A reliability data base for performance and predictive assessment of electric power systems

A RELIABILITY DATA BASE FOR PERFORMANCE AND PREDICTIVE ASSESSMENT OF ELECTRIC R Billinton University of Saskatchewan Canada ABSTRACT The Canadian Electrical Association CEA Equipment Rel

A RELIABILITY DATA BASE FOR PERFORMANCE AND PREDICTIVE ASSESSMENT OF ELECTRIC

R Billinton

University of Saskatchewan

Canada

ABSTRACT

The Canadian Electrical Association (CEA)

Equipment Reliability Information System

(ERIS) was initiated in 1975 and consists of

a comprehensive procedure for reporting

generation and transmission equipment

performance and developing reliability

parameters from the basic data A similar

procedure is now being developed for

distribution equipment These parameters

are the basic data used in reliability

assessment of generation, transmission and

distribution systems The CEA-ERIS data

base will be briefly described in this paper

and related to the predictive reliability

assessment performed by Canadian electric

power utilities

The CEA has also initiated the Electric

Power System Reliability Assessment (EPSRA)

procedure which is designed to provide data

on overall power system performance This

procedure is in the process of evolution and

at the present time contains systems for

compiling information on Bulk Electricity

Point performance and Customer Service

Continuity statistics This paper will

describe the EPSRA procedure and illustrate

some of the information generated The

EPSRA and ERIS data bases provide the basic

information to quantitatively assess the

past performance of the bulk electricity

system and the basic component data required

to predict future performance

INTRODUCTION

Overall reliability evaluation of an

electrical power system can be divided into

the two basic elements of past performance

assessment and future performance prediction

This paper illustrates the procedures which

have been instituted by the Consultative

Committee on Outage Statistics (CCOS) of the

Canadian Electrical Association (CEA) to

provide a uniform and consistent reporting

procedure for assessing past system

performance and providing the data required

to predict future system performance

Effective quantitative assessment of power

system reliability requires, in addition to

other basic information, suitable

mathematical models and realistic failure

and repair statistics for the relevant

system components The failure and repair

statistics are normally obtained from past

operating data of comparable components and

subcomponents in the system Data collection

is therefore an integral and indispensable

part of quantitative power system reliability

analysis All the major electric power

utilities in Canada participate in a single

data collection and analysis system called

M Oprisan and I M Clark Canadian Electrical Association Canada

the Zquipment Reliability Information System (ERIS) of the Canadian Electrical

Association (CEA) CEA started collecting data on generation and transmission outages

in 1977 and since then has published a number of reports in both areas (1) and (2) This paper illustrates the extent to which the CEA data base can be utilized in the development of basic reliability models and their utilization in the computation of reliability indices

CEA has also initiated an Electric Power System Reliability Assessment (EPSRA) procedure which is designed to provide data

on the past performance of an overall power system This procedure is in the process of evolution and at the present time contains systems for compiling information on bulk system disturbances, bulk system delivery point performance and customer service continuity statistics This paper illustrates the Electric Power System Reliability Assessment (EPSRA) procedure and its present level of development

A power system can be broadly categorized by three functional zones containing generation, transmission and distribution facilities These three zones can be combined to provide

a practical and consistent framework for predictive reliability evaluation containing three hierarchical levels Billinton et

al (7) Hierarchical Level one {HLI) is composed of only the generation facilities Hierarchical Level two (HLII) includes both the generation and the transmission facilities Hierarchical Level three (HLIII) encompasses all three functional zones The functional zones and hierarchical levels are shown in Figure 1 There is a long history

of reliability application at the HLI level, where the primary objective is the

determination of the adequacy of the total system generation to satisfy the total system load demand

generation hierarchical level I

transmission hierarchical level II facilities HL II

distribution hierarchical level III facilities | [®————— HL TII

Figure 1 Hierarchical levels in

power system reliability evaluation

256

HLII evaluation is still at its infancy and

while there is published material available

Billinton (3), (4), Allan et al (5) and (6),

there is still no consensus regarding

requirements, techniques and indices

Evaluation at the HLIII level is a major

task and if done at all is accomplished by

using the HLII indices as input to the

distribution functional zone

THE EQUIPMENT RELIABILITY INFORMATION

SYSTEM (ERIS)

