this
type of program is FlowMaster, which is useful for performing quick calculations on
small systems.
Other computer programs allow the user to develop a network of on-grade inlets,
inlets in sag, and gutters. These programs analyze the surface system using the same
methodologies as the simpler programs, but alterations in the loading of gutters and
inlets due to upstream bypass flows are automatically taken into account. Typically,
this type of program will also perform the subsurface hydraulic calculations. Many
software packages are available for this application, but one popular package is
StormCAD (Haestad, 2003b).
Flow in the subsurface system is modeled as one of two flow conditions (see
Table 2.1). When a pipe is not full and the water surface is open to the atmosphere, the
flow is modeled as open-channel flow. A friction loss equation such as the Manning
or Chézy equation may be used to determine flow depths and velocities. When a pipe
is flowing full (that is, it is surcharged), closed conduit pressure flow calculation
methods are used. The data requirements for these calculations are discharge, pipe
size, roughness, length, and slope.
In a storm sewer network, the minor energy losses that occur at inlets, at locations
where pipe size or alignment changes, and at outlets can be significant. Data require-
ments for computing these losses include the geometry of the structure and the
entrance and exit velocities. Several methods for computing minor losses are available
(see Sections 6.3 and 11.4), and computer models such as StormCAD can compute
these losses as part of a network analysis.
The operating characteristics of a pump, when installed in a piping system, can be
ascertained from graphs of pump and system curves. These characteristics include the
energy head produced by the pump, the corresponding pump discharge, the brake
horsepower requirements for the pump motor(s), and the net positive section head
required to avoid cavitation damage to the impeller and other internal parts. Methods
for determining these operating characteristics, for both single- and multiple-pump
installations, are presented in Section 13.3.
Stormwater Detention and Retention

A common objective in stormwater management is to maintain the peak runoff rate
from a developing area to a value no greater than the predevelopment rate. This prac-
tice will reduce local flooding, soil erosion, sedimentation, and pollution. Local gov-
erning bodies typically establish specific design criteria for peak flow attenuation.
Detention basins reduce the magnitude of peak flows by temporarily storing the run-
off and releasing it over an extended period of time. Additionally, ponds often func-
tion as sedimentation basins, reducing the concentration of suspended materials in the
discharge.
Detention Basin Types and Components. Wet ponds are designed to
maintain a minimum water depth between storms. Dry ponds are designed to com-
pletely discharge stored runoff between storms. The configuration of a detention pond
is generally dictated by site conditions. A dry detention pond is shown in Figure 2.12.
The most common detention facility designs are earthen basins or ponds located
downstream of the collection area and upstream of the discharge point. The detention
pond’s outlet structure controls the rate at which runoff may discharge from the facil-
ity. Common outlet structure devices include weirs, orifices, culverts, inlet boxes, and
standpipes. Most ponds are also equipped with an emergency overflow spillway to
handle flows if the primary outlet becomes clogged, or if a storm has a magnitude
greater than the pond’s design event.
Model Representation. Detention ponds are modeled using the routing proce-
dures described in Section 12.7. The basic data requirements are an inflow
hydrograph, the depth to volume relationship of the pond, and the hydraulic charac-
teristics of the outlet structure.
Figure 2.12
Dry detention pond
Capabilities of computer programs for stormwater detention modeling vary widely.

The most basic programs allow for hydrograph input into a single pond with a simple
outlet such as an orifice or weir, and the pond is routed using a technique such as
storage-indication routing or numerical integration. More powerful models, such as
PondPack (Haestad Methods, 2003a), can analyze networked watersheds and multiple
ponds, interconnected ponds, and complex outlet structures. Fully dynamic stormwa-
ter modeling programs allow storage facilities to be modeled together with other ele-
ments of the stormwater conveyance system.
Summary of Hydraulic Models for Various

Components of Stormwater Management Systems
A number of computer modeling software programs are available for virtually all of
the methods presented in this text, and some examples of these programs have been
mentioned in this chapter. However, because of the wide variety of structure types and
available calculation techniques used in the field of stormwater management, there is
great variation and specialization among the available programs. The primary thrust
of this text as it relates to computer modeling is to aid the reader in understanding the
methodologies underlying the programs. Such understanding is necessary if the engi-
neer is to apply the computer programs correctly and confidently.
Table 2.1 lists the numerical models frequently used to describe the behavior of vari-
ous components of stormwater conveyance systems, the input data required by these
models, and references to the sections of this book where each model is discussed. It
should be noted that most software programs can implement a collection of models
from this list. For instance, modeling software for storm sewer systems would likely
include solvers for uniform, gradually varied, and pressure flow; gutter flow; inlet
performance; and minor losses in manholes.
Table 2.1 References for common stormwater conveyance system component models
Component Modeling Methods Typical Input data Section of Text

Channel with uni- Manning equation, Chézy Channel roughness 7.3
form flow equation, or Darcy-Weisbach Channel geometry
equation Channel slope
Discharge or velocity
Channel with gradu- Direct-step method or Geometry of each cross section 7.8
ally varied flow standard-step method Starting water surface elevation
Channel with flow Muskingum Historical flood data or routing 5.8
routing coefficients
Inflow hydrograph
Modified Puls Geometry of outlet 5.8
Inflow hydrograph
Saint-Venant equations Channel geometry Appendix C
Inflow hydrograph
Culvert with inlet HDS-5 methodology Inlet geometry 9.5
control Discharge
Culvert with outlet HDS-5 methodology Pipe geometry 9.5
control Pipe material
Inlet geometry
Discharge
Tailwater elevation
Roadway overtop- HDS-5 methodology Roadway profile coordinates 9.6
ping Type of road
Roadway width
Discharge
Gutter flow Manning equation or Chézy Roadway slopes 10.1
equation Curb height
Pavement roughness
Discharge
Storm sewer pipes Manning equation or Chézy Pipe roughness 7.3; 7.4; 7.5; 7.8
with open channel equation; Saint-Venant Pipe geometry 11.5; Appendix C
flow equations Pipe slope
Storm sewer pipes Darcy-Weisbach, Pipe roughness 6.2; 6.3; 11.5;
with pressure flow Manning, Hazen-Williams, or Pipe geometry Appendix C
Chézy equation; Saint-Venant Pipe slope
equations Discharge or velocity
Tailwater conditions
Inlets Weir and orifice equations Discharge 10.2
Inlet type and geometry
Manholes Minor loss equations Entrance and exit velocities 6.3; 11.4
Structure geometry
Diversions/over- Weir and/or orifice equations Geometry 7.6; 7.7
flows Water surface elevations
Pumps Energy equation Pump head-discharge curves 13.3; 13.7
System equilibrium System curve(s)
System outlets Weir equations Boundary conditions such as tail- 6.3; 9.5; 11.7
Orifice equations water data or downstream chan-
Minor loss equations nel characteristics
Energy dissipation model Geometric characteristics of outlet
Detention ponds Hydrologic routing and Inflow hydrograph 12.6; 12.7
Hydraulic routing Depth vs. volume for pond
Geometry of outlet or stage-
discharge curve
Note: This table is repeated in Appendix D. That version lists the software program on the accompanying
CD-ROM that can be used with each type of model component.
2.3 THE DESIGN PROCESS

The urban stormwater infrastructure can be viewed as consisting of a minor system
and a major system. The minor system is designed to handle frequent events with
return periods typically on the order of 2 to 10 years. The minor system consists of,
for instance, roadway gutters, inlets, and pipes.
The major system consists of the pathways taken by flows in excess of the capacity of
the minor system. It can be thought of as inundated roadways, swales and depres-
sions, and natural and man-made open channels. A major system always exists, even
when a minor system does not. Unfortunately, major systems are often neglected and
do not receive the attention that they warrant, especially in light of the high probabil-
ity of operational failure of the minor system.
The overall process of designing a storm drainage system can be broken down into a
number of phases. These phases are
1. Master planning
2. Concept and preliminary design development
3. Detailed design
4. Preparation of construction drawings, specifications, and contract documents
Each of these phases is described more fully in the sections that follow.
Master Planning
Master planning, often conducted for an entire urban area by a drainage-system
authority or review agency, provides a holistic view of the urban drainage system and
how its various components interact. The need for master planning is particularly
acute where an urbanized area spans a number of different political jurisdictions, each
having its own objectives and design criteria. In such instances, master planning can
help ensure that systems that cross jurisdictional boundaries are consistent with one
another. Even if only one political jurisdiction is involved, master planning provides
useful guidance for constructing new drainage systems that are consistent not only
with one another, but also with existing facilities.
The issues that should be addressed by a master plan are, at a minimum, delineations
of major urban drainageways (whose floodplains might be mapped for flood insur-
ance purposes) and approximate limits and locations of storm sewers in the contribut-
ing drainage basin(s). The master plan should make a clear distinction between the
minor and major drainage systems, and it should also take whatever steps are politi-
cally and legally possible to ensure that the major system is functional and provides a
reasonable degree of protection from severe storm events.
Technically, a master plan should contain enough detail to be an effective guide to the
ultimate construction and/or rehabilitation of storm drainage facilities in the area cov-
ered by the plan. Hydrologic analyses should have sufficient detail to provide reason-
able estimates of required system conveyance capacities and should address estimates
of expected future flows in developing drainage basins. Zoning master plans, where
Section 2.3 The Design Process 37
available, are invaluable for providing estimates of ultimate land uses in a drainage
basin.
The master plan report and other documents should clearly state the design storm-
recurrence intervals for which planned drainage facilities are to be designed and
should be referred to frequently as individual components of the overall system are
designed and constructed. A listing of typical design storm return periods, as recom-
mended by the American Society of Civil Engineers, is presented in Table 2.2.
Table 2.2 Typical return periods for design of stormwater drainage systems (ASCE, 1992)
Return Period
Land Use (years)
Minor drainage systems
Residential 2 to 5
High-value generation commercial 2 to 10
Airports (terminals, roads, aprons) 2 to 10
High-value downtown business 5 to 10
Major drainage system elements Up to 100
In areas where existing storm-drainage facilities require rehabilitation and/or retrofit-

ting, and especially in cases where additional right-of-way acquisitions may be
required, the master plan should identify such needs and should provide estimates of
the costs associated with the upgrades.
Preliminary Design Development

As new subdivision developments are built or as drainage-improvement projects are
begun, development of preliminary designs should precede detailed final design activ-
ities. Preliminary designs should be consistent with the broad system outlines estab-
lished during the master planning phase, but they should provide additional detail on
locations of individual drainage structures and features. In the case of a new subdivi-
sion, for example, the preliminary design can be indicated on a preliminary plat show-
ing the proposed street and parcel layout of the development. This preliminary design
provides an opportunity for jurisdictional authorities to review the consistency of the
proposed facilities with the broader-scale master plan and ensure that both minor and
major drainage-system considerations have been dealt with.
Activities associated with preliminary design entail more detailed hydrologic analyses
than those conducted for master planning purposes. They should also explicitly
account for how the proposed facilities fit into larger-scale regional plans. Design
alternatives should receive detailed evaluation in this phase, as should methods for
dealing with upstream flows and discharges to downstream areas.
At a minimum, a preliminary design drawing and report should consist of topographic
mapping on which streets, land parcels, and proposed storm-drainage facilities have
been superimposed. The storm-drainage layout should indicate pipe, manhole, and
inlet locations, as well as locations of other features such as detention and/or retention
ponds. Drainage basins contributing to each inlet should be shown and referenced to
accompanying hydrologic estimates of design flows in the report. Finally, the prelim-
inary design report should explicitly address the major drainage system, including the
general pathways and directions of major system flow on the design drawing.
Detailed Design
Following the approval of the preliminary design by the appropriate jurisdictional
authorities, the engineer can proceed with the detailed design. As the name implies,
this phase involves detailed engineering analyses that will serve as the basis for the
design described by the construction plans and specifications.
Detailed hydrologic and hydraulic analyses must be performed at this stage to finalize
design discharges and determine the dimensions of hydraulic structures, such as
pipes, inlets, ponds, and energy dissipators. Structural analyses are necessary to deter-
mine required pipe strength classes or wall thicknesses and may also be required for
other project-specific hydraulic structures. Geotechnical analyses are necessary to
locate shallow bedrock that could affect construction costs; to determine groundwater
elevations that could affect trenching, buoyancy forces on submerged pipes and struc-
tures, and pipe infiltration; and to determine pipe-bedding requirements.
It is essential that complete and well-organized files be maintained to document each
activity and decision in the detailed design phase. Should unforeseen circumstances
occur during construction, or should litigation arise after construction of the system
has been completed, these files will provide a record of the adequacy of the design
and its adherence to accepted engineering practices.
Construction Drawings, Specifications, and Contract

Documents
Construction drawings should be prepared as the final design is completed and should
go hand in hand with that activity. Construction plans consist of plan and profile
sheets showing horizontal and vertical alignments of new and existing facilities,
including all dimensions required for a survey crew to lay out the system in the field.
Plan views should indicate the locations, elevations, and dimensions of all proposed
pipes, inlets, manholes, and other system features, and should show at least approxi-
mate locations of existing utilities that could pose potential conflicts. Existing land-
surface features, such as curbs, walks, and other structures requiring demolition and
subsequent replacement, should also be shown.
Profile views should include required pipe slopes and invert elevations, existing and
proposed ground-surface elevations, approximate limits of bedrock, and approximate
elevations of underground utilities. If the construction contractor is to perform utility
relocations, then construction requirements for the relocations must also be included
in the plans. It is sometimes desirable to show energy and hydraulic grade-line pro-
files on construction plans; they are certainly not needed for construction, but may be
required for regulatory submittals.
In addition to plan and profile drawings, construction plans should clearly identify
required materials and material classes and should provide construction details, such
Section 2.4 Integration of Stormwater Conveyance Modeling with CAD Software and GIS 39
as reinforcing steel layouts, for any special structures. Some construction details may
already be available in sets of standard plans published by, for example, transportation
departments. As a courtesy to project bidders, construction plans may also include
one or more sheets containing tabulations of estimated material quantities.
Technical specifications are the heart of any set of construction documents, spelling
out in detail the required construction methods and material types, classes, and testing
requirements. It is common to reference standard construction specifications pub-
lished by jurisdictional agencies, but care should be taken to avoid conflicting require-
ments in the plans and technical specifications. There may be conflicts between the
engineer’s specifications and standard specifications, between standard specifications
and construction plans, or both. There is no such thing as a “standard project.” Stan-
dardized technical specifications must be viewed with a wary eye.
The contract documents for a construction project consist of the plans, technical spec-
ifications, and other “boiler-plate” provisions. These additional provisions generally
consist of bidding documents, bonds, official notices to proceed and notices of project
acceptance, and general and special conditions of the contract. The general conditions
typically specify such things as owner-engineer-contractor responsibilities and job-
site requirements relating to issues such as materials storage, sanitation, site cleanup,
and conformance with applicable laws and regulations. The general conditions are
typically a standard document used for all construction contracts within a jurisdic-
tional area. The “Standard General Conditions of the Construction Contract” docu-
ment from the Engineers Joint Contract Documents Committee (EJCDC, 2002) is a
widely used set of general conditions in the United States Special conditions, which
are project-specific and assembled by the engineer, identify requirements other than
those specified in the general conditions, and can also be used to modify provisions of
standardized technical specifications.
2.4 INTEGRATION OF STORMWATER CONVEYANCE

MODELING WITH CAD SOFTWARE AND GIS
The integration of conveyance modeling with computer-aided design (CAD) pro-
grams and geographic information systems (GIS) has a number of benefits. Hydraulic
and hydrologic design and analysis software often includes both CAD and GIS-type
features. In some cases, analysis applications can run directly within an organization’s
preferred CAD or GIS platform. This section describes some of the benefits of lever-
aging CAD and GIS capabilities in hydraulic design and analysis.
CAD Integration
CAD software can be useful in both the initial and final stages of a hydraulic design
project. A CAD drawing typically provides the base plan for the layout of stormwater
management facilities such as sewer networks, channels, and detention ponds. After
the facilities have been designed, this plan is updated to reflect proposed structures for
inclusion in the construction documents.
Software applications interact with CAD systems in a variety of ways. A graphics-

based application such as StormCAD may be capable of displaying a background
CAD file for use in system layout, or the hydraulic analysis application may run
directly inside of a CAD program such as AutoCAD (see Figure 2.13) or Microsta-
tion. CAD features can be used to fine-tune system layout and supply data on the
topography and horizontal and vertical structure coordinates. Many CAD programs
used by civil engineers are equipped with special tools for assisting engineers in cre-
ating roadway grading plans and drawing sewer profiles.
Figure 2.13
Storm sewer plan and
profile created using
hydraulic design
software running
inside of AutoCAD
Engineers can choose from various approaches when integrating hydraulic design
with a CAD application. Several possible options, ordered from least to greatest
degree of integration, are as follows:
• No automated data exchange occurs between the CAD program and the
hydraulic modeling software. The engineer manually enters schematic data
on the hydraulic system into the analysis software and designs the system.
The system characteristics resulting from the design are then manually input
into the CAD drawing for inclusion in construction documents.
• A CAD drawing containing the background information necessary to lay out
the stormwater facilities is used as a background map in the hydraulic design
software. After the system’s layout is completed and its elements designed,
the layout is imported into the CAD drawing. Other element characteristics
are added to the drawing manually.
• The initial hydraulic system layout is created in the CAD drawing and
imported into the analysis program. Depending on the features of the CAD
software, the data imported may include, in addition to the horizontal layout
(that is, coordinate and/or connectivity data), elevation information, structure
types, preliminary structure sizes, and other characteristics. Once the design
is final, system data can be transferred back into the CAD application
through automated import/export methods.
• An analysis application that runs directly within the CAD platform is used.
In the design process, hydraulic structures are inserted directly into the draw-
ing file, and data are readily available from both the CAD program and
hydraulic design tool. This technique is the most efficient of the various inte-
gration approaches.
Typically, increased integration of CAD and hydraulic design and analysis applica-
tions translates to less data duplication, decreased opportunities for the introduction
of errors and omissions, and increased time savings.
GIS Integration
A geographic information system (GIS) is a powerful configuration of computer hard-
ware and software used for compiling, storing, managing, manipulating, analyzing,
and mapping (displaying) spatially referenced information. It combines the function-
alities of a computer graphics program, an electronic map, and a database.
Like CAD, a GIS can be used in many stages of stormwater conveyance modeling.
Although CAD programs provide an excellent format for developing technical draw-
ings, they generally do not provide a means of storing data associated with those
drawings. For example, a CAD drawing may show a stormwater pipe network for a
neighborhood. However, if the engineer is to model the network using information
from the drawing, the drawing must contain detailed annotations for each structure
describing characteristics such as pipe material, size, and inlet and outlet elevations.
A GIS, by comparison, can store model information in an internal database associated
with each GIS data layer. The database functions of a GIS also allow data to be stored
and maintained externally and accessed when needed through a dynamic database
link.
GIS software typically has tools for drawing (digitizing), map display, data storage,
and data analysis. The digitizing tools allow for conversion of hard-copy drawings to
an electronic format and are almost identical to those of CAD programs. Map display
tools allow users to change the appearance of different types of features (points, lines,
polygons, and text) according to values in the database. [For instance, within a single
layer, all 6-in. (150-mm) pipes may be shown in red, with concrete pipes displayed as
solid lines and corrugated steel pipes displayed as dashed lines.] The data storage
tools are similar to those of many database programs, with data lookup, query, and
sort functions.
The data analysis tools of GIS make it well-suited to supporting stormwater convey-
ance modeling. These tools allow users to consider both spatial and data relationships
to develop new data or to provide model inputs. For example, one may be trying to
evaluate the runoff response of a watershed by looking at information about a storm
sewer trunk line, which may be made up of multiple pipes of different sizes and mate-
rials. Data analysis tools in GIS could aid in determining the hydrograph’s time to
peak by spatially defining a connected flow path, and then using stored data on pipe
characteristics to solve the hydraulic equations necessary to calculate the runoff
response times.
As may be inferred from the previous example, two important ways in which GIS
software supports stormwater conveyance modeling are
• By assisting hydrologic modeling
• By storing data from stormwater system inventories.
GIS and Hydrologic Modeling. GIS can be integrated with hydrologic mod-
eling in a number of ways, depending on the type of work being performed. Many
local government agencies maintain extensive GIS systems, and these can be excel-
lent sources of hydrologic and hydraulic model input data, such as
• Data for physical features like rivers and streams, land cover, geology, and
soils
• Specialized topographic data including spot elevations, surveyed cross-
sections, contours, and digital elevation models (DEMs)
• Data for man-made features like roadways and other impervious areas,
bridges, storm sewers, and other utilities
• Tax parcel or property boundaries, addressed buildings, and ownership
information
• Governmental maps including municipal boundaries, census data, and
zoning
GIS software packages offer various specialized tools for managing the topographic
data used in delineating watersheds. However, actual watershed boundaries can differ
from those delineated using only surface topography (for instance, storm sewers may
cross surface watershed boundaries). For this reason, the engineer must take care to
address both natural and man-made features when delineating watersheds and ensure
that the boundaries are correctly placed in the GIS. Figure 2.14 shows a watershed
delineated using a GIS.
After the watershed boundary is delineated, the next task is often to model the runoff
response of the basin. GIS data analysis tools can assist in processing rainfall data and
determining area-weighted averages for parameters such as runoff coefficient (see
Section 5.2, page 118). For more complex coefficients based on multiple data layers,
the GIS can also use an overlay tool to develop new mapping for areas with unique
combinations of characteristics. For example, the curve number method predicts run-
off response based on both land use and hydrologic soil group (see Section 5.2, page
119). Areas with unique curve numbers can be delineated by overlaying a land use
map on a soils map and then using a lookup table.
Figure 2.14
Watershed delineated
using a GIS
GIS data can also play a critical role in finding suitable sites for new stormwater
structures and systems. As an area develops, land for stormwater management struc-
tures such as large detention ponds may become more scarce and difficult to find. A
GIS is ideally suited to performing the spatial and data analyses needed to locate these
sites. For example, hydrologic analysis may show that a 10-acre (4-ha) pond needs to
be located along a 3-mi (5-km) stream reach. Any site for the pond must be within
300 ft (90 m) of the stream, have a minimum of 15 ac (6 ha) of available space, and
possess suitable soils and proper zoning. Further, the pond cannot impact environ-
mental or historical sites. Through a series of data analysis steps, a GIS could identify
sites that meet these criteria and rank these sites according to anticipated land cost
based on the current tax assessment.
GIS for Stormwater System Inventories. GIS can also be used to store
the inventory information for stormwater management systems. Even simple storm-
water systems may have a variety of gutters, inlets, outlet, manholes, culverts, pipes,
channels, and other facilities. The database capabilities of a GIS allow each compo-
nent to have a unique database structure and setup for use in storing individual struc-
ture information.
GIS inventories can play an important role in both the hydraulic modeling and main-
tenance of a stormwater system. Data about structures that could be stored and used in
modeling would likely include
• Types
• Locations
• Materials
Data useful in maintenance efforts may include

• Inspection reports
• Maintenance logs
• Public works complaint records
The database linking capability of the GIS is important, because many of the mainte-
nance records may be stored in databases of several different local government
departments.
GIS integration with stormwater modeling, like CAD integration, can range from no
automated data exchange to complete integration within a GIS platform. However, the
data analysis and database capabilities that can be performed using a GIS are more
complex and, in many cases, can substantially reduce the number of data-entry steps.
The data exchange between a model and a GIS is two-way: a GIS can provide existing
information to a model, and data from a model of a new system can be imported into
a GIS to update a stormwater system inventory. Some hydraulic analysis applications
(for example, StormCAD) provide a means of automatically synchronizing data with
a GIS. Figure 2.15 shows storm sewer system data stored in a GIS.
Figure 2.15
Information on a
storm sewer system
can be stored in a GIS
2.5 Sources of Model Data

The typical stormwater conveyance project, whether it is a new design or a study of an
existing system, uses data from a variety of sources. At the beginning of a project, the
engineer should consider what the data requirements for the project will be, including
data types, quality, and quantity. Once these requirements have been defined, it is usu-
ally a straightforward effort to identify the data sources. When multiple sources are
Section 2.5 Sources of Model Data 45
available for a specific data type, the choice of which source to use is based on bal-
ancing cost with data quality.
Proposed Systems
For proposed systems, most of the model input is developed by the engineer as part of
the design process and will be subject to constraints dictated by project goals, site
conditions, and local design standards. Common sources of data utilized in the design
of new systems are described in the paragraphs that follow.
Surveys. For proposed systems, survey data are used as the base upon which the
proposed layout and grading plans, and therefore stormwater conveyance designs, are
developed. A survey also provides information on existing watercourses, drainage
systems, utilities, and various obstacles in the project area, and aids the engineer in
avoiding conflicts with these features. Of particular interest to stormwater conveyance
designers is survey information on the site’s outfall(s). The outfall location provides
the engineer with a controlling elevation that the proposed system must tie into. For
cases in which a proposed sewer system will tie directly to an existing sewer system,
information on existing structure characteristics is required.
Layout and Grading Plans for the Proposed Site. The design of the
stormwater conveyance system for a project is typically performed concurrently with
or following the development of the layout and grading plans. These plans serve as
the primary source of information for development of the preliminary drainage sys-
tem layout. The proposed grading reveals the site’s major catchment areas and pro-
vides the ground elevations required by the model.
Hydrologic Data Sources. Hydrologic data are available from a variety of
sources. For most designs, local rainfall data and design storm information can be
obtained from the local review agency. Transportation departments are another possi-
ble source of hydrologic data, as are governmental organizations such as the National
Weather Service and Natural Resource Conservation Service in the United States.
Often, data are readily available through an organization’s web site. Further informa-
tion on hydrologic data sources is presented in Chapters 4 and 5.
Local Design Standards. Local design standards provide a number of con-
straints in drainage design and may originate from more than one source (for instance,
in addition to city or county requirements, state environmental regulations and trans-
portation department standards may also be applicable to a single design project).
Requirements frequently include
• Minimum pipe or culvert size
• Acceptable pipe or culvert materials
• Approved inlet types
• Design storm event return periods
• Maximum gutter spread and depth
• Minimum inlet capture efficiency
• Maximum distance between manholes

• Minimum freeboard requirement
• Stormwater detention requirements
• Accepted analysis techniques and applicable procedures and coefficients
• Submittal requirements for permitting
When beginning a drainage design in an unfamiliar jurisdiction, it is best to contact
the local reviewing agency or agencies early in the design process to discuss the
applicable requirements and obtain a copy of the local stormwater regulations/design
guidelines, if available. This preparation can save considerable time later in the
process.
Existing Systems
The sources of information used in evaluating the performance of an existing storm-
water facility such as a sewer system, detention basin, or roadway culvert are often
the same as those used in new designs, but the approach to the project will differ
somewhat. On the surface, the analysis of an existing system may appear to be an eas-
ier task than designing new facilities: the geometry of the system is known, and topo-
graphic information can be taken directly from a survey rather than from a proposed
design. However, the analysis of an existing system often presents other challenges, as
such studies almost always stem from problems such as undersized piping and flood-
ing in the area. The modeler is faced with the forensic problem of trying to re-create
what happens in a real system, as opposed to simply sizing system components for a
synthetic design event.
Surveys. Survey data are relied on much more extensively in the analysis of an
existing system than in the design of a new system. For instance, a number of storm
sewer manholes and inlet structures must often be field-located and their covers
removed to determine pipe sizes, orientations, and invert elevations. Ponds and
depressions that serve as detention basins must be carefully surveyed, and these areas
are often wet or overgrown, making access difficult. Detention outlet structures must
also be surveyed carefully to model the basin accurately, including planned or
unplanned overflow spillways. If a culvert roadway crossing is being examined, the
culvert entrance type must be noted and the roadway profile surveyed for overtopping
analysis. A detailed topographic survey may also be needed to determine drainage
subbasin areas.
Site Reconnaissance. When conducting a study of an existing drainage system,
it is advisable for the engineer to make at least one site visit. A visit to the site pro-
vides an opportunity to
• Resolve ambiguities in the survey
• Examine and delineate contributing drainage areas that may lie outside of the
survey area
• Observe the condition of the existing infrastructure
Section 2.5 Sources of Model Data 47
• Look for project-related problems such as soil erosion, debris in drainage-

ways, and water quality issues
• Observe watershed and drainageway characteristics to determine model
parameters such as channel roughnesses, runoff coefficients, and culvert
entrance types
• Photograph the site to assist other engineers involved in the project, provide
answers to questions that may arise after the site visit, and provide a record
of site conditions
Original Plans and Design Calculations. A great deal of information use-
ful in the study of an existing system can be obtained from original construction doc-
uments and design calculations. The compliance of the constructed system with the
original plans must still be verified through a survey, but these plans do help the engi-
neer to understand the original intent of the designer. They are an excellent starting
point for the engineer in deciding what information the surveyor should gather. The
engineer can also use the plans to identify problems in the system resulting from dif-
ferences in the intent of the designer and the constructed system—problems that may
be alleviated by altering the site to bring it into closer agreement with the original
plan—or to identify deficiencies in the original design.
The original hydraulic and hydrologic design calculations can also be of help to the
engineer. The model input data and calculations can be verified for accuracy, which
can aid the discovery of why the system is not providing the anticipated level of pro-
tection. For instance, a contributing drainage area upstream of the project’s convey-
ance system may have been developed since the project’s construction and be
contributing more runoff than it was previously.
As-Built Surveys/Drawings. When available, as-built drawings can be an
excellent source of information when conducting a study of an existing system. These
drawings may be obtained from the property owner, the engineer who did the original
design, the surveyor who performed the as-built survey, or the local review agency.
If the accuracy of the as-built drawings can be verified through site reconnaissance,
and depending on the level of detail required by the study, as-built drawings can
greatly reduce or even eliminate the need for additional survey data. The engineer
must make allowances for the level of detail of the as-built drawings and any changes
made to the site since its construction when deciding how much to rely on this source
of information.
Structure Inventories and Atlases. Local public works departments or
stormwater utilities, and state transportation departments, often maintain mapping
and other data on stormwater conveyances, especially those within the public right-of-
way. Frequently, this information is stored within a GIS. These data can be useful
when performing a hydraulic analysis, especially to help the engineer understand
what is happening off-site and learn about the potential downstream effects of a
project.
Existing Flood Studies. Existing hydraulic studies are available for many
streams, especially in developed areas. In the United States, these studies may have
been performed by the U.S. Army Corps of Engineers or, increasingly, by private
consulting engineers under contract with government agencies or developers. In the

United States, these studies are usually submitted to the Federal Emergency Manage-
ment Agency (FEMA) and, if approved, become part of regulatory flood maps.
For projects involving large, previously studied conveyances, detailed reports and
hydrologic and hydraulic calculations can often be obtained through the contracting
or reviewing agency. The calculations can be used to understand the performance of
conveyances in and around the project area, and they can be modified to reflect
changes to these conveyances resulting from the proposed project.
When using existing flood study data, it is important to keep in mind the study’s level
of accuracy. Many of these studies are out-of-date and were subject to tight budgetary
constraints. Thus, the level of detail that often characterizes these large-scale studies
is typically insufficient when examined on the smaller scale of many stormwater con-
veyance projects. For instance, flows may have been computed using a regression for-
mula having a high degree of uncertainty associated with it, or stream cross-sections
may be very far apart.
Also important to be aware of when referring to an older study are changes in the state
of engineering practice since the study was performed. More accurate modeling tech-
niques may have become available since that time. Further, the study area may have
undergone significant changes since the study was performed, such as alterations to
land use, stream cross-sections, and roadway crossings.
Floodplain Modeling Using HEC-RAS (Dyhouse et al., 2003) provides detailed infor-
mation on flood studies.
2.6 CHAPTER SUMMARY

Stormwater drainage systems involve many types of structures; thus, a number of
mathematical models (both hydraulic and hydrologic) are necessary to describe these
systems adequately. Facility types used in stormwater management include open
channels, culverts, storm sewers, pump stations, and detention basins, and various
software packages are available for their analysis and design.
Hydraulic software that allows for integration with CAD, GIS, and databases already
used by an organization can greatly facilitate the model-building process, the creation
of construction documents, and the maintenance of records on stormwater facilities.
Data that can be used in modeling stormwater systems come from a variety of
sources. The sources chosen depend somewhat on whether the project consists of the
study of an existing system or the design of a new system. For proposed systems, data
come primarily from surveys, site plans, and local design standards. Surveys take on
increased importance when modeling existing systems, as do site reconnaissance and
record drawings. Original plans and design calculations and structure inventories may
also be used when available.
References 49
REFERENCES
American Society of Civil Engineers (ASCE). 1992. Design and Construction of Urban Storm Water Man-
agement Systems, New York: ASCE.
Dyhouse, Gary, J. A.Benn, David Ford Consulting, J. Hatchett, and H. Rhee. 2003. Floodplain Modeling
Using HEC-RAS. Waterbury, Connecticut: Haestad Methods.
Engineers Joint Contract Documents Committee (EJCDC). 2002. Standard General Conditions of the Con-
struction Contract, Reston, Virginia: ASCE.
Haestad Methods. 2003a. PondPack: Detention Pond and Watershed Modeling Software. Waterbury, Con-
necticut: Haestad Methods.
Haestad Methods. 2003b. StormCAD: Storm Sewer Design and Analysis Software. Waterbury, Connecticut:
Haestad Methods.
Normann, J.M., R.J. Houghtalen, and W.J. Johnston. 2001. Hydraulic Design of Highway Culverts. 2d ed.
Washington, D.C.: Federal Highway Administration (FHWA).
A hydraulic jump is
formed in a
laboratory tank.
For a short jump,
momentum is the
same upstream
(left) and
downstream (right).
Energy is lower
downsteam because
of dissipation by
turbulence.
C H A P T E R
3
Fundamental Laws and Units
Scientists and engineers have developed equations to describe hydrologic processes

and the hydraulic properties of stormwater conveyances. The engineering calculations
used in the design and evaluation of stormwater systems are, to a large extent, applica-
tions of the fundamental physical laws of conservation of mass, energy, and momen-
tum. The first three sections of this chapter provide reviews of these fundamental
laws. The fourth section addresses some of the common units of measurement
employed in hydrologic and hydraulic engineering.
3.1 CONSERVATION OF MASS

The familiar principle of conservation of mass simply states that matter is neither
created nor destroyed. The mass entering a system is equal to the mass leaving that
system, plus or minus the accumulation of mass (that is, storage) within the system.
Without doubt, this principle is the single most important concept that must be
applied in hydrologic and hydraulic engineering. In the engineering analyses that
must be performed, the properties of flow change both through space and over time.
For example, the amount of fluid present in a given channel reach may change over
time; thus, accounting for the accumulation (or loss) of fluid during any time interval
is essential.
As a practical matter, the application of the principle of conservation of mass is
simply an accounting procedure. Consider, for example, a bank account and assume
that the account contains $100 at the beginning of the month. Suppose also that
deposits made during the month amount to $50, and withdrawals amount to $70.
Clearly, the account balance at the end of the month is $80 (= $100 + $50 − $70). If
the ending balance had been specified instead of the monthly withdrawals, then a sim-
ilar calculation could be applied easily to determine the withdrawals. In any case, the
change in the account balance is equal to the difference between deposits and
withdrawals.
52 Fundamental Laws and Units Chapter 3
Now, consider an incompressible fluid (that is, a fluid having a constant density) that
flows through a fixed region (a control volume) in space and denote the volume of
fluid, or storage, in that region at time t by the quantity S(t). The fixed region might
correspond to a lake or reservoir, to a water tank, or to a length of river or stream. If
I(t) represents a volumetric flow rate into the region, then a volume I(t)∆t of
additional fluid flows into the fixed region during the time interval of duration ∆t.
Representing the volumetric flow rate out of the system as Q(t), a volume of fluid
equal to Q(t)∆t flows out of the region during the time interval. [Note that I(t) and
Q(t) may be functions of time.]
According to the principle of conservation of mass, at the end of the time interval
(that is, at time t + ∆t), the storage S(t + ∆t) of fluid in the fixed region will be equal to
what was there initially, plus the additional fluid that entered during that time, less the
amount of fluid that left during that time. Expressed mathematically (with example
units),
S (t + ∆t ) = S (t ) + I (t )∆t − Q(t )∆t (3.1)
where t = time (s)

S(t + ∆t) = storage at the end of the time interval (ft3, m3)
S(t) = storage at the beginning of the time interval (ft3, m3)
I(t) = volumetric inflow rate at time t (cfs, m3/s)
Q(t) = volumetric outflow rate at time t (cfs, m3/s)
A volumetric flow rate, typically denoted by Q, is referred to as a discharge in hydro-
logic and hydraulic engineering.
A rearrangement of Equation 3.1 yields
S (t + ∆t ) − S (t )
= I (t ) − Q(t ) (3.2)
∆t
As ∆t approaches zero, this equation becomes
dS
= I (t ) − Q(t ) (3.3)
dt
Equation 3.3 is the differential form of the mathematical expression for conservation
of mass. An integral form may be obtained by multiplying both sides of the expres-
sion by dt and integrating:
S2 t2 t2
∫ dS = S 2 − S1 = ∫ I (t )dt − ∫ Q (t )dt (3.4)
S1 t1 t1
If a flow is steady—that is, if the flow characteristics do not change with time—then
the time derivative in Equation 3.3 is zero, and conservation of mass may be
Section 3.1 Conservation of Mass 53
expressed by simply stating that the inflow and outflow discharges to and from a con-
trol volume are equal. In other words, I(t) = Q(t).
The volumetric flux of a fluid is defined as the volume of fluid that passes a unit cross-
sectional area per unit time. It can be expressed as
S
Volumetric Flux = (3.5)
At
where S = fluid volume (ft3, m3)

A = cross-sectional area (ft2, m2)
t = time (s)
Because Q is a fluid volume per unit time, the volumetric flux can be restated as
Q
Volumetric Flux = (3.6)
A
where Q = flow (cfs, m3/s)

It can be seen from this definition that volumetric flux has units of length per unit
time. Noting that velocity has the same units, one can express the average velocity of
flow as
Q
V= (3.7)
A
where V = average velocity (ft/s, m/s)

Comparing Equations 3.6 and 3.7, average flow velocity is the same as volumetric
flux.
In steady flow, the discharges into and out of a control volume must be equal, and
Equation 3.7 can be used to write
( AV )inlow = ( AV )outflow (3.8)
The reader should recognize that the equations presented in this section have been
cast in terms of fluid volumes and volumetric flow rates even though the discussion
concerns conservation of mass. The use of volumes and volumetric flow rates is valid
as long as the fluid in question is incompressible (has a constant density). This
assumption is valid for nearly all civil and environmental engineering applications
dealing with water. For instances in which this assumption is not valid, the volumes
and volumetric flow rates in the above equations should be replaced with fluid masses
and mass flow rates. A fluid mass is found by multiplying the fluid’s volume by its
density, and a mass flow rate is found similarly by multiplying the volumetric flow
rate by the density.
Example 3.1 – Determining Discharge and Velocity Based on Conservation of

Mass. A storm sewer is being designed for a location where the slope of the sewer must decrease
because of local topography. A manhole will be positioned at the location where the pipe slope
changes. Upstream of the manhole, the topography is relatively steep, and the pipe diameter is 600
mm. Downstream of the manhole, where the topography is flatter, a 750-mm diameter pipe is
required to convey the flow. No additional flow enters the sewer at the location of the manhole. The
pipes in the sewer are flowing full both upstream and downstream of the manhole, and the discharge
in the upstream pipe is 0.5 m3/s. The flow is steady. Determine the discharge and velocity in the
downstream pipe.
Solution: Consider the manhole as a control volume into which flow enters from the upstream pipe
and from which flow departs through the downstream pipe. Because the flow is steady, the inflow and
outflow discharges must be equal, and thus the discharge in the downstream pipe is 0.5 m3/s. The
velocity in the downstream pipe can be found using Equation 3.7. The cross-sectional area of the 750-
mm diameter downstream pipe is A = 0.442 m2, and thus the velocity is
V = Q/A = 0.5/0.442 = 1.13 m/s
Example 3.2 – Applying Conservation of Mass to a Detention Pond. A discharge

hydrograph is a graph showing how the discharge in a pipe or channel changes over time. The dis-
charge is usually shown on the vertical axis of the graph, and time is shown on the horizontal axis. For
any time interval on the horizontal axis, the area under the curve shown on the graph represents the
volume of the flow over that time interval.
Figure E3.2.1 shows the inflow and outflow hydrographs for a detention pond. The inflow hydrograph
is represented by the solid line, and the outflow hydrograph is shown as a dashed line. Assuming that
the pond is initially empty, the outflow hydrograph is triangular in shape, and the pond is again empty
at time tend, determine the following:
a) the total volume of water passing through the pond
b) the time tend
c) the time at which the storage in the pond is at a maximum
d) the maximum storage in the pond
Figure E3.2.1 Inflow and outflow hydrographs for a detention pond

Section 3.2 Conservation of Energy 55
Solution to (a): The total volume of water passing through the pond is equal to the total volume of
water entering the pond because it is empty at time zero and again at time tend. This volume of water
is equal to the area under the inflow or outflow hydrograph. In this case, the volume can be computed
using the formula for the area of a triangle as
∀inflow = 0.5(3 hr)(100 cfs)(3,600 s/hr) = 540,000 ft3

Solution to (b): The time tend at which the pond is again empty may be found using Equation 3.4 and
noting that S1 = S2 = 0 when one takes t1 = 0 and t2 = tend. Hence, the flow volumes represented by the
inflow and outflow hydrographs must be equal. The inflow volume was found in the solution to part
(a) to be 540,000 ft3; thus,
∀outflow = ∀inflow = 540,000 ft3 = 0.5(50 cfs)(tend)

Solving,
tend = 21,600 s = 6 hr
Solution to (c): Inspection of Figure E3.2.1 shows that during the time interval from 0 to 2 hr, the dis-
charge into the pond is greater than the discharge out of it. Thus, the storage in the pond must be
increasing during that time interval. After t = 2 hr, the outflow discharge is greater than the inflow dis-
charge, so the storage in the pond must be decreasing. Therefore, the time at which storage in the
pond is a maximum is t = 2 hr.
Solution to (d): Apply Equation 3.4 with t1 = 0 and t2 = 2 hr. Because the pond is empty at time zero,
S1 = 0. The maximum storage volume S2 at time t2 is equal to the difference between the volumes rep-
resented by the inflow and outflow hydrographs over the time interval from 0 to 2 hr:
S2 = [0.5(1 hr)(100 cfs) + 0.5(1 hr)(100 ftcfs + 50 cfs) – 0.5(2 hr)(50 cfs)](3600 s/hr)
S2 = 270,000 ft3
3.2 Conservation of Energy

In the most formal sense, a discussion of conservation of energy for a fluid would
begin with the First Law of Thermodynamics. That law states that the rate of change
of stored energy in a fluid system is equal to the rate at which heat energy is added to
the system, minus the rate at which the fluid system does work on its surroundings.
The stored energy is composed of kinetic energy due to the motion of the fluid,
potential energy due to its position relative to an arbitrary datum plane, and internal
energy.
In the vast majority of civil and environmental engineering applications, flow can be
considered to be steady and incompressible, and it is generally sufficient to apply the
principle of conservation of energy in a much simpler way using what is commonly
known as the energy equation. The expression most commonly applied expresses
energy on a unit weight basis (that is, as energy per unit weight of fluid) and may be
stated as
p1 α1V12p α V2
+ Z1 + = 2 + Z 2 + 2 2 + hL − hP + hT (3.9)
γ 2g γ 2g
where p = fluid pressure (lb/ft2, Pa)

γ = specific weight of fluid (lb/ft3, N/m3)
Z = elevation above an arbitrary datum plane (ft, m)
α = velocity distribution coefficient
V = fluid velocity, averaged over a cross section (ft/s, m/s)
g = acceleration of gravity (ft/s2, m/s2)
hL = energy loss between cross sections 1 and 2 (ft, m)
hP = fluid energy supplied by a pump between cross sections 1 and 2 (ft, m)
hT = energy lost to a turbine between cross sections 1 and 2 (ft, m)
The first three terms on each side of this equation represent, respectively, the internal
energy due to fluid pressure, potential energy, and kinetic energy. When the dimen-
sion of each term is given in units of length as in Equation 3.9 (resulting from division
of energy units by fluid weight units), the three terms are generally called the pressure
head, the elevation head, and the velocity head. The terms on the left side of the equa-
tion with the subscript 1 refer to an upstream cross section of the fluid, and those on
the right side with the subscript 2 refer to a downstream cross section.
Appendix A contains tabulations of the physical properties of water at standard atmo-
spheric pressure. The specific weight and other properties of water may be deter-
mined from those tables.
Strictly speaking, the energy equation applies only to a particular streamline within a
flow. Because the velocities along individual streamlines are generally different due
to the effects of pipe walls or channel sides, the velocity head terms in Equation 3.9
should be corrected using the velocity distribution coefficient (α) if one chooses to
use an average cross-sectional velocity. Chapter 6 discusses this coefficient in more
detail. Chapter 6 also provides more details related to the energy loss, pump, and tur-
bine terms of Equation 3.9.
Because the dimension of each term in Equation 3.9 has units of length, qualitative
aspects of the energy equation can be shown graphically. Figure 3.1 shows the profile
of a pipe in which a venturi meter has been installed to measure flow. The pipe diam-
eter downstream of the venturi is smaller than the pipe diameter upstream of the ven-
turi. An arbitrary datum is shown below the pipe profile.
In Figure 3.1, the elevation head (Z) represents the vertical distance from the datum to
the pipe centerline. The hydraulic grade line (HGL) represents the height to which a
column of water would rise in a standpipe placed anywhere along the length of the
pipe. The height of the HGL is sometimes called the piezometric head. The vertical
distance from the pipe centerline to the HGL is the pressure head, p/γ, and the dis-
tance from the HGL to the energy grade line (EGL) is the velocity head, V2/2g. The
vertical distance from the datum to the EGL is representative of the sum of the first
three terms on each side of Equation 3.9 and is called the total head. Flow always
occurs in a direction of decreasing total head (not necessarily decreasing pressure),
and hence is from left to right in the figure.
Section 3.2 Conservation of Energy 57
Figure 3.1
A venturi meter is
used to measure flow
in a pipe
According to conservation of mass, the velocity must be greater in the pipe

downstream of the venturi than in the upstream pipe; thus, the distance between the
HGL and EGL is greater for the downstream pipe. At the throat of the venturi itself,
where the flow velocity is the highest, the distance between the EGL and HGL is the
greatest. In the regions upstream and downstream of the venturi where the HGL is
above the pipe centerline, the fluid pressure (gauge pressure) in the pipe is positive. In
this example, the HGL falls below the pipe centerline at the venturi, and the fluid gage
pressure is negative. Note that although the fluid pressure just downstream of the ven-
turi is greater than that in the throat of the venturi, flow is still from left to right.
The EGL in the figure has a nonzero slope because energy is lost due to friction and
turbulent eddies as fluid moves along the length of the pipe. In other words, the total
head changes along the pipe and decreases in a downstream direction. In Figure 3.1,
because neither a pump nor turbine is present, the difference in the elevation of the
EGL between any two locations (cross sections) along the pipe is representative of the
head loss hL between those two cross sections. The loss rate of total energy in a fluid
flow increases with the velocity of the flow. Thus, in the figure, the magnitude of the
slope of the EGL is larger for the downstream pipe than for the upstream pipe.
Example 3.3 – Determining Flow Direction from the Energy Equation. Two cross
sections along a constant-diameter pipe are denoted by A and B. The pipe centerline elevation at cross
section A is 30 m, and the fluid pressure in the pipe at that location is 210 kPa. At cross section B, the
pipe centerline elevation is 21 m, and the fluid pressure is 240 kPa. Determine the direction of flow in
the pipe. The specific weight of water is 9.81 kN/m3.
Solution: The total energy head at each of the two cross sections is the sum of the elevation head (that
is, the pipe centerline elevation), the pressure head, and the velocity head. Flow occurs in the direction
of decreasing total energy head.
From Equation 3.9, the total energy head HA at cross section A is
pA α V A2 210 α V A2 α V A2
HA = + ZA + = + 30 + = 51.41 +
γ 2g 9.81 2g 2g
At cross section B, the total energy head is

2 2
240 α VB α VB
HB = + 21 + = 45.46 +
9.81 2g 2g
Because by continuity, the discharge and velocity must be the same at each of the two cross sections,
the velocity head terms on either side of Equation 3.9 are equal and cancel out. Thus, in this case,
flow is in a direction of decreasing piezometric head. Because the piezometric head of 51.41 m at
cross section A is higher than the piezometric head of 45.46 m at cross section B, flow is from A to B.
Note that this is true even though the fluid pressure is higher at cross section B.
3.3 CONSERVATION OF MOMENTUM

The principle of conservation of momentum states that the summation of the external
forces acting on a system is equal to the time rate of change of momentum for the sys-
tem. The physical principle of conservation of momentum is generally more difficult
to apply in practice than either conservation of mass or conservation of energy. The
added complexity is the consequence of momentum being a vector-valued quantity
(that is, it has both magnitude and direction), whereas mass and energy are scalars
(represented by magnitude only). Momentum equations must therefore be written for
each coordinate direction separately.
Section 3.3 Conservation of Momentum 59
In its most general form, conservation of momentum is expressed as
∂
∑ Fx = ∂t ∫ ρVx dS + ∑ (M out ) x − ∑ (M in ) x (3.10)
cv
where Fx =
force acting on the water in a control volume in the x-direction (lb, N)
t =
time (s)
cv control volume (ft3, m3)
=
ρ fluid density (slugs/ft3, kg/m3)
=
Vx =
x-component of the velocity of the fluid in the control volume (ft/s, m/s)
S fluid volume (ft3, m3)
=
(Mout)x =
x-component of momentum outflow rate from the control volume (lb,
N)
(Min)x = x-component of momentum inflow rate into the control volume (lb, N)
This expression, which is written for the coordinate direction x, can be written for
other coordinate directions as needed. It states that the sum of the external forces act-
ing on the water in a control volume is equal to the time rate of change of momentum
within the control volume, plus the net momentum flow rate (in the x-direction) from
the control volume. In this expression, the sums of Mout and Min account for the pos-
sibility of more than one inflow and/or outflow pathway to/from the control volume.
A momentum flow rate, in the x-direction, can be written as
Mx = βρ QVx (3.11)
where Mx = momentum flow rate in the x-direction (lb, N)

β = velocity distribution coefficient
Q = discharge (ft3/s, m3/s)
The numerical value of β is close to 1 for turbulent flow. Sections 6.1 and 6.4 provide
an explanation of turbulent flow and more detail on β.
If a fluid flow is steady, such that its characteristics do not change with time, Equation
3.10 reduces to
∑ Fx = ∑(Mout )x − ∑(Min )x (3.12)
Further, if there is only a single inflow stream (that is, a single inflow pipe) and a sin-
gle outflow stream from the control volume, Equation 3.12 can be rewritten as
∑ Fx = ρQ∆( βV )x (3.13)
Equation 3.13 states that the vector sum of the x-components of the external forces
acting on the fluid within a fixed control volume (including the pressure forces) is
equal to the fluid density times the discharge times the difference between the x-
components of the outgoing and incoming velocity vectors from and to the control
volume. As in Equation 3.9, the velocity terms should be modified using a velocity
distribution coefficient if one chooses to use the average velocity at a cross section.
Example 3.4 – Determining the Force Exerted on a Bend from the Momentum
Equation. A pipe with flow in a northerly direction has a bend in which flow is turned to 45
degrees east of north (see Figure E3.4.1). The discharge is 25 cfs, the pipe diameter is 24 in., and the
fluid pressures upstream (cross section 1) and downstream (cross section 2) of the bend are 25 psi and
22 psi, respectively. Assuming that the pipe is flowing full and β = 1, determine the force that the
water exerts on the pipe bend.
Figure E3.4.1 Forces on a pipe bend

Solution: Take a control volume as the region of the pipe between cross sections 1 and 2 shown in the
figure. Let the x-direction be taken as due east, and the y-direction as due north. External forces
exerted on the water in the control volume in the x- and y-directions consist of the forces F1 and F2
due to the fluid pressures upstream and downstream of the bend, and the forces Rx and Ry exerted on
the flow by the pipe. The fluid pressure forces can be found by multiplying the fluid pressures by the
pipe cross-sectional area:
F1 = p1A = (25 psi)(3.14)(12 in)2 = 11,300 lb
F2 = p2 A = 9,950 lb
The velocity in the pipe at both cross-sections is
V = Q/A = (25 cfs)/(3.14 x 1 ft2) = 7.96 ft/s
An application of Equation 3.13 in the x-direction leads to
ΣFx = Rx – 9,950 sin 45° = 1.94(25)(7.96 sin 45° − 0)
A similar application in the y-direction leads to
ΣFy = Ry + 11,300 – 9,950 cos 45° = 1.94(25)(7.96 cos 45° − 7.96)
Solving these expressions yields Rx = 7,310 lb and Ry = −4,380 lb. The negative sign on Ry means that
its direction is actually to the south rather than the assumed northerly direction (see figure). The net
force acting on the water has a magnitude of
R = 7,310 2 + 4,380 2 = 8,520 lb
The direction in which this net force acts is

θ = arctan(4,380/7,310) = 30.9° to the south of due east
The forces determined in this example are those acting on the water. The forces exerted on the pipe by
the water are equal to these, but in the opposite directions. Thrust blocks or special pipe jointing
methods may be required in some instances to resist these forces.
Section 3.4 Common Units of Measurement 61
3.4 COMMON UNITS OF MEASUREMENT

Hydrologic and hydraulic engineering is a field marked by many different units of
measurement; therefore, one must frequently convert from one set of units to another.
All of the fundamental dimensions of mass (or force), length, and time are routinely
used. For example, rainfall and rainfall losses are often expressed in units of inches or
millimeters of rainfall per hour (in/hr, mm/hr), and discharges in pipes and open chan-
nels are often expressed in units of cubic feet or meters per second (cfs, m3/s). Drain-
age basin areas might be expressed in units of acres (ac), hectares (ha), square miles
(mi2), or square kilometers (km2); and volumes of water might be expressed in cubic
feet or meters (ft3, m3), or in units such as acre-feet (ac-ft). Table 3.1 summarizes
some of the more frequently used units in stormwater applications, and Appendix B
presents conversion factors for changing from one set of units to another.
Table 3.1 Common units in stormwater applications

Measurement U.S. Customary Units SI Units
Length foot (ft), inch (in.), mile (mi) meter (m), millimeter (mm), kilometer (km)
Mass slug, pound-mass (lbm) kilogram (kg), gram (g)
Time hour (hr), minute (min), second (s) hour (hr), minute (min), second (s)
Temperature degree Fahrenheit (°F) degree Celcius (°C)
Force slug, pound (lb) Newton (N), kilonewton (kN)
Area sq. foot (ft2), acre (ac), square mile sq. meter (m2), hectare (ha), sq. kilometer
(mi2) (km2)
Volume cubic foot (ft3), gallon (gal), cubic meter (m3), liter (l or L)
acre-foot (ac-ft)
Pressure/Stress pounds per sq. ft (lb/ft2), kilonewtons per sq. m (kN/m2), grams per sq.
pounds per sq. inch (psi) meter (g/m2)
Flow cubic feet per second (cfs or ft3/s), cubic meters per second (m3/s), liters per sec-
gallons per minute (gpm) ond (L/s)
Concentration milligrams per liter (mg/l)
A common situation encountered in hydrologic and hydraulic engineering is the need

to determine the units associated with the area under a graphed curve (that is, with the
integral of some function). This problem was encountered in Example 3.2. The solu-
tion is accomplished by simply multiplying the units of the two quantities represented
on each axis of the graph. If, for example, a hydrograph shows discharge on the verti-
cal axis in ft3/s and time on the horizontal axis in seconds, an area under the plotted
hydrograph represents cubic feet of water.
Example 3.5 – Determining the Equivalent Depth of Runoff. The detention pond of
Example 3.2 is located at the outlet of a 140-ac drainage basin. Given the information in Example 3.2,
determine the equivalent depth of runoff from the drainage basin.
Solution: From Example 3.2, the volume of inflow to the detention pond is 540,000 ft3, or 12.4 ac-ft.
Dividing this volume of water by the area of the drainage basin yields the equivalent depth of runoff
as
d = 12.4 ac-ft/140 ac = 0.886 ft = 1.06 in.
This 1.06 in. of runoff may be viewed as the depth of effective precipitation causing the runoff to
occur. As will be described in Chapter 4, effective precipitation is what remains of total precipitation
after surface infiltration and other rainfall abstractions have been accounted for.
Example 3.6 – Determining the Volume of Water from a Surface Area versus
Elevation Function. The area of the water surface of a detention pond, graphed as a function of
the water surface elevation, can be used to determine the volume of water in the pond. Figure E3.6.1
shows water surface area as a function of water surface elevation for a pond with a bottom elevation
of 196.0 ft. The surface areas have been determined by digitizing a contour map of the pond at 2-ft
contour intervals (for example, the area shown for a water surface elevation of 200.0 ft has been found
by determining the area represented within the 200-ft contour on the map. Using the values shown in
the figure, determine the volume of water in the pond when the water surface elevation is 200.0 ft.
Figure E3.6.1 Water surface area versus water surface elevation

Section 3.5 Chapter Summary 63
Solution: An incremental volume dS of water in the pond can be expressed as

dS = A(h)dh
where A(h) = water surface area at elevation h
dh = incremental change in water surface elevation
Integration of both sides of this expression shows that the volume of water is simply the area under a
curve of surface area expressed in terms of elevation. Using the trapezoidal rule to approximate the
area under the curve shown in Figure E3.6.1, the volume of water in the pond when the water surface
elevation is 200.0 ft (represented by the shaded area in Figure E3.6.1) is found as
S = 0.5(2,000 ft2)(2 ft) + 0.5(2,000 ft2 + 5,500 ft2)(2 ft) = 9,500 ft3 = 0.22 ac-ft
3.5 CHAPTER SUMMARY

The calculations used in analyzing and designing stormwater conveyance systems are
typically derived from one or more of the fundamental physical laws of conservation
of mass, momentum, and energy.
As it applies to hydraulics, conservation of mass basically states that, for an incom-
pressible fluid, the amount of fluid in a control volume at the end of a time interval is
equal to the amount of fluid in the control volume at the beginning of the time inter-
val, plus the amount of fluid entering the control volume during the time interval, and
minus the amount leaving the control volume during the time interval.
If flow can be considered to be steady and incompressible and changes in thermal
energy negligible, the energy at a point is taken as the sum of the pressure head (inter-
nal fluid energy), elevation head (potential energy), and velocity head (kinetic
energy). Conservation of energy between two points can be applied in the following
form:
p1 α1V12 p2 α 2V22
+ Z1 + = + Z2 + + hL − hP + hT (3.14)
γ 2g γ 2g
According to the principle of conservation of momentum, the summation of the exter-

nal forces acting on a system is equal to the time rate of change of momentum for the
system. It is a vector-valued quantity that must be evaluated separately for each direc-
tion. For the x-direction, momentum flow rate is given by
M x = βρ QVx (3.15)
A variety of units of measurement are common in the field of hydrologic and hydrau-
lic engineering. Rainfall and rainfall losses are often expressed in units of inches or
centimeters of rainfall per hour (in/hr, cm/hr), and discharges are often expressed in
units of cubic feet or meters per second (cfs, m3/s). Drainage basin areas are fre-
quently in acres (ac), hectares (ha), square miles (mi2), or square kilometers (km2),
and volumes of water are often given in cubic feet (ft3), cubic meters (m3), or acre-ft
(ac-ft).
REFERENCES
Abbott, M.B. 1979. Computational Hydraulics: Elements of the Theory of Free Surface Flows. London :
Pitman.
Chow, V.T., D.L. Maidment, and L.W. Mays. 1988. Applied Hydrology. New York: McGraw-Hill
Prasuhn, A.L. 1980. Fundamentals of Fluid Mechanics. Englewood Cliffs, N. J.: Prentice-Hall.
Shames, I.H. 1992. Mechanics of Fluids. 3d ed. New York: McGraw-Hill.
Simon, A.L. 1981. Practical Hydraulics. 2d ed. New York: John Wiley & Sons.
White, F.M. 1999. Fluid Mechanics. 4th ed. New York: McGraw-Hill.
PROBLEMS
3.1 A pipe discharges from a pond at a rate of 40,000 gpm. What is the flow in cfs?
3.2 During an emergency release, the water level in a 12,000-ac reservoir dropped by 1 inch in 6 hours.
What was the discharge rate in cfs?
3.3 The last flow gaging station on the Susquehanna River before it enters the Chesapeake Bay is at the
Conowingo Dam. The average flow recorded at this station is 2,500 m3/s. What is the total amount
of water (in m3) that flows into the Chesapeake Bay from the Susquehanna River in an average year?
3.4 A square detention pond has a 100 × 100-ft, level bottom and 3H:1V side slopes. What volume of
water can the pond hold if the depth is 5 ft?
3.5 The discharge in a stream is 125 cfs. Downstream of the monitoring point, a 30-in. storm sewer is
discharging into the stream. The pipe is flowing full and the average velocity is 5 ft/s. What is the
total discharge in the stream downstream of the discharge?
3.6 The shape of a runoff hydrograph from a subdivision for a particular storm event can be approxi-
mated as a triangle. The duration of the event was 18 hours. The peak flow was 3.4 m3/s, and it
occurred 4 hours after the start of the runoff hydrograph. What was the total volume of runoff pro-
duced?
3.7 A rainfall event is quantified as follows:
• 30 minutes at 0.25 in/hr

The precipitation occurs over a 500-ac watershed. What are the total volume of rainfall (in ft3) and
the equivalent rainfall depth (in in.)?
3.8 Water is flowing in an open channel at a depth of 4 ft and a velocity of 7.5 ft/s. It then flows down a
chute into another open channel where the depth is 2 ft and the velocity is 30 ft/s. Assuming friction-
less flow and α = 1, determine the difference in elevation between the channels.
Problems 65
3.9 During a flood, runoff flows into a large reservoir where the velocity is 0. At the reservoir’s dam, the
water overtops a rectangular spillway at a depth of 0.6 m and a velocity of 2.4 m/s. The water then
flows down a long spillway and makes a smooth transition to a discharge channel. The bottom of the
discharge channel is 13.7 m below the top of the spillway, and the depth of flow is 3.7 m. The flow is
assumed to be frictionless.
a) Calculate the velocity head and total head at locations (1) upstream of the spillway (in the reser-
voir); (2) at the crest of the spillway; and (3) in the discharge channel.
b) Plot the hydraulic and energy grade lines over the path of the water.
3.10 A 36-in. diameter pipe has a 30° horizontal bend and a flow of 100 cfs. The inlet pressure is 35 psi
and the outlet pressure is 33 psi. Calculate the direction and magnitude of the forces required to sta-
bilize the bend.
Some rainfall is
intercepted by
plants, thereby
reducing the
fraction of rainfall
that becomes
runoff.
C H A P T E R
4
Modeling Rainfall
Rainfall data are a fundamental building block for determining the amount of storm-
water runoff generated during a particular event. This chapter provides an introduc-
tion to surface water hydrology concepts needed for the design and analysis of
stormwater conveyance systems, with emphasis on rainfall processes and their repre-
sentations. Chapter 5 presents methods for modeling runoff from rainfall data.
A simple but practical way of viewing the subject of hydrology is as a means of esti-
mating the loads (that is, the discharges) for which stormwater conveyance systems
must be designed. For example, when a structural engineer designs a beam, he or she
must know or estimate the loads that the beam will be required to support. In a similar
vein, the design of a storm sewer system or highway culvert requires an estimation of
the stormwater loads that the system will be required to convey. Hydrology is the sci-
ence by which these estimates can be made.
In general, hydrologic prediction methods can be classified as being either event-
based or continuous. Event-based methods are concerned with the prediction of dis-
charge and/or volume of runoff resulting from a single rainfall event. Continuous
methods simulate runoff on an uninterrupted basis and include both wet- and dry-
weather periods. Event-based methods are the traditional methods by which stormwa-
ter conveyance systems have been designed. However, with increasing attention being
paid to water quality issues, application of continuous simulation methods is becom-
ing more common. This book covers event-based runoff prediction methods.
Section 4.1 of this chapter reviews the various processes that make up the hydrologic
cycle, and Sections 4.2 and 4.3 describe the various type and characteristics of precip-
itation. Methods of obtaining rainfall data, including rainfall measurement and
sources of historic data, are presented in Section 4.4. Several formats for presenting
rainfall data are described in Section 4.5. Finally, Section 4.6 briefly summarizes the
rainfall data requirements of several runoff prediction methods.
68 Modeling Rainfall Chapter 4
4.1 THE HYDROLOGIC CYCLE

The amount of water on Earth is constant and finite and is in constant circulation
throughout a space called the hydrosphere. Although its dimensions vary from one
part of the globe to another, the hydrosphere extends approximately 9 mi (14 km) into
the atmosphere and about 0.6 mi (1 km) into the Earth’s crust. Circulation of water
through the hydrosphere, known as the hydrologic cycle, takes place through a num-
ber of pathways. Movement of water through the hydrologic cycle is driven by solar
energy, air currents, Coriolis forces (caused by the rotation of the Earth on its axis),
and gravitational and capillary forces. As water moves through the hydrologic cycle,
heat energy stored within the water also moves, and thus the hydrologic cycle is a
major factor in the Earth’s global heat balance.
Many different representations of the hydrologic cycle can be found in textbooks that
cover the topic. Some are pictorial (as shown in Figure 4.1), and others are more sys-
tem-oriented and take the form of link/node diagrams similar to flowcharts (see Fig-
ure 4.2). Both representations show in extremely simplified terms the various
pathways that water takes as it passes through the hydrologic cycle. However, not all
of these pathways need to be considered in every engineering problem. For example,
in the design of a stormwater drainage system for a paved parking lot, infiltration and
evaporation are usually ignored because of the imperviousness of the pavement mate-
rial and the negligible amount of evaporation expected to occur over the relatively
short time interval associated with a storm.
Figure 4.1
Pictorial
representation of the
hydrologic cycle
Credit: Chow, Maidment, and Mays (1988). Reproduced with permission of the McGraw-Hill Companies.
There is no beginning or end to the hydrologic cycle; water circulates through the
cycle continuously. Thus, a description of its various components could begin at any
point. It is common, however, to begin with precipitation. Precipitation can occur in
Section 4.1 The Hydrologic Cycle 69
Figure 4.2
Flowchart
hydrologic cycle
many forms, but the focus in this book will be rainfall only. When rainfall occurs over
a drainage basin, a certain amount of it is caught by vegetation before it reaches the
ground surface; this portion is referred to as interception. The rainfall that does make
its way to the ground surface can take any of a number of pathways, depending on the
nature of the ground surface itself. Some water may infiltrate into the ground, and
some may be trapped in low areas and depressions. The water that neither infiltrates
nor gets trapped as depression storage contributes to direct surface runoff. This sur-
face runoff may combine with subsurface flow and/or groundwater outflow (base
flow) to become streamflow supplying rivers, lakes, and oceans. Water is returned to
the atmosphere by evaporation from vegetation, soils, and bodies of water, and by
transpiration of vegetation. (Evaporation and transpiration from vegetation may
together be referred to as evapotranspiration.)
4.2 PRECIPITATION TYPES

Forms of precipitation include rainfall, hail, and snowfall. Precipitation occurs when
water vapor in the atmosphere is converted into a liquid or solid state. It takes place
only when water vapor is present in the atmosphere (which is virtually always true)
and a cooling mechanism is present to facilitate condensation of the vapor. The cool-
ing and resulting precipitation can take place in any of several ways, as described in
the subsections that follow. As stated previously, the focus in this text will be on rain-
fall only. Readers needing to address other forms of precipitation, especially snow, are
referred to Singh (1992) for useful introductory coverage.
Frontal Precipitation
Frontal precipitation occurs when a warm air mass, moved by wind currents and
atmospheric pressure gradients, overtakes and rises above a cooler air mass. The ris-
ing of the warm, moisture-laden air to a higher altitude causes it to cool (unless there
is a temperature inversion), and the water vapor condenses. The precipitation arising
from this process often extends over large areas, and the interface between the warm
and cool air masses is called a warm front, as illustrated in Figure 4.3.
A cold front occurs when a cold air mass overtakes a warmer one and displaces the
warm air upwards. Again, the rising and cooling of the warm air is what causes con-
densation to occur. Unlike warm fronts, the precipitation arising from a cold front is
frequently spotty and often covers relatively small areas.
Figure 4.3
Frontal precipitation
Cyclonic Precipitation
A cyclone is a region of low pressure into which air flows from surrounding areas
where the pressure is higher. Flow around the low pressure center moves in a counter-
clockwise direction in the northern hemisphere and in a clockwise direction in the
southern hemisphere. The direction of the rotation is determined by Coriolis forces,
which are forces caused by the tangential velocity of the Earth at the equator being
faster than its tangential velocity at other latitudes.
As air masses converge on the low pressure area, the incoming mass of air must be
balanced by an outgoing one. Because air is entering from all directions horizontally,
the outgoing air has no choice but to move vertically upward. The upward motion of
Section 4.2 Precipitation Types 71
the outgoing air mass leads to cooling and condensation of water vapor. Precipitation
of this type is called cyclonic precipitation (see Figure 4.4).
Figure 4.4
Cyclonic precipitation
Convective Precipitation
Convective precipitation tends to take place in small, localized areas (often no larger
than a square mile or a few square kilometers) and is caused by differential heating of
an air mass, as illustrated in Figure 4.5. This differential may occur in urban areas or
during summer months when air near the ground surface becomes heated and rises
with respect to the cooler surrounding air. This rise can be quite rapid and often
results in thunderstorms. These events are very difficult to predict because they occur
in short time frames. They can yield very intense rainfall.
Orographic Precipitation
Orographic precipitation occurs when an air mass is forced by topographic barriers to
a higher altitude where the temperature is cooler, as illustrated in Figure 4.6. This type
of precipitation is common in mountainous regions where air currents are forced up
and over the tops of the mountains by wind movement. When the air rises to a cooler
altitude, condensation occurs. Orographic precipitation can be quite pronounced on
the windward side of a mountain range (the side facing the wind), while there is often
a rain shadow where there is relatively little precipitation on the leeward side.
Figure 4.5
Convective
precipitation
Figure 4.6
Orographic
precipitation
4.3 BASIC RAINFALL CHARACTERISTICS

For the design engineer, the most important characteristics of rainfall are
• The depth, or volume, of rainfall during a specified time interval (or equiva-
lently, its average intensity over that time interval)
• The duration of the rainfall
• The area over which the rainfall occurs
• The temporal and spatial distributions of rainfall within the storm
• The average recurrence interval of a rainfall amount
Section 4.3 Basic Rainfall Characteristics 73
Depth and Intensity

The depth of rainfall is the amount of rainfall, in inches or millimeters, that falls
within a given duration or time period. The rainfall intensity represents the rate at
which rainfall occurs. The average intensity for a period is simply the rainfall depth
divided by the time over which the rainfall occurs.
Temporal Distributions
The temporal rainfall distribution, shown in Figure 4.7, is a good visualization tool
for demonstrating how rainfall intensity can vary over time within a single event. The
y-axis is represented by a simple rain gauge that fills over a certain period of time,
which is shown on the x-axis. Total depth is simply the final depth in the gauge. Aver-
age intensity during any segment of the storm is represented by the slope of the rain-
fall curve. A steeper slope for a given curve segment corresponds to a greater average
intensity during that segment.
The temporal distribution shown in Figure 4.7 can also be represented using a bar
graph that shows how much of the total rainfall occurs within each time interval dur-
ing the course of an event. A graph of this nature is called a rainfall hyetograph. Hye-
tographs can be displayed in terms of incremental rainfall depth measured within each
time interval (Figure 4.8), or as the average intensity calculated for each interval by
dividing incremental depth by the time interval (Figure 4.9). The latter approach,
shown in Figure 4.9, is recommended.
Figure 4.7
Temporal distribution
of rainfall
Figure 4.8
A hyetograph of
incremental rainfall
depth versus time
Figure 4.9
A hyetograph of
incremental rainfall
intensity versus time
Spatial Distributions
The spatial distribution of rainfall relates to the issue of whether the rainfall depths or
intensities at various locations in a drainage basin are equal for the same event. The
spatial aspect of rainfall is not covered in detail in this text because, in practice, spatial
variations for relatively small drainage basins can be neglected. However, areal rain-
fall reduction factors are presented later in this chapter to demonstrate the adjustment
of point rainfall values for application with large drainage areas.
Recurrence Interval
The probability that a rainfall event of a certain magnitude will occur in any given
year is expressed in terms of recurrence interval (also called return period or event
frequency). The recurrence interval is the average length of time expected to elapse
between rainfall events of equal or greater magnitude. It is a function of geographic
location, rainfall duration, and rainfall depth. Although recurrence interval is
expressed in years, it is actually based on a storm event’s exceedance probability,
which is the probability that a storm magnitude will be equaled or exceeded in any
Section 4.3 Basic Rainfall Characteristics 75
Time-Series Data for Rainfall

Statistical analyses of raw rainfall data are used to data points if there are n years of record.) In any case,
assign recurrence intervals to various rainfall depths or the partial-duration series approach explicitly recog-
intensities for selected storm durations. To accomplish nizes that more than one large event may occur in any
the statistical analyses, one must first extract either an year, and that the second or even third largest event in a
annual series of data or a partial-duration series of data year may be larger than the largest event in another
from the raw rainfall data records. year.
An annual series is constructed by extraction of the sin- In practice, the results of annual and partial-duration
gle largest depth or intensity within each year of record series analyses are found to lead to essentially the same
for a chosen duration. If one has n years of data, the results for recurrence intervals of 10 years or more.
resulting annual series contains n data points—one That is, the precipitation depth or intensity for a
from each year. To construct a partial-duration series, selected storm duration and recurrence interval is essen-
one begins by choosing a precipitation threshold value. tially the same regardless of the type of data series used.
Then, any depth or intensity exceeding the chosen However, for recurrence intervals smaller than 10 years,
threshold for the chosen duration is extracted from the partial-duration series analyses generally give rise to
records. Thus, if n years of record are available, one higher rainfall depths than do annual series analyses.
may end up with more or fewer than n data points Thus, in the interest of obtaining conservatively high
depending on the magnitude of the chosen threshold. estimates of rainfall amounts, an engineer might choose
(The U.S. National Weather Service implicitly chooses to use results based on partial-duration series analyses.
the threshold so that the partial-duration series has n
given year. The relationship between recurrence interval and exceedance probability
is given by
T = 1/ P (4.1)
where T=return period (years)

P=exceedance probability
For example, a 25-year return period represents a storm event that is expected to occur
once every 25 years, on average. This does not mean that two storm events of that size
will not occur in the same year, nor does it mean that the next storm event of that size
will not occur for another 25 years. Rather, a 4-percent chance of occurrence exists in
any given year.
Example 4.1 – Computing Recurrence Intervals. What is the recurrence interval for a
storm event that has a 20-percent probability of being equaled or exceeded within any given year? A
2-percent probability? A 200-percent probability?
Solution: For a storm event with a 20-percent probability of being equaled or exceeded in a given
year, the recurrence interval is computed as
1 / 0.20 = 5 yr
For an event with a 2-percent probability, the recurrence interval is
1 / 0.02 = 50 yr
For an event with a 200-percent probability, the recurrence interval is
1 / 2.00 = 0.5 yr = 6 months
4.4 OBTAINING RAINFALL DATA

A number of types of gauges are available for measuring rainfall, and these measure-
ments have varying degrees of accuracy. In most locations, rainfall data from gauging
stations are compiled by a government agency and are readily available to the engi-
neer for use in modeling. Very often, the raw rainfall data have already been pro-
cessed and used to develop simplified techniques for engineers to use in generating
design storms. This section presents information on methods of rainfall collection and
where to obtain rainfall data.
Rainfall Data Collection Methods

In the United States, as in most countries, a national network of gauging sites exists
through which rainfall data are measured and recorded on a regular basis. Historically,
these data were collected using non-recording gauges at a frequency of once per day.
Many gauging installations are now set up to collect data on an hourly basis using
automatic recording gauges. In some instances, gauging sites are set up to collect data
at even smaller time intervals (usually 15 minutes), but the number of these sites is
small compared to the numbers of hourly and daily recording sites.
The standard nonrecording gauge used by the U.S. National Weather Service is an
8-in. (20.3-cm) diameter cylinder with a funnel mounted inside its top. The funnel
directs rainwater into a cylindrical measuring tube where its depth is measured with a
dipstick. The cross-sectional area of the measuring tube is one-tenth that of the 8-in.
diameter gauge opening, so that 0.01 in. (0.254 mm) of precipitation over the gauge
appears as 0.10 in. (2.54 mm) of water in the measuring tube. Measurements of the
depth of accumulated water in the standard gauge are made once daily, usually by vol-
unteer observers. An illustration of the standard gauge is shown in Figure 4.10.
Three types of commonly used recording rain gauges are weighing gauges, float
gauges, and tipping-bucket gauges. Weighing gauges, shown in Figure 4.11, employ a
scale or spring balance to record the cumulative weight of the collection can and
precipitation over time. The weight of the precipitation is then converted into an
equivalent volume and depth of rainfall using the specific weight of water and the
cross-sectional area of the gauge opening.
Float gauges record the rise of a float as the water level rises in a collection chamber.
A tipping-bucket gauge, shown in Figure 4.12, operates by means of a two-compart-
ment bucket. When a certain quantity (depth) of rainfall fills one compartment, the
weight of the water causes it to tip and empty, and the second compartment to move
into position for filling. Similarly, the second bucket tips when full and moves the first
compartment back into position. This process continues indefinitely. Records of the
bucket tip times, combined with knowledge of the compartment volumes, are used to
estimate rainfall depths and rates.
Another emerging technology for measuring rainfall is the optical gauge or
disdrometer. The basic principle behind the optical gauge is the use of light sources
and line-scan cameras to measure the size/shape and velocity of raindrops. As shown
in Figure 4.13, the intersection of orthogonal light sources defines the measuring area.
Section 4.4 Obtaining Rainfall Data 77
Figure 4.10
A standard rainfall
gauge
Credit: Linsley, Kohler, and Paulus, 1982. Reproduced with permission of the McGraw-Hill Companies.
Figure 4.11
A weighing gauge
Credit: Chow, 1964. Reproduced with permission of the McGraw-Hill Companies.
The raindrops that fall through these beams cast a shadow that is recorded by the cam-
eras. The image is processed to determine the size and shape of the drops. Because the
two light sources are separated by a small distance, the vertical velocity of the rain-
drop can also be determined. The velocity and size and shape of the raindrop can then
be used to estimate the rainfall intensity.
Figure 4.12
A tipping bucket
gauge
Credit: Chow, 1964. Reproduced with permission of the McGraw-Hill Companies.
Figure 4.13
Camera and light
source configuration
for a two-dimensional
video disdrometer
Credit: A. Kruger, W.F. Krajewski, and M.J. Kundert, IIHR–Hydroscience and Engineering, The University of Iowa, Iowa City, Iowa.
Regardless of the type of rainfall gauge used, errors inevitably occur in rainfall mea-
surements. As a general rule, these errors lead to underestimation of the actual rain-
fall. Errors arise from many sources, including reading of depths for small amounts of
rain, dipstick displacement, recording of tipping times in tipping bucket gauges, evap-
oration, and frictional effects in weighing and float gauges. However, the greatest
source of error in rainfall gauging is the effect of wind, which may cause gauge catch
deficiencies of more than 20 percent (Larson and Peck, 1974). Windshields of various
designs have been developed to reduce the effects of wind on gauge catch, but errors
persist nevertheless.
In recent decades, much attention has been given to measurement of precipitation
using remote sensing technologies, the most prominent of which is radar (see Figure
Section 4.4 Obtaining Rainfall Data 79
4.14). The U.S. National Weather Service, as a part of its recent modernization pro-
gram, has installed a national network of radar sites. This network often provides
much better spatial resolution of the extent of precipitation events than gauges do, but
greater rainfall depth uncertainty exists with radar information. Radar is particularly
useful for tracking the movement of storms and tornadoes. Attention is also being
given to the use of radar data for flash flood prediction and flood warning systems
(Sweeney, 1992; Johnson et al., 1998).
Figure 4.14
U.S. National Weather
Service radar image
showing rainfall over
southern Florida
Rainfall Data Sources

Rainfall data can be obtained from a number of sources. Some sources provide raw
rainfall data that has not been subjected to analyses other than those related to quality
control. Others provide products that have been derived from various methods of pro-
cessing of raw data. Processed data are used in most engineering analyses, but it
should be recognized that many of the sources of processed data are now somewhat
dated. Raw data sources can be used to obtain detailed information about a particular
rainfall gauging location, and of course are the most basic information used in all
types of rainfall analyses.
Raw rainfall data typically can be obtained from the meteorological bureau of the
country of interest. In the United States, for example, raw rainfall data for any gaug-
ing site within the national network can be obtained from the National Climatic Data
Center (NCDC), which is operated by the National Oceanic and Atmospheric Admin-
istration (NOAA). The NCDC has made much of the raw data available on the Inter-
net. Global data are also available through the NCDC website.
In addition to national meteorological bureaus, other government organizations, as
well as private organizations, may be useful rainfall data sources. The U.S. Geologi-
cal Survey (USGS), for instance, as a part of its research programs, occasionally col-
lects rainfall data at a number of sites in a given geographical area. These data may
exist in “unit values” files and can be obtained by contacting the nearest office of the
USGS. Over the last decade or so, raw data also have been made available on CD-
ROM by private organizations such as EarthInfo, Inc. of Boulder, Colorado, a com-
pany that supplies climate and precipitation data for stations throughout the world.
For most practical engineering design and analysis purposes, the main sources of pre-
cipitation information consist of processed rainfall data. This processed information
may be available from government organizations involved in hydrologic analysis, or
from local institutions that conduct this type of research. In the United States, such
information may be found in several National Weather Service publications. For
example, the NWS publication TP 40 (Hershfield, 1961) presents maps showing pre-
cipitation depths in the United States for storm durations from 30 minutes to 24 hours
and for recurrence intervals from 1 to 100 years. TP 40 was partially superseded by a
later publication known as HYDRO-35 for the central and eastern United States (Fre-
derick, Myers, and Auciello, 1977), and by the NOAA Atlas 2 for the 11 coterminous
western states (Miller, Frederick, and Tracey, 1973). Again, these documents contain
maps showing precipitation depths for given durations and recurrence intervals. How-
ever, unlike TP 40, they can be used to obtain information for durations as short as 5
minutes. Figures 4.15 through 4.20 illustrate the six maps contained in HYDRO-35.
Note that TP 40, HYDRO-35, and the NOAA Atlas 2 are now rather dated. A number
of efforts to update the information presented in these publications have been initiated
in recent years. Durrans and Brown (2001) summarize a number of recent studies that
have been performed in the United States. Practitioners should be on the watch for
new developments in the locale(s) where they practice.
Figure 4.15
2-year, 5-minute
precipitation depths in
inches (Frederick,
Section 4.4
Myers, and Auciello,

1977)
Obtaining Rainfall Data
81
Figure 4.16
82
2-year, 15-minute
inches (Frederick,
1977)
Modeling Rainfall
Chapter 4
Figure 4.17
2-year, 60-minute
inches (Frederick,
Section 4.4

1977)
83
Figure 4.18
84
100-year, 5-minute
inches (Frederick,
1977)
Modeling Rainfall
Chapter 4
Figure 4.19
100-year, 15-minute
inches (Frederick,
Section 4.4

1977)
85
Figure 4.20
86
100-year, 60-minute
inches (Frederick,
1977)
Modeling Rainfall
Chapter 4
Section 4.5 Types of Rainfall Data 87
4.5 TYPES OF RAINFALL DATA

Information on rainfall is available in a variety of forms. IDF curves provide average
rainfall intensity information for particular storm events. Hyetographs provide rainfall
depths or intensities at successive time intervals over the course of a storm. They can
describe recorded rainfall events, or they may be generated using synthetic temporal
rainfall distributions. The subsections that follow describe rainfall data types in more
detail.
Intensity-Duration-Frequency Curves
For a selected storm duration, a rainfall intensity exists that corresponds to a given
exceedance probability or recurrence interval. A rainfall intensity-duration-frequency
curve (also referred to as an IDF curve) illustrates the average rainfall intensities cor-
responding to a particular storm recurrence interval for various durations.
Figure 4.21 shows sample IDF curves for 2-, 5-, 10-, 25-, 50-, and 100-year storms.
These curves are the result of the statistical analysis of rainfall data for a particular
area. Given the information on this graph, one can determine that the average one-
hour rainfall intensity expected to be equaled or exceeded, on average, once every 25
years is 3.0 in/hr (76 mm/hr).
Figure 4.21
IDF curves for 2-
through 100-year
return period events
IDF curves can also be used to determine the recurrence interval associated with an
observed storm rainfall intensity. That is, if one were to observe the intense gauged
rainfall event shown in Figure 4.22, a set of IDF curves for that locale (shown in Fig-
ure 4.21) could be used to estimate the recurrence interval associated with that storm
event. The storm in Figure 4.22 has an overall average intensity equal to 5 in./4 hr =
1.25 in/hr (32 mm/hr), but there would be periods during the 4-hour duration with
both higher and lower intensities than the average. From Figure 4.21, an average
intensity of 1.25 in/hr (32 mm/hr) over a duration of 4 hours is approximately a 50-
year event.
Figure 4.22
A sample gauged
rainfall event
Total Depth = 5.0 in.

5.0
4.0
Rain Depth, in.
Total Duration = 4.0 hr

3.0
2.0
1.0
0.0
0.0 1.0 2.0 3.0 4.0
Time, hr
Computer programs commonly access IDF data in the form of an equation. Several
forms have been developed to analytically describe the graphical IDF relationships.
The most common forms of these equations are (Haestad, 2003):
a
i= (4.2)
(b + D )
n
a ( RP )
m
i= (4.3)
(b + D )
n
i = a + b(ln D) + c(ln D)2 + d (ln D)3

(4.4)
where i = intensity of rainfall (in/hr, mm/hr)

D = rainfall duration (min)
RP = return period (yr)
a, b, c, d, m, and n are coefficients used to describe the IDF relationship
IDF Data Sources. As stated earlier in this chapter, hydrology can be thought of
as providing the loading for the design of stormwater conveyance systems. One of the
most widely used methods for estimating peak stormwater runoff rates for small
drainage basins is the rational method (see Section 5.4, page 140 for more informa-
tion). To use the rational method, one must have available a set of IDF curves devel-
oped for the specific geographic location where the conveyance system is to be built.
A single set of IDF curves may be applicable over a fairly large area (for instance, an
entire city or county), given that there are no significant topographic changes (such as
large ridges) that cause rainfall characteristics across the area to vary.
Many drainage jurisdictions and agencies can provide the engineer with IDF data rec-
ommended for their particular geographic location. Engineers should understand
when and by whom the IDF curves were created because more recently updated
resources may be available.
Creating I-D-F Curves from HYDRO-35. A widely used procedure for
developing IDF curves for the central and eastern United States is presented in
HYDRO-35. The procedure uses the six maps shown in Figures 4.15 through 4.20.
For a chosen location, one determines the depths of precipitation for the various com-
binations of storm durations and recurrence intervals represented by the six maps.
Storm depths for other durations and return periods are then found by interpolation
between the map values. For a fixed recurrence interval, rainfall depths for the 10-
minute duration are computed from those for 5- and 15-minute durations as follows:
P10 = 0.41P5 + 0.59P15 (4.5)
Similarly, for a given recurrence interval, depths for the 30-minute duration are com-
puted as
P30 = 0.51P15 + 0.49P60 (4.6)
When the storm duration is fixed, depths for recurrence intervals other than 2 or 100
years are computed as
PT-yr = aP2-yr + bP100-yr (4.7)
where a and b are coefficients given in Table 4.1.

The coefficients a and b in this last expression depend on the recurrence interval, T, as
shown in Table 4.1. The coefficients a and b in the table and the coefficients in the P10
and P30 estimation equations apply to all locations in the central and eastern United
States covered by HYDRO-35.
After the depths of rainfall are determined for various storm durations and recurrence
intervals by using the procedures described previously in this section, they can be
converted into average intensities by dividing each depth by its corresponding dura-
tion. The IDF curves are then constructed by graphing the average intensities as func-
tions of duration and recurrence interval.
Table 4.1 Coefficients for computation of IDF curves

(Frederick, Myers, and Auciello, 1977)
T (yr) a b
5 0.674 0.278
10 0.496 0.449
25 0.293 0.669
50 0.146 0.835
Example 4.2 – Developing IDF Curves from HYDRO-35 Isohyetal Maps. Using
the procedures outlined in HYDRO-35, develop a set of IDF curves for the U.S. city of Memphis,
Tennessee, for durations from 5 minutes to 1 hour and for recurrence intervals from 2 to 100 years.
Solution: From the six isohyetal maps contained in HYDRO-35 and shown in Figures 4.15 through
4.20, the following rainfall depths are obtained for Memphis (located at the southwest corner of the
state of Tennessee):
2-yr, 5-min depth: 0.49 in. 100-yr, 5-min depth: 0.83 in.
Table 4.2 contains these depths (underlined) as well as other depths that have been interpolated
between them using Equations 4.5 through 4.7. Within any of the last five columns of the table, pre-
cipitation depths for recurrence intervals other than 2 or 100 years are determined from the 2- and
100-year values using Equation 4.3 and the coefficients in Table 4.1. Depths in the third column are
then determined from the adjacent depths in the second and fourth columns by using Equation 4.1,
and depths in the fifth column are determined from the adjacent depths in the fourth and sixth col-
umns by using Equation 4.2.
Table 4.2 Total Rainfall Depths (in.)
Duration
T (yr) 5 min 10 min 15 min 30 min 60 min
2 0.49 0.80 1.01 1.39 1.79
5 0.56 0.93 1.18 1.69 2.22
10 0.62 1.02 1.30 1.90 2.53
25 0.70 1.17 1.49 2.22 2.97
50 0.76 1.28 1.64 2.46 3.31
100 0.83 1.40 1.79 2.70 3.65
Intensities are computed from the table of rainfall depths by dividing the rainfall depths in each col-
umn by their corresponding duration and expressing the result in units of inches per hour. These com-
putations lead to the results in Table 4.3.
Table 4.3 Intensity (in/hr)
Duration
T (yr) 5 min 10 min 15 min 30 min 60 min
2 5.88 4.78 4.04 2.78 1.79
5 6.73 5.55 4.71 3.38 2.22
10 7.39 6.13 5.22 3.81 2.53
25 8.39 7.01 5.97 4.43 2.97
50 9.18 7.69 6.57 4.92 3.31
100 9.96 8.38 7.16 5.40 3.65
A graph of the values in Table 4.3 is the desired set of IDF curves, shown in Figure E4.2.1.
Figure E4.2.1 Derived IDF Curves
Note that rainfall amounts obtained from TP 40, HYDRO-35, and the NOAA Atlas 2,
and hence from a set of IDF curves developed on the basis of those publications, rep-
resent rainfall depths and intensities at a single point in space. If one requires an aver-
age intensity over a large area (such as a large drainage basin), the point depths and
intensities should be reduced somewhat to reflect the reduced probability of a storm
occurring with the same intensity over this large area. Reduction percentages are indi-
cated in Figure 4.23 (U.S. Weather Bureau, 1958) and depend on the area and the
storm duration of interest. For example, the 10-year, 1-hour point rainfall intensity at
a particular location should be multiplied by a factor of approximately 0.72 if it is to
be applied to a drainage basin with an area of 100 mi2 (259 ha). It is generally
accepted that the curves shown in Figure 4.23 can be applied for any recurrence inter-
val and for any location in the United States, although some differences have been
noted between different regions.
Temporal Distributions and Hyetographs for Design

Storms
Some types of hydrologic analysis require the distribution of precipitation over the
duration of the storm. For example, the depth of a 100-year, 1-hour rainfall event in
Tuscaloosa, Alabama, is estimated from HYDRO-35 to be about 3.8 in. (96.5 mm).
But how much of this rainfall should be considered to occur in, say, the first 10 min-
utes of the storm? What about the second 10 minutes, or the last 10 minutes?
Figure 4.23
Reduction of point
precipitation based on
the area of the
drainage basin (U.S.
Weather Bureau,
1958)
A temporal distribution such as the one in Figure 4.7 shows the cumulative progres-
sion of rainfall depth throughout a storm. A rainfall hyetograph (Figures 4.8 and 4.9)
shows how the total depth (or intensity) of rainfall in a storm is distributed among var-
ious time increments within the storm. Figures such as these can be used to represent
the pattern of an actual recorded storm event.
When the design of a stormwater management facility requires a complete rainfall
hyetograph, the engineer is faced with selecting the appropriate storm distribution to
use in designing and testing for unknown, future events. In design situations, engi-
neers commonly use synthetic temporal distributions of rainfall. Synthetic distribu-
tions are essentially systematic, reproducible methods for varying the rainfall
intensity throughout a design event.
The selected length of the time increment, ∆t, between the data points used to con-
struct a temporal rainfall distribution, depends on the size (area) and other characteris-
tics of the drainage basin. As a rule of thumb, the time increment should be no larger
than about one-fourth to one-fifth of the basin lag time, or about one-sixth of the time
of concentration of the basin. The Natural Resource Conservation Service (NRCS,
formerly known as the Soil Conservation Service, or SCS) recommends using ∆t =
0.133tc = 0.222tL [see Section 5.3 for explanations of basin lag time (tL) and time of
concentration (tc)], but notes that a small variation from this is permissible (SCS,
1969). The smallest time increment for which rainfall data are generally available is
about 5 minutes. In small urban drainage basins where it is often necessary to use
time increments as small as 1 or 2 minutes, this data must be interpolated.
In addition to selecting an appropriate ∆t, the engineer must select the total duration
to be used when developing a design storm hyetograph. In many cases, the storm
duration will be specified by the review agency having jurisdiction over the area in
which a stormwater conveyance facility will be built; this approach promotes consis-
tency from one design to another.
When the design storm duration is not specified, the design engineer must select it.
The recommended way to accomplish this is to review rainfall records for the locality
of interest to determine typical storm durations for the area. Seasonal characteristics
of storm durations may exist, with longer storms tending to occur in one season and
shorter storms in another. An alternative approach is to perform runoff calculations
using a number of assumed design storm durations. With this approach, the engineer
can select the storm duration giving rise to the largest calculated flood peak and/or
runoff volume.
U.S. NRCS (SCS) Synthetic Temporal Distributions. Many methods
have been proposed for distributing a total rainfall depth throughout a storm to
develop a design storm hyetograph. The NRCS developed a set of synthetic distribu-
tions commonly used in the United States to create 24-hour synthetic design storms
(SCS, 1986). The dimensionless distribution data from the NRCS given in Table 4.4
provides fractions of the total accumulated rainfall depth over time for storms with
24-hour durations. (Figure 4.24 depicts this table graphically.) The storms are classi-
fied into various types, with each type being recommended for use in a certain U.S.
geographic region, as shown in Figure 4.25. If necessary, interpolation may be
employed to obtain values not shown in Table 4.4. Nonlinear interpolation methods
are recommended for this purpose, although many practicing engineers use simpler
linear interpolation instead. The NRCS (SCS, 1992) and many computer programs
provide data in increments of 0.1 hours.
Table 4.4 NRCS dimensionless storm distributions (adapted from SCS, 1992)
t (hr) Type I Type IA Type II Type III
0 0.000 0.000 0.000 0.000
1 0.017 0.020 0.011 0.010
2 0.035 0.050 0.022 0.020
3 0.054 0.082 0.035 0.031
4 0.076 0.116 0.048 0.043
5 0.100 0.156 0.063 0.057
6 0.125 0.206 0.080 0.072
7 0.156 0.268 0.099 0.091
8 0.194 0.425 0.120 0.114
9 0.254 0.520 0.147 0.146
10 0.515 0.577 0.181 0.189
11 0.623 0.624 0.235 0.250
12 0.684 0.664 0.663 0.500
13 0.732 0.701 0.772 0.750
14 0.770 0.736 0.820 0.811
15 0.802 0.769 0.854 0.854
16 0.832 0.801 0.880 0.886
17 0.860 0.831 0.902 0.910
18 0.886 0.859 0.921 0.928
Table 4.4 (cont.) NRCS dimensionless storm distributions (adapted from SCS, 1992)
t (hr) Type I Type IA Type II Type III
19 0.910 0.887 0.938 0.943
20 0.932 0.913 0.952 0.957
21 0.952 0.937 0.965 0.969
22 0.970 0.959 0.977 0.981
23 0.986 0.980 0.989 0.991
24 1.000 1.000 1.000 1.000
Figure 4.24
Graphical
representation of
NRCS (SCS) rainfall
distributions
Figure 4.25
Coverage of NRCS
(SCS) rainfall
distributions (SCS,
1986)
Example 4.3 – Developing a Design Storm Hyetograph from SCS Distribu-

tions. Develop a design storm for a 50-year, 24-hour storm in Boston, Massachusetts. Assume that
∆t = 0.1 hr is a reasonable choice for the drainage basin to which the design storm will be applied,
and that the total rainfall depth is 6.0 in.
Solution: Figure 4.25 illustrates that a Type III storm distribution is a reasonable choice for Boston.
Table 4.5 illustrates the calculation of the storm hyetograph.
The first column of the table is the time, in hours, since the beginning of the storm, and is tabulated in
1-hr increments for the total storm duration of 24 hours. In actuality, the ∆t used in the calculations
would be 0.1 hr; the 1-hr increment is used here for brevity.
The second column is the fraction of the total storm depth that has accumulated at each time during
the storm. These values are obtained from Table 4.4 for the Type III storm distribution. (The values
would be obtained by interpolation in the case of a 0.1-hr increment.) The third column contains the
cumulative rainfall depths for each time during the storm and is obtained by multiplying each fraction
in the second column by the total storm depth of 6.0 in. The fourth column contains the incremental
depths of rainfall for each time interval during the storm; these values are computed as the difference
between the concurrent and preceding values in the third column.
Table 4.5 50-year, 24-hour storm hyetograph for Boston, Massachusetts
t (hr) Fraction Cum. P (in.) Incr. P (in.)

0 0.000 0.000 —
1 0.010 0.060 0.060
2 0.020 0.120 0.060
3 0.031 0.186 0.066
4 0.043 0.258 0.072
5 0.057 0.342 0.084
6 0.072 0.432 0.090
7 0.091 0.546 0.114
8 0.114 0.684 0.138
9 0.146 0.876 0.192
10 0.189 1.134 0.258
11 0.250 1.500 0.366
12 0.500 3.000 1.500
13 0.750 4.500 1.500
14 0.811 4.866 0.366
15 0.854 5.124 0.258
16 0.886 5.316 0.192
17 0.910 5.460 0.144
18 0.928 5.568 0.108
19 0.943 5.658 0.090
20 0.957 5.742 0.084
21 0.969 5.814 0.072
22 0.981 5.886 0.072
23 0.991 5.946 0.060
24 1.000 6.000 0.054
The resulting graph of cumulative precipitation is shown in Figure E4.3.1, and the hyetograph is
shown in Figure E4.3.2. The height of each bar on the hyetograph is the average rainfall intensity dur-
ing that time interval, and the area of each bar is the incremental rainfall depth during that time inter-
val. Because the time increment on the graph is 1 hr (a 1-hr increment is shown in the graphic for
simplicity, though the actual ∆t is 0.1 hr) the value for the height of the bar (in units of in./hr) is equal
to the incremental depth for that time increment (in in.).
Figure E4.3.1 Graph of derived design storm cumulative precipitation
Figure E4.3.2 Derived design storm hyetograph

Alternating Block Method. Another method for developing design storm hye-
tographs is called the alternating block method and is illustrated in Example 4.4. With
this method, the resulting hyetograph is consistent with the IDF curves for a particular
location, and is in fact developed from those IDF curves. This method can be used for
cases where the storm duration is not 24 hours, as is required for the SCS method.
Example 4.4 – Developing a Design Storm Hyetograph Using the Alternating

Block Method. Using the IDF curves developed in Example 4.2, develop a design storm hyeto-
graph for a 25-yr, 1-hr storm in Memphis, Tennessee using the alternating block method. Use ∆t = 10
min for development of the hyetograph.
The first two columns in Table 4.6 are duration and intensity values from the Memphis IDF curves for
a recurrence interval of 25 years. The durations and intensities are tabulated every ∆t = 10 min for the
total storm duration of 60 min. Values in the third column are cumulative depths associated with each
duration and are equal to the duration multiplied by the intensity. Incremental depths are shown in the
fourth column and are obtained by subtracting sequential values in the third column.
Table 4.6 Calculation of incremental precipitation depths for 25-yr, 1-hr storm
Duration (min) Intensity (in./hr) Cum. P (in.) Incr. P (in)

10 7.02 1.17 1.17
20 5.26 1.75 0.58
30 4.42 2.21 0.46
40 3.84 2.56 0.35
50 3.36 2.8 0.24
60 2.97 2.97 0.17
Table 4.7 demonstrates the essence (and the source of the name) of the alternating block method. The
first column is a tabulation of each time interval during the 1-hr design storm, and the second is the
corresponding rainfall depth in each time interval. The second column is formed by taking the largest
incremental depth from the last column of Table 4.6 (1.17 in.) and placing it about one-third to one-
half of the way down the column (in this example, in the 20-to-30-min. time interval). The next larg-
est incremental depth from Table 4.6 (0.58 in.) is then placed immediately after the largest incremen-
tal depth (in this case, in the 30 to 40-min. time interval). The third largest depth (0.46 in.) is placed
before the largest depth (in the 10 to 20-min. interval). The remainder of the second column is filled
by proceeding in this fashion, taking the largest remaining value from the last column of Table 4.6
and placing it in an alternating way either after or before the other entries in Table 4.7. Finally, the last
column in Table 4.7 is computed by dividing each of the entries in the second column by ∆t and
expressing the result in units of in./hr. Figure E4.4.1 illustrates the resulting hyetograph.
Table 4.7 Formulation of the design storm hyetograph (alternating block method)
Time (min) Incr. P (in.) Intensity (in./hr)

0–10 0.24 1.44
10–20 0.46 2.76
20–30 1.17 7.02
30–40 0.58 3.48
40–50 0.35 2.1
50–60 0.17 1.02
Figure E4.4.1 Design hyetograph derived from the alternating block method
Other Synthetic Temporal Distributions. Other methods of developing

design storm hyetographs also exist. Notable among them are the quartile-based
storm patterns presented by Huff (1967) and the instantaneous intensity method pro-
posed by Keifer and Chu (1957). The engineer should use care when applying more
recently developed synthetic design storms, because they will typically result in sig-
nificantly different runoff hydrographs and may not be acceptable to the review
agency having jurisdiction.
The original studies by Huff were updated and expanded for the Illinois State Water
Survey Bulletin 71 (Huff and Angel, 1992), which provides rainfall data for nine
states in the U.S. Midwest: Illinois, Indiana, Iowa, Kentucky, Michigan, Minnesota,
Missouri, Ohio, and Wisconsin. Figure 4.26 shows some of the temporal distributions
from Bulletin 71. Because both axes are dimensionless, the engineer can vary both the
total depth of rainfall and the storm duration. The four temporal distributions corre-
spond to different duration ranges (td) and the quartile of the storm during which the
peak intensity occurs. As a general rule of thumb, these curves indicate that longer-
duration events typically yield a lower overall intensity (which corresponds to a flatter
slope on the distribution curve).
Figure 4.26
Synthetic temporal
distributions for the
U.S. Midwest (Huff
and Angel, 1992)
Credit: Huff, 1967. Reproduced/modified by permission of American Geophysical Union.
Example 4.5 – Developing a Design Storm Hyetograph from a Bulletin 71 Dis-

tribution. Develop a design storm hyetograph for an event with a duration td = 3 hr and a total
depth of 50 mm using the Bulletin 71 temporal distributions. Assume that ∆t = 0.15 hr is a reasonable
choice for the drainage basin to which the design storm will be applied. The dimensionless first-
quartile rainfall distribution is given in Table 4.8.
Table 4.8 Bulletin 71 1st-quartile distribution
t/td Fraction
0.00 0.00
0.05 0.16
0.10 0.33
0.15 0.43
0.20 0.52
0.25 0.60
0.30 0.66
0.35 0.71
0.40 0.75
0.45 0.79
0.50 0.82
0.55 0.84
0.60 0.86
0.65 0.88
0.70 0.90
0.75 0.92
0.80 0.94
0.85 0.96
0.90 0.97
0.95 0.98
1.00 1.00
Solution: First, the t/td column from Table 4.8 is multiplied by the duration (3 hr), and the rainfall
fractions are multiplied by the total rainfall depth of 50 mm to obtain cumulative rainfall depths. The
results are shown in Table 4.9.
Table 4.9 Cumulative depth for Example 4.5.
t (hr) Cum. P (mm)

0.00 0.0
0.15 8.0
0.30 16.5
0.45 21.5
0.60 26.0
0.75 30.0
0.90 33.0
1.05 35.5
1.20 37.5
1.35 39.5
1.50 41.0
1.65 42.0
1.80 43.0
1.95 44.0
2.10 45.0
2.25 46.0
2.40 47.0
2.55 48.0
2.70 48.5
2.85 49.0
3.00 50.0
Next, the incremental precipitation depths are obtained by subtracting successive values for cumula-
tive precipitation. Table 4.10 and Figure E4.5.1 show the resulting hyetograph.
Table 4.10 Incremental depth for Example 4.5
t (hr) Incr. P (mm)

0.00–0.15 8.0
0.15–0.30 8.5
0.30–0.45 5.0
0.45–0.60 4.5
0.60–0.75 4.0
0.75–0.90 3.0
0.90–1.05 2.5
1.05–1.20 2.0
1.20–1.35 2.0
1.35–1.50 1.5
1.50–1.65 1.0
1.65–1.80 1.0
1.80–1.95 1.0
1.95–2.10 1.0
2.10–2.25 1.0
2.25–2.40 1.0
Section 4.6 Rainfall Requirements for Modeling Runoff 101
Table 4.10 Incremental depth for Example 4.5
t (hr) Incr. P (mm)

2.40–2.55 1.0
2.55–2.70 0.5
2.70–2.85 0.5
2.85–3.00 1.0
0.40
0.35
0.30
Incremental P, in.
0.25
0.20
0.15
0.10
0.05
0.00
2.10
0.00
0.15
0.30
0.45
0.60
0.75
0.90
2.25
2.40
2.55
2.70
2.85
3.00
1.50
1.05
1.20
1.35
1.65
1.80
1.95
Time, hr
Figure E4.5.1 Hyetograph for Example 4.5
4.6 RAINFALL REQUIREMENTS FOR MODELING

RUNOFF
This chapter has presented several ways in which rainfall information may be repre-
sented for stormwater runoff modeling and management. Which method to use in any
system design or analysis depends on the approach used for transformation of the
rainfall input into a prediction of the resulting stormwater discharge and on local reg-
ulations. Chapter 5 reviews a number of different runoff prediction methods and their
input requirements.
Examples of various runoff prediction methods and their corresponding rainfall
requirements are as follows:
• Rational Method – This peak discharge method is commonly used for
smaller projects such as inlet design and simple storm sewers. Application of
the rational method requires IDF curves, either provided by a review agency
or created from other sources.
• SCS Peak Discharge Method – This method produces only a peak dis-
charge (no hydrograph). Its use requires the 24-hour total rainfall depths for
the selected recurrence interval.
• SCS Unit Hydrograph – Hydrographs are required for modeling detention
ponds and complex watershed systems. Any rainfall distribution can be used
with the SCS unit hydrograph method; however, the 24-hour total rainfall
depths and the 24-hour rainfall temporal distribution (Type I, IA, II, or III)
applicable to the geographic location of the project is most typical.
• General Unit Hydrograph Methods – A number of methods may be
applied to synthesize a unit hydrograph for a given watershed. Because unit
hydrograph analyses are time-based, applications of these methods require
use of a rainfall hyetograph or a temporal distribution of some type. As a
general rule, no limitations exist on the methods applied to estimate the hye-
tograph or temporal distribution. A number of “standardized” methods exist
for design storm development, such as the SCS method, but hydrograph
analyses are by no means limited to use with such methods. As a good prac-
tice, it is recommended that engineers compute stormwater runoff hydro-
graphs using a number of plausible rainfall distributions, as this will
enlighten them as to the sensitivities of their results and the corresponding
uncertainties in their designs and analyses.
4.7 CHAPTER SUMMARY

Before designing or evaluating any stormwater drainage system, the engineer must
first determine the flows, and therefore the rainfall event or events, that the system
must be capable of handling.
In evaluating a rainfall event, the rainfall characteristics that must be considered are
the depth/volume, duration, area, and average recurrence interval of the precipitation,
as well as its temporal and spatial distributions. Various types of rainfall gauges are
used to collect rainfall data and develop precipitation records. Rainfall data records
are available from agencies such as the NWS in the United States, and much informa-
tion is available on the Internet. Often, rainfall data has been processed into a form
more readily accessible to the engineer and is presented in publications such as
HYDRO-35 or TP-40, or as IDF curves for the locale of interest. IDF curves can also
be used to approximate the recurrence interval for an actual event, and equations
found in HYDRO-35 can be used to develop IDF curves for areas where little data is
available.
Temporal distributions describe the variation in rainfall intensity over the course of a
storm event. Temporal distributions may be actual gauged rainfall amounts, or they
may be synthetic in nature. Examples of synthetic storm distributions are the NRCS
(SCS) 24-hour distributions for the United States. Temporal distributions are used to
develop rainfall hyetographs, which show incremental rainfall depths or intensities for
storm events.
Once the appropriate hyetograph for a design or study has been developed, this infor-
mation can be used to compute the runoff flow rates that will be used in drainage sys-
tem analysis and design.
REFERENCES
Chow, V. T. 1964. Handbook of Applied Hydrology. New York: McGraw-Hill.

Chow, V. T., D. R. Maidment, and L. W. Mays. 1988. Applied Hydrology. New York: McGraw-Hill.
Durans, S. R. and P. A. Brown. 2001. “Estimation and Internet-Based Dissemination of Extreme Rainfall
Information.” Transporation Research Record no. 1743: 41–48.
Frederick, R. H., V. A. Myers, and E. P. Auciello. 1977. Five- to 60-Minute Precipitation Frequency for the
Eastern and Central United States.NOAA Technical Memorandum NWS HYDRO-35. Silver Spring,
Maryland: National Weather Service, Office of Hydrology.
Haestad Methods. 2003. StormCAD User Manual. Waterbury, Connecticut: Haestad Methods.
Hershfield, D. M. 1961. Rainfall Frequency Atlas of the United States for Durations from 30 Minutes to 24
Hours and Return Periods from 1 to 100 Years. Technical Paper No. 40. Washington, D.C.: Weather
Bureau, U.S. Dept. of Commerce.
Huff, F. A. 1967. “Time Distribution of Rainfall in Heavy Storms.” Water Resources Research 3, no. 4:
1007–1019.
Huff, F. A. and J. R. Angel. 1992. “Rainfall Frequency Atlas of the Midwest.” Illinois State Water Survey
Bulletin 71: 141.
Johnson, L. E., P. Kucera, C. Lusk, and W. F. Roberts. 1998. “Usability Assessments for Hydrologic Fore-
casting DSS.” Journal of the American Water Resources Association 34, no. 1: 43–56.
Keifer, C. J., and H. H. Chu. 1957. “Synthetic Storm Pattern for Drainage Design.” Journal of the
Hydraulics Division, ASCE 84, no. HY4: 1–25.
Larson, L. W., and E. L. Peck. 1974. “Accuracy of Precipitation Measurements for Hydrologic Modeling.”
Water Resources Research 10, no. 4: 857–863.
Miller, J. F., R. H. Frederick, and R.J. Tracey. 1973. Precipitation-Frequency Atlas of the Coterminous
Western United States. NOAA Atlas 2, 11 vols. Silver Spring, Maryland: National Weather Service.
Singh, V. P. 1992. Elementary Hydrology. Englewood Cliffs, NJ: Prentice Hall.
Soil Conservation Service (SCS). 1969. Section 4: Hydrology. In National Engineering Handbook. Wash-
ington, D.C.: U.S. Dept. of Agriculture.
Soil Conservation Service (SCS). 1986. Urban Hydrology for Small Watersheds. Technical Release 55.
Washington, D.C.: U.S. Department of Agriculture.
Soil Conservation Service (SCS). 1992. TR-20 Computer Program for Project Formulation Hydrology
(revised users’ manual draft). Washington, D.C.: U.S. Department of Agriculture.
Sweeney, T. L. 1992. Modernized Areal Flash Flood Guidance. NOAA Technical Memorandum NWS
HYDRO 44. Silver Spring, Maryland: National Weather Service, Office of Hydrology.
U.S. Weather Bureau. 1958. Rainfall Intensity-Frequency Regime, Part 2 – Southeastern United States.
Technical Paper 29. Washington, D.C.: U.S. Department of Commerce.
PROBLEMS
4.1 The total volume of water stored in the atmosphere is estimated to be 12,900 km3. The average rate
of global evaporation and transpiration is 505,000 km3/yr. What is the average amount of time (in
days per year) that moisture spends in the atmosphere?
4.2 Using the methods given by HYDRO-35 and the data given in the table below, construct a set of IDF
curves for return intervals of 2, 5, 10, 25, 50, and 100 years and durations of 5, 10, 15, 30, and 60
minutes.
Rainfall Intensity (in/hr) for duration (min)

T (yr) 5 15 60
2 5.1 3.3 2.4
100 8.4 6.6 3.4
4.3 Using Figures 4.15 through 4.20, complete the following table for Boston, Massachusetts.
Rainfall Intensity (in/hr) for duration (min)

T (yr) 5 15 60
2
100
Problems 105
4.4 Given the following data from a storm, plot the rainfall hyetograph (intensity versus time), calculate
the total rainfall depth for the storm, and calculate the average rainfall intensity.
Interval (min) Intensity (in/hr)

0–15 0.25
15–30 0.30
30–45 0.40
45–60 0.25
60–75 0.10
75–90 0.05
4.5 Use the appropriate 24-hour NRCS (SCS) synthetic rainfall distribution to develop a design storm
hyetograph for a 10 year, 24-hour storm in Atlanta, Georgia. Use a total rainfall depth of 8.5 in.
(from TP 40).
Depression storage,
such as the shallow
ponding in this
grassed area,
reduces the amount
of runoff
discharged from a
watershed.
C H A P T E R
5
Modeling Runoff
A number of procedures for developing estimates of stormwater runoff rates and other
quantities of interest in hydrology are available to the engineer, but a set of universally
accepted, “cookbook” procedures does not exist. Some government agencies have
developed their own hydrologic methods and computer modeling tools, and some pro-
fessional societies and organizations have developed manuals of practice, but ulti-
mately the design professional needs to judge the suitability of any particular method
for use in a given practical application. The professional must apply his or her knowl-
edge and understanding of the problem at hand in piecing together a set of analytical
tools to solve that problem. Experience clearly plays a large part in selecting and
applying the proper rainfall and runoff methods—activities that can prove difficult for
practitioners who are new to hydrology. Although the lack of strict procedures can be
unsettling, the resourcefulness required to perform hydrologic engineering makes it
an exciting field of practice.
The hydrology coverage in this text began in Chapter 4, which provided information
on developing and obtaining precipitation data. Precipitation information in formats
such as IDF curves or rainfall hyetographs is required input for the surface water
hydrology models presented in this chapter. The primary focus of this chapter is the
runoff generation processes and flow routing methods needed when determining the
loads used in stormwater conveyance system design and analysis. A number of exam-
ples detailing the hydrologic calculations for many of these methods are presented to
help the reader understand the theoretical concepts and assumptions inherent in the
various models. However, engineers usually perform these calculations with the aid of
computer programs.
Section 5.1 defines the various types of rainfall abstractions and presents some of the
equations used in quantifying them. Section 5.2 presents factors affecting the amount
of effective precipitation (runoff) and methods for quantifying it, as well as effective
rainfall hyetograph development. Methods for estimating the response time of a
drainage basin to a storm event are described in Section 5.3.
108 Modeling Runoff Chapter 5
Section 5.4 extends the concepts presented in earlier sections to estimation of the peak
discharge from a watershed. Often, the peak runoff rate for a given storm is sufficient
to load and analyze the performance of a conveyance such as a storm sewer or culvert.
Section 5.5 presents the application of the unit hydrograph concept to compute the
complete runoff hydrograph resulting from a storm. Runoff hydrographs are neces-
sary for situations in which both runoff volume and flow rate must be considered, as
in the design of detention ponds.
Section 5.6 introduces the concept of base flow in a stream, and Section 5.7 provides
a basic explanation of the concepts related to modeling the contribution of snowmelt
to runoff, which is an important consideration in many regions. The Muskingum
method, which is a technique for computing the effect of storage in a channel on the
shape of the downstream hydrograph, is presented in Section 5.8.
5.1 RAINFALL ABSTRACTIONS

Recalling the discussion of the hydrologic cycle in Section 4.1, only a portion of the
total rainfall occurring over a drainage basin contributes to surface runoff and stream
flow. For example, a simple comparison of rainfall and runoff records for most loca-
tions in the United States shows that the equivalent depth of runoff (that is, stream
flow) is typically about 30 to 50 percent of the precipitation depth on an annual basis.
During intense storm events, the equivalent depth of runoff is often a much larger
fraction of the total precipitation depth.
Rainfall that does not contribute to direct surface runoff may be intercepted by vegeta-
tion, infiltrated into the ground surface, retained as depression storage in puddles and
small irregularities in the land surface, or returned to the atmosphere through transpi-
ration and evaporation. Collectively, these losses of rainfall are called abstractions.
The rainfall that remains after abstractions have occurred comprises the surface runoff
and is called effective precipitation or effective rainfall. This section describes the var-
ious types of abstractions that occur in natural and urbanized drainage basins. Later
sections present several commonly used methods for quantifying abstractions and
effective precipitation.
Four basic types of abstractions are generally acknowledged, but only three of these—
interception, depression storage, and infiltration—typically need to be addressed in
the design and analysis of stormwater conveyance systems. The fourth abstraction
type is evaporation and transpiration (usually referred to as evapotranspiration).
Although they are not commonly considered in modeling stormwater conveyances,
losses due to evapotranspiration are taken into account in safe yield calculations for
facilities such as water supply reservoirs. Abstraction types are illustrated in
Figure 5.1.
Interception
Interception refers to the capture of rainfall on the leaves and stems of vegetation
before it reaches the ground surface. Water intercepted by vegetation is returned to the
atmosphere by evaporation during dry-weather periods.
Section 5.1 Rainfall Abstractions 109
Figure 5.1
Rainfall abstractions
On an annual basis, interception can be quite significant and in some areas may
approach 20 to 30 percent of the total rainfall (Helvey and Patric, 1965; Helvey, 1967;
Zinke, 1967). During the relatively short and intense storm events of interest for rain-
fall and runoff studies, the percentage is often much smaller. Although some attempts
to develop predictive formulas for computing interception have been made (Horton,
1919; Merriam, 1960), it is frequently assumed that interception is no more than an
average equivalent rainfall depth of about 0.1 to 0.3 in. (2.5 to 7.6 mm) over the drain-
age basin.
Interception tends to be greater for coniferous trees than for deciduous trees (Patric,
1966). At their maximum growth, grasses may intercept as much rainfall as trees dur-
ing individual storms (Merriam, 1961). In temperate regions, interception has a dis-
tinctly seasonal aspect because of the loss of leaves from vegetation during the cold
season.
Depression Storage
Excess water begins to pond on the land surface when the rainfall intensity exceeds
the infiltration capacity of the soil during a storm event. The ponded water fills small
depressions and irregularities in the ground surface, and additional water is held on
the surface through the phenomenon of surface tension. The water held in depressions
and on the surface because of surface tension is called depression storage, and it
either evaporates during dry-weather periods or infiltrates into the soil. Noninfiltrated
rainfall that remains after surface depressions and irregularities have been filled con-
tributes to surface runoff.
The depression storage capacity of a drainage basin is usually expressed in terms of

an equivalent average depth of rainfall over the basin. Values for depression storage
capacity range from about 0.01 in. (0.25 mm) for paved surfaces (Viessman, Knapp,
and Lewis, 1977) to about 0.3 in. (7.6 mm) for forest litter (ASCE, 1992). These val-
ues are for moderately sloped surfaces. Values tend to be higher for flatter surfaces
and lower for steeper surfaces.
Infiltration
When rainfall occurs on a pervious surface, some of the rainwater infiltrates into the
ground in response to gravitational and capillary forces. The infiltrated water may
contribute to groundwater recharge, or it may be taken up by the roots of vegetation
and subsequently transpired through stomata (openings) in leaves. Infiltrated water
may also be evaporated from the soil during dry-weather periods between storm
events, or it may move laterally through the near-surface soils and reappear as surface
water in a stream.
Infiltration capacity depends to a great extent on soil type. If the rate at which water
can infiltrate a soil is greater than the rate at which rainfall is supplied to the soil sur-
face, all rainfall is lost to infiltration. However, if the rainfall rate is greater than the
infiltration capacity of the soil, surface ponding and/or surface runoff occurs. Sandy
and gravelly soils generally have higher infiltration capacities than do silts and clays.
For all soils, the rate at which infiltration can occur decreases with time and
approaches a constant rate as the soil becomes wetter.
Two widely used methods for modeling infiltration, the Horton and Green-Ampt
methods, are presented in the following subsections.
Horton Equation. A widely used method of representing the infiltration capacity
of a soil is the Horton equation (Horton, 1939). The Horton method was empirically
developed to describe field observations reflecting an exponential decay of infiltration
rate over time as the soil becomes more saturated. For conditions in which the rainfall
intensity is always greater than the infiltration capacity (that is, when rainwater sup-
ply for infiltration is not limiting), this method expresses the infiltration rate f(t) as a
function of time, as follows:
f (t ) = f c + ( f 0 − f c )e − k (t − t0 ) (5.1)
where f(t) = the infiltration rate (in/hr, mm/hr) at time t (min or hr)
fc = a steady-state infiltration rate (in/hr, mm/hr) that occurs for sufficiently
large t
f0 = the initial infiltration rate at the time that infiltration begins
(in/hr, mm/hr)
k = a decay coefficient (min-1 or hr-1)
t0 = time at which infiltration begins (min or hr)
It can be shown theoretically that the steady-state infiltration rate fc is equal to the sat-
urated vertical hydraulic conductivity of the soil.
Note that the exponential term in Equation 5.1 is dimensionless. The units associated
with t, t0, and k may be taken in either minutes or hours, provided they are consistent.
The units on the remaining terms in the equation are completely independent of the
units used in the exponent term.
Estimation of the parameters fc, f0, and k in Equation 5.1 can be difficult because of
the natural variabilities in antecedent moisture conditions (that is, the amount of
moisture present in the soil prior to the rainfall event of interest) and soil properties.
Region-specific data may be available, such as the values recommended by Rawls,
Yates, and Asmusse (1976) in Table 5.1, but such tabulations should be used with cau-
tion. Singh (1992) recommends that f0 be taken as roughly 5 times the value of fc.
Table 5.1 Typical values of Horton infiltration parameters

(Rawls, Yates, and Asmusse, 1976)
f0 fc
Soil Type (in/hr) (mm/hr) (in/hr) (mm/hr) k (min-1)
Alphalpha loamy sand 19.0 483 1.4 36 0.64
Carnegie sandy loam 14.8 376 1.8 46 0.33
Dothan loamy sand 3.5 89 2.6 66 0.02
Fuquay pebbly loamy sand 6.2 157 2.4 61 0.08
Leefield loamy sand 11.3 287 1.7 43 0.13
Tooup sand 23.0 584 1.8 46 0.55
Often, the rainfall intensity during the early part of a storm is lower than the potential
infiltration capacity (rate) of the soil; thus, the supply of rainwater is a limiting factor
on the infiltration rate. During the time period when the water supply is limiting, the
actual infiltration rate is equal to the rate at which rainwater is supplied to the ground
surface. This effect is illustrated in Figure 5.2. Later in the storm, when the rainfall
rate is greater than the infiltration rate, the actual infiltration rate will be greater than
that predicted by Equation 5.1 because infiltration was limited early in the storm.
An integrated version of the Horton method can account for the underestimation of
the infiltration rate due to limiting rainfall intensity early in a storm (Viessman,
Knapp, and Lewis, 1977; Bedient and Huber, 1992; Chin, 2000), as can more compli-
cated infiltration models such as the Green-Ampt (Green and Ampt, 1911) model pre-
sented in the next subsection. Nevertheless, the simple Horton model represented by
Equation 5.1 is often used in practice because it yields a larger amount of effective
precipitation than does the integrated version of the Horton model and is thus conser-
vative for stormwater conveyance design. Depending on selected parameter values,
Equation 5.1 may or may not yield more effective rainfall than do other models (for
example, the Green-Ampt model).
Figure 5.2
Rainfall intensity can
limit infiltration rate
during the early part
of the storm
Rainfall Intensity
Rainfall Intensity, Infiltration Rate

Actual Infiltration Rate
Shaded area = actual depth of

infiltrated water
f(t) as predicted
by Equation 5.1
Time
Example 5.1 – Using the Horton Equation to Determine Infiltration Rate. Using
Horton’s equation, find the infiltration rate at time = 2.0 hours for a Dothan loamy sand. The rainfall
duration is 4 hours, and for this event, it takes 0.05 hours to reach the interception capacity, at which
time infiltration begins.
From Table 5.1,
f0 = 3.5 in/hr; fc = 2.6 in/hr; and k = 0.02 min-1 = 1.2 hr-1

Applying Equation 5.1,
f(2 hr) = 2.6 in/hr + (3.5 in/hr – 2.6 in/hr)e-1.2(2 0.05)

= 2.69 in/hr
The computed infiltration rate is between the values of f0 and fc, as it must be.
Green-Ampt Equation. In contrast to the empirically developed Horton equa-

tion, the Green-Ampt method is based on a theoretical application of Darcy’s law
(which relates flow velocity to the permeability of the soil) and conservation of mass.
The resulting equation inversely relates the infiltration rate f to the total accumulated
infiltration F as (Chow, Maidment, and Mays, 1988)
 ψ (θ s − θ i ) 
f = Ks  + 1 (5.2)
 F 
 
where f =
infiltration rate (in/hr, cm/hr)
Ks =
saturated hydraulic conductivity (permeability) (in/hr, cm/hr)
ψ =
capillary suction (in., cm)
θs =
volumetric moisture content (water volume per unit soil volume) under
saturated conditions
θi = volumetric moisture content under initial conditions
F = total accumulated infiltration (in., cm)
The benefit of the Green-Ampt method is that the infiltration rate can be calculated
based on physical, measurable soil parameters, as opposed to the more empirical
coefficients of Horton. For more information on these parameters, see Chow, Maid-
ment, and Mays (1988).
To calculate the infiltration rate at a given time, the total infiltration up to that time
must be calculated. This value can be determined by integrating Equation 5.2 with
respect to time (starting at t = 0) and solving for F.
 F 
F = K s t + ψ (θ s − θi ) ⋅ ln  1 +
 ψ (θ − θ ) 
(5.3)
 s i 
where t = time (hr)

Equation 5.3 cannot be explicitly solved and thus requires the application of a numer-
ical method such as the Newton-Raphson or bisection method to solve for F. Also, the
equation assumes that the rainfall intensity is always greater than the infiltration rate
at a given time step. If the intensity is less than the associated infiltration rate, then the
infiltration is equal to the rainfall amount for that time step.
Table 5.2 provides average values for ψ, Ks, and θs for the eleven U.S. Department of
Agriculture (USDA) soil textures, which can serve as initial estimates for these
parameters.
Table 5.2 Green-Ampt parameters (Rawls et al., 1993)

Volumetric Capillary Suction, ψ Saturated hydraulic conductivitya Ks
Moisture Content,
Soil Texture Class θs (in.) (cm) (in/hr) (cm/hr)
0.437 1.95 4.95
Sand 9.28 23.56
(0.374−0.500) (0.38−9.98) (0.97−25.36)
0.437 2.41 6.13
Loamy sand 2.35 5.98
(0.363−0.506) (0.53−11.00) (1.35−27.94)
0.453 4.33 11.01
Sandy loam 0.86 2.18
(0.351−0.555) (1.05−17.90) (2.67−45.47)
0.463 3.50 8.89
Loam 0.52 1.32
(0.375−0.551) (0.52−23.38) (1.33−59.38)
0.501 6.57 16.68
Silt loam 0.27 0.68
(0.420−0.582) (1.15−37.56) (2.92−95.39)
0.398 8.60 21.85
Sandy clay loam 0.12 0.30
(0.332−0.464) (1.74−42.52) (4.42−108.0)
0.464 8.22 20.88
Clay loam 0.08 0.20
(0.409−0.519) (1.89−35.87) (4.79−91.10)
0.471 10.75 27.30
Silty clay loam 0.08 0.20
(0.418−0.524) (2.23−51.77) (5.67−131.50)
0.430 9.41 23.90
Sandy clay 0.05 0.12
(0.370−0.490) (1.61−55.20) (4.08−140.2)
Table 5.2 (cont.) Green-Ampt parameters (Rawls et al., 1993)

Volumetric Capillary Suction, ψ Saturated hydraulic conductivitya Ks
Moisture Content,
Soil Texture Class θs (in.) (cm) (in/hr) (cm/hr)
0.479 11.50 29.22
Silty clay 0.04 0.10
(0.425−0.533) (2.41−54.88) (6.13−139.4)
0.475 12.45 31.63
Clay 0.02 0.06
(0.427−0.523) (2.52−61.61) (6.39−156.5)
a. Ks can be modified to obtain the Green-Ampt K. For bare ground conditions, K can be taken as Ks/2.
The following example illustrates common usage of the Green-Ampt method in cal-
culating the infiltration volume (expressed as depth) over the course of an unsteady
storm. For another example of the application of the Green-Ampt method, see Chow,
Maidment, and Mays (1988). Computer programs for performing these calculations
are available as well.
Example 5.2 – Computing Infiltration Volumes with Green-Ampt. Use the Green-
Ampt method to solve for the amount of rainfall infiltrated for each time step of the rainfall hyeto-
graph in Table 5.3.
Table 5.3 Rainfall Data for Example 5.2
Time (hr) Intensity (in./hr)
0.0 0.3
0.1 1.9
0.2 3.2
Assume the following soil parameters:

ψ = 3.5 in.
Ks = 0.13 in./hr
θs − θi = 0.434
Solution: The key to finding the infiltration volume is keeping track of the position on two timelines.
The first timeline is the rainfall hyetograph. The second timeline is the curve relating total infiltration
to time as generated by Equation 5.2. At t = 0, both of these timelines correspond.
If the rainfall intensity is less than or equal to the infiltration rate for a particular time step, then all
rainfall during that time step will infiltrate into the ground. If the intensity is greater than the infiltra-
tion rate during part or all of the time step, then ponding or runoff occurs during that time step. The
total accumulated infiltration volume can be calculated by using Equation 5.3. As the total infiltration
increases, the time on the second timeline increases, as well. If the intensity is greater than the infiltra-
tion rate, then the storm timeline and accumulated infiltration timeline proceed at the same pace. If
intensity is less than the infiltration rate, then the timelines proceed at different paces over the course
of the time step.
The steps taken in solving this example follow.
Time Step 1
a) Assume that all the rainfall infiltrates during the first time step:
0.3 in./hr × 0.1 hr = 0.03 in.
b) Calculate the infiltration rate at the end of the time step as if all the rainfall infiltrates using Equa-
tion 5.2.
 3.5 ( 0.434 ) 
f = ( 0.13)  + 1 = 6.71 in./hr
 0.03 
c) Because 6.71 in./hr > 0.3 in./hr up to an accumulated infiltration of 0.03 in., all rainfall during the
time step is infiltrated.
Time Step 2
a) Again, assume all the rainfall for the time step infiltrates:
1.9 in/hr ×0.1 hr = 0.19 in.
The total accumulated infiltration is then computed as
0.19 in. + 0.03 in. = 0.22 in.
b) The infiltration rate that would occur at the end of time step 2 assuming 0.22 in. of accumulated
infiltration is computed to be
 3.5 ( 0.434 ) 
f = (0.13)  + 1 = 1.03 in/hr
 0.22 
c) Because this infiltration rate is less than the intensity for time step 2 of 1.9 in/hr, ponding or runoff
must begin sometime between 0.1 hr and 0.2 hr. The accumulated infiltration for which the rainfall
intensity matches the infiltration rate must be computed. Substituting f = 1.9 in./hr and solving
Equation 5.2 for F:
 3.5 ( 0.434 ) 
1.9 = 0.13  + 1
 F 
F = 0.112 in.
d) The time between 0.1 hr and 0.2 hr when the intensity equals the infiltration rate must be deter-
mined. Intensity equals infiltration when the total accumulated infiltration is 0.112 in. Keeping in
mind that 0.03 in. infiltrated during the first time step, runoff begins at
(0.112 in. – 0.03 in.) /1.9 in./hr = 0.043 hr after the beginning of the second time step
e) Next, the position on the second teemingly when 0.112 in. of accumulated infiltration would occur
must be determined. Equation 5.3 is rearranged and solved for t.
1   0.112  
t= 0.112 − (3.5)(0.434) ⋅ ln 1 +   = 0.03 hr
0.13 
  3.5 ( 0.434 )  
f) To find the amount of infiltration during the time step, the time on the second timeline correspond-
ing to the end of the second time step of the storm must be determined. Runoff begins after
0.1 hr + 0.043 hr = 0.143 hr
when the total accumulated infiltration equals 0.112 in. This point corresponds to 0.03 hr on the
second timeline. There is
0.2 hr – 0.143 hr = 0.057 hr
remaining in the second time step for the storm timeline (first timeline). Thus, the end of time step
2 corresponds to a time of
0.03 hr + 0.057 hr = 0.087 hr

on the second timeline.
g) Now, solve for the total accumulated infiltration at 0.087 hr using Equation 5.3.
 F 
F = ( 0.13)( 0.087 ) + ( 3.5 )( 0.434 ) ⋅ ln  + 1
 3.5 ( 0.434 ) 
As mentioned above, F cannot be solved for explicitly and must be determined using a numerical
method. Using a root finder, F is computed to be 0.193 in. at the end of time step 2.
h) The infiltration that occurred during time step 2 is the difference between the total accumulated
infiltration at the end of step 1 and the total accumulated infiltration at the end of step 2:
0.193 in. – 0.03 in. = 0.163 in.
Time Step 3
a) Unlike the previous two time steps, the infiltration rate at the beginning of time step 3 is calculated
for F = 0.193 in.
 3.5 ( 0.434 ) 
f = (0.13)  + 1 = 1.153 in./hr
 0.193 
Because 3.2 in/hr > 1.153 in/hr, runoff must occur over the entire time step.
b) The next step is to determine the point on the second timeline that equates to the end of time step 3.
Because runoff occurs over the entire time step (unlike step 2), 0.1 hr can simply be added to the
point on the second timeline at the end of time step 2:
0.1 hr + 0.087 hr = 0.187 hr
Therefore, 0.187 hr on the second timeline corresponds to 0.3 hr on the storm hyetograph.
c) The total accumulated infiltration at 0.187 hr is computed as using Equation 5.3.
 F 
F = ( 0.13)( 0.187 ) + ( 3.5 )( 0.434 ) ⋅ ln  + 1 = 0.288 in.
 3.5 ( 0.434 ) 
d) The infiltration during the time step equals 0.288 in. – 0.193 in. = 0.095 in.
5.2 DETERMINATION OF EFFECTIVE PRECIPITATION

(RUNOFF)
During and shortly after a rainfall event, the total discharge flowing in a stream chan-
nel generally consists of direct runoff from contributing land surfaces, interflow or
subsurface storm flow resulting from lateral movement of water through shallow soil
layers, and base flow or groundwater outflow. During dry-weather periods between
storm events, a stream may not have a discharge at all. If a discharge does exist during
those periods, that discharge consists solely of base flow.
As noted earlier, effective precipitation is that portion of the total precipitation during
a storm event that is not lost to abstractions—it is the precipitation that becomes
direct runoff. The volume of direct runoff caused by a storm event is equal to the prod-
Section 5.2 Determination of Effective Precipitation (Runoff) 117
uct of the effective precipitation depth and the land surface area on which the precipi-
tation occurred (that is, the drainage basin area).
Design of storm sewers and other elements of stormwater conveyance systems usually
involves the estimation of direct runoff only, as stormwater conveyance elements
often do not have significant base flow discharges. Section 5.6 discusses base flow
estimation for cases in which it cannot be neglected.
The depth of effective precipitation may be determined in a number of ways; however,
in its most general form, the relationship of direct runoff (effective precipitation) to
total precipitation and abstractions is
Dr = Dp – Dli – Di – Ds – De (if Dr > 0) (5.4)
Otherwise,
Dr = 0 (5.5)
where Dr = total depth of direct runoff (effective precipitation)

Dp = total depth of precipitation (rainfall)
Dli = total initial loss, sometimes called initial abstractions
Di = total depth infiltrated after initial losses
Ds = total depression storage depth
De = transpiration and evaporation losses (often ignored for short duration
stormwater events)
These depths are presented graphically in Figure 5.3.
Figure 5.3
Precipitation,
abstraction, and
runoff volumes
As pointed out previously, the volume of direct runoff is expressed by
Vr = D r × A (5.6)
where Vr = direct runoff volume

A = contributing drainage area
The area term, A, in Equation 5.6 is the contributing drainage area, or watershed area,
for which the direct runoff volume is being evaluated. A watershed is a land area that
drains to a single point of discharge. Typically, contour maps are used to delineate
watershed boundaries and determine the area. Because water flows downhill, delin-
eating a watershed is simply a matter of identifying an outfall point of interest and
locating the watershed boundary such that any rain that falls within the boundary will
be directed toward that point of discharge. The delineated area is then measured using
a planimeter or by graphical or computer-aided methods.
Because river and stream systems collect water, a watershed may have any number of
subwatersheds within it. The focus of the analysis and the determination of whether
subwatersheds must be analyzed separately depend on the scope and purpose of the
project at hand. Figure 5.4 shows the collection channels for two typical natural
watersheds.
Figure 5.4
Typical natural
watersheds with
collection channels
Example 5.3 – Computing Direct Runoff Volume. A 2.5-hr storm deposits 50 mm of

rainfall over a drainage basin with an area of 1.5 ha. If the interception capacity is 8 mm, the depres-
sion storage depth is 5 mm, the calculated total infiltration is 18 mm, and evaporation is negligible,
find the total depth of effective precipitation and the total direct runoff volume for the drainage area.
Solution: Applying Equation 5.4,
Dr = 50 – 8 – 18 – 5 – 0 = 19 mm
Applying Equation 5.6,
Vr = (19 mm/1000 mm/m)(1.5 ha) = 0.0285 ha-m = 285 m3
The concept of computing the direct runoff volume by subtracting abstractions from a
precipitation depth is straightforward. However, the determination of the quantities
for the rainfall abstractions, and thus the amount of rainfall excess (that is, direct run-
off), is more complex. The following sections describe several possible methods for
accounting for rainfall abstractions. The method that the engineer chooses can depend
on a variety of factors, including the size of the site, land use and characteristics, data
availability and needs, and acceptable local practice.
Runoff Coefficient
A simple approach to estimating runoff is to apply a coefficient that represents the
ratio of rainfall that produces runoff. The use of the runoff coefficient C assumes that
the effective rainfall intensity (ie) is a fraction of the gross rainfall intensity (i), or
ie = C × i (5.7)
where C = runoff coefficient (0 ≤ C ≤ 1)

i = rainfall intensity (in./hr, mm/hr)
This expression also can be written in terms of rainfall depths:
Pe = C × P (5.8)
where Pe = effective precipitation depth (direct runoff depth) (in., mm)

P = total precipitation depth (in., mm)
A good way to visualize the runoff coefficient is to think of it as a percentage of rain-
fall. For example, a C of 0.85 would yield a direct runoff (effective precipitation)
depth that is 85 percent of the gross rainfall depth. The runoff coefficient is used with
the rational method for peak runoff rate estimation, which is described in Section 5.4
(page 140).
Example 5.4 – Using the Runoff Coefficient to Compute Effective Rainfall

Intensity. For a rainfall event with an average intensity of 50 mm/hr falling on a drainage area hav-
ing a runoff coefficient = 0.70, determine the effective rainfall intensity.
ie = Ci = 0.70 × 50 mm/hr = 35 mm/hr
The runoff coefficient is also used with the rational method to compute peak runoff
flow rate. A table containing ranges of values of C for various land uses, Table 5.11, is
included with the rational method discussion in Section 5.4.
NRCS (SCS) Curve Number Method

In the 1950s, the U.S. Department of Agriculture Soil Conservation Service (now the
NRCS) developed a procedure to partition the total depth of rainfall represented by a
design storm hyetograph into initial abstractions Ia, retention F, and effective rainfall
(runoff) Pe (SCS, 1969). These components are illustrated in Figure 5.5.
Initial abstractions consist of all rainfall losses occurring before the beginning of sur-
face runoff, including interception, infiltration, and depression storage. Retention
refers to the continuing rainfall losses following the initiation of surface runoff, which
are predominantly due to continuing infiltration. Conservation of mass requires that
F = P – Ia – Pe (5.9)
where F = equivalent depth of retention (in., mm)

P = total rainfall depth in storm (in., mm)
Ia = equivalent depth of initial abstractions (in., mm)
Pe = depth of effective precipitation (runoff) (in., mm)
Figure 5.5
Precipitation
components as
defined by the NRCS
(SCS) Curve Number
method
By rearranging the terms, Equation 5.9 yields
Pe = (P – Ia) – F (5.10)
This equation is very similar to the generalized infiltration and runoff Equation 5.4
(for that equation, Ia = Dli, F = Di, Ds = 0, and De = 0).
The curve number (CN) referred to in the NRCS curve number method is a parameter
used to estimate the maximum possible retention (S) of the soil in the area of interest.
Its value depends on factors such as soil type, land use, vegetative cover, and moisture
content prior to the onset of the storm event. S does not include initial abstractions
(Ia). Using the equations that follow and the curve number (typical values for which
are presented later in this section), the effective precipitation (runoff) resulting from a
storm event can be computed.
An assumption made in the development of the curve number method is
F Pe
= (5.11)
S P − Ia
where S = maximum possible retention (in., mm)

In essence, the assumption represented by Equation 5.11 is that the ratio of actual
retention to maximum possible retention of water during a storm is equal to the ratio
of effective rainfall to maximum possible effective rainfall (total rainfall less initial
abstractions). Substitution of Equation 5.9 into Equation 5.11 yields
( P − I a )2
Pe = (5.12)
(P − Ia ) + S
which is valid for values of P > Ia. Data analyzed by the NRCS indicated that Ia is
related to S, and on average supported the use of the relationship Ia = 0.2S. Thus,
Equation 5.12 becomes
( P − 0.2 S ) 2
Pe = (5.13)
P + 0.8S
when P > 0.2S (Pe = 0 when P ≤ 0.2S). Because the initial abstraction Ia consists of
interception, depression storage, and infiltration prior to the onset of direct runoff, it
may be appropriate in some applications to assume that Ia = 0.1S or Ia = 0.3S instead
of Ia = 0.2S. For example, the relationship Ia = 0.1S might be appropriate in a heavily
urbanized area where there is little opportunity for initial abstractions to occur. Equa-
tion 5.13 must be modified when the relationship between Ia and S is assumed to be
different from Ia = 0.2S.
The use of Equation 5.12 or 5.13 in estimating the depth of effective rainfall during a
storm requires an estimate of the maximum possible retention S. NRCS conducted
research to approximate S for various soil and cover conditions. To provide engineers
with tables having a manageable range of coefficients from 1 to 100, the original val-
ues for S were modified using the following simple relationship:
1000
CN = (5.14)
S + 10
where CN = runoff curve number

S = maximum possible retention (in.)
Practical values of the curve number CN range from about 30 to 98, with large values
being associated with impervious land surfaces. The NRCS has tabulated curve num-
bers as a function of soil type, land use, hydrologic condition of the drainage basin,
and antecedent moisture condition.
Rearranging Equation 5.14, S is related to the runoff curve number, CN, as
1000
S= − 10 (5.15)
CN
Figure 5.6 provides a graphical solution to Equation 5.13 for various rainfall depths
and curve numbers.
Figure 5.6
Graphical solution for
effective precipitation
(runoff) using the
curve number method
and assuming Ia =
0.2S
Soil Groups. Soils in the United States have been classified by the NRCS into
four hydrologic groups: A, B, C, and D. Group A soils have high infiltration rates (low
runoff potential), even when they are thoroughly wetted. Typical Group A soils are
well-drained sands and gravels. Group D soils are at the opposite end of the spectrum,
having low infiltration rates (high runoff potential). Typical Group D soils are clays,
shallow soils over nearly impervious material, and soils with a high water table.
Group B and Group C soils are in the midrange of the spectrum.
Information on the hydrologic soil group or groups present within a drainage basin in
the United States may be found by contacting the nearest NRCS office and obtaining
the soil survey of the county in which the project is located. If the hydrologic soil
group is not provided in the soil survey, it can be found based on the soil name using
Appendix A of TR-55 (SCS, 1986). When a drainage basin undergoes urbanization,
the hydrologic soil group may change due to compaction of the soil by heavy con-
struction equipment or mixing of soils as a consequence of grading operations.
Cover Type/Land Use. The surface conditions of a drainage area have a signif-
icant impact on direct runoff. For example, in the case of a sandy (Group A) soil com-
pletely paved with asphalt, the soil itself will have no impact on the amount of runoff.
Even for pervious conditions, cover type plays a significant role in the amount of
direct runoff from a site. For example, a heavily forested area will yield runoff vol-
umes that differ from those of a lawn or plowed field.
Hydrologic Condition. The hydrologic condition of rangeland, meadow, or
pasture is defined to be good if it is lightly grazed and has vegetative cover on more
than 75 percent of the area. Conversely, a poor hydrologic condition corresponds to a
heavily grazed area with vegetation covering less than 50 percent of the surface.
Sources of Soils Data for

Stormwater Modeling
Soils information plays an important role in stormwater provided (including links to PDF-format U.S. county
modeling and design. The ability of water to flow soil survey manuscripts when available), as well as
through the soil, typically indicated by the saturated online access to the State Soil Geographic Database
hydraulic conductivity, Ks, has an important impact on (STATSGO). STATSGO contains generalized soil infor-
how precipitation becomes runoff. Ks is used with mation on entire states at a level of detail appropriate
methods such as Green-Ampt to compute infiltration. for broad planning and management uses covering
Dry soils with high Ks values may produce little or no state, regional, and multi-state areas (map scale is
runoff even from heavy storms. Conversely, wet soils 1:250,000).
that are highly compacted will result in almost all pre-
The most detailed electronic information available from
cipitation becoming runoff.
the NRCS is the Soil Survey Geographic Data Base
Another important soil property descriptor used in (SSURGO) soils data. SSURGO data are the digital
stormwater modeling is the hydrologic soil group equivalent of information contained in the printed
(HSG), which has four soil classifications: A, B, C, and county soil surveys developed by the NRCS. The
D. The NRCS Curve Number Method uses the HSG NRCS is currently converting the tabular and mapping
and land use to determine the runoff curve number. The information from the soil surveys into the digital
table below shows the infiltration rate ranges and tex- SSURGO data. The SSURGO web site
tures associated with each hydrologic soil group (HSG). (www.ftw.nrcs.usda.gov/ssur_data.html) lists counties
that have had their soils data converted to a GIS format
and SSURGO database format.
Infiltration
Type Texture Information for soils at this detail level (scales ranging
Rate
from 1:12,000 to 1:63,360) is organized into Soil Map
A 0.30–0.45 in/hr Sands and gravels Units denoted by symbols like 10A or OcB. The first
B 0.15–0.30 in/hr Course to moder- part of the symbol (10 or Oc) represents the soils map
ately fine unit name, usually based on the abbreviated name of
one or several official soil series that make up that map
C 0.05–0.15 in/hr Moderately fine to
unit. The second part of the symbol (A or B) represents
fine
a slope class category for that county. One important
D < 0.05 in/hr Clays with high thing to remember when working with soil maps is that
swelling, high each county’s soil survey and resulting SSURGO data
water tables is unique to that county. Slope classes often differ
between counties, even neighboring counties, because
of the range of slopes in that county. Also, because soils
Values for Ks, HSG, and many other soil properties for surveys have been performed over a number of years
locations in the United States can be obtained from the and soil series standards have changed over time, simi-
NRCS. NRCS’s SOILS web site (soils.usda.gov) con- larly named map units from neighboring counties can
tains information and links to soils data at varying lev- have different characteristics.
els of detail. Information on obtaining local soils data is
Antecedent Moisture Condition. When rainfall events occur in quick suc-

cession, the time period between storms may be too short for the soils to dry to their
average or normal moisture conditions. When rainfall occurs on soils that are already
wet, the net result is that direct runoff volumes and peaks will be higher than normal.
The NRCS (SCS) method accounts for this possibility by allowing the curve number
to depend on an antecedent moisture condition (AMC). Three AMC classifications
exist. Normal conditions correspond to AMC-II. AMC-I corresponds to a drier condi-

tion, and AMC-III to a wetter condition. The National Engineering Handbook (NEH-
4) (SCS, 1969) provided guidance on AMC selection in its Table 4.2; however, AMC
is dependent on the size and location of the watershed, and this table was eliminated
in the 1993 edition of Chapter 4 (NRCS, 1993). Often, typical local practice will
influence AMC selection.
Table 5.4 is from Chapter 10 of NEH-4 (Mockus, 1972) and relates CN values for
AMC-II to corresponding CN values for AMC-I and AMC-III. Column 5 represents
the initial abstraction (Ia) for AMC-II conditions. The phrase “Curve starts where P
=” indicates the cumulative depth of precipitation that must occur for direct runoff to
begin (this occurs when P = Ia).
Alternatively, the curve numbers corresponding to AMC-I and AMC-III conditions
can be computed from AMC-II curve numbers using Equations 5.16 and 5.17 (Chow,
Maidment, and Mays, 1988). The computed CN may be rounded to the nearest whole
number.
4.2CN II
CN I = (5.16)
10 − 0.058CN II
23CN II
CN III = (5.17)
10 + 0.13CN II
where CNI, CNII, and CNIII = curve numbers for AMC-I, -II, and -III, respectively
Curve Number Tables. Tables 5.5 through 5.8 provide listings of curve num-
bers that account for both cover conditions and soil type for normal antecedent mois-
ture conditions (AMC-II) and Ia = 0.2S (Mockus, 1969). The curve numbers shown
for the urban and suburban land use conditions in Table 5.5 are based on the percent-
ages of directly connected impervious areas in the drainage basin as shown in the
table, and should be used with caution when the actual percentage of imperviousness
in a drainage basin differs from this assumed value. When necessary, a composite
curve number can be developed as an area-weighted average of individual curve num-
bers. The composite CN may be rounded to the nearest whole number.
Additional information on hydrologic condition and curve numbers for land uses
other than those contained in Tables 5.5 through 5.8 may be found in Chapter 9 of
NEH-4 (Mockus, 1969) and in TR-55 (SCS, 1986).
Table 5.4 Curve numbers (CN) and constants for the case Ia = 0.2S (from Mockus, 1972)
(1) (2) (3) (4) (5) (1) (2) (3) (4) (5)
CN for CN for Conditions Curvea CN for CN for Conditions Curvea

Condition S Valuesa Starts Where Condition S Valuesa Starts Where
II I III (in.) P = (in.) II I III (in.) P = (in.)
100 100 100 0 0.00 61 41 78 6.39 1.28
99 97 100 0.101 0.02 60 40 78 6.67 1.33
98 94 99 0.204 0.04 59 39 77 6.95 1.39
97 91 99 0.309 0.06 58 38 76 7.24 1.45
96 89 99 0.417 0.08 57 37 75 7.54 1.51
95 87 98 0.526 0.11 56 36 75 7.86 1.57
94 85 98 0.638 0.13 55 35 74 8.18 1.64
93 83 98 0.753 0.15 54 34 73 8.52 1.70
92 81 97 0.870 0.17 53 33 72 8.87 1.77

91 80 97 0.989 0.20 52 32 71 9.23 1.85
90 78 96 1.11 0.22 51 31 70 9.61 1.92
89 76 96 1.24 0.25 50 31 70 10.0 2.00
88 75 95 1.36 0.27 49 30 69 10.4 2.08
87 73 95 1.49 0.30 48 29 68 10.8 2.17
86 72 94 1.63 0.33 47 28 67 11.3 2.26

85 70 94 1.76 0.35 46 27 66 11.7 2.35
84 68 93 1.90 0.38 45 26 65 12.2 2.44
83 67 93 2.05 0.41 44 25 64 12.7 2.55
82 66 92 2.20 0.44 43 25 63 13.3 2.65
81 64 92 2.35 0.47 42 24 62 13.8 2.76
80 63 91 2.50 0.50 41 23 61 14.4 2.88

79 62 91 2.66 0.53 40 22 60 15.0 3.00
78 60 90 2.82 0.56 39 21 59 15.6 3.13
77 59 89 2.99 0.60 38 21 58 16.3 3.26
76 58 89 3.16 0.63 37 20 57 17.0 3.41
75 57 88 3.33 0.67 36 19 56 17.8 3.56
74 55 88 3.51 0.70 35 18 55 18.6 3.71
73 54 87 3.70 0.74 34 18 54 19.4 3.88
a. For CN in column 1
Table 5.5 Runoff curve numbers for urban areas (Mockus, 1969)a
Cover Description Average Curve Numbers for Hydrologic Soil Group
Percent
Impervious
Cover Type and Hydrologic Condition Areab A B C D
Fully developed urban areas (vegetation established)
Open space (lawns, parks, golf courses, cemeteries, etc.)c:
Poor condition (grass cover < 50%) 68 79 86 89
Fair condition (grass cover 50% to 75%) 49 69 79 84
Good condition (grass cover > 75%) 39 61 74 80
Impervious areas:
Paved parking lots, roofs, driveways, etc. (excluding right-of-way) 98 98 98 98
Streets and roads:
Paved; curbs and storm sewers (excluding right-of-way) 98 98 98 98
Paved; open ditches (including right-of-way) 83 89 92 93
Gravel (including right-of-way) 76 85 89 91
Dirt (including right-of-way) 72 82 87 89
Western desert urban areas:
Natural desert landscaping (pervious area only)d 63 77 85 88
Artificial desert landscaping (impervious weed barrier, desert shrub 96 96 96 96
with 1 to 2 in. sand or gravel mulch and basin borders)
Urban districts:
Commercial and business 85 89 92 94 95
Industrial 72 81 88 91 93
Residential districts by average lot size:
1/8 acre (506 m2) or less (town houses) 65 77 85 90 92
38 61 75 83 87
1/4 acre (1,012 m2)
30 57 72 81 86
1/3 acre (1,349 m2) 25 54 70 80 85
1/2 acre (2,023 m2) 20 51 68 79 84
1 acre (4,047 m2) 12 46 65 77 82
2 acres (8,094 m2)
Developing urban areas
Newly graded area (pervious areas only, no vegetation)e 77 86 91 94
Idle lands (CNs are determined using cover types similar to those in Table 5.6)
a. Average runoff condition, and Ia = 0.2S.

b. The average percent impervious area shown was used to develop the composite CNs. Other assumptions are as follows: impervious areas are directly
connected to the drainage system, impervious areas have a CN of 98, and pervious areas are considered equivalent to open space in good hydrologic
condition.
c. CNs shown are equivalent to those of pasture. Composite CNs may be computed for other combinations of open space cover type.
d. Composite CNs for natural desert landscaping should be computed using Figure 2.3 or 2.4 (in TR-55) based on the impervious area percentage (CN = 98)
and the pervious area CN. The pervious area CNs are assumed equivalent to desert shrub in poor hydrologic condition.
e. Composite CNs to use for the design of temporary measures during grading and construction should be computed using Figure 2.3 or 2.4 (in TR-55) based
on the degree of development (impervious area percentage) and the CNs for the newly graded pervious areas.
Table 5.6 Runoff curve numbers for cultivated agricultural landsa (Mockus, 1969)
Cover Description Curve Numbers for Hydrologic Soil Group
Hydrologic
Cover Type Treatmentb Conditionc A B C D
Bare soil -- 77 86 91 94
Fallow Poor 76 85 90 93
Crop residue cover (CR)
Good 74 83 88 90
Poor 72 81 88 91
Straight row (SR)
Good 67 78 85 89
Poor 71 80 87 90
SR + CR
Good 64 75 82 85
Poor 70 79 84 88
Contoured (C)
Good 65 75 82 86
Row crops
Poor 69 78 83 87
C + CR
Good 64 74 81 85
Poor 66 74 80 82
Contoured & terraced (C&T)
Good 62 71 78 81
Poor 65 73 79 81
C&T + CR
Good 61 70 77 80
Poor 65 76 84 88
SR
Good 63 75 83 87
Poor 64 75 83 86
SR + CR
Good 60 72 80 84
Poor 63 74 82 85
C
Good 61 73 81 84
Small grain
Poor 62 73 81 84
C + CR
Good 60 72 80 83
Poor 61 72 79 82
C&T
Good 59 70 78 81
Poor 60 71 78 81
C&T + CR
Good 58 69 77 80
Poor 66 77 85 89
SR
Good 58 72 81 85
Close-seeded or Poor 64 75 83 85
broadcast legumes C
or rotation meadow Good 55 69 78 83
Poor 63 73 80 83
C&T
Good 51 67 76 80
a. Average runoff condition, and Ia = 0.2S.
b. Crop residue cover applies only if residue is on at least 5% of the surface throughout the year.
c. Hydrologic condition is based on a combination of factors that affect infiltration and runoff, including density and canopy of vegetative areas, amount
of year-round cover, amount of grass or close-seeded legumes in rotations, and degree of surface roughness. “Poor” indicates that there are factors that
impair infiltration and tend to increase runoff. “Good” indicates that there are factors that encourage average and better than average infiltration and
tend to decrease runoff.
Table 5.7 Runoff curve numbers for other agricultural landsa (Mockus, 1969)
Cover Description Curve Numbers for Hydrologic Soil Group
Hydrologic
Cover Type Condition A B C D
Poor 68 79 86 89
Pasture, grassland, or range-continuous forage for grazingb Fair 49 69 79 84
Good 39 61 74 80
Meadow-continuous grass, protected from grazing and generally
30 58 71 78
mowed for hay
Poor 48 67 77 83
Brush—brush-weed grass mixture with brush the major elementc Fair 35 56 70 77
Good 30d 48 65 73
Poor 57 73 82 86
Woods-grass combination (orchard or tree farm)e Fair 43 65 76 82
Good 32 58 72 79
Poor 45 66 77 83
Woodsf Fair 36 60 73 79
Good 30d 55 70 77
Farmsteads—buildings, lanes, driveways, and surrounding lots 59 74 82 86
a. Average runoff condition, and Ia = 0.2S
b. Poor: less than 50% ground cover or heavily grazed with no mulch. Fair: 50 to 75% ground cover and not heavily grazed. Good: more than 75%
ground cover and lightly or only occasionally grazed
c. Poor: less than 50% ground cover. Fair: 50 to 75% ground cover. Good: more than 75% ground cover
d. Actual curve number is less than 30; use CN = 30 for runoff computations
e. CNs shown were computed for areas with 50% woods and 50% grass (pasture) cover. Other combinations of conditions may be computed from the
CNs for woods and pasture.
f. Poor: Forest litter, small trees, and brush are destroyed by heavy grazing or regular burning. Fair: Woods are grazed but not burned, and some forest
litter covers the soil. Good: Woods are protected from grazing, and litter and brush adequately cover the soil
Table 5.8 Runoff curve numbers for arid and semiarid rangelandsa (Mockus, 1969)
Cover Description Curve Numbers for Hydrologic Soil Group:
Hydrologic
Cover Type Ac B C D
Conditionb
Herbaceous-mixture of grass, weeds, and low-growing brush, Poor 80 87 93
with brush the minor element Fair 71 81 89
Good 62 74 85
Oak-aspen-mountain brush mixture of oak brush, aspen, moun- Poor 66 74 79
tain mahogany, bitter brush, maple, and other brush Fair 48 57 63
Good 30 41 48
Pinyon-juniper-pinyon, juniper, or both; grass understory Poor 75 85 89
Fair 58 73 80
Good 41 61 71
Sagebrush with grass understory Poor 67 80 85
Fair 51 63 70
Good 35 47 55
Desert shrub-major plants include saltbush, greasewood, creo- Poor 63 77 85 88
sote-bush, blackbrush, bursage, palo verde, mesquite, and cactus Fair 55 72 81 86
Good 49 68 79 84
a. Average antecedent moisture condition, and Ia = 0.2S. For range in humid regions, use Table 5.7.
b. Poor: less than 30% ground cover (litter, grass, and brush overstory). Fair: 30 to 70% ground cover. Good: more than 70% ground cover.
c. Curve numbers for group A have been developed for desert shrub.
Example 5.5 – Estimating Direct Runoff Depth and Volume Using the NRCS
(SCS) Curve Number Method (Modified from SCS, 1986). Estimate the curve num-
ber, depth of direct runoff (effective precipitation), and direct runoff volume for a 400-ha drainage
basin if the total depth of precipitation is 127 mm. All soils in the basin are in hydrologic soil group
C. The proposed land use is 50 percent detached houses with 0.1-ha lots; 10 percent townhouses with
0.05-ha lots; 25 percent schools, parking lots, plazas, and streets with curbs and gutters; and 15 per-
cent open space, parks, and schoolyards with good grass cover. Use an antecedent soil moisture con-
dition of AMC-III. The detached housing and townhouse areas have directly-connected impervious
area percentages corresponding to the ranges assumed in Table 5.5.
Solution: The composite curve number corresponding to AMC-II conditions is computed as a
weighted average of the curve numbers presented in Table 5.5.
Land Use Area (ha) CN Area × CN

Detached houses (0.1-ha lots) 200 83 16,600
Townhouses (0.05-ha lots) 40 90 3,600
Streets, plazas, etc. 100 98 9,800
Open space, parks, etc. 60 74 4,400
Sums 400 34,400
Thus, the composite CN is

CN = 34,400 / 400 = 86
From Equation 5.17, the curve number corresponding to AMC-III moisture conditions is found to be
23(86)
CN III = = 93
10 + 0.13(86)
Using Equation 5.15 and converting inches to millimeters, the maximum possible retention for this
basin at AMC-III is
S = (1/0.0394)(1000/93 – 10) = 19 mm
Initial abstractions are estimated to be
Ia = 0.2S = 3.8 mm
Because P > Ia, Equation 5.13 is used to estimate the depth of direct runoff (effective precipitation) as
Pe = [127 – 0.2(19)]2/(127 + 0.8(19)] = 107 mm

From Equation 5.6, the total volume of direct runoff is
Vr = (107 mm /1000 mm/m) (400 ha) = 42.68 ha-m
Directly Connected and Unconnected Impervious Area Adjustment

Factors. Some of the curve numbers given in Table 5.5 are actually composite val-
ues based on the assumption that a certain percentage of the drainage area is impervi-
ous area with a CN of 98, and the remainder of the area has a CN corresponding to
open space in good hydrologic condition [see footnote (b) of Table 5.5]. Further, it is
assumed that the impervious portions of the drainage area are directly connected,
meaning that runoff from these areas flows directly into the drainage system without
first crossing a pervious area as unconcentrated flow. A directly connected impervious
area can be contrasted with an unconnected impervious area, which is an impervious
area whose runoff has more opportunity to infiltrate because it must flow over a pervi-
ous area before it becomes shallow concentrated flow or enters the drainage system.
An example of an unconnected impervious area could be a tennis court surrounded by
a grassy park area.
An engineer faced with a drainage area for which all impervious areas are directly
connected, but the percentage of impervious area differs from the values in Table 5.5,
can compute an adjusted curve number, CNc, similar to the way a weighted runoff
coefficient is computed:
(CN P )( A − Ai ) + 98 Ai
CN c = (5.18)
A
whereCNc=adjusted curve number

CNp=curve number for pervious area
A=total area (ac, ha)
Ai=impervious area (ac, ha)
If the drainage area contains unconnected impervious areas, or both unconnected and
directly connected impervious areas, the adjusted curve number is computed as
A    Au  
 A
( )
CN c = CN p +  i  98 − CN p 1 − 0.5  
Ai  
(5.19)
 
where Ai = impervious area, connected and unconnected (ac, ha)

A = total area (ac, ha)
Au = unconnected impervious area (ac, ha)
Method Limitations. The curve number method is widely used in the United
States, mainly because of its simplicity and ease of use. It should be noted, however,
that the curve number method has some significant limitations. Perhaps the most seri-
ous of these is that the dimension of time is not explicitly considered. In Example 5.5,
for instance, it should matter whether the 5 in. (127 Mm) of rain fell in a time span of
24 hours or 2 hours, but the method has no way of accounting for this difference.
Kibler (1982) observed that the method should be used only with 24-hour rainfall
durations (for which it was originally developed) and advises caution when analyzing
incremental amounts of rainfall occurring during a storm.
The CN method has several recognized limitations:
• The method describes average conditions, which makes it useful for design
purposes, but the method’s accuracy decreases for historical events.
• The curve number equation is not time-dependent, and it thus ignores differ-
ences resulting from varying rainfall duration and intensity.
• The common assumption that Ia = 0.2S is generalized from data for agricul-
tural watersheds; it may overestimate losses for impervious areas and under-
estimate losses for surface depressions.
• The method is not applicable when computing runoff due to snowmelt or

rain on frozen ground.
• The method is less accurate for runoff depths of less than 0.5 in. (13 mm).
• The CN method only computes direct runoff; it does not consider subsurface
flow or groundwater effects.
• If the weighted CN is less than 40, a different procedure must be used.
Another criticism of the curve number method is that its functional form implies that
the infiltration rate f = dF/dt can be expressed as (Morel-Seytoux and Verdin, 1981)
dF iS 2
f = = (5.20)
dt ( P + 0.8S ) 2
where i = dP/dt, the rainfall intensity (in/hr, mm/hr)

Because P and S are constants for a particular storm and drainage basin, this equation
implies that the infiltration rate must rise and fall with the rainfall intensity. This
implication is clearly unrealistic from a physical perspective.
A final criticism of the curve number method is that dPe/dP = 1 for any curve number
and sufficiently large P. The implication is that the infiltration rate approaches zero
during sufficiently long storms, which again is unrealistic physically unless the water
table rises to the ground surface or the soil is very poorly drained. As noted earlier, the
infiltration rate should approach the saturated vertical hydraulic conductivity of a soil
during long storms.
Creating Direct Runoff (Effective Rainfall) Hyetographs

Provided the data are available, the most accurate method of accounting for rainfall
abstractions and developing a direct runoff (effective rainfall) hyetograph is to model
abstractions and infiltration within each time step for a rainfall event. The process
consists of subtracting the interception from the beginning of a gross rainfall hyeto-
graph, subtracting infiltration from what remains after interception has been
accounted for, and subtracting depression storage from what remains after both inter-
ception and infiltration have been accounted for. This approach is physically based
and is not limited to use with any particular storm duration or rainfall hyetograph
shape. The procedure is illustrated in the following example.
Example 5.6 – Computing an Effective Rainfall Hyetograph. Develop a direct run-

off (effective rainfall) hyetograph for a watershed with an initial loss (interception capacity) of 0.3 in.,
a depression storage capacity of 0.2 in., and Horton infiltration parameters of f0 = 1.5 in/hr, fc = 0.3
in/hr, and k = 0.04 min-1. The rainfall hyetograph, tabulated for both incremental depth and average
intensity, is as follows:
t (min) P (in.) Avg. i (in./hr)

0–10 0.24 1.44
10–20 0.46 2.76
20–30 1.17 7.02
30–40 0.58 3.48
40–50 0.35 2.10
50–60 0.17 1.02
Solution: The interception capacity of 0.3 in. is subtracted first. Because 0.24 in. of rainfall occurs
during the first 10 minutes of the storm, all of that rainfall plus an additional 0.06 in. of the rainfall
occurring in the second 10 minutes of the storm is lost to interception. The rainfall hyetograph after
accounting for interception is given in column 3 in the following table.
(1) (2) (3)

t (min) P (in.) P − 0.3 (in.)
0–10 0.24 0
10–20 0.46 0.4
20–30 1.17 1.17
30–40 0.58 0.58
40–50 0.35 0.35
50–60 0.17 0.17
The infiltration rate f(t) can be calculated and tabulated as a function of t using Equation 5.1, where t0
is the time at which rainwater first begins to infiltrate (t0 = 10 min in this example, because rainfall
prior to that time is lost to interception and hence is not available for infiltration).
– 0.04 ( t – 10 )
f ( t ) = 0.3 + 1.2 e
Column 1 in the table below is the time since the beginning of rainfall, and column 2 is the time since
the beginning of infiltration. Column 3 is the infiltration rate computed using Equation 5.1. Column 4
contains incremental infiltration depths for each 10-minute period during the storm. For example, the
first value of F is computed as the average of the current and preceding infiltration rates multiplied by
the time interval ∆t = 10 min = 1/6 hr or, 0.22 = [(1.50 + 1.10)/2]/6.
(1) (2) (3) (4)

t (min) t – t0 (min) f(t) (in/hr) Incr. F (in.)
0
10 0 1.5
20 10 1.1 0.22
30 20 0.84 0.16
40 30 0.66 0.13
50 40 0.54 0.10
60 50 0.46 0.08
Subtraction of the infiltration depth in each time interval (column 3) from the corresponding rainfall
depth remaining after interception (column 2) leads to the hyetograph in column 4 (any negative val-
ues produced should be set equal to zero) below. Finally, subtraction of the depression storage capac-
ity of 0.2 in. leads to the effective rainfall hyetograph in column 5. Column 6 shows the effective
rainfall hyetograph converted to intensities.
(1) (2) (3) (4) (5) (6)

t (min) P − 0.3 (in.) Incr. F (in.) P (in.) Pe (in.) ie (in./hr)
0−10 0 0 0 0
10–20 0.4 0.22 0.18 0 0
20–30 1.17 0.16 1.01 0.99 5.94
30–40 0.58 0.13 0.45 0.45 2.70
40–50 0.35 0.10 0.25 0.25 1.50
50–60 0.17 0.08 0.09 0.09 0.54
The rainfall hyetograph and the effective rainfall hyetograph are illustrated in Figure 5.6.1 and
Figure 5.6.2. Note that the effective rainfall intensities are less than the actual rainfall intensities. Note
also that effective rainfall, and hence direct runoff, does not begin (in this example) until 20 minutes
after the beginning of the storm.
Figure E5.6.1 Rainfall hyetograph for Example 5.6

Figure E5.6.2 Effective rainfall (runoff) hyetograph for Example 5.6
5.3 MEASURES OF BASIN RESPONSE TIME

The maximum amount of flow discharged from a watershed at its outlet is related to
the amount of time required for the entire watershed to be contributing to the flow. In
modeling stormwater conveyance systems, the basin outlet may be taken as the loca-
tion of an inlet or some other point of interest in the system. It can take minutes,
hours, or even days from the onset of a rainfall event for the water falling in some
parts of a watershed to be contributing to flow at a point of interest.
Because some points in a watershed are hydraulically closer to the outlet point than
others, flow originating from different locations in the watershed will have differing
travel times to the outlet. The response time of a drainage basin is usually considered
to be the largest of all possible travel times, although it is sometimes taken as an aver-
age of all possible travel times. Estimates of peak runoff rates resulting from a rainfall
event are quite sensitive to estimates of basin response time and vary inversely with
them. That is, all else being equal, long response times are associated with small peak
discharges and vice versa. Estimates of basin response time are also relevant to selec-
tion of the computational time step size ∆t used for runoff prediction, as discussed in
Section 4.5 (page 91).
The first of two common measures of basin response time is the time of concentra-
tion, denoted by tc. The most widely adopted definition of time of concentration is the
time required for a drop of effective rainfall falling at the most hydraulically remote
point in a drainage basin to reach the basin outlet. The most hydraulically remote
point from the outlet is usually, but not always, the most geographically remote point
in the drainage basin. The significance of the time of concentration is easily recog-
nized when one realizes that it is the minimum amount of time that must elapse before
all parts of the drainage basin contribute to the flow at the basin outlet.
Section 5.3 Measures of Basin Response Time 135
The second measure of response time commonly used in runoff estimation is the
basin lag time, denoted by tL. Often called simply the basin lag or lag time, this
response time can be thought of as an approximate average of the possible travel
times for runoff in a drainage basin. In practice, the basin lag is usually assumed to be
the amount of time between the center of mass of a pulse of effective rainfall and the
peak of the resultant direct runoff hydrograph (see Figure 5.7; tp denotes time of peak
discharge). The basin lag time is often used when estimating a complete runoff
hydrograph as opposed to merely the peak runoff rate.
Figure 5.7
Basin lag time
Basin Response Time Estimation Methods

Many methods for estimating the time of concentration and basin lag are given in var-
ious private, federal, and local publications. Although each of these methods is differ-
ent (in some cases only slightly), all are based on the type of ground cover, the slope
of the land, and the distance along the flow path. In most localities, there is also a
minimum tc (typically 5 to 10 minutes) recommended for small watersheds such as a
section of a parking lot draining to a storm sewer. Some methods predict the response
time directly, and others predict the velocity of flow. The predicted velocity coupled
with estimates of the flow path length can then be used to estimate the response time.
With few exceptions, methods for the prediction of basin response time are empirical
in nature. Consequently, large errors in response time estimates can be expected to
occur if these methods are not carefully selected and applied. These errors can signif-
icantly affect peak runoff estimates. The method selected for estimation of tc or tL
should be one that was developed for basin conditions similar to those existing in the
drainage basin for which an estimate is desired. McCuen, Wong, and Rawls (1984)
compared a number of methods for estimating the time of concentration and devel-
oped measures of their reliability. As a general rule, methods that compute individual
travel times for various types of flow segments (for example, overland flows and
channelized flows), and then sum the individual travel times to estimate the total
travel time, are thought to be the most reliable.
Physically, the response time of a drainage basin depends on, at a minimum, the
length of the flow path, the slope of the basin, and the surface roughness. Additional
factors included in some prediction methods are rainfall intensity and a measure of
the basin shape. Because urbanization of a watershed tends to reduce surface rough-
ness and often changes flow path lengths and slopes, a change in basin response time
(nearly always a decrease) and corresponding increases in peak runoff rates should be
expected as a consequence of urbanization. One way to help reduce the increase in
peak runoff rate caused by urbanization is to increase the time of concentration and/or
basin lag through practices such as terracing of land surfaces.
Table 5.9 lists several commonly used methods for estimating basin lag time.
Table 5.10 lists several commonly used methods for estimating the time of concentra-
tion of a drainage basin. Figure 5.8 illustrates average overland flow velocities as a
function of land use characteristics and surface slope. When flows are channelized in
gutters, open channels, or storm sewers, Manning’s equation may be used to estimate
the velocity of flow. For more information, see Section 6.2 (page 198), Section 7.2,
and Section 10.1.
Table 5.9 Commonly used methods for estimation of basin lag time, in hours
Equation Source Remarks
Snyder (1938), Ct = empirical coefficient [typical range between 1.8
t L = Ct ( LLca )0.3 Linsley (1943) (steeper basins) and 2.2 (flatter basins)], L = basin length
(mi), and Lca = length along main channel to a point adja-
cent to the basin centroid (mi).
Taylor and D = drainage density, L = basin length (mi), Lca = length
t L = 0.6 exp{0.212 D ( LLca ) −0.36 } / S 0.5 Schwarz (1952) along main channel to a point adjacent to the basin centroid
(mi), and S = average channel slope (ft/ft).
0.8 Soil Conservation Lw = length of drainage basin (ft), CN = curve number of
t L = Lw (1000 / CN − 9)0.7 / 1900 S 0.5 Service (1986) drainage basin, and S = average basin slope (percent).
Kent (1972) Lag is approximated as 0.6tc for use in SCS Unit
t L = 0.6tc
Hydrograph computations.
Table 5.10 Commonly used methods for estimation of the time of concentration, in minutes
Equation Source Remarks
t c = 60 LA 0.4
/ DS 0.2 Williams (1922) L = basin length (mi), A = basin area (mi2), D = diameter
(mi) of a circular basin of area A, and S = basin slope (per-
cent). The basin area should be smaller than 50 mi2.
t c = KL0.77 / S n Kirpich (1940) Developed for small drainage basins in Tennessee and
Pennsylvania, with basin areas from 1 to 112 ac. L = basin
length (ft), S = basin slope (ft/ft), K = 0.0078 and n = 0.385
for Tennessee; K = 0.0013 and n = 0.5 for Pennsylvania.
The estimated tc should be multiplied by 0.4 if the overland
flow path is concrete or asphalt, or by 0.2 if the channel is
concrete-lined.
t c = ( 2 LN / 3S 0.5 ) 0.47 Hathaway (1945), Drainage basins with areas of less than 10 ac and slopes of
Kerby (1959) less than 0.01. This is an overland flow method. L = over-
land flow length from basin divide to a defined channel (ft),
S = overland flow path slope (ft/ft), and N is a flow retar-
dance factor (N = 0.02 for smooth impervious surfaces;
0.10 for smooth, bare packed soil; 0.20 for poor grass, row
crops, or moderately rough bare surfaces; 0.40 for pasture
or average grass; 0.60 for deciduous timberland; and 0.80
for coniferous timberland, deciduous timberland with deep
ground litter, or dense grass).
Johnstone and Developed for basins in the Scotie and Sandusky River
t c = 300( L / S ) 0.5
Cross (1949) watersheds (Ohio) with areas between 25 and 1,624 mi2.
L = basin length (mi), and S = basin slope (ft/mi).
Izzard (1946) Hydraulically derived formula. I = effective rainfall inten-
tc = {(0.007 I + c) / S 0.33} × sity (in/hr), S = slope of overland flow path (ft/ft), L =
( IL / 43,200) − 0.67 L / 60 length of overland flow path (ft), and c is a roughness coef-
ficient (c = 0.007 for smooth asphalt, 0.012 for concrete
pavement, 0.017 for tar and gravel pavement, and 0.060 for
dense bluegrass turf).
Henderson and Based on kinematic wave theory for flow on an overland
t c = 0.94 I −0.4 ( Ln / S 0.5 ) 0.6
Wooding (1964) flow plane. I = rainfall intensity (in/hr), L= length of over-
land flow (ft), n = Manning’s roughness coefficient, S =
overland flow plane slope (ft/ft).
Federal Aviation Developed based on airfield drainage data. C = rational
t c = 1.8(1.1 − C ) L0.5 / S 0.333 Agency (1970) method runoff coefficient, L = overland flow length (ft), and
S = slope (percent).
Soil Conservation Time of concentration is developed as a sum of individual
1
tc =
60
∑ (L / V )i Service (1986) travel times. L = length of an individual flow path (ft) and V
= velocity of flow over an individual flow path (ft/s). V may
be estimated by using Figure 5.8 or by using Manning’s
equation.
Figure 5.8
Average overland
flow velocities as a
function of land use
characteristics and
surface slope (Kent,
1972)
Example 5.7 – Estimating Time of Concentration Using NRCS (SCS) Methods.

