A S H RA E
JOURNAL
The following article was published in ASHRAE Journal, February 1998. Copyright 1998 American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Inc. It is presented for educational purposes only. This article may not be copied and/or distributed electronically or in paper form without permission of ASHRAE.
Applying Process Hazard Analysis
To Laboratory HVAC Design
By John O. Varley, P.E.
Member ASHRAE
he chemical and oil processing industries have operated
under self-regulated best practice, including the use of
process hazard analysis (PHA) techniques, for several years.
The Occupational Safety and Health Administration (OSHA)
authored a federal mandate, the Process Safety Management
(PSM) of Highly Hazardous Chemicals...1 (29 CFR 1910.119)
in 1992, which provides regulatory motivation for engineers to
apply PHA techniques to new or revised processes. Although
laboratory facilities using hazardous materials are excluded
from this law, the procedures and guidelines listed in it provide
excellent guidance for the design and operation of these facilities. Regardless of the nature of the hazards (chemical or biological) contained in a laboratory, the engineers and operations
personnel responsible for laboratory design, construction and
maintenance can benefit by using PHA methods.
Several guidelines exist for the design and maintenance of
laboratory facilities. For example, regulatory requirements for
laboratories using hazardous chemicals are summarized in
Occupational Exposure to Chemicals in Laboratories 2 (29
CFR 1910.1450). In addition, federal mandates exist for nonclinical and pharmaceutical laboratories3,4 (21 CFR 58, 21
CFR 211). Finally, several industry guidelines and
standards5,6,7,8 (ASHRAE, 1995; NRC, 1981; NFPA, 1996;
ANSI/AIHA,1992) are available for laboratory HVAC design.
These tools offer prudent guidance for the design and operation of laboratories; but none of them specify the deliberate
methods of evaluating design listed in the PSM rule. Although
not intended to be used as a design tool, PHA offers designers
a systematic approach to engineering for safety. This article
will review PHA requirements, offer ideas to apply them to
HVAC system design and operation, and relate the benefits of
their use.
PHA Method Overview
Process hazard analysis is the study of a facility, process,
building, service or operation carried out to ensure that hazards
are identified, understood and properly controlled. A variety
of PHA methods are available, including:
Hazard and Operability (HAZOP) Studies.
What If analysis.
Failure modes and effects analysis.
Fault tree analysis.
Event tree analysis.
Cause-consequence analysis.
Human error analysis.
A more detailed overview of these approaches can be found
in the literature9 (AICHE, 1985). In this article, the HAZOP
method is reviewed9,10 (AICHE, 1985; Larkin, 1996).
54
ASHRAE Journal
Guide Words
Meaning
No
Design intent is not achieved.
Less
Result is less than the design intent.
More
Result is more than the design intent.
Part of
Result is qualitatively less than the design intent.
As well as
Result is in addition to the design intent.
Reverse
Result is the opposite of the design intent.
Other than
Complete substitution for design intent.
Table 1: HAZOP guide words.
HAZOP is only part of the total hazard study of a new or
revised process. A typical hazard study consists of several
phases:
1. This phase is performed during the initial and preliminary engineering stage of project development. Its purpose is
to ensure that the project, the process and related processing
materials are understood. This study will result in a cursory
understanding of the safety, health and environmental issues
associated with the project. It plays an important role in
enabling management to make key policy decisions and establishing rapport among the project team members. Upon completion of flow diagrams and general equipment layouts
significant hazards, their possible causes and protective measures to meet relevant criteria, are identified.
2. This phase is performed after detailed process and instrumentation diagrams (P&IDs) have been generated. It is
intended to identify hazards or obstacles to operability that
could arise from deviations in the design intent. In this phase
the HAZOP method is applied.
3. The post-construction phase consists of several steps: process commissioning, review of safety, health and environmental
protection, and documentation verification. The purpose of the
commissioning process is to verify that the building and/or process has been built to the intended design and that actions from
the previous hazard studies have been completed. Next, safety,
employee health and environmental protection on the site are
reviewed for conformance with corporate and legislative
requirements. Finally, all documentation associated with the
About the Author
John O. Varley, P.E., is a consulting engineer located in Bray
Village, Ohio. He is a member of ASHRAE TC 9.10, Laboratory
Systems. Varley has a Bachelor of Science in Mechanical Engineering from the University of Akron, Akron, Ohio, and a Master of Science in Chemical Engineering from Cleveland State
University, Cleveland, Ohio.
February 1998
L A B O RA T O R Y
project is reviewed and
AIR HANDLING UNIT
incorporated into the
EXHAUST FAN
operation of the facility.
Using a team-oriHWR
ented approach that
HWS
M
brainstorms potential
problems with a design,
EA
the HAZOP method
evaluates process hazOUTSIDE AIR
ards and identifies
potential
operability
T
issues. It achieves its
results by applying
guide words to specific
points (or nodes) in the
DA Damper actuator
Legend:
T
Temperature sensor
design.
