ML20092B544

Component Failure Data Handbook, Technical Evaluation Rept

ML20092B544
Person / Time
Site:	05000605
Issue date:	04/30/1991
From:	Gentillon C EG&G IDAHO, INC., IDAHO NATIONAL ENGINEERING & ENVIRONMENTAL LABORATORY
To:	NRC
Shared Package
ML20087F663	List: ML20092B544
References
CON-FIN-D-6152 EGG-EAST-8563, NUDOCS 9202110028
Download: ML20092B544 (165)

Text

__ _ _ - _ _ _ _ _ -

i m: a '

EGG-EAST-8563

- April 1991 l

7ECHNICAL EVALUATION REPORT

" Idaho -

National: COMPONENT FAILURE DATA HANDBOOK

~ Engineering Laboratory O

.b'h[ -

Department C.D. Gentillon of5nergy o

o s

=

i

=, --

?

s .

Y ln. Gk EGc '""'

Prepared for the

+ U.S. NUCLEAR REGULATORY COMMISSION Won perfonnen) under b DOE Contract No. DE-ACG7-MIDf'9570.

~

hO D A

_ . _ _ _ _ _ = _ _ _ _ - _ _ _ _ _ _

This document contains new concepts or the author (s) interpretation of new calculations and/or measurements; accordingly, EG&G Idaho, ' .is required by the United States Government to include the following disclaimer:

DISCLAIMER This report was prepared as an account of work soonsored by an agenty of the United States Government. Neither the Urwted States Govemment nor any agency thereof, nor any of their empioyees, manes any warranty, express or imphed, or assumes any legal liabhty or respons;bihty for the accuracy, comp 6eteness, or usefulness of any informaton, apparatus, pm or process disc!csed, or represents that its use would not intnnge pnvately owned ngnts. References herein to any specific commercial product, process, or Service by trade name, trademark, manufacturer. or othefwise, does not necessanly constitute or impty rts endorsement, recommendation, or favonng by the United States Govemment or any agency thereof. The views and opinions of authors expressed herein do not necessanly state or reflect those of the United States Govemment or any agency thereof.

EGG-EAST-8563 t

COMPONENT FAILURE DATA HANDBOOK C. D. GENTILLON 1

. APRIL 1991.

+

t IDAHO NATIONAL ENGINEERING LABORATORY EG&G IDAHO,'INC.

IDAHO FALLS, IDAHO 83415 PREPARED FOR THE U.S. NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555

.UNDER DOE CONTRACT NO. DE-AC07-76ID01570

, FIN NO.-D6152 l

ABSTRACT This report presents generic component failure rates that are used in reliability and risk studies of commercial- nuclear power plants. The rates are computed using plant-specific data from published probabilistic risk assessments supplemented by selected other sources. Each data source is described. For rates with four or more separate estimates among the sources, plots show the data that are combined. The method for combining data from different sources is presented. The resulting aggregated rates are listed with upper bounds that reflect the variability observed in each rate across the nuclear power plant industry. Thus, the rates are generic. Both per hour and per demand rates are included. They may be used for screening in risk assessments or for forming distributions to be updated with plant-specific data.

FIN N0. 06152 SPECIAL PROJECTS ii

I

SUMMARY

This handbook presents failure rates for components used in reliability and risk studies. It was developed by EG&G Idaho, Inc. (EG&G) at the Idaho National Engineering Laboratory (INEL) as a part of the Integrated Risk As-sessment Data Acquisition Program (IRADAP). The IRADAP, sponsored by the U.S. Nuclear Regulatory Commission's (USNRC's) Office for Analysis and Evalua-tion of Operational Data (AE0D), was to meet USNRC operational data needs through two primary activities. The first was to disseminate existing data;

{

the second was to develop taxonomies, statistical aggregation techniques, and software that permit the c'ata to be fully useful and traceable. The generic

component failure data set described herein is based primarily on published plant-specific sources of data, gathered in 1987. For each component and failure mode, available data were combined to form a single rate and bound for use in risk assessments, thus fulfilling the IRADAP objective.

During the sarne period in 1987 that the IRADAP was seeking component failure data, the Nuclear Computerized Library for Assessing Reactor Risk (NUCLARR) program at the INEL was expanding. Component failure data were to be added to the NUCLARR data base to supplement the existing human error probability data. The NUCLARR program, sponsored by the USNRC's Office of Research, provided the computer data base environment for storing the data acquired through IRADAP.

Selected component failure data have been entered into this data base.

The primary sources are plant-specific data from probabilistic risk assess-ments (PRAs). Failure rates and probabilities of failure on demand for me-chanical and electrical components and instrumentation are included, with failure modes that are typically used in PRAs. Care has been taken not to duplicate data that may be contained in more than one source. Because of l

l iii

such stipulations for the data reflected in this handbook, not all of the component failure data contained in NUCLARR are used.

To provide generic component failure data for use in screening probabi-listic risk assessments and in other risk analyses of U. S. commercial nuclear power plants, the capability of combining relevant data from many sources is required. The resulting aggregates are point estimates of either failure rates or probabilities of failure on demand for specified components and failure modes. Upper tolerance bounds are also calculated. The bounds are assessments of the variability in the occurrence rates or probabilities across the population of sites and specific components described in the data sources. The aggregation methods are implemented in NUCLARR.

Software was developed for IRADAP to display and summarize the component failure data rates. The NUCLARR data files and aggregation and plotting algo-rithms were accessed to produce tables showing the mean, median, upper toler-l ance bound, and error factor for each component and-failure mode. The tables also show how many data points contributed to each aggregation and the number of sources represented. Plots that show the data going into each aggregation containing at least four records from NUCLARR were also generated.

Descriptions including the key features of each source wert written to assist handbook users in judging the applicability of each rate. They ad-dress issues such as the purpose of the data collection effort and the consi-derations that went into processing the data for NUCLARR. Knowledge concern-ing the types of facilities that provided the data together with the varia-tions in the data shown in the plots will help the user to better understand the data. The recommended method for using data from a particular source ra-ther than an aggregate from several sources is to obtain access to NUCLARR.

In some cases, a user may also identify such values from the plots.

l l

l iv

t FOREWORD 1

The purpose of this foreword is to provide historical background informa-tion on the relationship between the Integrated Risk Assessment Data Acquisi-tion Program (IRADAP) at the Idaho National Engineering Laboratory (INEL) sponsored by USNRC's Office for Analysis and Evaluation of Operational Data (AE00) and the Nuclear Computerized Library for Assessing Reactor Risk (NUCLARR) program at the INEL sponsored by USNRC's Office of Research.

l Gathering generic component failure data in one place and developing me-thods to combine them were needs identified in the IRADAP's 1986 ranking of f data needs to suoport USNRC requirements for operational data. Work on this i task was authorized in early 1987 under the auspices of the IRADAP.

I During the same period in 1897 that IRADAP personnel were considering how to process component failure dcta, Dr. Thomas G. Ryan, the USNRC Techni-cal Monitor for the NUCLARR program, wanted the NUCLARR program at the INEL to expand to include such data. The benefits of the two programs working together were clear: IRADAP could supply component failure data and, from a review of the use of component tailure data in several probabilistic risk assessments (PRAs), could supply component failure data base taxnomy sugges-tions. Meanwhile, NUCLARR could provide a sof tware environment for storing the data and using them at comput - work stations. Ms. Bennett M. Brady, the USNRC Technical Monitor for the IRADAP, concurred.

In conjunction with IRADAP, a data base structure and taxonomy were de-vised for NUCLARR to contain failure rates with a variety of searchable infor-mation. In addition to the component type and failure mode, attributes such as component design, normal state, system, plant, failure severity, and data collection dates were included. For certain components, codes describe appli-cations or functions, such as the parameter being monitored fer instrumenta-tion, the internal environment, or the voltage level for an application. ,

Data origin information, including the type of records used to extract the failure data, and a comment field are also included.

v i

s

e Many of these attributes are not necessary for most base level relia--

bility analyses. For this generic component failure data handbook, only l attributes such as the component type, design, normal state, failure mode, l nature of the original data sources, and source documentation references are used. The additional attributes aid in fine tuning the failure rates for specific applications. This capability was envisioned for the IRADAP data

- collection effort but not for the handbook it; elf. Access to NUCLARR allows full use of these capabilities.