An urbanized drainage basin is shown in Figure 5.7.1. Three types of flow conditions exist from the
furthest point of the drainage basin to its outlet. Estimate the time of concentration based on the fol-
lowing data:
Reach Flow Description Slope (%) Length (ft)
A to B Overland (forest) 7 500

B to C Overland (shallow gutter) 2 900
Storm sewer with manholes, inlets, etc.
C to D 1.5 2,000
(n = 0.015, diam. = 3 ft)
Open channel, gunite-lined, trapezoidal
D to E 0.5 3,000
(B = 5 ft, y = 3 ft, z = 1:1, n = 0.019)
Solution: For the reach from A to B, the average flow velocity is V = 0.7 ft/s (from Figure 5.8). The
travel time for that reach is therefore
tAB = L/V = 500/0.7 = 700 s
Similarly, for the reach from B to C, the average flow velocity is V = 2.8 ft/s. The travel time for that
reach is therefore
tBC = L/V = 900/2.8 = 320 s
To compute the travel time in the storm sewer from C to D, Manning’s equation is employed to com-
pute the pipe-full velocity:
2/3 2/3
1.49  D  1.49  3 
V=   S1/ 2 =   (0.015)1/ 2 = 10 ft/s
n 4 0.015  4 
Figure E5.7.1 Flow paths in a drainage basin for calculation of tc

The travel time for that reach is therefore
tCD = L/V = 2000/10 = 200 s
Travel time in the open channel from D to E is computed using the bank-full velocity, again found via
the Manning equation:
1.49 2/ 3 1/ 2 1.49
V= R S = (1.78)2 / 3 (0.005)1/ 2 = 8.2 ft/s
n 0.019
The travel time for that reach is therefore

tDE = L/V = 3,000/8.2 = 370 s
The time of concentration is the sum of the four individual travel times and is
tc = 1,590 s = 0.44 hr
Time of Concentration and Basin Lag

Because the time of concentration is the longest of all the travel times in a drainage
basin, and because the basin lag time can be thought of as an average of all the possi-
ble travel times, the time of concentration is longer than the basin lag time. The
NRCS has suggested that tc = 1.67tL, or that tL = 0.6tc. Overton and Meadows (1976)
indicated that tc = 1.6tL.
The NRCS, in Technical Release 55 (TR-55) (SCS, 1986) describes a procedure for
estimating the time of concentration of a drainage basin as the sum of up to three
types of individual travel times consisting of sheet flow, shallow concentrated flow,
and channelized flow. The method presented therein is essentially the same as that
represented by the last entry in Table 5.10 in that it involves estimation of flow veloc-
ities and then deduces travel times by considering the length of the flow paths. Con-
ceptually, Example 5.7 illustrates the same type of procedure.
5.4 PEAK FLOW ESTIMATION

To design stormwater conveyance or detention systems, the engineer must first esti-
mate runoff rates, as these are the discharges for which the conveyance facilities must
be designed. Two basic levels of analysis exist. The first level is a peak flow calcula-
tion to determine the maximum runoff rate at a given point resulting from a storm
event. This level of analysis is often sufficient for designing storm sewers and culverts
whose only function is to convey runoff away from areas where it is unwanted. The
second level, which is more complex, consists of the generation of a runoff
hydrograph to provide information on flow rate versus time and runoff volume. This
type of information is necessary when drainage basins are too large or too complex to
be treated by peak flow estimation methods, or when the analysis of natural or artifi-
cial detention or retention facilities is required. Section 5.5 describes hydrograph esti-
mation methods.
A number of methods are commonly used for the estimation of peak runoff rates.
Three popular methods are (1) the rational method, (2) the NRCS (SCS) peak flow
estimation method, and (3) regression-based methods developed by the U.S. Geologi-
cal Survey. Each of these three methods is described in the subsections that follow.
Rational Method
The rational method, also called the Lloyd-Davies method in the United Kingdom,
was developed in 1851 by Mulvaney. It is an equilibrium-based approach to peak flow
Section 5.4 Peak Flow Estimation 141
estimation that uses rainfall intensity data and watershed characteristics to predict
peak flows for a rainfall event. This method was originally presented in American
hydrologic literature by Kuichling (1889) and has been a staple of American hydro-
logic practice since that time. The rational method is especially popular in storm
sewer design because of its simplicity, and because storm sewer design typically
requires only peak discharge data.
At the most fundamental level, the rational method assumes that an equilibrium (that
is, a steady state) is attained such that the effective rainfall inflow rate of water onto a
drainage basin is equal to the outflow rate of water from the basin. If one expresses
the volumetric effective inflow rate as the product of the basin area A and the effective
rainfall intensity ie, then the outflow rate Q is obtained as Q = ieA. Further, if one
accounts for abstractions using a runoff coefficient, then the effective intensity is a
product of the actual rainfall intensity and the runoff coefficient, resulting in
Q = CiA (5.21)
where Q = runoff rate (ac-in/hr, ha-mm/hr)

C = runoff coefficient (see Table 5.2)
i = rainfall intensity (in/hr, mm/hr)
A = drainage area (ac, ha)
Because 1 ac-in/hr = 1.008 cfs ≈ 1 cfs, engineers performing calculations by hand in
U.S. customary units typically ignore the conversion factor and simply assume that
the discharge Q is in units of cfs. This conversion factor is not ignored in computer
applications.
Table 5.11 lists recommended runoff coefficients corresponding to various land uses.
It should be noted, however, that some locales have developed runoff coefficient
tables that also consider soil type and/or storm return period. Coefficients should be
selected carefully for proper application to a particular locale.
When a drainage basin consists of a mixture of land uses, a composite runoff coeffi-
cient may be computed for the basin by weighting individual runoff coefficients for
each land use by their respective areas, as demonstrated in Example 5.8.
The time of concentration used to find ie is the smallest time for which the entire basin
is contributing runoff to the basin outlet; therefore, the storm duration must be at least
as long as the time of concentration if a steady-state condition is to be achieved. Also,
steady-state conditions dictate that the storm intensity be spatially and temporally uni-
form. It is not reasonable to expect that rainfall will be spatially uniform over a large
drainage basin, or that it will be temporally uniform over a duration at least as long as
the time of concentration when tc (and hence A) is large. Therefore, these conditions
limit the applicability of the rational method to small drainage basins. An upper limit
of 200 acres (80.9 ha) has been suggested by some, but the limit should really depend
on the storm characteristics of the particular locale. These local characteristics may
limit the applicability of the rational method to basins smaller than 10 acres (4 ha) in
some cases.
Table 5.11 Runoff coefficients for use in the rational method

(Schaake, Geyer, and Knapp, 1967)
Type of Area or Development C
Types of Development
Urban business 0.70–0.95
Commercial office 0.50–0.70
Residential development
Single-family homes 0.30–0.50
Condominiums 0.40–0.60
Apartments 0.60–0.80
Suburban residential 0.25–0.40
Industrial development
Light industry 0.50–0.80
Heavy industry 0.60–0.90
Parks, greenbelts, cemeteries 0.10–0.30
Railroad yards, playgrounds 0.20–0.40
Unimproved grassland or pasture 0.10–0.30
Types of Surface Areas
Asphalt or concrete pavement 0.70–0.95
Brick paving 0.70–0.80
Roofs of buildings 0.80–0.95
Grass-covered sandy soils
Slopes 2% or less 0.05–0.10
Slopes 2% to 8% 0.10–0.16
Slopes over 8% 0.16–0.20
Grass-covered clay soils
Slopes 2% or less 0.10–0.16
Slopes 2% to 8% 0.17–0.25
Slopes over 8% 0.26–0.36
Example 5.8 – Determining the Weighted Runoff Coefficient. Estimate the runoff
coefficient for a drainage basin that is made up of 6 ha of park and 12 ha of medium-density, single-
family housing.
Solution: From Table 5.11, the runoff coefficients for the park and residential areas are estimated to
be 0.20 and 0.40, respectively. The composite runoff coefficient for the entire drainage basin of 18 ha
is therefore
C = [6(0.20) + 12(0.40)]/18 = 0.33
Additional assumptions associated with the rational method are that the runoff coeffi-
cient is a constant and does not change during the duration of the storm, and that the
recurrence intervals of the rainfall and corresponding runoff are equal (see, for exam-
ple, Schaake, Geyer, and Knapp, 1967).
When several drainage basins (or subbasins) discharge to a common facility (such as
a storm sewer or culvert), the time of concentration should be taken as the longest of
all the individual times of concentration. Further, the total drainage area served (the
sum of the individual basin areas) should be no larger than the 200-acre limit (or
smaller where applicable) of the rational method.
The basic steps for applying the rational method are as follows:
Step 1: Estimate tc and apply I-D-F data. Develop or obtain a set of
intensity-duration-frequency (IDF) curves for the locale in which the drainage basin
resides (see Chapter 4). Estimate the time of concentration of the basin using one of
the techniques presented in Section 5.3 or other applicable method. Assume that the
storm duration is equal to the time of concentration and use the IDF curves to deter-
mine the precipitation intensity for the recurrence interval of interest. Note that the
assumption that the storm duration and time of concentration are equal is conservative
in that it represents the highest intensity for which the entire drainage area can con-
tribute.
The rational method is frequently used in the design of storm sewer systems, and an
important part of the hydraulic analysis is determining the time of concentration for
each component in the system so that design flows can be computed correctly. The
time of concentration for an inlet’s contributing drainage area is used to compute the
inlet design flow. However, finding the appropriate time of concentration for use in
computing the design flow for a pipe additionally involves the consideration of travel
time through any upstream piping. The longest total travel time to the pipe is the con-
trolling time of concentration used in design flow calculation. An illustration of this
approach is provided later in the text in Example 11.1 (page 424).
Step 2: Compute watershed area. The basin area A can be estimated using
topographic maps, computer tools such as CAD or GIS software, or by field recon-
naissance.
Step 3: Choose C coefficient. The runoff coefficient C can be estimated using
Table 5.11 if the land use is homogeneous in the basin, or a composite C value can be
estimated if the land use is heterogeneous (see Example 5.5).
Step 4: Solve for peak flow. Finally, the peak runoff rate from the basin can
be computed using Equation 5.21.
The following example illustrates the use of the rational method for several subbasins
draining into a common storm sewer system.
Example 5.9 – Computing Peak Flow with the Rational Method. A proposed
roadway culvert near Memphis, Tennessee is to drain a 5-ha area with a runoff coefficient of 0.65.
The time of concentration for the drainage area is 15 min. Use the rational method to compute the
peak discharge (in m3/s) for a 25-year design storm. From Example 4.2 (page 90), the 15-min inten-
sity for a 25-year storm in Memphis is 5.96 in/hr (151.4 mm/hr).
Solution: From Equation 5.21, the peak flow is computed as
Q = CiA = (0.65)(151.4 mm/hr)(5 ha) = 492.05 ha-mm/hr
Converting the units, Q is 1.37 m3/s.
Additional Duration and Area Considerations. Differences in land sur-

face characteristics within a drainage basin can give rise to peak discharges larger
than those computed assuming basin homogeneity. For example, the impervious sur-
faces within a basin can be subdivided into two parts, one consisting of directly con-
nected impervious areas and the other consisting of indirectly connected or
unconnected impervious areas. In a residential subdivision, the directly connected
impervious areas consist of roadway and driveway surfaces from which runoff flows
directly to stormwater inlets. Runoff from indirectly connected impervious areas
passes over pervious areas prior to reaching channels, gutters, or inlets; thus, an infil-
tration opportunity exists for runoff from these areas. Rooftop areas are a common
example of unconnected impervious areas in a residential subdivision.
One problem with using the rational method as previously described is that it may
under-predict peak flow for heterogeneous drainage basins. This under-prediction
occurs because the time of concentration for the basin as a whole is longer than that
for the directly connected impervious areas alone. Thus, the intensity for the directly
connected impervious area considered alone will be higher than that of the basin as a
whole. In fact, it may be so much higher that the peak flow computed for the directly
connected impervious area only is greater than that computed for the whole basin,
despite the fact that its area is smaller. The higher peak flow estimate should govern in
such cases. Example 11.2 (page 427) illustrates this procedure.
NRCS (SCS) Peak Flow Estimation

The NRCS (SCS) has developed a simple procedure for estimating peak runoff rates
for small drainage basins. Graphical and tabular solutions to this procedure are pre-
sented in Technical Release 55 (TR-55) (SCS, 1986), and tabular and regression-
based solutions are presented in the Federal Highway Administration’s Hydrologic
Engineering Circular No. 22 (HEC-22) (Brown, Stein, and Warner, 2001). The HEC-
22 regression procedure is described in this section.
For use of this method, the drainage basin should have a fairly homogeneous distribu-
tion of CN, and the composite CN should be at least 40. The basin should also have
one main channel, though branches are acceptable if they have approximately equal
times of concentration. Further, this method was developed for use with 24-hour
storm rainfall depths. Its use with other storm durations is not advisable.
This method is implemented by first applying Equation 5.12 or 5.13 to estimate the
depth of effective precipitation during a storm. The peak discharge is then estimated
as
Qp = quAPe (5.22)
where Qp = peak runoff rate (cfs, m3/s)

qu = unit peak runoff rate (cfs/mi2/in, m3/s/km2/cm)
A = drainage basin area (mi2, km2)
Pe = depth of effective precipitation (in., cm)
The unit peak runoff rate is estimated as
qu = C f ⋅10 K (5.23)
where Cf = conversion factor (1.00 for U.S. units; 4.30 × 10-3 for SI units)
K = exponent given by Equation 5.24
The exponent K is computed as
K = C0 + C1 log10 tc + C2 (log10 tc ) 2 (5.24)
where C0, C1, and C2 are coefficients listed in Table 5.12

tc = time of concentration (hr)
The values of C0, C1, and C2 are given in Table 5.12 as a function of the SCS 24-hour
design storm distribution type and the ratio Ia /P, where P is the 24-hour rainfall depth
and Ia is the depth of initial abstractions, as covered in Section 5.2 (page 119). For
convenience, the ratio Ia/P is tabulated in Table 5.13 as a function of the curve num-
ber. The ratio Ia/P should not be less than 0.1, nor larger than 0.5. Further, the time of
concentration should be in the range of 0.1 to 10 hours.
When ponding or swampy areas exist in a drainage basin, it is recommended that the
peak discharge produced by Equation 5.22 be reduced to account for the temporary
storage of runoff. In this case, an adjusted peak flow Qpa should be computed as
Qpa = FpQp (5.25)
where Qpa = adjusted peak flow (cfs, m3/s)

Fp = adjustment factor obtained from Table 5.14
Ponds and swampy areas lying along the flow path used for computation of the time
of concentration should not be included in the area percentage in Table 5.14. Neither
channel nor reservoir routing can be accommodated with this method.
Table 5.12 Coefficients for SCS (NRCS) peak discharge method by storm type Ia /P
(Brown, Stein, and Warner, 2001)
Storm Type Ia /P C0 C1 C2
0.10 2.30550 –0.51429 –0.11750
0.15 2.27044 –0.50908 –0.10339
0.20 2.23537 –0.50387 –0.08929
0.25 2.18219 –0.48488 –0.06589
I 0.30 2.10624 –0.45696 –0.02835
0.35 2.00303 –0.40769 –0.01983
0.40 1.87733 –0.32274 0.05754
0.45 1.76312 –0.15644 0.00453
0.50 1.67889 –0.06930 0
0.10 2.03250 –0.31583 –0.13748
0.15 1.97614 –0.29899 –0.10384
0.20 1.91978 –0.28215 –0.07020
0.25 1.83842 –0.25543 –0.02597
IA 0.30 1.72657 –0.19826 0.02633
0.35 1.70347 –0.17145 0.01975
0.40 1.68037 –0.14463 0.01317
0.45 1.65727 –0.11782 0.00658
0.50 1.63417 –0.09100 0
0.10 2.55323 –0.61512 –0.16403
0.15 2.53125 –0.61698 –0.15217
0.20 2.50928 –0.61885 –0.14030
0.25 2.48730 –0.62071 –0.12844
II 0.30 2.46532 –0.62257 –0.11657
0.35 2.41896 –0.61594 –0.08820
0.40 2.36409 –0.59857 –0.05621
0.45 2.29238 –0.57005 –0.02281
0.50 2.20282 –0.51599 –0.01259
0.10 2.47317 –0.51848 –0.17083
0.15 2.45395 –0.51687 –0.16124
0.20 2.43473 –0.51525 –0.15164
0.25 2.41550 –0.51364 –0.14205
III 0.30 2.39628 –0.51202 –0.13245
0.35 2.35477 –0.49735 –0.11985
0.40 2.30726 –0.46541 –0.11094
0.45 2.24876 –0.41314 –0.11508
0.50 2.17772 –0.36803 –0.09525
Table 5.13 Ia /P for selected rainfall depths and curve numbers (adapted from Brown, Stein, and Warner, 2001)
Curve Number, CN
P (in.) 40 45 50 55 60 65 70 75 80 85 90 95
0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.44 0.21
1.00 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.35 0.22 0.11
1.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.44 0.33 0.24 0.15 0.10
2.00 0.50 0.50 0.50 0.50 0.50 0.50 0.43 0.33 0.25 0.18 0.11 0.10
2.50 0.50 0.50 0.50 0.50 0.50 0.43 0.34 0.27 0.20 0.14 0.10 0.10
3.00 0.50 0.50 0.50 0.50 0.44 0.36 0.29 0.22 0.17 0.12 0.10 0.10
3.50 0.50 0.50 0.50 0.47 0.38 0.31 0.24 0.19 0.14 0.10 0.10 0.10
4.00 0.50 0.50 0.50 0.41 0.33 0.27 0.21 0.17 0.13 0.10 0.10 0.10
4.50 0.50 0.50 0.44 0.36 0.30 0.24 0.19 0.15 0.11 0.10 0.10 0.10
5.00 0.50 0.49 0.40 0.33 0.27 0.22 0.17 0.13 0.10 0.10 0.10 0.10
5.50 0.50 0.44 0.36 0.30 0.24 0.20 0.16 0.12 0.10 0.10 0.10 0.10
6.00 0.50 0.41 0.33 0.27 0.22 0.18 0.14 0.11 0.10 0.10 0.10 0.10
6.50 0.46 0.38 0.31 0.25 0.21 0.17 0.13 0.10 0.10 0.10 0.10 0.10
7.00 0.43 0.35 0.29 0.23 0.19 0.15 0.12 0.10 0.10 0.10 0.10 0.10
7.50 0.40 0.33 0.27 0.22 0.18 0.14 0.11 0.10 0.10 0.10 0.10 0.10
8.00 0.38 0.31 0.25 0.20 0.17 0.13 0.11 0.10 0.10 0.10 0.10 0.10
8.50 0.35 0.29 0.24 0.19 0.16 0.13 0.10 0.10 0.10 0.10 0.10 0.10
9.00 0.33 0.27 0.22 0.18 0.15 0.12 0.10 0.10 0.10 0.10 0.10 0.10
9.50 0.32 0.26 0.21 0.17 0.14 0.11 0.10 0.10 0.10 0.10 0.10 0.10
10.00 0.30 0.24 0.20 0.16 0.13 0.11 0.10 0.10 0.10 0.10 0.10 0.10
10.50 0.29 0.23 0.19 0.16 0.13 0.10 0.10 0.10 0.10 0.10 0.10 0.10
11.00 0.27 0.22 0.18 0.15 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10
11.50 0.26 0.21 0.17 0.14 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10
12.00 0.25 0.20 0.17 0.14 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10
12.50 0.24 0.20 0.16 0.13 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10
13.00 0.23 0.19 0.15 0.13 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
13.50 0.22 0.18 0.15 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
14.00 0.21 0.17 0.14 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
14.50 0.21 0.17 0.14 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
15.00 0.20 0.16 0.13 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
Table 5.14 Adjustment factor, Fp, for ponds and swampy areas
Pond and Swamp Area
(percent) Fp
0 1.00
0.2 0.97
1.0 0.87
3.0 0.75
5.0 0.72
Example 5.10 – Using the NRCS (SCS) Method to Compute Peak Discharge.
Compute the peak discharge from the 1,000-acre (1.56 mi2) drainage basin described in Example 5.9,
but assume AMC-II applies. The rainfall distribution is an SCS Type II, and the time of concentration
of the drainage basin is 0.9 hr. No ponds or swampy areas exist in the basin.
Solution: From Example 5.5, P = 5.0 in. and CN = 86 for the AMC-II condition. For this AMC-II
condition, the maximum possible retention S is 1.63 in., and the effective precipitation Pe is 3.46 in.
From Table 5.13, the ratio Ia/P is 0.10; and from Table 5.12, the applicable coefficients are C0 =
2.55323, C1 = –0.61512, and C2 = –0.16403. Substitution of these values into Equation 5.24 yields
K = 2.55323 – 0.61512 log10(0.9) – 0.16403[log10(0.9)]2 = 2.58104

From Equation 5.23, the unit peak runoff rate is
qu = 102.58104 = 381 cfs/mi2/in.

The peak discharge from the drainage basin is computed from Equation 5.22 as
Qp = 381(1.56)(3.46) = 2,057 cfs
USGS Regression-Based Formulas

The U.S. Geological Survey (USGS), in its efforts related to estimation of flood fre-
quencies for ungauged locations, developed formulas for the prediction of peak dis-
charges corresponding to various recurrence intervals. These formulas were
developed on a state-by-state basis but may vary from one region of a state to another.
Separate formulas are usually presented for rural and urbanized areas, and for differ-
ent recurrence intervals. The nearest office of the USGS may be contacted to obtain
information for a desired location. Jennings, Thomas, and Riggs (1994) present a
nationwide summary of the rural equations, including a computer program for their
implementation. Sauer et al. (1983) present equations permitting rural peak flows to
be converted to urbanized peak flows.
Failure Probability and Hydraulic

Structure Design
Although statistical methods, probability theory, and a very high probability of hydraulic failure of the sys-
risk analysis are beyond the scope of this text, it is tem during its lifetime. Conversely, if the probability
worthwhile to examine an application of such meth- of failure must be low, the recurrence interval
ods to stormwater management system design. As an becomes large, and the capital costs of system con-
illustration, a culvert having a design return period of struction are also large.
50 years and an expected life of 75 years can be
In practice, engineers often adopt design recurrence
shown using the binomial probability distribution to
intervals specified by the regulatory and/or review
have a failure probability of 0.78, or 78 percent. This agency having jurisdiction over the area in which a
probability does not reflect the likelihood of a struc-
conveyance system will be built. In the interest of
tural failure such as collapse; rather, it is the probabil-
minimizing the cost of any one system (so as to per-
ity that the design discharge capacity of the culvert mit resources to be allocated more or less equally
will be exceeded one or more times during its life-
among the many systems in the jurisdictional area),
time.
recurrence intervals are typically set fairly low. The
The probability of failure determined for the culvert recurrence interval used in the first example above is
in this example seems exceedingly high. It is certainly typical of real-world practice.
unacceptable for a structural engineer to design facili-
The high probability of operational failure, and thus
ties with such high structural failure rates. However, overflow, typically associated with stormwater con-
hydrologic and hydraulic engineering differs in that
veyance systems means that significant attention
requiring a small failure probability can be prohibi-
needs to be paid to the issue of what happens when a
tively expensive. If the culvert described previously flood does occur. The operation of the major drainage
were required to have a failure probability of 5 per-
system (the often unplanned, naturally occurring sys-
cent or less, then the design storm return frequency
tem that takes over when the minor flow system is
for an expected culvert life of 75 years would need to overtaxed), which historically has not received much
be at least 1,500 years!
attention, needs some careful investigation and plan-
It should be clear from these examples that a trade-off ning. It is not enough to simply call an exceedance of
must occur in the selection of the recurrence interval a system design capacity an Act of God, especially
for which a stormwater conveyance system is to be when such Acts have such a large probability of
designed. If the recurrence interval is small and the occurring.
system capital costs are correspondingly low, there is
To develop the predictive formulas, the USGS begins by estimating flood frequencies
at stream flow gauging locations throughout a state. These estimates are accom-
plished by fitting the log Pearson Type III probability distribution to data series of
annual flood peaks. Using the fitted distribution, flood discharges corresponding to
various recurrence intervals are predicted. It is then assumed that flood discharges for
a particular recurrence interval, and for any location in an assumed homogeneous
region, can be predicted on the basis of easily measured drainage basin characteris-
tics.
Stepwise multiple regression is used to develop the predictive relationship for each
recurrence interval and region. Predictor variables typically examined include basin
area, main channel slope and length, basin elevation, percentage of basin covered by
swamps and lakes, land use and soil conditions, and precipitation. The drainage basin
area is usually the most significant predictor, and is often the only one retained in the
relationship. For example, the predictive formulas developed for rural basins in north-
ern Alabama for the 2-, 5-, 10-, 25-, 50-, and 100-year storms are (Olin, 1984):
Q2 = 182A0.706 Q5 = 291A0.711 Q10 = 372A0.714
Q25 = 483A0.717 Q50 = 571A0.720 Q100 = 664A0.722
In these equations, QT is the peak discharge, in cfs, corresponding to the T-year recur-
rence interval, and A is the drainage basin area in mi2. These northern Alabama for-
mulas are limited to use for drainage basins with areas between 1 and 1,500 mi2 (259
and 388,500 ha) and have standard errors that range from 29 to 36 percent.
When applying regression-based formulas to compute peak runoff rates from a drain-
age basin, the engineer must check the range of areas used in formula development
[1 mi2 to 1,500 mi2 (259 and 388,500 ha) in the Alabama case] to ensure that the
equation is applicable to the basin being evaluated. If the basin area is outside of this
range, the predicted peak discharge for a relatively high recurrence interval can be
smaller than the predicted peak discharge for a lower recurrence interval. Similarly,
the predicted peak discharge for an urban area can be less than the predicted peak dis-
charge for a rural area. Finally, it should be recognized that the standard errors associ-
ated with regression-based formulas for rural basins are typically on the order of 30
percent (35 to 50 percent for urban basins). Therefore, significant errors may arise
when using this method.
Example 5.11 – Using a Regression Equation to Compute the Peak Discharge

Rate. Use the rural equations for northern Alabama to estimate the 50-year peak discharge from a
drainage basin with an area of 25 mi2. The standard error for this formula is 33 percent (Olin, 1984).
For this basin, what is the standard error of the estimate in cfs?
Solution: The peak discharge is
Q50 = 571(25)0.720 = 5,800 cfs

The standard error Se associated with the estimate is
Se = 0.33(5,800) = 1,900 cfs
5.5 HYDROGRAPH ESTIMATION

A hydrograph represents the runoff rate (discharge) as it varies over time at a particu-
lar location within a watershed. The integrated area under a hydrograph represents the
volume of runoff. Base flow, if present, represents subsurface flow from groundwater
that discharges into the conveyance channel. Many storm conveyance channels are
dry at the beginning of a rainfall event, which equates to a base flow of zero as long as
the water table does not rise into the channel during the storm. Figure 5.9 displays the
various components of a surface runoff hydrograph for a case in which the base flow
is zero and demonstrates that the area under the hydrograph is equal to the runoff vol-
ume as initially presented in Section 5.2.
Section 5.5 Hydrograph Estimation 151
Figure 5.9
Hydrograph definition
sketch
Estimation of a runoff hydrograph, as opposed to merely the peak rate of runoff, is

necessary to account for the effects of storage in a drainage basin. Because a
hydrograph accounts for volume and flow variations over an entire rainfall event, it is
useful for analyzing complex watersheds and designing detention ponds. Of particular
importance to the design engineer is assessment of the effects of storage associated
with natural ponds and lakes, and with constructed stormwater detention and/or reten-
tion facilities. Hydrograph estimation is also necessary in the assessment of the
impacts of storm duration and/or hyetograph shape on runoff production, and in cases
of two or more adjacent drainage basins (or subbasins) discharging to a common
stream.
Most approaches to hydrograph estimation are based on the concept of a unit
hydrograph, which is a hydrograph produced by a unit depth of runoff (usually 1 in.
or 1 cm) distributed uniformly over a basin for defined period of time. The unit
hydrograph must be convolved (combined) with an effective precipitation (runoff)
hyetograph to obtain the direct runoff hydrograph from a basin. Direct runoff
hydrograph estimation using unit hydrograph approaches is described in detail in this
section.
Several alternatives to the unit hydrograph method are used for runoff hydrograph
estimation, including hydraulic methods such as the kinematic wave model. This
method is a more physically-based, hydraulic approach to hydrograph estimation but
requires that the surface of a drainage basin be idealized as a set of overland flow
planes and channels. A brief introduction to the kinematic wave model is presented
later in this section (page 162).
Overview of Direct Runoff Hydrograph Estimation

Using a Unit Hydrograph
Development of a unit hydrograph is an important intermediate step in computing a
direct runoff hydrograph for design or analysis purposes. After a unit hydrograph has
been established for a particular basin, it can be used to compute the direct runoff
hydrographs resulting from different rainfall events.
Estimation of a direct runoff hydrograph using a unit hydrograph approach involves
several steps. The basic procedure is given below, with references to the specific sec-
tions of this text and examples detailing each step.
1. Establish a design storm rainfall hyetograph (see Section 4.5).
2. Convert the design storm rainfall hyetograph into a direct runoff (effective rain-
fall) hyetograph by subtracting rainfall abstractions (see Section 5.2, page 132)
and Example 5.6 (page 132).
3. Estimate a unit hydrograph for the drainage basin of interest (see the next section
and Example 5.12, page 157).
4. Merge the direct runoff (effective rainfall) hyetograph and the unit hydrograph
using discrete convolution to produce a direct runoff hydrograph, which repre-
sents the surface runoff hydrograph for the complex storm pattern represented by
the rainfall hyetograph [see Section 5.5 and Example 5.13 (page 161)].
5. If appropriate, add base flow to the direct runoff hydrograph to obtain the total
hydrograph representing both surface runoff and base flow (see Section 5.6).
Figure 5.10 illustrates the basic process of developing a hydrograph without baseflow.
Unit Hydrographs
The concept of the unit hydrograph was introduced to the hydrologic community by
Sherman (1932). The basic theory rests on the assumption that the runoff response of
a drainage basin to an effective rainfall input is linear; that is, it may be described by a
linear differential equation. Practically speaking, this means that the concepts of pro-
portionality and superposition can be applied. For example, the direct runoff volume
and discharge rates resulting from 2 in. of effective rainfall in a given time interval are
four times as great as those caused by 0.5 in. of effective rainfall in the same amount
of time. It also means that the total amount of time for the basin to respond to each of
these rainfall depths is the same. Consequently, the base length of each of the direct
runoff hydrographs is the same.
The unit hydrograph approach to runoff estimation is a spatially lumped approach,
meaning that it assumes that no spatial variability exists in the effective rainfall input
into a drainage basin. In cases where the effective rainfall input varies from one loca-
tion to another within a basin, the unit hydrograph approach requires the subdivision
of the basin into smaller subbasins, with routing of the runoff from each subbasin to
obtain the integrated basin response.
Figure 5.10
Hydrograph
development process
Visualizing and Developing Unit Hydrographs. By definition, the ∆t-

hour (or ∆t-minute) unit hydrograph of a drainage basin is the direct runoff
hydrograph produced by 1 in. (or 1 cm in the case of SI units) of effective rainfall
(runoff) falling uniformly in time and space over the drainage basin during a duration
∆t time units. By conservation of mass, the volume of direct runoff produced (repre-
sented by the area under a graph of the unit hydrograph) must be equal to 1 in. (or 1
cm) of runoff (effective rainfall) times the drainage basin area (see Figure 5.11). It
may be observed from this definition that the unit hydrographs for two drainage
basins must be different from one another, because the basin areas (as well as other
factors) are different.
Figure 5.12 helps to demonstrate the principles of a unit hydrograph. A sprinkler sys-
tem distributes water uniformly over an entire drainage area. Constant rainfall is mod-
eled by using the sprinklers to apply a constant rate of water until exactly one inch of
direct runoff (effective precipitation) is generated, at which time the sprinklers are
turned off. The duration of time required for the sprinklers to generate the required
depth of effective precipitation is shown as ∆t. A stream gauge is placed on the down-
stream end of the contributing drainage area to record the runoff hydrograph resulting
from the 1 in. (or 1 cm) of runoff. The recorded hydrograph represents the unit
hydrograph generated for the drainage area.
For a given drainage basin, there are an infinite number of durations over which 1 in.
of direct runoff (effective rainfall) could be generated depending on the intensity of
the rainfall being applied to the area. For instance, 1 in. of runoff could occur over
∆t = 30 minutes, over ∆t = 1 hour, or over some other time interval. Thus, for any
given drainage basin, one could speak of its 30-minute unit hydrograph, its 1-hour
unit hydrograph, or its unit hydrograph of any other duration. Note that the
duration ∆t
Figure 5.11
Unit hydrograph
resulting from 1 in. of
effective rainfall
(runoff) over time ∆t
Figure 5.12
Visualizing the unit
hydrograph concept
associated with a unit hydrograph is the duration over which the 1 in. (or 1 cm) of
effective rainfall occurs and not the duration over which the corresponding runoff
drains from the area.
Two basic categories of unit hydrographs exist: (1) unit hydrographs for gauged
watersheds and (2) synthetic unit hydrographs (described in the next section). When
concurrent rainfall and runoff records are available for a drainage basin, methods of
deconvolution (separation) may be applied to estimate a unit hydrograph for the basin
on the basis of those records (Singh, 1988). More commonly, rainfall and runoff
records for a drainage basin do not exist, and one must resort to synthesis of a unit
hydrograph based on information that can be gathered about the basin. Extensive lit-
erature exists on various ways to calculate synthetic unit hydrographs, including pro-
cedures proposed by Clark, Snyder, and Singh (Snyder, 1938; Singh, 1988). This text
provides an introduction to the development of synthetic unit hydrographs by apply-
ing the NRCS (SCS) method.
NRCS (SCS) Synthetic Unit Hydrographs. The NRCS (SCS) analyzed a
large number of unit hydrographs derived from rainfall and runoff records for a wide
range of basins and basin locations and developed the average dimensionless unit
hydrograph shown in Figure 5.13 (Snider, 1972). The times on the horizontal axis are
expressed in terms of the ratio of time to time of peak discharge (t/tp), and the dis-
charges on the vertical axis are expressed in terms of the ratio of discharge to peak
discharge (Q/Qp). Table 5.15 lists the dimensionless unit hydrograph ordinates.
Figure 5.13
NRCS (SCS)
dimensionless unit
hydrograph (from
Snider, 1972)
Application of the dimensionless unit hydrograph involves estimating the lag time tL
of the drainage basin. The lag time can be estimated by relating it to an estimate of the
time of concentration, or it can be estimated directly (see Section 5.3). The time to
peak of the synthetic unit hydrograph of duration ∆t is then computed as
∆t
tp = + tL (5.26)
2
The NRCS recommends that ∆t be equal to 0.133tc, or equal to 0.222tL (Snider,

1972). A small variation from this value is acceptable. The ∆t chosen for the develop-
ment of the synthetic unit hydrograph must be consistent with the ∆t chosen for devel-
opment of the design storm and effective rainfall hyetographs (see Section 4.5, page
91).
Table 5.15 Ordinates of the NRCS (SCS) dimensionless unit hydrograph (from Snider,
1972)
t/tp Q/Qp t/tp Q/Qp
0 0.000 1.7 0.460
0.1 0.030 1.8 0.390
0.2 0.100 1.9 0.330
0.3 0.190 2.0 0.280
0.4 0.310 2.2 0.207
0.5 0.470 2.4 0.147
0.6 0.660 2.6 0.107
0.7 0.820 2.8 0.077
0.8 0.930 3.0 0.055
0.9 0.990 3.2 0.040
1.0 1.000 3.4 0.029
1.1 0.990 3.6 0.021
1.2 0.930 3.8 0.015
1.3 0.860 4.0 0.011
1.4 0.780 4.5 0.005
1.5 0.680 5.0 0.000
1.6 0.560
The peak discharge Qp for the synthetic unit hydrograph is calculated as
C f KAQ
Qp = (5.27)
tp
where Qp = peak discharge (cfs, m3/s)

Cf = conversion factor (645.33 U.S., 2.778 SI)
K = 0.75 [a constant based on geometric shape of dimensionless unit
hydrograph (Snider, 1972)]
Q = runoff depth for unit hydrograph calculation (1 in. U.S., 1 cm SI)
A = the drainage basin area (mi2, km2)
tp = the time to peak (hr)
Simplifying yields
484 A
Qp = (for U.S. units) (5.28)
tp
or
2.08 A
Qp = (for SI units) (5.29)
tp
The coefficients 484 and 2.08 appearing in the numerators of Equation 5.28 and
Equation 5.29 include a unit conversion factor and are average values for many drain-
age basins. These values may be reduced to about 300 and 1.29, respectively, for flat
or swampy basins, or increased to about 600 and 2.58, respectively, for steep or
mountainous basins. Care should be taken when changing this coefficient, as the base
length and/or shape of the synthetic unit hydrograph must also be changed to ensure
that it represents a volume of water equivalent to 1 in. or 1 cm of effective rainfall
over the drainage basin area.
After tp and Qp are estimated using Equation 5.26 and Equation 5.27, the desired syn-
thetic unit hydrograph may be graphed or tabulated using the dimensionless unit
hydrograph shown in Figure 5.13 and Table 5.15.
Example 5.12 – Developing an NRCS (SCS) Synthetic Unit Hydrograph. Develop

a synthetic unit hydrograph for a 1.5-mi2 drainage basin in Memphis, Tennessee, having a time of
concentration of 90 min. Assume that the basin slopes are moderate so that the factor 484 can be
applied in computing Qp.
Employing the guidelines given above for estimation of ∆t, its value should be about
0.133tc = 0.133(90) = 12 min
A duration of ∆t = 10 min is selected, and the synthetic unit hydrograph will be a 10-min unit
hydrograph. The basin lag is estimated as
tL = 0.6tc = 54 min
Using Equation 5.26, the time to peak of the synthetic unit hydrograph is
tp = (10/2) + 54 = 59 min = 0.98 hr
The peak discharge is estimated using Equation 5.28 as
Qp = 484(1.5)/0.98 = 740 cfs
Ordinates of the synthetic unit hydrograph are tabulated in Table 5.16 and plotted in Figure 5.12.1.
The first column of the table is the time t, in minutes, and is tabulated in ∆t = 10 min intervals. The
second column is the dimensionless time ratio t/tp, where tp = 59 min. The third column is the dimen-
sionless discharge ratio, and is determined using the dimensionless time ratio and interpolation from
Table 5.15. The fourth and last column contains the ordinates of the 10-min unit hydrograph, which
are computed as the products of the dimensionless discharge ratios and Qp = 740 cfs.
Table 5.16 SCS (NRCS) synthetic unit hydrograph computations for Example 5.12
(1) (2) (3) (4)

t (min) t/tp Q/Qp Q (cfs)
0 0.00 0.00 0
10 0.17 0.08 59
20 0.34 0.24 178
30 0.51 0.49 363
40 0.68 0.79 585
50 0.85 0.96 710
60 1.02 1.00 740
70 1.19 0.93 688
Table 5.16 (cont.) SCS (NRCS) synthetic unit hydrograph computations for Example 5.12
(1) (2) (3) (4)
t (min) t/tp Q/Qp Q (cfs)
80 1.36 0.81 599
90 1.53 0.64 474
100 1.69 0.47 348
110 1.86 0.35 259
120 2.03 0.26 192
130 2.20 0.21 155
140 2.37 0.16 118
150 2.54 0.12 89
160 2.71 0.09 67
170 2.88 0.07 52
180 3.05 0.05 37
190 3.22 0.04 30
200 3.39 0.03 22
210 3.56 0.02 15
220 3.73 0.02 15
230 3.90 0.01 7
240 4.07 0.01 7
250 4.24 0.01 7
260 4.41 0.01 7
270 4.58 0.00 0
Figure E5.12.1 Synthetic unit hydrograph for Example 5.12
Discrete Convolution
A unit hydrograph represents the runoff hydrograph from a drainage basin subjected
to 1 in. or 1 cm of direct runoff (effective precipitation) applied over duration ∆t.
However, a runoff (effective rainfall) hyetograph typically contains many ∆t time
periods, each of which has its own associated runoff depth (see Example 5.6 on page
132). Discrete convolution is the process by which the direct runoff hydrograph
resulting from a complete rainfall hyetograph is computed by applying a unit

hydrograph to each discrete time step within the hyetograph.
To conceptualize the process of discrete convolution, refer to Figure 5.14 and
Figure 5.15. Figure 5.14 illustrates a runoff (effective rainfall) hyetograph consisting
of np = 3 rainfall pulses, each of duration ∆t = 10 min, with the depth of each pulse
denoted by Pi (i = 1, 2…np). Figure 5.15 shows a unit hydrograph with nu = 11 non-
zero ordinates shown at ∆t time intervals, with the non-zero unit hydrograph ordinates
denoted by Uj (j = 1, 2 … nu). [Note also that the ordinates of the unit hydrograph are
in units of cfs per inch of direct runoff (effective precipitation), and thus the notation
of this subsection is different from the previous one in which unit hydrograph ordi-
nates were denoted by Q in cfs.] The time increment ∆t used for development of the
runoff (effective rainfall) hyetograph must be the same as the ∆t duration of excess
rainfall used to create the unit hydrograph.
Figure 5.14
Effective rainfall
(runoff) hyetograph
Figure 5.15
Unit hydrograph
ordinates
Recall that the ordinates Uj of the unit hydrograph represent the basin response to 1
in. or 1 cm of direct runoff (effective rainfall) applied over a time of ∆t time units.
Because unit hydrograph theory assumes that drainage basins behave linearly, the
ordinates Qj of the runoff hydrograph produced by P1 inches of effective rainfall must
be equal to P1Uj. These ordinates are shown in the third column of Table 5.17 (the
first and second columns of the table reproduce the information in Figure 5.15). These
ordinates are calculated by multiplying each unit hydrograph ordinate by the runoff
depth for pulse P1 = 0.2 in. The values in column 3 are shifted down such that the
time at which the runoff hydrograph in column 3 begins is the same as the time at
which the first pulse of effective rainfall begins.
Columns 4 and 5 in Table 5.17 are computed by multiplying the unit hydrograph ordi-
nates in the second column by the effective rainfall depths P2 = 1.5 in. and P3 = 0.7
in., respectively. Again, the time at which each runoff pulse’s hydrograph begins cor-
responds to the times at which each runoff (effective rainfall) pulse begins. The sum
of the runoff hydrograph ordinates in each row of the table (shown in column 6) is an
ordinate of the direct runoff hydrograph caused by the complete rainfall event.
Table 5.17 Discrete convolution

(1) (2) (3) (4) (5) (6)
t (min) U (cfs/in) P1U (cfs) P2U (cfs) P3U (cfs) Q (cfs)
0 0 0
10 20 0
20 70 0 0
30 160 4 0 4
40 240 14 30 0 44
50 180 32 105 14 151
60 110 48 240 49 337
70 75 36 360 112 508
80 50 22 270 168 460
90 33 15 165 126 306
100 21 10 113 77 200
110 10 7 75 53 134
120 0 4 50 35 89
130 2 32 23 57
140 0 15 15 30
150 0 7 7
160 0 0
Figure 5.16 shows the runoff hydrograph resulting from each rainfall pulse and the
total runoff hydrograph for the entire rainfall event.
Figure 5.16
Summing of
hydrographs from
individual rainfall
pulses
The tabular process of discrete convolution illustrated in Table 5.17 can be summa-
rized in equation form for computation of the non-zero ordinates Qk of the direct run-
off hydrograph as
Qk =
∑ Pi Uk – i + 1 (5.30)
i=1
where k = 1, 2 … nq, and nq = np + nu – 1

To clarify Equation 5.30, Q1 = P1U1, Q2 = P1U2 + P2U1, Q3 = P1U3 + P2U2 + P3U1,
etc. These expressions are easily rationalized with reference to Table 5.15.
Example 5.13 – Computing a Runoff Hydrograph. Compute the direct runoff

hydrograph for the Memphis, Tennessee, drainage basin described in Example 5.12 on page 157. Use
the unit hydrograph developed for the basin in that example, and use the runoff (effective rainfall)
hyetograph developed in Example 5.6 on page 132.
Solution: There are a total of np = 4 runoff (effective rainfall) pulses, and a total of nu = 26 non-zero
unit hydrograph ordinates. Thus, there will be a total of nq = 29 non-zero direct runoff hydrograph
ordinates.
Table 5.18 illustrates the tabular calculation (discrete convolution). The first column is the time since
the beginning of rainfall in ∆t = 10 min increments (the duration of the unit hydrograph and the dura-
tion of the effective rainfall pulses). The second column contains the ordinates of the unit hydrograph
from Example 5.12. The third through the sixth columns are the unit hydrograph ordinates multiplied
by the effective rainfall depths from Example 5.6. Note that the first entry (the first zero) in each col-
umn corresponds to the time at which the corresponding effective rainfall pulse begins (t = 20 min for
the first pulse, t = 30 min for the second pulse, and t = 40 min for the third pulse; see Figure 5.6.2).
The seventh column is the sum of the previous four, and is the direct runoff hydrograph.
Table 5.18 Discrete convolution to obtain direct runoff hydrograph
(1) (2) (3) (4) (5) (6) (7)

t (min) U (cfs/in.) 0.99U 0.45U 0.25U 0.09U Q (cfs)
0 0 0
10 59 0
20 178 0 0
30 363 58 0 58
40 585 176 27 0 203
50 710 359 80 15 0 454
60 740 579 163 45 5 792
70 688 703 263 91 16 1073
80 599 733 320 146 33 1231
90 474 681 333 178 53 1244
100 348 593 310 185 64 1152
110 259 469 270 172 67 977
120 192 345 213 150 62 769
130 155 256 157 119 54 585
140 118 190 117 87 43 436
150 89 153 86 65 31 336
160 67 117 70 48 23 258
170 52 88 53 39 17 197
180 37 66 40 30 14 150
190 30 51 30 22 11 115
200 22 37 23 17 8 85
210 15 30 17 13 6 65
220 15 22 14 9 5 49
230 7 15 10 8 3 36
240 7 15 7 6 3 30
250 7 7 7 4 2 19
260 7 7 3 4 1 15
270 0 7 3 2 1 13
280 7 3 2 1 12
290 0 3 2 1 6
300 0 2 1 2
310 0 1 1
320 0 0
Kinematic Wave Model

Unit hydrograph-based approaches for runoff estimation tend to be straightforward
mathematical procedures with only a limited physical basis (but see Rodríguez-Iturbe
and Valdes, 1979, for a physical interpretation of them). A more physically based
approach to runoff estimation is the kinematic wave model, which involves applica-
tion of the fundamental laws of conservation of mass and momentum to overland and
channelized flows. This approach is hydraulic (as opposed to hydrologic) in nature, as
it applies conservation of mass and momentum to describe free-surface (open-
channel) flows.
As mentioned previously, use of the kinematic wave model requires that the surface of
a drainage basin be idealized as a set of overland flow planes and collector channels
(see Figure 5.17). One-dimensional, free-surface flow on an overland flow plane or in
a collector channel is then described using the Saint-Venant equations, which are a
pair of partial differential equations describing conservation of mass and momentum
(see Appendix C).
Occasionally, some of the terms appearing in the momentum equation can be elimi-
nated because they are small in comparison to the remaining terms. When all terms
are retained in the momentum equation, the Saint-Venant equations become what is
known as the dynamic wave formulation. When the so-called inertial acceleration
terms are dropped in the momentum equation, the diffusion wave formulation results.
When only the gravitational and frictional terms are retained in the momentum equa-
tion, the kinematic wave formulation results (see also Appendix C).
The Saint-Venant equations, when the kinematic wave simplification is invoked, take
the forms given in Equation 5.31 and Equation 5.32:
∂A ∂Q
+ = qo (5.31)
∂t ∂x
where A = cross-sectional area of flow (ft2, m2)

t = time (s)
Q = flow rate (ft3/s, m3/s)
x = distance along the flow path (ft, m)
qo = lateral discharge added to the flow per unit length of the flow path
(cfs/ft, m3/s/m)
A = α Qβ (5.32)
where α and β are coefficients representing the physical nature of the land surface or
stream channel and the type of constitutive relationship applied.
The lateral discharge qL in Equation 5.31 may be due to additional flow entering the
channel from its sides. It may also be due to rainfall directly onto the water surface, or
due to infiltration into the channel bottom (in which case qL would be negative).
Equation 5.32, which is a simplified version of the momentum equation, may be
invoked using a constitutive relationship for uniform flow. For example, Manning’s
equation can be expressed as
3/5
 QnP 2 / 3 
A= 1/ 2 
(5.33)
 1.49 So 
where n = Manning’s roughness coefficient

P = wetted perimeter (ft, m)
So = longitudinal slope
Comparing this form to Equation 5.32, α = (nP2/3/1.49So1/2)3/5 and β = 3/5.
Figure 5.17
Idealized drainage
basin for kinematic
wave model (U.S. qo
Army Corps of 1
Engineers, 1998) Lo Lo
1 2
qo
2
Flows from
Overland Flow
Elements
qo
1
Lo
2
qo
Lm
Watershed Boundary
Section 5.6 Base Flow 165
Substitution of Equation 5.32 into Equation 5.31 yields a single partial differential
equation in Q:
∂Q ∂Q
αβ Q β −1 + = qL (5.34)
∂t ∂x
The solution of this equation, which expresses Q as a function of location and time
within a flow path, may be approximated using a variety of numerical methods. The
interested reader is referred to Chow, Maidment, and Mays (1988) for an introductory
exposition, and to other references cited therein for more detail.
5.6 BASE FLOW

The total discharge in a stream channel is composed of surface runoff (direct runoff)
and subsurface flow (base flow) from groundwater, as shown in the hydrograph in
Figure 5.18. Hydrologists may also include interflow in the total flow of a stream,
where interflow is defined as the runoff that occurs within near-surface soils. Inter-
flow passes through the near-surface soils much faster than it does through deeper
soils, and thus reaches streams more quickly than does base flow. In this discussion, it
is assumed that interflow is lumped into surface runoff and therefore does not need to
be treated separately.
Base flow may or may not exist in a stream channel, or it may exist only during cer-
tain times of the year. A stream in which base flow is always present is called a peren-
nial stream, and one in which base flow either does not exist, or in which it exists only
some of the time, is called an ephemeral or intermittent stream. Perennial streams are
the norm in regions with an abundance of rainfall and where the elevation of the
groundwater table is above stream channel elevations. Ephemeral streams are com-
mon in arid and semi-arid regions. Ephemeral streams may also exist in humid
regions, particularly in locations where the contributing drainage basin is small and/or
the groundwater table lies below the bottom of the stream channel.
When base flow is present in a stream, the total discharge for which a water convey-
ance system is designed should include the base flow as well as the surface runoff
(direct runoff). For estimation of base flow to support the design of a stormwater con-
veyance system, the best course of action may be to periodically monitor and measure
the flow in the stream of interest using a current meter. Base flow should be measured
during dry-weather periods when enough time has elapsed following storm events for
the discharge to become nearly constant with respect to time.
Figure 5.18
Hydrograph for a
stream showing
contribution of base
flow
Base flow may change seasonally due to fluctuations in the groundwater table. Urban-
ization of an area may cause base flow to decrease, or it may initiate base flow in a
stream where none existed previously. The latter is especially true in semi-arid
regions, where the wastewater discharges, leaking water mains, irrigation, car wash-
ing, and so forth that accompany urbanization may cause local groundwater levels to
rise or perched groundwater conditions to develop.
Once a base flow discharge has been estimated, that discharge may be simply added
to each ordinate of a direct runoff hydrograph to determine the total runoff
hydrograph, as in Example 5.14 on page 172.
5.7 SNOWMELT
In many parts of the world, snow is the dominant form of precipitation. Water pro-
duced by melting snow supplies reservoirs, lakes, and rivers, and infiltrating meltwa-
ter recharges soil moisture and groundwater (Gray and Prowse, 1993). Because
snowmelt is an additional mechanism by which urban runoff may be generated, it
affects facilities such as storm sewers, drainage channels, culverts, and detention
ponds.
Although runoff flow rates from snowmelt are typically low, they may be sustained
over several days. Rainfall events superimposed on snowmelt baseflow will produce
higher runoff peak flows and volumes, and rainfall increases the melt rate of snow
(Huber and Dickinson, 1988). Simulation of snowmelt effects is typically more criti-
cal for facilities significantly impacted by an increase in runoff volume, such as
stormwater pump stations and detention facilities. Snowmelt can also have significant
effects on water quality in urban areas because stored pollutants such as
Section 5.7 Snowmelt 167
hydrocarbons, oil and grease, chlorides, sediment, and nutrients are flushed into
surface waters (Oberts, 1994).
Most of the techniques for modeling runoff generated by melting snow were devel-
oped for mountainous, rural, and agricultural watersheds. These same models have
recently been applied to urban watersheds. This section describes, in general terms,
the models for forecasting runoff from snowmelt in urban environments.
Runoff Potential
The physical properties of interest to modelers are depth, density and water equivalent
(Gray and Prowse, 1993). The potential quantity of runoff in snowpack is the snow-
water equivalent. It is the equivalent depth of water in snow cover, and is calculated as
SWE = dsρs/ρw (5.35)
where SWE = snow water equivalent (in, mm)

ds = depth of snow (in, mm)
ρs = density of snow (lb/ft3, kg/m3)
ρw = density of water (62.4 lb/ft3, 1,000 kg/m3)
Note that density is expressed in force units when U.S. units are used and in mass
units when SI units are used. The density of freshly fallen snow varies widely depend-
ing on the amount of air contained within the lattice of the snow crystals, but is com-
monly in the range of 3 to 7.5 lb/ft3 (50 to 120 kg/m3) (Gray and Prowse, 1993).
Following deposition, the density of snow increases due to metamorphism, which is a
change in the size, shape, and bonding of snow crystals due to temperature and water-
vapor gradients, settlement, and wind packing. Typical densities for settled snow are
12 to 19 lb/ft3 (200 to 300 kg/m3), but densities may be as high as 22 lb/ft3 (350
kg/m3) for hard wind slab (Gray and Prowse, 1993).
Snowmelt Models
The temperature-index or degree-day method is described in numerous references,
including the NEH-4 (Mockus, 1973), Huber and Dickinson (1988), and Semadeni-
Davies (2000). The meltwater rate, M, is a function of daily average air temperature:
M = 0, if Ta < Tm (5.36)
M = C(Ta – Tm), if Ta ≥ Tm (5.37)
where M = meltwater rate, (in/day, mm/day)

C = melt rate factor (in/°F/day, mm/°C/day)
Ta = daily average air temperature (°F, °C)
Tm = threshold temperature, (32°F, 0°C)
Values of C for rural areas range from 0.72 to 3.6 in/°F/day (1.4 to 6.9 mm/°C/day)
and from 1.7 to 4.1 in/°F/day (3 to 8 mm/°C/day) for urban areas (Huber and Dicken-
son, 1988). Gray and Prowse (1993) presented an empirical equation for C:
C = Kρs (5.38)
where C = melt rate factor, mm/°C/day

K = empirical coefficient equal to 0.00385 for U.S. Customary units or 0.011
for SI units
ρs = snow density (lb/ft3, kg/m3)
Although the degree-day method is convenient because it combines several complex
processes into a single constant C, it can yield misleading results. An equation such as
Equation 5.38 should only be used with great care. Ideally, developing the melt-rate
factor should be accomplished by calibration.
Equation 5.37 calculates the meltwater rate on a daily basis. The air-temperature data
typically available are the minimum and maximum values for each day, and Ta is
commonly taken as the average of these values. Incorporation of meltwater into con-
tinuous-simulation hydrologic models requires temperature data for more frequent
time steps. Procedures for developing hourly meltwater rates from this type of data
are presented by Huber and Dickenson (1988).
More detailed models for forecasting snowmelt runoff are available. Most of these are
based on the energy-balance method described by Anderson (1973). The units for
each energy budget term are energy/area-time (for instance, langleys/day, where 1
langley = 1 cal/cm2). The energy balance is presented in Equation 5.39 (Huber and
Dickenson, 1988).
∆H = Hrs + Hg + Hrl + Hc + He + Hp (5.39)
where ∆H = change in heat storage in snowpack

Hrs = net shortwave radiation entering snowpack
Hg = conduction of heat to snowpack from underlying ground
Hrl = net (incoming minus outgoing) longwave radiation entering snowpack
Hc = conductive exchange of sensible heat between air and snowpack
He = release of latent heat of vaporization by condensation of atmospheric
water vapor
Hp = advection of heat to snowpack by rain
Equation 5.39 assumes that the cold content of the snow pack is zero; that is, none of
the energy available to the snow pack is used to bring the snow pack up to the temper-
ature at which any additional energy will result in melt. Procedures for estimating
each term in Equation 5.39 are described by Huber and Dickenson (1988) and Gray
and Prowse (1993).
The meltwater rate may then be determined by Equation 5.40:
Section 5.8 Channel Routing 169
(10, 000)∆H
M = (5.40)
Lρw
where M = meltwater rate (mm/day)

∆H = change in heat storage in snowpack (langley/day)
L = latent heat of fusion (80 cal/g water)
Eighty calories are required to melt 1 g of water, or 80 langley/cm (8 langley/mm) of
snow depth. The rate of snowmelt is calculated by multiplying the meltwater rate
times the area of the watershed. Production of meltwater is not necessarily equivalent
to runoff, however, because snowpack is porous and free water moves downward
through the snowpack before runoff is generated. Huber and Dickinson (1988) use a
simple reservoir-type routing procedure to simulate this process. The free water hold-
ing capacity of snowpack is modeled as a fraction of the snow depth, usually between
0.02 and 0.05. In a continuous simulation, this depth is filled with meltwater before
runoff occurs.
Snow tends to insulate the soil beneath it. If the ground is frozen before snow falls, it
will usually remain frozen. Unfrozen soil tends to remain unfrozen underneath a
snowpack. Therefore, the soil properties related to runoff (infiltration, detention stor-
age, and so forth) are assumed to remain unchanged by snow.
The energy-balance method offers the advantage of accounting for factors like solar
radiation, wind speed, and albedo (the fraction of sunlight reflected from the surface
of the snowpack) that impact the meltwater rates. It can also account for melting that
is induced by rain. However, the application of this method to urban environments has
been limited. Part of the difficulty is that the energy balance method has extensive
data requirements. Also, urban snow data are sparse, which limits the possibility of
assessing or improving model performance.
Huber and Dickinson (1988) showed that the energy balance equation can be reduced
to the degree-day equation by making appropriate substitutions and assumptions.
Anderson (1976) reported that, under many conditions, results obtained using the
energy-balance model were not significantly better than those obtained using the
degree-day method. However, Matheussen and Thorolfsson (2001) stated that the
energy fluxes related to snowmelt are significantly altered in the urban environment,
and degree-day models are not suitable for urban areas. The HEC-1 model (U.S.
Army Corps of Engineers, 1998) allows the modeler to use either the degree-day or
energy-balance model. The energy-budget routine can be applied to heavily forested
areas.
5.8 CHANNEL ROUTING

Channel routing is a type of unsteady flow analysis necessary when rainfall occurs in
a subbasin some distance upstream of a point of interest, or when the runoff hydro-
graphs from multiple subbasins must be combined and routed downstream to a point
of interest. Channel routing is usually applied in cases where runoff hydrographs, as
opposed to just peak flows, are being used. In storm sewer designs that use the ratio-
nal method for peak discharge prediction, routing is not performed. Instead, the time
of concentration of the entire contributing drainage basin area is determined for each
point of interest in the system, and the peak discharge is computed as illustrated in
Example 5.9 on page 143.
Channel routing may be done by using one of many available hydrologic or hydraulic
methods. Hydrologic methods involve the principle of conservation of mass and a
rather simple storage-discharge relationship (Chow, Maidment, and Mays, 1988).
Hydraulic methods involve the simultaneous solution of the Saint-Venant equations
(see Appendix C). The next section introduces the concepts of channel routing by
demonstrating a hydrologic routing method called the Muskingum method (McCarthy,
1938). Additional information on unsteady flow modeling methods can be found in
Appendix C.
Muskingum Routing
This method derives its name from the Muskingum River basin in Ohio, where it was
first developed and applied. As described by Linsley, Kohler, and Paulhus (1982), an
expression for the volume S of water stored within a channel reach may be expressed
as
b
S=  XI m / n + (1 − X )O m / n  (5.41)
a
where
S = volume stored in channel reach (ft3, m3)
a, n = constants in the stage-discharge relationship for the reach
b, m = constants in the stage-storage relationship for the reach
X = a dimensionless weighting factor
I = inflow discharge into the reach (cfs, m3/s)
O = outflow discharge from the reach (cfs, m3/s)
The Muskingum method assumes that m/n = 1 and assigns K = b/a such that
S = K [ XI + (1 − X )O ] (5.42)
K is approximately equal to the travel time through the channel reach and, in the
absence of additional information, is often estimated that way. For most streams and
rivers, the constant X is in the range of about 0.1 to 0.4. Small values of X tend to be
associated with rivers having wide floodplains and therefore large amounts of storage.
Large values of X tend to be associated with narrow canyons where little storage is
available as the water level rises.
Section 5.8 Channel Routing 171
If conservation of mass is approximated over a time interval ∆t = t2 – t1, the following

expression results:
I1 + I 2 O + O2
S 2 − S1 = ∆t − 1 ∆t (5.43)
2 2
Substituting Equation 5.42 for each of the S terms in this expression and collecting
like terms yields
O2 = C0 I 2 + C1 I1 + C2 O1 (5.44)
In Equation 5.44, the coefficients C0, C1, and C2 are defined by Equations 5.45, 5.46,
and 5.47, respectively. Note that C0 + C1 + C2 = 1.
0.5∆t − KX
C0 = (5.45)
K (1 − X ) + 0.5∆t
KX + 0.5∆t
C1 = (5.46)
K (1 − X ) + 0.5∆t
K (1 − X ) − 0.5∆t
C2 = (5.47)
K (1 − X ) + 0.5∆t
K must be in the same time units as ∆t. With K, X, and ∆t established, these expres-
sions may be used to estimate C0, C1, and C2. Equation 5.44 can then be applied to
predict the outflow rate from the channel reach at time t2 on the basis of the current
inflow and the previous inflow and outflow rates.
The number of routing reaches to which the Muskingum method is applied is desig-
nated as NR. To ensure computational stability and accuracy of the computed outflow
hydrograph, the U.S. Army Corps of Engineers (1960) recommends choosing ∆t or
NR such that
1 K 1
≤ ≤ (5.48)
2(1 − X ) N R ∆t 2 X
If ∆t is fixed by the time step of an available inflow hydrograph, a value for NR can be
chosen to ensure that Equation 5.48 is satisfied. If, for example, NR = 2 satisfies Equa-
tion 5.48, the stream reach would be treated as a series of two reaches, each of which
would be routed using the preceding expressions. The inflow hydrograph to the sec-
ond (downstream) reach would be taken as the outflow hydrograph from the first
(upstream) reach, and so on. As an alternative to fixing the time step based on the
inflow hydrograph as just described, points on the inflow hydrograph can be interpo-
lated for a time step other than that for which it was initially tabulated.
Example 5.14 – Routing a Channel Using the Muskingum Method. Route a run-
off hydrograph through a channel reach for which X = 0.2 and K = 20 min. Use the hydrograph com-
puted in Example 5.13 (page 161), but add 25 cfs to each hydrograph ordinate to account for base
flow in the stream. Assume that flow in the channel is steady prior to the onset of runoff.
Note that ∆t = 10 min is the time step at which the inflow hydrograph is tabulated. Using Equation
5.48, the ratio K/NR∆t must be between 0.28 to 2.50. For K = 20 min and ∆t = 10 min, the relationship
is satisfied for NR = 1. Thus, the channel reach can be treated as a single reach, and the inflow
hydrograph will not need to be interpolated for a different time step.
The values of the coefficients C0, C1, and C2 are estimated to be
C0 = [0.5(10) – 20(0.2)]/[20(1 – 0.2) + 0.5(10)] = 0.048
C1 = [20(0.2) + 0.5(10)]/[20(1 – 0.2) + 0.5(10)] = 0.429
C2 = [20(1 - 0.2) – 0.5(10)]/[20(1 – 0.2) + 0.5(10)] = 0.523
A check confirms that C0 + C1 + C2 = 1.
Table 5.19 is set up for performing the hydrograph routing. The first column is the time, in minutes,
since the beginning of rainfall. The second column is the inflow hydrograph to the reach, in cfs, and
has been obtained by adding 25 cfs to each direct runoff hydrograph ordinate computed in Example
5.13. The first several entries in the third column (the outflow hydrograph column), up until a time of
20 min, are equal to 25 cfs because of the steady-state assumption.
At time t = 30 min, the discharge is computed as
Q2 = 0.048I2 + 0.429I1 + 0.523Q1 = 0.048(83) + 0.429(25) + 0.523(25) = 28 cfs
At time t = 40 min, the discharge is computed as
Q2 = 0.048(228) + 0.429(83) + 0.523(28) = 61 cfs
Proceeding in this fashion, the table is completed. The inflow and outflow hydrographs are plotted in
Figure 5.14.1.
Table 5.19 Muskingum channel routing
t (min) I (cfs) Q (cfs) t (min) I (cfs) Q (cfs)

0 25 25 170 222 389
10 25 25 180 175 307
20 25 25 190 140 242
30 83 28 200 110 192
40 228 61 210 90 152
50 479 153 220 74 122
60 817 325 230 61 98
70 1098 573 240 55 80
80 1256 831 250 44 68
90 1269 1034 260 40 56
100 1177 1142 270 38 49
110 1002 1150 280 37 44
120 794 1070 290 31 40
130 610 930 300 27 35
140 461 770 310 26 32
150 361 618 320 25 29
160 283 492
Section 5.9 Reservoir Routing 173
Figure E5.14.1 Inflow and outflow hydrographs for the channel in Example 5.14
When inflow and outflow hydrograph data are available for a channel reach, they may
be used to estimate the parameters K and X for that reach. Either the graphical
approach discussed by Chow, Maidment, and Mays (1988) may be applied for this, or
a constrained least-squares method using a Lagrange multiplier can be applied.
5.9 RESERVOIR ROUTING

Reservoir routing is used when one needs to assess the effects of a natural pond or
lake, or of a constructed facility such as a detention pond, on the characteristics of a
hydrograph. As in channel routing, either hydrologic or hydraulic methods may be
applied in reservoir routing. Level-pool hydrologic methods are most commonly
used.
In the design and analysis of stormwater conveyance systems, reservoir routing is
usually performed for facilities such as detention ponds. Therefore, descriptions of
reservoir routing methods are presented with detention pond design in Chapter 12.
5.10 SUMMARY OF METHODS FOR COMPUTING

FLOWS
Table 5.20 summarizes the basic characteristics of a number of methods for comput-
ing peak flows, developing runoff hydrographs, and routing hydrographs. The meth-
ods presented in this chapter are included, as well as a sampling of other popular
models.
Table 5.20 Descriptions of Selected Methods for Computing Peak Flows
174
Type Model Description Advantages Disadvantages

Peak Flow Rational This method estimates the peak flow from rainfall events of a specific Peak flow models are easy to formulate The lumped system approach of peak flow
These methods are generally statisti- duration and frequency occurring over a given catchment area. Infil- because there are an abundance of data for models does not allow the analysis of spatial
cally based or rely on an empirical tration losses, surface detention, and antecedent conditions are determining the values of the parameters. With variation in rainfall and catchment properties.
formula to estimate the peak runoff accounted for by a runoff coefficient applied to the product of the the rational method, multiple storms can be The rational method cannot be used on large
from a catchment. Coefficients and rainfall intensity and the catchment area. analyzed quickly and easily to determine the catchments or those which experience large
parameters used in the formulas are worse case scenario. Partial area effects can storage effects. Its runoff coefficient C is a
lumped to express catchment proper- SCS The depth of effective precipitation during a storm is generated using also be included to improve the quality of the lumped parameter taking into account infiltra-
ties such as surface detention and the SCS Curve Number method and the peak discharge is then esti- analysis. The SCS model is able to account for tion, surface detention, antecedent conditions,
infiltration. mated as the product of this effective precipitation, the unit peak flow storage within the catchment by using a pond etc.These factors cannot be effectively quanti-
rate, and the drainage basin area. The unit peak flow rate is based on and swamp adjustment factor. fied by a single parameter, and variation in the
regression analysis of a large number of computer runs of SCS Curve
Modeling Runoff
value of the runoff coefficient has been shown

Number method (TR-20). Consideration of storage effects in the to occur with more extreme events (Cordery
catchment is also achieved through the introduction of a pond and and Pilgrim, 1983). The SCS method was
swamp adjustment factor. developed for use with 24-hour storm rainfall
depths and its use for storm durations other
than this is not advised.
Unit Hydrograph Clark This method (Clark, 1945) uses the time-area approach to develop a The application of the unit hydrograph to the Unit hydrographs are not capable of estimating
Unit hydrograph methods are based unit hydrograph for the catchment. The unit hydrograph is convolved rainfall hyetograph to estimate the runoff runoff hydrographs on catchments that experi-
on the relationship of a pulse of rain- with the rainfall hyetograph to produce a runoff hydrograph, which is hydrograph is a simple process and takes ence a high level of nonlinearity caused by stor-
fall (1 mm or 1 in.) applied to a then routed through a linear storage concentrated at the catchment account of all the catchment characteristics that ages. Because they deal with lumped systems,
catchment and the resulting outlet so as to reproduce the effect of hydrograph attenuation and dif- influence the runoff hydrograph, such as infil- it is not possible to analyze the effect of spatial
response. With this relationship, the fusion. tration, abstraction, storage, etc. With the Clark rainfall. With the Snyder unit hydrograph, lag
actual rainfall hyetograph is con- unit hydrograph, the characteristics of the times may actually vary with the magnitude of
volved to produce a runoff Snyder Snyder (1938) developed this synthetic unit hydrograph from a study catchment (shape, hydraulic length, etc.) are the event and result in an underestimation of
hydrograph. There are two forms of in the Appalachian Highlands. Catchment parameters, such as length captured in the time-area histogram and there- the peak of the runoff hydrograph.
unit hydrograph: measured and syn- along the main stream and length along the main stream from the out- fore reflected in the unit hydrograph. Also with
thetic. A measured hydrograph is the let to the centroid, are used in an expression for the basin lag time. the Clark unit hydrograph, the use of the con-
result of deconvolving an observed Once this lag has been determined, other parameters affecting the centrated storage at the outlet addresses storage
runoff hydrograph and the hyeto- shape of the unit hydrograph can be estimated. The U.S. Army Corps effects experienced in the catchment.
graph. A synthetic hydrograph is of Engineers has developed empirical equations to further define the
developed from the catchment prop- shape of the unit hydrograph.
erties.
Time Area TRRL This approach is similar to the Clark method except that there is a It is possible to analyze spatial variations in the Storage effects within the catchment cannot be
A series of isochrones are used to nonlinear concentrated storage at the outlet instead of a linear stor- rainfall. The characteristics of the catchment accounted for because this is a translation
divide the catchment and produce a age. The time-area method is used to generate a hydrograph which is (shape, hydraulic length, etc.) are captured in method and therefore the resulting hydrographs
histogram of contributing area for then routed through a concentrated nonlinear storage at the outlet. the time-area histogram and therefore reflected lack diffusion and attenuation.
each time interval. The partial flow The model is also capable of performing simple routing of the runoff in the runoff hydrograph. An advantage of the
from each of these time-based areas hydrograph through pipe networks. TRRL method over the Clark method is that
is generated by multiplying the area nonlinearity can also be modeled. Also with
by the effective rainfall intensity for Modified RRL Manning equation is used to develop a storage relationship for the TRRL, diffusion and attenuation effects are
each time interval. Ordinates of the main drainage pipe in the catchment. The hydrograph generated by addressed by including the concentrated stor-
hydrograph at the outlet of the catch- the time-area method is then routed through this storage. age at the catchment outlet
ment are then developed by lagging ILLUDAS-SA, These models are derived from the TRRL implementation. The main
and summing these partial flows. ILLUDAS, and difference is that the concentrated storage at the outlet is neglected,
ILSAX resulting in only translation of the hydrograph.
Hydraulic Routing Kinematic Wave This model consists of various plane elements and channel sections Since this is a deterministic method, it may be Although the model has a sound mathematical
These methods generally involve the arranged and connected in a manner that depicts the subcatchment used on a catchment where no gauged data is basis and physical attribute of the catchment
representation of the overland and properties and channel network within the catchment. A lumped or available. are used to parameterize the equations, the
Chapter 5
stream flow within the catchment by distributed approach can be applied depending on whether the param- assumptions of one-dimensional sheet flow on
the unsteady, one-dimensional, grad- eters are kept constant or allowed to vary in space. the plane surfaces and uniform flow in channel
ually varied open-channel flow equa- are approximations to reality.
tions (continuity and momentum).
Type Model Description Advantages Disadvantages
The full form of the equations is EPA Runoff This method is similar to a kinematic wave. The main difference
known as the Saint-Venant equa- being in the formulation of the momentum equation, which is in a
tions. The kinematic wave equation, form that is very similar to Manning's equation.
which uses an approximation of the
full Saint-Venant momentum equa-
tion, is more commonly used for run-
off routing.
Storage Routing RSWM Regional Stormwater Model (Goyen & Aitken, 1979; Black and Cod- The arrangement of storage models within the The power function relationship that provides
Section 5.10
Storage routing models generally fall ner, 1979) is a distributed storage model where the catchment is catchment allows the physical characteristics of the nonlinear response is not always applicable
into two broad categories, distributed divided into 10 subareas using isochrones at equal time intervals. A the catchment to be modeled with accuracy. In to some catchments. Errors in establishment
and lumped storage models, and can concentrated nonlinear storage is placed at the outlet of each of these this way, the effects of urbanization can easily and calibration of nonlinear models can have a
be either linear or nonlinear. Distrib- subareas, and flow is routed through these storages in series to pro- be incorporated into the design. The ability to more pronounced effect on the results and can
uted models are capable of account- duce a cascading storage routing scheme. include various loss models further improves lead to underestimates for large events where
ing for spatial variation in rainfall the applicability of these models to the design data is not available.
and catchment properties, while non- RORB This model was developed by Laurenson and Mein (1983). A distrib- of stormwater facilities. The ability of storage With the Nash model, there is a tendency for
linear models better reflect runoff uted nonlinear network model is developed by dividing the catchment routing models to account for nonlinearity in the model to average disparities within the
variation that occur at different inten- into subcatchments, which are usually based on the major tributary storage is an added advantage for the analysis catchment, since the collective runoff from the
sities. network. Nodes are placed at the subcatchment centroids, the stream of urban catchments, which often contain stor- entire catchment is routed through the same
confluences, downstream of storages, etc. Catchment storage effects age facilities. storage. Also with the Nash model, the number
are represented by nonlinear concentrated storage placed midway With the WBNM model, as the storage charac- of storages and the storage coefficient has been
between the nodes with special concentrated storages used to repre- teristics and storage delay times of two differ- seen to vary from event to event, and it is diffi-
sent detention basins and reservoirs. ent types of storages are different, the provision cult to reproduce the response of sharp peaks in
WBNM The watershed bounded network model (WBNM) was developed by of these storages is physically realistic. observed hydrographs.
Boyd et al. (1979a, 1979b) and is similar to RORB, although it is The SBUH method directly computes a runoff The main problem with the Tank Model is that
based on more detailed consideration of geomorphological relations. hydrograph without going through an interme- it requires a considerable amount of expertise
The main difference is that WBRM has two different types of stor- diate process such as developing a unit to calibrate because the model structure and
ages for types of subcatchments: ordered basins and interbasin areas. hydrograph. parameters must be determined subjectively.
Ordered basins are complete subcatchments where no water flows With the Tank Model, accurate results can be
into the area across the watershed boundary and their storage repre- obtained as all the physical characteristics of
sents the transformation of rainfall excess into surface runoff at the the hydrological cycle are taken into consider-
downstream end of the catchment. Interbasin areas are subcatchments ation
with a stream draining upstream areas that flow through them. Inter-
basins transform rainfall into runoff and contain transmission stor-
ages that route upstream runoff through the stream in the
subcatchment.
Nash This method is based on the average rainfall excess over the catch-
ment being routed through a series of cascading linear storages con-
centrated at the catchment outlet.
SBUH The SBUH method is based on the curve number (CN) approach and
uses SCS equations for computing soil absorption and precipitation
excess. This method converts the incremental runoff depths into
instantaneous hydrographs that are then routed through a conceptual
reservoir with a time delay equal to the basin time of concentration.
Tank Model The tank model was developed by Sugawara and further refined as
reported in Sugawara et al. (1984). The main concept behind this
model is to represent the various components in the hydrologic cycle
by linear or nonlinear storages (“tanks”). Each tank has one or more
outlets located on the bottom or sides. The various parameters for
each tank are estimated by considering catchment conditions such as
evaporation, infiltration, and soil moisture, while calibrating to
observed data.
Summary of Methods for Computing Flows 175
5.11 CHAPTER SUMMARY

Many methods and levels of analysis are available to quantify runoff from a water-
shed area. Several of these methods are described within this chapter.
The portion of the rainfall occurring over the area of interest that does not contribute
to runoff is the rainfall abstraction. The total abstraction consists of several compo-
nents: interception, depression storage, infiltration, and evapotranspiration. The first
three components are typically accounted for in hydrologic analysis for stormwater
management facilities. Evapotranspiration can usually be neglected. Two possible
methods for modeling infiltration are Horton and Green-Ampt.
The portion of the total rainfall that contributes to runoff is called the effective precip-
itation. Effective precipitation, regardless of the method used for its determination, is
used as the basis for practical estimation of the runoff volume, for peak runoff flow
rate calculations, and for computing runoff flow rates over time (that is, a runoff
hydrograph). Computation of the peak flow rate or a runoff hydrograph requires an
estimate of the basin’s time of concentration or lag time.
Methods used to compute the volume of effective precipitation include the runoff
coefficient and the curve number method. The most accurate way to model effective
precipitation is to create a runoff hyetograph by subtracting interception, infiltration,
and depression storage from a rainfall hyetograph.
To compute the peak flow rate resulting from a rainfall event, the rational method,
curve number method, or regression formula developed for a particular area and
application may be used. For most approaches, it is necessary to compute the
response time of the drainage basin (either the time of concentration or basin lag) to
compute runoff flow rates.
A number of techniques for developing a runoff hydrograph exist; this text focuses on
the unit hydrograph approach. With this method, an effective rainfall hyetograph is
merged with the unit hydrograph for the basin through a process called discrete con-
volution. If applicable, base flow can be added to the resulting hydrograph. A more
physically-based approach to hydrograph development is the kinematic wave model.
Runoff from snowmelt is an important consideration in designing stormwater facili-
ties in many regions of the world. In determining the runoff potential for snowmelt, it
is necessary to calculate the snow water equivalent, which is a function of the snow
depth and density. The meltwater rate for snow is based on a melt rate factor and tem-
perature.
Channel routing and reservoir routing are processes through which the effects of stor-
age in a conveyance channel or basin within the watershed are accounted for. Reser-
voir routing is presented in more detail in Chapter 12.
References 177
REFERENCES
American Society of Civil Engineers (ASCE) 1992. Design and Construction of Urban Storm Water Man-
agement Systems. New York: American Society of Civil Engineers.
Anderson E. A. 1973. “National Weather Service River Forecast System – Snow Accumulation and Abla-
tion Model.” Report NWS 17. Washington, D.C.: U.S. Department of Commerce.
Anderson E. A. 1976. “A Point Energy and Mass Balance Model of Snow Cover.” Report NWS 19. Wash-
ington, D.C.: U.S. Department of Commerce.
Bedient, P. B., and W. C. Huber. 1992. Hydrology and Floodplain Analysis. 2nd ed. Reading, Mass.: Addi-
son-Wesley.
Black, D. C. and G. P. Codner. 1979. “Investigation of urban drainage strategy using simulation.” Pp 199–
204 in Hydrology and Water Resources Symposium 1979. Institute of Engineers Australia, National
Conference Publication No. 79/10.
Boyd, M. J., B. C. Bates, D. H. Pilgrim, and I. Cordery. 1979a. “A storage routing model based on catch-
ment geomorphology.” Journal of Hydrology 42: 209-230.
Boyd, M. J., B. C. Bates, D. H. Pilgrim, and I. Cordery. 1979b. “An improved routing model based on geo-
morphology.” Pp 189-193 in Hydrology and Water Resources Symposium 1979, Institute of Engineers
Australia, National Conference Publication No. 79/10.
Boyd, M. J., B. C. Bates, D. H. Pilgrim, and I. Cordery. 1987. “WBNM: a general runoff routing model.”
Pp 137-141 in 4th National Local Government Eng. Conference, Inst. Engrs Australia, National Confer-
ence Publication No. 87/9.
Brown, S. A., S. M. Stein, and J. C. Warner. 2001. Urban Drainage Design Manual. Hydraulic Engineering
Circular No. 22, FHWA-SA-96-078. Washington, D.C.: U.S. Department of Transportation.
Chin, D. A. 2000. Water-Resources Engineering. Upper Saddle River, New Jersey: Prentice Hall.
Chow, V. T., D. R. Maidment, and L. W. Mays. 1988. Applied Hydrology, New York: McGraw-Hill.
Clark, C. O. 1945. “Storage and the unit hydrograph.” Transactions, American Society of Civil Engineering
110: 1419-1446.
Cordery, I. and D. H. Pilgrim. 1983. “On the lack of dependences of losses from flood runoff on soil and
cover characteristics.” Pp. 187-195 in IAHR Publ. No. 140.
Federal Aviation Administration (FAA). 1970. Advisory Circular on Airport Drainage. Report A/C 150-
5320-58. Washington, D.C.: U.S. Department of Transportation.
Goyen, A. J. and A. P. Aitken. 1979. “A stormwater drainage model.” Pp 40-44 in Hydrology Symposium
1976. Institute of Engineers Australia, National Conference Publication No. 76/2.
Gray, D. M. and T. D. Prowse. 1993. “Snow and Floating Ice.” In Handbook of Hydrology, D.R. Maidment,
ed. New York: McGraw-Hill, Inc.
Green, W. H., and G. Ampt. 1911. “Studies of Soil Physics, Part I: The Flow of Air and Water Through
Soils.” Journal of Agricultural Science 4: 1–24.
Hathaway, G. A. 1945. “Design of Drainage Facilities.” Transactions, American Society of Civil Engineers
110: 697–730.
Helvey, J. D. 1967. “Interception by Eastern White Pine.” Water Resources Research 3: 723–730.
Helvey, J. D, and J. H. Patric. 1965. “Canopy and Litter Interception of Rainfall by Hardwoods of the East-
ern United States.” Water Resources Research 1: 193–206.
Henderson, F. M., and R. A. Wooding. 1964. “Overland Flow and Groundwater Flow from a Steady Rain of
Finite Duration.” Journal of Geophysical Research 69, no. 8: 1531–1540.
Horton, R. 1939. “Analysis of Runoff Plot Experiments with Varying Infiltration Capacity.” Transactions,
American Geophysical Union 20: 693–711.
Horton, R. 1919. “Rainfall Interception.” Monthly Weather Review 47: 602–623.

Huber, W. C. and Dickinson, R.E. 1988. Storm Water Management Model User’s Manual, Version 4. EPA-
600/3-88-001a (NTIS PB88-236641/AS). Athens, Ga.: U.S. EPA.
Izzard, C. F. 1946. “Hydraulics of Runoff From Developed Surfaces.” Proceedings, Highway Research
Board 26: 129–150.
Jennings, M. E., W.O. Thomas, Jr., and H.C. Riggs. 1994. Nationwide Summary of U.S. Geological Survey
Regional Regression Equations for Estimating Magnitude and Frequency of Floods for Ungaged Sites.
Water Resources Investigations Report 94-4002. U.S. Geological Survey.
Johnstone, D., and W.P. Cross. 1949. Elements of Applied Hydrology. New York: Ronald Press.
Kent, K. 1972. “Travel Time, Time of Concentration, and Lag.” National Engineering Handbook, Section
4: Hydrology, Chapter 15. Washington, D.C.: Soil Conservation Service, U.S. Department of Agricul-
ture.
Kerby, W. S. 1959. “Time of Concentration of Overland Flow.” Civil Engineering 60: 174.
Kibler, D. F. 1982. “Desk-Top Runoff Methods.” Pp. 87–135 in Urban Stormwater Hydrology, ed. D.F.
Kibler. Water Resources Monograph Series. Washington, D.C.: American Geophysical Union.
Kirpich, T. P. 1940. “Time of Concentration of Small Agricultural Watersheds.” Civil Engineering 10, no.
6: 362.
Kuichling, E. 1889. “The Relation Between the Rainfall and the Discharge of Sewers in Populous Dis-
tricts,” Transactions, American Society of Civil Engineers, vol. 20, pp. 1-60.
Laurenson, E. M. and R. G. Mein. 1983. RORB version 3: Runoff routing program-users manual, 2d ed.
Department of Civil Engineering, Monash University.
Linsley, R. K. 1943. “Application of the Synthetic Unit Graph in the Western Mountain States.” Transac-
tions American Geophysical Union 24:581–587.
Linsley, R. K., Jr., M. A. Kohler, and J. L. H. Paulhus. 1982. Hydrology for Engineers. 3d ed. New York:
McGraw-Hill.
Matheussen, B. R. and Thorolfsson, S. T. 2001. Urban Snow Surveys in Risvollan, Norway, Urban Drain-
age Modeling. Proceedings of the Specialty Symposium of the World and Water Environmental
Resources Conference. Alexandria, VA: American Society of Civil Engineers.
McCarthy, G. T. 1938. “The Unit Hydrograph and Flood Routing.” North Atlantic Division Conference.
U.S. Army Corps of Engineers.
McCuen, R. H., S. L. Wong, and W. J. Rawls. 1984. “Estimating Urban Time of Concentration.” Journal of
Hydraulic Engineering 110, no. 7: 887–904.
Merriam, R. A. 1960. “A Note on the Interception Loss Equation.” Journal of Geophysical Research 65:
3850–3851.
Merriam, R. A. 1961. “Surface Water Storage on Annual Ryegrass.” Journal of Geophysical Research 66:
1833–1838.
Mockus, V. 1969. “Hydrologic Soil-Cover Complexes.” National Engineering Handbook, Section 4:
Hydrology, Chapter 9. Washington, D.C.: Soil Conservation Service, U.S. Department of Agriculture.
Mockus, V. 1972. “Estimation of Direct Runoff from Storm Rainfall.” National Engineering Handbook,
Section 4: Hydrology, Chapter 10. Washington, D.C.: Soil Conservation Service, U.S. Department of
Agriculture.
Mockus, V. 1973. “Estimation of Direct Runoff from Snowmelt.” National Engineering Handbook, Section
4: Hydrology, Chapter 11. Washington, D.C.: Soil Conservation Service, U.S. Department of Agricul-
ture.
Morel-Seytoux, H. J., and J. P. Verdin. 1981. Extension of Soil Conservation Service Rainfall Runoff Meth-
odology for Ungaged Watersheds. Report FHWA/Rd-81/060. Washington, D.C.: U.S. Department of
Transportation, Federal Highway Administration.
References 179
Natural Resources Conservation Service (NRCS). 1993. “Storm Rainfall Depth.” National Engineering
Handbook, Part 630: Hydrology, Chapter 4. Washington, D.C.: U.S. Department of Agriculture.
Oberts, G. 1994. “Influence of snowmelt dynamics on stormwater runoff quality.” Watershed Protection
Techniques 1 no. 2: 55–61.
Olin, D. A. 1984. Magnitude and Frequency of Floods in Alabama. Water Resources Investigations Report
84-4191. U.S. Geological Survey.
Overton, D. E., and M. E. Meadows 1976. Stormwater Modeling. New York: Academic Press.
Patric, J. H. 1966. “Rainfall Interception by Mature coniferous Forests of Southeast Alaska.” Journal of
Soil and Water Conservation 21: 229–231.
Rawls, W., P. Yates, and L. Asmusse. 1976. Calibration of Selected Infiltration Equations for the Georgia
Coastal Plain. Technical Report ARS-S-113. Washington, D.C.: U.S. Dept. of Agriculture, Agricultural
Research Service.
Rawls, W. J., L. R. Ahuja, D. L. Brakensiek, and A. Shirmohammadi. 1993. “Chapter 5: Infiltration and
Soil Water Movement” in D.R. Maidment, ed. Handbook of Hydrology. New York: McGraw Hill.
Rodriguez-Iturbe, I., and J. B. Valdes. 1979. “The Geomorphologic Structure of Hydrologic Response.”
Sauer, V. B., W.O. Thomas Jr., V. A. Stricker, and K. V. Wilson. 1983. Flood Characteristics of Urban
Watersheds in the United States. Water Supply Paper 2207. U.S. Geological Survey.
Schaake, J. C., Jr., J. C. Geyer, and J. W. Knapp. 1967. “Experimental Examination of the Rational
Method.” Journal of the Hydraulics Division, American Society of Civil Engineers (ASCE) 93, no.
HY6.
Semadeni-Davies, A. 2000. “Representation of snow in urban drainage models.” J. Hydrologic Engineering
5 no. 4: 363–370.
Sherman, L. K. 1932. “Stream Flow from Rainfall by the Unit Graph Method.” Engineering News-Record
108: 501–505.
Singh, V. P. 1988. Hydrologic Systems, Rainfall-Runoff Modeling, vol. I. Englewood Cliffs, New Jersey:
Prentice Hall.
Singh, V. P. 1992. Elementary Hydrology. Englewood Cliffs, New Jersey: Prentice Hall.
Snider, D. 1972. “Hydrographs.” National Engineering Handbook, Section 4: Hydrology, Chapter 16.
Washington, D.C.: Soil Conservation Service (SCS), U.S. Department of Agriculture.
Snyder, F.F. 1938. “Synthetic Unit Graphs.” Transactions, American Geophysical Union 19: 447.
Soil Conservation Service (SCS). 1969. Section 4: Hydrology in National Engineering Handbook. Wash-
Sugawara, M., I. Watanabe, E. Ozaka, Y. Katsuyama. 1984. Tank model with snow component. Research
Notes of the National Research Center for Disaster Prevention No. 65. Ibaraki-Ken, Japan: Science and
Technology Agency.
Taylor, A. B., and H. E. Schwarz. 1952. “A unit hydrograph lag and peak flow related to basin characteris-
tics.” Trans. Am. Geophys. Union 33: 235.
U.S. Army Corps of Engineers. 1960. Routing of Floods Through River Channels. Engineering and Design
Manuals, EM 1110-2-1408. Washington, D.C.: U.S. Army Corps of Engineers.
U.S. Army Corps of Engineers Hydrologic Engineering Center (HEC). 1998. HEC-1 Flood Hydrograph
Package User’s Manual. Davis, Ca.: HEC.
Viessman, W., J. Knapp, and G. L. Lewis. 1977. Introduction to Hydrology, 2d ed. New York: Harper &
Row.
Williams, G. B. 1922. “Flood Discharges and the Dimensions of Spillways in India.” Engineering (London)
134: 321.
Zinke, P. J. 1967. “Forest Interception Studies in the United States.” Pp. 137–160 in W. E. Sopper and H. W.
Lull, eds. Forest Hydrology. Oxford: Pergamon Press.
PROBLEMS
5.1 Using the rational method, determine what is the composite runoff coefficient from a shopping cen-
ter with the following characteristics:
Description Area (ac) C

Roadways and parking lots 8.5 0.90
Rooftops 3.4 0.85
Landscape areas 1.2 0.20
5.2 Use the data below to plot the incremental rainfall excess over the duration of the storm. What is the
total effective precipitation, in inches?
Rainfall Intensity Infiltration Depression

Time (hr) (in/hr) Rate (in/hr) storage (in/hr)
0 to 1 0.41 0.2 0.2
1 to 2 0.49 0.16 0.14
2 to 3 0.32 0.13 0.04
3 to 4 0.31 0.09 0.02
4 to 5 0.22 0.08 0
5 to 6 0.08 0.07 0
6 to 7 0.06 0.06 0
7 to 8 0.04 0.04 0
8 to 9 0.04 0.03 0
5.3 Use the curve number method to calculate the effective precipitation from a 4-in. rainfall event over
a watershed for each of the three antecedent moisture conditions. The watershed characteristics are
given in the table.
Land Cover Soil Group Area (mi2)

Pasture in good hydrologic condition A 0.72
Residential (1/2 ac lots) B 1.52
Streets (paved with curbs and gutters) B 0.58
5.4 A subdivision has the following land uses: 160 ac of large-lot residential, 68 ac of good condition
open space, and 13 ac of gravel roads. Three percent of the subdivision is covered by swampy areas,
and the entire area overlays hydrologic soil type C. What are the effective rainfall and the peak dis-
charge from a 5-in. rainfall event? The time of concentration is estimated to be 0.6 hr. Use the curve
number method for a Type II rainfall distribution and assume AMC-II applies.
5.5 A stormwater collection system is to be designed for a subdivision consisting of subbasins A, B, C,

D, and E, each draining to an inlet with the same designation. The layout of the subbasins and collec-
tion system is shown. The characteristics of the subbasins, pipe flow times, and IDF data for the 25-
Problems 181
year design storm are given in the tables that follow. (Note that the IDF data is the same as that
developed in Example 4.2.) Assume that every inlet captures 100 percent of approaching flows and
use the rational method and the 25-year-storm IDF data from Example 4.2 to compute the design
peak discharge to each inlet and within each pipe. Complete the results tables at the end of the
problem.
Runoff Inlet tc
Subbasin Area (ac) Coefficient (min)
A 1.5 0.50 7
B 1.7 0.60 8
C 1.1 0.75 5
D 2.7 0.55 10
E 1.1 0.70 5
Flow Time
Pipe (min)
Pipe 1 0.85
Pipe 2 0.55
Pipe 3 0.54
Pipe 4 0.50
Intensity Data for 25-Year Storm (in./hr)

5 min 10 min 15 min 30 min 60 min
8.40 7.02 5.96 4.42 2.97
Peak Flow
Inlet (cfs)
A
B
C
D
E
Peak Flow
Pipe (cfs)
Pipe 1
Pipe 2
Pipe 3
Pipe 4
Outlet Pipe
5.6 Given the following inflow hydrograph for a channel reach, compute the outflow at a section 3 mi
downstream using the Muskingum method if K = 11 hr and X = 0.13.
Time (hr) Inflow (cfs) Outflow (cfs)
0 90 90
6 310
12 690
18 505
24 395
30 310
36 220
42 90
Problems 183
5.7 Construct a synthetic unit hydrograph using the NRCS (SCS) method for a watershed with a drain-
age area of 250 ac. The time of concentration has been computed as 0.5 hr. Assume that Equation
5.26 is valid for this watershed.
Discharge in this
force main is an
example of closed-
conduit flow. The
brick structure is
the pump station.
C H A P T E R
6
Flow in Closed Conduits
Flow of water through a conduit is said to be pipe flow, or closed-conduit flow, if it

occupies the entire cross-sectional area of the conduit. Conversely, if only a portion of
the total cross-sectional area is occupied by the flow such that there is a free water
surface at atmospheric pressure within the conduit, open-channel flow exists. Figure
6.1 shows a cross-sectional view of open-channel flow in a pipe.
Figure 6.1
Open-channel flow in
a pipe
Storm sewers and culverts are most commonly designed and constructed using com-
mercially available pipe, which is manufactured in a number of different materials
and cross-sectional shapes. Reinforced concrete and corrugated metal (steel or alumi-
num) are the most frequently used pipe materials, and circular is the most common
cross-sectional shape. In some cases, notably for large culverts and sewers, concrete
may be cast in place. For cast-in-place culverts, it is usually advantageous to use sim-
ple, rectangular cross-sectional shapes.
186 Flow in Closed Conduits Chapter 6
Although stormwater-related systems are primarily thought of as open-channel con-

veyances, closed-conduit flow occurs in a number of instances. The most obvious of
these are stormwater pumping facilities; however, storm sewers, culverts, and deten-
tion pond outlet structures can also experience closed-conduit flow conditions when
the hydraulic grade is sufficiently high.
In closed-conduit flow, the cross-sectional area of the flow is the same as that of the
inside of the conduit. However, in open-channel flow, the cross-sectional area of the
flow, and hence other flow parameters such as the velocity, depend not only on the
size and shape of the conduit, but also on the depth of flow. As a consequence, open-
channel flow is more difficult to treat from an analytical point of view than is pipe
flow. This chapter, therefore, focuses on closed-conduit flow only; Chapter 7
addresses open-channel flow.
The hydraulics of pipe flow are examined in this chapter by first reviewing the energy
equation from Section 3.2. Section 6.2 then covers the subject of flow resistance and
presents several commonly used constitutive relationships for estimating the energy
loss due to friction. Minor (or local) energy losses that occur at bends and other
appurtenances in pipe systems are addressed in Section 6.3.
When flow momentum is changed, either by a change in flow direction (such as at a
bend or manhole), or by a change in the magnitude of the velocity (due to a pipe
diameter change, for instance), the change in momentum is accompanied by a corre-
sponding hydraulic force that is exerted on the conduit or appurtenance. Prediction of
the magnitudes and directions of these forces is presented in Section 6.4. Calculation
of the predicted forces provides a basis for design of structural elements such as thrust
blocks or other force restraints.
6.1 THE ENERGY EQUATION

In Section 3.2, the energy equation was introduced as a simplification of the most for-
mal statement of conservation of energy, the First Law of Thermodynamics. The
energy equation is
p1 α V2 p α V2
+ Z1 + 1 1 = 2 + Z 2 + 2 2 + hL − hP + hT (6.1)
γ 2g γ 2g
where p = fluid pressure (lb/ft2, Pa)

γ = specific weight of fluid (lb/ft3, N/m3)
Z = elevation above an arbitrary datum plane (ft, m)
α = velocity distribution coefficient (see next section)
V = fluid velocity, averaged over a cross-section (ft/s, m/s)
g = acceleration of gravity (32.2 ft/s2, 9.81 m/s2)
hL = energy loss between cross-sections 1 and 2 (ft, m)
hP = fluid energy supplied by a pump between cross-sections 1 and 2 (ft, m)
hT = energy lost to a turbine between cross-sections 1 and 2 (ft, m)
Section 6.1 The Energy Equation 187
The first three terms on each side of Equation 6.1 are the pressure, elevation, and
velocity heads at a given cross-section of the flow. The subscript 1 applies to an
upstream cross-section, and the subscript 2 applies to a downstream cross-section.
The last three terms on the right side of Equation 6.1 represent energy losses and
gains that may occur along the flow path between cross-sections 1 and 2. In stormwa-
ter conveyance systems, it is rare for a turbine to be present; thus the term hT will be
assumed to equal zero throughout this text. Pumps are occasionally present in storm-
water conveyance systems and are discussed in Chapter 13. However, in this chapter,
it will be assumed that the term hP equals zero.
Energy losses between any two cross-sections in pipe flow are composed of distrib-
uted losses due to pipe friction (distributed because they occur all along the length of
a pipe, as opposed to occurring at one location) and localized or point losses (minor
losses) due to the presence of bends, pipe entrances and exits, and other appurte-
nances. Thus, the hL term in Equation 6.1 includes both frictional losees and minor
losses. Frictional losses are discussed in detail in Section 6.2. Minor losses are dis-
cussed in Section 6.3.
The Velocity Distribution Coefficient

The velocity term, V, in Equation 6.1 refers to the average cross-sectional velocity of
the fluid. This definition implies that velocity distribution is non-uniform over a cross
section. In fact, due to friction, the velocity of the fluid is zero at the walls of the con-
duit.
Strictly speaking, the energy equation applies only to a particular streamline in a flow.
The velocity distribution coefficient α, also known as the kinetic energy correction
factor, applied to the velocity head terms in Equation 6.1 should be included when
applying the equation to an averaged cross-sectional velocity (V = Q/A) rather than an
individual streamline velocity. By definition, the value of α at any cross section is
1 3
α=
AV A 3 ∫ v dA (6.2)
where A = total cross-sectional area of flow (ft2, m2)

V = cross-sectionally averaged velocity (ft/s, m/s)
v = velocity through an area dA of the entire cross-section (ft/s, m/s)
The integration is performed over the entire cross-sectional area.
An understanding of the basic types of flow that can exist in a conduit is essential to
determining the appropriate value of α for use with the average velocity in Equation
6.1. With turbulent flow, eddies form in the flow and cause flow mixing. In the case of
laminar flow, this intense mixing does not occur; in fact, the velocity for a steady flow
will be constant at a given location in the cross section over time. A parameter called
the Reynolds number (Re) (see page 192) is used in classifying flow as laminar or tur-
bulent. Between these classifications is a transitional range in which flow may be
either laminar or turbulent, depending on outside disturbances. The velocity distribu-
tions for typical turbulent and laminar flow conditions in a pipe are shown in Figure
6.2, as well as the idealized uniform velocity distribution in which the velocity is the
same across the entire cross section.
Figure 6.2
Cross-sectional
velocity distribution
for uniform, turbulent,
and laminar flow
conditions
Numerically, the velocity distribution coefficient α is greater than or equal to 1; it is

equal to 1 when the velocity distribution is uniform. Because the velocity distribution
in pipe flow is nonuniform, α is larger than 1 in pipe flow; however, it is often
assumed to be equal to 1 as an approximation. Table 6.1 lists recommended values of
α. Note that in fully turbulent flow, the flow distribution is approximately uniform
(see Figure 6.2), and α ≈ 1. For laminar flow, which has a parabolic velocity distribu-
tion, α = 2. (For an explanation of β, see Section 6.4.)
Table 6.1 Recommended values of α and β

Flow Description Re α β
Uniform velocity distribution N/A 1.0 1.00
Turbulent > 4,000 1.04–1.06 1.03–1.07
Transitional 2,000–4,000 1.1–1.9 1.1–1.3
Laminar < 2,000 2.0 1.33
Graphical Representation of the Energy Equation

Because each term in the energy equation has units of length and is referred to as
head, the equation can be graphically represented on a profile view of a pipe system.
Figure 6.3 illustrates such a profile with the various terms in the energy equation
depicted graphically.
Figure 6.3
Graphical
energy equation
Figure 6.3 shows two arbitrarily chosen flow cross sections, 1 and 2 (note that cross
section 1 is upstream of cross section 2), as well as an elevation datum. In practice,
cross sections are selected to correspond to locations where the flow behavior is of
interest (such as the entrance and outlet of a culvert). The datum may be taken as
mean sea level or any other convenient elevation.
At any cross section, the distance from the datum to the centerline of the pipe is
known as the elevation head and is denoted by the term Z on each side of the energy
equation. The pressure head, p/γ, is the distance from the pipe centerline to the
hydraulic grade line (HGL). The HGL elevation at a particular location represents the
height to which water would rise, due to pressure, if a standpipe were there. The dis-
tance from the datum to the HGL, which is the sum of the pressure and elevation
heads at any cross section, is called the piezometric head:
p
h= +Z (6.3)
γ
where h = piezometric head (ft, m)

The HGL represents the head due to the potential and internal energies only. Kinetic
energy is represented by the velocity head term, αV2/2g, which is the distance
between the HGL and the energy grade line (EGL) in Figure 6.3. The EGL represents
the total energy at any cross section. The distance from the datum to the EGL at any
section is called the total head:
p αV 2
H= +Z + (6.4)
γ 2g
where H = total head (ft, m)

Noting that H is equal to the sum of the first three terms on each side of the energy
equation (Equation 6.1), it can be seen that a difference in H between any two cross
sections is due to the head loss hL (and the pump head, if a pump exists) between
those cross sections. If no pump is present, H1 is always larger than H2 by the amount
hL because the total energy decreases due to frictional and minor losses as the flow
progresses in a downstream direction. In Figure 6.3, the energy loss is due to a combi-
nation of the frictional losses in the two pipes and the minor loss at the pipe expan-
sion.
In summary, the following statements about the graphical interpretation of the energy
equation can be made:
• The EGL represents the total head at a cross section relative to the chosen
datum, and always decreases in elevation as one moves in a downstream
direction (unless a pump is present).
• The HGL represents the height to which water would rise in a standpipe. In a
constant-diameter pipe, it is parallel to and below the EGL by the amount
αV2/2g. At locations where the flow velocity changes, the HGL becomes
either closer to or further from the EGL.
• The difference in the elevation of the EGL between any two cross sections is
equal to the head loss (and/or the pump energy supplied) between those
cross sections. For example, if the elevation of the EGL at one section is
known, and if the head loss between that section and a second cross section
is also known, then the elevation of the EGL at the second cross section can
be determined.
• The difference in elevation between the HGL and the pipe centerline is the
pressure head, p/γ. If this value is known, then the (gauge) fluid pressure in
the pipe can be found as the product of γ and the pressure head. (To convert
from gauge pressure to pressure head in feet if pressure is in psi, multiply the
pressure by 2.31. To obtain pressure head in meters, multiply the pressure in
kPa by 0.102.) The gauge pressure will be positive (that is, the absolute fluid
pressure will be greater than atmospheric pressure) if the HGL is above the
pipe centerline. Otherwise, the gauge pressure will be negative and the abso-
lute fluid pressure will be lower than the atmospheric pressure.
• The EGL slopes downward (in the direction of flow) along pipe reaches,
with the magnitude of the slope being dependent on the pipe diameter, fric-
tion factor, and flow velocity (see Section 6.2). At locations of minor losses,
the elevation of the EGL is usually assumed to drop abruptly, with the mag-
nitude of the drop being equal to the minor energy loss (see Section 6.3).
This set of observations is of great practical use for engineering calculations. Delinea-
tion of the HGL and EGL on design drawings illustrating the profile of a stormwater
conveyance system will clearly show the adequacy of the design, at least from an
energy point of view. Section 11.5 discusses this concept in detail.
Example 6.1 – Application of the Energy Equation. For the pipe flow situation shown
in Figure 6.3, the following information is given:
Z1 = 50 ft Z2 = 42 ft Q = 5 cfs
D1 = 12 in. D2 = 24 in. hL = 25 ft
p2/γ = 30 ft
At both cross sections, determine the velocity head, the elevations of the EGL and HGL, and the fluid
pressure in the pipe. Assume that α = 1 at both cross sections.
Solution: The pipe areas are required to determine the velocity at each cross section, which in turn is
necessary to compute the velocity head. At cross section 1, where the pipe diameter is 12 in., the area
is A1 = 0.785 ft2. At cross section 2, the area is A2 = 3.14 ft2. The velocity at cross section 1 is
V1 = Q/A1 = 5/0.79 = 6.33 ft/s
and the velocity at cross section 2 is V2 = 1.59 ft/s. The velocity heads at the two cross sections are
(αV2/2g)1 = 6.332/[2(32.2)] = 0.62 ft
(αV2/2g)2 = 1.592/[2(32.2)] = 0.04 ft

At cross section 2, the HGL elevation is the sum of Z2 and p2/γ, or
HGL2 = 42 ft + 30 ft = 72 ft
The EGL at cross section 2 is determined as the sum of the HGL and the velocity head, or
EGL2 = 72 ft + 0.04 ft = 72.04 ft
The EGL elevation at cross section 1 is found by adding the EGL at cross section 2 and the head loss
between the cross sections:
EGL1 = 72.04 ft + 25 ft = 97.04 ft
The HGL at cross section 1 is calculated by subtracting the velocity head from the EGL:
HGL1 = 97.04 ft – 0.62 ft = 96.42 ft
The pressure head at cross section 1 is found by subtracting the pipe centerline elevation from the
HGL elevation:
p1/γ = 96.41 ft – 50 ft = 46.42 ft
The fluid pressure at each cross section is found by multiplying the pressure head by the specific
weight of water:
p1 = (46.42 ft)(62.4 lb/ft3) = 2,900 lb/ft2 = 20.1 psi

p2 = (30 ft)(62.4 lb/ft3) = 1,870 lb/ft2 = 13.0 psi
6.2 RESISTANCE TO FLOW IN CLOSED CONDUITS

In most stormwater conveyance systems, the major cause of head losses is the friction
exerted on the fluid by the pipe walls. Minor losses, as the name implies, are often
small in comparison to frictional losses. This section discusses head losses caused by
pipe friction, and describes several of the methods that have been developed to quan-
tify pipe friction. Minor losses are discussed in Section 6.3.
The Reynolds Number

A set of hydraulic experiments carried out in the nineteenth century by Osborne Rey-
nolds (1883) examined the onset of turbulence in pipe flow. Reynolds varied the
velocity of flow through a glass tube into which a thread of dye was injected, and
observed that the thread of dye would break up and disperse by mixing when the
velocity became relatively large. By considering not only the velocity, but also the
pipe diameter and fluid viscosity, Reynolds was able to summarize his observations in
terms of a dimensionless parameter known as the Reynolds number, which is an indi-
cator of the relative importance of inertial forces to viscous forces. The Reynolds
number is computed as:
ρVD VD
Re = = (6.5)
µ ν
where Re = the Reynolds number (dimensionless)

ρ = fluid density (slugs/ft3, kg/m3)
V = flow velocity (ft/s, m/s)
D = pipe diameter (ft, m)
µ = dynamic viscosity of the fluid (lb-s/ft2, N-s/m2)
ν = kinematic viscosity of the fluid (ft2/s, m2/s)
In his experiments, Reynolds found that flow became turbulent when Re > 12,000.
However, in the commercially available pipes used for construction of stormwater
conveyance systems, flow is generally turbulent when the Reynolds number exceeds
4,000. Flow in a pipe is laminar when the Reynolds number is less than 2,000, and is
transitional (that is, between laminar and fully turbulent) when the value of Re is
between 2,000 and 4,000. Note that the Reynolds number of a flow increases when
either the pipe diameter or fluid velocity increases, or when the fluid viscosity
decreases.
Calculation of the Reynolds number is necessary to determine the friction factor to be
used in the Darcy-Weisbach equation for computing head loss, as described in the fol-
lowing subsection.
The Darcy-Weisbach Equation

The Darcy-Weisbach equation can be derived several ways and is an expression for
the head loss in a conduit due to friction. The head loss depends on both conduit and
flow properties and is based on the fact that, for steady flow, the frictional forces at
Section 6.2 Resistance to Flow in Closed Conduits 193
the pipe wall must be equal to the driving forces of pressure and gravity. Solving for
head loss due to friction, the Darcy Weisbach equation can be written as
L V2
hL = f (6.6)
D 2g
where hL = head loss due to friction (ft, m)
f = a friction factor depending on the Reynolds number of the flow and the
relative roughness of the pipe
L = pipe length (ft, m)
D = pipe diameter (ft, m)
V = cross-sectionally averaged velocity of the flow (ft/s, m/s)
g = acceleration of gravity (32.2 ft/s2, 9.81 m/s2)
If both sides of Equation 6.6 are divided by the pipe length, the slope of the energy
grade line Sf (that is, the energy gradient or friction slope) can be expressed as
hL 1 V2
Sf = = f (6.7)
L D 2g
This form of the Darcy-Weisbach equation presents the relationship between the
energy gradient and the flow velocity (or vice versa, by a simple algebraic rearrange-
ment). Equations of this nature, relating the energy gradient and flow velocity, are
called constitutive relationships. The Darcy-Weisbach equation can be argued to be
the most fundamentally sound constitutive relationship for pipe flow although many
others, such as the Manning equation, are also available and widely used.
As already noted, the friction factor f in the Darcy-Weisbach equation depends on the
Reynolds number and on the relative roughness of the pipe. The relative roughness is
defined as the ratio ε/D. Here, ε is the equivalent sand grain roughness of the pipe,
which must be experimentally determined. Table 6.2 presents recommended values of
ε for several common pipe materials. The values given are for new pipe. As pipe ages,
corrosion, pitting, scale build-up, and/or deposition usually occur, causing the equiva-
lent sand grain roughness to increase. Values for old pipes may be several orders of
magnitude larger than those shown in Table 6.2.
Table 6.2 Equivalent sand grain roughness for commercial pipes
Pipe Material ε (ft) ε (mm)

Corrugated metal 0.1–0.2 30–61
Riveted steel 0.003–0.03 0.91–9.1
Concrete 0.001–0.01 0.3–3.0
Wood stave 0.0006–0.003 0.18–0.91
Cast iron 0.0008–0.018 0.2–5.5
Galvanized iron 0.00033–0.015 0.102–4.6
Asphalted cast iron 0.0004–0.007 0.1–2.1
Steel, wrought iron 0.00015 0.046
Drawn tubing 0.000005 0.0015
With the Reynolds number and relative roughness known, the Moody diagram
(Moody, 1944), illustrated in Figure 6.4 on page 196, can be employed to obtain the
friction factor f. Note that for Reynolds numbers below 2,000, where the flow is lami-
nar, f depends only on the Reynolds number and is calculated as f = 64/Re. For large
Reynolds numbers where the curves on the Moody diagram become horizontal, it can
be seen that f depends only on the relative roughness of the pipe. Flow in this region is
considered to be fully turbulent or completely rough flow. In the transitional region
between laminar and fully turbulent flow, f depends on both Re and D.
Example 6.2 – Application of the Darcy-Weisbach Equation to Pipe Flow.

Determine the head loss in 75 m of 610-mm diameter concrete pipe conveying a discharge of 0.70
m3/s. Assume that the water temperature is 15° C.
Solution: Referring to Table 6.2, the equivalent sand grain roughness for the pipe can be taken as ε=
3 mm. The relative roughness is therefore ε/D = 0.005. Because the adopted value of ε is at the high
end of the range of recommended values, the computed head loss will also be high. It is often wise in
practical design to adopt high values of ε for head loss computations and to adopt low values of ε for
velocity calculations. When this is done, both computed head losses and computed velocities will
tend to be on the high side and thus conservative estimates.
The cross-sectional area of the 610-mm pipe is A = 0.292 m2, and the velocity of flow is V = Q/A =
2.40 m/s. Table A.1 (Appendix A) provides the kinematic viscosity for the water as ν = 1.13×10-6
m2/s. The Reynolds number is computed as
VD 2.4(0.61)
Re = = = 1.3 × 10 6
ν 1.13 × 10 −6
From the Moody diagram, f = 0.030. Applying the Darcy-Weisbach equation, the head loss in the pipe
is
 75  2.42
hL = 0.030   = 1.08 m
 0.61  2(9.8)
Example 6.3 – Determining Required Pipe Size Using Darcy-Weisbach. The

energy gradient available to drive the flow in a pipe is Sf = 0.015. If the required flow rate is Q = 50
cfs, and if the equivalent sand grain roughness of the pipe is ε = 0.001 ft, determine the required pipe
diameter.
Solution: Because the pipe diameter is unknown, the relative roughness of the pipe is also unknown.
The pipe cross-sectional area and flow velocity are unknown, and hence the Reynolds number and the
friction factor f are also unknown. Because of all the unknowns, a trial and error solution is required.
The solution proceeds by first assuming an f = 0.020. The pipe diameter is then computed by rear-
ranging Equation 6.7 to
1 Q2 8 fQ 2
Sf = f =
D 2 gA2 π 2 gD 5
or
1/ 5 1/ 5
 2   8(0.020)(50)2 
 8 fQ 
D=  =  = 2.43 ft
2 2
 π gS f   32.2π (0.015) 
With this initial estimate of the pipe diameter, the relative roughness and Reynolds number can be
computed to obtain a more accurate estimate of f:
ε 0.001
= = 0.0004
D 2.43
Q 50
V= = = 10.8 ft/s
A π (2.43) 2 / 4
VD 10.8(2.43)
Re = = = 2.2 ×106
ν 1.217 ×10−5
f = 0.016
With this improved estimate of f, the pipe diameter can be recomputed as
1/ 5
 8(0.016)(50) 
2
D=  = 2.32 ft
2
 32.2π (0.015) 
An additional iteration does not significantly change this diameter and thus the minimum required
diameter is 2.32 ft or 28 in. Rounding up to the next commercially available size, a 30-in. diameter
pipe should be selected.
Explicit Equations for f, Q, and D. Use of the Darcy-Weisbach equation for

pipe flow calculations can be inconvenient because of the need to consult the Moody
diagram. Further, as seen in the previous example, calculations are often iterative in
nature. Because of this, a number of efforts have been made to express the relation-
ship between the important variables in explicit mathematical form. This is especially
necessary if solutions are to be computerized.
Prandtl and von Kármán (Prandtl, 1952) used Nikuradse’s experimental results to
express the friction factor for flow in smooth pipes as
1  2.51 
= −2 log   (6.8)
f  Re f 
 
A smooth pipe is one in which the relative roughness ε/D is very small (see Figure
6.4). Prandtl and von Kármán also expressed the friction factor for completely rough
flow in pipes (where f depends only on ε/D) as
Credit: Moody, 1944. Used with permission.
1 ε /D
= −2 log   (6.9)
f  3.7 
For the transitional flow region where f depends on both Re and ε/D, Colebrook
(1939) developed the equation
1 ε /D 2.51 
= −2 log  +  (6.10)
f  3.7 Re f 
 
Figure 6.4
The Moody diagram
Note that Equations 6.8 and 6.9 are asymptotes of this Equation 6.10. Equation 6.10
reduces to Equation 6.8 when ε/D is small (smooth pipe), and to Equation 6.9 when
Re is large (fully turbulent flow).
The Colebrook equation, which was developed by Cyril Colebrook and Cedric White,
preceded the Moody diagram. It is useful but complicated by its implicit form.
Because f appears on both sides of the equation, a trial and error solution is required.
Recognizing this, Jain (1976) developed an explicit approximation to the Colebrook
equation as
1  ε / D 5.74 
= −2 log  + 
 3.7 Re0.9 
f (6.11)
This expression is valid for 10-6 ≤ ε /D ≤ 10-2 and 5,000 ≤ Re ≤ 108.

Swamee and Jain (1976) developed similar explicit equations for Q and D, which do
not require the iterative trial-and-error solutions of the Darcy-Weisbach equation.
Their equation for Q is
gDhL ε /D 1.784ν 
Q = −2.22D2 ⋅ log  +  (6.12)
L  3.7 D gDh / L 
 L 
A modified version of their equation for D is (Streeter and Wylie, 1979)
0.04
 4.75 5.2 
  LQ 2   L  
D = 0.66 ε 1.25   + ν Q9.4    (6.13)
 ghL   ghL 
   
 
These expressions are valid for 10-6 ≤ ε /D ≤ 10-2 and 5000 ≤ Re ≤ 108.
Noncircular Conduits. It is sometimes necessary in design and construction of
stormwater conveyance systems to employ conduits of noncircular shapes. The most
common of these shapes are elliptical, arch, and rectangular, although others exist as
well. Use of these shapes is most common in cases where the burial depth of the
stormwater conduit is limited. Rectangular shapes are often used in large cast-in-place
or precast concrete structures.
When the Darcy-Weisbach equation is applied to the analysis of flow in a noncircular
conduit, the pipe diameter D should be replaced by four times the hydraulic radius, or
D = 4R (6.14)
The hydraulic radius is defined as the ratio of the cross-sectional area to the wetted
perimeter of the flow, or R = A/P. Note that for full flow in a circular cross-section,
the area is simply the area of the circle (D2/4), and the wetted perimeter is the circum-
ference of the circle (P = 2πr = πD); thus, R = A/P = D/4.
Use of the substitution D = 4R means that the head loss, Reynolds number, and rela-
tive roughness expressions should all be modified to the following forms:
L V2
hL = f (6.15)
4R 2 g
4 ρVR 4VR
Re = = (6.16)
µ ν
ε
Relative roughness = (6.17)
4R
This substitution is an approximation, but the error is small if the ratio of the maxi-
mum to minimum cross-sectional dimensions of the conduit is less than about 4.0
(Potter and Wiggert, 1991).
The Manning Equation

Experiments conducted by the Irish engineer Robert Manning on flows in open chan-
nels and pipes led to the formulation of a constitutive relationship of the form (Man-
ning, 1891)
Cf
V = R 2 / 3 S01/ 2 (6.18)
n
where V = flow velocity (ft/s, m/s)

Cf = unit conversion factor (1.49 for U.S. Customary units; 1.00 for SI units)
n = friction factor (also known as Manning’s n)
R = hydraulic radius (ft, m)
So = channel slope (ft/ft, m/m)
Although it was primarily developed for open-channel flow problems, the Manning
equation has been used extensively for analysis of pipe flow. The equation was origi-
nally developed for uniform flow, where the pipe or channel bed slope is equal to the
energy gradient. However, it is also widely used for applications where So and Sf are
different from one another. When the energy gradient, Sf, is used for the slope term,
the equation can be rearranged as
( nV )2
Sf = (6.19)
C 2f R 4 / 3
In pipe flow applications, it is usually assumed that the value of Manning’s n depends
only on the pipe material, and that it is independent of the Reynolds number and the
pipe diameter. Therefore, the Manning equation should be considered to be valid only
for fully turbulent flow considering earlier discussions regarding the friction factor f
in the Darcy-Weisbach equation. It would also be expected that, in reality, the n value
should be different for large pipes than for small ones, even when the pipe materials
are the same. However, the same values of n are usually used in practice. Table 6.3
contains typical values of n for common pipe materials. Additional information for
selecting n values for corrugated steel pipe is provided in Section 11.1.
Table 6.3 Typical Manning’s n values

Pipe Material n
Corrugated metal 0.019–0.032
Concrete pipe 0.011–0.013
Concrete (cast-in-place) 0.012–0.017
Plastic (smooth) 0.011–0.015
Plastic (corrugated) 0.022–0.025
Steel 0.009–0.012
Cast iron 0.012–0.014
Brick 0.014–0.017
Clay 0.012–0.014
Example 6.4 – Application of the Manning Equation to a Culvert Flowing Full.

A 1200-mm diameter reinforced concrete pipe culvert conveys a discharge of 3.5 m3/s under a high-
way. The culvert has a length of 60 m. If n = 0.013 and the culvert is flowing full, determine the head
loss through the culvert. Neglect minor losses at the culvert entrance and outlet.
Solution: The cross-sectional area of the culvert is 1.13 m2 and the flow velocity is
V = Q/A = 3.10 m/s
The hydraulic radius of a circular cross section when flowing full is R = D/4, and hence R = 0.3 m.
Applying Equation 6.19, the energy gradient is
( 0.013 × 3.10 )2
Sf = = 0.0081
4/3
0.3
Multiplying the energy gradient by the culvert length yields the head loss as
hL = Sf L = 0.0081(60) = 0.486 m
The Chézy Equation

Around 1769, Antoine Chézy, a French engineer, developed what is probably the old-
est constitutive relationship for uniform flow in an open channel (Herschel, 1897).
The same relationship has also been used for pipe flow. The Chézy equation takes the
form
V = C RSo (6.20)

C = a resistance coefficient known as Chézy’s C (ft0.5/s, m0.5/s)
If one assumes that the slope can be replaced by the energy gradient, and the equation
is rearranged, one obtains a constitutive relationship of the form
V2
Sf = (6.21)
C2R
As is the case with the Manning equation, it is usually assumed in practice that the
resistance coefficient (C in this case) is a constant and is independent of the Reynolds
number and the relative roughness of the conduit. Chézy’s C can be related to Man-
ning’s n by equating the right sides of Equations 6.19 and 6.21 to obtain
C f R1 / 6
C= (6.22)
n
where Cf = unit conversion factor (1.49 for U.S. Customary units; 1.00 for SI units)
Example 6.5 – Application of the Chézy Equation to Pipe Flow. The head loss
along a 107-m long pipe is 1.2 m. If the pipe is 460 mm in diameter and the flow rate is 0.6 m3/s,
determine Chézy’s C for the pipe.
Solution: The energy gradient is the ratio of the head loss to the pipe length, or
Sf = 1.2 m/107 m = 0.01121
The hydraulic radius of the pipe is R = D/4 = 0.115 m, and the flow velocity is V = Q/A = 0.6/0.166 =
3.61 m/s. Rearranging Equation 6.21, Chézy’s C is determined as
V 3.61
C= = = 100.5 m1/2/s
RS f 0.115(0.01121)
The Hazen-Williams Equation

The Hazen-Williams equation was developed primarily for use in water distribution
design and has rarely been used for other applications. The equation is (Williams and
Hazen, 1933):
0.63 0.54
V = Cf Ch R Sf (6.23)

Cf = a unit conversion factor (1.318 for U.S. Customary units; 0.849 for SI
units)
Ch = Hazen-Williams resistance coefficient
Sf = energy gradient (ft/ft, m/m)
Some typical values of the resistance coefficient are given in Table 6.4.
Table 6.4 Typical values of the Hazen-Williams resistance coefficient (Viessman and
Hammer, 1993)
Pipe Material Ch
Cast iron (new) 130
Cast iron (20 yr old) 100
Concrete (average) 130
New welded steel 120
Asbestos cement 140
Plastic (smooth) 150
Chin (2000), in a discussion of the Hazen-Williams formula, points out that its use
should be limited to flows of water at 60°F (16°C), to pipes with diameters from 2 to
72 in. (50 to 1800 mm), and to flow velocities less than 10 ft/s (3 m/s). Further, the
resistance coefficient is dependent on the Reynolds number, which means that signifi-
cant errors can arise from use of the formula.
For high velocities in rough pipes, Walski (1984) gives a correction factor for C as
0.081
V 
Ch = Co  o  (6.24)
V 
where Co = measured Hazen-Williams resistance coefficient

Vo = velocity at which C was measured
V = velocity for which Ch is to be calculated
6.3 MINOR LOSSES AT BENDS AND APPURTENANCES

The previous section presented a number of methods for estimating the head loss
caused by friction as flow moves along the length of a pipe. This section discusses an
additional type of head loss, called minor loss, which occurs at locations of manholes,
pipe entrances and exits, and similar pipe appurtenances. As stated in Section 6.1, the
total head loss along a pipe is equal to the sum of the frictional and minor losses.
There are a number of methods by which minor losses can be predicted. The classical
approach, which is presented in this section, is to express the head loss as the product
of a minor loss coefficient and the absolute difference between the velocity heads
upstream and downstream of the appurtenance. This approach is reasonable for many
types of minor loss calculations. However, at stormwater inlets, manholes, and junc-
tion structures, where flow geometries are often complex, additional methods have
been developed for minor loss prediction and are presented by Brown, Stein, and
Warner (1996). These approaches are known as the energy-loss and power-loss meth-
ods, and they are covered in detail in Section 11.4.
Classical Method
As already noted, the classical approach to predicting minor losses at pipe entrances,
exits, and other appurtenances is to express the head loss as the product of a minor
loss coefficient, K, and the absolute difference between the upstream and downstream
velocity heads:
V22 − V12
hL = K (6.25)
2g
where hL = head loss due to minor losses (ft, m)

K = minor loss coefficient
V = velocity of flow (ft/s, m/s)
g = gravitational acceleration constant (32.2 ft/s2, 9.81 m/s2)
In cases where the upstream and downstream velocities are equal in magnitude, the
head loss can be expressed as the product of a loss coefficient and the velocity head:
V2
hL = K (6.26)
2g
Equation 6.25 reduces to Equation 6.26 when either V1 or V2 is small in comparison

to the other. An assumption that either V1 or V2 is small in comparison to the other is
often quite reasonable at, for example, culvert entrances and outlets.
Regardless of whether Equation 6.25 or Equation 6.26 is used, the numerical value of
the minor loss coefficient depends on the type of the appurtenance where the loss
occurs, as well as on the geometry of the appurtenance. The following subsections
summarize several appurtenance types and their associated minor loss coefficients.
Section 6.3 Minor Losses at Bends and Appurtenances 203
Pipe Entrances. The minor loss that occurs at a pipe entrance can be predicted
using Equation 6.25, where V1 is the velocity upstream of the entrance and V2 is the
velocity in the pipe. It is often reasonable to assume that V1 is small compared to V2;
thus, Equation 6.26 can be used. The minor loss coefficient at a pipe entrance depends
on the geometry of the pipe entrance, and particularly on conditions such as whether
the edges of the entrance are square-edged or rounded. The loss coefficient also
depends on whether the entrance is fitted with headwalls, wingwalls, or a warped
transition, and on whether the entrance to the pipe is skewed.
Table 6.5 provides a listing of entrance loss coefficients for culverts operating under
outlet control (see Section 9.5, page 338 for an explanation of outlet control). These
values are also applicable to entrances to other pipes, such as storm sewers, where the
water depth at the pipe entrance is controlled by downstream conditions. In cases
where inlet control governs (usually in cases where the pipe is steep and/or down-
stream flow depths are small), the water surface elevation upstream of a pipe entrance
should be determined using inlet control calculation procedures for culverts (see Sec-
tion 9.5).
Table 6.5 Entrance loss coefficients for pipes and culverts operating under outlet control
(Normann, Houghtalen, and Johnston, 2001)
Structure Type and Entrance Condition K
Concrete Pipe
Projecting from fill, socket or groove end 0.2
Projecting from fill, square edge 0.5
Headwall or headwall and wingwalls
Socket or groove end 0.2
Square edge 0.5
Rounded (radius = D/12) 0.2
Mitered to conform to fill slope 0.7
End section conforming to fill slope 0.5
Beveled edges (33.7° or 45° bevels) 0.2
Side- or slope-tapered inlet 0.2
Corrugated Metal Pipe or Pipe-Arch
Projecting from fill (no headwall) 0.9
Headwall or headwall and wingwalls (square edge) 0.5
Reinforced Concrete Box
Headwall parallel to embankment (no wingwalls)
Square-edged on 3 sides 0.5
Rounded or beveled on 3 sides 0.2
Wingwalls at 30° to 75° from barrel
Square-edged at crown 0.4
Crown edge rounded or beveled 0.2
Wingwalls parallel (extensions of box sides)
Pipe Outlets. The exit loss (or outlet loss) from a storm sewer or culvert can be
predicted by using Equation 6.25, although it is often assumed that V2 is small com-
pared to V1 and so Equation 6.26 is used instead. The minor loss coefficient at a pipe
outlet is nearly always assumed to be equal to 1.0, although it may be as small as 0.8
if a warped transition structure is provided at the pipe outlet. If there is no change in
the velocity of flow at a pipe outlet, the minor loss is equal to zero.
Pipe Bends. The head loss at a mitered bend in a pipe, or at a location where the
alignment of the pipe is deflected at a pipe joint, can be predicted using Equation
6.26. The U.S. Bureau of Reclamation (USBR, 1983) indicates that the minor loss
coefficient in this case depends on the pipe deflection angle, as shown in Figure 6.5.
Figure 6.5
The minor loss
coefficient at a pipe
bend depends on the
pipe deflection angle
(USBR, 1983)
The energy-loss and composite-energy-loss methods for computing losses in manhole

structures are presented in Section 11.4; however, the American Association of State
Highway and Transportation Officials (AASHTO, 1991) indicates that minor losses at
manholes also can be estimated using Equation 6.26. The value of K depends on the
pipe deflection angle ∆ (in degrees) as
K = 0.0033∆ (6.27)
where ∆ = pipe deflection angle (degrees)

This expression is limited to use when ∆ ≤ 90°, when there is only one pipe flowing
into the manhole and one pipe flowing out, and when the pipes are of equal diameter.
When these conditions are not met, the more complex prediction methods described
in Section 11.4 should be used.
Pipe Transitions. A pipe transition is a structure where the pipe cross-sectional
area (that is, the diameter) changes. A transition is called an expansion if the area
increases, and a contraction if the area decreases.
The minor loss coefficient for an expansion or contraction depends on the relative
sizes of the upstream and downstream pipes, and on the abruptness (as measured in
terms of the cone angle) of the transition (see Figure 6.6). Expansion and contraction
loss coefficients also depend on whether the flow is pipe flow or open-channel flow.
This section discusses minor loss coefficients for pipe flow; those for open-channel
flow are discussed in Section 7.2.
Figure 6.6
Pipe expansion and
contraction
The head loss in a sudden or gradual pipe expansion can be predicted using Equation
6.26 if the upstream (higher) velocity is used to compute the velocity head. Minor loss
coefficients are presented in Table 6.6 for sudden expansions (cone angle greater than
60 degrees) and in Table 6.7 for gradual expansions (cone angle 60 degrees or
smaller).
The head loss in a sudden contraction can also be estimated using Equation 6.26, but
the downstream (higher) velocity should be used to compute the velocity head. The
minor loss coefficient for a sudden contraction may be obtained from Table 6.8. The
head loss due to a gradual contraction may be estimated as one-half the head loss for a
similar pipe expansion or, alternatively, by using Equation 6.25 and the minor loss
coefficients presented in Table 6.9.
Table 6.6 Minor loss coefficients for sudden expansions in pipe flow [American Society of
Civil Engineers (ASCE), 1992]
D2 /D1
V1 (ft/s) 1.2 1.4 1.6 1.8 2.0 2.5 3.0 4.0 5.0 10 ∞
2 0.11 0.26 0.40 0.51 0.60 0.74 0.83 0.92 0.96 1.00 1.00
3 0.10 0.26 0.39 0.49 0.58 0.72 0.80 0.89 0.93 0.99 1.00
4 0.10 0.25 0.38 0.48 0.56 0.70 0.78 0.87 0.91 0.96 0.98
5 0.10 0.24 0.37 0.47 0.55 0.69 0.77 0.85 0.89 0.95 0.96
6 0.10 0.24 0.37 0.47 0.55 0.68 0.76 0.84 0.88 0.93 0.95
7 0.10 0.24 0.36 0.46 0.54 0.67 0.75 0.83 0.87 0.92 0.94
8 0.10 0.24 0.36 0.46 0.53 0.66 0.74 0.82 0.86 0.91 0.93
10 0.09 0.23 0.35 0.45 0.52 0.65 0.73 0.80 0.84 0.89 0.91
12 0.09 0.23 0.35 0.44 0.52 0.64 0.72 0.79 0.83 0.88 0.90
15 0.09 0.22 0.34 0.43 0.51 0.63 0.70 0.78 0.82 0.86 0.88
20 0.09 0.22 0.33 0.42 0.50 0.62 0.69 0.76 0.80 0.84 0.86
30 0.09 0.21 0.32 0.41 0.48 0.60 0.67 0.74 0.77 0.82 0.83
40 0.08 0.20 0.32 0.40 0.47 0.58 0.65 0.72 0.75 0.80 0.81
Table 6.7 Minor loss coefficients for gradual expansions in pipe flow (ASCE, 1992)
D2 /D1
Cone
Angle 1.1 1.2 1.4 1.6 1.8 2.0 2.5 3.0 ∞
2° 0.01 0.02 0.02 0.03 0.03 0.03 0.03 0.03 0.03
6° 0.01 0.02 0.03 0.04 0.04 0.04 0.04 0.04 0.05
10° 0.03 0.04 0.06 0.07 0.07 0.07 0.08 0.08 0.08
15° 0.05 0.09 0.12 0.14 0.15 0.16 0.16 0.16 0.16
20° 0.10 0.16 0.23 0.26 0.28 0.29 0.30 0.31 0.31
25° 0.13 0.21 0.30 0.35 0.37 0.38 0.39 0.40 0.40
30° 0.16 0.25 0.36 0.42 0.44 0.46 0.48 0.48 0.49
35° 0.18 0.29 0.41 0.47 0.50 0.52 0.54 0.55 0.56
40° 0.19 0.31 0.44 0.51 0.54 0.56 0.58 0.59 0.60
50° 0.21 0.35 0.50 0.57 0.61 0.63 0.65 0.66 0.67
60° 0.23 0.37 0.53 0.61 0.65 0.68 0.70 0.71 0.72
Table 6.8 Minor loss coefficients for sudden contractions in pipe flow (ASCE, 1992)
D2 /D1
V2 (ft/s) 1.1 1.2 1.4 1.6 1.8 2.0 2.2 2.5 3.0 4.0 5.0 10 ∞
2 0.03 0.07 0.17 0.26 0.34 0.38 0.40 0.42 0.44 0.47 0.48 0.49 0.49
3 0.04 0.07 0.17 0.26 0.34 0.38 0.40 0.42 0.44 0.46 0.48 0.48 0.49
4 0.04 0.07 0.17 0.26 0.34 0.37 0.40 0.42 0.44 0.46 0.47 0.48 0.48
5 0.04 0.07 0.17 0.26 0.34 0.37 0.39 0.41 0.43 0.46 0.47 0.48 0.48
6 0.04 0.07 0.17 0.26 0.34 0.37 0.39 0.41 0.43 0.45 0.47 0.48 0.48
7 0.04 0.07 0.17 0.26 0.34 0.37 0.39 0.41 0.43 0.45 0.46 0.47 0.47
8 0.04 0.07 0.17 0.26 0.33 0.36 0.39 0.40 0.42 0.45 0.46 0.47 0.47
10 0.04 0.08 0.18 0.26 0.33 0.36 0.38 0.40 0.42 0.44 0.45 0.46 0.47
12 0.04 0.08 0.18 0.26 0.32 0.35 0.37 0.39 0.41 0.43 0.45 0.46 0.46
15 0.04 0.08 0.18 0.26 0.32 0.34 0.37 0.38 0.40 0.42 0.44 0.45 0.45
20 0.05 0.09 0.18 0.25 0.32 0.33 0.35 0.37 0.39 0.41 0.42 0.43 0.44
30 0.05 0.10 0.19 0.25 0.29 0.31 0.33 0.34 0.36 0.37 0.38 0.40 0.41
40 0.06 0.11 0.20 0.24 0.27 0.29 0.30 0.31 0.33 0.34 0.35 0.36 0.38
Table 6.9 Minor loss coefficients for gradual

contractions in pipe flow (Simon, 1981)
Cone Angle K
20° 0.20
40° 0.28
60° 0.32
80° 0.35
Example 6.6 – Computing Friction, Entrance, and Exit Losses for a Culvert
Flowing Full. Consider again the culvert described in Example 6.4 with a diameter of 1200 mm, a
length of 60 m, and n = 0.013 (reinforced concrete pipe). The discharge is Q = 3.5 m3/s. The invert
elevation at the culvert outlet is 80 m, and the culvert slope is So = 0.002. The culvert entrance con-
sists of the groove end of a pipe joint at a headwall. No special transition is provided at the culvert
outlet. If the depth of flow just downstream of the culvert outlet is 1.5 m, determine the depth of water
and the water surface elevation just upstream of the culvert entrance. Assume that the velocities in the
channels upstream and downstream of the culvert are small and approximately equal to one another.
Solution: As determined in Example 6.4, the head loss due to friction in the culvert is 0.486 m.
Because the channel velocities upstream and downstream of the culvert are small, the minor losses
due to the culvert entrance and outlet can be determined using Equation 6.26, where V is taken as the
culvert barrel velocity (V = 3.10 m/s). Referring to Table 6.5, the entrance minor loss coefficient is
K = 0.2. For the outlet, K = 1.0. Thus, the total head loss through the culvert is
3.10 2 3.10 2
hL = 0.486 + 0.2 + 1.0 = 1.07 m
2(9.8) 2(9.8)
In open-channel flow, which exists both upstream and downstream of the culvert, the water surface
elevation is the hydraulic grade line (HGL), and the total head or energy grade line (EGL) lies above
the water surface by an amount equal to the velocity head (see Section 7.2). Because the velocity
heads upstream and downstream of the culvert in this example are approximately equal, the head loss
can be taken as the difference between the water surface elevations upstream and downstream of the
culvert.
The downstream water surface elevation, HGL2, is the sum of the channel invert elevation and the
water depth, or
HGL2 = 80.0 + 1.50 = 81.5 m
The upstream water surface elevation, HGL1, is the sum of the downstream water surface elevation
and the head loss in the culvert:
HGL1 = 81.5 + 1.07 = 82.57 m
The channel invert elevation, Z1, at the culvert entrance is equal to the invert elevation at the outlet
plus the change in elevation due to the culvert slope, or
z1 = z2 + SoL = 80.0 + 0.002 (60) = 80.12 m
Thus, the water depth, y1, at the culvert entrance is
y1 = 82.57 – 80.12 = 2.45 m
6.4 MOMENTUM AND FORCES IN PIPE FLOW

Section 3.3 presented the momentum equation as one of three fundamental physical
laws applied to flows in stormwater conveyance systems. The discussion pointed out
that analyses based on the principle of conservation of momentum are more complex
than are analyses based on conservation of mass and/or energy because momentum is
a vector-valued quantity. Thus, a separate equation must be written for each coordi-
nate direction.
The momentum equation must be applied in cases where one needs to estimate the
hydraulic forces acting on a component of a stormwater conveyance system. For
example, knowledge of hydraulic forces may be required to determine the need for
thrust blocks or restraints at pipe bends or other appurtenances. These types of
hydraulic forces occur whenever there is a change in the velocity of a flow. Since
velocity is a vector, a change in velocity means either a change in its magnitude or
direction, or both.
Conservation of momentum also plays an important role in the analysis of time-
variable pressures within a piping system as a consequence of hydraulic transients.
These transients, often called water hammer, can arise as a result from starting or
stopping of pumps. Hydraulic transients caused by pumps are discussed briefly in
Chapter 11.
In its most general form, conservation of momentum is expressed as
∂
∑ Fx = ∂t ∫ ρVx dS + ∑ (M out ) x − ∑ (M in ) x (6.28)
cv
Section 6.4 Momentum and Forces in Pipe Flow 209
where Fx = force in x-direction (lb, N)

t = time (s)
Vx = x-component of the velocity of fluid in the control volume (ft/s, m/s)
dS = incremental volume of fluid (ft3, m3)
(Mout)x = momentum outflow rate from the control volume in the x direction (lb, N)
(Min)x = momentum inflow rate into the control volume in the x direction (lb, N)
This expression, written for the coordinate direction x, also can be written for other
coordinate directions as needed. It states that the sum of the external forces acting on
the water in a control volume is equal to the time rate of change of momentum within
the control volume, plus the net momentum outflow rate (in the x direction) from the
control volume. The sums of Mout and Min account for the possibility of more than
one inflow and/or outflow pipe to and from the control volume. A momentum flow
rate (in the x direction) can be written as
M x = βρ QV x (6.29)
where β = a coefficient defined by Equation 6.30

Q = discharge (cfs, m3/s)
The term β, like α in the energy equation (Equation 6.1), is a coefficient accounting
for the effect of a nonuniform velocity distribution. The definition of β is
1 2
β=
AV 2
∫v dA (6.30)
A
where A = total cross-sectional area of flow (ft2, m2)

V = averaged cross-sectional velocity (ft/s, m/s)
v = velocity through an area dA of the entire cross section (ft/s, m/s)
Typical values of β are provided in Table 6.1. Note that β for turbulent flow in a pipe
is near 1, and thus it is often taken to be 1 in practice.
Evaluating Forces on Hydraulic Structures

For steady-state flow, the time derivative is equal to zero and Equation 6.28 reduces to
the simpler expression given by Equation 3.12 (repeated here):
∑ Fx = ∑(Mout )x − ∑(Min )x (6.31)
This steady-state case is adequate for use in most practical applications, although the
more complicated Equation 6.28 must be used in the analysis of hydraulic transients.
In analyzing the forces acting on a structure such as a manhole, the forces due to fluid
pressure and change in momentum must be computed and used to determine the reac-
tion forces, Rx and Ry, on the water in the structure. The pressure and reaction forces
are placed on the left side of Equation 6.31 for each coordinate direction, and the dif-
ference in the momentum of flows entering and leaving the structure is computed on
the right side of the equation.
The hydraulic forces exerted by the water on the manhole are equal, but opposite in
direction, to the computed reaction forces. Example 6.7 demonstrates the calculation
of the hydraulic forces exerted on a manhole.
Example 6.7 – Hydraulic Forces. Determine the hydraulic forces exerted on the manhole in
Figure E6.7.1. The hydraulic grade line elevations given in the figure should be used in computing the
pressure forces. (Note that these HGL elevations have already been computed using the energy-loss
method; these calculations will be presented in Example 11.3.) Assume that β = 1 for all pipes.
Solution: Figure E6.7.1 shows the various forces exerted on the water in the control volume. The con-
trol volume is taken as the circle in the figure, which represents the manhole. The x- and y-coordinate
directions assumed for this example are also shown in the figure. Note that there are two momentum
flows into the control volume, and one momentum flow out of the control volume.
As shown in Figure E6.7.1, the forces acting on the water in the control volume in the x-direction are
due to the pressures in Pipe 1 and the outflow pipe, plus the reaction force Rx on the water in the con-
trol volume. The pressure in Pipe 1 is determined by multiplying the difference in elevation between
HGL1 and the centerline elevation of Pipe 1 by the specific weight of water. The pressure force
exerted on the control volume by Pipe 1 is the product of its cross-sectional area and the pressure.
Thus, the force is
Px1 = γ (HGL1 – Z1)A1 = 62.4(104.12 – 102.00)(3.14) = 415 lb
Similarly,
Pxo = −γ (HGLo – Zo)Ao = -62.4(106.00 – 101.50)(7.07) = −1,985 lb
Note that Pxo is negative because it acts in the negative x-direction.
Forces acting on the water in the control volume in the y-direction are Py2 and the reaction force Ry.
The pressure force is
Py2 = γ (HGL2 – Z2)A2 = 62.4(106.57 – 102.25)(1.77) = 477 lb
Section 6.4 Momentum and Forces in Pipe Flow 211
Figure E6.7.1 Forces exerted on a manhole

Momentum flows in the x-direction are
(Min)x = ρQ1V1x = 1.94(42)(13.37) = 1,089 lb
(Mout)x = ρQoVox = 1.94(50)(7.07) = 686 lb
Momentum flows in the y-direction are
(Min)y = ρQ2V2y = 1.94(8)(4.53) = 70 lb
(Mout)y = 0
Applying Equation 6.28, and assuming steady-state conditions (that is, the time derivative is zero),
415 – 1985 + Rx = 686 – 1,089
or
Rx = 1,167 lb
Applying Equation 6.28 again, but writing it for the y-direction, leads to
477 + Ry = 0 – 70
or
Ry = −547 lb
Thus, the actual reaction force Ry has a direction opposite to that assumed and shown in Figure
E6.7.1.
It should be noted that the reaction forces Rx and Ry computed using the momentum equation are the
forces acting on the water in the control volume. As stated previously, the hydraulic forces exerted by
the water on the manhole are equal, but opposite in direction, to the computed reaction forces.
6.5 CHAPTER SUMMARY

A pipe or closed-conduit flow condition exists when the flowing water occupies the
entire cross-sectional area of a closed conduit. The energy equation in the form
p1 α1V12 p2 α 2V22
+ Z1 + = + Z2 + + hL − hP + hT (6.32)
γ 2g γ 2g
can be used to describe flow in closed conduits.