The basic structure of ERIS is shown in

Figure 2

The following is a brief description of the

above reporting systems with particular

emphasis on their ability to support

predictive reliability assessment

Generation Equipment Status Reporting System

The generation data reporting system (8)

employs a continuous state monitoring

procedure for each generating unit The

states in which a unit can reside have been

grouped into eleven categories These states

are numbered 11 through 16 for the available

states (operating or capable of operation)

and 21 through 25 for the unavailable

(outage) states These 11 states provide a

generalized eleven-state model for each

generating unit

States 12, 13, 15 and 16 provide a three

dimensional aspect, as these states also

include information on the extent of the

unit derating State 21 which is the full

forced outage state has an “outage type”

associated with it which designates the

outage as a sudden forced outage, an

immediately deferrable forced outage, a

deferrable forced outage or a starting

failure

The time spent in any of the states can be

easily calculated from the recorded data and

important generating unit parameters such as

the forced outage rate (FOR) and the derating

adjusted forced outage rate (DAFOR) can be

computed using the relevant state residence

times The continuous state monitoring

approach also provides detailed information

on transition rates associated with the

model These transition rates are essential

elements in frequency and duration (F&D)

analysis and in multi-state modeling of

generating units

Generating Unit Models

A generating unit is modeled to suit a particular application The amount of complexity involved will depend on the detail required for the particular application, the information available for the purpose and other considerations The more frequently used generating unit models and the ability

of ERIS to support these models are discussed

in the following sections

The Two State Model

The simplest representation of a generating unit is the two-state model shown in Figure

3 This model containing an UP and a DOWN state has been used extensively in power system reliability analysis to represent base load generating units

Unit

Up

Unit Down

Two-state model of a generating unit, Figure 3

The fundamental parameters required in the model are the failure and repair rates (A and#) These rates can be estimated from the number of departures from a particular state and the residence time in that state The departures and the residence times can be easily obtained from the ERIS data base If the probability of being in the DOWN state is estimated by the conventional FOR then the transition rates must be consistent with this value The most consistent approach is to estimate the failure rate from the basic data and to calculate the repair rate using the estimated FOR

The fundamental parameters required to complete the two-state model have been calculated for hydraulic and fossil units using the 1985-89 ERIS data and are shown in Table 1 which also includes the Derating Adjusted Forced Outage Rate (DAFOR) discussed

in the next section

ERIS

Generation Equipment

Status Reporting

System

Transmission Equipment Outage Reporting System

Distribution Equipment Outage Reporting System

Figure 2

Basic components of the ERIS

287

Hydraulic Units

MCR Class FOR DAFOR Failure Rate

24 - 99 4.06 4.69 3.31

100 - 199 3.61 3.87 3.65

200 - 299 7.90 7.94 6.45

300 - 399 4.39 4.84 5.62

400 ~ 499 3.10 3.38 2.70

Fossil Units

MCR Class FOR DAFOR Failure Rate

60 - 99 7.63 15.84 11.37

100 - 199 5.43 8.32 15.04

200 - 299 4.07 6,20 14,90

300 - 399 10.07 14.42 15.75

400 - 599 6.70 8.10 7.28

TABLE 1 - 1985-89 Base load

generation data

It can be seen from the data shown in Table

1, that the ERIS data base provides all the

necessary information associated with the

conventional two state model frequently used

in power system reliability evaluation

Derating Adjusted Two-State Model

The two-state representation is the

appropriate model for generating units that

exist in either the UP OF DOWN state (i.e

no derated states) Many generating units

can, however, reside in a range of partial

output (i.e derated) states In order to

recognize the time spent in these derated

states and still create a two state model,

the derated state residence times are

apportioned to the UP and DOWN states which

gives rise to a DERATING ADJUSTED two-state

model Table 1 also shows the derating

adjusted forced outage rates (DAFOR) values

The repair rate in this case can be computed

from the estimated failure rate and the

DAFOR The fundamental continuous monitoring

approach utilized by the ERIS provides the

ability to recognize all the generating unit

capability levels and the times associated

with them This procedure can therefore

generate all the information required to

produce detailed models or reduced models as

required

Derated Models (Three-State Model)