The calculation of aggregated rates from a set of component failure data records is another area in which the NUCLARR program support has been benefi--

cial for the IRADAP component failure data analyses. The algorithms were derived in 1987 primarily under NUCLARR funding and were implemented on a PC entirely under NUCLARR funding. The rates presented in this handbook would not have been possible without this support. -l 1

The NUCLARR documentation (NUREG/CR-4639, Volume 5, Part 3) includes a data sheht for each component failure data record in the NUCLARR data base; in addition, Volume-5 Part 4 contains aggregations of the data by component and failure mode. All records described in Volume 5 having a data entry date prior to May 25, 1989 and noted as being IRADAP Suitable were used in the aggregations for this handbook. The differences between the aggregations for this handbook and the NUCLARR aggregations are as follows:

1. Only data viewed as being "lRADAP Suitable" are included in this handbook. This results in an exclusion of

. Data from the Licensee Event Report (LER) Summaries, since con-sistent reporting of LERs among units is not expected and the denominators for the LER rates are rough estimates (particularly for per demand data) i

. Data from WASH-1400 vi i

b

. Data points that have been excluded for various reasons discussed in Section 3 of the main text.

y (Note that these data have been entered into NUCLARR, using NUCLARR funding, to provide coverage for a wide variety of data within NUCIARR. The philosophy for data entry for NUCLARR is that suffi-cient information be present to allow a failure rate to be classi-fied by component and failure mode. As a repository or library of-source data, NUCLARR does not have a list of restrictions for data entry such as those discussed for IRADAP. The philosophy for NUCLARR data is that the suitability for a selected data point in a ,

given application ultimately rests with the end user of the system.

While this responsibility exists for the user of data frcm any l source, the IRADAP data have additional safeguards to ensure their quality and applicability for risk assessment and thus are a subset of the NUCLARR component failure data.)

l l 2. The har.dbook displays provide additional information to assist the i user in judging the applicability of the data. The summary tables l

provide an indication of the sources of the data. Plots labelled by source allow the reader to see which studies gave rise to the high-est and lowest failure rate estimates. The data source descriptions .

in Appendix A provide information to help the user judge the useful-ness of the rates.

In summary, the IRADAP component failure data aggregations are more selective than those reported in the NUCLARR series, t

The NUCLARR program has an ongoing mission to collect component failure '

data as well as human error probabilitier. As more data become available, updates to this report might be feasible. However, the IRADAP was discon-tinued in 1988. A Special Projects Program under AEOD funding was instituted e in 1988; completion of a draft of this handbook in 1898 was one of its ini-tial chartered tasks. The current handbook, issued in 1991, incorporates in the text minor changes from an NRC peer review of the 1989 handbook.

vii

... . ..~ - -- ~.- . - . - . .- - . . . - - . . - - - . - . . . .. - . - -

The software developed under Special Projects funding to display the tables and plots of this report using the NUCLARR data base could be used at some future time.in the NUCLARR program to provide updates to this report as more plant-specific component failure data becomo available. NUCLARR personnel would in such a case apply the principles discussed herein in assessing the "IRADAP Suitable" indicator for data-from additional sources.

Documentation similar to the information in Appendix A would also be needed for each new source with suitable data.

I 4

i 3

I i

l viii

ACKNOWLEDGMENTS The afforts of the Integrated Risk Assessment Data Acquisition Program (IRADAP) staff at the Idaho National Engireering Laboratory (INEL) are appre-ciated. Special thanks are extended to Michael R. Groh who was instrumental in developing the component failure data taxonomy on which the handbook is based. Oren V. Hester's reviews of probabilistic risk assessment to provide a firm foundation for the taxonomy is appreciated. Helpful review comments on the taxonomy by William J. Galyean are also appreciated. Quinn R. Decker is thanked for his collection of data sources and development of a system for tracking them. Quinn R. Decker, Frank G. Farmer, and Bruce S, Anderson are f thanked for selecting and encoding data. Appreciation is extended to Osmond J. Call, who developed the software that produces tables and plots for this handbook.

In addition, the assistance of the Nuclear Computerized Library for Assessing Reactor Reliability (NUCLARR) staff is gratefully acknowledged; this includes computer software support from a team originally headed by Gretchen H. Beers and more recently by Osmond J. Call, data encoding and processing by B. Gay Gilbert and Wendy J. Reece, and overall program coordi-nation by Dr. David I. Gertman.

Sincere appreciation is expressed to Jm ?ph R. Fragola and the staff of his New York and Clearwater, Florida offices of Science Applications Inter-national Corporation (SAIC) for plant-specific component failure data from several nuclear power plant units and for advice in selecting and displaying the failure information.

l l Finally, the author would like to iaank Bennett M. Brady and Thomas G.

Ryan, both of the USNRC staff. Ms. Brady has providec support since 1986 as AE00 Program Manager for the IRADAP and for the Special Projects on behalf of I

the USNRC's Office for Analysis and Evaluation of Operational Data. Dr.

l Ryan's support as the Technical Monitor for the NUCLARR program, particularly l

in the development of aggregation algorithms, is gratefully acknowledged.

ix f

L i

CONTENTS Page I

ABSTRACT ............................................................. 11

SUMMARY

.............................................................. iii FOREWORD ............................................................. v ACKNOWLEDGMENTS ...................................................... ix NOMENCLATURE ......................................................... xii

1. INTRODUCTION ..................................................... I 1.1 Purpose ..................................................... 2 1.2 Background .................................................. 2

-2. SCOPE OF COMPONENT FAILURE DATA ................... .............. 5 2.1 Components and their Boundaries ............................. 5 l 6 2.2 Failure Modes ...............................................

2.3 Individual failure Data Records ............................. 10 l-

3. SELECTION OF SOURCES ............................................. 13

4. DATA ENTRY CONSIDERATIONS ........................................ 17 4.1 Data Base Quality ........................................... 17 4.2 Use of Judgement ............................................ 19

-4.3 Interface with NUCLARR ...................................... 22

5. DATA COMPUTATIONS ................................................ 23 5.1 Treatment of Single Data Points ............................. 23 5.2 Treatment of Several Data Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

-5.3 Adjustments to the NUCLARR-Calculated Means and Bounds ...... 26

6. GENERIC COMPONENT FAILURE DATA ................................... 27

7. QUALITY OF DATA IN THE HAND 300K .................................. 47

8. REFERENCES ....................................................... 49 APPENDIX A--COMPONENT FAILURE DATA S0VRCES............................ A-1 APPENDIX B--COMPONENT FAILURE DATA PLOTS..... .. ..................... B-1 APPENDIX C--AGGREGATION METHODS ..... ............. ....... .......... C-1 x

4 TABLES

1. Data Sources in NUCLARR and tFrtr Use in lhis Handbook ............ 14

2. Reasons for Excluding Data ........................................ 20

3. Generic Fail ure Rates for Mechanical Components . . . . . . . . . . . . . . . . . . . 30

4. Generic Failure Rates for Electrical / Instrumentation Components ... 33 j

5. Generic Failure rates for Mechanical Components with Various  ;

Designs and Normal States .................. ...................... 35

6. Generic Failure Rates for Electrical / Instrumentation Components with Various Designs and Normal States . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 1

a-1

.I l

l Xi l

l e

NOMENCLATURE DEFINITIONS Aaorecation: An aggregate is "any total or whole considered with refer-ence to its constituent parts" (The American Heritage Dictionary, Second y

College Edition). Most of the failure rates and per demand probabilities described in this handbook are aggregates, derived from more than one datum for a specified component and failure mode. Aggregation in the cur-rent context is the process of combining individual records from one or more data sources to form these estimates and their bounds.

Comoonent Boundary: The component boundary includes the basic component (i.e., valve) with its operator-(i.e., motor valve operator) and asso-ciated circuit breaker and control circuits. For a pump the boundary would include: the pump, its control circuitry, its driver (a motor or turbine), and the motor's circuit breaker and controller or the turbine's i stop and throttle valves and associated control mechanism.

Failure: A failure is a particular type of fault that represents an irre-i versible state of a component such that it must be repaired in order for it to perform its design function. Failures are further classified as primary or secondary failures:

a. A orimary failure is a failure that results from wearout or apparently random problems within a component. No clear external cause can be found for such failures,

b. A gcondary failure results when the component is subject to conditions that exceed its design envelope (for example, exces-sive voltage, pressure, shock, vibration, or temperature).