The hydraulic grade line elevation at a cross section is given by the sum of the pres-
sure head (p/γ) and elevation head (Z) terms; this quantity is also known as the piezo-
metric head. The energy grade line elevation or total head is given by the sum of the
piezometric head and the velocity head (V2/2g).
The primary cause of head loss in stormwater conveyance systems is pipe friction.
Several methods for computing energy losses are available, including the Darcy-
Weisbach, Manning, Chézy, and Hazen-Williams formulas.
In addition to friction losses, energy losses occur due to changes in flow velocity
through pipe entrances, outlets, transitions, bends, and other appurtenance. These
losses, known as minor losses, are computed by multiplying a coefficient K by the
velocity head or change in velocity head. Typical values for K under many conditions
are available.
In cases where forces on hydraulic structures need to be computed to determine
whether thrust blocks or restraints are needed, the momentum equation is applied.
Another application of the momentum equation is in the analysis of pressures due to
hydraulic transients.
Section References 213
REFERENCES
American Association of State Highway and Transportation Officials (AASHTO). 1991. Model Drainage
Manual, Washington, D.C.: American Association of State Highway and Transportation Officials.
agement Systems, New York: American Society of Civil Engineers.
Circular No. 22, FHWA-SA-96-078. Washington, D.C.: U.S. Department of Transportation., Federal
Highway Administration.
Chin, D.A. 2000. Water-Resources Engineering. Upper Saddle River, N. J.: Prentice Hall.
Colebrook, C. F. 1939. “Turbulent Flow in Pipes with Particular Reference to the Transition Between
Smooth and Rough Pipe Laws.” Journal of the Institution of Civil Engineers of London 11.
Herschel, C. 1897. “On the Origin of the Chézy Formula.” Journal of the Association of Engineering Soci-
eties 18: 363–368.
Jain, A. K. 1976. “Accurate Explicit Equation for Friction Factor.” Journal of the Hydraulics Division,
ASCE 102, no. HY5: 674–677.
Manning, R. 1891. “On the Flow of Water in Open Channels and Pipes. Transactions, Institution of Civil
Engineers of Ireland 20: 161–207.
Moody, L. F. 1944. “Friction Factors for Pipe Flow.” Transactions, American Society of Mechanical Engi-
neers (ASME) 66.
Normann, J. M., R. J. Houghtalen, and W. J. Johnston. 2001. Hydraulic Design of Highway Culverts, Sec-
ond Edition. Hydraulic Design Series No. 5. Washington, D.C.: Federal Highway Administration
(FHWA).
Potter, M.C., and D. C. Wiggert. 1991. Mechanics of Fluids. Englewood Cliffs, NJ: Prentice Hall.
Prandtl, L. 1952. Essentials of Fluid Dynamics. New York: Hafner Publ.
Reynolds, O. 1883. “An Experimental Investigation of the Circumstances Which Determine Whether the
Motion of Water Shall be Direct or Sinuous and of the Law of Resistance in Parallel Channels.” Philo-
sophical Transactions of the Royal Society 174: 935.
Simon, A. L. 1981. Practical Hydraulics. 2d ed. New York: John Wiley & Sons.
Streeter, V. L., and Wylie, E. B. 1979. Fluid Mechanics. 7th ed. New York: McGraw Hill.
Swamee, P. K., and A. K. Jain. 1976. “Explicit Equations for Pipe Flow Problems.” Journal of the Hydrau-
lics Division, ASCE 102, no. HY5: 657–664.
U.S. Bureau of Reclamation (USBR). 1983. Design of Small Canal Structures. Lakewood, Colorado:
USBR.
Viessman, W., Jr., and M. J. Hammer. 1993. Water Supply and Pollution Control. 5th ed. New York: Harper
Collins.
Walski, T. M. 1984. Analysis of Water Distribution Systems. New York: VanNostrand Reinhold.
Williams, G. S., and A. H. Hazen. 1933. Hydraulic Tables. 3d ed. New York: John Wiley and Sons.
PROBLEMS
6.1 The bottom of a 24-in. diameter pipe lies 8 ft below the ground surface. It is flowing full at a rate of
15 cfs. The pressure in the pipe is 35 psi. What is the total hydraulic energy at the centerline of the
pipe with respect to the ground?
6.2 A 30-in. pipe lies 10 ft above a reference level. The flow is 30 cfs, and the pressure is 45 psi. What is
the total energy in the pipe with respect to the reference level?
6.3 A 300-mm pipe drains a small reservoir. The outlet to the pipe is 6.1 m below the surface of the res-
ervoir, and it discharges freely to the atmosphere. Assuming no losses through the system, what is
the flow through the pipe and the pipe’s exit velocity?
6.4 Find the total head loss in 305 m of 460-mm diameter cast-iron pipe carrying a flow of 0.43 m3/s.
Use the Hazen-Williams formula to compute head loss assuming C = 130.
6.5 Water flows through a 12-in. diameter riveted steel pipe with ε = 0.01 ft and a head loss of 20 ft over
1,000 ft. Determine the flow.
6.6 What size pipe is required to transport 10 ft3/s of water a distance of 1 mile with a head loss of 6 ft?
Use new cast-iron pipe with an equivalent sand roughness of 0.00085 ft.
6.7 Water is to be pumped at a rate of 0.057 m3/s over a distance of 610 m. The pipe is level and the
pump head cannot exceed 18.3 m. Determine the required concrete pipe diameter. Use ε = 0.005 ft
(1.5 mm).
6.8 For the following figure, sketch the energy grade line between points 1 and 2.
6.9 An 8-in. concrete pipe is carrying 2.25 ft3/s and has a length of 200 ft. Calculate the head loss using
a) Darcy Weisbach equation using ε = 0.01 ft
b) Hazen-Williams formula using C = 100
c) Manning’s equation using n = 0.011
6.10 Revisit the circular, concrete culvert from Example 6.6 and assume that the headwater rises to a level
of 10 ft above the upstream invert of the culvert to an elevation 273.4 ft and determine the discharge
Problems 215
through the culvert given the characteristics in the following table. The velocities in the channels
upstream and downstream of the culvert are low and approximately equal.
Parameter Value
Diameter 48 in.
Length 200 ft
Manning’s n 0.013
Outlet invert elev. 263.00 ft
Slope 0.002
Entrance K 0.2
Exit K 1.0
Tailwater depth 5.0 ft
6.11 A 380-mm diameter new cast iron pipe connecting two reservoirs as shown carries water at 50°F.
The pipe is 45.7 m long, and the discharge is 0.57 m3/s. Determine the elevation difference between
the water surfaces in the two reservoirs.
Overflows from this
reservoir exit over
a spillway and
enter the rock
channel below.
C H A P T E R
7
Flow in Open Channels
Open channels are ubiquitous in both natural and constructed environments. They
consist of natural stream channels as well as natural swales and depressions along
which water runs following rainfall events. In the constructed or urbanized environ-
ment, an open channel may take the form of a street gutter, drainage ditch, pipe (if
there is a free water surface at atmospheric pressure), lined or unlined stormwater col-
lection channel, or natural or modified stream channel.
In open-channel flow, unlike closed-conduit flow, the cross-sectional area of flow is
not a constant that is governed by the shape and size of the conveyance. Rather, the
cross-sectional area and associated flow descriptors, such as the average flow veloc-
ity, are unknowns and must be determined as part of the solution to a problem. Further
complicating analyses of open-channel flow is the fact that there is not, in general, a
unique or one-to-one relationship between discharge Q and depth of flow y in an open
channel. In most instances of open-channel flow, the depth of flow for any specified
channel discharge depends on channel shape and roughness characteristics, as well as
on specified boundary conditions at flow control sections (that is, cross sections where
a known relationship between depth and discharge exists). Thus, open-channel flow is
analytically much more difficult to treat than is pressure flow, but the challenge
imposed by the additional difficulty also makes it more interesting.
This chapter presents methods for analysis of the hydraulics of steady flows in open
channels. Unsteady flows are viewed as an advanced topic and are introduced in
Appendix C. The reader may consult Chow (1959); Abbott (1979); Cunge, Holly, and
Verwey (1980); French (1985); Fread (1993); and Chaudhry (1993) for more detailed
coverage of unsteady flows in open channels.
Section 7.1 of this chapter describes the classification of the types of open channels
and gives an overview of their geometric properties. Energy and constitutive relation-
ships for the estimation of energy losses in open-channel flow are presented in Sec-
tion 7.2.
Sections 7.3 and 7.4, respectively, deal with the fundamentally important concepts of
uniform flow and critical flow in open channels—the only types of flow for which
there is a one-to-one relationship between channel discharge and flow depth. Uniform
218 Flow in Open Channels Chapter 7
flow is the basis for many types of channel design; the subject of design is taken up in
detail in Chapter 8. The concept of critical flow is fundamental to the analysis of var-
ied flows in open channels and to understanding many types of flow controls. Estab-
lishing critical flow in an open channel can also provide a basis for flow
measurement.
The momentum principle, as applied in the analysis of open-channel flow, is taken up
in Section 7.5. As in pipe flow, the momentum principle is applied in situations where
the engineer desires an estimate of the hydraulic forces acting on a flow obstruction
such as a bridge pier. The momentum principle is also fundamental to the analysis of
the rapidly varied flow phenomenon known as the hydraulic jump.
Sections 7.6 and 7.7 address the topics of weirs and orifices, respectively, which are
commonly used structures for flow control and measurement. Weirs and orifices may
be used as outlets for stormwater detention facilities. The weir and orifice equations
also play a role in describe the hydraulic performance of stormwater inlets and culvert
entrances.
Gradually varied flow is discussed in Section 7.8. Gradually varied flows are those in
which the depth of flow changes along the length of a channel as a consequence of
convective accelerations of the flow. These accelerations occur where flow controls
are such that the flow depth at the control is different from the depth of uniform flow
and forces on the water are not balanced. Such situations are common upstream of
culvert entrances, at channel entrances and exits, and at locations where channel geo-
metric properties change.
Spatially varied flows—flows for which the discharge is not a constant along the
length of a channel—are discussed briefly in Section 7.9. These flows may have
either an increasing or a decreasing discharge in the direction of flow, and different
methods exist for the analysis of each flow type. Spatially varied flows with a
decreasing discharge occur, for example, in gutters at stormwater inlet locations. Spa-
tially varied flows with an increasing discharge may occur in side overflow spillway
channels. Flow in a natural stream channel is usually spatially varied because of the
increase in contributing drainage basin area as one proceeds downstream.
7.1 CLASSIFICATION AND PROPERTIES OF OPEN

CHANNELS
Open channels may be natural or man-made and come in a wide variety of cross-sec-
tional shapes. For analytical purposes, channels may be classified as prismatic or non-
prismatic. In a prismatic open channel, the descriptive geometric variables such as its
width, side slopes, and longitudinal slope, do not change along the length of the chan-
nel. Practically speaking, one would typically not expect to encounter a perfectly pris-
matic channel outside of the laboratory, but many constructed channels can, for the
purposes of analysis, be considered prismatic. A concrete curb and gutter along the
edge of a paved roadway with a constant slope is one example of a prismatic channel.
Another example is shown in Figure 7.1.
Section 7.1 Classification and Properties of Open Channels 219
Figure 7.1
A riprap-lined
prismatic channel
In a nonprismatic channel the descriptive geometric variables (shape, roughness, etc.)

change with location along the channel length. Natural stream channels, which typi-
cally have variations in width and other properties, are examples of nonprismatic open
channels (see Figure 7.2). Coverage of open-channel flow in this text deals primarily
with prismatic channels. Nevertheless, certain principles associated with the hydrau-
lics of flow in nonprismatic channels will be addressed as necessary.
Open channels can also be classified on the basis of whether they are lined or unlined.
In a lined channel, the channel bottom and sides have been paved with an erosion-
resistant material, such as concrete, asphalt, metal, riprap, or vegetation. An unlined
channel is one in which the channel bottom and sides consist of natural geologic
materials such as earth, gravel, or rock. As shown in Chapter 8, the lining material (or
lack thereof) has practical significance in the design of the open channel.
Geometric Properties of Open Channels

The longitudinal slope of an open channel, denoted by So, is a critically important
geometric variable. An alternative way of describing the channel slope is by use of its
angle of inclination above the horizontal, denoted by θ (see Figure 7.3). Note that So
= tanθ. Note also that if θ is small (less than about 5 degrees), then So ≈ sinθ. A hori-
zontal open channel has a longitudinal slope of zero, and an adverse open channel has
a negative slope (that is, the channel bottom elevation increases in the downstream
direction).
Figure 7.2
A nonprismatic
natural channel
Figure 7.3
Depth of flow
measurements
Most geometric properties of open channels are those related to a channel’s cross sec-
tion. Strictly speaking, channel cross sections should be taken perpendicular to the
channel bottom (that is, normal to the direction of flow). The depth of flow measured
Section 7.1 Classification and Properties of Open Channels 221
perpendicular to the channel bottom is denoted by d in Figure 7.3. However, in most

open channels, the slope is small, and vertical channel cross sections can be employed
with little error. The depth of flow measured vertically is denoted by y (see Figure
7.3). The relationship between these depths is d = ycosθ.
The most common cross-sectional shapes of prismatic channels are rectangular, tri-
angular, trapezoidal, and circular. Other shapes are those associated with curbs and
gutters, which are discussed in Section 10.1. Figure 7.4 illustrates the dimensions
used to describe common channel shapes. B denotes the channel bottom width, y
denotes the depth of flow, D denotes the diameter of the channel (if a circular pipe), T
denotes the top width of flow, and z represents a measure of steepness of the channel
sides. For circular channels, φ represents the angle subtended by the intersections of
the water surface with the channel or conduit surface, as shown in the figure. Note
that rectangular and triangular channels are merely special cases of a trapezoidal
channel, the former in the case of z = 0, and the latter in the case of B = 0.
The cross-sectional area, wetted perimeter, hydraulic radius, top width T, and hydrau-
lic depth can be determined on the basis of the depth of flow y and the properties
shown in Figure 7.4. The hydraulic radius is defined as
R = A/P (7.1)
where R = hydraulic radius (ft, m)

The definition of hydraulic depth is
Dh = A/T (7.2)
where Dh = hydraulic depth (ft, m)

T = top width (ft, m)
Table 7.1 summarizes the relationships for each of the channel cross-sectional shapes
shown in Figure 7.4. Note that the formulas from triangular and trapezoidal channels
are only applicable to cross sections with equal side slopes, and the depth y in an open
channel is always taken as the greatest depth at any point in a cross section. The
hydraulic depth can be thought of as the average depth of flow in a cross section. Use
of the hydraulic depth can also be viewed as a type of shorthand, as the ratio Dh = A/T
occurs frequently in equations describing open-channel flow.
The cross-sectional properties of irregularly shaped channels can be determined by
dividing a cross section into a set of rectangles and/or triangles and applying basic
geometric formulae related to areas, perimeters, and slopes. Clearly, this process is
quite tedious and time-consuming; for that reason, computer-based methods for solv-
ing open-channel flow problems are especially useful for irregularly shaped channels.
Figure 7.4
Dimensions of
common channel
shapes
Table 7.1 Cross-sectional properties of prismatic open channels
Channel Shape Area (A) Wetted Perimeter (P) Hydraulic Radius (R) Top Width (T) Hydraulic Depth (Dh)
By
Rectangular By B + 2y B y
B + 2y
By + zy 2
Trapezoidal By +zy 2 B +2 y 1 + z 2 B +2 y 1 + z 2 B + 2zy
B + 2 zy
zy y
Triangular zy2 2 y 1 + z2 2zy
2 1 + z2 2
D 2 (φ − sin φ ) Dφ D  sin φ  φ  D  φ − sin φ 

Circulara 1 −  D sin    
8 2 4 φ  2 8  sin(φ / 2) 
a. Angle φ is measured in radians. 1 radian ≈ 57.3 degrees.

Section 7.2 Energy and Constitutive Relationships 223
Example 7.1 – Geometric Properties for a Trapezoidal Channel. An open channel

has a cross-sectional shape as shown in Figure E7.1.1. If the depth of flow is 4 ft, determine the cross-
sectional area, wetted perimeter, hydraulic radius, top width, and hydraulic depth.
Figure E7.1.1 Open channel for Example 7.1
Solution: Because the cross-sectional shape of the channel does not coincide with any of those shown
in Figure 7.4, the properties of the channel section must be derived on the basis of its geometric prop-
erties.
The cross-sectional area can be constructed as the sum of the areas of a rectangle and a triangle, and is
A = 12(4) + 0.5(4)[2(4)] = 64 ft2

The wetted perimeter can be computed as the sum of the water depth against the right-hand side of the
channel, the channel bottom width, and the length along the sloping side of the channel:
P = 4 + 12 + 42 + 82 = 24.94 ft
The hydraulic radius is
R = A/P = 64/24.94 = 2.57 ft
The top width is equal to
T = 12 + 2(4) = 20 ft
and the hydraulic depth is
Dh = A/T = 64/20 = 3.20 ft
7.2 ENERGY AND CONSTITUTIVE RELATIONSHIPS

In open-channel flow problems, the energy equation takes a somewhat different form
than that presented in Equations 3.9 and 6.1. The difference arises in the pressure
head term, which, instead of being written as p/γ, is written as dcosθ, where d is the
depth of flow normal to the channel bottom and θ is the angle of inclination of the
channel. Thus, the energy equation for flow in an open channel is
α V2 α V2
Z1 + d1 cos θ + 1 1 = Z 2 + d 2 cos θ + 2 2 + hL (7.3)
2g 2g
where Z = channel bottom elevation (ft, m)

d = depth of flow normal to the channel bottom (ft, m)
θ = channel slope angle
α = a velocity distribution coefficient defined by Equation 6.3
V = average flow velocity (ft/s, m/s)
g = gravitational acceleration constant (ft/s2, m/s2)
hL = energy loss between section 1 and section 2 (ft, m)
Figure 7.5 depicts the terms in Equation 7.3, as well as the next two equations.
Table 7.2 gives values of α and β (for an explanation of β, see Section 7.5) for open-
channel flows. (Estimation of α for compound channel sections is presented in Sec-
tion 7.3.)
Because d = ycosθ (see Figure 7.3), an alternative expression for the energy equation
for open-channel flow is
α V2 α V2
Z1 + y1 cos 2 θ + 1 1 = Z 2 + y2 cos 2 θ + 2 2 + hL (7.4)
2g 2g
where y1 and y2 are the flow depths (measured vertically) at sections 1 and 2, respec-
tively (ft, m).
When the angle of inclination, θ, is small, this equation simplifies to the version most
commonly applied in practice:
α V2 α V2
Z1 + y1 + 1 1 = Z 2 + y2 + 2 2 + hL (7.5)
2g 2g
Figure 7.5
Definition figure for
energy equation for an
open channel
Table 7.2 Typical values of α and β in open-channel flow (Chow, 1959)
α β
Channel Description
Min Avg Max Min Avg Max
Regular channels, flumes, spillways 1.10 1.15 1.20 1.03 1.05 1.07
Natural streams and torrents 1.15 1.30 1.50 1.05 1.10 1.17
Rivers under ice cover 1.20 1.50 2.00 1.07 1.17 1.33
River valleys, over flooded 1.50 1.75 2.00 1.17 1.25 1.33
Pressure Variations in Open-Channel Flow

Replacement of the pressure head term, p/γ, with dcosθ in Equation 7.3 means that the
fluid pressure at depth d in open-channel flow is equal to
p = γ d cos θ (7.6)
where p = fluid pressure at depth d (lb/ft2, N/m2)

d = depth measured normal to the flow streamlines (ft, m)
γ = fluid specific weight (lb/ft3, N/m3)
If one desires instead to use the vertical depth, y, the pressure is found to be
p = γ y cos 2 θ (7.7)
Because cosθ is less than unity in all nonhorizontal channels, this expression indicates
that the pressure at depth y in an open-channel flow is less than the pressure at the
same depth in a static fluid.
Example 7.2 – Pressure Variation over a Spillway. Flow occurs on a spillway with a
depth d = 1.00 m. The slope of the spillway is 15 percent. Determine the fluid pressure on the face of
the spillway and compare that pressure to the pressure existing at a depth of 1.00 m in standing water.
Solution: At a depth of 1.00 m in standing water, the pressure is found to be
p = γ d = 1,000(1.00) = 1,000 kg/m2

The pressure at the same depth on the face of the spillway can be evaluated using Equation 7.6, where
θ = arctanSo = arctan(0.15) = 8.53°.
p = 1,000(1.0)(cos 8.53°) = 988.9 kg/m2
The pressure reduction caused by the slope of the channel is slightly more than 1 percent in this case.
Constitutive Relationships for Open-Channel Flow

Energy losses represented by the hL term in the energy equation can be viewed as con-
sisting of localized and distributed losses. Local losses, often referred to as minor
losses, occur at locations where the flow velocity changes in either magnitude or
direction. They are described later in this section on page 230. Distributed losses are
due to friction exerted on the flow by the channel bottom and sides. Distributed losses
may be described by expressing the head loss for a channel reach as the product of the
energy gradient, Sf, and the length, L, of the reach:
hL = Sf L (7.8)
where Sf = energy gradient (ft/ft, m/m)

L = reach length (ft, m)
A number of constitutive relationships are available to express the energy gradient as
a function of flow and channel characteristics. The most commonly used relationships
are described in the subsections that follow.
The Manning Equation. The Manning equation (Manning, 1891) is an empiri-
cally derived expression obtained from observations of flows in laboratory channels.
It was developed as a formula for uniform flow but has also been successfully applied
for analyses of gradually and spatially varied flows. It is the most widely used of all
constitutive relationships for open-channel–flow analysis.
The Manning equation can be written as
Cf
V= R 2 / 3 S1/
f
2
(7.9)
n
where V = cross-sectionally averaged flow velocity (ft/s, m/s)

Cf = unit conversion factor (1.49 for US customary units, 1.00 for SI)
n = Manning’s roughness coefficient
R = hydraulic radius of the flow (ft, m)
Note that Sf is equal to the channel slope So in the case of uniform flow, and is differ-
ent from So in the case of varied flow. The numerical value of Manning’s n is indepen-
dent of the set of units employed.
Table 7.3 provides a listing of typical values of Manning’s n for various channel types
and surfaces. Note that n is not a constant for a particular channel; it depends on a
number of factors, including flow depth, channel alignment and irregularity, the pres-
ence of vegetation and flow obstructions, and the amount of sediment being trans-
ported by the stream. Chow (1959) provides a detailed discussion of the effects of
these factors.
Because computations of flow velocities and energy gradients via the Manning equa-
tion are fairly sensitive to the assumed value of n, an understanding of influential fac-
tors should enable one to make more informed choices regarding this important
parameter. For example, Chow (1959), in a theoretical discussion of channel rough-
ness and uniform flow formulas, showed that the value of Manning’s n is related to
the channel roughness height as:
R
n = ϕ   k1/ 6 (7.10)
k
where ϕ(R/k) = a value that is a function of the dimensionless ratio of the hydraulic
radius R to the roughness height k
k = channel roughness height (ft, m)
Table 7.3 Typical values of Manning’s n (adapted from Chow, 1959)

Channel Description Minimum n Average n Maximum n
Closed conduits flowing partly full
Brickwork 0.011 0.014 0.017
Cast iron, coated 0.010 0.013 0.014
Cast iron, uncoated 0.011 0.014 0.016
Clay, vitrified 0.011 0.014 0.017
Concrete, finished 0.011 0.012 0.014
Concrete, unfinished 0.012 0.014 0.020
Corrugated metal 0.021 0.024 0.030
Rubble masonry 0.018 0.025 0.030
Steel, welded 0.010 0.012 0.014
Steel, riveted and spiral 0.013 0.016 0.017
Wood stave 0.010 0.012 0.014
Lined or built-up channels
Asphalt 0.013 0.016 —
Brickwork 0.011 0.014 0.018
Concrete, troweled 0.011 0.013 0.015
Concrete, unfinished 0.014 0.017 0.020
Corrugated metal 0.021 0.025 0.030
Rubble masonry 0.017 0.029 0.035
Steel, smooth 0.011 0.013 0.017
Wood, planed and creosoted 0.011 0.012 0.015
Wood, planed and untreated 0.010 0.012 0.014
Excavated or dredged channels
Earth, straight and uniform 0.016 0.022 0.033
Earth, winding and sluggish 0.023 0.035 0.050
Overgrown with vegetation 0.040 0.080 0.140
Rock cuts, smooth and uniform 0.025 0.035 0.040
Rock cuts, jagged and irregular 0.035 0.040 0.050
Natural streams
Flood plains, pasture 0.025 0.033 0.050
Flood plains, cultivated 0.020 0.035 0.160
Flood plains, trees 0.050 0.080 0.200
Major streams (width > 100 ft) 0.025 — 0.100
Minor streams on plain areas 0.025 0.045 0.150
Mountain streams 0.030 0.045 0.070
The value of the function ϕ(R/k) does not vary much over a wide range of R/k. Strick-
ler (1923) showed that, on average, ϕ(R/k) = 0.0342 when the roughness height (in ft)
is taken as the median grain size of the channel bed material. (When the median grain
size is expressed in units of meters, ϕ(R/k) = 0.0281.) This relationship indicates that
Manning’s n is proportional to the one-sixth power of the median sediment size in a
stream.
Example 7.3 – Determining the Energy Gradient with the Manning Equation.
A discharge of 300 cfs occurs at a depth of 2.50 ft in a concrete-lined rectangular open channel with a
width of 10 ft and n = 0.013. What is the energy gradient Sf for this flow? If the channel bottom slope
is So = 0.001, is this flow uniform or varied?
Solution: The cross-sectional area, wetted perimeter, and hydraulic radius of the flow are
A = 10(2.50) = 25 ft2
P = 10 + 2(2.50) = 15 ft
R = A/P = 25/15 = 1.67 ft
V = Q/A = 300/25 = 12 ft/s
Substituting these values into the Manning equation and rearranging yields the energy gradient:
(nV )2 [0.013(12)]2
Sf = 2 4/3
= = 0.00555
1.49 R 1.492 (1.67) 4 / 3
Because the energy gradient is not equal to the channel slope, the flow must be varied.
The Chézy Equation. The Chézy equation, developed around 1769 by the
French engineer Antoine Chézy, is probably the oldest constitutive relationship for
open-channel flow (Herschel, 1897). Unlike the Manning equation, the Chézy equa-
tion can be derived theoretically by equating the force resisting a flow to the gravita-
tional force driving the flow (a condition that must be true for uniform flow). Because
the resisting force is proportional to the square of the flow velocity, the net result is
the expression
V = C RS f (7.11)
where C = a roughness coefficient known as Chézy’s C (ft0.5/s, m0.5/s)

Like Manning’s n, Chézy’s C depends on a number of factors and thus is not a con-
stant. In 1869, the Swiss engineers Ganguillet and Kutter (1869) published a formula
relating C to the roughness, slope, and hydraulic radius of a channel as
0.00281 1.811
41.65 + +
C= S n
0.00281  n (7.12)

k +  41.65 + 
 S  R
where S = channel slope (ft/ft, m/m)

n = Kutter’s n
k = a constant equal to 1 for U.S. Customary units or 1.811 for SI units
For practical purposes, Kutter’s n is the same as Manning’s n when the channel slope
is larger than 0.0001 and the hydraulic radius is between 1 and 30 feet (0.3 and 9 m).
Bazin (1897) published a formula for C in which slope is not considered. His equation
is
157.6
C=
m (7.13)
k+
R
where m = a roughness coefficient (see Table 7.4)
Table 7.4 Proposed values of Bazin’s m (Chow, 1959)

Channel Description m
Very smooth cement or planed wood 0.11
Unplaned wood, concrete or brick 0.21
Rubble masonry or poor brickwork 0.83
Earth channels in good condition 1.54
Earth channels in ordinary condition 2.36
Earth channels in rough condition 3.17
Although considerably simpler than the Ganguillet and Kutter (1869) formula,
Bazin’s is considered to be less reliable.
Powell (1950) suggested that Chézy’s C can be expressed as
 C ε
C = −42 log10  +  (7.14)
 4 Re R 
where C = Chézy’s C (ft0.5/s)

Re = Reynolds number
ε = measure of roughness having the tentative values shown in Table 7.5
R = hydraulic radius (ft)
For open-channel flow, the Reynolds number can be computed as
VR ρVR
Re = = (7.15)
ν µ
where ν = fluid kinematic viscosity (ft2/s, m2/s)

µ = fluid dynamic viscosity (lbf-s/ft2, N-s/m2)
The reader may want to compare Equation 7.15 to Equation 6.5, which is used to
compute the Reynolds number in the case of a circular pipe.
Table 7.5 Tentative values of Powell’s ε (Chow, 1959)

.
Channel Description ε
Neat cement 0.0002–0.0004
Unplaned plank flumes 0.0010–0.0017
Concrete-lined channels 0.004–0.006
Earth, straight and uniform 0.04
Dredged earth channels 0.10
The Powell formula, being an implicit function of C, requires a trial-and-error solu-

tion and therefore is less convenient than other approaches. It should be used with
caution, as further research is required for determination of proper values of the
roughness parameter ε.
Example 7.4 – Determining the Energy Gradient with Chézy’s Equation. Repeat
Example 7.3 using the Chézy equation to compute the energy gradient. Use the Ganguillet and Kutter
formula to determine the value of Chézy’s C. In Example 7.3, a discharge of 300 cfs occurs at a depth
of 2.50 ft in a concrete-lined rectangular open channel with a width of 10 ft and n = 0.013.
Solution: Using n = 0.013 and R = 1.67 ft from Example 7.3, the Ganguillet and Kutter formula may
be stated as
0.00281
180.96 +
Sf
C=
0.0000283
1.42 +
Sf
Substitution of this expression, V = 12 ft/s, and R = 1.67 ft into Equation 7.11 leads to an expression
that is implicit in Sf. Solving by trial and error,
Sf = 0.00534
Again, because the energy gradient is not equal to the channel slope, the flow must be varied.
Minor Losses in Open-Channel Flow

As indicated in Section 6.3, the classical method of predicting minor energy losses
caused by flow expansions and contractions is to express the energy loss as being pro-
portional to the absolute difference between the upstream and downstream velocity
heads:
V22 − V12
hL = K (7.16)
2g
where hL = energy loss (ft, m)

K = minor loss coefficient
Section 7.3 Uniform Flow and Normal Depth 231
The minor loss coefficient, K, depends on the geometry of the flow contraction or
expansion. For channel transitions such as entrances and exits to and from flumes,
tunnels, or siphons, Chow (1959) presents the recommended design values shown in
Table 7.6. Note that the loss coefficient for a flow expansion is always greater than the
coefficient for a flow contraction.
Table 7.6 Minor loss coefficients for open-channel flow (Chow, 1959)
Type of Channel Transition Contraction Expansion
Cylinder quadrant 0.15 0.25
Square-ended 0.30 or more 0.75
Straight line 0.30 0.50
Warped 0.10 0.20
In non-prismatic channels, where the cross-sectional area and hence the flow velocity
may change from one cross section to another, eddy losses occur and can be predicted
using Equation 7.16. For gradually converging and diverging reaches, respectively,
recommended values of K are about 0.1 and 0.3 [Chow, 1959; U.S. Army Corps of
Engineers (USACE), 1991].
7.3 UNIFORM FLOW AND NORMAL DEPTH

Uniform flow in an open channel occurs when the depth, velocity, cross-sectional
area, and discharge are constant along the channel length. As a practical matter, this
condition can exist only in a prismatic channel, and only if the flow is steady. When
uniform flow does exist, the constancy of the flow variables (such as velocity and
depth) can be used to recognize that the slopes of the channel bottom (So), water sur-
face (Sw), and energy grade line (Sf) are all equal.
Uniform flow occurs in a channel reach between two cross sections when all forces
acting on the water in the reach are in balance and there is no flow acceleration. With
reference to Figure 7.6, these forces are the hydrostatic pressure forces, P1 and P2,
acting at the upstream and downstream ends of the reach, the frictional force, Ff,
opposing the flow, and the gravitational force, Wsinθ (where W is the weight of the
water), that is driving the flow. Because the flow is uniform, P1 and P2 must be equal
and the frictional force must be equal to the driving force, or Ff = Wsinθ.
If the frictional force is expressed as the product of a shear stress and the area of the
channel bottom and sides to which that stress is applied, one can write Ff = τPL,
where τ is the shear stress, P is the wetted perimeter of the flow, and L is the length of
the channel reach (Figure 7.6). Similarly, if the weight of the water is expressed as
W = γAL, where A is the cross-sectional flow area, and if the slope So is approximated
by sinθ, then Wsinθ = γALSo. Equating the frictional and driving forces leads to Equa-
tion 7.17:
τ = γRSo (7.17)
where τ = average shear stress (lb/ft2, N/m2)

γ = fluid specific weight (lb/ft3, N/m3)
Thus, when flow is uniform, Equation 7.17 gives the average shear stress acting on
the channel bed and sides. This parameter is important in the design of unlined chan-
nels, as it has a direct bearing on whether the flow will mobilize soil particles on the
channel boundaries.
Other than its use for the derivation of Equation 7.17, the concept of the force balance
that must exist in uniform flow illustrates that uniform flow represents a condition of
equilibrium. The depth at which this equilibrium condition is reached is called the
normal depth and is denoted by yn. The normal depth depends on the discharge, chan-
nel cross-sectional geometry, channel roughness, and channel slope.
If a flow control exists that causes the actual depth of flow at a cross section to be dif-
ferent from the normal depth, then a varied flow profile will exist as a transition
between the depth at the flow control and the normal depth in the channel. In other
words, despite the existence of the flow control, there will be a region of the channel
in which the flow adjusts itself to seek out its equilibrium condition. Depending on
the length of the channel, the flow may or may not have the opportunity to actually
attain its normal depth condition.
Figure 7.6
Forces on a fluid
element for a uniform
flow condition
Uniform Flow in Simple Channels

A simple open channel is one with constant roughness and without overbank (or
floodplain) areas. For the purpose of this text, a simple channel will also be taken as
one without a gradually closing top, such as occurs in a pipe. In a simple open chan-
nel, the relationships between flow variables under conditions of uniform flow are
given by a constitutive relationship for which it is assumed that the energy gradient Sf
is equal to the channel bottom slope So. As discussed in Section 7.2, the most com-
mon constitutive relationships are the Manning and Chézy equations.
If either the Manning or the Chézy equation is multiplied by the cross-sectional area A
of flow, it can be expressed in terms of the discharge rather than the velocity of the
flow. That is,
Cf
Q= AR 2 / 3 So1/ 2 (7.18)
n
where Q = discharge (cfs, m3/s)

Cf = is a conversion factor equal to 1.49 for U.S. Customary units or 1.00 for
SI units
A = cross-sectional area of flow (ft2, m2)
or
Q = CA RSo (7.19)
where C = Chézy’s C (ft0.5/s, m0.5/s)

It can be seen from these expressions that, for a simple channel of fixed slope and dis-
charge, there is a unique value of flow depth (recall that R and A depend on depth)
that will balance each equation. That unique depth is the normal depth yn. Similarly, if
the slope and normal depth are fixed, there is a unique discharge that results. Finally,
if the discharge and normal depth are fixed, there is a unique channel slope that results
(recognizing that the slope represented by the negative square root can be disre-
garded). That slope is called the normal slope of the channel for the given discharge
and depth, and is denoted by Sn.
Computation of uniform flow involves determining the depth, discharge, or slope that
satisfies Equation 7.18 or 7.19 when the other variables are specified. As illustrated in
the following examples, determination of the normal depth generally requires a trial-
and-error solution. Determination of either the discharge or the normal slope can be
accomplished without iterating.
Example 7.5 – Determining Discharge with the Manning Equation. A trapezoidal

open channel has a bottom width B = 10 ft, side slopes with z = 3 (see Figure E7.5.1), a slope So =
0.002, and n = 0.035. Determine the discharge in the channel if the flow is uniform and its depth is
3.20 ft. Use the Manning equation.
Figure E7.5.1 Trapezoidal channel for Example 7.5
Solution: From the expressions given in Table 7.1, the cross-sectional area and wetted perimeter of
the channel can be computed as
A = 10(3.20) + 3(3.20)2 = 62.72 ft2
P = 10 + 2(3.20)(1 + 32)1/2 = 30.24 ft

The hydraulic radius is therefore R = 62.72/30.24 = 2.07 ft. Substitution into Equation 7.18 yields
1.49
Q= (62.72)(2.07)2 / 3 (0.002)1/ 2 = 194 cfs
0.035
Example 7.6 – Determining Normal Slope with the Manning Equation. Deter-
mine the normal slope of the channel in Example 7.5 if the discharge is Q = 350 cfs.
Solution: By algebraic rearrangement, Equation 7.18 can be expressed in the form
2
 nQ 
Sn =  2/3 
 1.49 AR 
Substituting the given values, the normal slope is
Sn = 0.00651
Example 7.7 – Determining Normal Depth with the Manning Equation. A rect-
angular open channel has a width of B = 3.6 m, n = 0.013, and So = 0.015. Determine the normal
depth of flow if the discharge is Q = 7.0 m3/s.
Solution: The cross-sectional area of flow can be written as
A = 3.6y
The wetted perimeter can be written similarly as
P = 3.6 + 2y
Substitution of these expressions into Equation 7.18 yields
1.00 5⁄3 –2 ⁄ 3 1⁄2
Q = 7.0 = ------------- ( 3.6y ) ( 3.6 + 2y ) 0.015
0.013
or
5⁄3 –2 ⁄ 3
( 3.6y ) ( 3.6 + 2y ) = 0.74
By trial and error, the value of y that balances this equation is yn = 0.42 m.
Uniform Flow in a Channel with a Gradually Closing

Top
A pipe, whether it is circular, elliptical, or another shape, is an example of a channel
with a gradually closing top. In a channel with a gradually closing top, there may be
more than one depth of flow that satisfies the normal depth criterion of equilibrium.
Furthermore, the maximum discharge that can be conveyed by the channel under uni-
form flow conditions occurs at a depth of flow smaller than that corresponding to full
flow in the conduit.
Figure 7.4 on page 222 illustrates a circular pipe of diameter D in which the depth of
flow is denoted by y. The angle φ defined by the edges of the free water surface is also
shown. Note that φ = 0 when y = 0, and that φ = 2π radians when y = D. The relation-
ship between y, D, and φ (in radians) is given by
D  φ 
y= 1 − cos    (7.20)
2  2 
From Table 7.1, the cross-sectional area and wetted perimeter are given by
D2
A= (φ − sin φ ) (7.21)
8
Dφ
P= (7.22)
2
Substitution of these expressions into the Manning equation yields
5/3 −2 / 3
C f  D2   Dφ 
Q=  (φ − sin φ )  2  So1/ 2 (7.23)
n  8   

Cf = unit conversion factor (1.49 for US customary units, 1.00 for SI)
n = Manning’s n
D = diameter of pipe (ft, m)
φ = angle defined by edges of free water surface of circular channel (see
Figure 7.4)
For chosen values of D, n, and So, Equation 7.23 may be applied to develop a graph of
Q as a function of φ (or as a function of y by rearranging Equation 7.20 to solve for φ
and substituting this expression into Equation 7.23). Thus, φ (or y) could be obtained
from the graph if Q were specified. As shown in Figure 7.7, which illustrates such a
graph, two possible depths of flow (y-values) exist for any possible discharge greater
than or equal to the full-flow discharge (that is, the discharge when y = D). It can also
be seen from the graph that the maximum discharge occurs when y ≈ 0.94D.
Although Figure 7.7 is useful for demonstrating that there may be more than one pos-
sible depth of uniform flow in a channel with a gradually closing top, it was devel-
oped for the case of a specific pipe diameter, roughness, and slope. Figure 7.8 depicts
a more useful dimensionless graph that can be used for any combination of pipe geo-
metric variables. In this figure, the relative depth of flow, y/D, is shown as a function
of the ratio of actual flow to full flow and the ratio of actual velocity to full velocity.
Example 7.8 illustrates the use of Figure 7.8.
Figure 7.7
Graph of discharge
versus normal depth
for a representative 1-
ft diameter pipe
Figure 7.8
Dimensionless graph
of discharge and
velocity fractions
versus depth fraction
for circular pipe
Example 7.8 – Depth and Velocity in a Circular Pipe. A 36-in. diameter pipe with
n = 0.013 and So = 0.01 conveys a discharge of 45 cfs. Determine the depth and velocity of flow in the
pipe. If the pipe flows at its maximum capacity (y ≈ 0.94D), what is the discharge?
Solution: For full-flow conditions, the pipe’s cross-sectional area and wetted perimeter are, respec-
tively, 7.07 ft2 and 9.42 ft. Applying the Manning equation, the full-flow discharge is Qf = 67 cfs. The
full-flow velocity is
Vf = 67/7.07 = 9.46 ft/s
The actual discharge in the pipe is Q = 45 cfs, so
Q/Qf = 45/67 = 0.67
From Figure 7.8, the relative depth of flow is y/D = 0.60. Thus, the actual depth of flow is
D(y/D) = 3(0.60) = 1.80 ft.
Using y/D = 0.60, Figure 7.8 shows that V/Vf = 1.09. Thus, the actual flow velocity is
Vf (V/Vf ) = 9.46(1.09) = 10.3 ft/s
At maximum capacity, the actual discharge is about 1.09 times the full-flow discharge (see Figure
7.8). Therefore, the maximum discharge is
1.09(67) = 73 cfs
Dimensionless graphs of geometric elements similar to the one shown in Figure 7.8
can be developed for pipe shapes other than circular. The Concrete Pipe Design Man-
ual (American Concrete Pipe Association, 2000) and similar publications developed
by associations of pipe manufacturers may be consulted to obtain graphs for other
common pipe shapes.
Uniform Flow in a Channel with Composite Roughness

In a simple channel with composite roughness, the roughness characteristics vary
from one part of the channel boundary to another. For example, a constructed pris-
matic channel might have an earthen bottom and riprap side slopes. A laboratory
channel might have a steel bottom and glass sides. In such channels, a uniform flow
formula such as the Manning equation can be applied if the n-value used is an equiva-
lent n-value computed to represent an average measure of the channel roughness.
There are several ways in which an equivalent n value may be determined. As shown
in Figure 7.9, a channel cross section may be subdivided into a number of subareas N,
each of which has its own roughness ni, area Ai, wetted perimeter Pi, hydraulic radius
Ri, and average depth yi, where i = 1, 2 … N. The wetted perimeter of each subarea is
computed using only that part of the perimeter that coincides with the channel bottom
or sides.
Figure 7.9
Division of channel
for computation of the
equivalent roughness
coefficient
If one assumes that the velocity in each subarea is equal to the mean velocity for the
cross section as a whole, the equivalent n value may be determined as (Horton, 1933;
Einstein, 1934)
2/3

ne = 
∑ Pi ni3 / 2  (7.24)
 ∑ Pi 
where ne = equivalent Manning’s n
Pi = wetted perimeter for the i-th channel subarea (ft, m)
ni = Manning’s n for the i-th channel subarea
If one assumes that the total force resisting the flow is equal to the sum of the resisting
forces in each of the subareas, the equivalent n value is (Pavlovskii, 1931; Muhlhofer,
1933; Einstein and Banks, 1951)
1/ 2

ne = 
∑ Pi ni2  (7.25)
 ∑ Pi 
Lotter (1933) showed that the equivalent n value could be expressed as the following,
assuming that the total discharge is equal to the sum of the discharges in the individ-
ual subareas:
PR5 / 3
ne =
 P R5 / 3  (7.26)
∑  i i 
 ni 
 
Finally, if the velocity distribution in the cross section is assumed to be logarithmic,

the equivalent n value is (Krishnamurthy and Christensen, 1972)
ln ne =
∑ Pi yi3 / 2 ln ni (7.27)
∑ Pi yi3 / 2
where yi = average depth of the i-th subarea
In an evaluation of these alternative expressions based on data from 36 actual channel
cross sections, Motayed and Krishnamurthy (1980) concluded that Equation 7.24 per-
formed the best.
Example 7.9 – Composite Roughness in a Rehabilitated Channel. Figure E7.9.1

illustrates the cross section of an old masonry tunnel whose hydraulic performance is to be improved
by paving the invert with finished concrete. If the depth of flow in the tunnel is 4 ft both before and
after the rehabilitation, determine the amount by which the tunnel discharge capacity increases fol-
lowing the change. The tunnel slope is So = 0.002, and the n values for rough masonry and finished
concrete are 0.030 and 0.013, respectively. Use Equation 7.24 to determine the equivalent n value fol-
lowing the rehabilitation.
Figure E7.9.1 Tunnel cross section for Example 7.9

Solution: In both cases, the cross-sectional area of flow is A = 16 ft2 and the wetted perimeter is P =
12 ft. The tunnel capacity prior to the change is
1.49
Q= (16)5 / 3 (12)−2 / 3 (0.002)1/ 2 = 43 cfs
0.030
The equivalent n value following the rehabilitation is
{4(0.030) }
2/3
3/ 2
+ 4(0.013)3/ 2 + 4(0.030)3/ 2
n =
e = 0.025
122 / 3
Using this equivalent value, the capacity following the rehabilitation is
1.49
Q= (16)5 / 3 (12)−2 / 3 (0.002)1/ 2 = 52 cfs
0.025
In this instance, the capacity of the tunnel is increased by over 20 percent.
Uniform Flow in Compound Cross-Sections

Figure 7.10, which is a simplified representation of a flooded river valley, illustrates a
compound channel cross-section. It consists of a main channel and overbank (or
floodplain) areas on either side of the main channel that are inundated by high flows.
Roughness is typically higher in the overbank areas than in the main channel, with the
effect that flow velocities are typically lower in the overbank areas than in the main
channel. This condition causes either of two possible problems to arise in the analysis
of the flow.
Figure 7.10
A compound channel
cross-section
To illustrate the first type of problem, assume that at a given cross section the water
surface elevations are the same in both the main channel and the overbank areas.
Under this assumption, there must be three different energy grade line elevations
within the cross section: one for the main channel and one for each overbank area.
The multiple energy grade lines occur because the velocities, and hence the velocity
heads, are different in each area of the cross section.
Alternatively, one might assume that the energy grade line elevation at a cross section
is the same for all three parts of the flow area. In this case, again because of velocity
differences, one comes to realize that the water surface elevations must be different in
the main channel and overbank areas.
To resolve these issues and arrive at both a single water surface elevation for a cross
section and a single energy grade line elevation, the velocity distribution in the cross
section must be accounted for. The velocity distribution coefficient α is used for this
purpose.
If a channel cross section is subdivided into N subareas, each of which has an area Ai
and average velocity Vi, where i = 1, 2 … N, then α can be computed using an approx-
imation of Equation 6.3:
α=
∑ AiVi3 (7.28)
AVm3
where α = velocity distribution coefficient

Ai = cross-sectional area of the i-th subarea (ft2, m2)
Vi = average velocity of the i-th subarea (ft/s, m/s)
A = total cross-sectional area (ft2, m2)
Vm = the mean velocity in the entire cross section (ft/s, m/s)
Vm is computed as
1
Vm =
A
∑ AiVi (7.29)
Using these expressions, the velocity head at a cross section may be expressed as
αVm2/2g.
The total discharge at a cross section may be computed as
Q = ∑ Qi = ∑ AiVi (7.30)
where Q = total discharge at a channel cross section (cfs, m3/s)

Qi = discharge at a cross-sectional subarea (cfs, m3/s)
If one assumes that the energy gradient Sf is equal for each subarea, it may be shown
that
Qi = κ i S1/
f
2 (7.31)
where κi = the conveyance of the i-th subarea (cfs, m3/s)

and
Q = κ S 1/f 2 (7.32)
where
κ = ∑ κi (7.33)
In these expressions, κi denotes the conveyance of the i-th subarea of the cross sec-
tion. If the Manning equation (Equation 7.9) is employed as the constitutive relation-
ship for the flow, then the conveyance can be written as
Cf
κi = Ai Ri2 / 3 (7.34)
ni
where Cf = unit conversion factor (1.49 for U.S. customary units, 1.00 for SI)
Dividing Equation 7.31 by 7.32, one obtains
κi
Qi = Q (7.35)
∑ κi
This expression shows that the discharge in the i-th subarea is a fraction of the total
discharge, where the fraction is computed as the ratio of the conveyance for the sub-
area in question to the total conveyance for the cross section.
Use of these equations is typically made in either of two ways. If one desires to find
the depth of uniform flow for a specified set of discharge and channel characteristics,
the solution involves iterative calculations using assumed water surface elevations.
For an assumed water surface elevation, the velocity and area of each subarea can be
determined, and a total discharge Q can be computed using Equations 7.30 and 7.31.
If the computed flow rate is different from the known discharge, then a new water sur-
face elevation is assumed, and the procedure is repeated until agreement is attained.
After the water surface elevation is found, the mean velocity of flow can be deter-
mined as Vm = Q/A, and ι can be determined by using Equation 7.28. Finally, the
velocity head is computed as αVm2/2g and added to the water surface elevation to
obtain the energy grade line elevation.
A second way in which these equations can be applied occurs when the discharge,
channel characteristics, and energy grade line elevation are known at a cross section.
This situation occurs in gradually varied flow profile analysis via the standard-step
method (see Section 7.8). As in the previous case, iterations are made on assumed
water surface elevations. For an assumed water surface elevation, one can determine
the areas and conveyances of each of the subareas of the channel section. The dis-
charge in each subarea can then be determined using Equation 7.35, and the velocity
in each subarea can be found as the ratio of the discharge to the area. The mean veloc-
ity can be found by using Equation 7.29, and α can be found by using Equation 7.28.
Finally, the velocity head can be computed as αVm2/2g and added to the assumed
water surface elevation to obtain a computed energy grade line elevation. If the com-
puted energy grade line elevation is not the same as the known elevation, a new water
surface elevation is assumed, and the procedure is repeated.
Example 7.10 – Uniform Flow in a Compound Cross Section. Figure E7.10.1 illus-
trates a compound channel section consisting of two rectangular elements. Determine the depth of
uniform flow in this channel section if the total discharge is Q = 450 cfs and the channel slope is So =
0.005. Also, determine the velocity distribution coefficient α for this uniform flow. Manning’s n val-
ues for the main channel and overbank area are 0.020 and 0.030, respectively.
Figure E7.10.1 Compound channel cross section for Example 7.10

Solution: For any assumed depth of flow y1 greater than 3 ft, the areas and wetted perimeters of the
two subareas can be found as
A1 = 10y1
A2 = 12y2 = 12(y1 – 3)
P1 = 13 + y1
P2 = 12 + y2 = 12 + (y1 – 3)
where the subscripts 1 and 2 refer to the main channel and overbank, respectively.
These expressions can be used with Equations 7.31 and 7.34 to find the discharge in each subarea, and
then a total discharge as the sum of the two subarea discharges. Because the flow is uniform, So is
used in place of Sf in Equation 7.31. By trial and error, the correct depth is found when the computed
discharge is equal to the known discharge.
If the initial assumption for depth is y1 = 4 ft, then A1 = 40 ft2, A2 = 12 ft2, P1 = 17 ft2, and P2 = 13 ft2.
Therefore, from Equations 7.31 and 7.34,
2/3
Cf 1.49  40 
κ1 =
n1
A1R12 / 3 =
0.020
( 40 )  17  = 5, 272 cfs
 
1/ 2
Q1 = κ1So1 / 2 = ( 5272 )( 0.005 ) = 373 cfs
2/3
Cf 1.49  12 
κ2 =
n2
A2 R22 / 3 =
0.030
(12 )  13  = 565 cfs
 
1/ 2
Q2 = κ 2 So1 / 2 = ( 565 )( 0.005 ) = 40 cfs
So, according to Equation 7.30,

Q = ∑ Qi = 373 + 40 = 413 cfs
Because 413 cfs is less than the given flow of 450 cfs, a slightly higher water surface elevation must
be assumed and the flow computed again until the correct elevation is found. For this example, trial
and error leads to yn = 4.18 ft.
For the depth of 4.18 ft, the individual subareas and velocities are
A1 = 41.8 ft2
A2 = 14.16 ft2
V1 = 9.52 ft/s
V2 = 3.67 ft/s
Therefore, from Equations 7.28 and 7.29:
1 1
Vm = ∑ AV
i i = ( 41.8)( 9.52 ) + (14.16 )( 3.67 )  = 8.04 ft/s
A 41.8 + 14.16  
α=
∑ AV
i i
3
=
( 41.8)( 9.52 )3 + (14.16 )( 3.67 )3 = 1.26
AVm3 ( 41.8 + 14.16 )(8.04 )3
As shown, the velocity distribution coefficient is 1.26.
7.4 SPECIFIC ENERGY AND CRITICAL FLOW

Critical flow, like uniform flow, is a case in open-channel flow for which a unique
relationship between depth and discharge exists. The occurrence of critical flow cor-
responds to the point of minimum specific energy for a given discharge and channel
geometry; thus, the study of critical flow is facilitated by an understanding of specific
energy concepts.
Specific Energy
Specific energy is the energy head relative to the channel bottom elevation. For a
channel of small slope, the specific energy E is
αV 2 α Q2
E = y+ = y+ (7.36)
2g 2 gA2
where E = specific energy (ft, m)

y = depth of flow (ft, m)
V = cross-sectionally averaged flow velocity (ft/s, m/s)
g = gravitational acceleration (32.2 ft2/s, 9.81 m2/s)
Section 7.4 Specific Energy and Critical Flow 245
Clearly, the specific energy at a channel cross section consists of the sum of the pres-
sure head (the depth of flow) and velocity head terms from the energy equation
(Equation 7.25).
For a given discharge Q and channel cross-sectional geometry, inspection of Equation
7.36 reveals that E depends only on the depth of flow. A graph showing the depen-
dence of E on y for a given discharge and channel geometry is called a specific energy
curve. Qualitatively, the characteristics of a specific energy curve can be deduced
from Equation 7.36. For small values of y, the velocity will be large and the second
term on the right side of the equation will be much larger than the first. Thus, the spe-
cific energy curve will approach the asymptote E = αV2/2g. Conversely, for large val-
ues of y, the velocity will be small and the first term will be much larger than the
second. In that case, the curve will asymptotically approach E = y. Figure 7.11 illus-
trates the characteristics of a typical specific energy curve.
Because a specific energy curve shows the dependence of E on y for a fixed discharge
and channel cross-sectional geometry, a different curve must arise if either the dis-
charge or geometry changes. (Note, however, that changing only the slope simply
means moving another point on the same curve.) It is difficult (if not impossible) to
state any generally applicable effects of geometric changes, but Equation 7.36 shows
that E increases when the discharge increases. That is, if two specific energy curves
for a given channel geometry were plotted on the same graph, with one curve corre-
sponding to a discharge Q1 and the second corresponding to a discharge Q2 > Q1, the
result would be a shift of the curve for Q2 to higher values of E. This shift is shown in
Figure 7.11.
Figure 7.11
Example of specific
energy curves for two
flows in the same
channel
Inspection of a specific energy curve reveals that there are two possible depths of flow
corresponding to any specific energy greater than the minimum specific energy level
denoted by Ec. These two depths are called alternate depths. The smaller of the two
depths corresponds to a shallow and rapid flow known as supercritical flow. The
larger of the two depths corresponds to a deep and tranquil flow known as subcritical
flow. The two limbs of a specific energy curve join at a point where the specific
energy is Ec and there is a single corresponding value of depth denoted by yc (see Fig-
ure 7.11). The depth yc is called the critical depth, and a flow at this depth is corre-
spondingly called a critical flow.
Example 7.11 – Specific Energy in a Rectangular Channel. Compute and graph a

specific energy curve for a rectangular open channel with B = 10 ft and Q = 200 cfs.
Solution: To graph the curve, one may assume depths of flow and compute the corresponding E val-
ues. For an assumed depth y, the cross-sectional area is A = 10y and the velocity is
V = Q/A = 20/y
The specific energy (assuming α = 1) is therefore
(20 / y) 2
E = y+
2g
Figure E7.11.1 shows this specific energy curve for this example.
Figure E7.11.1 The specific energy curve for Example 7.11
Critical Depth and Critical Flow

As shown in the preceding section, critical depth and the corresponding critical flow
occur when the specific energy is a minimum for a fixed discharge and channel cross-
sectional geometry. Computation of the critical depth yc can be accomplished in a
trial-and-error manner to minimize E. An alternative (and more enlightening) way of
computing critical depth is to recognize that it occurs when dE/dy is equal to zero (see
Figure 7.11). Differentiating Equation 7.36 and assuming α = 1, one obtains
dE Q 2 dA Q 2T
= 1− = 1− (7.37)
dy gA3 dy gA3
where T = dA/dy = top width of the flow

Equating this expression to zero, critical depth occurs when
Q 2T
=1 (7.38)
gA3
The Froude number of an open-channel flow is a dimensionless quantity defined as
1/ 2
 Q 2T  V
Fr =   = (7.39)
 gA3  gDh
 
where Fr = the Froude number

Dh = hydraulic depth (see Equation 7.2) (ft, m)
Comparing Equations 7.38 and 7.39, it can be seen that critical depth and critical flow
occur when the Froude number is equal to unity. Subcritical flow occurs when Fr < 1,
and supercritical flow occurs when Fr > 1.
An additional fact that can be deduced from Equations 7.38 and 7.39 is that when a
flow is at critical depth, the velocity head is equal to one-half of the hydraulic depth:
V 2 Dh
= (7.40)
2g 2
This relationship can be quite useful for practical problem solving, particularly when
a channel is rectangular in shape.
The reader should note that the hydraulic depth Dh of a channel is generally not the
same as the depth of flow y. For common prismatic channel shapes, the hydraulic
depth is smaller than the depth of flow, and is equal to the depth of flow only in rect-
angular channels.
Example 7.12 – Finding Critical Depth and Minimum Specific Energy in a Rect-
angular Channel. Determine the critical depth and minimum specific energy for the channel in
Example 7.11.
Solution: The criterion for critical depth is that Equation 7.38 must be satisfied. For the channel of
interest, B = T = 10 ft, Q = 200 cfs, and A = 10y. Substituting and solving for y, one obtains the critical
depth as
2002
yc = 3 = 2.32 ft
32.2(10)2
From Equation 7.40, the velocity head for critical flow is one-half the hydraulic depth, or 1.16 ft in
this example. Thus, the minimum specific energy is
Ec = 2.32 + 1.16 = 3.48 ft
Example 7.13 – Finding the Critical Depth in a Trapezoidal Channel. A trapezoi-

dal open channel has a bottom width of B = 3 m and side slopes with z = 2. Determine the critical
depth if Q = 7 m3/s.
Solution: The cross-sectional area of flow, in terms of the depth, is
A = 3y + 2y2
The top width is
T = 3 + 4y
Substituting into Equation 7.38, one obtains
2 2
Q
---------T- = --------------------------------------
7 ( 3 + 4y )
- = 1
3 2 3
gA 9.81 ( 3y + 2y )
By trial and error, the depth that satisfies this expression is yc = 0.70 m.
Compound Channel Sections

In a compound channel section, consisting of a main channel and overbank areas,
there may be more than one critical depth for a given discharge. Further, determina-
tion of the number of critical depths and their values is more complex than simply
locating all local minima on a specific energy curve, because a critical flow condition
may exist even in situations for which the specific energy is not at a local minimum
(Chaudhry and Bhallamudi, 1988).
The procedure described by Chaudhry and Bhallamudi (1988) for determining the
critical depths in a compound channel is presented here. The procedure first deter-
mines the number of critical depths that occur within a channel section. It then deter-
mines each of the critical depths. It assumes that the channel cross section is
symmetrical about its centerline, with dimensional terminology as shown in Figure
7.12.
Based on these dimensions, one can define a number of dimensionless parameters:
y
yr = (7.41)
yf
where yr = ratio of the depth y to the vertical distance from the bottom of the over-
bank to the bottom of the main channel yf
Bf
br = (7.42)
Bm
where br = ratio of the width of an overbank Bf to the width of the main channel Bm
Figure 7.12
determining critical
depths in a compound
channel
Credit: Chaudry, 1993. Reproduced by permission of Pearson Education, Inc., Upper Saddle River, N.J.
Bf
bf = (7.43)
yf
where bf = ratio of the width of an overbank Bf to the vertical distance from the bot-
tom of the overbank to the bottom of the main channel yf
nm
nr = (7.44)
nf
where nr = ratio of Manning’s n for the main channel nm to Manning’s n for an

overbank nf
As described by Chaudhry and Bhallamudi (1988), one can use these dimensions and
dimensionless ratios to write an expression of the form
gBm2 y3f
C= (7.45)
Q2
where
C = c1 + c2 (7.46)
and
1  m 2  1 − m 2 1 
 
c1 =   +    (7.47)
yr + 2br ( yr − 1)  yr   yr − 1  2br 
 
2m(1 − m)  5 2   m 1  1 − m  
c2 =  −  −   (7.48)
3[ yr + 2br ( yr − 1)]  yr ( yr − 1) b f + yr − 1   yr 2br  yr − 1  
1
m= (7.49)
1 + 2nr ( A2 / A1 )5 / 3 ( P1 / P2 ) 2 / 3
where A1 = flow area for the main channel

A2 = flow area for one of the overbanks
P1 = wetted perimeter for the main channel (2yf + Bm)
P2 = wetted perimeter for one of the overbanks (y − yf +Bf)
For specified channel dimensions, roughness values, and a discharge, one can deter-
mine a target value of C using Equation 7.45. Various flow depths (and hence values
of yr) can be assumed so that Equations 7.46 through 7.49 can be applied and C
graphed as a function of yr. The values of yr for which C is equal to the target value
(the right side of Equation 7.45) are the critical depths for flows over the floodplain.
An additional critical depth, depending on the discharge, may also occur at a depth
smaller than y = yf. Example 7.14, adapted from Chaudhry (1993), illustrates the pro-
cedure for computing critical depths in a compound channel.
Example 7.14 – Determining the Critical Depths in a Compound Channel.

Determine the critical depths for Q = 1.7 and 2.5 m3/s in a channel having the following dimensions:
Bm = 1.0 m, Bf = 3.0 m, yf = 1.0 m, and nr = 0.9 (see Figure 7.12).
Solution: Consider first the possibility that a critical depth might occur within the main channel (i.e.,
at a depth smaller than yf). Applying Equation 7.38 with T = Bm and Q = 1.7 m3/s, one finds a critical
depth of yc = 0.67 m. Applying the same equation, but with Q = 2.5 m3/s, one finds a critical depth of
yc = 0.86 m. Thus, for both of these discharges, a critical depth occurs within the main channel. It
must now be determined if additional critical depths occur once the flow rises to a depth greater than
yf.
Figure E7.14.1 is a graph of C versus yr obtained by applying Equations 7.46 through 7.49. Note that
the minimum value of yr employed is 1.00, which corresponds to an actual depth of y = yf. Note also
that C approaches 1 when yr approaches 1 m, and that C attains a maximum value of Cmax = 2.88
when yr = 1.033 m.
Using Equation 7.45 with Q = 1.7 m3/s yields C = 3.39. This is shown as the upper of the two hori-
zontal lines in Figure E7.14.1. Because it does not intersect the C- yr curve in that figure, there is no
critical depth larger than yf when the discharge in the channel is 1.7 m3/s.
Using Equation 7.45 with Q = 2.5 m3/s yields C = 1.57. This is shown as the lower of the two hori-
zontal lines in Figure E7.14.1. Because it intersects the C- yr curve in the figure twice, two critical
depths larger than yf exist when the discharge in the channel is 2.5 m3/s. Those critical depths are
1.002 m and 1.12 m.
Figure E7.14.2 shows specific energy curves computed for each of the two discharges in the channel.
Note that when the discharge is Q = 1.7 m3/s, only one critical depth exists: 0.67 m. However, when
the discharge is Q = 2.5 m3/s, there are three critical depths, at values of 0.86, 1.002, and 1.12 m. As
shown in the figure, not all of the critical depths correspond to local minima on the specific energy
curve.
Figure E7.14.1 C versus yr for the compound channel in Example 7.14
Figure E7.14.2 Specific energy curves for two discharges in a compound channel for Example 7.14
Flow Control
A control section in open-channel flow is said to occur at a channel location for which
there is a known or unique value of the depth of flow for a specified discharge. The
depth of flow at a control section is essentially a boundary condition that provides a
unique solution to the differential equation describing a varied flow in a channel. This
concept is discussed further in subsequent sections of this chapter.
Because a unique relationship exists between discharge and flow depth for a critical
flow, a channel cross section at which critical flow occurs is necessarily a flow control
section. Common occurrences of such critical flow control sections occur at free over-
falls and at locations where a mildly sloping channel ends and becomes a steeply
sloping channel. Rouse (1936) has shown that, at a free overfall, critical depth actu-
ally occurs at a distance of about 3yc to 4yc upstream of the brink of the overfall.
In open-channel flow, flow control sections can exist at locations other than critical
flow sections. For example, the flow depth created at the upstream end of a culvert is
often a flow control section, as is the depth of a supercritical flow issuing from
beneath a sluice gate. Many other examples of flow control sections can be found.
Flow Measurement
Critical flow is an instance in which there is a unique definitive relationship between
the depth of flow and discharge in an open channel. Thus, if one can cause a flow to
pass through its critical depth, measurement of that depth can be used to deduce the
discharge.
There are a number of ways that one can force a flow to pass through critical depth.
One way is to install a weir in a channel (see Section 7.6). Weirs, however, tend to
cause large head losses in the flow, which may not be acceptable in some circum-
stances. Critical depth can be caused by a flow contraction, and it occurs naturally at a
free overfall or at a location where a mild and steeply sloping channel join. A Parshall
flume is another widely used device for measurement of open-channel flow, though
critical depth may or may not occur in such a flume.
Analysis of Flow in Channel Transitions

A transition occurs in an open channel at a location where the cross-sectional geome-
try of the channel changes. For example, a transition might correspond to a change in
the width of a rectangular channel, or in the diameter of a pipe (such as at a manhole).
A transition may also be as simple as a step or drop in the bottom elevation of the
channel, with no corresponding change in cross-sectional geometry.
When there is a reduction in cross-sectional area and a corresponding increase in
velocity at a transition, the flow is contracted. A flow expansion occurs when the
cross-sectional area increases and the velocity decreases.
Consider a transition consisting of a simple step in the bottom elevation, as shown in
Figure 7.13. The discharge, Q, is constant through the transition, and the cross-
sectional geometry of the channel is the same both upstream and downstream of the
step. If the depth of flow at a cross section downstream of the step is known and des-
ignated by y2, the corresponding specific energy E2 can easily be found. If one
assumes that there is no energy loss in the transition (that is, that the energy grade line
has a negligible slope in the transition), the specific energy E1 at a cross section
upstream of the step can be determined as E1 = E2 + ∆Z, where ∆Z is the height of the
step. A specific energy relationship can then be applied to determine the two alternate
depths of flow at the upstream section.
Figure 7.13
A simple step channel
transition and the
resulting specific
energy curve
Determining which of the two alternate depths is the correct upstream depth requires
one to consider the concept of attainability. Because neither the channel geometry nor
the discharge changes in the transition, the single specific energy curve shown in Fig-
ure 7.13 applies to both the upstream and downstream portions of the transition. To
move from one location on the curve (represented by a particular combination of
depth and specific energy) to another location on the curve requires that one traverse
along the path of the curve. Point A on the curve represents the known depth and spe-
cific energy downstream of the transition, and points B and C represent the alternate
depths upstream of the transition. Note that movement along the curve from point A
to point B requires an increase in specific energy, which occurs in this situation since
E1 is larger than E2 by the amount ∆Z. On the other hand, movement along the curve
from point A to point C requires that the specific energy first decrease to Ec, and then
increase to the amount E1.
Noting that specific energy may be viewed graphically as the distance from the
energy grade line to the channel bottom, it is seen in Figure 7.13 that in moving from
the downstream to the upstream cross section, the specific energy only increases. It
does not decrease and then increase, and thus point C on the specific energy curve is
not attainable from point A in this instance. Thus, the correct upstream depth is the
larger of the two alternate depths. Had there been a hump or sill in the channel bottom
between the two cross sections such that the specific energy at the hump were equal to
Ec, then the smaller of the two alternate depths would have been the correct solution.
This concept of attainability is applicable to any type of flow in a channel transition,
whether it is subcritical or supercritical, provided that neither the discharge nor the
channel geometry changes in the transition. When the geometry changes, the concept
of attainability remains valid, but its analysis becomes more complicated. The com-
plication arises because the specific energy relationships upstream and downstream of
the transition are different from one another, and there is generally a continuum of
specific energy relationships due to the gradually changing geometry of the transition
along its length.
In the preceding discussion, it was assumed that energy losses in a transition were
negligible. In truth, the energy losses occurring in a channel transition consist of fric-
tional losses along the length of the transition and minor losses caused by velocity
changes. Because channel transitions are often short in length, the frictional losses are
usually small compared to the minor losses and can be safely ignored. Minor loss
coefficients applicable to channel transitions were presented in Section 7.2 (page
230).
Section 7.5 Momentum in Open-Channel Flow 255
Incorporation of energy losses into an analysis of flow through a transition requires a

trial-and-error approach. To illustrate, if the downstream depth and specific energy are
known, a first estimate of the upstream depth and specific energy can be made using
the procedures discussed previously. With an initial estimate of the upstream depth,
the velocities at the upstream and downstream cross sections can be determined, and
the minor energy loss can be computed using Equation 7.16. A refined estimate of the
upstream specific energy is then determined by recognizing that the minor loss causes
the energy grade line elevation to increase by the amount hL. This refined estimate of
the upstream specific energy can be used to obtain a refined estimate of the upstream
depth, and the procedure is repeated until convergence is obtained.
Example 7.15 – Bottom Transition in a Rectangular Channel. A sudden drop of

0.50 ft occurs in the channel described in Examples 7.11 and 7.12. Recall that the channel is rectangu-
lar with B = 10 ft and Q = 200 cfs. The depth of flow downstream of the drop is y2 = 4.80 ft. Deter-
mine the depth upstream of the drop assuming that energy losses are negligible.
Solution: The specific energy downstream of the drop is
{ }
2
V22 200 / (10 )( 4.8 ) 
E2 = y2 + = 4.8 + = 5.07 ft
2g 2 ( 32.2 )
Based on the assumption of zero energy loss, the specific energy upstream of the 0.5-ft drop must
therefore be
E1 = 5.07 – 0.5 = 4.57 ft
The alternate depths upstream of the drop are the two positive solutions of the following expression:
(200 /10 y1 )2
E1 = y1 + = 4.57 ft
2(32.2)
The two alternate depths are found, by trial and error, to be y1 = 4.22 ft and y1 = 1.40 ft.
From Example 7.12, the critical depth in this channel is 2.32 ft, and the corresponding minimum spe-
cific energy is 3.48 ft. Because the specific energy in the channel decreases from 5.07 to 4.57 ft, but
never drops to a value as small as 3.48 ft, the larger of the two alternate depths is the only attainable
one. Therefore, the upstream depth of flow is y1 = 4.22 ft.
7.5 MOMENTUM IN OPEN-CHANNEL FLOW

Equation 3.12, the momentum principle, as applied to a coordinate direction x for a
steady flow, states that the sum of the external forces acting on the water within a con-
trol volume is equal to the net rate of momentum outflow from the control volume.
This principle is applied when determining the upstream and downstream depths for a
hydraulic jump and when evaluating forces on submerged structures.
Figure 7.14(a) illustrates a reach of channel (a control volume) of length L between an
upstream cross section 1 and a downstream cross section 2. The coordinate direction x
is taken as positive in the downstream direction and parallel to the channel bottom.
External forces acting on the water in the control volume consist of the pressure force
P1 acting at cross section 1, the pressure force P2 acting at cross section 2, the fric-
tional force Ff opposing the flow, and the gravitational force Wsinθ driving the flow.
Unlike the case in uniform flow, the pressure forces P1 and P2 shown in Figure 7.14
are not necessarily equal. The sum of the forces is equal to the change in momentum
across the control volume.
Figure 7.14
forces acting on a
fluid element with
non-uniform flow
Substituting the sum of the external forces into Equation 3.12 and equating the sum to
the net momentum outflow rate, one obtains
∑ Fx = P1 −P2 + W sin θ − F f = ρ Q( β 2V2 − β1V1 ) (7.50)
where ∑Fx = sum of external forces in the x direction

P1 and P2 = forces exerted on cross section due to hydrostatic pressure (see
Equation 7.53) (lb, N)
W = weight of fluid element (lb, N)
θ = angle that channel bed makes with horizontal
Ff = frictional force (lb, N)
β1 and β2 = velocity distribution coefficients as defined by Equation 6.30
V1 and V2 = average flow velocities (ft/s, m/s)
Typical values of β for open-channel flow are presented in Table 7.2. In compound
channels, where the velocity may vary greatly from one part of the cross section to
another, an estimate of β may be obtained as
n
∑ AiVi2 (7.51)
i =1
β=
AV 2
where β = velocity distribution coefficient

Ai = cross-sectional area of the i-th subarea (ft2, m2)
Vi = average velocity of the i-th subarea (ft/s, m/s)
n = number of subareas
A = total cross-sectional area (ft2, m2)
V = average velocity for the entire cross section (ft/s, m/s)
The pressure force P exerted at a cross section in an open-channel flow (see Equation
7.50) can be expressed as
P = γ zA (7.52)
where γ = fluid specific weight (lb/ft3, N/m3)

z = distance from the free water surface to centroid of cross-sectional area
[see Figure 7.14(b)] (ft, m)
A = cross-sectional area of flow (ft2, m2)
Table 7.7 summarizes expressions for z for rectangular, trapezoidal, triangular, and
circular cross-sectional shapes (for dimension definitions, see Figure 7.4). (Note that
both side slopes must be equal for triangular and trapezoidal channels if these equa-
tions are used.) For other channel shapes, z can be computed as
n
∑ Ai zi (7.53)
i =1
z=
A
where zi = vertical distance from free water surface to centroid of cross section of
i-th subarea (ft, m)
If other forces are exerted on the water in the control volume, such as by a step in the
channel bottom or by a bridge pier, then those forces should be included in Equation
7.50. This point is illustrated in the “Forces on Submerged Structures” discussion
later in this section.
Table 7.7 Centroidal distance z for prismatic channelsa

Channel shape Area z
y
Rectangular By ---
2
y
Triangular zy2 ---
3
3
y ( 3B – 2zy )
Trapezoidal By + zy2 --------------------------------
2
6 ( By + zy )
φ φ φ
2
D ( φ – sin φ ) 2D sin  --- – D sin φ cos  --- D cos  ---
Circulara --------------------------------  2  2  2
8 ----------------------------------------------------------------- – ----------------------
3 ( φ – sin φ ) 2
Specific Force
If Equations 7.50 and 7.52 are applied to a short reach of a horizontal channel, the
frictional force Ff will be small and Wsinθ will be zero. Thus, if β1 and β2 are
assumed to be 1, the momentum equation simplifies to
 Q2   Q2 
γ  z1 A1 +  = γ  z2 A2 +  (7.54)
 gA1   gA 
  2 
or
Fs1 = Fs2 (7.55)
where
 Q2 
Fs = γ  zA +  (7.56)
 gA 

and Fs = specific force (lb, N)

The quantity Fs expressed by Equation 7.56 is known as the specific force in an open-
channel flow. It consists of two terms, one representing the hydrostatic pressure force,
and the second representing the additional dynamic force caused by the flow momen-
tum.
Specific force, like specific energy, depends on the flow depth, channel geometry, and
discharge. For a specified discharge and channel shape, a graph of specific force can
be developed as a function of flow depth. This graph is known as a specific force
curve.
Like a specific energy curve, a specific force curve contains two limbs: the first corre-
sponds to subcritical flow, and the second corresponds to supercritical flow. The spe-
cific force curve also has a minimum force given by Fsc, which occurs at the critical
depth yc. Note that for any value of the specific force greater than Fsc, there are two
possible depths of flow. Again, this is exactly analogous to the case of a specific
energy curve. However, in the case of a specific force curve, the two depths are called
sequent depths (some texts refer to them as conjugate depths).
Example 7.16 – Specific Force in a Trapezoidal Channel. Develop a specific force

curve for a trapezoidal open channel with B = 8 ft, z = 2, and Q = 250 cfs.
Solution: Using expressions for A and z from Table 7.7, one can substitute into Equation 7.56 to
obtain an expression for Fs as a function of flow depth y:
 By 2 zy 3 Q2 
Fs = γ  + + 
 2 3 g ( By + zy 2 
)
 
A graph of the resulting relationship is shown in Figure E7.16.1, which is the specific force curve.
Note that the specific force has a minimum at the critical depth of 2.51 ft.
Figure E7.16.1 Graph of specific force versus depth for channel in Example 7.16
The Hydraulic Jump

The hydraulic jump is a phenomenon that occurs when flow in an open channel
increases from a small supercritical depth to a larger subcritical depth with a corre-
sponding decrease in energy caused by turbulent eddies. This loss of energy can be
considerable. For this reason, hydraulic jumps are often employed to dissipate other-
wise harmful energy in a channel.
To illustrate the characteristics of a hydraulic jump, consider a jump in a horizontal
trapezoidal channel with dimensions and discharge as specified in Example 7.16 on
page 259. Assume that the depth of flow upstream of the jump is y1 = 1.00 ft (0.3 m),
as shown in Figure 7.15, and that the length of the jump is short. Because the channel
is horizontal and the jump is short, conditions assumed at the beginning of the preced-
ing subsection (that is, Ff is small and Wsinθ is 0) are applicable, and one can con-
clude that Fs1 = Fs2. Using the information in Figure E7.16.1 with y1 = 1.00 ft (0.3 m)
yields Fs1 = Fs2 = 12,403 lb (5,625 kg) and y2 = 4.93 ft (1.5 m). Thus, for a hydraulic
jump in a horizontal channel, the relationship between the depths upstream and down-
stream of the jump is given by the sequent depths on a specific force curve.
The energy loss in the hydraulic jump can be assessed by computing the specific
energy upstream and downstream of the jump. Upstream of the jump where the depth
is y1 = 1.00 ft (0.3 m), the corresponding specific energy (from Equation 7.36) is E1 =
10.70 ft (3.3 m). Downstream of the jump, where the depth is y2 = 4.93 ft (1.5 m), the
specific energy is E2 = 5.06 ft (1.54 m). The difference between the two specific ener-
gies is the head loss in the jump, which is hL = 5.64 ft or 1.72 m (see Figure 7.15).
Note that the alternate depths do not play a role in the analysis of a hydraulic jump,
but the specific energy curve does provide information relevant to the head loss that
occurs in the jump.
Figure 7.15
Definition figure for a
hydraulic jump
As described in the previous paragraphs, the characteristics of a hydraulic jump in a

horizontal channel of any cross-sectional shape can be determined by use of specific
force and specific energy curves. However, in the special case of a horizontal, rectan-
gular channel, one can show that the sequent depths are related as
y2 1 
=  1 + 8Fr12 − 1 (7.57)
y1 2  
where y1, y2 = flow depths at sections 1 and 2, respectively (ft, m)

Fr = Froude number
and
V2
Fr12 = 1 (7.58)
gy1
where g = gravitational acceleration constant (32.2 ft/s2, 9.81 m/s2)

In such a channel, one can also show that the energy loss is
( y2 − y1 )3
hL = E1 − E2 = (7.59)
4 y1 y2
where hL = head loss in hydraulic jump (ft, m)

E1, E2 = total energy (ft, m)
If a hydraulic jump occurs in a sloping channel such that the term Wsinθ in the
momentum equation cannot be neglected, the specific force approach to estimating
the relationship between the depths upstream and downstream of the jump cannot be
applied. Instead, one must rely on an experimental relationship, such as that illus-
trated in Figure 7.16. The figure illustrates the ratio of the downstream depth to the
upstream depth (y2/y1 measured vertically or d2/d1 measured normal to the channel
bottom) as a function of the channel slope (So) and the Froude number of the
upstream flow (Fr1). For this figure, a rectangular cross-section shape is assumed.
The length of a hydraulic jump is a characteristic that must be evaluated on the basis
of experimental data. Figure 7.17, which is based on data collected by the U.S.
Bureau of Reclamation (1955), illustrates the length L of a jump as a function of the
downstream depth, the channel slope, and the Froude number of the upstream flow.
This figure was developed for rectangular channels; however, in the absence of other
data, it can be used to approximate jump length in nonrectangular channels.
Figure 7.16
Ratio of downstream
to upstream depths for
a hydraulic jump in a
rectangular channel as
a function of the
channel slope and
upstream Froude
number
Credit: Chow, 1959.
Figure 7.17
Graph for determining
the length of a
hydraulic jump
Credit: Chow, 1959.

Example 7.17 – Hydraulic Jump in a Rectangular Channel. A hydraulic jump forms

in a horizontal rectangular open channel with B = 3 m and Q = 6 m3/s. The depth of the approaching
flow is y1 = 0.40 m. Determine the depth downstream of the jump, the energy loss in the jump, and the
length of the jump.
Solution: Because the channel is horizontal and rectangular, Equation 7.57 may be used to determine
the downstream depth. The Froude number of the approaching flow is
V1 6 ( 3 × 0.04 )
Fr = ------------ = ---------------------------- = 2.52
gy 1 9.81 × 0.4
Therefore, applying Equation 7.57, the downstream depth is

y1 2
y 2 = ----- ( 1 + 8Fr – 1 ) = 1.24 m
2
The energy loss in the jump may be determined using Equation 7.59, and is
3 3
( y2 – y1 ) ( 1.24 – 0.40 )
hL = ----------------------
- = ------------------------------------ = 0.30 m
4 y1 y2 4 × 1.24 × 0.40
The length of the jump can be estimated using Figure 7.17. From that figure, the dimensionless ratio
L/y2 is estimated to be about 4.8. Therefore,
L = 4.8y2 = 4.8(1.24) = 5.95 m
Forces on Submerged Structures

When an open-channel flow occurs past a submerged structure, such as a bridge pier,
the flow exerts a hydraulic force on that structure. The structure in turn exerts an equal
and opposite force on the flow. The force exerted on the flow by the structure must be
considered as an external force on the flow when applying the momentum equation.
Consider flow in an open channel of small slope and short length where there is a sub-
merged structure (such as a bridge pier) in the flow path. In this case, the weight of
the water and frictional forces may be neglected, and the momentum equation
becomes
γ z1 A1 − γ z2 A2 − Fp = ρ Q( β 2V2 − β1V1 ) (7.60)
where Fp = force exerted on the flow by the pier (lb, N)

Note that it is assumed in this expression that the force exerted by the pier opposes the
flow direction (that is, it is in the negative x-direction).
A comparison of Equation 7.60 to Equations 7.54 and 7.55 indicates that
Fs1 = Fs2 + Fp (7.61)
Thus, the pier force Fp can be determined graphically from a specific force curve if
the depths of flow y1 and y2 upstream and downstream of the pier are known. The
depth change due to the pier may be found through an application of the energy equa-
tion based on an estimate of the energy loss due to the pier.
Example 7.18 – Force in a Channel Transition. Figure E7.18.1 illustrates a channel tran-
sition consisting of a step of height ∆Z. Determine the force exerted on the step by the water if the
channel is rectangular with B = 6 m, y1 = 2 m, Q = 12 m3/s, and ∆Z = 0.25 m. Assume that the energy
loss in the transition is negligible.
Figure E7.18.1 Channel transistion for Example 7.18
Solution: First, the downstream depth y2 must be determined. The specific energy upstream of the
step is
2
V12 12 / ( 2 )( 6 ) 
E1 = y1 + = 2+   = 2.05 ft
2g 2 ( 9.81)
and the specific energy downstream of the step is
E2 = E1 – ∆Z = 2.05 – 0.25 = 1.80 m
The equation for E2 can then be written as
2
12 / ( 6 y2 ) 
1.8 = y2 +  
2 ( 9.81)
By trial and error, the downstream depth is y2 = 1.74 m.
Assuming that the step exerts a force on the flow in the upstream direction (the negative x-direction),
that force may be determined from Equation 7.61 as
Fp = Fs1 – Fs2
or, substituting from Equation 7.56,
 Q2   Q2 
Fp = γ  z1 A1 +  − γ  z2 A2 + 
 gA1   gA2 
 
  122   1.74 122  

Fp = 9,810  1(12) + − (1.74)(6) +   = 26,800 N
9.81(12) 2 9.81(1.74)(6)
     
Because the sign of this result is positive, the assumed direction of the force exerted by the step is cor-
rect. Therefore, the force exerted on the step by the water is in the downstream direction and has a
magnitude of 26,800 N.
Section 7.6 Weirs 265
7.6 WEIRS
A weir in an open channel is a structure that may be employed for flow measurement
or as a flow control structure. Weirs may be used as outlet structures for water storage
facilities such as detention ponds or reservoirs. Weir flows over roadway embank-
ments may also occur during high flow events at culvert crossings. Finally, some
hydraulic structures, notably stormwater inlets, behave as weirs during some types of
flow conditions.
Weir Characteristics
Weirs are classified as being either sharp-crested or broad-crested. A sharp-crested
weir is one in which a relatively thin metal, concrete, or plastic plate forms the crest
of the weir. For a broad-crested weir, the weir crest thickness is relatively large, and
flow over the weir crest is critical.
As illustrated in Figure 7.18, additional classifications for weirs concern the attributes
of the crest cross-sectional shape (such as rectangular, triangular, or trapezoidal), and
whether there are contractions of the flow at the ends of the weir. A contracted weir
[Figure 7.18(b)] is a weir type that causes the flow to contract. Flow over a suppressed
or uncontracted weir [Figure 7.18(a)], however, has no end contractions.
Figure 7.18
Definition figures for
rectangular and
triangular weirs
Credit: Chin, 2000

The jet of water that issues in a downstream direction from the crest of a sharp-crested
weir is called the nappe. Ideally, the nappe should spring from the crest and should
have atmospheric pressures prevailing at both its upper and lower surfaces. If the
lower face of the nappe is at a pressure below atmospheric, the nappe may cling to the
downstream face of the weir and change its hydraulic performance. If the possibility
of a clinging nappe presents itself in a particular application (usually it is only a prob-
lem with suppressed weirs), vents should be installed so as to enable atmospheric
pressure to exist on the underside of the nappe [Figure 7.18(a)].
Sharp-Crested Weirs
Theoretically, the discharge over a sharp-crested weir may be determined via the
energy equation. The procedure amounts to equating the total energy head upstream
of the weir to the total energy head on the weir crest. The resulting equation for dis-
charge in terms of the upstream energy level depends on the cross-sectional shape of
the weir, and inevitably also involves some simplifying assumptions. An empirical
weir discharge coefficient, Cd, is introduced to account for the simplifying assump-
tions and to correct the theoretical discharge. (Note that a “weir coefficient” is some-
times used instead of the discharge coefficient Cd described here. The weir coefficient
is equivalent to Cd(2g)1/2. Whereas Cd is unitless, the weir coefficient value depends
on the units being used.)
Two types of sharp-crested weirs, rectangular and triangular, are discussed in this sec-
tion.
Rectangular Weirs. The discharge Q over a suppressed rectangular sharp-

crested weir [Figures 7.18(a) and 7.19] may be expressed in terms of the head H on
the weir as
2
Q= Cd 2 g LH 3 / 2 (7.62)
3

Cd = discharge coefficient determined using Equation 7.63
g = gravitational acceleration constant (32.2 ft/s2, 9.81 m/s2)
L = length of the weir crest (ft, m)
H = height of the energy grade above weir crest (ft, m)
Figure 7.19
Rectangular sharp-
crested weir
If the velocity upstream of the weir is low, as in the case of a pond spillway, the veloc-
ity head is often considered negligible, and H is taken as the difference in elevation
between the upstream water surface and the weir crest. H should be measured far
enough upstream of the weir that drawdown effects are avoided.
Fundamentally, the discharge coefficient in Equation 7.62 depends on the Reynolds
and Weber numbers of the flow, and on the dimensionless head H/Hw, where Hw is the
height of the weir crest above the bottom of the channel (Figure 7.18). However, for
practical purposes, experiments have shown that the discharge coefficient can be rep-
resented as (Rouse, 1946; Blevins, 1984)
H
Cd = 0.611 + 0.075 (7.63)
Hw
where Hw = height of weir crest above bottom of channel (ft, m)

Equation 7.63 is valid for H/Hw less than 5 and is approximate up to H/Hw = 10. For
H/Hw greater than 15, the discharge can be computed from the critical flow equation
(see Section 7.4) by assuming that yc = H (Chaudhry, 1993).
When flow occurs over a contracted rectangular weir [see Figure 7.18(b)], the end
contractions cause the effective crest length of the weir to be reduced from the actual
crest length. In this instance, the head-discharge relationship for the weir may be
expressed as
2
Q= Cd 2 g ( L − 0.1NH ) H 3 / 2 (7.64)
3
where N = number of weir end contractions (N = 1 or 2; usually, N = 2)

This equation assumes that L is greater than 3H. Note that the effective length of the
weir is L – 0.1NH, which is 0.1NH shorter than the actual length of the weir crest.
A Cipolletti weir is a contracted weir that, rather than a rectangular cross-section, has
a trapezoidal cross section with 1:4 (horizontal:vertical) side slopes to compensate for
the flow contraction. Denoting the base width of the trapezoid (that is, the weir crest
length) as L, Equations 7.62 and 7.63 can be used to determine the head-discharge
relationship. (Note that Equation 7.64 is not used in this case because the sloped sides
of the trapezoid offset the effective crest length reduction that otherwise occurs with a
contracted weir.)
Example 7.19 – Discharge over a Contracted Rectangular Weir. A rectangular

weir with contractions on both sides (N = 2) has a crest length L = 3 m, Hw = 2.5 m, and H = 0.30 m.
Determine the discharge over the weir.
Solution: Using Equation 7.63, the discharge coefficient is
0.30
Cd = 0.611 + 0.075 = 0.62
2.5
Applying Equation 7.64, the discharge is
2
Q= (0.62) 2(9.81) {3 − 0.1(2)(0.30)} (0.30)3/ 2 = 0.88 m3 s
3
Triangular Weirs. A triangular, or v-notch, weir has a v-shaped cross section

[see Figures 7.18(c) and 7.20]. This type of weir can control flow more accurately
than a rectangular weir when the discharge and head on the weir are small. Because
the discharge over a triangular weir for any given weir head H is much smaller than
that for a rectangular weir with the same head, triangular weirs are usually used only
in instances where the discharge is small.
Figure 7.20
Triangular sharp-
crested weir
The discharge over a sharp-crested triangular weir can be shown to be
8 θ 
Q= Cd 2 g tan   H 5 / 2 (7.65)
15 2
where θ = angle formed by vertex of v-shaped weir cross section (radians)

Values of the discharge coefficient for use in this expression are shown in Figure 7.21.
The solid curves on the graph should be used for nearly all design purposes, espe-
cially when the head on the weir is 2 in. (5 cm) or more. For smaller heads, both vis-
cous and surface-tension effects are important determinants of weir behavior, and the
dashed lines reflect this combination of effects. Although the dashed lines extend
above H = 0.2 ft in the graph, it is recommended that the solid lines be used for larger
values of H. As a practical matter, the difference in using the dashed or solid curves
will be small, as changes in Cd are typically less than 0.01.
Broad-Crested Weirs
Broad-crested weirs, as the name implies, have weir crests that are broad as opposed
to sharp-edged. For example, a roadway embankment containing a culvert behaves as
a broad-crested weir if the water is high enough to flow over the roadway surface.
Theoretically, the flow profile over a broad-crested weir can be evaluated through the
concepts of specific energy and critical depth. Figure 7.22(a) illustrates a broad-
crested weir of height Hw = ∆Z with an approaching flow of discharge Q and depth y1.
The thickness of the weir crest in the direction of flow is ∆x, and the cross section is
rectangular with a weir crest length of L normal to the plane of the page.
Figure 7.21
Discharge coefficient
values for triangular
weirs (Franzini and
Finnemore, 1997)
Credit: Finnemore and Franzini, 2002. Reproduced with permission of the McGraw-Hill Companies.
Figure 7.22
Definition figure and
specific energy curve
for a broad-crested
weir
The specific energy upstream of the weir is E1. Assuming no energy loss, the specific
energy on the crest of the weir is (E1 − ∆Z) = E2, from which the depth y2 may be
determined using the specific energy curve [Figure 7.22(b)]. Thus, if the weir crest
height Hw = ∆Z is not excessively large and the assumption of zero energy loss is
made, then the flow profile over the weir may be evaluated using a specific energy
curve. Further, the specific energy E3 and depth y3 downstream of the weir will be
equal to E1 and y1, respectively, if the channel bottom elevation is the same upstream
and downstream of the weir. If desired, a minor loss can be introduced to account for
the energy loss that actually takes place.
If, assuming fixed values for E1, y1, and Hw = ∆Z, E2 is found to be less than Ec, the
assumptions described in the preceding paragraph are not valid. This situation is
impossible because Ec represents the lowest energy on the specific energy curve [see
Figure 7.22(b)]. What actually occurs is that the flow is choked, meaning that a back-
water effect is created by the weir. In such a case, one must set y2 = yc and E2 = Ec.
The specific energy upstream of the weir is then found as E1 = Ec + Hw, and the depth
y1 is determined from the specific energy curve.
When a flow is choked (that is, when a structure behaves as a broad-crested weir), the
discharge may be computed as a function of the head H = E1 − Hw as
3/ 2
 2H 
Q = Cd g L  
(7.66)
 3 
where values of the discharge coefficient can be estimated from
0.65
Cd = (7.67)
(1 + H / H w )1/ 2
These expressions are valid provided that 0.08 < (y1 − Hw)/∆x < 0.50. Head losses
across the crest of the weir cannot be neglected if (y1 − Hw)/∆x < 0.08. If (y1 − Hw)/∆x
> 0.50, streamlines over the weir crest are not horizontal. In these cases, the flow pro-
file should be evaluated using gradually varied flow profile analysis methods (see
Section 7.8).
7.7 ORIFICES
Like a weir, an orifice functions as a hydraulic control. For example, in stormwater
systems, an orifice may be part of a detention pond outlet structure configuration in
the form of a restrictor plate or an opening in a riser pipe or inlet box (see Chapter
12). Culvert entrances and stormwater inlets may also function hydraulically as ori-
fices.
The discharge through a single orifice opening depends on the area of the opening, the
effective head on the orifice, and the type of edge around the opening. The relation-
ship between these variables, which are shown in Figure 7.23, is expressed as:
Q = CdA0(2gh0)1/2 (7.68)
Section 7.8 Gradually Varied Flow 271

Cd = discharge coefficient depending on the orifice edges
A0 = area of the orifice opening (ft2, m2)
g = acceleration of gravity (ft/s2, m/s2)
h0 = effective head on the orifice (ft, m)
Brown et al. (1996) indicate that Cd can be taken as 0.60 if the edges of the orifice
opening are uniform and square-edged. For ragged edges, such as those resulting
from use of an acetylene torch to cut an opening, a value of 0.40 should be used.
The effective head on an orifice depends on whether it has a free or submerged dis-
charge. If the discharge is a free outfall, the effective head is the difference between
the upstream water surface elevation and the elevation at the centroid of the orifice
opening. If tailwater submerges the outlet of an orifice, the effective head is equal to
the difference between the water surface elevations immediately upstream and down-
stream of the orifice.
Figure 7.23
Orifice flow diagram
7.8 GRADUALLY VARIED FLOW

For gradually varied flow in an open channel, the depth and velocity of a steady flow
change along the length of the channel. As the name implies, the rates of change of
depth and velocity are gradual and there are no abrupt changes such as hydraulic
jumps. Because the curvatures of streamlines are small, hydrostatic pressures can be
assumed to exist in the flow (that is, the pressure head can be taken as the depth of
flow), and the constitutive relationships developed for uniform flow, such as the Man-
ning equation, can be assumed to be valid for analyses of gradually varied flow.
Gradually varied flow occurs in an open channel when a flow control causes the depth
of flow at a cross section to be different from normal depth in the channel. This situa-
tion can occur, for instance, upstream of culverts, at locations where the slope of a
channel changes, and at channel entrances. Depending on the depth at the control sec-
tion relative to the normal depth, the flow may be accelerated or decelerated in the
channel.
The evaluation of gradually varied flow requires computations to determine flow pro-
file type and flow depths at multiple locations along the length of the channel. This
section presents an example in which a gradually varied flow profile is evaluated.

However, because of the number of calculations involved, engineers typically use
spreadsheets or other computer programs (such as StormCAD or HEC-RAS) when
performing this type of analysis.
Water Surface Profile Equation

For a channel of small slope, the total energy head H at a cross section consists of the
sum of the channel bottom elevation Z, the depth of flow y, and the velocity head
αV2/2g. Taking the x-direction to be horizontal and positive in the downstream direc-
tion, the derivative dH/dx is
dH dZ dy d  α Q 2 
= + +   (7.69)
dx dx dx dx  2 gA2 

where H = total energy head at a cross section (ft, m)

x = distance along channel (ft, m)
Z = channel bottom elevation (ft, m)
y = vertical flow depth (ft, m)
g = gravitational constant (32.2 ft2/s, 9.81 m2/s)
A = area of cross section (ft2, m2)
Noting that the cross-sectional area A depends on the depth of flow y, the chain rule
can be applied to rewrite this expression as
dH dZ dy  Q 2T  dZ dy
dx
= + 1 − α
dx dx  gA3
= +
 dx dx

(
1 − α Fr 2 ) (7.70)
where Fr = Froude number

Because dH/dx = −Sf is the slope of the energy grade line, and because dZ/dx = −So is
the channel bottom slope, this equation can be further simplified to read
dy So − S f
= (7.71)
dx 1 − α Fr 2
where So = channel bottom slope (ft/ft, m/m)

This equation is the governing equation of gradually varied flow in an open channel.
It describes how the depth changes with distance along the channel as a function of
channel and flow characteristics. A unique solution to this differential equation that
describes the flow profile for a particular case of interest requires knowledge of a
boundary condition (that is, a known depth of flow at a particular channel cross
section).
Classification of Flow Profiles

For a specified discharge Q in a prismatic open channel of known dimensions, one
can apply the procedures described in Section 7.3 and Section 7.4 to determine the
normal and critical depths yn and yc. Knowledge of these two depths, as well as the
depth at a flow control section, enables one not only to classify the channel and flow
type, but also to determine the direction in which numerical calculations of the flow
profile must proceed.
When one compares the normal and critical depths for a particular channel geometry
and discharge, several possibilities arise. If one finds that yn is greater than yc, then the
channel is said to have a hydraulically mild (M) slope. If yn is equal to yc, the channel
has a critical (C) slope. If yn is less than yc, the channel is said to have a hydraulically
steep (S) slope. If the channel has a slope of zero, then normal depth does not exist,
and the channel is said to be horizontal (H). Finally, if the channel bottom elevation
increases as one moves in a downstream direction, the channel is said to be adverse
(A), and the normal depth is imaginary (that is, it involves the square root of a nega-
tive quantity).
Generally, the range of possible flow depths in an open channel can be subdivided
into three subranges, called zones. Zone 1 corresponds to all flow depths that are
larger than both the normal and critical depths. Zone 2 corresponds to flow depths
between normal and critical depths, and Zone 3 corresponds to flow depths smaller
than both normal and critical depths. For example, in a hydraulically mild (M) chan-
nel, where yn is greater than yc, Zone 1 corresponds to y > yn, Zone 2 corresponds to
yc < y < yn, and Zone 3 corresponds to y < yc. In a critical (C) channel, where yn
equals yc, Zone 2 does not exist. Similarly, in horizontal (H) and adverse (A) chan-
nels, where normal depth is either non-existent or imaginary, Zone 1 does not exist.
By comparing a known depth of flow at a cross section with the normal and critical
depths for that discharge, one can conclude the flow profile type. For example, if the
known depth at a particular cross section in a mild (M) channel lies in Zone 1, the
flow profile is of the M1 type. If the known depth of flow at a cross section in a steep
(S) channel lies in Zone 3, the flow profile is of the S3 type. Figure 7.24 illustrates the
various flow profile types that may exist in prismatic channels and indicates the
shapes of the water surface profiles in terms of their curvatures.
Classification of the flow profile type in a channel for a particular discharge is an
important first step that must take place before the actual computation of the water
surface elevations along the length of the channel. The classification aids in the under-
standing of what the profile should look like, and hence provides a check on subse-
quent calculations. Understanding the flow profile type also helps in terms of
classifying the flow as being subcritical or supercritical. Regardless of the channel
slope classification (M, C, S, H, or A), a flow is subcritical when y is greater than yc
and supercritical when y is less than yc.
Knowledge of whether a flow profile is subcritical or supercritical has implications
for profile computations. Computation of a subcritical profile should always proceed
in an upstream direction from a known depth at a downstream control section. Con-
versely, computation of a supercritical profile should always proceed in a downstream
direction from a known depth at an upstream control section. This distinction arises
because of fundamental differences in the ranges of influences in open-channel flow,

particularly with reference to the celerities (velocities) of small gravity waves caused
by flow disturbances.
Figure 7.24
Flow profile types for
prismatic channels
Credit: Chow, 1959.

Profile Calculations
Equation 7.71 is the fundamental equation describing a gradually varied flow profile.
Given a known flow depth at a control section (that is, a boundary condition), a solu-
tion of that equation to define depths at additional locations (cross sections) of interest
can be accomplished in a number of ways. Chow (1959) provides a discussion of
many of the historically used methods, as well as of methods that continue to be used
today. More modern methods involving numerical integration of the governing equa-
tion can also be developed.
There are two basic classifications of methods for solving Equation 7.71. In the first,
one solves for values of x—the cross section locations—corresponding to assumed
flow depths y. This type of method is generally useful only for prismatic channels
because the cross-sectional geometry is independent of location. The direct-step
method is an example of this type of procedure.
The second solution method classification is useful for non-prismatic channels. In this
instance, methods for solution of Equation 7.71 are designed to determine the flow
depth y at each specified channel location x. The most widely used of these
approaches is called the standard-step method. Further discussion of this method can
be found in texts on the subject of open-channel flow such as those by Chow (1959)
and Chaudhry (1993).
Because the focus of this text is prismatic channels, the remainder of the discussion is
limited to the direct-step method. To develop the direct-step method, the form of the
energy equation given in Equation 7.5 is written for a reach of channel as
Z1 + E1 = Z 2 + E2 + hL (7.72)
where Z1 and Z2 = channel bed elevation at sections 1 and 2, respectively (ft, m)

E1 and E2 = total energy at sections 1 and 2, respectively (ft, m)
hL = head loss due to friction (ft, m)
If the length of the channel reach between the two cross-sections is denoted by ∆x,
then
hL = Sf∆x (7.73)
where Sf = friction slope

∆x = distance between cross sections 1 and 2
and
Z1 = Z2 + So∆x (7.74)
where So = channel bed slope

Making these substitutions, Equation 7.72 can be rewritten in terms of ∆x as
E2 − E1
∆x = (7.75)
So − S f
This last expression, while not resembling the differential Equation 7.70, provides the
basis for the flow profile computation. For a given discharge Q and channel cross-
sectional geometry, and for known or assumed flow depths y1 and y2, one can deter-
mine the specific energies E1 and E2. Using a constitutive relationship such as the
Manning equation, one can also determine the friction slopes Sf1 and Sf2 at each of the
two cross sections. If the average of those two slopes is used in Equation 7.74, one
can finally find the distance ∆x between the two cross sections. The procedure is best
explained through an example.
Example 7.20 – Gradually Varied Flow Analysis of a Channel Slope Transi-

tion. A slope change occurs in a trapezoidal open channel with B = 5 m and z = 2. Upstream of the
break point, the channel slope is So = 0.001, and downstream of the break point the slope is
So = 0.020. Figure E7.20.1 illustrates the channel profile. If n = 0.025 and Q = 15 m3/s, compute the
flow profile in the vicinity of the slope change. Assume that the channel lengths upstream and down-
stream of the slope change are long.
Figure E7.20.1 Channel profile for Example 7.20
Solution: The critical depth in the channel is found, by trial and error, to be yc = 0.86 m. Upstream of
the slope break point, the normal depth is yn1 = 1.48 m, and downstream of the slope break point the
normal depth is yn2 = 0.65 m. Thus, upstream of the break point the channel is mild (M), and down-
stream of the break point it is steep (S).
In this situation, the control section occurs at the break point and the depth at the control section is the
critical depth of 0.86 m. Moving in an upstream direction from the break point, the flow profile depth
should converge toward the normal depth, and the flow profile is of the M2 type. Downstream of the
break point, the flow profile is of the S2 type. Because an M2 profile is subcritical and an S2 profile is
supercritical, the entire flow profile must be computed in two parts. The M2 profile is computed in an
upstream direction from the control section, and the S2 profile is computed in a downstream direction
from the control section.
Calculations for the M2 profile are illustrated in Table 7.8. The first column contains assumed flow
depths, beginning with the known depth at the flow control section and proceeding in an upstream
direction. [Note that as one moves upstream, the depth increases until it becomes the normal depth
(see Figure 7.21). Assumed flow depths were developed by subtracting from the normal depth in 0.1-
m increments.] The second through fourth columns are computed on the basis of the depth and chan-
Section 7.9 Spatially Varied Flow 277
nel geometry. The fifth column, velocity, is computed as the ratio of the discharge to the cross-sec-
tional area, and the specific energy in the sixth column is computed using Equation 7.36. Friction
slopes, in the seventh column, have been computed for this example using the Manning equation.
Table 7.8 Direct-step flow profile calculations
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
y (m) A (m2) P (m) R (m) V (m/s) E (m) Sf Avg Sf ∆x (m) Σ∆x (m)
0.86 5.78 8.85 0.65 2.60 1.203 0.007427 0
0.98 6.82 9.38 0.73 2.20 1.226 0.004624 0.006026 4.6 4.6
1.08 7.73 9.83 0.79 1.94 1.272 0.003238 0.003931 15.4 20.1
1.18 8.68 10.28 0.85 1.73 1.332 0.002334 0.002786 33.7 53.8
1.28 9.68 10.72 0.90 1.55 1.402 0.001722 0.002028 68.5 122.3
1.38 10.71 11.17 0.96 1.40 1.480 0.001297 0.001510 152.1 274.4
1.48 11.78 11.62 1.01 1.27 1.563 0.000995 0.001146 565.7 840.1
The average friction slope for successive rows in the table is shown in column eight, and the corre-
sponding ∆x, determined from Equation 7.75, is shown in column nine. A running summation of ∆x
values is shown in the last column of the table. Those values represent the distances upstream of the
slope break at which the corresponding flow depths occur.
Table 7.9 is identical to Table 7.8, except that it pertains to the steep reach of the channel downstream
of the slope break point. (Assumed flow depths were developed by adding to the normal depth in 0.1-
m increments.) Again, the summed ∆x values shown in the last column of the table represent locations
at which the corresponding depths occur. In this case, however, the distances are measured down-
stream rather than upstream from the slope break point. The computed water surface profile is illus-
trated in Figure E7.20.1.
Table 7.9 Direct-step flow profile calculations
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
y (m) A (m2) P (m) R (m) V (m/s) E (m) Sf Avg Sf ∆x (m) Σ∆x (m)
0.86 5.78 8.85 0.65 2.60 1.203 0.007427 0

0.80 5.28 8.58 0.62 2.84 1.211 0.009633 0.008530 0.7 0.7
0.75 4.88 8.35 0.58 3.08 1.233 0.012134 0.010884 2.3 3.0
0.70 4.48 8.13 0.55 3.35 1.271 0.015510 0.013822 6.3 9.3
0.65 4.10 7.91 0.52 3.66 1.334 0.020163 0.017837 28.9 38.2
7.9 SPATIALLY VARIED FLOW

With spatially varied flow in an open channel, the discharge varies along the channel
length, even in otherwise steady flow conditions. This situation occurs when flow is
either added to or diverted from the channel along its length. Flow may be added, for
example, in a side-channel spillway. A stormwater inlet on a continuous grade (dis-
cussed in Section 10.2) is an example where flow is diverted from the channel as it
passes the inlet opening.
Methods for dealing with spatially varied flow are distinguished in terms of whether
flow is added to or diverted from the channel. When flow is added to a channel, as in
the case of a side channel spillway, turbulence generated in the flow is such that
energy losses are greater than those due to friction alone, and the energy equation is
not easily applied. Therefore, in a channel with an increasing discharge, momentum
rather than energy should be used in computing the flow profile. In the case of a chan-
nel with a decreasing discharge, experience has shown that the energy equation can be
applied for flow profile computations because of the general lack of turbulence gener-
ated in this sort of flow.
Spatially Varied Flow with an Increasing Discharge

A derivation of the fundamental differential equation describing a spatially varied
flow with an increasing discharge is accomplished by applying the momentum princi-
ple to the flow in a channel reach. Assuming that the coefficients α and β are unity,
the resulting equation is (Chow, 1959)
2Qq
So − S f −
dy gA2 (7.76)
=
dx 1 − Fr 2
where y = vertical depth of flow (ft, m)

x = distance along channel bottom (ft, m)
So = channel bottom slope (ft/ft, m/m)
q = dQ/dx (cfs/ft, m3/s/m)
g = gravitational acceleration (32.2 ft/s2, 9.81 m/s2)
A = cross-sectional area of channel (ft2, m2)
Fr = Froude number
Note that this equation is identical to Equation 7.71 with the exception of the third
term in the numerator. In the third term, Q, is the discharge, which depends on the
location x along the length of the channel, and q = dQ/dx is the rate at which flow is
added to the channel along its length. Note that q has units of cfs per foot of channel
length, or of m3/s per meter of channel length. The friction slope Sf appearing in the
equation, which accounts for friction along the channel bottom and sides, may be
evaluated using a constitutive relationship such as the Manning equation.
If the depth of flow is known at a control section, a numerical solution of Equation
7.76 may be used to determine the water surface profile in a channel. French (1985)
has shown that the location xc of the control section in a channel (measured from
upstream end of the reach) with Q = 0 when x = 0 is
8q 2
xc =
3
 gn2 Pc4 / 3  (7.77)
gTc2  So −

 C f Tc Ac1/ 3 
where xc = location of control section (ft, m)

Tc = top width of flow at control section (ft, m)
Pc = wetted perimeter at control section (ft, m)
Cf = unit conversion factor (2.21 U.S. Customary units, 1.0 SI units)
Ac = cross-sectional area of flow at control section (ft, m)
Critical depth occurs at the control section location determined by Equation 7.77.
Solution of Equation 7.77 must be accomplished in an iterative, trial-and-error man-
ner. One begins by assuming a value xc, from which Q = qxc. For this discharge, one
determines the corresponding critical depth yc, and then the values of Ac, Pc, and Tc.
Finally, Equation 7.76 is applied to compute a new value of xc. If the computed value
of xc agrees with the assumed one, convergence is attained. Otherwise, the procedure
is repeated with a new trial value.
If the location xc of the control section lies within the length L of the channel that is
subjected to lateral inflow, then flow upstream of that point will be subcritical, and
flow downstream of that point will be supercritical. If xc is greater than L, then flow in
the channel will be subcritical throughout and will be controlled by conditions exist-
ing at the location x = L. Even if xc is less than or equal to L, the critical section may
be “drowned” by downstream conditions.
Spatially Varied Flow with a Decreasing Discharge

The equation describing spatially varied flow with a decreasing discharge is derived
using the energy principle, as shown by Chow (1959). Assuming that the coefficients
α and β are unity, the equation is
Qq
So − S f −
dy gA2 (7.78)
=
dx 1 − Fr 2
Note that this expression is identical to Equation 7.76, except for the coefficient on
the last term in the numerator. Note also that in the case of spatially varied flow with a
decreasing discharge, the quantity q = dQ/dx is negative.
As in the case of Equation 7.75, a numerical method may be employed to solve Equa-
tion 7.78 to yield a flow profile.

Open channels are extremely common in both natural and constructed environments,
and include streams, swales, ditches, gutters, and gravity-flow pipes. Unlike closed-
conduit flow analysis, descriptors such as cross-sectional flow area and average
velocity are unknown variables when evaluating flow in open channels. Their values
depend on channel shape, slope, and roughness characteristics, and on channel bound-
ary conditions.
Open channels may be classified as either prismatic or non-prismatic. Common cross-

sectional shapes for prismatic channels are rectangular, trapezoidal, triangular, and
circular. Channels may also be classified on the basis of whether they are lined or
unlined.
When evaluating open-channel flow, the energy equation takes the form
α1V12 α 2V22
Z1 + d1 cos θ + = Z 2 + d 2 cos θ + + hL (7.79)
2g 2g
Uniform flow in an open channel means that the flow depth, discharge, and other flow
properties are constant. This condition can occur in a sufficiently long prismatic chan-
nel. The depth at which uniform flow occurs is called normal depth. Constitutive rela-
tionships such as the Manning and Chézy equations can, in addition to being used to
compute friction losses, be used to predict the normal depth for a given channel and
discharge.
Critical flow is another condition for which a unique relationship exists between
depth and discharge. It also corresponds to the point of minimum specific energy (the
energy relative to the channel bottom) for a given channel geometry and discharge.
Critical flow often occurs at channel control sections, such as at free overfall loca-
tions. The concept of specific energy of flow can also be applied to the evaluation of
flow properties at channel transitions such as expansions, contractions, and steps.
Momentum and specific force concepts are used in evaluating the depths upstream
and downstream of hydraulic jumps, and in computing the forces on hydraulic struc-
tures such as channel transitions and bridge piers. The momentum equation for open-
channel flow is written as
∑ Fx = P1 −P2 + W sin θ − F f = ρ Q( β 2V2 − β1V1 ) (7.80)

References 281
For a short, horizontal channel reach, specific force can be computed as
 Q2 
Fs = γ  zA +  (7.81)
 gA 

Weirs are used to control or measure flow in open channels. Weirs may be classified
as either sharp- or broad-crested. They may also be classified by shape; common
shapes are rectangular and triangular. Finally, weirs can be classified as to whether the
weir flow is contracted or suppressed. Equations and coefficients used in computing
the flow or head for some types of weirs were presented in the chapter.
When depth and velocity gradually change along the length of a channel, the condi-
tion is termed gradually varied flow. Evaluation of gradually varied flow involves
classifying the type of profile based on the relative values of normal depth and critical
depth for the channel, and on the location of the water surface relative to these depths.
Once the profile type is known, corresponding flow depths and locations along the
channel reach are determined.
If discharge varies along a channel reach, spatially varied flow exists. If the discharge
increases along the channel reach, momentum concepts (rather than energy) should be
used to evaluate the flow with a trial-and-error procedure. If the discharge is decreas-
ing, energy concepts can be used.
REFERENCES
Abbott, M. B. 1979. Computational Hydraulics: Elements of the Theory of Free Surface Flows. London:
Pitman.
American Concrete Pipe Association (ACPA). 2000. Concrete Pipe Design Manual. 13th Printing
(revised). Irving, Texas: ACPA.
Bazin, H. 1897. “A New Formula for the Calculation of Discharge in Open Channels.” Mémoire No. 41,
Annales des Ponts et Chausées, Paris 14, Ser. 7, 4th trimester: 20–70.
Blevins, R. D. 1984. Applied Fluid Dynamics Handbook. New York: Van Nostrand Reinhold.
Chow, V. T. 1959. Open-Channel Hydraulics. New York: McGraw Hill.
Chaudhry, M. H. 1993. Open-Channel Flow. Englewood Cliffs, New Jersey: Prentice Hall.
Chaudhry, M. H., and S. M. Bhallamudi. 1988. “Computation of Critical Depth in Compound Channels.”
Journal of Hydraulic Research 26, no. 4: 377–395.
Cunge, J. A., F. M. Holly, and A. Verwey. 1980. Practical Aspects of Computational River Hydraulics.
London: Pitman.
Einstein, H. A. 1934. “The Hydraulic or Cross Section Radius.” Schweizerische Bauzeitung 103, no. 8: 89–
91.
Einstein, H. A., and R. B. Banks. 1951. “Fluid Resistance of Composite Roughness.” Transactions of the
Finnemore, E. J. and J. Franzini. 2000. Fluid Mechanics with Engineering Applications. New York:
McGraw-Hill.
Franzini, J. and E. J. Finnemore. 1997. Fluid Mechanics, 9th Ed. New York: McGraw-Hill.
Fread, D. L 1993. “Flow Routing” in Maidment, D. R., ed. Handbook of Hydrology. New York: McGraw-
Hill.
French, R. H. 1985. Open-Channel Hydraulics. New York: McGraw-Hill.
Ganguillet, E., and W. R. Kutter. 1869. “An Investigation to Establish a New General Formula for Uniform
Flow of Water in Canals and Rivers” (in German), Zeitschrift des Oesterreichischen Ingenieur und
Architekten Vereines vol. 21, no. 1: 6–25; no. 2: 46–59. Translated into English as a book by R. Hering
and J.C. Trautwine, Jr., A General Formula for the Uniform Flow of Water in Rivers and Other Chan-
nels, John Wiley and Sons, New York, 1888.
eties 18: 363–368.
Horton, R. 1933. “Separate Roughness Coefficients for Channel Bottom and Sides.” Engineering News-
Record 111, no. 22: 652–653.
Krishnamurthy, M., and B. A. Christensen. 1972. “Equivalent Roughness for Shallow Channels.” Journal
of the Hydraulics Division, ASCE 98, no. 12: 2257–2263.
Lotter, G. K. 1933. “Considerations of Hydraulic Design of Channels with Different Roughness of Walls.”
Transactions All Union Scientific Research, Institute of Hydraulic Engineering 9: 238–241.
Manning, R. 1891. “On the Flow of Water in Open Channels and Pipes.” Transactions, Institution of Civil
Motayed, A. K., and M. Krishnamurthy. 1980. “Composite Roughness of Natural Channels.” Journal of the
Hydraulics Division, ASCE 106, no. 6: 1111–1116.
Muhlhofer, L. 1933. “Rauhigkeitsunteruchungen in einem stollen mit betonierter sohle und unverkleideten
wanden.” Wasserkraft und Wasserwirtschaft 28(8): 85–88.
Pavlovskii, N. N. 1931. “On a Design Formula for Uniform Movement in Channels with Nonhomogeneous
Walls” (in Russian), Izvestiia Vsesoiuznogo Nauchno-Issledovatel’skogo Insituta Gidrotekhniki 3: 157–
164.
Powell, R. W. 1950. “Resistance to Flow in Rough Channels.” Transactions, American Geophysical Union
31, no. 4: 575–582.
Rouse, H. 1936. “Discharge Characteristics of the Free Overfall.” Civil Engineering 6(7): 257.
Rouse, H. 1946. Elementary Fluid Mechanics. New York: Wiley.
Strickler, A. 1923. “Some Contributions to the Problem of the Velocity Formula and Roughness Factors for
Rivers, Canals and Closed Conduits” (in German). Mitteilungen des Eidgenossischen Amtes fur Wasser-
wirtschaft 16.
U.S. Army Corps of Engineers (USACE). 1991. HEC-2, Water Surface Profiles, User's Manual. Davis,
California: USACE Hydrologic Engineering Center.
U.S. Bureau of Reclamation (USBR). 1955. Research Studies on Stilling Basins, Energy Dissipators, and
Associated Appurtenances, Hydraulic Laboratory Report No. Hyd-399. Lakewood, Colorado: USBR.
Problems 283
PROBLEMS
7.1 From the natural channel cross-section data given in the table, determine the area, wetted perimeter,
hydraulic radius, top width, and hydraulic depth if the depth of flow is 4 ft.
Station (ft) Elevation (ft)

0 10.1
5 7.8
15 5.5
18 6.0
22 8.2
26 9.8
30 10.7
7.2 Determine the area of flow, wetted perimeter, hydraulic radius, top width, and hydraulic depth for
each channel described. The flow depth for all channels is 1.22 m.
a) rectangular channel with a bottom width of 1.5 m
b) triangular channel with side slopes of 2 horizontal to 1 vertical
c) semi-circular channel with a radius of 2.1 m
d) trapezoidal channel with a bottom width of 3.0 m and side slopes of 3 horizontal to 1 vertical
7.3 Demonstrate that the hydraulic depth Dh in a rectangular channel is equal to the actual depth of flow.
Also demonstrate that the hydraulic depth is smaller than the actual depth of flow in trapezoidal and
triangular channels.
7.4 Determine the discharge in a 3-ft diameter sewer pipe if the depth of flow is 1 ft and the slope of the
pipe is 0.0019. Use a value of 0.012 for the Manning roughness coefficient.
7.5 An asphalt-lined trapezoidal channel has the dimensions indicated in the figure. The channel slope is
0.001. Use the Chézy equation to find the discharge if B = 4.9 m and y = 1.4 m.
7.6 By comparing the Manning and Chézy equations, show that the relationship between n and C is C =
1.49R1/6/n if R is in feet.
7.7 Explain why uniform flow cannot occur in frictionless and horizontal channels.
7.8 Determine the normal and critical depths in a triangular channel if Q = 5 cfs, n = 0.045, and
So = 0.008. Is uniform flow in this channel subcritical or supercritical?
7.9 A 36-in. diameter pipe with n = 0.013 and So = 0.01 conveys a discharge of 25 cfs. Determine the
depth and velocity of flow in the pipe.
7.10 Determine the critical depth in a trapezoidal channel if the discharge is 14.2 m3/s. The width of the
channel bottom is 6.1 m, and the side slopes are 1:1.
7.11 In the triangular section shown, the flow is 350 cfs. Using n = 0.012, calculate the critical depth and
the critical velocity.
7.12 The discharge is 7.1 m3/s in a 3-m-wide rectangular section. Plot the specific energy as a function of
depth of flow.
7.13 Calculate the critical depth of the channel in Problem 7.12 using Equation 7.38.
7.14 Determine the sequent depth and length of a hydraulic jump in a rectangular channel with B = 5 ft, Q
= 150 cfs, n = 0.013, and So = 0.05. Flow upstream of the jump is uniform. What is the energy loss in
the jump?
7.15 Water flows through a rectangular section at a rate of 14.2 m3/s. The channel is 6.1 m wide, and the
upstream depth is 0.94 m. The flow undergoes a hydraulic jump as indicated in the diagram. Com-
pute the:
a) downstream depth
b) energy loss in the hydraulic jump
Problems 285
7.16 A sharp-crested, contracted, rectangular weir (N = 2) is 20 ft long. The height of the weir crest above
the channel is 8 ft. Develop and plot a rating curve (that is, a relationship between Q and H) for this
weir. Use H values ranging from 0 to 3 ft to develop the rating curve.
7.17 Water flows in a 6.1-m-wide rectangular concrete channel at a rate of 14.2 m3/s. The slope changes
from 0.0125 to 0.05. Plot the water surface profile through the transition.
Stormwater is
conveyed through
this rectangular,
concrete channel,
which is fenced for
safety reasons.
C H A P T E R
8
Design of Open Channels
Chapter 7 presented methods of open-channel flow analysis. It focused on fundamen-

tal hydraulic principles and how they could be applied to analyze the characteristics of
a flow in a specified channel. This chapter focuses on channel design, with a corre-
sponding emphasis on determining appropriate channel geometry and materials. The
design methods presented in this chapter are applicable to open channels such as
ditches, gullies, and streams. (Chapter 9 covers culvert design, and Chapters 10 and
11 cover the design of storm sewer systems.) The techniques presented in this chapter
are further limited to uniform flow in prismatic channels. For instances in which the
flow is varied (non-uniform), design modifications, such as increasing the channel
depth, must be made in order to ensure adequate channel capacity.
Section 8.1 of this chapter discusses the various issues that should be considered
when designing an open channel. Some of the issues relate to channel hydraulics, and
others relate to public safety, right-of-way requirements, maintenance, and channel
alignment.
The concept of the best hydraulic section as a basis for channel design is briefly dis-
cussed in Section 8.2. Channel design using this approach provides optimal hydraulic
performance in that the cross-sectional area of the channel is maximized for a fixed
wetted perimeter, thus minimizing the resistance to flow. Channels designed using
this method, however, must be carefully examined for feasibility in light of other
design considerations.
For design purposes, a channel may be classified according to three basic types: lined
channels, unlined channels, and grassed channels. Lined channels have erosion-
resistant materials lining their bottom and sides. Most commonly, they are con-
structed using concrete or asphalt as the lining material. In unlined channels, also
referred to as erodible channels, the channel bottom and sides consist of unprotected,
natural soil materials. Grassed channels have intermittent flows such that grass or
similar vegetation can establish itself in the channel during no-flow periods. Various
combinations of these three basic types also exist; for instance, a channel may have a
concrete bottom for frequent, lower flows and, above the concrete, grassed sides for
higher flows. Design methods for lined, erodible, and grassed channels are covered in
288 Design of Open Channels Chapter 8
Sections 8.3, 8.4, and 8.5, respectively. Design of riprap lined channels is discussed in
Section 8.6
Frequently, open-channel design involved modifying an existing channel rather than
creating a completely new drainageway. Most often, the purpose of the modifications
is to increase channel capacity. Information on designing modifications to existing
channels is presented in Section 8.7.
8.1 DESIGN CONSIDERATIONS

Proper design of an open channel involves careful hydraulic analysis, as well as an
understanding that the constructed channel will be only a part of the larger urban
infrastructure. Therefore, the design of an open channel should involve not only con-
sideration of technical criteria related to flow velocities and the like, but also issues
such as public safety, aesthetics, and maintainability. Brief discussions of a number of
design considerations are provided in the following paragraphs.
Velocity of Flow
The velocity at which various flows will travel in an open channel is an important
design consideration for a number of reasons. Excessively high velocities pose a pub-
lic safety hazard. They also tend to cause channel erosion even in lined channels, par-
ticularly if the water is transporting a significant amount of sediment or debris. If
velocities are too small [less than about 2 to 3 ft/s (0.6 to 0.9 m/s)], sediments and
debris may settle out of the water column, with the net effect being that the channel
will require regular cleaning. Biological slime and/or vegetation growth may also
become problematic if flow velocities are too small.
Section 8.1 Design Considerations 289
Froude Number
The Froude number of a flow should always be checked when designing an open
channel. Except for cases in which drop structures, spillways, or other grade-change
facilities are being addressed, channel design generally seeks to establish a subcritical
flow regime where Fr < 1. Also, because critical or near-critical flow tends to be
unstable and may be accompanied by a wavy water surface, the Froude number corre-
sponding to a particular design should not be close to unity (that is, it should not be
between approximately 0.9 and 1.1−1.2).
Side Slopes
The side slopes of a channel should be selected with regard to maintainability, safety,
and slope stability. Channels that require periodic mowing or other maintenance must
have side slopes that are flat enough to accommodate equipment ingress and egress.
In many areas, side slopes of 3 horizontal to 1 vertical are used as a rule of thumb for
channel design to facilitate mower operation. If side slopes are too steep, and espe-
cially if flow velocities are also high, it may be difficult for someone who falls into
the channel to get out. A 3 to 1 slope is fairly mild and fairly easy to navigate when it
is dry, but may present more of an obstacle when it is wet or, worse still, when water
is flowing.
Soil materials have a natural angle of repose (the maximum slope for which they are
stable) that should not be exceeded. This angle may even have to be reduced when
designing channel side slopes due to seepage forces that develop as the soils drain fol-
lowing a large flow event.
Longitudinal Slope
The longitudinal slope of a channel is governed in large part by local topography, but
there are other considerations to bear in mind as well. For example, if an existing nat-
ural channel is straightened and/or reshaped to improve its hydraulic efficiency, the
velocity of flow usually increases and may pose erosion problems. In such instances,
a designer may wish to decrease the channel slope and provide one or more drop
structures where necessary to make up required channel bed elevation changes.
The engineer must also consider that a channel conveys a range of flows, not just a
single design discharge. Even if a channel is designed to convey the design discharge
with a subcritical flow regime, supercritical flow may still occur at other discharges.
The longitudinal slope should be selected such that all possible flows are subcritical.
Freeboard
The total depth of an open channel from the bed to the top of either bank should, of
course, be greater than the depth of the design flow. Freeboard is difference in height
between the lower of the banks and the free water surface, and is essentially a factor
of safety. While various sources may provide some guidance on freeboard require-
ments, the selection of a freeboard height remains largely a professional judgment
issue. Freeboard heights are usually at least 1 ft (0.3 m). This value often increases
when flows have large velocities or depths, or if there is a curve in the horizontal
alignment of the channel that causes the water surface to superelevate (that is, to be
higher on one side of a cross section).
Lining Material
Selection of the channel lining material or type of grass cover, if any, is a major con-
sideration in channel design. Channels lined with concrete, asphalt, or other imperme-
able materials are hydraulically efficient, but are also more costly to construct, are
frequently less aesthetically pleasing, and have little value as habitat. They may also
require consideration of uplift forces such as buoyancy and frost heave. In grass-lined
channels, selection of the grass species is a design consideration. Sod-forming
grasses, once established, provide uniform protection to the channel bottom and sides.
Bunch-forming grasses such as alfalfa may channelize flows between grass bunches,
causing localized erosion.
Cost and Maintenance

The life-cycle cost of a channel design includes not only the capital costs associated
with engineering and construction, but also long-term maintenance costs. Evaluations
of alternative designs should seek to minimize the total life-cycle cost, but design
objectives and/or constraints other than cost may ultimately dictate the final design
selection.
Safety
A design issue of paramount importance, especially in urban and suburban settings, is
public safety. Channels (and water in general) often become magnets for leisure and
recreational activities, with the net effect being that humans often come into contact
with flowing waters. Children especially are attracted to open streams, and are more
vulnerable than adults to the hazards posed by large flow velocities and/or depths. As
mentioned previously, whenever possible, channel sides slopes should be flat enough
for a person to climb out. Steps or ladder rungs similar to those in sewer manholes can
be installed if necessary to facilitate egress. It may be appropriate to use fencing along
larger channels, channels that experience high velocities, and channels with steep
sides in order to keep people away, especially in urban areas. The U.S. Bureau of Rec-
lamation (1974) and Aisenbrey et al. (1978) provide some safety-related guidance.
Aesthetics
More and more frequently, open channels and other types of water bodies are becom-
ing centerpieces for parks and similar cultural centers. Greenways with trails and bike
paths often follow streams. In such instances, the aesthetics of a design are typically
among the most significant evaluation criteria, and other issues such as hydraulic effi-
ciency are of less concern.
Section 8.2 Best Hydraulic Section 291
The aesthetics of a channel can be enhanced to make it more pleasing and natural in
appearance. Materials such as stones and boulders can be placed in the bottom and on
the sides. Where possible, the vertical alignment can be designed to allow for drops
such as waterfalls, and a naturally meandering horizontal alignment can be used. A
liberal amount of vegetation along the channel edges adds to its natural character. A
landscape architect can assist in the design of a channel for which aesthetic consider-
ations are important.
Right-of-Way Requirements
In urban and suburban areas where land values may be high and space at a premium,
right-of-way requirements can be a major consideration in channel design. This is true
both for areas that are just becoming urbanized and for already-established areas
where retrofits or upgrades are being made to the stormwater conveyance system.
For areas where available right-of-way is limited, one should clearly seek to develop a
channel design with a small top width. A rectangular cross section may be used, but
typically the channel must then be lined with concrete, which can increase the chan-
nel construction cost considerably. This type of channel may also give rise to prob-
lems with maintainability and safety.
Dry-Weather Flows
Open channels must be capable of conveying large design storm discharges, but they
also should be designed with due regard for relatively small dry-weather flows (also
called base flows). Particularly in park-like or greenbelt settings where grassed chan-
nels are often used for conveyance, it may be desirable to construct a small low-flow
channel or trickle channel using riprap or other lining material. This practice helps to
define the path of low flows, preventing them from meandering and shifting with time
across the bottom of the larger channel in which they reside.
Clearly, not all of the considerations described above will arise in every channel
design. In some cases, there may be additional considerations not cited here. The
important point is that there are many issues that must be considered in design, and
tradeoffs occur regularly. A critical aspect of the design process is to understand and
weigh the tradeoffs and associated consequences so that a well-balanced design may
be selected for implementation.
8.2 BEST HYDRAULIC SECTION

If a channel’s longitudinal slope and roughness characteristics are fixed, maximiza-
tion of its hydraulic efficiency amounts to either maximizing the cross-sectional area
for a fixed wetted perimeter or, equivalently, minimizing its wetted perimeter for a
fixed cross-sectional area. This is observed by noting that, with all other things equal,
a large cross-sectional area yields a greater flow capacity than does a small area. Sim-
ilarly, a large wetted perimeter exerts a greater frictional resistance to flow than does a
small wetted perimeter. The best hydraulic section is defined as a cross section having
the highest possible ratio of cross-sectional flow area to wetted perimeter.
Of all possible geometric shapes that could be selected as the cross section for an
open channel, a semicircle with the flat face corresponding to the water surface is the
“best” hydraulically. The semicircular shape results in the largest possible cross-
sectional area for a fixed wetted perimeter, or the smallest possible wetted perimeter
for a fixed cross-sectional area.
When one considers non-semicircular shapes such as rectangles, triangles, and trape-
zoids, one finds that the best dimensions for these shapes are those that most closely
approximate a semi-circle. For example, the best rectangular cross section is one that
is the bottom half of a square (that is, one for which the depth of flow is one-half the
channel width). The best triangular cross section is the bottom half of a square tilted
to resemble a diamond shape (that is, one with side slopes inclined at 45 degrees). The
best trapezoidal cross section is the bottom half of a hexagon, such that its side slopes
are inclined 60 degrees from the horizontal. Figure 8.1 illustrates these cross sections.
Figure 8.1
Best hydraulic cross-
sections for various
channel shapes
Because the best hydraulic section approximates a semicircle, its side slopes are quite
steep, if not vertical. Unfortunately, this geometry limits the practical use of the best
hydraulic section to instances in which the channel is either to be lined or cut through
solid rock. Even in instances where the side slopes can be steep or vertical, design
considerations such as safety and maintainability may limit the practicality of the best
hydraulic section.
Example 8.1 – Computing the Best Hydraulic Section. Use the concept of the best
hydraulic section and the Manning equation (see Section 7.2, page 226) to design a concrete-lined
trapezoidal channel to convey Q = 3.0 m3/s with n = 0.013 and So = 0.0015. Check the velocity and
Froude number of the flow, and add 20 percent of the computed depth of flow as freeboard.
Solution: The best trapezoidal cross section is half of a hexagon. Letting B denote the cross-section
bottom width (that is, the length of one side of the hexagon), y the flow depth, and P the wetted
perimeter, it can be observed from Figure E8.1.1 that
P = B + B + B = 3B
Section 8.2 Best Hydraulic Section 293
y = B sin 60° = 0.866B
A = By + (B/2)y = 1.299B2
T = B + 2(B/2) = 2B
Figure E8.1.1 Channel cross section for Example 8.1
Substituting into the Manning equation (Equation 7.9), the discharge can be expressed as
1 2/3 1/ 2 1
Q =3= AR So = (1.299B2 )5/3 (3B)−2/ 3 (0.0015)1/ 2 = 2.215B8/ 3
n 0.013
Solving this equation,
B = (3/2.215)3/8 = 1.12 m
The other hydraulic dimensions are
P = 3.36 m
y = 0.97 m
A = 1.63 m2
T = 2.24 m
The average flow velocity is
V = 3/1.63 = 1.84 m/s
This velocity is not excessively high, and hence the design would likely be considered adequate on
the basis of this consideration.
The Froude number is computed as (from Equation 7.39):
Q 2T 32 ( 2.24 )
Fr = = = 0.69
3 3
gA 9.81(1.63 )
This is well into the subcritical range and again leads to a positive evaluation of this design.
Using 20 percent of the depth as the freeboard criterion, the freeboard height is computed to be
0.2 × 0.97 = 0.19 m
Thus, the total channel depth is
0.97 + 0.19 = 1.16 m
8.3 Design of Lined Channels

From the point of view of determining channel cross-section dimensions, the design
of a lined channel is the most straightforward. The steps that follow outline a design
procedure for a specified design discharge Q.
1. Determine Manning’s n and the channel slope So. These selections will depend
on the lining material chosen, as well as on topography.
2. Using the design discharge and the values determined in step 1, rearrange the
Manning equation (Equation 7.9) such that the unknown variables describing
channel geometry are on the left side and known variables are on the right. Com-
pute the channel section factor AR2/3.
nQ
AR 2 / 3 = (8.1)
C f So
where A = cross-sectional area of channel (ft2, m2)

Cf = unit conversion factor (1.49 for U.S. customary units; 1.00 for SI)
3. Select the channel dimensions B and/or z and write an expression for the section
factor AR2/3 in terms of those dimensions and the depth of flow. Figure 8.2
(repeat of Figure 7.4) and Table 8.1 (repeat of Table 7.1) may be consulted to aid
in this step.
4. Equate the value of the section factor determined in step 2 to the expression writ-
ten in step 3, and solve for the normal depth of flow, yn. This solution usually
must be accomplished by trial and error.
5. Check the flow velocity and Froude number for acceptability and add freeboard
to obtain the required channel depth.
6. Develop rating curves or tables to check velocity and flow depth for a wide range
of possible flow rates, and consider whether a low-flow channel is needed.
For lined channels constructed using unreinforced concrete, ASCE (1992) has recom-
mended that the flow velocity not exceed 7 ft/s (2.1 m/s). In reinforced concrete chan-
nels, the flow velocity should not exceed 18 ft/s (5.5 m/s). In either case, the velocity
should not be less than about 2 to 3 ft/s (0.6 to 0.9 m/s). The Froude number should be
in the subcritical range, and should not be close to unity (1.0).
Section 8.3 Design of Lined Channels 295
Figure 8.2
Common open-
channel shapes
Table 8.1 Cross-sectional properties of prismatic open channels
Channel Shape Area (A) Wetted Perimeter (P) Hydraulic Radius (R) Top Width (T) Hydraulic Depth (Dh)
By
Rectangular By B + 2y B y
B + 2y
By + zy 2
Trapezoidal By +zy 2 B +2 y 1 + z 2 B +2 y 1 + z 2 B + 2zy
B + 2 zy
zy y
Triangular zy2 2 y 1+ z2 2zy
2 1 + z2 2
D 2 (φ − sin φ ) Dφ D  sin φ  φ  D  φ − sin φ 

Circulara 1 −  D sin    
8 2 4 φ  2 8  sin(φ / 2) 
Although a general expression for required freeboard height has not been developed,
Chow (1959) indicates that the U.S. Bureau of Reclamation has suggested that an
estimate may be made using the expression
F = Cy (8.2)
where F = required freeboard height (ft)

C = coefficient based on discharge
y = depth of flow (ft)
The coefficient C ranges from 1.5 for a discharge of about 20 cfs (0.6 m3/s) to 2.5 for
discharges in excess of 3,000 cfs (85 m3/s). This estimate is based on average U.S.
Bureau of Reclamation practice and should not be expected to apply to all conditions.
Freeboard requirements for lined channels are usually at least 1 ft (0.3 m) (ASCE,
1992).
At channel bends, where superelevation of the free water surface occurs, freeboard
usually needs to be increased. For subcritical flow, the superelevation height of the
water surface can be approximated as (Chow, 1959)
V 2T
hs = (8.3)
grc
where hs = superelevation height (ft, m)

V = flow velocity (ft/s, m/s)
T = top width (ft, m)
g = gravitational acceleration (32.2 ft/s2, 9.81 m/s2)
rc = radius of curvature of the channel center line (ft, m)
Example 8.2 – Lined Channel Design. Design a concrete-lined open channel to convey a
maximum design discharge of 500 cfs. The channel is to be trapezoidal in shape with z = 2, and its
slope will be So = 0.001.
Solution: Referring to Table 7.3, a typical n value for concrete of 0.013 is selected. This value, along
with the stated channel slope and discharge, are used to compute the section factor as
0.013(500)
AR 2 / 3 = = 138
1.49 0.001
Using a trial channel bottom width of B = 10 ft, one can use the expressions in Table 8.1 to write the
section factor as
2/3
 10 y + 2 y 2 
2/3 2
AR = 138 = (10 y + 2 y )   = (10 y + 2 y 2 )5 / 3 (10 + 2 y 5 ) −2 / 3
 10 + 2 y 1 + 22 
 
Solving this expression by trial and error, one obtains
yn = 4.04 ft
For this depth of flow, other hydraulic dimensions are
A = 73.0 ft2
T = 26.2 ft
The flow velocity is
V = 500/73.0 = 6.85 ft/s
and the Froude number is
Section 8.4 Design of Erodible Channels 297
Q 2T 5002 ( 26.2 )
Fr = = = 0.72
3 3
gA 32.2 ( 73.0 )
Both of these values are within acceptable ranges. Applying linear interpolation to the range of values
for C provided in the text (for Equation 8.2), C is computed to be 1.7, and the required freeboard
height is estimated as
F = [1.7(4.04)]1/2 = 2.62 ft
Thus, the total channel depth should be about 6.7ft from its bottom to the tops of the banks. Figure
E8.2.1 illustrates the designed cross section.
Figure E8.2.1 Channel design for Example 8.2
The step-by-step procedure for a lined channel design may be repeated for a number
of alternatives using different lining materials and different channel dimensions and
shapes. The engineer evaluates the alternative designs, and then chooses the best one
for implementation.
8.4 Design of Erodible Channels

The bottom and sides of an erodible channel consist of natural soil materials that are
not protected by a rigid lining or vegetation. Depending on flow characteristics, sedi-
ments may be either scoured from or deposited onto the channel boundaries. One can
therefore speak of design of channels that scour but do not silt, channels that silt but
do not scour, or channels that silt and scour simultaneously. Only the first of these
three types is considered here. The latter two types require that sediment transport
issues be addressed, which is beyond the scope of this text.
Two approaches for designing erodible channels are used by hydraulic engineers. The
first approach is based on the idea that excessive velocities are the cause of scour and
seeks to design a channel on the basis of a maximum permissible velocity. The second
approach, which is based on a more mechanistic understanding of sediment transport
processes, seeks to design a channel on the basis of a maximum permissible shear
stress, tractive force, or tractive tension. Each of these design approaches is discussed
in the subsections that follow.
Method of Permissible Velocity

The permissible velocity in an open channel is defined as the maximum average
velocity for which soil particles on the channel boundaries will not be set into motion.
The permissible velocity depends on many factors and therefore has significant uncer-
tainty associated with it. It tends to be higher for old channels than for new ones, and
higher for deep channels than for shallow ones.
The permissible velocity depends strongly on the characteristics of the soil making up
the channel boundaries, and on whether the water is clear or is transporting colloidal
silts. For noncohesive soils such as sands, the permissible velocity depends mainly on
the median grain size. On the other hand, the permissible velocity for a cohesive soil
depends on the void ratio of the material, as well as on cohesive and/or electrostatic
forces within the soil material.
Table 8.2 lists permissible velocities assembled by Fortier and Scobey (1926) using
data from canals. Values given in the table are applicable to aged channels with small
slopes. Also shown in the table are suitable Manning’s n values for the soil materials
listed. Note that the permissible velocities for water transporting colloidal silts are
greater than the corresponding velocities for clear water. Figure 8.3 illustrates
U.S.S.R. data on permissible velocities for cohesive soils (Chow, 1959). It can be seen
that the soil type itself has some influence on the permissible velocity, but the voids
ratio (or degree of soil compaction) is the most important parameter.
An alternative to Table 8.2 and Figure 8.3 is to employ Figure 8.4, which shows the
permissible velocity as a function of mean sediment size. This latter figure is based
mainly on the work of Hjulstrom (1935) and has been suggested for use by the ASCE
Sedimentation Task Committee (ASCE, 1977). Note that the permissible velocity is at
a minimum for a mean sediment size of approximately 0.2 mm. Sediments smaller
than 0.2 mm tend to be associated with cohesive materials, which have cohesive and/
or electrostatic forces that tend to resist particle motion. Larger sediments tend to cor-
respond to larger permissible velocities because, by virtue of their mass, they are
more resistant to motion.
Table 8.2 Permissible velocities in open channels (Fortier and Scobey, 1926)
Permissible Velocity
Clear Water Water with Colloidal Silts
Material n ft/s m/s ft/s m/s
Fine sand, colloidal 0.02 1.50 0.46 2.50 0.76
Sandy loam, noncolloidal 0.02 1.75 0.53 2.50 0.76
Silt loam, noncolloidal 0.02 2.00 0.61 3.00 0.91
Alluvial silts, noncolloidal 0.02 2.00 0.61 3.50 1.07
Ordinary firm loam 0.02 2.50 0.76 3.50 1.07
Volcanic ash 0.02 2.50 0.76 3.50 1.07
Stiff clay, very colloidal 0.025 3.75 1.14 5.00 1.52
Alluvial silts, colloidal 0.025 3.75 1.14 5.00 1.52
Shales and hardpans 0.025 6.00 1.83 6.00 1.83
Fine gravel 0.02 2.50 0.76 5.00 1.52
Graded loam to cobbles, noncolloidal 0.03 3.75 1.14 5.00 1.52
Graded silts to cobbles, colloidal 0.03 4.00 1.22 5.50 1.68
Coarse gravel, noncolloidal 0.025 4.00 1.22 6.00 1.83
Cobbles and shingles 0.035 5.00 1.52 5.50 1.68
Figure 8.3
U.S.S.R. data on
permissible velocities
for cohesive soils
Credit: Chow, 1959
The scatter of the data points in Figure 8.4, combined with the differences in the illus-
trated curves, reinforce the fact that the permissible velocity is highly uncertain and is
affected by variables other than just sediment size. For design purposes, one may rea-
sonably use the mean Hjulstrom curve shown in the figure.
The permissible velocities indicated in Table 8.2 and in Figures 8.3 and 8.4 are appli-
cable to depths of flow equal to about 3 ft (0.9 m). For depths greater or smaller than
this, the permissible velocity should be modified using a correction factor such that
Vc = kV (8.4)
where Vc = corrected permissible velocity (ft/s, m/s)

k = correction factor
V = permissible velocity for flow depth of 3 ft (0.9 m)
Figure 8.5, from U.S.S.R. data (Chow, 1959), shows values for the correction factor k
as a function of average depth. These correction factors are applicable to both cohe-
sive and noncohesive soil materials. It can be seen that the correction factor is greater
than unity for depths of more than 3 ft (0.9 m), and less than unity for depths less than
3 ft. Therefore, depths shallower than 3 ft have lower permissible average velocities,
and those deeper than 3 ft have higher permissible average velocities. This can be
rationalized in terms of the typical shape of the velocity distribution in an open chan-
nel; other factors being equal, the average velocity will be higher in a deeper channel.
Mehrota (1983) indicated that the correction factor, denoted by k, can be computed as
k = y1/6 (8.5)
where y = depth of flow (m)
Figure 8.4
Permissible velocity
as a function of mean
sediment size
Credit: ASCE, 1977. Reproduced by permission of ASCE.
Figure 8.5
U.S.S.R. data on
correction factor (k)
that may be applied to
permissible velocity
Credit: Chow, 1959

The permissible velocities described previously are for straight channels. In sinuous
channels, flow can impinge on the outside bank in bends, and thus the permissible
velocity should be reduced. Lane (1955) suggested a 5-percent reduction for mildly
sinuous channels, a 13-percent reduction for moderately sinuous channels, and a 22-
percent reduction for highly sinuous channels.
Based on the estimate of permissible velocity described previously, the design of a
trapezoidal erodible channel can proceed through the following steps:
1. Estimate Manning’s n and the channel side slope z to be used in the design.
Table 8.2 and/or Table 7.3 may be used to guide a selection of Manning’s n. Rec-
ommended maximum side slopes for various soil materials are presented in
Table 8.3.
2. Using the estimated permissible velocity V and the values of n and z determined
in step 1, apply the Manning equation (Equation 7.9) to compute the required
hydraulic radius:
3/ 2
 nV 
R=  (8.6)
 C f So 
 
where R = hydraulic radius (ft, m)

V = average velocity (ft/s, m/s)
Cf = unit conversion factor (1.49 for U.S. customary units; 1.00 for SI units)
3. Compute the values of the required cross-sectional area and wetted perimeter as
A = Q/V and P = A/R, where Q is the design discharge.
4. Write expressions for A and P using Table 8.1 and substitute the previously deter-
mined values of A, P, and z into these expressions. Solve the resulting two expres-
sions simultaneously for the values of B and yn.
5. Check the Froude number, add freeboard, and modify the channel bottom width
B for practicality.
Example 8.3 illustrates this procedure.
Table 8.3 Recommended maximum side slope, z, in various materials (Chow, 1959)
Material Maximum side slope (z)
Rock Nearly vertical
Muck and peat soils 0.25
Stiff clay or earth with concrete lining 0.5–1
Earth with stone lining, or earth for large channels 1
Firm clay or earth for small ditches 1.5
Loose sandy earth 2
Sandy loam or porous clay 3
Example 8.3 – Erodible Channel Design with the Permissible Velocity

Method. Using the method of permissible velocity, design a trapezoidal channel to convey a dis-
charge of Q = 400 cfs. The channel bottom and sides will be composed of non-colloidal coarse gravel,
and the channel slope will be So = 0.0015.
Solution: Because coarse gravel is similar to a sand, the maximum allowable side slope is estimated
from Table 8.3 as z = 2. Manning’s n and the permissible velocity are estimated from Table 8.2 as n =
0.025 and V = 4 ft/s. In the interest of developing a conservative design, the water is assumed to be
clear.
The required hydraulic radius is computed from Equation 8.6 and is
3/ 2
 0.025(4) 
R=  = 2.28 ft
 1.49 0.0015 
The required area and wetted perimeter are

A = 400/4 = 100 ft2
P = 100/2.28 = 43.8 ft
Using the expressions for a trapezoidal channel in Table 8.1, the following equations may be written:
A = 100 = By + 2y2
Py = 43.8 y = By + 2 y 2 5
Note that the value of z = 2, determined in step 1, has been used in these equations. Note also that the
equation for wetted perimeter has been multiplied by y. Simultaneous solution of these equations can
be accomplished by subtracting the second from the first to obtain
2 2
A – Py = 100 – 43.8y = 2y ( 1 – 5 ) = – 2.47y
or
2.47y2 – 43.8y + 100 = 0
The two roots of this quadratic equation are y = 2.69 ft and y = 15.04 ft. Using each of these two val-
ues of y, and knowing that the cross-sectional area should equal 100 ft2, one can determine the chan-
nel bottom width B. For the depth y = 2.69 ft, the required value of B is
2 2
A – zy 100 – 2 ( 2.69 )
B = ----------------- = ------------------------------------ = 31.8 ft
y 2.69
For the depth y = 15.04 ft, the corresponding value of B is –23.4 ft, clearly an invalid result. There-
fore, one concludes that the channel bottom width should be approximately 32 ft and the depth of uni-
form flow will be about yn = 2.69 ft. The Froude number of this flow is approximately 0.46.
From Equation 8.2, the required freeboard for this channel is estimated to be about 2 ft; thus, the total
channel depth should be about 4.7 ft. Figure E8.3.1 illustrates the completed design.
Figure E8.3.1 Channel design for Example 8.3

Note that because the depth of flow in this example is fairly close to 3 ft (0.9 m), there is no need to
correct the permissible velocity using Figure 8.5 or Equation 8.4. However, had one decided to cor-
rect the permissible velocity, a second iteration of the illustrated calculations would yield a refined
design.
Method of Tractive Force

Equation 7.17 showed that the average shear stress acting on the boundary of an open
channel can be expressed as
τ = γ RSo (8.7)
where τ = average shear stress (lb/ft2, N/m2)

γ = specific weight of water (lb/ft3, N/m3)
The shear stress, also called the tractive force, is the force applied per unit area of the
channel bottom and sides. This force will mobilize individual soil particles and lead to
scour if it is sufficiently large. As channel width increases (relative to depth), the
value of the hydraulic radius approaches that of the flow depth, and Equation 8.7 can
be rewritten as
τ = γ ySo (8.8)
where y = vertical flow depth (ft, m)

The distribution of shear stress is not uniform on the bottom and sides of a channel.
Olsen and Florey (1952) and Lane (1955) found that the shear stress distribution in
trapezoidal channels resembles that shown in Figure 8.6. Further, they related the
maximum shear stresses on the channel bottom and sides to the shear stress for an
infinitely wide channel having the same depth and slope, γySo (see Figure 8.7). Figure
8.6 also illustrates the calculation of the maximum values for shear stress on the bot-
tom and sides of the representative channel given the data presented in Figure 8.7.
Figure 8.6
Shear stress
distribution in a
trapezoidal channel
(Chow, 1959)
The maximum permissible shear stress acting on the bottom of a channel is the criti-
cal shear stress that causes incipient motion of a soil particle. Its value depends on
sediment size and shape, sediment particle density, fluid properties, and the sus-
pended sediment concentration in the water column. Based on data from a number of
sources, the U.S. Bureau of Reclamation (1987) recommended use of Figure 8.8 for
determining the permissible shear stress on the bottom of a straight channel with non-
cohesive soil. In this figure, the permissible shear stress is referred to as “critical trac-
tive force” and is given in units of g/m2 (1 g/m2 = 0.00981 N/m2 = 0.000205 lb/ft2).
The large number of curves shown in Figure 8.8, and the corresponding wide range of
permissible shear stresses for any mean sediment size, again reinforces the fact that
many variables affect the permissible stress. The engineer must apply his or her best
judgment in selecting which of the many cases to use.
Figure 8.7
Maximum shear stress
(unit tractive force) on
the bottom and sides
of a trapezoidal or
rectangular channel as
a function of channel
geometry (Chow,
1959)
Credit: Chow, 1959.
Figure 8.8
Figure for
determining the
permissible shear
stress (critical tractive
force) on the bottom
of a straight channel
with noncohesive soil
(USBR, 1987)
The permissible shear stress on the bottom of a straight channel in a cohesive soil can
be obtained from Figure 8.9. It is called “unit tractive force” in this figure and is given
in lb/ft2 (1 lb/ft2 = 47.9 N/m2). Lane (1955) suggested that the permissible values
should be reduced by about 10 percent for slightly sinuous channels, by 25 percent for
moderately sinuous channels, and by 40 percent for very sinuous channels.
Figure 8.9
Determining the
permissible shear
stress (unit tractive
force) on the bottom
of a straight channel
with cohesive soil
(Chow, 1959)
The permissible shear stress on the sides of a channel is smaller than the permissible
stress on the channel bottom. This difference is due to the addition of the gravitational
force that acts on a soil particle and tends to roll it down the channel side slope. The
tractive force ratio expresses the relationship between permissible stresses for the
channel bottom and sides and can be used to find the permissible shear stress on the
sides of the channel if the permissible shear stress for the bottom of the channel is
known. It is defined as
τs
K= (8.9)
τb
where K = tractive force ratio

τs = permissible stress for the channel sides (e.g., lb/ft2, kg/m2)
τb = permissible stress for the channel bottom (e.g., lb/ft2, kg/m2)
The value of the tractive force ratio can be computed as
 tan 2 ξ   sin 2 ξ 
K = cos ξ  1 −  = 1 − 2  (8.10)
 tan 2 ς   sin ς 
   
where ξ = angle of inclination of the channel side slopes = arctan(1/z) (z is the H:V
ratio of the side slope)
ζ = angle of repose for the soil material
The angle of repose is the maximum angle of rest for dry sediment. This angle is a
function of soil particle size and shape; in general, it increases with size and angular-
ity. Figure 8.10 can be used to find the angle of repose for a particular soil (Chow,
1959).
In summary, design of a channel by the method of tractive force involves a determina-
tion of the permissible stresses on the channel bottom and sides, followed by the
dimensioning of the channel such that the maximum actual stresses on the bottom and
sides do not exceed the permissible values. Stresses on the sides are usually the con-
trolling factor in a design, but this should always be checked. Example 8.4, which is
adapted from Yang (1996), illustrates the procedure.
Figure 8.10
Determining the angle
of repose for soil
material
Credit: Chow, 1959.
Example 8.4 – Erodible Channel Design with the Tractive Force Method. Design
a straight trapezoidal channel using the tractive force method. The discharge is 60 m3/s, the channel
slope is 0.001, and Manning’s n = 0.015. The soil material in which the channel is to be constructed is
slightly rounded gravel with a median diameter of 40 mm.
Solution: The soil size of 40 mm is equivalent to 1.57 in. From Figure 8.10, the angle of repose is ζ =
37°. The channel side slope must be flatter than this, and so it is chosen as 1.5:1 (H:V). The side slope
angle is therefore ξ = arctan(1/1.5) = 33.7°.
The permissible tractive force on the channel bottom is obtained from Figure 8.8 as τb = 4 kg/m2. The
tractive force ratio is computed using Equation 8.10 as
1/ 2
 tan 2 33.7° 
K = cos 33.7°1 −  = 0.388
 tan 2 37° 
Therefore, the permissible tractive force on the channel sides is

τs = K × τb = 0.388 × 4 = 1.55 kg/m2.
Determination of the actual channel cross-section dimensions requires an assumption as to whether
the channel sides or bottom will control the design. One also needs to assume a channel width to
depth ratio, B/y, for use with Figure 8.7. Once the dimensions have been determined on the basis of
these assumptions, the reasonableness of the assumptions can be checked.
Assuming that the sides will control the design, and that the ratio B/y = 4, Figure 8.7 can be used to
express the maximum stress on the channel sides as
τs = 0.76γySo = 0.76(1000)(0.001)y = 0.76y
Note that γ in this case is expressed in kg/m3 rather than N/m3 because the permissible shear stress is
expressed in kg/m3. Equating this expression to the permissible stress on the channel sides, deter-
mined earlier to be 1.55 kg/m2, yields y = 2.04 m. Further, using the assumed width to depth ratio,
B = 4y = 8.2 m
To check the validity of the assumption that the sides control the design, the maximum shear stress on
the channel bottom is determined using Figure 8.7 as
τb = 0.96γySo = 0.96(1000)(2.04)(0.001) = 1.96 kg/m2

Because this value is smaller than the permissible tractive force on the bottom, the validity of the
assumption is established.
A check of the validity of the width to depth ratio assumption requires that the channel discharge be
computed on the basis of the determined dimensions and compared to the actual design discharge. For
the determined channel dimensions of z = 1.5, y = 2.04 m, and B = 8.2 m, an application of the Man-
ning equation yields Q = 62.8 m3/s, which is very close to the design discharge of 60 m3/s. If more
accuracy is needed, the B/y ratio can be changed and the calculations repeated.
Completion of the design in this example would include checking the velocity and Froude number
and adding a suitable amount of freeboard to the channel depth.
8.5 DESIGN OF GRASSED CHANNELS

Grassed channels are often preferable to lined or erodible channels in areas where
aesthetics are a concern and for cases in which flow in the channel is intermittent.
They are widely used as swales, in park settings, and for borrow and median ditches
along highway rights-of-way. Grassed channels are generally not appropriate for use
in channels that convey continuous flows.
Resistance to flow in a grassed channel depends on grass type and length, on the sea-
son of the year (because of its influence on the grass), and on the depth of flow rela-
tive to the grass height. Experiments conducted by the U.S. Soil Conservation Service
Section 8.5 Design of Grassed Channels 309
(Ree, 1949; Ree and Palmer, 1949; Stillwater Outdoor Hydraulic Laboratory, 1954)
showed that, for a specific type and condition of grass, Manning’s n could be related
to VR, the product of the mean flow velocity and the hydraulic radius. The relation-
ship between n and VR was found to be different for different types of grasses, and
even for the same type of grass under different growing conditions. The concept of
retardance of a grass was defined and introduced to account for these differences, and
different grass species and growing conditions were classified into retardance catego-
ries from A (very high retardance) to E (very low retardance).
High retardances are generally associated with long and stiff grasses that are not eas-
ily flattened by flowing water. Low retardances, on the other hand, are associated with
short grasses and those that are easily flattened out. Table 8.4 provides a listing of a
number of common grass species with their retardance classifications for various
growing conditions. Figure 8.11 shows the relationship between n and VR for each
retardance category. The curves shown in Figure 8.11 are average relationships for all
of the grasses within each category. Chow (1959) provides more detailed curves
showing the variations from one grass species to another.
Permissible velocities in grassed channels depend on the grass species, growing con-
dition, channel slope, and soil erodibility. Well-maintained grasses with high densities
can generally tolerate higher velocities than can poorly maintained grasses with poor
ground cover. Permissible velocities suggested by the Stillwater Outdoor Hydraulic
Laboratory (1954) are presented in Table 8.5.
Recommended freeboard requirements for grassed channels are different than for
lined and erodible channels. The 1-ft (30-cm) minimum recommended by ASCE
(1992) should be maintained, but ASCE has also suggested that the minimum free-
board should be 6 in. (152 mm) plus the velocity head of the flowing water. Addi-
tional freeboard should be provided in curves where superelevation of the water
surface occurs.
Design of a grassed channel begins with selection of the grass species to be used. This
selection depends in large part on the climate and soil conditions where the channel is
to be constructed. Sod-forming grasses, such as Kentucky bluegrass, centipede, and
Bermuda grass, are generally preferable to bunch-forming grasses, particularly where
discharges are high and/or channel slopes are steep. In newly constructed channels,
consideration may be given to using annual grasses to provide temporary erosion pro-
tection until the permanent grass species is well established.
The second issue to be addressed in grassed channel design is selection of a channel
side slope, z. Grassed channels typically require periodic mowing and other mainte-
nance such as fertilization, and thus must be accessible by mowers and similar
machinery. Slopes steeper than about 3 or 4:1 (horizontal:vertical) may be difficult to
maintain.
Table 8.4 Degree of retardance for various grasses (Stillwater Outdoor Hydraulic
Laboratory, 1954)
Retardance Grass Type Condition
Weeping love grass Excellent stand, tall [avg. 30 in. (76 cm)]
A: Very high
Yellow bluestem ischaemum Excellent stand, tall [avg. 36 in. (91 cm)]
Kudzu Very dense growth, uncut
Bermuda grass Good stand, tall [avg. 12 in. (30 cm)]
Native grass mixture (little Good stand, unmowed
bluestem, blue grama, and other
long and short Midwest U.S.
grasses)
B: High Weeping love grass Good stand, tall [avg. 24 in. (61 cm)]
Lespedeza sericea Good stand, not woody, tall [avg. 19 in. (48 cm)]
Alfalfa Good stand, uncut [avg. 11 in. (28 cm)]
Weeping love grass Good stand, mowed [avg. 13 in. (33 cm)]
Kudzu Dense growth, uncut
Blue grama Good stand, uncut [avg. 13 in. (33 cm)]
Crabgrass Fair stand, uncut [10 to 48 in. (25 to 122 cm)]
Bermuda grass Good stand, mowed [avg. 6 in. (15 cm)]
Common lespedeza Good stand, uncut [avg. 11 in. (28 cm)]
Grass-legume mixture – sum- Good stand, uncut [6 to 8 in. (15 to 20 cm)]
C: Moderate mer (orchard grass, redtop, Ital-
ian rye grass, and common
lespedeza)
Centipede grass Very dense cover [avg. 6 in. (15 cm)]
Kentucky bluegrass Good stand, headed [6 to 12 in. (15 to 30 cm)]
Bermuda grass Good stand, cut to 2.5-in. (6-cm) height
Common lespedeza Excellent stand, uncut [avg. 4.5 in. (11 cm)]
Buffalo grass Good stand, uncut [3 to 6 in. (8 to 15 cm)]
Grass-legume mixture – fall to Good stand, uncut [4 to 5 in. (10 to 13 cm)]
D: Low
spring (orchard grass, redtop,
Italian rye grass, and common
lespedeza)
Lespedeza sericea After cutting to 2-in. (5-cm) height; very good
stand before cutting
Bermuda grass Good stand, cut to 1.5 in. (4-cm) height
E: Very low
Bermuda grass Burned stubble
Figure 8.11
Relationship between
Manning’s n and VR
for each retardance
category
Credit: Chin, 2000
Table 8.5 Permissible velocities in grassed channels (Stillwater Outdoor Hydraulic

Laboratory, 1954)
Erosion-Resistant Soils Easily Eroded Soils
Grass Type Slope (%) ft/s m/s ft/s m/s
0–5 8 2.4 6 1.8
Bermuda grass 5–10 7 2.1 5 1.5
> 10 6 1.8 4 1.2
0–5 7 2.1 5 1.5
Buffalo grass, Kentucky blue-
grass, smooth brome, blue 5–10 6 1.8 4 1.2
grama
> 10 5 1.5 3 0.9
0–5 5 1.5 4 1.2
Grass mixture 5–10 4 1.2 3 0.9
Do not use on slopes > 10%
Lespedeza sericea, weeping 0–5 3.5 1.1 2.5 0.8
love grass, ischaemum (yellow Do not use on slopes > 5%, except for side slopes in a compound
bluestem), kudzu, alfalfa,
channel
crabgrass
Annuals: used on mild slopes 0–5 3.5 1.1 2.5 0.8
or as temporary protection
Use on slopes > 5% is not recommended
until permanent covers are
established – common lespe-
deza, Sudan grass
Procedure for Design of a Grassed Channel

Once the grass species and side slope have been selected, design of a grassed channel
is usually treated as a two-stage procedure. In the first stage, it is assumed that the
retardance category for the grass is relatively low (for example, due to winter condi-
tions), with the net result being that the velocity and erosion potential in the channel
will be relatively high. Thus, the intent of the first stage is to develop a channel cross
section such that the permissible velocity will not be exceeded. In the second stage of
the process, the retardance is assumed to be relatively high, with a corresponding low
velocity and large depth of flow. The intent of this stage is to ensure that the channel
depth is adequate to convey the design discharge in conditions where, for example,
the grass has not been mowed in some time. Following is a description of the steps
involved in each of the two design stages.
Stage 1.
1. Determine the design discharge and channel slope, and select the grass species
and channel side slope to be used.
2. Assume a value for n and determine the corresponding value of VR from Figure
8.11. Use the curve indicative of a relatively low degree of retardance for the
selected grass species.
3. Determine the permissible channel velocity from Table 8.5, and use it with the
value of VR obtained in step 2 to determine the required hydraulic radius R.
4. Using the Manning equation, compute the product VR as
1.49 2 / 3 1/ 2
VR = R So (8.11)
n
Compare this computed value to the value read from Figure 8.11 in step 2.
5. If the values of VR obtained in steps 2 and 4 do not agree, repeat steps 2 through
4 with different assumed n values until agreement is obtained.
6. Compute the required cross-sectional area as A = Q/V.
7. Use the values of A and R determined in steps 3 and 6 to determine the required
cross-sectional dimensions.
Stage 2.
1. Using the cross-sectional dimensions determined in Stage 1 and an assumed
depth of flow y, compute the cross-sectional area A and hydraulic radius R.
2. Compute the values V = Q/A and VR.
3. Using the curve in Figure 8.11 for a relatively high retardance for the chosen
grass, determine the value of Manning’s n.
4. Compute the flow velocity using Manning’s equation and compare it to the value
determined in step 2.
5. If the velocities in steps 2 and 4 do not agree, repeat steps 1 through 4 with differ-
ent assumed flow depths until agreement is obtained.
6. Add freeboard to the computed depth to obtain the total required depth of the
channel.
Example 8.5, from Chow (1959), illustrates the procedure.
Example 8.5 – Grassed Channel Design. Design a grassed channel to convey a discharge
of 50 cfs on a 4-percent slope. The soil is erosion-resistant, and the grass selected is a grass-legume
mixture. Assume that the channel will be trapezoidal with side slopes of 3:1 (H:V).
Solution: Referring to Table 8.4, the grass-legume mixture has a low (D) retardance in the fall, win-
ter, and spring, and a moderate (C) retardance during summer months. Thus, the D curve will be used
in the first stage of the design, and the C curve will be used in the second stage. For the first stage, the
permissible velocity is found from Table 8.5 to be 5 ft/s.
The trial-and-error steps in the first stage of the design are summarized in the following table.
Trial n VR R VR = (1.49/n)R5/3S1/2
1 0.04 3.0 0.60 3.2
2 0.038 3.5 0.70 4.3
3 0.042 2.5 0.50 2.2
4 0.041 2.75 0.55 2.7
The required cross-sectional area and hydraulic radius are therefore A = 50/5 = 10 ft2 and R = 0.55 ft.
Thus, using Table 8.1, one can write the equations
A = 10 ft2 = By + 3y2
R = A/P = 0.55 ft = (By + 3y2)/[(B + 2y(1+32)1/2]

Solving these equations simultaneously yields B = 14 ft and y = 0.62 ft. Thus, the channel cross sec-
tion will have a bottom width of 14 ft, side slopes of 3:1, and a low-retardance depth of 0.62 ft.
The intent of Stage 2 is to determine the depth of flow that will occur under high-retardance (summer)
conditions. The trial-and-error calculations are shown in the following table:
Trial y A R V VR n V = (1.49/n)R2/3S1/2
1 0.65 10.4 0.57 4.81 2.74 0.049 4.18
2 0.70 11.3 0.61 4.44 2.70 0.049 4.37
The correct depth is therefore y = 0.70 ft.

