Power System Reliability Analysis with Distributed Generators

by
Dan Zhu
Thesis submitted to the faculty of
Virginia Polytechnic Institute and State University
in partial fulfillment of the requirements for the degree of
Master of Science
In
Electrical Engineering

Approved:
Dr. Robert P. Broadwater, Chairman

Dr. Ira Jacobs

Dr. Timothy Pratt

May, 2003
Blacksburg, VA

Keywords: power system reliability, distributed generator, set, segment,

circuit traces.

Power System Reliability Analysis with Distributed Generators

by
Dan Zhu
Committee Chairman: Dr. Robert P. Broadwater,
Electrical Engineering

Abstract
Reliability is a key aspect of power system design and planning. In this
research we present a reliability analysis algorithm for large scale, radially
operated (with respect to substation), reconfigurable, electrical distribution
systems. The algorithm takes into account equipment power handling constraints
and converges in a matter of seconds on systems containing thousands of
components.

Linked lists of segments are employed in obtaining the rapid

convergence. A power flow calculation is used to check the power handling

constraints. The application of distributed generators for electrical distribution
systems is a new technology. The placement of distributed generation and its
effects on reliability is investigated. Previous reliability calculations have been
performed for static load models and inherently make the assumption that system
reliability is independent of load. The study presented here evaluates improvement
in reliability over a time varying load curve. Reliability indices for load points and
the overall system have been developed. A new reliability index is proposed. The
new index makes it easier to locate areas where reliability needs to be improved.
The usefulness of this new index is demonstrated with numerical examples.

Acknowledgements
I would like to acknowledge the invaluable guidance, concern and support
of my advisor, Dr. Robert Broadwater. During this research, he always accepted
my ideas with an open mind and gave me the maximum opportunity to contribute
to the program. His advice really helped me to refine the application.
I would like to thank Electric Distribution Design (EDD) Inc. for providing
facilities to finish this research work, and Electric Power Research Institute (EPRI)
Distribution Engineering Workstation (DEW) for benchmark analysis of the power
flow calculations.
Thanks are also due to Dr. Jacobs and Dr. Pratt for serving on my
committee. They both helped to review my thesis paper.
My husband, Max, deserves special thanks. His unselfish support and
encouragement has allowed me to keep my perspective during this time.

iii

Table of Contents
1. Introduction ... 1
1.1. Introduction ..... 1
1.2. Objective of the Research 1
1.3. Distributed Generators.. 2
1.4. Literature Review . 3
1.5. Definition of Power System Reliability 4
1.6. Reliability Assessment Techniques.. 5
2. Measuring Service Quality.. 7
2.1. Definitions of Performance Indices.. 7
3. Comparison of Different System Designs.. 9
3.1. Simple Radial Distribution System 9
3.2. Alternative Feed Distribution Arrangement..10
3.3. Alternative Feed Arrangement with DR .. 10
4. Switching Operations....... 12
5. Reliability Analysis Sets... 14
5.1. Segment..... 14
5.2. Reliability Analysis Sets............... 15
6. Pointer and Circuit Traces...................... 21
6.1. Workstation Circuit Model............21
6.2. Pointers...... 22
6.3. Circuit Traces............................... 24
7. Computer Algorithm............................... 28
7.1 Introduction .. 28
7.2 Algorithm...28

iv

7.3 Power Flow Calculation. 36

7.4 Software Design. 38
8. Reliability Indices .42
8.1. Functional Characterization.......... 42
8.2. Reliability Indices Calculation...... 43
8.3. Relative Reliability Index............. 45
9. Distributed Generator Placement ...... 48
10. Case Studies........ 49
10.1. Introduction. 49
10.2. Case Study One.. 49
10.3. Case Study Two.. 58
10.4. Case Study Three 61
11. Conclusions and Further Research...... 65
11.1. Conclusions 65
11.2. Further Research. 66
12. References .......... 67
Appendix A .............. 69
Vita............................ 80

List of Figures
Figure 1.1 Subdivision of System Reliability.....5
Figure 3.1 Simple Radial Distribution System.......9
Figure 3.2 Alternative Feed Distribution Arrangement ...10
Figure 3.3 Alternative Feed Arrangement with DR .....11
Figure 4.1 Sample Circuit.............13
Figure 5.1. Sample segment .........15
Figure 5.2. Reliability Analysis Sets .......16
Figure 6.1 Sample Circuit ............25
Figure 7.1 Illustrating Selection of Alternative Feed ...34
Figure 7.2 Reliability Analysis Algorithm Sequence Diagram 39
Figure 8.1 Example Circuit for Relative_CAIDI .........46
Figure 10.1 System 1 for Case Study One ...................50
Figure 10.2 System 2 for Case Study One: Adding an Alternative Feed .53
Figure 10.3 System 3 for Case Study One: Adding a Distributed
Generator .............56
Figure 10.4 System for Case Study Two .....58
Figure 10.5 Addition of Substation and DG to System Shown
in Figure 10.4 .......... 59
Figure 10.6 DG at the End of Circuit .......60
Figure 10.7 Circuit for Case Study Three ....................62
Figure 10.8 Down Time Variation with Varying Load of L_C32 .. 63
Figure 10.9 Down Time with Varying Load of L_C32 ...64

vi

List of Tables
Table 6.1 DEW Component Trace Structure Element 23
Table 7.1 Summary of Traces Used to Develop the RA Sets ..36
Table 7.2 Summary of Messages in the RA Sequence Diagram .40
Table 10.1 Equipment Index Table ..51
Table 10.2 Improvement of Reliability ....55
Table 10.3 Comparison of Reliability Improvements ..57
Table 10.4 System Reliability Improvement for Case Study Two .. 61

vii

1. Introduction
1.1. Introduction
The economic and social effects of loss of electric service have
significant impacts on both the utility supplying electric energy and the end
users of electric service. The cost of a major power outage confined to one
state can be on the order of tens of millions of dollars. If a major power
outage affects multiple states, then the cost can exceed 100 million dollars.
The power system is vulnerable

[1]

to system abnormalities such as control

failures, protection or communication system failures, and disturbances,

such as lightning, and human operational errors. Therefore, maintaining a
reliable power supply is a very important issue for power systems design and
operation.
This thesis presents the research efforts and the software
implementation of a reliability analysis algorithm for electrical power
distribution systems. This algorithm is used to study reliability
improvements due to the addition of distributed generators. This algorithm
also takes into account system reconfigurations.

1.2. Objective of the Research

One objective of this research is to evaluate power system reliability
analysis improvements with distributed generators while satisfying
equipment power handling constraints. In this research, a computer
algorithm involving pointers and linked list[2] is developed to analyze the

power system reliability. This algorithm needs to converge rapidly as it is to

be used for systems containing thousands of components. So an efficient
object-oriented computer software design and implementation is
investigated.
This algorithm is also used to explore the placement of distributed
generators and how the different placements affect system reliability, which
has not been done in previous research. This exploration makes possible the
comparison of alternative system designs to discover systems yielding
desired reliability properties.
In this paper, variation of power system reliability with the varying
loads is also investigated. Other publications of distribution system
reliability analysis associated with time varying loads have not been found.
1.3. Distributed Generators
Distributed generators (also known as Distributed Resources) come in
many forms including gas turbine driven synchronous generators, wind
powered induction generators, fuel cells with inverter circuitry, and others.
The use of distributed resource generation is projected to grow. This growth
is due to cost reductions available with distributed generators. The cost
reductions may be the result of released system capacity or reductions in
generation costs at peak conditions.

1. 4. Literature Review
Prior to the 1960s, the reliability of proposed power systems was
often estimated by extrapolating the experience obtained from existing
systems and using rule-of-thumb methods to forecast the reliability of new
systems[3].
During the 1960s considerable work was performed in the field of
power system reliability and some excellent papers were published. The
most significant publications were two company papers by a group of
Westinghouse Electric Corporation and Public Service and Gas Company
authors[3],[4]. These papers introduced the concept of a fluctuation
environment to describe the failure rate of transmission system components.
The techniques presented in these papers were approximations which
provided results within a few percent of those obtained using more
theoretical techniques, such as Markov processes. The application of
Markov Chains in the power system reliability field was illustrated in
Reference [5]. The Markov approach is limited in application because of
computer storage requirements and the rounding errors which occur in the
solution of large systems.
Most previous publications have focused on transmission system
reliability. This research focuses on distribution system reliability. This work
extends previous research[5], which demonstrated sets used in describing
power system reliability calculations. Reference [6] presented the first
application to apply linked lists and pointer concepts to reliability analysis. It

took into account constraints associated with switching operations, but it was
relatively slow due to running numerous power flow calculations.
One aspect investigated h ere is the effect of Distributed Generators
(DG) on power system reliability. Standards for connecting DGs into
distribution systems are just being developed. Reference [7] deals with
issues related to existing DG interconnection practices. An investigation of
eleven utilities and industry interconnection standards was performed to
identify the key requirements for a DG connection. The results of this
investigation led to the development of a unified approach for determining
interconnection requirements. Reference [8] considers many aspects of DGs
in distribution systems, including protection, harmonics, transients, voltage
and frequency control. A Power flow calculation based on the positive
sequence model of the distribution circuits was presented.
1.5. Definition of Power System Reliability
The function of an electric power system is to satisfy the system load
requirement with a reasonable assurance of continuity and quality. The
ability of the system to provide an adequate supply of electrical energy is
usually designated by the term of reliability. The concept of power-system
reliability is extremely broad and covers all aspects of the ability of the
system to satisfy the customer requirements. There is a reasonable
subdivision of the concern designated as system reliability, which is
shown in Figure 1.

Figure 1.1 Subdivision of System Reliability

Figure 1 represents two basic aspects of a power system: system
adequacy and security. Adequacy relates to the existence of sufficient
facilities within the system to satisfy the consumer load demand. These
include the facilities necessary to generate sufficient energy and the
associated transmission and distribution facilities required to transport the
energy to the actual consumer load points. Security relates to the ability of
the system to respond to disturbances arising within that system. Security is
therefore associated with the response of the system to perturbations[9]. Most
of the probabilistic techniques presently available for power-system
reliability evaluation are in the domain of adequacy assessment. The
techniques presented in this paper are also in this domain.
1.6 Reliability Assessment Techniques
Reliability analysis has a wide range of applications in the engineering
field. Many of these uses can be implemented with either qualitative or
quantitative techniques. Qualitative techniques imply that reliability
assessment must depend solely upon engineering experience and judgment.