Failure Mode: The failure mode is the specification of the aspect of the failure process that is of concern and includes a description of how the failed component / system state differs from the desired state ("...the effect by which a failure is observed to occur..." --IEEE Std 500-1984, Appendix A).

Hgmogeneous: This term applies to a population of components whose fail-ure rate per unit time or failure probability per demand for a particular failure mode can reasonably be assumed to be cc:.stant. This implies that, for per hour rates, there is no reason to expect the number of failures not to be Poisson. For per demand probabilities, this term applies when-ever there is no evidence to show that the number of failures in a fixed number of demands is not described by the binomial distribution.

Recair: Repair is the restoration of a component to an operable state following the failure of that component. To repair a component requires that the component be removed from service (i.e., made unavailable).

Repair may be as_ simple as adjusting the setpoint or as drastic as the total replacement of the component.

xii

1 ACRONYMS / ABBREVIATIONS AE00 Office for Analysis and Evaluation of Operational Data (USNRC)

BWR Boiling water reactor CFD Component failure data CRPS Component Reliability Parameter System '

EB Empirical Bayes EG&G EGLG Idaho, Inc.

FID Facility identifier IBM International Business Machines Corporation IEEE Institute of Electrical and Electronics Engineers, Inc.

INEL Idaho National Engineering Laboratory IRADAP Integrated Risk Assessment-Data Acquisition Program  ;

LER Licensee Event Report LWR Light water reactor NUCLARR Nuclear Computerized Library for Assessing Reactor Reliability PC Personal computer l

PWR Pressurized water reactor f SAIC Science Applications International Corporation USNRC United States Nuclear Regulatory Commission f

xiii

)

COMPONENT FAILURE DATA HANDBOOK

)

I l

1, INTRODUCTION i

This handbook contains failure rates and probabilities for failure on demand for components that are typically used in United States commercial f nuclear power plants. These data are provided from sources identified in 1987 in the USNRC's Integrated Risk Assessment Data Acquisition Program (IRADAP) administered by the Office fo" Analysis and Evaluation of Opera-

) tional Data (AE00).

(

l This handbook provides the following information:

1. A discussion of the scope of the component failure data, including l the components and failure modes considered

2. Information on the methods used for selecting data sources

3. A discussion of data entry considerations

4. Notes on the calculations for medians, means and upper bounds

5. Tables listing the component failure data by component and failure mode and then more specifically by component design and normal state as well as failure mode

6. A brief discussion of the quality of the data

7. An appendix describing each data source used in the handbook, inclu-ding the purpose of the data collection effort and any assumptions used in screening the data for the handbook

(

l i

l l

(

l l

8. An appendix providing plots of the individual records that were ag-g egated to form the rates listed in the tables (for all casas in which at least four records were combined to form a rate) and

9. An appendix describing the aggregation methods.

With this information in a handboot, selected recommended component failure data and supporting information are available in a portable, convenient for-mat.

In the subsections below, the purpose of the handbook is restated and background information for this project is provided.

1.1. PURPOSE The purpose of this handbook is to present in a single place generic component failure data for various tasks.

1.2. BACKGROUND Several types of operational data are needed for risk assessments of commercial nuclear power plants. These include basic generic component failure data, data on the frequency and duration of both corrective and preventive maintenance outages during plant operation, common cause data, data on internal and external initiating events, and probabilities for successful recovery actions by plant operators. In addition, data sets on root causes and the effect of aging are needed to assess lifetime extension issues and prevent the recurrence of problems that impact reactor safety.

In the 1986 IRADAP ranking of the USNRC's data needs,1 generic failure rate data were ranked second in importance in satisfying USNRC pro-grant,atic operational data requirements in support of reliability and risk analysis. Only common cause data had a slightly higher importance ranking. l l

2 l

1

(

l l

In comparison to commen cause data, the generic failure rate data set was  !

chosen to be built first-since it could be developed efficiently and in a cost-effective manner by combining existing data sets.

In 1987, the Nuclear Computerized Library for Assessing Reactor Relia-( bility (NUCLARR) program added a module to store, process and display 2

component failure data (CFD) in addition to human error probabilities in its PC-based computer work station. Sponsored by the USNRC Office of Nuclear Regulatory Research, this software tool makes the raw data available in a form that facilitates sensitivity studies and evaluations of subsets. From inc'eption, both the development of a component failure data handbook through IRADAP and the development of a hardware component failure data module for NUCLARR have been complementary efforts, each program benefitting fror the other. The IRADAP program provided data and, from a review of the use of com-ponent failure data in several probabilistic risk assessments (PRA-1, compo-nent failure data base taxnomy suggestions for NUCLARR. Conversely, NUCLARR l has provided IRADAP a software environment in which the data can be stored and aggregated.

Further documentation for some of the topics discussed in this nandbook can be f~und in the NUCLARR series of NUREGs (NUREG/CR-4639; Reference 2).

In particular, the scope of the component failure data and the nature of the aggregation methods are discussed in. Volume 4: User's Guide. Part 3:

NUCLARR System Descriotion. Data entry considerations are also included.

However, they are discussed from a different perspective in the NUCLARR re-ports because NUCLARR is intended to be a library or repository for data.

Data entry considerations for NUCLARR focus on whether the operational data provide enough information to permit their classification in the NUCLARR sys-tem. The "IRADAP Suitable" field in NUCLARR allows the data within NUCLARR

[ -.

that are flagged for the component failure data handbook to be distinguished from.the remaining data.

Additional documents from the NUCLARR series would be of use to those who would like to study the component failure handbook data in more depth.

3

F

s. Hardware _(omoonent Failure Data (HCFD) of Volume 5 contains a page for every component failure data entry in NUCLARR, including those used for this handbook. This document is updated periodically as new data are entered into NUCLARR.

The best method for further study of these data is to obtain access to I the NUCLARR system itself. The system allows one to identify subsets of com-ponent failure data and form aggregations on-line, so that the effect of se-lecting particular data sources or individual records to form a failure rate estimate can be immediately seen. On-line plots of such data also can be ob-tained. Names and phone numbers for obtaining further information regarding the NUCLARR System are as follows:

Thomas G. Ryan U. S. Nuclear Regulatory Commission--RES Reliability and Human Factors Branch 5650 Nicholson Lane, NL/N-316 Rockville, MD 20852 USA Phone: 301-492-3550 (FTS 492-3550)

David I Gertman NUCLARR Data Clearinghouse Idaho National Engineering Laboratory P. O. Box 1625 idaho Falls, ID 83415-2405 USA '

Phone: 208-526-0652 (FTS 583-0652)

{

4

2. SCOPE OF COMPONENT FAILURE DATA l

The types of data regarded as basic " generic component failure data" are described below.

l 2.1. COMPONENTS AND THEIR BOUNDAP.IES The components selected for the component failure data handbook (CFDH) are those components that are typically involved in basic events in fault tree models of nuclear power plant systems. Thus, they are for the most part components of the type found in safety systems. For convenience, the compo-nent list has been divided into two sections, mechanical and electrical /-

instrumentation. Forty-two general categories of mechanical components and 33 categories of electrical components have been identified. These component types have been further characterized by their design. The components and designs are listed in the con.ponent failura data section (Section 5) for cases in which data exist. A complete list, including NUCLARR data base codes, can be found in Volume 4 of the NUCLARR NUREG series (Reference 2).

The component list is ir. tended to be fairly complete. The availability of failure data for every component / design combination in the list is not expected, at least until more operating experience is acquired and improved data collection schemes are available. The existence of the oterall list is beneficial in determining where additional data collection efforts for NUCLARR might be warranted.

To maintain consistency between different data sources contained within the generic failure rate data, some general discussion on component boun-daries for apportioning the failures is appropriate. If possible, each component's boundary should include the physical Soundary of the component itself plus that of any associated, dedicated auxi'liary equipment. This 5

(

boundary would include closely associated equipment that is physically cou-pled, such as generators or pumps with their drivers (diesel engines, motors, etc.) and valves or dampers with their operators (motor, pneumatic, etc.).

Also, this boundary would include dedicated control circuits and circuit breakars associated with pump motors, valvt aperators and diesel generators.

For diesel generators it would additionally include dedicated air starting systems and Jeoicated batteries. Similarly, lube oil cooling failures would be included in the data set for those components that provide their own lube oil cooling function.