The minimum amount of freeboard provided should be the larger of either 1 ft or 6 in. plus the veloc-
ity head. In Stage 2, the velocity was found to be 4.4 ft/s, so the velocity head plus 6 in. is 0.80 ft.
Therefore, the required freeboard is determined to be 1 ft, and the total channel depth should be about
1.7 ft.
Open-Channel Stabilization Options

Many channel lining options are available for consideration and sediment control regulations, primarily for site construc-
in the design of open channels. The choice of lining primarily tion. The FHWA also provides guidance on designing chan-
depends on regulations and site-specific issues. Applicable nels with flexible linings (Chen and Cotton, 2000).
regulations may include highway design standards, environ- Channel linings range from simple grass vegetation to heavily
mental regulations related to concerns such as erosion and armored, concrete, engineered structures. The table below
sediment control and stream stabilization. The features of a shows approximate ranges for shear stress for different chan-
site that must be considered include slopes, other drainage nel lining types. The second table provides descriptions of
systems, soil erodibility, aesthetics, and existing riparian veg- channel lining materials and suggested references for further
etation. information.
States and localities typically have design handbooks or Design standards for all of these lining types are very product
guidelines to make the process of selecting the channel lining specific. Although shear stress is one consideration, a factor
more straightforward. For example, the Virginia Erosion and such as permissible velocity, side slope, soil erodibility, or
Sediment Control Handbook (Virginia Dept. of Conservation lining decomposition time may be of greater concern for
and Recreation, 1992) details applicable regulations, channel some lining materials.
design criteria, and lining types that satisfy Virginia erosion
Maximum Allowable Shear Stress Ranges (Burchett, 1998)

Lining Material Shear Stress (lb/ft2) Shear Stress (Pa)
Bare soil 0.1–2 0.15–3.0
Erosion control blanket over bare soil 1–5 1.5–7.4
Turf reinforcement mat over bare soil 3–8 4.5–11.9
Unreinforced Vegetation 0.5–4 0.7–11.9
Turf reinforcement mat, reinforced vegetation 4–10 6.0–14.9
Rock rip-rap 8–20 11.9–29.8
Gabions Design Dependent
Lining Type Lining Description

Erosion control blankets: Excelsior mat Excelsior (dried, shredded wood) covered with fine paper net covering
Degradable blankets used to pro- Jute mesh Mat lining woven of jute yarn that varies in diameter
vide channel protection until per-
manent vegetation is established Combination Photo-degradable plastic netting that covers and is entwined with a natu-
mats or blankets ral organic or man-made mulching material such as straw or coconut
Permanent turf reinforcement Fiberglass mat Fine, loosely woven glass fiber mat similar to furnace air-filter material;
mats: Mats designed to stay in different fiber densities allow or inhibit underlying particular plant
place that provide a matrix to help growth
support roots and vegetation; per- Fiberglass A lightly bound ribbon of continuous glass fibers, applied using a special
manent vegetation will grow roving nozzle attached to an air compressor
through the lining openings
Permanent armoring: Erosion- Rip-rap Permanent, erosion-resistant ground cover of large, loose, angular stone;
resistant structures composed of graded rip rap contains stones of various sizes, while uniform riprap con-
hard materials tains stones that are fairly close in size
Gabions Rectangular, rock-filled wire baskets; the amount of channel armoring
needed affects rock and wire specifications
Interlocking Interlocking concrete blocks that may have extra space for soil for the
concrete blocks establishment of permanent vegetation
Section 8.6 Riprap-Lined Channels 315
8.6 RIPRAP-LINED CHANNELS

A special case of lined channels has the side slopes (and possibly the bottom also) of
the channel lined with riprap. Construction of channels in this manner is a common
practice, and may be viewed as a compromise between a concrete-lined channel and
an erodible or grassed channel. In some cases, a trickle channel may be lined with
riprap, while the main body of the conveyance may be vegetated or unlined.
While a riprap-lined channel may be attractive in terms of its capital cost, the designer
and review agency should bear in mind that such channels are more difficult to main-
tain and may require more frequent inspections than other channel types. Design of
riprap-lined channels requires detailed attention to required rock dimensions, not only
to withstand the erosive forces of moving water, but also the “erosive” effect of neigh-
borhood children who revel in the splashes caused by throwing the rocks into the
stream. The availability or cost of obtaining suitably sized rock may make a riprap-
lined channel an unsuitable alternative. It is foolhardy to use rock that is too small for
a particular application simply because larger rock is not available or too expensive,
as the riprap lining will fail in short order.
Design of riprap-lined channels is a rather specialized topic and is beyond the scope
of this text. The reader desiring more information on this topic is referred to Volume 2
of Denver's Urban Storm Drainage Criteria Manual (UDFCD, 2000) and to other ref-
erences cited therein.
8.7 MODIFYING EXISTING CHANNELS

Many open channel design projects involve making modifications to existing natural
channels, typically to increase channel capacity. These modifications may be neces-
sary because of increased runoff from new impervious areas, or because of flooding
of existing developments. Other options for dealing with these types of issues, such as
stormwater detention and floodproofing structures in the floodplain, may be viable
and should be considered prior to undertaking a channelization project.
Channelization measures that increase channel capacity are the focus of this section.
The basic ways in which capacity can be increased are:
• Increasing the cross-sectional area of the channel be making it deeper or
wider, or by installing levees or floodwalls
• Reducing the roughness of the channel by removing debris and woody vege-
tation, straightening the channel, or installing a smooth lining such as
concrete
• Increasing the slope of the channel, which is often accomplished in concert
with channel straightening
• Removing or enlarging downstream controls that cause flow to back up in
the channel
In addition to hydraulic capacity, an engineer designing channel improvements should
consider factors such as habitat development and preservation, aesthetics, recreation,
and sediment transport. Channel stability is another important issue, except in special
cases such as a stream flowing over bedrock or a channel lined with paving material
or large boulders.
Traditional channelization projects have often led to undesirable responses, such as

increased erosion and/or sedimentation (McCarley et al., 1990; Brookes and Shields,
1996). Designing an open channel involves more than simply drawing a straight line
between two points of known elevation and applying the Manning equation using the
design flow rate. Engineers designing projects dealing with alluvial conditions need
to understand fluvial geomorphology—the science that assesses the shape and form of
a watercourse and the contributing physical processes—or have someone on the
design team with expertise in channel erosion and sedimentation (Knighton, 1998;
Nunnally, 1978).
A current trend is to move away from designing a simple channel shape, such as a
trapezoid, to carry peak flow. Instead, there is an increasing use of more complex
geometries that perform better over a wide range of flows. Various governmental
guidelines reflect this trend (Federal Interagency Stream Corridor Restoration Work-
ing Group, 2001; U.S. Army Corps of Engineers, 1989). Stream restoration has also
received considerable attention from engineers in recent years, including ASCE con-
ferences such as the Joint Conference on Water Resources Engineering and Water
Resources Planning and Management in 2000 (ASCE, 2000) and the Wetlands Engi-
neering and River Restoration conference in 2001 (Hayes, 2001).
An improved cross section often consists of a main channel to handle smaller flows
and overbank areas that convey flows from larger storm events. This approach is
being used in the construction of new channels, improvements to existing channels,
and restoration of channels previously impacted by development. Figure 8.12 con-
trasts a simple channel design that considers only peak flow with a more complex
cross section that includes a floodway/overbank area to be used for recreational pur-
poses or as habitat. In some cases, this overbank area may be large enough for hiking
and bike paths or even ball fields, as with Cherry Creek in Denver, Colorado.
Figure 8.12
A simple channel
design versus a
complex channel
design with a smaller
main channel and
floodway
For natural channels, it is important to evaluate the channel at the peak design flow as
well as at other important flow rates. An important flow is the dominant discharge, or
channel-forming flow, that corresponds to a one- or two-year return interval flow for
humid areas (Richards, 1982) and slightly longer periods in arid areas (U.S. Army
Corps of Engineers, 1994). The channel-forming flow is not necessarily the same as
the bankfull discharge, which is the discharge at which the channel begins overflow-
ing onto the floodplain (Copeland et al., 2001). If a channel remains stable for long
periods of the time, the bankfull discharge should be close to the dominant discharge
because the channel has adjusted to hydraulic and sediment loads (Shields et al.,
2003). Computer programs can assist the engineer in creating a rating curve for evalu-
ation of a channel over a range of flows.
Section 8.7 Modifying Existing Channels 317
Modeling Focus: Channel Design

Computer programs can assist in the design of open channels, The assumed n for grass of 0.03 was used for stations 0+00.0
but there is great variation in the level of analysis among to 0+12.0 and 0+18.0 to 0+30.0. The n value for concrete,
available programs. In the design of smaller prismatic chan- 0.013, is used between stations 0+12.0 and 0+18.0. The engi-
nels, it is frequently reasonable to assume that the channel neer can then solve for the flow characteristics in the channel
operates under normal flow conditions. FlowMaster is an for a flow of 125 ft3/s.
example of a program that can be used in the design of open Note: To apply Manning’s equation to a channel such as this
channels where a normal depth assumption is appropriate. one, which has a composite roughness, a method for comput-
This example outlines a design for a proposed channel to con- ing a weighted n value must be employed. In this example, the
vey the peak discharge of 125 cfs for the 25-year storm. It is method proposed by Pavlovskii, by Muhlhofer, and by Ein-
assumed that the flow in the channel will be close to normal stein and Banks (Chow, 1959) is used. Software programs
depth. The lower portion of the channel will be designed to may vary in the method or methods available for weighting
handle a low flow of 31 cfs. This portion of the channel will Manning’s n.
be lined with concrete (Manning’s n = 0.013) and will be trap- The program computes the following:
ezoidal with a 3-ft-wide bottom and 1:1 side slopes. The
Flow depth 3.54 ft Wetted perim. 20.16 ft
banks of the channel above the elevation of the concrete will
Flow area 31.5 ft2 Velocity 3.96 ft/s
be Bermuda grass on a 3:1 slope. The longitudinal slope of
the channel is 0.25 percent. A freeboard of 1 ft should be pro- Checking the initial assumption of n = 0.03 for the grass, VR
vided above the 25-year storm flow depth. is computed as (3.96)(31.5)/20.16 = 6.19 ft2/s. From Figure
8.11, n = 0.026. With this n for the grass portions of the chan-
The engineer can start by designing the trapezoidal, concrete,
nel, the following output is obtained:
low-flow portion of the channel. The following data can be
entered into a program such as FlowMaster to solve for depth: Flow depth 3.36 ft Wetted perim. 19.03 ft
Flow area 28.3 ft2 Velocity 4.41 ft/s
Manning’s n 0.013 Bottom width 3 ft
Long. Slope 0.0025 Discharge 31 cfs VR is now found to be 6.6 ft2/s, and our assumed n = 0.026 is
Side slopes 1:1 acceptable. Checking the velocity with the permissible veloci-
ties for Bermuda grass on slopes less than 5 percent, the
The normal depth is computed to be 1.37 ft. The flow is sub-
velocity of 4.41 ft/s is acceptable, even for easily eroded soils.
critical, with an average velocity of 5.2 ft/s. Based on this
computed depth, the engineer decides the make the concrete Next, the design must be checked assuming a high retardance.
portion of the channel 1.5 ft deep. The highest retardance for Bermuda grass is category B. The
initial value for n is selected as 0.07 after consulting Figure
Next, the engineer must design the upper (grass) portion of 8.11, and this value corresponds to VR = 3 ft2/s. Updating the
the channel. A value for n corresponding to a low retardance
input data for this n, the results are as follows:
for Bermuda grass must be selected. The lowest retardance
category for Bermuda grass is category E. From Figure 8.11, Flow depth 4.89 ft Wetted perim. 28.68 ft
an n value of 0.03 is assumed for the initial calculation. The Flow area 61.6 ft2 Velocity 2.03 ft/s
corresponding value of VR is 2 ft2/s. Thus, VR is computed to be 4.36. Referring again to Figure
The engineer assumes a conservative maximum depth and 8.11, the new value for n is read as 0.065. Updating this in the
enters the geometry of the full cross-section into the computer input data and solving, the results are as follows:
program being used. For this example, the following cross- Flow depth 4.75 ft Wetted perim. 27.82 ft
section points were used: Flow area 58.0 ft2 Velocity 2.15 ft/s
Station (ft) Elevation (ft) Station (ft) Elevation (ft) In this iteration, VR is computed as 4.48 ft2/s, and our
0+00.0 5.5 0+16.5 0.0 assumed value of n is acceptable. For the final design, the
0+12.0 1.5 0+18.0 1.5 engineer adds an additional 1 ft to the flow depth for free-
0+13.5 0.0 0+30.0 5.5 board, making the total depth of the channel 5.75 ft. A plot of
the cross-section for the final channel design is shown below.
In natural channels, the roughness, particularly in the overbanks, may change over
time due to the growth of vegetation, the addition of obstructions such as buildings
and fences, and maintenance practices or lack of maintenance. The extent of the
effects of such changes can be predicted by performing additional analyses to investi-
gate the sensitivity of the channel to changes in the roughness factor.
The channel geometry of bends differs from straight sections of channels. The chan-
nel is deeper at the outside of the bend where the maximum erosion occurs, and the
channel is shallower with slower velocity on the inside of the bend where a gravel or
sand bar usually forms. Special protection measures may be necessary on the outside
bank to prevent erosion.
Projects that involve straightening of streams have the effect of increasing velocity
and therefore erosion. Armored riffle sections or drop structures can enable the flow
to remain subcritical through most of the reach, and may be preferable to lining the
entire length of the channel.
8.8 CHAPTER SUMMARY

Many factors must be considered in the design of open channels. The most obvious is
the hydraulic capacity of the channel, but other factors such as flow velocity, Froude
number, side slopes, longitudinal slope, freeboard, lining material, construction cost,
maintenance cost, aesthetics, right-of-way requirements, and dry-weather flow per-
formance should be considered as well.
The best hydraulic section is a channel section for which the wetted perimeter is min-
imized for a given cross-sectional area. However, the steepness of this section’s side
slopes limit its practical application.
Open channels may be classified as lined, erodible, or grassed. Some channels have
characteristics of more than one of these types. The design of erosion-resistant lined
channels is the most straightforward, and is primarily concerned with conveyance
capacity.
When designing erodible channels, one must pay particular attention to the forces that
could cause the channel to scour. Two methods for designing channels in order to pre-
vent scour are described in the text. The first method uses a maximum permissible
velocity, and the second uses a maximum permissible shear stress.
Grassed channels are often used in instances where there is intermittent flow. The
flow retardance properties of various grass types in different conditions are consid-
ered, as well as the maximum allowable velocity and the suitability of different grass
species for particular climates and soil conditions.
Often, channel design involves planning modifications to an existing channel rather
than construction of an entirely new conveyance. In addition to ensuring that the
channel will have adequate capacity for the design flow, the engineer should evaluate
flow levels that will occur more frequently and should consider other factors such as
the effect on aquatic habitat, channel stability, and the potential for erosion.
References 319
REFERENCES
Aisenbrey, A. J., Jr., R. B. Hayes, H. J. Warren, D. L. Winsett and R. B. Young. 1978. Design of Small
Canal Structures. U.S. Department of Interior, Bureau of Reclamation.
American Society of Civil Engineers (ASCE). 1977. Sedimentation Engineering. ASCE Manuals and
Reports on Engineering Practice, No. 54. New York: ASCE.
agement Systems. New York: ASCE.
American Society of Civil Engineers (ASCE). 2000. Joint Conference on Water Resources Engineering
and Water Resources Planning and Management. Minneapolis, Minn.: ASCE.
Armortec. 2003. Product literature for Armorflex concrete mats. (www.armortec.com).
Brookes, A. and F. D. Shields. 1996. River Channel Restoration: Guiding Principles for Sustainable
Projects. New York: John Wiley & Sons.
Burchett, C. 1998. “Open Channel Linings: Choosing the Right Erosion Control Proudcts.” Burns &
McDonnell Tech Briefs 1998, no. 3. Kansas City, Missouri: Burns & McDonnell (www.burnsmcd.com).
Chen, Y. H. and B. A. Cotton. 2000. Design of Roadside Channels with Flexible Linings, Hydraulic Engi-
neering Circular no. 15 (HEC-15). McLean, Virginia: Federal Highway Administration (FHWA).
Chow, V. T. 1959. Open-Channel Hydraulics. New York: McGraw-Hill.
Copeland, R. R., D. N. McComas, C. R. Thorne, P. J. Soar, M. M. Jonas, and J. B. Fripp. 2001. Hydraulic
Design of Stream Restoration Projects. Technical Report ERDC/CHL TR-01-28. Vicksburg, Missis-
sippi: U.S. Army Corps of Engineers, Engineer Research and Development Center.
Federal Interagency Stream Corridor Restoration Working Group. 2001. Stream Corridor Restoration:
Principles, Processes and Practices. GPO 0120-A. (http://www.usda.gov/stream_restoration/).
Fortier, S., and F. C. Scobey. 1926. “Permissible Canal Velocities.” Transactions, American Society of Civil
Engineers 89: 940–956.
Hayes, D. F., ed. 2001. “Wetlands Engineering and River Restoration.” ASCE Conference Proceeding.
Reno, Nevada: ASCE.
Hjulstrom, F. 1935. “The Morphological Activity of Rivers as Illustrated by River Fyris,” Bulletin of the
Geological Institute 25, chap. 3.
Knighton, D. 1998. Fluvial Forms and Processes. Arnold.
Lane, E. W. 1955. “Design of Stable Channels.” Transactions, American Society of Civil Engineers 120:
1234–1260.
McCarley, R. W., J. J. Ingram, B. J. Brown, and A. J. Reese. 1990. Flood Control Channel National Inven-
tory. MP HL-90-10. Vicksburg, Mississippi: US Army Engineer Waterways Experiment Station.
McWhorter, J. C., T. G. Carpenter, and R.N. Clark. 1968. “Erosion control criteria for drainage channels.”
Study conducted for the Mississippi State Highway Department and the Federal Highway Administra-
tion. State College, Mississippi: Department of Agricultural Engineering, Mississippi State University.
Mehrota, S. C. 1983. “Permissible Velocity Correction Factor.” Journal of the Hydraulics Division, ASCE
109, no. HY2: 305–308.
Nunnally, N. R. 1978. “Stream Renovation: An Alternative to Channelization.” Environmental Manage-
ment 2(5): 403.
Olsen, O. J. and Q. L. Florey. 1952. Sedimentation Studies in Open Channels: Boundary Shear and Velocity
Distribution by Membrane Analogy, Analytical, and Finite-Difference Methods. U.S. Bureau of Recla-
mation, Laboratory Report Sp-34, August.
Petratech. 2003. Product literature for Petraflex interlocking blocks. (www.petraflex.com).
Ree, W. O. 1949. “Hydraulic Characteristics of Vegetation for Vegetated Waterways.” Agricultural Engi-
neering 30, no. 4: 184–187 and 189.
Ree, W. O. and V. J. Palmer. 1949. Flow of Water in Channels Protected by Vegetative Lining. U.S. Soil
Conservation Service, Technical Bulletin No. 967.
Richards, K. 1982. Rivers: Form and Processes in Alluvial Channels. New York: Metheun and Co.
Shields, F. D., R. R. Copeland, P. C. Klingeman, M. W. Doyle, and A. Simon. 2003. “Design for Stream
Restoration.” Accepted for ASCE Journal of Hydraulic Engineering.
Stillwater Outdoor Hydraulic Laboratory. 1954. Handbook of Channel Design for Soil and Water Conser-
vation. U.S. Soil Conservation Service, SCS-TP-61.
Urban Drainage and Flood Control District (UDFCD). 2001. Urban Storm Drainage Criteria Manual, Vol.
2, Major Drainage. Denver, Colorado: UDFCD.
U.S. Army Corps of Engineers. 1989. Environmental Engineering for Local Flood Control Channels. Engi-
neering Manual 1110-21601. Washington, D.C.: USACE.
U.S. Army Corps of Engineers. 1994. Hydraulic Design of Flood Control Channels. Engineering Manual
1110-21601. Washington, D.C.: USACE.
U.S. Bureau of Reclamation (USBR). 1987. Design of Small Dams. Lakewood, Colorado: USBR.
Virginia Department of Conservation and Recreation (DCR). 1992. Virginia Erosion and Sediment Control
Handbook, 3rd ed. Richmond, Virginia: DCR.
Yang, C. T. 1976. “Minimum Unit Stream Power and Fluvial Hydraulics.” Journal of Hydraulic Engineer-
ing. 105 (HY7): 769.
Yang, C. T. 1996. Sediment Transport: Theory and Practice. New York: McGraw-Hill.
PROBLEMS
8.1 Determine the discharge in a trapezoidal, concrete channel with a bottom width of 2.4 m and side
slopes of 1 to 1. Uniform depth is 1.8 m, the bottom slope is 0.009, and Manning’s n = 0.013.
8.2 A concrete-lined channel with a triangular cross-section conveys a flow of 250 cfs. The slope is
0.005. The side slopes are 2H:1V and Manning’s n is 0.012. Find the uniform flow depth.
Problems 321
8.3 What is the uniform depth for a flow of 4.25 m3/s in a 1.8-m wide, rectangular, wood-planed channel
(n = 0.012) with a bottom slope of 0.002?
8.4 Use the maximum permissible velocity method to design an erodible channel to transport 250 cfs.
The channel bottom and the side slopes are to be constructed from sandy loam. Use a trapezoidal
cross section and a slope of 0.00033, and design for water with colloidal silts. Sketch your design
cross section.
8.5 Use the tractive force method to design a channel lined with 2-in., slightly angular gravel. The
design discharge is 500 cfs and the channel slope will be 0.002. Use a Manning's n = 0.020. Sketch
your design cross section.
8.6 A trapezoidal channel is lined with a good stand of uncut buffalo grass that is 75 to 150 mm high.
The channel slope is 0.01, the bottom width is 0.8 m, and the side slopes are 3H:1V. Find the channel
capacity and the flow velocity for a depth of 0.5 m.
A triple-barrel box
culvert with small
wingwalls in a
riprap channel
conveys flow under
a roadway.
C H A P T E R
9
Culvert Design
A culvert is a hydraulically short underground water conveyance conduit that allows

water to flow through an obstruction such as a highway or railway embankment (see
the preceding page). The orientation of a culvert’s longitudinal axis is usually at a
right angle, or nearly so, to the centerline of the obstruction under which it passes. In
some instances, structures constructed as culverts may be considered bridges if their
spans exceed a certain dimension. Transportation departments in the United States
usually consider any culvert with a total span (counting all culvert barrels, if there are
more than one) greater than 20 ft (6 m) to be a bridge. It may sometimes be appropri-
ate to hydraulically model a structure classified as a “bridge” by using culvert analysis
methods; however, the techniques used to model bridge hydraulics generally differ
from those used with culverts and are beyond the scope of this book. [The reader is
referred to Floodplain Modeling Using HEC-RAS (Dyhouse et al., 2003) for informa-
tion on bridge modeling.]
Typically, a culvert is designed to convey the peak flow from a specified design storm,
and its performance is checked for a storm having a longer recurrence interval. The
purpose of this check is to assess the magnitude of roadway or embankment overtop-
ping expected to occur for the larger event. A local or regional review agency gener-
ally specifies the recurrence interval for a design. Chapters 4 and 5 present hydrologic
procedures for calculating discharges for both the design event and check event, and
Section 9.1 of this chapter presents a discussion of design flow estimation as it relates
specifically to culverts.
As stated in Section 2.2 (page 22), culverts are available in various shapes, sizes, and
materials. Materials and shapes are described in detail in Section 9.2. Section 9.3 pro-
vides a discussion of various types of culvert end treatments and their hydraulic per-
formance characteristics and describes special inlet configurations that can be used
when allowable headwater depths are limited.
The horizontal and vertical alignments of a culvert should closely follow those of the
channel or stream. This typical alignment is sometimes modified for reasons relating
to hydraulic performance, cost, or other constraints. Possible modifications include
the use of a doglegged alignment or the use of a drop at the culvert entrance to reduce
324 Culvert Design Chapter 9
the barrel slope. Issues associated with such modifications are presented in Section
9.4.
Section 9.5 presents the detailed hydraulic performance characteristics of a culvert.
The discussion includes the controls that affect hydraulic performance, the potential
effects of tailwater elevation, and detailed computational procedures for flow profile
estimation. Also discussed are methods for estimating outlet velocity, which is neces-
sary for design of energy dissipation measures.
During storm events having recurrence intervals longer than that of the culvert design
storm, or in cases where the culvert becomes clogged with debris, overtopping of the
roadway or embankment is expected to occur. Prediction of the weir flow characteris-
tics of roadway overtopping is presented in Section 9.6. Section 9.7 describes the
development of culvert performance curves, which provide a convenient way of
assessing the headwater elevation at a culvert for any discharge of interest.
Because the flow exiting the downstream end of a culvert often has a velocity higher
than what can be tolerated in the downstream channel, or because of the acceleration
of flow at the entrance, erosion protection measures are generally needed at the outlet
and/or inlet of a culvert. This protection may consist of a concrete apron, protective
riprap, or a geosynthetic material for reinforcement of soil and vegetation. Procedures
for riprap erosion protection design at storm sewer outlets are presented in Section
11.7, and these procedures can be applied to culverts as well. Section 9.8 discusses
outlet protection considerations specific to culvert design.
Because culverts often convey water in natural streams that may support fish and
other aquatic wildlife, minimizing the disruption of migration pathways may be a
design consideration. A discussion of designing for some of these environmental fac-
tors is provided in Section 9.9.
Section 9.1 Culvert Design Flows 325
9.1 CULVERT DESIGN FLOWS

In most cases, culverts are designed to convey a discharge corresponding to a chosen
recurrence interval. In such instances, one must estimate only the peak runoff rate.
However, for cases in which a culvert causes significant ponding upstream of its
entrance, the reduction of the peak discharge caused by detention effects may need to
be considered. In such cases, the engineer should develop and route a complete runoff
hydrograph. Chapter 5 outlines procedures for peak flow and hydrograph estimation.
Section 12.7 discusses flow routing procedures.
A government review agency usually specifies the recurrence interval for which a
design discharge or hydrograph should be determined. This criterion and others are
provided in the interest of promoting consistency from one design to another. In high-
way design applications, culvert design flows are often estimated for a 50-year event
based on FHWA design frequency recommendations (Brown, Stein, and Warner,
2001). Lower or higher recurrence intervals may be justified in other applications.
In addition to developing the design discharge, the engineer should estimate a check
flow. The recurrence interval associated with the check flow is longer (and thus the
discharge greater) than that of the design flow, because the check flow is used to
assess performance when embankment overtopping is expected to occur. Again, local
or state review agencies usually specify the check flow recurrence interval that should
be used; in highway design applications, it is often the 100-year event.
For cases in which a performance curve is needed for a culvert design, a range of
discharges corresponding to various recurrence intervals should be estimated. The
low end of the discharge range should be lower than the design discharge. The high
end of the range should be equal to or greater than the check flow discharge. In
performance curve development (Section 9.7), the headwater elevation is estimated
and graphed for each of these estimated discharges.
An alternative to basing design discharge on recurrence interval criteria is to base it
on an economic analysis. Figure 9.1 shows the costs associated with a circular culvert
as a function of its diameter. All costs are expressed as a present worth using appro-
priate discounting formulas. The capitalized cost curve represents the costs of design,
construction, and maintenance of the culvert over its lifetime. Culverts with large
diameters are clearly more costly than culverts with small diameters, and thus the cap-
italized cost increases with the diameter. The damage curve represents the present
value of flood damages anticipated to occur when the culvert’s capacity is exceeded.
Because the capacity of a smaller culvert will be exceeded more frequently, the result-
ing damages will be higher than those expected for a larger culvert. Thus, the damage
curve has a negative slope. The total cost associated with a culvert is the sum of the
capitalized cost and the damages. It is shown as a third curve in Figure 9.1. The diam-
eter corresponding to the minimum cost on the total cost curve represents the least-
cost culvert design.
Figure 9.1
Graph of culvert costs
vs. diameter
An economic analysis such as that just described is usually not accomplished in

practice because of difficulties associated with estimating the damage curve. The
analysis for an area can be further complicated by the presence of multiple culverts
providing differing levels of protection. Nevertheless, performing an economic analy-
sis is an enlightening exercise, as the true total cost of a culvert becomes evident.
Such an analysis also shows that most culvert designs are likely less than optimal in
economic terms.
9.2 CULVERT TYPES AND MATERIALS

Culverts, like storm sewers, come in a variety of shapes, sizes, and materials. Cross-
sectional shapes associated with pipe culverts are circular, elliptical (horizontal or
vertical), and arched. Culverts may also be constructed from precast or cast-in-place
rectangular box sections, or from field-assembled structural-plate-arch materials.
These latter shapes are common for large-capacity culverts. Materials used for culvert
design are most commonly reinforced precast concrete, cast-in-place concrete, and
corrugated steel. Section 11.1 discusses available materials and sizes. Information
that is more detailed can be obtained from pipe manufacturer associations such as the
American Concrete Pipe Association (ACPA) and the American Iron and Steel Insti-
tute (AISI). Other materials used for culverts include corrugated aluminum, polyeth-
ylene, and polyvinyl chloride (PVC). Aluminum and plastic materials offer
advantages over steel in terms of corrosion, but are more limited in size availability.
They also do not offer the structural strength of concrete or steel.
In many cases, to address a situation where there is limited cover over the top of a cul-
vert, or to approximate the full width of a stream channel, multiple culvert barrels
may be placed side by side. Such a configuration can minimize flow contractions at
the culvert entrance and make the installation more hydraulically efficient. For
instances in which multiple culvert barrels are used, it is usually assumed during
Section 9.3 Culvert End Treatments 327
design that the total discharge is divided equally among the individual barrels. Excep-
tions to this practice are cases in which the individual barrels have different sizes, or
where they have different invert elevations or slopes. In these latter cases, a trial-and-
error procedure is required to determine the discharge through each barrel. Computer-
based modeling tools are particularly useful in such instances.
In the absence of review agency criteria, which would otherwise govern, the mini-
mum recommended diameter for a culvert is 12 in. (300 mm). A higher minimum
may be desirable for relatively long culverts, which are more difficult to maintain than
short ones. Culverts are susceptible to clogging, especially in areas where sediment
transport is high and where debris such as fallen leaves and branches is significant.
When non-circular materials are used, it is recommended that the shortest cross-
sectional dimension be at least 12 in. (300 mm).
9.3 CULVERT END TREATMENTS

The entrance and exit of a culvert are locations where flow is accelerated or deceler-
ated, often causing erosion of the channel bed and embankment side slopes. Erosion
protection measures are generally required in these locations and usually take the
form of a riprap blanket and/or special culvert end treatments. The section focuses on
the various types of end treatments that are used; Section 9.8 and Section 11.7 discuss
riprap protective measures.
Culvert end treatments do more than provide protection against erosion; they help to
stabilize embankment slopes around culvert entrances and exits and blend the culvert
ends into the embankment slope. The type of culvert end treatment used in a design
also has hydraulic ramifications, as some types may actually enhance the hydraulic
conveyance capacity of the structure. For instance, all other factors being equal (cul-
vert dimensions, material, discharge, tailwater elevation, etc.), some culvert entrance
types will result in lower upstream water surface elevations (that is, lower headwater
depths); these entrance types are considered to be more hydraulically efficient.
Figure 9.2 illustrates the most commonly used types of culvert end treatments. Each
of these types can be used at either end (entrance or exit) of a culvert. A plain or
projecting culvert [Figure 9.2(a)] has no special treatment. The culvert barrel simply
projects from the embankment slope. This type of end treatment offers no protection
against erosion of the embankment slope; its primary advantage is that it is
inexpensive. Hydraulically, a projecting end is neither efficient nor inefficient if the
wall thickness of the culvert pipe is significant (as in the case of concrete pipe); it may
be viewed as a “middle-of-the-road” treatment. If corrugated metal pipe (which has a
small wall thickness) is used, a projecting end is hydraulically inefficient.
A mitered end [Figure 9.2(b)] provides a visual improvement over the projecting end
and usually incorporates an erosion protection blanket or apron around the mitered
end. However, from a hydraulic capacity point of view, a mitered-end configuration is
inefficient. The culvert entrance loss coefficient associated with a mitered end is often
higher than that of other types of end treatments (see Table 6.5 on page 203).
Figure 9.2
Typical culvert end
treatments
Concrete and corrugated metal pipe manufacturers can supply, in addition to pipe
materials, specially formed end sections such as the precast end section shown in Fig-
ure 9.2(c). These end sections are usually tapered to gradually accelerate or decelerate
the flow, and are shaped to facilitate the blending of the culvert end into the embank-
ment slope. Entrance loss coefficients shown in Table 6.5 indicate that special end
sections have approximately the same efficiency as a square-edged headwall or
square-edged projecting end.
Headwalls and wingwalls [see Figure 9.2(d)] offer perhaps the best slope stability and
erosion protection, but are generally the most expensive of treatment types of those
shown in Figure 9.2. A headwall is a concrete wall at the inlet or outlet of a culvert
through which the pipe projects. A headwall located at the outlet end of a culvert is
sometimes referred to as an endwall. Wingwalls may be attached to either side of a
headwall to provide additional stability or improve hydraulic performance. Headwalls
and wingwalls, which may be precast or cast-in-place, are essentially retaining walls
and should be designed as such by considering factors such as earth pressure, loading,
and soil properties. The engineer should consider requiring a pipe joint near the head-
wall to allow for possible differential settlement or movement between the headwall
Section 9.3 Culvert End Treatments 329
and pipe. Headwalls and wingwalls should be provided with weep holes to allow
seepage water trapped behind them to escape.
Any type of end treatment illustrated in Figure 9.2 can be enhanced hydraulically by
rounding or mitering the edges of the entrance to the culvert barrel. Placing the socket
or groove end of a pipe joint at the entrance has essentially the same effect as round-
ing or mitering. The hydraulic enhancement derives from a reduction in flow contrac-
tion effects as the water enters the culvert. A square-edged entrance causes a large
contraction of the flow streamlines, resulting in an effective reduction of cross-
sectional flow area. Rounding or mitering the entrance edges as shown in Figure 9.3
reduces the contraction. Table 6.5 shows that loss coefficients for rounded, mitered, or
socket-end edges are always less than those for square-edged entrances.
Figure 9.3
Square-edge and
curved-edge culvert
entrances (Norman,
Houghtalen, and
Johnston, 2001)
At locations where available headwater depths are too limited to allow adequate cul-
vert performance, special culvert entrance configurations may be used to increase the
hydraulic efficiency of the culvert inlet. Side-tapered or slope-tapered inlets, as illus-
trated in Figure 9.4, may be employed to fulfill this need. (Note that the slope-tapered
inlet in this figure also has a side taper.) These special inlet configurations are more
costly to construct but can be advantageous in locations where other alternatives fail
to provide the required performance. Harrison et al. (1972) provide further informa-
tion on the hydraulic performance for these types of culverts.
Regardless of the type of end treatment used for a culvert, it is recommended that a
concrete cradle or cutoff wall (see Figure 9.5) be installed around and under the ends
of the pipe. The cradle is especially important at the culvert outlet to prevent under-
cutting of the pipe and subsequent collapse. Of course, if a headwall end treatment is
used, the headwall itself performs this function. If a concrete apron is used at the
entrance or exit of a culvert, a cutoff wall should be integrated into it to prevent under-
mining.
Figure 9.4
Side-tapered and
slope-tapered culvert
entrances
Credit: FHWA, 1985
Figure 9.5
Concrete cradle at a
pipe outlet
Credit: FHWA, 1985
It is frequently desirable to provide trash racks (sometimes called safety racks)

secured to the wingwalls at culvert entrances and/or exits to prevent unauthorized
entry and/or clogging by debris. Trash racks should have clear openings between bars
no greater than about 6 in. (150 mm) and should be constructed in modular sections to
facilitate removal for maintenance of the culvert. The rack itself should receive peri-
odic maintenance to clear trapped debris. Installing the rack so that the bars are
Section 9.4 Alignments of Culverts 331
inclined from the vertical can help keep the rack clear because the flowing water will
push floating debris up the slope of the rack. Inclining the rack also makes it easier for
a person trapped against the rack to climb out.
Although guidelines and recommendations vary, a trash rack should be structurally
designed to withstand the hydraulic loads that would occur under partial or full clog-
ging of the rack. Hydraulically, a trash rack causes a head loss to occur, but the loss
essentially can be neglected if the total clear opening area of the rack is much greater
(about four times) than that of the cross-sectional area of the culvert barrel. Racks
should be placed far enough upstream of the culvert entrance to ensure that the veloc-
ity of the accelerating flow is low enough for a person to escape.
9.4 ALIGNMENTS OF CULVERTS

Generally, the horizontal and vertical alignments of a culvert should conform as
closely as possible to those of the channel. However, it may be necessary or desirable
in some cases to depart from this ideal. Discussions of alignment problems and solu-
tions are presented in the subsections that follow.
Horizontal Alignment
If a stream channel is very sinuous, it can be difficult or impossible to develop a cul-
vert alignment that provides a smooth transition of the flow at both the upstream and
downstream ends of a culvert. In such instances, it may be desirable to realign the
channel to provide for smooth transitions. However, such efforts should be undertaken
with great care because the newly aligned channel may be unstable, resulting in sedi-
mentation and/or erosion problems upstream or downstream of the culvert. Brice
(1981) concluded that channel realignments are generally successful, except in
instances where the channel was already unstable.
An alternative to channel realignment is to use a curved or doglegged alignment for
the culvert barrel (see Figure 9.6). Of course, such a configuration will result in addi-
tional energy losses due to the curve or bend. Another disadvantage is that debris
tends to build up and cause culvert blockages at bend locations. If a culvert operates
under inlet control [see Section 2.2 (page 24) and Section 9.5 (page 335)], additional
head losses caused by a curve or bend will not affect the headwater elevation. The
headwater elevation will be affected if the culvert operates under outlet control.
Changes in the horizontal alignment of a culvert should ideally be accomplished
using a curved alignment. In cast-in-place concrete box culvert construction, curved
concrete forms may be used. In pipe culvert construction, small deflections at individ-
ual pipe joints according to the manufacturer’s recommendations can be accom-
plished to approximate a smooth curve. If a sudden bend is used instead of a curved
alignment, the bend deflection angle should be limited to no more than approximately
15 degrees, and such bends should be separated from one another by a distance of at
least 50 ft (15 m) (ASCE, 1993).
Figure 9.6
Curved and doglegged
alignments
Vertical Alignment
In a sag culvert, one or more of the straight barrel segments has an adverse slope, as
illustrated in Figure 9.7(a), which usually results in part of the culvert flowing full and
under pressure. A sag culvert is especially prone to clogging by trapped sediments
and debris, but it can be the best option if its use prevents, for instance, a costly utility
relocation. Sag culverts are also used frequently at roadway crossings of irrigation
canals.
Broken-back culverts [Figure 9.7(b)] have bends in their vertical alignments. The
bend in a broken-back culvert may cause the upstream part of the culvert to have a
smaller slope than the downstream part, or vice versa. A broken-back culvert may be
used to minimize excavation, avoid a rock outcropping, or reduce the exit velocity
from the culvert.
Section 9.4 Alignments of Culverts 333
Figure 9.7
Sag and broken-back
culverts (Norman,
Houghtalen, and
Johnston, 2001)
If a broken-back culvert has a steeper slope downstream of the bend, the possibility
exists that the hydraulic grade will drop below the crown of the culvert in the vicinity
of the bend although it is otherwise above the crown. This phenomenon causes sub-
atmospheric pressure to develop within the culvert at the bend location and can poten-
tially lead to structural and hydraulic problems. If this possibility exists, a vent should
be provided to allow atmospheric pressure to exist at the bend location.
Hydraulic analyses of broken-back and sag culverts are complicated by the fact that
they usually involve regions of both full (pipe) flow and open-channel flow. Further,
open-channel flow may be subcritical in one part of the culvert and supercritical in
another. Thus, the potential exists for many flow controls, and each situation must be
treated differently. Computer-based modeling tools can quickly evaluate flow regimes
for complicated culvert geometries, and hence can be used to optimize designs.
Skewed Culverts
Positioning a culvert alignment to approximate the flow path of the stream channel
often involves placing the culvert at an angle that is not perpendicular to the embank-
ment centerline, as shown in Figure 9.8, parts (b) and (c). This culvert alignment is
referred to as skewed in relation to the normal embankment alignment. In the case of
skewed culverts, headwalls may be skewed with respect to the barrel alignment, as
illustrated in Figure 9.8(c), to better align with embankment slopes (that is, the culvert
is skewed with respect to the embankment, and the headwalls are skewed by an offset-
ting amount with respect to the culvert so that they will still align with the embank-
ment slopes). If the headwall on a culvert is skewed with respect to the barrel, an
additional entrance loss occurs; however, this loss is usually small and may be
neglected in many cases. This additional loss can be estimated by using the design
charts published in FHWA’s Hydraulic Design Series Number 5, Hydraulic Design of
Highway Culverts (Normann, Houghtalen, and Johnston, 2001), also known as HDS-
5.
Figure 9.8
Skewed culvert
alignments (Normann,
Houghtalen, and
Johnston, 2001)
9.5 CULVERT HYDRAULICS

The hydraulics of a culvert are generally complex because of the short conduit length
and the several flow control types that may exist. A culvert may flow less than full
over part or all of its length, or it may flow full (closed-conduit flow) over its entire
length. Culverts are frequently not long enough for uniform flow to be achieved and
maintained; therefore, the entire flow profile often consists of gradually and/or rap-
idly varied flow. The flow control may also change with changes in discharge (that is,
the culvert may operate under one hydraulic control for some discharges, and under
one or more other controls for other discharges).
In design, the head loss (which approximately equals the difference between the water
surface elevations upstream and downstream of the culvert) that can be tolerated in a
culvert is often limited. The most widely used design approach is to evaluate the per-
formance of a trial culvert design for several different flow control types. The worst-
performing control condition (the condition resulting in the greatest head loss) is then
used to evaluate the acceptability of the proposed design. This method is a conserva-
tive, worst-case design approach that ignores conditions that may result in better per-
formance. The benefits of this approach are its ease of use and the assurance that the
culvert will perform adequately under the most adverse conditions.
Section 9.5 Culvert Hydraulics 335
Two general methods can be used to predict the hydraulic performance characteristics
of a culvert, and both are presented in this section. The graphical method involves the
use of nomographs that relate the headwater elevation at the culvert entrance to fac-
tors such as the length, size, and roughness of the barrel and the tailwater elevation
downstream of the outlet. This method is relatively easy to use and provides results
that are adequate for most purposes. It should not be used for culverts containing hor-
izontal curves or bends unless the additional losses are accounted for separately. It
should almost never be used for a broken-back or sag culvert.
The second method for predicting culvert performance is to compute the flow profile
using the gradually varied flow techniques presented in Section 7.8. This method is
much more laborious than the first but can yield much more accurate predictions of
culvert performance in many cases. This method is usually required for broken-back
and sag culverts. Because of the number of calculations involved, computer programs
such as CulvertMaster are especially beneficial when using this method.
Flow Controls
Research by the National Bureau of Standards (NBS) and the FHWA in the United
States divides culvert flow controls for a straight, uniformly shaped culvert into two
basic classes depending on the location of the control section: inlet control and outlet
control (Normann, Houghtalen, and Johnston, 2001). With inlet control, the culvert’s
entrance characterisics determine its capacity. Conversely, with outlet control, the
inlet can accept more flow than the culvert can carry, because of either the friction
losses in the barrel or the tailwater elevation. Descriptions of the various flow regimes
encountered under both control types follow. Additional flow controls may exist in
broken-back or sag culverts, and in those with improved inlet configurations.
Inlet Control. For a culvert operating under inlet control, the culvert barrel is
capable of conveying a greater discharge than the inlet will accept. The flow control
section is just inside the culvert barrel at its entrance, and the flow profile passes
through critical depth at this location. The culvert is hydraulically steep, and the flow
regime downstream of the control section is supercritical with an S2 profile [see Sec-
tion 7.8 (page 274) and Figure 7.24]. Conditions downstream of the entrance have no
effect on the culvert capacity.
Figure 9.9 illustrates four cases in which a culvert operates under inlet control. Note
the drawdown of the S2 water surface profile near the upstream end of the barrel in
each case.
Case A in Figure 9.9 represents a situation in which neither the inlet nor the outlet of
the culvert are submerged by the headwater or tailwater. The flow approaches normal
depth as it moves along the culvert, and, depending on the flow depth in the down-
stream channel, a hydraulic jump may form downstream of the culvert outlet.
Case B illustrates a situation in which the outlet is submerged but the inlet is not sub-
merged. The culvert is hydraulically steep. An S2 water surface profile forms inside
the culvert entrance, and flow approaches normal depth. Submergence of the down-
stream end of the culvert results in pressure flow near the outlet end of the barrel. A
hydraulic jump forms inside the culvert barrel and provides a transition between the
supercritical and pressure flow regimes.
Figure 9.9
Inlet control flow
conditions (Norman,
Houghtalen, and
Johnston, 2001)
In Case C, the inlet is submerged by the headwater, but the outlet is not submerged.
As in Case A, the flow approaches normal depth as it moves along the culvert. A
hydraulic jump may form in the downstream channel.
Case D shows that a culvert may flow partly full even for cases in which both the inlet
and the outlet are submerged. The hydraulically steep nature of the culvert promotes
the S2 profile near the entrance end of the barrel. A hydraulic jump forms a transition
between the upstream supercritical flow and the downstream pressure flow. A vent,
such as a median drain, should be provided in a culvert anticipated to operate in this

mode to permit atmospheric pressure to exist within the culvert.
Other than the discharge, factors affecting headwater depth for a culvert operating
under inlet control include the cross-sectional area and shape of the barrel and the
inlet’s edge configuration. The barrel slope also influences the headwater depth, but
its effect is small. A sharp inlet edge, such as that occurring on a thin-walled metal
pipe with a projecting end, causes a large flow contraction at the culvert entrance. A
square-edged inlet, or better yet, an inlet with a rounded or pipe socket edge, would
reduce the headwater elevation in this case.
For a given allowable headwater elevation, the capacity of a culvert can be enhanced
by depressing the invert of the culvert barrel below the channel bed (with the fill
retained by wingwalls), thereby increasing the headwater depth. The height of the
depression T, as shown in Figure 9.10, is called the fall.
Figure 9.10 shows that most of the fall T occurs over the front portion of the culvert
approach apron, and this portion of the apron should have a minimum horizontal
length of 2T. The remainder of the fall occurs over the second portion of the approach
apron, which consists of a projection of the culvert slope for a distance equal to half
the height D of the culvert.
Figure 9.10
Fall at culvert
entrance (Norman,
Houghtalen, and
Johnston, 2001)
If not done properly, provision of a depressed inlet can lead to headcutting and degra-
dation of the upstream channel. Grade control structures in the form of buried riprap
or concrete walls placed at various intervals along the channel bottom may become
necessary to provide a measure of protection against upstream headcutting and degra-
dation. Also, the approach apron should be paved. When a depressed inlet is used,
sufficient elevation change must exist to ensure a positive slope between the upstream
and downstream ends of the culvert.
Outlet Control. Outlet control occurs when the slope of a culvert is hydraulically
mild or when the tailwater elevation is high enough to affect the headwater elevation.
In such cases, the inlet is capable of accepting more flow than the barrel and/or down-
stream flow conditions will permit. The control section is at or near the outlet point,
and the flow condition in the barrel near the outlet is critical, subcritical, or full (pres-
sure) flow.
For a culvert operating under outlet control, many factors affect the headwater depth.
These factors include discharge; barrel cross-sectional area, shape, length, roughness,
and slope; inlet edge configuration; and tailwater depth. Figure 9.11 illustrates seven
cases in which outlet control may occur.
Case A, shown in Figure 9.11, represents a situation in which the culvert barrel flows
full and under pressure over its entire length. Both the inlet and the outlet are sub-
merged. This condition is often assumed to exist in outlet control calculations, but sel-
dom occurs in actual culvert performance.
Case B is a situation in which the outlet is submerged but the headwater elevation is
only slightly higher than the crown of the inlet. In this instance, because of the low
submergence of the inlet and the flow contraction that occurs at that location, the
crown of the inlet is exposed where the water enters the barrel.
Case C represents a rare situation in which the inlet is submerged, the outlet is not
submerged, and the barrel flows full over its entire length. A large headwater depth is
required to cause a culvert to flow full over its entire length when the outlet is not sub-
merged. The exit velocity in this case is quite high and can cause severe erosion in the
downstream channel.
Case D occurs more commonly than does Case C. Like Case C, it represents a situa-
tion in which the inlet is submerged but the outlet is not. If the tailwater elevation is
below critical depth, the flow passes through critical depth at the outlet end of the cul-
vert barrel. If the tailwater is between critical depth and normal depth for the barrel,
profile calculations begin at the tailwater depth. Because of the hydraulically mild
slope of the barrel, an M2 water surface profile forms upstream of the critical depth
(control) section. The M2 profile intersects the crown at some distance from the out-
let, and upstream of that point the culvert flows full and under pressure.
Case E is similar to Case D. If the tailwater is below critical depth, flow passes
through critical depth at the outlet and profile calculations begin at critical depth. If
the tailwater depth is between normal depth and critical depth, profile calculating
begin at the tailwater elevation. An M2 water surface profile occurs upstream of that
point. The normal depth is smaller than the height of the barrel, and the flow profile
approaches it asymptotically. Flow in the culvert remains less than full throughout its
length.
Figure 9.11
Outlet control flow
conditions (adapted
from Norman,
Houghtalen, and
Johnston, 2001)
Cases F and G can occur when the tailwater at the outlet is above normal depth for a
culvert operating with subcritical flow conditions. Depending on the culvert entrance
characteristics, the resulting backwater profile may take precedence over inlet control
conditions.
Graphical Analysis of Culvert Performance

A traditional approach to culvert performance analysis is graphical in nature and
makes use of nomographs published in HDS-5 (Normann, Houghtalen, and Johnston,
2001). The nomographs are published for both inlet and outlet control conditions, for
different barrel shapes and materials, and for different entrance edge configurations.
Because of the large number of possible combinations of these influential factors,
many nomographs are needed.
The following subsections describe and illustrate the use of the charts for both inlet
and outlet control conditions. When designing a culvert, the headwater depths com-
puted using inlet and outlet control assumptions are compared, and the larger value is
used to determine whether the design is acceptable.
Inlet Control Charts. Figure 9.12 is an inlet control nomograph for concrete
pipe culverts with circular cross sections. It is only one of the many inlet control
charts available for different combinations of pipe material, shape, and entrance edge
configuration. This particular chart illustrates the relationship that exists between the
culvert diameter, discharge, and headwater depth for a circular concrete culvert with
three possible entrance types: (1) square edge of pipe with headwall; (2) groove end
of pipe with headwall; and (3) groove end of pipe projecting from fill.
For a specified discharge, culvert diameter, and entrance edge configuration, the
nomograph in Figure 9.12 can be used to estimate HW/D, which is a dimensionless
ratio of the required headwater depth to the diameter. As illustrated with the example
given in the figure, a straight line can be drawn from a point on the pipe diameter
scale, through the appropriate point on the discharge scale, to a point on HW/D scale
(1). The value on the HW/D scale defines the required depth ratio if the culvert inlet
configuration consists of a headwall with a square-edged barrel entrance. If the inlet
configuration is instead a groove (socket) pipe end with a headwall or a projecting
groove pipe end, a horizontal line should be projected from the HW/D scale (1) inter-
section point to scale (2) or scale (3), respectively.
Example 9.1 – Inlet Control for a Circular Culvert. A circular concrete pipe culvert has
a diameter of 36 in., a length of 100 ft, and a slope of 0.50 percent, as illustrated in Figure E9.1.1. The
invert of the culvert will match the channel bottom elevation, which is 125.30 ft at the upstream end.
Assuming that the culvert operates under inlet control, use Figure 9.12 to estimate the headwater ele-
vations corresponding to discharges of 30 and 60 cfs. The culvert inlet consists of a headwall with a
square edge at the barrel entrance.
Solution: A line projected to scale (1) from the points representing 36 in. and 30 cfs on the diameter
and discharge scales yields HW/D = 0.90. The headwater depth is therefore
HW = 0.90D = 0.90(36/12) = 2.70 ft
The headwater elevation is the sum of the channel bed elevation and the headwater depth, or
125.30 + 2.70 = 128.00 ft
For the discharge of 60 cfs, HW/D = 1.65, and the headwater depth is therefore
HW = 1.65(3.00) = 4.95 ft
The headwater elevation is
125.30 + 4.95 = 130.25 ft
Figure E9.1.1 Culvert for Example 9.1
The next example shows that simply changing the entrance edge configuration can
turn an otherwise unacceptable design into an acceptable one, thus illustrating the sig-
nificant influence that the entrance edge has on the performance of a culvert operating
under inlet control.
Example 9.2 – Inlet Control for a Circular Culvert with Different Entrances.
Consider the same culvert described in the previous example, but assume that the maximum allowable
headwater elevation is 129.65 ft when the discharge is 60 cfs. Clearly, the square edge with headwall
design is not acceptable on this basis. Without changing the culvert diameter, determine whether an
alternative type of culvert entrance configuration can be used to lower the headwater elevation to
below 129.65 ft.
Solution: As described in Example 9.1, the value of HW/D found on scale (1) for the discharge of 60
cfs is 1.65. Extending a horizontal line on the chart from that point, one determines values of HW/D
on scales (2) and (3) to be 1.44 and 1.47, respectively. The respective headwater depths are 4.32 and
4.41 ft, and the respective headwater elevations are 129.62 and 129.71 ft. Comparing these values to
the allowable headwater depth of 129.65, it is concluded that use of a groove end with a headwall will
achieve the desired performance, whereas a projecting groove end (marginally) will not.
Outlet Control Charts. An outlet control nomograph such as the one shown in
Figure 9.13 is used to predict the culvert head, H, for a culvert operating under outlet
control. The head is equal to the total head loss through the culvert, but, in practice, it
can usually be taken as the difference between the headwater and tailwater elevations.
The variables that impact the outlet control head are culvert discharge, cross-section
geometry, length, entrance loss coefficient, and roughness. The nomograph presented
here is for circular concrete culverts with an n value of 0.012. As with the inlet control
nomograph from Figure 9.12, it is only one of many such charts published for differ-
ent combinations of culvert materials and shapes.
To use the nomograph shown in Figure 9.13, the engineer must first determine which
of the two length scales shown should be used based on the entrance loss coefficient
ke. The descriptions and entrance loss coefficient values provided in Table 9.1 (which
is a repeat of Table 6.5) can be used to determine which length scale (in this example,
ke = 0.2 or ke = 0.5) should be used.
Figure 9.12
Nomograph to
compute headwater
depth for circular
concrete culverts with
inlet control
(Normann,
Houghtalen, and
Johnston, 2001)
After the appropriate length scale has been determined, a straight line connecting the
applicable points on the diameter and length scales is drawn. Then, the head value is
determined by drawing a second straight line from the appropriate value on the dis-
charge scale, through the intersection point of the first line and the “turning line,” to
the head scale. An example of how these lines are constructed is shown on the chart.
Finally, an energy balance for full flow through a culvert with both the inlet and outlet
submerged can be written to express the headwater depth as
Figure 9.13
Nomograph for
computing head on
circular concrete pipes
with outlet control
(Normann,
Houghtalen, and
Johnston, 2001)
HW= ho + H − SoL (9.1)

where HW = headwater depth (ft, m)
ho = tailwater depth (ft, m)
H = total head loss through the culvert (ft, m)
So = culvert barrel slope (ft/ft, m/m)
L = culvert length (ft, m)
This expression assumes that the velocity heads in the upstream and downstream
channels are equal, which is a reasonable approximation in most instances.
Table 9.1 Entrance loss coefficients for pipes and culverts operating under outlet control
Structure Type and Entrance Condition ke
Concrete pipe
Projecting from fill, socket or groove end 0.2
Projecting from fill, square edge 0.5
Headwall or headwall and wingwalls
Socket or groove end 0.2
Square edge 0.5
Rounded (radius = D/12) 0.2
Corrugated metal pipe or pipe-arch
Projecting from fill (no headwall) 0.9
Headwall or headwall and wingwalls (square edge) 0.5
Reinforced concrete box
Headwall parallel to embankment (no wingwalls)
Square-edged on 3 sides 0.5
Rounded or beveled on 3 sides 0.2
Crown edge rounded or beveled 0.2
Wingwalls parallel (extensions of box sides)
One method for determining the tailwater depth in Equation 9.1 is to approximate it
using a uniform flow formula such as the Manning equation. This approximation
requires information on the downstream channel cross-sectional dimensions, slope,
and roughness. Also, this approach assumes that the flow in the downstream channel
is approximately uniform and that flow controls downstream of the culvert do not
cause backwater or similar effects. An alternative (and more accurate) method for
determining the tailwater depth is to perform a flow profile analysis for the down-
stream channel. This method is clearly more laborious than the uniform flow method
just described, but is required when backwater or other effects cause the tailwater
depth to differ appreciably from that of uniform flow.
Example 9.3 – Outlet Control for a Circular Culvert with Tailwater Effects.
Repeat Example 9.1 assuming that the culvert operates under outlet control. The invert elevation at
the downstream end of the culvert is 124.80 ft, as illustrated in Figure E9.3.1. Backwater effects due
to an obstruction in the downstream channel cause the tailwater depth to be ho = 4.20 ft when Q = 30
cfs, and ho = 4.70 ft when Q = 60 cfs. By comparing the inlet and outlet control solutions, determine
whether the culvert operates under inlet or outlet control for discharges of 30 and 60 cfs.

Solution: From Example 9.1, the circular concrete culvert has a diameter of 36 in., a length of 100 ft,
and a slope of 0.50 percent. The channel bed elevation at the upstream end of the culvert is 125.30,
and the inlet configuration is a headwall with a square-edged barrel entrance.
From Table 9.1, the entrance loss coefficient for the this type of inlet configuration is 0.5. Using a
diameter of 36 in. and a length of 100 ft (on the scale representing ke = 0.5), a straight-edge can be
used on Figure 9.13 to position a point on the turning line. Using that point and the discharge of 30
cfs, a line can then be projected to the head scale to determine H = 0.60 ft. When the discharge is 60
cfs, the head loss is similarly found to be H = 2.45 ft.
Application of Equation 9.1 for the case when Q = 30 cfs yields
HW = 4.20 + 0.60 – 0.005(100) = 4.30 ft
The headwater elevation is therefore
125.30 + 4.30 = 129.60 ft
Because this elevation is higher than the headwater elevation of 128.00 ft computed for inlet control,
conservative practice dictates that the outlet control headwater elevation of 129.60 ft governs for a
discharge of 30 cfs.
For the case of Q = 60 cfs, Equation 9.1 yields
HW = 4.70 + 2.45 – 0.005(100) = 6.65 ft
The headwater elevation is therefore
125.30 + 6.65 = 131.95 ft
Because this value is again higher than the headwater elevation determined for the assumption of inlet
control (130.25 ft), outlet control again governs and the headwater elevation is 131.95 ft for a dis-
charge of 60 cfs.
Figure 9.13 and other similar charts are based on the assumption that the culvert flows
full over its entire length and is submerged at both its upstream and downstream ends.
Nevertheless, such charts can be used to approximate outlet control behavior when the
tailwater elevation drops to or below the crown of the culvert outlet (Cases C through
G in Figure 9.11). The procedure yields the best approximation, however, when the
culvert flows full over at least a portion of its length. When the culvert flows only
partly full over its entire length, as in Case E, the method deteriorates as the headwa-
ter elevation drops further below the top of the culvert pipe. Adequate results may be
obtained for headwater depths as small as 0.75D, where D is the height of the culvert
barrel. Accurate calculations require flow profile analysis, as described in the next
section.
If one chooses to use the approximate nomograph analysis of outlet control in Case D
or E, the tailwater depth used in Equation 9.1 should be the larger of (1) the actual
tailwater depth or (2) the arithmetic average of the critical depth in the barrel and the
height of the barrel [(yc + D)/2]. The critical depth in the barrel can be determined by
using the methods described in Section 7.4, page 246, or it can be estimated for circu-
lar pipes using Figure 9.14. As noted in this figure, in no case can the critical depth be
greater than the height of the barrel.
Example 9.4 – Outlet Control for a Circular Culvert with Smaller Tailwater
Effects. Repeat Example 9.3 (page 344) assuming that the obstruction in the downstream channel
has been removed and lowering the tailwater depths to ho = 1.80 ft when Q = 30 cfs and ho = 2.70 ft
when Q = 60 cfs, as illustrated in Figure E9.4.1. By comparing the inlet and outlet control solutions,
determine whether the culvert operates under inlet or outlet control for each of the discharges.

Solution: From Example 9.3, H = 0.60 ft when Q = 30 cfs. Using Figure 9.14, the critical depth in the
culvert barrel for this discharge is about 1.75 ft. When using the outlet control nomographs with Case
D or E flow, the tailwater depth should be taken as the larger of the actual tailwater depth or the aver-
age of tailwater depth and barrel height. Averaging this depth with the culvert diameter yields 2.38 ft.
Comparing this value with the actual tailwater depth of 1.80 ft and taking the larger of the two leads to
ho = 2.38 ft. Substitution into Equation 9.1 yields
HW = 2.38 + 0.60 – 0.005(100) = 2.48 ft
This value is approximately 83 percent of the height of the culvert barrel, and hence the result, while
only an approximation, may be adequate. The headwater elevation is
125.30 + 2.48 = 127.78 ft
Comparing this elevation to the inlet control headwater elevation of 128.00 ft found in Example 9.1
(page 340) shows that inlet control governs the performance. The headwater elevation for this dis-
charge would therefore be 128.00 ft.
When Q = 60 cfs, H = 2.45 ft (from Example 9.3). The critical depth for this discharge is about 2.50 ft
(from Figure 9.14). Averaging the critical depth with the culvert barrel height yields
(3.00 + 2.50)/2 = 2.75 ft
Because 2.75 ft is larger than the actual tailwater depth, ho = 2.75 ft. Substitution into Equation 9.1
yields
HW = 2.75 + 2.45 – 0.005(100) = 4.70 ft
Because this depth is larger than the culvert diameter, the culvert flows full over part of its length;
therefore, the quality of the approximation is better than it is for the smaller discharge of 30 cfs. The
headwater elevation is
125.30 + 4.70 = 130.00 ft
Because this elevation is again lower than the headwater elevation determined for the assumption of
inlet control (130.25 ft), inlet control governs and the headwater elevation is 130.25 ft for a discharge
of 60 cfs.
Figure 9.14
Graphs for estimating
critical depths for
circular pipes
Credit: FHWA, 1985

Flow Profile Analysis of Culvert Performance

As shown in the previous sections and examples, culvert performance analysis using
the nomographs published in HDS-5 (Normann, Houghtalen, and Johnston, 2001) is
convenient, but doing so yields only approximate solutions in many practical cases.
More accurate analysis requires computation of flow profiles and makes use of the
closed-conduit and open-channel flow relationships presented in Chapter 6 and Chap-
ter 7. The subsections that follow describe flow profile analysis methods for both inlet
and outlet control. The tedious nature of these methods makes the use of computer-
based modeling tools attractive. The methods described herein are the basis for the
algorithms used in many modern computer tools.
Inlet Control. The entrance to a culvert operating under inlet control behaves as a
weir when the headwater depth is smaller than the culvert entrance height. It operates
as an orifice if the headwater submerges the entrance. This behavior is shown by the
two curves in Figure 9.15. Between the two curves representing weir flow and orifice
flow is a transition zone representing an intermediate type of flow. A graph such as
Figure 9.15 is a key to inlet control performance evaluation using flow profile analy-
sis.
When the culvert entrance operates as a weir, the HW/D ratio can be expressed as
M
HW Ec  Cf Q 
= + K   − 0.5So (9.2)
D D 1/ 2
 AD 
An alternative expression is
M
HW  Cf Q 
= K   (9.3)
D 1/ 2
 AD 
where HW = headwater depth (ft, m)

D = culvert barrel height (ft, m)
Ec = specific energy (ft) corresponding to the barrel critical depth
(Ec = yc + Vc2/2g), (ft, m)
A = full cross-sectional area of the barrel (ft2, m2)
So = barrel slope (ft/ft, m/m)
Cf = conversion factor (1.0 U.S. Customary, 1.811 SI)
K and M are constants determined from hydraulic model tests
Values of K and M are provided in Table 9.2, along with an indication of whether
Equation 9.2 or 9.3 should be used in the calculation. Equation 9.2 is theoretically
better than Equation 9.3, but Equation 9.3 is easier to apply in practice. The slope
term expressed as (–0.5So) in Equation 9.2 should be changed to (+0.7So) if the barrel
entrance is mitered. Equations 9.2 and 9.3 are applicable for values of Q/AD1/2 less
than or equal to about 3.5 for U.S. Customary units, or 1.93 for SI units.
Figure 9.15
Transition from weir
to orifice control in a
culvert (Norman,
Houghtalen, and
Johnston, 2001)
If Q/AD1/2 is between about 3.5 (1.93 SI) and 4.0 (2.21 SI), transitional flow occurs,
and the headwater value must be interpolated as shown in the transition zone in Figure
9.15. Above this range, the culvert entrance acts as an orifice, and the HW/D ratio can
be expressed as (Normann, Houghtalen, and Johnston, 2001)
2
HW  Cf Q 
= c  + Y − 0.5So (9.4)
D 1/ 2 
 AD 
where c and Y are constants determined from hydraulic model tests

Values of the constants c and Y are provided in Table 9.2. Equation 9.4 is valid for val-
ues of Q/AD1/2 greater than or equal to about 4.0. Below this range, transitional flow
occurs. Again, the slope term (−0.5So) should be replaced with (+0.7So) if the barrel
entrance is mitered.
Table 9.2 Constants for inlet control equations (Normann, Houghtalen, and Johnston, 2001)
Unsubmerged (Weir Flow) Submerged (Orifice Flow)
Culvert Shape Inlet Edge Equation
and/or Material Description in text K M c Y
Square edge with headwall .0098 2.0 .0398 0.67
Circular, concrete Groove end with headwall (9.2) .0018 2.0 .0292 .74
Groove end projecting .0045 2.0 .0317 .69
Headwall .0078 2.0 .0379 .69
Circular, CMP Mitered to slope (9.2) .0210 1.33 .0463 .75
Projecting .0340 1.50 .0553 .54
Circular Beveled ring, 45° bevels (9.2) .0018 2.50 .0300 .74
Beveled ring, 33.7° bevels .0018 2.50 .0243 .83
30° to 75° wingwall flares .026 1.0 .0347 .81
Rectangular box 90° and 15° wingwall flares (9.2) .061 0.75 .0400 .80
0° wingwall flares .061 0.75 .0423 .82
Rectangular box 45° wingwall flares, d = 0.043D (9.3) .510 .667 .0309 .80
18° to 33.7° wingwall flares, d = 0.083D .486 .667 .0249 .83
90° headwall with ¾-in. chamfers .515 .667 .0375 .79
Rectangular box 90° headwall with 45° bevels (9.3) .495 .667 .0314 .82
90° headwall with 33.7° bevels .486 .667 .0252 .865
45° skewed headwall; ¾-in. chamfers .545 .667 .0505 .73
Rectangular box 30° skewed headwall; ¾-in. chamfers (9.3) .533 .667 .0425 .705
15° skewed headwall; ¾-in. chamfers .522 .667 .0402 .68
10-45° skewed headwall; 45° bevels .498 .667 .0327 .75
45° nonoffset wingwall flares .497 .667 .0339 .803
Rectangular box
18.4° nonoffset wingwall flares (9.3) .493 .667 .0361 .806
with ¾-in. chamfers
18.4° nonoffset wingwall flares; 30° .495 .667 .0386 .71
45° wingwall flares, offset .497 .667 .0302 .835
Rectangular box
33.7° wingwall flares, offset (9.3) .495 .667 .0252 .881
w/ top bevels
18.4° wingwall flares, offset .493 .667 .0227 .887
90° headwall .0083 2.0 .0379 .69
Corrugated metal
Thick wall projecting (9.2) .0145 1.75 .0419 .64
boxes
Thin wall projecting .0340 1.5 .0496 .57
Square edge w/ headwall .0100 2.0 .0398 .67
Horizontal ellipse,
Groove end w/ headwall (9.2) .0018 2.5 .0292 .74
concrete
Square edge w/ headwall .0100 2.0 .0398 .67
Vertical ellipse,
Groove end w/ headwall (9.2) .0018 2.5 .0292 .74
concrete
90° headwall .0083 2.0 .0379 .69
Pipe arch, CM,
Mitered to slope (9.2) .0300 1.0 .0463 .75
18-in. corner radius
Projecting .0340 1.5 .0496 .57
Projecting .0300 1.5 .0496 .57
Pipe arch, CM,
No bevels (9.2) .0088 2.0 .0368 .68
33.7° bevels .0030 2.0 .0269 .77
Projecting .0300 1.5 .0496 .57
Pipe arch, CM,
No bevels (9.2) .0088 2.0 .0368 .68
33.7° bevels .0030 2.0 .0269 .77
90° headwall .0083 2.0 .0379 .69
Arch, CM Mitered to slope (9.2) .0300 1.0 .0463 .75
Thin wall projecting .0340 1.5 .0496 .57
Under inlet control conditions, flow in the barrel just downstream of the culvert
entrance is supercritical and forms an S2 water surface profile that asymptotically
approaches normal depth. The direct-step method, described in Section 7.8 on page
275, can be used to compute the profile.
If the tailwater level is high and submerges the culvert outlet as shown in Cases B and
D of Figure 9.9, one of two possible situations occurs. In the first situation (which is
illustrated for Cases B and D in Figure 9.9), pressure flow occurs in the most down-
stream part of the culvert and a hydraulic jump forms inside the barrel. In the second
situation (not illustrated), the S2 water surface profile extends all the way through the
culvert to its outlet and a hydraulic jump may form downstream of the outlet. In either
case, formation of a hydraulic jump is contingent on a momentum balance being
achieved (see Section 7.5, page 260).
Example 9.5 – Inlet Control Equations. Assume that the culvert described in Example 9.1,
illustrated in Figure E9.5.1, operates under inlet control. Use Equations 9.2, 9.3, and/or 9.4 to deter-
mine the headwater depths for discharges of 30 cfs, 45 cfs, and 60 cfs. Compare the headwater depths
for the discharges of 30 cfs and 60 cfs to those determined in Example 9.1.

Solution: From Example 9.1, a circular concrete culvert with a square-edge headwall entrance has
L = 100 ft, D = 36 in., So = 0.5%, and an upstream invert elevation of 125.30 ft. For each of the three
discharges, the values of Q/AD1/2 and the corresponding flow types (based on criteria given in the
preceding sections) are as follows:
Q (cfs) Q/AD1/2 Flow Type

30 2.45 Weir flow
45 3.68 Transitional flow
60 4.90 Orifice flow
Consider first the discharge of 30 cfs, for which weir flow occurs. For a circular culvert with head-
walls and a square-edged barrel entrance, Table 9.2 indicates that the constants K1 and M1 are equal to
0.0098 and 2.0, respectively, and that these values should be used with Equation 9.2. By trial and
error, the critical depth in the barrel corresponding to this discharge is yc = 1.77 ft, and the corre-
sponding specific energy is Ec = 2.51 ft. Substitution of these values into Equation 9.2 yields
HW = 2.51 + 3.00(0.0098)(2.45)2 – 0.5(0.005)(3.00) = 2.68 ft

This value compares quite favorably with the depth of 2.70 ft determined in Example 9.1.
Orifice flow occurs for the discharge of 60 cfs; thus, Equation 9.4 applies. From Table 9.2, the values
of K2 and M2 are 0.0398 and 0.67, respectively. The computed headwater depth is
HW = 3.00(0.0398)(4.90)2 + 3.00(0.67) – 0.5(0.005)(3.00) = 4.87 ft

Again, this compares favorably with the depth of 4.95 ft determined in Example 9.1.
To determine the headwater depth for a discharge of 45 cfs (for which transitional flow occurs) a lin-
ear interpolation procedure will be used. Equation 9.2 applies for Q/AD1/2 values up to 3.5, whereas
Equation 9.4 applies for Q/AD1/2 values greater than 4.0. Thus, the largest discharge that Equation 9.2
applies to for this culvert is 43 cfs, and the smallest discharge that Equation 9.4 applies to is 49 cfs.
Using Equation 9.2, the headwater depth corresponding to a discharge of 43 cfs is found to be 3.48 ft.
Similarly, using Equation 9.4, the headwater depth corresponding to a discharge of 49 cfs is found to
be 3.91 ft. Linear interpolation between the coordinate pairs (Q = 43 cfs, HW = 3.48 ft) and (Q = 49
cfs, HW = 3.91 ft) produces a headwater depth of HW = 3.62 ft for a discharge of 45 cfs.
Outlet Control. The type of flow profile analysis performed for a culvert operat-
ing under outlet control depends on the profile condition from Figure 9.11 that is
being analyzed. Equation 9.1 can be applied directly if the culvert outlet is submerged
and the flow corresponds to either Case A (a full-flow condition) or Case B (full flow
except for the contraction at the culvert entrance) from Figure 9.11. Note, however,
that this equation assumes that the velocity heads in the upstream and downstream
channels are approximately equal and thus cancel one another out in the energy equa-
tion.
Case C is a case of full flow that occurs when the critical depth in the culvert barrel is
equal to the height of the barrel (it can be no larger) and the tailwater depth is smaller
than the height of the barrel. In this case, the hydraulic grade line elevation at the out-
let is equal to the elevation of the crown of the barrel. Assuming that the velocity head
in the upstream channel is negligible, the headwater elevation is greater than the outlet
crown elevation by an amount consisting of the sum of the velocity head in the barrel,
the frictional loss along the length of the barrel, and the entrance loss. The entrance
loss can be expressed as the product of the velocity head in the barrel and a minor loss
coefficient ke (see Table 9.1). If the Manning equation is applied for evaluation of
frictional losses, the friction slope Sf can be estimated using Equation 6.19. The prod-
uct of that slope and the barrel length is the total head loss due to friction. Thus, the
headwater elevation can be expressed as
V2
ELhw = ELo + D + S f L + (1 + ke ) (9.5)
2g
where ELhw = headwater elevation (ft, m)

ELo = invert elevation at the outlet end of the culvert (ft, m)
D = culvert barrel height (ft, m)
Sf = slope of the energy grade line (ft/ft, m/m)
L = culvert barrel length (ft, m)
ke = culvert entrance loss coefficient
V = full-flow velocity inside the culvert barrel (ft/s, m/s)
Cases D and E in Figure 9.11 occur when both the critical depth and the tailwater
depth are smaller than the height of the barrel. A flow profile analysis for either of
these cases begins with the determination of whether the tailwater depth is greater
than or equal to critical depth or less than critical depth. If the tailwater depth is
greater than or equal to critical depth, the depth at the control section at the down-
stream end of the barrel is equal to the tailwater depth, and computation of the M2
flow profile begins at this depth and proceeds in the upstream direction. If the tailwa-
ter depth is smaller than the critical depth, the M2 flow profile calculation starts at
critical depth and proceeds upstream.
A Case D condition occurs if the M2 water surface profile reaches the crown of the
culvert at some distance upstream of the outlet. Upstream of the point where the water
surface intersects the crown, the culvert flows full and the friction slope can evaluated
using Equation 6.19. An extension of the friction slope to the inlet of the culvert
yields the hydraulic grade line elevation just inside the culvert entrance. The entrance
loss and the full-flow velocity head can be added to this value to obtain the headwater
elevation.
A Case E condition occurs if the M2 flow profile extends all the way to the culvert
inlet without intersecting the crown. In this case, the hydraulic grade line elevation
just inside the barrel entrance is equal to the computed water surface elevation at that
point. The headwater elevation is determined by adding the velocity head just inside
the barrel entrance and the entrance loss to this computed water surface elevation.
Example 9.6 – Determination of Outlet Control with Gradually Varied Flow. A

rectangular concrete box culvert has a span of 1.2 m and a rise (height) of 1.0 m. It is 60.0 m in length
and has a slope of 0.0015, as illustrated in Figure E9.6.1. Assuming that the culvert operates under
outlet control and the tailwater depth is 0.7 m, determine the headwater elevation for a discharge of 2
m3/s. Use Manning’s n = 0.013. The invert elevation at the culvert outlet is 30.5 m, and the culvert
entrance has a loss coefficient of 0.2.

Solution: Because the tailwater depth is smaller than the barrel height, the outlet is not submerged.
The critical depth inside the rectangular barrel is found to be
yc = {22/[9.81(1.2)2]}1/3 = 0.66 m
Because this value is smaller than the barrel height, the flow corresponds to either Case D or Case E
in Figure 9.7. The normal depth in a rectangular channel with the width, slope, and roughness of the
culvert barrel is yn = 1.06 m. Because this depth is larger than the culvert height, the flow might corre-
spond to Case D. However, profile calculations are necessary to confirm this. Had the normal depth
been less than the barrel height, one could say without qualification that the flow would correspond to
Case E.
Computation of the M2 flow profile using the direct-step method (see Section 7.8 on page 275) is
summarized in the following table. Because the tailwater depth of 0.7 m exceeds critical depth, calcu-
lations begin at this elevation.
Depth (m) Distance from Outlet (m)

0.70 0
0.76 12.18
0.82 28.70
0.88 51.07
0.94 81.71
It can be seen from this summary that the water surface profile does not intersect the crown of the cul-
vert; thus, the flow profile corresponds to Case E.
More detailed M2 water surface profile calculations indicate that the depth of flow at the upstream
end of the culvert just inside its entrance is about 0.9 m. The corresponding velocity is
V = 2/[0.89(1.20)] = 1.85 m/s
and the velocity head is 0.18 m. The headwater depth is the sum of the depth inside the entrance, the
velocity head, and the entrance loss:
HW = 0.9 + (1 + 0.2)(0.18) = 1.116 m
The culvert invert elevation at its entrance is
30.5 + 60(0.0015) = 30.59 m
and thus the headwater elevation is
30.59 + 1.116 = 31.706 m
Outlet Velocity
It is necessary to evaluate the velocity of flow emerging from a culvert in order to
design measures to protect against erosion at the outlet (see Section 9.8). If the outlet
is submerged, or if the barrel flows full over its entire length, then the outlet velocity
is simply the ratio of the discharge to the cross-sectional area of the barrel. As shown
in Figures 9.9 and 9.11, this situation can occur for culverts operating under either
inlet or outlet control. When the outlet is not submerged, the procedure for determin-
ing the outlet velocity is more involved.
For a culvert operating under inlet control, an accurate estimate of the outlet velocity
requires a flow profile analysis. However, the outlet velocity can be approximated
using a simpler procedure that is conservative in that it always yields a value greater
than or equal to the actual velocity. This procedure assumes that normal depth exists
in the culvert at the outlet. Because normal depth may be smaller than the actual
depth, the cross-sectional area of the flow is assumed to be smaller than it may be in
reality, which in turn results in a higher calculated velocity. Determination of the nor-
mal depth and the associated cross-sectional area is a trial-and-error process that may
be accomplished using the procedures described in Section 7.3.
When a culvert operates under outlet control and the outlet is not submerged, the
depth of flow at the outlet should be taken as the larger of the tailwater depth and the
critical depth in the culvert barrel. The cross-sectional area of the flow in the barrel
Section 9.6 Roadway Overtopping 355
may be computed based on the chosen depth, and the outlet velocity may be deter-
mined.
Example 9.7 – Outlet Velocity Determination. Estimate the outlet velocity for the 4-ft by
3-ft culvert from Example 9.6 (page 353).
Solution: As described in Example 9.6, the flow is 2 m3/s, the outlet is not submerged, and the tailwa-
ter depth is 0.70 m. This depth is greater than the critical depth in the barrel, and thus is the depth at
which flow exits the barrel. The cross-sectional area of the flow at the barrel outlet is
A = 0.70(1.2) = 0.84 m2
The outlet velocity is therefore
V = Q/A = 2/0.84 = 2.38 m/s
9.6 ROADWAY OVERTOPPING

If the actual discharge approaching a culvert exceeds the design discharge for the
structure, or if the culvert has become clogged by accumulated sediment or debris, the
headwater elevation at the upstream end of the culvert will rise and may lead to over-
topping of the embankment, or flow over the embankment. Overtopping can cause a
hazard to the traveling public and, in severe cases, can cause the entire embankment
to fail and be washed out.
Flow over a roadway embankment usually occurs at the low point of a sag vertical
curve and is similar to the flow over a broad-crested weir. Equation 9.6 relates dis-
charge to the head on the weir:
Q = CdL(HWr)3/2 (9.6)

Cd = discharge coefficient (ft1/2/s, m1/2/s)
L = crest length normal to the direction of flow (ft, m)
HWr = head on the weir (ft, m)
HWr is computed as the difference between the headwater elevation and the roadway
embankment elevation. The discharge coefficient is determined as
C d = k tC r (9.7)
where kt = correction factor for submergence

Cr = base coefficient value (ft1/2/s, m1/2/s)
The value of Cr depends on HWr, the breadth of the embankment in the direction of
flow (Lr), and whether the embankment has a paved or gravel surface. The term kt is a
correction factor to account for the effect of potential submergence of the weir and is
a function of the ratio of the tailwater height ht to HWr. Numerical values of Cr and kt
may be obtained by using Figure 9.16. It is important to note that the value of Cr
obtained from Figure 9.16 will be in ft1/2/s. This value must be converted to m1/2/s
before solving Equation 9.6 using SI units.
Figure 9.16
Graphs for
determining the
broad-crested weir
discharge coefficient
used in computing
flow over a roadway
(Normann,
Houghtalen, and
Johnston, 2001)
The weir’s crest length and elevation can be difficult to estimate when the crest is
defined by a sag vertical curve in a roadway profile. An approximate solution in this
situation is to subdivide the total crest length into a series of horizontal segments. For
the example shown in Figure 9.17(a), the roadway crest is approximated as three weir
segments of lengths X1, X2, and X3. The discharge for each segment can be computed
separately, and then the total discharge can be computed as the sum of the individual
segment discharges.
Normann, Houghtalen, and Johnston (2001) reported that it is often adequate in cul-
vert design to represent a sag vertical curve by a single horizontal weir segment [see
Figure 9.17(b)]. The length of the weir X in this case can be taken as the distance
between the points on the sag vertical curve having the same elevation as the headwa-
ter. The head on the weir can be approximated as half of the difference between the
headwater elevation and the elevation of the lowest point on the vertical curve.
Section 9.6 Roadway Overtopping 357
Figure 9.17
Methods of
computing
overtopping
discharges at a sag
vertical curve
(Normann,
Houghtalen, and
Johnston, 2001)
Credit: FHWA, 1985
If a total discharge exceeding the design capacity of a culvert is specified, a trial-and-

error approach is required to determine how much of the flow passes through the cul-
vert barrel and how much overtops the roadway embankment. Clearly, the sum of the
two individual discharges must be equal to the specified total. If a performance curve
has been developed and graphed for the culvert installation, it can be used to obtain a
direct, graphical solution to a problem of this type.
Example 9.8 – Roadway Overtopping. For the culvert described in Example 9.1, assume
that the roadway embankment is paved and has a breadth of 70 ft, as illustrated in Figure E9.8.1. The
lowest roadway elevation at the sag point of a vertical curve is 130.70 ft. The sag vertical curve crest
of the roadway embankment is described by the parabolic equation
y = 130.70 + 0.000075x2
where y = crest elevation (ft)
x = distance from the lowest point in the sag (ft)
Estimate the discharge over the roadway embankment when the headwater elevation is 131.20 ft.
Assume that a single horizontal segment can be used to approximate the weir and the tailwater eleva-
tion downstream of the culvert does not submerge the weir.
Figure E9.8.1 Culvert and roadway profile for Figure 9.8

Solution: The maximum depth of submergence of the roadway embankment (weir) is the difference
between the headwater elevation and the minimum crest elevation, or
131.20 − 130.70 = 0.50 ft
The average head on the weir is HWr = 0.25 ft (one-half the maximum depth of submergence). The
length of the weir crest across which flow occurs is found by computing the values of x in the para-
bolic equation that satisfy y = 131.20. These values are –81.65 ft and +81.65 ft; thus, the total crest
length is L = 163.30 ft.
The ratio of HWr to Lr is
HWr/Lr = 0.25/70.0 = 0.00357
Section 9.7 Performance Curves 359
From Figure 9.16, kt = 1.00 and Cr = 2.97 ft1/2/s, so
Cd = ktCr = 2.97 ft1/2/s

Substituting this value into Equation 9.6, along with the crest length and average head, yields the dis-
charge over the embankment as
Q = 2.97(163.30)(0.25)3/2 = 61 cfs
9.7 PERFORMANCE CURVES

A performance curve for a culvert is a graphical relationship between the discharge
and the headwater elevation (or depth). As shown in Figure 9.18, the overall perfor-
mance curve for a culvert is constructed from individual performance curves repre-
senting inlet control, outlet control, and roadway overtopping. It therefore illustrates,
at a glance, the type of flow control that exists for any discharge of interest, the corre-
sponding headwater elevation, and the depth of flow over the roadway (if any). If
detention effects upstream of the culvert are significant and are thus taken into consid-
eration in optimizing the culvert design, then the performance curve is the same as the
stage-discharge curve required for the detention basin routing calculations (see Sec-
tion 12.6, page 505).
Figure 9.18
Construction of the
overall performance
curve for a culvert
(Normann,
Houghtalen, and
Johnston, 2001)
An inlet control performance curve such as the one illustrated in Figure 9.18 is con-
structed by assuming various discharges, determining the corresponding headwater
elevations, and plotting the headwater versus discharge data on a graph. Either the
nomographs or the inlet control equations presented in Section 9.5 (page 348) can be
used for this purpose.
Assumed discharges used in developing the curve should range from zero to a value
that is greater than the design discharge. If the inlet control equations are used instead
of nomographs, care should be taken to represent the transitional range of flows as
well as the weir and orifice flow ranges.
The outlet control performance curve is also graphed by assuming various discharges
and determining the corresponding headwater elevation for each. The chosen range of
discharges should span the design discharge. The curve can be developed by using
nomographs, but it is more accurate to use flow profile analysis as described in Sec-
tion 9.5 (page 352).
After the inlet and outlet control performance curves have been plotted on the same
graph, the overall performance curve is formed from those parts of the individual
curves that have the highest headwater for a given discharge. For example, at loca-
tions on the graph where the inlet control curve is higher than the outlet control curve,
the overall performance curve is the same as the inlet control curve. In a similar vein,
the overall performance curve is the same as the outlet control curve where it is higher
than the inlet control curve (see Figure 9.18). This procedure for delineation of the
overall performance curve applies only to the portion for which the headwater eleva-
tion is at or below the roadway crest.
For headwater elevations above the crest, roadway overtopping must also be consid-
ered. The head on the weir represented by the sag vertical curve is equal to the differ-
ence between the headwater elevation and the roadway crest elevation. Knowing this
head, the discharge over the weir can be determined by using Equation 9.6. This dis-
charge, added to the discharge that passes through the culvert for that same headwater
elevation, defines the overall performance curve for the region where the headwater
elevation is above the roadway crest. As shown in Figure 9.18, this portion of the
overall performance curve is generally quite flat compared to the rest of the curve,
indicating that the water surface elevation rises much more slowly with increasing
discharge once roadway overtopping has begun.
9.8 EROSION PROTECTION

The velocity at which water exits a culvert is usually much higher than the velocity in
the channel and can cause objectionable scour and erosion if the excess kinetic energy
associated with the flow is not dissipated. Section 9.5 (page 354) describes how to
estimate the velocity at which a flow exits a barrel. That exit velocity, along with geo-
metric information related to the culvert design, can be used to predict the size and
extent of required riprap erosion protection at the outlet (see Figure 9.19). Section
11.7 describes this process.
Section 9.8 Erosion Protection 361
Figure 9.19
Riprap outlet
protection
For instances in which the kinetic energy caused by the exit velocity of the flow is too
great to be safely dissipated using a riprap blanket, several options are available. The
exit velocity may be reduced by modifying the culvert barrel near the outlet. For
example, one might design a broken-back culvert with a smaller slope at the outlet
than at the inlet, thereby reducing the outlet velocity. Another approach is to increase
the barrel roughness near the outlet by using a material such as corrugated metal
instead of concrete in this area. The rougher corrugated metal increases the depth and
reduces the velocity near the barrel outlet.
If barrel modifications such as those just described cannot reduce the outlet velocity
to an acceptable limit, one must resort to use of a type of stilling basin structure at the
outlet, such as an impact-type energy dissipator (which uses a baffle in front of the
water jet to dissipate energy). Corry et al. (1983) provide further direction on this sub-
ject.
In culvert design, it is often desirable to provide erosion protection measures at the
entrance as well as the outlet. As flow enters a culvert, it must usually contract into a
smaller area than that of the upstream channel. The result is increased velocity, which
in turn leads to localized erosion of the channel bed and banks. The erosion protection
necessary at an entrance is typically not as extensive as that required at the outlet—a
riprap blanket, concrete apron, or slope paving is usually sufficient.
9.9 ENVIRONMENTAL CONSIDERATIONS

Culverts can often be used in lieu of bridges for small streams; however, culverts can
pose obstacles to migration pathways if the stream supports fish and similar aquatic
organisms. For instances in which fish passage is a concern, it is essential that input
and requirements of state and/or federal fish and wildlife protection officials be
obtained early in the design process. In some instances, a bridge may be required even
though a less expensive culvert could provide the required hydraulic performance.
To ensure that culverts are suitable for fish passage, additional design criteria may be
introduced to govern minimum flow depths within the barrel and minimum and/or
maximum flow velocities. Seasonalities of fish migration and streamflow are impor-
tant considerations here. The species of fish present in the stream are also important,
as some species are less tolerant of certain flow depths and velocities than are other
species. Another consideration is culvert lighting. Fish may not enter dark culverts
even when conditions are otherwise favorable. Long culverts may be constructed in
sections to allow natural light to enter. Artificial lighting may also be an option in
some circumstances.
Perhaps the most significant requirements for fish passage relate to the need to simu-
late the natural streambed environment within the culvert structure. If an open-bottom
culvert is used, such as a field-assembled corrugated arch structure, natural channel
materials can be left in place along the culvert bottom. This type of culvert requires
adequate foundation support for the culvert arch bearing ends. Localized scour or
general channel bed degradation that could undermine the foundations is another con-
cern. If a culvert pipe is used instead of an open-bottom structure, the barrel may be
oversized as shown in Figure 9.20. The barrel can then be partially filled with natural
streambed materials.
Figure 9.20
Partial filling of a
culvert pipe with
natural streambed
materials
In some cases, special baffles and/or weirs are installed in culvert barrels to simulate
natural riffles and pools that may exist in undisturbed areas of the stream. Design pro-
cedures for these types of installations, and for fish passage facilities in general, can
be found in the Model Drainage Manual (AASHTO, 1991) and in Chang and Nor-
mann (1976). Fish and wildlife agencies also may provide specific design guidance.
Section 9.9 Environmental Considerations 363
Modeling Focus: Supplemental Culvert Design
An existing roadway stream crossing consists of a 6-ft (span) ments. Because the existing culvert is in good condition, it
by 4-ft (high) concrete box culvert with the following charac- will remain in place but will be supplemented by a new cul-
teristics: vert in parallel.
• Upstream invert elevation = 124.75 ft The supplemental culvert must be capable of conveying
• Downstream invert elevation = 124.55 ft 295 cfs – 236 cfs = 59 cfs at the allowable headwater of
131.55 ft. The engineer decides to evaluate the feasibility of
• Length = 60.0 ft
supplementing with a circular, concrete culvert section. The
• Headwall with 30° to 75° wingwall flares, Ke = 0.5 existing headwalls will be extended to accommodate the new
The minimum elevation of the roadway sag vertical curve is culvert.
132.55 ft. The culvert outfalls to a downstream trapezoidal Assuming that the inverts and length for the new culvert will
channel with the following characteristics: match the existing culvert and using a “square edge with head-
• Channel invert elevation matches culvert outlet invert wall” entrance type, the engineer enters the needed capacity of
• Bottom width = 6 ft 59 cfs and allowable headwater elevation of 131.55 ft and
allows the program to solve for the minimum commercially
• Side slopes are = 1.5 H to 1.0 V
available size that meets the criteria-in this case, a 36-in. pipe.
• Manning’s n = 0.045
The 36-in. pipe is inserted into the program's worksheet for
• Channel slope = 0.0075 ft/ft the existing system, and the program performs iterative calcu-
A computer program such as CulvertMaster can be used to lations to solve for a headwater elevation of 131.44 ft on both
analyze the capacity of the culvert and generate a rating curve. culverts for design flow conditions. This design therefore
For the culvert and channel characteristics described above meets the minimum 1-ft freeboard requirement. The program
and a maximum headwater elevation of 132.55 ft (the road- may also report the exit velocities, which are 9.6 cfs for the
way overtopping elevation), the culvert capacity is found to be box culvert and 9.1 cfs for the 36-in. culvert. Outlet protection
269 cfs. However, allowing for 1 ft of freeboard between the such as a riprap apron (see Section 11.7) should be provided.
maximum headwater elevation and the minimum roadway ele- The engineer is also asked to check the design for the 100-
vation (i.e., the maximum headwater elevation is 131.55 ft), year storm event, which corresponds to a peak discharge of
the culvert capacity is 236 cfs. The rating curve for the culvert 336 cfs. The headwater on the culverts is found to be 132.46 ft
is shown below. for the 100-year storm, which is slightly below the minimum
roadway elevation (i.e., the road will not be overtopped for the
134 100-year check storm).
Headwater Elevation, ft
132
The individual and composite rating curves for the proposed
culvert system are shown below.
130
136
128
134
126 Supplemental
48-in. Culvert
Headwater Elevation (ft)
124 132
0 50 100 150 200 250 300 Existing Culvert
Discharge, cfs
130
Improvements are being made to the roadway, and the ability Both culverts
of the existing culvert to meet current needs must be evalu-
ated. The culvert was originally designed to convey the 50- 128
year peak flow from the NRCS (SCS) Type II storm with 1 ft
of freeboard; however, substantial development has occurred 126
in the area since the original structure was installed.
Assuming that the watershed will be fully developed (subject 124
to zoning regulations) within 5 years, the 50-year design flow 0 50 100 150 200 250 300 350
is found to be 295 cfs. It is apparent that the capacity of the Discharge (cfs)
existing culvert is insufficient to meet the design require-

A culvert is a relatively short conveyance conduit that passes through an obstruction
such as a roadway embankment. Culverts can have a variety of cross-sectional shapes,
including circular, elliptical, arched and rectangular geometries, and may consist of
multiple barrels. Available materials include concrete, corrugated steel, PVC, corru-
gated aluminum, and polyethylene.
Culverts are typically designed to convey the peak flow resulting from a design storm
of a particular recurrence interval (for instance, the peak flow resulting from a 24-
hour, 25-year storm event) with a headwater elevation that does not exceed an allow-
able level. If significant upstream ponding occurs, the detention effects will result in a
reduction of the design storm peak flow that should be considered in the evaluation.
As an alternative to the design-storm approach, culverts may be designed using eco-
nomic analysis. The goal of this approach is to minimize total cost if both flood dam-
ages and the capitalized cost of the culvert are considered.
Culvert performance may also be checked for a storm event of greater magnitude than
the design storm (often, the 100-year event). This evaluation typically involves an
analysis of roadway overtopping in addition to culvert hydrualics.
Various end treatments are used at culvert entrances and exits to provide erosion con-
trol and improve hydraulic efficiency. The culvert barrel may simply project from the
embankment, or it may be mitered to conform to the embankment slope. Precast end
sections are often available from pipe manufacturers. Alternatively, headwalls—with
or without wingwalls—may be used at one or both ends. Finally, specialized inlet
types, such as side-tapered and slope-tapered inlets, may be necessary if certain geo-
metric constraints are present.
Although culvert barrels are typically straight and roughly perpendicular to the road-
way or obstruction they cross, there are many exceptions. A skewed culvert is a cul-
vert that is not perpendicular to its obstruction, typically in order to better align with
the natural direction of the stream. A horizontal alignment may be curved or dog-
legged to avoid obstacles such as rock outcroppings. A broken-back culvert has bends
in its vertical alignment, and may be used to reduce exit velocities, for instance. With
a sag culvert, a portion of the culvert barrel is lower than the culvert entrance and exit.
Such culverts tend to clog, but they may be required in, for example, certain drainage
canal crossings.
Culvert hydraulics are complex, and several flow control types may exist for a single
culvert and discharge. The basic approach to hydraulic evaluation of a culvert is to
analyze its performance for each of the possible flow control types and allow the
assumption that results in the worst performance (that is, the highest headwater) to
govern the design. Two general methods of evaluation exist: (1) the graphical method,
which uses nomographs to find the headwater elevation for a particular set of condi-
tions; and (2) the method of calculating the complete flow profile, as presented in
Section 7.8. The second method is used by most culvert-analysis computer programs.
It is the more computationally intensive of the two approaches, but it is more accurate
in many situations.
Flow in a culvert can be classified as functioning under either inlet control or outlet
control. Under inlet control, the barrel of the culvert is capable of conveying a greater
discharge than the inlet will accept. For an outlet-control situation, the inlet is capable
of accepting more flow than the barrel characteristics or downstream tailwater condi-
tion will permit.
A performance curve graphically shows the relationship between discharge and head-
water elevation. For a culvert, the performance curve is constructed by plotting the
rating curves that would result for both inlet- and outlet-control assumptions. Which-
ever curve has the higher headwater elevation for a given discharge is assumed to gov-
ern the design and is used to define the performance curve. If the discharge is
increased to the point where the corresponding headwater is higher than the culvert
embankment, the resulting flow over the roadway (for example) should be reflected in
the performance curve. For hydraulic analysis purposes, the roadway is typically
assumed to be a broad-crested weir.
Culvert outlet velocities should be evaluated to determine what energy dissipation and
erosion protection measures are necessary, if any. Riprap can be used to both dissipate
energy and provide erosion protection. Another option is to modify the culvert barrel
to reduce the outlet velocity. Other protective measures include impact-type dissipa-
tors and stilling basins.
Other factors that should be considered in culvert design relate to possible environ-
mental impacts. For instance, a culvert may be an obstacle to fish passage. To mini-
mize these impacts, the culvert may be designed to maintain a certain minimum water
level and/or minimum or maximum velocity, and to simulate the natural streambed
environment.
REFERENCES
Manual. Washington, D.C.: AASHTO.
agement Systems. ASCE Manuals and Reports on Engineering Practice, No. 77. New York: ASCE.
Brice, J. C. 1981. Stability of Relocated Stream Channels. Report No. FHWA/RD-80/158. U.S. Geological
Survey.
Circular No. 22, 2d ed., FHWA-SA-96-078. Washington, D.C.: U.S. Department of Transportation.
Chang, F. M., and J. M. Normann. September, 1976. Design Considerations and Calculations for Fishways
Through Box Culverts, Unpublished Report, Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C.
Corry, M. L, P. L. Thompson, F. J. Watts, J. S. Jones, and D. L. Richards. 1983. Hydraulic Design of Energy
Dissipators for Culverts and Channels. Hydraulic Engineering Circular No. 14. Washington, D.C.: Fed-
eral Highway Administration, U.S. Department of Transportation.
Dyhouse, Gary, J. A. Benn, David Ford Consulting, J. Hatchett, and H. Rhee. 2003. Floodplain Modeling
Harrison, L. J., J. L. Morris, J. M. Normann, and F. L. Johnson. 1972. Hydraulic Design of Improved Inlets
for Culverts. Hydraulic Engineering Circular No. 13. Washington, D.C.: Federal Highway Administra-
tion, U.S. Department of Transportation.
Normann, J. M., R. J. Houghtalen, and W. J. Johnston. 2001. Hydraulic Design of Highway Culverts, 2d ed.
Washington, D.C.: Hydraulic Design Series No. 5, Federal Highway Administration (FHWA).
PROBLEMS
Note: Problems 1, 3, 4, 5, and 11 may be solved with the nomographs provided in this book. Problems 2, 8,
and 10 may be solved using the nomographs in HDS 5. Alternatively, the CulvertMaster software provided
with this book can be used.
9.1 Considering only inlet control, determine the flow in a 24-in. concrete pipe culvert with a square-
edge headwall. The headwater elevation is 6 ft. Assume Manning’s n = 0.013.
9.2 Considering only inlet control, what is the headwater elevation required to convey 5.7 m3/s in a
1200-mm corrugated metal pipe with a mitered entrance? Assume Manning’s n = 0.02.
9.3 A 60-in. concrete pipe is to be used as a culvert to convey 400 cfs under a road. Calculate the reduc-
tion in headwater elevation resulting from replacing a square-edge inlet (K = 0.5) at the headwall
with a groove-end inlet (K = 0.02).
9.4 A culvert is to be designed to carry 250 cfs under a roadway. The culvert is 75 ft long, and the
upstream invert elevation is 100 ft. A 60-in., groove-end concrete pipe is to be installed at the natural
streambed slope of 0.01. Assume that there is a free fall at the exit (no tailwater). Determine if the
culvert is operating under inlet or outlet control and report the headwater elevation.
9.5 Repeat Problem 9.4 for a culvert with a tailwater at an elevation of 4 ft above the crown of the down-
stream end.
Problems 367
9.6 A 1200-mm CMP culvert is operating under inlet control. The inlet type is a headwall, and the slope
is 0.008. Determine the flow type and the headwater depth for flows of 1.70 m3/s and 3.4 m3/s.
Assume Manning’s n = 0.02.
9.7 The 50-year flow at a design site is 11.3 m3/s. The maximum allowable headwater elevation for a
proposed culvert is 4.57 m. The culvert will have a 90° headwall with chamfers. Design the smallest
rectangular concrete barrel that will pass the peak flow without exceeding the allowable maximum
headwater. Use n = 0.012. Because of structural considerations, the box culvert must be 1520 mm
high. Other pertinent design data are:
• Stream bed slope = 0.01
• Culvert length = 61 m
• Tailwater elevation = 2.74 m above the culvert outlet invert
9.8 Determine the outlet velocity of a 4-ft–diameter culvert conveying 150 cfs and operating under inlet
control. Use the conservative approximation method.
9.9 A performance chart for a roadway culvert is shown below. The roadway at the culvert location has a
low point of 37.2 m. The vertical alignment of the road is described by the parabolic equation y =
37.2 + 1.52 × 10-3 x2 (x and y are in meters), where y is the crest elevation and x is the distance from
the location of the low point. Assume a 10-m-wide paved road surface. Calculate the total flow
through the culvert and over the roadway when the headwater elevation is 38.1 m.
9.10 Construct a performance curve for a CMP culvert with a diameter of 4 ft and a length of 150 ft. The
pipe entrance is projecting from the fill, and the pipe slope is 0.01. Assume that Manning’s n = 0.024
and that there is free flow away from the exit under all flow conditions. Use a flow range from 0 to
200 cfs.
9.11 Construct a performance curve for a 60-in. circular concrete pipe culvert. The pipe is 100 ft long and
the slope is 0.008. The entrance is a groove end with a headwall, and Manning’s n = 0.013. The tail-
water is 8 ft above the invert of the exit end for all flow conditions. Use a flow range of 0 to 400 cfs.
This combination
inlet in a local
depression captures
runoff in a parking
lot.
C H A P T E R
10
Gutter Flow and Inlet Design
Storm sewer systems, which are most commonly designed and constructed for service
in urbanized areas, are an integral part of the infrastructure. They are typically
designed as conveyance systems for smaller, more-frequent storms in order to prevent
the nuisances and flood damages that would otherwise accompany them. Runoff from
larger, less-frequent events is generally not conveyed entirely by storm sewers; rather,
it flows over the land surface in roadways and in natural and constructed open chan-
nels.
Stormwater enters the subsurface sewer conveyance system through inlets in roadway
gutters, parking lots, depressions, ditches, and other locations. These surface compo-
nents are critical, and they must be properly designed to ensure that the overall system
functions as intended. For example, an engineer may design a pipe system capable of
handling the 10-year storm event, but the level of protection provided will be lower
(often much lower) if the system’s inlet capacities are not sufficient to intercept the
10-year flows. Inadequate attention to roadway drainage and inlets can cause undue
hazards to the traveling public and can compromise the degree of protection afforded
to private and public properties located along the sides of a roadway.
This chapter is the first of two chapters on the design and hydraulic analysis of storm
sewer systems. It focuses on surface drainage—specifically, the physical and hydrau-
lic characteristics of roadway gutters and grate, curb, combination, and slotted-drain
inlets. The various types of gutter sections, including detailed information on the
hydraulic analysis of uniform and composite gutters, are presented in Section 10.1.
Section 10.2 covers the various inlet types and methods for evaluating the hydraulic
performance of grate, curb, combination, and slotted-drain inlets. Determination of
inlet design flows and proper placement of inlets are discussed in Section 10.3.
Chapter 11 continues the presentation on storm sewer systems with information on
the analysis and design of the subsurface portion of the system.
370 Gutter Flow and Inlet Design Chapter 10
10.1 ROADWAY GUTTERS

During design storm events, the depth of open-channel flow in roadway gutters must
be limited to prevent overflowing of curbs and consequent flooding of adjacent prop-
erties. The spread (that is, the top width) of the gutter flow should also be limited to
keep water from extending too far into roadway travel lanes. Excessive spread can
cause safety hazards such as vehicular hydroplaning and limited visibility due to
splash and spray. At locations where the depth and/or spread of the flow into the travel
lane begins to exceed the allowed value, stormwater inlets or catch basins are required
to intercept some or all of the water and convey it into the subsurface storm sewer sys-
tem.
The relationship between the depth of flow in a gutter and its spread depends on the
gutter type and the cross slope of the pavement and/or gutter. As illustrated in Figure
10.1, gutters fall into two basic categories: conventional curb and gutter sections and
shallow swale sections. With conventional curb and gutter, the flow cross-section
extends from the curb toward the roadway centerline. The basic types of curb and gut-
ter illustrated in the figure are uniform, composite, and curved sections. For streets
where curbs are not required, shallow swale sections along the side of the road or in
the median can be used instead. Swale sections include V-shape gutters, V-shape
median swales, and circular swales. In urbanized areas, composite gutters and V-
shape gutters are the most common types.
Figure 10.1
Typical gutter cross
sections
Section 10.1 Roadway Gutters 371
To maintain proper drainage, roadway longitudinal slopes (slopes along the length of
the road) are usually specified to be no smaller than about 0.4 percent. It is difficult to
maintain positive drainage with smaller slopes, resulting in localized puddling. Pave-
ment cross slopes (measured perpendicular to the road direction) necessary to induce
flow off of the pavement surface and into the gutter range from 1.5 percent to as high
as about 4 percent, but are typically specified at around 2 percent (ASCE, 1992).
The allowable spread of open-channel flow in a gutter section is generally a function
of the roadway classification, the design traffic speed, and whether that portion of the
gutter is in a sag location (that is, at the low point of a roadway vertical curve). Design
criteria suggested by the FHWA (Brown, Stein, and Warner, 2001) are presented in
Table 10.1; other criteria may be specified by local jurisdictions.
Table 10.1 Suggested maximum flow spread in roadway sections

Roadway Classification Design Speed/Location Maximum Spread
< 45 mph (70 km/h) Shoulder + 3 ft (0.9 m)
High volume, divided,
> 45 mph (70 km/h) Shoulder
or bidirectional
Sag point Shoulder + 3 ft (0.9 m)
< 45 mph (70 km/h) ½ driving lane
Collector streets > 45 mph (70 km/h) Shoulder
Sag point ½ driving lane
Low traffic ½ driving lane
Local streets High traffic ½ driving lane
Sag point ½ driving lane
Flow in Uniform Gutter Sections

In a uniform gutter section (Figure 10.2), the depth of flow is related to the spread and
the pavement cross slope as
d = TSx (10.1)
Figure 10.2
Uniform gutter cross
section
The cross-sectional area of the flow is
A = Td/2 = T 2Sx /2 (10.2)
and the wetted perimeter is
P = T(1 + Sx) ≈ T (10.3)
where d = depth of flow (ft, m)

T = spread (ft, m)
Sx = pavement cross slope (ft/ft, m/m)
A = cross-sectional flow area (ft2, m2)
Substitution of these expressions into the Manning equation yields
Cf
Q= S x5 / 3T 8 / 3 So1/ 2 (10.4)
n
where Q = flow in gutter section (cfs, m3/s)

Cf = 0.56 for U.S. customary units or 0.376 for SI units
So = gutter longitudinal slope (ft/ft, m/m)
Manning’s n values for use in Equation 10.4 are presented in Table 10.2. The values
in this table may need to be modified to account for obstructions in the gutter, such as
parked cars. Note that Equation 10.4 can be used to determine the discharge if the
spread is specified, and vice versa.
Table 10.2 Manning’s n for streets and gutters (Brown, Stein, and Warner, 2001)
Type of Guttera or Pavement Manning’s n

Concrete gutter, troweled finish 0.012
Asphalt pavement
Smooth texture 0.013
Rough texture 0.016
Concrete gutter with asphalt pavement
Smooth 0.013
Rough 0.015
Concrete pavement
Float finish 0.014
Broom finish 0.016
a. For gutters with small slopes, where sediment may

accumulate, increase above n values by 0.002.
Equation 10.4 is a modified form of the Manning equation presented in Section 6.2
(page 198) and Section 7.2 (page 226). The modification is necessary because the
hydraulic radius does not adequately describe the gutter section, particularly when the
spread may be more that 40 times the depth at the curb face (ASCE, 1992).
Example 10.1 – Computing Gutter Capacity. An asphalt-paved street has a cross slope
Sx = 0.02 and a longitudinal slope So = 0.03. As shown in Figure E10.1.1, the gutter cross section is
uniform with a 6-in.-high concrete curb. The street is in a residential area, and the distance from the
street centerline to the face of the curb is 18 ft. Assuming that the street width allows for a 10-ft-wide
driving lane and 8 ft of parking space on either side of the centerline, estimate the conveyance capac-
ity of each side of the street such that maximum spread criteria are not violated.
Figure E10.1.1 Uniform gutter cross section for Example 10.1

Solution: From Table 10.1, the maximum recommended spread is one-half of a driving lane, or 5 ft in
this instance. Adding to that the 8 ft of parking space along the curb gives an allowable spread of T =
13 ft. Note that the depth of flow at the curb face is
d = TSx = 13(0.02) = 0.26 ft
which is smaller than the height of the curb. Therefore, curb overtopping will not occur, and the
spread is the controlling criterion for the allowable depth of flow.
For Manning’s n = 0.015 (Table 10.2), the conveyance capacity of each side of the street is estimated
from Equation 10.4 to be
0.56
Q= 0.025 / 3138 / 30.030.5 = 8.9 cfs
0.015
Thus, the total capacity of the street (both gutters) is 2(8.9) = 17.8 cfs.
Flow in Composite Gutter Sections

As shown in Figure 10.3, the total discharge in a composite gutter section is parti-
tioned into the discharge within the depressed (gutter) section and the discharge in the
roadway above the depressed section:, as expressed in Equation 10.5.
Q = Qw + Qs (10.5)
where Q = total discharge (cfs, m3/s)

Qw = discharge within the depressed (gutter) section (cfs, m3/s)
Qs = discharge within the roadway above the depressed section (cfs, m3/s)
Figure 10.3
Composite gutter
cross section
If the spread, T, of the flow is also partitioned, it can be written as
T = W + Ts (10.6)
where W = width of the gutter depression (ft, m)

Ts = width of flow in the roadway above depressed section (ft, m)
Note that for this type of gutter section, the actual depth of flow at the curb face is
y = d + a = TSx + a (10.7)
where y = depth of flow at the curb face (ft, m)

d and a are as shown for the composite section in Figure 10.3 (ft, m)
Note further that the dimension a is not the difference in elevation between the lowest
point of the cross section and the top of the gutter depression; rather, it is smaller than
that difference by the amount WSx. Thus, the gutter cross slope is
Sw = Sx + a/W (10.8)
where Sw = gutter depression cross slope (ft/ft, m/m)

To determine the total discharge in a composite gutter section, the discharge Qs can be
determined using Equation 10.4 if the spread T is replaced by Ts, or
Cf
Qs = S x5 / 3Ts8 / 3 So1/ 2 (10.9)
n
If the discharge in the depressed gutter section is expressed as
Q w = E oQ (10.10)
where Eo = fraction of the total discharge that flows in the depressed gutter section
(0 ≤ Eo ≤ 1)
then, substituting for Qw from Equation 10.5 and solving for Q gives
Qs
Q= (10.11)
1 − Eo
The fraction Eo of the total discharge that flows in the depressed gutter section is
given by Brown, Stein, and Warner (2001).
−1
 
 
 Sw / S x 
Eo = 1 +  (10.12)
8/3
  Sw / S x  
 1 + (T / W ) − 1  − 1 
   
Equations 10.5 through 10.12 permit an easy determination of the discharge in a com-
posite gutter section when the spread is specified. Unfortunately, a trial-and-error
solution of these equations is necessary to determine the spread when the discharge is
known. This solution is accomplished by assuming different spread values until the
computed discharge agrees with the given value. A computer program can also be
used to perform this calculation.
It should be noted that when a roadway is rehabilitated and existing pavement is over-
laid with new asphalt, the existing concrete gutter may be paved over as well. In this
way, a composite gutter section may be transformed into a uniform section.
Example 10.2 – Composite Gutter Capacity. Repeat Example 10.1 to find the capacity of
the gutter if the gutter has a composite cross section with a gutter depression width of W = 2 ft. As
shown in Figure E10.2.1, the curb is again 6 in. in height, and the vertical distance from the lowest
point in the cross section to the top of the gutter depression is 1.5 in. (0.125 ft).
Figure E10.2.1 Composite gutter section

Solution: With the substitutions T = 13 ft and W = 2 ft, Equation 10.6 yields Ts = 11 ft.
The gutter cross slope is
Sw = 0.125/2 = 0.0625
The gutter depression, from Equation 10.8, is
a = W(Sw – Sx) = 2(0.0625 – 0.02) = 0.085 ft
The depth of flow at the curb face is
y = TSx + a = 13(0.02) + 0.085 = 0.345 ft
Again, the spread, rather than curb overtopping, is the limiting criterion. Substituting these values into
Equation 10.9 yields the discharge Qs as
0.56 5/3 8/3 1/ 2
Qs = 0.02 11 0.03 = 5.7 cfs
0.015
The fraction Eo of the total discharge that flows in the depressed gutter section is found from Equation
10.12 as
−1
 
 
 0.0625 / 0.02 

Eo = 1+  = 0.43
  8/3 
0.0625 / 0.02 
 1+  −1 
 
  (13 / 2 ) −1  

The total conveyance capacity from Equation 10.11 is
Q = 5.7/(1 – 0.43) = 10 cfs
Considering both sides of the street, the conveyance capacity is 20 cfs.
Section 10.2 Inlets 377
Other Gutter Sections

For gutter-section shapes other than uniform and composite, such as shallow triangu-
lar swale sections, the Manning equation presented in Section 7.2 (page 226; Equa-
tion 7.9) can be applied. The FHWA (Brown, Stein, and Warner, 2001) also provides
guidance for gutter sections not detailed here.
10.2 INLETS
A stormwater inlet is a structure for intercepting stormwater on the ground surface or
in a roadway gutter and conveying it into the storm sewer piping system. It consists of
an inlet opening at the surface and an inlet box, which is a subsurface box, manhole,
or other structure that supports the inlet opening and connects to the subsurface piping
system. Additional information on inlet boxes and manholes is provided in
Section 10.1.
All basic inlet types can be classified further as either continuous-grade inlets (also
referred to as inlets on grade) or sag inlets. A sag inlet is constructed in the low point
of a road or other surface where water drains to the inlet from all directions. With a
continuous-grade inlet, surface water can flow away from the inlet opening in one or
more directions. The cutaway storm sewer system profile in Figure 10.4 shows inlets
in a roadway gutter for both continuous-grade and sag installations.
Figure 10.4
Cutaway view of a
storm sewer system
profile (vertical scale
exaggerated)
Inlet opening geometries vary considerably from one locale to another, but all can be
classified as one of four general types, as illustrated in Figure 10.5. Grate inlets con-
sist of an opening in a gutter or swale, with the opening covered by a grate. Curb-
opening inlets consist of an opening in the face of a curb. Combination inlets have
both grate and curb openings and combine the advantages of each of the individual
inlet types. Finally, slotted drain inlets consist of a grate opening oriented along the
longitudinal axis of a pipe and manufactured integrally with the pipe.
In most cases, the tops of stormwater inlets and inlet openings are constructed of cast-
in-place reinforced concrete, and grates, if present, are cast or fabricated metal. In
many locales, precast concrete inlet structures are also available.
Figure 10.5
Common inlet types
Hydraulic Performance Concepts

From a hydraulic performance standpoint, a particular inlet type may be more suitable
than another for use with a given installation. Grate inlets have been found to perform
well over a wide range of longitudinal gutter slopes. Their capacity decreases as the
gutter slope increases, but this loss of capacity is not as great as it is for curb-opening
inlets. If necessary, a curved vane or tilt-bar grate, which has its bars oriented perpen-
dicular to the direction of flow and tilted in the upstream direction, can be used to
increase the interception capacity of a grate inlet on a steep slope. Grate inlets on
grade are able to intercept more water than combination inlets on grade because the
grate lies in the path of the water flowing in the gutter, but they suffer from a propen-
sity to be easily clogged by trash and debris. They can also present hazards to bicy-
clists and persons wearing heeled footwear, although bicycle-safe grate designs have
been developed and are widely used. Because grate inlets clog easily, they are not
generally recommended for use in sag locations.
Curb-opening inlets are less apt to clog than are grate inlets and are particularly
attractive for use in sags and other locations where the gutter slope is relatively flat.
Because flowing water must change direction to enter a curb-opening inlet, its inter-
ception capacity decreases as the gutter slope and approach velocity increase. For
public-safety reasons, the height of the opening should be no greater than about 6 in.
(15 cm).
Combination inlets, which have both grate and curb openings, offer the advantages of
both inlet types. The curb opening is usually the same length as the grate, but it may
extend beyond the grate in either or both directions. When the curb opening is longer
than the grate, it becomes effective for intercepting trash and debris before it has a
chance to clog the grate opening. For this reason, so-called sweeper inlets with this
configuration are often recommended for use in sags where excessive ponding due to
debris accumulation could pose a safety hazard.
Slotted-drain inlets, because of their longer lengths, can be very effective in intercept-
ing sheet flows at locations such as loading docks and garage entrances. They are also
effective for use in locations, such as parking lot entrances, where they can intercept
flow before it enters the roadway. Slotted drains are very susceptible to clogging and
are therefore not recommended for use in sags or other locations where trash and
debris loadings may be considerable. Slotted drains are also susceptible to crushing
loads because of their shallow burial depths, and thus require great care in pipe bed-
ding and backfilling operations.
Given a specific set of inlet dimensions, the performance of a particular type of inlet
depends on a number of factors. Among them are the depth and velocity of the flow in
the approach gutter(s), the longitudinal and cross slopes of the roadway, and the Man-
ning roughness coefficient for the gutter. The performance also depends significantly
on whether the inlet is installed in a depression, such as the sag point of a vertical
curve, or on a continuous grade. Inlets installed in sag points should be sized to cap-
ture the entire design discharge approaching them. Inlets on continuous grades may
be designed to intercept either all or part of the design flow in the approach gutter.
The efficiency of an inlet is defined as the ratio of the discharge intercepted by the
inlet to the total discharge approaching the inlet:
E = Qint/Q (10.13)
where E = efficiency of the inlet

Qint = discharge intercepted by the inlet (cfs, m3/s)
Q = total discharge approaching the inlet (cfs, m3/s)
Bypass flow, or carryover, is the term given to any discharge not intercepted by an
inlet and is defined as
Qb = Q – Qint (10.14)
where Qb = bypass flow (cfs, m3/s)

The following subsections present procedures for estimating inlet performance char-
acteristics for each of the four inlet types. These procedures are based on research
conducted by the Bureau of Reclamation and Colorado State University on behalf of
the Federal Highway Administration, and are also presented by Brown, Stein, and
Warner (2001). Some locales have developed their own standard inlet designs, and
their design procedures may vary from those described here.
Grate Inlets
The methods for evaluating the capacity of a grate inlet (see Figure 10.6) differ
depending on whether the inlet is located on a continuous grade or in a sag. The sub-
sections that follow present information on determining grate capacity for both condi-
tions.
Capacity of a Grate Inlet on a Continuous Grade. A grate inlet’s
interception capacity on a continuous grade depends on the depth of flow, the size and
configuration of the grate, and the velocity of the approaching flow. The FHWA has
developed a number of standardized grate designs, which are described in Table 10.3
and illustrated in Figures 10.7 through 10.12 (Brown, Stein, and Warner, 2001;
FHWA, 1977). These grates may be used with uniform or composite gutter sections,
or in the bottom of a channel such as a median swale or roadside ditch.
Figure 10.6
Grate inlet
Table 10.3 FHWA standard grate types (Brown, Stein, and Warner, 2001; FHWA, 1977)
Grate type Description Figure Reference
Reticuline “Honeycomb” pattern of lateral and longitudinal bearing bars 10.7
P-50 Parallel bar grate with bar spacing 48 mm (1.875 in.) on center 10.8
Parallel bar grate with bar spacing 48 mm (1.875 in.) on center,
P-50 × 100 and with 10-mm (0.375-in.) diameter lateral rods spaced at 102 10.8
mm (4 in.) on center
P-30 Parallel bar grate with bar spacing 29 mm (1.125 in.) on center 10.9
Curved-vane grate with longitudinal bar spacing 82 mm (3.25
Curved vane in.) on center and transverse bar spacing 114 mm (4.5 in.) on 10.10
center
45° tilt-bar grate with longitudinal bar spacing 57 mm (2.25 in.)
45°–60 tilt bar 10.11
on center and transverse bar spacing 102 mm (4 in.) on center
45°–85 tilt bar 10.11
30°–85 tilt bar 10.12
Figure 10.7
Reticuline grate
Figure 10.8
P-50 and P-50 × 100
grates
Figure 10.9
P-30 grate
Figure 10.10
Curved-vane grate
Figure 10.11
45°–60° tilt bar and
45°–85° tilt-bar grates
Figure 10.12
30°–85° tilt-bar grate
Analysis of the interception capacity of a grate inlet on a continuous grade first

requires the determination of the frontal flow, Qw, which is the discharge in the gutter
over the width W of the grate, and side flow, Qs, which is the discharge in the gutter
above the width of the grate. If the gutter section is uniform, the ratio Eo of frontal
flow to total flow is given by
2.67
 W (10.15)
Eo = 1 −  1 − 
 T 
With a composite gutter section, the grate width is often equal to the width of the gut-
ter depression. In this instance, the definitions of frontal flow and side flow are the
same as those presented in the discussion of composite gutters in Section 10.1. There-
fore, the ratio Eo for the composite gutter when the spread is greater than the grate
width is determined using Equation 10.16:
−1
 
 
 Sw / S x 
Eo =1 +  (10.16)
8/3
  Sw / S x  
  1 + (T / W ) − 1  − 1 
   
where Eo = ratio of frontal flow to total gutter flow (Qw/Q)

Sw = cross slope for depressed portion of gutter
Sx = cross slope for portion of gutter above gutter depression
T = spread (ft, m)
W = width of grate (ft, m)
For a composite gutter section that has a grate width smaller than the width of the gut-
ter depression, Eo is first computed using Equation 10.12, assuming that W is the full
width of the depressed portion of the gutter (that is, in the same way that it was com-
puted in Section 10.1). An adjusted value for the ratio, Eo′ , is then computed using
Equation 10.17. This adjusted value should be used in place of Eo for Equations 10.19
through 10.26.
 A′ 
Eo′ = Eo  w  (10.17)
 Aw 
where Eo′ = adjusted frontal-flow area ratio for grates in composite cross-sections
Aw′ = cross-sectional gutter flow area for the grate width (ft2, m2)
Aw = cross-sectional gutter flow area in the depressed gutter width (ft2, m2)
For a grate inlet in the bottom of a trapezoidal channel, as shown in Figure 10.13, the
Manning equation may be used to solve for the flow depth in the channel (see
Section 7.2, page 226). Eo is then given by
W
Eo = (10.18)
B + dz
where B = bottom width of the trapezoidal channel (ft, m)

d = depth of flow (ft, m)
z = measure of the steepness of the channel sides
Figure 10.13
Cross section of a
trapezoidal channel
with a grate inlet
Once Eo (or Eo′ ) is known, the gutter flow and side flow approaching the grate are
given by the equations
Q w = E oQ (10.19)
and
Qs = (1 – Eo)Q (10.20)
The frontal flow intercepted by the inlet, denoted by (Qw)int, is
(Qw)int = RfQw (10.21)
where
1 − K c (V − Vo ) ; V ≥ Vo
Rf =  (10.22)
 1; V ≤ Vo
Kc = 0.09 for U.S. customary units or 0.295 for SI units

V = Q/A, the velocity of flow in the gutter (ft/s, m/s)
Vo = the velocity at which splash-over occurs (ft/s, m/s)
Note that Rf cannot be greater than one nor less than zero. Vo, the splash-over velocity,
is the minimum velocity at which some of the frontal flow passes over the top of the
grate without being intercepted. This value depends on the grate type and its length in
the direction of flow. Vo (and Rf) can be determined from Figure 10.14.
Figure 10.14
Vo and Rf for common
grate types
The side flow intercepted by a grate inlet on a continuous grade, denoted by (Qs)int, is
(Qs)int = RsQs (10.23)
where
−1
 K V 1.8 
Rs = 1 + s 2.3  (10.24)
 SxL 
Ks = 0.15 for U.S. customary units or 0.0828 for SI units

Sx = roadway cross slope (ft/ft, m/m)
L = grate length (ft, m)
Note that Rs as computed by Equation 10.24 will always be less than one, and hence
some of the side flow will always bypass a grate inlet on a continuous grade regard-
less of the flow velocity or inlet length. This is a deficiency of the equation, but it is
not of great concern for practical purposes.
Combining the above equations gives the total discharge intercepted by a grate inlet
on a continuous grade as
Qint = (Qw)int + (Qs)int = RfQw + RsQs = Q[RfEo + Rs(1- Eo)] (10.25)
The efficiency of the inlet is, therefore,
E = RfEo + Rs(1 − Eo) (10.26)

The orientation of an inlet grate can have a significant influence on the interception
capacity of the structure. Analyzing the capacity of an inlet grate requires that the
relationships used are appropriate for the intended orientation in which the grate will
be installed.
Example 10.3 – Interception Capacity of a Grate Inlet on a Grade. Estimate the

interception capacity of a P-50 inlet grate on a continuous grade if the grate width is W = 2 ft and the
grate length is L = 3 ft. The gutter has the composite cross section from Example 10.2, which is
shown in Figure E10.3.1. Assume that the inlet is placed at a location with the maximum permissible
spread, so that the total discharge approaching the inlet is Q = 10 cfs.
Figure E10.3.1 Composite gutter section

Solution: From Example 10.2, the ratio of frontal flow to the total gutter flow is Eo = 0.43. Therefore,
Qw = 4.3 cfs and Qs = 5.7 cfs. The total cross-sectional area of the flow is
 2(11)(0.02) + 0.125  2 2
A=  2 + 0.5(11) (0.02) = 1.27 ft
 2 
The gutter flow velocity is therefore

V = 10/1.27 = 7.9 ft/s
Figure 10.14 shows that the splash-over velocity for a P-50 grate with a length of 3 ft = 0.9 m is
approximately Vo = 3 m/s = 10 ft/s. Since V ≤ Vo, according to Equation 10.22, Rf = 1 and all of the
frontal flow will be intercepted by the inlet.
The fraction of the side flow that will be intercepted is estimated using Equation 10.24 as
−1
 0.15 ( 7.9 )
1.8

Rs =  1 +  = 0.039
 0.02 ( 3 )
2.3
 
The total flow intercepted by the inlet is, from Equation 10.25,
Qint = 10[1(0.43) + 0.039(0.57)] = 4.5 cfs
and its efficiency is, therefore,
E = 4.5/10 = 0.45
The bypass flow is Qb = 5.5 cfs.
Capacity of a Grate Inlet in a Sag. A grate inlet in a sag location behaves

hydraulically like a weir at small ponding depths and like an orifice at large depths.
The depth at which the transition from weir to orifice behavior occurs depends on the
grate’s bar spacing and overall dimensions. The transition depth increases as the over-
all size of the grate increases.
The length of the perimeter of the grate opening is the critical performance variable
when the inlet behaves as a weir. When the inlet behaves as an orifice, the cross-sec-
tional area of the grate opening is the critical variable. Clogging of a grate can reduce
the value of either of these variables, and thus grate inlets often are not recommended
for use in sag locations.
The capacity of a grate operating as a weir can be estimated as
Q = CwPd3/2 (10.27)
where Q = capacity of the grate operating as a weir (cfs, m3/s)

Cw = weir coefficient, use 3.0 for U.S. customary units or 1.66 for SI units
P = length of the grate’s effective perimeter (ft, m)
d = average depth of flow across the grate (ft, m) (see Figure 10.15)
When one edge of a grate inlet is placed against a curb face, the length of that edge
should be disregarded when computing P.
The capacity of a grate operating as an orifice can be estimated as
Q = 0.67Ag(2gd)1/2 (10.28)
where Q = capacity of the grate operating as an orifice (cfs, m3/s)

Ag = clear opening area of the grate (ft2, m2)
g = acceleration due to gravity (ft/s2, m/s2)
Figure 10.15
Depth d for a grate
inlet in sag
The clear-opening area of a grate used in Equation 10.28 is equal to the total grate
area less the area occupied by the longitudinal and transverse bars when projected
onto a horizontal plane. When the clear opening area is computed in this manner, it
can be increased by about 10 percent for curved-vane and 30° tilt-bar grates (Brown,
Stein, and Warner, 2001). However, curved-vane and tilt-bar grates are not recom-
mended for use in sag locations where there is a chance that a grate inlet would oper-
ate as an orifice.
In an analysis of the capacity of a grate inlet in a sag, a conservative practice is to esti-
mate the capacity using both Equations 10.27 and 10.28 and to then adopt the smaller
of the two estimates. For design problems in which the required inlet dimensions are
being determined, the larger of the sizes obtained by solving the two equations should
be used.
Example 10.4 – Capacity of a Grate Inlet in Sag. Determine the required size of a
grate inlet in a sag if the roadway has a uniform cross slope and dimensions as specified in Example
10.1, as shown again in Figure E10.4.1. Assume that the design discharge for the inlet is Q = 3.5 cfs
and that the width of the grate is W = 2 ft.
Figure E10.4.1 Uniform gutter cross section for Example 10.1

Solution: As shown in Example 10.1, the maximum permissible spread is T = 13 ft. The depth d1 is
d1 = (13 – 2)0.02 = 0.22 ft
and depth d2 at the curb face is
d2 = 13(0.02) = 0.26 ft
Therefore, the average depth is computed as
d = (d1 + d2)/2 = (0.26 + 0.22)/2 = 0.24 ft
Assuming that the inlet behaves hydraulically as a weir, Equation 10.27 yields the required perimeter
of the grate as
Q 3.5 2
P = ------------------ = ------------------------------- = 9.9ft
3⁄2 1⁄2
Cw d 3.0 ( 0.26 )
Because one edge of the grate is adjacent to the curb face and is therefore ineffective, the required
length of the grate is
L = P – 2W = 9.9 – 2(2) = 5.9 ft
If the inlet is assumed to behave hydraulically as an orifice, its required clear-opening area is given by
Equation 10.28:
Q 3.5 2
A g = --------------------------------- = ---------------------------------------------------------- = 1.3 ft
1⁄2 1⁄2
0.67 ( 2gd ) 0.67 [ 2 ( 32.2 ) ( 0.24 ) ]
If a P-50 grate has a length of 5.9 ft and a width of 2 ft, then the clear-opening area is well in excess of
1.3 ft2 (shown in Figure 10.8); therefore, weir behavior controls the design.
To provide a factor of safety against clogging of the grate, one might assume (for example) that its
perimeter should be increased by 25 percent. Applying this assumption gives the required perimeter
as
P = 1.25(9.9) = 12.4 ft
Therefore, the inlet length should be
L = 12.4 – 2(2) = 8.4 ft
Curb-Opening Inlets
Like those of grate inlets, the methods used to compute the capacity of a curb-opening
inlet (see Figure 10.16) are different if the inlet is on a continuous grade than if it is in
a sag location. The subsections that follow present methods for evaluating both condi-
tions.
Figure 10.16
Curb-opening inlet
Capacity of a Curb-Opening Inlet on a Continuous Grade. The

depth of flow at the curb face and the length of the inlet opening are the primary vari-
ables governing the interception capacity of a curb-opening inlet on a continuous
grade. If the gutter cross-slope is increased in the vicinity of the inlet so as to corre-
spondingly increase the depth of flow at the curb face, then the capacity of the inlet
can be greatly improved.
When curb-opening inlets must be provided with top-slab supports for structural rea-
sons (see Figure 10.17), the supports should be rounded and set back several inches
from the curb face. If the supports are placed flush with the curb face, tests have
shown that they can reduce the interception capacity of the downstream opening by as
much as 50 percent. The capacity of the downstream opening may be reduced to near
zero if trash and debris are collected by a structural support.
Figure 10.17
Curb-opening inlet
with supports
As noted earlier in Section 10.2, the height of the opening in a curb-opening inlet
should be no more than about 6 in. (150 mm). The length of the opening required for
total interception of the gutter flow when the gutter has a uniform cross-section is
0.6
 1 
LT = C f Q 0.42 So0.3   (10.29)
 nS x 
where LT = length of curb opening necessary for interception of the total gutter dis-
charge (ft, m)
Q = gutter discharge (cfs, m3/s)
So = longitudinal slope of the gutter (ft/ft, m/m)
n = Manning roughness coefficient for the gutter
Sx = cross slope (ft/ft, m/m)
Cf = 0.6 for U.S. customary units or 0.817 for SI units
The efficiency of a curb-opening inlet of length L less than LT is
1.8
 L 
E = 1 − 1 −  (10.30)
 LT 
where E = efficiency of the curb-opening inlet

L = length of the curb opening (ft, m)
If the gutter cross slope is increased in the vicinity of a curb-opening inlet to create a
gutter depression (see Figure 10.18), either along the entire gutter or locally at the
inlet, then the interception capacity of the inlet improves significantly. In such cases,
the length required for total interception can be computed using Equation 10.29 if an
equivalent cross slope is used in place of Sx.
The equivalent cross slope is
Se = S x + S wp Eo (10.31)
where Se = equivalent cross slope (ft/ft, m/m)

Sx = pavement cross slope (ft/ft, m/m)
Swp = a/W, the slope of the depression relative to the pavement cross slope Sx
(see Figure 10.18)
Eo = fraction of the total gutter flow within the gutter width W
The value of Eo can be determined using Equation 10.12.
Equation 10.30 can again be used to determine the inlet efficiency if its actual length
is shorter than the length required for total interception.
Figure 10.18
Depressed curb-
opening inlet
Example 10.5 – Determining the Capacity of a Curb-Opening Inlet on a

Grade. Determine the required length of a curb-opening inlet on a continuous grade (So = 0.01) if it
is to intercept a total approaching flow of Q = 0.14 m3/s. Assume that the gutter is depressed in the
inlet area so that the opening height is 13 cm and the top slab is 8 cm thick. (Thus, the total curb
height at the inlet opening is 21 cm.) The gutter width is W = 0.6 m, the depression dimension is a =
0.08 m (Figure 10.18), and the pavement cross slope is Sx = 0.02. Assume that the ratio Eo = 0.40 and
Manning’s n = 0.015.
Solution: The equivalent cross slope of the roadway is determined from Equation 10.31 as
Se = 0.02 + (0.08/0.6)0.40 = 0.07
Substitution of this result into Equation 10.29 yields the required inlet length as
0.6
 1 
LT = 0.817(0.14)0.42 (0.01)0.3   = 5.5 m
 0.015(0.07) 
As a factor of safety, this inlet length should be increased by an assumed amount to allow for
clogging.
Capacity of a Curb-Opening Inlet in a Sag. A curb-opening inlet in a

sag behaves hydraulically as a weir when the depth of water ponding at the curb face
is less than or equal to the height h of the curb opening. For ponding depths greater
than 1.4h, the inlet opening behaves as an orifice. For intermediate depths, a transition
between weir and orifice behavior occurs. In inlet capacity analysis, a conservative
approach is to estimate the capacity assuming both weir and orifice behaviors, and to
then adopt the smaller of the two capacities. Similarly, a conservative approach to
design is to adopt the larger of the two required inlet sizes.
The weir equation for a curb-opening inlet without a gutter depression yields an esti-
mate of the inlet capacity as
Q = CwLd3/2 (10.32)
where Q = inlet capacity (cfs, m3/s)

Cw = weir coefficient equal to 3.0 for U.S. customary units or 1.66 for SI units
If the gutter in the vicinity of a curb-opening inlet in a sag is depressed to increase the
depth at the curb face, then the location of the weir behavior moves from the curb face
to the lip of the gutter and its length is effectively increased. In such cases, the weir
equation for the inlet capacity takes the form
Q = Cw(L + 1.8W)d3/2 (10.33)
where W = width of the gutter (ft, m)

d = depth at curb measured from the normal cross slope (ft, m)
In this expression, the depth d at the curb face is measured relative to the normal road-
way cross slope (that is, d = TSx).
For curb-opening inlets with lengths in excess of 12 ft (3.6 m), Equation 10.32 pro-
duces a higher estimate of the capacity than does Equation 10.33. This is a deficiency
of the equations, as an inlet with a depressed gutter will perform at least as well as one
without the depression. It is therefore recommended that Equation 10.32 be used in
lieu of Equation 10.33 for inlets with lengths greater than 12 ft (3.6 m), regardless of
whether there is a gutter depression. In this case, the value of d in Equation 10.32 is
equal to the total depth at the curb face, including the depression.
When a curb-opening inlet behaves as an orifice, the effective depth of flow at the curb
face, do, is considered in the capacity calculation. The determination of do depends on
the type of throat opening at the inlet. Horizontal, inclined, and vertical throat config-
urations, and the method used to determine do for each, are illustrated in Figure
10.19. Note that di is the depth at the lip of the curb opening.
The capacity of a curb-opening inlet behaving as an orifice can be estimated as
Q = 0.67hL(2gdo)1/2 (10.34)
where h = curb opening height (ft, m)

do = effective depth of flow at the curb face (see Figure 10.19)
Figure 10.19
Curb-opening inlet
throat types
Example 10.6 – Capacity of a Curb Inlet in a Sag. A roadway has a composite gutter
section with a pavement cross slope Sx = 0.02, a normal gutter cross slope Sw = 0.0625, and a curb-
face height of 15 cm. The gutter width is W = 0.61 m. Thus, the top of the curb is normally 11 cm
higher than the elevation of the lip of the gutter. A curb-opening inlet with a horizontal throat is to be
designed at a sag point in the roadway with a design discharge of Q = 0.3 m3/s. The gutter is
depressed at the inlet location so that an opening height of 12 cm and a top slab thickness of 8 cm
yield a total curb height of 20 cm. The allowable spread at the inlet is T = 4 m. Determine the required
curb-opening length.
Solution: Because the gutter is depressed at the inlet, Equation 10.33 can be used to estimate the
required inlet length under the assumption that it behaves as a weir. Setting
d = TSx = 4(0.02) = 0.08 m
gives the required length as
Q 0.3
L= − 1.8W = − 1.8(0.61) = 6.89 m
Cw d 3/ 2 1.66(0.08)3 / 2
Because this length is greater than 3.6 m, Equation 10.32 should be applied instead. The depth d used
in Equation 10.32 is the depth at the curb face. This value is equal to the depth at the curb face at a
normal gutter cross-section, plus an additional 5 cm for the depressed gutter. [The 5-cm depression is
computed as the difference between the normal curb height (15 cm) and the height at the inlet (20
cm).] At a normal gutter section, the curb-face depth is
y = (T – W)Sx + WSW = (4 – 0.61)(0.02) + 0.61(0.0625) = 0.11 m
and thus the depth for use in Equation 10.32 is
d = 0.11 + 0.05 = 0.16 m

Therefore, if weir behavior controls the design, the required inlet length is
Q 0.3
L= = = 2.82 m
Cwd 3/ 2 1.66(0.16)3/ 2
If orifice behavior controls the design, then the required inlet length is determined using Equation
10.34 with h = 13 cm = 0.13 m, which gives
do = 0.16 – (0.12/2) = 0.10 m
Substituting this result into Equation 10.34 and solving for L gives
Q 0.3
L= = = 2.66 m
0.67 h(2 gd o )1/ 2 0.67(0.12)[2(9.81)(0.10)]1/ 2
Comparing the lengths shows that weir behavior controls the design, and the required length is L =
2.58 m. This length should be increased by an assumed amount in order to account for clogging of the
inlet opening.
Combination Inlets
As previously discussed, a combination inlet is a configuration in which a grate inlet
and curb inlet are used together (see Figure 10.20). Like both calculations for grate
and combination inlets, the method used to compute combination inlet capacity
depends on whether the inlet is on grade or in a sag.
Figure 10.20
Combination inlet
Capacity of a Combination Inlet on a Continuous Grade. When the

length of the curb opening is the same as the length of the grate for a combination
inlet on a continuous grade, the capacity of the inlet is assumed to be no greater than
the capacity of the grate alone. Thus, the interception capacity of the inlet should be
determined using the procedure for grate inlets on grade.
If a combination inlet on a continuous grade has a sweeper configuration, which
means that the curb opening extends upstream of the grate, then the interception
capacity is the sum of the capacity of the curb opening upstream of the grate and the
grate capacity. When performing the calculations, the frontal flow Qw that reaches the
grate should be reduced by the amount that is intercepted by the curb opening.
Capacity of a Combination Inlet in a Sag. The procedure for determin-

ing the capacity of a combination inlet in a sag depends on the length of the curb
opening relative to the length of the grate. If the inlet has a sweeper configuration in
which the curb opening is longer than the grate, then the curb opening can be very
effective in capturing trash and debris before it has a chance to clog the grate.
Under weir-flow conditions, the capacity of a combination inlet with equal grate- and
curb-opening lengths is essentially that of the grate alone. However, if the inlet has a
sweeper configuration, it is reasonable to assume that the inlet’s capacity is equal to
the sum of the weir capacities of the grate and the portion of the curb opening that
extends beyond the grate. When the capacity of a sweeper inlet is computed in this
manner, it is suggested that the weir length applicable to the grate be computed as
P=L+W (10.35)
where P = applicable weir length for the grate (ft, m)

L = length of the grate (ft, m)
W = width of the grate (ft, m)
Note that the edge of the grate adjacent to the extended curb opening is neglected in
this determination of P, as is the edge of the grate adjacent to the curb face.
Under orifice-flow conditions, the capacity of a combination inlet in a sag is the sum
of the capacities of the grate and the curb opening. The capacity of the grate is given
by Equation 10.28 and the capacity of the curb opening by Equation 10.34.
Slotted-Drain Inlets
Slotted-drain inlets may be installed either along the length of the gutter, as shown in
Figure 10.21, or perpendicular to the flow direction to intercept sheet flows. The cal-
culation of hydraulic capacity for both installation types is presented in this section.
Capacity of a Slotted-Drain Inlet on a Continuous Grade. The inter-
ception capacity of a slotted-drain inlet on a continuous grade depends significantly
on the orientation of the longitudinal axis of the slotted drain with respect to the sur-
face-flow direction. When slotted drains are installed in the normal fashion (along the
length of the gutter), as illustrated in Figure 10.21, and when the slot width is at least
1.75 in. (45 mm), FHWA tests show that the required length of a slotted drain for total
interception of the gutter flow is given by Equation 10.29. In other words, the slotted
drain performs similarly to a curb-opening inlet in this instance. If the actual length of
the slotted drain is shorter than the length required for total interception, then its effi-
ciency is given by Equation 10.30.
If a slotted drain is installed in a configuration for sheet-flow interception, its capacity
can be approximated by treating it as a grate inlet with a grate length equal to the
width of the slotted opening. Thus, its capacity can be approximated using Equations
10.21 and 10.22, where the frontal flow is the total sheet-flow discharge approaching
the slotted drain opening. Use of this approximation requires some judgment about
the appropriate value of the splash-over velocity. Although no specific recommenda-
tion is made here, intuition suggests that because of the generally shallow nature of
sheet flow, the factor Rf can often be taken as unity.
Figure 10.21
Slotted-drain inlet
Credit: McCuen, 1998. Reprinted by permission of Pearson Education, Inc., Upper Saddle River, NJ.
Capacity of a Slotted-Drain Inlet in a Sag. When installed in a sag con-

dition, a slotted drain will perform as a weir for flow depths up to about 0.2 ft (0.06
m), although this depth depends on the slot width. For such operating conditions, the
capacity of the slotted drain can be estimated as
Q = CwLd3/2 (10.36)
where Cw = weir coefficient