Quantitative

methodologies

use

statistical

approaches

to

reinforce

engineering judgments. Quantitative techniques describe the historical

performance of existing systems and utilize the historical performance to
predict the effects of changing conditions on system performance. In this
research, quantitative techniques combined with theoretical methods are
used to predict the performance of designated configurations. The systems
considered in this research are radially operated[10] with respect to
substations, but are reconfigurable.

2. Measuring Service Quality Performance Indices

A basic problem in distribution reliability assessment is measuring the
efficacy of past service. A common solution consists of condensing the
effects of service interruptions into indices of system performance. The
Edison Electric Institute (EEI), the Institute of Electrical and Electronics
Engineers (IEEE), and the Canadian Electric Association (CEA) have
suggested a wide range of performance indices[11]. These indices are
generally yearly averages of interruption frequency or duration. They
attempt to capture the magnitude of disturbances by load lost during each
interruption.
2.1. Definitions of Performance Indices
SAIDI (system average interruption duration index) is the average
interruption duration per customer served. It is determined by dividing the
sum of all customer interruption durations during a year by the number of
customers served.

SAIDI =

sum of customer interruption durations

total number of customers

CAIDI (customer average interruption duration index) is the average

interruption duration for those customers interrupted during a year. It is
determined by dividing the sum of all customer interruption durations by the
number of customers experiencing one or more interruptions over a one-year
period.

CAIDI =

sum of customer interruption durations

total number of customer interruptions

These two performance indices express interruption statistics in terms of

system customers. A customer here can be an individual, firm, or
organization who purchases electric services at one location under one rate
classification, contract or schedule. If service is supplied to a customer at
more than one location, each location shall be counted as a separate
customer.

3. Comparison of Different System Designs

Of paramount interest in any reliability study is ensuring a good
quality of service to customers defined as a combination of availability of
the energy supply and the quality of the energy available to the customers
(Medjoudj, 1994). In the following sections we will discuss the reliability of
the power supply for three kinds of situations. We will show how
reconfiguration and alternative sources improve the reliability of the power
system.
3.1. Radial Distribution System
Figure 3.1 shows a simple Radial Distribution System. In this system
a single incoming power service is received and distributes power to the
facility.

Figure 3.1 Simple Radial Distribution System

There is no duplication of equipment and little spare capacity is
typically included. Failure of any one component in the series path between
the source and the load will result in a power interruption to at least all loads
downstream of the failed component.

3.2. Alternative Feed Distribution Arrangement

A second distribution arrangement is used for facilities requiring more
reliable power. Figure 3.2 is a diagram representing this system arrangement.
Part of the load is connected to one source and the other part of the load is
connected to a second power source.

Figure 3.2 Alternative Feed Distribution Arrangement

The circuits (one circuit fed by S0 and the other fed by S1) are tied
together through a normally open tie-switch, with both power sources
energized. The electrical equipment is designed to accommodate 100% of
the facility load. For instance, when a failure occurs in source S0, after the
failure is isolated by opening the circuit breaker, the tie-switch is closed
allowing the complete load to be served from a single source until the
problem is corrected. Most customers can be restored immediately and dont
have to wait until S0 is repaired.
3.3. Alternative Feed Arrangement with DR
We can improve the reliability further by adding a distributed resource
(DR) into the circuit, as indicated in Figure 3.3. In case the failure occurs on

10

the left hand side of SW0, we can open SW0 and close SW3, so that the DR
can pick up the rest of the circuit, which was originally fed by S0. Without
the DR, we have to draw the power from S1. Such operation might violate
system constraints or degrade the quality of the power supply, especially
when the customer load reaches a peak value.

Figure 3.3 Alternative Feed Arrangement with DR

11

4. Switching Operations
Reliability analysis for a power system also leads to more reliable and
cost-effective operation, since power restoration analysis is a subset of the
calculations performed for reliability analysis. Here we assume switch
operation time is less than repair time, so loads that have lost power may be
restored faster by appropriate switching operations, or reconfiguration of the
system.
There are two kinds of switching operations of interest. One is
isolating the failure point so that a load point of interest which has lost
power may be re-supplied from the original source. The other is to again
isolate the failure point and to feed a load point of interest from an alternate
source, if an alternate source is available. For example, in Figure 4.1, if a
fault happens in component 5, we can open switch SW4 to isolate
component 5 from the rest of the system. The original source S0 can still
supply power to all the customers, except those on the downstream of switch
SW4.

12

Figure 4.1 Sample Circuit

The second kind of switching operation isolates the failure point and
interrupts the original power supply to the load point of interest. In this case
we need an alternate feed to restore power to the load point of interest. For
instance, if component 2 in the example circuit has a permanent fault, the
fault can be isolated by opening B1 and SW14. In case there is no alternate
source, all the segments downstream of the failed zone can only be restored
after the fault is repaired. Since we have an alternate source S1 (assuming S1
can supply the power and the alternative feed path can carry the power),
downstream of SW14 can be restored by closing SW25. The restoration time
for this part of the system is shorter with switching operations than with the
repair operation.

13

5. Reliability Analysis Sets

5.1. Segment
In essence, there are two configurations in a distribution system. One
consists of lines, transformers, and other components that are directly
responsible for transmitting power from the distribution substation to
customers. The second one consists of fuses, reclosers, circuit breakers, etc.
This interrelated network is designed to detect unusual conditions on the
power delivery system and isolate the portions of system that are responsible
for these conditions from the rest of the network. The location of protection
or isolation components on the distribution system and their response to
failures can have an important impact on the reliability indices. We will
sectionalize the distribution system into segments by these protection and
isolation components. In the following pages, the power system is not
modeled in terms of components but segments. A segment is a group of
components whose entry component is a switch or a protective device. This
sectionalizing device isolates groups of components into indivisible sections.
Each segment has one and only one switch or protective device.
In Figure 5.1, the only protection on the feeder is the station breaker.
The failure of any of the components in this segment can cause an
interruption at load point 1. It is the same for the other load points (2, 3, 4,
and 5). No temporary restoration is possible. For this configuration, the
reliability of all the load points (1, 2, 3, 4, and 5) is identical.

14

Figure 5.1. Sample segment

A segments name is the same as that of its sectionalizing device. In
Figure 5.1, there is only one segment, which is segment B. Breaker B and
components 1, 2, 3, 4 and 5 all belong to segment B.
Modeling the power system in terms of segments speeds up the
reliability index calculations. The algorithm can be programmed to run faster
since only the sectionalizing devices are processed without processing the
intermediate components.

5.2. Reliability Analysis Sets

In order to analyze the reliability of distribution systems, the Electric
Power Research Institute (EPRI) defined sets

[11]

needed for calculating the

reliability of a given load point. Figure 5.2 illustrates the relation among
these sets.

15

Figure 5.2. Reliability Analysis Sets

In reliability analysis, the failure of all elements that can cause a loss
of service to a particular load point must be considered. (This load point will
be presented in terms of a segment, which is the segment of interest S.) All
system components are either located on the continuous path between the
source and the segment of interest, or not located on the path. The failure of
all continuous path components can cause an interruption at the load point.
And the failure of components not in the path can also cause an interruption
at the load point, unless the component is separated from the path by a
protective device that responds automatically to the component failure. The
effects of nonseries elements and temporary restoration are now considered
in the sets shown in Figure 5.2, as will now be explained.

16

The L set shown in Figure 5.2 contains all segments within a circuit
whose failure can cause loss of power to the segment of interest S. This L set
includes all segments that are not separated from the continuous path
between the source (substation, generator, etc.) and the segment of interest S
by an automatic protection device.
Now we partition the L set into the sets SSL and NSSL:
The SSL set consists of the segments that may be isolated from the
continuous path between S and the original source
The NSSL set consists of the segments that cannot be switched away
from the continuous path between S and the original source.
The SSL set contains any segments separated from the continuous
path by manually operated switches. If any element of this set fails, the
segment of interest S can be temporarily restored from the original source
before the failed component is repaired or replaced.
Examining those segments that cannot be separated from the
continuous path, we can further partition the set NSSL into SL and NSL:
The SL set consists of the segments that can be switched away from
the segment of interest S, so that if the failure occurs in the SL set, S
may be fed by an alternate source

17

 The NSL set consists of the segments that cannot be switched away
from the segment of interest S. That is the segment of interest itself,
so this set only contains the element {S}.
If any thing fails in the NSL set, all the components within that
segment have to experience the full repair or replacement time of the failed
component. Temporary restoration is not possible.
Considering the SL set, we can divide it into SAF and NSAF:
For the SAF set, if the failed component lies in these segments, it is
possible to restore power to S by an alternate source
For the NSAF set, if the failed segment belongs to this set, the
segment of interest S cannot be temporarily restored from an alternate
feed.
The set SAF contains the segments that can be isolated from both the
segment of interest S and the alternative source, which make the temporary
restoration topologically possible. Sometimes, system constraints may limit
the restoration options; the alternate source might not have the capacity to
support the particular load point that of interest. So the set SAF is partitioned
into SF and NSF:
The SF set consists of all segments that can be isolated from S and an
alternative source, allowing power to be restored to S from the

18

alternative source (for segments in this set, system constraint

violations do not occur during the restoration)
The NSF set consists of all segments which may be isolated from S
and an alternative source, but for which it is not possible to restore
power to S because of violating system constraints.
The set L, including all the segments for calculating the reliability indices, is
decomposed into a number of sets as given by
L=SSL NSSL;

(5.1)

NSSL=SL {S};

(5.2)

SL= SAF NSAF;

(5.3)

SAF= SF NSF

(5.4)

Equation (5.1), (5.2), (5.3) and (5.4) yield

L=SSL SF {S} NSAF NSF

(5.5)

To sum up, if the failed component from the L set is placed in the SSL
set, it is possible to restore power to the load point of interest S from the
original source. If the failure occurs in the SF set, the power can be restored
to S from an alternative source without violating system constrains. But if

19

the failed component locates in either {s}, NSAF or NSF sets, then the failed
component must be completely repaired before power can be restored to S.
We use several additional reliability analysis (RA) sets to calculate the
sets of Equation (5.5), as given by
SIC = a set of all the segments in the circuit
SW = a set of all the sectionalizing devices in the circuit
AF = a set of available alternate sources
IS = a set of sectionalizing devices that will isolate the segment of interest S
from the original sources
NIS = a set of switches that do not isolate the original source from the
segment of interest
EC = a set of ending components for the circuit
PD = a set of protective devices in the circuit that isolate a load point of
interest from its source.