Note that the rates in this handbook are combinations or aggregates of rates from various sources, not all of which conform to the component boun-dary definition just stated. For PRA purposes, most of them do. In any case, component boundary guidelines are provided here for any future data collection efforts that might contribute data for updates to NUCLARR.

L 2 . P. . FAILURE MODES O

This handbook contains generic failure rates and probabilities of fail-ure on demand that are used in PRAs. Thus, these rates and probabilities apply to the types of events that typically are found in the basic events of fault trees. These events are unplanned and render the components

~

unavaliable to perform their function. Thus, " catastrophic" failures are of interest rather tnan degradeu or incipient conditions that may cause some reduction of function. If plant personnel have some flexibility in deciding when to repair the components and do not have to declare the component inoperable, the basic function of the component must be available.

More specifically, the following failure modes are the primary failure I modes considered for the generic component failure data:

Fails to operate Fails to start Fails to run Fails to open Fails to close 6

l l

(

L j Fails to energize j fails to de-energize Fails to transfer electrically Spurious operation Spurious start Spurious open (transfer open)

Spurious close (transfer closed)

Spurious energize Spurious de-energize Spurious transfer electrically Leakage External leakage / rupture Internal leakage Blockage Plugged The " fails to operate" mode applies to nearly all components. This mode includes cases where data sources say " operational failure," " improper opera-tion," "no output," or, just " fails." In all of these cases, the mode ap-plies to failures per hour. However, per demand probabilities can also be associated with this mode, as in " fails to operate on demand." Specific examples of componerts whose catastrophic failure mode will for consistency always be denoted " fails to operate" are:

static inverters batteries electrical busses battery chargers transformers

" Fails to start", " fails to run", and the other modes listed above under

" fails to operate" are more sp'cific descriptions o) le failure events. For

• standby components, these mode, may have both per hour and per demand rates, depending on whether the mechanisms et ciated with the failures are asso-ciated with the component's use or ar~ > 4,q mechanisms that occur even when the component is not being used. The modes also typically apply to particu-lar component groups; for example, " fails to start" and " fails to run" apply to pumps and generators while " fails to open" and " fails to close" apply to valves and breakers and fails to energize" and " fails to deenergize" apply 7

l

to relays and other electrical components. The " fails to transfer electri-cally" mode applies to auto bus transfer (ABT) devices. For relief valves, the per demand rate, " fails to close", describes failure to reseat.

For components such as fans and compressots, " failure to run" and " fail-ure to operate" may both be used to desc'ibe failure events in the fail to operate (per hour) group. For cone"tency, the fail to run mode will be used wnenever the component also har, a ' to start" failure mode; otherwise, the " fail to operate" mode will be used.

The "spuriouc operation" failure mode group is similar in that it, too, contains modes that '.ypically apply to particular components. The spurious categor includes failures that are described in the data sources as "prema-ture open," " function without signal," "normally open fails closed," and

" fails high." Unlike the " fails to operate" category, all spurious operation failures give rise to per hour rates only.

Failure modes associated with leakage apply to components that form a pressure boundary; primarily valves. External leakage is typically leakage through valve packing; internal leakage k leakage across the valve seat.

Hourly rates for failure to seat or rescat are described by the leakage fail-ure mode. The " internal leakage" mode for check valves is back leakage. All leakage rater per hour.

Blockage, like leakage, applies to certain hourly rates for components that fcrm a pressure boundary. The " plugged" failure mode applies when the blockage is known to be complete.

A feature of the failure mode scheme described above is that it is hierarchical. That is, more detailed modes are listed under general modes so that differer' ..ypes of data can be accot.odated. Thus, for example, sources that provide C for the " fails to open/ fails to close" failure mode can be used as well as those that procide separate characterizations for a compo-nent's failure to open and failure to close.

8 l

b _ _ _ _ - _ _ _ _ _ - - - -_ ---___ J

n i

1.:ree other general comments apply to the f ailure nodes and the events they describe. The first is that the failure counts for the numerators of

) the failure rates or demand probabilities for operating components are counts of failures that occurred during operation. Failures that have no effect on the components during operation (for example, failures detecteo and corrected during preoperational testing) generally do not count. The same statement applies to the denominators. ,

lhe second comment is related to the first: the mode of operation of a a component is important. The normal state of a component can have a major in-fluence on its failure rates. For example, normally operating pumps general-ly have a lower probability for fa'. lure to start than normally stand' omps have. Where possible, the normal state for components is noted in col ecting g data for this handbook. The following normai st..Les are considered in ,

NUCLARR:

x-Cad.g Definition Shorter form used in data tables -

NO Normally open Open ,

NC Normally close Closed NE Normally energized Energized ND Normally de-energized De-energized NR Normally running Running or oper.

NS Normally in standby Standby '

NA Normally alternating Alt.

NL Normally locked-out Locked-out  %-

XX Unknown (unspecified) Unsp.

The normal state codes have ~aecial interpretations when applied to passive components such as pipes and heat exchangers. For these components, running or operating indicates wet flowing; standby indicates a wet stagnant condi-tion, not flowing; locked out indicates a dry condition; and alterna+ing is between running and standing by. In cases where the normal state is not stated, it is of ten implied by the failure mode; for example, the " fails to clo;e" mode is typically associated wi+h a normally open valve. Separate aggregations are performed for differing normal states.

The final general comment is that secondary f allures should be exclu-ded. That is, failures that are caused by common support systems or improper 9

inputs are not included in the failure rates if possible because these are modeled separately in most risk assessments. For example, consider a front ,

line system pump failure caused by a loss of the pump's lobe oit cooling due to a failure in the component cooling water system. Ideally, no front line system pump failure would be counted even though the front line system pump experienced a f:ilure that required repair. Such failures should not be in-cluded in the failure data because their occurrence is entirely dependent on the specific designs of the systems involved. They are not needed in the failure data sets because PRA system models will identif" what components will fail due to the loss of each support system. Thus, the ideal failure rate data would be dependent only on the component and its immediate environ-I ment and application and would be independent of the specific plant and sys-tem designs. Sources that include secondary failures in the basic failure i rate data will have higher failure rate estimates; however, data from these sources are still included in the dcta base because in most cases the impact of these failures is minimized by the protective devices commonly installed on components.

2.3. INDIVIDUAL FAILURE DATA ReconDs Component failure data records are designed to supply information for failure rates (failures / hour) or probabilities of failure on demand (fail-ures/ demand). The level of detail for which the operational experience al-lows for distinguishing separate rates within a component type is another aspect of the scope. Far example, failures could be counted separately for each pump in a plant, resulting in a failure record unique to each pump, or the operating experience could be combined to form an overall pump failure rate for all pumps. As another example, separate rates for each year of operational experience could be reported. The desired scope of the data in the generic failure rate data se to a certain extent depende on the infor-mation available. The two possibilities are discussed below.

Ideally, this gereric failure rate data set will contain data at a level of detail such that the failure rate can reasonably be assumed to be 10 l

l

l constant. Thus, data for components having the same design, function, appli-cation, and operating environment may be combined. If there is any reason to expect differences in the rates (e.g., perhaps a particular component was re-cently replaced), the preferred approach is to separate the data into sepa-rate records. On the other hand, since the focus is primarily on generic failure rate data, individual component identifications are not a part of the data stored. In summary, the optimal data record supporting the component failure data handbook is a statement of the number of failures of a particu-lar type and corresponding operating history (times or demands) for a set of homogeneous components.

if this level of de ;l is not available, the failure data will describe a collection or populatio of components whose failure rates vary. For such data, an important featurt ss to supply tolerance intervals that describe this population variability. If no tolerance intervals are available, the data will still be accepted and the variability will be characterized through the combining (aggregating) of data from several sourccs.

f k

11 I

l

/

4 12

A-

- - w g(:V w . _ _ - _ . _ . _ _ _ _ _ _ _ .

E

g. 3. SELECTION OF SOURCES w:

Generic failure rates that are based solidly on operational experience were obtained for this manual by combining or aggregating plant-specific data from several plants across the it.dustry. A bibliography of data sources for use in risk assessment3 completed in the FYd987 IRADAP activities served as a starting point for gathering these data. This document describes 38 studies, five nonnuclear sources, six foreign data bases, and tht, data i efforts documented in 27 U.S. commercial nuclear power plant probabilistic risk assessments (PRAs). Among these sources, available recent PRAs with component failure data are the prime generic failure data source. The PRA

'ata are plact-specific, and are for the most part collected at a level of a detail to support variations in equipment design and use since they were collected specifically for risk assessment purposes.