LAB MODULE
GEA General exhaust air
EOA Exhaust outside air
TE Temperature element
The HAZOP study
EA Exhaust air
MA Makeup air
FT Flow transmitter
requires basic informaInstrument
PT
Pressure transmitter
FV Face velocity
Use
tion prior to commenceSA Supply air
SP Static pressure
ment
including
a
preliminary process and
instrumentation diagram Figure 1: Sample AF&ID.
(P&ID), material data
safety sheets (MSDS)
EXHAUST FAN
and preliminary operatNODE 2
AIR HANDLING UNIT
ing procedures. The
NODE 3
project manager typically drives the HAZOP
HWR
NODE 6-EXHAUST
HWS
process by assembling
MANIFOLD
M
the data, establishing the
EA
schedule and identifying
the team members.
Team
membership
OUTSIDE AIR
NODE 5
should include a trained
T
HAZOP team leader, a
scribe to record the
NODE 4
results of the study, releNODE 1
vant operating personnel
Control
LAB MODULE
and the project manager.
Volume of
The team meetings
Node
begin after the process
or facility has been
divided into its discrete Figure 2: Breakdown of AF&ID into nodes.
nodes. Under the tutelage of the study leader, the team identi- same materials as a hazardous chemical tion of process hazard analysis methods
fies the design intent of each node and or biochemical process facility, but in to laboratory design.
then applies guide words (Table 1) to smaller and more varied quantities.
Prior to the HAZOP session, the
Therefore, the hazards within the labora- design team will have gathered the folreview the process.
The use of guide words prompts dis- tory facility can be as severe as a process lowing information:
covery of deviations from the design facility, as seen in recent catastrophic
The airflow and instrumentation diintent. The team then discusses the con- accidents11,12,13 (Dowdy, 1995; Lhotka, agram (AF&ID). This drawing is similar
sequences and causes, suggests recom- 1995; Reed, 1996). Further, the HVAC to the P&ID described earlier and promendations and records results. The system associated with the laboratory is vides the basis for the structure of the
results are recorded. An example of this a primary means of hazard mitigation. HAZOP (Figure 1).
process is discussed later in this article. This suggests that application of the
Relevant material data safety sheets
HAZOP method is a logical extension of
Applying HAZOP
(MSDS).
the design process. Although a literature
Preliminary equipment list.
Regardless of the nature of work search uncovered discussion of risk
Sequence of operations for the conhoused in a laboratory (industrial, aca- assessment14 (West, 1978) for laboratodemic or institutional), it will contain the ries, no work was found on the applica- trol system.
TE
FT
MA
DA
OA
DA
SA
PT
DA
DA
EOA
SP
EA
DA
GEA
FT
FV
YY
XX
TE
MA
FT
DA
DA
OA
SA
PT
DA
SP
EA
DA
DA
EOA
GEA
FT
FV
February 1998
ASHRAE Journal
55
Guide Word
Deviation
Causes
Consequences
No Flow
VAV Valve failure, blockage, power failure
Less
Less Flow
Same as no. Hoods all open Chemical exposure
at the same time.
2. Establish procedure for low flow situation.
Reverse flow
Fan Failure
Chemical exposure
3. Establish procedure for this failure.
4. Evaluate failure mode of damper.
5. Minimize leakage areas in duct.
6. Explore possibility of installing LEL detectors.
7. Investigate whether explosion venting for the duct
should be installed.
Late valve action Slow controller response
Same as no
8. Establish procedure to verify hood performance prior to
use.
9. Establish appropriate controller response time.
Spill in hood as
well as normal
hazards
Chemical exposure,
fire and explosion
hazard
Reverse
Late
As well as
Spill
Chemical exposure
Recommendations
No
1. Evaluate failure modes of VAV valve.
10. Establish procedure for clean-up.
12. Investigate installing emergency devices in hood and
room.
13. Establish emergency response procedure for incident in
hood or outside of hood.
Table 2: HAZOP example.
The project manager (owners representative) will then assemble the study
team. The team members should consist
of the following:
The hazard study team leader.
Scribe.
The local safety and health manager.
Commissioning manager.
Facilities representative.
The affected lab users.
The engineer.
The project manager.
The meeting membership (with the
exception of the hazard study leader and
scribe) should be composed of individuals who have working knowledge of the
proposed system. For example, the portion of the study involving mechanical
space should involve facilities personnel. On the other hand, the study of operations within the lab module should
involve lab personnel. Typically, executive management is excluded from hazard study meetings.
Prior to the meeting, the project manager and engineer will reduce the airflow
and instrumentation diagram into its
nodes (Figure 2). This will allow the
team to focus on discrete parts of the
process to simplify the discussion.
The study team participants should
allow ample time to complete the study.
Although time requirements vary according to the complexity of the drawings, a
typical rule-of- thumb is to allow thirty
minutes per node. Common nodes can
receive a cursory study, thereby eliminating some labor. In addition, the deliberate
quality of the study will require many
breaks in the meeting. To maintain a high
56
ASHRAE Journal
caliber of discussion, the meetings should
not extend to a point where team members can no longer concentrate. Four hour
meetings seem to provide optimal results.