L = length of the slot (ft, m)
d = depth at curb measured relative to the normal cross slope (i.e. d = TSx)
(ft, m)
The weir coefficient Cw in this expression depends on the flow depth and the slot
length, but a typical value is 2.48 for U.S. customary units or 1.4 for SI units.
Equation 10.36 is based on the assumption that the slotted drain is installed next to a
curb face, so that only one edge of the slot behaves as a weir. The capacity of a slotted
drain when both edges of the slot serve as weirs should be twice that predicted by
Equation 10.36, though it is expected that the weir coefficient would require modifi-
cation in such cases. Manufacturers of slotted-drain products can typically provide
guidance on the required length of drain for various types of installations.
For flow depths greater than about 0.4 ft (0.12 m), a slotted drain behaves hydrauli-
cally as an orifice, with its capacity given by
Q = 0.8LW(2gd)1/2 (10.37)
where W = slot width (ft, m)

d = depth of water at the slot opening (ft, m)
10.3 INLET LOCATIONS AND DESIGN FLOWS

The surface portion of a stormwater conveyance system cannot be viewed as simply a
collection of individual inlets. Rather, it is a network of interdependent inlets con-
nected by gutters and other conveyances. The gutter downstream of a particular inlet
on grade must have sufficient capacity to carry both that inlet’s bypass flow and any
additional flow entering the gutter. The next downstream inlet on grade will intercept
only a portion of the total gutter flow—again, some flow is carried over to the next
downstream inlet. This interdependency continues in a downstream direction until the
flow reaches a sag inlet, which captures 100 percent of the inflow.
Determining the locations where inlets are required and their corresponding design
discharges is at least in part a trial-and-error process. Nevertheless, required locations
of many inlets can be determined simply by referring to plans and profiles for the
roadway. According to Brown, Stein, and Warner (2001), inlets should be provided at
the following locations, at a minimum:
• At all low points (sag points) in gutters
• Immediately up-grade of median breaks, entrance/exit ramp gores, and street
intersections (that is, at any location where water could flow onto the trav-
eled surface of a roadway)
• Immediately up-grade of bridges, to prevent drainage onto bridge decks, and
immediately down-grade of bridges, to intercept bridge-deck runoff [see
Young, Walker, and Chang (1993) for information on drainage of bridge
decks]
• Immediately up-grade of roadway cross-slope reversals
• Immediately up-grade of pedestrian cross-walks
• Behind curbs, shoulders, or sidewalks as necessary to drain low areas
Section 10.3 Inlet Locations and Design Flows 401
Inlet locations based on the above criteria should be plotted on a plan view of the
roadway system. Topographic information and roadway profile drawings can then be
used to delineate the boundaries of the drainage basin contributing runoff to each
inlet. Estimates of design flows to each inlet are then computed for a selected design-
storm (see Table 2.2, page 37). The design flow in the approach gutter to each inlet
can be compared with the allowable gutter conveyance capacity to check that flow
depth and spread criteria are not violated. In all cases, one should strive to place inlets
using projections of land-parcel or lot lines so the inlets will not interfere with poten-
tial driveway locations. In sag vertical curves, such placement often requires adjust-
ment of roadway grades and/or vertical curve lengths.
Provided that the gutter capacity is not exceeded at an inlet, the selection of an inlet
type and the inlet design can proceed as outlined in the previous subsections. Designs
of sag inlets should be such that the entire design flow approaching the inlet is inter-
cepted. Conversely, inlets on continuous grades may be designed to intercept only a
fraction of the approaching flow, leaving the bypass flow to proceed to the next down-
stream inlet.
For any inlet, the total design flow consists of the runoff from its local drainage basin
plus any bypass flow from upstream inlets. However, if the bypass flow is large, or if
the times of concentration are significantly different for the bypass flow and the local
runoff to an inlet, then the total design discharge may be computed using the proce-
dure outlined by Guo (1997).
In the event that the discharge approaching an inlet is in excess of the conveyance
capacity of the gutter section, a trial-and-error approach is required to determine loca-
tions for additional inlets. These inlets are virtually always continuous grade, but are
usually necessary only on extended lengths of roadway where inlets are not already
required on the basis of the geometric criteria previously listed.
The trial-and-error approach entails selection of a trial inlet location. Based on the
trial location, drainage basins are again delineated and design discharges computed. A
comparison of the design discharges with allowable gutter capacities is then made to
check the adequacy of the selected inlet locations. Where possible, inlet positions
determined in this manner should coincide with locations where projections of lot or
parcel lines intersect the gutter. As noted above, this practice avoids potential con-
flicts with driveways or access roads.
In roadway sag locations, especially where there is no outlet for stormwater other than
through the storm sewer system, flanking inlets should be provided on either side of
the sag inlets. These inlets provide relief if the sag inlet becomes clogged. This prac-
tice should not be necessary in newly developed areas if proper attention has been
given to the major storm-drainage system, but it may be required in retrofit or rehabil-
itation projects. Flanking inlets should be designed to intercept all of the design flow
without violating allowable spread criteria if the sag inlets are completely clogged.
Modeling Focus: Location of

Inlets on Grade
Computer programs can be useful tools when deter- decides to locate inlet B so that its drainage area is
mining where inlets should be placed to meet design roughly half that of the original area. The flow
constraints such as allowable gutter spread and depth. approaching inlet B is calculated as 2.1 cfs.
A number of programs have been written to perform Because 2.1 cfs is less than 3.0 cfs (the maximum
these types of calculations, including the FlowMaster
allowable gutter capacity), the location of inlet B
software packaged with this book. Like the methods
appears acceptable from the standpoint of upstream
presented in this chapter, FlowMaster’s computational gutter capacity. The computer program or hand calcu-
methods are based on FHWA’s HEC-22 publication
lations can now be used to evaluate the performance
(Brown, Stein, and Warner, 2001).
of inlet B. It is determined that inlet B intercepts
Consider that an engineer is designing a subdivision 1.4 cfs and bypasses 0.7 cfs to inlet A. (The gutter
and plans to use curved-vane grate inlets measuring spread upstream of inlet B is found to be 7.0 ft.)
2 ft long by 2 ft wide as the standard inlet size. He or Summing the bypass flow from inlet B and the flow
she begins by proposing inlets at the minimum num- from inlet A’s additional contributing drainage area,
ber of locations using the guidelines provided in
the engineer determines that the flow approaching
Section 10.3. The next step is to evaluate the perfor-
inlet A will be 0.7 cfs + (4.0 cfs − 2.1 cfs) = 2.6 cfs.
mance of the gutters and inlets, given these inlet loca- Because this flow is less than 3.0 cfs, the gutter spread
tions.
constraint is met. Inlet A is computed to intercept 1.6
The first inlet the engineer examines, inlet A, is the cfs and bypass 1.0 cfs. (The gutter spread immedi-
most-upstream inlet on a section of roadway. The ately upstream of the inlet is 7.6 ft.)
engineer delineates the contributing drainage area for
this inlet and calculates that the peak runoff flow rate
to the inlet for the design storm is 4.0 cfs.
This section of roadway has a 3-percent longitudinal
slope and a 2-percent cross slope. A composite gutter
section with a a 2-ft–wide depressed gutter having a
cross slope of 3-percent will be used on this project.
According to local regulations, the maximum amount
of gutter spread permitted for this roadway is 8 ft.
The engineer uses a computer program or hand calcu-
lations to determine whether the gutter can handle a
flow of 4.0 cfs without violating the maximum-spread
constraint. Assuming a spread of 8 ft and solving for
flow, the engineer determines that the maximum
capacity of this gutter is 3.0 cfs. The design of the remaining inlets in the subdivision
proceeds in a similar fashion, working downstream
Based on this calculation, the engineer knows that an
from the most upstream inlets, transferring bypass
additional inlet must intercept, at a minimum, (4.0 cfs
flows downstream, checking gutter spread, and add-
− 3.0 cfs) = 1.0 cfs.
ing or enlarging inlets as needed. Note that in this
simple example, the peak bypass flow and the peak
additional flow from the drainage area are assumed to
be additive. If the times of concentration for these
flows differed significantly, this would not be the
Anticipating that some of the flow from the additional case.
inlet (inlet B) will be bypassed to inlet A, the engineer

Inlets provide a means for stormwater to enter the subsurface sewer system. Most
inlets are located in roadway, driveway, or parking lot gutters. These gutters collect
runoff and convey it to the inlet. The two basic types of gutters are the conventional
curb and gutter section (including uniform, composite, and curved gutters) and shal-
low swale sections (including V-shape, V-shape median, and circular gutters).
Gutter-capacity calculations are performed to ensure that allowable gutter spread and
depth criteria are not violated for the system design. These calculations are typically
performed using the modified versions of the Manning equation presented in this
chapter.
Common inlet types are grate, curb-opening, combination, and slotted-drain. These
inlets are found on both continuous grades and in sag locations. For an inlet on a con-
tinuous grade, the inlet may only intercept a portion of the flow and bypasses remain-
ing flow to the next downstream inlet.
Common grate inlet types include parallel bar, vaned, tilt-bar, and reticuline of vari-
ous shapes and sizes. For a grate inlet on grade, the flow captured by the inlet is com-
puted as the sum of the intercepted frontal flow and side flow. The frontal flow
intercepted is a function of the cross-sectional geometry of the gutter, the splash-over
velocity for the inlet (the velocity at which frontal flow begins to bypass the inlet),
and the flow velocity. The side flow intercepted is a function of flow velocity, road-
way cross slope, and grate length. A grate inlet in a sag functions hydraulically as a
weir for shallow flows and as an orifice for deeper flows.
The determination of capture efficiency for a curb inlet on grade is based on the inlet
length required to capture 100 percent of the flow. Inlet performance can be signifi-
cantly increased by increasing the gutter cross slope. Similar to a grate inlet, a curb
inlet in sag operates as a weir for shallow flows and as an orifice for deeper flows. The
orifice behavior of a curb-opening inlet is influenced by the type of throat opening
(horizontal, inclined, or vertical).
Because a combination inlet consists of a grate inlet and curb inlet together, the calcu-
lation methods presented for both of these inlet types apply. However, for a combina-
tion inlet on a continuous grade or for a combination inlet in sag behaving as a weir,
the inlet capacity is assumed to be the same as that of the grate acting alone if the curb
opening and the grate are the same length. If the curb opening is longer than the grate,
the capacity for the length of curb opening extending beyond the grate should be con-
sidered. For a combination inlet in sag operating as an orifice, the capacity is com-
puted as the sum of the grate and curb opening capacities.
The capacity of a slotted-drain inlet on a continuous grade depends on the orientation
of the drain relative to the direction of surface flow. When a slotted drain is installed
along the length of a gutter, the slotted drain performs similarly to a curb-opening
inlet on a continuous grade. If the direction of the drain is normal to the flow direc-
tion, it is treated as a grate inlet in capacity calculations. If the slotted-drain inlet is
installed in a sag, it is analyzed as a weir or an orifice, depending on the flow depth.
Various geometric considerations on a site control the initial placement of inlets. For
instance, inlets should be placed in low points and immediately upstream of
intersections. Inlets are then added or resized as necessary to meet gutter-flow con-
straints. A computer program that performs gutter and inlet calculations can be of
great assistance in this trial-and-error process.
REFERENCES
Circular No. 22, Second Edition, FHWA-SA-96-078. Washington, D.C.: U.S. Department of Transpor-
tation.
Federal Highway Administration (FHWA). 1977. Bicycle Safe Grate Inlets Study: Vol. 1 and 2—Hydraulic
and Safety Characteristics of Selected Grate Inlets on Continuous Grade. FHWA-RD-77-44. Washing-
ton, D.C.: U.S. Department of Transportation.
Guo, J. C. Y. 1997. Street Hydraulics and Inlet Sizing Using the Computer Model UDINLET. Highlands
Ranch, Colorado: Water Resources Publications.
McCuen. 1998. Hydrologic Analysis and Design. Upper Saddle River, New Jersey: Pearson Education.
Young, G. K., S. E. Walker, and F. Chang. 1993. Design of Basic Bridge Deck Drainage. Hydraulic Engi-
neering Circular No. 21, GHWA-SA-92-010. Washington, D.C.: U.S. Department of Transportation.
PROBLEMS
10.1 A roadway gutter has a longitudinal slope of 0.01 and a cross slope of 0.02. Using a Manning’s n of
0.016, find:
a) the spread at a flow of 1.8 cfs
b) the gutter flow at a spread of 8.2 ft
10.2 A composite curb and gutter section is located on a roadway with a longitudinal slope of 0.01 and a
cross slope of 0.02. The gutter width is 2 ft, and the depression is 2 in. Assume Manning’s n = 0.016.
a) What is the gutter flow if the spread is 8.2 feet?
b) What is the spread if the flow is 4.2 cfs?
10.3 Find the interception capacity of a 610-mm–wide by 914-mm–long P-50 grate located in a compos-
ite gutter section having a longitudinal slope of 0.01 and a roadway cross-slope of 0.02. The width W
of the depressed portion of the gutter is 0.6 m, the depth of the depression a is 50 mm, and the gutter
cross slope is 0.103. The flow in the gutter is 0.065 m3/s, and Manning’s n is 0.016.
10.4 A uniform roadway gutter has a grated inlet in a depressed sag configuration. The flow is 0.20 m3/s,
and the depth of flow at the curb is 122 mm. Determine the length of 610-mm–wide, P-50-type grate
needed to intercept the flow in a sag configuration.
Problems 405
10.5 A gutter has the following characteristics:

• S0 = 0.01
• Sx = 0.03
• Q = 2 cfs
• n = 0.016
Find Qint for a 10-ft–long curb opening.
10.6 A 3.05-m curb-opening inlet is located on a grade with a slope of 0.01 and n = 0.012. The gutter sec-
tion is depressed, and its dimensions are:
• Sx = 0.03
• Swp = 0.05
• a = 51 mm
• W = 0.91 m
Determine the flow intercepted by the inlet if its efficiency if 50 percent. (Hint: use the included
FlowMaster software to graph efficiency versus discharge.)
Workers install a
section of concrete
pipe for a storm
sewer.
C H A P T E R
11
Storm Sewer Pipe System and
Outlet Design
This chapter is the second of two chapters related to the design and analysis of storm
sewer systems. Chapter 10 focused on the gutter and inlet components that make up
the surface portion of a system, whereas this chapter covers the subsurface portion of
the system.
Storm sewers are typically dendritic (tree-like) subsurface systems that collect and
direct flows from inlets into larger and larger conveyance structures (typically pipes)
or systems as one moves downstream in the watershed. The downstream ends of
storm sewers, called outfalls, are generally located at treatment and/or detention or
retention facilities, receiving waters, or at locations where alternate conveyance facil-
ities, such as open channels, are more practical.
Chapter 2 briefly introduced many of the basic physical components and materials
found in storm sewer systems. Section 11.1 of this chapter expands on that discussion
by providing more detailed information on storm sewer pipe materials and sizes, and
manhole construction.
The horizontal and vertical alignments of storm sewers are governed by a number of
factors, as will be discussed in Section 11.2. These factors include topography, stan-
dardization of horizontal locations within roadway rights of way, subterranean condi-
tions, such as the presence of bedrock and other utilities, and minimum and maximum
flow velocities. Storm sewer alignments, both horizontal and vertical, are often shown
in construction documents on “plan and profile” sheets such as the one shown in Fig-
ure 11.1
Estimation of design flows in storm sewers and determination of the pipe sizes
required for conveyance of these flows are usually accomplished using the rational
method and the Manning equation. This procedure was introduced in Example 5.8
and is elaborated on in Section 11.3.
408 Storm Sewer Pipe System and Outlet Design Chapter 11
Figure 11.1
Plan and profile sheet
for storm sewer pipe
runs
In many cases, plotting the hydraulic and energy grade lines for design flows on a
profile of a storm sewer is desirable in order to ensure that the design is adequate.
This verification is particularly important when one or more parts of the system are
surcharged. To compute the hydraulic and energy grades, it is necessary to evaluate
not only the hydraulic profile and friction losses in the pipes, but also minor losses in
inlet boxes, manholes, and junction structures. The classical approach to computing
minor losses through various appurtenances was presented in Section 6.3, and this
method can be applied to sewer systems. However, other techniques have been devel-
oped specifically for common sewer conditions and configurations. Two of these
methods, the energy-loss method and composite energy-loss method, are described in
Section 11.4.
Section 11.5 covers final design, and presents procedures for computing and plotting
the hydraulic and energy grade lines through a storm sewer system for both sur-
charged and open-channel flow conditions. The latter part of this section provides dis-
cussions on additional design considerations, including structural loads on sewer
pipes and flow routing of storm sewers. Because of the number of calculations
required to design or analyze storm sewer hydraulics, computer programs are typi-
cally used for this purpose. Section 11.6 outlines a typical procedure for designing a
storm sewer system with the aid of a computer.
The energy associated with an excessive flow velocity at a storm sewer outfall point
should be dissipated to prevent potentially destructive and unsightly erosion from
occurring at that location. As described in Section 11.7, erosion protection may con-
sist of a simple riprap blanket, a concrete energy dissipator, or a stilling basin.
Section 11.1 Storm Sewer Types, Components, and Materials 409
11.1 STORM SEWER TYPES, COMPONENTS, AND

MATERIALS
This section provides information on the components and materials commonly used
in the construction of storm sewer systems. Most of these components were first pre-
sented in Chapter 2, but are they are covered in more detail here.
It is important to note that storm sewers, as considered in this book, are separate from
systems intended to convey sanitary wastes. Combined sewers, which are not
addressed in this book, convey both storm runoff and sanitary flows. The choice of
cross-sectional shape and material for a sewer conduit can be influenced by whether
the system is combined or separate. In a combined sewer, for example, it may be
desirable to provide a cunette (smaller channel) in the sewer invert to handle low
flows. Also, materials such as concrete or steel may be undesirable in combined sew-
ers because of the potential for corrosion caused by hydrogen sulfide gas released
from raw sewage.
The subsections that follow first present information on pipe materials frequently
used in storm sewer system design and then describe common appurtenances such as
inlets, catch basins, manholes, junctions, and outlets.
Storm Sewer Pipes

A variety of pipe materials is used in the construction of storm sewers. Concrete and
steel are by far the most popular and are discussed in detail in this section. Another
material that may be used as an alternative to corrugated steel is corrugated alumi-
num, which is resistant to corrosion. Because of its lower stiffness, however, alumi-
num pipe requires more care during construction. Other pipe materials include high-
density polyethylene (HDPE) and polyvinyl chloride (PVC), as well as various forms
of cast-in-place concrete sections.
Concrete Pipe. Concrete is the most-used pipe material for storm sewer con-
struction. It is available both with and without reinforcing steel and comes in a range
of wall thicknesses and strength classes. In addition to pipe, precast concrete man-
holes, inlets, outlet sections, and special structures, such as tees and bends, are avail-
able. In the United States, these materials are manufactured in accordance with
applicable standard specifications of the American Society for Testing and Materials
(ASTM). The American Concrete Pipe Association (ACPA) publishes considerable
technical data, as well as software that can be downloaded from the Internet to assist
in specifying appropriate concrete-pipe products. The NSSN web site (www.nssn.org)
provides links to standards information from more than 600 international, national,
and regional standards organizations.
As illustrated in Figure 11.2, concrete pipe is manufactured in a variety of cross-
sectional shapes: circular, elliptical, arch, and rectangular (box). Plain or nonrein-
forced circular concrete pipe is available in inside diameters ranging from 4 to 36 in.
(100 to 900 mm). Typically, plain concrete pipes are available in 2-in. (50-mm) incre-
ments for diameters of 4 to 12 in. (100 to 300 mm), and in 3-in. (75-mm) increments
for diameters of 12 to 36 in. (300 to 900 mm). Reinforced-concrete pipe is more com-
monly used and is available in diameters ranging from 12 to 180 in. (300 to
4500 mm). Generally, the minimum recommended diameter for storm sewer applica-
tions is 12 in., due to the potential for clogging by debris and for maintenance require-
ments, but some jurisdictions may specify a larger diameter as the minimum
acceptable size.
Table 11.1 provides a listing of the commonly available sizes of circular reinforced-
concrete pipe and corresponding available pipe-wall thickness classifications A, B,
and C. The thinnest wall is designated as A and the thickest wall is designated as C.
Pipe manufacturers typically provide additional information on the permissible loads
for the various classfications.
Figure 11.2
Cross-sectional
shapes available from
concrete pipe
manufacturers
Circular Horizontal Ellipse
Vertical Ellipse Arch
Rectangular
Reproduced by permission of the American Concrete Pipe Association.
Table 11.1 Standard sizes of circular reinforced-concrete pipe with tongue and groove
joints (adapted from ACPA, 2000)
Min. Wall Thickness, in. (mm)

Internal Diameter
Wall A Wall B Wall C
in. mm in. mm in. mm in. mm
12 300 13/4 44 2 50
15 375 17/8 48 21/4 56
18 450 2 50 21/2 63
21 525 21/4 56 23/4 69
24 600 21/2 63 3 75 33/4 94
27 675 25/8 66 31/4 82 4 100
30 750 23/4 69 31/2 88 41/4 106
33 825 27/8 72 33/4 94 41/2 113
36 900 3 75 4 100 43/4 119
42 1050 31/2 88 41/2 113 51/4 131
48 1200 4 100 5 125 5 /4
3
144
54 1350 4 /2
1
113 5 /2
1
138 6 /4
1
156
60 1500 5 125 6 150 6 /4
3
169
66 1650 51/2 138 61/2 163 71/4 182
72 1800 6 150 7 175 7 /4
3
194
78 1950 6 /2
1
163 7 /2
1
188 8 /4
1
207
84 2100 7 175 8 200 8 /4
3
219
90 2250 71/2 188 81/2 213 91/4 232
96 2400 8 200 9 225 9 /4
3
244
102 2550 8 /2
1
213 9 /2
1
238 10 /4
1
257
108 2700 9 225 10 250 10 /4
3
269
114 2850 91/2 238
120 3000 10 250
126 3150 10 /2
1
263
132 3300 11 275
138 3450 111/2 288
144 3600 12 300
150 3750 12 /2
1
313
156 3900 13 327
162 4050 131/2 339
168 4200 14 350
174 4350 14 1/2 363
180 4500 15 375
Concrete pipe is also manufactured in elliptical (horizontal or vertical) and arch-

shaped cross sections for use in applications where vertical or horizontal clearances
are limited. Rectangular concrete box sections are also available. For noncircular sec-
tions, the pipe or box size is specified by both its internal horizontal dimension (span)
and its internal vertical dimension (rise). Standard product sizes are provided in vari-
ous ACPA publications.
Two basic types of joints are used to connect segments of concrete pipe. Bell-and-
spigot joints are common for smaller pipe diameters (4 to 36 in. or 100 to 900 mm),
and tongue-and-groove joints are used with larger pipe sizes (12 to 180 in. or 300 to
4500 mm). Joints are usually sealed with a bituminous mastic material supplied in
rope form by the pipe manufacturer, but mortar, compression rubber gaskets, or O-
ring rubber gaskets can also be used. These latter types of joints are useful in situa-
tions where water-tightness is a concern. In high-pressure applications, such as force
mains, joints can be manufactured with steel end rings and O-ring gaskets.
Corrugated-Steel Pipe. Corrugated-steel pipe is another frequently used mate-
rial in storm-sewer construction. Like concrete pipe, it is manufactured in a number of
cross-sectional shapes and sizes and in strict accordance with applicable standard
specifications. The American Iron and Steel Institute (AISI) publishes a number of
design documents, some of which are available for download from their website.
Circular corrugated-steel pipe has either annular (ringed) or helical (spiral) corruga-
tions, and is available in diameters ranging from 4 to 144 in. (100 to 3600 mm). Stan-
dard sizes and thicknesses for steel pipes of various corrugation sizes are shown in
Table 11.2. Like concrete pipe, 12-inch-diameter (300-mm) steel pipe is generally the
minimum recommended size for storm-sewer applications. For corrosion protection,
corrugated-steel pipe may be specified as galvanized, bituminous-coated, or coated
and asphalt-paved. Steel thicknesses should be specified to withstand the loads
expected to occur in a particular application.
In addition to circular pipe sections, corrugated-steel products include arch pipes,
structural-plate pipe arches, and structural-plate arches. Structural-plate products con-
sist of galvanized plates that require bolted assembly in the field, as illustrated in Fig-
ure 11.3. The shape of steel arch pipe is similar to that of concrete arch pipe. A
structural-plate arch is an arch span similar to a typical bridge, in that it has no bot-
tom. Other available steel drainage-system products include bends, tees, wyes, transi-
tions, manhole risers, and end sections.
Table 11.2 Standard sizes of corrugated-steel pipe (adapted from AISI, 1999)
Internal
Diameter Available Steel Thicknesses
in. mm in. mm in. mm in. mm in. mm in. mm in. mm
11⁄2 ×1⁄4-in. (38 × 6.5-mm) Corrugations
6 150 0.052 1.32 0.064 1.63
8 200 0.052 1.32 0.064 1.63
10 250 0.052 1.32 0.064 1.63
12 300 0.052 1.32 0.064 1.63
15 375 0.052 1.32 0.064 1.63
18 450 0.052 1.32 0.064 1.63
2 × 1⁄2-in. (68 × 13-mm) Corrugations
12 300 0.052 1.32 0.064 1.63 0.079 2.01
15 375 0.052 1.32 0.064 1.63 0.079 2.01
18 450 0.052 1.32 0.064 1.63 0.079 2.01
21 525 0.052 1.32 0.064 1.63 0.079 2.01
24 600 0.052 1.32 0.064 1.63 0.079 2.01
30 750 0.052 1.32 0.064 1.63 0.079 2.01
36 900 0.052 1.32 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51
42 1050 0.052 1.32 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51
48 1200 0.052 1.32 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
54 1350 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
60 1500 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
66 1650 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
72 1800 0.109 2.77 0.138 3.51 0.168 4.27
78 1950 0.109 2.77 0.138 3.51 0.168 4.27
84 2100 0.109 2.77 0.138 3.51 0.168 4.27
90 2250 0.109 2.77 0.138 3.51 0.168 4.27
96 2400 0.138 3.51 0.168 4.27
3 × 1-in. or 5 × 1-in. (75 × 25-mm or 125 × 25-mm) Corrugations
54 1350 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
60 1500 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
66 1650 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
72 1800 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
78 1950 0.109 2.77 0.138 3.51 0.168 4.27
84 2100 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
90 2250 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
96 2400 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
102 2550 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
108 2700 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
114 2850 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
120 3000 0.064 1.63 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
126 3150 0.079 2.01 0.109 2.77 0.138 3.51
132 3300 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
138 3450 0.079 2.01 0.109 2.77 0.138 3.51 0.168 4.27
144 3600 0.109 2.77 0.138 3.51 0.168 4.27
Figure 11.3
Assembly of
structural-plate pipe
arches for an
underground
stormwater-detention
facility
Reproduced by permission of the American Iron and Steel Institute.
Manning’s roughness values for corrugated-steel pipe vary considerably depending

on the type of corrugation, pipe size, how full the pipe is, and whether or not it has
asphalt lining. Table 11.3 provides data for selecting n-values for corrugated-steel
pipe.
Joints for corrugated-steel pipes usually consist of bands that overlap at the ends of
adjoining pipe sections. Mastic, sleeve-type, or O-ring gaskets are used with the
bands to seal against leakage. For unusual conditions such as high pressures or in
locations with high disjointing forces, alternate types of joints can be specified.
Table 11.3 Values of Manning’s roughness coefficient (n) for standard corrugated-steel pipe (AISI, 1999)
Condition
Corrugation Type and Pipe Size (if applicable) Full Unpaved 25% Paved Part Full Unpaved
Annular 2 ⁄3in. (68 mm), circular
2 0.024 0.021 0.027
Helical 1 ⁄2 × ⁄4 in. (38 × 6.5 mm), circular

1 1
8 in. (200 mm) 0.012 — —

10 in. (250 mm) 0.014 — —
Helical 22⁄3 × 1⁄2 in. (68 × 13 mm), circular
12 in. (300 mm) 0.011 — 0.012
15 in. (375 mm) 0.012 — 0.013
18 in. (450 mm) 0.013 — 0.015
24 in. (600 mm) 0.015 0.014 0.017
30 in. (750 mm) 0.017 0.015 0.019
36 in. (900 mm) 0.018 0.017 0.020
42 in. (1050 mm) 0.019 0.018 0.021
48 in. (1200 mm) 0.020 0.020 0.022
54 in. (1350 mm) and larger 0.021 0.018 0.023
Annular 22⁄3 in. (68 mm), arch 0.026 — 0.029
Helical 22⁄3 ×1⁄2 in. (68 ×13 mm), arch
17 × 13 in. (430 × 330 mm) 0.013 — 0.018
21 × 15 in. (530 × 380 mm) 0.014 — 0.019
28 × 20 in. (710 × 510 mm) 0.016 — 0.021
35 × 24 in. (885 × 610 in.) 0.018 — 0.023
42 × 29 in. (1060 × 740 mm) 0.019 — 0.024
49 × 33 in. (1240 × 840 mm) 0.020 — 0.025
57 × 38 in. (1440×970 mm) 0.021 — 0.025
64 × 43 in. (1620×1100 mm) and larger 0.022 — 0.026
Annular 3 × 1 in. (75 × 25 mm), circular 0.027 0.023 —
Helical 3 × 1 in (75 × 25 mm), circular
36 in. (900 mm) 0.022 0.019 —
42 in. (1050 mm) 0.022 0.019 —
48 in. (1200 mm) 0.023 0.020 —
54 in. (1350 mm) 0.023 0.020 —
60 in. (1500 mm) 0.024 0.021 —
66 in. (1650 mm) 0.025 0.022 —
72 in. (1800 mm) 0.026 0.022 —
75 in. (1950 mm) and larger 0.027 0.023 —
Annular 5 × 1 in. (125 × 25 mm), circular 0.025 0.022 —
Helical 5 × 1 in. (125 × 25 mm), circular
48 in. (1200 mm) 0.022 0.019 —
54 in. (1350 mm) 0.022 0.019 —
60 in. (1500 mm) 0.023 0.020 —
66 in. (1650 mm) 0.024 0.021 —
72 in. (1800 mm) 0.024 0.021 —
75 in. (1950 mm) and larger 0.025 0.022 —
All pipe with smooth interior a All diameters 0.012
a. Includes full-paved, concrete-lined, spiral-rib, and double-wall pipe.

Storm Sewer Appurtenances

Storm sewer systems consist of a number of appurtenances other than the sewer con-
duits (pipes). Among these appurtenances are inlet boxes and catch basins; manholes
(access holes) and junction structures; and outlet structures and associated energy-
dissipation devices. Increasingly, additional types of appurtenances are finding their
way into storm sewer systems, primarily in response to water-quality concerns.
Many of the elements described in this section were introduced in Section 2.2, which
included several photographs. This section provides more detailed descriptions.
Inlet Boxes. An inlet box is a structure that supports an inlet opening and con-
nects it to the subsurface piping system. Inlet boxes also allow access to the subsur-
face system for cleaning and inspection. The walls of the inlet box may be precast or
cast-in-place concrete, brick, masonry block, or vertically mounted corrugated pipe.
Three typical inlet box configurations are shown in Figure 11.4.
Ideally, the base of an inlet box will be shaped and formed to more smoothly channel
flows from the inlet pipe(s) to the outlet pipe. As described in Section 11.4 (page
437), shaping of the base minimizes minor energy losses due to flow expansion and
contraction within the structure. In new construction, where concrete is used almost
exclusively, experienced construction personnel can pour and finish a shaped base
very quickly and without concrete forms. When a precast-concrete structure is used,
ready-mix concrete can be poured into the base and shaped after the precast section
has been set in place.
Catch Basins. Although the two terms are often used synonymously, stormwater
inlets and catch basins are not the same. Both are intended to convey surface water
into the storm sewer pipe system, and from the surface they appear to be identical.
However, a catch basin differs from an inlet in the important respect that it has a
sump, which is a portion of the subsurface structure that is lower than the invert of the
outlet pipe. Also, a catch basin sometimes has one or more baffles (such as the hinged
cast iron hood in Figure 11.5) to enhance capture of suspended sediment and other
debris in the stormwater before it can enter the outlet pipe. In some cases, outlet pipes
from catch basins may be equipped with traps to prevent objectionable sewer odors,
but this is normally required only for combined sewers. Catch basins offer water-
quality benefits not offered by regular inlets, but they require more regular mainte-
nance to keep them free of trapped debris.
Manholes. Like inlets and catch basins, manholes (or access holes) can provide
access to the sewer system for routine inspection and maintenance. Manholes are usu-
ally installed where there is a change in horizontal (plan-view) pipe direction or pipe
slope, where several pipes join, or where the pipe size changes. However, manholes
are typically used for these purposes only where use of an inlet box would be inappro-
priate. Manholes or inlet boxes should also be installed to provide regular access
intervals along straight sections of sewer. Recommended maximum distances
between access locations depend on pipe diameter and are indicated in Table 11.4.
Figure 11.4
Typical inlet box
configurations
(Brown, Stein, and
Warner, 2001)
Figure 11.5
Catch basin
Table 11.4 Manhole (or inlet box) spacing criteria (adapted from Brown, Stein, and Warner,
2001)
Pipe Diameter Maximum Spacing
in. mm ft m
12–24 300–600 300 91
27–36 675–900 400 122
42–54 1050–1350 500 152
60 and larger 1500 and larger 1000 305
Illustrations of precast-concrete manhole assemblies are provided in ACPA (2000)

(see Figure 11.6). The riser or barrel of a manhole typically has a circular cross sec-
tion, though not necessarily so. In old construction, manhole barrels were usually
constructed of brick and mortar. The risers of new manholes are almost exclusively
precast concrete with bituminous (mastic) joint fillers, though some are constructed of
corrugated pipe. Internal diameters of manhole risers are usually 4 ft (1.2 m), but pre-
cast sections are available in other diameters. Manhole steps (usually aluminum)
should be provided to facilitate access.
In new construction, the upper portion of the subsurface structure usually consists of a
precast eccentric cone to transition between the diameter of the riser and the diameter
of the opening. For applications where the depth of cover over the sewer is limited,
precast flat-slab tops may be used instead. Precast-concrete grade rings or courses of
brick and mortar may be placed on top of either the cone or flat slab to serve as final
grade adjustments for the cast-iron manhole frame and cover. Where applicable,
frame and cover should be rated for heavy-duty service and should be able to with-
stand heavy vehicular traffic loadings.
Typically, a manhole base is made of precast concrete, and it has holes that have been
appropriately sized and located for connection of the storm sewer piping. As with
inlet boxes, it is preferable for the base of the manhole to be formed to minimize
energy losses through the structure. A transition section connects the base to the riser.
For large-diameter sewers, manholes may be formed using special tee sections in the
sewer pipe (see Figure 11.7). In such cases, minor energy losses through the manhole
structure are negligible, particularly when the sewer flow is less than full.
Junction Structures. Junction structures, like manholes, are used where two or
more pipes join one another. However, junction structures tend to be associated with
large-pipe installations for which more-conventional precast manhole sections are not
large enough, or with installations having complicated connecting-pipe alignments
and geometries. Junction structures are usually constructed of cast-in-place reinforced
concrete, although specialized corrugated-steel junctions can also be fabricated. Many
junction structures do not provide direct access to the surface; however, manhole
structures may be integrated for this purpose.
Figure 11.6
Standard Precast-concrete
Manhole manhole assemblies
Standard Frame and
Manhole Cover
Frame and
Cover Grade Rings
or Brick
Grade Rings Eccentric
or Brick Cone
Flat Slab Top
Base
Base Standard
Manhole
Standard Frame and
Manhole Cover
Frame and
Grade Rings
Cover
or Brick
Grade Rings
Eccentric
or Brick
Cone
Transition
Section
Concentric
Cone
Base
Base
Standard Standard
Manhole Manhole
Frame and Frame and
Cover Cover
Grade Rings Grade Rings
or Brick or Brick
Eccentric
Concentric Cone
Cone
Riser Riser
Riser Riser
Riser
Transition
Section
Tee Section
Base
Credit: Reproduced by permission of the American Concrete Pipe Association.
Figure 11.7
Large-diameter
manhole construction
Credit: Steel and McGhee, 1979. Reproduced with permission of the McGraw-Hill Companies.
Outlet Structures. Exit velocities at storm sewer outfalls are often high enough
to cause erosion problems and even undermining of the outfall pipe. Outlet structures
can be designed to limit or prevent undermining by dissipating flow energy. As
described in Section 11.7 (page 464), an outlet structure may consist of a simple con-
crete endwall with riprap and a filter blanket downstream of the outfall point, or a
more complex energy dissipator or stilling basin configuration may be used. Prefabri-
cated, flared end-sections are also available in both concrete and corrugated-pipe
materials. In any case, it is recommended that a concrete toe-wall (also called a cradle
or cut-off wall) similar to that in Figure 9.5 (page 330) and Figure 11.11 (page 465) be
placed at the outfall point to prevent undermining of the pipe and/or outlet structure.
Sometimes, a storm sewer may outfall to a water body whose water-surface elevation
may, at times, be high enough to back up and flood the system. This backup can be
prevented by a valve at the outlet, such as a flap gate or duckbill check valve.
Other Structures. Many types of structures and facilities other than those
described in the previous subsections can be included in storm sewer systems. Under-
ground detention basins, for example, may be fitted with manifold-like distribution
chambers to distribute incoming flows to multiple storage chambers or to collect out-
flows from multiple storage chambers. Increasing environmental and water-quality
concerns have given rise to the use of water-quality-enhancement structures, includ-
ing filters and vortex-type oil and sediment separators.
Section 11.2 Alignment of Storm Sewers 421
11.2 ALIGNMENT OF STORM SEWERS

In an urbanized setting where much of the utility infrastructure is buried in public
rights of way, a degree of standardization for utility locations within the right of way
can minimize potential conflicts between the alignments of the various utility lines
and can make maintenance operations more efficient. Locations of utilities such as
sanitary sewers, potable-water lines, electrical lines, and gas lines are more easily
standardized because they are present in nearly every public street right-of-way and,
except for sanitary sewers, are not gravity-driven. Nevertheless, consideration should
also be given to standardizing storm sewer alignments.
Horizontal Alignment
The horizontal (plan-view) alignment of a storm sewer normally should be straight
between manholes, inlets, and other structures. This restriction may necessitate very
close manhole spacing in roadway curves with small radii.
Curved storm sewers may be acceptable in some instances if local jurisdictions permit
them, but should not be used with small-diameter pipes because they tend to cause
maintenance problems. Curved sewer alignments are usually constructed by deflect-
ing pipe joints when they are assembled. Associations of pipe manufacturers provide
design guidance on maximum deflection angles.
Storm sewers may be located behind the curb and outside of paved roadway areas, or
they may be located under the pavement at a standard distance from the curb line.
Installation behind the curb means less disruption when excavating and repairing the
sewer; however, this option often requires that temporary or permanent easements be
obtained from private properties adjacent to the public right of way.
Some review agencies suggest that a minimum horizontal distance of about 10 ft
(3 m) be maintained between the centerlines of potable water and storm sewer pipes.
In cases of large-diameter pipes, this centerline-to-centerline distance should be
increased to maintain a minimum separation distance of 5–6 ft between the pipes.
Vertical Alignment
The vertical alignment of a storm sewer should maintain a depth of cover sufficient to
prevent the crushing of pipes by traffic loads. Methods for developing detailed esti-
mates of loads and pipe strengths are published by associations of pipe manufacturers,
although a depth of cover of about 3 ft (0.9 m) is common. The vertical alignment of a
storm sewer should be shown on a profile drawing and should be laid out to minimize
conflicts with other buried utilities. Vertical separation distances between storm sew-
ers and other buried utility lines should be at least 1 ft (0.3 m) to allow for differential
settlement. Where groundwater tables are high, or where bedrock is encountered at
shallow depths, the alignment should also strive to minimize potential buoyancy prob-
lems and/or construction difficulties caused by these factors.
There are a few approaches for setting the relative invert elevations of pipes entering
and leaving a manhole or junction structure. For many engineers, the traditional rule
of thumb has been to match the crown elevations of incoming and outgoing pipes.
Benefits of this approach are said to be that the drop in elevation helps to offset energy
losses through the structure, and the likelihood of having pressurized flow in a smaller
upstream pipe is reduced. An alternative is to match pipe invert elevations, or provide
a small minimum drop in elevation across the structure. When inverts are matched,
the bottom of the structure can be formed to convey some flows with less energy loss.
Finally, others assert that the most hydraulically efficient configuration is to match
pipe springline (centerline) elevations.
In reality, the hydraulic benefit of any of these strategies over another is probably neg-
ligible. From a least-cost perspective, if a minimum cover constraint is driving the ele-
vations of pipe inverts, then matching pipe crowns will typically result in the least
amount of excavation. However, if the ground surface is fairly flat, or if the storm
sewer slopes in the direction opposite that of the ground slope, then minimum pipe
slope requirements will typically drive the invert elevations. In this situation, match-
ing pipe inverts will minimize excavation. In the absence of regulations to the con-
trary or extenuating circumstances, the engineer may choose to match inverts to
minimize construction costs in this case.
Sewer Slopes
The slope of a storm sewer pipe should be sufficient to maintain a self-cleansing flow
velocity, thus preventing sediment deposition that can lead to maintenance problems.
According to ASCE (1992), the minimum flow velocity should be about 2 to 3 ft/s
(0.6 to 0.9 m/s) when the sewer is flowing full. To prevent erosion of the pipe interior
by suspended sediment and debris, the full-flow velocity should be less than approxi-
mately 15 ft/s (4.5 m/s).
For any given pipe diameter and material, a constitutive relationship such as the Man-
ning equation can be applied to determine the minimum pipe slope required to main-
tain a self-cleansing, full-flow velocity. Storm sewers flow full only occasionally, but
it is typically only during storm events that they convey a flow at all. Therefore,
checking minimum velocities for non-full-flow conditions is not as important as it is,
for example, in sanitary sewer design. Table 11.5 is a listing of minimum pipe slopes
required to maintain full-flow velocities of 2 ft/s (0.6 m/s) and 3 ft/s (0.9 m/s) for con-
crete and corrugated-metal pipes. Manning’s n values used to construct the table are
n = 0.013 for concrete and n = 0.024 for corrugated metal. Note that, with few excep-
tions, all of the slopes in the table are quite small—less than minimum slopes nor-
mally specified for construction quality control. Thus, in most cases of practical
interest, self-cleansing velocity criteria will be satisfied regardless of the pipe slope
specified by the designer.
11.3 INITIAL PIPE SYSTEM DESIGN

Following the selection of a design storm return period, which will frequently be
determined by the local regulating authority, and the computation of flows intercepted
by each inlet in the storm sewer system (see Section 10.3), one can proceed with an
initial design of the pipes in the system. The purpose of the initial design is to esti-
Section 11.3 Initial Pipe System Design 423
mate the design flow for each pipe in the system and make an initial selection of the
size required for each pipe.
Table 11.5 Minimum slope required to maintain pipe cleansing velocities for pipes
flowing full
Pipe Slope Required to Maintain Vmin
Concrete Pipe (n = 0.013) Corrugated-Metal Pipe (n = 0.024)

Internal Diameter Vmin = 2 ft/s Vmin = 3 ft/s Vmin = 2 ft/s Vmin = 3 ft/s
in. mm (0.6 m/s) (0.9 m/s) (0.6 m/s) (0.9 m/s)
12 300 0.0019 0.0044 0.0066 0.0149
15 375 0.0014 0.0032 0.0049 0.0111
18 450 0.0011 0.0025 0.0039 0.0087
21 525 0.0009 0.0021 0.0031 0.0071
24 600 0.0008 0.0017 0.0026 0.0059
30 750 0.0006 0.0013 0.0020 0.0044
36 900 0.0004 0.0010 0.0015 0.0034
42 1050 0.0004 0.0008 0.0012 0.0028
48 1200 0.0003 0.0007 0.0010 0.0023
54 1350 0.0003 0.0006 0.0009 0.0020
60 1500 0.0002 0.0005 0.0008 0.0017
66 1650 0.0002 0.0005 0.0007 0.0015
72 1800 0.0002 0.0004 0.0006 0.0014
78 1950 0.0002 0.0004 0.0005 0.0012
84 2100 0.0001 0.0003 0.0005 0.0011
90 2250 0.0001 0.0003 0.0005 0.0010
96 2400 0.0001 0.0003 0.0004 0.0009
The initial design begins with the most upstream pipe(s) and progresses in a down-
stream direction. The design discharge for each pipe includes any flow from the inlet
at the pipe’s upstream end and the flow from the upstream piping system. Typically,
initial pipe sizes are determined by assuming that the pipe slope will be approxi-
mately the same as that of the ground surface, subject to some minimum slope con-
straint. The Manning equation or another method is then used to solve for the required
pipe size given the pipe roughness, design discharge, and slope. The calculated size is
then rounded up to the next commercially available size that is at least as large as the
minimum allowable size.
The required pipe sizes are determined as each pipe is considered in turn. When siz-
ing pipes for a storm sewer, the diameter of a pipe exiting a structure should not be
smaller than the largest incoming pipe, regardless of its conveyance capacity. In other
words, pipe diameters should never decrease downstream, even when pipe slopes may
be such that a smaller downstream pipe would have sufficient capacity.
Design discharges are usually determined using the rational method (see Section 5.4,
page 140), although other methods, such as NRCS (SCS) methods, may be used
instead. Example 11.1 presents a simplified but realistic calculation of storm sewer
design flows and initial pipe sizing for a typical case using the rational method. In this
example, the travel time calculations are simplified by assuming the pipes flow full.
The next two subsections clarify the initial design process for special cases involving
additional considerations in the computation of design flows. Example 11.2 provides
a detailed initial design, including inlet flow calculations, inlet design, pipe sizing,
and more accurate travel time calculations that consider velocities for partially full
pipes.
Based on hydraulic and energy grade line computations and possible constraints on
pipe slopes and vertical alignments, the initially selected pipe sizes may require
changes during the final design process, which is presented in Section 11.5.
Example 11.1 – Computing Flows for Storm Sewer Pipes with the Rational
Method. Figure E11.1.1 is a plan view of a storm sewer system draining three subbasins. Use the
rational method to determine the peak discharge in each pipe given that the inlets do not bypass any
flow, and size each pipe assuming that the pipes flow full. The tables that follow the figure list the IDF
data and subbasin and pipe characteristics. Assume also that the pipes will be concrete with
n = 0.013.
Subbasin A
Pipe 1
Subbasin B Subbasin C
Pipe 2
Pipe 3
To Outfall
Figure E11.1.1 System for Example 11.1
Table 11.6 IDF data for Example 11.1
Duration (min) Intensity (mm/hr)

5 213.0
10 178.0
15 151.0
30 112.0
60 75.0
Table 11.7 Subbasin characteristics for Example 11.1
Subbasin A (ha) C tc (min)

A 2.5 0.6 20
B 1.5 0.8 10
C 2.0 0.8 15
Table 11.8 Pipe characteristics for Example 11.1
Pipe Length (m) Slope (%)

1 150 1.0
2 120 1.2
3 150 0.9
Flow into Pipe 1 occurs from Subbasin A only. Using the time of concentration as the storm duration,
the 25-year rainfall intensity is obtained from Table 11.6 as i = 133.6 mm/hr. The peak discharge used
in sizing Pipe 1 is therefore
Q = 0.6(2.5)(133.6) = 200.4 ha-mm/hr = 0.56 m3/s

Assuming that Pipe 1 is flowing full, its required diameter D may be found using Manning’s equation
as follows:
3/8 3/8
 Qn   .56(0.013) 
D= 1/ 2 
=   = 0.58 m
 0.312 S   0.312(0.01)1/ 2 
 
Rounding up to the next commercially available size, Pipe 1 should have a diameter of 600 mm.
The cross-sectional area of Pipe 1 is therefore 0.28 m2, and the average velocity in Pipe 1 is
V = Q/A = 0.55/0.28 = 2.0 m/s
The travel time in Pipe 1 is
t = L/V = 150/2.0 = 75 s = 1.3 min
Note that the travel time is roughly estimated by assuming that the pipe flows full.
Pipe 2 is treated the same way as Pipe 1, recognizing that runoff from Subbasin C only enters Pipe 2.
The peak discharge from Subbasin C is Q = 0.62 m3/s, and the required diameter of Pipe 2 is D = 600
mm. The travel time in Pipe 2 is t = 54 s = 0.9 min. The cross-sectional area of Pipe 2 is A = 0.28 m2.
Pipe 3 must be sized to handle the runoff from all three of the subbasins, which have a total area of 6.0
ha. The runoff coefficient for the combined areas is computed as a composite value, and is
2.5(0.6) + 1.5(0.8) + 2(0.8)
C= = 0.72
6
The time of concentration is computed as the longest of the travel times to the upstream end of Pipe 3.
These travel times are:
The time of concentration of Subbasin B (10 minutes).
The time of concentration of Subbasin A plus the travel time in Pipe 1 (20 + 1.3 = 21.3 min).
The time of concentration of Subbasin C plus the travel time in Pipe 2 (15 + 0.9 = 15.9 min).
Thus, the time of concentration for Pipe 3 is 21.3 min, and the corresponding rainfall intensity is
134.6 mm/hr. The peak discharge for Pipe 3 is
Q = 0.72(134.6)(6) = 581.5 ha-mm/hr = 1.62 m3/s

The required diameter of Pipe 3 is 880 mm. Rounding up to the next available size yields a diameter
of 900 mm.
Multiple Design-Storm Return Periods

Frequently, a storm sewer system serves a drainage basin in which multiple land uses
are present. For example, one part of a storm sewer system may serve a residential
area, while another branch of the same system may serve a commercial area such as a
local shopping center. For such a case, regulations might state that the residential part
of the system should be designed for a 2-year return period, and the commercial por-
tion for a 10-year period.
Because high-return-period flows are larger than short-return-period flows, the por-
tion of the overall system that must convey runoff from both types of land uses should
be designed using the higher flows. Portions of the system that serve only single land
uses should be designed according to their respective return periods.
In the final design phase when hydraulic and energy grade lines are computed, calcu-
lations for each design storm should be performed. First, the entire system should be
analyzed for lower-return-period discharges, even though some parts of the system
were initially designed for higher discharges. The final set of computations then ana-
lyzes the entire system for high-return-period discharges. In this calculation run, it is
expected that the portion of the system designed for low-return-period discharges will
be inadequate to convey the higher flows, and therefore the major drainage system
will be activated in that portion of the system. This multi-step procedure is necessary
to ensure that the portion of the system serving multiple land uses is capable of con-
veying overall system inflows corresponding to the highest applicable design storm
return period.
Directly and Indirectly Connected Areas

The rational method for design flow estimation, as presented in Example 11.1, is
applicable in instances where the drainage basin contributing runoff to an inlet is
homogeneous in terms of its land use characteristics and small enough that the ratio-
nal method is considered appropriate (local practices vary regarding acceptable maxi-
mum basin size). However, many urban drainage areas are non-homogeneous in the
sense that the impervious areas of the drainage basin contributing to each inlet can be
subdivided into directly connected areas and indirectly connected areas. Runoff from
directly connected impervious areas such as streets and driveways proceeds directly
to the nearest conveyance (storm sewer, drainage ditch, etc.) without traversing a per-
vious area such as a lawn. Runoff from indirectly connected impervious areas; how-
ever, flows across impervious areas where it has the chance to at least partially
infiltrate before reaching an inlet. An example of an indirectly connected impervious
area is a roof drained by a downspout that discharges onto a lawn.
Although the directly connected impervious area of a drainage basin is smaller than
the total basin area, the peak rate of runoff from the directly connected area alone may
be larger than the peak rate of runoff from the basin as a whole. This discrepancy
occurs because the time of concentration associated with the directly connected area
is usually smaller than the time of concentration for the basin as a whole, which trans-
lates to a higher design storm intensity. In such an instance, design flows to each inlet
and pipe should be computed twice, once assuming that only the directly connected
area contributes runoff, and a second time assuming that the entire drainage basin
contributes runoff. The larger of the runoff estimates may then be taken as a conserva-
tive estimate of the design discharge.
The initial design process, including consideration of directly connected impervious
areas, is illustrated in Example 11.2.
Example 11.2 – Inlet Design and the Initial Piping Design Process. The layout of
a proposed storm sewer system to serve three city blocks of single-family residential lots is illustrated
in Figure E11.2.1. Design the system for a 2-year event using the IDF data in the table below.
Washington Street
Connecticut Avenue
Delaware Avenue
Colorado Avenue
Dakota Avenue
Basin A Basin B Basin C
Pipe 1 Pipe 2
Inlet 1 Inlet 2 Inlet 3
Jefferson Street
Pipe 3 Pipe 4
Basin D
Inlet 4 Outfall
Open Channel
Figure E11.2.1 Storm sewer system for Example 11.2
Duration (min) Intensity (in/hr)

5 5.88
10 4.80
15 4.04
30 2.78
60 1.79
Drainage area data

Each of the city blocks (that is, Basins A through C) is 330 ft by 660 ft, which corresponds to a total
drainage area of 5.0 ac for each, of which impervious areas (C = 0.95) total 1.7 ac. The remaining area
of each block consists of lawns and landscaped areas with C = 0.15. Of the 1.7 ac of impervious area,
1.0 ac is directly connected. The time of concentration is 20 minutes for each 5.0 ac area, or 10 min-
utes if only the directly connected impervious area is considered.
Basin D consists of one-half of the right-of-way of Jefferson St. between Colorado Ave. and Dakota
Ave. It has an area of 0.45 ac, a runoff coefficient of 0.95, and a time of concentration of 5 minutes.
Gutter and inlet design criteria
Street slopes are So = 0.50 percent in the directions shown in Figure E11.2.1. The composite gutters
have a width W = 2 ft, a curb height of 6 in., and a gutter cross slope of Sw = 6.25 percent (see Figure
E11.2.2). Street cross slopes are Sx = 2 percent. The maximum allowable spread T is 15 ft. Manning’s
n = 0.015 for the gutters.
All of the inlets shown in Figure E11.2.1 are curb-opening inlets in sag locations. Additional inlets, if
required on the basis of gutter flow calculations, will be continuous-grade inlets. The inlets have hor-
izontal throats with h = 5 in., and the gutters are depressed an additional 2 in. at inlet locations (see
Figure E11.2.3). Required inlet lengths should be increased by 25 percent to account for potential
clogging. The minimum permissible inlet length is 3 ft. The inlet throat bottom elevations are given in
the following table.
Figure E11.2.2 Gutter cross-section for Example 11.2
Figure E11.2.3 Curb inlet for Example 11.2
Inlet Label Throat Elevation (ft)

Inlet 1 212.20
Inlet 2 211.73
Inlet 3 213.50
Inlet 4 211.85
Pipe design criteria

Pipes are to be reinforced concrete with n = 0.013. The minimum acceptable pipe diameter, as dic-
tated by local regulations, is 15 in. Pipe crowns should be matched at junctions where incoming and
outgoing pipes are of different diameters. Based on the flatness of the local topography, pipe slopes
should be no greater than the minimum allowable slope of 0.5 percent specified by the local review
agency. Pipe 4 outfalls to an open channel with an invert elevation of 205.73 ft. This controlling out-
fall elevation is used with the pipe slopes and crown matching criteria to establish pipe invert eleva-
tions throughout the sewer system. Pipe lengths are provided in the following table.
Pipe Label Length (ft)

Pipe 1 70
Pipe 2 315
Pipe 3 40
Pipe 4 40
Solution: The design proceeds by determining, first, the design flow rates and street conveyance
capacities in order to evaluate whether the placement of inlets shown in Figure E11.2.1 is adequate.
Inlets are then designed to intercept the design flows, and pipes are designed to convey flows to the
system outfall point.
Step 1: Compute design flows for inlets
Because the drainage basins A, B, and C are essentially identical in terms of their hydrologic charac-
teristics, the peak runoff rate computed for one is also applicable to the other two. Noting that imper-
vious areas in each basin consist of both directly-connected and indirectly-connected areas, the peak
runoff calculations must be performed twice.
The first set of peak runoff calculations for basins A, B and C assumes that only the directly-con-
nected impervious areas are contributing runoff to the basin outlet. Thus, the area and time of concen-
tration for use in the rational method are A = 1.0 ac and tc = 10 min. The rainfall intensity
corresponding to a tc of 10 minutes is i = 4.80 in/hr, and the peak runoff rate is
Q = CiA = 0.95(4.80 in/hr)(1.0 ac) = 4.6 cfs
The second set of peak runoff calculations for Basins A through C computes the runoff from the
entire basin area of A = 5.0 ac. The composite runoff coefficient for the entire basin is
C = [(1.7 ac)(0.95) + (5.0 ac − 1.7 ac)(0.15)]/5.0 = 0.42
The time of concentration of the entire basin is 20 minutes; thus, i = 3.62 in/hr. The peak runoff rate is
Q = CiA = 0.42(3.62 in/hr)(5.0 ac) = 7.6 cfs
The larger of the two peak discharges, 7.6 cfs, controls the design of the gutters, inlets, and pipes.
For Basin D, which consists only of the right-of-way of Jefferson St., A = 0.45 ac, C = 0.95, and tc = 5
minutes. The rainfall intensity corresponding to a 5-minute duration is i = 5.88 in/hr. The peak dis-
charge from Basin D is therefore
Q = CiA =0.95(5.88 in/hr)(0.45 ac) = 2.5 cfs
Step 2: Perform gutter capacity calculations
The total discharge of 7.6 cfs approaching each of the Inlets 1 through 3 arrives at each inlet by two
routes. For example, in Basin A, flows approach Inlet 1 via both Jefferson St. and Connecticut Ave.
Assuming that the flows in each approaching curb are more or less equal (a reasonable assumption for
this example because of the basin geometries), the flow in each approach gutter is about 3.8 cfs.
Based on similar reasoning, the flow approaching Inlet 4 from each direction in Jefferson St. is about
1.25 cfs. The gutter capacities must be checked to ensure that these discharges do not exceed permis-
sible flow depth or spread limits.
Design criteria provide for a maximum permissible spread of T = 15 ft. With a gutter width of W = 2
ft, the allowable side flow width from (10.6) is
Ts = T − W = 15 – 2 = 13 ft
The discharge capacity Qs in the roadway area can be determined using Equation 10.9 as
0.56 5 / 3 8 / 3 1 / 2 0.56
Qs = S T So = (0.02)5 / 3 (13)8 / 3 (0.005)1 / 2 = 3.6 cfs
n x s 0.015
The total discharge capacity Q of the gutter can be determined using Equations 10.11 and 10.12. The
fraction Eo is determined from (10.12) to be
−1 −1
   
   
 S /S
w x   0.0625 / 0.02 
E = 1 +  = 1 +  = 0.372
o 8/3 8/3
  S /S 
w x    0.0625 / 0.02  
 1 +  −1   1 + (15 / 2) − 1  − 1 
  (T / W ) − 1      
Then, from (10.11), the total discharge capacity is found as

Q 3.6 cfs
s
Q= = = 5.7 cfs
1 − Eo 1 − 0.372
Because the gutter capacity of 5.7 cfs is greater than the peak runoff rate (3.8 cfs or 1.25 cfs, depend-
ing on location), the adequacy of the gutters is established and no additional inlets are required.
Step 3: Perform Inlet Design

Because the gutter is depressed at the inlet location, (10.33) should be used to estimate a required
inlet length under the assumption that it behaves as a weir. The weir coefficient is Cw = 3.0, and the
gutter width is W = 2 ft. The depth d required for the expression (see Figure E11.2.3) is
d = TSx = 15(0.02) = 0.30 ft
Therefore, rearranging (10.32) to solve for L, the required length for Inlets 1 through 3 (Q = 7.6 cfs) is
Q 7.6
L= − 1.8W = − 1.8 × 2 = 11.8 ft
Cw d 3/ 2 3.0 ( 0.30 )
3/ 2
The required length of Inlet 4 (Q = 2.5 cfs) is similarly calculated as
Q 2.5
L= − 1.8W = − 1.8 × 2 = 1.5 ft
Cw d 3/ 2 3.0 ( 0.30 )
3/ 2
To find the required inlet length assuming orifice flow, it is necessary to first compute the flow depth
at the curb face di, which can be used to find the depth of the centroid of the inlet throat opening, do
(see Figure E11.2.3):
di = TsSx + WSw + (2-in. inlet depression) = 13(0.02) + 2(0.0625) + 2/12 = 0.55 ft
do = di − h/2 = 0.55 − (5/2)/12 = 0.34 ft
Then, the required length of Inlets 1 through 3 can be determined by solving (10.34) for length:
Q 7.6
L= = = 5.8 ft
1/ 2
0.67 h(2 gd o ) 0.67(5 /12)[2 × 32.2 × 0.34]1/ 2
The required length of Inlet 4 is
Q 2.5
L= 1/ 2
= = 1.9 ft
0.67 h(2 gd o ) 0.67(5 /12)[2 × 32.2 × 0.34]1/ 2
For each inlet, the longer of the two calculated inlet lengths should be used. Thus, for Inlets 1 through
3, weir behavior controls the design, and the required length of each inlet is 11.8 ft. Orifice behavior
controls the length of Inlet 4; its required length is 1.9 ft. Increasing required inlet lengths by 25 per-
cent to account for clogging, the required length (rounded to the nearest foot) for Inlets 1 through 3 is
L = 1.25(11.8) = 15 ft
and for Inlet 4 is
L = 1.25(1.9) = 2 ft
Invoking the design criterion regarding minimum permissible inlet length, an inlet of length L = 3 ft
will be used for Inlet 4.
Step 4: Design Pipes

Pipes 1 and 2 are terminal pipes in the system in the sense that each receives inflow from only one
inlet. The design discharge for each of these pipes is therefore Q = 7.6 cfs, and the pipes can be sized
to flow full under design flows. Using concrete pipes with n = 0.013 and So = 0.005, the required pipe
diameters can be found by rearranging the form of the Manning equation given in (6.18):
Cf
V= R 2 / 3 So1/ 2
n
The velocity V can be rewritten as
Q Q
V= =
A π D2 / 4
where D is the pipe diameter, and the hydraulic radius R is
R = D/4
Rearranging to solve for diameter D and using substitution, the Manning equation becomes:
3/8 3/ 8
 nQ   0.013(7.6) 
D=  =  = 1.51 f
 0.464 S1/ 2   0.464(0.005)1/ 2 
 o   
Rounding up to the next commercially available size, Pipes 1 and 2 should have diameters of 21 in.
Pipe 3 carries flows from both pipes 1 and 2 (that is, Basins A and C), plus additional flow entering
locally at that inlet from Basin B. Its design discharge can be found by applying the rational method
to the total area of Basins A, B, and C (15 ac) with a composite runoff coefficient and a controlling
time of concentration. Because Basins A, B, and C are nearly identical, the C value of 0.42 computed
previously applies.
The controlling time of concentration is the longest travel time to the upstream end of Pipe 3, which
in this case would be the longest of the following:
a) The time of concentration to Inlet 1 plus the travel time of flow in Pipe 1
b) The time of concentration to Inlet 3 plus the travel time of flow in Pipe 2
c) The time of concentration to Inlet 2
The time of concentration to Inlet 2 is not larger than the others, the velocities in Pipes 1 and 2 can be
assumed to be equal in this case, and Pipe 2 is longer than Pipe 1. Therefore, the longest travel time
approaching Inlet 2 must be from Basin C/Inlet 3, and the velocity in Pipe 2 must be determined.
The velocity in a circular pipe flowing partially full can be determined with the aid of Figure 7.8,
which is shown again here as Figure E11.2.4. Using the Manning equation, the full-flow velocity Vfull
for a 21-in. pipe is
2/3
1.49 2 / 3 1/ 2 1.49  21/12 
V full = R So =   0.0051/ 2 = 4.67 ft/s
n 0.013  4 
and the full-flow capacity Qfull is
Qfull = AVfull = 4.67[π(21/12)2/4] = 11.2 cfs

Thus, for pipes 1 and 2,
Q/Qfull = 7.6/11.2 = 0.68
From Figure E11.2.4, Q/Qfull = 0.68 corresponds to y/D = 0.60 and V/Vfull = 1.08. Assuming open
channel flow, the actual flow velocity is therefore
V = 1.08(4.66) = 5.03 ft/s
Pipe 2, which has a length of 315 ft, has a travel time of
tt = (315 ft)/(5.03 ft/s) = 63 s = 1.0 min
Thus, the travel time to Pipe 3 is
(Basin C tc) + (Pipe 2 tt) = 20 min + 1 min = 21 min
Interpolating from the 2-year IDF table, the rainfall intensity for a duration of 21 min. is i = 3.54
in/hr. The design discharge for Pipe 3 is
Q = CiA = 0.42(3.54)(15.0) = 22 cfs
Applying Manning’s equation with n = 0.013 and So = 0.005, the required pipe diameter (after round-
ing to the next commercially available size) is D = 30 in.
1.2
1.0
V/Vfull
0.8
Q/Q full and V/V full
0.6
Q/Q full
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
y/D
Figure E11.2.4 Graph for determining velocity in a circular pipe flowing partially full
Pipe 4 must convey the total discharge from all four basins. Its design discharge can be found using a
composite runoff coefficient for the entire contributing area and the longest flow time. Each of the
Basins A through C has a composite C = 0.42 and A = 5.0 ac, and Basin D has C = 0.95 and A = 0.45
ac. Therefore, the composite C for all four basins is computed as
3(5)(0.42) + 0.45(0.95)
C= = 0.44
15.45
The longest travel time to the upstream end of Pipe 4 can be seen to be equal to the time of concentra-
tion for Basin C plus the travel times in Pipes 2 and 3; thus, it is necessary to find the flow velocity in
Pipe 3. Using Manning’s equation, the full-flow capacity of a 30-in. pipe is computed as Qfull = 29.1
cfs, and the full-flow velocity is Vf = 5.93 ft/s.
Q/Qfull = 22/29.1 = 0.76
Therefore, from Figure E11.2.4, V/Vf = 1.10. Assuming open-channel flow, the actual flow velocity is
V = 1.10(5.93) = 6.52 ft/s
Because Pipe 3 has a length of 40 ft, this translates into a travel time of 6 sec = 0.1 min. The travel
time in Pipe 2 was computed as 1 min. Thus, the total travel time to Pipe 4 is
20 min + 1 min + 0.1 min = 21.1 min
Interpolating from the IDF data, the design rainfall intensity is i = 3.53 in/hr. The design discharge for
Pipe 4 is therefore
Q = CiA = 0.44(3.53)(15.45) = 24 cfs
Applying the Manning equation with n = 0.013 and So = 0.005, the required pipe diameter (after
rounding to the next commercially available size) is D = 30 in.
Step 5: Set Initial Invert Elevations
If pipe crowns are matched at Inlet 2, then the invert elevation of Pipe 3 at that structure should be 9
in. lower than the invert elevations of Pipes 1 and 2 where they enter the structure. At Inlet 4, where
the incoming and outgoing pipes are of the same diameter, an additional invert elevation drop of 0.1 ft
is provided through the structure per standard practice in the project locale. Figure E11.2.5 is a profile
drawing showing the completed initial design.
Figure E11.2.5 Profile of the initial design for Example 11.2

11.4 MINOR LOSSES IN JUNCTION STRUCTURES

The completion of a storm sewer design, which will be presented in Section 11.5,
requires the evaluation of the hydraulic grade line through the system. To do this, the
head losses throughout the system, including both frictional and minor losses, must be
computed.
In Section 6.3 (page 202), the classical method for computing minor head losses due
to pipe entrances, exits, transitions, and various appurtenance was introduced. The
classical approach computes head loss as the product of a minor loss coefficient and
the absolute difference between the velocity heads upstream and downstream of an
appurtenance, and it can be applied to computing minor losses through junction struc-
tures in storm sewer systems such as manholes and inlet boxes. However, at stormwa-
ter inlets, manholes, and junction structures, where flow geometries are often
complex, additional methods have been developed for minor loss prediction. Two of
these methods—the energy-loss and composite energy-loss methods—are presented
by Brown, Stein, and Warner (2001).
Energy-Loss Method
The incoming pipe(s) to an inlet, manhole, or junction structure in a stormwater con-
veyance system can occur in a number of different configurations with respect to their
invert elevations. The various configurations that can occur are
a) All incoming pipe invert elevations lie below the elevation of the predicted
depth of water in the structure.
b) All incoming pipe invert elevations lie above the elevation of the predicted
depth of water in the structure.
Section 11.4 Minor Losses in Junction Structures 435
c) One or more of the incoming pipe invert elevations lie above the elevation of
the predicted depth of water in the structure, and one or more incoming pipe
invert elevations lie below the elevation of the predicted depth of water in the
structure.
Note that a stormwater inlet capturing water on the ground surface and discharging
into an outlet pipe corresponds to configuration (b), provided that the free water depth
inside the structure lies below the elevation of the stormwater inlet opening. If a
stormwater inlet not only captures water at the ground surface for discharge to the
outlet pipe but also acts as a junction between the outlet pipe and another incoming
pipe, it corresponds to configuration (c).
For structures in which all of the incoming pipe invert elevations lie above the pre-
dicted free water surface elevation within the structure (that is, plunging flow occurs
from all of the incoming pipes), the outlet pipe behaves hydraulically as a culvert. In
that case, the water surface elevation within the structure can be predicted using the
methods presented in Chapter 9, and the water surface elevations in each of the
upstream pipes can be determined independently of one another as free outfalls.
The energy-loss method for estimation of head losses at inlets, manholes, and junc-
tions is applicable only to configuration (a) and to the incoming pipes in configuration
(c) whose invert elevations lie below the predicted free water surface elevation within
the structure. When one or more incoming pipes meet these criteria, the method may
be applied to each to determine the corresponding head loss.
For incoming pipes to which the energy-loss method is applicable, the head loss
through the structure can be computed using (6.26), repeated here as (11.1)
Vo2
hL = K (11.1)
2g
where hL = minor loss (ft, m)

Vo = velocity in the outlet pipe (ft/s, m/s)
K = adjusted minor loss coefficient
g = gravitational acceleration (32.2 ft2/s, 9.81 m2/s)
Laboratory research has shown that K can be expressed as
K = K o CD1CD 2CQ CP CB (11.2)
where Ko=initial head loss coefficient based on relative size of structure

CD1=correction factor for pipe diameter
CD2=correction factor for flow depth
CQ=correction factor for relative flow
CP=correction factor for plunging flow
CB=correction factor for benching
Equations 11.1 and 11.2 can be applied to each of several incoming pipes at a struc-
ture, provided their invert elevations lie below the predicted water surface elevation
within the structure.
The initial head loss coefficient, Ko, depends on the size of the structure relative to the
outlet pipe diameter, and on the angle θ between the inlet and outlet pipes (see Figure
11.8):
0.15
 B   B 
K o = 0.1  (1 − sin θ ) + 1.4   sin θ (11.3)
 Do   Do 
where B = structure diameter (ft, m)

Do = outlet pipe diameter (ft, m)
θ = angle between inlet and outlet pipes
Figure 11.8
Plan view of junction
structure
If the structure is not circular, an equivalent structure diameter should be used. The
equivalent diameter is defined as the diameter of a circular structure having the same
area as the actual non-circular one. For example, if the inside dimensions (plan view)
of a stormwater inlet were 3 ft by 6 ft (A = 18 ft2), the equivalent diameter used in
(11.3) would be B = (4A/π)1/2 = [4(18)/3.14]1/2 = 4.79 ft.
Computing CD1 (Correction for Relative Pipe Diameter). The correc-
tion factor for pipe diameter, CD1, needs to be considered only in cases where the pre-
dicted water depth d in the structure is at least 3.2 times as great as the outlet pipe
diameter. When applying this and other correction factors, the depth of water d should
be taken as the difference between the hydraulic grade line elevation at the upstream
end of the outlet pipe and the invert elevation of the outlet pipe.
In such cases, CD1 is determined as
3
D 
CD1 =  o  (11.4)
 Di 
where Di = inlet pipe diameter (ft, m)

When the predicted depth in the structure is not at least 3.2 times the outlet pipe diam-
eter, CD1 = 1.
Computing CD2 (Correction for Flow Depth). The correction factor for
flow depth, CD2, needs to be considered only in cases where the predicted depth in the
structure is less than 3.2 times the outlet pipe diameter; otherwise, CD2 = 1. The cor-
rection factor is determined as
0.6
 d 
CD 2 = 0.5   (11.5)
 Do 
where d = water depth in the structure computed as the difference in HGL and
invert at the upstream end of the outlet pipe (ft, m)
Computing CQ (Correction for Relative Flow). The correction factor for
relative flow, CQ, is required in applications where there are two or more incoming
pipes at a structure. When there is only one incoming pipe, CQ = 1. The correction
factor depends on the angle θ between the outlet pipe and the incoming pipe of inter-
est, and on the ratio of the discharge in the incoming pipe of interest to the outlet pipe:
0.75
 Q 
CQ = 1 + (1 − 2sin θ ) 1 − i  (11.6)
 Qo 
where Qi = discharge in the incoming pipe of interest (cfs, m3/s)

Qo = discharge in the outflow pipe (cfs, m3/s)
Computing CP (Correction for Plunging). When one or more of the
incoming pipes to a structure have invert elevations higher than the elevation of the
free water surface in the structure d, they are said to have free outfalls, and the water
plunges into the structure. Plunging flow can also enter a structure from the inlet
opening. The resulting turbulence and energy dissipation within the structure has an
effect on the head loss for other incoming pipes whose flow is not plunging (that is,
their invert elevations lie below the free water surface). The coefficient CP is com-
puted and applied in the head loss calculations only for the pipe(s) whose flow is not
plunging. The coefficient is determined as
 h  h − d 
C P = 1 + 0.2    (11.7)
 Do   Do 
where h = difference in elevation between the highest incoming pipe invert and the
centerline of the outlet pipe (ft, m)
For cases in which no plunging incoming pipes are present, or in which h ≤ d, the
value of the coefficient is CP = 1.
Selecting CB (Correction for Benching of Structure Invert). Benching
of the invert of a structure (see Figure 11.9) can reduce head losses by effectively
directing the path of the flow through the structure. In practice, the base of a manhole
can be benched by pouring a cast-in-place base that is shaped to the pipes connecting
to the structure. Benching can also be achieved by pouring fresh concrete into the
inside base of a precast manhole section and shaping it to the required form. The cor-
rection factor for benching, CB, is presented in Table 11.9 and depends on the depth of
water in the structure relative to the outlet pipe diameter. For the use of that table, sub-
merged flow is considered to occur if d/Do ≥ 3.2, and unsubmerged flow is considered
to occur if d/Do ≤ 1.0. For values of d/Do between 1.0 and 3.2, linear interpolation
between the tabulated values may be performed.
Figure 11.9
Benching the invert of
a structure
Table 11.9 Correction factors for benching

Bench Type Submerged Unsubmerged
Flat or depressed floor 1.00 1.00
Half bench 0.95 0.15
Full bench 0.75 0.07
Example 11.3 – Applying Minor Losses. Consider a manhole with two inlet pipes, as
shown in Figure E11.3.1. The inlet pipe invert elevations have been set so that their crown elevations
match the crown of the outlet pipe. If the HGL elevation at the upstream end of the outlet pipe is
32.0 m and is controlled by downstream conditions, determine the EGL and HGL elevations for all
three pipes. Assume that no benching is provided in the manhole.
Figure E11.3.1 Manhole configuration for Example 11.3

Solution: The velocity in the outlet pipe is computed as
Vo = Qo/Ao = 1.42/0.64 = 2.23 m/s.
The EGL at the upstream end of the outlet pipe is therefore
EGLo = HGLo + Vo2/2g = 32.0 + 2.322/2(9.81) = 32.27 m

For Pipe 1, θ = 180° so, using (11.3),
0.15 0.15
 B   B   1.5   1.5 
Ko1 = 0.1  (1 − sin θ ) + 1.4   sin θ = 0.1  (1 − sin180° ) + 1.4   sin180° = 0.17
 DO   DO   0.90   0.90 
Similarly, for Pipe 2, where θ = 90°,
0.15
 1.5   1.5 
Ko2 = 0.1  (1 − sin 90°) + 1.4   sin 90° = 1.51
 0.90   0.90 
The depth, d, of flow in the manhole structure is the difference in elevation between the outlet pipe
HGL and its invert, or d = 1.5 m. Also, d/Do = 1.64. Because d/Do < 3.2, the correction factor for pipe
diameter is equal to 1 for both incoming pipes:
(CD1)1 = (CD1)2 = 1
The correction factor for flow depth is determined using (11.5) for each of the incoming pipes:
0.6 0.6
 d   32.0 − 30.5 
( CD2 )1 = ( CD2 )2 = 0.5 D  = 0.5  = 0.67
 o  0.90 
For application of (11.6), the angle θ1 = 180° and sin θ1 = 0. Also, θ2 = 90° and sin θ2 = 1. The coef-
ficients CQ for each of the two incoming pipes are therefore
0.75 0.75
 Q   1.19 
CQ1 = 1 + (1 − 2sin θ ) 1 − 1  = 1 + 1 −  = 1.26
 Qo   1.42 
0.75 0.75
 Q   0.23 
CQ2 = 1 + (1 − 2sin θ ) 1 − 2  = 1 + (1 − 2) 1 −  = 0.12
 Qo   1.42 
Because no plunging flow is in this example, CP1 = CP2 = 1. Further, there is no benching in the struc-
ture so CB1 = CB2 = 1.
The adjusted loss coefficient for Pipe 1 is computed using (11.2) as
K1 = K o1 (CD1 )1 (C D 2 )1CQ1CP1CB1 = 0.17(1)(0.67)(1.26)(1)(1) = 0.14
Similarly,
K 2 = K o 2 (C D1 )2 (CD 2 )2 CQ 2CP 2CB 2 = 1.51(1)(0.67)(0.12)(1)(1) = 0.12
The head loss for Pipe 1 is computed using (11.1) as
Vo2 2.232
hL1 = K1 = 0.14 = 0.04 m
2g 2(9.81)
and the head loss for Pipe 2 is
Vo2 2.232
hL 2 = K 2 = 0.12 = 0.03 m
2g 2(9.81)
The EGL at the downstream end of Pipe 1 is the sum of the EGL at the upstream end of the outlet pipe
plus the head loss, or
EGL1 = EGLo + hL1= 32.27 + 0.04 = 32.31 m
Similarly, for Pipe 2,
EGL2 = EGLo + hL2 = 32.27 + 0.02 = 32.28 m
The HGL in Pipe 1 is computed as
2 2 2
V
HGL1 = EGL 1 – -----1- = 32.31 – [---------------------------------------------
1.19 ⁄ ( π × 0.3 ) ] - = 31.41 m
2g 2 × 9.81
For Pipe 2,
2 2 2
V
HGL2 = EGL 2 – -----2- = 32.29 – [--------------------------------------------------------
0.23 ⁄ ( π × ( 0.225 ) ) ] - = 32.22 m
2g 2 × 9.81
Composite Energy-Loss Method

The composite energy-loss method can be used in situations similar to those for
which the energy-loss method, discussed in the previous section, applies. However,
the composite method is better suited to analyzing losses in structures with many
inflow pipes. It is only applicable to subcritical flows in pipes.
The method is used to compute a unique head loss through the junction structure for
each of the incoming pipes. This head loss is then added to the energy grade at the
upstream end of the outlet pipe (EGLo) to obtain the EGL at the downstream end of
the incoming pipe (EGLi). The velocity head for the incoming pipe can then be sub-
tracted from EGLi to obtain the hydraulic grade, HGLi. The head loss through the
structure for a particular inflow pipe is again given by (6.26), repeated here as (11.8):
Vo2
hL = K (11.8)
2g
where K = the adjusted minor loss coefficient for the composite energy-loss
method
The adjusted minor loss coefficient K is defined as
K = (C1C2C3 + C4)CB (11.9)
where C1 = coefficient for the relative access-hole diameter

C2 = coefficient for the water depth in the access hole
C3 = coefficient for lateral flow, the lateral angle, and plunging flow
C4 = coefficient for the relative pipe diameters
CB = coefficient for benching (given in Table 11.9)
Empirical equations describing these coefficients were developed from laboratory
studies and analyses. These equations are presented in the subsections that follow.
Computing C1 (Relative Access-Hole Diameter Coefficient). The
energy-loss coefficient relative to the access-hole diameter is related to the ratio of the
access-hole diameter, B, to the outlet-pipe diameter, Do, and is given by
  B 
 0.9  
  Do  , B < 4
 B Do
C1 =  6 + (11.10)
 Do
 B
0.36, ≥4
 Do
Thus, the value of this coefficient increases with the ratio B/Do until that ratio is equal
to 4. For B/Do ≥ 4, C1 has the constant value C1 = 0.36. Note that the value of C1 is the
same for all incoming pipes.
Computing C2 (Water Depth in Access Hole Coefficient) and C3
(Lateral Flow, Lateral Angle, and Plunging Flow Coefficient). The
coefficients C2 and C3 represent the composite effect of all the inflow pipes, the out-
flow pipe, and the access hole. Their calculation is affected by and also affects the cal-
culation of the access-hole water depth, d. Because of this interdependence, an
iterative method is used to calculate these coefficients. The first step is to compute an
initial estimate of d with the equation
Vo2
d = HGLo + C1C B − Zo (11.11)
2g
where HGLo = hydraulic grade elevation at the upstream end of the outlet pipe
(ft, m)
Zo = elevation of the outlet pipe invert, (ft, m)
This value is used to calculate an estimate of C2 as
 2 3
   
0.24  d  − 0.05  d  ,
d
≤3
  Do   Do  D0
C2 =  (11.12)
 d
0.82, D > 3
 o
The analysis of the factors affecting energy losses for lateral flows resulted in an
equation for C3 that is the most complex of any of the coefficients. For a simple, two-
pipe system with no plunging flow, C3 = 1.0. (A pipe has plunging flow if the critical
flow depth elevation in the pipe, yc + Zi, is higher than the access-hole depth eleva-
tion, d + Zo.) Otherwise, C3 is expressed in the form
C3 = 1 + C3A + C3B + C3C + C3D (11.13)
where the individual terms in (11.13) are given by Equations 11.14 through 11.18.
The calculations for C3 consider the angle θi between the inlet and outlet pipes. As
this angle deviates from 180 (straight-line flow), the energy loss increases because
the flow cannot smoothly transition to the outlet pipe. All inflow pipe angles are mea-
sured clockwise from the outlet pipe. The calculation of C3 accounts for inlet-flow
plunging by considering the inlet as a fourth, synthetic inflow pipe with the corre-
sponding angle set to 0 .
C3 can have a value ranging from 1 for no lateral flow to potentially very high values
for greater plunge heights. Because empirical studies do not support this result, a
value of 10 is set as a realistic upper limit on C3.
The term C3A represents the energy loss from plunging flows and is valid for three
inlet pipes plus the plunging flow from the inlet:
4 Q  0.75  0.3 0.3 

   
C3A = ∑  i  1 + 2  Zi − d   Zi   (11.14)
i =1  Qo 
  Do Do   Do  
 
where Q1, Q2, Q3 = discharge from inflow pipes 1, 2, and 3 (cfs, m3/s)
Q4 = discharge into the access hole from the inlet (cfs, m3/s)
Z1, Z2, Z3 = invert elevations of the inflow pipes relative to the outlet pipe
invert (ft, m)
The term C3B represents the energy loss due to change in direction between the inlet
and outlet pipes. If the horizontal momentum check HMCi as computed using (11.15)
is less than 0, then the flow is assumed to be falling from such a height that horizontal
momentum can be neglected and C3B = 0. Otherwise, HMCi is given by (11.15). Note
that the inlet flow is not considered in this calculation.
0.75
 Z  Q 
HMCi = 0.85 −  i  i  (11.15)
 Do  Do 
If HMCi ≥ 0, C3B is computed as
3 cosθi × HMCi
C3B = 4∑
0.3 (11.16)
i =1  d 
 
 Do 
where θ1, θ2, θ3 = angle between the outlet main and inflow pipes 1, 2, and 3,
degrees
HMCi = horizontal momentum check for pipe i computed using (11.15)
If there is more than one inflow pipe, C3C is calculated for all combinations of inflow
pipes with HMCi > 0. The pair that produces the highest value is then used for the cal-
culations of C3C and C3D.
Z Z 
C3C = 0.8  A − B  (11.17)
 Do Do 
where ZA, ZB = invert elevation, relative to the outlet pipe invert, for the inflow
pipes that produce the largest value of C3D
0.75 0.75
Q  Q 
C3D =  A  sin θ A +  B  sin θ B (11.18)
 Qo   Qo 
where QA, QB = discharges for the pair of inflow pipes producing the largest value
of C3D
θA, θB = angle between the outlet main and inflow pipes for the pair of
inflow pipes producing the largest value of C3D
With the initial estimates of C2 and C3, the access-hole depth is recalculated with
V02 V2
d = HGLo + + C1C2C3CB 0 − Z o (11.19)
2g 2g
where HGLo = hydraulic gradeline at the upstream end of the inlet pipe
Zo = invert elevation of the outlet pipe at the upstream end
This new value of d is compared to the previous estimate, and the entire procedure for
calculating C2, C3, and d continues until the method converges on the value of d.
Computing C4 (Relative Pipe Diameter Coefficient). The last coeffi-
cient to be determined in (11.9), the coefficient related to relative pipe diameters, C4,
is computed for each inflow pipe using
Q A V2
C4i = 1 +  i + 2 i cos θi  i (11.20)
2
 Qo Ao  Vo
where Ai, Ao = cross-sectional areas of the inlet and outflow pipes (ft2, m2)
θi = angle between the outlet pipe and inflow pipe i
Vi, Vo = velocity of flow in the inlet pipe and outflow pipes (ft/s, m/s)
Because this term represents an exit loss from each inflow pipe, C4i is set to 0 for each
pipe with plunging flow. In addition, if θi is less than 90° or greater than 270° for any
inlet pipe, the corresponding term cos θi is set to 0 in (11.20). The upper limit for the
value of C4 is 9.0.
Because the complex, iterative nature of the composite energy-loss method, its appli-
cation is not well suited to manual calculations and thus it is not illustrated here
through an example problem. However, it is readily adaptable to computer solutions.
11.5 FINAL DESIGN OF THE PIPE SYSTEM

Following completion of the initial design process as illustrated in Example 11.2, one
proceeds to computation of the hydraulic and energy grade lines (the HGL and EGL)
in order to ensure that the system performs adequately. Computation of the HGL and
EGL elevations is especially important in cases where the storm sewer system outlet
may be submerged, or where the subsurface drainage structures are shallow in depth.
When computing hydraulic and energy grade lines in a system, the direction (that is,
upstream or downstream) in which the calculations should proceed depends on
whether a pipe has surcharging or open-channel flow conditions. This is discussed
further in the subsections and examples that follow.
Surcharged Pipes
In a surcharged pipe, the HGL is above the crown of the pipe (there is pressure flow)
and calculations proceed upstream from a known tailwater condition at the outfall
point. Checks must be made throughout the calculations to ensure that the HGL
remains above the crown of the pipe. If it drops below the pipe crown at any point, the
condition should be treated as open-channel flow.
Energy losses, whether they refer to frictional losses along pipes or to minor losses at
structures, apply to the EGL rather than the HGL. Therefore, computation of the HGL
elevations in a storm sewer system is actually accomplished by first computing the
EGL elevations. Once these elevations are known, HGL elevations may be found by
subtracting the velocity head V2/2g. EGL elevations must always decrease in a down-
stream direction, and must always increase in an upstream direction. However, HGL
elevations, depending on relative velocities in the pipes connected to one another at a
structure, may increase or decrease at structure locations regardless of the direction
considered. For example, an HGL elevation may increase in a downstream direction
within a pipe if there is a hydraulic jump.
Section 11.5 Final Design of the Pipe System 445
In design of a surcharged storm sewer system, the HGL must be maintained at eleva-
tions below the ground surface, especially at locations of manholes and inlets. If the
HGL elevation is higher than the ground surface at a manhole or inlet location, it indi-
cates that, for the design storm event, stormwater would exit the system through the
top of the structure and flow onto the land surface. In extreme cases, these high HGL
elevations can cause popping of manhole covers due to excessive pressures in the
storm sewer. At inlet locations, it is recommended that the HGL elevation be no
higher than about 6 in. (15 cm) below the grate elevation or the elevation of the bot-
tom of the throat. [Local regulations may require more than 6 in. 15 cm).] Otherwise,
the hydraulic behavior of the grate or throat may be impacted. When making HGL
checks at inlet locations, conservative design practice is to use the highest HGL eleva-
tion computed for the structure.
Example 11.4 illustrates final design calculations for the storm sewer system origi-
nally presented in Example 11.2, where discharges and pipe sizes were computed.
Because of the high tailwater on the system, all of the pipes will operate under pres-
sure-flow conditions. The basic steps involved in the final design calculations for the
pipe system are:
1. Compute the EGL in the outfall channel and the EGL at the downstream end of
the outfall pipe. Consider minor losses at the pipe exit. Typically, exit losses are
computed using (6.25) with K = 1.0:
V22 − V12
hL = 1.0
2g
2. Compute the velocity head in the most downstream pipe and subtract this value
from the EGL to compute the HGL at the downstream end of the pipe.
3. Compute the head loss in the most downstream pipe, and add this value to the
EGL at the downstream end of the pipe to compute the EGL at the upstream end
of the pipe. Subtract the velocity head from step 2 to find the HGL at this locaton.
4. Use a method such as the energy-loss or composite energy-loss method to com-
pute the head loss through the structure for each of the incoming pipes. Add these
values to the EGL of the outgoing pipe to determine the EGL at the downstream
end of each of the incoming pipes.
5. Compute the velocity head for each of the incoming pipes, and subtract this value
from the EGL at the downstream end of each pipe to determine the HGL at these
locations.
6. If an HGL elevation is higher than the allowable maximum elevation necessary to
maintain minimum freeboard at a structure, increase the size of the outgoing
pipe, redo the affected calculations, and recheck the HGL elevations.
7. If the structure has more than a single incoming pipe, select one of the incoming
branches and continue the calculations by repeating steps 3 through 7 until the
EGL and HGL elevations for the entire branch have been determined.
8. Use Equation 6.25 with K = 0.5 to compute the pipe entrance loss at terminal
upstream structures, and compute the EGL and HGL at this location. Compare
this value for the HGL (which was computed assuming outlet-control conditions)
with the inlet-control HGL computed using methods presented in Section 9.5.
The higher of these elevations controls the value of the HGL at this location.
9. Repeat steps 3 through 8 until the entire system has been evaluated.
10. Check to ensure that all velocities are within acceptable limits and adjust pipe
sizes if necessary.
Example 11.4 – Final Design Process for a System with Pressure Flow. Com-
pute the HGL and EGL elevations for the storm sewer system from Example 11.2 (page 427) and
shown again in Figure E11.4.1. Backwater from a culvert downstream of the outfall causes the tailwa-
ter at the storm sewer outfall point to be at an elevation of 209.70 ft. The tailwater velocity in the
channel downstream of the outfall is 2 ft/s. Use the energy-loss method to evaluate minor losses at
structures, and assume that the inlets are constructed with no benching of their inverts.
Washington Street
Connecticut Avenue
Delaware Avenue
Colorado Avenue
Dakota Avenue
Basin A Basin B Basin C
Pipe 1 Pipe 2
Inlet 1 Inlet 2 Inlet 3
Jefferson Street
Pipe 3 Pipe 4
Basin D
Inlet 4 Outfall
Open Channel
Figure E11.4.1 Plan view of storm sewer system for and Example 11.4
Pipe data
The pipe lengths, diameters, discharges, and invert elevations given or computed in Example 11.2 will
be used in the calculations. This information is presented again in the following table.
Upstream Invert Downstream Invert Discharge Q

Pipe Label Length L (ft) Diameter (in.) Elevation (ft) Elevation (ft) (cfs)
Pipe 1 70 21 207.33 206.98 7.6
Pipe 2 315 21 208.56 206.98 7.6
Pipe 3 40 30 206.23 206.03 22.0
Pipe 4 40 30 205.93 205.73 24.0
Inlet data
The curb inlets designed in Example 11.2 will be used for the calculations. The inlet boxes are rectan-
gular with a length equal to the inlet throat opening length and a width (perpendicular to the gutter
flow line) of 3 ft. Information for individual inlets is given in the following table.
Bottom of Throat Equivalent

Inlet Label Elev. (ft) Length L (ft) Diameter Ba (ft)
Inlet 1 212.20 15 7.6
Inlet 2 211.73 15 7.6
Inlet 3 213.50 15 7.6
Inlet 4 211.85 3 3.4
a. Equivalent diameter B for inlet boxes is computed as B = (4A/π)1/2, where A = 3L for

3-ft-wide boxes.
Solution: Because the sewer outfall is submerged by the tailwater, all or part of the storm sewer sys-
tem will be surcharged and calculations should proceed upstream through the system from the outfall
point. Following are the calculations for HGL and EGL through the system, and the results are tabu-
lated in the table following the calculations.
EGL and HGL at outfall
The HGL immediately downstream of the outfall point is the same as the tailwater elevation in the
channel, or HGL = 209.70 ft. Thus, the EGL immediately downstream of the outfall point is
EGL = HGL + V2/(2g) = 209.70 + 22/[2(32.2)] = 209.76 ft

The exit loss at the outfall point can be determined using (6.25), where V1 is the velocity in Pipe 4 and
V2 = 2 ft/s is the tailwater velocity. The minor loss coefficient K = 1.0. Because Pipe 4 will be flowing
full due to the tailwater elevation, its velocity is
Q 4Q 4(24)
V= = = = 4.89 ft/s
A π D 2 2.52 π
Therefore, the exit loss is
V22 − V12 4.892 − 22

hL = K = 1.0 = 0.31 ft
2g 2(32.2)
The EGL immediately upstream of the outfall point (at the downstream end of Pipe 4) is equal to the
EGL downstream of the outfall point plus the head loss, or
EGL = 209.76 + 0.31 = 210.07 ft
The velocity head for Pipe 4 is
V2/(2g) = 4.892/[2(32.2)] = 0.37 ft

The HGL at the downstream end of Pipe 4 is therefore
HGL = EGL − V2/(2g) = 210.07 – 0.37 = 209.70 ft

EGL and HGL at Inlet 4
The EGL at the upstream end of Pipe 4 (at the outlet of Inlet 4) is computed as the sum of the EGL
elevation at the downstream end of Pipe 4, plus frictional losses that occur in Pipe 4. The slope of the
EGL for Pipe 4 can be evaluated by rearranging the Manning equation ((7.9)), and the frictional loss
can be determined from (7.8) as the product of the EGL slope and the pipe length:
(nV )2 (nV )2 [0.013(4.89)]2

Sf = = = = 0.00341
1.492 R 4 / 3 1.492 ( D / 4)4 / 3 2.22(2.5 / 4)4 / 3
hL = SfL = 0.00341(40) = 0.14 ft
The EGL elevation at the upstream end of Pipe 4 is

EGL = 210.07 + 0.14 = 210.21 ft
Because the velocity head in Pipe 4 is 0.37 ft, the HGL elevation at the upstream end of Pipe 4 is
HGL = EGL − V2/(2g) = 210.21 − 0.37 = 209.84 ft

This value is higher than the pipe crown elevation, confirming that Pipe 4 flows full for its entire
length.
The minor loss at Inlet 4 is evaluated using Equations 11.1 and 11.2. The initial loss coefficient Ko is
found using (11.3) with Do = 30 in. = 2.5 ft and θ = 180°. The equivalent diameter of the inlet box is
3.4 ft. Therefore,
0.15
 B   B 
K o = 0.1  (1 − sin θ ) + 1.4   sin θ
 DO   DO 
0.15
 3.4   3.4 
= 0.1  (1 − sin180° ) + 1.4   sin180° = 0.14
 2.5   2.5 
The pipe diameter correction factor CD1 = 1 because the predicted depth in the inlet will not be
greater than 3.2 times the outlet pipe diameter. The correction factor for flow depth, CD2, is evaluated
from (11.5) with d being the difference in elevation between the HGL and pipe invert elevations at the
upstream end of Pipe 4:
CD2 = 0.5(d/Do)0.6 = 0.5[(209.84 – 205.93)/2.5]0.6 = 0.65

The correction factor CQ is equal to 1 because there is only one inflow pipe to Inlet 4.
Because flow plunges into Inlet 4 from its throat opening, the factor CP must be determined using
(11.7). In that expression, d is 3.91 ft (as above) and h is the elevation difference between the bottom
of the inlet throat and the centerline of the outlet pipe, or
h = 211.85 – 207.18 = 4.67 ft
Therefore,
 h  h − d   4.67   4.67 − 3.91 

CP = 1 + 0.2  
  = 1 + 0.2  2.5    = 1.11
 Do  Do  2.5 
The correction factor for benching is CB = 1 because no benching is provided in the inlet.
Applying Equations 11.1 and 11.2, the minor loss in Inlet 4 is
Vo2
hL = K oCD1CD 2CQ CP CB = 0.14(1)(0.65)(1)(1.11)(0.37) = 0.04 ft
2g
The EGL elevation at the downstream end of Pipe 3 is equal to the EGL at the upstream end of Pipe 4,
plus the minor loss, and is therefore
EGL = 210.21 + 0.04 = 210.25 ft
The diameter of Pipe 3 is 30 in. = 2.5 ft. The velocity for this pipe is
Q 4Q 4(22)
V= = = = 4.48 ft/s
A π D 2 2.52 π
The HGL at the downstream end of Pipe 3 is then computed as
HGL = EGL – V2/2g = 210.25 – 4.482/[2(32.2)] = 209.94 ft

The highest of the HGL elevations computed at Inlet 4 is HGL = 209.94 ft. This elevation is 1.91 ft
below the bottom of the throat elevation; therefore, the hydraulic behavior of the inlet opening will
not be affected.
HGL and EGL at Inlet 2

The HGL and EGL elevations at the upstream end of Pipe 3 (at the outlet of Inlet 2) are higher than
the corresponding values for the downstream end of Pipe 3 by the amount hL = SfL. The pipe length is
L = 40 ft and, from the Manning equation,
(nV )2 [0.013(4.48)]2
Sf = = = 0.00286
2 4/3
1.49 ( D / 4) 1.492 (2.5 / 4)4 / 3
Therefore, hL = 0.11 ft and the EGL and HGL elevations are
EGL = 210.25 + 0.11 = 210.36 ft
HGL = 210.36 − 4.482/[2(32.2)] = 210.05 ft

Minor losses in Inlet 2 are evaluated using Equations 11.1 through 11.7 with
d = 210.05 – 206.23 = 3.82 ft
and
h = 211.73 – 207.48 = 4.25 ft
The minor losses must be evaluated separately for each of the two inlet pipes.
For inlet Pipe 1, θ = 90°, Ko = 1.65,CD1 = 1, CD2 = 0.64, CP = 1.06, and CB = 1.
The factor CQ is evaluated using QI = 7.6 cfs and Qo = 22 cfs:
0.75 0.75
 Q   7.6 
CQ = 1 + (1 − 2sin θ )  1 − I  = 1 + (1 − 2sin 90°) 1 −  = 0.27
 Qo   22 
Therefore, the minor loss to be used for Pipe 1 is
hL = 1.65(1)(0.64)(0.27)(1.06)(1)[4.482/2(32.2)] = 0.09 ft
and the EGL elevation at the downstream end of Pipe 1 is
EGL = 210.36 + 0.09 = 210.29 ft
The diameter of Pipe 1 is 21 in. = 1.75 ft. The velocity for this pipe is
Q 4Q 4(7.6)
V= = = = 3.16 ft/s
A π D 2 1.752 π
The HGL elevation at the downstream end of Pipe 1 is therefore
HGL = EGL – V2/2g = 210.45 – 3.162/[2(32.2)] = 210.29 ft

For Pipe 2, θ = 90°, Ko = 1.65, CD1 = 1, and CD2 = 0.64. The factor CQ is evaluated using QI = 7.6
cfs and Qo = 22 cfs, and is 0.27. The factor CP = 1.06 and CB = 1. Therefore, the minor loss to be
used for Pipe 2 is
hL = 1.65(1)(0.64)(0.27)(1.06)(1)[4.482/2(32.2)] = 0.09 ft
and the EGL elevation at the downstream end of Pipe 2 is
EGL = 210.36 + 0.09 = 210.45 ft
The diameter, flow, and therefore velocity are the same for Pipe 2 as for Pipe 1. The HGL elevation at
the downstream end of Pipe 2 is
HGL = EGL – V2/2g = 210.45 – 3.162/2(32.2) = 210.29 ft

Note that, in this example, the minor loss applicable to each of the incoming pipes (Pipes 1 and 2) was
identical. In general, this will not be true (see Example 11.3), but, in this case, the inlet pipes were of
the same size, with the same flow rates and identical geometric orientations relative to the outflow
pipe.
The highest of the HGL elevations in Inlet 2 is HGL = 210.29 ft. This elevation is 1.44 ft below the
inlet throat elevation, and thus the hydraulic behavior of the throat will not be affected.

The EGL and HGL elevations immediately downstream of Inlet 1 (at the upstream end of Pipe 1) are
determined by computing frictional losses in Pipe 1, hL = SfL. First, the friction slope is computed as
( nV ) 2 [0.013(3.16)]2
Sf = = = 0.00227
1.49 2 ( D / 4) 4 / 3 1.49 2 (1.75 / 4) 4 / 3
The head loss in Pipe 1 is therefore
hL = SfL = 0.00227(70) = 0.16 ft
Thus, the HGL and EGL elevations are
EGL = 210.45 + 0.16 = 210.61 ft
HGL = 210.61 − 3.162/[2(32.2)] = 210.45 ft
The minor loss due to the pipe entrance within the inlet can be evaluated using (6.26). Assuming that
the pipe entrance has a square edge, Table 6.5 on page 203 yields K = 0.5. Thus, the minor loss is
hL = KV2/(2g) = 0.5(3.15)2/[2(32.2)] = 0.08 ft
The EGL elevation inside Inlet 1 is therefore
EGL = 210.61 + 0.08 = 210.69 ft
Assuming, conservatively, that the velocity of the flow inside Inlet 1 is negligible, the HGL elevation
is the same as the EGL, and is HGL = 210.69 ft.
To check that inlet control does not control the HGL elevation within Inlet 1, the inlet control chart
(Figure 9.12) is employed with D = 21 in = 1.75 ft and Q = 7.6 cfs to obtain HW/D = 0.85 from the
scale representing a squared-edged entrance. Thus, under inlet control, the water depth in the inlet
would be
HW = 0.85D = 0.85(1.75) = 1.49 ft
Adding this to the invert elevation of the inlet yields a HGL elevation of HGL = 208.82 ft. This is
lower than the value of HGL = 210.69 ft determined above, and thus the actual HGL is not governed
by inlet control.
The controlling HGL in Inlet 1 is governed by downstream conditions, and HGL = 210.69 ft. This is
1.51 ft below the throat elevation, and hence the hydraulic behavior of the throat will not be affected.
The EGL and HGL elevations immediately downstream of Inlet 3 (at the upstream end of Pipe 2) are
determined by computing frictional losses in Pipe 2. The friction slope for Pipe 2 is the same as for
Pipe 1, so Sf = 0.00227. The head loss in Pipe 2 is therefore
hL = SfL = 0.00227(315) = 0.72 ft
and thus the HGL and EGL elevations are
EGL = 210.45 + 0.72 = 211.17 ft
HGL = 211.17 − 3.162/[2(32.2)] = 211.01 ft
The minor loss due to the pipe entrance within the inlet can be evaluated using (6.26). Assuming that
the pipe entrance has a square edge, Table 6.5 yields K = 0.5. Thus, the minor loss is
hL = 0.5(3.16)2/[2(32.2)] = 0.08 ft
The EGL elevation inside Inlet 1 is therefore
EGL = 211.17 + 0.08 = 211.25 ft
Assuming, conservatively, that the velocity of the flow inside Inlet 3 is negligible, the HGL elevation
is the same as the EGL and is HGL = 211.25 ft.
A check of inlet control yields HW = 1.49 ft and HGL = 210.05 ft. This is lower than the controlling
HGL = 211.25 ft, and thus inlet control does not govern the HGL elevation. The governing HGL =
211.25 ft is 2.25 ft below the throat elevation, and hence the hydraulic behavior of the throat will not
be affected. The following table summarizes the results of this example. Figure E11.4.2 is a profile of
the storm sewer system illustrating the computed HGL and EGL
elevations.
Junction Downstream Velocity Upstream Velocity

EGL (ft) HGL (ft) EGL (ft) HGL (ft)
Structure Structure (ft/s) Structure (ft/s)
Outfall
Outfall 209.76 2.00 209.70 Pipe 4 210.07 4.89 209.70
Channel
Inlet 4 Pipe 4 210.21 4.89 209.84 Pipe 3 210.25 4.48 209.94
Pipe 1 210.45 3.15 210.29
Inlet 2 Pipe 3 210.36 4.48 210.05
Pipe 2 210.45 3.15 210.29
Inlet 1 Pipe 1 210.61 3.15 210.45 None 210.69 0.00 210.69
Inlet 3 Pipe 2 211.17 3.15 211.01 None 211.25 0.00 211.25
Figure E11.4.2 Profile of HGL for system in Example 11.4

Open-Channel Flow
If a pipe conveys open-channel flow, calculations proceed in an upstream direction if
the flow is subcritical or in a downstream direction if the flow is supercritical.
Hydraulic jumps, if they occur in the system, may be analyzed using the concept of
specific force discussed in Section 7.5 on page 258. Calculations are complicated by
the need to compute gradually-varied flow profiles, which require substantial amounts
of information on cross-sectional areas, hydraulic radii, and so forth for various
depths of flow. These relationships are complex for pipe cross-sections, and make the
HGL and EGL calculations quite laborious. Computer modeling tools capable of per-
forming the tedious calculations are obviously desirable in such instances.
Example 11.5 demonstrates the evaluation of the EGL and HGL for the final design of
a storm-sewer system with open-channel flows. The basic steps are as follows:
1. Compute normal depth and critical depth for every pipe in the system for use in
classifying the profile type for each pipe.
2. Determine whether the flow profiles for the system will be subcritical, supercriti-
cal, or mixed. Subcritical flow profiles will be evaluated from the downstream
end to the upstream end; the reverse is true for supercritical profiles.
3. If the flow in the outfall pipe is subcritical, start at the downstream end of the sys-
tem and compute the EGL in the outfall channel and the EGL at the downstream
end of the outfall pipe. Consider minor losses at the pipe exit. Typically, exit
losses are computed using (6.25) with K = 1.0:
V22 − V12
hL = 1.0
2g
If the flow is supercritical, assume that the depth at the upstream end of the pipe
is equal to critical depth. Determine the entrance losses at the upstream end using
the appropriate entrance loss coefficient.
4. Compute the HGL and EGL at the downstream (upstream for supercritical) end
of the pipe.
5. Use the direct-step or standard-step method (from Section 7.8) to evaluate the
flow profile through the pipe.
6. Compute the HGL and EGL at the upstream (downstream for supercritical) end
of the pipe.
7. Use a method such as the energy-loss or composite energy-loss (subcritical only)
method to compute the head loss through the structure for each of the incoming
pipes. Use these values to determine the EGL and HGL at the downstream end of
the incoming pipe (upstream end of outgoing pipe for supercritical).
8. If an HGL elevation is higher than the allowable maximum elevation necessary to
maintain minimum freeboard at a structure, increase the size of the outgoing
pipe, redo the affected calculations, and recheck the HGL elevations.
9. For subcritical flow, if the structure has more than a single incoming pipe, select
one of the incoming branches. Continue the calculations by repeating steps 4
through 9 until the EGL and HGL elevations for the entire branch have been
determined.
10. Use (6.25) with K = 0.5 to compute the pipe entrance loss at terminal upstream
structures, and compute the EGL and HGL at this location. Compare this value
for the HGL (which was computed assuming outlet-control conditions) with the
inlet-control HGL computed using methods presented in Section 9.5. The higher
of these elevations controls the value of the HGL at this location.
11. Repeat steps 3 through 11 until the entire system has been evaluated.
12. Check to ensure that all velocities are within acceptable limits and adjust pipe
sizes if necessary.
Example 11.5 – Final Design Process for a System with Open-Channel Flow.
Compute the EGL and HGL elevations for the system from Examples 11.2 on page 427 and 11.4 on
page 446. However, assume this time that the tailwater depth in the downstream channel is 2 ft. The
tailwater velocity is 4.5 ft/s.
Solution: In this example, the tailwater elevation is lower than it was in Example 11.3, where all
pipes operated under pressure flow conditions. For this example, the flow profile types must be classi-
fied and the HGL determined using procedures for evaluating open-channel flow profiles with the
Manning equation and minor losses with the energy-loss method. Profiles showing HGL through the
system are given in Figure E11.5.3.
Computing normal and critical depths
The solution procedure begins with the determination of normal and critical depths for each pipe.
This determination permits the pipes to be classified as to whether they are hydraulically steep, mild,
etc., and hence provides guidance as to the types of flow profiles that will occur in the pipes.
Pipe 1 (D = 1.75 ft) conveys a discharge of Q = 7.6 cfs with n = 0.013 and So = 0.005. Geometric
properties of the circular pipe were provided in Table 7.1 and are repeated here in Figure E11.5.1 and
in Table 11.10.
Figure E11.5.1 Definition figure for open-channel flow in a circular channel
Table 11.10 Geometric properties of circular open channels
Area (A) Wetted Perimeter (P) Hydraulic Radius (R) Top Width (T) Hydraulic Depth (Dh)
D 2 (φ − sin φ ) Dφ D  sin φ  φ 
D sin   D  φ − sin φ 
1 −   
8 2 4 φ  2 8  sin (φ / 2 ) 
Applying (7.23) from page 235 to first determine φ (trial and error required), the normal depth yn can
be determined using (7.20). (7.23) is
5/3 −2 / 3
1.49  D 2   Dφ 
Q= 
n  8
( φ − sin φ )    So1/ 2
  2 
and (7.20) is
D  φ 
y= 1 − cos   
2   2 
Alternatively, Figure 7.8 (repeated here as Figure E11.5.2) can be used to find normal depth from the
full-flow capacity and velocity for the pipe. For instance, the diameter of Pipe 1 is 21 in., its discharge
(from Example 11.4) is 7.6 cfs, and its slope is 0.005. Applying the Manning equation, the full-flow
capacity of the pipe is
2/3
1.49 1.49 π (1.752 )  1.75 
Q full = AR 2 / 3So1/ 2 =   0.0051/ 2 = 11.23 cfs
n 0.013 4  4 
The ratio Q/Qfull is 7.6/11.23 = 0.68. Reading from Figure E11.5.2, y/D = 0.605. Therefore,
y = 0.605(1.75) = 1.06 ft
1.2
1.0
V/Vfull
0.8
Q/Q full and V/V full
0.6
Q/Q full
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
y/D
Figure E11.5.2 Dimensionless graph of discharge and velocity fractions vs. depth fraction for
circular pipe
(7.20) and (7.23) can be used to verify that 1.06 ft is the normal depth. The same procedure is fol-
lowed to obtain the normal depths for the other three pipes. These depths are given in the table that
follows.
The critical depth in each pipe is found using (7.38) (page 247):
Q 2T
=1
gA3
By assuming a trial depth, one can use (7.20) to determine φ, followed by use of table to determine the
top width and cross-sectional area. These values are then inserted into (7.38), and the solution pro-
ceeds in a trial-and-error fashion until (7.38) converges. For example, for Pipe 1, critical depth is
assumed to be 1.0. (7.20) is used to find the corresponding value of φ:
1.75   φ 
1= 1 − cos   
2   2 
The value of φ is found to be 196.43º = 3.428 rad. A and T can then be computed as
D 2 (φ − sin φ ) 1.752 (3.428 − sin 3.428)

A= = = 1.42 ft 2
8 8
φ   3.428 
T = D sin   = 1.75sin   = 1.73 ft
2  2 
Checking this assumption,
Q 2T 7.6 2 (1.73)
= = 1.08
3 32.2(1.42)
gA
Because 1.08 is not equal to 1.0, another depth should be assumed and the process repeated. In this
manner, yc for Pipe 1 is found to be 1.02 ft. The same procedure can be used to compute the critical
depths for the remaining pipes, which are listed in the following table.
Pipe yn (ft) yc (ft)

1 1.06 1.02
2 1.06 1.02
3 1.63 1.60
4 1.73 1.67
Note that the normal depth is larger than the critical depth in all four pipes; thus, all four pipes have
hydraulically mild slopes. Gradually varied flow profiles in the pipes will be of the M types (M1, M2,
or M3). M1 and M2 profiles are subcritical flows, and hence are controlled by downstream conditions
(calculations proceed upstream). An M3 profile has supercritical flow, and is controlled by upstream
conditions (calculations proceed downstream). Detailed water surface profile calculations follow for
each pipe and structure.
EGL and HGL at the Outfall
The tailwater depth is 2 ft and the tailwater velocity is 4.5 ft/s. Its velocity head is
V2/2g = 4.52/2(32.2) = 0.31 ft

The HGL (water surface) elevation just downstream of the outfall point is
HGL = 205.73 + 2.00 = 207.73 ft
and the EGL elevation is
EGL = 207.73 + 0.31 = 208.04 ft
The minor energy loss that occurs at the outfall can be evaluated using (7.16) (page 230) with K = 1.0.
Writing Bernoulli’s equation ((7.3), page 223) for points just upstream (section 1) and just down-
stream (section 2) of the outfall, and assuming α = 1 and V1 > V2, one obtains
V 12 V 2 V 22 − V 12
y1 + = y2 + 2 + K
2g 2g 2g
V 12 4 .5 2 4 .5 2 V 2
y1 + = 2 .0 0 + + 1 .0 − 1
2g 2 (3 2 .2 ) 2 (3 2 .2 ) 2g
V12 V2
y1 + = 2.00 + 0.31 + 1 − 0.31
2g 2g
y1 = 2.00 ft
Cancelling the velocity head for section 1 from both sides, this expression yields y1 = 2.00 ft. There-
fore, A1 = 4.21 ft2 and V1 = 24/4.21 = 5.70 ft/s, which is greater than the downstream channel veloc-
ity of 4.5 cfs, so the assumption that V1 > V2 is correct.
Immediately upstream of the outfall point in Pipe 4, the water surface (HGL) elevation is HGL =
207.73 ft and the EGL elevation is
EGL = 207.73 + 5.702/2(32.2) = 208.23 ft

EGL and HGL through Pipe 4
Because Pipe 4 has a mild (M) slope and the flow depth of y = 2.00 ft at its downstream end is larger
than both normal and critical depths, the water surface profile within Pipe 4 is of the M1 type (see
Section 7.8). Calculation of the M1 profile can be accomplished via the direct-step method, and is
summarized in the following table (see Example 7.20 for an illustration of the detailed calculations
required).
Distance
y (ft) V (ft/s) E (ft) Upstream (ft)
2.00 5.70 2.50 0
1.98 5.76 2.49 7
1.96 5.81 2.48 15
1.94 5.87 2.48 22
1.92 5.93 2.47 30
1.90 6.00 2.46 37
1.88 6.06 2.45 45
Pipe 4 has a length of 40 ft, and thus the depth and specific energy at its upstream end are y = 1.89 ft
and E = 2.46 ft. The water surface (HGL) elevation at its upstream end is
HGL = 205.93 + 1.89 = 207.82 ft
and the EGL elevation is
EGL = 205.93 + 2.46 = 208.39 ft
The minor energy loss in Inlet 4 can be evaluated using Equations 11.1 through 11.7 with θ = 180°, d
= 1.89 ft, h = 211.85 – 207.18 = 4.67 ft, and B = 3.4 ft. The factor Ko = 0.14 and CD1 = 1. The factor
CD2 = 0.42 and CQ = 1. Because there is a plunging flow from the inlet throat, the factor CP = 1.42.
With no benching of the invert, CB = 1. The velocity for a depth of 1.89 ft is 6.02 ft/s. Therefore, the
minor loss is
hL = 0.14(1)(0.42)(1)(1.42)(1)(6.022/64.4) = 0.05 ft
The EGL elevation at the outlet of Pipe 3 is
EGL = 208.39 + 0.05 = 208.44 ft
Subtracting the invert elevation of Pipe 3, the specific energy at the downstream end of Pipe 3 is
E = 208.44 – 206.03 = 2.41 ft
By trial and error, the depth at the downstream end of Pipe 3 corresponding to this specific energy
level is y = 1.97 ft. The HGL elevation at the downstream end of Pipe 3 is therefore
HGL = 206.03 + 1.96 = 208.00 ft
Because the depth of 1.90 ft at the downstream end of the pipe is larger than both the normal and crit-
ical depths, the profile is of the M1 type. Profile calculations are summarized in the following table:
Distance
1.97 5.30 2.41 0
1.95 5.36 2.40 6
1.93 5.41 2.38 12
1.91 5.47 2.37 18
1.89 5.53 2.36 24
1.87 5.59 2.35 30
1.85 5.65 2.35 36
1.83 5.71 2.34 42
Because Pipe 3 is 40 ft long, the depth at its upstream end is y = 1.84 ft and the specific energy at its
upstream end is E = 2.34 ft. Therefore, the HGL and EGL elevations at its upstream end are
HGL = 206.23 + 1.84 = 208.07 ft
EGL = 206.23 + 2.34 = 208.57 ft
At Inlet 2, there are two incoming pipes (Pipes 1 and 2). The minor energy loss needs to be evaluated
separately for each.
For Pipe 1, θ = 90°, d = 1.84 ft, B = 7.6 ft, and
h = 211.73 – 207.48 = 4.25 ft
The factor Ko = 1.65 and CD1 = 1. The factor CD2 = 0.42 and CQ = 0.27. Because there is plunging
flow from the inlet throat, the factor CP = 1.36. With no benching of the invert, CB = 1. Therefore, the
minor loss is
hL = 1.65(1)(0.42)(0.27)(1.36)(1)(5.682/64.4) = 0.13 ft
In this example, Pipes 1 and 2 have identical discharges relative to the outlet discharge and identical
geometries. Therefore, the minor energy losses for these pipes are the same.
The EGL elevation at the outlets of Pipes 1 and 2 is therefore
EGL = 208.57 + 0.13 = 208.70 ft
Subtracting from this the invert elevations of those pipes, the specific energy at the downstream end of
each pipe is E = 1.72 ft. By trial and error, the depth at the downstream ends of Pipes 1 and 2 corre-
sponding to this specific energy level is y = 1.54 ft. The HGL elevation at the downstream ends of
Pipes 1 and 2 is therefore
HGL = 206.98 + 1.54 = 208.52 ft
The depth y = 1.54 ft at the downstream end of Pipe 2 is larger than both the normal and critical
depths, so the profile is of the M1 type. The calculations are summarized in the following table.
Because the length of Pipe 2 is 315 ft, which is longer than the 150-ft M1 profile, the actual depth of
flow at points further than 150 ft from the downstream end of the pipe is y = yn = 1.06 ft. Therefore,
the HGL and EGL elevations at the upstream end of Pipe 2 are
HGL = 208.56 + 1.06 = 209.62 ft
EGL = 208.56 + 1.45 = 210.01 ft
Distance
1.54 3.39 1.72 0
1.49 3.48 1.68 14
1.44 3.59 1.64 28
1.39 3.71 1.60 41
1.34 3.85 1.57 55
1.29 4.00 1.54 69
1.24 4.17 1.51 83
1.19 4.36 1.49 98
1.14 4.58 1.47 114
1.09 4.83 1.45 133
1.06 (= yn) 4.99 1.45 150

Because the HGL and EGL elevations at the downstream end of Pipe 1 are the same as those for Pipe
2, and because the pipes and their discharges are identical, the water surface profile computed for
Pipe 2 is also applicable to Pipe 1.
Pipe 1 has a length of 70 ft, and thus the depth at its upstream end is 1.28 ft. The specific energy at its
upstream end is 1.53 ft. The HGL and EGL elevations at its upstream end are
HGL = 207.33 + 1.29 = 208.62 ft
EGL = 207.33 + 1.54 = 208.87 ft
The minor loss at the pipe entrance in Inlet 1 can be evaluated using (7.16) (page 230) with K = 0.50
(square-edged pipe entrance). The velocity in Pipe 1 immediately downstream of the inlet is 4.03 ft/s.
Therefore, the minor loss is
hL = KV2/2g = 0.50(4.00)2/2(32.2) = 0.12 ft

Adding this value to the EGL at the upstream end of Pipe 1, and assuming (conservatively) that the
velocity inside the inlet is negligible, the HGL and EGL elevations inside Inlet 1 are
HGL = EGL = 208.87 + 0.12 = 208.99 ft
From Example 11.4, the HGL and EGL elevations in Inlet 1, if inlet control were governing, would be
HGL = EGL = 208.82 ft. Because this elevation is lower than the computed value of 208.99 ft, inlet
control does not govern, and the final HGL and EGL for Inlet 1 are 208.99 ft.
The minor loss at the pipe entrance in Inlet 3 can be evaluated using (7.16) with K = 0.50 (square-
edged pipe entrance). The velocity in Pipe 3 immediately downstream of the inlet is 4.99 ft/s. There-
fore, the minor loss is
hL = 0.50(4.99)2/2(32.2) = 0.19 ft
Adding this value to the EGL at the upstream end of Pipe 2, and assuming (conservatively) that the
velocity inside the inlet is negligible, the HGL and EGL elevations inside Inlet 3 are
HGL = EGL = 210.01 + 0.19 = 210.20 ft
From Example 11.4 (page 446), the HGL and EGL in Inlet 1, if inlet control were governing, would
be HGL = EGL = 210.05 ft. Because this elevation is lower than the computed value of 210.20 ft, inlet
control does not govern, and the final HGL/EGL elevation for Inlet 1 is 210.20 ft.
The example results are summarized in the following table. Profiles showing the HGL through the
system are given in Figure E11.5.3.
Downstream Upstream
Junction Downstream Velocity Upstream Velocity
Structure Structure EGL (ft) (ft/s) HGL (ft) Structure EGL (ft) (ft/s) HGL (ft)
Outfall
Outfall 208.04 4.50 207.73 Pipe 4 208.23 5.70 207.73
Channel
Inlet 4 Pipe 4 208.39 6.02 207.82 Pipe 3 208.44 5.30 208.00
Pipe 1 208.70 3.39 208.52
Inlet 2 Pipe 3 208.57 5.68 208.07
Pipe 2 208.70 3.39 208.52
Inlet 1 Pipe 1 208.87 4.00 208.62 None 208.99 0.00 208.99
Inlet 3 Pipe 2 210.01 4.99 209.62 None 210.20 0.00 210.20
Figure E11.5.3 Profiles showing HGL for Example 11.5

External Forces
External forces on a buried pipe are caused by a combi- of live loads that might be applied at the ground sur-
nation of dead loads due to the weight of the soil and face. The external forces and stresses also depend on
other materials overlying the pipe, and live loads whether the pipe is installed in a trench excavated into a
caused by traffic loading on the ground surface. These natural ground surface, or whether it is installed in a
external forces may cause both compressive and bend- trench excavated into a constructed embankment.
ing stresses to develop in pipe walls, and can cause a Because of the variety of pipe materials and installation
pipe to collapse if its strength is insufficient. Because of
methods encountered in practice, estimation of the
these adverse effects, some review agencies specify cri-
required pipe strength is generally an empirical pro-
teria pertaining to minimum and/or maximum pipe cess. For additional information on this subject, readers
burial depths.
are referred to literature and design guidelines pub-
The external forces and corresponding stresses on a lished by associations of pipe manufacturers such as the
buried pipe depend on many things, including the pipe American Concrete Pipe Association (ACPA) and the
diameter and/or shape, the pipe material, the width of American Iron and Steel Institute (AISI). Manufactur-
the trench, the burial depth, the specific weight of the ers and suppliers of plastic pipe materials may also be
backfill material and its degree of compaction, the way consulted for guidance.
in which the pipe is bedded in the trench, and the types
Structural Loads and Stresses in Pipes

In addition to determining the required diameters and layout of the pipes and/or cul-
verts in a project, the engineer must consider loads and stresses on the conduits so that
they will be structurally sound. In particular, the engineer needs to consider issues
such as pipe materials and pipe wall thicknesses. In the case of concrete pipe, for
example, the engineer needs to specify the pipe class (a measure of its strength) and
whether it should be reinforced.
Fundamental to the problem of specifying pipe materials and strength classes is an
analysis of the loads on a pipe and the resulting stresses on the pipe walls. Loads on
pipes may be caused by internal forces (fluid pressure or other momentum-related
hydraulic forces; see “Internal Pressure Forces” in Chapter 13 on page 557), which
tend to cause tensile stresses in pipe walls, or by external forces, which tend to cause
compressive and/or bending stresses to develop in pipe walls. External forces may
arise from dead loads due to overlying soil used to backfill pipe trenches, or from live
loads due to traffic on the ground surface. Stresses in pipes may also develop as a con-
sequence of thermal effects, wherein temperature changes may cause axial, that is,
longitudinal, forces to develop in pipe walls. When pipes are installed above the
ground surface, or are exposed within a pumping station, gravitational forces can
cause the pipe to behave much like a beam. Differential settlement of soil and pipe
bedding materials can also cause beam loads to develop in buried pipes.
Flow Routing in Storm Sewers

The storm sewer design process, as described in this chapter, is based on estimates of
peak discharges for each element (inlet, pipe, etc.) in the system, with hydraulic siz-
Section 11.6 Using Computer Models for Storm Sewer Design and Analysis 461
ing of elements to accommodate the peak discharge estimates. In most instances,

storm sewer systems are small enough that there is so little in-pipe storage compared
with the flow through the pipe, so minimal flow attenuation occurs in the system.
Flow routing is not necessary in these cases.
An alternative—but much more complex—approach is to route runoff hydrographs
through an existing or proposed sewer system to check its performance. Observations
of unacceptable system performance attributes, based on the results of flow routing
simulations, provide guidance for design modifications and/or needed retrofits. Flow
routing in storm sewers is necessary in continuous simulation where inflow hydro-
graphs, as opposed to just peak flow rates, are computed or specified. This may be
necessary for large drainage basins where the rational method is not appropriate, in
instances where temporary water storage is a concern, or in instances where one
desires to simulate water quality attributes as well as flows.
A number of computer algorithms are available for dynamic simulation of storm
sewer systems. This is accomplished this by solving the complete Saint-Venant equa-
tions expressing conservation of mass and momentum (see Appendix C). There are
two practical approaches for solving these equations: explicit and implicit solutions.
The widely distributed SWMM is an example of an explicit solution. However, most
commercial solutions rely on implicit schemes to solve these equations since explicit
schemes exhibit problems with robustness and stability for large time steps. Implicit
solutions being more robust have proven reliable in complex situations (Price, 1994).
ASCE (1992) provides descriptions of several additional models with flow routing
capabilities, and also provides direction to a number of published reviews of available
models.
11.6 USING COMPUTER MODELS FOR STORM SEWER

DESIGN AND ANALYSIS
The calculations required to analyze the hydraulics of storm system design can be
complicated and tedious. Computer programs have greatly simplified the design pro-
cess, allowing the engineer to focus on making design decisions instead of becoming
mired in repetitive computations. In addition, computer programs allow the engineer
to quickly view and report results in a variety of useful formats, such as plan and pro-
file drawings, graphs, and tables.
The following steps outline the basic process of using a model for design.
1. Obtain a computer program capable of performing the necessary hydraulic and
hydrologic calculations.
2. Obtain topographic data for the project area from surveys, construction drawings,
topographic maps, or digital elevation models (see Section 2.5).
3. Identify the roadway and gutter cross-section geometry, preferred inlet types,
preferred pipe material, minimum pipe size and slope, and design storm charac-
teristics. (Design storms are discussed in Chapter 4.)
4. Lay out the storm sewer system, including preliminary locations of inlets, man-
holes, outfalls, and pipes. Identify inlet opening elevations.
5. For each inlet, delineate the drainage area and compute the time of concentration
and other parameters necessary for runoff calculations (for example, runoff coef-
ficient or curve number; see Chapter 5).
6. Use the computer program to compute the peak design flow to each inlet and
check the capacities of the surface structures (see Chapter 10). Resize the inlets,
add inlets, or alter roadway and gutter cross-sectional geometry if necessary. In
some instances (such as with StormCAD), the computer program may be capable
of automatically adjusting inlet sizes.
7. Establish the horizontal layout for the sewer pipes based on the inlet layout.
8. If you are using a computer program capable of automatically sizing pipes and
setting invert elevations, the following steps apply:
a. Establish pipe design constraints (for example, cover, slope, size, and flow
velocity requirements)
b. Assume all pipes are the minimum allowable size. Using the design storm,
perform a design run on the network to set pipe inverts, increase pipe sizes as
necessary, and perform hydraulic calculations.
9. If you are using a program that does not have automated design capabilities, do
the following:
a. Approximate pipe slopes based on the slope of the ground surface or mini-
mum slope requirement, whichever is greater. Estimate the design flow for
each pipe based on cumulative upstream area characteristics, controlling
time of concentration, and system flow time if significant.
b. For each pipe, use the Manning equation or a similar method (see Sections
6.2 and 7.2) to solve for the pipe size required to convey the peak flow.
Assign initial pipe sizes based on next larger commercially available size in
the desired material, subject to the minimum size constraint.
c. Set initial pipe invert elevations based on design constraints such as mini-
mum slope and cover.
d. Run the program to calculate hydraulic grades, velocities, and so forth.
10. Evaluate the acceptability of the initial design, considering factors such as
hydraulic grade, too much or too little capacity, whether design constraints are
met, and vertical alignment issues (for instance, minimum distances at utility
crossings and cover at intermediate locations along the pipe length). Problems
with pipes having excessive velocity and/or slope may be alleviated by using
drop structures.
11. Make adjustments to the design and run the calculations again. Re-examine the
acceptability of the design, make more adjustments if necessary, and recalculate.
Repeat this process until an acceptable design is obtained. Some programs have
features such as color coding, annotation, and profiling that can help in identify-
ing problem areas.
12. If multiple design alternatives exist (such as a choice in pipe material or storm
sewer route), compare these alternatives by considering factors such as perfor-
mance, cost, and durability, and develop a design recommendation.
Section 11.6 Using Computer Models for Storm Sewer Design and Analysis 463
Figure 11.10 is a flowchart showing the typical process for the design of a storm
sewer system.
Figure 11.10
Flowchart of storm
sewer system design
process
11.7 OUTFALL DESIGN AND ENERGY DISSIPATION

The discharge exiting a storm sewer system at its outfall point often has a velocity
high enough to cause destructive and unsightly erosion. Protection against erosion is
therefore necessary at storm sewer outfalls. Methods of erosion protection vary, but
riprap aprons are frequently used. Riprap is angular stone of a size and gradation suf-
ficient to resist the mobilizing forces of flowing water. For extreme instances in which
exit velocities are very high, a special type of energy dissipator structure may be
required, such as an impact-type dissipator or stilling basin. However, because of the
potentially high cost of special energy dissipators, designers should give consider-
ation to system design modifications that reduce exit velocities and thus eliminate the
need for them. Guidance for design of energy dissipators can be obtained from the
U.S. Bureau of Reclamation (USBR, 1987; USBR, 1983).
This section provides guidance on design of riprap erosion protection and is based on
a procedure used in the Denver, Colorado Urban Drainage and Flood Control District
(UDFCD, 2001). This procedure has been suggested for use in applications where all
of the following are true.
1. The outlet Froude number is less than or equal to 2.5 [that is, the values of param-
eters Q/D2.5 or Q/WH1.5 are less than or equal to 14 ft0.5/s (7.7 m0.5/s)].
2. The outlet pipe slope is approximately the same as that of the downstream riprap
apron.
3. The invert of the outlet pipe is flush with the level of the riprap surface.
In number (1) above,

D = diameter of a circular outlet pipe (ft, m)
W = width of rectangular conduit (ft, m)
H = height of rectangular conduit (ft, m)
Plan and profile views showing the general configuration of a riprap apron for a trap-
ezoidal channel are illustrated in Figure 11.11. The recommended maximum side
slope for the channel is 4:1 (horizontal:vertical), and a slope that should not exceed
3:1. The height of the riprap should extend to the height of the culvert or normal depth
in the channel, whichever is smaller. The thickness of the riprap in the channel sides
and bottom should be 1.5 times the median size of the riprap, d50. Note the extra
thickness of the riprap (2d50) in the area adjacent to the pipe outlet, where the hydrau-
lic forces tending to mobilize the riprap are generally the largest. At the pipe outlet,
either a concrete cradle (see Figure 9.6) or a headwall should be provided to prevent
undermining and possible settlement or failure of the pipe. At the downstream end of
the riprap apron, an end slope is provided on the riprap to accommodate potential deg-
radation of the downstream channel.
Section 11.7 Outfall Design and Energy Dissipation 465
Figure 11.11
A riprap apron
The required size of the rock used in the riprap layer may be determined with the aid
of Figures 11.12 and 11.13, with the former used for circular outlet pipes and the lat-
ter for rectangular conduits. The variable yt used in the figures denotes the tailwater
depth in the channel at the downstream end of the riprap apron. If yt is unknown, or if
it is suspected that a hydraulic jump occurs at the outlet, it is recommended that one
use yt /D = yt /H = 0.4 with these figures. The gradations for the indicated rock types
(L, M, and so forth) are provided in Table 11.11.
Figures 11.12 and 11.13 are based on an assumption that the flow in the outlet pipe is
subcritical. Nevertheless, these figures can be used for supercritical pipe flows if the
values of D or H are modified and replaced, respectively, by Da or Ha as given in
Equations 11.21 and 11.22.
D + yn
Da = (11.21)
2
H + yn
Ha = (11.22)
2
where yn = normal depth of the supercritical flow in the conduit (ft, m)

Maximum values of Da and Ha, respectively, are D and H.
Figure 11.12
gradation for riprap
sin
outlet protection for
60 Ba
g
circular conduits e in
et at
n cr sip
is
Co D
s e rgy
U ne
E
40
Q / D1.5
H
p eV
Ty
e H
Typ
e M
20 Typ
Type L
0 **
0.0 0.2 0.4 0.6 0.8 1.0
Yt / D
Use Da instead of D whenever flow is supercritical in the barrel.

** Use Type L for a distance of 3D downstream
Figure 11.13
gradation for riprap
outlet protection for
rectangular conduits
Table 11.11 Classification and gradation of ordinary riprap (adapted from UDFCD, 2001)
Rock Size Median Size, d50

% Smaller by
a
Type Weight in. mm in. mm
70–100 12 300
50–70 9 230
VL 6 150
35–50 6 150
2–10 2 50
70–100 15 380
50–70 12 300
L 9 230
35–50 9 230
2–10 3 80
70–100 21 530
50–70 18 460
M 12 300
35–50 12 300
2–10 4 100
100 30 760
50–70 24 610
H 18 460
35–50 18 460
2–10 6 150
100 42 1070
50–70 33 840
VH 24 610
35–50 24 610
2–10 9 230
a. Bury types VL and L with native topsoil and revegetate to protect from vandalism.
The length L of the riprap apron (see Figure 11.11) depends on the degree of desired
protection. The greater the difference between the pipe outlet velocity and the down-
stream tailwater velocity (that is, the required velocity reduction), the longer the
required length of the apron. The procedure recommended by the UDFCD (2001)
assumes that the velocity at any cross section downstream from the pipe outlet point is
related to the angle of the lateral expansion, θ, of the jet. Assuming that the expanding
jet has a rectangular shape, the required apron length can be determined as
1  At 
L=  −W  (11.23)
2 tan θ  yt 
where L = length of apron (ft, m)

W = width of the outlet pipe (use W = D for circular conduits) (ft, m)
yt = tailwater depth (ft, m)
θ = expansion angle of the jet
At = the tailwater cross-sectional area, Q/Vt (ft2, m2)
Vt = allowable non-eroding tailwater velocity (ft/s, m/s)
The expansion factor, 1/(2 tanθ), in (11.23) can be obtained from Figure 11.14 or
11.15, depending on whether the outlet conduit is circular or rectangular. For exam-
ple, Figure 11.14 is read by finding the value of the expansion factor corresponding to
the intersection of yt/D with the line for Q/D2.5.
Figure 11.14
Graph for obtaining ! = Expansion Angle
the expansion factor 8
for circular conduit
for use with (11.23)
7
2.5
Q/D
6
Expansion factor, 1/(2tan!)
5
0.5 1.0
2.0
4 3.0
4.0 5.0
6.0
3
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Tailwater Depth / Conduit Height, yt / D
Figure 11.15
Graph for obtaining
the expansion factor
for rectangular
conduit for use with
(11.23)
In some cases, (11.23) will yield unreasonable results. For circular conduits, the
required length L should not be less than 3D, nor more than 10D if Q/D2.5 is less than
6 ft0.5/s (3.3 m0.5/s). If Q/D2.5 is greater than 6 ft0.5/s (3.3 m0.5/s), then the apron
length is computed as
L = 10D + (Q/D2.5Cf − 6)(D/4) (11.24)
where Cf = 1 (U.S. Customary) or 1.811 (SI)

For rectangular channels, the required length L should not be less than 3H, nor more
than 10H if Q/WH1.5 is less than 8 ft0.5/s (4.4 m0.5/s). If Q/WH1.5 is greater than
8 ft0.5/s (4.4 m0.5/s), then the apron length is computed as
L = 10H + (Q/WH1.5Cf − 8)(H/4) (11.25)
To minimize piping of natural soil materials into the riprap layer, a filter fabric or
granular filter layer, or both, should be placed under the riprap. Filter fabric is prone
to tear during the dumping of rock to form the actual riprap layer, and also provides
less resistance to the mobilizing forces of the flowing water than does a granular filter
material. Nevertheless, filter fabrics have proven to be an adequate replacement for
granular filter materials, particularly in mildly sloping channels. If a granular filter is
used, its gradation should be determined using an appropriate geotechnical filter
design method. The UDFCD (2001) provides guidance on this, as do many
geotechnical engineering texts (such as Holtz and Kovacs, 1981).
Example 11.6 – Riprap Blanket Design. Determine the required length and gradation of a
riprap apron downstream of the storm sewer system designed in Example 11.5. Recall that the down-
stream channel velocity is Vt = 4.5 ft/s, the downstream depth is yt = 2.00 ft, the pipe exit velocity is V
= 5.70 ft/s, the discharge is Q = 24 cfs, and the pipe diameter is D = 30 in. = 2.5 ft.
Solution: Because the depth of flow in the exit pipe is larger than critical depth, the flow is subcritical
and the Froude number is less than unity. The design procedure described is therefore applicable.
The required riprap gradation can be determined using Figure 11.12 with
Q/D1.5 = 24/2.51.5 = 6.1 and yt/D = 2 /2.5 = 0.80

From the figure, Type L riprap (Table 11.11) should be selected. Because this stone is small enough to
be subject to vandalism, it is recommended that it be buried using natural channel materials after it
has been placed.
The expansion factor, 1/(2 tanθ), in (11.23) can be obtained from Figure 11.14 with
Q/D2.5 = 24/2.52.5 = 2.4 and yt/D = 2 /2.5 = 0.80
The expansion factor is therefore
1/(2 tanθ) = 6.7
The required length of the apron is found using (11.23) as
 24 / 4.5 
L = 6.7  − 2.5  = 1 ft
 2.00 
Because this length is shorter than the minimum length of 3D, the apron length is set equal to
L = 3(2.5) = 7.5 ft
The geometric configuration of the riprap layer should be as shown in Figure 11.11.
Modeling Focus: Storm Sewer Design

A storm sewer system is being designed for a commercial site, and the proposed horizontal layout for the system is shown in the
figure below. However, pipe sizes and invert elevations must still be determined, and the performance of the inlets and gutters must
be checked.
A6
D2L
520 92.5 ft 13.0 ft D2
13.0 ft
D1L D2R
125.0 ft
C1 13.0 ft D1 520
A5 13.0 ft
16.0 ft
36.0 ft D1R
A4 110.0 ft 515
105.0 ft 13.0 ft
B1L A3 55.5 ft
13.0 ft
A2 510
B1 13.0 ft
B1R 236.5 ft
A1
The system is to be designed for the 10-year storm using the rational method. Ten-year IDF data is given in the table below.
Duration (min) 5 10 15 20 25 30 35 45 60
Intensity (in./hr) 4.00 3.00 2.65 2.30 2.05 1.80 1.65 1.40 1.10
Pipes should be designed to flow full and pipe crown elevations should be matched at inlet locations. Other design constraints are
provided in the table on the left.
All gutters have a uniform cross-section with a 2-percent cross slope, and all inlets are combination-type with the characteristics
shown on the right (note that these characteristics correspond to a DI-10A inlet type in StormCAD's default inlet library): All pipes
will be concrete (Manning’s n = 0.013) and circular with the lengths shown in the figure above.
Criterion Value Parameter Value
Pipe cover 2 ft min., 10 ft max. Inlet box dimensions 3 ft x 3 ft
Flow velocity 2 ft/s min., 15 ft/s max. Curb opening characteristics Vertical throat
Slope 1% min, 10% max. Height 4.5 in.
Gutter spread 6 ft max. Length 3 ft
Gutter depth 6 in. max. Local depression depth 1 in.
Inlet capture efficiency 50% min. Local depression width 28 in.

Grate characteristics 45° tilt bar grate
Width 20 in.
Length 3 ft
Information on the node structures (inlets, manholes, and outlet) is provided in the next table, including hydrologic characteristics,
whether inlets are on grade or in sag locations, “bypass targets” that define the next downstream inlet for inlets on grade, roadway
slopes, and head loss calculation data.
(continued on facing page)
Time of Long HEC-22

Structure Ground Area Conc. Inlet Bypass Slope Head Loss Benching Head Loss
Label Type Elev, (ft) (ac) Inlet C (min) Location Target (ft/ft) Method Method Coeff.
A1 Outlet 510.60 N/A N/A N/A N/A N/A N/A N/A N/A N/A
A2 Manhole 518.20 N/A N/A N/A N/A N/A N/A HEC-22 Energy Flat N/A
A3 Inlet 516.80 0.60 0.45 5.00 In Sag N/A N/A HEC-22 Energy Flat N/A
A4 Manhole 516.90 N/A N/A N/A N/A N/A N/A HEC-22 Energy Flat N/A
A5 Inlet 517.00 0.39 0.45 10.00 In Sag N/A N/A HEC-22 Energy Flat N/A
A6 Inlet 517.80 0.75 0.41 10.00 In Sag N/A N/A Standard N/A 0.50
B1 Manhole 518.40 N/A N/A N/A N/A N/A N/A HEC-22 Energy Flat N/A
B1L Inlet 518.20 0.20 0.42 7.00 On Grade A5 0.020 Standard N/A 0.50
B1R Inlet 518.20 0.20 0.90 5.00 On Grade A3 0.020 Standard N/A 0.50
C1 Inlet 517.00 0.37 0.63 8.00 In Sag N/A N/A HEC-22 Energy Flat N/A
D1 Manhole 517.90 N/A N/A N/A N/A N/A N/A HEC-22 Energy Flat N/A
D1L Inlet 517.70 0.46 0.35 10.00 On Grade C1 0.010 Standard N/A 0.50
D1R Inlet 517.70 0.08 0.90 5.00 On Grade A3 0.010 Standard N/A 0.50
D2 Manhole 519.20 N/A N/A N/A N/A N/A N/A HEC-22 Energy Flat N/A
D2L Inlet 518.80 0.25 0.42 5.00 On Grade D1L 0.010 Standard N/A 0.50
D2R Inlet 518.80 0.14 0.56 5.00 On Grade D1R 0.010 Standard N/A 0.50
All of the input data, including design constraints, are entered into a design program capable of performing automatic design,
such as StormCAD. The results are summarized in the following tables.
Intercepted Bypassed Capture Gutter Gutter Upstream

Label
Flow (cfs) Flow (cfs) Efficiency (%) Spread (ft) Depth (ft) HGL (ft)
A3 1.34 0.00 100.0 3.87 0.08 514.19
A5 0.57 0.00 100.0 1.75 0.01 514.68
A6 0.92 0.00 100.0 2.24 0.04 515.51
B1L 0.27 0.04 86.3 3.66 0.07 515.45
B1R 0.53 0.20 73.1 5.04 0.10 515.56
C1 0.93 0.00 100.0 2.25 0.04 514.67
D1L 0.41 0.13 76.1 5.15 0.10 515.01
D1R 0.34 0.08 80.0 4.32 0.08 514.96
D2L 0.34 0.08 80.0 4.69 0.09 516.08
D2R 0.27 0.05 84.3 4.19 0.08 516.05
Sec. Size U.S. Invert D.S. Invert Const. Total Flow Ave. Vel. U.S. Cover D.S. Cover
Label (in.) Elev. (ft) Elev. (ft) Slope (%) (cfs) (ft/s) (ft) (ft)
A2-A1 15 512.39 508.60 1.60 5.84 6.45 4.57 0.75
A3-A2 15 512.94 512.39 1.00 5.86 5.37 2.61 4.57
A4-A3 15 513.07 512.94 1.00 4.88 4.03 2.58 2.61
A5-A4 12 513.84 513.32 1.44 3.25 4.57 2.16 2.58
A6-A5 12 514.80 513.84 1.04 1.90 3.23 2.00 2.16
B1-A4 12 515.07 513.32 1.67 0.74 1.94 2.33 2.58
B1L-B1 12 515.20 515.07 1.00 0.27 1.59 2.00 2.33
B1R-B1 12 515.20 515.07 1.00 0.53 2.26 2.00 2.33
C1-A5 12 514.00 513.84 1.00 0.93 1.50 2.00 2.16
D1-A4 12 514.42 513.32 1.00 1.08 2.33 2.48 2.58
D1L-D1 12 514.70 514.42 2.15 0.41 1.82 2.00 2.48
D1R-D1 12 514.70 514.42 2.15 0.28 1.52 2.00 2.48
D2-D1 12 515.67 514.42 1.00 0.60 2.22 2.53 2.48
D2L-D2 12 515.80 515.67 1.00 0.34 1.87 2.00 2.53
D2R-D2 12 515.80 515.67 1.00 0.27 1.66 2.00 2.53

The basic components of storm sewers are pipes, inlet boxes, manholes, junction
structures, catch basins, and outlet structures. Some storm systems may include other
facilities such as detention structures and pumping stations, which are covered in
Chapters 12 and 13, respectively.
The most common materials used for storm sewer piping are concrete and corrugated
steel. The most common cross-sectional shape is circular, but shapes such as boxes,
arches, and ellipses are also available. Inlet boxes support the inlet opening, connect
the surface drainage system to the subsurface piping system, and can serve as junc-
tions where pipe alignment or size changes. Catch basins are modified versions of
inlet boxes for capturing suspended sediment and other debris in the bottom of the
structure. Manholes also serve as pipe junctions, and they provide access to the sewer
system at locations where no additional surface flow enters. Manholes and inlets
should be located at regular intervals to provide access for cleaning and maintenance
of the system. Other types of junction structures may be used instead of manholes in
large or more complex installations. Storm sewer outlet structures commonly consist
of a concrete endwall and some type of erosion protection, such as riprap. Frequently,
energy dissipation devices are also required.
Storm sewer pipes typically run in straight lines to connect inlets, manholes, and other
structures. A certain amount of pipe curvature can be accomplished by deflecting pipe
joints. Pipes must be placed to avoid conflicts with other utilities and obstructions.
Minimum cover depths on pipes must be maintained to prevent crushing, and mini-
mum and maximum slope criteria should be met. It is advisable to check flow veloci-
ties to help ensure that pipes can be self-cleansing.
Once the design flows for each pipe have been determined by hydrologic analysis and
inlet design calculations, and a horizontal layout of the system is complete, the engi-
neer can perform the initial design of pipe sizes and invert elevations. Typically, pipe
slopes are assumed to be roughly the same as the slope of the ground surface, and an
equation for computing pipe capacity, such as the Manning equation, is used to com-
pute minimum pipe size requirements assuming a full-flow condition. These sizes are
then rounded to the next available size, and pipe inverts are assigned based on mini-
mum/maximum cover and slope criteria.
The final design calculations for a storm sewer system involve determining the
hydraulic grade through the system. In addition to considering the head losses due to
pipe friction, these calculations must also account for minor losses through structures
such as manholes and inlet boxes. Two methods for computing these minor losses
presented in FHWA’s HEC-22 publication are the energy-loss method and the com-
posite energy-loss method. These methods consider factors such as the junction
geometry, shaping of the structure to minimize losses, the relative elevations of
incoming pipes, and so forth.
For systems or portions of systems operating under pressure-flow conditions, the
hydraulic grade can be computed using the equations presented in Chapter 6. In the
case of systems with open-channel flow conditions, the water surface profile evalua-
tion techniques described in Chapter 7 should be used. In either case, the computed
hydraulic grade elevations must be compared with the elevations of the tops of inlet
References 473
structures and manholes to ensure that no flooding will occur for the design event.
Final design calculations can be quite cumbersome to do by hand, so computerized
modeling techniques are generally used.
Other considerations when designing a storm sewer system include structural loads
and stresses in pipes and proper outfall design. Riprap is a common measure for pre-
venting erosion and dissipating flow energy at storm sewer outlets. Various proce-
dures are available for estimating riprap apron dimensions and the gradation of stone
required, such as the Denver, Colorado Urban Drainage and Flood Control District
approach described in this chapter.
REFERENCES
American Concrete Pipe Association (ACPA). 2000. Concrete Pipe Design Manual. 13th Printing (revised).
Irving, Texas: ACPA.
American Iron and Steel Institute (AISI). 1999. Modern Sewer Design., 4th ed. Washington, D.C.: AISI.
Chang, F. M., R. T. Kilgore, D. C. Woo, and M. P. Mistichelli. 1994. Energy Losses through Junction Man-
holes, Vol. 1, Research Report and Design Guide. McLean, Virginia: Federal Highway Administration.
Holtz, R. D. and W. D. Kovacs. 1981. An Introduction to Geotechnical Engineering. Englewood Cliffs,
New Jersey: Prentice Hall.
Steel, E. W. and T. J. McGhee. 1979. Water Supply and Sewerage. 5th ed. New York: McGraw-Hill.
USBR.
U.S. Bureau of Reclamation (USBR). 1987. Design of Small Dams. 3d ed. Lakewood, Colorado: USBR.
2, Major Drainage. Denver, Colorado: UDFCD.
Yen, B. “Hydraulics of Storm Sewer Systems.” In Stormwater Collection System Design Handbook, ed.
Mays, L. New York: McGraw Hill.
PROBLEMS
11.1 A stormwater collection system is being designed for a small subdivision. The layout of the system
is shown schematically in the figure, and the table presents hydrologic and ground elevation data.
(Catchment and ground elevation information for each row apply to the “From” inlet.)
The design is based on a time of concentration of 20 min. to the most upstream inlets (A-6, C-2, and
B-1), and then governing system flow times (from estimated flow velocities in the system) as the
design proceeds downstream. Using the rational method and the 10-year rainfall data provided, cal-
culate the design flow in each segment of the system. Then, appropriately size concrete pipes
(n = 0.013) for the system and set pipe invert elevations. Assume that the pipes can flow full, the
minimum allowable pipe size is 12 in., the minimum allowable slope is 0.005, and pipe crown eleva-
tions should match at each structure. Record your answers in the results table.
.
10-Year Intensity
Duration (min) (in./hr)
5 7.44
10 6.18
15 5.20
30 3.80
60 2.53
Accumulated
Inlet Numbers Section Area System Flow Average Ground
From To Length (ft) To Inlet (ac) Accumulated (ac) Time (min) Runoff Coeff. Elevation (ft)
A6 A5 500 2.69 2.69 20.0 0.30 98.6
A5 A4 500 3.61 6.30 21.4 0.30 94.9
A4 A3 400 3.88 10.18 22.6 0.41 91.8
C2 C1 500 5.59 5.59 20.0 0.30 96.5
C1 A3 500 6.44 12.03 21.1 0.30 92.3
A3 A2 400 3.92 26.13 23.8 0.39 89.7
B1 A2 400 2.51 2.51 20.0 0.37 92.9
A2 A1 400 32.5 32.5 24.9 0.43 89.5
A1 Outfall 150 38.04 38.04 26.0 0.45 88.7
Problems 475
Results Table
Upstream Inv. Downstream Inv.
Pipe Discharge (cfs) Pipe Size (in.) Elevation (ft) Elevation (ft)
A6-A5
A5-A4
A4-A3
C2-C1
C1-A3
A3-A2
B1-A2
A2-A1
A1-Outfall
11.2 Use the Energy-Loss Method to compute the head loss through a manhole with the characteristics
given in the table. The structure has one incoming pipe, matching pipe crown elevations, and half
benching.
Structure Diameter (B) 3 ft

Outlet Pipe Diam. (Do) 24 in.
Incoming Pipe Diam. 18 in.
Angle between pipes (θ) 4.2 radians
Water depth in structure (d) 1.2 ft
11.3 Develop the hydraulic and energy grade lines for the three segments of the conveyance system
described in the table and figure below. The water level in the receiving body is 22.9 m. The total
flow through the system is 5.7 m3/s. All pipes are 60-inch concrete pipe (n = 0.013). Assume that the
adjusted loss coefficient through the access holes is K= 0.5. What is the elevation of the water sur-
face in Access Holes 1 and 2?
Nodes Upstream Inv. Downstream Inv.