Other sources are regarded as supplements to this basic set of sources.

For certain cononents, little plant-specif',c data from nuclear power plants exist. Generic sources including some nonnuclear sources are used in this case. Care is taken to avoid the use of sources that duplicate each other.

The principle is to combine data from the best possible sources for each specified component and failure mode.

Table 1 describes the data sources currently contained in the hardware component failure data iaodule of NUCLARR, the program that identified the source, the prograc that encoded data for the source, and whether data from the source tre used in this handbook. Each data source used in this handbook is describeG in Appendix A. Complete bit,liographic information for these sources it orovided there.

Processing of sources that become available in the future is expected under the NUCLARR program. Ongoing risk assessments may generate data that are solidiy based on maintenance work requests and clant-specific analyses of the r.vnDer of demands; these would be ideal inputs to the NUCLARR Clearing-house as they become available.

13

TAnLE 1. DATA SOURCES IN NUCLARR AND THEIR USE IN THIs HANDBOOK' Program Program Use Identifying Coding in Study ID Data _ Source _

Source Data Handbook" PRA Sources BRPl-PRA Big Rock Point 1 1RADAP IRADAP Yes HNPI-PRA Haddam Neck Plant (Conn. IRADAP IRADAP Yes Yankee)

HNSI-PRA Millstone 1 1RADAP IRADAP Yes IPS2-PRA Indian Point 2 1RADAP IRADAP Yes ZIS-PRA Zion Station IRADAP IRADAP Yes NEE 3-PRA Oconee 3 IRA 9AP 1RADAP Yes X-PRA Plant "X" 1RADAP 1RADAP Yes Y-PRA Plant "Y" 1RADAP IRADAP Yes EUR-PRA European plant IRADAP IRADAP Yes A-PRA Plant "APRA" Note c NUCLARR Noe B-PRA Plant "BPRA" Note c NUCLARR Na C-PRA Plant "CPRA" Note c NUCLARR No D-PRA Plant "DPRA" Note c NUCLARR No E-PRA Plant "EPRA" Note c NUCLARR No F-PRA Plant "FPRA" Note c NUCLARR No G-PRA Plant "GPRA" Note c NUCLARR No Additional Data Sources IEEE-500 IEEE-Standard 500-1984 IRADAP IRADAP Yes IPRDS In-Plant Reliability Data IRADAP 1RADAP Yes System (pumps, valves, electrical components)

SWD-HBK Swedish Reliability Data IRADAP Both Yes Book RAC Reliability Analysis Center IRADAP 1RADAP Yet 4 (Nonelectronic components)

NSAC-108 Reliability of Emergency IRADAP NUCLARR Yes Diesel Generators PBS Pipe Break Frequency IRADAP 1RADAP Yes WASH-1400 Reactor Safety Study NUCLARR NUCLARR No LER-SUMM LER Summaries (pumps valves, NUCLARR NUCLARR No channels, inverters)g A-REL Reliability Study "A" Note c NUCLARR No B-REL Reliability Study "B" Note c NUCLARR No

a. This table addresses data sources for the hardd!re co[bonent failure data module of NUCLARR as of June 8, 1989. It does 4,t de ; ribe the human error probability module. DataentrywasaNUCL/R1fuacti(c>exceptfor data from the first six PRAs listed above. For the Swedish Reliability Data Book, some records were coded by NUCLARR personnel and some were coded by IRADAP personnel .

14 l

f l TAstc 1. DATA S0uRcES IN NUCLARR AND THEIR USc IN THIS HANDBOOK' (continued)

) Program Program Use Identifying Coding in Study 10 Data Source Source _DalA_. llan@ galb .

(continuation of notes)

b. Yes: some data from the source were used for this handbook (indivi--

dual data points may still be excluded). No: no data from the sourt e were used for the handbook. The "lRADAP Suitable" field in NUCLARR re-cords from these sources do not contain "Y". Exclusions are due to consi-derations discussed in Section 2.

c. Data from Plants APRA through CPRA and the A and B reliability studies were provided to NUCLARR in May of 1989, too late to be included in this handbook.

( d. These are Data Sununaries of Licensee Event Reports (LERs) from U. S.

commercial nuclear power plants (NUREG/CR-1205, -1363, -1740, and -3867).

They contain failure rates estimated from LERs as recent as 1982. They are not included in the handbook because LERs ware not complete in plant maintenance records and because plant-specific information on the number of demands or the components was not available.

d I

15 1

l . - _ - - - _ _ - _ . - _ _ _ . _ _. . . _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _

I k

16 ,

1 l

l l

4. DATA ENTRY CONSIDERATIONS This section provides further information about the considerations that influence the selection of specific records for the CFDH. These considera-tions affect both the data identified for NUCLARR by the IRADAP team and the assessment of the "lRADA' Suitable" flag for any records subsequently added to the data base by NUCL .1 personnel. Section 2 discussed the handbook data scope in broad terms; mere detailed statements of selection and encoding principles are provided here.

4.1. DATA BASE QUALITY As stated in Section 3, the primary data source is plant-specific data from probabilistic risk assessments of nuclear power plants. One would ex-f pect most such data to be applicable and appropriate for the component fail-ure data base. Among other sources, some contain commercial nuclear power plant operational data that are directly suitable for PRA applications and others have generic data that may not be applicable. For most of the generic sources, only a selected subset of the data presented has been entered into NUCLARR to fill gaps in plant-specific data. The Component Reliability Para-r.,eter System (CRPS) develv.ed by Science Applications International Corpora- a tion provided guidance for the selection of particular rates from sour-3 ces such as IEEE-500 and RAC. Generally, generic data were only used for com-ponent type / design / normal state /f ailure mode combinations that have less than two nuclear power plant stations represented among the plant-specific raw data or less than three plant-specific data points. Another consideration that influenced the selection of sources and general guidelines for entry of data from sources is the desire to avoid duplication of data sets. For each data source, general criteria for selection of data to enter are presented in Appendix A.

Once a source is selected and preliminary data criteria have been '

established, all data meeting the criteria are ente "er, there still 17

may remain records among the set entered from a source whose validity is du-bious. For example, some PRAs present plant-specific data and then later dis-count selected data points because of some defect, such as the belief that the station records do not contain enough accessible information to assess the number of demands for a particular component and failure mode with suf-ficient accuracy. The "lRADAP Suitable" field in NUCLARR is an indicator that permits such records to be flagged and excluded from the aggregations that form the generic failure rates for this handbook.

In addition to known defects in the data, the IRADAP flag may be set for data points that are believed to be not applicable even though they were in-cluded in the initial assessment of data from a source, for example, a source may provide plant-specific data for a particular component for all U.S. commercial nuclear power plants. The IRADAP flag will be set for data from known atypical facilities, such as Ft. St. Vrain (a gas-cooled reactor),

unless the INEL IRADAP team knows from other sources that the component appli-cation at such facilities is similar to a generic PRA application. Known atypical component boundaries would be another reason for setting the IRADAP flag.

A final reason for setting the IRADAP flag is that the corresponding data value (failure rcte or demand probability) is significantly different from other sources as shown by plots of the data. This test can only be made for component / failure mode groups that have at least four data points with bounds (the bounds are discussed more in the next chapter). When a data value and its bounds stand out in comparison to values and bounds from other sources, one suspects either an atypical application or a data quality con-cern. In some cases, the difference can be explained. For example, if one source reports one or more failures and has much more data (such more service time) than other sources that report no failures, the source with failures will have a distinctly higher failure rate estimate than the others, and yet will still be valid.. When such explanations are lacking, atypical data points are excluded from the aggregations that form data for this handbook.

I 18 N

i, Other fields are present in the data base that could influence the selec-tion of data for aggregations that contribute to the rates presented herein.

For example, the data could be restricted to U.S. domestic data, to data based on primary failures only, to data for which the boundaries include con-trol circuits, to data based solely on nuclear grade components, or to data having only nuclear power plant operating experience as the failure data source. However, these restrictions are not made because the sparsity of the resulting data would reduce their usefulness and quality in characterizing the population variability that exists among failure rates at different sites. Furthermore, many sources do not give clear guidance on some of these topics; this would contribute even more to the data sparsity if records having "not specified" in some of these fields had to be excluded. These fields are present in the data base for possible future use when more data are available, and for current use in sensitivity studies directly accessing NUCLARR.