20

6. Pointer and Circuit Traces

6.1. Workstation Circuit Model
Electric

Power

Research

Institutes

Distribution

Engineering

Workstation, DEWorkstation, provides an engineering environment that is

focused on the design and analysis of electric distribution systems[12].
DEWorkstation is used in the research here.
Reliability analysis is complicated by a number of factors. One of
these is the size of distribution systems. Large metropolitan areas may
contain thousands of devices with several separate circuits supplied by
different substations. Calculation of reliability for a system is an extensive
logistical problem. Fundamental to reliability improvement is manipulation
of large amounts of interrelated data. This data includes distribution system
configuration, system fault protection, customer density, failure rate and
repair time. The methods with which this data is stored, displayed and
modeled determine the effectiveness of the computerized method. In
DEWorkstation, information about the distribution system under study is
permanently stored in data base tables. Initialization of the environment
results in the most commonly used circuit model data being loaded into the
workstation active memory[13]. This data is immediately available to and
shared by application modules, such as the reliability analysis application. In
this way, the number of accesses to the relational database is minimized. The
most commonly used application modules run entirely in high speed
memory and do not have to access the hard disk. This approach provides
rapid graphical display operations and engineering analysis computation.

21

6.2. Pointers
With large amounts of data in active memory, data structure
manipulation is a primary concern. A feature of the C language which has a
significant impact on this problem is the pointer. The pointer is a variable
that holds the address of a data element. Pointers permit the construction of
linked lists of data elements in computer memory [14]. In DEWorkstation,
pointers are used for all data objects. Applications share circuit information
via pointers, and also use pointers to manipulate data objects hidden inside
the applications.
In distribution systems, a single circuit model may contain over 5000
components, and an entire system model consisting of hundreds of circuits
may contain over a million components. With such large systems, modeling
methods have a direct impact on the ability to perform engineering analysis.
Use of pointers in linked lists allows system interconnects and equipment
parameters to be directly available for analysis without repetitive search
algorithms. Intrinsic in the graphical creation of the circuits is the creation of
linked lists. The DEWorkstation memory model links together sources and
components of each circuit[15]. In this way, it is possible to trace from circuit
to circuit, through an individual circuit, or through a particular branch of a
circuit.
Application programmers work with DEWorkstation defined objects.
These objects are manipulated and accessed via pointers and indices into
arrays of pointers. The links provided that pertain to component traces
involved in reliability analysis are:

22

 Forward Pointerforward direction for doubly linked list of circuit

components
Backward Pointer backward direction for doubly linked list of
circuit components
Feeder Path Pointer for a radial system, the feeder path pointer of a
given component is the next component toward the reference
substation that feeds the given component
Brother Pointer a given components brother pointer points to the
first component connected in its forward path which is not fed by the
given component. (It is used to detect dead ends or physical jumps in
connectivity.)
Because of these contained links and pointers, each components data
object is known as a trace structure. Table 1 lists the elements in the
trace component structure that are related to the reliability analysis
module. Each trace structure contains 198 data elements, including
pointers to other structures.
Table 6.1 DEW Component Trace Structure Element
Element Name

Data Type

Circuit number

Short integer

Substation number

Short integer

Equipment index number

Short integer

Component type number

Short integer

Component name

String

Forward Pointer

Pointer

23

Backward Pointer

Pointer

Feeder Path Pointer

Pointer

Brother Pointer

Pointer

//Elements added for reliability analysis module

Segment Pointer

Pointer

Forward Segment Pointer

Pointer

Backward Segment Pointer

Pointer

Feeder Path Segment Pointer

Pointer

.
Due to the large size of the trace structure, only the elements which

are employed by the reliability analysis module are listed in Table 1. Several
segment trace pointers are included in the structure. The Segment Pointer is
used to find the primary sectionalizing device for a component.
Sectionalizing devices in a circuit are linked in a doubly linked list via the
Forward Segment Pointer and the Backward Segment Pointer. Sectionalizing
devices are also linked with the Feeder Path Segment Pointer, which is
similar to the Feeder path pointer for components, except that only
sectionalizing devices are processed.

6.3. Circuit Traces

Circuit traces are applied in determining the reliability analysis (RA)
sets shown in Figure 5.2. Circuit traces employ pointers and linked lists

24

discussed previously. Circuit traces represent the order in which an

algorithm processes the components of the system. As indicated earlier, a
circuit analysis program must efficiently manage large quantities of system
and equipment data. The pointers and linked lists compact the data storage
and reduce algorithm execution time.
Here we provide an overview of using circuit traces. Figure 6.1 is an
example circuit used to illustrate the application of circuit traces. Source S0
is the original source of the circuit of interest, and S1 is the alternate source.
S1 is separated from the circuit of interest by the open switch SW25.

Figure 6.1 Sample Circuit

Each circuit trace represents a particular linked list tracing through the
components of a circuit. We will apply four types of component circuit

25

traces. These traces along with the notation used to indicate the trace, are
defined as follows:
FTm = forward component trace beginning with component m (if m is not
specified, FT begins from the substation). FT in the example circuit is
given by
FT= B1 2 3 SW4 5 6

(6.1)

BTm= backward component trace beginning with m; as illustrated by

BT15 = 15 SW14 13 12 SW11 10 9

(6.2)

FPTm = component ms feeder path component trace, as illustrated by

FPT15 = 15 SW14 6 2 B1.

(6.3)

ECT = ending component trace, here for the example circuit is given by
ECT=5 9 13 17 18 19

(6.4)

The circuit traces discussed above are basic circuit traces. For reliability
analysis, it is more efficient to work with pointers to segments and to
perform traces based on these pointers. The segment circuit traces used in
this research are as follows:

26

FSTm = forward segment trace from segment m, (if m is not specified, the
forward trace will begin with the substation). In the example circuit,
FST is given by
FST= B1 SW4 Fu7 SW14

(6.5)

FPSTm = feeder path segment trace (It is performed relative to a given

segment m). For instance, if we trace from the segment of interest,
segment SW16, FPSTSW16 is given by
FPSTSW16 = SW16 SW14 B1.

(6.6)

AFT = alternative feed trace. In the example circuit, there is only one
alternative source, so AFT is given by
AFT = SW25

(6.7)

If there is more than one alternative feed for the circuit, then AFT would
consist of the linked list of all alternative feeds.

27

7. Computer Algorithm
7.1 Introduction
This chapter presents the computer algorithm used to develop the
reliability analysis (RA) sets. The algorithm is implemented with linked lists.
A notation in terms of linked lists is introduced to describe the algorithm. A
software design for implementing the algorithm is also discussed. Along
with the presentation of the algorithm, the example circuit illustrated in
Figure 6.1, is used to explain the development of the RA sets.
7.2 Algorithm
In what follows, we assume for the example circuit that the segment
of interest is given by
{S} = {SW16}

(7.1)

We first conduct a forward component trace (FCT), beginning with

the substation, so that we can determine the SW set and set up segment
pointers. This can be expressed as
FCT

SW, pFSeg, pBSeg, pSeg

(7.2)

where
pFSeg = pointer to forward segment (in the example circuit, segment B1s
pFSeg pointer is pointed to segment SW14)
pBSeg = pointer to backward segment (in the example circuit, segment
SW14s pBseg pointer is pointed to segment B1)

28

pSeg = pointer to segment device for component (In the example circuit,
all the components in segment SW16, components 17, 18 and 19,
have their pSeg pointed to SW16)
The expression (7.2) is read as the Forward Component Trace (FCT) yields
the SW set and sets the pointers pFSeg, pBSeg, and pSeg. Note that the
notation used here is always to have pointers begin with a small p.
For the example circuit,
SW = {B1, SW4, Fu7, SW11, SW14, SW16, SW23, Fu26, SW25}

(7.3)

In the FCT, we can also find the ending components that make up the
EC set, by using the following condition
If a components forward pointer points to its brother pointer[6], then
this component is an ending component.
Thus,
FCT EC

(7.4)

There is a set of pointers representing the list of existing alternate

feeds, AF, which can be set up during the FCT as well. If a components
adjacent component, say component A, belongs to another circuit and is fed
by another substation, it means the original circuit is connected to an
alternative feed. Once such a component as A is found, the source for A can

29

be traced via a FPST. In this way, we can collect all the available alternate
sources. Thus
FCT AF

(7.5)

Note that for each segment stored in the AF set, there are two ending
components. One corresponds to a component in the EC set, and the other
component exists in the adjacent circuit.
Since IS consists of all the sectionalizing devices in the feeder path of
S, we can use a FPSTs to obtain the IS set, as well as the PD (protective
device) set, as given by
FPSTs IS, PD

(7.6)

For the segment of interest S in the example circuit

IS= {SW16, SW14, B1}
PD = {B1}

(7.7)
(7.8)

The logic used to develop the L set is as follows:

Perform a FST. When the FST encounters a segment whose primary
protective device belongs to the PD set, this segment is in the L set.
Otherwise, when the FST encounters a segment whose primary
protective device does not belong to the PD set, the segment is not in
the L set.

30

Thus,
FST L

(7.9)

Following the steps described above, we obtain the L set for the segment of
interest S
L= {B1, SW4, SW11, SW14, SW16, SW23}

(7.10)

The segments in the SSL set may be isolated from S and the original source,
so that the power can be restored from the original source. SSL is given by
the following set operations
SSL=L NIS

(7.11)

where NIS= SW IS.

Applying Equation (7.11) in the example circuit, and using expressions (7.3),
(7.7) and (7.10), we obtain
SSL= {SW4, SW11, SW23}

(7.12)

The NSL set has only one element the segment of interest S. All the failed
components in the segment of interest must be completely repaired before
power can be restored to S.