The derating adjusted two-state model of a

generating unit is an approximation which

provides a pessimistic appraisal of the load

carrying capabilities of a unit or a system

of units A detailed multistate model

provides a more valid representation A

generating unit can exist in a large number

of derated states and therefore complete

modeling can be a difficult, if not

impossible, task As a first approximation,

the derating levels can be grouped in ten-

percent-of-capacity intervals giving rise to

an l1i-state model which includes 9 derated

states This model can be easily obtained

from the ERIS data base In many cases, the

three-state model shown in Figure 4 which

contains only one derated state is adequate

It has been found Billinton and Allan (9)

that a three-state (UP, DOWN and one DERATED

state) model offers most of the benefits of a multistate model, while keeping the

computational burden to a minimum There is

a growing tendency to represent large generating units by a three state representation rather than the conventional two-state model There is, however, no unique criterion for locating the single derating level used in the model Figure 4 shows the transition rates required ina three~state model These data can also be extracted from the ERIS data base A procedure for obtaining the transition rates and state probabilities associated with the Billinton et al (10)

Figure 4 Three~state generating unit model

Adequacy Assessment at Hierarchical Level I Generating capacity adequacy evaluation is a common procedure in most major utilities throughout North America Recent surveys have shown that virtually all Canadian utilities now use probabilistic rather than deterministic criteria to evaluate the adequacy of installed and planned generating facilities (7) The basic probabilistic techniques and the ability of ERIS to provide the required data can be briefly summarized

as follows

Loss of Load Expectation (9) Loss of load expectation (LOLE) is the most commonly used generation adequacy index (7) and is normally defined as the expected number of days in a year in which the load will exceed the generation The basic generating unit parameter used in the calculation of the LOLE is the unit FOR This is sometimes replaced by the DAFOR for large generating units which tends to give a pessimistic appraisal of the unit and therefore of the system adequacy Asa result, there is a growing tendency to utilize the three-state generating unit model shown in Figure 4 The ERIS data base provides all the basic generating unit information required for the calculation of the conventional LOLE index

Loss of Energy Expectation (9) Several Canadian utilities use loss of energy expectation (LOEE) as their HLI adequacy index This parameter is the expected energy that will be curtailed due to inadequate capacity As in the calculation of a LOLE index, either two, three or multistate generating unit models can be used to compute the LOEE, All the basic generating unit data for the LOEE technique can be readily obtained from the ERIS data base

258

Frequency and Duration Analysis (9) Line-Related Sustained Outages

The LOLE indicates the expected number of Class Km years (per 100 Unavailability

days or hours that load will be curtailed (kv) (km.a) km.a } (8)

The LOEE on the other hand estimates the

indices do not provide any indication of the 110-149 180,131 1.0609 0.076

number of times that load will be curtailed 150-199 5,846 0.4105 0.004

nor the extent of the curtailment and the 200-299 154,209 0.5434 0.096

average duration of such a curtailment 300-399 41,837 0.2988 0.037

This information can be obtained from a 500-599 39,126 0.4038 0.047

additional parameters required for such an

analysis are the transition rates associated

with the appropriate generating unit model

The ERIS data base contains all the

Line-Related Transient Outages

Kilometre generating unit state information and

therefore it is possible to obtain transition Voltage Class Gen a) ( or d00 ka a)

Reference (10) illustrates a direct

information for a three state model 150 _ 199 neo eae , 02000 `

Security Assessment at Hierarchical Level I 20g — s39 39 126 nh

A11 Canadian utilities are concerned with 600 ~ 792 30,218 0-0836

security assessment within their own system

and in the interconnected configuration

within which their own system resides These Terminal-Related Sustained Outages

assessments are done primarily using methods Voltage Terminal

which utilize deterministic criteria (7) Class Years Frequency Unavailability

Some work has been done in the area of (kv) (t.a} (per a) (3%)