Table 2 provides a sunmary of the data quality considerations. In each case where the IRADAP flag is set, an explcnation is prcvided in the record's comrrent field.

4.2. USE OF JUDGEMENT An issue in processing the component failure data is the treatment of fields for which the data source information is incomplete. Although the data base has been designed to accomodate data that lack detail, its useful- ...

ness increases if some of the gaps in fields such as component design, sys-tem, nuclear grade, and component application can be filled. Using data from many plants to establish population variability curves is desirable; this is inhibited if the data for a particular component type and failure mode are split into a common design category for the components and a "not specified" category, when in reality most of the latter data are probably also from the same common design category. Examples of specific situations of this type, and their treatment, are discussed below.

19 o

l TABLE 2. REASONS FOR EXCLUDING DATA Reason _

Not Acolicable Poor Ouality Other Reasons Data Source: Commercial nuclear power plant operational datt (PRA data or other sources that support PRAs)

Known atypical Underestimated Plot differences (see text) component number of demands boundary Avoiding duplication of data Listed but not Knewn atypical used in PRA facility Any known defect in the data D3ta Source: Other (aeneric) sources Not catastrophic Any known defect Flot differenses Atypical Avoiding duplication of data Existence of at least 3 plant- -

specific data points from at least -

2 stations Two fields that have largely been encoded using judgement are normal state and safety grade. The normal state can of ten be inferred from a know-ledge of the system and the failure mode. For example, auxiliary feedwater pumps are standby at nearly all commercial nuclear power plants. Another example is that the presence of a " spurious open" failure mode for a set of valves and no " spurious close" mode implies that the associate valves are normally closed.

The safety grade field is included in the data base because the design, inspection procedures, and test and maintenance environment of a component can be influenced by whether it is safety grade. Most components that are covered by technical specifications are safety grade. Thus, separate evalua-tions of components in safety-grade systems and non-safety grade systems may 20

l I

be useful. However, few data sources disting'Jish whether a component is safe-ty grade. Therefore, the IRADAP team has assigned safety grade codes fw se-lected records based on their system. Components in the class IE power sys-tems, the emergency onsite power supply system, systems that provide engineer-ed safety features, and systems that actuate safety functions such as the plant protection system and engineered safety features actuation system are l

flagged as safety grade, while components in balance of plant systems are flagged as not being safety grade. For the remaining records, the safety grade remains unspecified unless it is provided in the source,

~

for data from PRAs, turbine-driven and diesel-driven pumps are nearly always separated out; these pumps are expected to have higher failure rates r than motor-driven pumps because their supporting systems are more compli-cated. Therefore, in encoding PRA data, pumps whose drivers were not speci-fled were assumed to be motor driven unless the pumps were in the following systems that are known to often have other types of drivers for pumps: auxil-iary/ emergency feedwater, high pressure coolant injection, reactor core isola-f tion cooling, fire protection, and main feedwater.

The voltage application of main electrical distribution system breakers has in several cases been inferred by the IRADAP team from information con-tained in the report, Data Summaries of Licensee Event Reports of Relays and Circuit Breakers at U.S. Commercial Nuclear Power Plants. January 1. 1976 to .

December 31. 1983 (draft NUREG/CR-4126, January, 1985, by S. R. Brown).

In these cases, the codes assigned are not always compared with plant sources such as drawings and safety analysis reports; thus, they are not known to be absolutely accurate. However, the assignments are reasonable and they are accurate in most cases. The increased usefulness of the data base '

makes these assessments worthwhile.

I 21

4.3. INTERFACE WITH NUCLARR Two methods have been used for input of data for the component failure data handbook to NUCLARR. For the first six FRAs listed in Table 1, PC-based data files were formed containing the data. A part of the NUCLApR data input software processed these files. Data from all the other sources have been entered using the regular NUCLARR data entry software, which displays a se-ries of forms on a computer screen for entering the information needed for each record. Prior to data entry, coding sheets (printed versions of the blank forms) were filleo out for each record.

For each of the 42 mechanical component types and each of tne 33 electri-cal / instrumentation component types defined wit 51n NUCLARR, the NUCLARR data base can support a maximum of 250 records. The entry of records describing motor-driven pumps for each plant unit from the LER Summaries virtually filled this area of the data base. To enter further plant-specific data from Plant X and the Swedish Reliability Data Book, some of the LER-based records were pooled. The LER-based records are regarded as having lower quality be-cause LER reporting re lected in the Data Summaries depended on plant-speci-fic technical specifications and thus varied among plants; also, plant-speci-fic estimates of how often the pumps were started were not available for those studies.

l 22 1

i

5. DATA COMPUTATIONS 1

Two types of calculations are performed by the NUCLARR system for the component failure data. The first is processing of individual records as they are entered into the system, while the second concerns the combining or aggregating of failure rates from several sources for a particular type of

( failure. A brlef discussion of these computations and adjustments for the component failure data h.?ndbook will help the handbook user better appreciate the significance of the medians, means, and bounds presented in the generic component failure data tables.

5.1. TREATMENT OF SINGLE DATA POINTS On data entry, the NUCLARR system computes a median, upper bound, and lower bound for each data point (if sufficient information is available).

The error factor is defined as the upper bound divided by the median, and the lower bound is the median divided by the error factor. If a record supplies raw data (numbers of failures and corresponding operating hours or demand counts) and meets the criteria established in the NUCLARR system fnr homoge-neity, upper tolerance bounds are computed using the raw data. The calcula-tions are described in Volume 4 Part 3 of the NUCLARR series [ Reference 2d]. .

The results of these computations are approximately as follows:

Number Assessed Upper Bound of Divided by Median Failures (Error _ factor) 0 8.4 1 3.3 2 2.6 3 2.2 4 2.0 7 1.7 15 1.5 20 1.4 50 1.3 (This table contains ratios of 95th percentiles to 50th percentiles for 23 l

chi-square distributions with two times the number of failures plus one as the degrees of freedom). Since approximately 90% of the data fall into this raw data category, bounds from most of the single-record failure rates re-flect these conditions. Individual bounds show in the plots in Appendix B; most of the cases for which the median to upper bound span is one cycle on the logarithmic scale (and thus the error factor is approximately 10) are cases where no failures were observed.

In cases where no raw data are provided, bounds for individual points are computed from bounds in the data sources if these are available. The bounds are adjusted so that the upper bounds are approximately 95% bounds.

In risk assessment, knowledge of upper bounds on failure rates is important in analyses the seek to limit the risk. One is seldom concerned about how small a failure rate might be. Therefore, lower bounds are simply plotted as medians divided by error factors, and do not appear in the generic component failure data tables.

The NUCLARR system retains point estimates of failure rates such as estimates directly cited in a data source or numbers of failures divided by numbers of demands. However, the means and medians used in this handbook are generated by the routines used in OLARR's ad hoc aggregation. This means that they fit a legnormal distribution.

5.2. TREATMENT OF SEVERAL DATA POINTS Appendix C describes the methods implemented in NUCLARR for cambining and aggregating data from difference systems or plants. Some consequences of these methods that influence the interpretation of the upper bounds are presented here.

A common feature of the plots in Appendix B is that many points lie outside, and particularly above, the plotted aggregated bounds. Three aspects of the methodology account for this phenomenon:

24 l

. Points showing high failure rates are often cases where little exper-ience is available (that is, of ten such points are high because the l denominators are low) l l . In aggregations using raw data, often the data are pooled. The ag-I gregate is based or, the combined experience of the various sources.

In cases where many of the sources show no failures, the resulting average (sum of numerators divided by sum of denominators) is lower than any of the values contributing to it The sources are weighted in proportion to the amount of experience that they represent. That is, individual rates (numbers of failures divided by numbers of demands or operating hours) having large denom-inators are most influential in the results. These rates are often,

! though not always, among the lower values shown in the plots.

Two situations in the aggrega.: nm algorithms can give risq to low error factors. The first is the presence of a data point showing several failures and millions of operating ho:1rs or demands. A large number of failures drives down the uncertainty in the failure rate, as shown above. When these failures are associated with a data point showing a large amount of exper-ience, other data sources showing a variety of rates have little influence due to the weighting of tre dsta. Thi:; problem is compounderi by the fact _

that data points that reflect a great deal of experience often have beer p pooled across several components; it is possible that these components as a I group do not have a constant failure rate. ,

Second, lower error factors may also be the result of an anomoly in the upper bound calculation for the aggregation algorithm itself. As explained in Appendix C, the algorithm attempts to characterize the distribution of failure rates by considering different estimates of a rate rather than the uncertainty estimated for each individual rate. The Bayesian procedures esti-mate a variance between separate estimates of failure rates, subtracting out I the variation associated with the individual estinates. Unusually low error 25 1

X factors result if, for example, two points are close together even though tneir individual uncertainty bounds are large.