31

The segments in the SL set can be switched away from the segment of
interest S, so that if the failure occurs in the SL set, S may be fed from an
alternative source. The SL set is given by the following set operation
SL= L IS {S}

(7.12)

In the example circuit, applying expressions (7.1), (7.7) and (7.10), we

obtain
SL= {B1, SW14}

(7.13)

If the failed component lies in the SAF set, it is possible to restore

power to S when system constraints are not violated. The system constraints
that are of interest here are the power handling capabilities of the equipment.
Of particular interest is the remaining power handling capability of each
piece of equipment. In order to find the SAF set, we conduct feeder path
segment traces both from an alternate source and the segment of interest S,
FPSTAF and FPSTS, respectively. When these traces encounter a common
path, then the SAF set is not empty. The SAF set includes the segments in
the common path except the first segment that the feeder path traces meet in
the common path. Thus,
FPSTAF, FPSTS

32

SAF

(7.14)

In the example circuit,

SAF= {B1}

(7.15)

The NSAF set includes all the segments for which it is not possible to
restore power to S from an alternative source. All the failed components in
these segments must be completely repaired before restoring power to S.
The NSAF set is given by set operation:
NASF = SL SAF

(7.16)

In the example circuit, using expression (7.13) and (7.15), we get

NSAF= {SW14}

(7.17)

The segments in the SF set may be isolated from S and an alternative

source, so that power can be restored to S from the alternative source
without violating system constraints.
The NSF set includes all the segments which may be isolated from S
and an alternative source, but for which it is not possible to restore power to
S because of system constraint violations. All the failed components in these
segments must be completely repaired before power can be restored to S.
To achieve the SF set, the power required by S must be compared to
the minimum remaining capacity of the components along the feeder path

33

from the alternative feed (AF). If there is more than one alternative feed in
the system, the minimum capacities encountered in the feeder path
component traces FPTAF for all the available sources in the AF set must be
compared. For instance, there are n alternative feeds in the system. Let
CAFk = minimum remaining component power capacity in the FPTAF for the
kth alternative feed, k =1, 2, 3 n
CAFm = max
{CAFk}
k

(7.18)

(7.19)

Thus CAFm represents the greatest minimum remaining capacity

available among the alternative sources. For example, as demonstrated in
Figure 7.1, there are two alternative sources, AF1 and AF2. The segment of
interest is marked as S. As indicated in the figure, the power required by S is
5 KW. The numbers on the alternative feed components stand for the
remaining capacity (units of KW) of the components.

Figure 7.1 Illustrating Selection of Alternative Feed

34

According to Equation (7.18) and (7.19),

CAF1= min {10, 5, 30} = 5
CAF2 min {40, 20, 20, 10} = 10
CAFm = max { CAF11 , CAF21}
=max {5, 10}
=10
So
AFm = AF2

(7.20)

Even though the minimum remaining capacity on the feeder path from
AF1 is equal to the required power in S, pulling the power from AF1 to S will
fully utilize component AF12. Thus AF2 is chosen since it has more
remaining capacity on the feeder path.
In the general case, the segment of interest is not directly connected to
the alternative feeds as shown in Figure 9. So FPT traces in the circuit of
interest are also required to determine remaining power handling capabilities.
In essence, component traces from the segment of interest to all alternative
sources are required to check power handling capacities.
In summary, the Circuit traces which yield the reliability analysis (RA)
sets are shown in Table 7.1

35

Table 7.1 Summary of Traces Used to Develop the RA Sets

Algorithm Steps

Traces in the Circuit Model

Step 1

FCT

Step 2

FPSTs IS, PD

Step 3

FST L

Step 4

FPSTAF, FPSTS

Step 5

FPTAF

SW, pFSeg, pBSeg, pSeg, EC, AF

SAF

SF or NSF

7.3 Power Flow Calculation

In order to get the required power or remaining capacity of a
component, the power flow needs to be calculated. The Power Flow
algorithm is based on the two-port element model and the tree traverse [8]. It
is carried out by several iterations. Every iteration consists of a backward
traverse, followed by a forward traverse of all the elements. The backward
traverse calculates the currents through all the elements. The forward
traverse will calculate the voltage drops across elements. These calculations
are represented by the following equations.
*
S load
I j = Im + *
Vj

V j = Vi I j Z j

36

(7.21)
(7.22)

where
Ij = current through element j
Im = current through directly connected downstream element fed by
elements j
Sload = load attached to element j
Vj = voltage at downstream port of element j
Vi = voltage at upstream port of element j
Zj = the impedance of element j.
The sequential algorithm for the Radial Power Flow is given as follows:
1.

Starting from an ending element, backward traverse the tree

element-by-element. Equation (7.21) is applied to calculate
the current for each element.

2.

Starting from the source or root element, forward traverse the

tree element-by element. Equation (7.22) is applied to
calculate voltages for each element.

3.

Check the convergence criteria. If converged, stop; otherwise,

go back to Step 1.

Once the power flow calculation is completed, then

FPTAF

SF or NSF

(7.23)

In the example circuit, assuming system constraints are not violated,

SF= {B1}
37

(7.24)

7.4 Software Design

Figure 7.2 shows a sequence diagram which describes a software
implementation of the reliability analysis algorithm. It illustrates the
interactions among the objects and packages involved in the calculations.
Two objects, RA of type Reliability Analysis, and PF of type Power Flow
Analysis, and four packages- Circuit Model, RA Sets, Indices Calculation,
and Reliability Data- are illustrated in the sequence diagram. This diagram
visualizes the dynamic aspects of the reliability analysis software application.
As shown in Figure 7.2, after the user selects the segment of interest
with the message Pick _Seg( ), the Reliability Analysis object sends the
FCT( ) message repeatedly (as indicated by * ) to the Circuit Model package,
corresponding to Step 1 in Table 7.1. Note that messages are named after the
traces that are performed. Signatures of messages shown in Figure 7.2 are
defined in Table 7.2. In essence, FCT( ) provides a specialized iterator that
implements the Forward Component Trace. The FCT( ) message called
repeatedly, returns component pointers in the order of the FCT trace. Please
refer to Table 7.1 for the details of the component structure. Reliability
Analysis uses the returned components to set up segment pointers and the
sets SW, EC and AF.

38

Figure 7.2 Reliability Analysis Algorithm Sequence Diagram

39

Table 7.2 Summary of Messages in the RA Sequence Diagram

Messages

Return Value

FCT ( )

Component pointer

FPST ( )

Component pointer

FST ( )

Component pointer

Min_Cap ( )

Double representing the minimum remaining power

capacity of the components on the alternative feed
feeder path

Max_Cap( )

Double representing the maximum of the minimum

remaining capacities available among all the
alternative feeds

GetCus ( )

Integer representing the number of customers attached

to a component

setOperation_Org( ) RA sets for the original circuit

setOperation_AF( )

RA sets for the alternative source.

Get_Sets( )

Arrays of component pointers

Get Data ( )

Array of doubles representing annual failure rate,

repair time for a component and switch operation time

Reliability Analysis sends the message FPST(S) (S is the segment of

interest passed in as a parameter) repeatedly to the Circuit Model,
corresponding to Step 2 in Table 7.1. Circuit Model traces through the whole
circuit and returns segment pointers in the order encountered in the FPST,
and these segments are used to set up the PD set and IS set.

40

Corresponding to Step 3 in the Table 7.1, the message FST( ) is sent

repeatedly to the Circuit Model to set up the L set. Then set operations are
performed to derive the sets NIS, SSL, NSSL, SL, and NSL. The
development of these sets depends only on the original circuit, regardless of
whether alternative sources are available or not.
If there are alternative feeds, via the message FPST (AF), the Circuit
Model can achieve the matched components for reliability analysis to set up
the SAF set. This is the fourth step shown in Table 7.1. Once the SAF set is
available, the power flow calculation is called to check the system
constraints. The Message Min_Cap( ) is sent repeatedly to the circuit of
interest and all alternative feed circuits. In order to determine the remaining
power handling capability, PF sends the message FPT(AF) to conduct the
feeder path traces from all the alternative feeds. Then applying Equation
(7.19), the maximum remaining capacity is obtained. The SF set and NSF set
now be determined.
Then the message SetOperation_AF( ) is used to determine the rest of
the reliability analysis sets. Once all the sets of Figure 5.2 are determined
and the number of customers in each segment is obtained, reliability indices
can be calculated. The computation of reliability indices will be described in
the next chapter.

41

8. Reliability Indices
This analysis relies on two general classes of information to estimate
the reliability: component reliability parameters and system structure. Using
system structure and component performance data, we can evaluate the
reliability of specific load points or the whole distribution system. The
structure information is achieved by the circuit traces presented previously.
In the following paragraphs the performance data is discussed.
Predictive reliability techniques suffer from data collection difficulties.
Simplifying assumptions (default values) are required for practical analysis
of distribution systems.
8.1. Functional characterization
The availability of component functionally is characterized by the
following indices:
Annual Failure Rate = the annual average frequency of failure
Annual Down Time = the annual outage duration experienced
at a load point.
The failure rate for segment i, FRi , is the sum of the failure rates of all the
components contained in the segment i as given by
n

FRi = Fr j
j =1

where
Frj = the failure rate for component j

42

(8.1)

n = the number of components in segment i.

The average repair time for a segment i, REPi , can be calculated by

REPi =

Fr
j =1

Re p j

(8.2)

Fr
j =1

where
Frj = the failure rate for component j
Repj = the average repair time for component j
n = the number of components in segment i.
These indices are computed for each segment in the feeder. All load
points within a segment experience the same failure rate and down time.
In the reliability analysis program, failure rates and repair times from
field data are preferred. When this data is not available, default values are
fetched from a table in the relational database which has generic average
failure rates and repair times for each type of device.
8.2. Reliability Indices Calculation
After finding the reliability analysis sets for the segment of interest S,
we can calculate the reliability indices. First assume there is a single failure
incident.

43

The down time for the segment S, DTS , is given by

DTS =

FR REP + FR SOT

iNSL ,
NSAF ,
NSF

iSSL ,
SF

(8.3)

where
SOTi = switch operation time to re-supply segment S due to the failure of
segment i.
Note that the reliability analysis algorithm presented here assumes that
switch operations can always be performed faster than repairs.
The customer average interruption duration index (CAIDI) for a
segment is the same as DTs
CAIDI = DTs

(8.4)

Once the down time for each segment is calculated, and given the
number of customers attached to each segment, the total customer down
time, DTC, for a given circuit can be calculated by
DTC =

DT C

icircuit

where Ci = the number of customers attached to segment i.