Upstream Downstream Length (m) Elev. (m) Elev. (m)
N/A Access Hole 2 152 22.9 22.2
Access Hole 2 Access Hole 1 122 22.2 21.0
Access Hole 1 Outfall 91 21.0 20.4
Water exits a
stormwater
detention pond
through a weir box
outlet structure.
Vertical bars
extending from the
top of the weir keep
debris from
entering the outlet,
but require frequent
cleaning.
C H A P T E R
12
Stormwater Detention
Stormwater detention is the temporary storage of runoff in underground containers,

ponds, basins, or depressions. Stormwater detention is typically necessary in new
developments because of the higher runoff rates caused by the increased impervious
area (streets, driveways, roofs, sidewalks, etc.). Higher runoff rates from a site typi-
cally increase the severity and frequency of flooding for downstream properties and
conveyance structures. If runoff is collected in a detention facility, it can be released
to downstream properties in a controlled manner such that downstream flooding and
other adverse impacts are prevented or alleviated. Stormwater detention is a widely
practiced approach to stormwater management in urbanized areas of the United
States.
This chapter provides basic information on stormwater detention design and analysis.
Section 12.1 presents an overview of stormwater detention concepts, and Section 12.2
discusses the basic classifications of detention facilities. Section 12.3 provides a con-
ceptual review of stormwater runoff hydrographs and how they are shaped and altered
by urbanization and detention facilities.
Section 12.4 covers design criteria and other considerations, and Section 12.5
describes the various physical components and geometries of detention facilities.
Included are discussions of pond outlet structures and overflow spillways. The section
also addresses other design considerations, such as benching of pond side slopes and
facility maintenance.
Section 12.6 discusses hydraulic relationships for detention facilities. These relation-
ships include detention pond stage versus surface area, stage versus storage, and stage
versus discharge relationships. Detention routing is a process that uses these relation-
ships to transform a detention inflow hydrograph into an outflow hydrograph. Section
12.7 presents the storage-indication method of routing, a numerical integration rout-
ing method, and modeling examples incorporating these techniques. An overview of
the iterative process of designing detention facilities is provided in Section 12.8.
478 Stormwater Detention Chapter 12
12.1 OVERVIEW OF STORMWATER DETENTION

Figure 12.1 illustrates the basic concepts involved in stormwater detention analysis.
An inflow runoff hydrograph from one or more contributing drainage areas is directed
to a storage facility—in this case, a detention pond. (Section 5.5 covers the develop-
ment of runoff hydrographs.) Runoff is then released from the facility at a controlled,
lower rate through a properly sized outlet structure, such as a culvert. The result is a
pond outflow hydrograph that is substantially flatter (that is, has a lower peak dis-
charge) than the inflow hydrograph.
Figure 12.1
Conceptual drawing
of a detention pond
The Advantages of Implementing Detention Ponds

Stormwater detention is mandated by law for new land development in many jurisdic-
tions within the United States. The intent of required detention is to reduce flood
damages by limiting post-development peak discharges to be equal to or less than
those that occurred prior to development, or to a rate based on other criteria specified
by the applicable stormwater authorities. However, even if these peak discharge
objectives are met, most detention ponds do not mitigate the increase in runoff volume
caused by land development. Thus, with detention, the peak flow from a site will not
be higher than it was prior to development, but the flow will occur over a longer
period.
In addition to peak discharge reduction, stormwater detention improves the water
quality of stormwater runoff. The residence time resulting from the low velocity of
Section 12.2 Types of Stormwater Detention Facilities 479
flow through a detention pond may provide sufficient time for suspended particulate
matter and adsorbed contaminants to settle out of the water column. As a best man-
agement practice (BMP), detention ponds can help reduce pollutant loads to receiving
waters (see Section 15.3). Achievement of significant loading reductions often
requires that increased attention be paid to the selection of pond shape, pond depth,
maintenance plans, and vegetation.
Stormwater Detention versus Retention

The use of the terms stormwater detention and stormwater retention has generated
much confusion, as many engineers use the terms synonymously. For the purposes of
this text, retention facilities are distinguished from detention facilities by the absence
of outlet structures. Stormwater retention involves capturing all runoff from a drain-
age basin, with no release of water onto downstream lands. Retention ponds are some-
times used on mining sites or other areas where no discharge is allowed for water
quality reasons. They may also be used to avoid increasing the runoff volume from a
site. The mechanisms of infiltration and evaporation serve as the only outflows from a
retention facility.
Unlike retention ponds, detention ponds have outlet structures. Stormwater detention
involves a temporary storage of runoff, with a peak release rate to downstream proper-
ties that is usually substantially lower than the detention facility peak inflow rate. The
mechanisms of infiltration and evaporation play a role in stormwater detention, but
they are often negligible compared to the inflow and outflow rates and are usually
ignored. This chapter addresses only detention facility design, which is more widely
used in urban land development than retention, and focuses on the peak discharge
reduction aspect of detention facility design.
Additional References on Stormwater Detention

The hydraulics of detention pond sizing have been the subject of considerable
research, and the reader can turn to numerous publications for more details, including
Akan (1989); Aron and Kibler (1990); Glazner (1998); Kamaduski and McCuen
(1979); Kessler and Diskin (1991); and McEnroe (1992). In addition to these papers,
extensive guidance on the design and construction of the impoundments can be found
in Task Committee (1985); ASCE (1982); ASCE (1992); Guo (2001); Loganathan,
Kibler and Grizzard (1996); Paine and Akan (1993); Urbonas and Roesner (1993);
and Urbonas and Stahre (1993).
12.2 TYPES OF STORMWATER DETENTION FACILITIES

Several types of stormwater detention facilities exist. They may be classified on the
basis of whether they are dry or wet during dry-weather periods, and whether they are
on-line or off-line. Each of these types of ponds has strengths and weaknesses with
respect to a number of issues. A review of various pond types and their inherent
strengths and weaknesses is presented in this section.
Dry versus Wet Ponds

A dry pond has an outlet that is positioned at or below the lowest elevation in the
pond, such that the pond drains completely between storm events. A wet pond, how-
ever, has its lowest outlet at an elevation above the pond bottom. Water remains in the
pond during dry-weather periods between storm events and can be depleted only by
infiltration into the soils and/or by evaporation from the water surface. Wet ponds can
be provided with a gated low-level outlet that can be opened to permit drainage of the
pond for maintenance purposes, but that outlet is kept closed during normal operation.
Figure 12.2 demonstrates the distinction between wet ponds and dry ponds.
Considerations in pond-type selection should include, at minimum, hydrologic, aes-
thetic, and water quality issues. A wet pond is typically suitable if inflows into the
pond maintain a minimum water level by exceeding the rate of loss caused by infiltra-
tion and evaporation. This inflow can come from either a continuous base flow
through the wet pond or from frequent rainfall events. However, it should be recog-
Figure 12.2
Wet detention pond
versus dry detention
pond
nized that a wet pond can decline to low or completely empty levels during extended
dry-weather periods. Dry ponds should be considered in areas where base flow and/or
abundant rainfall are not present.
Aesthetically, a wet pond is usually more desirable than a dry one. A wet pond can
provide a focal point for a land development project and usually can be integrated eas-
ily into required open-space areas. Nevertheless, a wet pond can be a haven for mos-
quitoes and a source of odors and may have unsightly floating debris and scum. Also,
wet ponds are more subject to filling with sediment because they allow more time for
particles to settle.
Dry ponds tend to trap transported debris and require frequent maintenance to keep
from becoming eyesores. The bottoms of dry ponds are often soggy, waterlogged
areas where weeds and other vegetation thrive, creating a hospitable environment for
stench and vermin.
From a water quality perspective, wet ponds are more desirable than dry ponds. Dur-
ing storm events, the water in a properly designed wet pond will be displaced in
whole or in part by incoming runoff. The new stormwater resides in the pond until the
next runoff event. This extended residence time provides an opportunity for sediments
and adsorbed pollutants to settle out of the water column. Vegetation in the pond and
along its banks can also provide water quality benefits during dry-weather periods.
Aboveground versus Underground Detention

Detention ponds can be designed in an almost unlimited number of shapes and sizes.
An aboveground facility usually consists of a depressed or excavated area in the land
surface. An earthen dam or berm may or may not be required. In cases where there is
not ample surface area available to meet storage requirements, underground detention
may be necessary. Underground detention can be as simple as a series of appropri-
ately sized pipes or pre-fabricated custom chambers manufactured specifically for
underground detention.
Aboveground and underground detention ponds function in the same manner; they
detain water and attenuate the outflow rate. The same routing fundamentals apply for
each type of pond, as well. The main difference in the two analyses is the way in
which pond volume data are computed. For aboveground earthen ponds, contour data
are typically used to determine the volume characteristics of the pond. For sloped
underground pipes or custom structures, the calculations depend on the particular
geometry of the storage facility.
On-Line versus Off-Line Detention

On-line and off-line detention ponds are distinguished from one another on the basis
of whether they are positioned along the alignment of a main stormwater drainage
pathway. An on-line pond, as its name suggests, is positioned along the pathway and
all runoff from the upstream drainage basin must pass through it. An off-line pond is
located away from the main drainage pathway, and only a portion of the total contrib-
uting stormwater runoff is diverted through it. The advantages of off-line detention
are that the outlet structure can be smaller and, in some cases, less storage volume is
required (see Section 12.5).
Figure 12.3 illustrates the configuration of on-line and off-line detention facilities.
Only a fraction of the main channel’s flow is diverted into the pond with the off-line
configuration.
Off-line ponds are not as effective as on-line ponds (whether wet or dry) for water
quality enhancement because only a fraction of the total runoff is diverted into them.
Figure 12.3
On-line and off-line
detention ponds
Regional versus On-Site Detention

In some instances, the runoff from several land development projects may be directed
into a regional detention pond rather than individual on-site detention ponds for each
development project. In the case of regional detention, the objective is to mitigate the
peak flow rate for a large combined watershed area, whereas the objective for on-site
detention is to mitigate the peak outflow from each specific development site.
Figure 12.4 illustrates the concepts of regional and on-site detention facilities for a
watershed. The regional ponds are designed to handle flows from relatively large con-
tributing areas, and on-site ponds are designed on a site-by-site basis while typically
ignoring the overall watershed characteristics. Because development increases runoff
volumes as well as discharges, and because detention ponds affect the timing charac-
teristics of runoff hydrographs (see Section 12.3), the combined hydrographs from all
the on-site detention ponds will not necessarily be effective in mitigating the flow rate
for the overall watershed. In fact, on-site detention near the watershed outfall may in
some cases increase the peak flow rate from the watershed as a whole because it can
delay the hydrograph peak from the site to coincide more closely with that of the
larger watershed.
Figure 12.4
Regional versus on-
site detention
Both regional and on-site detention philosophies have advantages and disadvantages.
If only regional detention is used, then the overall watershed peak discharge for that
region can be attenuated, but the conveyance systems upstream of the ponds usually
must be larger to handle the increased, unattenuated flows. Regional ponds are often
more difficult to implement for political and economic reasons. The design, planning,
and construction of regional ponds usually involve multiple property owners, devel-
opers, and possibly even multiple drainage jurisdictions. Conversely, if only on-site
detention is used, better water quality will be achieved on an on-site basis, and many
intermediate conveyance systems will be smaller because of the on-site attenuation
from the ponds. However, the attenuation of the peak flow from the overall watershed
will be in question because the watershed’s pond system was developed in a piece-
meal fashion. Further complicating the question of whether to implement regional or
on-site detention is the fact that the designated “regional” watershed is actually a sub-
watershed of an even larger region. When one considers this larger picture, what were
considered regional detention facilities begin to look like the on-site ponds of the
original view, with similar advantages and disadvantages. The issues are simply trans-
ferred to a larger scale.
A hybrid system of regional and on-site ponds offers the strengths of both regional
and on-site detention. This approach makes it possible to design for each individual
site, while also considering the larger watershed picture. Figure 12.5 demonstrates
how both philosophies can be combined within the same watershed.
Figure 12.5
Combination of
regional and on-site
detention
On-site retention is another approach to mitigating some of the adverse effects of

development. It involves retention of at least a portion of the runoff in depressions and
artificial storage devices, allowing the water to slowly percolate into the soil over an
extended period of time. In theory, if the infiltration capacity of the soils is large
enough, the detention storage requirements for an overall development project can be
greatly reduced or even eliminated. However, more research and pilot programs on
retention are needed to address potential water quality issues surrounding the recharge
of water into the groundwater system, facility life expectancy, maintenance require-
ments, and other issues.
Independent versus Interconnected Ponds

Ponds are considered to be interconnected if they are directly connected and interact
hydraulically. However, ponds may be connected in series without being considered
“interconnected” from a hydraulic modeling standpoint. For instance, Figure 12.6
illustrates a situation in which each pond can be routed separately using a procedure
Section 12.3 Stormwater Detention Modeling Concepts 485
such as the storage indication method, which is discussed in Section 12.7. The ponds
are connected, but the downstream pond does not influence the hydraulic behavior of
the outlet structure for the upstream pond because the downstream water surface
never submerges it. Storage indication routing is applicable to each pond because
their outlets function independently of one another.
Figure 12.7 illustrates an interconnected pond scenario in which a tidal water surface
rises high enough to affect the water surface elevation in the downstream pond, which
in turn affects the water surface elevation in the upstream pond as the culvert outfall is
submerged. In this case, the hydraulic behaviors of the two ponds are interdependent.
With interconnected conditions, reverse flow through the outlet structure can occur if
the downstream water surface elevation rises above the elevation in the connected
upstream pond. If desired, flap gates can be installed to prevent reverse flow in these
situations.
Figure 12.6
Ponds in series
Figure 12.7
Interconnected ponds
12.3 STORMWATER DETENTION MODELING CONCEPTS

Stormwater detention design requires the estimation of a complete runoff hydrograph,
as opposed to the single peak discharge typically used in storm sewer design. Section
5.5 presents procedures for inflow hydrograph estimation. This section provides a
conceptual review of stormwater hydrographs and discusses how they are shaped and
altered by the introduction of detention facilities.
Figure 12.8 illustrates the basic hydrograph concepts associated with detention. The
inflow hydrograph represents the runoff from a watershed. Its principal attributes are
the peak discharge rate and the time from the beginning of the storm to the peak dis-
charge. The area under the hydrograph represents the total volume of runoff resulting
from this storm event.
Figure 12.8
Attenuation of an
inflow hydrograph
and lag in hydrograph
peak due to detention
routing
The outflow hydrograph represents the discharge from a detention facility. The out-
flow hydrograph is obtained by performing a routing analysis (see Section 12.6 and
Section 12.7) on a given detention pond, outlet configuration, and inflow hydrograph.
Attenuation (reduction) of the peak discharge resulting from a storm event is accom-
plished by a detention facility or any type of temporary storage. With the reduction of
the peak comes a lag in the time at which the peak occurs. Figure 12.8 illustrates
attenuation and lag. Note that the peak discharge for the detention pond inflow
hydrograph is larger and occurs earlier than peak discharge for the outflow
hydrograph.
If pond infiltration and evaporation losses occurring in the detention facility during
the storm event are negligible, the areas under the two hydrographs (the runoff vol-
umes) must be equal for ponds that do not provide retention. The shaded area in Fig-
ure 12.8 under the inflow hydrograph represents the maximum volume of water
present in the pond during the storm event. If this volume is greater than the storage
capacity of the pond, the pond will overflow. (A detention pond will typically be
equipped with an emergency spillway to handle the overflow.)
The following three scenarios summarize the relationship of a pond’s inflow rate, out-
flow rate, storage, and water surface elevation for any given time interval within a
stormwater runoff event:
Section 12.4 Detention Design Considerations 487
• If the average inflow rate is greater than the average outflow rate during a
time interval, the volume of stored water in the pond increases during the
time interval, and the water surface elevation in the pond increases.
• If the average inflow rate is equal to the average outflow rate during a time
interval, the volume of stored water does not change during the interval, and
the water surface elevation remains constant.
• If the average outflow rate is greater than the average inflow rate during a
time interval, the volume of stored water decreases during the time interval,
and the water surface elevation in the pond decreases.
Most detention ponds have uncontrolled outlets (their outlets have no valves or gates
that can be opened or closed). For a pond with an uncontrolled outlet into a free out-
fall, the peak of the outflow hydrograph from the pond will occur at the point where
the outflow hydrograph intersects the receding limb of the inflow hydrograph (see
Figure 12.8).
This concept can be used as a check on the validity of a set of routing calculations,
and can be reasoned as follows. Prior to the time of the intersection point, the inflow
rate is larger than the outflow rate, and the volume of water in the pond is increasing.
After that intersection time, the volume is decreasing because the outflow rate is
greater than the inflow rate (see Figure 12.8). Thus, the maximum stored volume of
water in the pond occurs at the time where the two hydrographs intersect. Because the
outflow rate from a pond increases as the headwater depth in the pond increases, it
follows that the maximum outflow rate must occur at the maximum depth, which also
corresponds to the maximum storage during the event. Finally, at the point of intersec-
tion, the outflow is equal to the inflow rate, resulting in a zero rate of change of stored
volume at the moment in time when the water surface crests in the pond.
12.4 DETENTION DESIGN CONSIDERATIONS

When designing a detention facility, the engineer must meet certain design criteria.
These design criteria are typically provided by one or more governmental regulating
bodies, such as the local public works department and the state environmental depart-
ment. Before beginning the design of a detention facility, the engineer should contact
the relevant authorities to obtain their requirements.
A common criterion for detention design is that the peak discharge from a site after
development cannot exceed that of the undeveloped site, but the recurrence interval or
intervals for which the criterion must be met vary among jurisdictions. Another fre-
quent requirement is that the discharge from the site should not exceed the capacity of
downstream conveyances for a particular storm event.
An additional factor that the engineer may need to consider is the effect of detention
on the shape of the discharge hydrograph. For instance, detention at a site that is near
the outlet of a watershed may actually delay the hydrograph peak for the site such that
it coincides more closely with the peak of the overall basin. A possible result is that
the peak discharge from the larger basin will increase even though pre- versus post-
development criteria for the particular site are met. The effects of the increased vol-
ume discharging from a site due to development should also be considered.
Pre-Development versus Post-Development Criteria

As a watershed is developed, the peak discharge and the amount of runoff will tend to
increase, and the time to peak will typically decrease. The intent of stormwater deten-
tion is typically to limit post-development peak discharges from a site to discharges
no greater than those that occurred prior to development. For instance, the post-devel-
opment peak discharges from the site for the 2-, 5-, 10-, and 25-year storms should be
equal to or less than the peak discharges for those same storms under pre-developed
site conditions. Thus, the conveyance capacities of existing downstream structures
will still be sufficient if these structures were adequately sized prior to development.
A variation of the pre- versus post-development criteria is to attenuate post-
development peak discharges for a given set of storms to be significantly less than
pre-development peak flows for those same storms. For instance, regulations might
require that the 25-year post-development peak not exceed the 5-year pre-
development peak. Pond outflows are significantly attenuated, but the discharge will
occur over a longer time.
Downstream Systems
Sometimes, allowable detention facility discharges may be based on criteria other
than pre-development flows. For instance, the pond may need to release flows at rates
significantly lower than pre-development values to avoid exceeding the capacity of
one or more existing downstream conveyance structures. It is important to note, how-
ever, that even if peak discharge objectives are met, detention ponds typically do not
mitigate the increase in runoff volume caused by land development. Although the
peak flow rate from a site for a given storm event may be the same as it was prior to
development, the duration of the discharge can be substantially longer (see page 491).
Mandated Release Rates

Some drainage jurisdictions mandate a specific allowable peak outflow rate based on
site area. For example, if the mandated release rate for a locale is 0.1 cfs/ac and the
development site is 100 ac, then the allowable release rate from the site is 10 cfs.
Design Storms
The pre- versus post-development criteria described in the preceding sections must be
met for one or more specified design storm events. For most localities, these design
events will be specified in the local regulations. For example, the regulations may
specify that post-development peak flows be less than pre-development peaks for 24-
hour synthetic storms having return frequencies of 2, 5, and 10 years. It is advisable,
however, to check the proposed design for additional storms of various durations. It is
also good practice to provide an emergency overflow structure capable of handling
flows from the 100-year storm (for instance). The magnitude of the largest storm that
the facility must be designed to accommodate may depend on risk factors such as site
area and land use.
Comparing Storm-Event Magnitudes
The 2-year return event is often used as the smallest runoff depth for the 100-year event. Thus, if a deten-
storm evaluated when designing stormwater facilities. tion pond were designed for the 2-, 5-, 10-, 25-, 50-,
However, with respect to 24-hour total rainfall depths, and 100-year events, the volume of water impounded
the 2-year event is not really as “small” as it may for the 2-year storm could be expected to be about 30
sound; it often represents more than 40 percent of the percent of the 100-year impounded volume (that is,
100-year event. The table shown below compares the pond would be about 30 percent full, not including
northwest Illinois’ (United States) rainfall depths for freeboard).
various return frequencies to its 100-year rainfall These comparisons demonstrate that design engineers
event to demonstrate the significant size of events
need additional rainfall data (less than the 2-year
with “lower” return frequencies.
event) to understand and evaluate rainfall events that
A second comparison in this table applies the NRCS are similar in magnitude to those that stormwater
(SCS) Curve Number method (see Section 5.2, page facilities will often experience (that is, multiple times
119) for computing runoff volume for a developed in a typical year).
area with CN = 85 to demonstrate the resulting runoff
The following table compares rainfall depths and run-
depths for these events. In this example, the runoff
off volumes for 24-hour storm events in northwest
depth for the 2-year event is about 30 percent of the
Illinois (Huff and Angel, 1992).
100-Year, 24-Hour Percent of 100- Percent of 100-

Return Runoff Depth for
Year Rainfall Year Runoff
Event Rainfall Depth* Depth
CN=85 and Ia=0.35
Depth
2 month 1.40 in. (35.6 mm) 19% 0.39 in. (9.9 mm) 7%
6 month 2.08 in. (52.8 mm) 28% 0.85 in. (21.6 mm) 15%
1 year 2.57 in. (65.3 mm) 35% 1.24 in. (31.5 mm) 22%
2 year 3.11 in. (79.0 mm) 42% 1.68 in. (42.7 mm) 30%
5 year 3.95 in. (100.3 mm) 54% 2.41 in. (61.2 mm) 43%
10 year 4.63 in. (117.6 mm) 63% 3.03 in. (77.0 mm) 54%
25 year 5.60 in. (142.2 mm) 76% 3.93 in. (99.8 mm) 70%
50 year 6.53 in. (165.9 mm) 89% 4.81 in. (122.2 mm) 86%
100 year 7.36 in. (186.9 mm) 100% 5.60 in. (142.2 mm) 100%
Peak Flow Timing Issues

Figure 12.9 illustrates problems that can arise from changes in the times at which
peak runoff rates occur and demonstrates the need to consider regional as well as local
effects in detention facility design and implementation. The provision of detention
can actually lead to increases in downstream discharges if timing effects are not care-
fully considered. To counter this effect, some jurisdictions will suspend the detention
requirement in certain cases, and may ask that other conveyance system improve-
ments be made in lieu of detention.
The figure shows hydrographs for two tributaries (A and B) to the main stem of a
stream. A detention facility is to be provided for a development site on tributary B.
Two scenarios are shown for the total combined flow downstream of the confluence
of the two tributaries. One scenario represents the total combined flow for post-
development conditions if the detention pond is provided. The other scenario repre-
sents the total combined flow for post-development conditions assuming that no
detention is provided. Both scenarios are compared to the pre-developed hydrograph
for the overall watershed (areas A and B).
Figure 12.9
Effect of detention lag
on overall basin
hydrograph
Although the scenario with the detention pond successfully attenuates the post-
developed flow from tributary B to a peak discharge lower than the pre-developed
condition for that tributary, it actually increases the peak discharge of the total com-
bined flow downstream of the confluence. Conversely, the scenario without detention
has little impact on the combined flow downstream of the confluence during the
period when the runoff hydrograph from the overall watershed is peaking, but it does
greatly increase the discharges on the rising limb of the overall hydrograph until most
of the water from tributary B has drained from the system. Removal of detention from
tributary B also negates any possible water quality improvement that might be associ-
ated with detention.
Bypassing Flows from Off-Site

In most cases, some discharge from off-site drainage areas will flow onto the project
site. The drainage system on the site can be designed either to convey and detain off-
site discharges with the on-site flow, or to keep the flow separated, diverting it around
or through the site. This situation is illustrated in Figure 12.10.
Figure 12.10
Off-site flows are
bypassed through a
site
With some detention designs, it may be advantageous to bypass flows from off-site if
they would greatly increase the required size of the detention pond outlet structure.
However, due to the timing issues previously discussed, the resulting peak discharge
from the site (computed by summing the bypass and detention outflow hydrographs)
can actually be higher in some instances than if there were no detention at all.
Volume Mitigation Issues

When a drainage basin is urbanized, the addition of impervious surfaces, such as
roofs and pavement, and the compaction of soils result in an increase in the volume of
runoff produced for a given storm event. If evaporation and pond infiltration losses are
neglected, and if the amount of water in a pond is the same after a storm event as
before the event, then the total pond inflow volume must be equal to the total outflow
volume. That is, the areas under the two hydrographs in Figure 12.8 are equal. For this
reason, the duration for which the outflow hydrograph is nonzero must be longer than
the duration of the inflow hydrograph. This increased flow duration can lead to ero-
sion and/or sedimentation problems downstream of the pond.
Figure 12.11 illustrates a comparison of a hydrograph for pre-developed conditions
with post-development inflow and outflow hydrographs for a detention pond.
Although the peak discharge on the outflow hydrograph from the pond is slightly
smaller than the peak discharge on the pre-development hydrograph, a relatively high
discharge is sustained for a long period of time on the outflow hydrograph because it
represents a larger volume of runoff.
Figure 12.11
Comparison of pre-
development
hydrograph, post-
development pond-
inflow hydrograph,
and post-development
pond outflow
hydrograph
Figure 12.12 shows the increase in runoff volume that resulted from the development.
The detention pond has been designed to prevent the peak of the outflow hydrograph
from exceeding that of pre-developed conditions, but the pond still must pass the
entire volume of post-development runoff downstream. The shaded area represents
the increase in the volume of runoff and can be viewed as the additional volume of
water that was lost to rainfall abstractions under pre-development conditions.
Figure 12.12
Net increase in runoff
volume resulting from
development
The classical focus of detention facility design has been to reduce the post-
development peak discharge of runoff to no greater than the peak discharge that
occurred prior to development. Recognizing the increase in runoff volume caused by
urbanization, review agencies may in some cases require that the post-development
peak discharge be limited to some fraction of the pre-development peak discharge to
reduce the likelihood of pond failure.
Section 12.5 Components of Detention Facilities 493
12.5 COMPONENTS OF DETENTION FACILITIES

Detention facilities, whether dry or wet, on-line or off-line, usually consist of an
aboveground pond that is excavated and graded into the land surface. In some
instances, they may consist of shallow ponding areas in parking lots or on rooftops. In
areas where available space is at a premium, it may be economically advantageous to
use underground storage consisting of pipes and/or vaults. Thus, the physical compo-
nents of individual detention facilities can be quite variable, but some general obser-
vations can be made. The following subsections provide discussions of the most
common elements of detention facilities.
Dam Embankments
In most cases, a stream or localized depression is enlarged and/or dammed to create a
detention pond. The design of the dam should conform to accepted geotechnical engi-
neering practices. Governmental dam safety and permitting requirements such as the
Dam Safety Act in the United States may apply if the embankment height or storage
volume exceeds a regulatory limit.
Freeboard
Freeboard is defined as the vertical distance from the maximum water surface eleva-
tion (stage) for a storm event to the top of the dam or to the lowest pond bank eleva-
tion. The storage available within the freeboard elevation range adds a “safety”
volume to the required detention storage volumes. Some jurisdictions and reviewing
agencies mandate freeboard requirements for the maximum design event.
Pond Bottoms and Side Slopes

Side slopes of ponds, especially for wet ponds that become attractions for people,
should be gentle. Side slopes should also be mowable for maintenance of dry ponds
and dry elevations of wet ponds. Barrier vegetation, such as cattails, can be installed
along the perimeter of a wet pond to limit access to the water’s edge. Provision of a
shelf in wet pond side slopes limits water depths in the vicinity of the banks, provid-
ing an additional safety measure. An extended shelf width near the inlet of a wet pond
can effectively act as a forebay in which incoming velocities are reduced and much of
the large particulate matter settles out.
Detention pond bottoms should have a slope that is sufficient to ensure complete
drainage between storm events in dry ponds, and to facilitate pond drainage for main-
tenance in the case of wet ponds. For dry ponds, the engineer should consider provid-
ing a low-flow channel along the bottom of the pond from its inlet to its outlet
structure to assist in draining the pond bottom.
A common mistake made in erosion control design around ponds is to assume that the
impounded water within the pond will serve as a natural energy dissipation device for
incoming flows from conveyance elements such as channels, pipes, and culverts. This
may in fact be true during the course of an event when the pond is storing a significant
volume of water. However, in the early part of a high-intensity storm, conveyance ele-
ments may discharge significant flows into a pond before it has impounded much
water. This situation results in exposed outfalls with the potential to cause significant
erosion. For both wet and dry ponds, riprap or similar protective measures should be
provided near the inlet to prevent inflows from eroding the side slopes of the pond.
Outlet Structures
A detention facility’s outlet structure allows flow to discharge from the pond at a con-
trolled rate. Early detention ponds were usually designed to control only a single run-
off event having a specified recurrence interval; thus, these ponds had very simple
outlet structures. This practice provided little to no benefit during smaller or larger
storm events.
Modern detention facility design typically focuses on control of multiple storm events
and often requires several outlet openings at different pond stages. These multiple
openings can consist of multiple pipe outlets of differing diameters or invert eleva-
tions, orifices at various levels, overflow weirs, or any combination of these elements.
Figure 12.13 depicts a profile view of an outlet structure configuration consisting of
several structural components.
Figure 12.13
Composite detention
outlet structure
If pipe outlets, such as culverts, are used, they can be integrated into the side slopes of
a pond perimeter by using headwalls or special end sections. For safety reasons, large-
diameter pipes should be equipped with trash racks or screens to prevent entry by
children (see Section 9.3). In dry ponds, it is recommended that the invert of the low-
est pipe be set at an elevation below the bottom of the pond, and that a fall similar to
that illustrated in Figure 9.4 be employed at the pipe inlet. In wet climates, the fall
will help drain the soils in the pond bottom and prevent it from becoming water-
logged.
Section 12.5 Components of Detention Facilities 495
Orifice openings, of various sizes and/or at various elevations, are often provided in
the walls of riser pipes, also called standpipes or inlet box structures (see Figure
12.14) in detention facilities. Orifices are an effective means of controlling small flow
rates. However, orifices can become clogged by floating or suspended debris, espe-
cially if the openings are small, and should be checked regularly for blockages.
Figure 12.14
Inlet box structure as
detention pond outlet
Weirs of various shapes are also effective outflow control devices and are almost uni-
versally used in wet detention facilities. Weirs can be provided in the bank of a pond
at its outlet, or they can consist of the top edge of a riser pipe or inlet box structure. If
riser structures are used, their tops should be fitted with anti-vortex devices to prevent
objectionable flow patterns and trash racks or screens to prevent debris from entering
the structure. Ladder rungs or similar safety devices should also be provided inside of
risers.
Regardless of the type of flow control device used in a detention facility, consider-
ation should be given to providing an anti-seepage collar around any outlet pipe con-
veying flow through a dam embankment. The collar will help minimize piping of the
surrounding soil particles caused by water leakage, which could ultimately lead to
embankment failure.
In the case of a wet pond, a low-level outlet might be provided so that the pond can be
completely drained for maintenance purposes. The low-level outlet should be fitted
with a valve or gate to prevent the water in the pond from draining through it. The
valve or gate should be periodically opened for short time periods to prevent it from
seizing so that it will be operable when maintenance is needed. An alternative is to
pump the pond dry when maintenance is required. Requirements for pump size, pump
quantity, and pumping time should be computed at the time of the pond design in
order to estimate the cost and level of effort required for the pumping procedure.
Overflow Spillways
All detention ponds, regardless of type, should be fitted with an overflow spillway.
The purpose of an overflow spillway is to safely convey discharge when the design
storage capacity is exceeded during a large storm event. Overflow spillways are com-
monly referred to as emergency spillways because they are designed to operate under
extreme rainfall conditions to prevent dam overtopping.
Many types of outlet devices are lumped into the general definition of spillways,
including broad-crested weirs, side overflow weir spillways, and ogee spillways.
Overflow spillways for detention ponds range from little more than localized depres-
sions in the pond embankment to elaborate structural facilities. Erosion protection
such as riprap or paving helps to prevent excessive damage to the spillway due to high
velocity flows. For ponds with earthen dam embankments, erosion during overtop-
ping can cause the embankment to fail. Therefore, adequate erosion protection for a
spillway located in a dam is critical and may take the form of riprap or concrete pav-
ing.
The storm recurrence interval for which an overflow spillway should be designed is
an issue that requires an exercise of professional judgment, as well as consideration of
the adverse effects of downstream flooding or damages. In areas where the major
drainage system is evaluated for a designated storm recurrence interval (frequently
the 100-year storm), the same recurrence interval might be used for overflow spillway
design. If failure of a dam embankment could cause considerable damage or loss of
life downstream, regulatory requirements pertaining to dams may be appropriate. For
detailed coverage of spillway hydraulics, the reader is referred to USBR (1987).
Outfall Sheet Flow Conditions

In some predevelopment situations, such as a property that resides within an open
meadow on the side of a hill, the overall flow pattern of runoff exiting the property is
sheet flow in nature. If the property is developed and detention is implemented, dis-
charges from the property become concentrated at a single outfall point. Directing this
concentrated flow onto a natural sheet flow area with no channel can lead to signifi-
cant erosion.
In these cases, an appropriate outfall solution must be designed. Two alternative solu-
tions are:
• Acquire an easement across the downstream property and design and con-
struct a surface channel or underground pipe system to convey the water
safely to the next viable downstream conveyance system.
• Design a flow diffusor that will emulate a sheet flow discharge onto the
downstream property. Care should be taken to ensure that prolonged dis-
charges do not erode the downstream property, and that slope stability on the
downstream property is not compromised by saturation of the soils.
Section 12.6 Routing Data: Storage and Hydraulic Relationships 497
Diversion Structures and Conveyances

Designs of off-line ponds (those constructed adjacent to the main drainage pathway)
must provide for diversion of flows to the pond facility. A diversion structure is
placed along the main drainage pathway upstream of the pond, and pipes or open
channels convey diverted water from this structure to the pond, and from the pond
outlet back to the main drainage pathway.
The diversion structure for an off-line detention facility can be one of two types. The
first type diverts a fixed percentage of the total discharge to the pond, regardless of
whether the total discharge is small or large. This type of diversion can be
accomplished practically using a flow splitter. The fraction of the total flow that is
diverted to the pond is a decision that is made during the design process.
The second type of diversion structure diverts all or part of the total discharge above
some base flow level. Flows below that base level are not diverted to the detention
pond. During the design process, the base level and the percentage of flow that is
diverted once the base flow level is exceeded are determined.
Nix and Durrans (1996) found that diversion of discharges above a base level is more
effective at reducing the required size of a detention facility than is diverting a speci-
fied fraction of the total discharge. This makes sense as the threshold diversion
method delays the use of pond storage capacity until the threshold discharge is
reached, rather than using some storage beginning with the onset of the event.
12.6 ROUTING DATA: STORAGE AND HYDRAULIC

RELATIONSHIPS
Hydrograph routing through a detention facility is the process of computing an out-
flow hydrograph and water surface elevations in the pond for a specified inflow
hydrograph. In addition to the inflow hydrograph, pond routing requires two key
types of data to describe the pond’s storage and hydraulic characteristics:
1. Stage versus storage data to describe the pond’s shape and size (may be devel-
oped from stage versus surface area data)
2. Stage versus discharge data to describe the hydraulic behavior of the pond’s out-
let structure(s)
The following subsections describe these relationships.
Stage versus Surface Area

A stage versus surface area relationship for a detention facility is represented by a
graph or table showing the water surface area of the pond or other storage facility as a
function of the water surface elevation (stage) or water depth. This relationship is
used as an intermediate step in calculating the stage versus storage relationship. For a
storage facility having a regular shape (such as storage in underground pipes), geo-
metric equations may be applied to determine the surface area corresponding to any
stage of interest. For ponds that are excavated and graded, contour areas may be
acquired by using a planimeter or CAD drawings. If several ponds will be routed, a
separate stage versus surface area relationship must be developed for each pond.
The choice of whether to use stage (water surface elevation) or water depth to develop
the relationship is largely a matter of personal preference. Either approach will yield
the same routed outflow results. However, stage is normally used because it can be
directly related to site elevations and construction drawings. If ponds are connected to
other conveyance elements, it is important to know their relative vertical positions.
If possible, stage versus water surface area information should begin at the dry invert
for a pond, even if it is designed to be a wet pond. By following this rule, any water
surface elevation within the pond can be modeled during the routing process. For
example, it would permit a wet pond to be modeled for a scenario in which a pro-
longed drought has occurred such that the water surface elevation in the pond at the
beginning of the runoff event is below the lowest pond outlet. In cases of existing wet
pond facilities where contour data is not available below the water surface, the normal
pool elevation can be taken as the “invert” of the pond. Stage versus water surface
area information should extend from the pond invert to elevations at least as high as
the top of the dam embankment or to the tops of the pond banks.
In this chapter, the water surface elevation (or alternatively, the depth of water) in a
pond is denoted by h. The horizontal area occupied by the water surface is denoted by
A. For most ponds, A is a function of h. An exception is a pond with vertical sides, for
which A is constant.
Example 12.1 – Stage versus Surface Area Relationship. An underground storage

vault is to be constructed to serve as a detention pond for a heavily urbanized area. The walls of the
vault will be vertical, but its floor will be sloped along its length to promote drainage. In plan view,
the vault is rectangular with a length of 23 m and a width of 12 m. The slope of the floor is 2 per-
cent along the 23-m dimension. The invert of the lower end of the vault is at an elevation of 37 m.
The elevation of the top of the storage vault is 39 m. Determine the stage versus surface area rela-
tionship for the vault.
Solution: The difference in the elevation of the floor along the length of the vault is 23(0.02) =
0.46 m. Because the floor elevation at its lower end is at an elevation of 37 m, the elevation at its
high end is 37.46 m. For any stage between these elevations, the rectangular water surface area is
equal to the width of the vault (W = 12 m) times the length of the water surface area. That length
depends on the depth of water in the vault and can be expressed in terms of the stage, h, as
L = (h – 37)/0.02
The corresponding water surface area is
A = WL = 12(h – 37)/0.02
These expressions apply to values of h ranging from 37 m to 37.46 m. For stages above 37.46 m,
the surface area is constant because of the vertical walls and is equal to
A = 12(23) = 276 m2
The following table provides tabular data on the stage (elevation) versus surface area relationship
between elevations 37 m and 39 m. Figure E12.1.1 graphs the relationship.
Elevation (m) Area (m2)

37 0
37.1 60
37.2 120
37.3 180
37.4 240
37.46 276
37.5 276
38 276
39 276
39.50
39.00
38.50
Elevation, m
38.00
37.50
37.00
36.50
0 50 100 150 200 250 300
2
Surface Area, m
Figure E12.1.1 Elevation versus surface area relationship for underground vault in Example 12.1
Stage versus Storage Volume

The stage versus storage relationship for a detention facility relates pond water sur-
face elevation (the stage) or water depth (h) to the volume of water stored (S). It pro-
vides a geometric description of the pond that is used during routing to determine the
rise or fall of the water surface elevation given a change in the volume of stored water.
Storage volumes are determined incrementally over a range of pond stages using the
stage versus surface area relationship described previously. The precision of the stage
versus storage curve increases as the incremental elevation change between stages
decreases. The designer must judge the appropriate degree of approximation when
setting the stage increment size to be used.
Figure 12.15 demonstrates how stage-storage data are used in routing to determine the
relative rise or fall in water surface for a given change in volume. In this example, h1
and S1 are the elevation and volume, respectively, at the beginning of the time inter-
val, and ∆S is the change in volume during the time interval calculated from the dif-
ference between the integrated inflow and outflow hydrographs. At the end of the
time interval, the stored volume is S2 = S1 + ∆S. The stage h2 at the end of the time
interval can be found from the curve based on S2. In Figure 12.15, the net change in
storage is positive, so the net change in storage elevation is also positive (that is, the
water is rising).
Figure 12.15
Relationship of stage
(elevation) versus
storage data to change
in volume
Several approaches to developing stage-storage relationships exist. If a detention

facility has a regular geometric shape (for example, it is a cylindrical tank), a geomet-
ric equation can be applied to describe the stage-storage relationship. In most cases,
however, the engineer must use a stage-surface area relationship to develop stage-
storage data. Three methods for determining the stage-storage relationship—the aver-
age-end-area method, the conic method, and numerical integration—are presented in
this section.
Average-End-Area Method. The average-end-area method, also known as
the trapezoidal method, uses the same concept frequently used in computations for
estimating earthwork volumes. In this method, the incremental storage volume ∆S
available between two water surface elevations is computed as:
A1 + A2
∆S = (h2 − h1 ) (12.1)
2
where ∆S = change in volume during the time interval

h1 = water surface elevation at 1 (ft, m)
h2 = water surface elevation at 2 (h2 > h1) (ft, m)
A1 = water surface area corresponding to stage h1 (ft2, m2)
A2 = water surface area corresponding to stage h2 (ft2, m2)
Computation of incremental storage amounts for a series of known stages and surface
areas, followed by summing of the incremental storages, yields the desired relation-
ship between h and S.
A limitation of the average-end-area method is that it applies a linear averaging tech-
nique to describe the surface area of a pond, whereas, for a pond with sloped sides,
the surface area is actually a second-order (or higher) function. However, the error
associated with this approximation is well within the error bounds of other key design
assumptions, such as design storm determination and watershed homogeneity. This is
especially true if the selected stage increment size (h2 – h1) is small. If more accuracy
is needed, the conic method described in the next section may be applied to represent
second-order effects.
Example 12.2 – Computing Storage Volume Using the Average-End-Area

Method. Use the average-end-area method to calculate and graph the stage versus storage relation-
ship for the elevation-area data obtained in Example 12.1.
Solution: The first two columns of the following table contain the stage versus surface area data
from Example 12.1. To compute the stage-storage relationship using the average-end-area method,
adjacent area values in column 2 are added and then divided by the stage increment for those areas.
The steps used in computing the stage-storage relationship are described as follows, and the results
are reported in the table. The relationship is plotted in Figure E12.2.1.
Column 3 = [Column 2 (current row)] + [Column 2 (previous row)]
Column 4 = [Column 1 (current row)] – [Column 1 (previous row)]

Column 5 = [Column 3 (current row) ÷ 2] × [Column 4 (current row)]
Column 6 = [Column 5 (current row)] + [Column 6 (previous row)]
(1) (2) (3) (4) (5) (6)

An + An–1 hn − hn–1 Cumulative S
Stage (m) Area (m2) (m2) (m) ∆S (m3) (m3)
37.00 0 0 0.0 0 0
37.10 60 60 0.1 3 3
37.20 120 180 0.1 9 12
37.30 180 300 0.1 15 27
37.40 240 420 0.1 21 48
37.46 276 516 0.06 15 63
37.50 276 552 0.04 11 75
38.00 276 552 0.5 138 213
39.00 276 552 1.0 276 489
39.50
39.00
38.50
Elevation, m
38.00
37.50
37.00
36.50
0 200 400 600 800 1000 1200 1400
3
Storage, m
Figure E12.2.1 Plot of stage versus storage for Example 12.2
Conic Method. This method applies a conic geometry to each stage increment to
approximate the nonlinear relationship between stage and water surface area, as illus-
trated in Figure 12.16. The incremental storage volume between two stages h1 and h2
(h2 > h1) is defined as
h2 – h1
∆S =  ----------------- ( A 1 + A 2 + A 1 A 2 ) (12.2)
 3 
Figure 12.16
Illustration of conic
method for computing
pond volume
Example 12.3 – Computing the Storage Volume Using the Conic Method. Use
the conic method to calculate and graph the stage (elevation) versus storage relationship for a pond
with the elevation versus surface area relationship given in the following table.
Stage (ft) Area (acre)

100.00 0.158
100.50 0.170
101.00 0.182
101.50 0.194
102.00 0.207
102.50 0.221
103.00 0.234
103.50 0.248
104.00 0.263
104.50 0.278
105.00 0.293
105.50 0.309
106.00 0.325
106.50 0.342
107.00 0.359
Solution: The following table is used to determine the cumulative volume at each elevation. Each
column represents a step in the calculations used for the conic method. A plot of the elevation-stor-
age relationship is given in Figure E12.3.1.
Columns 1 and 2 are the elevation vs. surface area relationship, and are taken from the previous
table.
Column 3 = {[Column 1 (current row)] – [Column 1 (previous row)]} / 3
Column 4 = [(Column 2 (current row)] + [Column 2 (previous row)] + {[Column 2 (current row)]
× [Column 2 (previous row)]}0.5
Column 5 = (Column 3) × (Column 4)
Column 6 = [Column 6 (previous row)] + [Column 5 (current row)]
(1) (2) (3) (4) (5) (6)

A1 + A2 + A1 A2 Cumulative S
Stage (ft) Area (ac) (h2 − h1)/3 (ft) (ac) ∆S (ac-ft) (ac-ft)
100.00 0.158 0.000 0.000 0.000 0.000
100.50 0.170 0.167 0.492 0.082 0.082
101.00 0.182 0.167 0.528 0.088 0.170
101.50 0.194 0.167 0.564 0.094 0.264
102.00 0.207 0.167 0.602 0.100 0.364
102.50 0.221 0.167 0.641 0.107 0.471
103.00 0.234 0.167 0.682 0.114 0.585
103.50 0.248 0.167 0.723 0.121 0.705
104.00 0.263 0.167 0.766 0.128 0.833
104.50 0.278 0.167 0.811 0.135 0.968
105.00 0.293 0.167 0.856 0.143 1.111
105.50 0.309 0.167 0.903 0.150 1.262
106.00 0.325 0.167 0.951 0.158 1.420
106.50 0.342 0.167 1.000 0.167 1.587
107.00 0.359 0.167 1.051 0.175 1.762
Figure E12.3.1 Stage versus storage relationship for Example 12.3
Numerical Integration. Another approach to development of a stage-storage

relationship is similar to the average-end-area method, but involves numerical or
graphical integration of a stage versus surface area relationship. If the water surface
elevation of a pond changes by a small increment dh, the corresponding increment of
storage change is dS, where dS = A(h)dh. Integrating, the volume of storage between
two stages h1 and h2 (h2 > h1) is
h2
∆S = ∫h A( h ) dh
1
(12.3)
That is, the incremental amount of storage between two stages is equal to the corre-
sponding area under a stage versus surface area curve. That area can be determined
either by applying a trapezoidal rule (which reduces to Equation 12.1), by using a
more sophisticated numerical integration algorithm (see Chapra and Canale, 1988),
by counting squares, or by digitizing.
Stage versus Discharge

The discharge (Q) from a detention facility depends on the stage (h) of the water sur-
face and the geometric and hydraulic characteristics of the outlet(s), and possibly on
tailwater effects that may influence outlet hydraulics. In cases where interconnected
ponds are routed, the discharge from one pond into another depends on the stages in
the ponds and the hydraulic characteristics of the channel or conduit connecting them.
The hydraulic characteristics of a pond’s outlet structure are described by a stage ver-
sus discharge relationship, which is applied during routing to determine the increase
or decrease in outflow given a change in the stage (water surface elevation). Figure
12.17 demonstrates how stage-discharge information is related to routing. The rela-
tionship is determined by computing the discharge for various water surface stages.
The precision of the curve increases as the incremental change in stage used to com-
pute the relationship decreases. The designer must judge the appropriate degree of
approximation when selecting the stage increment size for development of the rating
curve.
Figure 12.17
Stage versus
discharge relationship
In Figure 12.17, h1 and Q1 are the stage and discharge, respectively, at the beginning
of a time interval. The stage at the end of the time interval, h2, is determined from the
stage-storage curve based on the change in storage volume. Using h2, the discharge
Q2 at the end of the time interval can be obtained from the stage versus discharge rela-
tionship. In this illustration, the change in stage was positive (rising), so the change in
discharge is also positive (increasing).
Most dry detention facilities have both low- and high-level outlets. The low-level out-
let may consist of a pipe or culvert positioned with its invert either at or near the bot-
tom of the pond, one or more orifices, or a weir. The high-level outlet is usually an
overflow spillway, and is expected to function only during extreme runoff events with
recurrence intervals that exceed the pond’s design storm recurrence interval. When
ponds are designed to control runoff rates for several different recurrence intervals,
intermediate outlets are usually provided as well. They are similar to low-level outlets
in design but are installed at higher elevations.
The stage versus discharge relationship for a pond is developed using hydraulic rela-
tionships applicable to the number and types of individual outlets. The relationships
for several common outlet structure types have already been presented in this book,
including weirs (Section 7.6), orifices (Section 7.7), and culverts (Section 9.5). This
section describes overflow spillway structures, which are commonly found in deten-
tion ponds and are a specific application of the weir equation. The development of
composite rating curves for situations in which multiple structures are present is also
described.
Composite Stage Versus Discharge Relationship for Structures in
Parallel. The composite stage versus discharge relationship for a detention pond
represents the total discharge from a pond (possibly occurring through several outlets)
as a function of the water surface elevation (or depth). Many detention ponds have
several outlet structures that work independently of one another. In this situation, the
discharges through the various outlets are summed to obtain the total discharge for
each water surface elevation (stage) selected to compute the composite stage versus
discharge relationship. Figure 12.18 illustrates a set of outlet structures working in
parallel, and Figure 12.19 shows the stage versus discharge relationship for this outlet
configuration. Example 12.4 illustrates how the composite relationship for a compos-
ite structure consisting of a weir and an orifice is developed.
Figure 12.18
Composite outlet with
outlet structures
discharging in parallel
Figure 12.19
Rating curve for
composite outlet with
outlet structures
Example 12.4 – Stage versus Discharge for a Composite Outlet Structure. A

simple composite outlet structure is being designed with an orifice and a weir operating in parallel.
The orifice is a 6-in. diameter orifice plate attached to the entrance of a culvert with an invert ele-
vation of 100.00 ft and an orifice coefficient of 0.60. (The culvert is large enough that it does not
affect the hydraulic capacity of the orifice; the orifice can be assumed to have a free outfall.) The
concrete rectangular contracted weir has a weir coefficient approximated as 0.49 for the entire
range of headwaters, a weir length of 15 ft, and a crest elevation at 105.00 ft. It operates under free
outfall conditions.
Figure E12.4.1 Outlet structures for Example 12.4

Calculate and graph the elevation vs. discharge rating curve from elevations 100.00 ft to 107.00 ft
for each 0.50 ft increment. (Note that a smaller increment would yield a smoother, more precise
rating curve.)
Solution: From Chapter 7, the rectangular weir equation for a weir (Equation 7.64) with two con-
tractions is
2
Q= Cd 2 g ( L − 0.2 H ) H 3/ 2
3
The orifice equation (Equation 7.68) is
Q = CdAo(2gho)1/2
The following table shows the stage-storage calculation results described by the following steps.
Figure E12.4.1 is a plot of the composite structure rating curve.
Orifice Centroid Elevation = 100.00 ft + (1/2 × 0.5 ft) = 100.25 ft
Orifice Area = πD2/4 = 0.196 ft2
Column 1 is the range of rating table elevations (100.0 ft to 107.0 ft, in 0.5-ft increments)
Column 2 is the head on the orifice = (Column 1) – (Orifice Centroid Elevation)
Column 3 is the orifice discharge = Solution to orifice equation for Cd = 0.60, Ao = 0.196,
ho = Column 2.
Column 4 is the head on the weir = (Column 1) – (Weir Crest Elevation)
Column 5 is the weir discharge = Solution to rectangular weir equation for Cd = 2.6, L = 15.0,
H = Column 4
Column 6 is the composite structure discharge = (Column 3) + (Column 5)
(1) (2) (3) (4) (5) (6)

Centroid = 100.25 ft Crest = 105.00 ft Total Discharge
Elevation (ft) Head (ft) Discharge (cfs) Head (ft) Discharge (cfs) (cfs)
100.00 0.00 0.00 0.00 0.00 0.00
100.50 0.25 0.47 0.00 0.00 0.47
101.00 0.75 0.82 0.00 0.00 0.82
101.50 1.25 1.06 0.00 0.00 1.06
102.00 1.75 1.25 0.00 0.00 1.25
102.50 2.25 1.42 0.00 0.00 1.42
103.00 2.75 1.57 0.00 0.00 1.57
103.50 3.25 1.70 0.00 0.00 1.70
104.00 3.75 1.83 0.00 0.00 1.83
104.50 4.25 1.95 0.00 0.00 1.95
105.00 4.75 2.06 0.00 0.00 2.06
105.50 5.25 2.16 0.50 13.79 15.95
106.00 5.75 2.26 1.00 39.00 41.27
106.50 6.25 2.36 1.50 71.65 74.01
107.00 6.75 2.45 2.00 110.31 112.76
Figure E12.4.2 Rating curve for outlet structure in Example 12.4

Section 12.7 Storage Routing Calculations 509
Composite Stage Versus Discharge Relationship for Structures in

Series. When outlet structures are in series, it is possible that backwater effects
caused by hydraulic controls at downstream structures will affect the discharges
through connecting upstream structures. In these cases, hydraulic grade line calcula-
tions are required to determine the interactions between individual structure compo-
nents.
Figure 12.20 illustrates a pond outlet with structures in parallel and in series. In this
example, at lower stages, the water surface elevation in the culvert remains low
enough that it does not influence the hydraulic performance of the v-notch weir, and
the v-notch weir controls the discharge. As the water surface rises above 4 ft, water
flows over the crest of the inlet box, and the culvert still has enough capacity for the
discharge to be simply the sum of the flows through the v-notch weir and over the top
of the box. When the water surface reaches an elevation of approximately 6 ft, how-
ever, the v-notch weir/inlet box combination becomes capable of passing more flow
than the culvert. The inlet box riser becomes submerged, and the culvert controls the
discharge. For elevations of more than 8 ft, water flows over the spillway, and the
spillway and culvert together control the discharge.
Figure 12.20
Composite outlet with
outlet structures
and in series
12.7 STORAGE ROUTING CALCULATIONS

Routing is the process of analyzing the difference between flow entering and leaving
the pond for a series of time increments in order to determine the change in volume
and water surface elevation. Figure 12.22 graphically depicts an inflow hydrograph
and the routed outflow hydrograph for a specific pond geometry (stage versus storage
relationship), outlet structure (stage versus discharge relationship), and initial water
surface elevation.
Figure 12.21
Rating curve for
composite outlet with
outlet structures
and in series
Figure 12.22
Change in pond
volume for a routing
time step
Fundamentally, routing of a pond amounts to a solution of the conservation of mass

equation
dS
= I (t ) − O(t ) (12.4)
dt
where S = volume of water in the pond at time t (ft3, m3)

I(t) = specified inflow rate at time t (cfs, m3/s)
O(t) = outflow rate at time t (cfs, m3/s)
Equation 12.4 states that the change in storage during a time period is equal to the dif-
ference between the inflow and the outflow. A number of routing methods are avail-
able for evaluating detention facilities. The most common of these is the storage-
indication method, which is described in the next section. This method is based on a
finite-difference approximation of Equation 12.4. A second approach to storage rout-
ing is based on the numerical integration of a modified form of Equation 12.4. This
integration can be performed using any one of a number of alternative numerical inte-
gration algorithms; in this book, a fourth-order Runge-Kutta algorithm is used.
Storage-Indication Method
Figure 12.22 shows the inflow and outflow hydrographs for a pond. The graph magni-
fies a time increment (step) for which the inflow is greater than the outflow. The net
change in pond volume during the time increment is computed as the area between
the hydrographs for that time increment (that is, the difference in flow multiplied by
the time step). Because the inflow is greater than the outflow for this time increment,
the change in volume is positive, which means that the pond water surface (stage) is
rising.
At the point where the inflow equals the outflow (that is, where the inflow and out-
flow hydrographs intersect), the change in storage is zero, the water surface elevation
(stage) is constant, and the pond is cresting. After the pond has crested, the outflow
rate is greater than the inflow rate, resulting in a net loss in volume that causes the
water surface (stage) to fall during a time step. The shaded area of Figure 12.22 is the
bounded area for which inflow is greater than outflow (that is, for the period when the
pond is filling); this area is equal to the maximum volume stored.
Figure 12.23 shows a portion of the inflow and outflow hydrographs for a pond for a
period in which the inflow discharge is greater than the outflow discharge. This figure
is useful for understanding the relationships given in Equations 12.5 through 12.7,
which are applicable for any time period, regardless of the relative magnitudes of the
inflow and outflow discharges.
Figure 12.23
Definition sketch for
terms in Equation
12.5
Over a time interval ∆t = tn+1 − tn, the inflow varies from In to In+1, and therefore the
average inflow discharge over the time interval can be approximated as (In + In+1)/2.
Over the same time interval, the average outflow discharge can be approximated as
(On + On+1)/2. The change in storage over the time interval is ∆S = Sn+1 − Sn, and the
time rate of change of storage may be approximated as ∆S/∆t = (Sn+1 – Sn)/∆t. Substi-
tution of the average inflow and outflow rates, and of the approximate rate of change
of storage, into Equation 12.4 yields:
Sn +1 − Sn I n + I n +1 On + On +1
= − (12.5)
∆t 2 2
where Sn+1 = storage at time tn+1 (ft3, m3)

Sn = storage at time tn (ft3, m3)
∆t = change in time, tn+1 − tn (s)
In = inflow at time tn (cfs, m3/s)
In+1 = inflow at time tn+1 (cfs, m3/s)
On = outflow at time tn (cfs, m3/s)
On+1 = outflow at time tn+1 (cfs, m3/s)
Because the inflow hydrograph is known in detention pond routing, the terms In and
In+1 are known for every time interval in the routing process. For the first time inter-
val, at the beginning of a pond inflow event, the terms On and Sn are also initially
known. Note that On and Sn are related, as for each pond stage hn a corresponding Sn
and a corresponding On exist. Rearranging Equation 12.5 with unknown quantities on
the left and known quantities on the right yields the storage indication equation:
 2 Sn +1   2S 
 + On +1  =  n − On  + ( I n + I n +1 ) (12.6)
 ∆ t   ∆ t 
Thus, knowledge of the quantities Sn, On, and In at the beginning of a time interval (at
time tn), as well as knowledge of the inflow discharge In+1 at the end of the time inter-
val, can be used to compute the value of the left side of Equation 12.6. However the
left side of the equation has two unknown variables (On+1 and Sn+1). Additional infor-
mation must be established to expose the relationship between On+1 and Sn+1 so that
these two values can be determined.
A storage indication curve is a graph that permits the two unknowns to be determined
by illustrating the relationship between O and (2S/∆t + O). Such a graph can be con-
structed because for any stage h in a pond, corresponding values of O and S exist, and
thus corresponding values of O and (2S/∆t + O) also exist.
In performing storage indication calculations for a specified time interval, Equation
12.6 is applied to compute the numerical value of the quantity (2Sn+1/∆t + On+1).
Next, the value of (2Sn+1/∆t − On+1) for the time interval must be found. Because the
value of the left side of Equation 12.6 (2Sn+1/∆t + On+1) is known, and because this
value can be used with the storage-indication curve to find On+1, (2Sn+1/∆t − On+1)
can be computed as
 2 Sn +1   2S 
 − On +1  =  n +1 + On +1  − 2On +1 (12.7)
 ∆t   ∆t 
The term on the left side of the equation can now be used in the next time step to com-
pute (2Sn+2/∆t + On+2), and the process repeats through every time step.
The following section describes the full procedure used in storage indication routing
and presents an example problem. It should be noted that the storage indication rout-
ing procedure discussed is not applicable to interconnected ponds. For interconnected
pond analysis, a more complex algorithm is required because the tailwater conditions
for the outlet are time-variable. For information on routing interconnected ponds, the
reader is referred to Glazner (1998).
Stormwater Detention Analysis Procedure. This chapter has introduced
each of the basic components required for modeling how a pond functions during a
runoff event. A determination of the detailed behavior of a pond during a runoff event
can be accomplished through the following steps if the storage indication method is
used.
1. Calculate the inflow hydrograph to the pond. In most design cases, this inflow
hydrograph will be the post-development runoff hydrograph from a development
site. Methods for calculating hydrographs are presented in Section 5.5.
2. Calculate the stage versus storage relationship for the selected pond design.
Stages used for development of the relationship should range from the dry invert
of the pond to the top of the dam or pond sides.
3. Calculate the stage versus discharge relationship for the pond’s outlet struc-
ture(s). Again, stages used for development of the relationship should range from
the dry invert of the pond to the top of the dam or pond sides, and should be the
same stages used for development of the stage vs. storage relationship in step 2.
This holds true even if the pond discharges corresponding to low stages are zero.
4. Establish the relationship between stage, storage, and discharge by creating a

storage indication curve for 2S/t + O versus O.
5. Use the inflow hydrograph from step 1 and the storage indication curve from step
4 to route the inflow hydrograph through the pond.
Example 12.5 – Storage Indication Routing. Route the inflow hydrograph given in the
following table through the pond described by the stage versus storage data computed in Example
12.3 and the stage versus discharge data computed in Example 12.4. Determine the maximum
water surface elevation for the detention basin, and the maximum storage volume, maximum out-
flow rate, and peak lag resulting from detention.
Time (minutes) Inflow (cfs)

0 0.00
10 0.58
20 5.94
30 15.44
40 20.75
50 20.27
60 17.44
70 14.19
80 11.42
90 9.44
100 8.12
110 7.27
120 6.61
130 5.49
140 3.32
150 1.52
160 0.66
170 0.29
180 0.12
190 0.05
200 0.02
210 0.00
Solution: The results from Examples 12.3 and 12.4 are combined in the following table. Note the
importance of using the same elevations for storage and discharge in order to directly correlate
flow and storage at each elevation.
Elevation Total Volume Total Flow

(ft) (ac-ft) (cfs)
100.00 0.000 0.00
100.50 0.082 0.47
101.00 0.170 0.82
101.50 0.264 1.06
102.00 0.364 1.25
102.50 0.471 1.42
103.00 0.585 1.57
Elevation Total Volume Total Flow

(ft) (ac-ft) (cfs)
103.50 0.705 1.70
104.00 0.833 1.83
104.50 0.968 1.95
105.00 1.111 2.06
105.50 1.262 15.95
106.00 1.420 41.27
106.50 1.587 74.01
107.00 1.762 112.76
Perform the five steps outlined for storage indication pond routing, as follows:
a) Compute the post-developed pond inflow hydrograph. This can be accomplished using the concept
of rainfall hyetographs discussed in Section 4.5 (page 91) in combination with infiltration, runoff,
and hydrograph procedures discussed in Chapter 5. In this example, the final inflow hydrograph is
given.
b) Calculate the stage versus storage curve from the pond grading plan or storage structure geometry.
This information was calculated in Example 12.3 on page 503 (no calculations necessary for this
example).
c) Calculate the stage versus outflow curve for the pond outlet structure. This information was calcu-
lated in Example 12.4 on page 507 (no calculations necessary for this example).
d) Calculate and graph the storage indication curve that correlates (2S/t + O) and O. To calculate this
curve, the rating elevations for the stage versus storage curve must be identical to those used for
the stage versus outflow curve. Use a 10-minute time step in this example. The results are shown in
the following table, where Column 4 values were computed using:
Column 4 = {2.0 × [(Column 2) × 43,560 ft2/ac] ÷ (10 min × 60 s/min)} + (Column 3)
(1) (2) (3) (4)

Elevation (ft) Volume (ac-ft) Outflow (cfs) 2S/∆t + O (cfs)
100.00 0.000 0.00 0.00
100.50 0.082 0.47 12.38
101.00 0.170 0.82 25.50
101.50 0.264 1.06 39.39
102.00 0.364 1.25 54.10
102.50 0.471 1.42 69.81
103.00 0.585 1.57 86.51
103.50 0.705 1.70 104.07
104.00 0.833 1.83 122.78
104.50 0.968 1.95 142.50
105.00 1.111 2.06 163.38
105.50 1.262 15.95 199.19
106.00 1.420 41.27 247.45
106.50 1.587 74.01 304.44
107.00 1.762 112.76 368.60
Because the maximum inflow is about 20 cfs, we already know that the maximum outflow value
should be equal to or less than 20 cfs. The storage indication curve shown in Figure E12.5.1 is an
expanded view for the outflow range below 20 cfs, which will make it easier to read the smaller
flow values during the routing tabulation.
Figure E12.5.1 Storage indication curve for Example 12.5

e) Use the inflow hydrograph from Step 1 and the storage indication curve from Step 4 to route the
hydrograph using the systematic worksheet approach as shown in the table following the column
descriptions
Columns 1 and 2 represent the given inflow hydrograph.
Under initial conditions, columns 3 through 6 have values of 0. Because the pond is initially empty,
the initial elevation (column 6) is 100.00 ft.
For subsequent time steps,
[Column 3 (current row)]= [Column 2 (previous row)]+ [Column 2 (current row)]
[Column 5 (current row)] = [Column 3 (current row)] + [Column 4 (previous row)]
Determine column 6 values by using the column 5 values and solving for the associated outflow
from the storage indication curve calculated in Step 4.
Solve column 7 (current row) using stage vs. outflow curve created in Step 3.
Finally, to complete the current row, solve
[Column 4 (current row)] = [Column 5 (current row)] – 2 × [Column 6 (current row)].
The tabulated results shown stop at t = 300 minutes into the routing. To complete the routing, the
calculations need to be continued until the pond is fully drained (the routed outflow in column 6
recedes to zero). The graph for the inflow and outflow hydrographs shown in Figure E12.5.2 dis-
plays the entire routing time.
(1) (2) (3) (4) (5) (6) (7)

Time Inflow I1 + I2 2S/∆t − O 2S/∆t + O Outflow Elevation
(min) (cfs) (cfs) (cfs) (cfs) (cfs) (ft)
0 0 0 0.00 0.00 0.00 100.00
10 0.58 0.58 0.54 0.58 0.02 100.02
20 5.94 6.52 6.52 7.06 0.27 100.28
30 15.44 21.38 26.18 27.90 0.86 101.09
40 20.75 36.19 59.69 62.37 1.34 102.26
50 20.27 41.02 97.36 100.71 1.68 103.40
60 17.44 37.71 131.26 135.07 1.90 104.31
70 14.19 31.63 158.77 162.89 2.06 104.99
80 11.42 25.61 163.97 184.38 10.21 105.29
90 9.44 20.86 164.07 184.83 10.38 105.30
100 8.12 17.56 163.35 181.63 9.14 105.25
110 7.27 15.39 162.70 178.74 8.02 105.21
(1) (2) (3) (4) (5) (6) (7)

Time Inflow I1 + I2 2S/∆t − O 2S/∆t + O Outflow Elevation
(min) (cfs) (cfs) (cfs) (cfs) (cfs) (ft)
120 6.61 13.88 162.22 176.58 7.18 105.18
130 5.49 12.1 161.71 174.32 6.30 105.15
140 3.32 8.81 160.86 170.52 4.83 105.10
150 1.52 4.84 159.78 165.70 2.96 105.03
160 0.66 2.18 157.86 161.96 2.05 104.97
170 0.29 0.95 154.73 158.81 2.04 104.89
180 0.12 0.41 151.11 155.14 2.02 104.80
190 0.05 0.17 147.29 151.28 2.00 104.71
200 0.02 0.07 143.41 147.36 1.98 104.62
210 0 0.02 139.52 143.43 1.95 104.52
220 0 0 135.65 139.52 1.93 104.42
230 0 0 131.84 135.65 1.91 104.33
240 0 0 128.07 131.84 1.89 104.23
250 0 0 124.34 128.07 1.86 104.13
260 0 0 120.66 124.34 1.84 104.04
270 0 0 117.03 120.66 1.82 103.94
280 0 0 113.45 117.03 1.79 103.85
290 0 0 109.92 113.45 1.77 103.75
300 0 0 106.44 109.92 1.74 103.66
Figure E12.5.2 Pond inflow and outflow hydrographs

From the routing table:
Maximum stage = 105.30 ft, which is above spillway crest
Maximum storage = 1.2 ac-ft (determined from the stage versus storage data for an elevation of
105.30 ft)
Maximum outflow = 10.38 cfs
Outflow peak time = 90 min
Peak flow was delayed 50 min from inflow peak time of 40 min
Modeling Focus: Detention Pond Analysis

A proposed pond is to attenuate peak flows from a new
100-ac residential development with 1-ac lots. Accord- Contour Data
ing to local criteria, the peak outflows from the pond Elevation (ft) Area (ac)
may not exceed pre-development peak flows from the 450.0 1.607
site for the 2-, 25-, and 100-year events. In addition, the 450.5 1.645
local reviewing agency is requiring that the 100-year
451.0 1.684
peak flow not exceed the maximum capacity of the
downstream conveyance system of 32 cfs, and that a 451.5 1.723
minimum freeboard of 1 ft be maintained. The analysis 452.0 1.762
should be performed using an NRCS (SCS) 24-hour, 452.5 1.801
Type II rainfall distribution and the hydrologic parame- 453.0 1.842
ters given in the tables below.
453.5 1.882
24-Hr Rainfall 454.0 1.923
Storm 454.5 1.965
Depth, in.
2 year 3.11 455.0 2.006
25 year 5.30 455.5 2.049
100 year 7.36
Outlet Structure Data
Watershed Area Culvert entrance type Groove end projecting
CN tc (hr)
Condition (ac) Diameter 24 in.
Predevelopment 50 61 1.0 Upstream invert elevation 450.0 ft
Developed 50 68 0.5 Downstream invert elevation 449.5
A variety of computer programs are available for per- Length 38 ft
forming hydrologic calculations. The design storms are Manning’s n 0.013
created by applying the synthetic storm distribution to
Entrance loss coefficient, Ke 0.20
24-hr rainfall depths for the 2-, 25-, and 100-year events
in local area. [Note that, in the case of the PondPack PondPack is an example of a computer program capable
software that accompanies this text, the NRCS (SCS) of performing not only hydrograph calculations, but also
Type II rainfall distribution has already been entered.] the hydraulic calculations necessary to describe the out-
The times of concentration, drainage areas, and curve let structure rating curve and the detention pond routing
numbers for pre-developed and developed conditions calculations. HEC-HMS is an example of a program
are then entered, and the program computes the hydro- capable of computing hydrographs and performing rout-
graphs. ing; however, the stage-storage relationship for the out-
The proposed pond outlet structure is a single, 24-in. let structure must be developed externally. Other
circular concrete culvert with a free outfall. Data programs may compute only the hydrographs, making it
describing the culvert and the pond volume (in the form necessary to perform hydraulic and routing calculations
of contour areas) are provided in the tables that follow. externally, for instance with the aid of a spreadsheet.
For this example, all calculations are performed using

PondPack. Therefore, the next step is to input the eleva-
tion-area data describing the pond volume and the phys-
ical data describing the culvert outlet. The analysis is run
for the entire system, with the storage-indication method
being used for the routing calculations. The computed
pond volume curve and outlet structure rating curve are
shown, as well as the pre-development hydrograph and
pond inflow and outflow hydrographs.
The results of the analysis are summarized in the follow-

ing table. The allowable outflow is taken as the smaller
of the pre-development peak flow and the downstream
capacity of 32.0 Cfs, and the freeboard is the difference
between maximum water surface elevation and the top
of the pond (455.50 ft). The pond meets all design crite-
ria. Note that the engineer could add an emergency spill-
way with an invert above elevation 454.05 ft (100-year
maximum pond elevation), which will provide addi-
tional emergency outflow capacity but not affect pond
performance for events equal to or smaller than the 100-
year design storm.
Predevelopment Developed Conditions Developed Conditions (Pond

Conditions Allow. (Pond Inflow) Outflow)
Return Q Peak Volume Outflow Q Peak Volume Q Peak Max. Freeboard
Freq. Tp (hr) (cfs) (ac-ft) (cfs) Tp (hr) (cfs) (ac-ft) (cfs) Elev. (ft) (ft)
2 years 12.65 6.73 1.70 6.73 12.20 24.03 2.85 2.73 450.79 4.71
25 years 12.55 41.86 7.26 32.00 12.20 96.87 9.65 18.73 452.51 2.99
100 years 12.55 75.22 12.35 32.00 12.20 157.55 15.43 28.88 454.05 1.45
Pond Routing by Numerical Integration

Use of an accurate numerical integration method for pond routing requires that the
governing equation of conservation of mass be expressed in a slightly different form.
Note that a small change in storage dS is accompanied by a small change in stage dh:
dS = A(h)dh (12.8)
where S = storage volume (ft3, m3)

h = stage (ft, m)
A(h) = water surface area at stage h (ft2, m2)
Note also that the outflow rate O depends on h. Thus, Equation 12.4 can be rewritten
as
dh I (t ) − O(h)
= (12.9)
dt A(h)
where I(t) = inflow rate at time t (cfs, m3/s)

O(h) = outflow rate at stage h (cfs, m3/s)
Recalling that a derivative is the slope of a line tangent to a curve, the right side of this
expression is simply the slope of a graph of h as a function of t. Thus, if the stage hn at
the beginning of a time interval (at time tn) is known, and if the slope m of the rela-
tionship between h and t can be found, then the value of the stage hn+1 at the end of
the time interval is
hn +1 = hn + m∆t (12.10)
where hn = stage at beginning of time interval n (ft, m)

hn+1 = stage at end of time interval n (ft, m)
m = slope of the relationship between h and t
∆t = length of time interval (s)
Noting that, from Equation 12.9,
I (t ) − O (t )
m= (12.11)
A(h)
one can write
 I (t ) − O(h) 
hn +1 = hn +   ∆t (12.12)
 A(h) 
This last expression permits values of h to be computed at the end of each of a number
of sequential time intervals. From those computed stages, corresponding outflow rates
from the pond may be computed from a stage versus discharge relationship. Note that
this approach to routing does not require development of a storage versus discharge
relationship; it requires only a stage versus surface area relationship and a stage ver-
sus discharge relationship.
Applying Equation 12.12 requires an approximation of the slope term (the term in
parentheses) on the right side, and alternative numerical integration methods are
available for computing this approximation. One such method, called the 4th-order
(or classical) Runge-Kutta method, was described by Chapra and Canale (1988) and
is outlined below. It approximates the slope m as
1
m= ( k1 + 2k2 + 2k3 + k4 ) (12.13)
6
where
I (tn ) − O(hn )
k1 = (12.14)
A(hn )
 ∆t   k ∆t 
I  tn +  − O  hn + 1 
2  2 
k2 =   (12.15)
 k ∆t 
A  hn + 1 
 2 
 ∆t   k ∆t 
I  tn +  − O  hn + 2 
2 2 
k3 =    (12.16)
 k ∆t 
A  hn + 2 
 2 
I ( tn + ∆t ) − O ( hn + k3 ∆t )
k4 = (12.17)
A(hn + k3 ∆t )
Note that the value of k2 depends on k1, k3 depends on k2, and so forth. Thus, determi-
nation of the stage h of the pond as a function of time involves application of Equation
12.12 and Equations 12.14 through 12.17 for each time interval. Once h has been
determined as a function of time, the outflow rate O is found using the stage versus
discharge relationship. The procedure is illustrated in the following example.
Example 12.6 – Detention Routing with Numerical Integration Using 4th-order

Runge-Kutta approximation. The hydrograph given in the following table is to be routed
through a reservoir. The reservoir has a water surface area of 50 ac, and has sides that are nearly
vertical rock walls. The base flow is assumed to be 25 cfs, and the reservoir is initially under
steady-state conditions where the inflow and outflow rates are equal. Outflow occurs over a spill-
way on a dam at the downstream end of the reservoir. The spillway is a 100-ft-long contracted weir
with the outflow rate given by
O(h) = 0.5(2g)1/2(100 − 0.2h)h3/2

where h is the height (ft) of the reservoir water surface above the weir crest.
Time (minutes) Inflow (cfs)

0 25
10 25
20 25
30 125
40 400
50 866
60 1304
70 1540
80 1529
90 1338
100 1024
110 750
120 541
130 388
140 283
150 207
160 153
170 115
180 89
190 70
200 57
210 48
220 41
230 37
240 33
250 29
260 28
270 27
280 26
290 25
Solution: Details of the solution are presented here and tabulated in the table of results following
the calculations. In the table, Column 1 is simply the time index n, and Columns 2 and 3 are the
values of tn and In (the inflow hydrograph).
For Column 4, an initial value of hn is computed for the assumed steady-state weir flow of 25.0 cfs:
25.0 = 0.5[2(32.2)]1/2(100 − 0.2h0)h03/2

By trial and error, h0 is found to be 0.157 ft, and this value is placed in the first row of Column 4
(note that the initial assumed outflow of 25.0 cfs was placed in Column 14).
The water surface area of the pond A is constant at 50 ac (2,178,000 ft2). Thus, the value of k1 for
the first row of Column 5 is computed using Equation 12.14 as
I (tn ) − O(hn ) 25.0 − 25.0

k1 = = =0
A 2,178, 000
The value of k2 (Column 8) is given by Equation 12.15 as
 ∆t   k ∆t 
I  tn + − O  hn + 1 
 2   2 
k2 =
A
The two terms in the numerator of this equation are given in columns 6 and 7. The values in col-
umn 6 are computed by averaging the inflow for the current time step, n, and the next time step, n
+ 1. The value in Column 7 is computed using the discharge equation for the weir:
O(h) = 0.5(2g)1/2(100 − 0.2h)h3/2

assuming that the head h on the weir is equal to the quantity
k1∆t
hn +
2
For the first time increment, this quantity is
0.157 + (0)(600)/2 = 0.157 ft
noting that ∆t is in seconds. The value in Column 7 is therefore 25.0 cfs.
With the values for Columns 6 and 7 computed, the value of k2 for the first row is found to be
k2 = (25.0 − 25.0)/2,178,000 = 0
The value of k3 (Column 10) is given by Equation 12.16:
 ∆t   k ∆t 
I  tn +  − O  hn + 2 
 2   2 
k3 =
A
The first term in the numerator is already given in Column 6. The value of the second term is given
in Column 9. This value is computed using the discharge equation for the weir, assuming that the
head h on the weir is equal to the quantity
k2 ∆t
hn +
2
For the first row, this quantity is 0.157 + 0 = 0.157 ft. The value in column 9 is therefore 25.0 cfs.
Column 10 for the first row can then be computed as
k3 = (25.0 − 25.0)/2,178,000 = 0
The value of k4 (Column 12) is given by Equation 12.17:
I ( tn + ∆t ) − O ( hn + k3 ∆t )
k4 =
A
The value of the first term in the numerator is simply the flow from the inflow hydrograph in Col-
umn 3 for the next time step (tn+1 = tn + ∆t). The value of the second term is given in Column 11.
This value is computed using the discharge equation for the weir, assuming that the head h on the
weir is equal to the quantity
hn + k3 ∆t
For the first row, this quantity is 0.157 + 0 = 0.157 ft. The value in Column 11 is therefore 25.0 cfs.
Column 12 for the first row can then be computed as
k4 = (25.0 − 25.0)/2,178,000 = 0
The slope m (Column 13) is then found using Equation 12.13:
1
m=
6
( k1 + 2k2 + 2k3 + k4 )
Therefore, m = 0 for the first row. The head on the weir for the next row, hn+1, is then computed as
hn+1 = hn + m∆t
For the first row, the value in Column 13 is computed as
hn+1 = 0.157 + (0)(600) = 0.157 ft
The outflow from the reservoir (Column 14) for the second row, O(hn+1), is therefore given by
O(hn+1) = 0.5(2g)1/2(100 − 0.2hn+1)hn+13/2 = 25.0 cfs

The remainder of the columns in the second row are completed by repeating the procedure used for
the first row, and these steps are repeated until the entire outflow hydrograph is computed. Note
that this solution was made simpler by the fact that the water surface area for the reservoir is con-
stant. In the case of a reservoir with a changing water surface area, the denominators for k1, k2, k3,
and k4, as given in Equations 12.14 through 12.17, would have to be calculated for each row.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
tn I(tn+∆t/2) O(k2) O(k3) O(k4)
n (min) In (cfs) hn (ft) k1 (cfs) (cfs) k2 (cfs) k3 (cfs) k4 m On (cfs)
0 0 25.0 0.157 0.000E+00 25.0 25.009 -4.251E-09 25.009 -4.111E-09 25.009 -3.981E-09 -3.451E-09 25.0
1 10 25.0 0.157 -4.024E-09 25.0 25.008 -3.892E-09 25.008 -3.896E-09 25.008 -3.768E-09 -3.895E-09 25.0
2 20 25.0 0.157 -3.768E-09 75.0 25.008 2.295E-05 26.668 2.219E-05 28.249 4.442E-05 2.245E-05 25.0
3 30 125.0 0.171 4.440E-05 262.5 31.661 1.060E-04 36.543 1.037E-04 45.083 1.630E-04 1.045E-04 28.3
4 40 400.0 0.233 1.629E-04 633.0 60.118 2.630E-04 69.960 2.585E-04 97.048 3.531E-04 2.598E-04 45.2
5 50 866.0 0.389 3.529E-04 1085.0 139.612 4.341E-04 150.024 4.293E-04 208.382 5.030E-04 4.304E-04 97.3
6 60 1304.0 0.648 5.029E-04 1422.0 285.659 5.217E-04 288.696 5.203E-04 376.353 5.343E-04 5.202E-04 208.7
7 70 1540.0 0.960 5.343E-04 1534.5 474.262 4.868E-04 465.254 4.909E-04 561.900 4.440E-04 4.890E-04 376.3
8 80 1529.0 1.253 4.444E-04 1433.5 652.818 3.584E-04 634.723 3.667E-04 714.898 2.861E-04 3.635E-04 561.1
9 90 1338.0 1.471 2.867E-04 1181.0 776.821 1.856E-04 754.267 1.959E-04 800.485 1.026E-04 1.921E-04 713.5
10 100 1024.0 1.586 1.034E-04 887.0 822.232 2.974E-05 805.469 3.743E-05 815.722 -3.018E-05 3.460E-05 798.7
11 110 750.0 1.607 -2.958E-05 645.5 807.709 -7.448E-05 797.543 -6.981E-05 782.868 -1.111E-04 -7.153E-05 814.4
12 120 541.0 1.564 -1.107E-04 464.5 757.372 -1.345E-04 752.097 -1.320E-04 723.550 -1.541E-04 -1.330E-04 782.1
13 130 388.0 1.485 -1.539E-04 335.5 689.744 -1.626E-04 687.857 -1.618E-04 653.521 -1.701E-04 -1.621E-04 723.1
14 140 283.0 1.387 -1.700E-04 245.0 617.723 -1.711E-04 617.499 -1.710E-04 582.348 -1.723E-04 -1.711E-04 653.4
15 150 207.0 1.285 -1.723E-04 180.0 547.573 -1.688E-04 548.282 -1.691E-04 514.810 -1.661E-04 -1.690E-04 582.3
16 160 153.0 1.183 -1.661E-04 134.0 482.698 -1.601E-04 483.853 -1.606E-04 453.315 -1.553E-04 -1.605E-04 514.8
17 170 115.0 1.087 -1.554E-04 102.0 424.560 -1.481E-04 425.892 -1.487E-04 398.772 -1.422E-04 -1.485E-04 453.4
18 180 89.0 0.998 -1.423E-04 79.5 373.553 -1.350E-04 374.827 -1.356E-04 351.117 -1.291E-04 -1.354E-04 398.8
19 190 70.0 0.916 -1.291E-04 63.5 329.178 -1.220E-04 330.379 -1.225E-04 309.824 -1.161E-04 -1.224E-04 351.2
20 200 57.0 0.843 -1.161E-04 52.5 290.894 -1.095E-04 291.971 -1.099E-04 274.263 -1.039E-04 -1.098E-04 309.9
21 210 48.0 0.777 -1.039E-04 44.5 257.989 -9.802E-05 258.905 -9.844E-05 243.670 -9.305E-05 -9.831E-05 274.3
22 220 41.0 0.718 -9.307E-05 39.0 229.648 -8.753E-05 230.477 -8.791E-05 217.381 -8.282E-05 -8.780E-05 243.7
23 230 37.0 0.666 -8.284E-05 35.0 205.362 -7.822E-05 206.028 -7.853E-05 194.759 -7.427E-05 -7.843E-05 217.4
24 240 33.0 0.618 -7.428E-05 31.0 184.361 -7.041E-05 184.899 -7.066E-05 175.118 -6.709E-05 -7.059E-05 194.8
25 250 29.0 0.576 -6.710E-05 28.5 166.046 -6.315E-05 166.576 -6.340E-05 158.094 -5.973E-05 -6.332E-05 175.1
26 260 28.0 0.538 -5.974E-05 27.5 150.286 -5.638E-05 150.723 -5.658E-05 143.400 -5.344E-05 -5.651E-05 158.1
27 270 27.0 0.504 -5.345E-05 26.5 136.633 -5.057E-05 136.996 -5.073E-05 130.634 -4.804E-05 -5.068E-05 143.4
28 280 26.0 0.474 -4.805E-05 25.5 124.734 -4.556E-05 125.038 -4.570E-05 119.477 -4.338E-05 -4.566E-05 130.6
29 290 25.0 0.446 119.5
Inspection of the table shows that the peak outflow discharge is about 814 cfs, and that it occurs at
time t = 110 min.
Section 12.8 Iterative Pond Design 525
12.8 ITERATIVE POND DESIGN

The previous sections of this chapter explained the fundamental elements of stormwa-
ter detention modeling. The examples presented in the chapter up to this point have
been detention facility analyses (that is, the final design configurations have been
specified for the facility shape, size, and outlet structure components).
Design of a new detention pond involves an iterative (trial and error) design proce-
dure. Figure 12.24 provides a flowchart for the basic steps in design of a detention
facility. First, the inflow hydrograph is developed, and can be used with data on
allowable outflows to estimate the required detention volumes for the design storm
event(s). From this estimate, the engineer can develop a grading plan for the pond
area, and then develop stage-storage data for the proposed pond. Based on the stage-
storage data, the engineer can project what stage the pond will reach for each design
storm volume estimate.
Next, a proposed outlet structure is selected and its stage-discharge relationship devel-
oped. Typically, the engineer can tell if the outlet structure configuration is feasible by
checking to see if the discharge values corresponding to the predicted pond stages for
the various design storms roughly match allowable outflow values. If not, the engi-
neer should resize the outlet as needed to increase or decrease capacity.
Once a likely outlet structure configuration is found, the final routing calculations for
the pond can occur. If the design criteria are met, the design can be finalized. Other-
wise, the engineer must go back and adjust the pond volume and/or outlet configura-
tion as necessary and perform the routing again. Many iterations of the procedure may
be required to develop an acceptable design. Although the example in this text shows
how detention pond routing can be performed manually, engineers typically use com-
puter programs to perform these calculations.
Figure 12.24
Flowchart illustrating
steps in designing
stormwater detention

Stormwater detention is used to mitigate the effects of increased runoff rates caused
by development. Several facility types or classifications exist, including:
• Dry ponds versus wet ponds
• Aboveground versus underground detention
• On-line versus off-line detention
• Regional versus on-site detention
• Individual ponds versus interconnected ponds
A detention facility has the effect of attenuating an inflow hydrograph, meaning that
the peak flow is delayed and reduced. A common design requirement in facility
design is that the peak outflow from a site under developed conditions be equal to or
less than the peak outflow under undeveloped conditions. The design storms to which
this requirement applies vary. However, even with detention, the total runoff volume
increases when a site is developed.
The basic parts of a typical detention pond are the dam, pond bottom and sides, outlet
structure(s), and emergency spillway. Also, a freeboard requirement (for example, 1
ft) may exist to ensure some minimum distance between the maximum water surface
elevation and the top of the dam for a particular storm event (often the 100-year
storm).
The stage versus storage volume and stage versus outlet discharge relationships must
be known before an inflow hydrograph can be routed through a detention facility
using the storage indication method. The basic steps of detention facility analysis
using the storage indication method are:
1. Calculate the inflow hydrograph to the pond.
2. Calculate the stage versus storage relationship for the selected pond design.
3. Calculate the stage versus discharge relationship for the pond’s outlet struc-
ture(s).
4. Establish the relationship between stage, storage, and discharge by creating a
storage indication curve for (2S/t + O) versus O.
5. Use the inflow hydrograph from step 1 and the storage indication curve from
step 4 to route the inflow hydrograph through the pond.
A fourth-order Runge-Kutta approximation can also be used to perform rating calcu-

lations.
Detention facility design is a trial-and-error process that typically involves multiple
iterations of the preceding steps until an acceptable combination of storage facility
geometry and outlet structure is found.
REFERENCES
Akan, A. O. 1989. “Detention Pond Sizing for Multiple Return Periods.” Journal of Hydraulic Engineering
115(5): 650.
ASCE. 1982. Stormwater Detention Facilities. Reston, Va.: ASCE.
ASCE Task Committee. 1985. Detention Outlet Control Structures. Reston, Va: ASCE.
Aron, G. and Kibler, D. F. 1990. “Pond Sizing for Rational Formula Hydrographs.” Water Resources Bulle-
tin,26(2): 255.
Chapra, S. C., and R. P. Canale. 1988. Numerical Methods for Engineers. 2d ed. New York: McGraw-Hill.
Glazner, Michael K. 1998 “Pond Routing Techniques: Standard vs. Interconnected.” Engineering Hydrau-
lics and Hydrology. ed. by Strafaci, A. M. Waterbury, Connecticut: Haestad Methods.
Guo, J. C. Y. 2001. “Design of Off-Line Detention Systems.” In Stormwater Collection Systems Design
Handbook, ed. L. W. Mays. New York: McGraw-Hill.
Kamaduski, G. E. and R. H. McCuen. 1979. “Evaluation of Alternative Stormwater Detention Policies.”
Journal of Water Resources Planning and Management 105(WR2): 171.
Kessler, A. and M. H. Diskin. 1991. “The Efficiency Function of Detention Reservoirs.” Water Resources
Research 27(3): 259.
Loganathan, G. V., D. F. Kibler, and T. J. Grizzard. 1996. “Urban Stormwater Management.” In Water
Resources Handbook, ed. L. W. Mays. New York: McGraw-Hill.
McEnroe, B. M. 1992. “Preliminary Sizing of Detention Reservoirs to Reduce Peak Discharge.” Journal of
Hydraulic Engineering 118(11): 1540.
Nix, S. J. and S. R. Durrans. 1996. “Off-Line Storm water Detention Systems.” Journal of the American
Water Resources Association 32, no. 6: 1329–1340.
Paine, J. N. and A. O. Akan. “Design of Detention Systems.” In Stormwater Collection Systems Design
Handbook, ed. L. W. Mays. New York: McGraw-Hill.
Urbonas, B. R. and L. A. Roesner. 1993. “Urban Design and urban Drainage for Flood Control.” In Hand-
book of Hydrology, ed. D. Maidment. New York: McGraw-Hill.
Urbonas, B. and P. Stahre. 1993. Stormwater Best Management Practices including Detention. Englewood
Cliffs, New Jersey: Prentice-Hall.
U.S. Bureau of Reclamation (USBR). 1987. Design of Small Dams. Washington, D.C.: U.S. Department of
the Interior, USBR.
Problems 529
PROBLEMS
12.1 Use the average-end-area method to develop and graph an elevation versus storage relationship for a
reservoir having the elevation versus surface area relationship given in the following table.
Elevation (m) Area (ha) Volume (m3)

37.6 0.00
37.8 0.08
38.0 0.20
38.2 0.31
38.4 0.42
38.6 0.62
38.8 0.68
39.0 0.75
39.2 0.90
39.4 1.12
39.6 1.42
12.2 Develop and graph an elevation versus discharge relationship at 0.5-ft increments for a dry detention
pond with a dam elevation of 248 ft. The outlet structure consists of the following:
• Two 6-in. orifices (Cd = 0.6), both having invert elevations of 234.3 ft, which is equal
to the elevation of the bottom of the pond
• Three 6-in. diameter orifices (Cd = 0.6) with invert elevations of 240.1 ft
• A rectangular, sharp-crested, contracted overflow weir with a crest elevation of 244.0 ft
and a length of 10 ft (Cd = 0.62)
12.3 A trapezoidal detention pond has a base with dimensions W =15.2 m and L = 30.5 m, side slopes of
3 horizontal to 1 vertical (that is, Z = 3), and a maximum depth (D) of 4.5 m. Tabulate the stage ver-
sus storage volume (V) relationship at 0.1-m increments.
Hint: According to HEC-22 (Brown, Stein, and Warner, 2001), the volume of a trapezoidal pond
with equal side slopes can be computed as
V = LWD + (L + W)ZD2 + (4/3)z2D3
12.4 The outlet structure of a wet detention pond consists of a vertical riser pipe with an open top. The
diameter of the riser is 600 mm, the top of the riser is at an elevation of 7.5 m, and the elevation of
the top of the pond is 10.5 m. Develop an elevation-discharge relationship at 0.1-m increments for
the outlet structure of the elevation range of 7.5 to 10.5 m. Evaluate the structure as both a weir (Cd
= 0.62) and an orifice (Cd = 0.6), and assume that the assumed structure type yielding the lower dis-
charge for a particular elevation controls.
12.5 The detention pond described in Problem 12.3 is used to reduce peak runoff. The bottom of the pond
is at an elevation of 6.0 m. The outlet structure described in Problem 12.4 is used to control dis-
charge. Assume that, at the start of the storm, the water surface elevation in the basin is equal to the
elevation of the top of the riser (7.5 m). The pond inflow hydrograph is given in the table that fol-
lows.
Use the storage indication method to route the hydrograph through the detention facility and com-
plete the hydrograph table provided. What is the peak pond outflow? What is the highest water sur-
face elevation in the pond during the storm?
Time (Harmon) Pond Inflow (m3/s) Pond Outflow (m3/s)

12:00 0.00
12:05 0.01
12:10 0.21
12:15 0.36
12:20 0.54
12:25 0.99
12:30 1.47
12:35 2.43
12:40 2.64
12:45 2.58
12:50 2.37
12:55 1.71
13:00 1.44
13:05 1.17
13:10 0.96
13:15 0.78
13:20 0.66
13:25 0.45
13:30 0.33
13:35 0.27
13:40 0.09
13:45 0.00
13:50 0.00
13:55 0.00
14:00 0.00
12.6 Repeat Problem 12.5 using a 4th-order Runge-Kutta method to perform the routing calculations.
12.7 A detention pond has the elevation-area relationship provided for computing its volume. The pond’s
outlet structure consists of an orifice plate with two 6-in. diameter orifices (Cd = 0.6), both with
invert elevations of 150.00 ft, and a 10-ft-long, rectangular, contracted weir (Cd = 0.62) with a crest
elevation of 155.00 ft. All outlet structures have a free outfall tailwater condition. The pond inflow
hydrograph is provided.
a) Use the conic method to graph and tabulate the elevation–volume relationship for an elevation
increment of 0.5 ft.
b) Graph and tabulate the elevation-discharge relationship for the composite outlet structure from
elevation 150.00 ft to 156.00 ft at 0.5-ft increments.
c) Route the inflow hydrograph through the pond using the storage-indication method and tabulate
the outflow hydrograph. What are the peak pond outflow, water surface elevation, and storage
volume? Approximately how long is the pond outflow peak delayed compared to the inflow
hydrograph time to peak? How much freeboard does the pond have for this event?
Problems 531
Elevation (ft) Area (ac) Volume (ft3)

150.00 0.119
150.50 0.133
151.00 0.147
151.50 0.162
152.00 0.178
152.50 0.194
153.00 0.212
153.50 0.230
154.00 0.248
154.50 0.268
155.00 0.288
155.50 0.309
156.00 0.331
Time (min) Pond Inflow (cfs) Pond Outflow (cfs)

0 0.00
10 0.59
20 4.64
30 10.40
40 12.85
50 11.98
60 10.00
70 7.98
80 6.33
90 5.18
100 4.42
110 3.93
120 3.56
130 2.95
140 1.78
150 0.81
160 0.35
170 0.15
180 0.06
190 0.03
200 0.01
210 0.00
220 0.00
Stormwater
normally flows
under the levee
through the culvert
on the far left, and
additionally
through the
remaining culverts
for larger flows.
When the water
behind the levee is
high, the culverts
are closed, and
flow is pumped to
the other side.
C H A P T E R
13
Stormwater Pumping
In most cases, storm sewers are designed as gravity systems that roughly parallel nat-
ural drainage pathways. They accept water at inlets located throughout a drainage
basin and transport it to an outfall point at a lower elevation in the basin. There is no
need to add energy to the flow to move the water in the desired direction. However,
sometimes stormwater discharge must be pumped to a higher elevation in a drainage
basin. Examples of such cases are:
• Natural or constructed levees that block drainage from reaching stream or
river channels
• Highways or roadways that are depressed with respect to surrounding
ground elevations
• Obstructions that necessitate flow diversion
• Detention ponds with bottom elevations that are lower than an available out-
fall discharge point
The total capitalized cost of a stormwater conveyance system with pumping can be
high compared to an otherwise similar system without pumping. Elements that con-
tribute to the higher cost include pump maintenance, debris-removal requirements,
added energy costs, and the need to periodically retire and replace pumps. Design
issues affecting these additional costs are discussed in Section 13.1.
Section 13.2 provides a description of the various types of pumps and installation
methods commonly used for stormwater pumping applications, including discussions
on pump wet wells and dry wells, sizing of pumping wells, and debris handling and
removal.
The operating characteristics of a pump, when installed in a piping system, can be
ascertained on the basis of pump and system curves. These characteristics include the
energy head produced by the pump, the corresponding pump discharge, the brake
horsepower requirements for the pump motor(s), and the pump installation require-
ments for avoiding cavitation damage to the impeller and other internal parts. Section
13.3 presents methods for determining these operating characteristics for single- and
multiple-pump installations.
534 Stormwater Pumping Chapter 13
Section 13.4 discusses piping-system requirements for pumping. Considerations in

the selection of piping materials include estimates of internal fluid pressures, back-
flow prevention, and requirements related to maintenance and vibration isolation. The
startup and stopping of a pump (or pumps) at a pumping station are usually controlled
by float or pressure-sensing devices. Section 13.5 describes typical control devices
and methods.
Peak discharges of stormwater runoff from a drainage basin are often larger than flow
rates that can be economically pumped. Because of this disparity, temporary water-
storage facilities (essentially, detention basins) are often required at pumping stations.
Section 13.6 presents methods for developing an estimate of the required storage vol-
ume. This estimate of the storage volume, along with other trial-design selections
involving pump models and piping-system components, must then be refined through
a trial-and-error process that evaluates alternative designs. Section 13.7 describes this
iterative design process.
In critical locations where temporary pump failures due to power outages cannot be
tolerated, it may be necessary to provide standby power facilities. Pump stations
should also be aesthetically pleasing and have appropriate safety and monitoring
devices. Brief summaries of these miscellaneous aspects of pump-station design are
presented in Section 13.8.
13.1 NEED FOR STORMWATER PUMPING

As noted, stormwater pumping is generally undesirable, but is unavoidable in some
instances. Pumps require maintenance and repair, periodic replacement, provisions
for an electrical or fuel power supply, and regular removal of trapped debris. Over a
long period of service, the operation and maintenance costs associated with a storm-
water pumping installation can be significant. Thus, at the planning stage of a design,
one should carefully consider alternative system routes and designs that eliminate the
need for pumping. The long-term costs associated with pumping can often justify a
considerably higher initial capital cost for a nonpumped system.
The most obvious alternative to a stormwater pumping installation is to use a gravity
system that outfalls at the nearest available location. This outfall may be some dis-
tance away, requiring more pipe and associated storm sewer appurtenances like man-
holes. In such instances, downhill pipe slopes will generally be adverse to the ground
surface slope, with the net effect that pipes must be deeply buried and excavation and
backfilling costs may be high. Other alternatives to stormwater pumping include
recharge basins, siphons, or tunnels.
Pumps are often used to transport tributary streams across levees or flood walls when
the main channel is at flood stage (or high tide). The figure before the start of this
chapter is an example of this type of application. When the main stream is at normal
elevation, drainage can occur under gravity-flow conditions, as shown in Figure
13.1(a). When the main stream is at flood stage, the flood gates through which the
tributary stream flows must be closed to prevent backflow, resulting in flooding due to
interior drainage behind the levee. Through a combination of ponding and pumping,
this flooding can be minimized, as shown in Figure 13.1(b). Figure 13.2 shows
pumped flow exiting through a tidal gate in a levee.
Section 13.1 Need for Stormwater Pumping 535
Figure 13.1
Stormwater pumping
for interior drainage
Figure 13.2
Pump discharge
through a tide gate
Another common location for a stormwater pump station is at a low point within a
highway interchange that has grade separation if the surrounding terrain is flat.
Because these stations are usually built in conjunction with new highway work, the
designer has the opportunity to develop a grading plan that minimizes the drainage
area contributing to the station. Therefore, the drainage areas for highway pump sta-
tions are usually small, and peak inflow can be estimated using the rational method.
If stormwater detention is required, it may be necessary to limit the peak flow from
the station the pre-development peak flow or other specified flow rate. In such cases,
the station may be equipped with a “storage unit” in addition to the wet well (Smith,
2001). Providing detention storage also reduces the required pumping capacity. These
storage units can also be used to contain hazardous material spills that may occur on
the highway. Because design flows are often associated with large storms that can
knock out power, standby generators, either permanent or portable, are usually
required.
Pumping station installation is often an inexpensive alternative in terms of construc-
tion costs, but it may not be advantageous when future operation and maintenance
costs are considered. Further, highway pumping station sites are often congested and
difficult to access. U.S. Federal Highway Administration (FHWA) guidance states,
“…the use of stormwater pumping stations is recommended only where no other
practicable alternative is available. Alternatives to pumping stations include siphons,
recharge basins, deep and long storm drain systems, and tunnels” (Smith, 2001).
When there is no option but to pump stormwater, a great amount of care and thought
is required from the designer. Each situation is unique and requires careful pump
selection and station design to maximize operating efficiencies and minimize costs.
Standardized designs and equipment are certainly desirable to the extent that they can
be implemented, but the engineer must take care to ensure that a design is optimal for
the site in question. The U.S. Army Corps of Engineers publishes a series of manuals
on flood control pumping station design (USACE, 1989, 1995, 1999).
13.2 PUMP TYPES AND INSTALLATION METHODS

Pumps are one of the most common machine types in existence. At the most funda-
mental level, a pump and its motor are machines for conversion of electrical or fuel
energy into mechanical energy (in this case, rotation of a drive shaft), and subse-
quently into fluid energy (in this case, increased energy head). Turbine generators are
the opposites of pumps; they convert fluid energy into mechanical and then electrical
energy.
Pump stations can be configured in a number of ways. For most stormwater pumping
applications, pumps are submerged in a wet well, which is an underground chamber or
vault for storing stormwater. If the wet well is relatively shallow, it may be possible to
locate the pumps above ground, connecting them to the wet well through suction pip-
ing. With larger pump stations, the pumps may be located in a separate dry well adja-
cent to the wet well.
Pump Types
Pumps come in a wide variety of types and sizes, with each type best suited to a par-
ticular range of applications. The two broad classifications of pumps are centrifugal
and positive displacement. Centrifugal pumps have rotating impellers and are further
classified by impeller type. Axial-flow, radial-flow, and mixed-flow pumps are the
pump types most commonly used in stormwater applications. In the U.S., these three
Section 13.2 Pump Types and Installation Methods 537
types are considered centrifugal pumps, although, strictly speaking, the term applies
to radial-flow pumps only.
An axial-flow pump has an impeller that resembles the propeller on a boat. Flow
moves axially along the drive shaft, which is usually encased within a protective
shield. Axial-flow pumps do not handle debris particularly well, as it can damage the
impeller. Fibrous materials also tend to wind around the drive shaft and clog the
pump. These pumps are best suited to applications involving high discharges and low
energy heads.
A radial-flow pump has an impeller that is shrouded within a volute (spiral) casing.
Water typically enters the pump near the center of the impeller, and vanes on the
impeller cast it radially outwards towards the casing wall. Depending on the impeller
configuration, radial-flow pumps can handle debris quite well and are available for a
wide range of operating conditions. They are, however, best suited to low-discharge,
high-energy-head applications. Figure 13.3 a shows a small end suction radial flow
pump located in a dry well.
Figure 13.3
Radial flow pump
Mixed-flow pumps, as the name implies, are hybrids with aspects of both axial- and
radial-flow pumps. The flow direction caused by the impeller has both axial and
radial components. Mixed-flow pumps handle debris better than axial-flow pumps but
worse than radial-flow pumps. They are best suited for intermediate discharge and
energy-head applications.
Positive-displacement pumps may also be used for stormwater pumping. These
pumps work by using a moving element to displace water from the pump casing and
simultaneously raise its pressure. For stormwater applications, screw pumps are the
most commonly used type of positive-displacement machinery, and may be used to
lift flow over a levee or floodwall. A screw pump has a helical impeller reminiscent of
a corkscrew that causes water to be displaced axially along the impeller as it rotates.
Other types of positive-displacement pumps include reciprocating-piston, rotary, and
pneumatic pumps (Sanks, 1998).
One way to approach the problem of selecting the pump type is to use the pump char-
acteristic known as specific speed. In the United States, specific speed is commonly
defined as
ω Q
Ns = (13.1)
hP3 / 4
where Ns = specific speed (gal1/2/min3/2ft3/4, but units are usually unstated)

ω = rotational speed of the pump (rpm)
Q = pumping rate (gpm)
hP = energy head produced by the pump (ft)
Outside of the United States, specific speed is typically defined as
ω Q
Ns = (13.2)
( ghP )3 / 4
where Ns = specific speed (dimensionless)

ω = angular speed of the pump (rad/s)
Q = pumping rate (m3/s)
g = 9.81 m/s2 = acceleration of gravity
hP = energy head produced by the pump (m)
In these equations, both Q and hP are taken at the pump’s best efficiency point. These
definitions imply that high specific speeds tend to be associated with pumps deliver-
ing high discharges at low heads and, conversely, low specific speeds tend to be asso-
ciated with pumps delivering small discharges at high heads. Figure 13.4 relates pump
efficiency to specific speed and discharge. It also shows the specific speed range for
various impeller shapes, indicating that axial-flow pumps tend to have relatively high
specific speeds, radial-flow pumps tend to have relatively low specific speeds, and
mixed-flow pumps have specific speeds somewhere in between. In stormwater pump-
ing applications, discharges are often high while head requirements are low, indicat-
ing a common requirement for high-specific-speed pumps.
Figure 13.4
Pump efficiency as
related to specific
speed and discharge
Reprinted from Pumping Station Design, 2nd ed., ISBN 0750694831, Robert L Sanks, Ph.D., P.E., Copyright 1998, with permission from Elsevier.
Example 13.1 – Selecting Pump Type Use Figure 13.4 to determine the type of pump
(axial-, radial-, or mixed-flow) best suited to an application where the pumping rate is to be Q =
10 cfs, the pumping head is hP = 20 ft, and the motor speed is ω = 1,800 rpm.
Solution: The discharge of 10 cfs is equivalent to 4,500 gpm. Using Equation 13.1, the specific speed
is found
ω toQbe 1800 4500
Ns = = = 13, 000
hP 3/ 4 203 / 4
From Figure 13.4, an axial-flow pump appears to be the best choice for this installation.
Pump Installations
There are two general types of pump installation: wet-well and wet/dry-well. Pump-
ing-station installations include the pumps, motors, and other associated mechanical
and electrical equipment. They frequently also include water-storage facilities neces-
sary to temporarily store accumulated stormwater runoff until it can be pumped to a
discharge location.
In wet-well installations, the pumps are submerged within a pool of water in the
sump, with the pump motors and other controls usually installed above the pump and
connected via a vertical drive shaft. Vertically installed axial-flow pumps, as shown in
Figure 13.5, are well suited to this type of installation. Figure 13.6 shows a typical
motor room for a vertical pump installation with the motors on the right and the motor
control center on the left. Submersible pumps, which are available in large sizes for
stormwater pumping applications, are also installed in wet wells. These types of
pumps are designed for easy removal and often include rail systems to facilitate this
task.
In wet/dry-well installations, which are most common with larger pumping stations,
the pumps are usually mounted on the floor of a dry chamber. Horizontal piping sup-
plies the pumps from a separate wet well or temporary water-storage facility (such as
a detention pond or outdoor forebay).
Figure 13.5
Vertically installed
axial-flow pumps in a
large wet-well
installation
Figure 13.6
Motor control center
and motor for vertical
pumps
Centrifugal pumps are often used in dry-well installations, although axial-flow pumps
can be mounted with horizontal or vertical drive shafts. Motors are generally located
adjacent to the pumps on the floor of the dry well. A typical layout of a dry-well
installation is shown in Figure 13.7. Note that a small sump pump, in addition to the
stormwater pumps, should be provided in a wet/dry-well installation. The sump pump
discharges any water that accumulates in the bottom of the dry well, either from leak-
age or from groundwater infiltration. Wet/dry-well installations may have higher con-
struction costs, but offer several advantages over wet-well installations. In particular,
pump access is easier, which reduces operation and maintenance costs.
In some instances, pumps may be mounted in the open at ground level in wet/dry-well
installations. This option can be particularly attractive when pumping from an open
detention pond. Pump enclosures or security fencing are recommended with installa-
tions of this type to preclude unauthorized access to pumps and appurtenances. In
such applications, care must be taken to ensure that the available net positive suction
head (NPSH) on the pumps is adequate (see Section 13.3).
Regardless of the type of installation, the spacing of pumps and the positioning of
their suction inlets should be consistent with the manufacturer’s recommendations
and with the standards published by the Hydraulic Institute (1998). Of particular
importance are minimum sump dimensions near the pump intake points, which
should be adhered to in order to prevent vortexing and/or cavitation damage. Also,
manufacturers can provide minimum submergence criteria for submersible pumps.
The distribution of flow to the pumps should be as uniform as possible; water should
not have to flow past one pump to reach another. Figure 13.8 illustrates one type of
sump design, and Figure 13.9 provides corresponding guidance on the dimensions
shown.
The design of the intakes is often complex in that it must account for high and low
flow conditions. For example, Figure 13.10 shows a series of culverts on the left that
pass flow to the river through a levee. The intake screens for the pump station are on
the right. When the river (behind the levee) is high, the culverts are closed and all flow
diverted is to the pump station.
Figure 13.7
Dry-well pump
installation
Figure 13.8
Design of a sump
Reproduced by permission of the Hydraulic Institute.

Figure 13.9
Recommended sump
design dimensions
Reproduced by permission of the Hydraulic Institute.
Figure 13.10
Intake for flood
control pump station
Section 13.3 Pump Performance and Selection 545
13.3 PUMP PERFORMANCE AND SELECTION

The operating characteristics of a pump installed in a piping system can be deter-
mined from pump performance curves and piping system curves. These characteris-
tics include the energy head produced by the pump for a corresponding pump
discharge, the efficiency and brake horsepower requirements for the pump motor(s),
and the pump installation requirements for avoiding cavitation damage to the impeller
and other internal parts. The following subsections describe various ways of quantify-
ing these pump performance characteristics for single- and multiple-pump
installations.
Pump Performance Curves

The operating characteristics of a pump, in isolation from a piping system, are typi-
cally summarized in a set of performance curves. The pump manufacturer prepares
these curves based on hydraulic bench testing and publishes them in pump-selection
manuals. Figure 13.11 illustrates a typical set of pump performance curves. Note that
the actual appearances of pump performance curves vary from one manufacturer to
another.
The horizontal axis of the graph in Figure 13.11 is the discharge Q through the pump
and the left-hand vertical axis is the energy head hP produced by the pump. The
energy head hP is also called the total dynamic head (TDH), and it corresponds to the
difference in head across the pump. For a given pump and impeller diameter, Q and
hP are related: as Q increases, hP decreases, and as hP increases, Q decreases. The
concave-downward curves in Figure 13.11 are called head characteristic curves, and
they indicate the relationship between Q and hP. Usually, as in Figure 13.11, several
such curves are shown with each pertaining to a different impeller diameter. Note that
relatively large impellers yield higher energy heads for a given discharge than do
small impellers.
Superimposed over the Q versus hP curves are pump-efficiency contours, which, for
this type of graph, have circular or elliptical shapes that form an approximate bulls-
eye with a center corresponding to maximum efficiency. For a given point on a Q ver-
sus hP curve, the efficiency ε of the pump can be estimated by interpolating between
the contours. Pumps should be selected for maximum efficiency, as low efficiencies
typically result in increased power requirements. The pump efficiency is defined as
Pout
ε= (13.3)
Pin
where ε = pump efficiency

Pout = water power produced by the pump (ft-lb/s, HP, W, etc.)
Pin = power supplied to the pump by its drive shaft (ft-lb/s, HP, W, etc.)
The efficiency is always less than unity (or 100 percent) because power is lost in bear-
ing friction, turbulence, and other energy-dissipating mechanisms. In some cases,
such as with close-coupled, submersible pumps, the efficiency is not simply the pump
efficiency, but rather the overall efficiency of the pump and motor combined (also
known as the wire-to-water efficiency).
Figure 13.11
Sample set of pump
performance curves
Used with permission of the Peerless Pump Company
The water power produced by a pump, denoted by Pout in Equation 13.3, can be
expressed in terms of the discharge, energy head, and specific weight of the fluid as
Pout = Qγ hP (13.4)
where Pout = water power produced by the pump (ft-lb/s, W)

γ = specific weight of the fluid (lb/ft3, N/m3)
hP = head supplied by the pump (ft, m)
To convert from units of ft-lb/s to horsepower, divide the result by 550.
The power provided to a pump by its drive shaft is known as the brake power (often
the brake horsepower) and is represented in Equation 13.3 by Pin. Rearranging that
expression and substituting Equation 13.4 yields the brake power requirement for a
pump motor as
Qγ hP
Pin = (13.5)
ε
where Pin = brake power (ft-lb/s, W)

This brake power requirement is sometimes shown on a set of pump curves as a func-
tion of the discharge. If pump curves do not contain a discharge-versus-power curve
or curves (see Figure 13.11), the power requirement can be computed using
Equation 13.5.
Another type of information usually shown on a set of pump curves is the required net
positive suction head (NPSH). This parameter is related to potential cavitation prob-
lems that could ultimately damage the pump. The required NPSH for a pump is usu-
ally shown, as in Figure 13.11, as a function of the discharge. Cavitation and NPSH
are discussed in more detail later in this section.
The performance curves for a pump, summarized in the preceding paragraphs, depend
on fluid properties and on the rotational speed of the drive shaft. Pump performance
curves are nearly always published for water at a standard temperature of 20° C
(68° F), with the rotational speed usually noted on the chart. Pump performance
curves should be approximately correct if actual water temperatures do not signifi-
cantly differ from standard temperature, but will be incorrect for large temperature
departures or for motor speeds other than that shown on the chart. If the motor speed
is different, pump affinity laws may be applied to develop an accurate set of perfor-
mance curves.
The pump curves shown in Figure 13.11 are typical of those from pump catalogs
where a number of curves are presented for each pump for a range of impeller sizes.
The engineer must select the correct impeller size for the system head curve. Another
type of pump curve, shown in Figure 13.12, depicts the results of a pump test con-
ducted by the manufacturer before a specific pump is being shipped out for installa-
tion.
The reader is referred to Sanks (1998), Karassik et al. (2000), and U.S. Army Corps
of Engineers (1989, 1995, 1999) for more information on pumps, pump selection, and
the design of pumping stations.
Example 13.2 – Determining Water Power and Brake Horsepower Required.

A pump with an operating efficiency of 80 percent delivers a discharge of 19 m3/min at a total head of
6 m. Determine the water power produced by the pump and the brake horsepower required for the
motor to drive the pump.
Solution: The discharge of 19 m3/min is equivalent to 0.32 m3/s. From Equation 13.4, the water
power produced by the pump is
Pout = QγhP = 0.32(9810)(6) = 18,835.2 W = 25.2 HP
The brake horsepower requirement for the motor is equal to the water horsepower produced by the
pump divided by the pump’s operating efficiency:
Pin = QγhP/ε = 25.2/0.80 = 31.5 HP
In practice, one would select a commercially available motor providing at least this much brake
horsepower.
Figure 13.12
Results of certified
pump test
System Curves
System curves for a pump installation depict the relationship between the discharge
and the energy head that must be supplied by the pump(s) to maintain that discharge.
System curves reflect the effects of head losses caused by friction and minor losses in
the piping system, as well as the effect of the difference between the total energy
heads at the system intake and discharge points.
The energy equation is applied to develop and plot a system curve for a given piping
configuration. The following description refers to the piping layout in Figure 13.13.
The upstream point 1 for the energy equation is at the free water surface of the lower
(suction) reservoir, and the downstream point 2 is at the free water surface at the
upper (discharge) reservoir. Because p1 = p2 = 0 and V1 = V2 = 0, the energy equation
can be written as
hP = ∆Z + hL (13.6)
where ∆Z = the difference between the two free water surface elevations, (ft, m)
hL = energy losses in the piping system due to frictional and minor losses (ft, m)
Figure 13.13
Piping system for
illustration of system
curve development
If, for example, frictional energy losses are expressed using the Darcy-Weisbach
equation, Equation 13.6 can be expressed as
L Q2 L Q2 Q2
hP = ∆Z + f1 1 + f2 2 + ∑ Ki (13.7)
D1 2 gA12 D2 2 gA22 2 gAi2
where f1 and f2 = Darcy-Weisbach friction factors for pipes 1 and 2, respectively

L1 and L2 = length of pipes 1 and 2, respectively (ft, m)
D1 and D2 = diameter of pipes 1 and 2, respectively (ft, m)
A1 and A2 = cross-sectional area of pipes 1 and 2, respectively (ft2, m2)
Ki = minor loss coefficient for fitting i
Ai = cross-sectional area of fitting i
The second and third terms on the right-hand side of this expression are the frictional
losses in the pipes upstream and downstream of the pump. The last term represents
the sum of all the minor losses that occur in the two pipes.
It can be seen from Equation 13.7 that the energy head hP that must be supplied by the
pump increases with the square of the discharge. For a given piping system, a range of
values of Q may be substituted into Equation 13.7 to compute corresponding values
of hP. The resulting graph of hP versus Q is the system curve.
In most stormwater pumping, the lift provided by the pump (that is, the z term in
Equation 13.7) is usually larger and more variable than the other terms. Therefore,
rather than a single system head curve, there is actually a family of system head
curves corresponding to various water levels in the wet well and receiving water. For
example, a pump may come on when the water level in the receiving stream is 2 ft
above the wet well, but may also need to pump against 20 ft of head difference during
the height of a flood. The entire range of system head curves must be developed.
Example 13.3 – Developing System Head Curves. A centrifugal pump is installed in a

dry well. Suction piping leading to the pump is ductile iron (absolute roughness = 0.00085 ft) with a
length of 10 ft and a diameter of 12 in. The entrance to the suction pipe is rounded and its length
includes a gate valve. The discharge pipe has a free discharge to the atmosphere at its downstream end
and is ductile iron with a diameter of 10 in. and a length of 100 ft. It includes one 90-degree bend, a
gate valve, and a check valve. Minor loss coefficients for the various valves and fittings are provided
in the following table (Mays, 2001).
Fitting Type Minor Loss Coefficient, K

Pipe entrance (rounded) 0.20
Gate valves (fully open) 0.10
Check valve 1.00
90° bend 0.30
Exit loss 1.00
Develop the system curve for a 2-ft elevation difference between the water level in the suction reser-
voir and the pipe centerline at the discharge point, which will be the elevation when the pump first
comes on. Repeat the calculation for an elevation difference of 20 ft, which corresponds to the pump’s
“off” elevation. Assume a water temperature of 60ºF (ν = 1.217 × 10−5 ft2/s).
Figure E13.3.1 Piping system configuration for Example 13.3

Solution: The cross-sectional area of the 12-in. suction pipe (Pipe 1) is
A1 = πD2/4 = π/4 = 0.7854 ft2

The cross-sectional area of the 10-in. discharge pipe (Pipe 2) is
A2 = πD2/4 = π(10/12)2/4 = 0.5454 ft2

A tabulation of the calculations of the two required system curves is provided in the table following
the calculations. The first column contains assumed discharges for development of the curve. The sec-
ond and third columns contain the Reynolds number of the flow in each of the two pipes, which are
computed as
Re1 = VD/ν = [QD/A]/ν = (Q/0.7854)/(1.217 × 10−5) = 104,620Q
Re1 = VD/ν = [QD/A]/ν = (0.8333Q/0.5454)/(1.217 × 10−5) = 125,544Q

The fourth and fifth columns contain the friction factors computed using Equation 6.11:
1  ε / D 5.74 
= −2 log  + 
f  3.7 Re0.9 
−2 −2
 ε /D    5.74  
f1 =  −2 log  1 + 5.74  =  −2 log  0.0002297 + 0.9  
  3.7 Re10.9    
     Re1  
−2 −2
 ε /D    5.74  
f 2 =  −2 log  2 + 5.74  =  −2 log  0.0002757 + 0.9  
  3.7 Re0.9    Re2  
  2   
Columns 6 and 7 contain the frictional losses in each of the two pipes, which are computed as
L1 Q 2 10Q 2
hLf 1 = f1
D1 2 gA12
= f1
2(32.2)(0.7854)2
(
= f1 0.2517Q 2 )
L2 Q 2 100Q 2
hLf 2 = f 2
D2 2 gA22
= f2
2(32.2)(0.5454)2
= f 2 5.220Q 2 ( )
Columns 8 and 9 contain the sums of the minor losses in each of the pipes:
V12 (Q / 0.7854)2
hLm1 = ∑ K1 = ( 0.20 + 0.10 ) = 0.007552Q 2
2g 2(32.2)
V22 (Q / 0.7854)2
hLm 2 = ∑ K 2 = (1.00 + 0.30 + 0.10 + 1.00 ) = 0.01812Q 2
2g 2(32.2)
Column 10 is the total energy-head requirement for an elevation difference Z = 2 ft; this was com-
puted by adding 2 ft to the sum of the entries in Columns 6 through 9. Column 11 contains the total
energy-head requirements for an elevation difference of 20 ft; these were computed by adding 20 ft to
the sums of the entries in columns 6 through 9. Note that the entries in Column 11 are equal to those
in Column 10 plus the 18 ft of additional elevation difference. Head requirements for other ∆Z values
could be obtained in a similar fashion.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
hP (ft) hP (ft)
Q (cfs) Re1 Re2 f1 f2 hLf1 (ft) hLf2 (ft) hLm1 (ft) hLm2 (ft) (∆Z = 2 ft) (∆Z = 20 ft)
0 0 0 0.000 0.000 0.00 0.00 0.00 0.00 2.00 20.00
1 104,674 125,641 0.022 0.022 0.01 0.14 0.01 0.13 2.28 20.28
2 209,348 251,282 0.021 0.021 0.02 0.53 0.03 0.50 3.08 21.08
3 314,023 376,923 0.020 0.021 0.05 1.17 0.07 1.13 4.40 22.40
4 418,697 502,564 0.020 0.020 0.08 2.05 0.12 2.01 6.25 24.25
5 523,371 628,205 0.020 0.020 0.12 3.19 0.19 3.14 8.63 26.63
6 628,045 753,847 0.020 0.020 0.18 4.57 0.27 4.52 11.52 29.52
7 732,720 879,488 0.019 0.020 0.24 6.20 0.37 6.15 14.94 32.94
8 837,394 1,005,129 0.019 0.020 0.31 8.08 0.48 8.03 18.88 36.88
9 942068 1,130,770 0.019 0.020 0.39 10.20 0.61 10.16 23.34 41.34
10 1,046,742 1,256,411 0.019 0.020 0.49 12.58 0.76 12.55 28.33 46.33
11 1,151,417 1,382,052 0.019 0.020 0.59 15.20 0.91 15.18 33.84 51.84
12 1,256,091 1,507,693 0.019 0.020 0.70 18.07 1.09 18.07 39.87 57.87
13 1,360,765 1,633,334 0.019 0.020 0.82 21.19 1.28 21.20 46.42 64.42
14 1,465,439 1,758,975 0.019 0.020 0.95 24.55 1.48 24.59 53.50 71.50
15 1,570,113 1,884,616 0.019 0.020 1.09 28.16 1.70 28.23 61.09 79.09
Figure E13.3.2 illustrates the two required system curves and is a graph of the energy-head require-
ments in columns 10 and 11 of the preceding table versus the corresponding discharges in column 1
of that table.
Figure E13.3.2 System head curves for Example 13.3
Operating Point and Pump Selection

The reader should observe from the two preceding subsections that a pump perfor-
mance curve and a piping-system curve both show hP as a function of Q. Both curves
can be plotted on a single graph, as illustrated in Figure 13.14. The point where the
two curves intersect, the operating point, is the actual combination of discharge and
energy head that will be produced by the pump when installed in that piping system.
Figure 13.14
Plot of pump curve
and system curve to
determine pump
operating point
After the hP and Q of the operating point have been determined, the pump’s effi-
ciency, brake power requirement, and required NPSH at that operating point can be
determined from their respective curves, as described earlier. All of these operating
characteristics, as well as other issues addressed in this chapter, should be considered
when judging the acceptability of a pump’s performance. Generally, one seeks to
maximize the efficiency and minimize the power requirement, while meeting a
desired discharge goal.
The operating point will move as the suction and discharge water levels (and hence
system head curves) change. A pump can therefore operate at a very high flow rate
when the discharge side head is low but a much lower rate as the water level increases.
The engineer must consider operation over this full range of conditions. A pump will
run efficiently at one operating point but will lose efficiency as it moves away from
that point.
Cavitation and Limits on Pump Location

The physical properties of water, such as density and viscosity, depend on both tem-
perature and pressure. For most civil and environmental engineering problems, fluid
pressures are not expected to depart significantly from atmospheric pressure, and one
can neglect pressure variations and treat the fluid properties as functions of tempera-
ture only.
Pumps represent one instance in which neglect of pressure considerations can cause
problems with a design. As water enters a pump and nears the impeller, its velocity
increases. To compensate for this increase in kinetic energy, the fluid pressure (poten-
tial energy) must drop. If the absolute fluid pressure falls below the vapor pressure
corresponding to the fluid’s temperature, the fluid vaporizes [even though its tempera-
ture is typically well below the “normal” boiling temperature of 212°F (100°C)].
Water-vapor pockets form in the fluid as it passes through the low-pressure area.
When these vapor pockets subsequently reach a high-pressure area within the pump,
they implode. This process is called cavitation. The vapor-pocket implosions, com-
bined with any particulate matter in the fluid, erode the impeller and other internal
parts of the pump. Left unchecked, the pump will perform poorly and eventually fail.
During bench testing of a pump by its manufacturer for preparation of its performance
curves, tests are conducted to define its cavitation characteristics. These are usually
quantified through a parameter known as the required net positive suction head
(NPSH), defined as
p V2 p
NPSH = s + s − v (13.8)
γ 2g γ
where NPSH= required net positive suction head (ft, m)

ps = absolute fluid pressure at the entrance to the suction side of the pump
(lb/ft2, Pa)
Vs = flow velocity at the suction side of the pump (ft/s, m/s)
pv = vapor pressure of the fluid (lb/ft2, Pa)
To prevent cavitation damage in a pump installation, the available net positive suction
head must be at least as great as the required net positive suction head.
This can be accomplished by rearranging Equation 13.8 to yield the required mini-
mum absolute fluid pressure at the suction side of the pump:
 V2 
ps = pv + γ  NPSH − s  (13.9)
 2 g 

If the actual absolute fluid pressure is lower than the amount given by this expression,
cavitation can be expected and the design should be revised. Often, the pump can be
installed at a lower elevation (relative to the water level in the suction reservoir),
which will increase ps. If that is not possible, the suction piping should be redesigned
to reduce the energy losses, or a different pump should be selected, or both.
Example 13.4 – Determining Cavitation Limits. A pump installed in a dry-well draws

water from a suction reservoir having a water-surface elevation of 61 m. At its operating point, the
pump delivers a discharge of 0.12 m3/s. Energy losses in the suction piping for this discharge, includ-
ing frictional losses and all minor losses, total 1.7 m. If the pump installation is near sea level with
atmospheric pressure of about 101.3 kPa and if the NPSH requirement for the pump is 4.6 m, deter-
mine the maximum elevation at which the pump can be installed to avoid possible cavitation. The
diameter of the pump entrance is 300 mm. The water temperature is 15.5° C.
Solution: The cross-sectional area of the pump entrance is 0.07 m2, so the velocity at that point is
Vs = 0.12/0.07 = 1.71 m/s
From Appendix A, the vapor pressure of the water is 1.79 kPa. Using Equation 13.9 gives the fluid
pressure at the suction side of the pump to be at least
 1.712 
ps = 1790 + 9810  4.6 −  = 45.5 kPa
 2 ( 9.81) 
 
The required gage pressure is equal to this amount less the atmospheric pressure of 101.3 kPa or
ps = 45.5 − 101.3 = –55.85 kPa
The minimum required pressure head at the pump entrance is therefore
ps/γ = –55.84/9810 = –5.70 m.
The maximum permissible pump mounting elevation can be found by subtracting the suction-pipe
energy losses and the minimum pressure-head requirement from the suction-reservoir water-surface
elevation:
Max. permissible elevation = 61 – (−5.70) – 1.7 = 65 m
Cavitation should not occur if the pump is installed at an elevation lower than 65 m. Of course, one
would probably design the pump in this installation to be below the suction-reservoir water-surface
elevation of 61 m, thus minimizing the likelihood of cavitation.
Installations with Multiple Pumps

Multiple-pump installation facilities are often desirable because of the potential for
individual pump failure and because a pumping station may be required to handle a
wide range of flow rates. For cases in which the length of discharge piping is short
and the required pump head is small, it is economical to provide a separate piping sys-
tem for each pump. In this case, the pumps and their associated piping systems are
constructed in parallel and can be designed and analyzed independently, as previously
described.
For instances in which the length of discharge piping downstream of the pumps is
long, it may be economical to install a manifold such that two or more pumps
installed in parallel discharge to a common pipe. In either of these situations, an
equivalent pump performance curve must be derived from the individual pump per-
formance curves to determine the operating point for simultaneous pump operation.
The manner in which the equivalent pump curve is derived depends on the number of
pumps that operate simultaneously and whether the pumps are identical.
When two pumps are installed in parallel, energy considerations dictate that the
energy heads produced by the pumps must be equal. If for a given pump head hP one
pump delivers a discharge of Q1 and the second pump delivers a discharge of Q2, then
the total discharge delivered by the pair of pumps is Q = Q1 + Q2. The equivalent
pump curve for a system of two pumps in parallel is illustrated in Figure 13.15.
Again, this result is easily generalized to the case of three or more pumps.
Figure 13.15
Equivalent pump
curve for a system of
two pumps in parallel
Figure 13.16 shows a system curve, a pump curve for a single pump, and an equiva-
lent pump curve for two pumps in parallel. For this example, it is assumed that the
parallel installation is constructed with identical pumps and that there is a single dis-
charge piping system. From inspection of the operating points in the figure, it can be
seen that the parallel installation does not double the discharge delivered to the sys-
tem. In other words, the marginal benefit of the second pump, in terms of the increase
in discharge yielded, is less than the discharge produced by the operation of one pump
alone. If a third pump were added, its marginal benefit would be less than that of the
second pump, and so on. The cause of this behavior is the increased frictional and
minor losses in the piping system resulting from the increase in discharge. This is evi-
dent in Figure 13.16 from the increasing slope of the system curve.
Figure 13.16
Comparison of
performance of single
pump and pumps in
parallel
13.4 PIPING SYSTEM REQUIREMENTS

Because the pump discharge piping conveys stormwater under pressure, there are
stresses on the pipes that can, if not accounted for, result in leaks or even system fail-
ure. Therefore, pipe material, joint, and installation requirements must be given care-
ful consideration. Additional design issues center on valves and backflow prevention
and on pipe restraints and vibration isolation.
Pipe Requirements
As described in Section 11.1, pipe materials used in storm sewer design and construc-
tion are typically concrete or corrugated metal. Either of these materials can be used
for those portions of a pumping system where the pipe is buried below ground level.
The pipe, however, must be specified so that it will have the strength necessary to
Section 13.4 Piping System Requirements 557
resist the external and internal pressures expected to develop during pump operation
(see “External Forces” on page 460, “Internal Pressure Forces” on this page, and Sec-
tion 6.4, page 210). An appropriate pipe strength class and/or wall thickness must be
selected. Pipe-jointing methods should also be given consideration, as leakage could
cause erosion of soil materials surrounding the pipe and, ultimately, system failure.
Alternatives to concrete and corrugated metal pipes that are more desirable for use
within a pumping station are steel, ductile iron, and PVC pipes. Economy, consis-
tency, and ease of construction might even dictate that they be used for the entire
pumping system. Major considerations in pipe-material selection include size avail-
ability, jointing methods, and the need for special linings or coatings. Pressure pipe
standards published by the American Water Works Association (AWWA) may be con-
sulted for further guidance.
Within a pumping station, exposed pipes should be jointed, supported, and braced to
form a rigid structure capable of withstanding the hydrodynamic forces developed by
momentum changes in bends and fittings. In buried portions of the piping system,
concrete thrust blocks or similar restraints may be necessary at pipe bends. Sections
3.3 and 7.5 present methods for determining the magnitudes and directions of hydro-
dynamic forces.
Internal Pressure Forces

The figure is a free-body diagram of the left half of a The derivation of the previous equation assumes that the
circular pipe with diameter D and wall thickness t. A pipe walls are thin, or t < D/10. For thick-walled pipes
force balance in the horizontal direction in the figure (t > D/10), the tensile stress can be computed as (Seely
requires that the force caused by the internal pressure, p, and Smith, 1959)
must be equal to the force induced by the tensile stress,
2 2
σ, in the pipe walls. Considering a unit length of pipe pr i  r o
(into the page), this may be expressed as pD = 2tσ, or as σ = ---------------
2 2
 1 + ----2
ro – ri  r 
σ = pD
-------
2t where ri = inside radius of the pipe
ro = outside radius of the pipe
r = radius associated with the location where the
stress is desired (ri ≤ r ≤ ro)
It may be seen from this expression that the stress is the
largest at the inside face of the pipe wall, and that it is
the smallest at the outside face of the pipe wall.
In storm sewers and culverts, where internal fluid pres-
sures are usually small, the tensile stress predicted by
either of the two equations is usually negligible. How-
ever, internal fluid pressures in storm sewers and cul-
verts can, in some instances, give rise to a need for
special pipe joints to resist the internal pressures and
prevent leakage into the surrounding soils. Non-
negligible internal pressures and corresponding tensile
stresses in pipe walls can occur when pumps are placed
This tensile stress may be counteracted by a compres- in a system.
sive stress due to external loading on the pipe.
Valves and Fittings

The following subsections describe standard fittings used in pump installations, with
descriptions of their functions. Figure 13.17 shows a possible arrangement of these
elements in a pump station.
Figure 13.17
Example of a possible
arrangement of
equipment in a pump
station
Shutoff Valves. Shutoff valves, such as gate or butterfly valves, serve to isolate a
pump from the main piping network. They are often installed immediately upstream
and downstream of the pump so that it can be easily removed from service for mainte-
nance and eventual replacement. These valves should remain either completely open
or completely closed; they should not be used to throttle the flow.
Reducers and Flexible Connections. Reducers are required if the pump
inlet and discharge diameters are different from the piping system diameters and
should be eccentric (flat on one edge) to minimize air trapping. Reducers consisting
of flexible connections minimize the transfer of vibration energy from the motor and
pump to the piping system and may eliminate the need for expansion and contraction
joints if large temperature fluctuations will be encountered.
Flap Gates. Flap gates and/or check valves, which restrict flow to one direction,
are required for backflow prevention. Flap gates may be installed on the downstream
end of the pump’s discharge pipe and are desirable if tailwater conditions downstream
of that point could cause backups into the system. Check valves, installed between the
pump and the shutoff valve in the downstream discharge pipe, may be desirable if the
pipe is long and/or large in diameter and if its geometric configuration is such that
water would flow backwards during nonpumping times. Check valves are also
required to prevent backflows when two or more pumps are connected via a manifold
to a common discharge pipe. Check valves should be installed in horizontal portions
of the piping system and should be designed to prevent valve slamming, which can
sudden velocity changes and, consequently, hydraulic transients (water hammer).
Air-Release Valves. Air-release or vacuum valves permit trapped air to escape
from pipes during pump startup. Similarly, when pumps shut off, vacuum valves
enable air to enter the pipe to prevent its possible collapse by external forces.
Pressure Gauges and Clean-Outs. For purposes of operation and mainte-
nance, consideration might be given to installation of a pressure gauge on the pump
discharge pipe. The gauge can be used to periodically check pump performance and
can provide warnings of impending problems or failures. Clean-outs, which serve as
access points for easy removal of debris and blockages, can also be useful in storm-
water applications.
Section 13.5 Pump Controls 559
13.5 PUMP CONTROLS

Pump controls start and stop pumps based on water-level sensors in the wet-well or
temporary-storage facility. Sensors include floats, pressure transducers, and other
devices, and they have the capability to control several pumps simultaneously. Careful
consideration must be given to the design of pump controls, as an inadequate design
can lead to excessive pump cycling and premature motor failure. The pump-control
scheme also has implications for required temporary water-storage volumes; a poor
scheme may result in a larger storage volume than necessary. It is recommended that
designers work closely with pump manufacturers to ensure that a design does not vio-
late any of the pump’s mechanical constraints.
The American Association of State Highway and Transportation Officials Model
Drainage Manual (AASHTO, 1991) describes two types of pump-control methods
commonly used in stormwater pumping applications. In the common off-elevation
scheme, individual pumps start at successively higher water-level elevations, but they
all stop simultaneously at the same elevation (see Figure 13.18). This scheme is
advantageous if large amounts of suspended matter (such as sediment) are present in
the stormwater, as it limits the amount of time that solids have to settle.
Figure 13.18
Multiple-pump
control schemes
In the successive start/stop scheme, the start elevation for one pump is the stop eleva-
tion for the subsequent pump. In other words, the start elevation for pump 1 is the stop
elevation for pump 2, the stop elevation for pump 2 is the start elevation for pump 3,
and so on, as depicted in Figure 13.18. A benefit to this scheme is that pumps are suc-
cessively shut off as the water surface in the suction reservoir drops; thus, no more
pumps are operated than is necessary.
In addition to these schemes and their infinite variations, control schemes are
employed to equalize wear on the individual pumps in a system. Some of these
schemes simply alternate the first pump to start, while schemes that are more complex
continuously alternate all pumps during operation.
As indicated previously, start/stop elevations for individual pumps should be devised

to avoid short pump-cycling times. Because of the high power requirement and conse-
quent heat generation during pump startup, sufficient time must be provided between
starts to prevent problems with pump-motor overheating and failure. Temporary
water-storage volumes must be large enough to allow minimum cycling times; how-
ever, overly large storage volumes promote settling of suspended solids in the storm-
water, with consequent maintenance requirements. Although the pump manufacturer
should be consulted first, some guidance on minimum cycling times is provided in
Table 13.1.
Table 13.1 Guidelines for minimum pump cycling times (Brown, Stein, and Warner, 1996)
Motor HP Motor kW Min. Cycling Time (min)
0–15 0–11 5
20–30 15–22 6.5
35–60 26–45 8
65–100 49–75 10
150–200 112–149 13
13.6 STORAGE REQUIREMENTS

Because stormwater runoff rates and volumes tend to be large and quickly generated,
temporary storage (that is, detention) of the runoff is often required. The discharge
that can be economically pumped may be much lower than the peak rate of runoff;
thus, temporary storage offers a means to reduce the required pumping capacity. The
storage of stormwater can also provide water quality benefits because a portion of the
particles suspended in the water has time to settle out. When temporary storage is
needed (as it usually is), the designer must estimate not only the peak discharge from
the drainage basin, but also the time distribution of runoff in the form of a
hydrograph. The runoff hydrograph may be estimated using synthetic unit hydrograph
procedures (see Section 5.5), or, in the case of small drainage basins, using a
hydrograph-based application of the rational method.
As described in more detail in Section 13.7, determination of the required volume of a
temporary-storage facility is accomplished through a trial-and-error process that also
includes pump selection, piping diameters, and control strategies. However, to begin
evaluating the performance of alternative designs, Baumgardner (1981) presented a
technique that can be used to make a first estimate of the required temporary-storage
volume. As shown in Figure 13.19, the method is graphical and simply involves draw-
ing a horizontal line through the incoming runoff hydrograph at a discharge equal to
the average pumping rate. The volume of water under the hydrograph and above the
horizontal line is an estimate of the required storage capacity. If the inflow
hydrograph can be reasonably approximated as triangular (see Figure 13.20), similar
triangles can be used to estimate the required storage volume as
1  tb  2
S=   ( ∆Q ) (13.10)
2  Q p 

Section 13.6 Storage Requirements 561
where S = required storage volume (ft3, m3)

tb = base time of the inflow hydrograph (s)
Qp = peak inflow rate (cfs, m3/s)
∆Q = difference between peak inflow and outflow rates (ft3/s, m3/s)
Figure 13.19
Graphical estimation
of required storage
volume
Figure 13.20
Triangular
approximation of
inflow hydrograph
and required storage
volume
Example 13.5 – Determining Required Storage with Baumgardner’s

Technique. An inflow hydrograph is shown in columns 1 and 2 of the following table. Determine
the required storage volume if the maximum pumping rate is 0.071 m3/s.
Solution: The solution to this problem could be obtained by graphing the inflow hydrograph and out-
flow rate, as shown in Figure 13.19, and determining the volume between the two curves. However, it
will be solved instead using the numerical procedure illustrated in the following table.
Entries in column 3 in the table are the outflow hydrograph (i.e., the pumping rate); they are equal to
the inflow rate when the inflow rate is less than 0.071 m3/s and are equal to 0.071 m3/s when the
inflow rate is larger than that amount. Entries in column 4 are equal to the difference between entries
in columns 2 and 3; note that these values are zero when the inflow rate is less than or equal to
0.071 m3/s. Averages of two successive entries in column 4 are listed in column 5. Those averages,
multiplied by the time-step interval (5 min, or 300 sec), yield the incremental storage volumes in col-
umn 6. Finally, column 7 contains running sums of the incremental volumes in column 6. The last
entry in column 7, 682.8 m3, is an estimate of the required temporary-storage volume.
(1) (2) (3) (4) (5) (6) (7)

t (min) I (m3/s) Q (m3/s) I - Q (m3/s) Avg I - Q (m3/s) Incr. Vol. (m3) Cum. Vol. (m3)
0 0 0 0 0
5 0.003 0.003 0 0 0 0
10 0.006 0.006 0 0 0 0
15 0.009 0.009 0 0 0 0
20 0.011 0.011 0 0 0 0
25 0.014 0.014 0 0 0 0
30 0.017 0.017 0 0 0 0
35 0.020 0.020 0 0 0 0
40 0.023 0.023 0 0 0 0
45 0.025 0.025 0 0 0 0
50 0.028 0.028 0 0 0 0
55 0.031 0.031 0 0 0 0
60 0.034 0.034 0 0 0 0
65 0.071 0.071 0 0 0 0
70 0.127 0.071 0.056 0.028 8.4 8.4
75 0.326 0.071 0.255 0.156 46.7 55.1
80 0.538 0.071 0.467 0.361 108.3 163.4
85 0.609 0.071 0.538 0.503 150.8 314.1
90 0.481 0.071 0.410 0.474 142.2 456.3
95 0.340 0.071 0.269 0.340 101.9 558.2
100 0.184 0.071 0.113 0.191 57.3 615.5
105 0.142 0.071 0.071 0.092 27.6 643.1
110 0.113 0.071 0.042 0.057 17.0 660.0
115 0.099 0.071 0.028 0.035 10.5 670.5
120 0.093 0.071 0.022 0.025 7.5 678.0
125 0.076 0.071 0.005 0.014 4.1 682.1
130 0.071 0.071 0.000 0.003 0.8 682.8
135 0.065 0.065 0 0 0 682.8
140 0.059 0.059 0 0 0 682.8
145 0.057 0.057 0 0 0 682.8
150 0.054 0.054 0 0 0 682.8
Section 13.7 Pumping Station Design 563
Baumgardner’s technique is quite simple and does not take into account the actual
variation of pumping rate caused by system-head changes and pump controls, during
a storm event. The FHWA (1982) and Brown, Stein, and Warner (1996) have pre-
sented a mass curve routing procedure for estimation of the required storage capacity.
This routing procedure is a vast improvement over the simple graphical method, but it
still suffers from a lack of consideration of system head change effects on discharges.
The discharge delivered by a pump is typically quite sensitive to changes in head (see
Figure 13.11), and thus the mass curve routing method may yield inaccurate storage-
volume estimates. Use of an analysis method such as a fourth-order Runge-Kutta
approximation method as discussed in Section 13.7 will yield more accurate results.
13.7 PUMPING STATION DESIGN

Design of a pumping facility is a trial-and-error process involving evaluations of mul-
tiple combinations of design variables, such as storage volume, pump model, piping
diameters, and control system. None of these design variables should be determined
independently, as they are all related. Further, because of the large number of combi-
nations of design variables, experience and professional judgment play significant
roles in developing the alternative designs to be evaluated. The following sections
provide a general overview of the necessary design tasks.
Hydrology
One of the first tasks in pumping-station design is to estimate the design inflow
hydrograph to the pumping facility. Hydrologic methods discussed in Chapter 5 can
be used for this purpose. In locations such as highway or roadway depressions, storm-
water pumping installations are typically designed to convey 50-year design storm
flowrates (FHWA, 1982; Brown, Stein, and Warner, 1996), but it is also recom-
mended that checks of station performance and flooding limits be made for the 100-
year design storm event. Designs for more frequent return periods may be acceptable
in some cases, but should be agreed upon between the designer, owner, and appropri-
ate review agencies. In areas where storms occur quite frequently, water-storage facil-
ities may not drain completely between storms, and continuous simulation may be
necessary to assess the adequacy of a design.
Development of a Trial Design

The initial trial-design variables can be selected after the design inflow hydrograph
has been established. As noted earlier, the trial-design variables that must be selected
for any evaluation include:
• Piping diameters and layouts
• Pump models
• Temporary water-storage volume
• Pump-control strategies
All of these variables are interrelated and should not be chosen independently. For
example, choosing excessively small pipe diameters or a small storage volume will
necessitate the use of large pumps; poor design of the control strategy may result in
excessive pump cycling times; and so on.
The first design variable that should be selected is the size and layout of the piping
system. Piping-system layout is usually governed by legal right-of-way and
discharge-point considerations. Piping diameters should be chosen to maintain flow
velocities within reasonable bounds. Reasonable velocities are important not only for
hydrodynamic reasons (such as the hydraulic forces that develop at bends and fit-
tings), but also to minimize the effects of hydraulic transients (such as water ham-
mer). The designer should also consider at this stage whether to use a dry/wet- or dry-
well installation, as this choice will affect the economics of the trial design.
Subsequently, system curves can be developed from the piping-system design. Based
on these system curves and an estimate of the desired maximum pumping rate, manu-
facturer guidance and pump performance curves are used to select an initial trial
pump model. Maximum pumping rates are determined by downstream system capac-
ities, legal requirements limiting downstream peak discharges to predevelopment lev-
els, or economics.
Pumping stations are usually equipped with, at a recommended minimum, two or
three pumps, although guidelines vary. Consideration should be given to oversizing
the pumps to partially compensate for pump failure. For example, in a two-pump sys-
tem, the pumps may be sized so that each is capable of handling 70 percent of the
total design flow. In a three-pump system, each pump might be capable of handling
50 percent of the design flow. The firm capacity of a pumping installation is generally
taken as the maximum discharge that could be delivered by the facility if one pump
were out of service. When designing pumping stations with multiple pumps, it is
advantageous if all the pumps are identical to facilitate ease of maintenance.
An initial estimate of the storage-volume requirement can be obtained graphically, as
described in Section 13.6. If storage is to be provided underground, this estimate can
be used to select trial storm-sewer pipe and/or storage-tank dimensions. If the storage
is to be provided in a surface-water pond, a topographic map can be used to contour
the pond to provide the necessary volume.
Development of a trial control strategy is based on the design inflow hydrograph, the
selected pump make and model, and the storage-volume characteristics. As already
noted, the objective here should be to maintain minimum pump-cycling times.
Trial Design Evaluation

After a set of trial-design variables has been selected, as described in the preceding
subsection, the task of routing the design inflow hydrograph through the facility can
proceed. This routing process, carefully performed, will provide information on
actual pumping rates and efficiencies, pump-cycling times, storage-volume adequacy,
and other characteristics. These results can then be reviewed to assess the acceptabil-
ity of a design and develop cost estimates that can be compared to those of other
alternatives.
Section 13.7 Pumping Station Design 565
Because of the relative complexity of flow routing in a pumping facility (as compared
to a simpler storage facility without pumping, such as a detention pond), flow routing
should be accomplished using a numerical integration approximation method such as
the Runge-Kutta method presented in Section 12.7 rather than the storage indication
method. Runge-Kutta algorithms are also described by Chow, Maidment, and Mays
(1988) and by Chapra and Canale (1988). Application of the Runge-Kutta method
requires the development of stage versus discharge and stage versus water-surface
area relationships for the trial design.
As with detention analysis, the stage-discharge relationship, O(h), describes the dis-
charge O from a storage facility as a function of the water-surface elevation (or stage)
h of the suction reservoir or wet-well. In the context of pumping-station design, this
relationship is dictated by the pump-control strategy and by the loci of operating
points defined by the pump and system curves. Note here that multiple system curves
must be considered, corresponding to the different suction-reservoir elevations
expected to occur during passage of a stormwater runoff event through a facility. It is
important to consider the effects of the suction reservoir or wet-well stage on the
actual pumping rate, because the pumping rate is typically quite sensitive to energy-
head changes. It is also important to recognize that the O(h) relationship will exhibit
sudden changes (discontinuities) at stages where pumps start or shut off. The relation-
ship may also be different during periods when the reservoir level is receding than
when it is rising (called a hysteresis phenomenon), depending on the pump-control
strategy.
A stage versus water-surface-area relationship, As(h), describes the suction-reservoir
water-surface area, As, as a function of the water-surface elevation h. This relationship
can be developed from geometric relationships for underground storage facilities, or
from a topographic map depicting contours of a surface-water facility.
Given the design inflow hydrograph I(t), and the O(h) and As(h) relationships, routing
of flows through a pumping facility is accomplished using the following form of the
continuity equation:
dh I (t ) − O(h)
= (13.11)
dt As (h)
where h = stage (ft, m)

t = time (s)
I(t) = inflow at time t (cfs, m3/s)
O(h) = outflow at stage h (cfs, m3/s)
As(h) = water surface area at stage h (ft2, m2)
Integration of this equation (that is, determination of the relationship between h and t)
can be accomplished using the Runge-Kutta method. Once h is known as a function of
t, the discharge Q and other variables of interest can also be found as a function of t.
13.8 MISCELLANEOUS DESIGN CONSIDERATIONS

When designing a pump station, the design engineer must consider a number of fac-
tors beyond selection of the pump size and arrangement and the sizing of the wet well.
Among these are the station’s architecture, power requirements, debris removal facili-
ties, monitoring devices, safety, and accessibility.
Architectural Considerations
In some cases, such as when submersible pumps are employed, stormwater pumping
stations are completely below ground level. Because the ground around the pump sta-
tion may become saturated during storm events, the designer should consider the pos-
sibility that the station may become buoyant. Adding weight to the station or
providing valves that drain the groundwater into the wet-well can keep the station
from floating. The structural design of pump stations can vary from completely
underground stations to large, attractive structures as shown in Figure 13.21.
Figure 13.21
Flood control pump
station structure
If the pump station is above ground, it is often designed as an integral, nonobtrusive

part of the surrounding neighborhood. Pump stations can be designed to be aestheti-
cally pleasing without significantly increasing their cost through the use of clean lines
and textured finishes (such as textured concrete). Strategically placed screening and/
or landscaping features can further improve the station’s appearance and can help
reduce noise propagation during periods when the station is operating. Because of the
need for frequent maintenance of pump stations, adequate access roadways and park-
ing space must be provided for maintenance personnel.
Section 13.8 Miscellaneous Design Considerations 567
Power and Control Requirements

Pump stations are usually powered by electricity, but they can also be powered by die-
sel, gasoline, or natural or LP gas engines. Electrical power, however, is usually the
most cost-effective and reliable option if electrical service is readily available. Storm-
water pumps usually are equipped with constant-speed, three-phase induction motors.
Motor voltages depend on the size of the pump, but are generally 440 to 575 V, with a
maximum recommended motor capacity of 300 HP (225 kW).
Power-system failures during storms can cause pumps to stop, with consequent flood-
ing of adjacent lands and improved areas. In locations where temporary flooding can-
not be tolerated, a standby power supply must be provided. In some instances, two
independent electric feed lines may be supplied to the station, with an automatic
transfer switch controlling which feed is used. In most cases, a standby generator is
required, which can be powered by a diesel, gasoline, or LP or natural gas engine.
In addition to high-voltage power wiring, pump stations can contain a great deal of
low-voltage control wiring connecting various sensors to control equipment and the
SCADA (supervisory control and data acquisition) system. Control wiring should be
shielded from power wiring per the manufacturer requirements. The wiring behind a
typical control panel is shown in Figure 13.22.
Trash Racks and Grit Chambers

Stormwater often conveys considerable amounts of trash and debris, which, if not
removed from the flow, could damage or clog the pumps. Trash racks to filter out
debris can be manufactured from steel bars. Bar spacing should be about 1.5 inches,
but may depend on the maximum size of a sphere that can be passed by the pumps.
Trash racks should be installed in modular units to facilitate easy removal and should
be inclined from the vertical to facilitate self-cleaning as water levels rise and fall.
Clear-opening areas of trash racks should be large enough to not cause excessive
energy-head losses in the flow. In some cases, screening may be accomplished at

inlets to the stormwater system, negating the need for additional screens at the pump
intakes. A typical trash rack for a large pump station is shown in Figure 13.23.
Figure 13.22
Wiring inside a
typical control panel
Figure 13.23
Trash rack for pump
station
In areas where roadways are routinely sanded during winter storms or in other loca-
tions where stormwater may contain significant accumulations of sediment, it may be
desirable to provide a grit chamber where solids can settle out of the flow. The grit
chamber may be designed to be either within or upstream of the wet-well, depending
on its ease of accessibility for required sediment removal. An efficient means for sed-
iment removal should be provided.
Monitoring Devices
Power outages and mechanical problems are routinely encountered at pumping facili-
ties. Warning lights and remote alarms can be employed to notify appropriate mainte-
nance personnel, who can then take necessary actions. Alarms are provided for power
supplies, pump operation, high water levels, presence of explosive gases, and/or sta-
tion security.
Safety and Ventilation Requirements

A safety engineer should perform a thorough review of a pumping-station design.
Safety shrouds should be provided around moving parts, and adequate space must be
provided for maintenance personnel to access motors, pumps, and electrical panels.
Ladders and stairwells should be provided, as necessary, and lighting, heating, and
ventilation should enable personnel to perform their duties in a safe and efficient man-
ner. In some cases, a pumping station may be classified as a confined space for which
special access and safety requirements are defined by code.
Wet- and dry-wells should always be provided with ventilation equipment such as
exhaust fans to ensure a safe working environment for maintenance personnel.
Exhaust fans should be switched on at least 10 minutes prior to entry to a wet- or dry-
well. In some cases, it may be desirable to also provide automatic switching equip-
ment for fans. For example, a sensor could be provided to start the fans if explosive
gases were detected.
Doorways, Hatches, and Monorails

Because of the need to periodically remove pumps, motors, or other equipment for
maintenance or replacement, all doorways and hatches in a pumping station should be
large enough and appropriately positioned to enable these operations.
In some cases, especially in small stations, portable hoisting units can be used for
equipment removal. Often, however, it is desirable to provide monorails and lifting
tackle as apart of the station infrastructure.