5.3. ADJUSTMENTS TO THE NUCLARR-CALCULATED M2ANS AND BOUNDS As noted above, the aggregation algorithms in some cases result in ano.aolies; for example, unusually low et ror factors have been observed.

Causes for the low error factors were discussed above. Some variety is expected for generic rates that will apply across the nuclear power plant industry; thus, the heavy influence of just one point or two points that are close together at described above may not be desirable. Because, in both of these situations, the assumptions being used to model the data processing appear unrealistic, the software that generates tables for this handbook from NUCLARR adjusts los error factors. Specifically, error factors that are less than 1.3 are set equal to 1.5 and the upper bounds and mean values (based on a fittad lognt,rmal dis'r. Jtion) are recomputed.

l' 26

l 6. GENERIC COMPONENT FAILURE DATA Generic aggregated failure rates for components and failure modes based on data gathered in late 1987 and entered into the NUCLARR data base in 1988 and early 1989 are presented in a series of four tables following the text in this section. Two levels of detail are provided. Tables 3 and 4 present, respectively, data for mechanical and for electrical / instrumentation compo-nents by just component type and failure mode. Tnese tables contain the most aggregated rates, with little detail required to qualify a record for inclu-sion in the aggregate. For many component type groups, only a single failure rate or probability for each failure mode is required for gross screening purposes, and Tables 3 and 4 provide this level of detail. The entries are listed in order by the codes used in the NUCLARR data base; this corresponds to an alphabetical ordering in most cases. Note that, for miscellaneous groups such as " electrical functicn items," no group-level aggregate is formed.

In Taules 3 and 4, seven pieces of information are given for the corres-ponding aggregate. They are: the units [per hour (h) or per demsnd (d)],

the number of data base records contributing to the rate, four parameters from the derived lognormal distribution, and a comment on th data source.

The lognormal parameters are the mean, median, upper 95% bound, and error factor. The Source column contains data source acronyms or " Study-ids" from Appendix A if less than three sources are involved. Where there are two sources, numbers in narentheses after the study-ids indicate the number of records from each source contributing to the aggregated data. If three or more sources are involved, the Source column contains the number of sources followed by "/ Plant" for operational data from commercial nuclear power plants, "/ Generic" for generic data from other sources, or "/Both." Data are classified as operational data from commercial nuclear power plants if the

" Failure Data Origin" field in NUCLARR is "PLNT"; for "IRADAP Suitable" records this means that maintenance work orders or other in-depth plant records were examined to determine the failure rates.

27 l

For applications requiring more detail than just component type and fail-ure mode, Tables 5 and 6 are provided. These tables are in the same format as Tables 3 and 4, but characterize the rates by component design and normal state in addition to the obvious component type and failure mode.

In generating Tables 5 and 6 from NUCLARR, the term " operating" was sub-stituted for the normal state in place of " running" in some cases to increase $

clarity. This change was made whenever the normal state was listed as run- t ning and the failure mode was " fail to operate" or " spurious operation."

, "Not specified" is a possible value for the design (these are denoted as

" Unknown" in NUCLARR). Similarly, " Unspecified" appears in these tables when the normal state is unknown. In addition, " general (preaggregated)" is a pos-sibility for the component design. This designation indicates that the corre- ,

sponding data were classified as preaggregated; i.e., as reflecting a mixture 1

of the component designs. NUCLARR records from sources such as IEEE-500 that supply data combined across component designs contribute to these entries.

Some overlap exists in Tables 3 and 5 and in Tables 4 and 6. Records )

for component types having only one design defined, such as battery chargers, contribute in the same way to both the generic failure data tables and the more detailed tables. The same is true in cases where a group-level aggre-gate is formed and all the data contributing to that aggregate come from one component design. On the other hand, only Tables 3 and 4 have aggregates across component designs and only Tables 5 and 6 have data for component designs that are in groupings for which a component-type-level aggregate is not meaningful. Specifically, the following are only in Tables 5 or 6:

Elbows Nozzles Auctionneers, averagers, and other specific computation modules l Inverters, controlled rectifiers, and other solid state power elec-tronics Some applications may require additional detail, such as knowledge of the specific system or a particular plant unit. Tables containing aggregated rates for data grouped by one of these variables in addition to the variables 28 i

in Tables 5 and 6 are not presented herein, because of the extreme sparsity of such data. Many records in NUCLARR lack plant unit information, and even more lack system details.

For applications requiring such additional detail, Part 3. Hardware Com-Donent Failure Datq, of Volume 5 of the NUCLARR documentation [ Reference 2e]

lists the supporting data that contribute to each aggregate. It contains a data sheet for each of the records in the NUCLARR hardware component failure data base. The records are listed in order by component type, design, fail-ure mode, and normal state. Virtually ali information stored in the data base is listed. The sheets which have a "Y" for the "lRADAP Sultable" entry, a data input date before June 8, 1989, and an earlier year than "1989" as the NUCLARR document reference year are included in the aggregates shown in tais handbook.'

in addition, the plots of the data included in Appendix B provide fur-ther information for component / failure mode / design / normal state groupings having at least four records. The reader can form more tailored aggregations using these data. Access to NUCLARR allows such aggmegations to be performed automatically.

Note that, although " fail to open" and " fail 1:s close" are special cases of the " fail to operate" failure mode, the data tables do not contain data aggregated across failure modes. The presentation of this type of aggrega-tion in the tables is cumbersome since such aggregates must be distinguished from rates that are based solely on the more general failure mode cate-gories. Also, the user might want to sum such rates rtither than average them. Again, reference to the NUCLARR documents or access to NUCLARR will allow the user to form such rates if desired.

u

8. rec 8ll f rom Iable 1 thdt data were received for entry to NUCL ARR too late to be included in this handbook.

29 l

w- _ _ _ _ - _ - _

TABLE 3. GENERIC FAILURE RATES FOR MECHANICAL COMPONENTS

n ,.

of tJpper Error a b Component Type Failure Mode (I Rec. Mean Median Bound Factor Source Air Conditioning (Jnits/ Chillers Fatis to Start /d 1 1.5E-002 1.3E 002 2.9E-002 2.3E<000 NEE 3-PRA.

Falls to Run /h 2 1.4E-005 1.1E-005 3.6E-005 3.2E+000 NE E3-PR A.

8 lowers / Compressors Fatis to Start /d 2 1.lE-001 1.1E-001 1.EE-001 1.5E4000 X-PRA.

Falls to Run /h 3 1.lE-004 9.1E-005 2.6E-004 2.8E+000 X-PRA(2),

RAC-85(1).

Control Rods Falls to Operate /h 1 4.5E-008 3.5E-008 1.1E-007 3.2E+000 SWD-HSK.

Control Rod Drives fails to Operate lh 2 5.2E-004 4.TE-004 1.0E-003 2.1E+000 SW-HBK. I Dampers, Exclusive of Operators ratis to Operate /h 1 1.9E-006 1.4E-C06 5.1E-006 3.7E+000 IEEE-500.

Dampers, Operator Not Specified Fatis to Operate /d 1 3.0E-003 2.7E-003 6.1E-003 2.3E+000 ZIS-PRA.

Fan Cooler tinits Fatis to Start /d 2 1.9E-003 1.8E-003 2.7E-003 1.5E+000 1PS-PR A(1),

4 ZIS-PRA(1).

Fatts to Run /h 2 8.8E-006 7.2E-006 2.lE-005 2.9E+000 ZIS-PRA(1).

IPS-PRA(1). {

w O Filters P lugged /h 2 4.8E-006 3.1E-006 1.5E-005 4.7E+000 IEEE-500.

Fans, w-ntilators Fatis to Start /d 1 1.5E-003 8.6E-004 4.BE-003 5.6E+000 HNP!-PRA.

Falls to Run /h 9 4.6E-006 1.7E-006 1.7E-005 9.9E+000 3/Both Heat Exchangers Internal Leakage /h 7 4.SE-005 1.0E-005 1.9E-004 1.9E+001 3/Both Plugged /h 4 1.6E-006 9.2E-001 5.1E-006 5.6E+000 IPS-FRA(2),

ZIS-PRA(2).