44

(8.5)

Since the failure rate and down time is known at each segment on the
feeder, the system index SAIDI (system average interruption duration index)
is then given by

SAIDI =

DTC
Ci

(8.6)

icircuit

The average restoration time for segment S is computed as

RTs =

DTs
FRi

(8.7)

iL

8.3. Relative Reliability Index

A new measure of reliability referred to as Relative_CAIDI is
introduced here. Relative_CAIDIj helps to identify the areas that need
improvement. Relative_CAIDIj is given by

Relative _ CAIDI j =

CAIDI ckt
CAIDI j

where
CAIDIckt = average CAIDI for the circuit of interest
CAIDIj = CAIDI for segment j.

45

(8.8)

Thus
If Relative_CAIDIj = 1, then the customers in segment j have average
reliability
If Relative_CAIDIj < 1, then the reliability of the customers in
segment j is less than average
If Relative_CAIDIj > 1, then customers in segment j have reliability
better than average.

Figure 8.1 Example Circuit for Relative_CAIDI

In Figure 8.1, the number attached to each sectionalizing device is the
Relative_CAIDIj for that segment. We can see segments such as P11, P12,
P2, P31, and P4, have reliabilities greater than the average level of Circuit

46

C1, while segments such as P52, P71, P72, P63, have reliabilities poorer
than the average value.

47

9. Distributed Generator Placement

In the evolving energy industry, emerging distributed generator
technologies have the potential to provide attractive, practical, and
economical generation options for energy companies and their customers.
Distributed resource technologies range in size from 3-10 kW for residential
systems to 50-500 kW for commercial users to 1-50 MW in the industrial
market segment. Primary opportunities lie in using these technologies to
(1) improve the service and delivery of energy to end users
(2) support the operation and management of transmission and
distribution systems.
This work does not consider the islanding of distributed generators
(that is the generator operating without substation supply).
A distributed generator is often placed at a substation because no
further land purchases are needed. However, locating generators at
substations, distributed generator acts only as a back up power source, which
may not contribute significant reliability improvement as far as the entire
system is concerned. Instead, generators located further out on a circuit can
often significantly affect system reliability. It is necessary to evaluate the
effects of different placements of distributed generators. In case studies in
the next chapter we will see that locating the DG at the end of the circuit
produces more reliability improvement than placing it at the substation.

48

10. Case Studies

10.1 Introduction
Reliability is affected by the following
Varying loading
Switch/protective device placement
Switch operation times
Available alternative feeds
Equipment current limits
Equipment failure rates
Equipment repair times.
The following examples will illustrate the effect of some of these variations.
In this chapter the improvement of reliability by distributed generators
is demonstrated through three case studies. The first case study uses a test
circuit developed to show the influence of various factors on system
reliability. The second case study tests the reliability analysis program on a
large scale system. It also shows how different DG placements affect the
reliability of the system. The third example demonstrates how reliability
changes with system load variation.
10.2. Case Study One
System 1 is presented in Figure 10.1. This is a system with only one
substation Sub1, and 31 customers. The number attached to each
sectionalizing device is the number of customers in that segment.

49

Lp611

31

Figure 10.1 System 1 for Case Study One

Line Lp611 is assumed failed and switch p61 is assumed to have opened.
Thus the set of segments losing power due to the operation of p61 is
{p71, p72, p8, p62, p63}
Assume that segment p62 is the segment of highest priority. Applying set
Equations (1) - (8) relative to segment p62 gives
L= {p11, p12, p31, p61, p62}
SSL= {p31}

50

NSSL= {p11, p12, p61, p62}

NSL= {p62}
SL= {NULL}
and SF= NSF= NSAF= {NULL}
Using the default failure rate and repair time in Table 10.1, we can calculate
the annual down time for segment p62 as 0.355 hours. Since there are no
alternate feeds in the system, only the failure occurring in the SSL set, which
is p31 in this example, can be switched away; for the failure in the rest of Set
L, segment p62 has to experience the restoration time for the failing
component being completely repaired.
Table 10.1 Equipment Index Table
Equipment

Default Repair Time

Component Type

Default Failure Rate

Substation

0.1

Disconnect switch

0.001

Load break switch

0.001

Supervisory switch

0.001

Cutout Switch

0.001

Fuse

0.001

Recloser

0.001

Electronic recloser

0.001

Index

51

(Hrs/Yr)

Remotely set recloser

0.001

Sectionalizer

0.001

10

Breaker

0.001

11

Network protector

0.001

13

Remotely set relay

14

Reclosing device

0.001

15

Fixed tap transformer

0.01

16

Distribution transformer

0.01

17

Network transformer

0.01

18

Regulating transformer

0.01

19

Voltage regulator

0.01

20

Fixed shunt capacitor bank

0.01

21

Switched shunt capacitor bank

0.01

33

3-Phase line

0.01

34

2-Phase line

0.01

35

1-phase line

0.01

37

3-Phase cable

0.01

38

2-Phase cable

0.01

39

1-Phase cable

0.01

41

3-phase underground cable

0.01

42

2-Phase underground cable

0.01

43

1-Phase underground cable

0.01

44

Arrester

0.001

45

Current transformer

46

Potential transformer

47

Communication transmitter

48

Communication receiver

49

Combination switch

0.001

50

Motor operator

52

52

Ground relay

53

Phase Imbalance Relay

54

Elbow Switch

0.001

56

Cable, Station Pole

0.001

59

Normally Open Point Location

60

Pole Top Switch

0.001

In Figure 10.2, an adjacent circuit C2 is added to the system. This circuit has
some remaining capacity, which means it is possible for it to supply some
power to circuit C1.

31

L68

Figure 10.2 System 2 for Case Study One: Adding an Alternative Feed

53

Again, applying the set equations we get

L= {p11, p12, p31, p61, p62}
SSL= {p31}
NSSL= {p11, p12, p61, p62}
NSL= {p62}
SL= {p11, p12, p61}
SAF= {P11, p12, p61}
NSAF= {NULL}
SF= {p11, p12, p61}
NSF= {NULL}
If the failure happens in the set SAF, p62 can be restored from circuit C2
without violating system constraints, because Sub2 has plenty of capacity to
support its adjacent circuit. The set NSF is empty, so SF=SAF.
With the alternate source, the annual down time for p62 is reduced to 0.055
hours, nearly one seventh of the annual down time in the original system.

54

The significant drop comes from power being restored from Sub2, and p62
does not need to wait for the failing component to be completely repaired. In
this case, the down time will be the switch operation time instead of the
repair time for the failing component. The alternate source also improves the
reliability of the entire system. Table 10.2 shows a comparison of reliability
indices for System 1 and System 2.
Table 10.2 Improvement of Reliability
Reliability

System1 without

System 2 with

Percent

Indices

Alternate Feed

Alternate Feed

Improvement

SAIDI(Hrs/yr)

0.002

0.001

50%

CAIDI(Hrs/yr)

0.305

0.176

42%

If the load on circuit C2 becomes heavier, substation Sub2 might lose the
capacity to pick up the load on C1. For example, when we lengthen line L68
or add 5600kw load to it, pushing the load near to the overload point for the
line, the annual down time for segment p62 will jump back to 0.355 Hrs/yr,
and the system CAIDI will also go back to 0.305 Hrs/yr. It means the load
point of interest cannot be restored from the alternate source because system
constraints will be violated. Now we can see how the availability of alternate
feeds and the change of the system loading impact the system reliability.
Next we will illustrate how a distributed generator enhances the reliability of
the system. As it is illustrated in Figure 10.3, a distributed generator DR0 is
added near to the point where circuits C1 and C2 have their tie switch.

55

31
9

L68

Figure 10.3 System 3 for Case Study One: Adding a Distributed

Generator
When the load in circuit C2 grows so that substation Sub2 can no longer
pick up any load in circuit C1, the distributed generator DR0 will be
activated. This provides a source of power that can also be used to supply
loads switched from C1 to C2. The reliability of circuit C1 will increase due
to the availability of DR0. Table 10.3 shows the improvement in annual
down time for the segments in circuit C1.

56

Table 10.3 Comparison of Reliability Improvements

Segment Name

Down Time (Hrs/yr)

Improvement

Without DR0

With DR0

p63

0.405

0.095

77%

p62

0.355

0.085

76%

p61

0.31

0.13

58%

p71

0.36

0.18

50%

p72

0.41

0.23

44%

p12

0.22

0.13

41%

p8

0.46

0.28

39%

p31

0.265

0.175

34%

p4

0.27

0.18

33%

p2

0.27

0.18

33%

p51

0.28

0.19

32%

p52

0.37

0.28

24%

p32

0.465

0.375

19%

P11

0.13

0.13

0%

From Table 10.3, we notice that the segments close to DR0 (etc. P62, P63)
have more improvement than those (etc. P52, P32) far from DR0. The
segment P11, which is next to source Sub1, has no improvement at all. This
is because as the distance between the segment of interest and the alternative
feed increases, the alternative source needs to supply more and more power
to its adjacent circuit in order to restore the segment of interest, and its
remaining capacity becomes less and less. When it reaches the point that the
alternative source has no spare capacity to pick up any more load, the
57

availability of DR0 will not make any additional contribution to the

reliability of its adjacent circuit.
10.3. Case Study Two
Figure 10.4 illustrates a large scale system. It has two circuits
consisting of 5,421 components. The overall system contains 222 segments.
Using the RA program, it takes about half of a second to calculate the
system reliability indices on a personal computer (Pentium 4 CUP 2.40GHz,

Figure 10.4 System for Case Study Two

58

512MB of RAM). A reliability analysis report for the system shown in

Figure 10.4 is shown in Appendix A.
Figure 10.5 is part of the system shown in Figure 10.4. A small circuit
C3 fed by substation Sub2 is added to the original system in Figure 10.5.

L_C1

Sub 2
DG
C3

L_C32
L_C31

Figure 10.5 Addition of Substation and DG to System Shown in

Figure 10.4
A DG is placed next to Sub2, which has the same effect as putting it
in the substation, because there is not any load between Sub2 and the DG.
Line L_C1 is the component that exists in the original system (prior to the
addition of Sub2 and C3) and is very close to Circuit C3.
When circuit C3 is heavily loaded, C3 is not able to supply any power
to its adjacent circuit. Under this condition, the down time for line L_C1 in
the original system is 0.805 Hr/Yr. Now we activate the DG and recompute
the down time for L_C1. We get the same result 0.805 Hr/Yr, which means
59

locating the DG in Sub2 does not improve the reliability of L_C1 at all.
From Case Study One, we can predict that placing the DG in Sub2 will not
increase the reliability of the rest of the original system either (the segments
further away from the adjacent circuit have less improvement).