Operating reserve assessment and

probabilistic evaluation of transient

stability (9), Billinton and Chu (11) The 110-149 7,165.5 0.1173 0.004

ERIS data base can provide almost all the 150-199 337.0 0.0534 0.002

basic data required in a probabilistic 200-299 4,301.0 0.1269 0.006

operating reserve assessment Billinton et 300-399 1,086.5 0.0782 0.005

Transmission Equipment Outage Reporting

The CEA-ERIS defines transmission equipment

as being rated at 110 kV and above and

associated with the transmission of electric Class Components Frequency Duration

Forced outages of transmission lines and

cables have been categorized into line/cable~ _

related outages and terminal-related outages 150-189 6n ng 0 0657 oe

The transmission line performance statistics 200-299 3,862.0 0.0683 61.9

are expressed on a “per 100 kilometre-year” 300-399 1 038,0 0.0511 839.2

(km.a) basis for line-related outages and on 500-599 "968.5 0.0723 433.0

classified according to their duration

Outages of less than one minute are

classified as transient outages and those of TABLE 3 - 1984-88 Transformer outage

one minute and longer are called sustained statistics involving

terminal eguipment

The reporting system on transmission

equipment is limited to forced outages, their

duration, primary cause and subcomponents

involved Other parameters necessary for

reliability analysis such as repair or

replacement times of any component can be

easily computed from the recorded data

Data for transmission lines for the 1984-88

period are shown in Table 2 The ERIS also

includes data on transformers, circuit

breakers, cables, synchronous and static

compensators, shunt reactors and shunt and

series capacitor banks A sample of

transformer data is shown in Table 3 Table

4 provides the circuit breaker outage

statistics for the year 1984-88,

259

Class Components Frequency Duration (kV) Years (a) (per a) (h)

110-149 10,680.0 0.0326 167.1

200-299 7,232.0 0.0441 83.4 300-399 2,703.0 0.0448 139.6 500-599 1,112.0 0.0990 152.5 600-799 2,219.5 0.0955 217.6

TABLE 4 - 1984-88 Circuit breaker

outage statistics involving integral subcomponents and terminal eguipment

It can be seen from the data shown in the above Tables 2, 3 and 4 that the ERIS data base provides the information required in the conventional two-state models frequently used in power system reliability evaluation

Distribution Equipment Outage Reporting System

CEA is now in the process of implementing this new reporting system which will complete the ERIS mandate as a system which covers equipment used by power utilities from generation to distribution (Figure 2)

Distribution equipment is defined as being rated at less than 110 kV, however, higher voltage equipment will also be included if

it is specifically associated with the distribution of electric power

Similarly to the transmission system, the reporting system on distribution equipment

is limited to forced outages, their duration, primary cause and subcomponents involved

ERIS data base on distribution equipment will provide, when completed, the necessary information for the conventional two-state models used in power system reliability

‘evaluation

ELECTRIC POWER SYSTEM RELIABILITY ASSESSMENT (EPSRA)

The basic structure of EPSRA is shown in Figure 5

The bulk electricity system (BES) parameters provide valuable HLII data The service continuity statistics collected at the distribution system level provide overall HLIII indices The following is a brief description of the salient features of the EPSRA components

BES - Significant Power Interruptions This is a relatively simple reporting procedure in which each participating utility provides annual data on the frequency and severity of significant power interruptions

on its system Typically these events will involve wide spread customer interruptions or localized power interruptions of an extended duration It is expected that these events will occur infrequently Disturbance severity is defined as the unsupplied energy

in an event and is measured in MW - minutes

This is transformed into “System Minutes” by dividing the unsupplied energy by the annual peak system load

The disturbances are grouped by severity as follows:

Degree 1 - an incident with a Severity

from 1-9 System Minutes Degree 2 - an incident with a Severity

from 10-99 System Minutes Degree 3 - an incident with a Severity

from 100-999 System Minutes

The basic definition of a significant power interruption is that is an event which originates on the BES and is of Degree 1 or higher severity

The reporting process was initiated on January 1, 1985 and two reports have now been released, the latest results being based on

37 utility years of data for the 1985-88 period

BES - Delivery Point Interruptions

This reporting procedure was initiated on January 1, 1988 and is intended to provide a centralized source of BES delivery point performance data at the national level which will allow utilities to compare their performance with that of other Canadian utilities The measurement system will focus

on the collection of all delivery point interruptions due to problems within the BES and therefore will include both adequacy and security considerations The interruptions will be divided into momentary and sustained events It is expected that this reporting procedure will provide some very important information for Canadian utilities and will

be an important area of discussion in the CEA Power System Reliability Subsection