This discussion suggests that the effort
associated with a rigorous HAZOP
should not be underestimated. As a result,
the study team leadership must receive
the support of executive management.
This will provide positive reinforcement
to the team membership to contribute the
requisite time for this endeavor.
The effort of documenting the study
can be significantly reduced by using
software readily available on the market.
Other meeting tools, such as an overhead projector, can also improve discussion and productivity.
With team and tools assembled at the
meeting place, the actual process of
HAZOP begins. Following the HAZOP
guide word method, each node is evaluated for possible deviations to the design
intent. The study depicted in Table 2
involves the design of a manifolded VAV
exhaust system. The node illustrated here
(Node 1 of Figure 2) includes a fume
hood, fume hood sensor, controller and
exhaust valve. The design intent, as determined by the team, is to provide a hood to
enclose weigh-up and mixing operations,
and solvent cleaning of lab apparatus.
The study leader must ensure that all
nodes are identified and discussed during
the session. The project manager is responsible for responding to all recommendations (sometimes referred to as findings).
Although the HAZOP method is ideally suited for evaluation of AF&IDs, it
can also be used to review building con-
struction practices. In addition, the other
aforementioned PHA techniques may
offer expedient and effective approaches
to the analysis of the process and building.
Conclusion
Process Hazard Analysis enhances the
opportunity to create inherently safe
designs by identifying alternatives to processing, raw materials used in processing,
construction materials and procedures. In
addition, PHA forces the design team to
rigorously analyze their design assumptions and clearly document the intent of
their design. Although design improvements generated by PHA significantly
lower the hazards of the process, the identification of procedural and educational
requirements for the design may provide
even greater benefits. As seen in the earlier example, these needs are identified
and resolved. Thus, the PHA process can
significantly reduce the hazards created
by human error.
By using a multi-discipline team, these
studies tend to identify design improvements. Although HAZOP requires an
expenditure of time and money, it can
result in even greater savings in first costs
and operating costs by improving the system design. As a result, processes that
have been evaluated using PHA are inherently more value-driven than those
designs that ignore the technique. Further,
PHA reduces costly additions to a project
during construction by enhancing communication among team members during
conceptual and preliminary engineering.
Finally, the process expedites the
commissioning and turn-over phase of a
project. Operating personnel involved
February 1998
L A B O RA T O R Y
with the study receive early training relative to the design. This results in a
higher degree of involvement by the
users of the design and ultimately greater
buy-in. In addition, the HAZOP provides complete documentation of the
design intent of the process, thereby
facilitating the commissioning of control
loops and other operations.
References
1. American Institute of Chemical Engineers, Guidelines for Hazard Evaluation Procedures, New York, NY, 1985.
2. ANSI/AIHA, Standard Z 9.5-92, Laboratory Ventilation, American Industrial
Hygiene Association, Fairfax, VA, 1992.
3. ASHRAE, Laboratory Systems, Chapter 13, Applications Handbook, 1995.
4. Code of Federal Regulations, Current
Good Manufacturing Practice for Finished
Pharmaceuticals, 21 CFR 211.
5. Code of Federal Regulations, Good
Laboratory Practice for Non-clinical Laboratory Studies, 21 CFR 58.
6. Code of Federal Regulations, Occupational Exposure to Chemicals in Laboratories, 29 CFR 1910.1450.
7. Code of Federal Regulations, Process
Safety Management of Highly Hazardous
Chemicals, 29 CFR 1910.119.
8. Dowdy, Zachary R., Chemical Explosion Burns Two Lab Technicians, Boston
Globe, p. 24, col 3, Aug. 30, 1995.
9. Larkin, Felim, HAZOP Study from
Theory to Practice, Process Engineering,
vol 27, p. 26-7, March 1996.
10. Lhotka, William C., Technician Succumbs to Blast Injuries, St. Louis-Dispatch,
Sec. D, p. 12, col. 1, Sep. 17, 1995.
11. National Research Council, Prudent
Practices for Handling Hazardous Chemicals in Laboratories, National Academy
Press, Washington, D.C., 1981.
12. NFPA, Standard 45, Fire Protection for
Laboratories Using Chemicals, National Fire
Protection Association, Quincy, MA, 1996.
13. Reed, Mack, Firm Agrees to Record
Fine in Fatal Blast, Los Angeles Times, Sec.
A, p. 3, col.5, April 9, 1996.
14. West, D., Assessment of Risk in the
Research Laboratory: A Basis for Facility
Design, ASHRAE Transactions, Vol. 84,
Part 1B, 1978.
Advertisement in the print
edition formerly in this space.
Advertisement in the print
edition formerly in this space.
Please circle the appropriate number
on the Reader Service Card at the back
of the publication.
Extremely Helpful ........................ 462
Helpful ....................................... 463
Somewhat Helpful ....................... 464
Not Helpful................................. 465
February 1998
ASHRAE Journal
57