L_C1
Sub 2
DG

C3
L_C32
L_C31

Figure 10.6 DG at the End of Circuit

If the DG is placed at the end of circuit C3 connecting to L_C32, as
shown in Figure 10.6, the down time for line L_C1 drops to 0.360 Hr/Yr.
This significant change in the reliability of L_C1 is due to the change of the
DGs placement. When the DG is located in Sub2, C3 dose not have enough
remaining capacity to support its adjacent circuit. Placing the DG at the end
of C3 provides capability to pick up the load on line L_C1 if the failure
occurs in the original system. So the reliability of L_C1 dramatically

60

increases. Table 10.4 shows the system reliability improvement after adding
the alternative source Sub2 and distributed generator DG.
Table 10.4 System Reliability Improvement for Case Study Two
Without Alternate

With Alternate

Feed

Feed and DG

SAIDI(Hrs/yr)

0.72

0.54

25%

CAIDI(Hrs/yr)

9.12

6.03

34%

System indices

Improvement

10.4 Case Study Three

Previous reliability calculations have been performed for static load
models and inherently make the assumption that system reliability is
independent of load. In this case study, we investigate the reliability
improvement over a time varying load curve.

61

Sub 2

L_C1
C3
L_C32

Figure 10.7 Circuit for Case Study Three

Figure 10.7 shows the same part of the circuit that we studied in Case Two.
Now we look into the load curve of line L_C32 for a weekday in January.
As illustrated in Figure 10.8, the estimated load of L_C32 fluctuates during
the 24-hour period, and reaches its peak value around 6pm to 7pm, when
most of people return home and turn on their electric utilities.

62

0.8

0.6
0.5
0.4

Down Time (Hr/yr)

0.7

0.3
0.2
0.1

Estimated Load

Down Time

Figure 10.8 Down Time Variation with Varying Load of L_C32

Along with time, the variation of the reliability of line L_C1, which is
reflected by its down time, is also shown in Figure 10.8. We can see that
from 12am to 4pm, when the load in line L_C32 remains relatively low, the
down time for line L_C1 stabilizes at 0.36Hr/Yr. When the load of line
L_C32 rapidly grows in the evening, it triggers a dramatic increase in the
down time, which jumps from 0.36Hr/Yr to 0.805Hr/Yr. After that summit
period, from 6pm to 9pm, the down time of line L_C1 decreases to
0.36Hr/Yr again. This change of the reliability of line L_C1 with the
variation of load in line L_C32 is because the reliability depends on the
availability of the alternative source. While the load in the adjacent circuit

63

goes up to the point close to being over loaded, it has no remaining

capability to support the original circuit any more. From the reliability
analysis sets explained in Chapter 5, set SAF is empty in this case. If any
failure occurs in the circuit except in set SSL, the load point of interest (here
it is L_C1) will experience the interruption for the entire repair time of the
failure point. This is why it annual down time significantly increases.
If the time varying load on line L_C32 is reduced such that its peak
dose not exceed 4000KW, as illustrated in Figure 10.9. In this case the
reliability of line L_C1 remains at a high level throughout the load cycle,
which is indicated by the constant down time 0.36Hr/Yr.

0.4

0.2

0.1

Estimated Load

Down Time

Figure 10.9 Down Time with Varying Load of L_C32

64

Down Time (Hr/Yr)

0.3

11. Conclusions and Further Research

11.1. Conclusions
In this work, we have presented a reliability analysis algorithm. Set
calculations coupled with circuit traces are used to calculate the reliability of
a given load point and an entire system. An application has been developed
to implement this algorithm. The placement of distributed generation and its
effects on reliability is investigated. An evaluation of reliability over time
varying load curves is also presented. Three case studies are demonstrated,
where reliability indices produced by the reliability analysis program for
particular segments and the entire system provide concrete figures to assess
reliability improvements.
Conclusions from the investigations are:
The created reliability analysis algorithm is fast enough on large
systems to be used in interactive design studies
A new reliability index, Relative_CAIDI, has been proposed which
makes it easier for a design engineer to find circuit locations in need
of improvement
Placing distributed generators further out on a circuit, instead of
locating them in the substation, can help enhance a systems reliability
It is practical to estimate reliability as a function of time (loading).

65

11.2. Further Research

Besides adding a distributed generator, there are also other ways to
enhance a systems reliability. For example, we can change the system
structure by adding more protective devices or by moving sectionalizing
devices forward or backward. And then if we recalculate the reliability
indices for the entire system, and compare them with the original values, we
can see whether the change improves the reliability or not.
We have proposed a new reliability index, Relative_CAIDI. If the
Relative_CAIDI for a given load point is less than 1, it means the reliability
of the affected customers is less than average. Further research could focus
on automated system structure modifications which are base upon the value
of Relative_CAIDI.

66

12. Reference
[1]

C.C. Liu, G.T. Heydt, A. G. Phadke et al, The Strategic Power

Infrastructure Defense (SPID) System, IEEE Control System
Magazine, Vol. 20, Issue 4, August 2000, pp. 40 - 52.

[2]

Joel Adams, Sanford Leestma, Larry Nyhoff, C++ Introduction to

computing (second edition), pp. 762-775.

[3]

D.P. Gaver, F.E. Montmeat, A.D. Patton, Power system reliability:

IMeasures of reliability and methods of calculation. IEEE Trans.
Power Apparatus Syst., Vol. 83, pp. 727-737, July, 1964.

[4]

C. E. Montmeat, A.D. Patton, J. Zemkowski, D. J. Cumming, Power

system reliability IIApplications and a computer program, IEEE
Trans. Power Apparatus Syst., Vol. PAS-87, pp. 636-643, July, 1965.

[5]

Roy Billinton, Kenneth E. Bollinger, Transmission system Reliability

Evaluation Using Markov Processes, IEEE Trans. Power Apparatus
Syst., Vol. PAS-87, no. 2, pp. 538-547, Feb. 1968.

[6]

Robert P. Broadwater, Hesham E. Shaalan, Distribution system

reliability and restoration analysis, Electric Power System Research,
29 (1994) pp. 203-211

[7]

Barker, P., et al., Integration of Distributed Resources in Electric

Utility Systems: Current Interconnection Practice and Unified
approach, Power Technologies, Inc., EPRI Report TR-111489.

[8]

Dick, E. P., et al., Integration of Distributed Resources in Electric

Utility Distribution Systems: Distribution System Behavior Analysis
for Suburban Feeder, Ontario Hydro, EPRI Report TR-111490.

67

[9]

Roy. Billinton, Ronald N. Allan, Power-system Reliability in

Perspective, IEE J. Electron. Power, vol.30, pp.231-236, March 1984.

[10] Robert P. Broadwater, Murat Dilek, Jeff C. Thompson and D. Lloyd,

Design of Integrated software for Reconfiguration, Reliability, and
Protection System Analysis, IEEE, pp. 1011-1015, 2001.
[11] Development of Distribution System Reliability and Risk Analysis
Models, EPIR Rep. No. EL-2018, Vols.2 and 3, Electric Power
Research Inst., Palo Alto, CA Aug. 1981.
[12] Distribution Engineering Workstation Data Schema, Version 1.01,
EPIR EL-7249-V5 Project 3079-03 Final Report, Sep. 1995.
[13] Robert Broadwater, Jeff Thompson, Mike Ellis, Harry Ng, Nand
Singh, Darrel Loyd, Application Programmer Interface for the EPRI
Distribution Engineering Workstation, IEEE Transaction On Power
Systems, Vol. 10, No 1, pp 499-505, February 1995.
[14] R.P. Broadwater, J.C. Thompson, T.E. McDermott, Pointers and
linked lists in Electric Power Distribution Circuit Analysis, Proc.
IEEE Power Industry Computer Applications (PICA) Conf. pp. 16-21,
MD, 1991.
[15] Charlie Alan Jones, Operational Extensions to a Power Distribution
Design Workstation for Enhanced Emergency Restoration, VA Dec.
1990.

68

Appendix A: Reliability Analysis Report for the System Shown in Figure 10.4

=========================================
====== Reliability Analysis Report ======
=========================================
Report created at Mon Apr 07 10:28:39 2003
==========================================
Reliability Indices Of The system Analyzed
==========================================
System Name
=============
SUB_A1_001

SAIDI
=======
0.54

CAIDI
=======
6.03

============================================
Reliability Indices Of The Circuits Analyzed
============================================
Circuit Local Name
====================
F9818_C1
F9786_C2
no local name

GLN_X
GLN_Y
========== ==========
180481
261676
180384
261614
185831
248589

SAIDI
=======
0.17
1.25
0.20

CAIDI
=======
3.49
14.52
0.07

69

============================================
Reliability Indices for all the Segments
============================================
Circuit :
F9818_C1
=================================================================================================================
Segment Local Name
GLN_X
GLN_Y
CAIDI_Rel
# of Customers
DownTime
CustomerDownTime
RestorationTime
---------Hrs
---Hrs
=================== ======== ======== =========== ================ ========== ================== ================
F9818_C1
180481
261676
0.000
0
0.000
0.000
0.000
CB1
180505
261736
2.830
0
1.235
0.000
1.153
hPF_J324
182021
262866
2.138
96
1.635
156.960
1.527
NO
181254
262970
2.138
0
1.635
0.000
1.527
hPF_J202
182326
263081
1.864
124
1.875
232.500
1.751
NO
181209
263506
1.864
0
1.875
0.000
1.751
f_25K
183922
263853
2.523
0
1.385
0.000
1.258
R3_560A_2A2D_
185601
264352
1.363
0
2.565
0.000
1.187
hPF_J784
193816
264539
1.228
0
2.845
0.000
1.317
no local name
195047
264544
1.211
0
2.885
0.000
1.335
no local name
193814
264778
1.302
0
2.685
0.000
1.242
hPF_J931
193808
265595
1.179
0
2.965
0.000
1.372
DscSw_192544:265150 192560
265168
1.179
0
2.965
0.000
1.372
DscSw_192544:265150 192539
265135
1.133
0
3.085
0.000
1.428
DscSw_192544:265150 192509
265152
1.163
0
3.005
0.000
1.391
NO
192424
265123
1.163
0
3.005
0.000
1.391
f_65K
192683
266154
0.923
94
3.785
355.790
1.590
hPF_2698
191455
265912
1.077
60
3.245
194.700
1.502
NO
190839
266018
1.077
0
3.245
0.000
1.502
hPF_1642
191254
266519
1.133
59
3.085
182.015
1.428
NO