In some locations, it may be necessary to use a pump station to remove stormwater.
For instance, stormwater may have to be pumped through a levee or floodwall. Pump
types used in stormwater applications include axial-, radial-, and mixed-flow pumps,
and positive-displacement pumps such as screw pumps.
Initial pump selection is typically based on the specific speed characteristic. High spe-
cific speeds correspond to pumps delivering high discharges at low heads, and low
specific speeds correspond to pumps that produce lower discharges at high heads.
Axial-flow pumps tend to have high specific speeds, and radial-flow pumps tend to
have lower ones.
Pump stations can have either a wet-well or wet/dry-well configuration. With wet-
well installations, the pump is submerged in a pool of water in the sump. In a wet/dry-
well installation, the pump is not submerged in the wet well; rather, it is located in an
adjacent dry well and is connected to the wet well with piping. Wet/dry-well installa-
tions are usually associated with larger pump stations.
Pump performance curves are used to determine how the pump will operate under a
range of conditions. A pump characteristic curve shows the relationship between flow
rate and pump head. Other curves show pump efficiency at various operating points
and the minimum NPSH required by the pump to prevent cavitation. A system curve
that reflects the head loss in the pumping system for a range of discharges can be plot-
ted with the pump characteristic curve; the intersection of these curves is the operat-
ing point of the pump for the assumed upstream and downstream boundary conditions
and flow rate.
Because of the possibility of pump failure and the wide range of flow rates that a
pump station must be able to handle, multiple pumps may be installed. An equivalent
pump performance curve for both pumps operating at the same time should be devel-
oped to see how the system will perform under this condtion.
In addition to the pumps themselves, stormwater pumping systems include system
piping and several types of valves and fittings. The engineer must ensure that the
specified pipes have sufficient strength to resist the internal and external forces that
they will be subjected to. Typical pumping systems include shutoff valves, reducers,
flap gates, air-release valves, gauges, cleanouts, and pump on/off controls.
Several methods exist for estimating the amount of storage needed in the wet-well. A
simple approach for preliminary volume estimation is Baumgardner’s method, in
which the required volume is taken as the area under the inflow hydrograph in excess
of the average pumping rate.
Prior to development of a trial pump station design, the inflow hydrograph for one or
more design storms must be developed. Methods such as those presented in Chapter 5
may be used for this purpose. Other elements of the trial design are pipe sizing and
layout, pump selection, wet-well volume, and a pump control scenario. Once a trial
design has been developed, it is evaluated by considering stage-discharge and stage-
surface area (or volume) relationships and applying a routing technique to solve the
continuity equation. The principles involved are similar to those used in routing
detention ponds, which were presented in Chapter 12.
Additional considerations in developing the final design of the pumping station sta-
tion architecture and power requirements, facilities for removing grit and debris, mon-
itoring devices to alert personnel to malfunctions, safety measures, adequate
ventilation, and pump accessibility.
References 571
REFERENCES
Baumgardner, R. H. 1981. “Estimating Required Storage to Reduce Peak Pumping Rates at Storm Water
Pumping Stations.” Washington, D.C.: Federal Highway Administration Workshop Paper, Office of
Engineering.
Chapra, S. C. and R. P. Canale. 1988. Numerical Methods for Engineers, 2d ed. New York: McGraw-Hill.
Chow, V. T., D. R. Maidment, and L. W. Mays. 1988. Applied Hydrology. New York: McGraw-Hill.
Federal Highway Administration (FHWA). 1982. Highway Storm Water Pumping Stations, Vol. 1 and Vol.
2. FHWA–IP-82-17. Washington, D.C.: U.S. Department of Transportation.
Hydraulic Institute. 1998. Centrifugal/Vertical Pump Intake Design. Parsippany, New Jersey: Hydraulic
Institute.
Karassik, I. J., J. P. Messina, P. Cooper, and C. C. Heald. 2000. Pump Handbook, 3d ed. New York:
McGraw-Hill.
Mays, L. W. 2001. Water Resources Engineering. New York: John Wiley and Sons.
Sanks, R. L. 1998. Pumping Station Design, 2d ed. Boston: Butterworth-Heinemann.
Seely, F. B., and J. O. Smith. 1959. Advanced Mechanics of Materials. 2nd ed. New York: John Wiley and
Sons.
US Army Corps of Engineers. 1995. General Principles of Pumping Station Design and Layout. EM 1110-
US Army Corps of Engineers. 1989. Structural and Architectural Design of Pumping Stations. EM 1110-2-
3104. Washington, D.C.: USACE.
US Army Corps of Engineers. 1999. Mechanical and Electrical Design of Pumping Stations. EM 1110-2-
PROBLEMS
13.1 A stormwater storage system consists of a 158-m–long, 1220 mm pipe and a 6.1-m–diameter wet
well as shown in the following figure. The storage volumes of the two components as a function of
elevation are given in the table. Construct a stage-storage curve showing the storage in the system as
a function of the elevation.
Volume (m3)
Elevation (m)
Pipe Tank
0.00 0.0 0.0
0.15 1.3 4.9
0.30 7.1 9.8
0.46 19.0 14.7
0.61 37.7 19.6
0.76 62.6 24.5
0.91 90.2 29.4
1.07 118.0 34.3
1.22 143.5 39.2
1.37 163.4 44.1
1.52 176.3 49.0
1.68 183.0 53.9
1.83 184.9 58.8
1.98 184.9 63.7
2.13 184.9 68.6
13.2 A pump at a small stormwater pump station is designed to pump from a sump with a water level that
can vary from 510 to 515 ft to the other side of a levee with water level during a flood that can vary
from 518 ft to 525 ft. The design flow rate is approximately 2,500 gpm (5.58 cfs). The suction piping
has a diameter of 12 in. and is 20 ft long. The discharge piping also has a diameter of 12 in. and a
length of 150 ft. All pipes have a relative roughness of e = 0.0015. The sum of all minor loss coeffi-
cients in the station is K = 15.
Compute and plot the system head curves for the minimum (3 ft) and maximum (15 ft) on the pump
curves in Figure 13.11. Select the impeller diameter that will yield close to 2,500 gpm for the maxi-
mum lift.
What are the flow, head, horsepower, and NPSH required for the maximum and minimum lift?
Record your answers in the table provided.
Property Max Lift Min Lift

Flow, gpm
Head, ft
NPSH required, ft
Power, hp
13.3 A pump delivers runoff from a detention pond to a spillway through an 800-ft–long, 24-in.–diameter
concrete pipe. The performance curve for the pump is shown. The pump starts when the water sur-
face elevation in the detention pond is 70 ft and continues to operate until the pond is empty with a
bottom elevation of 60 ft. What power must be supplied to the pump when the water surface eleva-
tion at the spillway is 100 ft and the elevation in the detention pond is 60 ft? Use a relative roughness
height ε = 0.005 ft.
Problems 573
13.4 The pump and system curves for a proposed station are given by the coordinates in the table. What is
the operating point for a single pump in this system? For two identical pumps in parallel?
Q (L/s) Pump Head (m) System Head (m)

0.0 40.2 15.2
31.5 39.0 20.7
63.1 33.8 37.8
94.6 21.3 70.1
13.5 Figure 13.18 illustrates two types of pump control schemes described in this chapter. For this prob-
lem, four pumps are located at the bottom of a 20-by-20–ft wet well. The bottom of the wet well is at
an elevation of 0.0, and, for the distances shown in Figure 13.18, h1 = h2 = h3 = h4 = 6 ft. Assume
that the pumps are identical, and that each has its own piping system. A constant flow of 10 cfs
enters the wet-well for a period of 60 minutes.
If the discharge produced by any single pump depends on the wet-well water surface elevation h as
Q = 400 + 25h (where Q is in gpm and h is in ft), compute and plot stage-discharge curves for each
of the two control strategies. In other words, graph the discharge produced by the combination of
four pumps as a function of the water surface elevation in the wet well. Be careful to note that the
relationships may exhibit hysteresis.
Stormwater
regulations are
concerned with
protecting both
natural and
constructed
environments from
damage by flooding
and pollution.
C H A P T E R
14
Regulatory and Environmental
Issues
Stormwater conveyance systems must be designed to meet minimum requirements

specified by local, regional, state, or national review agencies. These requirements
relate to both stormwater quality and quantity. This chapter provides a broad overview
of regulatory and environmental issues, but cannot accommodate all of the differences
from one political jurisdiction to another.
Section 14.1 provides an international overview of water resources guidelines and
regulations. World Bank water resources policies for international development and
aid, policies of the European Union, and water-related environmental regulation in
Canada are discussed. A summary of the U.S. federal, state, and local laws and regu-
lations applicable to urban stormwater management is presented in Section 14.2.
14.1 INTERNATIONAL OVERVIEW OF REGULATORY

GUIDELINES AND POLICIES
Stormwater management practices throughout the world vary widely. Industrialized
countries typically have policies concerning flood control and water quality in place,
but many developing countries do not. Further, the ways in which stormwater can be
effectively managed differ with the level of urbanization, local climate and topogra-
phy, and a host of other factors.
Because of the enormous range of differences in stormwater management needs and
practices around the globe, it is difficult to present even a summary in the space of a
chapter. The goals of this section are to provide the reader with an overview of inter-
national stormwater-related policies, some insight into stormwater management in
developing countries, illustrative examples of stormwater practices in the European
Union, and an overview of stormwater regulations in Canada.
576 Regulatory and Environmental Issues Chapter 14
Milestones in International Water Resources Policies

The major milestones in the development of current international water resources pol-
icies are summarized in Table 14.1 and briefly reviewed here. The 1977 United
Nations Mar del Plata Conference (United Nations, 1977) was the first to identify the
need for Integrated Water Resources Management (IWRM). The assessments pre-
sented at this conference showed that the lack of adequate water supply and sanitation
caused many more premature deaths and much more illness than previously believed.
Consequently, a call was made to designate the 1980s as the “International Drinking
Water Supply and Sanitation Decade.” Furthermore, it was concluded that lack of
international coordination was a major reason for the poor state of the world's water
affairs. The conference report is known as the Mar del Plata Action Plan (United
Nations, 1977).
Table 14.1 Milestones in international water resources policies

Year Agreement Milestone
1977 Mar del Plata Action Plan • Identified need for integrated water resources management
1987 Brundtland Commission – “Our
• Shifted focus to sustainable development
Common Future”
• Promotion of a dynamic, interactive, iterative, and multi-
sectoral approach to water resources management
1992 Agenda 21
• Projects and programs that are both economically efficient
and socially appropriate
• Defined problems in urban areas to include air, soil, and
1996 Brown Agenda
water pollution, and noise and traffic congestion
• Calls for dedicated local actions to address environmental
1997 Local Agenda 21
problems
1997 Charter of European Cities and
Towns Towards Sustainability • Consensus among cities and towns on sustainability issues
(Aalborg Charter)
In 1987, the World Commission on Environment and Development (popularly known

as the Brundtland Commission) presented its report Our Common Future (WCED,
1987). The creation and general acceptance of this report was a watershed event in
that it marked the first time the world community agreed on the need to reverse cur-
rent development trends and instead focus on sustainable development. The report
called for an international conference on environment and development—a confer-
ence that eventually took place in Rio de Janeiro in 1992. The major outcome of the
1992 United Nations Conference on Environment and Development was Agenda 21
(United Nations, 1992), a framework for the implementation of sustainable develop-
ment throughout the world during the twenty-first century.
Agenda 21 and Water Management

Agenda 21 states that the general objective of water management is to make certain
that adequate supplies of good-quality water are maintained for the entire human pop-
ulation, while preserving the hydrological, biological, and chemical functions of eco-
systems; adapting human activities to remain within the capacity limits of the natural
Section 14.1 International Overview of Regulatory Guidelines and Policies 577
environment; and combating vectors of water-related diseases. In order to fully utilize

the earth's limited water resources and to safeguard those resources against pollution,
innovative new technologies, as well as existing indigenous technologies, are needed.
Among the principal objectives of Agenda 21 (United Nations, 1992) are:
• The promotion of a dynamic, interactive, iterative, and multisectoral
approach to water resources management that integrates technological,
socioeconomic, environmental, and human health conditions; and
• The design, implementation and evaluation of projects and programs that are
both economically efficient and socially appropriate within clearly defined
strategies, and that calls upon the full participation of indigenous people and
local communities (including women and children) in water management
policy-making and decision-making.
The promotion of integrated water, sanitation, drainage, and solid waste management
(environmental infrastructure) is cited as central to the efforts for sustainable urban
development. Agenda 21 also stresses the importance of adequate environmental and
social impact assessments. It furthermore advocates the application of precautionary
measures in regard to water resources management with a focus on waste minimiza-
tion and prevention.
Chapter 18 of Agenda 21 lays down the fundamental principles for sustainable water
resources management. The chapter builds on the Mar del Plata Action Plan, but adds
relevant lessons from the Water Supply and Sanitation Decade, which suffered from
poorly targeted intervention projects and a lack of coordination and integration. It has
been widely accepted and forms the basis of guidelines and policies for the UN family
of agencies and for most international development aid agencies. Many countries are
also in the process of adopting these principles in their environmental and water
resources legislation.
Local Agenda 21. In 1997, a worldwide Local Agenda 21 (United Nations,
1997) was formulated. This document makes a plea for dedicated local actions, in
both industrialized and developing countries, that combine the reduction of environ-
mental degradation with improvements in local socioeconomic conditions. It was
adopted by the UN General Assembly Special Session on the Environment on June
23, 1997.
The European version of Local Agenda 21 is called the Charter of European Cities
and Towns Towards Sustainability (ECSCT, 1994), or Aalborg Charter. The Aalborg
Charter aimed at building consensus among its members' cities and towns on 14 key
sustainability issues. Many local authorities in Europe are now developing Local
Agenda 21s, and ponds for stormwater treatment and daylighting of stormwater con-
veyance systems are frequent components of these plans.
Additional International Guidelines

The Brown Agenda. The 1996 United Nations Conference on Human Settle-
ments, in Istanbul, commonly referred to as Habitat II, stated that health problems
related to adverse environmental conditions and ineffective and inadequate health ser-
vices exact a heavy toll on the quality of life and the overall contribution to society of
millions of people (United Nations, 1996). Particular adverse environmental condi-
tions identified at the conference were lack of access to safe water and sanitation,
inadequate waste management, poor drainage, air pollution, and exposure to exces-
sive noise levels. The unique environmental problems of urban areas identified in the
Brown Agenda include air, soil, water, and noise pollution, and traffic congestion.
These have a direct and immediate impact on human health and safety, especially for
the poor, and on business productivity. Habitat II concluded that implementing the
Brown Agenda (improving water supply, drainage, sanitation, solid waste manage-
ment, etc.) remains a priority for most cities in developing countries.
Policies and Guidelines in International Development Aid. The
World Bank (WB) is the largest funding institution for international development
projects, and its policies and guidelines are adopted by many other international
development agencies. WB policies are widely criticized by NGOs and others, and
since the 1990s, public pressure brought to bear against the WB has led it to adopt
and/or strengthen a number of environmental policies that govern its lending. The WB
is in the process of converting its old Operational Directives into Operational Policies,
commonly known as the 4.0 series.
The Operational Policies differentiate between, and outline requirements for, sectorial
approaches and integrated approaches to development projects. In the sectorial
approach to a water project, for example, only water-resource-related issues need be
considered. The integrated approach requires an environmental assessment (EA) to
review a project’s effect on the natural environment, human health, and societal/cul-
tural establishments. WB policies require an EA to be conducted by independent
experts for those projects likely to have potentially significant adverse environmental,
human and/or societal/cultural impacts.
The WB developed a water resources policy paper entitled Water Resources Manage-
ment (World Bank, 1993). This paper promotes a major shift in water resources man-
agement, from a fully segregated approach to a comprehensive and integrated
approach. It specifically calls for linking land-use management to sustainable water-
resource management and for addressing quantity and quality concerns in an inte-
grated manner.
The current WB philosophy is that the world cannot afford to view the environment as
a technical issue to be addressed independently from overall municipal and industrial
strategic decision-making. This thinking is best summed up as “environmental man-
agement, not just pollution control” (World Bank, 1998). This means that WB poli-
cies are meant to affect, among other things, inter alia stormwater management.
However, it is debatable whether the WB has been able to implement this new
approach. Consequently, in February 2003, the WB’s executive board approved and
released The New Water Resources Strategy (World Bank, 2003). This strategy is
strongly based on the 1993 Water Resources Management policy paper and intends to
vigorously implement its policies.
Much of the WB’s urban sanitation work is done in cooperation with the United
Nations Development Programme’s (UNDP) Water and Sanitation Program (WSP).
The WSP promotes the Strategic Sanitation Approach (Wright, 1997), a key principle
of which is that development should be demand-led. This principle provides that

projects should develop those types of systems that beneficiaries are asking for and
are prepared to pay for. This approach demands extensive and informed dialogue
between the beneficiaries and the project developers in order to achieve success.
The UNDP states that some of the most important results that could presently be pro-
duced in the field of international development involve water. The organization is in
the process of shifting the direction of its efforts to be in line with Agenda 21. To this
end, the UNDP has produced a Strategy Framework Document (UNDP, 1998).
The United Nations Environment Programme (UNEP) has found that technology and
technology selection for improving water quality, although very important, cannot be
detached from other equally important factors for successful project implementation.
A framework is required that considers integrated waste management, transsectorial
issues, and the effect on all stakeholders, while granting each a representative voice.
Simply solving the problem of wastewater drainage without taking into account solid
waste and stormwater drainage will not provide improvements sufficient to protect
either public health or the environment. To promote sustainable urban development,
the International Environmental Technology Centre (IETC) of the UNEP is producing
an International Source Book on Environmentally Sound Technologies for Wastewater
and Stormwater Management (IETC, 2002), to be published by the International
Water Association. The IETC has also developed a Framework for Wastewater and
Stormwater Management, which can be found on UNEP’s web site (www.unep.org).
The European Commission Directorate-General for Development is another major
source of development aid for projects that implement the Agenda 21 principles. The
protection and enhancement of natural resources are recognized as important dimen-
sions to the development support provided by the European Union (EU). Towards
Sustainable Water Resources Management (European Commission, 1998) states that
attention must be paid to the improvement of urban drainage, the prevention of floods,
and the extension of capacity to withstand periodic droughts.
These directives and documents complement the guiding principles for international
development projects. As such they represent a conglomeration of “soft themes and
policies” coupled to “hard technical realities” in developing water resources. This is a
direct consequence of the complexity and subtle nuances in addressing combined
human and environmental issues rather than just technical implementation issues.
Naturally, the technical implementation of water-related development projects is rec-
ognized as important, and it is the central necessity of integrated management systems
to be able to interlink the “soft themes” with the “hard technical realities.”
The Urban Stormwater Initiative (Environment Australia, 2002) is an example of a
combined “hard” and “soft” effort aimed at enhancing water quality in the waterways
of major coastal cities by improving stormwater management. It promotes the con-
struction of infrastructure that is innovative and exemplifies best-practice techniques
in managing the quality of urban stormwater. Particular attention is given to projects
that turn a problem into a resource, such as reduction of water shortages through the
capture, treatment, and reuse of urban stormwater. Strong industry source control,
community education, and water-sensitive urban design are also an integral part to
this effort.
Effective interagency and intersectorial planning is essential. Therefore, intersectorial

policy and planning weaknesses need to be identified, and water should form a part of
overall urban planning. Water resources management in the twenty-first century will
become increasingly influenced by the principles outlined in Agenda 21, especially in
developing countries. The complete integration of technical realities with the limita-
tions on the environmental and human capacity for change in short periods of time
will become a central aspect of water resource management and development.
Drainage Management in Developing Countries

In developing countries, urban drainage channels often serve as trash dumps during
dry weather conditions. This has two significant consequences. First, when precipita-
tion events occur, there is a great risk of channels clogging, leading to flooding and
the dispersal of human refuse. Second, the runoff entrains the garbage, contributing to
pollution in receiving waters. In addition, many storm conveyance systems function
as combined sewers transporting untreated sanitary wastewater. These factors have
severe effects on public health that can be prevented through urban drainage infra-
structure improvements. In order to develop efficiently functioning urban drainage
systems, both an urban development plan and a master drainage plan are necessary.
Most cities in developing countries grow quickly and with little coherent planning. A
lack of adequate urban land-use planning and enforcement, coupled with very limited
available funds for the provision of basic water services, make the development of
efficient drainage systems a difficult task. To overcome these difficulties and success-
fully implement improvements, the technical tools and management principles devel-
oped in industrialized countries must be adapted to a developing country's present
conditions and needs. The following examples from Thailand and India illustrate
these points.
Thailand. A 2002 WB report on environmental degradation in Thailand (World
Bank, 2002) illustrates commonly encountered legal and institutional problems asso-
ciated with improving environmental conditions in developing countries. Although
Thailand has made significant progress in establishing environmental laws and insti-
tutions, polluters have had little incentive to control and abate pollution. Implementa-
tion of regulatory measures has suffered from weak monitoring and enforcement. The
report notes that Thailand's existing command and control measures have not been
particularly effective. Implementation of existing regulations has been further ham-
pered by duplication and overlap of responsibilities between the numerous agencies
and departments involved in pollution control.
This situation makes a joint approach to urban drainage and solid waste management
both difficult and vitally important. Because existing drains are frequently open chan-
nels misused for garbage disposal, the regulatory, enforcement, and technical
approaches toward both water resources management and solid waste management
must be fully integrated in order to be effective.
Indore, India. The World Resources 1996-97 Report (World Resources Institute,
1997) underlined the importance of a community's perception of health risks, such as
polluted water, inadequate drainage, or lack of garbage collection, to the successful
development of an efficient water resources infrastructure. The report cites the experi-
ence and behavior of the inhabitants of Indore's slums as an example. The majority of
the slums are located along the banks of the many rivers and canals that crisscross the
city and on the city's floodplains. Monsoon waters regularly flood these communities.
Although the residents are acutely aware of the fact that the diseases they continu-
ously suffer are tied to the flooding, social and economic factors prevent them from
leaving these high-risk areas. City planners designed a project to improve health con-
ditions in the slums by reducing flooding in these areas and the city proceeded to con-
struct closed drainage channels and upgrade sanitary conditions. Despite these
improvements, many community members now perceive flooding problems to be
worse than before.
An important source of dissatisfaction with the new drainage system is the reduced
ability of the community to apply its own risk-reduction strategies. Because the drains
are closed, residents can no longer predict the severity of an approaching flood. With
the open drains, community members would adjust their flood response behavior
according to both rainfall intensity and the volume of water in the open drains. Fur-
ther, unlike the wider open channels, the closed channels and their inlets are easily
blocked by plastic bags and trash.
Surat, India. Surat, a city located in the Indian state of Gujarat, suffered from an
epidemic outbreak of plague (a disease caused by the bacterium Yersinai pestis) in
1994. The faulty drainage system could not cope with an enduring rain period, and the
result was large-scale waterlogging. Hundreds of cattle and other animals died
because of the floods, and the municipal authorities were not prompt enough in clean-
ing the city, which led to massive sanitation problems.
Despite being one of the richest civic bodies in the country, the Surat Municipal Cor-
poration had failed to provide basic infrastructure to a majority of the city's popula-
tion. Over 70 percent lacked sewers, and in the northern part of the city there were no
sewers at all. The incident led to a complete overhaul of the city administration, and a
massive cleaning operation involving government, community, and private organiza-
tions was undertaken. Two years later, the concerted efforts of the partners had suc-
cessfully transformed the city from one of the filthiest to one of the cleanest in India.
Regulation of International Watercourses

The 1966 Helsinki Rules on the Uses of the Waters of International Rivers
(SOURCE) have until recently been the primary legal basis for transboundary issues
related to international rivers. The Helsinki Rules encompass two principles: (1) the
prohibition of appreciable harm to any party through the deprivation of water rights,
pollution, or other means; and (2) the right of each basin state along an international
waterway to a reasonable and equitable utilization of the international waterway.
In 1974, the Helsinki Convention on the Protection of the Marine Environment of the
Baltic Sea was signed by all of the then seven Baltic Sea States. It has been in force
since 1980 and is the first international instrument to deal with all sources of pollution
to the Baltic Sea. The major aim is to minimize the load of pollution entering the Bal-
tic Sea from land-based sources.
The UN Economic Commission for Europe established the Convention on the Protec-
tion and Use of Transboundary Watercourses and International Lakes (SOURCE),
which came into force in Europe in 1996. The convention is based on the principles of
Agenda 21 and seeks to protect the quality and supply of fresh water throughout
Europe.
In 1997, the UN General Assembly adopted the Convention on the Law of the Non-
navigational Uses of International Watercourses (United Nations, 1997). This conven-
tion drew from the earlier Helsinki Rules and the Convention on the Protection and
Use of Transboundary Watercourses and International Lakes. In particular, it defines
the process by which watercourse states come to an agreement, the principle of equi-
table and reasonable utilization and participation, and the obligation not to cause sig-
nificant harm. Numerous regional agreements for specific river basins, lakes, or
inland seas, have developed from or moved under the auspices of the Convention of
the Non-Navigational Uses of International Watercourses. Among these are regional
agreements for the Danube, Indus, Niger, Rhine, and Zambezi rivers; the Baltic Sea;
and Lake Victoria.
In 1816, one of Europe's oldest treaties established the Rhine as a navigable waterway
(Weber, 2000). The necessity for transboundary management of the river Rhine was
recognized for shipping as early as the end of the eighteenth century. The transbound-
ary pollution aspect was given attention in 1932, when the Dutch government pro-
tested against emissions of residual salt into the French part of the river. After World
War II, the pollution of the river increased, and, in 1950, the International Rhine Com-
mission started to study wastewater and water quality problems. The international
cooperation to abate pollution was strengthened by the draft resolution International
Commission for the Protection of the Rhine against Pollution, released in Berne in
1963. This was followed by the European Council Decision to incorporate the Con-
vention for the Protection of the Rhine Against Chemical Pollution (77/586/EEC) into
the Berne resolution of 1963. In order to accelerate ecological improvements in the
Rhine, ministers from riparian countries (countries that abut a natural stream, river, or
lake) decided in 1986 to establish the Rhine Action Programme, with the
following aim:
The ecosystem of the Rhine must become a suitable habitat to allow the return to
this great European river of the higher species which were once present here and
have since disappeared (such as salmon). (ICPR, 1992)
The International Commission for the Protection of the Danube River was modeled
after the Rhine Commission. Its task is to implement the Danube River Protection
Convention (SOURCE), which was signed in 1994 and came into force in 1998. Cur-
rently, action is directed toward
• Improvement of the ecological and chemical status of the river;
• Prevention of accidental pollution events; and
• Minimization of the effects of floods.
One of the major areas of effort is in the reduction of pollution from municipal dis-
charges of both stormwater and sanitary wastewater.
Additional international agreements of importance to water resources management
include those on climate change, biological diversity, wetlands, and desertification.
Climate change is related to stormwater management as climate change is expected to
effect precipitation patterns and water levels. This issue was addressed at the Confer-
ence on Environment and Development in Rio de Janeiro (UNCED, 1992). The Con-
vention on Biological Diversity was also signed at UNCED, Rio de Janeiro in 1992.
The Ramsar Convention, which was signed in Ramsar, Iran in 1971, defines the con-
cept of wetland and calls for their protection (Carp, 1971). It is relevant to stormwater
management because poor stormwater management is a threat to biological diversity
in receiving water bodies. The United Nations Conference on Drylands was held in
Nairobi, Kenya in 1977. Stormwater management as it affects soil erosion was identi-
fied as an important process in desertification. In June 1994, the United Nations Con-
vention to Combat Desertification was adopted in Paris, France (United
Nations, 1992).
Water and Environment in the European Union

The EU aims to integrate environmental aspects into all development activities. This
integration became a legal obligation under the Treaty on European Union (Maas-
tricht Treaty).
When the European Community was created in 1956, the water management systems
in different member states could be classified broadly into three categories:
• Decentralized local action
• Coordinated regional action
• Uniform national standards systems
Due to the prevailing sectorial approach, most countries have numerous laws related
to stormwater management. For example, a country might have an environmental act,
a water act, a water utility act, a town and country planning act, and/or building acts
and regulations.
The rules and regulations at the national and regional levels tend to be legally binding,
while the local rules are more subject to political decision-making. In most countries,
the environmental agency or equivalent ministry provides guidelines on stormwater
management. In addition, local stormwater and wastewater utilities have their own
rules concerning pre-treatment of water that is discharged into their networks. In Ger-
many and Sweden, some water utilities have developed separate stormwater fees to
encourage developers to install on-site facilities for stormwater treatment.
Environmental management standards like the European Union's Eco-Management
and Audit Scheme (EMAS) (Hillary, 1994) and the International Organization for
Standardization's ISO 14000 have become fashionable among water utilities in sev-
eral European countries. Utilities that have adopted these standards are becoming
increasingly wary about what they allow into their networks. Information on EMAS,
ISO 14000, and other environmental management standards can be found on the
International Network for Environmental Management's web site (www.inem.org).
For a long time, stormwater management focused on safeguarding land or property
from damages resulting from excessive moisture and flooding. Since the mid-nine-
teenth century, the dominant idea has been that stormwater and other wastewater
should be collected in urban areas and disposed of outside of the urban environment
as quickly and efficiently as possible. However, stormwater-induced overflows from
combined sewer systems have since become an issue, as wastewater treatment plants
typically have a design capacity to treat only three to four times the dry-weather
wastewater flow. Therefore, the dominant European strategy has turned to developing
separate drainage systems for stormwater that discharge the stormwater directly to a
surface water body. However, both the quality and the quantity of unmitigated storm-
water discharges can have significant environmental impacts that must be considered.
Among the problems encountered in the development of separate drainage systems
for stormwater are the impracticality of installing them in old, crowded city centers
and ensuring the complete disassociation of all sanitary sewer connections from the
old conduit, which generally becomes the stormwater conduit in the separate system.
Only a few missed sanitary sewer connections could make stormwater runoff hygien-
ically unsafe.
Although stormwater detention ponds are popular in the United States and Canada,
stormwater infiltration/percolation and on-site detention/retention methods have gen-
erally been preferred in Europe. For instance, Switzerland's federal water pollution
control law now requires that stormwater be infiltrated whenever possible. A permit
must be obtained for cases in which stormwater is discharged into surface water.
In many German states, infiltration of stormwater runoff has become the state of the
art for sustainable urban storm drainage. In Northrhine-Westphalia, for example, legal
regulation demands that newly developed properties infiltrate rainwater into the
groundwater or discharge it to receiving waters. The technical and economic advan-
tages are the use of smaller diameter storm sewers and more efficient operation of
wastewater treatment plants because the inflow during wet-weather conditions is uni-
form. However, local geology and housing density have to be duly considered. Imper-
vious soils or high groundwater tables make infiltration inappropriate (Coldewey et
al., 2001). In densely populated areas, the capability for full disconnection from the
sanitary sewer system may be as low as 2 to 3 percent of all stormwater connections,

while the capability may be 50 percent at larger commercial sites and housing
associations.
More recently, a significant European interest in vegetated ponds and open convey-
ance systems for stormwater treatment has emerged. The aesthetic and biodiversity
advantages especially have promoted this interest. Vegetated roofs are also seen to
have beneficial impacts on stormwater runoff quantity and quality.
Each of these options has significant development potential, but the major challenge
is combining them with the existing infrastructure in a creative and productive way so
that development moves in a sustainable direction. As the practical understanding of
integrated urban drainage is limited, a learn-by-doing approach is recommended in
which small, innovative interventions are followed by evaluation and feedback.
The European Union Regulatory Approach. Until very recently, the EU
has applied an approach whereby standards are uniformly applied in a nation. This
approach has caused concerns about high costs, and questions have been raised about
the efficiency of the overall system in meeting water quality goals.
The uniform standards approach is, in fact, a departure from the river basin approach
that was widely used in national systems in Western Europe. Germany, France, Spain,
and the United Kingdom have river basin authorities of one kind or another, and those
systems are now changing to come into compliance with EU requirements. Neverthe-
less, they still have some flexibility within their own areas of authority. Such flexibil-
ity allows for the setting of priorities and realistic targets consistent with available
resources.
The 1976 Dangerous Substances Directive (76/464/EEC) was one of the first water-
related directives to be adopted, and it had the ambitious objective of regulating
potential aquatic pollution for thousands of chemicals. Among other directives of
importance to stormwater management are the Bathing Water Directive (76/160/
EEC), the Fish Water Directive (78/659/EEC), and the Shellfish Water Directive (79/
923/EEC).
In December 2000, the EU implemented new regulations for water management
through the EU Water Framework Directive (2000/60/EC). The major innovations
were a call for a river basin approach encompassing both point and nonpoint pollu-
tion, and replacement of the uniform standards approach with a combined approach.
Under this combined approach, pollution is controlled at the source through emission
limits, and environmental quality standards are applied to the chemical and physical
properties and the ecologic status of receiving waters.
The directive also requires holistic and integrated management of groundwater, sur-
face water, and coastal waters. River basins are lumped into districts, for which there
should be a “competent authority” to provide regulation. These authorities are
required to be operational by December 2003.
The EU will decide the minimum standards for good water quality. These standards
will be based on eco-regions and water categories: rivers, lakes, transitional waters,
coastal waters, artificial surface water bodies, or heavily modified surface water
bodies. The standards also state that, in addition to existing EU directives, other inter-
national conventions should be honored.
The competent authorities (defined by member states) are to classify the status of
their waters and to report this information to the EU. For waters not meeting the
requirements for “good” status, management plans for achieving “good” status within
15 years must be developed.
The directive calls for two kinds of monitoring: surveillance monitoring to assess
water status and operational monitoring to monitor the impacts of rehabilitation
measures.
The major objectives of the directive are to
• Prevent further deterioration of, protect, and enhance the status of aquatic
ecosystems and, with regard to their water needs, terrestrial ecosystems and
wetlands directly dependent on the aquatic ecosystems.
• Promote sustainable water-use based on long-term protection of available
resources.
• Achieve enhanced protection and improvement of the aquatic environment.
• Ensure a progressive reduction of groundwater pollution.
• Contribute to mitigating the effects of floods and droughts.
European Experiences. Stormwater management in Europe is affected by fac-
tors such as local climate, regulations, and institutional settings. In this section, the
stormwater management practices of selected countries are reviewed.
England and Wales. In 1974, the British Parliament decided to transfer the responsi-
bility for water and sewage services from local authorities to regional water authori-
ties. These regional authorities were responsible for the supply of water as well as
sanitary and stormwater drainage and treatment. The boundaries of these authorities
were set in accordance with the watershed areas. This approach formed a nationalized
industry, organizationally and constitutionally.
The regional water authorities were also responsible for environmental protection and
regulated water use in agriculture and industry. However, in doing so the regional
water authorities became both service providers and regulatory bodies.
In 1989 the regional water authorities were privatized. The regulatory authority
remained in public hands through a complicated departmental system. The principal
attributes of this system are as follows:
• The setting of policies and standards is the responsibility of the politically
elected government, and these policies and standards are executed through
the Department of Environment, Transport and Regions.
• Protection of the water environment is the responsibility of the Environment
Agency (formerly the National Rivers Authority), which took over the regu-
latory roles of the regional water authorities.
Economic regulation is carried out by the Director General of Water Services, who is
supported by the Office of Water Services (OFWAT).
On a local level, many municipalities have adopted Local Agenda 21s to promote sus-
tainable urban development. A Local Agenda 21 for the Thames River Basin, Thames
21, has been developed for catchment and land-use planning.
The first edition of the Urban Pollution Management Manual (UPM Manual) (Foun-
dation for Water Research, 1994) was published in November 1994 following a major
research and development program sponsored by the entire UK water industry. Many
of the planning concepts and enabling tools in the UPM Manual were substantially
new at that time and addressed issues that were, and continue to be, of great impor-
tance to the industry. The intervening period has seen widespread adoption and appli-
cation of the manual's procedures throughout the UK, particularly in areas with acute
combined sewer overflow problems.
The UPM Manual points out the need to consider the wastewater system, which con-
sists of the drainage system, the treatment plant, and the receiving water, as a single
entity. A change to one part of the wastewater system will have implications for other
parts of the system that must be taken into account in the planning process.
A common misconception that has developed is that an Urban Pollution Management
(UPM) study means costly, detailed simulation modeling. In reality, many UPM stud-
ies can be undertaken using relatively simple tools. The more complex and costly
tools are only employed when the economic case for using them has been proven.
Some three years after publication of the original UPM Manual, the sponsors agreed
that it was appropriate to update the original guidance in light of the experience
gained over the intervening period. Thus, the second edition of the UPM Manual
(UPM 2) was published (Foundation for Water Research, 1998).
UPM 2 identifies two approaches for the protection of river-based aquatic life from
wet-weather episodes, as described below.
• Fundamental Intermittent Standards (FIS) directly relate to the characteris-
tics of events that cause stress in river ecosystems. These standards are
expressed in terms of concentration-duration thresholds with an allowable
return period or frequency.
• High-percentile standards (such as the 99th percentile where water quality
standards are exceeded in one case in a hundred) are based on the extrapola-
tion of the 90/95 percentile thresholds from the Environment Agency's Eco-
system (River Ecosystem) classes used to plan and assess river quality.
Significant advances in UPM 2 include:
• The document is no longer focused on the specific circumstances pertaining
to England and Wales. The guidance presented is equally applicable else-
where. References to specific planning documents which have a finite life
span have been removed to avoid rapid dating of the document.
• A more comprehensive range of environmental standards appropriate to the
management of urban discharges is identified. More emphasis is placed on
the potential role of high-percentile standards, while the range of Fundamen-

tal Intermittent standards is expanded and the previous Derived Intermittent
Standards are abandoned.
The Construction Industry Research and Information Association (CIRIA) is a UK-
based research association concerned with improving performance in the area of con-
struction and the environment. CIRIA has published a best practices manual for sus-
tainable urban drainage, as well as sustainable urban drainage manuals for Scotland,
Northern Ireland, England, and Wales.
The Chartered Institution of Water and Environmental Management (CIWEM) is a
multidisciplinary professional and examining body for engineers, scientists, and other
qualified personnel engaged in water and environmental management. CIWEM has
produced a number of manuals of interest to stormwater management.
Scotland. In Scotland, local authorities own the water utilities. So far it has not been
possible to privatize the water industry. Instead, the water and sewer services were
transferred to three independent, government-established organizations in 1996. The
1996 Scottish reorganization corresponds to the 1974 reorganization in England and
Wales in that it created three multi-purpose water authorities to take over the duties of
local authorities and water boards.
Under the current system, the ultimate responsibility for the regulation of urban drain-
age rests with the Scottish Environment Protection Agency (SEPA). At the outset,
SEPA policy was designed to protect water quality from pollution by surface water
runoff through active regulation by SEPA. However, current policy is to use coopera-
tive approaches as much as possible.
SEPA promotes Sustainable Urban Drainage Systems (SUDS) that allow water to be
treated prior to release to Scottish waters. To promote and develop SUDS, SEPA and a
range of partner organizations formed the Sustainable Urban Drainage Scottish Work-
ing Party (SUDSWP) in 1997. This group has been instrumental in changing attitudes
toward sustainable urban drainage systems in Scotland.
The establishment of SUDSWP was based on the recognition that a partnership
approach is essential to resolving the problems of urban drainage, and that no single
organization can act alone to secure improvements in water quality, prevent flooding,
or improve local amenity value. An important achievement, partly funded by SUD-
SWP, was the publication of the Sustainable Urban Drainage Systems Design Manual
for Scotland and Northern Ireland (Martin, 2000).
Current SEPA policy does not necessitate a discharge permit for water from a SUD
system if the system has been designed in accordance with the Sustainable Urban
Drainage Systems Design Manual for Scotland and Northern Ireland. However, it is
recommended that conditional prohibition notices be served to the developer prior to
construction to ensure that the SUD system is constructed in accordance with the
agreed design.
France. In France, municipalities and their territorial services are under the control of
the Prefects who represent the State. With the development of government environ-
mental regulations and economic policies, the Prefects have ever-increasing responsi-
bilities regarding water policy issues. This system fosters serious problems if
individual Prefects give more weight to economic development than to environmental

issues.
The French rely extensively on separate sanitary and stormwater drainage systems.
However, these sewers have not always been carefully designed. Further, no specific
policy for the control of stormwater quality or quantity exists. The French government
has decided to fund some general Agenda 21 activities by levying a tax on all pollut-
ing activities. However, a ramification of this tax is that the Agences lose some of
their income—income that could be used to develop badly needed intervention capac-
ity in the area of diffuse pollution.
Germany. Germany is a federal republic with laws at both the federal and state levels.
In the past, Germany had developed fairly autonomous river basin regulatory organi-
zations. However, because of Germany's adoption of many of the top-down
approaches prescribed by the EU, much of the flexibility at the local level has been
lost.
An interesting feature of the German system is a law regarding “balanced develop-
ment” that was enacted in 1987. Balanced development can be interpreted as “no
impact” development. Any negative impact expected to endure longer than five years
must be mitigated. The consequence of this law with regard to stormwater regulation
is that stormwater from newly developed areas must be controlled in terms of both
quality and quantity so that no negative impact compared to the undeveloped situation
occurs.
The Netherlands. Because the Netherlands has significant areas lying below sea level,
drainage has been very important to the country. The Netherlands is at the forefront of
cooperative water management. The famous Waterschappen act as water boards and
are responsible for local flood protection and water quality and quantity management,
including wastewater treatment. The central government formulates the main lines for
strategic policies on water issues at the national level, and the provincial governments
define their strategic policies for non-state-controlled waters and the regional frame-
work for flood protection.
The Ministry of Housing, Spatial Planning, and the Environment is responsible for
general environmental policy; setting of water quality targets and emission standards;
and executing laws concerning drinking water, sewerage, and spatial planning.
The water boards and the municipalities are responsible for the operational manage-
ment and the actual implementation of policy issues. The water boards are also
responsible for local flood protection and water quantity and quality management,
including wastewater treatment. The water management task at the municipal level is
limited to the management of sanitary drainage systems, which is usually conducted
by local public works departments or joint agencies.
Canadian Stormwater Regulations

In Canada, development proponents are typically subject to legislation at the provin-
cial level. Federal-provincial/territorial agreements exist to permit cursory review and
approval of development applications with regard to federal legislation. This subsec-
tion presents a framework for Canadian stormwater-related regulations to demonstrate

how the different levels of government interact and uses Ontario as an example of
how stormwater management operates on a provincial level.
Canada Water Conservation Assistance Act. Canadian stormwater regu-
lations developed with a focus on flood control. The first federal legislation dealing
with water resources management was the Canada Water Conservation Assistance
Act, which was passed in 1953. Prior to this law, the provinces were charged with the
responsibility of managing their own natural resources, and the federal government
was responsible for fisheries, navigable waterways, and irrigation, among other
duties. Flood control was not dealt with specifically, and provinces encountered diffi-
culty when entering areas of federal control.
The Canada Water Conservation Assistance Act enabled the federal government to
assist the provinces by providing up to 37.5 percent of the cost of water conservation
or water control works that would normally exceed the financial means of the prov-
ince and municipalities. This assistance was provided through federal-provincial/ter-
ritorial cost-sharing arrangements.
Canada Water Act of 1970. Eventually, it was felt that the Canada Water
Conservation Assistance Act was too restrictive as it focused on works rather than the
non-structural solutions that later became an integral part of the water resources man-
agement strategy. As a result, this Act was superseded by the Canada Water Act of
1970, which reflected a change in the attitudes of both professionals and the public.
This Act provided for more public consultation, comprehensive planning that
included all water uses, and comprehensive analysis that considered larger geographi-
cal areas for a more holistic approach. Most important, the consideration of non-struc-
tural alternatives became a critical component in the development of the federal
regulations.
Flood Damage Reduction Program. The traditional model for flood man-
agement in place in the 1970s employed structural works for prevention and financial
disaster assistance in times of need. Payouts continued to escalate as assistance was
provided to new developments that were encouraged by the false sense of security
engendered by structural works (Environment Canada, 1993). In 1975, the federal
government established the Flood Damage Reduction Program. The program repre-
sented a shift from how to deal with problems after they occurred to how to prevent
flooding damage in the first place. The main premise of this legislation was to keep
floodplains free of development. Similar programs already in place in the provinces
of British Columbia, Ontario, and Alberta aided the national program by building
public support through the information already available. In 1999, the federal govern-
ment discontinued the Flood Damage Reduction Program. Currently, the mapping of
floodplains is conducted at the provincial and sometimes local level.
Other Federal Environmental Legislation. Other major federal environ-
mental legislation includes the Canadian Environmental Protection Act, the Fisheries
Act, and the Navigable Waters Protection Act. The Canadian Environmental Protec-
tion Act can be viewed as “umbrella” legislation that deals with a broad range of envi-
ronmental issues (Research Advisory Panel, 2001). The Fisheries Act is administered
by the Department of Fisheries and Oceans (DFO) and protects against “harmful
alteration, disruption, or destruction of fish habitat (R.S., c. F-14. s-1).” The Naviga-
ble Waters Act is administered by Transport Canada and is concerned with the protec-
tion of waterways for ship passage.
Stormwater Regulation in Practice: Ontario. In practice, most interac-
tion with government regulation is at the provincial level; each province or territory
has differing legislation. The regulatory framework and application of the policies
described in this subsection provide an introductory overview for Ontario. Develop-
ment proponents should fully investigate or retain professionals with specific knowl-
edge of the pertinent approval framework and procedures, as no two projects will
follow the same road to final approval.
Examples of provincial legislation in Ontario are as follows (Research Advisory
Panel, 2001).
• The Ontario Water Resources Act governs all aspects of water and wastewater
activities.
• The Environmental Protection Act provides guidelines for ensuring a thorough
review of proposed infrastructure, municipal servicing, and stormwater man-
agement facilities prior to granting Certificates of Approval.
• The Environmental Bill of Rights deals with the public’s access to information.
• The Environmental Assessment Act works in conjunction with the Bill of Rights.
Under the Conservation Authorities Act (1946), regional authorities were given the
task of establishing and undertaking the management of natural resources for lands
deemed to be under their jurisdiction. The Conservation Authorities (CAs) were given
the responsibility of carrying out flood control measures and enforcing the floodplain-
free development principles described previously. During the review process for a
development proposal, municipalities work in conjunction with the CAs. Local-level
governments such as municipalities and regional municipalities issue design stan-
dards and criteria that deal primarily with the detailed design of stormwater manage-
ment facilities and conveyance systems.
CAs and provincial government agencies may, in many cases, be empowered to
review and comment on behalf of federal authorities. As an example, the Ontario
Ministry of Natural Resources administers and enforces some of the sections of the
Fisheries Act (federal legislation) regarding habitat on behalf of the Department of
Fisheries and Oceans, where the CA does preliminary review of proposed stormwater
management schemes. For scenarios in which stormwater is discharged directly into a
watercourse, the CA will recommend the level of review that must be satisfied
depending on the assessed level of disturbance to fish habitat (including both physical
disturbances and thermal impacts on receiving waters). A “fill construction” permit
may be requested if material is proposed for addition or removal in fill-regulated areas
on the subject lands.
Permits are required for construction of sewers, water mains, and appurtenances
under the Ontario Water Resources Act, which is administered by the Ontario Minis-
try of the Environment (MOE). Permits for operation and maintenance are not typi-
cally required, as municipalities will assume responsibility for infrastructure
management. However, the current trend in Ontario is toward private operators for
water and wastewater infrastructure management.
Municipalities will review and provide comments on development proposals on sewer
allocation, capacity, and availability through their sewer use bylaws. Quantity con-
trols may be imposed at this stage or through the CA policies. As discussed earlier,
the notion of stormwater management in Canada was founded on flood control; thus,
CA policies may reflect this by imposing quantity control restrictions for stormwater
discharges. The regulating authority for the development will be influenced by loca-
tions of the site and storm sewer outfall.
The municipality may have quality control objectives implemented as part of an Offi-
cial Plan or sewer use by-laws that are more stringent than CA quality control goals.
Quality control levels are governed by the more stringent requirements of the CA or
municipality. Quality criteria may be met through on-site controls such as infiltration
trenches, grassed swales, oil-grit separators, or some combination of these or other
best management practices. (For information on water quality best management prac-
tices, see Chapter 15.) Permits are not specifically required for stormwater discharge
pollutant control, though the MOE has outlined provincial water quality objectives
that are targets for the protection of aquatic life.
Projects located in the Great Lakes drainage basin may be subject to additional regu-
lation. The International Joint Commission has designated 42 Areas of Concern
(AOC), 16 of which are in Canada. Remedial Action Plans (RAPs) have been devel-
oped to restore beneficial uses to the AOCs. Implementation of the RAPs may require
preparation of Pollution Prevention and Control Plans for control of contamination
from urban runoff and sewage.
14.2 REGULATORY REQUIREMENTS IN THE UNITED

STATES
Historically, stormwater management programs in almost all U.S. communities
addressed drainage issues only. These reluctant communities were only able and will-
ing to participate in stormwater management at the most basic and minimal level,
with the frequent argument that scarce local resources were needed for more pressing
issues.
However, with the development of new and increasingly complex laws and regula-
tions concerning water quality, stormwater quality management is no longer a luxury
seen only in rapidly growing affluent communities that have the resources and
motives. Cities and municipalities around the country are now being mandated to
assume roles as water resources managers and enforcers.
A host of laws, regulations, and ordinances exist related to the planning, design, and
construction of stormwater management systems. They stem from all branches of
government, from federal to local levels. Further, they are in a continual state of
change and place a corresponding obligation on designers to keep abreast of their
developments. Summaries of some of the more important U.S. laws and regulations
are presented in the following subsections. The focus in this text is federal laws and
Section 14.2 Regulatory Requirements in the United States 593
regulations. Less emphasis is given to state and local laws and regulations because of
the many differences that exist from one political jurisdiction to another. Neverthe-
less, state and local laws and regulations must be adhered to in any particular project,
and the onus falls on the designer to ensure that their requirements are met.
Regulations pertaining to stormwater generally make a distinction between point and
nonpoint sources of pollution. Point sources of water pollution are defined in Section
502 of the 1972 Amendments to the Water Pollution Control Act as “any discernible,
confined, and discrete conveyance, including but not limited to any pipe, ditch, chan-
nel, tunnel, well, discrete fissure, container, rolling stock, concentrated animal feed-
ing operation, or vessel or other floating craft, from which pollutants are or may be
discharged.” Nonpoint sources are the remaining pollutant sources not included in this
definition. This definition of point source appears to include almost all water dis-
charges, but important court actions have been necessary to clarify it further. The
chief contributors to nonpoint source pollution are runoff from agricultural, mining,
and forestry operations, and runoff from urban areas.
Federal Laws and Regulations

Federal laws and regulations concerning stormwater drainage systems, pollutants, and
receiving waters are extensive. The following subsections present the development of
these regulations, and include discussions of recent and emerging regulations.
Table 14.2 summarizes the major federal legislation addressing clean water.
Table 14.2 Clean Water Act and major amendments (Copeland, 1999)
Year Act Public Law
1948 Federal Water Pollution Control Act P.L. 80-845
1956 Water Pollution Control Act of 1956 P.L. 84-660
1961 Federal Water Pollution Control Act Amendment P.L. 87-88
1965 Water Quality Act of 1965 P.L. 89-234
1966 Clean Water Restoration Act P.L. 89-753
1970 Water Quality Improvement Act of 1970 P.L. 91-224, Part I
1972 Federal Water Pollution Control Car Amendments P.L. 92-500
1977 Clean Water Act of 1977 P.L. 95-217
1981 Municipal Wastewater Treatment Construction Grants Amendment P.L. 97-117
1987 Water Quality Act of 1987 P.L. 100-4
The Water Pollution Control Act of 1948. Before 1948, almost all water
pollution control authority was vested in state and local governments. The legal pow-
ers of different state agencies varied greatly. The Public Health Service Act of 1912
authorized the investigation of pollution in navigable waters, and the Oil Pollution
Act of 1924 was enacted to prevent oil discharges into navigable waters. The Water
Pollution Control Act (WPCA, PL 80-845), passed in 1948, established some federal
authority in abating interstate water pollution.
Because the 1948 Act focused on receiving-water standards instead of effluent stan-
dards, and because it was enforced ineffectively, it was not an immediate success. The
receiving-water standards only considered the current uses of the water and were dif-
ficult to enforce against any single polluter. It usually was not possible to determine
who was responsible when a stream standard was exceeded. The early regulations
also placed much of the burden of proof on enforcement agencies.
Most of the water pollution control legislation enacted since 1948 has been in the
form of amendments to the 1948 Act. Important amendments include:
1956: Made the legislation permanent and funded construction grants for pub-
licly owned treatment works (POTWs)
1961: Increased funding for water quality research and construction grants
1965: Increased construction grants and started research concerning combined
sewer overflows
1966: Removed the dollar limit on construction grants
1972: Enacted the most important advances to date, as described more fully
below
1977: Extended some of the deadlines established in the 1972 amendments
1988: Required discharge permits for stormwater
The WPCA Amendments of 1972. The Refuse Act of 1899 (33 USC 407)
was used in 1970 to establish a discharge permit system. The Refuse Act prohibited
the discharge of any material, except sewage and runoff, into navigable waterways
without a permit from the Department of the Army. It was written during the “pro-
gressive conservation era,” when multiple uses of natural resources first became seri-
ously considered. The U.S. Supreme Court upheld the applicability of the law for
pollution control in 1966. However, a later decision in 1971 invalidated the program.
Because of these difficulties, the WPCA amendments of 1972 (PL 92-500) contained
a permit program called the National Pollutant Discharge Elimination System
(NPDES). The NPDES is an enforcement scheme through which effluents from point
sources are controlled. The permits are required for all point sources and establish
discharge limits based on the available control technology. The discharge limits set
for each industry were based on a series of studies that characterized each industry's
waste effluents and existing controls. Future discharge limits were reduced as all
facilities were to obtain the “best available technology economically achievable”
(BATEA) by July 1983. New sources were to obtain discharges representative of a
“standard of performance” that was much more restrictive. Nonpoint sources were
originally exempt from the NPDES permit program, but stormwater discharges are
included in the NPDES system as part of CFR 122, 123, 124, and 504, as published in
the Federal Register of December 7, 1988.
The NPDES was intended to achieve Congress’s goal of no pollutant discharges what-
soever by 1985. Other goals of PL 92-500 included the protection and propagation of
fish, shellfish, and wildlife and recreational uses of water by July 1983; the prohibi-
tion of the discharge of toxic pollutants; the continuation of funding for publicly-
owned treatment works (POTWs); the development of area-wide wastewater treat-
ment management plans; the funding of a major resource and demonstration effort to
improve treatment technology; and the protection of the rights of the states to reduce
pollution and to plan their water resources uses.
The Clean Water Act of 1977. The Clean Water Act of 1977 (PL 95-217)
extended some of the deadlines, but no waivers were allowed for toxic pollutant dis-
charges. The amendments also allowed for ad valorem taxes by municipalities to fund
treatment projects, provided incentives for use of innovative technologies, and placed
increased emphasis on the area-wide treatment planning studies (Section 208).
Section 208 planning studies recognized the need to control nonpoint pollution in
order to meet Congressional goals. This section was an incentive to local governments
to develop their own plans, with minimal federal input. These plans were to character-
ize all point and nonpoint pollutant discharges in designated areas and to develop
treatment schemes that would allow the goals to be met. The results of these plans
affect the issuance of NPDES permits. Unfortunately, most of these plans were con-
ducted in short time periods with limited technical success. Control measures were
recommended with few local demonstrations of their potential success (especially for
nonpoint pollution control). Recognizing these technical shortcomings, Congress
authorized the Nationwide Urban Runoff Program (NURP) to demonstrate the appli-
cability of various urban runoff control measures in about 30 cities. These studies
were completed in 1983, and with these results, the Section 208 plans were to be
revised under Section 205g.
EPA Stormwater Control Regulations. In the United States, regulations
promulgated under the Clean Water Act are administered by the Environmental Pro-
tection Agency (EPA). Many of the laws allow for the EPA to transfer permitting and
regulatory authority to the states. The NPDES permit program is an example of this.
In general, states must adopt regulations that are at least as stringent as the federal
regulations before they are granted regulatory authority.
The EPA regulations to control stormwater runoff were first published in the Decem-
ber 7, 1988 issue of the Federal Register. These regulations initiated a permit process
for urban runoff, but the schedules and required reporting information varied depend-
ing on land use and the size of the community. The EPA was required by Section 405
of the Water Quality Act of 1987 to establish permit application requirements for
municipalities having populations greater than 250,000 and industrial concerns
(including construction operations) by February 4, 1989. Permit application require-
ments for municipalities having populations between 100,000 and 250,000 were to be
established by February 4, 1991. The first applications for industrial concerns and
large cities were to be submitted by February 4, 1990. The applications for smaller
cities were to be filed by February 4, 1992. Permits are now required for cities with a
population of less than 100,000 as well, as part of the second phase of the NPDES
stormwater permit program.
Phase I Regulations. The Phase I general application requirements stressed descrip-
tive information concerning the drainage area, with minimal runoff monitoring
requirements. The permit applications relied mainly on the use of simple models to
predict annual discharges of pollutants, and on field analyses for detecting illicit con-
nections and illegal dumping. Permit applications also required a description of any
locally required stormwater and construction-site runoff controls. Local municipali-
ties were also to establish an authority for managing stormwater.
Phase II Regulations. The Clean Water Act 402(p)(6) initial Phase II rule (for small
municipalities) was published on August 7, 1995. Its purpose was to designate addi-
tional sources of stormwater that needed to be regulated to protect water quality. It
requires operators of small municipal separate storm sewer systems (MS4s) to apply
for NPDES permits starting in March 2003. Phase II regulations also apply to entities
such as universities, departments of transportation, and military bases that operate
MS4s. It also affects millions of industrial/commercial facilities and almost all con-
struction activities. Little, if any, monitoring is required by Phase II communities, but
they are still responsible for ensuring that water quality regulations are met.
The Stormwater Phase II Rule superseded the August 1995 regulation. The final rule
was published in the Federal Register on December 8, 1999. The schedule for imple-
mentation of the Phase II regulations calls for general permits to be issued in 2003
with an in-compliance schedule that extends up to five years. Some commercial and
industrial groups have also been working with the EPA for special consideration and
scheduling of implementation of regulations for their group members.
The Phase II Rule defines a stormwater management program for a small MS4 as a
program composed of six elements that, when implemented together, are expected to
reduce pollutants discharged into receiving water bodies to the maximum extent pos-
sible. These six program elements, or minimum control measures, are
• Public education and outreach on stormwater impacts
• Public involvement/participation
• Illicit discharge detection and elimination
• Construction site runoff control
• Post-construction stormwater management in new development and
redevelopment
• Pollution prevention/good housekeeping for municipal operations
For each minimum control measure, permittees will select and implement best man-
agement practices and measurable goals that comprehensively address the specific
stormwater problems in their respective areas.
Significant Court Cases. Several court cases have considered the question of whether
urban runoff is a point source (and therefore required to have a NPDES permit under
the original program) or a nonpoint source (and therefore exempt from the permit pro-
cess, but still subject to applicable discharge standards). Urban runoff usually enters
receiving waters through a conduit or ditch. The Natural Resources Defense Council
(NRDC) v. Costle et al. (568 F.2D 1369, 1977) case found that uniform NPDES dis-
charge limitations were not a necessary precondition for the inclusion of agricultural,
silvicultural, and stormwater runoff point sources into the NPDES program. However,
the regulations (40 CFR Section 125.4, 1975) did specifically exempt separate storm
sewers containing only storm runoff uncontaminated by any industrial or commercial
activity.
The NRDC vs. Costle et al. case validated the NPDES program by setting a clear pre-
cedent for considering any pollutant discharges as unlawful. The legal premise of the
inherent right to utilize the nation’s waterways for the purpose of disposing of wastes
was eliminated; however, due in part to technological limitations pollutant discharges
are continuing. Also, the stormwater runoff exemption of 40 CFR Section 125.4 was
maintained because it was felt that requiring permits on the approximately 100,000
urban runoff point sources would tie up the ability of the EPA to effectively adminis-
ter the more important industrial and municipal point sources. State authorities were
encouraged to implement a plan to use general permits to cover these stormwater dis-
charges. The District of Columbia Court of Appeals concluded that Congress’s intent
was to require permits for all point sources, but that the EPA was to have flexibility in
structuring the permits in the form of general or area permits.
In Pedersen et al. v. Washington State Department of Transportation (611 P.2D 1293,
1980), the court found that “Separate storm sewers, as defined in this section, are
point sources subject to the NPDES permit program. Separate storm sewers may be
covered either under individual NPDES permits or under the general permit.” The
State of Washington Court of Appeals found that states are not required to implement
general permit programs, and that individual permits can be required if appropriate
general permits are not available.
The decision in the United States of America v. Frezzo (642 F.2D 59, 1981) case
stated that the intent of the NPDES regulations “is to exclude from the NPDES permit
program all natural runoff from agricultural land which results from precipitation
events.” This case further stated, “When water pollution from irrigation ditches results
from precipitation events, that pollution is nonpoint in nature. However, when dis-
charges from irrigation ditches result from the controlled application of water by any
person, that pollution is considered point source and subject to the program proposed
herein.”
These cases established that collected urban runoff is a point source, while agricul-
tural ditches are nonpoint sources during rain events and point sources during irriga-
tion events.
Total Maximum Daily Load (TMDL) Program. Statewide Watershed

Assessment Reports mandated by the Clean Water Act and other legislative and regu-
latory programs identify water quality deficiencies and prioritize watersheds for
assessment. The 305(b) report, or State Water Quality Assessment, identifies and pri-
oritizes water bodies with known violations of water quality standards or criteria. The
305(b) reports are prepared by the states on a biennial basis.
When states identify impaired water bodies through 305(b) assessments, they are
required to identify the waters and to develop total maximum daily loads (TMDLs)
for them. In general, a TMDL is a quantitative assessment of water quality problems,
contributing sources, and pollution reductions needed to attain water quality stan-
dards. The TMDL specifies the amount that pollution and/or other stressors need to be
reduced by in order to meet water quality standards It further allocates pollution con-
trol or management responsibilities among sources in a watershed and provides a
science-based policy for taking action to restore a water body.
Pollutant loading from stormwater is considered in developing the TMDL. Stormwa-
ter discharges that are regulated under the Phase I or Phase II of the NPDES stormwa-
ter program are point sources and are included in the Waste Load Allocation portion
of the TMDL. Stormwater discharges that do not require an NPDES permit are con-
sidered nonpoint sources and are included in the load allocation of the TMDL. When
the TMDL is established, it may result in requirements for monitoring and/or treat-
ment. For non-permitted discharges, treatment may involve implementation of one or
more Best Management Practices (see Chapter 15).
By mid-1998, all states had EPA-approved Section 303(d)(1) lists, but the content and
scope of these lists varied greatly. Development of TMDLs has been initiated at an
increasing pace in some states, but most TMDLs are still incomplete. Many of the
waters still needing TMDLs are impaired by contributions from both point and non-
point sources. The EPA has undertaken a variety of steps to strengthen and define the
TMDL program. The EPA should be consulted for updates on the list of impaired
waters, the status of TMDLs, and policy guidance.
Dredge and Fill Permit Program. Section 404 of the Clean Water Act
requires a permit, issued by the U.S. Army Corps of Engineers, for discharge of fill or
dredge materials into the waters of the United States. This jurisdiction includes not
only navigable waters, but also their tributaries and associated wetlands. The Supreme
Court has held that this jurisdiction extends to lands supporting plant growth typical
of wetlands, such as willow trees.
The scope of activities covered by Section 404 is much more broad than the tradi-
tional dredging and filling of navigable ship channels. Regulated activities include,
for example, construction of bridges, dams, buildings, roads, flood control facilities,
and shellfish operations.
Section 404 brings a wide range of federal environmental law to bear on any project
requiring a permit. In its issuance of a Section 404 permit, the Corps must comply
with the Fish and Wildlife Coordination Act, Endangered Species Act, Wild and Sce-
nic Rivers Act, Coastal Zone Management Act, National Environmental Policy Act,
and other laws that govern federal activities.
State Laws and Regulations

In the United States, power to regulate the use and control of waters is vested largely
in individual states, although there are federal requirements relating to factors such as
navigability and reserved water rights. State constitutions or statutes often contain
provisions relating to water laws (statutory laws), but most applied water laws are
based on precedent. As a result, laws can vary significantly from one state to another,
and even within a state from incorporated municipal areas to adjacent unincorporated
areas.
Despite the fixed chemical composition of pure water and the interconnections among
the various stores in which water exists, the legal system tends to compartmentalize
water into various types, with each type being subject to somewhat different sets of
water laws. For example, groundwater is generally subject to a different set of laws
than surface water in a stream. Diffused water, which results from runoff that has not
reached a well-defined channel with a bed and banks, is subject to yet another set of
laws. A summary of the main legal doctrines and rules for surface streams and dif-
fused waters is provided in the following subsections. Groundwater laws are not dis-
cussed here, as they are not as important in the context of urban stormwater
management. The interested reader is referred to Getches (1997) for a thorough intro-
duction to water law.
Stormwater Regulations in the

State of Virginia
Stormwater in Virginia is driven by federal and state laws. such as dam safety and erosion and sediment control. The
National permits, like the EPA's NPDES Phase I and II VA Department of Health plays a role in shellfish sanita-
permits, are implemented through programs under the Vir- tion issues and protecting drinking water supplies. The
ginia Department of Environmental Quality (DEQ). One VA Department of Game and Inland Fisheries deals with
characteristic of Virginia's stormwater laws is that Virginia non-tidal fisheries and boating issues. The VA Marine
is a “Dillon Rule” state. This means that a locality needs to Resources Commission (VMRC) is the state agency that
have enabling state legislation to develop ordinances for manages the state portion of the U.S. Army Corps of
different topics of concern. If localities want to develop Engineers’ Joint Permit Application for any work includ-
more stringent regulations than the current state mini- ing construction, dredging, filling, or excavation in the
mums, the Dillon Rule allows them to develop those ordi- water or in wetlands. The VA Department of Emergency
nances. In Virginia, the state has enabling legislation for Management deals with natural disasters and administers
erosion and sediment control ordinances, stormwater man- FEMA grant programs for areas such as flood mitigation.
agement ordinances, and stormwater utility ordinances. Finally, the VA Department of Transportation (VDOT)
This does not force a community to enact these ordinances, plays an important role in establishing road drainage
but simply provides a framework for developing them. guidelines. Roads must meet VDOT design guidelines,
Most Virginia local government have enacted erosion and including stormwater drainage, in order for the road to be
sediment control ordinances, but not stormwater ordi- maintained by VDOT or for communities to receive state
nances. money to maintain their VDOT-approved roads. There-
There are many different Virginia state agencies that play fore, the VDOT design guidelines are often used for most
stormwater design standards.
a role in stormwater management. The DEQ plays the pri-
mary environmental regulatory role. The Chesapeake Bay Most of these state agencies have regional offices through-
Local Assistance Department addresses specific water out the state that cover multiple counties and cities. These
quality issues and regulations in the portion of Virginia agencies’ regions are not the same, so often designers must
that drains to the Chesapeake Bay. The VA Department of get approval from VDOT in one location, DEQ in another
Conservation and Recreation plays an important role by location, DCR in a third location, and VMRC at the state
managing state lands and providing guidance in areas capital.
Surface Water Doctrines. There are two main doctrines of water law. These
doctrines are known as the riparian and prior appropriation doctrines. Some states
have adopted hybrid systems consisting of elements from each of these doctrines,
whereas others, such as Louisiana and Hawaii, have almost completely different sys-
tems. Most states east of the 100th meridian have adopted the riparian doctrine,
whereas most of the western states adhere to the doctrine of prior appropriation. (For
further discussion, see Novotny and Olem, 1994.)
Riparian Doctrine. Landowners whose property abuts a natural stream, river, or lake
hold riparian rights. By virtue of their physical location, they are entitled to rights not
held by nonriparians. When a property is sold or conveyed, the associated riparian
rights are transferred to the new owner. Riparian rights entitle a landowner to the full
natural flow of a stream, both in quantity and quality. However, riparian rights also
require a landowner to use the water reasonably level and without infringing upon the
rights of downstream property owners. Riparian rights can be sold as property, but
water cannot be transferred from one drainage basin to another. During water short-
ages, property owners have equal rights to reasonable use of the water, and the avail-
able supply must be shared.
Doctrine of Prior Appropriation. During settlement of the American West, it became
apparent that the riparian doctrine was not practical in the relatively water-poor west.
This gave rise to a first-in-time, first-in-right approach to water resources. Under this
doctrine, seniors (those making early appropriations) have priority of water use over
juniors (those with later appropriations) during times of shortage; no requirement
exists for water sharing. An appropriation is made by giving notice of intent to divert
water, diverting water from its natural course, and applying it to a beneficial use. Non-
use of a water right can result in loss of the right. Often, water rights are term limited
and must be reappropriated. Water rights can be sold as property, and water can be
transferred from one drainage basin to another, but restrictions are placed on the
amount of water that can be transferred out of a basin so as to protect other
appropriators.
Diffused Water Rules Diffused water is water from precipitation or snowmelt
that runs across the land surface without a well-defined course having a channel bed
and banks. Two major rules have been applied by the courts to resolve disputes con-
cerning diffused waters: the civil law rule and the common enemy rule. Both of these
rules may be viewed as extreme stances, and thus there has been some modification to
bring them closer together through a reasonable use rule.
Civil Law Rule. This rule is based on the perpetuation of natural drainage. It places a
servitude upon a lower elevation landowner for the drainage of surface water in its
natural course and flow from a higher elevation landowner. The higher landowner is
not entitled to discharge runoff at a location other than that at which flow occurred
historically, nor is the higher landowner entitled to increase the amount of the flow
discharged to the lower proprietor.
Common Enemy Rule. Under this rule, diffused waters are viewed as the common
enemy of all, and each landowner may take whatever actions are deemed necessary to
protect their own interests. Owners whose properties may be damaged have no
recourse against their neighbors.
Reasonable Use Rule. Under this rule, each landowner can make reasonable use of his
land, even though the flow of diffused water may be altered, thereby causing harm to
others. Liability occurs when the interference or alteration of the flow is unreason-
able, as judged by a balancing test. Generally, questions concern issues such as
whether there was a reasonable necessity to alter the flow to make use of one's land,
whether the alteration itself was accomplished in a reasonable manner, and whether
the utility of the actor’s conduct reasonably outweighs the gravity of harm to others.
Local Regulations
Federal legislation mandates stormwater quality management at most levels of gov-
ernment. These mandates have arisen mainly from the NPDES program and from the
Phase I and Phase II regulations that have been promulgated by the EPA. Thus, many
of the local laws and regulations concern issues such as erosion and sedimentation,
stormwater quality controls, and various codes, regulations, and ordinances relating to
detention facilities, land-use zoning, and subdivision development requirements.
Codes, ordinances, and subdivision regulations are usually adopted and published by
municipal and/or county governments with the purpose of establishing minimum cri-
teria that must be met by developers of unimproved land parcels. These criteria estab-
lish a degree of uniformity from one development to another, especially in terms of
the transportation and utility infrastructures. Regulations are often set forth with
regard to stormwater management systems, including design-storm frequencies and
minimum system requirements.
Zoning regulations are also commonly adopted and published by municipalities and
county governments. Zoning maps and master plans are often prepared to guide future
development patterns. This planning has advantages not only with respect to ensuring
that objectionable development patterns do not occur (for example, a factory next to a
prime residential area), but also for the planning of transportation and utility systems.
In the context of urban stormwater management, zoning maps may be used to project
future land uses in undeveloped areas upstream of a project locale, and hence can help
to ensure that the project facilities will meet future needs.
Many states and cities throughout the United States have regulations that require
developers to construct runoff control facilities. Many of these regulations were ini-
tially used only to regulate runoff flow rates or to reduce erosion due to construction.
Some state court cases have led to the requirement for runoff controls as part of the
environmental impact statement process (such as Maryland v. U.S. Postal Service,
487 F. 2D 1029, 1973, and Veterans Administration Hospital case in Tennessee 48 FR
11551, 1983). In the opinion for Parsippany v. Costle (503 F. Supp. 314, 1979), it was
recommended that municipalities within the drainage basin adopt regulations that
stipulate the use of detention and sedimentation basins.

Stormwater conveyance systems must be designed to meet minimum requirements
specified by local, regional, state, or national review agencies.
In 1977, the UN Mar del Plata Conference placed water on the international agenda.
In 1987, the World Commission on Environment and Development agreed on the
need to reverse current development trends and popularized the notion of sustainable
development. This new awareness resulted in the United Nations Conference on Envi-
ronment and Development in Rio de Janeiro in 1992, whose major outcome was to
create a framework for the work toward sustainable development during the twenty-
first century. A broad consensus exists that sustainable development requires the sus-
tainable use and management of water resources in an environmental, economic, and
social context.
In developing countries, insufficient urban drainage systems are a major public health
concern. Existing channels are often misused as trash dumps during dry weather con-
ditions. When rain occurs, there is a considerable risk of clogging the channels, which
results in flooding, spreading of garbage, and pollution of water bodies. A lack of ade-
quate urban land-use planning and enforcement in most developing countries means
that technical tools and principles developed for industrialized countries must be
adapted to the developing country's present conditions.
The Helsinki Convention on the Protection of the Marine Environment of the Baltic
Sea was signed by all of the seven Baltic Sea States in 1974 and has been in force
since 1980. It is the first international instrument to deal with all of the different
sources of pollution in the Baltic Sea. A major aim is to minimize the load of pollu-
tion entering the Baltic Sea from land-based sources.
Until recently, the European Union (EU) has applied a uniform standards approach.
This approach has caused concerns about high costs, and questions have been raised
about the efficiency of the overall system in meeting water quality goals. Germany,
France, Spain, and the United Kingdom have river basin authorities of one kind or
another, and those systems are now changing to come into compliance with EU
requirements. Nevertheless, they still have some flexibility within their own areas of
authority. Such flexibility allows for the setting of priorities and realistic targets con-
sistent with available resources.
In Canada, development proponents are typically subject to legislation at the provin-
cial level. Federal-provincial/territorial agreements exist to permit cursory review and
approval of development applications with regard to federal legislation. Canadian
stormwater regulations developed, as they did elsewhere, with a focus on flood con-
trol.
Historically, stormwater management programs in almost all U.S. communities
addressed drainage issues only. Stormwater quality management was brought to the
attention of most medium- to large-size cities with the passage of the Clean Water Act
and its various amendments, and the associated NPDES permit program.
REFERENCES
Carp, E. ed. 1972. Proceedings, International Conference on the Conservation of Wetlands and Waterfowl,
Ramsar, Iran, 30 January - 3 February 1971. Slimbridge, UK: International Wildfowl Research Bureau.
Coldewey, W.G., P. Göbel, W.F. Geiger, C. Dierkes, and H. Kories. 2001. “Effects of Stormwater Infiltra-
tion on the Water Balance of a City.” in New Approaches Characterizing Groundwater Flow, Proceed-
ings of the XXXI. International Association of Hydrogeologists Congress, München, 10–14 September
2001. ed. K. P. Seiler and S. Wohnlich. Balkema, Lisse.
Copeland, C. 1999. Clean Water Act: A Summary of the Law, CRS Issue Brief for Congress, RL 30030.
Environment Australia. 2002. Urban Stormwater Initiative. Canberra: Environment Australia.
Environment Canada. 1993. Flooding: Canada’s Water Book. Ottawa, Canada: Ecosystem Sciences and
Evaluation Directorate, Economics and Conservation Branch.
European Commission. 1998. “Towards Sustainable Water Resources Management – A Strategic
Approach.” DG Development and DG External Relations.
European Conference of Sustainable Cites and Towns (ECSCT). 1994. Charter of European Cites and
Towns Towards Sustainability (Aalborg Charter). Aalbord, Denmark: ECSCT.
Foundation for Water Research. 1994. Urban Pollution Management Manual. Marlow, Great Britain: Foun-
dation for Water Research.
Foundation for Water Research. 1998. Urban Pollution Management Manual. 2nd ed. Marlow, Great Brit-
ain: Foundation for Water Research.
Getches, D. H. 1997. Water Law. 3d ed. St. Paul, Minnesota: West Publishing Co.
Hillary, R. 1994. The Eco-Management and Audit Scheme: A Practical Guide. Trans-Atlantic Designs.
International Commission for the Protection of the Rhine (ICPR). 1992. Ecological master plan for the
Rhine. Coblenz: International Commission for the Protection of the Rhine against Pollution.
International Environmental Technology Center (IETC). 2002. Environmentally Sound Technology for
Wastewater and Stormwater Management: An International Source Book. London: IWA Publishing.
International Rivers Network (IRN). 2003. IRN Press Release. 27 Feb. 2003.
Martin, P. et al. 2000. Sustainable Urban Drainage Systems – Design Manual for Scotland and Northern
Ireland. CIRIA Publications,.
Novatny, V. and H. Olem. 1984. Water Quality, Prevention, Identification and Management of Diffuse Pol-
lution. New York: Van Nostrand Reinhold.
Research Advisory Panel: The Walkerton Inquiry. 2001. “Wastewater Collection and Treatment – Issue
Papar 9.” Research Paper Advisory Panel, The Walkerton Inquire. Toronto, ON: Delcan Corporation.
United Nations. 1977. Report of the United Nations Conference. Mal del Plata. 14–25 March 1977. New
York: United Nations.
United Nations. 1992. Intergovernmental Negotiating Committee For The Elaboration Of An International
Convention To Combat Desertification In Those Countries Experiencing Serious Drought And/Or
Desertification, Particularly In Africa.
United Nations. 1992. Report of the United Nations Conference on Environment and Development
(UNCED). Rio de Janeiro, June 1992. New York: United Nations.
United Nations, 1996, Report of the United Nations Conference on Human Settlements (Habitat II), Istan-
bul, 3–14 June 1996. New York: United Nations.
United Nations. 1997. Report of the Commission on Sustainable Development on Preparations for the Spe-
cial Session of the General Assembly for the Purpose of an Overall Review and Appraisal of the Imple-
mentation of Agenda 21. New York: United Nations.
References 605
United Nations Development Program (UNDP). 1998. Capacity Building for Sustainable Management of
Water Resources and the Aquatic Environment. New York: United Nations.
U.S. Environmental Protection Agency. 1991. Guidance for Water Quality-Base Decisions: The TMDL
Process. EPA 440/4-91-001.
Weber, U. 2000. The 'Miracle' of the Rhine Courier. Unesco, Paris.
World Bank. 1993. Water Resources Management. Washington, D.C.: International Bank for Reconstruc-
tion and Development/The World Bank.
World Bank. 1998. Pollution Prevention and Abatement Handbook 1998: Toward Cleaner Production.
Washington, D.C.: International Bank for Reconstruction and Development/The World Bank.
World Bank. 2002. Thailand Environmental Monitor 2001: Water Quality. Washington, D.C.: The World
Bank Group.
World Bank. 2003. The New Water Resources Strategy. Washington, D.C.: World Bank Group.
World Commission on Environment and Development (WCED). 1987. Our Common Future. Oxford,
Great Britain: Oxford University Press.
World Resources Institute. 1997. The World Resources 1996–97. Oxford University Press.
Wright, A.M. 1997. Toward a Strategic Sanitation Approach: Improving the Sustainability of Urban Sani-
tation in Developing Countries. Washington, D.C.: International Bank for Reconstruction and Develop-
ment/The World Bank.
Measures should be
taken to protect
drainageways in
and around
construction sites
from sediment-
laden runoff due to
erosion of disturbed
soils.
C H A P T E R
15
Stormwater Quality Management
Sources of water pollution can be classified as either point sources or nonpoint

sources. With point-source pollution, the discharge of a contaminant can be attributed
to a specific location such as a wastewater discharge or chemical spill. In the case of
nonpoint-source pollution, the contamination source is distributed over a large area or
a large number of point sources (Chin, 2000). Nonpoint sources, such as runoff from
urbanized areas, irrigated lands, surface mining operations, and construction sites, are
frequently the largest uncontrolled contributors to water pollution. This chapter
focuses on the negative effects of stormwater runoff from urbanized areas and pre-
sents various techniques, often referred to as best management practices (BMPs),
used in mitigating these effects.
The effects of stormwater on receiving waters are a function of the types and concen-
trations of contaminants in the runoff and the frequency, intensity, and duration of
runoff flows. The nature of water-body impairment includes trash accumulation, ero-
sion and sedimentation, and contamination by harmful microorganisms and various
chemicals and toxins. Control measures designed to temporarily detain, reroute, or
treat stormwater can mitigate these effects, as can measures that eliminate contact
between stormwater and pollution sources. Stormwater quality controls have proven
to significantly contribute to improving the quality of urban runoff when implemented
as part of a comprehensive stormwater management plan.
Section 15.1 of this chapter summarizes the extent to which stormwater affects the
quality of receiving waters and describes various environmental effects of stormwater
runoff. Section 15.2 discusses hydrology as it relates to stormwater quality, as storm
events of different magnitudes tend to impact receiving-water quality in different
ways.
Section 15.3 introduces BMPs and presents descriptions of commonly used nonstruc-
tural and structural BMPs. The use of BMPs in stormwater management programs for
newly developed and existing communities and for construction projects is discussed
in Section 15.4.
608 Stormwater Quality Management Chapter 15
Although many similarities exist between stormwater controls implemented in differ-

ent regions, a particular type of stormwater control (such as a filter strip or grassed
swale) may be appropriate in one location but not another. It may also be appropriate
for some types of sites (for example, residential), but not others. For this reason, and
because individual state and local governments develop differing management plans,
the discussion of controls in this chapter is rather general. Much of the material pre-
sented here is modeled after the plan implemented in the Denver, Colorado (USA)
metropolitan area and published by the Urban Drainage and Flood Control District
(UDFCD, 1999). These discussions are presented only as examples to illustrate the
general process of control implementation. For another example, the reader is referred
to the manual published by the Water and Rivers Commission of Western Australia
(Water and Rivers Commission, 1998), which is accessible on the Internet.
15.1 ENVIRONMENTAL EFFECTS OF STORMWATER

RUNOFF
In the United States, the effect of urban runoff and storm sewers on water quality is
described in the 2000 National Water Quality Report (U.S. EPA, 2002). The report
summarizes the assessments of approximately 33 percent of U.S. waters, including
rivers, streams, lakes, and estuaries, relative to the water quality standards set forth for
each body. Waters not meeting water quality standards were classified as impaired.
For each impaired body, the leading cause of impairment was identified.
Table 15.1 shows that the urban runoff/storm sewers category was the leading source
of impairment for 13 percent of rivers and streams, 18 percent of lakes, and 32 percent
of estuaries in the United States. Relative to other sources, urban runoff/storm sewers
was the second leading source of impairment for estuaries, third leading source for
lakes, and fourth leading source for rivers and streams. Urban stormwater is a signifi-
cant cause of impaired water in many parts of the world, as well. Environment Canada
maintains the website for the United Nations Environment Programme’s (UNEP)
Global Environment Monitoring System (GEMS) Freshwater Quality Programme.
Sixty-nine nations participate in this program, whose activities are aimed at under-
standing freshwater quality issues. Information on water quality in participating coun-
tries is available through the website, including an “Annotated Digital Atlas of Global
Water Quality” that lists concentrations for a number of constituents in many major
rivers and lakes for the years 1976 through 1990.
Table 15.1 Impaired waters in the United States attributed to urban runoff/storm sewers (U.S. EPA, 2002)
Total Impairment Urban runoff/storm sewers as leading source
Receiving Body
Type Units Amount % of Total Assessed Amount % of Impaired Ranka
Rivers and streams miles 269,258 39 34,871 13 4
Lakes acres 7.7 million 45 1.37 million 18 3
Estuary mi2 15.7 million 51 5.05 million 32 2
a. Other sources: municipal point sources, industrial discharges, atmospheric deposition, agriculture, hydrologic modifications, and
resource extraction
Section 15.1 Environmental Effects of Stormwater Runoff 609
The adverse effects of stormwater pollutants on receiving waters include acute and
chronic toxic effects and secondary effects such as eutrophication. Stormwater also
causes changes to the physical properties and flow regime of streams. Table 15.2 lists
typical receiving-water problems associated with the long-term accumulation of pol-
lutants and short-term storm events (Pitt, 1995a, b). Many of these problems have
been found in urban receiving waters worldwide. Because these problems are so
diverse, a variety of individual stormwater controls must be used together to be effec-
tive.
Table 15.2 Typical receiving water problems in heavily developed urban drainage basins
(Pitt, 1995a, b)
Long-term problems associated with accumulations of contaminants:
Trash accumulation along receiving water corridors
Accumulation of debris and sediment in stormwater conveyance systems
Accumulation of fine-grained and contaminated sediment in receiving waters
Nuisance algal growths from nutrient discharges
Inedible fish, undrinkable water, and shifts to aquatic organisms less sensitive to heavy metals and
organic toxicants
Short-term problems associated with high pollutant concentrations or frequent high flows
(event related):
Swimming beach closures because of pathogenic microorganisms
Water-quality violations
Property damage from flooding and conveyance system failures
Habitat destruction caused by frequent high flow rates (bed scour, bank erosion, flushing of organisms
downstream, etc.)
Pollutants in Urban Stormwater

Although stormwater is often viewed by the public as being as clean as rain, it in fact
contains significant quantities of the same types of constituents more commonly asso-
ciated with municipal and/or industrial wastewater. A summary of some of the com-
mon constituents in urban stormwater runoff, their sources, and their effects is
provided in Table 15.3.
An example of the effects that stormwater may have on water quality in urban receiv-
ing waters is shown in Table 15.4. This table lists the average stormwater constituent
concentrations for the major watersheds in the Atlanta, Georgia area. Clearly, the
stormwater in this area has elevated concentrations of many important pollutants, with
some direct effects. These elevated concentrations, coupled with increased runoff vol-
umes, account for massive quantities of pollutant discharges.
Table 15.5 shows the contaminant percentage contributions of the different flow
sources for the South River Basin in Atlanta. Stormwater accounts for most of the
pollutant discharges for all but one of the listed constituents. (The exception is fecal
coliform bacteria, which are predominantly from CSO storage overflows at one facil-
ity.) This example is typical in that stormwater discharges usually account for the
majority of pollutant discharges in completely developed urban watersheds.
Table 15.3 Urban runoff pollutants (U.S. EPA, 1993)
Constituents Sources Effects
Sediments — TSS, turbidity, dis- Construction sites, urban/agricul- Habitat changes, stream turbid-
solved solids tural runoff, landfills, septic fields ity, recreation and aesthetic loss,
contaminant transport, bank ero-
sion
Nutrients — nitrate, nitrite, Lawn/agricultural runoff, land- Algal blooms, ammonia toxicity,
ammonia, organic nitrogen, phos- fills, septic fields, atmospheric nitrate toxicity
phate, total phosphorus deposition, erosion
Pathogens — total and fecal Urban/agricultural runoff, septic Ear/intestinal infections, shell-
coliforms, fecal streptococci systems, illicit sanitary connec- fish bed closure, recreational/aes-
viruses, E. Coli, enterooccus tions, domestic and wild animals thetic loss
Organic enrichment — BOD, Urban/agricultural runoff, land- Dissolved oxygen depletion,

COD, TOC, DO fills, septic systems odors, fish kills
Toxic pollutants — metals, Urban/agricultural runoff, pesti- Toxicity to humans and aquatic
organics cides/herbicides, underground life, bioaccumulation in the food
storage tanks, hazardous waste chain
sites, landfills, illegal disposals,
industrial discharges
Salts — sodium chloride Urban runoff, snowmelt Contamination of drinking water,
harmful to salt-intolerant plants
Note: TSS = total suspended solids, BOD = biochemical oxygen demand, COD = chemical oxygen
demand, TOC = total organic carbon, DO = dissolved oxygen
Table 15.4 Representative average stormwater and combined sewer overflow (CSO) constituent concentrations for
major land uses in Atlanta, Georgia (CH2M-Hill, 1998)
Fecal
Coliform Suspended Ammonia
Land Use BOD Bacteria Solids Nitrogen Phosphorus Copper Lead Zinc
Category (mg/L) (per 100 mL) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L)
Forest/open space 5 500 70 0.2 0.1 0.00 0.00 0.04
Residential 15 8,700 180 0.2 0.4 0.03 0.03 0.10
Institutional 10 1,400 140 0.5 0.3 0.01 0.02 0.14
Commercial 10 1,400 130 0.5 0.3 0.01 0.02 0.12
Industrial 14 2,300 90 0.4 0.3 0.02 0.02 0.16
Highway 14 1,400 140 0.5 0.3 0.01 0.02 0.14
Construction 13 1,000 1,000 0.2 0.2 0.05 0.02 0.40
CSOs 13 175,000 170 1.3 0.8 0.04 0.04 0.20
Table 15.5 Distribution of pollutant sources for the South River drainage basin, as
percentages of total loading (CH2M-Hill, 1998)
CSO Storage CSO

Constituent Stormwater Base Flow Overflow Effluenta
Flow 61 21 14 5
BOD 54 4 32 6
Fecal coliform bacteria 10 6 84 <1
Suspended solids 76 1 24 1
Ammonia nitrogen 37 19 28 16
Phosphorus 55 10 31 4
Copper 64 N/A 30 6
Lead 67 N/A 32 1
Zinc 67 N/A 29 4
a. CSO effluent undergoes some treatment prior to discharge.
Clearly, high levels of combined sewer overflow (CSO) control are needed in areas
having combined sewers. But urban stream bacterial levels are generally high even in
areas affected only by stormwater (fecal coliform levels greater than 10,000 per 100
mL are common), preventing any contact recreation uses. Unfortunately, bacteria lev-
els in stormwater are not amenable to sufficient reduction, although controls such as
wet detention ponds and filters may reduce the levels by about 50 to 90 percent (com-
pared to desirable reductions of much greater than 90 percent).
Ponds and lakes, by virtue of their long residence times relative to streams, are partic-
ularly susceptible to the adverse effects of contamination in urban runoff. They take
longer to recover from pollution episodes and they effectively act as sinks for nutri-
ents, metals, and sediments. Nutrient enrichment, in particular, can result in the unde-
sirable growth of algae and aquatic plants, which, in turn, can cause additional
problems related to dissolved oxygen levels and odors.
Effects on Streams
In addition to the increases in various pollutant concentrations, urbanization results in
increases in the frequency and magnitude of runoff events and decreases in stream
base flows. These changes result in secondary effects that include alteration of the
channel cross-section and stream alignment, increased erosion and stream sediment,
and, ultimately, degradation of the aquatic habitat. The subsections that follow
describe the effects of urban runoff on streams in more detail.
Flow Regime Alteration. Urbanization decreases the availability of pervious
surfaces and surface storage capacity, resulting in an increase in the frequency and
severity of flooding. Hydrograph peak flows in urban streams tend to be higher and
occur earlier than hydrographs for streams in undeveloped areas that are otherwise
similar. Also, the total volume of runoff substantially increases under developed con-
ditions as less rainfall infiltrates into the soil. These flow alterations result in changes
in flow regime that can have numerous effects on receiving streams (U.S. EPA, 1999;
U.S. EPA, 1997).
Increased peak flows can greatly accelerate the process of streambank erosion, and
the increased volume and velocity can overwhelm the natural capacity of the stream
network. As an area is urbanized, its streams typically experience an increase in the
frequency of bank-full flows. Because bank-full flows are highly erosive, streams
may become wider and straighter. Other effects on streams attributable to increased
flows include undercut and fallen banks, felled bushes and trees along the banks, and
exposed sewer and utility pipes. Sediments from eroding banks and other sources are
deposited in areas where the water slows, causing buildup, destruction of benthic
(bottom) habitat, and decreased capacity for flood conveyance. The decreased capac-
ity ultimately results in greater potential for further erosion.
Urbanization also affects dry-weather flows in streams. Because groundwater
recharge is reduced, the water table may be lowered. This change in watershed
hydrology reduces the contribution of base flow to the total flow in the stream, which
translates into decreased dry-weather flows, pool depths, and ripple intensities.
Habitat Effects. Changes in flow regime, the resulting changes in stream mor-
phology, and the increase in pollutants generally degrade the habitat for aquatic biota.
The quality of an aquatic habitat may be measured by the number of organisms, num-
ber of species, and types of species present and surviving in the water.
Effects of urban stormwater include a reduction in the total number and diversity of
macroinvertebrates and the emergence of more pollutant-tolerant species. A study in
the U.S. state of Delaware found that approximately 70 percent of the macroinverte-
brate community in streams in undeveloped, forested watersheds consisted of pollu-
tion-sensitive mayflies, stoneflies, and caddisflies, as compared to 20 percent in
urbanized watersheds (Maxted and Shaver, 1997).
The health of an ecosystem may also be measured by the abundance and variety of
fish species present and the presence of native species. A study compared fish popula-
tions in urbanized and nonurbanized sections of Coyote Creek, California (Pitt, 1995).
The results showed that the native fish were generally replaced by non-native fish in
the urbanized section of the stream. Limburg and Schmidt (1990) compared the abun-
dance of fish eggs and larvae with different land uses in New York. The researchers
concluded that significant declines in abundance were noted in watersheds having an
imperviousness of 10 percent or more.
Receiving-Water Considerations
The goals of an urban stormwater control program may be developed in terms of flow
control, pollutant removal, and pollutant source reduction (U.S. EPA, 1999). One of
the most difficult issues associated with stormwater management is that of determin-
ing reasonable receiving-water goals. It is not practical to believe that stormwater
management will enable natural conditions to be attained by receiving waters in
highly developed urban watersheds. The radical alteration of the hydrologic cycle and
the receiving-water habitat likely prevent any complete restoration to natural condi-
tions. In large watersheds where the urbanized portions comprise a small fraction of
the area, nearby streams of excellent quality may exist.
Typically, the greater the extent of urbanization in a watershed, the greater the nega-
tive effects on receiving waters. Beyond a certain point, receiving-water uses are
severely limited and are not likely to be restored, even with extensive stormwater
management. However, in some cases restoration efforts have significantly improved
receiving-water conditions. It is practical to expect only limited, but important, uses
of urban streams in heavily developed areas. Aesthetically pleasing waterways, some
habitat protection for relatively hardy aquatic and riparian life, and safe flow condi-
tions should be the minimum objectives. The basic uses that should be available
include:
• Noncontact recreation associated with stream-side parks and greenways. Nui-
sance odors, noxious sediment accumulations, and trash, for example, should be
rare. Hazardous debris should be nonexistent.
• A reasonable assemblage of tolerant aquatic life and riparian animals and plants
should be able to exist in the receiving water and surrounding area.
• Dangerous flows and floods should be controlled to minimize property damage
and prevent loss of life.
Without extensive knowledge and continuous monitoring, it is foolish to encourage
fishing and swimming in waters receiving substantial urban stormwater or at locations
near stormwater outfalls in large receiving waters. There is a small but definite health
risk associated with pathogens in the runoff. It may be possible for the effects of
urbanization to be sufficiently moderated using effective stormwater management
tools to allow downstream high-level uses such as fishing and swimming.
The determination of the level of stormwater control required for a receiving body to
achieve a specified quality level is a challenging process. In the United States, a Total
Maximum Daily Load (TMDL; see Section 14.2, page 598) is a calculation of the
maximum amount of a pollutant that a water body can receive and still meet water
quality standards. Based on this amount, pollution control responsibilities in a water-
shed are allocated among pollutant sources. Water quality standards are set by states,
territories, and tribes. They identify the uses for each water body [for example, drink-
ing water supply, contact recreation (swimming), and aquatic life support (fishing)],
and the scientific criteria to support that use.
15.2 HYDROLOGY FOR STORMWATER QUALITY

CONTROL
Small, frequently occurring storm events account for a significant portion of the total
annual pollutant load from urban drainage basins. Thus, capture and treatment of the
stormwater from such events is the recommended design approach for water-quality
enhancement, as opposed to facility designs that focus on less-frequent, larger events.
A design that not only controls the large events, but that also acknowledges the impor-
tance of small ones, is both possible and desirable.
The basis for developing appropriate stormwater controls is the understanding that
specific receiving-water problems are associated with specific rain-depth categories.
To identify which rain-depth categories are important for which receiving-water prob-
lems, long-term evaluations are needed. To illustrate the importance of relatively
small events in stormwater quality management, Pitt et al. (2000) conducted long-
term, continuous simulations for many locations in the United States.
An example from these simulations uses Atlanta, Georgia, rainfall data with the
SLAMM (Source Loading and Management Model) computer program (Pitt, 1986;
Pitt and Voorhees, 1995). The Atlanta simulations were based on eight years of rain-
fall records for the years 1985 through 1992, containing about 1,000 individual
storms. Dry periods of at least six hours were used to separate adjacent storm events,
and all rainfall depths were used, with the exception of trace values. (Similar analyses
were made using inter-event definitions ranging from 3 to 24 hours, with little differ-
ence in the conclusions.) These historical rainfall data were modeled in SLAMM with
typical medium-density and strip-commercial developments.
The results of the Atlanta simulations are illustrated in Figure 15.1. The “Accumula-
tive Rain Count” data for a particular rain depth yields the percentage of rainfall
events having a total depth less than or equal to that amount. The “Accumulative
Commercial Runoff Quantity” and “Accumulative Residential Runoff Quantity” plots
provide information on the proportion of runoff from these respective sources for
given rainfall depths.
The points that follow summarize the Atlanta evaluation. It should be noted that in
most areas of the United States (and in many parts of the world), the specific rain
depths shown below would be substantially smaller than those for Atlanta.
• Less than 0.3 in. (8 mm). These rains account for most of the events (about 60
percent), but little of the runoff volume (5 to 9 percent), and are therefore the eas-
iest to control. They produce much lower pollutant mass discharges and probably
have less effect on receiving waters than do other rains. However, the runoff pol-
lutant concentrations likely exceed regulatory standards for several categories of
critical pollutants, especially bacteria and some total recoverable metals. These
rains also cause large numbers of overflow events in uncontrolled combined sew-
ers and are very common, occurring once or twice a week. Rain depths less than
about 0.1 in. (3 mm) do not produce noticeable runoff.
• 0.3 to 4 in. These rains account for the majority of the runoff volume (about 85
percent) and produce moderate to high flows. They make up about 40 percent of
annual rain events. These rains occur approximately every two weeks, on aver-
age, and subject the receiving waters to frequent high-pollutant loads.
Section 15.2 Hydrology for Stormwater Quality Control 615
Figure 15.1
Modeled rainfall and
runoff accumulative
probability
distributions for
Atlanta, Georgia for
the years 1985 to
1992 (Pitt et al., 2000)
• More than 4 in. These rains probably produce the most damaging flows from a
habitat destruction standpoint and occur, on average, once every several months.
These recurring high flows establish a high energy gradient for the stream and
cause stream banks to become unstable. Fewer than 2 percent of rainfall events
are in this category, but they are responsible for about 5 to 9 percent of the annual
runoff and pollutant discharges.
• Very large rains. This category is seldom represented in field studies due to the
rarity of large events and the typically short duration of most field observations.
These rains occur only, on average, once every several decades or less often, and
produce extremely large flows. For example, the 8-year monitoring period for
Atlanta included a large rain of about 7 in. (an event with less than a 1-percent
probability of occurring in any one year, if the rainfall duration is about 1 day).
These extreme rains produce only a very small fraction of the annual average dis-
charge. However, when they do occur, great property and receiving-water dam-
age results. The receiving-water damage primarily consists of habitat destruction,
sediment scouring, and the flushing of organisms great distances downstream and
out of the system. The receiving water can conceivably recover naturally to
before-storm conditions within a few years.
These findings and others have resulted in recognition that the design of stormwater-
treatment facilities should be based on flows and volumes different from those used in
the design of conveyance structures. The volume used to size treatment facilities is
called the water quality capture volume, which corresponds to the volume of runoff
expected to result from a frequently occurring storm event, where the specific fre-
quency is typically defined by local regulations.
In the state of Georgia, for example, the water quality volume (WQV) sizing criterion
specifies a treatment volume equal to the amount of runoff generated by an 85th-
percentile storm event, which corresponds to 1.2 in. (30.5 mm) of rainfall for Atlanta.
Thus, all runoff from 85 percent of the storms occurring during the course of an aver-
age year is treated, as well as a portion of the runoff from all rainfall events greater
than 1.2 in. (30.5 mm).
The calculated volume for WQv is directly related to the amount of impervious cover
on a site (Atlanta Regional Commission, 2001). It is equivalent to a rainfall depth of
1.2 in. multiplied by the volumetric runoff coefficient (Rv) and the site area. RV is
calculated as
RV = 0.05 + 0.009I (15.1)
where RV = volumetric runoff coefficient

I = percentage of area with impervious cover (%)
The water quality volume is then calculated as
1.2R V A
WQ V = -----------------
- (15.2)
12
where WQV = water quality volume (ac-ft)
A = site area (ac)
For SI units, the equation is
WQ V = 305R V A (15.3)
where WQV = water quality volume (m3)

A = site area (ha)
In the state of New York, the water quality volume WQV is sized to capture and treat
90 percent of the average annual stormwater runoff volume (New York State Depart-
ment of Environmental Conservation, 2001), which corresponds to a precipitation
depth P ranging from 0.8 to 1.3 in. (20 to 33 mm) over the state. It is calculated as
PR V A
WQ V = -------------- (15.4)
12
where WQV = water quality volume (ac-ft)

P = depth of 90-percent rainfall event (mm)
A = site area (ac)
For SI units, the equation is
WQ V = 10PR V A (15.5)
where WQV = water quality volume (m3)

P = depth of 90-percent rainfall event (mm)
A = site area (ha)
Section 15.3 Stormwater Quality Best Management Practices 617
15.3 STORMWATER QUALITY BEST MANAGEMENT

PRACTICES
Best management practice (BMP), as applied to urban runoff management, is a term
adopted in the 1970s to represent actions and practices that could be used to reduce
urban runoff flow rates and constituent concentrations (WEF, 1998). An effective
stormwater management program consists of a series of BMPs that act in concert to
reduce the negative impacts of stormwater.
BMPs may be classified as structural or nonstructural. Structural BMPs are facilities
designed to reroute, temporarily detain, or treat stormwater prior to discharge into a
receiving body. Examples of structural BMPs are detention ponds, wetlands, porous
pavements, filter strips, and grassed swales. Structures designed to increase infiltra-
tion, such as porous pavement, are located on-site, but detention and treatment struc-
tures may be located away from the point of generation. Structural BMPs also include
temporary controls used during construction activities to limit sediment loads in run-
off, such as hay bales and silt fences. Stormwater treatment and control structures are
typically designed to operate in a passive manner; they have no moving parts and do
not require the attention of an operator. However, periodic maintenance is required.
Nonstructural BMPs include a variety of institutional and educational practices
designed to reduce the pollutant loads entering the stormwater system (Urbonas,
1999). Some nonstructural BMPs address land development and redevelopment pro-
cesses, and others focus on educating the public to modify behavior that contributes to
pollutant deposition on urban landscapes. Still others search out and disconnect illicit
wastewater connections, contain accidental spills, and enforce violations of ordi-
nances designed to prevent the deposition of pollutants on the urban landscape and its
uncontrolled transport downstream.
Detailed information on the design of stormwater controls may be found in references
such as the Urban Runoff Quality Management (WEF, 1998), Design and Construc-
tion of Urban Stormwater Management Systems (ASCE, 1992), or the Urban Storm
Drainage Criteria Manual (Urban Drainage and Flood Control District, 1999). The
U.S. EPA maintains the “National Menu of Best Management Practices for Storm
Water Phase II” on their website as required by the Storm Water Phase II rule (see
Section 14.2, page 595). For each BMP, the menu contains a description of the prac-
tice, its applicability and effectiveness, and references. Design equations and engi-
neering sketches are also provided for structural BMPs. Though the menu is intended
for small municipal, separate storm-sewer systems, the material is applicable to
stormwater management in any urban or suburban area.
The National Stormwater BMP Database was developed by a team of stormwater
experts associated with the Urban Water Resources Research Council of the Ameri-
can Society of Civil Engineers under a grant from the U.S. Environmental Protection
Agency and is available on the Internet. The data fields included in this database have
undergone intensive review by many experts and encompass a broad range of parame-
ters, including test-site location; watershed characteristics; climatic data; BMP design
and layout; monitoring instrumentation; and monitoring data for precipitation, flow,
and water quality. This database is one component of a broader project to identify fac-
tors that affect BMP performance, develop measures for assessing BMP performance,
and use the findings to implement design improvements (UWRRC, 2001).
Structural BMPs
Structural best management practices may be viewed as reactive approaches to storm-
water quality management, in that they seek to reduce peak runoff rates, reroute
flows, and/or improve stormwater quality during a runoff event. Descriptions of the
more commonly used structural BMPs are presented here. Engineers must ensure that
their designs accommodate local conditions such as weather, soils, and stream flow
patterns.
Porous Pavement and Porous Pavement Detention. Modular porous
pavement, which is shown in Figure 15.2, consists of open-void concrete block units
placed on gravel bedding. Surface voids in the blocks are filled with sand or a sandy
turf. The purpose of porous pavement is to reduce the imperviousness of a site by
encouraging rain falling directly on the porous pavement to infiltrate. A variation of
this approach is stabilized turf pavement, which consists of plastic rings affixed to a
filter fabric and underlain with gravel bedding. Voids between the plastic rings are
filled with sand and grass sod or seed. These types of controls are intended for use in
low-traffic areas, such as driveways and small parking lots, where traffic loading and
frequency do not require more conventional materials such as concrete or asphalt
pavement. Porous pavements can significantly reduce runoff rates and volumes.
Figure 15.2
Modular block porous
pavement
Credit: Photo courtesy of Unilock
Experience with porous pavement has shown that, when it is properly installed,
freeze-thaw cycles in cold climates do not adversely affect its performance. Porous
pavement is quite effective at removing sediment and associated constituents such as
oil, grease, and metals, but removal rates for dissolved constituents are low to moder-
ate. It has an uneven surface and therefore can be hazardous to those wearing heeled
footwear. It can also be rather costly to replace when the gravel and sand layers
become clogged.
Porous pavement detention (see Figure 15.3) is similar to porous pavement in that
many of the same types of modular block are used, the openings in the blocks are
filled with sand, and a granular bedding is used. However, porous pavement designed
to function as detention differs from basic porous pavement in a few ways. With
porous pavement detention, the porous area is slightly depressed [on the order of 1 to
2 in. (25 to 50 mm)] with respect to the surrounding pavement to create a shallow
ponding area for storage of the water quality capture volume. In addition to treating
the rain that falls directly on the pavement, the detention area may collect runoff from
other parts of the site. The water slowly infiltrates through the void spaces and exits
through a perforated pipe underdrain in the bedding material.
Aside from the benefits of additional detention and treatment of larger volumes of
stormwater, the advantages and disadvantages of porous pavement detention, as well
as its pollutant removal efficiencies, are similar to those for porous pavement.
Figure 15.3
Porous pavement
detention (from Urban
Drainage and Flood
Control District,
2002)
Grass Buffer Areas. A grass buffer, or filter strip, is a uniformly graded and
densely vegetated area of turf grass. Sheet flows over the grass tend to infiltrate, and
contaminants in the runoff tend to filter out and settle. Nevertheless, these strips pro-
vide only marginal pollutant-removal capability and generally should be used in con-
junction with other controls. Grass buffers are particularly appropriate between
detached sidewalks and curbs and along edges of parking lots where sheet flows occur
across the pavement. In addition to stormwater discharge and pollutant control, buff-
ers provide green space and can be aesthetically pleasing. Trees and shrubs can be
planted to improve their appearance and provide wildlife habitat.
Over a period of time, the accumulation of sediment in the portion of a grass buffer
adjacent to an impervious area (such as a parking lot) can cause runoff to pond on the
impervious surface. When this occurs, a portion of the buffer needs to be removed and
replaced.
Grass Swales. Grass swales, such as the one shown in Figure 15.4, are small
drainageways that convey concentrated flows. They may be used to convey flow along
roadway edges in lieu of curb and gutter, through park settings, and away from park-
ing areas. Grass swales should be densely vegetated with turf grass and should have
relatively flat side and longitudinal slopes. Swales should force the flow to be slow
and shallow, thereby facilitating infiltration and sedimentation. Berms and/or check
dams may be placed along swales at regular intervals to further slow flows and pro-
mote settling and infiltration.
Figure 15.4
Grass swale
Credit: Urban Drainage and Flood Control District, 2002
Some disadvantages of grass swales are that they can become soggy areas, provide
breeding areas for mosquitoes, and may require more right-of-way than a curb and
gutter or storm sewer. They are also prone to erosion during large storm events if flow
velocities are high. Positive attributes are that they are generally less-expensive to
construct and maintain than curbs and gutters or sewers and that they provide green
space. If wetland species of vegetation establish themselves in a swale, they can con-
tribute to pollution control. Like grass buffers, swales do not provide high pollutant-
removal efficiencies, though moderate removal rates of suspended sediments can
occur if flow velocities are low.
Stream Buffers and Greenways. Establishing greenways along urban
stream corridors (such as those shown on the plan in Figure 15.5) is becoming popular
in many cities. These greenways can provide important recreational benefits and help
protect the stream riparian zone and aquatic habitat. Although greenways are com-
monly described as areas that provide water quality treatment by filtering stormwater,
little filtering occurs in actuality because most stormwater enters the stream through
channels and underground pipes. However, buffer areas can provide shade that mod-
erates stream temperatures and provides protection for the stream habitat.
Figure 15.5
Greenway Master
Plan for Fletcher,
North Carolina
Credit: Used with permission from Town of Fletcher, North Carolina
Porous Landscape Detention. Like porous pavement detention, porous land-

scape detention consists of slightly depressed areas for temporary detainment of all or
a portion of the water quality capture volume. As its name implies, however, porous
landscape detention is provided in landscaped rather than paved areas. The low-lying
vegetated (or otherwise landscaped) area is underlain by a sand filter bed and perfo-
rated-pipe underdrain system. The underdrains discharge the filtered stormwater to a
suitable location.
Porous landscape detention facilities (see Figure 15.6) can be easily integrated into
islands in parking lots, roadway medians, and roadside swale features. Experience has
shown that they work well with common grass species such as bluegrass, whereas
extended detention basins (see next section) often become too soggy to maintain blue-
grass. Pollutant removal rates can be significant and should be at least equal to those
afforded by sand filters. Disadvantages of porous landscape detention facilities are
that they can become clogged by excessive amounts of sediment. They must also be
kept away from building foundations or other areas where flooding could create prob-
lems.
Extended Detention Basins. For development sites or drainage basins of
moderate area, with significant water quality capture volume, use of an extended
detention basin (see Figure 15.7) may be appropriate. A basin of this type is designed
with basically the same approach used to design dry detention ponds (see Chapter
12); however, the pond outlet is usually much smaller than would be required to meet
peak outflow design criteria, such as maintaining predevelopment peak flow rates.
With an extended detention basin, the outlet is sized to provide the residence time
necessary for suspended sediments in the water quality capture volume to settle.
Figure 15.6
Porous landscape
detention
Figure 15.7
Plan and profile views
of an extended
detention basin

Extended detention basins can also be used to control peak runoff rates for larger
storm events if they provide additional storage volume above the pool level associated
with the water quality capture volume. The outlet(s) used for this purpose should be
located above the pool level corresponding to the water quality capture volume.
Extended detention basins are quite effective in removing sediments, associated
heavy metals, and other constituents that have an affinity for sediments, but removal
rates for nutrients are low to moderate. Nutrient removal can be enhanced by provid-
ing wetland vegetation in the bottom of the basin or through the use of additional con-
trols. Because extended detention basins are designed to permit water to drain very
slowly, they may be inundated much of the time during dry-weather periods. This
inundation is a potential safety hazard and may create a mosquito-breeding habitat.
Sand Filter Extended Detention Basins. A sand filter extended detention
basin is an extended detention basin with the addition of a sand filter bed and a gravel
layer with an underdrain system beneath. The basin is sized to accept the water qual-
ity capture volume and infiltrate it. Pollutant removal is accomplished through the
joint mechanisms of settling and filtration. Because the sand bed can clog over time,
this type of control is best suited to sites where sediment loads are expected to be low.
For instance, a sand filter extended detention basin should not be put into operation
during the construction phase of a project. This type of control should also not be
used in locations with a base flow.
Wet Detention Ponds. A wet detention pond (see Figure 15.8 and Chapter 12)
is a structural control appropriate for large drainage basins. This type of pond has a
permanent pool of water below the invert of its outlet structure that can be depleted
only through evaporation and infiltration (that is, water below the level of the outlet is
retained). Temporary extended detention storage is provided above the permanent
pool level to capture runoff and facilitate sedimentation. The water in the permanent
pool is partially or completely replaced during a stormwater runoff event.
Maintenance of the permanent pool requires a continual or nearly continual base flow
to replenish losses due to evaporation and infiltration. For this reason, a wet detention
pond is best suited for use in regional applications where the drainage area is large
enough to establish a reliable base flow. At the individual site level, this type of con-
trol may not work well unless a constant source of water is provided artificially.
Wet detention ponds can create wildlife habitat opportunities; they offer amenities
related to recreation, aesthetics, and open space; and they can be part of a larger flood
peak control storage basin. They can also achieve moderate-to-high removal efficien-
cies for many urban pollutants. Their primary disadvantages include safety concerns,
floating litter, scum and algae problems, odors, and mosquito-breeding areas.
Removal of accumulated sediment is also more difficult than with alternative types of
basins that completely drain between storm events.
Constructed Wetland Basins. A constructed wetland basin is a shallow
retention pond or wet detention pond. This type of control is appropriate for large
drainage areas where a continual or nearly continual base flow is present to sustain
the growth of rushes, willows, cattails, and similar wetland species. It is an on-line
control that treats runoff through settling and biological uptake.
Figure 15.8
Plan and section of a
wet detention pond
In the United States, regulations intended to protect natural wetlands have recognized
a distinction between natural and constructed wetlands. Constructed wetlands gener-
ally are not permitted on receiving water bodies, nor can they be used to mitigate
losses of natural wetlands. Constructed wetlands can be disturbed for maintenance
purposes, though the U.S. Army Corps of Engineers has established criteria for the
maximum area that may be disturbed. It is recommended that any activity that could
disturb a constructed wetland be cleared with the Corps to ensure that it is covered by
some type of Section 404 permit (see Section 13.2).
Constructed Wetland Channels. A constructed wetland channel, like a con-

structed wetland basin, is a control with a continual base flow that sustains reeds, wil-
lows, rushes, and similar wetland species. A wetland channel, unlike a wetland basin,
does not attempt to temporarily store the water quality capture volume, but rather is
designed to promote a slow and shallow flow conducive to settling solids. Because of
the slow flow, residence times within the channel reach may be long enough to pro-
mote biological uptake by the vegetation. A wetland can be established within an oth-
erwise man-made channel, which can then act as both a conveyance facility and a
water-quality-enhancement facility. This design is particularly suited to wide, gently
sloping channels.
Like constructed wetland basins, constructed wetland channels are man-made, and, in
the United States, they are regulated under a different set of rules than are natural wet-
lands. However, any planned disturbance of a constructed wetland channel in the
United States should first be cleared with the U.S. Army Corps of Engineers.
Covering of Storage and Handling Areas. In locations where toxic or
non-toxic potential pollutants can accumulate on exterior surfaces, such as loading
docks or outdoor storage and stockpiling facilities, covering the area can eliminate or
greatly reduce stormwater wash-off or wind erosion of the accumulated materials.
Covers can be temporary or permanent, and can consist of tarpaulins, roofing, com-
pletely enclosed structures, or other such features.
Spill Containment and Control. Accidental spills are prone to occur at facil-
ities where bulk liquids are handled and dispensed, such as gasoline filling stations.
Spills can have devastating effects on water quality and aquatic ecosystems; therefore,
spill containment and control facilities are required on such sites. These facilities can
consist of berms or curbs to contain spills on the ground surface where they occur, or
devices such as oil and grease traps that take advantage of the different densities of
the spilled contaminants and water. In any case, spill-containment facilities should be
readily accessible for removal of pollutants. Stormwater collected in the containment
structure during a spill event may require treatment.
Critical Source Area Control. Special controls are needed in many critical
source areas (ultra-urban areas) such as vehicle service areas, heavy-equipment stor-
age and maintenance yards, scrap yards, and areas having frequent, high levels of
vehicle activity (convenience stores, for example). These areas usually have substan-
tially greater unit-area stormwater pollutant discharges than other areas in the water-
shed and should therefore be controlled near the site. Typical critical source area
controls include stormwater filters, subsurface multichambered treatment tanks, grit
chambers, and oil/water separators. Some of these controls (the filters and multiple
treatment units, for example) can provide very high levels of pollutant reduction on
the order of 90 percent or greater for many critical pollutants. Poor performance is
usually associated with greatly undersized units, bad designs that hinder needed
maintenance, and improper understanding of stormwater characteristics.
Source Area Biofiltration Devices. In many situations, significant stormwa-
ter control is possible by altering the microscale drainage paths around structures to
allow enhanced infiltration. Rain gardens are an example in which roof runoff is
directed to depressed lawn areas. These gardens are typically only several inches deep
and are planted with moisture-tolerant vegetation. They can be designed to easily
infiltrate all roof runoff from small to moderate rains. Another biofiltration approach
is to amend urban soils with compost or other organic matter to enhance infiltration,
capture runoff pollutants, and reduce the adverse effects of soil compaction associated
with construction.
Use of Cisterns and Rain Barrels. On-site temporary storage of runoff from
relatively clean areas (such as most roofs and residential driveways) in underground
cisterns or simple rain barrels allows this water to be used for irrigation during dry
periods. In many areas, it is possible to provide several weeks of landscaping irriga-
tion with the stored water while significantly reducing the amount of runoff during
rains.
Effectiveness of Structural BMPs

The first concern when investigating stormwater treatment methods is determining the
needed level of stormwater control. This issue should be considered carefully as it has
a great effect on the cost of the stormwater management program. Problems that need
to be addressed range from sewer maintenance issues to protection of many receiving-
water uses.
For example, Laplace et al. (1992) recommend that all particles greater than about 1
to 2 mm in diameter be removed from stormwater to prevent deposition in sewers.
The specific value is dependent on the energy gradient and hydraulic radius of the
flow in the conveyance system. This treatment objective can be achieved easily using
a number of cost-effective source-area and inlet treatment practices. In contrast, far
greater levels of control are typically required to prevent excessive receiving-water
degradation. Specific treatment goals usually specify a reduction in suspended solids
concentration of about 80 percent. In most stormwaters, this level of treatment would
require the removal of most particulates greater than about 10 µm in diameter.
The selection of a treatment goal must be done with great care. The Engineering
Foundation/ASCE Mt. Crested Butte Conference held in 1993 included many presen-
tations describing receiving-water impacts associated with stormwater discharges
(Herricks, 1995). Similarly, Pitt (1996) summarized numerous issues concerning
potential groundwater impacts associated with subsurface stormwater disposal. These
references illustrate the magnitude and variation of typical problems resulting from
untreated stormwater. Because of these variations, control programs must be area-
specific.
It is important to understand which stormwater controls are suitable for specific site
conditions and can achieve the required treatment goals. This knowledge will assist in
the realistic evaluation of each practice for technical feasibility, implementation costs,
and long-term maintenance requirements and costs. Unfortunately, the reliability and
performance characteristics of many of these controls are not well established, with
most still in the development stage. Emerging controls can be effective, but there is
not a large amount of historical data on which to base designs and be confident that
performance criteria will be met under local conditions. The most promising and best
understood stormwater control practices are wet detention ponds. Less reliable in
terms of predicting performance, but showing promise, are stormwater filters, wet-
lands, and percolation basins (Roesner et al., 1989). Grass swales have also shown
great promise in the EPA’s Nationwide Urban Runoff Program (NURP) (U.S. EPA,
1983) and other research projects.
A study of 11 types of stormwater quality- and quantity-control practices used in
Prince George’s County, Maryland (Metropolitan Washington Council of Govern-
ments, 1992) was conducted to examine their performance and longevity. That report
concluded that several types of practices had either failed or were not performing as
well as intended. Generally, wet ponds, artificial marshes, sand filters, and infiltration
trenches achieved moderate to high levels of removal for both particulate and soluble
pollutants. Only wet ponds and artificial marshes demonstrated an ability to function
for a relatively long time without frequent maintenance. They were also much more
robust and functioned adequately under a wider range of marginal conditions. Infiltra-
tion basins, porous pavements, grass filters, swales, smaller “pocket” wetlands, dry
extended detention ponds, and oil/grit separators were found to perform poorly.
Stormwater infiltration techniques had high failure rates, which often could be attrib-
uted to poor initial site selection and/or lack of proper maintenance. Other perfor-
mance problems were likely a function of poor design, improper installation,
inadequate maintenance, and/or unsuitable placement of the control. Greater attention
to these details would probably reduce the failure rate of these practices.
In addition to performance, important design considerations include
• Safety for maintenance access and operations
• Possible hazards to the general public (for example, drowning)
• Possible public nuisances (such as mosquito breeding)
• Acceptability to the public (For example, does the project enhance area aesthetics
and property values?)
The majority of stormwater treatment processes are most effective for the removal of
particulates only, especially the settleable-solids fraction. Removal of dissolved or
colloidal pollutants is minimal; therefore, pollution prevention or controls at the
source (which are frequently nonstructural BMPs) offer a more effective way to
reduce dissolved pollutant concentrations. Fortunately, most toxic stormwater pollut-
ants (heavy metals and organic compounds) are also associated with stormwater par-
ticulates (Pitt et al., 1995). Therefore, removal of the solids will also remove a large
portion of the pollutants of interest. Notable exceptions of potential concern include
nitrates, chlorides, zinc, pathogens, 1,3-dichlorobenzene, fluoranthene, and pyrene.
As shown in Table 15.6, structural controls can be very effective for the reduction of
stormwater pollutant concentrations if they are correctly designed, constructed, and
maintained. Many of these devices also reduce runoff volumes, significantly lowering
pollutant mass discharges. Grass buffers, grass swales, porous pavements, and biofil-
tration devices (porous landscape detention) can all be designed to provide up to a 90-
percent reduction in runoff volume. The mass discharge reductions for these devices
can therefore be much greater than shown in Table 15.6, which considers concentra-
tion reductions alone. Combinations of stormwater controls are even more effective,
but removal efficiencies reported in the table should not be viewed as additive.
Table 15.6 Pollutant concentration removal efficiencies, in percent, for common stormwater controls (UDFCD, 2002)
Type of Control TSS TP TN TZ TPb BOD Bacteria
Grass buffer 10−50 0−30 0−10 0−10 N/A N/A N/A
Grass swale 20−60 0−40 0−30 0−40 N/A N/A N/A
Modular block porous pavement 80−95 65 75−85 98 80 80 N/A
Porous pavement detention 8−96 5−92 −130−85 10−98 60−80 60−80 N/A
Porous landscape detention 8−96 5−92 −100−85 10−98 60−90 60−80 N/A
Extended detention basin 50−70 10−20 10−20 30−60 75−90 N/A 50−90
Constructed wetland basin 40−94 −4−90 21 −29−82 27−94 18 N/A
Retention pond 70−91 0−79 0−80 0−71 9−95 0−69 N/A
Sand filter extended detention 8−96 5−92 −129−84 10−98 60−80 60−80 N/A
Constructed wetland channel 20−60 0−40 0−30 0−40 N/A N/A N/A
Notes: TSS = total suspended solids, TP = total phosphorus, TN = total nitrogen, TZ = total zinc, TPb = total lead,
BOD = biochemical oxygen demand. Negative values reflect an increase in pollutant concentration.
Noticeably lacking in Table 15.6 are performance values for nonstructural stormwater
controls. Unfortunately, nonstructural controls are difficult to evaluate, even though
substantial research has been conducted on some public works practices (street clean-
ing and catch basins, for example). Although the direct environmental benefits of
many nonstructural controls appear to be limited, some practices (such as public edu-
cation) can play an important role in increasing overall support for a stormwater con-
trol program.
Nonstructural BMPs
Nonstructural controls are often called source controls or pollution prevention
because they seek to reduce or eliminate the introduction of contaminants into storm-
water. They cannot wholly eliminate pollution—and thus must be implemented along
with structural controls—but they can make structural controls more effective by
reducing the loadings that structural controls must handle. The principal disadvantage
of nonstructural controls is that they require changes in the activities, behaviors, and
attitudes of people. Such changes are very difficult to achieve and require sustained
efforts on the part of those attempting to implement them. The main benefit, however,
is that nonstructural controls are usually less expensive than structural controls.
Nonstructural controls take many forms, but each can be categorized by the purpose it
serves:
• Waste minimization
• Good housekeeping
• Preventive maintenance
• Exposure minimization
• Spill prevention
• Public education
• Mitigation
Within these categories, some controls are more appropriate for industrial and com-
mercial sites than for residential sites, and some apply to construction activities. The
applicability of various structural and nonstructural BMPs in different situations is
discussed further in Section 15.4.
Public works practices are commonly considered to be nonstructural stormwater con-
trols. These practices, which should be considered in all areas, include the use of
catch basins with frequent cleaning (with the frequency depending on sump volume
and drainage-area characteristics), street cleaning in industrial areas and for leaf con-
trol, and field screening procedures to identify and repair inappropriate discharges to
the storm drainage system. The subsections that follow discuss these practices in
detail.
Catch Basin Cleaning. As described in Section 11.1 (page 416), the typical
catch basin configuration consists of an appropriately sized sump with a hood over the
outlet pipe. Stormwater bed load and a low to moderate amount of suspended solids
(about 30 to 45 percent of the annual load) are trapped in the sump. The larger frac-
tion of the sediment in the flowing stormwater will be trapped in preference to finer
material, which has greater amounts of associated pollutants.
The catch basin configuration is robust and should be used in most areas. In almost all
full-scale field investigations, this design withstood extreme flows with little scour
loss, no significant differences between the quality of the supernatant water (the upper
portion of the water column in the structure that is discharged through the outlet) and
runoff quality, and minimal insect problems. Catch basins are designed for hydraulic
capacity using conventional procedures for the design of grate and outlet dimensions,
and the sump is designed based on the desired cleaning frequency (Lager et al., 1997).
Street Cleaning. Basic street cleaning should be conducted about four times a
year in industrial and other areas having obvious high dirt loadings in streets. Addi-
tional street cleaning should be scheduled in heavily vegetated areas to remove leaves
before rains wash them into the drainage system. In areas having long dry periods
before the onset of seasonal rains (such as central and southern California), street
cleaning may be able to reduce runoff pollutant loadings by a low to moderate amount
(possibly up to 30 percent). However, in areas having more consistent rains, very
small pollutant reductions would likely result. Recent innovations in street-cleaning
technology that allow more efficient removal of fine particles should result in higher
levels of stormwater pollution control, especially in areas having large amounts of
pavement that can be cleaned.
Inappropriate Discharge Control. Screening procedures (Pitt et al., 1993)
for identifying and correcting inappropriate discharges into storm drainage systems
are also needed. Many separate storm drainage systems have flowing water during dry
weather, with some discharging highly polluted water such as raw sewage, cleaning
waters, or industrial wastewaters. These waste discharges are inappropriate and need
to be identified and controlled. Communities with successful control programs for
inappropriate discharges have seen substantial improvements in receiving-water
quality.
Public Education Programs. Public education is an important and common

nonstructural component of many stormwater control programs. Education efforts
commonly include advertising to inform the general public about nonpoint-source
water-pollution problems and help them make proper decisions concerning the pro-
tection of runoff quality. Public use of pesticides and fertilizers and disposal of pet
wastes and household toxic wastes are examples of areas in which such education can
be beneficial.
Unfortunately, monitoring and quantifying the direct water-quality benefits of these
programs are difficult. Reported results include decreases in the number of instances
of improper disposal of wastes in gutters and inlets and greater care in disposing of
pet wastes. However, the greatest benefits of public education programs are support
for better enforcement of construction-site erosion and stormwater regulations, and
support for increased fees to help pay for these programs.
15.4 APPROACHES TO APPLYING BEST MANAGEMENT

PRACTICES
The BMPs described in this chapter may be used to develop an effective stormwater
management program for a municipality or regional government. An integrated set of
BMPs that includes source controls and on-site and off-site practices must be custom-
ized for application to a specific area. In the selection and site-specific design of
BMPs, many factors must be considered, including costs, performance efficiency,
applicable regulations, land-area requirements, public safety, and aesthetics.
The total design, construction, and maintenance costs should be considered. Gener-
ally, nonstructural controls are more cost-effective than structural controls. However,
use of nonstructural controls shifts the cost burden from the municipality to property
Section 15.4 Approaches to Applying Best Management Practices 631
owners and residents, and procedures must be implemented to ensure that participa-
tion levels are high. Structural BMPs are generally more cost-effective when included
in the planning stages of new developments.
The sections that follow describe programs that are suitable for newly developed
areas, retrofitting existing areas, and construction projects.
Implementation of Controls in Newly Developing

Areas
Stormwater controls established at the time of development can take advantage of
simple grading and site layout options that provide significant stormwater benefits at
little cost. In addition, the design of the storm drainage system can be easily modified
to meet pollution-control objectives as well as drainage objectives. Most important,
regional stormwater control facilities can be effectively located and sized without
much interference with existing development. The following list describes stormwater
management scenarios likely to be effective under various conditions for new devel-
opments (Pitt et al., 2000):
• Low- and very low-density residential developments [greater than 2 ac (0.8
ha) lot sizes]. Most stormwater should be infiltrated on-site by directing runoff
from paved and roof areas to small bioretention areas (retention areas consisting
of a planted soil bed over an underdrained sand layer). Disturbed soil areas
should use compost-amended soils and should otherwise be constructed to mini-
mize soil compaction. Roads should have grass swale drainage to accommodate
moderate to large storms.
• Medium-density developments [¼ to 2 acre (0.1 to 0.8 ha) lot sizes]. Again,
most stormwater should be infiltrated on-site by directing runoff from paved and
roof areas to small bioretention areas. Paved areas should be minimized, and
porous pavements and paver blocks should be used for walkways, driveways,
overflow parking areas, and so forth. Disturbed soil areas should have compost-
amended soils and be constructed to minimize soil compaction. Grass swales
should be used where possible to accommodate runoff from moderate to large
storm events. Use of grass swales in commercial and industrial areas must also
consider groundwater contamination potential and available space. Runoff from
commercial and industrial areas should be controlled using wet detention ponds,
and critical source areas that have excessive pollution-generating potential should
have special controls.
• High-density developments. Combined sanitary- and storm-sewer systems can
be effectively used in these areas, provided they are legally acceptable. On-site
infiltration of the least-contaminated stormwater (such as that from roofs and
landscaped areas) is needed to minimize wet-weather flows. On-site storage of
sanitary wastewaters during wet weather (Preul, 1996), extensive use of in-line
and off-line storage, and the use of effective high-rate treatment systems can min-
imize the damage associated with any combined sewer overflows (CSOs). The
treatment of the wet-weather flows at the wastewater treatment facility would
likely result in less pollutant discharge in these areas than if conventional sepa-
rate wastewater collection systems were used.
Denver’s Urban Drainage and Flood Control District (UDFCD, 2002) recommends a
four-step process for the selection of structural controls for newly developed and rede-
veloped urban areas. Selection of controls for a development site should be made col-
laboratively between the developer, design engineer, and local jurisdiction. The
subsections that follow summarize this process.
Step 1: Employ runoff reduction practices. One of the most effective
ways of reducing runoff peaks and volumes from urbanizing areas is to minimize the
directly connected impervious areas that contribute to the stormwater conveyance sys-
tem. In so doing, infiltration of rainfall is promoted, with consequent reductions in
pollution and conveyance system costs.
Low-impact development (LID) uses a variety of strategies intended to minimize the
amount of stormwater and stormwater pollution from newly developing areas. Reduc-
tion of the directly connected impervious areas in a drainage basin can be accom-
plished in a number of ways. Narrower roadways save not only on the initial capital
cost of construction, but also on long-term maintenance costs. For driveways and
other low-traffic and parking areas, modular-block porous pavement or reinforced turf
can be used in lieu of asphaltic or concrete pavement. Grass buffers, such as narrow
landscaped areas between detached sidewalks and curbs, also encourage infiltration
and essentially make the walks indirectly connected impervious areas. Use of grass
swales instead of curb and gutter (and even storm sewers themselves in some
instances) also slows runoff and promotes infiltration. With LID strategies, it is possi-
ble in low- and medium-density residential areas to produce runoff rates, volumes,
and pollutant mass discharges comparable to those of predevelopment conditions.
Step 2: Provide water quality capture volume. Provision of tempo-
rary storage for stormwater (that is, detention storage) is a fundamental requirement
for any site where stormwater quality is to be effectively addressed. The purpose of
temporary storage is to permit settling of suspended sediments and to provide suffi-
cient residence time for the settling to be effective. As previously discussed, deten-
tion, for the purpose of water quality enhancement, should focus on small, frequently
occurring storm events. Additional detention volume for the management of storm
runoff peaks may be provided either separately or in a surcharge pool area above the
water quality pool. The types of structural controls that may be employed to provide
the water quality capture volume were described in Section 15.2; they are porous
pavement detention, porous landscape detention, extended detention basins, sand fil-
ter extended detention basins, constructed wetland basins, and retention ponds.
Step 3: Stabilize drainageways. Erosion of drainageways, whether natural
or man-made, is a major source of contaminating sediments and associated pollutants
such as phosphorus. Increases in peak rates and volumes of runoff caused by urban-
ization can lead to uncontrolled bed and bank erosion. Stabilization of channels may
be necessary to minimize these potential problems, and can be accomplished in three
ways:
• Construct a grass-, riprap-, or concrete-lined channel. This method is the tra-

ditional engineering approach and can nearly eliminate bed and bank erosion
completely. Unfortunately, these channel stabilization practices do not provide
water-quality enhancement or wetland habitat. The Urban Drainage and Flood
Control District does not recommend the use of riprap or concrete channels for
flood conveyance, but it does recommend the use of rock-lined, low-flow chan-
nels for conveyance of small base flows.
• Avoid bed degradation through the use of grade control structures and by
reshaping sharp channel bends and steep banks. These stabilization efforts
should be accomplished in such a way as to minimize disturbances to desirable
vegetation and habitat.
• Employ a constructed wetland channel. This approach gives special attention
to stormwater quality, and is one form of a structural control.
Step 4: Consider the need for additional controls. If a site is being
developed or redeveloped for industrial or commercial purposes, specialized types of
controls may be needed. Included among these are spill containment and control facil-
ities and protection (that is, covering) of storage and handling areas. Discussions of
these controls were presented in Section 15.3.
Stormwater Controls Suitable for Retrofitting in

Existing Areas
The ability to construct new stormwater controls in developed areas is severely lim-
ited by both the availability of suitable controls and the extent of control that can be
accomplished. In addition, retrofitting a control is typically much more costly than if
the same control were installed at the time of development. Retrofitting stormwater
controls is generally limited to the practices shown in the following list. In general,
items near the top of this list are more popular; however, they may not be the least
expensive or most effective.
• Public works practices
– Enhanced street cleaning
– Increased catch basin cleaning
– Repairing trash hoods in catch basins
– Installing catch basin inserts
– Rebuilding inlets to create catch basins
– Litter control campaigning, with increases in availability of trash receptacles
and trash pickup frequency
– Public environmental education campaigning
– Pet waste control enforcement
– Household toxicant collection
– Enhanced enforcement of erosion control requirements
– Modification of public place landscaping maintenance
• Infiltration practices at source areas

– Routing rooftop drains toward pervious areas
– Amending soils with compost
– Modification of residential and commercial landscaping (such as construct-
ing small bioretention areas)
– Providing French drains for collection of roof runoff where pervious area is
limited
– Replacement of pavement with porous paver blocks
• Outfall wet detention ponds
– Modification of existing dry detention ponds for enhanced pollution control
– Enhancement of performance at existing wet ponds (options include modify-
ing the outlet structure, constructing a forebay to trap sediment near the inlet,
and providing additional treatment using media filtration or wetland treat-
ment)
– Construction of new wet detention ponds in available areas
• Control runoff at critical source areas (such as vehicle service facilities, scrap
yards, etc.)
– Use of sand perimeter filters
– Construction of underground sedimentation/filtration units
• Other controls
– Replacement of exposed galvanized metal with non-polluting material
– Providing new regional stormwater controls downstream of existing devel-
oped areas
Construction Project Best Management Practices

Construction activities that disturb natural soils and protective vegetation make soils
more prone to erosion by wind and water. Eroded sediments may be transported onto
roadway surfaces and then spread through the site’s neighborhood by passing vehi-
cles. The sediment can then clog newly constructed inlets, storm sewers, and natural
channels and can cause disruptions to natural aquatic life. Construction BMPs help
limit erosion and sedimentation and provide protection to natural and man-made
drainageways and stream channels.
Design engineers and construction contractors are often required to develop and
obtain approval on erosion and sedimentation control plans for a site before construc-
tion begins. Ultimately, however, responsibility for erosion and sedimentation control
falls on the owner of the site. Local review agencies should be consulted for guidance
and information on specific requirements.
An erosion and sediment control plan consists of three major elements: erosion con-
trol measures for disturbed areas, sediment-transport control measures, and drainage-
way protection. Commonly used practices for addressing each element are described
in the subsections that follow. More comprehensive discussions on sediment and ero-
sion measures for construction activities may be found in Best Management Practices
for Erosion and Sediment Control (Eastern Federal Lands Highway Design, 1995),
Urban Drainage and Flood Control District (2002), and the U.S. EPA’s “National
Menu of Best Management Practices for Storm Water Phase II” web site.
Erosion Control Measures. Erosion control measures should be used to limit
erosion of soil from disturbed areas on a construction site. Mulching is the application
of plant residues or other materials to the soil surface to provide temporary stabiliza-
tion. Mulch may be applied to surfaces that are exposed for a short time or during
seeding after final grades are established. Mulch protects the soil from rainfall impact,
retards overland flow, and promotes the growth of vegetation by protecting the seed
and fostering germination. Long-stemmed grass hay or cereal straw are typically used
in mulch, but synthetic chemical materials are also available. On steeper slopes, mats,
blankets, and nets are sometimes used to anchor the mulch.
Revegetation is the establishment of a viable plant community. Temporary vegetation
is typically required on surfaces that will be exposed for one year or longer. Perma-
nent vegetation is required on all areas that will not be covered by buildings, pave-
ment, or other structures. Specifications for topsoil, fertilizer, seed mixtures, and
seeding techniques depend on the use of the site and the local climate.
Road cuts, road fills, and parking areas on a construction site are susceptible to ero-
sion after initial grading. The aggregate base course for the pavement should be
applied promptly on these areas following grading to prevent erosion. Another result
of grading is soil stockpiles; these should be seeded with a temporary or permanent
grass cover within 14 days of stockpile completion if the stockpile will remain on the
site for more than 60 days. Mulching is recommended to ensure vegetation establish-
ment. If a stockpile is located near a drainageway, additional sediment control mea-
sures, such as a temporary diversion dike or silt fence, should be provided (see next
subsection).
Sediment Transport Control Measures. Sediment transport control mea-
sures are used to limit transport of sediment to off-site properties and downstream
receiving waters. One such control measure is a check dam, which consists of a small
temporary obstruction in a ditch or waterway used to prevent sediment transport by
reducing the velocity of flow. Check dams are most commonly constructed of loose
rock, logs, or compacted earth.
A temporary slope drain is a flexible or rigid conduit used to effectively transport run-
off down disturbed slopes. These structures are used during grading operations until
the permanent drainage structures are installed and slopes are permanently stabilized.
Corrugated plastic pipe or flexible tubing are commonly used. For flatter, shorter
slopes, a polyethylene-lined channel may be installed.
Diversions are measures used temporarily or permanently to divert water around an
area that is either under construction, being stabilized, or prone to erosion. The diver-
sion of water is accomplished with channels, berms, or temporary culverts.
A silt fence is a temporary barrier used to filter sediment from sheet flow. The barrier
causes water to pond upstream of the fence and deposit suspended sediment. This is
the most common type of perimeter control for removing sediment before it leaves the
site. The fence is constructed with a synthetic filter fabric mounted on posts and
embedded in the ground.
Temporary sediment traps are small impoundments that detain runoff from small
drainage areas [less than 5 ac (2 ha)] so that sediments can settle out. They are usually
constructed in drainageways, where a depression is created through excavation and
construction of an embankment. The discharge is controlled by a rock spillway or
other outlet that slows the release of runoff. Another type of sediment trap involves
excavating a depression around a storm sewer inlet structure to detain stormwater and
cause sediments to settle before water enters the storm sewer system.
A sediment basin is a temporary impoundment used to detain runoff from an area of
up to approximately 100 ac (40 ha) to settle sediments before discharging stormwater
from the construction site. The design of a sediment basin is similar to that of the
extended detention basins and wet detention ponds described in Section 15.3. The
outlet is designed to ensure a sufficiently long detention time, and frequently consists
of a perforated riser pipe. If permanent detention facilities are planned for the site, the
sediment basin can often be constructed so that it can be easily adapted to this purpose
by reconfiguring the outlet structure when construction controls are no longer neces-
sary.
Drainageway Protection. Drainageway protection measures prevent damage
to streams and other drainageways that can result from erosion and sedimentation on a
construction site. Crossing streams with construction vehicles should be avoided, if
possible, but if a temporary stream crossing is required, it should be properly con-
structed to stabilize streambanks and minimize stream damage and sediment loading.
Three options for temporary stream crossings are bridges, culverts, and fords. Bridges
cause the least amount of disturbance to a stream’s banks and habitat. Culvert cross-
ings should be of sufficient size to pass expected peak flows. Fords should be con-
structed from a stabilizing material such as large rocks, and should only be used if
normal flows are shallow and intermittent across a wide channel.
Temporary channel diversions can be used to keep flowing water away from construc-
tion activities, thus significantly reducing sediment movement. Diversion channels
may be constructed as lined channels or with pipes.
Storm sewer inlets that are in operation during construction must be protected to pre-
vent sediment in construction site runoff from entering the conveyance system. The
most common technique is to construct a temporary filter around inlets using straw
bales, filter fabric, or rocks. These filters may require cleaning during the construction
project. As discussed in the previous subsection, a temporary sediment trap around
the inlet is another option.
To protect downstream areas from erosion due to concentrated flows from a construc-
tion zone, the outlets of slope drains, culverts, sediment traps, and sediment basins
must include some type of erosion protection. The design of outlet protection was dis-
cussed in Sections 9.8 and 11.7.
Resources for Stormwater Management

Although the concern about pollution from stormwa- The American Society of Civil Engineers (ASCE) is
ter originated in Europe, much of the current develop- preparing an extensive database containing data from
ment is taking place in the United States, Australia, many research studies that have investigated stormwa-
and New Zealand. In these countries, there are several ter controls throughout the United States (www.bmp-
websites giving guidance on the design of environ- database.org). The ongoing effort of this project
mentally sound drainage systems. Some examples are involves gathering before-and-after water quality
as follows: information and presenting it with simple data analy-
www.rmi.org – This site is maintained by the Rocky sis tools and search engines. As the database expands,
it should become an extremely useful tool for select-
Mountain Institute, an entrepreneurial, non-profit
ing the most appropriate control practice for a specific
organization that fosters the efficient and restorative
use of resources to create a more secure, prosperous, situation. However, a thorough understanding of the
limitations of available controls is needed for proper
and life-sustaining world.
selection; one cannot rely only on test results from
www.metrocouncil.org/environment/Watershed/ other locations, even though those results are indica-
bmp/manual.htm – The Metropolitan Council of tive of what may be useful.
Minneapolis-St. Paul, Minnesota, presents an Urban
Small Sites Best Management Practices Manual. The proceedings of the 1997 Engineering Foundation
Conference in Malmo, Sweden, contains numerous
www.wrc.wa.gov.au/public/WSUD_manual – examples of what could be labeled European Best
Western Australia’s Water and Rivers Commission Management Practices. The proceedings from this
provides a manual on managing urban stormwater conference, entitled Sustaining Urban Water
quality. Resources in the 21st Century (Rowney, Stahre and
Very helpful guidance on the selection of stormwater Roesner, 1999), are available through the American
controls (and many other topics related to stormwater Society of Civil Engineers (ASCE) Bookstore
management) is available from the Center for Water- (www.asce.org/publications).
shed Protection (www.cwp.org). They have recently Many other websites containing stormwater informa-
developed a website called The Stormwater Man- tion are available. The most useful and unbiased
ager’s Resource Center at information is often obtained from academic and gov-
www.stormwatercenter.net. These websites contain ernmental agency sites. The stormwater manuals for
vast amounts of useful information, from past review the U.S. states of Texas and Maryland are two good
articles from Watershed Techniques, to current storm- examples, and these can be found at
water management ordinances. Stormwater character- www.txnpsbook.org, and at www.mde.state.md.us/
ization, effects, sources, and control information is environment/wma/stormwatermanual.
also presented. These sites are therefore good
resources for obtaining a general understanding for The U.S. EPA provides much information concerning
stormwater. The National Menu of Best Management
stormwater issues that are important when developing
Practices for Storm Water Phase II can be found at
a stormwater control program.
cfpub.epa.gov/npdes/stormwater/menuofbmps/
Some professional associations work with urban menu.cfm. Information on recent nonpoint-source
drainage. The International Association for Hydraulic water pollution issues can be found at www.epa.gov/
Research (IAHR) and the International Water owow/nps/urban.html. Links to other sites are
Association (IWA, formerly IAWQ) have a joint included there. The EPA’s Office of Research and
Committee on urban storm drainage, which has been Development maintains a comprehensive on-line
active since 1978. The Committee organizes library of historical and recent peer-reviewed research
conferences on Urban Storm Drainage, with the ninth reports at www.epa.gov/clariton/clhtml/
conference held in 2002. Furthermore, it produces a pubord.html, including many addressing stormwater.
newsletter that can be downloaded at the IAHR In addition, the EPA’s Urban Watershed Management
website (www.iahr.org). Branch sponsors many research projects, and recent
peer-reviewed reports are available at www.epa.gov/
ednnrmrl/newsmain.htm.