Electric Power System Reliability Assessment

Bulk Electricity Systems (a) Significant Power Interruptions (b) Delivery Point Interruptions

Distribution Systems

(a) Service Continuity Statistics

Figure 5

Basic structure of EPSRA

260

Service Continuity Statistics

Service continuity statistics have been collected by many Canadian utilities for over twenty years The 1989 report covers 8.6 million electric system customers

Table 5 shows the basic service continuity statistics on a national basis

System Average Interruption Frequency Index = 3.61 int/syst-cust

System Average Interruption Duration Index = 4.32 hrs/syst-cust

Customer Average Interruption Duration Index = 1.20 hrs/cust

Index of Reliability = 0.999507

TABLE 5 - Basic Canadian service

continuity statistics for 1989

These indices are HLIII parameters and provide a valuable indication of customer service The report also provides a breakdown of the primary causes of interruption

CONCLUSION

This paper has presented an overview of the CEA reporting procedures for assessing overall power system performance and providing generation, transmission and distribution outage statistics which can be used to predict future performance

The Equipment Reliability Information System (ERIS) is presented with particular emphasis

on the ability of the system to provide the necessary reliability parameters to support the conventional models used in reliability evaluation The Electric Power System Reliability Assessment (EPSRA) procedure is briefly described This procedure, when fully implemented, should provide an extremely important source of comparative utility performance data

REFERENCES

1 1990, Canadian Electrical Association Equipment Reliability Information System, 1989 Annual Report on Generation Equipment Status, Canadian Electrical Association, Montreal

2 1990, Canadian Electrical Association Equipment Reliability Information System, Forced Outage Performance of Transmission Equipment 1984-88, Canadian Electrical Association, Montreal

3 Billinton, R., 1972, “Bibliography On The Application of Probability Methods

in Power System Reliability Evaluation”, IEEE Transactions on Power Apparatus and Systems, Vol PAS-91, pp 649-660

10

11

12

Transactions PWRS-l,

1978, IEEE Subcommittee on the Application of Probability Methods, Power System Engineering Committee,

“Bibliography On The Application of Probability Methods in Power System Reliability Evaluation 1971-1977", IEEE Transactions on Power Apparatus and Systems, Vol PAS-97, pp 2235-2242

Allan, R N., Billinton, R and Lee, S H., February 1984, “Bibliography

on the Application of Probability Methods in Power System Reliability Evaluation 1977-1982", IEEE Transactions

on Power Apparatus and Systems, Vol PAS~103, No 2, pp 275-282 Allan, R N., Billinton, R., and Shahidapour, S M., 1988,

“Bibliography on the Application of Probability Methods in Power System Reliability Evaluation 1982-1987", IEEE Winter Power Meeting, New York

Billinton, R., Kirby, J J and Asgar-Deen, M L., 1987, “Generating Capacity Adequacy Criteria Used by Canadian Utilities for Planning Purposes”, CEA Transactions, Volume 26

Singh, C

1985, The Consultative Committee on Outage Statistics, Canadian Electrical Association, Instruction Manual for Reporting Generation Equipment Outage Data, Canadian Electrical Association, Montreal

Billinton, R and Allan, R N., 1984, Reliability Evaluation of Power Systems, Longman, London (England)/Plenum Publishers, New York

Billinton, R., Wee, C and Debnath, K.,

1987, “A Procedure and Digital Computer Program for Derated State Modeling of Generating Units Using the C.E.A Data

Base”, CEA Transactions, Volume 26 Billinton, R and Chu, K., November

1986, “Transient Stability Evaluation

in an Undergraduate Curriculum - A Probabilistic Approach”, IEEE

No 4, pp 171-178 Billinton, R., Debnath, K., Oprisan, M and Clark, I M., April 1990, “The Canadian National Data Base for Reliability Evaluation and Assessment”, Proceedings -.1lth Advances in

Reliability Technology Symposium, Liverpool, UK