190670

266297

1.133

70

3.085

0.000

1.428

R3_200A_2A2D_K2
f_65K
hPF_3292
NO
hPF_3199
hPF_3135
NO
hPF_3204
(NO)
iPS_J172
f_100K_K3
jCF_J716
jCF_J529
R3_200A_2A2D_R
hPF_J049
NO
f_100K_R6
jCF_J1125
jCF_J1460
hPF_J959
hPF_J773
hPF_J1590
hPF_J1334
jCF_J1445
hPF_J3371
f_80K_R5
hPF_3144
hPF_J1031
hPF_3296
NO
hPF_3189
hPF_3268

191727
193050
193356
193801
193363
193726
193558
193747
193885
193959
192414
192853
192264
191243
191212
190529
190225
188786
188992
189012
189261
189374
189938
190007
190256
188742
188500
188666
189107
188759
188460

266675
267149
268085
267769
267715
267334
267708
267334
267735
267517
266904
267895
267670
266593
266710
266865
266632
265519
265583
265619
265664
265784
265865
265885
266516
268071
268180
267806
268378
268436
268864

1.039
0.893
0.948
0.948
0.928
0.959
0.959
0.928
1.039
1.003
0.756
0.737
0.737
0.991
0.928
0.928
0.793
0.772
0.772
0.752
0.733
0.752
0.746
0.772
0.752
0.959
0.890
0.721
0.919
0.919
0.948

0
56
37
0
46
36
0
29
0
0
91
1
1
0
0
0
0
1
2
1
6
2
1
1
3
0
57
110
68
0
34

3.365
3.915
3.685
3.685
3.765
3.645
3.645
3.765
3.365
3.485
4.625
4.745
4.745
3.525
3.765
3.765
4.405
4.525
4.525
4.645
4.765
4.645
4.685
4.525
4.645
3.645
3.925
4.845
3.805
3.805
3.685

0.000
219.240
136.345
0.000
173.190
131.220
0.000
109.185
0.000
0.000
420.875
4.745
4.745
0.000
0.000
0.000
0.000
4.525
9.050
4.645
28.590
9.290
4.685
4.525
13.935
0.000
223.725
532.950
258.740
0.000
125.290

1.294
1.444
1.417
1.417
1.448
1.401
1.401
1.448
1.294
1.340
1.594
1.636
1.636
1.207
1.289
1.289
1.280
1.315
1.315
1.350
1.385
1.350
1.362
1.315
1.350
1.215
1.308
1.659
1.303
1.303
1.262

188435

268876

0.919

41

3.805

156.005

1.303

71

hPF_3223
NO
f_65K_R1
(DS)
jCF_J1524
(NO)
jCF_J1712
f_50K_R7
hPF_J050
hPF_J1640
jCF_J1144
hPF_J1626
hPF_2655
hPF_2630
hPF_J930
DscSw_191924:264803
DscSw_191924:264803
DscSw_191924:264803
jCF_J875
hPF_J774
NO
R3_200A_2A2D_L
f_65K_L1
f_10K
f_10K
jCF_J962
jCF_J1193
jCF_J173
f_8OK_K1
hPF_J325
hPF_J723
NO

188908
188819
189406
190136
190128
190796
190679
189769
190307
189497
189662
189945
191313
191372
193121
191890
191917
191942
194039
193787
195410
193643
193157
193344
193034
192954
193366
193628
185916
181266
180874

268937
268888
268063
268587
270064
269062
268962
268288
267212
267479
267049
266715
266318
266145
265824
264796
264829
264804
265657
264306
264687
263158
261239
260335
260907
261404
261853
262768
264389
262368
262515

0.969
0.969
0.609
0.457
0.450
0.457
0.457
0.578
0.523
0.928
0.959
0.928
1.014
1.133
1.039
1.039
1.026
1.003
1.246
1.283
1.283
1.179
1.104
1.039
1.054
1.118
1.133
1.133
0.860
2.138
1.991

0
0
43
102
1
0
0
0
41
0
2
2
85
38
1
0
0
0
0
0
0
0
0
3
2
1
0
1
48
75
81

3.605
3.605
5.735
7.655
7.775
7.655
7.655
6.045
6.685
3.765
3.645
3.765
3.445
3.085
3.365
3.365
3.405
3.485
2.805
2.725
2.725
2.965
3.165
3.365
3.315
3.125
3.085
3.085
4.065
1.635
1.755

0.000
0.000
246.605
780.810
7.775
0.000
0.000
0.000
274.085
0.000
7.290
7.530
292.825
117.230
3.365
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
10.095
6.630
3.125
0.000
3.085
195.120
122.625
142.155

1.234
1.234
1.521
2.030
2.062
2.030
2.030
1.526
1.688
1.289
1.248
1.289
1.594
1.428
1.557
1.557
1.576
1.613
1.298
1.261
1.261
1.277
1.341
1.401
1.386
1.346
1.329
1.329
1.652
1.527
1.639

180624

262450

1.991

1.755

0.000

1.639

72

hPF_J634
hPF_J633
NO
hPF_J635
NO
hPF_J636
hPF_J722

180875
180862
180539
180892
180414
180912
180866

262913
262895
264296
263000
263470
262981
262334

2.138
1.947
1.947
2.138
2.138
2.138
2.138

70
82
0
34
0
89
62

1.635
1.795
1.795
1.635
1.635
1.635
1.635

114.450
147.190
0.000
55.590
0.000
145.515
101.370

1.527
1.676
1.676
1.527
1.527
1.527
1.527

Circuit :
F9786_C2
=================================================================================================================
Segment Local Name
GLN_X
GLN_Y
CAIDI_Rel
# of Customers
DownTime
CustomerDownTime
RestorationTime
---------Hrs
---Hrs
=================== ======== ======== =========== ================ ========== ================== ================
F9786_C2
180384
261614
0.000
0
0.000
0.000
0.000
CB2
180321
261620
1.123
0
12.925
0.000
1.010
hPF_J1594
179665
260846
1.058
25
13.725
343.125
1.072
NO
180129
260782
1.058
0
13.725
0.000
1.072
no local name
179553
260754
1.127
0
12.881
0.000
1.006
f_100K
179396
260609
1.041
0
13.941
0.000
1.009
hPF_J743
179060
260222
0.906
61
16.021
977.281
1.159
NO
179105
258904
0.906
0
16.021
0.000
1.159
hPF_J742
178961
260125
0.913
85
15.901
1351.585
1.150
hPF_J820
179509
258517
1.028
37
14.121
522.477
1.103
NO
178402
258403
1.000
0
14.521
0.000
1.134
hPF_J819
179438
258765
1.000
0
14.521
0.000
1.134
hPF_J1426
179746
257632
1.011
56
14.361
804.216
1.122
NO
181605
258344
1.011
0
14.361
0.000
1.122
hPF_J327
179718
256941
1.064
52
13.641
709.332
1.066
NO
178606
257875
1.043
0
13.921
0.000
1.087
hPF_J328

179689

258095

1.043

73

13.921

0.000

1.087

hPF_J1226
NO
f_25K
f_80K
hPF_J1045
NO
f_65K
(NO)
f_65K
f_65K
jCF_J1749
f_65K
f_65K
hPF_J029
NO
f_65K
(DS)
Sect_L1
hPF_J1420
jCF_J1128
f_100K_L7
f_65K
Sect_L2
hPF_J3344
hPF_J3345
NO
(DS)-BA
f_65K
hPF_J288
hPF_J289
hPF_J287
NO

180497
181011
181566
184676
184907
185280
185481
186165
186672
187084
187101
188299
188718
189090
189177
189283
192270
192138
189685
189726
190089
193362
193832
194513
194499
195191
194165
196336
197790
198014
197271

256315
258103
256304
256512
255449
254494
256542
263497
256555
256696
257018
256620
256583
256631
258814
256627
256608
256341
253519
252750
253913
256429
256351
256309
256391
254911
256420
256765
257111
257133
257008

0.963
0.963
1.114
1.098
1.072
1.069
1.089
1.127
1.038
1.060
1.057
1.077
1.101
0.978
0.978
1.101
1.117
1.049
1.031
1.040
1.045
1.079
1.087
0.970
1.043
1.043
1.017
0.969
0.965
0.950
0.997

70
0
4
8
0
0
5
0
10
5
0
7
2
41
0
4
0
9
2
1
0
7
5
48
0
0
12
8
97
84
0

15.081
15.081
13.031
13.221
13.541
13.581
13.331
12.881
13.981
13.691
13.731
13.481
13.181
14.841
14.841
13.181
13.001
13.841
14.081
13.961
13.891
13.451
13.361
14.961
13.921
13.921
14.281
14.981
15.041
15.281
14.561

1055.670
0.000
52.124
105.768
0.000
0.000
66.655
0.000
139.810
68.455
0.000
94.367
26.362
608.481
0.000
52.724
0.000
124.569
28.162
13.961
0.000
94.157
66.805
718.128
0.000
0.000
171.372
119.848
1458.977
1283.604
0.000

1.178
1.178
1.016
1.022
1.046
1.049
1.034
1.006
1.074
1.056
1.059
1.043
1.025
1.159
1.159
1.025
1.016
1.081
1.100
1.091
1.084
1.043
1.044
1.169
1.087
1.087
1.116
1.158
1.175
1.194
1.137

197155

258029

0.997

14.561

0.000

1.137

74

hPF_J286
NO
f_25K
f_25K
jCF_J886
f_65K
jCF_J414
(NO)-BA
Sect_L3
hPF_J910
hPF_J611
jCF_JO30
jCF_J187
hPF_J3347
NO
jCF_J777
hPF_J3360
hPF_J1402
hPF_J1180
hPF_J1357
hPF_J1358
hPF_J1427
hPF_J1428
hPF_J1471
NO
hPF_J1472
NO
hPF_J1181
OpenSwtDev_4_1
hPF_J1179