Urban stormwater can negatively affect receiving waters by serving as a source of
contaminants and trash and by contributing to erosion. A variety of measures known
as “best management practices” or “BMPs” are available to help mitigate the adverse
impacts of stormwater. Some controls are more appropriate than others for a specific
location and situation, and, because of the wide range of stormwater-related prob-
lems, it is often necessary to use multiple controls at a specific location. Contaminants
that may be present in stormwater include sediments, nutrients, organic matter, toxic
pollutants, and salts.
Small, frequently occurring storms account for a significant portion of the pollutant
load. Consequently, designs for water-quality enhancement, unlike typical designs for
controlling discharge rate, focus on capturing and treating these smaller events. A siz-
ing criterion known as the water quality capture volume or water quality volume may
be used in the design of treatment facilities.
Actions and practices used to reduce runoff flow rates and constituent concentrations
are called best management practices, or BMPs. BMPs are classified as either struc-
tural or nonstructural. Structural BMPs consist of constructed facilities for detaining,
rerouting, or treating stormwater. Examples of structural BMPs include porous pave-
ment, grass buffers and swales, stream buffers and greenways, porous landscape
detention, extended detention basins, wet detention ponds, constructed wetlands, spill
containment facilities, critical source area controls, biofiltration devices, and rain
barrels.
Nonstructural BMPs are institutional and educational practices for producing behav-
ioral changes that result in decreases in pollutant loads. Nonstructural BMPs include
catch basin and street cleaning, preventing inappropriate discharges, and educating
the public about protecting water quality.
To determine which BMPs to use in a particular case, the level of control required
must be established. The more stringent the requirement, the more expensive it will be
to achieve, so cost-effectiveness must be considered. Once the treatment goal has
been determined for a particular site, the appropriate BMPs can be selected. The
selection process is complicated by the fact that the performance characteristics for
many BMPs are poorly understood, and because different measures are more applica-
ble to developing areas, existing areas, or construction sites. A number of websites
provide up-to-date information on stormwater best management practices.
REFERENCES
Atlanta Regional Commission. 2001. Georgia Stormwater Manual Volume 2: Technical Manual. Atlanta,
Georgia: Atlanta Regional Commission.
Circular No. 22, FHWA-SA-96-078. Washington, D.C.: Federal Highway Administration, U.S. Depart-
ment of Transportation.
References 639
Center for Environmental Research Information. 1995. EPA/625/R-95/003. Cincinnati, Ohio: U.S. Envi-
ronmental Protection Agency.
CH2MHill (with W. L. Jorden and R&D Environmental). 1998. East Watershed Impacts Assessment (also
West Watershed Impacts Assessment). Atlanta, Georgia: Metro Atlanta Watersheds Initiative, City of
Atlanta, Department of Public Works.
Chin, D. A. 2000. Water-Resources Engineering. Upper Saddle River, New Jersey: Prentice Hall.
Eastern Federal Lands Highway Design. 1995. Best Management Practices for Erosion and Sediment Con-
trol. Report No. FHWA-FLP-64-005. Washington, D.C.: U.S. Department of Transportation.
Herricks, E. E., ed. 1995. Stormwater Runoff and Receiving Systems: Impact, Monitoring, and Assessment.
Boca Raton, Florida: Engineering Foundation and ASCE, CRC/Lewis.
Lager, J. A., W. G. Smith, W. G. Lynard, R. M. Finn, and E. J. Finnemore. 1997. Urban Stormwater
Management and Technology: Update and Users' Guide. EPA-600/8-77-014. U.S. Cincinnati, Ohio:
Environmental Protection Agency.
Laplace, D., A. Bachoc, Y. Sanchez, and D. Dartus. 1992. “Trunk sewer clogging development—descrip-
tion and solutions.” Water Science and Technology 25, no. 8: 91–100.
Limburg and R. E. Schmidt. 1990. “Patterns of Fish Spawning in Hudson river Tributaries: Response to an
Urban Gradient?” Ecology 71, no. 4.
Maxted, J, and E. Shaver. 1997. The Use of Retention Basins to Mitigate Stormwater Impacts on Aquatic
Life, Effects of Water Development and Management on Aquatic Ecosystems. ed. by L. A. Roesner. New
York: American Society of Civil Engineers.
Metropolitan Washington Council of Governments (Anacostia Restoration Team). 1992. A Current Assess-
ment of Urban Best Management Practices: Techniques for Reducing Nonpoint Source Pollution in the
Coastal Zone. Department of Environmental Programs, Metropolitan Washington Council of Govern-
ments, Water Resources Planning Board.
New York State Department of Environmental Conservation. 2001. Stormwater Management Design Man-
ual.
Pitt, R. 1986. “The incorporation of urban runoff controls in the Wisconsin Priority Watershed Program.”
Pp. 290–313 in Advanced Topics in Urban Runoff Research, ed. by B. Urbonas and L. A. Roesner. New
York: Engineering Foundation and ASCE.
Pitt, R. E., R. I. Field, M. M. Lalor, D. D. Adrian, and D. Barbé. 1993. Investigation of Inappropriate Pollutant
Entries into Storm Drainage Systems: A User's Guide. Rep. No. EPA/600/R-92/238, NTIS Rep. No. PB93-
131472/AS. Cincinnati, Ohio: US EPA, Storm and Combined Sewer Pollution Control Program, Risk
Reduction Engineering Lab.
Pitt, R., R. Field, M. Lalor, and M. Brown. 1995. “Urban stormwater toxic pollutants: assessment, sources
and treatability.” Water Environment Research 67, no. 3: 260–275.
Pitt, R., and J. Voorhees. 1993. Source loading and management model (SLAMM). Seminar Publication,
National Conference on Urban Runoff Management, Enhancing Urban Watershed Management at the
Local, County, and State Levels.
Pitt, R. 1995a. “Biological Effects of Urban Runoff Discharges.” Pp. 127–162 in Stormwater Runoff and
Receiving Systems: Impact, Monitoring, and Assessment, ed. by E. E. Herricks. Boca Raton, Florida:
Engineering Foundation and ASCE, CRC/Lewis.
Pitt, R. 1995b. “Effects of Urban Runoff on Aquatic Biota.” Pp. 609–630 in Handbook of Ecotoxicology,
ed. by D. J. Hoffman, B. A. Rattner, G. A. Burton, Jr. and J. Cairns, Jr. Boca Raton, Florida: Lewis Pub-
lishers/CRC Press.
Pitt, R. 1996. Groundwater Contamination from Stormwater. Chelsea, Michigan: Ann Arbor Press.
Pitt, R., S. Nix, J. Voorhees, S. R. Durrans, and S. Burian. 2000. Guidance Manual for Integrated Wet
Weather Flow (WWF) Collection and Treatment Systems for Newly Urbanized Areas (New WWF Sys-
tems). Second year project report: model integration and use, Wet Weather Research Program. U.S.
Environmental Protection Agency, Cooperative agreement #CX 824933-01-0.
Preul, H. C. 1996. “Combined sewage prevention system (CSCP) for domestic wastewater control.” Pp.
193–198 in Proceedings of the 7th International Conference on Urban Storm Drainage. Hannover, Ger-
many.
Roesner, L. A., B. R. Urbonas, and M. B. Sonnen, eds. 1989. Design of Urban Runoff Quality Controls,
Proceedings of an Engineering Foundation Conference on Current Practice and Design Criteria for
Urban Quality Control. American Society of Civil Engineers.
Urban Drainage and Flood Control District. 1999. Urban Storm Drainage Criteria Manual Vol. 3 – Best
Management Practices. Denver, Colorado: Denver Regional Council of Governments.
Urban Water Resources Research Council (UWRRC) of ASCE and Wright Water Engineers. 2001.
National Best Management Practices (BMP) Database. www.bmpdatabase.org.
U.S. Environmental Protection Agency. 1983. Final Report for the Nationwide Urban Runoff Program,
Water Planning Division. Washington, D.C.
U.S. Environmental Protection Agency. 1993. Handbook on Pollution Prevention and Control Planning.
625/R-93-004.
U.S. Environmental Protection Agency. 1997. Urbanization and Streams: Studies of Hydrologic Impacts.
EPA-841-R-97-009.
U.S. Environmental Protection Agency 1999. Preliminary Data Summary of Best Management Practices.
EPA-821-R-99-012.
U.S. Environmental Protection Agency. 2002. 2000 National Water Quality Inventory Report. EPA-841-R-
02-001.
Water and Rivers Commission. 1998. A Manual for Managing Urban Stormwater Quality in Western Aus-
tralia. Perth.
Water Environment Federation. 1998. Urban Runoff Quality Management. WEF Manual of Practice No.
23. Alexandria, Virginia: Water Environment Federation.
A P P E N D I X
A
Physical Properties of Water
The physical properties of water depend on the ambient temperature and pressure of
the fluid. In the majority of civil and environmental applications, where ambient pres-
sures of interest do not span a large range, it can be safely assumed that the physical
properties of liquid water depend on temperature only. The tables below contain the
physical properties of water for both the U.S. customary and the S.I. systems of units.
Table A.1 Physical properties of water at atmospheric pressure (U.S. customary units)
Dynamic
Temperature Density Specific Weight Kinematic Viscosity Vapor Pressure Bulk Modulus
(°F) (slug/ft3) (lb/ft3) Viscosity (ft2/s) (lb-s/ft2) (psia) (psi)
40 1.940 62.43 1.664 x 10–5 3.229 x 10–5 0.12 2.94 x 105
50 1.940 62.41 1.410 x 10–5 2.735 x 10–5 0.18 3.05 x 105
60 1.938 62.37 1.217 x 10–5 2.359 x 10–5 0.26 3.11 x 105
70 1.936 62.30 1.059 x 10–5 2.050 x 10–5 0.36 3.20 x 105
80 1.934 62.22 0.930 x 10–5 1.799 x 10–5 0.51 3.22 x 105
90 1.931 62.11 0.826 x 10–5 1.595 x 10–5 0.70 3.23 x 105
–5 –5
100 1.927 62.00 0.739 x 10 1.424 x 10 0.96 3.27 x 105
110 1.923 61.86 0.667 x 10–5 1.284 x 10–5 1.28 3.31 x 105
120 1.918 61.71 0.609 x 10–5 1.168 x 10–5 1.69 3.33 x 105
130 1.913 61.55 0.558 x 10–5 1.069 x 10–5 2.22 3.34 x 105
642 Physical Properties of Water Appendix A
Table A.2 Physical properties of water at atmospheric pressure (S.I. units)

Dynamic
Temperature Density Specific Weight Kinematic Viscosity Vapor Pressure Bulk Modulus
3 3
(°C) (kg/m ) (N/m ) Viscosity (m2/s) (N-s/ft2) (kN/m2) (N/m2)
5 1000.0 9806 1.519 x 10–6 1.519 x 10–3 0.87 2.06 x 109
–6 –3
10 999.7 9802 1.308 x 10 1.308 x 10 1.23 2.11 x 109
15 999.1 9797 1.141 x 10–6 1.140 x 10–3 1.70 2.14 x 109
20 998.2 9786 1.007 x 10–6 1.005 x 10–3 2.34 2.20 x 109
25 997.1 9777 0.897 x 10–6 0.894 x 10–3 3.17 2.22 x 109
30 995.7 9762 0.804 x 10–6 0.801 x 10–3 4.24 2.23 x 109
35 994.1 9747 0.727 x 10–6 0.723 x 10–3 5.61 2.24 x 109
–6 –3
40 992.2 9730 0.661 x 10 0.656 x 10 7.38 2.27 x 109
45 990.2 9711 0.605 x 10–6 0.599 x 10–3 9.55 2.29 x 109
50 988.1 9689 0.556 x 10–6 0.549 x 10–3 12.33 2.30 x 109
A P P E N D I X
B
Unit Conversions
The physical units used in quantitative hydrologic and hydraulic analyses and designs
have the fundamental dimensions of mass (M), length (L), and time (T). Some quanti-
ties have these fundamental dimensions, whereas other quantities have derived units
made up of various combinations of the fundamental dimensions. Many of the derived
units have their own names, such as miles, acres, Joules, and Watts.
Table B.1, below, contains factors to convert from the U.S. customary system of units
to the S.I. system of units. Table B.2 contains factors to convert from one type of unit
to another within a given system of units. To convert a quantity having dimensions
shown in the left-hand columns of the tables to dimensions shown in the right-hand
columns, multiply by the conversion factors in the center columns.
Table B.1 Conversion factors from U.S. customary units to S.I. units
Parameter U.S. Customary Unit Conversion S.I. Unit
ft 0.3048 m
in. 25.4 mm
Length
mi (statute) 1609.3 m
mi (nautical) 1852 m
2
ft 0.0929 m2
Area
in2 645.2 mm2
ft3 0.02832 m3
Volume 3
in 16,387 mm3
ft/s 0.3048 m/s
Velocity mi/hr 0.44704 m/s
knot (naut. mi/hr) 0.51444 m/s
Acceleration ft/s2 0.3048 m/s2
Mass slug 14.594 kg
Density slug/ft3 515.38 kg/m3
644 Unit Conversions Appendix B
Table B.1 Conversion factors from U.S. customary units to S.I. units
Parameter U.S. Customary Unit Conversion S.I. Unit
Weight lb 4.4482 N
Specific Weight lb/ft3 157.09 N/m3
lb/ft2 47.88 N/m2
Pressure
psi 6894.8 N/m2
ft-lb 1.3558 J (N-m)
Energy, Work
Btu 1055.1 J (N-m)
ft-lb/s 1.3558 W (N-m/s)
Power
hp 745.69 W (N-m/s)
2
Dyn. Viscosity lb-s/ft 47.88 N-s/m2
Kin. Viscosity ft2/s 0.0929 m2/s
Table B.2 Conversion factors for units within unit systems

Parameter Unit Conversion Unit
ft 12 in
Length
mi (statute) 5280 ft
ac 43,560 ft2
2
Area mi 640 ac
ha 10,000 m2
Volume ac-ft 43,560 ft3
A P P E N D I X
C
Unsteady Stormwater Flow
Although most collection system design is based on steady state hydraulics, flow in
sewers is inherently unsteady due to the nature of precipitation and snow melt events.
This is further complicated by pump cycling and changing water levels in detention
structures. Routing is usually not important in small systems, such as culverts, or in
steep systems where the hydrograph is essentially translated downstream.
Changes in flow do not propagate instantaneously through a channel or sewer system,
but rather move through the system gradually. Consider what occurs in a pipe or as
the flow increases at the upstream end. The increase in flow manifests itself as an
increase in water level and velocity, which in turn moves downstream. In addition, the
flow regime may change. As a result of a storm event, flow in a storm sewer or open
channel begins empty or partially full, changes to a full flow condition, and may even
become pressurized as the tailwater rises at the downstream end or a restriction down-
stream limits the flow.
C.1 BASICS OF UNSTEADY FLOW ANALYSIS

The process of analyzing an unsteady flow event through a sewer system is referred to
as routing the flow. The National Engineering Handbook (Mockus and Styner, 1972)
defines routing as “computing the flood at a downstream point from the flood at an
upstream point, taking storage into account.” As a flow disturbance moves down-
stream, the peak of the disturbance (the hydrograph) tends to flatten. For instance, in
the case of an increasing flow, some of the flow goes toward filling the storage vol-
ume corresponding to the increase in flow depth in the channel. This process is known
as attenuation.
For steady, uniform flow, a unique rating curve relates depth and flow rate. However,
for unsteady flow, the depth deviates from normal depth because the slope of the
water surface is greater than that of the channel slope when the flow is increasing, and
is flatter than that of the channel slope during decreasing flow. The depth-flow rela-
tionship in unsteady flow situations is a looped rating curve.
646 Unsteady Stormwater Flow Appendix C
A variety of methods exist for evaluating unsteady flow in conveyance systems, rang-
ing from the most sophisticated numerical solutions of the Saint-Venant equations to
simpler approximation methods. Routing methods are grouped into two general cate-
gories: hydraulic and hydrologic. Hydraulic models analyze both flow and hydraulic
grade through the system. Hydrologic models analyze flow only and use a simplified
approximation to account for momentum effects. Hydrologic models are also referred
to as lumped models because they calculate flow at one location-the downstream end
of a reach. Hydraulic routing models are referred to as distributed models because
they calculate flow and depth at several cross sections simultaneously (Fread, 1993).
Attenuation of flows is due to temporary storage in a pipe or channel reach. Hydraulic
methods explicitly consider the depth of water in determining the volume stored, but
the calculations can be quite complex. In many cases, storage in a reach can be
approximated using simplified storage equations. Hydrologic methods assume that
knowing the inflow into the reach at the current and previous time steps and the out-
flow at the previous time step is sufficient to determine outflow at the current time
step. Once the inflow and outflow are known, flow depth, if needed, can be deter-
mined using gradually varied flow evaluation methods, or even normal depth calcula-
tions in some cases.
Hydrologic routing methods rely on the storage equation, which is given by
dS
I −Q = (3.1)
dt
where S = storage (ft3, m3)

I = inflow (cfs, m3/s)
Q = outflow (cfs, m3/s)
t = time (s)
Various methods for calculating the outflow from a channel or pipe reach based on the
inflow have been developed. Such methods are generally accurate, provided that
downstream conditions do not cause flow to back up in the reach. In other words, tail-
water depth and flow restrictions should not significantly reduce the downstream flow
rate if a hydrologic method is to be applied. If a downstream restriction does exist,
outflow hydrographs predicted by hydrologic routing methods tend to have higher
peak flows that arrive at the outlet earlier than those predicted by hydraulic models.
This difference is due to the fact that hydrologic models do not adequately account for
all factors leading to attenuation in such cases.
The hydraulic routing methods, on the other hand, are based on partial differential
unsteady flow equations that allow the flow rate and depth to be computed as func-
tions of both space and time, rather than time only as with the lumped hydrologic
methods.
Section C.2 Hydrodynamic Equations 647
C.2 HYDRODYNAMIC EQUATIONS

The most commonly used equations for modeling unsteady flow in sewers are based
on the assumption that the flow is one-dimensional, the pressure distribution is hydro-
static, the length of the channel is much greater than the flow depth, and the water
density is constant. Disturbances in a fluid under such conditions are referred to as
shallow-water or translatory waves (Linsley, Kohler, and Paulhus, 1982). The equa-
tions used to describe this type of flow are attributed to Jean-Claude Barre de St.
Venant. Because of the complexity of these equations, numerous approximations have
been developed, some of which are described in Section C.3.
The St. Venant equations consist of two partial differential equations, the first describ-
ing continuity and the second describing momentum. These equations can take a vari-
ety of forms. For one-dimensional, unsteady flow in a nonprismatic channel or
conduit, the equations can be written as (Jin, Coran, and Cook, 2002):
∂Q ∂ ( A + Ao )
+ −q = 0 (3.2)
∂x ∂t
∂Q ∂ ( β Q 2 / A)  ∂y 
+ + gA  − So + S f + Si  + L = 0 (3.3)
∂t ∂x  ∂x 

x = distance along the longitudinal axis of the channel (ft, m)
A = active cross-sectional area of flow (ft2, m2)
Ao = inactive (off-channel storage) cross-sectional area of flow (ft, m)
t = time (s)
q = lateral inflow or outflow (ft2/s, m2/s)
β = coefficient for nonuniform velocity distribution in cross-section
y = flow depth (m, ft)
So = slope of channel bed in longitudinal direction
Sf = friction slope
Si = slope due to severe local expansion effects
L = momentum effect of lateral flow (ft3/s2, m3/s2)
To solve these equations, initial conditions and boundary conditions are required. The
boundary conditions usually consist of the inflow hydrograph, the tailwater condition
at the outlet. The gradually varied flow profile or normal depth based on some prede-
termined flow is used at the initial condition. Because the St. Venant equations cannot
be solved analytically, a variety of numerical solutions have been developed over the
years. Papers on available numerical solutions include those by Abbott (1979); Chow
(1973); Cunge, Holley and Verwey (1980); Fread (1993); French (1985); NOAA
(2000); Ponce, Li and Simons (1978); Price (1973); Roesner, Aldrich and Dickinson
(1989); and Yen (1996, 2001).
C.3 APPROXIMATIONS OF THE HYDRODYNAMIC

EQUATIONS
Because of the difficulty in solving the St. Venant equations, numerous approxima-
tions have been developed, which usually involve eliminating terms from the momen-
tum equation. It is possible to rearrange and simplify the St. Venant momentum
equation to express discharge as a function of the discharge under normal flow, as
shown in Equation C.4 (Weinmann and Laurenson, 1979; French, 1985).
1/ 2
 1 ∂y u ∂u 1 ∂u 
Q = QN 1 − − −  (3.4)
 S x ∂x So g ∂x So g ∂t 
where Q = actual unsteady flow (cfs, m3/s)

QN = flow under normal conditions (cfs, m3/s)
u = velocity in the longitudinal direction (ft/s, m/s)
When all of the terms in the parentheses are considered, then Equation C.4 corre-
sponds to the St. Venant equations. When the last term in the parentheses is neglected,
then the models are referred to as a diffusion analogy. When only the first term is con-
sidered, the model determines normal depth only and is called kinematic wave rout-
ing. With each successive level of approximation, the equations become easier to
solve but lose some of their ability to represent real phenomena.
Kinematic Wave Routing

If the conditions during the unsteady flow event will not deviate significantly from
normal flow, then most terms on the right side of Equation C.4 can be eliminated to
yield the kinematic wave solution:
∂Q ∂Q
+c =0 (3.5)
∂t ∂x
where c = kinematic wave celerity (velocity), dQ/dA (ft/s, m/s)

Kinematic wave routing works well when flow changes are slow and the stage-dis-
charge relationship is not significantly looped.
Muskingum Routing
A commonly used form of hydrologic routing is the Muskingum method, which
assumes that the storage in the reach can be given by
S = K[XI + (1 - X)Q] (3.6)
where K = storage constant usually taken as the travel time through the reach (s)
X = relative importance of inflow and outflow in determining storage
This method is presented in more detail in Section 5.8 (page 170).
Section C.3 Approximations of the Hydrodynamic Equations 649
Muskingum-Cunge Routing
A variation in the Muskingum method proposed by Cunge (1969) allows the routing
coefficients to change based on changes in flow and the top width of the water sur-
face, and is referred to as the Muskingum-Cunge method. This method combines the
desirable features of the kinematic wave method with the benefits of the Muskingum
method in that it allows for a limited amount of attenuation. These routing methods
are discussed in Ponce, Li, and Simons (1978), U.S. Army Corps of Engineers (1998)
and Yen (2001).
Because the Muskingum-Cunge method changes the kinematic-type Muskingum
method to a method based on the diffusion analogy. Capable of predicting hydrograph
attenuation, the Muskingum-Cunge method can be effectively used as a distributed
routing technique. This method uses the following equations for K and X:
∆x
K= (3.7)
c
1 Q 
X = 1 −  (3.8)
2  BcS0 ∆x 
where B = width of water surface (ft, m)

The depth can be determined by backwater calculations. The Muskingum-Cunge
method is not suitable for fast-rising hydrographs and for flows where backwater
effects exist.
Convex Routing
Another variation in hydrologic routing is convex routing, which uses only flows from
the previous time step to determine outflow from a reach (Mockus and Styner, 1972).
The flow equation can be written as
Q2 = cI1 + (1 − c)Q1 (3.9)
where c = convex routing coefficient

Note that the convex routing coefficient c is not the same as the wave speed c
described previously. Rather, it is the ratio of the time step size to the travel time in the
reach. The time step size should be less than the travel time (i.e., c < 1). The travel
time, however, is not constant, but is a function of velocity. It is usually determined
for a velocity corresponding to 50 to 75 percent of maximum flow. Therefore, c is
defined by
V ∆t
c= (3.10)
L
where V = representative velocity (ft/s, m/s)

∆t = time step size (s)
L = length of pipe or channel reach (ft, m)
Weighted Translation Routing

Generally, the travel time should be less than the time step size; however, in some
cases, very short pipes would force time step sizes to be unrealistically small. There-
fore, it is necessary to have a method that works when c > 1. The method presented
below (Haestad Methods, 2001) is useful in such situations.
1  1
Q2 = I1 + 1 −  I 2 (3.11)
c  c
Translation Routing
In small systems with steep slopes, disturbances essentially move through the pipe
with little to no attenuation. To determine the flow at a given time, it is only necessary
to look upstream to the flow one time step earlier at position V∆t. This value will most
likely not be at a location between points where the flow is known. Therefore, inter-
polation will be needed to determine the flow to be passed downstream.
C.4 ROUTING METHOD COMPLICATIONS

Although the preceding equations can account for unsteady flow in channels and
pipes, real conveyance systems can include phenomena not accounted for in these
flow routing methods. Some of these phenomena include:
• Manholes, inlet boxes, and other junction structures
• Overflows at inlets and manholes
• Regulators and diversions
• Parallel channels and pipes
• Impoundments, especially with multiple outlets
• Flow reversal
• Dry pipes
• Drop structures
• Tidally influenced outlets
Each of these complications requires an adjustment to the flow routing. The adjust-
ments depend on the routing method being used.
REFERENCES
Pitman.
Chow, V. T. 1973. Open Channel Hydraulics. New York: McGraw Hill.
Cunge, J. A. 1969. “On the subject of a flood propagation computation method (Muskingum method).”
London: Pitman.
Fread, D. L. 1993. “Flow Routing.” In Maidment, D. R. ed. Handbook of Hydrology. New York: McGraw-
Hill.
Haestad Methods. 2001. SewerCAD: Sanitary Sewer Modeling Software. Waterbury, Connecticut: Haestad
Methods.
Jin, M., S. Coran, and J. Cook. 2002. “New One-Dimensional Implicit Numerical Dynamic Sewer and
Storm Model.” Proceedings of the 9th International Conference on Urban Drainage. American Society
of Civil Engineers (ASCE).
Linsley, R. K., Jr.; M. A. Kohler; and J. L. H. Paulhus. 1982. Hydrology for Engineers. 3rd ed. New York:
McGraw-Hill.
Mockus, V. and W. Steiner. 1972. “Flood Routing.” National Engineering Handbook, Section 4 Hydrology,
Chapter 17. Washington, D.C.: Soil Conservation Service, U.S. Department of Agriculture.
National Oceanic and Atmospheric Administration, National Weather Service. 2000. FLDWAV Computer
Program, Version 2-0-0.
Ponce, V. M., R. M. Li, and D. B. Simons. 1978. “Applicability of Kinematic and Diffusion Models.” Jour-
nal of Hydraulic Division, ASCE 104, HY12: 1663.
Price, R. K. 1973. “A comparison of four numerical methods for flood routing.” Journal of the Hydraulics
Division, ASCE 100, no. 7: 879–899.
Roesner, L. A., J. A. Aldrich, and R. E. Dickinson. 1989. “Storm Water Management Model User’s Manual
Version 4: Extra Addendum.” EPA 600/3-88/001b.
U.S. Army Corps of Engineers (USACE). 1998. HEC-1 Flood Hydrograph Package User’s Manual. Davis,
Ca.: USACE Hydrologic Engineering Center.
Weinmann, D. E. and E. M. Laurenson. 1979. “Approximate Flood Routing methods: A Review.” Journal
of Hydraulics Division, ASCE. 105, HY12: 1521.
Yen, B. C. 1996. “Hydraulics for Excess Water Management.” In L. W. Mays, ed. Water Resources Hand-
book. New York: McGraw-Hill.
Yen, B. C. 2001. “Hydraulics of Storm Sewer Systems.” In L. Mays, ed. Stormwater Collection System
Design Handbook. New York: McGraw Hill.
A P P E N D I X
D
References for Common
Stormwater Conveyance System
Component Models and Software
Component Modeling Methods Typical Input data Section of Text Software

Channel with uni- Manning equation, Chézy Channel roughness 7.3 FlowMaster
form flow equation, or Darcy-Weisbach Channel geometry
equation Channel slope
Channel with gradu- Direct-step method or Geometry of each cross-section 7.8 StormCAD
ally varied flow standard-step method Starting water surface elevation
Channel with flow Muskingum Historical flood data or routing 5.8 PondPack
routing coefficients
Inflow hydrograph
Modified Puls Geometry of outlet 5.8 PondPack
Inflow hydrograph
Culvert with inlet HDS-5 methodology Inlet geometry 9.5 CulvertMaster
control Discharge
Culvert with outlet HDS-5 methodology Pipe geometry 9.5 CulvertMaster
control Pipe material
Inlet geometry
Discharge
Tailwater elevation
Roadway overtop- HDS-5 methodology Roadway profile coordinates 9.6 CulvertMaster
ping Type of road
Roadway width
Discharge
Gutter flow Manning equation or Chézy Roadway slopes 10.1 FlowMaster,
equation Curb height StormCAD
Pavement roughness
Discharge
654 References for Common Stormwater Conveyance System Component Models and Software
Component Modeling Methods Typical Input data Section of Text Software

Storm sewer pipes Manning equation or Chézy Pipe roughness 7.3; 7.4; 7.5; 7.8 StormCAD
with open channel equation; Saint-Venant Pipe geometry 11.5; Appendix C
flow equations Pipe slope
Storm sewer pipes Darcy-Weisbach, Pipe roughness 6.2; 6.3; 11.5; StormCAD
with pressure flow Manning, Hazen-Williams, or Pipe geometry Appendix C
Chézy equation; Saint-Venant Pipe slope
equations Discharge or velocity
Tailwater conditions
Inlets Weir and orifice equations Discharge 10.2 FlowMaster,
Inlet type and geometry StormCAD
Manholes Minor loss equations Entrance and exit velocities 6.3; 11.4 StormCAD
Structure geometry
Diversions/over- Weir and/or orifice equations Geometry 7.6; 7.7 StormCAD
flows Water surface elevations
System outlets Weir equations Boundary conditions such as tail- 6.3; 9.5; 11.7 FlowMaster,
Orifice equations water data or downstream chan- StormCAD
Minor loss equations nel characteristics
Energy dissipation model Geometric characteristics of outlet
Detention ponds Hydrologic routing and Inflow hydrograph 12.6; 12.7 PondPack
Hydraulic routing Depth vs. volume for pond
Geometry of outlet or stage-
discharge curve
Bibliography
Pitman.
Akan, A. O. 1989. “Detention Pond Sizing for Multiple Return Periods.” Journal of Hydraulic Engineering
115 no. 5: 650.
Manual, 2000 Metric Edition. Washington, D.C.: AASHTO.
American Concrete Pipe Association (ACPA). 2000. Concrete Pipe Design Manual, 13th Printing
(revised). Irving, Texas: ACPA.
American Iron and Steel Institute (AISI). 1980. Modern Sewer Design. Washington, D.C.: AISI.
American Iron and Steel Institute (AISI). 1999. Modern Sewer Design. 4th ed. Washington, D.C.: AISI.
American Society of Civil Engineers (ASCE). 1977. Sedimentation Engineering. ASCE Manuals and
Reports on Engineering Practice, No. 54. New York: ASCE.
American Society of Civil Engineers (ASCE). 1996. Hydrology Handbook. 2d ed.,ASCE Manual of Prac-
tice No. 28. New York: ASCE.
Armortec. 2003. Product literature for Armorflex concrete mats. (www.armortec.com).
Anderson E.A. 1973. “National Weather Service River Forecast System – Snow Accumulation and Abla-
tion Model.” Report NWS 17. Washington, D.C.: U.S. Department of Commerce.
Anderson E.A. 1976. “A Point Energy and Mass Balance Model of Snow Cover.” Report NWS 19. Wash-
ington, D.C.: U.S. Department of Commerce.
Ang, A. H.-S., and W.H. Tang. 1975 and 1984. Probability Concepts in Engineering Planning and Design,
Vol. I, Basic Principles, and Vol. II, Decision, Risk, and Reliability. New York: John Wiley.
Aron, G. and D. F. Kibler. 1990. “Pond Sizing for Rational Formula Hydrographs.” Water Resources Bulle-
tin 26 no. 2: 255.
ASCE Task Committee. 1985. Detention Outlet Control Structures. Reston, Virginia: ASCE.
ASCE. 1982. Stormwater Detention Facilities. Reston, Virginia: ASCE.
ASCE. 2000. Joint Conference on Water Resources Engineering and Water Resources Planning and Man-
agement. Minneapolis, Minnesota: ASCE.
Baumgardner, R. H. 1981. “Estimating Required Storage to Reduce Peak Pumping Rates at Storm Water
Pumping Stations.” Washington, D.C.: Federal Highway Administration Workshop Paper, Office of
Engineering.
656 Bibliography
Bazin, H. 1897. “A New Formula for the Calculation of Discharge in Open Channels.” Mémoire No. 41,
Annales des Ponts et Chausées. Paris 14, Ser. 7, 4th trimester: 20–70.
Bedient, P.B., and W.C. Huber. 1992. Hydrology and Floodplain Analysis. 2nd ed. Reading, Massachusetts:
Addison Wesley.
Blake, M.E. 1947. Ancient Roman Construction in Italy from the Prehistoric Period to Augustus. Washing-
ton, D.C.: Carnegie Institution of Washington.
Blevins, R.D. 1984. Applied Fluid Dynamics Handbook. New York: Van Nostrand Reinhold.
Bobée, B. 1975. “The Log Pearson Type 3 Distribution and Its Applications in Hydrology,” Water
Resources Research 14, no. 2: 365–369.
Bobée, B., and F. Ashkar. 1991. The Gamma Distribution and Derived Distributions Applied in Hydrology.
Littleton, Colorado: Water Resources Publications.
Brice, J. C. 1981. Stability of Relocated Stream Channels. Report No. FHWA/RD-80/158. U.S. Geological
Survey.
Brookes, A. and F. D. Shields. 1996. River Channel Restoration: Guiding Principles for Sustainable
Projects. New York: John Wiley & Sons.
Burchett, C. 1998. “Open Channel Linings: Choosing the Right Erosion Control Products.” Burns &
McDonnell Tech Briefs 1998, no. 3. Kansas City, Missouri: Burns & McDonnell (www.burnsmcd.com).
Burian, S. J., S. J. Nix, S. R. Durrans, R. E. Pitt, C. Y. Fan, and R. Field. 1999. “Historical Development of
Wet-Weather Flow Management,” Journal of Water Resources Planning and Management 125, no. 1:
3–13.
Carp, E. ed. 1972. Proceedings, International Conference on the Conservation of Wetlands and Waterfowl.
Ramsar, Iran, 30 January–3 February 1971. Slimbridge, UK: International Wildfowl Research Bureau.
Center for Environmental Research Information, U.S. Environmental Protection Agency, EPA/625/R-95/
003, Cincinnati, Ohio, pp. 225–243, April, 1995.
CH2MHill (with W.L. Jorden and R&D Environmental). September, 1998. East Watershed Impacts Assess-
ment (also West Watershed Impacts Assessment). Prepared for the Metro Atlanta Watersheds Initiative,
City of Atlanta. Atlanta, Georgia: Department of Public Works.
Chang, F. M., and J. M. Normann. September, 1976. Design Considerations and Calculations for Fishways
Through Box Culverts. Unpublished Report. Washington, D.C.: Federal Highway Administration, U.S.
Department of Transportation.
Chang, F. M., R. T. Kilgore, D. C. Woo, and M. P. Mistichelli. 1994. Energy Losses through Junction Man-
holes, Vol. 1, Research Report and Design Guide. McLean, Virginia: Federal Highway Administration.
Chapra, S. C. and R. P. Canale. 1988. Numerical Methods for Engineers, 2d ed. New York: McGraw-Hill.
Chapra, S. C. 1997. Surface Water-Quality Modeling. New York: McGraw-Hill Book Co.
Chaudhry, M. H. and S. M. Bhallamudi. 1988. “Computation of Critical Depth in Compound Channels.”
Chaudhry, M.H. 1993. Open-Channel Flow. Englewood Cliffs, New Jersey: Prentice Hall.
Chesbrough, C. S. 1858. “Chicago Sewerage Report of the Results of Examinations Made in Relation to
Sewerage in Several European Cities in the Winter of 1856–57.” Chicago, Illinois.
Chen, Y. H. and B. A. Cotton. 2000. “Design of Roadside Channels with Flexible Linings.: Hydraulic Engi-
neering Circular no. 15 (HEC-15). McLean, Virginia: Federal Highway Administration (FHWA).
Chin, D.A. 2000. Water-Resources Engineering. Upper Saddle River, New Jersey: Prentice Hall.
Chow, V. T. 1959. Open-Channel Hydraulics. New York: McGraw Hill.
657
Chow, V.T., D.L. Maidment, and L.W. Mays. 1988. Applied Hydrology, New York: McGraw-Hill.
Coldewey, W. G., Göbel, P., Geiger, W. F., Dierkes, C. & Kories, H. 2001. “Effects of stormwater infiltra-
tion on the water balance of a city.” In New Approaches Characterizing Groundwater Flow, Proceedings
of the XXXI. International Association of Hydrogeologists Congress, ed K. P. Seiler and S. Wohnlich.
Balkema, Lisse.
Colebrook, C. F. 1939. “Turbulent Flow in Pipes with Particular Reference to the Transition Between
Smooth and Rough Pipe Laws.” Journal of the Institution of Civil Engineers of London 11.
Copeland, C. 1999. “Clean Water Act: A Summary of the Law.” CRS Issue Brief for Congress, RL 30030.
Copeland, R. R., D. N. McComas, C. R. Thorne, P. J. Soar, M. M. Jonas, and J. B. Fripp. 2001. Hydraulic
Design of Stream Restoration Projects. Technical Report ERDC/CHL TR-01-28. Vicksburg, Missis-
sippi: U.S. Army Corps of Engineers, Engineer Research and Development Center.
Corry, M. L, P. L. Thompson, F. J. Watts, J. S. Jones, and D. L. Richards. 1983. Hydraulic Design of Energy
Dissipators for Culverts and Channels. Hydraulic Engineering Circular No. 14. Washington, D.C.: Fed-
eral Highway Administration, U.S. Department of Transportation.
Cunge, J. A. 1969. “On the subject of a flood propagation computation method (Muskingum method).”
London: Pitman.
Cunnane, C. 1978. “Unbiased Plotting Positions – A Review.” Journal of Hydrology 37: 205–222.
David, H. A. 1981. Order Statistics. 2d ed. New York: John Wiley.
Doyle, E. 2001. “Wastewater Collection and Treatment – Issue Paper 9.” Toronto, Ontario: Research Paper
Advisory Panel, The Walkerton Inquire, Delcan Corporation.
Durrans, S. R. and P. A. Brown. 2001. “Estimation and Internet-Based Dissemination of Extreme Rainfall
Information” Transportation Research Record no. 1743: 41–48.
Durrans, S. R., T. B. M. J. Ouarda, P. F. Rasmussen, and B. Bobée. 1999. “Treatment of Zeroes in Tail
Modeling of Low Flows.” Journal of Hydrologic Engineering, ASCE 4, no. 1: 19–27.
Durrrans, S.R. 1996. “Low-Flow Analysis with a Conditional Weibull Tail Model.” Water Resources
Research 32, no. 6: 1749–1760.
Dyhouse, Gary, J. A. Benn, David Ford Consulting, J. Hatchett, and H. Rhee. 2003. Floodplain Modeling
Einstein, H. A. 1934. “The Hydraulic or Cross Section Radius.” Schweizerische Bauzeitung 103, no. 8: 89–
91.
Einstein, H. A. and R. B. Banks. 1951. “Fluid Resistance of Composite Roughness.” Transactions of the
Engineers Joint Contract Documents Committee (EJCDC). 2002. Standard General Conditions of the Con-
struction Contract. Reston, Virginia: ASCE.
Environment Australia. 2003. Urban Stormwater Initiative.
Environment Canada. 1993. Flooding: Canada’s Water Book. Ottawa, Canada: Ecosystem Sciences and
Evaluation Directorate, Economics and Conservation Branch.
European Commission. 1998. Towards Sustainable Water Resources Management – A Strategic Approach.
European Conference of Sustainable Cites & Towns (ECSCT). 1994. Charter of European Cites and Towns
Towards Sustainablility (Aalborg Charter). Aalborg, Denmark.
Federal Aviation Agency (FAA). 1970. Advisory Circular on Airport Drainage. Report A/C 150-5320-58.
Washington, D.C.: U.S. Department of Transportation.
658 Bibliography
Federal Highway Administration (FHWA). 1977. Bicycle Safe Grate Inlets Study: Vol. 1 and 2 – Hydraulic
and Safety Characteristics of Selected Grate Inlets on Continuous Grade. FHWA-RD-77-44. Washing-
ton, D.C.: U.S. Department of Transportation.
Federal Highway Administration (FHWA). 1982. Highway Storm Water Pumping Stations, Vol. 1 and Vol.
2, FHWA-IP-82-17. Washington, D.C.: U.S. Department of Transportation.
Federal Highway Administration (FHWA). 1985. Hydraulic Design Series No. 5 (HDS-5). Washington,
D.C.: U.S. Department of Transportation.
Federal Interagency Stream Corridor Restoration Working Group. 2001. Stream Corridor Restoration:
Principles, Processes and Practices. GPO 0120-A. http://www.usda.gov/stream_restoration/.
Field, R., M. Borst, M. Stinson, C-Y. Fan, J. Perdek, D. Sullivan, and T. O'Connor. 1996. Risk Management
Research Plan for Wet Weather Flows. Edison, New Jersey: U.S. Environmental Protection Agency.
Field, R., M. P. Brown, and W. V. Vilkelis. 1994. Storm Water Pollution Abatement Technologies. U.S.
Environmental Protection Agency, Office of Research and Development. Cincinnati, Ohio: EPA/600/R-
94/129.
Fortier, S., and F. C. Scobey. 1926. “Permissible Canal Velocities.” Transactions, American Society of Civil
Engineers 89: 940–956.
Foundation for Water Research. 1994. Urban Pollution Management Manual. Marlow, Great Britain: Foun-
dation for Water Research.
Foundation for Water Research. 1998. Urban Pollution Management Manual. 2d ed. Marlow, Great Britain:
Foundation for Water Research.
Finnemore, E. J. and J. Franzini. 2000. Fluid Mechanics with Engineering Applications. New York:
McGraw-Hill.
Franzini, J. and E. J. Finnemore. 1997. Fluid Mechanics. 9th ed. New York: McGraw-Hill.
Fread, D. L. 1993. “Flow Routing.” In Maidment, D. R. ed. Handbook of Hydrology. New York: McGraw-
Hill.
Frederick, R. H., V. A. Myers, and E. P. Auciello. 1977. Five- to 60-Minute Precipitation Frequency for the
Eastern and Central United States. NOAA Technical Memorandum NWS HYDRO-35. Silver Spring,
Maryland: National Weather Service, Office of Hydrology.
Ganguillet, E., and W. R. Kutter. 1869. “An Investigation to Establish a New General Formula for Uniform
Flow of Water in Canals and Rivers” (in German), Zeitschrift des Oesterreichischen Ingenieur und
Architekten Vereines 21, no. 1: 6–25; no. 2: 46–59. Translated into English as a book by R. Hering and J.
C. Trautwine, Jr. 1888 A General Formula for the Uniform Flow of Water in Rivers and Other Channels.
New York: John Wiley and Sons.
Glazner, Michael K. 1998. “Pond Routing Techniques: Standard vs. Interconnected.” In A. M. Strafaci, ed.
Engineering Hydraulics and Hydrology. Waterbury, Connecticut: Haestad Methods.
Goodman, A. S. 1984. Principles of Water Resources Planning. Englewood Cliffs, NJ: Prentice Hall.
Gray, D. M. and T. D. Prowse. 1993. “Snow and Floating Ice.” In R. Maidment, ed. Handbook of Hydrol-
ogy. D New York: McGraw-Hill, Inc.
Green, W. H., and G. Ampt. 1911. “Studies of Soil Physics, Part I: The Flow of Air and Water Through
Soils.” Journal of Agricultural Science 4: 1–24.
Greenwood, J. A., J. M. Landwehr, N. C. Matalas, and J. R. Wallis. 1979. “Probability Weighted Moments:
Definition and Relation to Parameters of Several Distributions Expressible in Inverse Form.” Water
Resources Research 15, no. 6: 1049–1054.
Grigg, N. S. 1996. Water Resources Management: Principles, Regulations, and Cases. New York:
McGraw-Hill.
659
Gringorten, I. I. 1963. “A Plotting Rule for Extreme Probability Paper.” Journal of Geophysical Research
68, no. 3: 813–814.
Gumbel, E. J. 1958. Statistics of Extremes. New York: Columbia University Press.
Guo, J. C. Y. 1997. Street Hydraulics and Inlet Sizing Using the Computer Model UDINLET. Highlands
Ranch, Colorado: Water Resources Publications.
Guo, J. C. Y. 2001. “Design of Off-Line Detention Systems.” In L. W. Mays, ed. Stormwater Collection
Systems Design Handbook. New York: McGraw-Hill.
Haan, C. T. 1977. Statistical Methods in Hydrology. Ames, Iowa: Iowa State University Press.
Haestad Methods. 2001. SewerCAD: Sanitary Sewer Modeling Software. Waterbury, Connecticut: Haestad
Methods.
Haestad Methods. 2003a. PondPack: Detention Pond and Watershed Modeling Software. Waterbury, Con-
necticut: Haestad Methods.
Haestad Methods. 2003b. StormCAD: Storm Sewer Design and Analysis Software. Waterbury, Connecticut:
Haestad Methods.
Harrison, L. J., J. L. Morris, J. M. Normann, and F. L. Johnson. 1972. Hydraulic Design of Improved Inlets
for Culverts. Hydraulic Engineering Circular No. 13. Washington, D.C.: Federal Highway Administra-
tion, U.S. Department of Transportation.
Hathaway, G. A. 1945. “Design of Drainage Facilities.” Transactions, American Society of Civil Engineers
110: 697-730.
Hayes, D. F., ed. 2001. “Wetlands Engineering and River Restoration.” ASCE Conference Proceeding.
Reno, Nevada: ASCE.
Helvey, J. D and J. H. Patric. 1965. “Canopy and Litter Interception of Rainfall by Hardwoods of the East-
ern United States.” Water Resources Research 1: 193-206.
Helvey, J. D. 1967. “Interception by Eastern White Pine.” Water Resources Research 3: 723–730.
Henderson, F. M. and R. A. Wooding. 1964. “Overland Flow and Groundwater Flow from a Steady Rain of
Finite Duration.” Journal of Geophysical Research 69, no. 8: 1531–1540.
Herricks, E.E., ed.1995. Storm Water Runoff and Receiving Systems: Impact, Monitoring, and Assessment.
Boca Raton, Florida: Engineering Foundation and ASCE, CRC/Lewis.
eties 18: 363-368.
Hershfield, D. M. 1961. Rainfall Frequency Atlas of the United States for Durations from 30 Minutes to 24
Hours and Return Periods from 1 to 100 Years. Technical Paper No. 40. Washington, D.C.: Weather
Bureau, U.S. Department of Commerce.
Hillary, R. 1994. The Eco-Management and Audit Scheme; A Practical Guide. Trans-Atlantic Designs.
Hjulstrom, F. 1935. “The Morphological Activity of Rivers as Illustrated by River Fyris.” Bulletin of the
Geological Institute 25, chap. 3.
Holtz, R. D. and W. D. Kovacs. 1981. An Introduction to Geotechnical Engineering. Englewood Cliffs,
New Jersey: Prentice Hall.
Horton, R. 1919. “Rainfall Interception.” Monthly Weather Review 47: 602–623.
Horton, R. 1933. “Separate Roughness Coefficients for Channel Bottom and Sides.” Engineering News-
Record 111, no. 22: 652–653.
Horton, R. 1939. “Analysis of Runoff Plot Experiments with Varying Infiltration Capacity.” Transactions,
American Geophysical Union 20: 693–711.
Hosking, J. R. M. 1990. “L-moments: Analysis and Estimation of Distributions Using Linear Combinations
of Order Statistics,” Journal of the Royal Statistical Society, Series B 52, no. 2105–124.
660 Bibliography
Hosking, J. R. M. and J. R. Wallis. 1997. Regional Frequency Analysis, An Approach Based on L-moments.
Cambridge University Press, 1997.
Huber, W. C. and Dickinson, R. E. 1988. Storm Water Management Model User's Manual. Version 4. EPA-
600/3-88-001a (NTIS PB88-236641/AS). Athens, Georgia: U.S. EPA.
Huff, F. A. 1967. “Time Distribution of Rainfall in Heavy Storms.” Water Resources Research 3, no. 4:
1007–1019.
Huff, F. A. and J. R. Angel. 1992. “Rainfall Frequency Atlas of the Midwest.” Illinois State Water Survey
Bulletin 71: 141.
Hydraulic Institute. Engineering Data Book. Parsippany, New Jersey: Hydraulic Institute.
Hydraulic Institute. Standards for Vertical Turbine Pumps. Parsippany, New Jersey: Hydraulic Institute.
Hydraulic Institute. 1998. Centrifugal/Vertical Pump Intake Design. Parsippany, New Jersey: Hydraulic
Institute.
Hydraulic Institute. 1990. Hydraulic Institute Engineering Data Book, 2d ed. Parsippany, New Jersey:
Hydraulic Institute.
Interagency Advisory Committee on Water Data (IACWD). 1982. Guidelines for Determining Flood Flow
Frequency. Bulletin 17B, IACWD.
International Commission for the Protection of the Rhine (ICPR). 1992. Ecological master plan for the
Rhine. Coblenz: International Commission for the Protection of the Rhine against Pollution.
International Rivers Network (IRN). 2003. IRN Press Release, 27 February 2003.
International Rivers Network, (IRN). 2003. World Bank “Water Strategy Is Dishonest and Reactionary.”
Berkeley, CA: International Rivers Network.
Izzard, C. F. 1946. “Hydraulics of Runoff from Developed Surfaces.” Proceedings, Highway Research
Board 26: 129–150.
Jain, A. K. 1976. “Accurate Explicit Equation for Friction Factor.” Journal of the Hydraulics Division,
ASCE 102, no. HY5: 674–677.
Jenkinson, A. F. 1955. “The Frequency Distribution of the Annual Maximum (or Minimum) Value of Mete-
orological Elements.” Quarterly Journal of the Royal Meteorological Society 81: 158–171.
Jennings, M. E.; W. O. Thomas, Jr.; and H. C. Riggs. 1994. Nationwide Summary of U.S. Geological Survey
Regional Regression Equations for Estimating Magnitude and Frequency of Floods for Ungaged Sites.
Water Resources Investigations Report 94-4002. U.S. Geological Survey.
Jin, M., S. Coran, and J. Cook. 2002. “New One-Dimensional Implicit Numerical Dynamic Sewer and
Storm Model.” Proceedings of the 9th International Conference on Urban Drainage. American Society
of Civil Engineers (ASCE).
Johnson, L. E., P. Kucera, C. Lusk, and W. F. Roberts. 1988. “Usability Assessments for Hydrologic Fore-
casting DSS.” Journal of the American Water Resources Association 34, no. 1: 43–56.
Johnstone, D. and W. P. Cross. 1949. Elements of Applied Hydrology. New York: Ronald Press.
Kamaduski, G. E. and R. H. McCuen. 1979. “Evaluation of Alternative Stormwater Detention Policies.”
Journal of Water Resources Planning and Management 105(WR2): 171.
Kapur, K. C., and L. R. Lamberson, 1977. Reliability in Engineering Design. New York: John Wiley.
Karassik, I. J., J. P. Messina, P. Cooper, and C. C. Heald. 2000. Pump Handbook, 3d ed. New York:
McGraw-Hill.
Keifer, C. J. and H. H. Chu. 1957. “Synthetic Storm Pattern for Drainage Design.” Journal of the Hydrau-
lics Division, ASCE 84, no. HY4: 1-25.
Kennedy, W. J., Jr., and J. E. Gentle. 1980. Statistical Computing. New York: Marcel Dekker.
661
Kent, K. 1972. “Travel Time, Time of Concentration, and Lag.” National Engineering Handbook, Section
4: Hydrology, Chapter 15. Washington, D.C.: Soil Conservation Service (SCS), U.S. Department of
Agriculture.
Kerby, W. S. 1959. “Time of Concentration of Overland Flow.” Civil Engineering 60: 174.
Kessler, A. and M. H. Diskin. 1991. “The Efficiency Function of Detention Reservoirs.” Water Resources
Research 27 no. 3: 259.
Kibler, D. F. 1982. “Desk-Top Runoff Methods.” pp. 87–135 in D.F. Kibler, ed. Urban Stormwater Hydrol-
ogy. Water Resources Monograph Series. Washington, D.C.: American Geophysical Union.
Kirby, W. 1974. “Algebraic Boundedness of Sample Statistics.” Water Resources Research 10, no. 2: 220–
222.
Kirpich, T. P. 1940. “Time of Concentration of Small Agricultural Watersheds.” Civil Engineering 10, no.
6: 362.
Knighton, D. 1998. Fluvial Forms and Processes. Arnold.
Kottegoda, N. T. and R. Rosso. 1997. Statistics, Probability, and Reliability for Civil and Environmental
Engineers. New York: McGraw-Hill.
Krishnamurthy, M. and B. A. Christensen. 1972. “Equivalent Roughness for Shallow Channels.” Journal of
the Hydraulics Division, ASCE 98, no. 12: 2257–2263.
Kroll, C. N. and J. R. Stedinger. 1996. “Estimation of Moments and Quantiles Using Censored Data.”
Kuczera, G. 1982. “Robust Flood Frequency Models.” Water Resources Research 18, no. 2: 315–324.
Kuichling, E. 1889. “The Relation Between the Rainfall and the Discharge of Sewers in Populous Dis-
tricts.” Transactions, American Society of Civil Engineers 20: 1–60.
Lager, J. A., W. G. Smith, W. G. Lynard, R. M. Finn, and E. J. Finnemore. 1977. Urban Stormwater Man-
agement and Technology: Update and Users' Guide. U.S. EPA-600/8-77-014. Cincinnati, Ohio: Envi-
ronmental Protection Agency.
Lane, E. W. 1955. “Design of Stable Channels.” Transactions, American Society of Civil Engineers 120:
1234–1260.
Laplace, D., A. Bachoc, Y. Sanchez, and D. Dartus. 1992. “Truck Sewer Clogging Development–Descrip-
tion and Solutions.” Water Science and Technology 25, no. 8: 91–100.
Larson, L. W. and E. L. Peck. 1974. “Accuracy of Precipitation Measurements for Hydrologic Modeling.”
Linsley, R. K. 1943. “Application of the Synthetic Unit Graph in the Western Mountain States.” Transac-
tions American Geophysical Union 24:581–587.
Linsley, R. K., Jr.; M. A. Kohler; and J. L. H. Paulhus. 1982. Hydrology for Engineers. 3rd ed. New York:
McGraw-Hill.
Loganathan, G. V., D. F. Kibler, and T. J. Grizzard. 1996. “Urban Stormwater Management.” In L. W.
Mays, ed. Water Resources Handbook. New York: McGraw-Hill.
Lotter, G.K. 1933. “Considerations of Hydraulic Design of Channels with Different Roughness of Walls.”
Transactions All Union Scientific Research, Institute of Hydraulic Engineering 9: 238–241.
Loucks, D. P., J. R. Stedinger, and D. A. Haith. 1981. Water Resource Systems Planning and Analysis.
Englewood Cliffs, N.J.: Prentice Hall.
Manning, R. 1891. “On the Flow of Water in Open Channels and Pipes.” Transactions, Institution of Civil
Matheussen, B. R. and Thorolfsson, S. T. 2001. “Urban Snow Surveys in Risvollan, Norway, Urban Drain-
age Modeling.: Proceedings of the Specialty Symposium of the World and Water Environmental
Resources Conference. Alexandria, VA: ASCE.
662 Bibliography
Mays, L. W. 2001. Water Resources Engineering. New York: John Wiley and Sons.
McCarley, R. W., J. J. Ingram, B. J. Brown, and A. J. Reese. 1990. Flood Control Channel National Inven-
tory. MP HL-90-10. Vicksburg, Mississippi: U.S. Army Engineer Waterways Experiment Station.
McCarthy, G. T. 1938. “The Unit Hydrograph and Flood Routing.” North Atlantic Division Conference.
U.S. Army Corps of Engineers.
McCuen, R. H. 1998. Hydrologic Analysis and Design, 2nd ed. Upper Saddle River, NJ: Prentice Hall.
McCuen, R. H., S. L. Wong, and W. J. Rawls. 1984. “Estimating Urban Time of Concentration.” Journal of
Hydraulic Engineering 110, no. 7: 887-904.
McEnroe, B. M. 1992. “Preliminary Sizing of Detention Reservoirs to Reduce Peak Discharge.” Journal of
Hydraulic Engineering 118, no. 11: 1540.
McWhorter, J. C., T. G. Carpenter, and R.N. Clark. 1968. “Erosion control criteria for drainage channels.”
Study conducted for the Mississippi State Highway Department and the Federal Highway Administra-
tion. State College, Mississippi: Department of Agricultural Engineering, Mississippi State University.
Mehrota, S. C. 1983. “Permissible Velocity Correction Factor.” Journal of the Hydraulics Division, ASCE
109, no. HY2: 305–308.
Merriam, R. 1960. “A Note on the Interception Loss Equation.” Journal of Geophysical Research 65:
3850–3851.
Merriam, R. A. 1961. “Surface Water Storage on Annual Ryegrass.” Journal of Geophysical Research 66:
1833-1838.
Merritt, F. S., ed. 1976. Standard Handbook for Civil Engineers, 2nd ed. New York: McGraw-Hill.
Metcalf, L. and H. P. Eddy. 1928. Design of Sewers. Vol. 1 of American Sewerage Practice. New York:
McGraw-Hill.
Metropolitan Washington Council of Governments (Anacostia Restoration Team). 1992. A Current Assess-
ment of Urban Best Management Practices: Techniques for Reducing Nonpoint Source Pollution in the
Coastal Zone. Department of Environmental Programs, Metropolitan Washington Council of Govern-
ments, Water Resources Planning Board.
Miller, J. F., R. H. Frederick, and R. J. Tracey. 1973. Precipitation-Frequency Atlas of the Coterminous
Western United States. NOAA Atlas 2, 11 vols. Silver Spring, Maryland: National Weather Service.
Mockus, V. 1969. “Hydrologic Soil-Cover Complexes.” National Engineering Handbook, Section 4:
Hydrology, Chapter 9. Washington, D.C.: Soil Conservation Service (SCS), U.S. Department of Agri-
culture.
Mockus, V. 1972. “Estimation of Direct Runoff from Storm Rainfall.” National Engineering Handbook,
Section 4: Hydrology, Chapter 10. Washington, D.C.: Soil Conservation Service (SCS), U.S. Depart-
ment of Agriculture.
Mockus, V. 1973. “Estimation of Direct Runoff from Snowmelt.” National Engineering Handbook, Section
4: Hydrology, Chapter 11. Washington, D.C.: Soil Conservation Service (SCS), U.S. Department of
Agriculture.
Mockus, V. and W. Steiner. 1972. “Flood Routing.” National Engineering Handbook, Section 4 Hydrology,
Chapter 17. Washington, D.C.: Soil Conservation Service, U.S. Department of Agriculture.
Moody, L. F. 1944. “Friction Factors for Pipe Flow.” Transactions, American Society of Mechanical Engi-
neers 66.
Morel-Seytoux, H. J., and J. P. Verdin. 1981. Extension of Soil Conservation Service Rainfall Runoff Meth-
odology for Ungaged Watersheds. Report FHWA/Rd-81/060. Washington, D.C.: Federal Highway
Administration.
Motayed, A.K., and M. Krishnamurthy. 1980. “Composite Roughness of Natural Channels.” Journal of the
Hydraulics Division, ASCE 106, no. 6: 1111-1116.
663
Muhlhofer, L. 1933. “Rauhigkeitsunteruchungen in einem stollen mit betonierter sohle und unverkleideten
wanden.” Wasserkraft und Wasserwirtschaft 28(8): 85-88.
National Oceanic and Atmospheric Administration, National Weather Service. 2000. FLDWAV Computer
Program, Version 2-0-0.
National Research Council. 1991. Opportunities in the Hydrologic Sciences. Washington, D.C.: Committee
on Opportunities in the Hydrologic Sciences, Water Science and Technology Board, National Academy
Press.
National Science and Technology Council. 1997. Committee on Environment and Natural Resources: Fact
Sheet, Natural Disaster Reduction Research Initiative.
Natural Resources Conservation Service (NRCS). 1993. “Storm Rainfall Depth.” National Engineering
Handbook, Part 630: Hydrology, Chapter 4. Washington, D.C.: U.S. Department of Agriculture.
Nix, S. J. and S. R. Durrans. 1996. “Off-Line Stormwater Detention Systems.” Journal of the American
Water Resources Association 32, no. 6: 1329–1340.
Normann, J. M., R. J. Houghtalen, and W. J. Johnston. 2001. Hydraulic Design of Highway Culverts, 2nd
ed. Washington, D.C.: Hydraulic Design Series No. 5, Federal Highway Administration (FHWA).
Nunnally, N. R. 1978. “Stream Renovation: An Alternative to Channelization.” Environmental Manage-
ment 2, no. 5: 403.
Oberts, G. 1994. “Influence of Snowmelt Dynamics on Stormwater Runoff Quality.” Watershed Protection
Techniques 1,g no. 2: 55–61.
Olin, D. A. 1984. Magnitude and Frequency of Floods in Alabama. Water Resources Investigations Report
84-4191. U.S. Geological Survey.
Olsen, O. J. and Q. L. Florey. 1952. Sedimentation Studies in Open Channels: Boundary Shear and Velocity
Distribution by Membrane Analogy, Analytical, and Finite-Difference Methods. U.S. Bureau of Recla-
mation, Laboratory Report Sp-34, August.
Overton, D. E., and M. E. Meadows. 1976. Stormwater Modeling. New York: Academic Press.
Paine, J. N. and A. O. Akan. “Design of Detention Systems.” In L. W. Mays, ed. Stormwater Collection
Systems Design Handbook. New York: McGraw-Hill.
Patric, J. H. 1966. “Rainfall Interception by Mature Coniferous Forests of Southeast Alaska.” Journal of
Soil and Water Conservation 21: 229–231.
Pavlovskii, N. N. 1931. “On a Design Formula for Uniform Movement in Channels with Nonhomogeneous
Walls.” (in Russian) Izvestiia Vsesoiuznogo Nauchno-Issledovatel'skogo Insituta Gidrotekhniki 3: 157–
164.
Petratech. 2003. Product literature for Petraflex interlocking blocks. (www.petraflex.com).
Pitt, R. 1986. “The incorporation of urban runoff controls in the Wisconsin Priority Watershed Program.”
Pp. 290–313 in B. Urbonas and L. A. Roesner, ed. Advanced Topics in Urban Runoff Research. New
York: Engineering Foundation and ASCE.
Pitt, R. 1995a. “Biological Effects of Urban Runoff Discharges.” Pp. 127–162 in E. E. Herricks, ed. Storm
water Runoff and Receiving Systems: Impact, Monitoring, and Assessment. Boca Raton, Florida: Engi-
neering Foundation and ASCE, CRC/Lewis.
Pitt, R. 1995b. “Effects of Urban Runoff on Aquatic Biota” Pp. 609–630 in D. J. Hoffman; B. A. Rattner;
G. A. Burton, Jr.; and J. Cairns, Jr., ed. Handbook of Ecotoxicology. Boca Raton, Florida: Lewis Pub-
lishers/CRC Press.
Pitt, R. 1996. Groundwater Contamination from Storm water. Chelsea, Michigan: Ann Arbor Press.
Pitt, R. and J. Voorhees. 1993. “Source loading and management model (SLAMM).” Seminar Publication,
National Conference on Urban Runoff Management, Enhancing Urban Watershed Management at the
Local, County, and State Levels, March 30–April 2.
664 Bibliography
Pitt, R. E., R. I. Field, M. M. Lalor, D. D. Adrian, and D. Barbé. 1993. Investigation of Inappropriate Pol-
lutant Entries into Storm Drainage Systems: A User's Guide. Rep. No. EPA/600/R-92/238, NTIS Rep.
No. PB93-131472/AS, U.S. EPA, Storm and Combined Sewer Pollution Control Program (Edison, New
Jersey). Cincinnati, Ohio: Risk Reduction Engineering Lab.
Pitt, R., R. Field, M. Lalor, and M. Brown. 1995. “Urban storm water toxic pollutants: assessment, sources
and treatability.” Water Environment Research 67, no. 3 260–275.
Pitt, R., S. Nix, J. Voorhees, S. R. Durrans, and S. Burian. 2000. Guidance Manual for Integrated Wet
Weather Flow (WWF) Collection and Treatment Systems for Newly Urbanized Areas (New WWF Sys-
tems). Second year project report: model integration and use, Wet Weather Research Program, U.S.
Environmental Protection Agency. Cooperative agreement #CX 824933-01-0.
Ponce, V. M., R. M. Li, and D. B. Simons. 1978. “Applicability of Kinematic and Diffusion Models.” Jour-
nal of Hydraulic Division, ASCE 104, HY12: 1663.
Potter, M. C., and D. C. Wiggert. 1991. Mechanics of Fluids. Englewood Cliffs, NJ: Prentice Hall.
Powell, R. W. 1950. “Resistance to Flow in Rough Channels.” Transactions, American Geophysical Union
31, no. 4: 575–582.
Prandtl, L. 1952. Essentials of Fluid Dynamics. New York: Hafner Publ.
Prasuhn, A.L. 1980. Fundamentals of Fluid Mechanics. Englewood Cliffs, New Jersey: Prentice-Hall.
Preul, H. C. 1996. “Combined sewage prevention system (CSCP) for domestic wastewater control.” Pp.
193–198 in Proceedings of the 7th International Conference on Urban Storm Drainage. Hannover, Ger-
many.
Price, R. K. 1973. “A comparison of four numerical methods for flood routing.” Journal of the Hydraulics
Division, ASCE 100, no. 7: 879–899.
Rasmussen, P. F. and D. Rosbjerg. 1991. “Prediction Uncertainty in Seasonal Partial Duration Series.”
Rasmussen, P. F., and D. Rosbjerg. 1989. “Risk Estimation in Partial Duration Series.” Water Resources
Research 25, no. 11: 2319–2330.
Rawls, W. J., Ahuja, L. R., Brakensiek, D. L., and Shirmohammadi, A. 1993. “Chapter 5: Infiltration and
Soil Water Movement” in D. R. Maidment, ed. Handbook of Hydrology. New York: McGraw Hill.
Rawls, W., P. Yates, and L. Asmusse. 1976. Calibration of Selected Infiltration Equations for the Georgia
Coastal Plain. Technical Report ARS-S-113. Washington, D.C.: U.S. Dept. of Agriculture, Agricultural
Research Service.
Ree, W. O. 1949. “Hydraulic Characteristics of Vegetation for Vegetated Waterways.” Agricultural Engi-
neering 30, no. 4: 184–187 and 189.
Ree, W. O. and V. J. Palmer. 1949. Flow of Water in Channels Protected by Vegetative Lining. U.S. Soil
Conservation Service, Technical Bulletin No. 967.
Reynolds, O. 1883. “An Experimental Investigation of the Circumstances Which Determine Whether the
Motion of Water Shall Be Direct or Sinuous and of the Law of Resistance in Parallel Channels.” Philo-
sophical Transactions of the Royal Society 174: 935.
Richards, K. 1982. Rivers: Form and Processes in Alluvial Channels. New York: Methuen and Co.
Rodriguez-Iturbe, I., and J. B. Valdes. 1979. “The Geomorphologic Structure of Hydrologic Response.”
Roesner, L. A., J. A. Aldrich, and R. E. Dickinson. 1989. “Storm Water Management Model User’s Manual
Version 4: Extra Addendum.” EPA 600/3-88/001b.
Roesner, L. A., B. R. Urbonas, and M. B. Sonnen (editors). 1989. “Design of Urban Runoff Quality Con-
trols,. Proceedings of an Engineering Foundation Conference on Current Practice and Design Criteria
for Urban Quality Control. American Society of Civil Engineers (ASCE).
665
Rosbjerg, D., H. Madsen, and P. F.Rasmussen. 1992. “Prediction in Partial Duration Series With General-
ized Pareto-Distributed Exceedances.” Water Resources Research 28, no. 11: 3001.
Rouse, H. 1936. “Discharge Characteristics of the Free Overfall.” Civil Engineering 6, no. 7: 257.
Rouse, H. 1946. Elementary Fluid Mechanics. New York: Wiley.
Sanks, R. L. 1998. Pumping Station Design, 2nd ed. Boston: Butterworth-Heinemann.
Sauer, V. B.; W. O. Thomas, Jr.; V. A. Stricker; and K. V. Wilson. 1983. Flood Characteristics of Urban
Watersheds in the United States. Water Supply Paper 2207. U.S. Geological Survey.
Schaake, J. C., Jr.; J. C. Geyer; and J. W. Knapp. 1967. “Experimental Examination of the Rational
Method.” Journal of the Hydraulics Division, American Society of Civil Engineers 93, no. HY6.
Seely, F. B., and J. O. Smith. 1959. Advanced Mechanics of Materials. 2nd ed. New York: John Wiley and
Sons.
Semadeni-Davies, A. 2000. “Representation of Snow in Urban Drainage Models.” J. Hydrologic Engineer-
ing 5, no. 4: 363-370.
Shames, I. H. 1992. Mechanics of Fluids, 3d ed. New York: McGraw-Hill.
Sherman, L. K. 1932. “Stream Flow from Rainfall by the Unit Graph Method.” Engineering News-Record
108: 501-505.
Shields, F. D., R. R. Copeland, P. C. Klingeman, M. W. Doyle, and A. Simon. 2003. “Design for Stream
Restoration.” Accepted for ASCE Journal of Hydraulic Engineering.
Simon, A. L. 1981. Practical Hydraulics. 2d ed. New York: John Wiley & Sons.
Singh, V. P. 1988. Hydrologic Systems, Rainfall-Runoff Modeling, vol. I. Englewood Cliffs, New Jersey:
Prentice Hall.
Singh, V. P. 1992. Elementary Hydrology. Englewood Cliffs, New Jersey: Prentice Hall.
Snider, D. 1972. “Hydrographs.” National Engineering Handbook, Section 4: Hydrology, Chapter 16.
Washington, D.C.: Soil Conservation Service (SCS), U.S. Department of Agriculture.
Snyder, F. F. 1938. “Synthetic Unit Graphs.” Transactions, American Geophysical Union 19: 447.
Soil Conservation Service (SCS). 1969. Section 4: Hydrology in National Engineering Handbook. Wash-
Soil Conservation Service (SCS). 1992. TR-20 Computer Program for Project Formulation Hydrology
(revised users' manual draft). Washington, D.C.: U.S. Department of Agriculture.
Stedinger, J. R., and Cohn, T. A. 1986. “Flood Frequency Analysis With Historical and Paleoflood Infor-
mation.” Water Resources Research 22, no. 5: 785.
Steel, E. W. and T. J. McGhee. 1979. Water Supply and Sewerage. 5th ed. New York: McGraw-Hill.
Stillwater Outdoor Hydraulic Laboratory. 1954. Handbook of Channel Design for Soil and Water Conser-
vation. U.S. Soil Conservation Service, SCS-TP-61.
Strickler, A. 1923. “Some Contributions to the Problem of the Velocity Formula and Roughness Factors for
Rivers, Canals and Closed Conduits” (in German). Mitteilungen des Eidgenossischen Amtes fur Wasser-
wirtschaft 16.
Swamee, P. K., and A. K. Jain. 1976. “Explicit Equations for Pipe Flow Problems.” Journal of the Hydrau-
lics Division, ASCE 102, no. HY5: 657-664.
Sweeney, T. L. 1992. Modernized Areal Flash Flood Guidance. NOAA Technical Memorandum NWS
HYDRO 44. Silver Spring, Maryland: National Weather Service, Office of Hydrology.
Taylor, A.B., and H.E. Schwarz. 1952. “A unit hydrograph lag and peak flow related to basin characteris-
tics.” Trans. Am. Geophys. Union 33: 235.
666 Bibliography
U.S. Army Corps of Engineers (USACE). 1959. Flood Hydrograph Analysis and Computations, Engineer-
ing and Design Manuals. EM 1110-2-1405. Washington, D.C.: USACE.
U.S. Army Corps of Engineers (USACE). 1960. Routing of Floods Through River Channels. Engineering
and Design Manuals, EM 1110-2-1408. Washington, D.C.: USACE.
U.S. Army Corps of Engineers (USACE). 1989. Structural and Architectural Design of Pumping Stations.
EM 1110-2-3104. Washington, D.C.: USACE.
U.S. Army Corps of Engineers (USACE). 1998. HEC-1 Flood Hydrograph Package User’s Manual. Davis,
Ca.: USACE Hydrologic Engineering Center.
U.S. Army Corps of Engineers (USACE). 1991. HEC-2, Water Surface Profiles, User's Manual. Davis, Ca:
USACE Hydrologic Engineering Center.
U.S. Army Corps of Engineers (USACE). 1995. General Principles of Pumping Station Design and Lay-
out. EM 1110-2-3102. Washington, D.C.: USACE.
U.S. Army Corps of Engineers (USACE). HEC-RAS River Analysis System. Davis, California: USACE
Hydrologic Engineering Center.
U.S. Army Corps of Engineers. 1989. Environmental Engineering for Local Flood Control Channels. Engi-
neering Manual 1110-21601. Washington, D.C.: USACE.
U.S. Army Corps of Engineers. 1994. Hydraulic Design of Flood Control Channels. Engineering Manual
U.S. Army Corps of Engineers Hydrologic Engineering Center (HEC). 1998. HEC-1 Flood Hydrograph
Package User’s Manual. Davis, Ca.: HEC.
U.S. Army Corps of Engineers. 1999. Mechanical and Electrical Design of Pumping Stations. EM 1110-2-
U.S. Bureau of Reclamation (USBR). 1955. Research Studies on Stilling Basins, Energy Dissipators, and
Associated Appurtenances. Hydraulic Laboratory Report No. Hyd-399. Lakewood, Colorado: USBR.
USBR.
U.S. Bureau of Reclamation (USBR). 1987. Design of Small Dams. 3rd ed. Lakewood, Colorado: USBR.
U.S. Bureau of Reclamation (USBR). 1987. Design of Small Dams. Washington, D.C.: U.S. Department of
the Interior, USBR.
U.S. Environmental Protection Agency. 1991. Guidance for Water Quality-Base Decisions: The TMDL
Process. EPA 440/4-91-001.
U.S. Environmental Protection Agency. December, 1983. Final Report for the Nationwide Urban Runoff
Program. Washington, D.C.: Water Planning Division.
U.S. Environmental Protection Agency. September, 1993. Handbook on Urban Runoff Pollution Prevention
and Control Planning. 625/R-93-004.
U.S. Weather Bureau. 1958. Rainfall Intensity-Frequency Regime, Part 2 – Southeastern United States.
Technical Paper 29. Washington, D.C.: U.S. Department of Commerce.
United Nations Development Program (UNDP). 1998. Capacity Building for Sustainable Management of
Water Resources and the Aquatic Environment. New York: United Nations.
United Nations. 1977. Report of the United Nations Conference, Mal del Plata, 14–25 March 1977.
United Nations. 1992. Report of the United Nations Conference on Environment and Development
(UNCED), Rio de Janeiro, June 1992.
United Nations. 1997. Report Of The Commission On Sustainable Development On Preparations For The
Special Session Of The General Assembly For The Purpose Of An Overall Review And Appraisal Of
The Implementation Of Agenda 21.
667
2, Major Drainage. Denver, Colorado: UDFCD. November 1982.
3, Best Management Practices. Denver, Colorado: Urban Drainage and Flood Control District. Septem-
ber, 1999.
Urbonas, B. and P. Stahre. 1993. Stormwater Best Management Practices Including Detention. Englewood
Cliffs, New Jersey: Prentice-Hall.
Urbonas, B. R. and L. A. Roesner. 1993. “Urban Design and Urban Drainage for Flood Control.” In D.
Maidment, ed. Handbook of Hydrology. New York: McGraw-Hill.
van der Vink, G., R.M. Allen, J. Chapin, M. Crooks, W. Fraley, J. Krantz, A.M. Lavigne, A. LeCuyer, E.K.
MacColl, W.J. Morgan, B. Ries, E. Robinson, K. Rodriquez, M. Smith, and K. Sponberg. 1998. “Why
the United States is Becoming More Vulnerable to Natural Disasters,” EOS, Transactions of the Ameri-
can Geophysical Union 79, no. 44: 533–537.
Viessman, W., J. Knapp, and G. L. Lewis. 1977. Introduction to Hydrology. 2nd ed. New York: Harper &
Row.
Viessman, W., Jr., and M. J. Hammer. 1993. Water Supply and Pollution Control. 5th ed. New York: Harper
Collins.
Virginia Department of Conservation and Recreation (DCR). 1992. Virginia Erosion and Sediment Control
Handbook. 3rd ed. Richmond, Virginia: DCR.
Vogel, R. M., and N. M. Fennessey, “L-moment Diagrams Should Replace Product Moment.”
Wallis, J. R., N. C. Matalas, and J. R. Slack. 1971. “Just a Moment!” Water Resources Research 10, no. 2:
211–219.
Walski, T. M. 1984. Analysis of Water Distribution Systems. New York: VanNostrand Reinhold.
Wang, Q. J. 1990. “Estimation of the GEV Distribution from Censored Samples by Method of Partial Prob-
ability Weighted Moments.” Journal of Hydrology 120: 103–114.
Wang, Q. J. 1997. “LH Moments for Statistical Analysis of Extreme Events.” Water Resources Research
33, no. 12: 2841–2848.
Weber, U. 2000. “The ‘Miracle’ of the Rhine” Courier, June 2000, Unesco, Paris.
Weinmann, D. E. and E. M. Laurenson. 1979. “Approximate Flood Routing methods: A Review.” Journal
of Hydraulics Division, ASCE. 105, HY12: 1521.
White, F.M. 1999. Fluid Mechanics. 4th ed. New York: McGraw-Hill.
Williams, G. B. 1922. “Flood Discharges and the Dimensions of Spillways in India.” Engineering (London)
134: 321.
Williams, G. S., and A. H. Hazen. 1933. Hydraulic Tables. 3rd ed. New York: John Wiley and Sons.
World Bank. 1993. Water Resources Management. Washington, D.C.: International Bank for Reconstruc-
tion and Development/The World Bank.
World Bank. 1998. Pollution Prevention and Abatement Handbook 1998: Toward Cleaner Production.
Washington, D.C.: International Bank for Reconstruction and Development/The World Bank.
World Bank. 2002. Thailand Environmental Monitor 2001: Water Quality. Washington, D.C.: The World
Bank Group.
World Bank. 2003. The New Water Resources Strategy. Washington, D.C.: World Bank Group.
World Commission on Environment and Development (WCED). 1987. Our Common Future. Oxford,
Great Britain: Oxford University Press.
Yang, C. T. 1976. “Minimum Unit Stream Power and Fluvial Hydraulics.” Journal of Hydraulic Engineer-
ing. 105 (HY7): 769.
668 Bibliography
Yang, C. T. 1996. Sediment Transport: Theory and Practice. New York: McGraw-Hill.
Yen, B. C. 1996. “Hydraulics for Excess Water Management.” In L. W. Mays, ed. Water Resources Hand-
book. New York: McGraw-Hill.
Yen, B. C. 2001. “Hydraulics of Storm Sewer Systems.” In L. Mays, ed. Stormwater Collection System
Design Handbook. New York: McGraw Hill.
Young, G. K., S. E. Walker, and F. Chang. 1993. Design of Basic Bridge Deck Drainage. Hydraulic Engi-
neering Circular No. 21, GHWA-SA-92-010. Washington, D.C.: U.S. Department of Transportation.
Zinke, P.J. 1967. “Forest Interception Studies in the United States.” Pp. 137–160 in W. E. Sopper and H. W.
Lull, ed. Forest Hydrology. Oxford: Pergamon Press.
Index
contribution to hydrographs, 165fig
A estimation of, 165–166

hydrograph, 165fig
See also dry-weather flows; subsurface flow
Aarlborg Charter. See Charter of European Cities and base flows, 165–166
Towns Towards Sustainability estimation of, 165–166
adverse open channels, 219 basin lag time. See drainage basin lag time
aesthetic considerations, in design of open channels, 290– basin response time, measures of, 134–135
292 Bazin's equation for Chézy’s C, 229, 229t
affinity laws of pumps, 547 values for Bazin’s m, 229t
Agenda 21, 576–577, 576t, 579 bell-and-spigot pipe joints, 412
objectives, 577 benching, of junction structures, correction factors for
See also Local Agenda 21 minor loss coefficients, 438t
agricultural runoff, 597 best hydraulic section of an open channel, 291–293, 292fig
air release valves, 558 calculation (example), 292–293
albedo, 169 best management practices (BMPs), 607, 617–628
angle of inclination, 219 applications, 630–636
angle of repose, of sediments, 306–307, 307fig for construction projects, 633–634, 634–636
antecedent soil moisture conditions, 111 effectiveness in treating stormwater, 626–628, 628t
curve numbers for various moisture conditions, 125t implementation in developing areas, 631
effect on curve number method, 123–124 information sources, 617–618, 637sb
assumptions, rational method of peak flow estimation, 122 sources of information, 617–618, 637sb
Atlas 2, 80, 91 steps for selecting structural controls, 632–633
attenuation, of flow, 645, 646 stormwater controls for existing areas, 633–634
average cross-sectional velocity, 187–188 See also nonstructural best management practices;
adjustment equation, 187 stormwater quality management; structural best
energy equation adjustment for, 187–188 management practices
momentum equation adjustment for, 188t, 209–210 bioretention areas, 631
values for different velocity distributions, 188t BMPs. See best management practices (BMPs)
See also velocity boundary condition in gradually varied flow, 272
brake power of pump, equation for, 547
brake power of pumps, 546–547
determination of (example), 547
bridge modeling, 323
broad-crested weirs, 268–270
B
Brown Agenda, 577–578
Brundtland Commission–Our Common Future, 576t
Bulletin 71, 98, 101
for design hyetograph distributions (example), 99–
bankfull discharge, 316, 316fig
101
base flow, 116, 150, 165–166
butterfly valves, 558
contribution to hydrograph, 165–166
670 Index
bypass flow of inlets, 379 Clean Water Act of 1977, 575, 595, 603
legislative history, 593–595, 593t
See also water law (U.S.)
clean-outs, in pumping stations, 558
climate change, stormwater management and, 583
closed-conduit flow
C Chézy equation for uniform flow in open channels

(and pipes), 200
Darcy-Weisbach equation, 192–195, 197–198
Canada Darcy-Weisbach equation for head loss due to
environmental legislation, 590–591 friction, 192–195, 197–198
stormwater management regulations, 589–592, 603 defined, 185–186
See also Ontario energy equation for, 186–191, 189fig
Canada Water Act of 1970, 590 Hazen-Williams equation for flows in distribution
Canada Water Conservation Assistance Act, 590 systems, 201
carryover. See bypass flow of inlets head losses due to friction, 192–201
catch basins, 416, 417fig Manning equation, 198–199
cleaning of, 629 minor losses, 202–208
cavitation in pumps. See pump cavitation momentum and forces, 208–211
channel classification, 218–219 Moody diagram for estimating friction factor, 194–
channel 196, 196fig
design, habitat floodway, 316fig Reynolds number (Re), 192
diversions, 636 See also Chézy equation; Hazen-Williams equation;
geometric properties, 219–223 pipe flow
cross-sectional properties of prismatic channels, cold fronts, 70, 70fig
222t Colebrook equation, 197
cross-sectional shapes, 221, 222fig combination inlets, 377fig, 378fig, 397–398, 397–399
inclination angle, 219 combined sewers, 409
longitudinal slope, 219, 220fig composite energy-loss method
trapezoidal channel property calculations (ex- for junction structures, 440–444
ample), 223 coefficients, 440–441
improvements, 315–316, 316fig correction factors, 441–444
lined, 219 See also energy-loss method
lining materials, 290 composite runoff coefficient, 141
modification, 315–317 concrete aprons, 361
nonprismatic, 219, 219fig, 220fig concrete pipes, 409–412
prismatic, 218–219, 219fig, 222t gaskets, 412
roughness height, 226–228 joints, 412
routing, of runoff hydrographs, 169–173 shapes, 409–410, 410fig, 411t
routing of runoff hydrograph, Muskingum routing sizes, 409–410, 411t
method, 170–173 standards, 409–412
section factor, 294 Conference on Environment and Development (1992, Rio
slope. See open-channel geometric properties de Janeiro), 583, 603
unlined, 219 conservation of energy, 55, 63
channel-forming flow, 316 flow and, 55
channelization, 315 See also elevation head; energy; energy equation;
Charted Institution of Water and Environmental pressure head; velocity head
Management, 588 conservation of mass, 51–53, 62fig
Charter of European Cities and Towns Towards application of, 51–53
Sustainability, 576t, 577 detention pond application (example), 54–55, 54fig
check dams, 635 discharge determination from (example), 54
check flows for culverts, 325 equation for detention pond routing, 511–513, 512fig
check valves, 558 equations for, 52–53
Chézy equation flow velocity determination from (example), 54
for closed-conduit flow, 200 principle of, 51
for closed-conduit flow velocity, 200 conservation of momentum, 58–59, 63
for open-channel flow, 228–230 See also momentum equation
for uniform flow in pipes, 200 constructed wetlands
for uniform flow velocity, in open channels, 228 basins, 623–624
Chézy’s resistance coefficient, 200 channels, 625
calculation (example), 200 Construction Industry Research and Information
for closed-conduit flow (example), 200 Association (CIRIA), 588
for open-channel flow, 228–230 construction project best management practices, 634–636
Bazin’s equation for, 229, 229t, 229t drainageway protection, 636
Ganguillet and Kutter’s equation, 228–229 erosion control measures, 635
Powell’s equation for, 229–230 sediment control measures, 635–636
Chézy’s resistance coefficient for open-channel flow, construction projects
Powell's equation for, 229–230 best management practices, 633–634
Cipolletti weir, 267 drainageway protection, 636
cisterns, 626 erosion control, 635
cisterns and rain barrels, 626 sediment control, 635–636
Clark unit hydrograph method, 174t construction-site runoff, regulations concerning, 595, 596,
671
602 culvert end treatments, 327–331

continuous grade inlets, 377, 377fig cradles for pipes, 329–330, 330fig
contracted weirs, 265fig curved-edge entrances, 329fig
control sections, in open-channel flow, 252 cutoff walls, 329
Convention on the Law of the Non-Navigational Uses of depressed inlets, 337–338, 337fig
International Watercourses, 582 for erosion prevention, 327–331
Convention on the Protection and Use of Transboundary headwalls, 328
Watercourses and International Lakes, 582 improving hydraulic performance of inlets, 329–330,
convex routing, 649–650 337–338, 337fig
conveyance of an open-channel, 241–242 mitered ends, 327, 328fig
Coriolis forces, 70 precast end section, 327, 328fig
corrugated-steel pipes, 412–414 projecting barrels, 327, 328fig
joints, 414 to protect banks, 327–331
Manning’s roughness coefficient for, 414, 415t rounding inlet edges, 329, 329fig
shapes, 412, 414fig side-tapered entrances, 329–330, 330fig
sizes, 412, 413t slope-tapered entrances, 329–330, 330fig
standards, 412–414 square-edge entrances, 329fig
cost considerations in open-channel design, 290 wingwalls, 328, 328fig
covering storage and handling areas, 625 culvert entrances, loss coefficients, 203t, 341–342
cradles culvert flow, 334–355
for culvert outlets, 329–330, 330fig head loss, 334
for storm water sewers, 420, 464 inlet flow controls, 335–338, 336fig
critical depth outlet flow controls, 338–340, 339fig
of open-channel flow, 245fig, 246, 246–247 tailwater effects, 342–346
in circular culverts, 346–347, 347fig in circular culverts with smaller effects (exam-
in compound channels, 248–251 ple), 346
in compound channels (example), 250–251 in circular culverts with submerged inlets and
Froude number, 247 outlets (example), 344–345
of gradually varied flow, 273 culvert head, 341
rectangular channels (example), 247–248 culvert inlet control
trapezoidal channels (example), 247–248 calculation of headwater depths (example), 351–352
critical open-channel flow, 244–255 constants for different culvert materials and shapes,
alternate depths, 245fig, 246 350t
in compound channels, 248–251 equations for, 348–351
flow control section, 252 nomographs for, 340, 342fig
See also open-channel flow; specific energy culvert inlets. See culvert end treatments; pipe entrances
critical source area control, 625 culvert outlet flow control
critical tractive force, 305 calculation of headwater elevation of gradually
cross slopes of roadways, 371 varying flow in a rectangular culvert, 353–354
culvert alignments, 331–334 calculation of velocity (example), 355
alternative to channel realignment, 331 equations for headwater elevation, 352
bend deflection angle limitations, 332fig flow profile conditions, 339fig, 352–353, 352–354
broken-back, 332, 333, 333fig nomographs for, 341–346, 343fig
curved, 331, 332fig selection of flow profile analysis methods, 352–353
doglegged, 331, 332fig culvert outlet velocity, 354–355
horizontal, 331 calculation of velocity (example), 355
sag, 332–333, 333fig culvert outlets. See culvert end treatments; pipe outlets
skewed, 333–334, 334fig culvert performance analysis, 340–354
vertical, 332–333 flow profiles, 348–355
culvert bridge, 323 nomographs, 340–346, 342fig, 343fig
culvert cross-sectional shapes, 326 headwater elevation calculation inlet control
culvert design, 323–324 (example), 340–341
alignments. See culvert alignments selection of treatment to lower headwater eleva-
culvert design flows. See culvert design flows tion with inlet control (example), 341
culvert types and materials, 326–327 See also culvert inlet control equations; outlet
definition of a bridge, 323 velocity of culverts
end treatments. See culvert end treatments CulvertMaster, 653–654t
environmental considerations, 362 culverts
fish passage, 262fig, 362 alignment, 331–334
erosion protection, 360–361 with natural streambed materials, 362, 362fig
hydraulics. See culvert flow See also pipes
performance curves of culverts See performance curb-opening inlets, 377fig, 378fig, 392–396, 392fig
curves, of culverts in a sag, 394–397
roadway overtopping. See roadway overtopping capacity behaving as a weir with gutter depres-
use of multiple barrels, 326–327 sion, 395
culvert design flows, 325–326 capacity behaving as a weir without gutter de-
based on economic analysis, 325–326, 326fig pression, 395–396
based on recurrence interval, 325 capacity behaving as an orifice, 395–396
check flows, 325 capacity (example), 396–397
See also under performance curves weir equation, 395
culvert diameter, minimum recommended, 327 behaving as an orifice, capacity of, 393
culvert diameter, minimum recommended, 327 combination, 397fig
672 Index
effective depth of inlet behaving as an orifice, 395– data sources

396 rainfall, 76, 79–80
on a continuous grade soils, 123sb
efficiency of, 393 dead loads on pipes, 460
equivalent gutter cross slope, 393–394, 394fig degree-day method, for modeling snowmelt, 167–168
gutter depression, 393, 394fig depression storage, 108, 108fig, 109–110
increasing capacity of inlet, 392 depth of flow, 220–221, 220fig
inlets with supports, 392–394, 393fig design criteria, for detention facilities, 487–492
interception capacity, 392–394 design flows
interception capacity (example), 394 for culverts, 325–326
interception capacity of inlet, 393 for gutters and inlets, 369
length of opening required to intercept total gut- for storm sewer pipes, 423, 425–426
ter flow, 393 design procedure, for detention ponds, 525–526, 526fig
throat types, 396, 396fig design storm criteria
See also inlets for culverts, 325
curve number method for detention facilities, 488
adjustments for impervious areas, 129–130, 130 design storm hyetographs
assumptions, 120–121 alternating block method, 97–98
curve numbers example, 97–98, 97t, 98fig
for antecedent moisture conditions, 123–124, alternating block method (example), 97–98, 97t,
125t 98fig
for arid and semiarid rangelands, 128t coverage of NRCS rainfall distributions by storm
for cultivated agricultural lands, 127t type in U.S., 94fig
defined, 120 coverage of NRCS rainfall distributions in U.S.,
definition of, 120 94fig
equation for, 121 dimensionless distribution factors for U.S. Midwest,
for other agricultural lands, 128t 99fig
tables of, 124 dimensionless distributions for U.S. Midwest, 98
for urban areas, 126t dimensionless NRCS distribution factors for 24-
for various antecedent moisture conditions, hours by storm type, 93–94t, 94fig
123–124, 125t dimensionless NRCS distributions for 24-hours by
direct runoff depth and volume estimation (example), storm type, 93–94t, 94fig
129 instantaneous intensity method, 98
effective precipitation NRCS method, 93–96
equation for, 121 example, 93–96
graphical solution for, 122fig quartile-based storm pattern method, 98–101
equation for effective precipitation, 121 example, 99–101
graphical solution for effective precipitation, 122fig synthetic, development (examples), 95–101
hydrologic condition, defined, 122 detention
hydrologic soil group, 120–121, 122, 123sb on porous landscape, 621
hydrologic soil groups, 122 porous pavement, 618–619, 619fig
initial abstractions of rainfall, 119–120 detention basins
limitations of method, 129–131 sand filters, 623
maximum possible retention of rainfall by See also detention ponds
infiltration, 120–121 detention pond outlets, 494–495
precipitation components, 119–120, 120fig anti-seepage collars, 495
curve number tables, 124, 125t, 126t–128t anti-vortex devices, 495
cyclones, 70 composite structure, 494, 494fig
orifices, 495
overflow weirs, 495
pipes, 494
riser pipes, 495
spillways, 496
D See also outlet structures

detention ponds
aboveground vs. underground detention, 481–482
dam embankments, 493 basic concepts of operation, 478fig, 486fig
Danube River, 582, 583 by-passing flows from off-site, 490–491
Darcy-Weisbach equation, 192–195, 197–198 components, 493–497
application to pipe flow (example), 194 conservation of mass and (example), 54–55, 54fig
for energy gradient, 193 conservation of mass equation for routing of storage,
for energy gradients, 193 511–513, 512fig
explicit equations for friction factor, 195–197 dam embankments, 493
explicit Swamee-Jain equations for pipe flow and design criteria, 487–492
diameter, 197 design procedure, 525–526, 526fig
for head loss in circular conduits, 192–193 design storm criteria, 488
for head loss in noncircular conduits, 197–198 distinction between stormwater detention and
for pipe size requirement (example), 194–195 retention, 479
for pump system curves, 549 diversion structures, 497
for required pipe size (example), 194–195 drainage basin runoff depth (example), 62–63
See also Colebrook equation; head losses; Swamee- dry ponds, 480–481, 480fig
Jain equation erosion control problems, 493–494
673
freeboard, 493 equation for, 117

hydraulic analysis using computer models, 518– hyetograph calculations (examples), 132–134, 161–
519sb 162
independent vs. interconnected ponds, 484–485, direct surface runoff volume, 117
485fig calculation (example), 118
inflow-outflow hydrographs, 54fig equation for, 117–118
information sources, 479 discharge, conservation of mass and, 54
inlet box, 495fig discharge coefficient
inlet box structures, 495fig of broad-crested weirs, 270
modeling concepts, 485–487 for orifices, 270–271
offsite flows, 491fig for sharp-crested weirs, 266
on-line vs. off-line, 482, 482fig of suppressed rectangular sharp-crested weir, 267
operation concepts, 486–487 of triangular sharp-crested weirs, 268–269, 269fig
operational concepts, 478–479 discharges, of waste into storm drainage, 629
outlet structures, 494–495, 494fig, 495fig, 506–510, discrete convolution, process for computing direct runoff,
506fig, 507fig, 509fig 158–159
peak flow timing issues, 489–490, 490fig discrete convolution process for computing direct runoff
pond bottoms, 493 concept of, 159fig
predevelopment vs. postdevelopment criteria, 488 equation for summing hydrographs from individual
rating curves for discharge of, 505–506, 505fig rainfall pulses, 161
regional vs. on-site, 482–484, 483fig, 484fig runoff hydrograph (example), 161–162
routing with storage indication method, 509–524 discrete convolution process for computing direct runoff
runoff volume mitigation, 491–492, 492fig hydrographs, 158–162, 160fig, 160t
sheetflow at outlets, 496 distributed energy losses, 187
side slopes, 493 in open channels, 225–230
stage vs. discharge, 505–510, 505fig, 507fig, 509fig, distributed losses, in open-channel flow, 225–226
510fig diversion structures for detention ponds, 497
developing rating curves for outlet structures in diversions for sediment control, 635
parallel (example), 507–508 dominant discharge, in natural channels, 316
rating curves for outlet structures in series, drainage area, 118
510fig drainage basin lag time, 135–139, 135fig, 136t
stage vs. storage volume, 398–505, 500fig, 503fig time of concentration and, 135
average-end-area method (example), 501–502 drainage basin response time estimation
conic method, 503fig basin lag time, 135–139, 135fig, 136t
conic method (example), 503–504, 503fig time of concentration and basin lag time, 135
stage vs. surface area, 398–499 drainage basin response time estimation methods, 135–
stage vs. surface area (example), 498–499 138, 137t
storage indication routing method, 509–524 overland flow velocities as a function of land use and
storage indication routing method (example), 514– slope, 138fig
517 time of concentration, 137t
storage routing using computer models, 518–519sb time of concentration estimation (example), 138–139
storage volume, 499 drainage basin response time measures, 134–135
stormwater detention analysis procedure, 513–514 response time definition, 134
surface area, 497, 497–498 time of concentration, 134
surface area vs. elevation, 62–63, 62fig travel time to outlet, 134
types of facilities, 479–485 drainage systems, inappropriate discharges into, 629
uncontrolled outlets, 487 drainageway protection, 636
volume changes, for routing time steps, 510fig temporary channel diversions, 636
volume of water calculation (example), 62–63, 62fig temporary filters, 636
wet ponds, 480–481, 480fig, 623, 624fig temporary stream crossings, 636
See also structural best management practices Dredge and Fill Permit Program, 598–599
detention storage requirements dry detention pond, 480, 480fig
for pumping facilities, 561fig dry detention ponds, See also detention ponds
Baumgardner's technique for temporary-storage dry-weather flows
volume, 561fig considerations in design of open channels, 291
detention storage requirements for pumping facilities, See also base flow
560–563 dynamic effect of flow, 645
Baumgardner's technique to estimate storage dynamic wave formulation. See Saint-Venant equations
(example), 562
Baumgardner's technique to estimate temporary-
storage volume, 560–563, 561fig
developing countries
drainage management in, 580–581
urban drainage in, 578
development types, runoff coefficients for, 142t E
diffused-water rules, 599, 601
civil law rule, 601 EarthInfo, Inc., source of rainfall data, 80
common enemy rule, 601 economic analysis of culvert design, 325–326, 326fig
reasonable-use rule, 601 eddy losses, 231
diffusion wave formulation, 163 effective precipitation, 108, 117, 122fig
direct surface runoff, 116 hyetograph calculations (example), 132–134
creating hyetographs, 132, 158–162 See also direct surface runoff
674 Index
effective rainfall. See effective precipitation Environmental Protection Agency

effective rainfall, 119–121 EPA. See U.S. Environmental Protection Agency
effective rainfall intensity, calculation (example), 119 EPA runoff model, 175t
EGL. See energy grade line ephemeral stream, 165
elevation, 187 ephemeral streams, 165
elevation datum, 189, 189fig equations
elevation head, 56, 189, 189fig conservation of energy, 55–56
endwalls, for culverts. See headwalls conservation of mass, for incompressible fluids,
energy 52–53
in fluid systems, 55–56 conservation of momentum, 59
types of, 55 depth of runoff from a drainage basin, 62
See also conservation of energy; elevation head; detention pond water volume, 62–63, 62fig
energy equation; pressure head; velocity head energy conservation between two points, 63
energy dissipation momentum flow rate, 59, 63
impact-type dissipators at culvert outlets, 361 volumetric flux, 53
riprap blanket design at stormwater outfalls (ex- equivalent Manning roughness coefficient, for composite
ample), 469–472 roughness, 238–239
impact-type dissipators at storm sewer outlets, 464 equivalent pump performance curves, 555–556, 555fig,
source of design guidance, 464 556fig
energy equation, 55–56, 63, 186–187 comparison of pumps, 555–556, 556fig
for closed-conduit flow, 186–191, 189fig equivalent sand grain roughness, 193
applications (examples), 191 effects of pipe aging, 193
energy equation, 186 effects on value of pipe aging, 193
graphical representation, 188–191, 189fig values for commercial pipe materials, 193t
system curve development for pumping values for pipe materials, 193t
stations, 548–552 equivalent structure diameter, 436
velocity distribution coefficient, 187–188 erodible, 287
determining flow direction from (example), 57–58 erodible open channels, 219, 287, 297
for open-channel flow, 223–225, 224fig permissible velocity method
energy equation, 223–225, 224fig correction factor for permissible velocity, 299–
open-channel flow modifications, 223–225 301, 300fig
See also conservation of energy design (example), 302
energy grade line (EGL), 57fig, 189fig, 190 for design of, 297–302
in closed conduit flow, 189fig, 190 design procedure for a specific discharge, 301
in open-channel flow, 224fig maximum velocities for various channel materi-
energy gradient, 193 als, 298, 298t
Chézy equation for, 200 maximum velocity as a function of mean sedi-
of close-conduit flow, Chézy equation, 200 ment size, 298–300, 300fig
of closed-conduit flow permissible velocities for cohesive soils, 298,
Darcy-Weisbach equation, 193 299fig
Hazen-Williams equation, 200 tractive force method
Manning equation, 199 for design of, 303–308
Darcy-Weisbach equation for, 193 design of (example), 307–308
in Hazen-Williams equation, 201 maximum permissible shear stress, 303–306,
Manning equation for, 199, 226 304fig
of open-channel flow, 225–230 maximum permissible shear stress as a function
energy gradient of open-channel flow, Manning's equation of mean sediment size, 304fig
for (example), 228 maximum permissible shear stress for cohesive
energy head of pumps. See total dynamic head (TDH) soils, 305fig
energy losses shear stress on channel boundaries, 303–304,
in closed conduit flow 303fig, 304fig
defined, 187 See also open channels
general equation for, 187 erodible open-channel design
defined, 187 permissible velocity method (example), 302
general equation for, 187 tractive force method (example), 307–308
in open-channel flow, 225–226, 225–230 erosion control
equation for, 226 at construction sites, 635
See also frictional losses; head losses; minor losses at culvert outlets, 360–361, 361fig
energy-balance method for modeling snowmelt, 168–170 erosion control measures, 635
energy-loss method at culvert outlets, 360–361, 361fig
for minor losses in junction structures, 434–439 erosion-control blankets for grass channels, 314sb
storm sewer junction structures, benching correction Europe, sewer systems in, 584
factors, 438t European Commission, Directorate-General for
See also composite energy-loss method Development, 579
England European Union, 579
regional water authority system, 586 Dangerous Substances Directive, 585
stormwater management regulations in, 586–588 Eco-Management and Audit Scheme, 584
Urban Pollution Managment Manual, 587–588 environmental policies, 583–585, 603
Environment Australia, Urban Stormwater Initiative, 579 national river basin authorities, 585
environmental concerns with culverts, 362 regulatory approach to water management, 585–586,
environmental policies, international, 578–580 603
Environmental Protection Agency. See U.S. Water Framework Directive, 585–586
675
water management directives, 585 grassed open channels, 309

evaporation, 108, 108fig open channels, 289–290
evapotranspiration, 108fig friction factor
event frequency. See recurrence interval in closed-conduit flow
exeedence probability, 74–75 Darcy-Weisbach friction factor, 193
extended detention basins, 621–623, 622fig explicit equations for, 195–197
external forces, 460, 460sb Manning’s friction factor, 199, 199t
external forces on pipes, 460, 460sb Moody diagram for, 194, 196fig
Colebrook equation, 197
Darcy-Weisbach friction factor, 193
explicit equations for, 195–197
Manning’s friction factor, 199, 199t
Moody diagram for, 194, 196fig
F friction slope. See energy gradient

frictional forces, in open-channel flow, 231–232, 232fig
frictional losses, 192–201
failure probability, of hydraulic structure design, 149sb in closed-conduit flow, 192–201
fall of a depressed culvert barrel, 337–338, 337fig in open-channel flow, 225–230
Federal Highway Administration, Hydrologic Circular See also energy losses; head losses
No. 22, 144 Froude number
filter fabric used to prevent piping of soils, 468fig, 469 defining characteristics of hydraulic jumps,
filter strips. See grass buffer areas 261–263, 262fig, 263fig
firm capacity of a pumping installation, 564 in design of open channels, 289, 294
First Law of Thermodynamics, 55 in governing equation of gradually varied flow,
fish passage through culvert pipes, 362, 362fig 272
fittings of open-channel flow, 247
for pumping stations, 558 See also hydraulic depth
flanking inlets in a sag, 401
flap gates, 558
flexible connections, 558
Flood Damage Reduction Program (Canada), 590
flooding, monsoon waters, 581
floodplains, 240
flow. See closed-conduit flow; culvert flow; gutter flow; G
open-channel flow; pipe flow; unsteady flow
flow Ganguillet and Kutter’s equation, for Chézy’s resistance
conservation of energy and, 55–56 coefficient, 228
depth of, 220–221, 220fig Ganguillet and Kutter’s equation for Chézy’s C, 228
flow control sections gate valves, 558
in culverts, 335–340 gauged rainfall event, 73fig, 87fig
in open channels, 217 GEMS. See Global Environment Monitoring System
flow control sections in open channels, 217 Germany
flow control structures. See weirs stormwater fees in, 584
flow measurements in open channels, 252 stormwater management regulations in, 589
Parshall flume, 252 Global Environment Monitoring System (GEMS),
weirs, 252 Freshwater Quality Program, 608
flow profile analysis governing equation of gradually varied flow, 272
of culverts, 348–354 gradually varied flow in open channels, 271–277
of open channels with gradually varied flow, boundary conditions, 272
271–277 classification of flow profiles, 273–274, 274fig
flow velocity. See velocity computation of profiles, 275–277
flow velocity channel slope transition analysis (example),
determinining from conservation of mass, 54 276–277
for pipe cleansing, 423t direct-step method, 275–276, 279
in pipes, 56–57, 57fig standard-step method, 275
volumetric flux and, 53 use of computer programs for computations, 272
See also velocity water surface profile equation, 272
FlowMaster, 653t–654t grass buffer areas, 619–620
fluid pressure at depth in open-channel flow, 225 grass swales, 620, 620fig
fluid pressure at depth in open-channel flow grassed open channels, 287, 308
(equation), 225 design (example), 313
fluid systems, stored energy in, 55–56 design procedure, 312–313
fluids freeboard requirements, 309
conservation of mass and, 52–53 grass selection, 309
volumetric flux, 53 permissible velocities in, 309–311, 311t
fluvial geomorphology applied to channel modification, retardance of various grasses
316 side slope selection, 309
France, stormwater management in, 588–589 grasses
free body diagram of pipe, 557sb retardance levels, 310t
free outfalls, 437 retardance of, 308–311, 310t
freeboard in design of See also grassed open channels
detention ponds, 493 grate inlets, 377fig, 378fig
676 Index
in a sag, 390–392 head characteristics curves of pumps, 545, 546fig

capacity behaving as a weir, 390 head losses
capacity behaving as an orifice, 390 in closed-conduit flow
capacity (example), 391–392 calculation of total head loss (example),
curve vane, 380t, 383fig 207–208
on a continuous grade, 380–389 Darcy-Weisbach equation for head loss in a cir-
efficiency of, 388 cular conduit due to friction, 193
frontal flow, 386–389 Darcy-Weisbach equation for head loss in a
interception capacity, 386–389 noncircular conduit due to friction, 198
example, 389 explicit equations, 195–197
parallel bar, 380t, 381fig, 382fig general equation for, 187, 189fig
reticuline, 380t, 381fig, 386fig Manning equation for (example), 199–200
side flow, 386, 386–389, 387 in culverts, 334
slash-over velocity, 387, 388fig Darcy-Weisbach equation for head loss in a circular
tilt bar, 380t, 383fig, 384fig, 385fig conduit due to friction, 193
performance, 378 Darcy-Weisbach equation for head loss in a
grates, curved vane, 378 noncircular conduit due to friction, 198
gravitational force, 231–232, 232fig due to friction, Manning equation (example), 199–
Great Lakes, International Joint Commission, 592 200
Green-Ampt equation in full-flowing culverts (example), 207–208
for infiltration volumes (example), 114–116 general equation for, 187, 189fig
for modeling infiltration, 112–116 in open-channel flow
Green-Ampt equation for modeling infiltration, 112–114 Chézy equation, 228–230
assumptions, 113 equation for, 226
equation, 112 through storm sewer structures, 435
estimating infiltration parameters, 113–114t See also energy losses
for infiltration volumes (example), 114–116 headwalls
greenways. See stream buffers and greenways for culverts, 328, 328fig
greenways, 620, 621fig for storm sewer outfalls, 464, 465fig
groundwater laws, 599 HEC-22, 144
groundwater outflow, 116 HEC-RAS use in analysis of gradually varied flow, 272
gutter depressions, 393–394, 394fig Helsinki Convention on the Protection of the Marine
gutter flow, 370 Environment of the Baltic Sea, 582, 603
in composite sections, 373–375 Helsinki Rules on the Uses of International Rivers, 582
depth, 374 HGL. See hydraulic grade line; hydraulic grade line
gutter capacity (example), 375–376 (HGL)
Manning’s equation, 374 horizontal culvert alignment, 331
spread equation, 374 horizontal open channel, 219
cross slope guidelines, 371 Horton equation for modeling infiltration, 110–112
discharge fraction in depressed section, 375 estimating Horton infiltration parameters, 111
longitudinal slope guidelines, 371 infiltration rate calculation (example), 112
Manning’s n for, 272–273, 372t integrated version of method, 111
in shallow-swale triangular sections, 377 integrated version of method more accurate, 111
spread of, 370, 371t overestimates effective precipitation, 111
in uniform sections, 371–373 typical values of infiltration parameters for different
depth, 371 soils, 111fig
gutter capacity calculation (example), 373 values of infiltration parameters for soils, 111fig
Manning's equation, 372 hydraulic depth, 221, 222t, 247
spread equation, 372 See also Froude Number
gutter sections, 370, 370fig hydraulic depth of channels, 221
composite, 370, 370fig hydraulic forces in pipes
curved, 370, 370fig estimated by momentum equation, 208
uniform, 370, 370fig forces exerted on manholes (example), 210–211
gutters momentum equation for, 208
design flows for, 369 hydraulic grade line (HGL), 57fig, 189, 189fig, 224fig
V-shape, shallow-swale sections, 370, 370fig in closed conduits, 189, 189fig
See also roadway gutters hydraulic jump, 218, 260fig
characteristics in horizontal rectangular channels
(example), 263
length, 262fig
ratio of upstream and downstream depths, 262fig
hydraulic jumps, 260–263
H length, 261–263
ratio of upstream and downstream depths, 261–262,
262fig
Habitat II, 577–578 hydraulic radius
Hazen-Williams equation for flows in distribution of channels, 221
systems, 201 of pipes, 197–198
Hazen-Williams resistance coefficient, 201 hydraulic routing models, 174t
correction factor for rough pipes with high for unsteady flow, 646–647
velocities, 201 See also kinematic wave model
values for various pipe materials, 201t hydraulic structure design, failure probability, 149sb
677
hydraulic transients, 208 See also data sources

HYDRO-35 maps, 80, 81fig–85fig, 89, 91–92, 103 inlet boxes of storm sewer systems, 416, 417fig
hydrodynamic equations, for unsteady flow, 647–648 inlet design, 401
hydrograph estimation, 150–165 use of flanking inlets in a sag, 401
defined, 150 See also combination inlets; curb-opening inlets;
definition, 151fig grate inlets; inlet design flows; inlet locations;
direct runoff hydrograph computed from unit slotted-drain inlets
hydrograph and rainfall hyetograph, 158–162 inlet design flows, 401
direct runoff hydrograph estimation procedure based inlet flow control, of culverts, 335–338, 336fig
on unit hydrograph, 152 inlet locations, 400–401
kinematic wave model, 162–165 locations on grade, 401
NCS synthetic unit hydrograph, 155–158 minimum criteria, 400–401
synthetic unit hydrograph (example), 157–158 inlet openings, 377–378, 377fig, 378fig
unit hydrographs, 151–158 combination inlets, 377–379, 378fig
hydrographs, 150, 151fig continuous-grade, 377–378, 377fig
base flow contribution in, 165fig curb-opening, 377, 378fig
detention pond inflows and outflows, 54fig grate, 377, 378, 378fig
hydrologic cycle, 68–69, 68fig, 69fig sag, 377, 377fig
hydrologic models, for unsteady flow, 646, 648–650 slotted-drain, 377, 378fig, 379
hydrologic soil group, 122, 123sb inlet performance, 378–379
hydrologic soil groups, 120–121 bypass flow, 379
hydrology, of stormwater, 614–616 of combination inlets, 379
hydrosphere, 68 of curb-opening inlets, 378–379
hydrostatic pressure forces, 231–232, 232fig efficiency of flow interception, 379
of gate inlets, 378
slotted-drain inlets, 379
inlets, 377–378, 377fig
boxes, 377
defined, 377
I on grade, 377, 377fig

openings, 377
See also curb-opening inlets
IDF curves. See intensity-duration-frequency (IDF) curves intensity of rainfall. See rainfall intensity
Illinois State Water Survey intensity-duration-frequency (IDF) curves, 87–91, 87fig
Bulletin 71, 98, 101 adjustment of point depths and intensities for average
for design hyetograph distributions (example), 99– data, 92fig
101 adjustment of point depths and intensities for average
ILLUDAS method, for runoff hydrographs, 174t over drainage basin area, 92fig
ILLUDAS-SA method, for runoff hydrographs, 174t calculation (example), 90–92, 91fig
ILSAX method, for runoff hydrographs, 174t calculation using HYDRO-35, 89–91
impervious areas equation for, 87–88
directly connected, 130, 143 gauged rainfall event, 87fig
unconnected, 130 interpolation coefficients for calculation of, 90t
India interpolation coefficients for computation of, 90t
Indore, drainage management in, 580–581 interpolation of storm depths from HYDRO-35
Sturat, drainage management in, 581 maps, 89
infiltration, 108, 108fig, 110–116, 143–144 sources of data, 89
modeled by the Green-Ampt equation, 112–114 use of curves in the rational method of peak flow
modeled by the Horton equation, 110–112 estimation, 143
infiltration capacity, 110–116 use of in the rational method of peak flow estimation,
impact of soil type, 110 143
infiltration rates, rainfall intensity and, 111fig interception, 108–109, 108fig
information resources interflow, 116, 165
Charted Institution of Water and Environmental intermittent stream. See ephemeral stream
Management, 588 internal forces. See internal pressure forces
Construction Industry Research and Information internal pressure energy. See pressure head
Association (CIRIA), 588 internal pressure forces, 460, 557sb
International Network for Environmental free-body diagram, 557sb
Management, 584 tensile stress, 557sb
from U.N. Environment Programme, 579 equation for thick walled pipes, 557sb
information sources equation for thin walled pipes, 557sb
best management practices, 637sb internal pressure forces in pipes, 557sb
energy dissipators design guidance, 464 international development aid, policies and guidelines,
“EPA National Menu of Best Management Practices 578–580
for Storm Water Phase II,” 617 International Environmental Technology Centre, 579
National Stormwater BMP Database, 617–618 international watercourses, regulation of, 582–583
pipe strength guidelines, 460sb irrigation ditches, discharge from, 597
routing model descriptions, 461
soils data, 123sb
standards, 409–412
stormwater detention references, 479
stormwater management, 637sb
678 Index
J M
junction structures maintenance in design of open channels, 290
minor losses in, 434–444 manholes
in storm sewers, 418 minor losses of, 204
spacing criteria, 416–418
in stormwater systems, 416–418, 419fig, 420fig
manifold for multiple pumps, 555
Manning equation
for closed-conduit flow, 163–165, 198–199
K energy gradient equation, 199

flow velocity equation, 198–199
head loss in a culvert (example), 199–200
kinematic wave model, 162–165, 175t for open-channel flow, 225–228
discrete convolution process, 158–162 energy gradient equation, 228
for hydraulic routing of runoff hydrographs, energy gradient (example), 228
162–165, 175t Manning equation for flow, 163–165, 198–199
idealization of surface drainage, 162fig energy gradient equation, 199
Saint-Venant equations with kinematic wave flow velocity equation for, 198–199
simplification, 163–165 head loss in a culvert flowing full (example), 199–
time-to-peak equation, 155 200
for unsteady flow, 648 Manning’s roughness coefficient, 199
See also hydraulic routing models channel roughness height relationship, 226–228
kinetic energy, 55 for channel types and surfaces, 226–228, 227t
See also velocity head for corrugated-steel pipes, 414–415, 415t
Kutter’s n, 228–229 equivalent n value for composite roughness, 238–240
flow in a rehabilitated channel (example),
239–240
hydraulic radius and, 308
n, values for different pipe materials, 199t
for pipe materials, 199, 199t
L relation to channel roughness height, 226–228

values for different pipe materials, 199t
velocity and, 308
lag time. See drainage basin lag time Manning’s roughness coefficient n
laminar flow, 187, 188fig, 188t Mar del Plata Conference, 576t, 577, 603
land use mass. See conservation of mass
composite runoff coefficient, 141 measurement, unit conversion factors, 643, 643t, 644t
runoff coefficients for, 142t measurement units, for stormwater applications, 61, 61t,
land use/land cover impact on direct runoff, 122 63
laws of conservation. See conservation of energy; meltwater rate, 167–168
conservation of mass; conservation of momentum metamorphism of snow, 167
lined channels, riprap, 315 minor loss coefficients
lined open channels, 219, 287 for gradual contractions in pipe flow, 205fig, 207t
design (example), 296–297 for gradual expansions in pipe flow, 205fig, 206t
design procedure, 294 for manholes, 204
estimate of freeboard height, 296 for open-channel flow, 230–231, 231t
estimate of freeboard height at bends, 296 for pipe and culvert outlets, 204
velocity recommendations for concrete linings, 294 for pipe bends with various deflection angles, 204,
lining material considerations in design of open channels, 204fig
290 classic method, 203
linings of open channels, maximum allowable shear composite energy-loss method, correction fac-
stresses, 314sb tors, 441–444
Lloyd-Davies method. See rational method energy-loss method, 435–436, 436fig
loads on pipes. See structural loads on pipes energy-loss method correction factors, 436,
Local Agenda 21, 576t, 577 438fig
See also Agenda 21 for pipe and culvert entrances operating under outlet
local losses. See minor losses control, 203t
longitudinal slopes for sudden contractions in pipe flow, 205, 205fig,
of channels, 219 207t
in design of open channels, 289 for sudden expansions in pipe flow, 205fig, 205t,
of roadway gutters, 371 206t
looped rating curves, 645 of storm sewer junction structures
minor losses, 202–208, 225
at bends and appurtenances, 202–208
calculating friction, entrance, and exit losses for a
culvert flowing full (example), 207–208
classical method of estimation, 203
in closed-conduit flow, 202–208
679
calculation of losses for a culvert (example),
N
207–208
classical method of estimation, 203
of energy losses at bends and appurtenances,
202–208 nappe, of sharp-crested weirs, 266
at manholes, 204 Nash storage routing model, 175t
at pipe and culvert entrances, 203, 203t National Climatic Data Center (NCDC), 80
at pipe and culvert outlets, 204 National Engineering Handbook, 120–121, 122, 123, 124,
at pipe bends, 204, 204fig 645
eddy losses, 231 National Oceanic and Atmospheric Administration
entrance losses for full-flowing culverts (example), (NOAA), 80
207–208 Atlas 2, 80, 91
exit losses for full-flowing culverts (example), See also National Weather Service
207–208 National Pollution Discharge Elimination System
friction losses for full-flowing culverts (example), (NPDES). See NPDES permit program
207–208 National Stormwater BMP Database, 617–618
at manholes, 204 National Weather Service (NWS), 80
in open-channel flow, 225, 230–231 HYDRO-35 Maps, 80, 81fig–85fig, 89, 91, 103
equation for, 230 source of rainfall data, 80
at pipe and culvert entrances, 203, 203t Technical Publication 40, 80, 91
at pipe and culvert outlets, 204 natural channels, 316–317
at pipe bends, 204, 204fig Natural Resources Conservation Service (NRCS)
in storm sewer junction structures, 434–444 curve number method. See curve number method
classic method, 202–208 rainfall distribution factors by storm type, 93–94t
composite energy-loss method, 440–444 synthetic rainfall temporal distributions, 93–96
coefficients, 440–441 Technical Release 55, 122, 124, 139, 144
corrections for access-hole water depth, Natural Resources Conservation Service (NRCS) curve
441–442 number method. See curve number method
corrections for lateral flow, lateral angle, Natural Resources Defense Council, 597
and plunging flow, 441–443 NCDC. See National Climatic Data Center
corrections for relative access-hole diame- net positive suction head (NPSH) of pumps, 547
ter, 441 equation for, 553
corrections for relative pipe diameter, The Netherlands, stormwater management regulations
443–444 in, 589
lateral angle correction, 443 The New Water Resources Strategy (World Bank), 578
lateral flow correction, 442–443 NOAA. See National Oceanic and Atmospheric
plunging flow correction, 442 Administration
energy-loss method, 434–440 nomographs
coefficients, 435–436 to calculate head on culvert with outlet control, 341–
corrections for benching of structure 346, 343fig
invert, 437–438, 438fig, 438t to calculate headwater depth in culverts with inlet
corrections for flow depth, 436–437 control, 340, 342fig
corrections for plunging flow, 437 nonpoint source pollution, 593, 607
corrections for relative flow, 437 nonprismatic open channels, 218–219, 219fig, 220fig
corrections for relative pipe diameter, 436 eddy losses, 231
energy-loss method (example), 438–440 nonstructural best management practices, 617, 628–630
See also energy losses catch basin cleaning, 629
mitered culvert ends, 327, 328fig control of inappropriate discharges into storm
modeled by the Green-Ampt equation, 112–114, 112–116 drainage systems, 629
momentum. See conservation of momentum inappropriate discharge control, 629
momentum equation, 257 lack of performance evaluations, 628
in closed-conduit flow public education programs, 630
general form, 208–209 street cleaning, 629
for steady-state flow, 209–210 types of, 628
general form, 208–209, 256fig See also best management practices, stormwater
for steady-state flow, 209–210 quality management
See also conservation of momentum normal depth
momentum flow rate, 59, 63 of gradually varied flow, 273
Moody diagram, 194, 196fig of uniform flow, 232, 233
use to obtain friction factor, 194, 196fig normal slope of uniform flow, 233
Muskingum routing method, 170–173, 653t North Carolina, Fletcher, greenway master plan, 621fig
routing a runoff hydrograph (example), 172–173 NPDES permit program, 575, 595–598, 601, 603
for unsteady flow, 648–649 NPSH. See net positive suction head (NPSH) of pumps
Muskingum-Cunge Routing method, 649 NRCS. See Natural Resources Conservation Service
NRCS curve number method. See curve number method
NRCS peak discharge method, 144–148, 174t
adjustment of peak discharge for temporary storage,
145–148
calculation (example), 148
conditions for use, 144
equation for peak discharge, 144
680 Index
equation for unit peak runoff rate, 145 velocity of flows, 288
values of coefficients for peak discharge equation for See also erodible open channels; grassed open
storm types and selected ratios of initial rainfall channels; lined open channels
abstractions to 24-hr rainfall depths, 146t open-channel flow
values of the ratio of initial rainfall abstractions to 24- channel transitions, force exerted by (example), 264
hr rainfall depth for selected rainfall depths and Chézy equation, 228–230
curve numbers, 147t in closed conduits, 185, 185fig
NRCS synthetic unit hydrograph. See synthetic unit control sections, 253
hydrographs energy equation for, 223–225, 224fig
NWS. See National Weather Service energy gradient, 225–230
Chézy equation for (example), 230
Manning equation for (example), 228
fluid pressure at depth (equation), 225
forces on submerged structures, 263–264
calculation (example), 264
O gradually varied flows, 271–277

head losses due to friction, 225–230
hydraulic jumps, 260–263
Ontario hydrostatic pressure force
Conservation Authorities, 591, 592 pressure variations over spillways (example),
stormwater management regulations in, 591–592 225
water quality control, 592 variations in, 225
See also Canada at a cross section, 257–258
open channel equation, 225
adverse, 219 Manning equation, 225–228
angle of inclination, 219 minor losses in, 230–231, 231t
best hydraulic section, 291–293, 292fig momentum and forces, 255–264, 256fig
circular, 221, 222fig orifices, 270–271
classification, 218–219 pressure variations in, 225
common shapes, 221, 222fig pressure variations over spillways (example), 225
cross-sectional shapes. See open-channel geometric specific forces, 258–259
properties in trapezoidal channels (example), 259
erodible, 287 subcritical, 246
geometric properties of, 219–223, 222fig, 222t transitions, 252–255, 253fig
grassed, 287, 308 attainability concept, 253–254
horizontal, 219 flow depth in rectangular channels with bottom
hydraulic depth, 221 transition (example), 255
hydraulic radius, 221 uniform flow, 226, 231–234
lined, 219, 287, 294–297 varied flow, 226
longitudinal slope, 219 See also critical open-channel flow; gradually varied
nonprismatic, 219, 219fig, 220fig flow; spatially varied flow; uniform open-channel
prismatic, 218–219, 219fig, 222t flow
rectangular, 221, 222fig, 222t open-channel geometric properties, 219–223
retardance of grasses in, 308–309, 310t cross-sectional properties of prismatic channels, 222t
shapes of, 295fig cross-sectional shapes, 221, 222fig
slope classification of, 273 inclination angle, 219
trapezoidal, 221, 221–223, 222fig, 222t longitudinal slope, 219, 220fig
geometric properties (example), 223 wetted perimeter, 221, 222t
triangular, 221, 222fig, 222t open-channel routing of runoff hydrograph, 169–173
unlined, 219 Muskingum routing method, 170–173
See also erodible open channels open-channel stabilization, lining options, 314sb
open-channel design, 287 operating point of pumps, 513fig, 552–553
aesthetics, 290–292 orifices, 270–271
best hydraulic section, 291–293, 292fig discharge coefficient, 271
calculation (example), 292–293 discharge equation, 270–271
channel modification flow diagram, 271fig
application of fluvial geomorphology, 316–317 outfalls of storm sewers, 407
channelization to increase capacity, 315 outlet structures
of natural channels, 316–317 cradles, 329–330, 330fig
other channel improvements, 315–316, 316fig for detention ponds, 494–495
channel modifications, 315–317 anti-seepage collars, 495
considerations and issues, 288–291 anti-vortex devices, 495
cost and maintenance, 290 composite structure, 494, 494fig
dry-weather flows, 291 orifices, 495
freeboard, 289–290 overflow weirs, 495
Froude number, 289 pipes, 494
lining material, 290 riser pipes, 495
lining options, 314sb spillways, 496
longitudinal slopes, 289 for stormwater sewers, 420
right-of-way requirements, 291 backup of flow prevention, 418
safety, 290 energy dissipators, 420
side slopes, 289 stilling basins, 420
681
See also detention pond outlets measuring flow in, 56–57, 57fig
outlet velocity of culverts. See culvert outlet velocity minimum slopes for cleansing flow velocities, 423t
overbank area of channel, 240 stresses in, 460
overland flow velocities, estimate as a function of land use structural loads on, 460
and slope, 138fig See also concrete pipes; corrugated-steel pipes;
overtopping of roadways. See roadway overtopping culverts; storm sewer pipes
point source pollution, 593, 607
urban runoff as, 597–598
pollutants. See urban stormwater pollutants
pollution prevention. See nonstructural best management
practices
P PondPack, 653t–654t
application to detention pond analysis, 518sb–519sb
porous landscape detention, 621
Parschall flumes, 252 porous pavement, 618, 618fig
pavement porous pavement detention, 618, 618–619, 619fig
porous, 618, 618fig potential energy. See elevation head
porous materials for detention, 618–619, 619fig Powell’s equation for Chézy’s C, 229–230, 230t
peak discharge equation, unit hydrograph, 156 values for Powell’s epsilon, 230t
peak discharge method, 174t precipitation, 68–69
adjustment of peak discharge for temporary storage, See also rainfall; snowmelt
148t precipitation depth
coefficients for peak discharge equation for storm maps, 81fig–85fig
types and selected ratios of initial rainfall maps of, 80
abstractions to 24-hr. rainfall depths, 146t precipitation types, 69, 70–71
conditions for use, 144 convective, 71, 72fig
equation for peak discharge, 144 cyclonic, 70–71, 71fig
equation for unit peak runoff rate, 145 frontal, 70, 70fig
ratios of initial rainfall abstractions to 24-hr. rainfall orographic, 71, 72fig
depth for selected rainfall depths and curve pressure, 187
numbers, 147t pressure forces in open-channel flow, 257–258
peak flow estimation, 140 pressure gauges, in pumping stations, 558
comparison of models, 174t pressure head, 56
NRCS method, 144–148 in closed conduits, 189, 189fig
rational method, 140–144 in open channels, 224fig, 225
USGS regression-based formulas, 148–150 pressure variations, on spillways (example), 225
perennial stream, 165 prismatic open channels, 218, 219fig, 295fig
performance curves cross-sectional properties, 295t
of culverts, 359–360 flow profile types, 274fig
of pumps, 545–547 properties of, 222t, 295t
permanent armoring of channels, 314sb shapes, 295fig
permanent turf reinforcement of channels, 314sb profile analysis. See flow profile analysis
permissible shear stresses on channel linings, 314sb projecting end of culvert, 327, 328fig
permissible velocities in open channels. See velocity public education programs, 630
piezometric head, 57fig, 189 public health, inadequate drainage and, 581
equation for, 189 pump cavitation, 553–554
pipe bends determining pump location limits (example), 554
force exerted on (example), 60–61, 60fig pump controls, 559–560
minor losses, 204 common off-elevation schemes, 559, 559fig
pipe cleaning velocities in storm sewers, 422, 423t minimum cycling times, 560, 560t
pipe deflection angle successive start/stop schemes, 559, 559fig
effects on minor losses, 204, 204fig pump cycling times, 560, 560t
minor loss coefficient of, 204, 204fig pump efficiency
pipe entrances, minor losses, 203, 203t equation for, 545
pipe flow relation to specific speed and discharge of pumps,
gradual contraction, 206t 538, 539fig
gradual expansion, 205 pump installations. See pumping stations
momentum and forces in, 208–211 pump operating characteristics
sudden contractions, 205 affinity laws, 547
sudden expansions, 202, 205 brake power, 546–547
transitions, 204–208 comparison of performance of multiple pumps, 555–
See also closed-conduit flow 556, 556fig
pipe friction, 187, 192–201 determination of water power and brake power
pipe joints required (example), 547
for concrete pipes, 412 efficiency, 538–539, 539fig, 545
for corrugated-steel pipes, 414 equivalent performance curves for multiple pumps,
pipe outlets, minor losses, 204 555, 555fig
pipe transitions, minor losses of, 204–205 head characteristic curves, 545–547
pipes net positive suction head (NPSH), 547
equivalent sand grain roughness for various pipe operating points, 513fig, 552–553, 555fig
materials, 193t performance curves, 545–547
internal pressure forces, 557sb total dynamic head (TDH), 545–547
682 Index
water power, 546 radial-flow pumps, 537fig

pump performance, 545–556 RAFTS storage routing model, 175t
test results, 548fig rain barrels, 626
See also pump operating characteristics rain gauges, 76–78
pump performance curves, 545–547 float, 76
pump performance tests, 548fig measurement errors, 78
pump piping requirements, 556–558 optical, 76–77, 78fig
system curves, 548–552 standard nonrecording, 76, 77fig
valves and fittings, 558 tipping-bucket, 76, 78fig
pump speed. See specific speed of pumps weighing gauges, 76, 77fig
pumping facilities, detention storage requirements, 561fig See also rainfall data
pumping station design, 563–569 rainfall, 69
architectural considerations, 556 abstractions, 108–116
design inflow hydrograph estimate, 563 data collection methods, 76–79
doorways, 569 See also rain gauges
firm capacities, 564 data types, 87–98
flow routing, Runge-Kutta method, 565 depth, 72, 73, 73fig
grit chambers, 568, 569 See also precipitation depth
hatches, 569 event and runoff cumulative probability
monitoring devices, 569 distributions, 614–615, 615fig
monorails, 569 hyetographs, 74fig
power, 567 depth, 74fig
safety, 569 intensity, 74fig
stage-discharge relationship of storage facility, 565 intensity, 73, 73fig
stage-water surface-area relationship of storage losses. See rainfall abstractions
facility, 565 spatial distribution, 74
trash racks, 567, 567–568 temporal distribution, 73, 73fig, 87fig
trial and error process, 563–564 selecting storm duration in constructing synthet-
trial facility design evaluation, 564–565 ic distribution, 92–93
trial facility design process, 563–564 selecting time increment in constructing syn-
ventilation, 569 thetic distribution, 92, 92–100, 93–94
pumping stations, 540–544 selecting time increment in constructing syn-
controls, 559–560, 559fig thetic distribution (example), 95–96
with multiple pumps, 555–556 synthetic distribution construction, 91–101,
pipe requirements, 555–557 92–100
station equipment layout, 558fig time series, 75sb
sump design, 540–544, 543fig, 544fig See also precipitation
system curve development (example), 550–552 rainfall abstractions, 108–116
system curves, 548–552 rainfall characteristics, 72–75
wet/dry-wells, 541, 542fig rainfall data
wet-wells, 540, 540fig collection methods, 76–79
See also detention storage requirements for pumping radar measurements, 79, 79fig
facilities; pump operating characteristics; pumping sources, 79–80
station design; valves and fittings time series, 75sb
pumps, 536–539 See also rain gauges
centrifugal, 536–538 rainfall depth. See precipitation depth
axial-flow, 536–537 rainfall hyetographs, 73, 74fig
mixed-flow, 536–537, 538 depth, 74fig
radial-flow, 536, 537, 537fig intensity, 74fig
operating points. See pump operating characteristics See also design storm hyetographs
performance. See pump operating characteristics rainfall intensity, 73fig
performance curves, 546fig rainfall losses. See rainfall abstractions
See also pump operating characteristics rainfall recurrence intervals, 74–75
positive-displacement, 538 calculating (example), 75
screw, 538 rainfall spatial distribution, 74
relation of specific speed to pump efficiency and rainfall temporal distribution, 73fig, 87fig
discharge, 538, 539fig selecting storm duration in constructing synthetic
selecting pump types (example), 539 distribution, 92–93
specific speed, 538–539 selecting time increment in constructing synthetic
specific speed factor in selecting (example), 539 distribution, 92
submersible, 540 synthetic distribution construction, 92–96, 98–101
system curves. See pumping stations rainstorms, exceedence probability, 74–75
wire-to-water efficiency of, 545 Ramsar Convention, 583
rating curves, in uniform flow, 645
rating curves for discharge of detention ponds. See under
stage vs. discharge
rational method
land use coefficients, 142t
R peak flow calculation (example), 143

rational method of peak flow estimation, 89, 140–144,
174t
radar measurements, of precipitation, 79, 79fig assumptions, 120–121, 141, 142
683
equation for, 141 357fig

limitations of application, 141–142 velocity calculation (example), 358–359
problem of under-predicting peak flow, 143–144 RORB storage routing model, 175t
runoff coefficient weighted by land use/land cover routing methods
(example), 141 hydraulic, 646
runoff coefficients, development types and, 141, 142t hydrologic, 646
steps in applying method, 123–124, 143 unsteady flow and, 650
under-predicting peak flows, 143–144 See also storage routing models; names of specific
rectangular, weirs, 265fig methods or models
recurrence interval, 74–75 Runge-Kutta routing method for pumping facilities, 565
calculation of (example), 75 Runge-Kutter algorithm, applied to storage routing, 511
equation, 75 runoff, water problems in urban basins, 609t
selection in hydraulic structure design, 149sb runoff coefficient, 118–119
reducers, for pumping stations, 558 composite, 141
regulations. See under water law (U.S.) for effective rainfall intensity (example), 119
reinforced-concrete pipe. See concrete pipe equation for, 119
relative roughness of pipes for peak flow estimation in rational method, 141, 143
defined for circular conduits, 193 weighted by land use/land cover (example), 141
modified for noncircular conduits, 198 runoff coefficients
reservoir routing, 173 equation for, 119
restraints for given development types, 142t
at pipe bends and other appurtenances, 208 use of in rational method to estimate peak flow, 143
retention ponds, 479 runoff hydrographs, time-area methods compared, 174t
return period. See recurrence interval See also TRRL method
Reynolds number (Re), 187, 192 runoff models, 107–108
for closed-conduit flow compared, 174–175t, 653–654t
Darcy-Weisbach equation, 192–197 general hydrograph methods, 102
equation for circular conduits, 192 generalized unit hydrograph methods, 102
modified for noncircular conduits, 198 NRCS peak discharge method, 102, 144–148
relation to the Darcy-Weisbach equation, NRCS unit hydrograph, 102, 155–158
193–194 peak discharge method, 102, 144–148
use in Colebrook equation, 195–197 rainfall information requirements, 101–102
use in Moody diagram, 194 rational method, 101, 101–102, 140–144
Darcy-Weisbach equation, 192–197 unit hydrograph methods, 102
for open-channel flow, equation, 229, 267 runoff potential from snowmelt, 167
velocity distribution values, 188t
Rhine River, 582–583
right-of-way requirement considerations for open
channels, 291
Rio de Janeiro conference. See United Nations,
Conference on Environment and Development
riprap S
at culvert inlets, 361
at culvert outlets, 360–361, 361fig safety considerations in design of open channels, 290
at storm sewer outfalls, 464 sag inlets, 377, 377fig
riprap apron. See riprap erosion protection guidance Saint-Venant equations, 163–165, 170, 653–654t
riprap blankets, 361 sand filter extended detention basins, 623
riprap erosion protection, prevention of piping using filter sanitation
fabric, 469 in developing countries, 578–579
riprap erosion protection guidance, 464–472 inadequate drainage and, 581
adjustments for supercritical flow, 465 SBUH storage routing model, 175t
apron configuration for trapezoidal channel, 464– Scotland
465, 465fig stormwater management regulations in, 588
conditions for application, 464 Sustainable Urban Drainage Systems (SUDS), 588
expansion factor for outlet jet, 467–468 scour channels. See erodible open channels
for circular conduits, 468fig SCS (Soil Conservation Service). See Natural Resources
for rectangular conduits, 468fig Conservation Service (NRCS)
length of riprap apron, 467–469 SCS curve number method. See curve number method
rock size in riprap layer at outlets, 465–467 sediment basin, 636
size of rock in riprap layer at outlets sediment control, at construction sites, 635–636
used for circular conduits, 466fig, 467t sediment control measures, 635–636
used for circular rectangular conduits, 466fig, check dams, 635
467t diversions, 635
riprap lined channels, 315 sediment basins, 636
See also permanent armoring silt fences, 635–636
rise of rectangular pipes, 412 slope drains, 635
risers in manhole assemblies, 418, 419fig temporary sediment traps, 636
roadway gutters sediment transport in open channels, 297
longitudinal slopes, 371 sediment traps, 636
See also gutters sediments, angles of repose for, 306–307, 307fig
roadway overtopping, 355–359 shallow swale sections, 370, 370fig
culvert discharge calculation, 355–357, 356fig, circular swales, 370, 370fig
684 Index
V-shape gutters, 370, 370fig composite relationship for outlet structures in

V-shape median swales, 370, 370fig series, 509, 509fig, 510fig
sharp-crested weirs, 266–269, 266fig rating curve development (example), 507–508
nappe, 266 pumping station wells, 565
sharp-crested, 266fig stage vs. storage volume of detention ponds, 499–505,
triangular, discharge coefficient values, 269fig 500fig
shear stress in open channels, 232, 303–306 average-end-area method, 500–502
equation for, 231–232, 303 average-end-area method (example), 501–502
See also tractive force conic method, 502–504, 503fig
sheetflow at outlets of detention ponds, 496 numerical integration method, 504–505
shutoff valves, 558 stage vs. surface area of detention ponds, 497–499
side slopes, 289 calculation (example), 498–499
maximum values for various materials, 301t standards for water management
selection for grassed open channels, 309 Eco-Management and Audit Scheme (European
silt fences, 635–636 Union), 584
silting channels, 297 ISO 14000, 584
slope drain for sediment control, 635 Urban Pollution Managment Manual (U.K.),
slope paving at culvert inlets, 361 587–588
slotted-drain inlets, 377, 378fig, 398–399, 399fig stemflow, 69fig
in a sag stilling basins at culvert outlets, 361
capacity behaving as a weir, 399 storage indication curves, 513
capacity behaving as an orifice, 399 storage indication routing method, 509–524
on a continuous grade, inception capacity, 398 analysis procedure, 513–514
snow, metamorphism of, 167 example, 514–517
snow water equivalent, 167 storage routing models, 175t
snowmelt, 167–170 See also routing methods; Runge-Kutta routing
energy-balance method of modeling, 168–170 method; storage indication routing method
models, 166–179 storm sewer alignments, 407–408, 408fig, 421–423
degree-day method, 167–168 curved, 421
energy-balance method, 168–170 horizontal, 421
runoff potential, 167 separation from portable water pipes, 421
snow water equivalent calculation, 167 location of, 421
snowmelt modeling, degree-day method, 167–168 maximum deflection angle guidance, 421
Snyder (Alemeda) unit hydrograph method, 174t, 176t vertical, 421–422
Soil Conservation Service (SCS). See Natural Resources separation from utilities, 421–422
Conservation Service (NRCS) See also storm sewer slopes
soil materials. See sediments storm sewer appurtenances, 416–420
soil moisture. See antecedent soil moisture conditions catch basins, 417fig
soils data, sources, 123sb inlet boxes, 416, 417fig
source area biofiltration devices, 625–626 junction structures, 418
source controls. See nonstructural best management manholes, 416–418, 419fig, 420fig
practices bases, 419fig
Source Loading and Management Model (SLAMM), 614 eccentric cones, 419fig
span of rectangular pipes, 412 grade rings, 419fig
spatially varied flow in open channels, 277–279 risers, 418, 419fig
with decreasing discharge, 279 spacing of, 418t
equation for, 279 transition sections, 419fig
with increasing discharge, 278–279 outlet structures, 420
equation for, 278 underground detention, 420
location of control section, 278 water-quality enhancement structures, 420
specific energy, 244–248 storm sewer design using computer models, 470sb
calculation for rectangular channel (example), 246, storm sewer outfall design, 464–472
247–248 See also energy dissipation
of channel transitions, 252–255, 253fig storm sewer pipe design, 407–408
curves, 245–246, 245fig alignment, 421–422
equation for, 244 appurtenances, 416–420
specific force design storm return periods, 425–426
curves, 259 directly and indirectly connected areas, 426
in open channels, 258–259 energy dissipation. See energy dissipation
specific speed of pumps, 538–539 external forces, 460sb
equations for, 538, 538–539 flow routing in storm sewers, 460–461
relation to pump efficiency and discharge, 538, source of information on models, 461
539fig minor losses in junction structures. See minor losses
spill containment and control, 625 in storm sewer junction structures
spillways outfall design. See storm sewer outfall design
for detention ponds, 496 pipes, 409–414
pressure variations on (example), 225 structural loads and stresses, 460
stage relationships for detention ponds, 497–509 See also storm sewer pipe final design; storm sewer
stage vs. discharge pipe initial design
detention ponds, 505–510, 505fig storm sewer pipe final design
composite relationship for outlet structures in minor losses in junction structures, 434–444
parallel, 506–508, 506fig, 507fig open-channel flow, 452–459
685
design process (example), 453–459 motor control center, 541fig

design steps, 452–453 trash racks, 568fig
performance testing, 444–459 See also pump operating characteristics
pressure flow, 444–451 stormwater quality management
design process (example), 446–451 best management practices (BMPs), 617–628
design steps, 445–446 application, 630–636
surcharged pipes, 444–451 impacts of stormwater runoff on streams, 612–613
See also storm sewer pipe design importance of small runoff events, 614–615
storm sewer pipe initial design pollutants in urban stormwater, 609–611
design process (example), 427–433 program goals, 613
design steps (in example), 429–433 receiving-water goals, 613
directly and indirectly connected areas, 426 runoff impacts on streams, 612–613
energy dissipation. See energy dissipation small runoff events’ importance, 614–615
flows using rational method (example), 424–425 water quality impacts of stormwater runoff, 608–609,
use of multiple design storm return periods for 608–613
different land uses, 425–426 water quality treatment design criteria, 615–616
See also storm sewer pipe design See also best management practices; nonstructural
storm sewer pipes, 409–414 best management practices; structural best
standard specifications, 409–414 management practices
See also concrete pipes; corrugated-steel pipes stormwater retention. See retention ponds
storm sewer slopes, 422 stormwater runoff. See runoff
minimum pipe cleansing flow velocities, 422, 423t stream buffers and greenways, 620
storm sewer systems, 369 street cleaning, 629
appurtenances. See storm sewer appurtenances stress in pipes, 460
catch basins, 416 tensile stress from internal forces equation for thick-
outfall design, 464–472 walled pipes, 557sb
pipe system design. See storm sewer pipe design tensile stress from internal forces equation for thin
pipes. See storm sewer pipes walled pipes, 557sb
storm sewers structural best management practices, 617, 618–628
alignment, 421–423 cisterns and rain barrels, 626
combined sewers, 409 constructed wetland basins, 623–624
Storm Water Management Model (SWMM), 461 constructed wetland channels, 625
StormCAD, 653t–654t covering storage and handling areas, 625
for analysis of gradually varied flow, 272 critical source area control, 625
application to storm sewer design, 470sb effectiveness in treating stormwater, 626–628, 628t
stormwater extended detention basins, 621–623, 622fig
control programs grass buffer areas, 619–620
goals, 613 grass swales, 620, 620fig
receiving water goals, 613 greenways, 620
detention. See detention ponds; detention storage porous landscape detention, 621
requirements for pumping facilities porous pavement, 618, 618fig
detention analysis procedure, 513–514 porous pavement detention, 618, 619fig
detention models. See detention ponds rain barrels, 626
hydrology, 613, 614–616 receiving-water problems, rain-depth and, 614–615
storm magnitudes compared, 489sb sand filter extended detention basins, 623
inlets. See inlets source area biofiltration devices, 625–626
models. See runoff models spill containment and control, 625
pumping, 533–534 stream buffers and greenways, 620
alternatives to, 534 wet detention ponds, 623, 624fig
applications of, 535fig See also best management practices; detention
design considerations, 534, 536, 566–569 ponds; stormwater quality management
pump selection, 538–539 structural loads in pipes, 460, 460sb
example, 539 structural loads on pipes, 460, 460sb
pump types, 536–538 structural loads within pipes, 460
pumping stations, 540–541, 540fig, 542fig submersible pumps, 540
station design, 563–565 subsurface flow, 165
storage requirements, 560–563 See also base flow
sump design, 540–544, 543fig, 544fig subsurface storm flow, 116
stormwater inlets. See inlets successive start/stop pump control schemes, 559, 559fig
stormwater management sump design, 540–544, 543fig, 544fig
climate change and, 583 sumps. See catch basins
information sources, 637sb supercritical open-channel flow, 246
international aspects, 575 superelevation of water surface at bends in open channels,
stormwater modeling. See runoff models 296
Stormwater Phase II Rule, 596–597 suppressed weirs, 265fig
stormwater pumping surface runoff, 110, 165
applications, 533, 534–536 See also effective precipitation
through a tide gate, 535fig sustainable development, 575
stormwater pumping stations, 566fig Sustainable Urban Drainage Systems Working Party, 588
alternatives to, 536 sustainable water resource management, 579
control panel wiring, 568 swales. See shallow swale sections
inlets, 544fig Swamee-Jain equation
686 Index
for pipe diameter, 197 TRRL method (for runoff hydrographs), 174t
for pipe flow rate, 197 modified method, 174t
See also Colebrook equation; Darcy-Weisbach turbulent flow, 187, 188fig, 188t, 192
equation
Sweden, stormwater fees in, 584
synthetic unit hydrographs, 155–158
developing (example), 157–158
dimensionless, 155–156, 155fig, 156fig
peak discharge equation, 156
time-to-peak equation, 155 U
system curves for pumping stations, 548–552
application of energy equation, 548–549, 549fig uniform flow, 188fig, 188t
curve development (example), 550–552 rating curves, 645
use of Darcy-Weisbach equation, 549 uniform flow profiles
in closed conduits, 188fig, 188t
in open-channel flow, 231–232
See also varied flow profiles
for prismatic open channels, 274fig
in simple channels, normal slope, 233
T uniform open-channel flow, 226, 231–234
in channels with composite roughness, 238
flow in rehabilitated channel (example), 239–
tailwater 240
depth in a culvert with submerged inlet and outlet, in compound channel cross sections, 240–244,
342–344 240fig
flow controls for a circular culvert (example), in compound channels
344–345 velocity, flow depth, and velocity distribution
flow controls for a circular culvert with smaller coefficients (example), 243–244
tailwater depths (example), 346 defined, 231, 232fig
tank storage routing model, 175t discharge, Manning equation (example), 234
TDH. See total dynamic head (TDH) normal depth, Manning equation (example), 234
temperature-index method, for modeling snowmelt, normal slope, Manning equation (example), 234
167–168 in pipes, 235–237
temporary filters for drainageway protection, 636 dimensionless graph of discharge and velocity
tensile stress in pipes vs. relative depth for circular pipe, 236fig
thick-walled pipe equation, 557sb dimensionless graphs of geometric elements for
thin-walled pipe equation, 557sb noncircular pipe shapes, 237
Thailand, drainage management in, 580 discharge vs. depth in a circular pipe, 235–236,
thrust blocks, need for at pipe bends and other 236fig
appurtenances, 208 full-flow discharge, 235–236, 236fig
tilt-bar grates, 378 Manning equation (example), 237
time of concentration Manning equation for discharge, 235
drainage basin flow paths, 139fig in simple channels, 233–234
in drainage basins, 134 Chézy equation for discharge, 233
estimating (example), 138–139 Manning equation for discharge, 233
time-area methods, for runoff hydrographs, compared, normal depth, 233
174t normal slope, 233
time-series analysis, of rainfall, 75sb unit hydrographs
TMDL. See total maximum daily load comparison of models, 174t
tongue-and-groove pipe joints, 412 defined, 151, 153, 154fig
total discharge, components, 116 for gaged watersheds, 154–155
total dynamic head (TDH), across pumps, 545 NRCS synthetic unit hydrographs, 155–158
total head, 56–58, 189fig, 190 peak discharge equation, 156
equation for, 190 synthetic, 155–158
in pipes, 189fig, 190 See also hydrographs
equation for, 190 United Kingdom, stormwater management regulations in,
total maximum daily load (TMDL), 613 586–588
Total Maximum Daily Load (TMDL) Program, 598 United Nations
TP-40, 80–91 Conference on Environment and Development
TR-55, 122, 124 (1992, Rio de Janeiro), 583, 603
tractive force, 305 Conference on Human Settlements (1996, Istanbul).
See also shear stress See Habitat II
tractive force ratio, 305fig Convention on the Law of the Non-Navigational
tractive tension. See shear stress Uses of International Watercourses, 582
transboundary waters, 582–583 Convention to Combat Desertification, 583
transitional flow, 188t, 192, 195 Mar del Plata Conference, 576t, 577, 603
translation routing, 650 United Nations Development Programme (UNDP), 578
transpiration, 108, 108fig Water and Sanitation Programme, 578–579
trash racks United Nations Economic Commission for Europe, 582
for culverts, 330 United Nations Environment Programme (UNEP), 579
in pumping stations, 568fig information resources from, 579
trickle channels, 291 International Environmental Technology Centre, 579
687
U.S. Army Corps of Engineers, Dredge and Fill Permit calculation (example), 243–244
Program, 598–599 velocity distributions, 187–188, 188fig
U.S. Environmental Protection Agency (EPA) values for, 188t
National Water Quality Report, 608 velocity head, 56, 187, 189fig, 190
runoff model, 175t See also kinetic energy
stormwater control regulations, 595–599 Virginia, stormwater regulations, 600sb
Total Maximum Daily Load (TMDL) Program, 598 volumetric flux
U.S. Geological Survey (USGS), 80, 148 flow velocity and, 53
regression-based formulas for predicting peak of fluids, 53
discharge, 148–150
source of rainfall data, 80
unlined open channels. See erodible open channels
unsteady flow, 645
attenuation, 645, 646
hydrodynamic equations, 647–648
looped rating curves, 645 W
routing method complications, 650
routing methods, 646 Wales
Urban Stormwater Initiative, Environment Australia, 579 regional water authority system, 586
urban stormwater pollutants, 609–611, 610t stormwater management regulations in, 586–588
urbanization warm fronts, 70, 70fig
flow regime alteration, 612 water, physical properties, 641, 641t–642t
habitat degradation, 612–613 Water and Sanitation Programme, 578–579
impacts on streams, 612–613 Strategy Framework Document, 579
USGS. See U.S. Geological Survey (USGS) water hammer. See hydraulic transients
USGS regression-based formulas for predicting peak water law (U.S.), 600–602, 603
discharge, 148–150 compartmentalization of, 499
limitations in use of, 150 court cases, 597–598, 602
regression equations applications (example), 150 diffused-water rules, 601
Dredge and Fill Permit Program, 598–599
EPA stormwater control regulations, 595–599
federal regulatory requirements, 592–593
local regulations, 601–602, 603
NPDES permit program, 575, 595–597, 601, 603
V prior appropriation, 601
riparian rights, 600–601
state laws and regulations, 599, 600sb
vacuum valves, 558 Stormwater Phase II Rule, 596–597
valves stormwater regulations, 592–593
for pumping stations, 558 surface water doctrines, 600–601
See also names of specific types of valves Total Maximum Daily Load (TMDL) Program, 598
valves and fittings See also Clean Water Act of 1977; European Union;
clean-outs, 558 under names of specific countries
flexible connectors, 558 water pollution, in Europe, 582, 583
varied flow, 226 Water Pollution Control Act Amendments of 1972, 594
varied flow profiles, 232 Water Pollution Control Act of 1948, 593–594
velocity water power produced by pumps, 546
average cross section, 187 determination of (example), 547
adjustment equation, 187 equation for, 546
for different velocity distributions, 188t water problems
energy equation adjustment for, 187–188 from stormwater runoff, 609 , 609t
momentum equation adjustment for, 188t, in urban areas, 609
209–210 water quality
conservation of mass and, 54 capture volume, 615
consideration in design of open channels, 288 of combined sewer outflow, 609–611
Manning equation for, 198–199 impairment attributed to urban runoff, 608t
in open channels, Manning equation for, 198 management, 607
in open pipes, Manning equation for, 198 standards, 613
permissible velocities in erodible channels, 298, 298t of stormwater, 609–611
as a function of mean sediment size, 298–300, of stormwater and combined sewer outflow, 610t,
300fig 611t
cohesive soils, 298, 299fig treatment volume sizing criterion, 615–616
permissible velocities in grassed open channels with Water Quality Act of 1987, 595
different grasses and slopes, 309–311, 311t water quality control, in Ontario, 592
pipe cleaning in storm sewers, 422, 423t water resources policies
See also average cross-sectional velocity; flow international milestones, 2, 576t
velocity World Bank, 578
velocity distribution coefficient Water Supply and Sanitation Decade, 577
in closed-conduit flow, 187–188, 188t, 209–210 water surface profile equation for gradually varied flow in
in open-channel flow, 225t, 257 open channels, 272
in uniform open-channel flow with compound cross watersheds, 118, 118fig
sections, 241 WBRM routing model, 174t
688 Index
weighted translation routing, 650

weir coefficients, 270
for curb-opening inlets, 395
for grate inlets, 390
for slotted-drain inlets, 399
weirs, 252
broad-crested, 268–270
discharge coefficients, 270
discharge equation, 270
specific energy, 268–270, 269fig
specific energy curves, 269fig, 270
Cipolletti weirs, 267
contracted, 265fig
sharp-crested, 266
discharge coefficient, 266, 266–269
rectangular, 265fig, 266fig, 267-268
rcontracted discharge equation, 267
contracted, discharge calculation
(example), 267
suppressed, discharge coefficients, 267
triangular, 265fig, 268–269
discharge coefficients, 268-269
specific energy curves, 269fig, 270
suppressed, 265fig
triangular sharp-crested, 268fig
discharge equation, 268
wet detention ponds. See detention ponds
wetlands, dredge and fill permits for, 598
wetted perimeter, 221, 222t
wingwalls, of culverts, 328, 328fig
World Bank, 578
Operational Policies, 578
water resources policies, 578
World Commission on Environment and Development,
576t, 577, 603
Z
zoning regulations, 602
EULA version 200501
END USER LICENSE AGREEMENT FOR BENTLEY SOFTWARE
IMPORTANT – READ CAREFULLY: This End-User License Agreement (“EULA”)

is a legal agreement between you (either an individual or a single entity) and Bentley
Systems, Incorporated ("Bentley") for the Bentley software and associated documentation
that accompanies this EULA, which includes the associated media and Bentley internet-
based services (“Software”).
YOU AGREE TO BE BOUND BY THE TERMS OF THIS EULA BY

DOWNLOADING, INSTALLING, COPYING, OR OTHERWISE ACCESSING
OR USING THE SOFTWARE. YOUR ACCEPTANCE OF ALL OF THE TERMS
AND CONDITIONS OF THIS EULA IS A CONDITION TO THE GRANT OF
LICENSE BELOW. THIS EULA, AS MAY BE MODIFIED BY ANY
APPLICABLE SIGNED WRITTEN AGREEMENT BETWEEN YOU AND
BENTLEY, REPRESENTS THE ENTIRE SET OF TERMS AND CONDITIONS
GOVERNING YOUR USE OF THE SOFTWARE AND SUPERSEDES ALL
PRIOR OR CONTEMPORANEOUS ORAL OR WRITTEN
COMMUNICATIONS, PROPOSALS AND PRESENTATIONS WITH RESPECT
TO THE SOFTWARE OR THE SUBJECT MATTER OF THE EULA.
If this agreement is translated into a language other than English and there is a conflict of
terms between the English and the other language, the English version will control. You
should keep a copy of this EULA for your records. The latest version of this EULA
appears in its entirety on http://www.bentley.com/legal/eula_en.txt. Bentley may update
or amend the EULA at any time without notice to you; however, the form of EULA in
effect at the time of the Software acquisition will apply.
Section 1: Terms and Conditions
1. CERTAIN DEFINITIONS.
1.1. “Academic Related Use” means the use of designated Software in object code
form solely for internal classroom instruction or research of your teaching staff
and/or students matriculated in a degree program and not to include student use
in a paid employment setting or any other use prohibited under this EULA.
1.2. “Academic Software” means Software that is identified as “Academic
Edition” or “Academic License” (or words of similar meaning).
1.3. “CAL” means client access license.
1.4. “Device” means a single personal computer, workstation, terminal, hand held
computer, pager, telephone, personal digital assistant, Server or other
electronic device used by a User.
1.5. “External User” means any individual (not an organization) who is not: (i)
one of your full-time, part-time or temporary employees; or (ii) agency
temporary personnel or an independent contractor on assignment at your place
of business or work-site.
1.6. “License Key” means the document furnished to you by Bentley in electronic
or such other format, as determined in Bentley’s sole discretion, that sets forth
a unique serial number for the Software and authorizes use of the Software.
1.7. “Production Use” means use of the Software in object code form by a single
User or a Device, as applicable, solely for internal production purposes in
support of one Site.
1.8. “Site” means the discrete geographic location where you first install or use the
Software.
1.9. “Time Clocks” means any time clocks, copy-protection mechanisms, or other
security devices embedded in the Software which may deactivate the Software
after expiration of any applicable subscription or termed license period.
1.10. “User” means any individual or entity that is not an External User.
2. GRANT OF LICENSE. As and for so long as you comply with all of the terms of
this EULA, Bentley grants you the right to (a) install and use one copy of the
Software for Production Use in the country where the Software is first obtained and
(b) use the documentation that accompanies the Software for internal, non-
commercial reference purposes only.
3. RESERVED RIGHTS. You acknowledge and agree that the Software is a
proprietary product of Bentley or its suppliers, distributors and unrelated third parties
("Suppliers") protected by copyright and other applicable intellectual property laws
and treaty provisions. You further acknowledge and agree that the entire right, title
and interest in and to the Software including associated intellectual property rights,
shall remain with Bentley or its Suppliers. This license grant may be made by Bentley
on behalf of Suppliers as third party beneficiaries of the license rights provided
herein. Bentley retains all rights not expressly granted to you in this EULA. THE
SOFTWARE IS LICENSED NOT SOLD.
4. REGISTRATION. You acknowledge that registration or activation may be required
in order for you to utilize the full benefits of the Software.
5. NO RENTAL OR COMMERCIAL HOSTING. Software is licensed for
Production Use only. You may not rent, lease, lend or provide commercial hosting
services with the Software. You may also not use the Software to provide fee or
transaction based services. Contact Bentley for the availability of alternate pricing if
you desire to use the Software in such fashion.
6. NO "MULTIPLEXING" OR POOLING. Use of software or hardware that
reduces the number of electronic devices directly monitored or managed by the
Software or directly accessing or utilizing the Software (sometimes called
"multiplexing" or "pooling" software or hardware) does not reduce the number of
licenses required; the number of licenses required would equal the number of distinct
inputs to the multiplexing or pooling hardware/software "front end."
7. LIMITATIONS ON REVERSE ENGINEERING. You may not decode, reverse
engineer, reverse assemble, reverse compile, or otherwise translate the Software
except and only to the extent that such activity is expressly permitted by applicable
law notwithstanding this limitation. To the extent that you are expressly permitted by
law to undertake any of the activities listed in the previous sentence, you will not
exercise those rights until you have provided Bentley with thirty (30) days prior
written notice of your intent to exercise such rights.
8. DATA CAPTURE AND USE. You agree that Bentley may collect and utilize
technical information gathered as part of Software support services that may be
provided to you. Data capture in this form will only be used to improve Bentley’s
products and/or provide customized services to you and will not be disclosed or
disseminated to third parties except in an aggregated form.
9. ARCHIVAL COPY. You may make one copy of the Software on media appropriate
for your single Device for the express purpose of backup in the event that the
Software media is damaged or destroyed, provided that you reproduce and include on
the backup copy all the information appearing on the original labels.
10. RESTRICTIONS ON CERTAIN SOFTWARE. Software identified as demo,
evaluation, BDN, Beta or “NFR” (or “Not for Resale” or with words of similar
meaning) may not be sold, bartered or otherwise transferred. Such Software may not
be used for any purpose other than your testing or evaluation unless specified
otherwise pursuant to a separate agreement signed by both you and Bentley.
11. ACADEMIC SOFTWARE. For Academic Software, Bentley hereby grants you a
non-exclusive right and license to use in object code form such Academic Software
for Academic Related Use only. You may not sell, barter or otherwise transfer
Academic Software. Special Note Applicable to Academic Software: If you have
covered the Academic Software subject to this EULA pursuant to a valid BEN
SELECT Agreement (or successor agreement) with Bentley then you may be entitled
to additional and incremental licensing benefits to those set forth in this EULA by
virtue of that relationship. In the event that Academic Software is no longer covered
by a valid BEN SELECT agreement due to termination of the BEN SELECT
agreement or any other reason, then you will lose those incremental benefits, and
your license rights will only be as set forth in this EULA.
12. TIME CLOCKS. Bentley’s default licensing term is perpetual unless otherwise
specifically identified for the Software licensed. If you have licensed the Software
subject to this EULA for a term shorter than a perpetual license, you acknowledge
that the Software may be delivered to you with embedded Time Clocks. You agree
that Time Clocks are not considered a defect of the Software and you release Bentley
from any and all claims, however characterized, arising from or related to Time
Clocks or their operation.
13. TRANSFER. Internal. You may transfer the Software and the EULA to a different
Device at the same Site, provided you completely remove the Software from all prior
Devices. You may also make a one-time transfer of a CAL to another of your Users
or Devices located at the same Site. In order to accomplish these transfers you may
need to contact Bentley. External. You may not transfer the Software and license
granted under this EULA, or a CAL, to a third party without Bentley's prior written
consent. If such consent is obtained, you may permanently transfer the Software and
the license granted under this EULA, or the CAL, provided you transfer the Software
and all and media to such third party, and you do not retain any copies. The recipient
of such transfer must agree to all terms and conditions of the EULA. Any purported
sublicense, assignment, transfer or encumbrance is void without Bentley's prior
consent.
14. UPGRADES. You may not use any Software identified as an upgrade unless you are
properly licensed to use Software which Bentley has identified as being eligible for
an upgrade. After installing an upgrade, you may use the original Software product
that was eligible for an upgrade provided that at any one time you use only the
upgraded Software or the prior Software version subject to the upgrade.
15. NO EXTENSION OF CAPABILITIES. You may develop your own applications
that interoperate or integrate with the Software. Bentley prices its Software, among
other factors, based on capabilities that we expose to you. You may not extend the
Software to enable or unlock capabilities of the Software not specifically identified by
Bentley as forming part of the specified end user functionality.
16. SEPARATION OF COMPONENTS. The Software is licensed as a single product.
Component parts of the Software may not be separated and installed or used on
multiple Devices.
17. TERMINATION. If you breach the terms and conditions of this EULA, Bentley
may terminate this EULA without prejudicing any of its other rights. In such event
you must destroy and remove all copies of the Software from your Device(s).
Sections 1, 3, 13, 20, 21, 23, 25, 26, 27, 28 and 29 specifically survive termination.
18. NO AUTOMATED USE. A license for the Software may not be shared or used
concurrently on different Devices, nor to support multiple User or operational
requests as indicated above. As a result, you may not use the Software in an
automated, unattended, non-interactive server application or component (including
ASP) where: (i) multiple User requests from different Users are queued for
processing; or (ii) multiple requests from one User are queued for processing but
acting against content created or edited by other Users. Examples which would
violate this Section 18 include but are not limited to use as a plot server, file
translator, print server or other applications using or employing similar methods.
19. LIMITED WARRANTY. Except for Software which is identified as no-charge,
free, demo, evaluation, BDN, Beta or NFR, which is provided to you “AS-IS” and
specifically without warranty of any kind, for sixty (60) days from the date of first
installation (the "Warranty Period"), Bentley warrants that (i) the Software will
perform substantially in accordance with the functional specifications in the
documentation which accompanies the Software; and (ii) the media on which the
Software is distributed meets generally accepted industry standards. It is understood
that neither Bentley nor its Suppliers are responsible for your use of the Software or
the results from such use. It is further understood that there may be errors or
omissions in the information contained in the Software, that the information
contained in the Software may not be current or complete and that defects in
hardware or software may prevent you from gaining access to the Software. This
limited warranty is offered by Bentley alone, and is not extended to any software
code that may be contributed to the Software by our Suppliers. Any supplements or
updates to the Software (including but not limited to fixes, work in progress builds, or
subsequent updates) provided to you after the expiration of the Limited Warranty
period above are not covered by any warranty or condition, express, implied or
statutory.
20. DISCLAIMER. THE FOREGOING LIMITED WARRANTY STATES THE SOLE
AND EXCLUSIVE REMEDIES FOR BENTLEY’S OR ITS SUPPLIER’S
BREACH OF WARRANTY. EXCEPT FOR THE LIMITED WARRANTY AND
TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, BENTLEY
AND ITS SUPPLIERS PROVIDE THE SOFTWARE AS IS AND WITH ALL
FAULTS, AND TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE
LAW IN YOUR JURISDICTION, BENTLEY AND ITS SUPPLIERS DISCLAIM
ANY AND ALL OTHER WARRANTIES, FOR ITSELF AND FOR ALL
SUPPLIERS, EITHER STATUTORY, EXPRESSED OR IMPLIED, INCLUDING,
WITHOUT LIMITATION, WARRANTIES OF GOOD TITLE, WARRANTIES
AGAINST INFRINGEMENT, AND THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THIS
LIMITED WARRANTY GIVES YOU SPECIFIC RIGHTS; YOU MAY HAVE
OTHER RIGHTS, WHICH VARY AMONG JURISDICTIONS.
21. HIGH RISK ACTIVITIES. The Software is not fault tolerant and is not designed,
manufactured or intended for use or resale as control equipment in hazardous
environments requiring fail-safe performance, such as in the operation of nuclear
facilities, aircraft navigation or communication systems, air traffic control, direct life
support machines, or weapons systems, in which the failure of the Software could
lead directly to death, personal injury, or severe physical or environmental damage
("High Risk Activities"). Accordingly, Bentley and its Suppliers specifically disclaim
any express or implied warranty of fitness for High Risk Activities.
22. END USER REMEDIES. If a defect in the Software appears that constitutes a
breach of the above Limited Warranty, Bentley shall, at its sole option, repair the
Software, refund the price you paid for the Software or replace the defective item(s),
provided that: (i) you notify Bentley of the defect during the Warranty Period; (ii) the
Software is not modified, changed, or altered by anyone other than Bentley, unless
authorized by Bentley in writing; (iii) your computer equipment is in good operating
order and the Software is installed in an officially supported environment; and (iv) the
non-conformity is not caused by a third party or by you, your agents, employees or
contractors. Repaired, corrected, or replaced Software shall be covered by this limited
warranty for the period remaining under the warranty covered by the original
Software, or if longer, for thirty (30) days after the date: (a) of installation by you of
the repaired or replaced Software, or (b) Bentley advised you how to operate the
Software so as to achieve the functionality described in the documentation. YOU
AGREE THAT THE FOREGOING CONSTITUTES YOUR SOLE AND
EXCLUSIVE REMEDY FOR BREACH BY BENTLEY OF THE LIMITED
WARRANTY MADE IN THIS EULA. Outside the United States, neither these
remedies nor any product support services offered by Bentley are available without
proof that you acquired the accompanying copy of the Software from an authorized
source outside the United States.
23. LIMITATION OF LIABILITY. Regardless of whether any remedy set forth herein
fails of its essential purpose by law, in no event will Bentley or its Suppliers be liable
for indirect, special, incidental, economic or consequential damages, regardless of the
nature of the claim, including without limitation lost profits, costs of delay,
interruption of business, loss of use, costs of lost or damaged data or documentation
or liabilities to third parties arising from any source, even if Bentley has been advised
of the possibility of such damages. In no event shall the liability of Bentley or its
Suppliers exceed the amount paid by you (in the currency used to purchase) for the
Software. Some jurisdictions do not allow the exclusion or limitation of implied
warranties or limitation of liability for incidental or consequential damages, so the
above limitation or exclusion may not apply to you. THE PROVISIONS OF THIS
EULA ALLOCATE THE RISKS BETWEEN BENTLEY AND YOU. BENTLEY’S
PRICING REFLECTS THIS ALLOCATION OF RISK AND THE LIMITATION
OF LIABILITY SPECIFIED HEREIN.
24. STATUTORY CONSUMER RIGHTS. Nothing in this EULA is meant to
contravene statutory rights that consumers may have pursuant to local law.
25. EXPORT CONTROLS. The Software has been manufactured or developed in the
United States of America and accordingly may be subject to U.S. export control laws,
regulations and requirements. Regardless of any disclosure made by you to Bentley of
an ultimate destination of the Software, you must not export or transfer, whether
directly or indirectly, the Software, or any portion thereof, or any system containing
such Software or portion thereof, to anyone outside the United States (including
further export if you took delivery of the Software outside the United States) without
first complying strictly and fully with all export controls that may be imposed on the
Software by the United States Government or any country or organization of nations
within whose jurisdiction you use the Software. The countries subject to restriction by
action of the United States Government are subject to change, and it is your
responsibility to comply with the United States Government requirements as they
may be amended from time to time. You shall indemnify, defend and hold Bentley
harmless for any breach of your obligations pursuant to this Section.
26. U.S. GOVERNMENT RESTRICTED RIGHTS. If the Software is acquired for or
on behalf of the United States of America, its agencies and/or instrumentalities ("U.S.
Government"), it is provided with restricted rights. The Software and accompanying
documentation are "commercial computer software" and "commercial computer
software documentation," respectively, pursuant to 48 C.F.R. 12.212 and 227.7202,
and "restricted computer software" pursuant to 48 C.F.R. 52.227-19(a), as applicable.
Use, modification, reproduction, release, performance, display or disclosure of the
Software and accompanying documentation by the U.S. Government are subject to
restrictions as set forth in this Agreement and pursuant to 48 C.F.R. 12.212, 52.227-
19, 227.7202, and 1852.227-86, as applicable. Contractor/Manufacturer is Bentley
Systems, Incorporated, 685 Stockton Drive, Exton, PA 19341-0678.
27. GOVERNING LAW. This EULA will be governed by and construed in accordance
with the substantive laws in force in the Commonwealth of Pennsylvania. The state
courts located in Chester County, Pennsylvania and the federal courts located in
Philadelphia, Pennsylvania shall have exclusive jurisdiction over all disputes relating
to this Agreement. To the maximum extent permitted by applicable law, the parties
agree that the provisions of the United Nations Convention on Contracts for the
International Sale of Goods, as amended, and of the Uniform Computer Information
Transactions Act, as it may have been or hereafter may be in effect in any jurisdiction
shall not apply to this Agreement.
28. SEVERABILITY. The provisions of this EULA shall be deemed to be separable
and the invalidity of any provision hereof shall not affect the validity of the remainder
of this Agreement.
29. NOTICES. Please send all notices under this EULA to Bentley Systems,
Incorporated, Attn: General Counsel, 685 Stockton Drive, Exton, PA 19341-0678.
30. QUESTIONS. Should you have any questions regarding this EULA, please contact
the Bentley subsidiary serving your country, or write to: Bentley Systems,
Incorporated, Legal Department, 685 Stockton Drive, Exton, PA 19341-0678.
31. RE-DISTRIBUTION OF BENTLEY® VIEW™. If you are interested in re-
distributing Bentley View either internally or externally to your organization, please
complete the online Bentley View Distribution Agreement found at:
http://www.bentley.com/bentleyview/redistribute.html.
Section 2: Terms Applicable to Server Software
This section details the installation, use and licensing of Server Software and associated
CALs as well as External Connector licenses.
1. CERTAIN ADDITIONAL DEFINITIONS.
1.1. “Client Software” means software that allows a Device to access or utilize
Server Software (and, also where applicable to utilize certain aspects of the
Software when disconnected from the Server).
1.2. "External Connector" means a separately licensable module for specific Server
Software which authorizes use of the Server Software by External Users.
1.3. "Per Processor" is a Server Software licensing mode where you are licensed to
utilize the Server Software on one or more physical or virtual processors within a
designated Server.
1.4. “Per Server” is a Server Software licensing mode where you are licensed to
utilize the Server Software on all processors physically contained within the
designated Server.
1.5. "Per User" or “Per Device” are licensing modes that requires you to license a
separate CAL for each unique User or Device that accesses or utilizes the Server
Software.
1.6. “Server” means any one of your computers that can run Server Software.
1.7. “Server Software” means Software that provides services or functionality to
your Server(s).
1.8. "Server/CAL" is a licensing mode where the maximum number of Users (or
Devices, if applicable) which may access or utilize the Server Software at a given
interval is less than or equal to the number of CALs that you have acquired and
designated for use exclusively with that Server Software.
2. LICENSING MODES.
2.1. Servers. Bentley licenses Server Software on a Server/CAL basis with either
Device or User CALs and/or on a Per Processor basis. Some Server Software
may be eligible for External Connector licensing. Server Software may be
limited, even in Server/CAL licensing mode, as to the total number of Devices
and/or Users that may access a designated Server Product. Unless explicitly
specified in the documentation accompanying the Server Software, the default
licensing mode for all Server Software is Server/CAL with User CALs.
2.2. CALs. CALs are licensed by Bentley on a per Server, per Device or per User
basis. A User CAL permits one User (using any Device) to access or use the
Server Software. A Device CAL permits one Device to access or use the Server
Software. Special Note Applicable to CAL Licensing: If you have covered the
CALs and Server Software subject to this EULA pursuant to a valid SELECT
Agreement with Bentley then you may be entitled to additional and incremental
licensing benefits to those set forth in this EULA by virtue of that relationship.
In the event your SELECT Agreement with Bentley terminates or you otherwise
elect to remove CALs and/or their associated Server Software from coverage
pursuant to a SELECT Agreement with Bentley then you will lose those
incremental benefits, and your license rights will only be as set forth in this
EULA.
2.3. External Connectors. Certain Server Products may be licensed to support
External Users by virtue of an External Connector license.
3. GRANT OF LICENSE FOR SOFTWARE IN SERVER/CAL MODE. As and
for so long as you comply with all of the terms of this EULA, Bentley grants you the
following rights:
3.1. Installation and Use.
(a) Server Software. You may install and use one copy of the Server Software
for Production Use on a single Server in the country where the Server
Software was first obtained. You may also use the documentation that
accompanies the Server Software for internal, non-commercial reference
purposes only.
(b) Client Software. You may install and use the Client Software on the
aggregate total number of Devices to which you have dedicated a CAL for
such usage.
(c) CALs. A separate CAL is required for each User or Device that accesses or
uses Server Software on any of your Servers. A CAL grants a User (using
any Device) or a Device the right to access or otherwise utilize the Server
Software you have deployed on your Server. The maximum number of
Users or Devices that may access or use Server Software installed on a
particular Server at a given interval equals the number of CALs (of either
type) that you acquire and designate for use exclusively with that Server.
CALs cannot be pooled among internal Users or Devices and are tied to one
Server Software installation.
(d) Passive Fail-Over Server. If the Server Software is used in a clustered
environment, you may use the Server Software on a temporary basis on a
Server that is employed only and exclusively for fail-over support.
4. GRANT OF LICENSE FOR SERVER SOFTWARE IN PER SERVER MODE.
As and for so long as you comply with all of the terms of this EULA, Bentley grants
you the following rights:
4.1. Installation and Use
(a) Server Software. You may install and use one copy of the Server Software
for Production Use on a single Server in the country where the Server
Software was first obtained. You may also use the documentation that
accompanies the Server Software for internal, non-commercial reference
purposes only. In Per Server licensing mode, the Bentley default is that you
may use the Server Software on all processors physically contained within
that one Server. If the Server Software is licensed in Per Processor mode,
you may only use the Software on the authorized and licensed number of
processors (physical or virtual) within that Server.
(b) Client Software. In Per Server licensing mode, unless Bentley licenses the
Server Software with a limitation on the maximum number of Devices or
Users that may access it, you may install the Client Software on any Device
in support of any number of Users so long as the Client Software is being
used only in conjunction with the Server Software.
(c) CALs. In Per Server licensing mode, unless Bentley licenses the Server
Software with a limitation on the maximum number of Users or Devices that
may access it, an unlimited number of Users or Devices may access and use
the Server Software. CALs are not required for individual Users or Devices
in the Per Server licensing mode.
(d) Passive Fail-Over Server. If the Server Software is used in a clustered
environment, you may use the Server Software on a temporary basis on a
Server that is employed only and exclusively for fail-over support.
5. GRANT OF LICENSE FOR EXTERNAL CONNECTOR LICENSES. As and
for so long as you comply with all of the terms of this EULA, Bentley grants you the
following rights:
5.1. Installation and Use
(a) External Connector License. You may install and use one copy of the
External Connector on one Device and connect that Device to the designated
Server Software whether or not located at the same Site, but always within
the same country as the Server Software installation.
(b) External Users. The default licensing mode for an External Connector
authorizes you, for each External Connector license that you acquire, to
permit any number of External Users to access or use a single copy of the
designated Server Software for which the External Connector License has
been obtained without the need for you to acquire a CAL for each External
User. If any User does not clearly qualify as an External User then you will
need to properly license such use and access by the User of the Server
Software by a method other than via the External Connector. Certain
External Connector licenses only authorize a limited number of External
Users to connect via that External Connector, please check your Product
documentation and License Key for specific details, limitations and
qualifications.
(c) Passive Fail-Over External Connector. If the External Connector is
installed on a Device used in a clustered environment, you may use the
External Connector on a temporary basis on a Server or Device that is
employed only and exclusively for fail-over support.

SCMD 2017

Uploaded by

SCMD 2017

Uploaded by

STORMWATER CONVEYANCE

MODELING AND DESIGN

Exton, Pennsylvania USA

Copyright © 2003 by Haestad Press

Library of Congress Control Number: 2007925244

Bentley Systems, Inc. Phone (US): 800-225-2613

“Haestad Press has once again published a book that is one of

“Stormwater Conveyance and Design was written by persons

Stormwater Conveyance Modeling and Design is the culmination of the efforts of

Stormwater Conveyance Modeling and Design represents a collaborative effort that

Muneef Ahmad, P. Eng.

Thomas E. Barnard, Ph.D., P.E.

Peder Hjorth, Ph.D.

Roger T. Kilgore, P.E.

Chapter 1 Introduction to Stormwater Conveyance

Chapter 2 System Components, Models,

2.2 Hydraulic Components 19

Chapter 3 Fundamental Laws and Units 51

Chapter 4 Modeling Rainfall 67

4.3 Basic Rainfall Characteristics 72

Chapter 5 Modeling Runoff 107

5.8 Channel Routing 169

Chapter 6 Flow in Closed Conduits 185

Chapter 7 Flow in Open Channels 217

Compound Channel Sections ...................................................... 248

Chapter 8 Design of Open Channels 287

8.5 Design of Grassed Channels 308

Chapter 9 Culvert Design 323

Chapter 10 Gutter Flow and Inlet Design 369

10.3 Inlet Locations and Design Flows 400

Chapter 11 Storm Sewer Pipe System and

Chapter 12 Stormwater Detention 477

12.3 Stormwater Detention Modeling Concepts 485

Chapter 13 Stormwater Pumping 533

13.5 Pump Controls 559

Chapter 14 Regulatory and Environmental Issues 575

Chapter 15 Stormwater Quality Management 607

15.3 Stormwater Quality Best Management Practices 617

Appendix A Physical Properties of Water 641

Appendix B Unit Conversions 643

Appendix C Unsteady Stormwater Flow 645

Appendix D References for Common Stormwater

Design and analysis of an urban stormwater management system is a fascinating and

Designers and managers of urban stormwater conveyance systems are

This text includes elements of design pertaining to stormwater quality,

Stormwater management professionals are increasingly being required to address

Continuing Education and Problem Sets

The CD in the back of this book contains academic versions of Bentley's

1.1 NEED FOR STORMWATER CONVEYANCE SYSTEMS

1.2 HISTORY OF STORMWATER MANAGEMENT

Post-Roman Era to the 1700s

The Nineteenth Century

where A = required cross-sectional area of drainage pipe (ft2)

The Twentieth Century

twentieth century consisted of the development of unit hydrograph theory by L. K.

Summary and Outlook

1.3 OVERVIEW OF STORMWATER CONVEYANCE

1.4 CHAPTER SUMMARY

Stormwater is managed with a network of structures designed to collect, convey,

2.1 Hydrologic Components

2.2 Hydraulic Components

Prismatic versus Nonprismatic Channels. Open channels exist in a wide

Channel routing is a procedure by which the outflow at the downstream end of a

oped by the Federal Highway Administration and published as “Hydraulic Design

• The flow condition in a culvert can be characterized as either full (pressure)

Storm Sewer Systems

In most cases, stormwater management systems are designed as gravity systems.

A steady-state analysis is typically sufficient for design applications that do not

Stormwater Detention and Retention

Capabilities of computer programs for stormwater detention modeling vary widely.

Summary of Hydraulic Models for Various

Component Modeling Methods Typical Input data Section of Text

2.3 THE DESIGN PROCESS

In areas where existing storm-drainage facilities require rehabilitation and/or retrofit-

Preliminary Design Development

Construction Drawings, Specifications, and Contract