196960
196891
195565
195355
195186
194944
194158
193825
193056
192975
193021
194250
194246
193007
193034
194287
192801
193878
194195
195667
197076
196887
195852
195805
195760
196946
195729
197864
195886
196207
194686

256944
258465
253759
257717
257919
258085
258802
259165
257105
257621
258383
260251
260455
258955
254880
254440
251459
251296
251334
251476
251187
251295
253722
253530
253327
253401
253134
254245
251498
249831
251389

0.994
0.994
0.935
1.006
1.008
0.996
1.008
1.017
1.103
1.074
1.040
1.022
1.031
1.031
0.978
0.978
1.106
0.970
0.994
1.002
0.963
1.037
0.914
1.002
1.043
1.043
1.034
1.028
1.093
1.093
1.074

0
0
14
2
0
3
1
0
0
1
4
1
1
1
48
0
1
88
74
51
78
20
86
51
21
0
0
6
0
0
0

14.601
14.601
15.531
14.431
14.401
14.581
14.401
14.281
13.161
13.521
13.961
14.201
14.081
14.081
14.841
14.841
13.121
14.961
14.601
14.481
15.081
14.001
15.881
14.481
13.921
13.921
14.041
14.121
13.281
13.281
13.521

0.000
0.000
217.434
28.862
0.000
43.743
14.401
0.000
0.000
13.521
55.844
14.201
14.081
14.081
712.368
0.000
13.121
1316.568
1080.474
738.531
1176.318
280.020
1365.766
738.531
292.341
0.000
0.000
84.726
0.000
0.000
0.000

1.141
1.141
1.190
1.125
1.125
1.134
1.125
1.116
1.028
1.056
1.091
1.109
1.100
1.100
1.159
1.159
1.025
1.169
1.141
1.131
1.178
1.094
1.241
1.131
1.087
1.087
1.097
1.103
1.037
1.037
1.056

NO

195965

249755

1.074

13.521

0.000

1.056

75

hPF_J1178
hPF_J1401
NO
OpenSwtDev_4_2
NO
hPF_J3361
NO
no local name
hPF_J3346
jCF_J1374
jCF_J1023
hPF_J1067
jCF_J585
hPF_J028
f_65K
jCF_J1293
f_80K
f_65K
jCF_J1657
R3_140A_2A2D_K
f_10K
f_65K
f_65K
f_65K
f_40K
hPF_J1554
NO
(DS)
f_40K
f_65K
hPF_J1055
f_65K

194513
194536
195332
196178
196307
194085
194051
192853
193024
191328
190568
190170
189971
188566
187845
188061
184119
183789
183408
182751
182717
182414
182478
182528
180334
182566
182353
182633
182699
182611
182591

251371
251365
250996
249793
248871
251321
249806
253908
255164
256614
256372
256657
256610
256636
256738
258484
256590
256512
256545
256315
253681
251751
251321
250604
249969
248692
249013
247634
247416
246991
246781

0.992
1.043
1.002
1.002
1.002
1.061
1.061
1.117
0.970
1.117
1.117
1.087
1.117
0.916
1.074
1.065
1.085
1.027
1.117
1.005
0.995
0.992
0.965
0.916
0.899
0.970
0.970
1.011
0.990
0.951
0.912

72
0
0
0
0
0
0
0
61
1
0
12
1
69
5
4
10
16
1
0
1
2
13
10
1
0
0
0
6
7
5

14.641
13.921
14.481
14.481
14.481
13.681
13.681
13.001
14.961
13.001
13.001
13.361
13.001
15.841
13.511
13.631
13.381
14.131
13.001
14.441
14.591
14.641
15.041
15.841
16.141
14.961
14.961
14.361
14.661
15.271
15.911

1054.152
0.000
0.000
0.000
0.000
0.000
0.000
0.000
912.621
13.001
0.000
160.332
13.001
1093.029
67.555
54.524
133.810
226.096
13.001
0.000
14.591
29.282
195.533
158.410
16.141
0.000
0.000
0.000
87.966
106.897
79.555

1.144
1.087
1.131
1.131
1.131
1.069
1.069
1.016
1.169
1.016
1.016
1.044
1.016
1.237
1.043
1.053
1.037
1.083
1.016
1.017
1.025
1.028
1.050
1.094
1.110
1.054
1.054
1.011
1.028
1.052
1.096

183088

246426

0.980

14.811

74.055

1.036

76

f_65K
Sect_K1
(DS)
f_65K
f_65K
jCF_J545
f_65K_K6
(NO)
jCF_J1141
jCF_J180
jCF_J442
f_65K_K5
f_40K
(NO)
f_40K
hPF_J985
hPF_J610
NO
f_65K_K8
f_40K
f_25K
hPF_J609
(NO)
f_65K
f_65K
hPF_J1555
f_65K
f_65K
hPF_J1070
f_65K
hPF_J1553
(DS)

183232
183202
182613
183830
185184
184413
184190
184554
183083
184196
184374
184538
190600
190791
190617
184750
186768
187257
188179
192893
193377
187738
187554
186628
185773
186979
185615
184724
184121
182545
182547

246017
245624
245625
245502
243814
237521
237422
232668
236052
239887
240293
240273
240019
234982
240058
240866
245476
244649
245916
246124
246126
245639
247735
245584
243408
243496
243030
242553
245743
248543
249756

0.980
1.014
0.992
0.960
0.993
1.005
0.872
0.872
0.866
1.005
1.005
0.884
0.792
0.792
0.797
0.983
0.994
0.994
0.954
0.935
0.926
0.970
1.014
0.941
0.965
0.957
0.948
0.938
1.014
0.995
0.986

6
0
1
6
0
1
25
0
1
1
4
10
11
0
22
5
0
0
3
0
1
20
0
12
7
2
16
32
1
2
1

14.811
14.321
14.641
15.121
14.621
14.441
16.651
16.651
16.771
14.441
14.441
16.421
18.321
18.321
18.221
14.761
14.601
14.601
15.221
15.521
15.671
14.961
14.321
15.421
15.051
15.171
15.321
15.471
14.321
14.591
14.721

88.866
0.000
14.641
90.726
0.000
14.441
416.275
0.000
16.771
14.441
57.764
164.210
201.531
0.000
400.862
73.805
0.000
0.000
45.663
0.000
15.671
299.220
0.000
185.052
105.357
30.342
245.136
495.072
14.321
29.182
14.721

1.036
1.008
1.031
1.053
1.025
1.017
1.133
1.133
1.142
1.017
1.017
1.123
1.221
1.221
1.216
1.039
1.028
1.028
1.058
1.075
1.083
1.054
1.008
1.069
1.047
1.056
1.064
1.072
1.008
1.025
1.037

183121

251251

0.968

15.001

45.003

1.056

77

f_40K
jCF_J879
jCF_J948
hPF_J825
jCF_J1247
(NO)
hPF_J1143
hPF_J1071
hPF_J1389
hPF_J1237
f_65K
f_100K
f_100K
hPF_J1225
jCF_J1285
PTS
hPF_J432
hPF_J431
jCF_J1250
f_80K_H3
f_40K
f_40K
R3_70A_2A2D_R
f_40K
f_40K_R1
jCF_J1022
jCF_J1050
jCF_J1068
hPF_J964
(NO)
f_25K
jCF_J516

184241
190695
190089
188516
187107
189473
186186
184768
183680
182780
182804
182751
182500
180295
180171
177599
178140
178392
179032
179341
179452
179589
179636
179897
179677
182023
181613
179953
180008
177518
178435

251255
248868
249115
249758
248078
252523
251300
251276
250833
252614
253846
256586
256475
256310
256331
256325
256344
256358
256353
256097
253930
251116
250890
247918
245736
245669
245614
244043
244494
246849
246905

0.958
0.960
0.960
0.952
0.960
0.968
0.952
0.952
0.952
0.989
0.975
1.050
1.006
0.963
1.117
1.127
1.067
1.096
1.117
1.007
1.004
0.977
0.922
0.888
0.803
0.798
0.798
0.798
0.788
0.922
0.914

1
2
0
6
0
0
2
4
4
2
4
12
18
78
2
0
26
16
0
10
0
10
8
6
10
1
3
2
9
0
1

15.151
15.121
15.121
15.241
15.121
15.001
15.241
15.241
15.241
14.681
14.891
13.831
14.431
15.081
13.001
12.881
13.601
13.241
13.001
14.411
14.461
14.861
15.741
16.341
18.071
18.191
18.191
18.191
18.431
15.741
15.891

15.151
30.242
0.000
91.446
0.000
0.000
30.482
60.964
60.964
29.362
59.564
165.972
259.758
1176.318
26.002
0.000
353.626
211.856
0.000
144.110
0.000
148.610
125.928
98.046
180.710
18.191
54.573
36.382
165.879
0.000
15.891

1.065
1.065
1.065
1.073
1.065
1.056
1.073
1.073
1.073
1.034
1.042
1.065
1.101
1.178
1.016
1.006
1.062
1.034
1.016
1.097
1.100
1.124
1.173
1.207
1.288
1.296
1.296
1.296
1.314
1.173
1.181

180165

247327

0.915

15.861

63.444

1.182

78

jCF_J904
hPF_J1425
f_25K
f_65K
hPF_J1593

179289
179744
179524
179501
179718

255367
257875
258788
259522
261024

0.999
1.019
1.114
1.101
1.052

3
57
1
6
17

14.531
14.241
13.031
13.181
13.805

43.593
811.737
13.031
79.086
234.685

1.107
1.112
1.016
1.025
1.078

Circuit :
no local name
=================================================================================================================
Segment Local Name
GLN_X
GLN_Y
CAIDI_Rel
# of Customers
DownTime
CustomerDownTime
RestorationTime
---------Hrs
---Hrs
=================== ======== ======== =========== ================ ========== ================== ================
no local name
185831
248589
0.000
0
0.000
0.000
0.000
no local name
186059
248471
0.667
1
0.100
0.100
5.000
no local name
185811
247887
0.667
0
0.100
0.000
5.000

79

Vita
Dan Zhu was born on September 4, 1977 in Jiangmen, Guangdong
Province, China. She received her BSEE degrees from South China Normal
University, Guangzhou, Guangdong Province, China, in June 2000.
Before she came to Virginia Tech to pursue her master degree, she was
employed as a Communication Engineer by Electric Power Bureau of
Jiangmen, Guangdong, China.

80