Ptpe, per system External Leakage /;upture /h 42 1.4E-007 8.0E-008 4.4E-007 5.5E+000 PBS. '

4 Pumps. Diesel Driven Falls to Start /d 9 9 3E-003 3.4E-403 3.5E-002 1.0E+001 S/ Plant Falls to Run /h 1 7.6E-002 6.4E- 02 1.7E-001 2.6E+000 ZIS-PRA. .

1 Pumps, Motor Ortven Fells to Start /d 81 4.1E-003 2.7E-003 1.2E-002 4.5E+000 9/ Plant Falls to Run /h 9/ 3.5E-005 9.3E-006 1.4E-004 1.5E+001 11/ Plant Pumps, Steam Turbine Driven Fatis to Start /d 9 2.0E-002 1.7E-002 4.3E-002 2.6E*000 8/ Plant Falls to Dun /h 9 1.!E-004 2.5E-005 4.3E-00a 1.7E+001 1/ Plant Pumps Driver hot Specified Falls to Run /h 6 2.6E-005 1.5E-006 8.4E-006 5.6E+000 IPS-PRA(2).

ZIS-PRA(4).

l m.= == .

D NN g; Nm .

ww . . . . ww me . . .

y b

et er a(

+.s tg (g ct u .=. me & CE e4

(. m 04 ee *d CE CK M at at

+**J ** **

%m -

• ' M W 4 4 8** W em

- 4JC C 4 6 tb C ww C4 w w C k CL C a C C C - C , Ccmwm Cww d'% 8 4 sg9- e e me e s.a) s*e e et at g o 4( g e ieeg g

==4 ==. - O--

et- eeg

- 03 .-

aC g g6 e ts e et M LE **

• O e - fr O n=, no CE Eg -M EE CE gr n.CL CL e uJ w'tCL- 4 CL CL uJ h EL CL (gCL 3 se.

ibAN(rs C 4 (L, CL (E. CL Cs 4 y (L M (L (L LE W e s s F we w s unmazac e a x >. sma u.s a e %

x > ~ cu a. m - m m x sg w s.

c a s ,J h (L s a h xuw.--w>~

g Z W OOOO O- O OO O DO COQO O -00 O-O 000 M b OOOO 00 O CO O OO OOOO O OOO OOO C5 OO y b O OOOO OO O OO O CO OOOO O OOO 000 O00 v Ou + + + + + + + ++ + + + + + + + 6 +++ + + + + + +

g b (#

L e LaJ 4k kJ haJ kJ tot uJ ti2 hJ LeJ w 40 hd tad w kJ knJ kJ LeJ kJ L. I teJ k.J w kJ W uA LJ M4

() wg e. Ch. 40. i:D. @ C.

. . wJ. m C. N. m. N. e. to. A. N. N. w er. P=.

. e. N. w.

nrg Ch er) e-t att e arl N@ Ch e ** m .* foi ett N a.e m M N ** N a#1 m -

g

• m eme mee000 emem we eOO me e Nm OO Nece OOOO 6

~i e ~

O00 OOOO O 00 O 300 LT CODO OO C O, O OO 0000 O 000 0, 00 000 N C e a e a e s O. e w

a e e e a e e s e a e e s e i h 2 ww 6J **J w w 4nJ laJ w kJ kJ kJ w kJ 6.J 6 J 6.e Jw w uJ uJ U uJ cc.W 4. ks.J

o. n qa. N. er.

o o. Cn. eo O. N. n.

o N. c) m. ==. . a0. -+ ar. . M.

Z gp

,. N Nu e 4

mo -m m. O. Un-. e e m =m m w-g 2 nn w mw w mm mene w ,wn n,w wnn

() e n,w8 000 OO O DO O 00 0000 o OOO 000 000 A -e O.O. O.O. O. O. O. O.O. O. O. O. O.O. O. O. O. O. O. O. G.o. O. O. O. O.

E wwww w w u. w uJ w wwww w w ud w www www

v. e. e. ~. 8. ~

w. > e e. ,. 0 y

Q 3 e ,. -. .- -~. m. . Cn.

. m.

m. w. o. n.

m-e- m

. m.

- - o. ,..n.~.

, , , - ~m~

sp s .ri , ~e e meOOe mm me~, e ,e~ ~me m~~

U O Ce O OO O O OO OOOO O OOO OOO OOO

< e. O. O. O. O. O. O. O. O. O. O. O. O.

00 0 0 O. 0 0 0, 0, 0 0 0 0, 0, U . had MJ ksJ %J ki kb ko isJ MJ %d LeJ %J QJ6ed MJ y Led %J Led %J %i QJ La.i %sh ksa %J K e0

~. w m. f*)

M W N. e-o. N.

. th. W. 0%. N. e. P%. N. O. >=. P=. w. (D. ~. N. N. O. at. . .

2 ~ ** " #" ** ** " ~ * N '" #**"" * * ~ * * * '" " ' ' * ** **

m e-m .n - ., o o N to m irs m* N w-O.e

,. .n m .

m m.4 O an- er wt M . N .

U E O CE w

.E*. reum er rss er .C um merc  : ver rur sss rrr al sss s ss s ss ssss s sss sss e

o La.

N 8 8 C

O W . . . . +*. ** . . . . es . .- .

y so sa ee . . 89e J tn e e.

e e.*

e . e

-M se O

e .s g e as

• • e.*s.

e et . e W

( $J k b C vt L b e b b C tt b C an b b b vt b b e.

4 C eA 2 -

5 jOpS E y SZ K ES K Opot R &

U .J O Qu UCOU &- O

& E*S Ohu $11 ES OOO OUJ 2 -

W 9 00 00 * * -

3 9 0 Oe O OO

+e *a O e ,* em O O e en OO e e ta -

3 3 g 6 +e +* ** 4J +* ** 3 ** ed 3 3 3 ++ +J 3 3 w ea 3 E O C <> OOO OO O

-OO r y en e e en +-

L b.

e em

• 5 e

--ee -et.+. 4

+

i.

an e+

e en -

u b- C, d - 3 ea +s .e- 3 .s= --s- += 3 3 3 - p. 3 . bg 3 3 es N 9 9 9 4 CL C 9 4 C1 4 ee 9 CL C1 Ch e e (*1 CL -e e C1 Q. n 6466 A- h. La. W La u. u hmeM 6 4 Mm 64M m W .C.e 8lC LA.

O M L L E O O Q .

es e s.

sa m

2 Ck A

b 4

O O

• e b

P* Cr ** e Q O 4 b

+d g C b. L.

tu -U CA O +-LJ O -

4 4 L 9 N 3 V

b

+*

J 3 Ca .

W  % O e- E W W I 3. E . a

.J -

M e.

e.

48 4J 4 J

,f .m GA a

P

- Cg p

W >

N e

>w

@ r A

31 k .

TABLE 3. GENERIC FAILURE RATES Fon HECHANICAL Cui.;F02.9TS (CONTINUED)

No.

of Upper Error a b Component Type Fallu e Mode U Rec. Mean Median Bound Factor Source Valves, Relief Falls to Operate /h 13 8.5E-008 4.9E-008 2.8E-001 5.6E+000 X-PRA.

Falls to Operate /d 20 1.2E-003 1.0E-003 2.8E-003 2.9E+000 3/P lant Falls to Open /d 5 3.6E-003 2.7E-003 9.2E-C03 3.4E+000 3/ Plant falls to Close /d 5 7.2E-004 6.4E-004 1.4E-003 2.2E+000 3/ Plant Spurious Operatton /h 1 7.0E-007 5.4E-007 1.8E-006 3.2E+000 IPR 05.

Spurious Oper. /h 5 2.0E-006 2.0E-006 3.0E-006 1.5E+000 $/ Plant Internal leakage /h 3 1.8E-007 1.4E-007 4.5E-007 3.2E+000 3/ Plant Valves, Solenoid Operator Falls to Operate /h 9 1.0E-006 5.5E-007 3.1E-006 5.6E+000 Y-PRA(4),

X-PRA(5).

Falls to Operate /d 9 1.1E-003 5.7E-004 3.8E-003 6.7E+000 3/ Plant Spurious Operation /h 1 1.7E-005 1.3E-005 4.3E-005 3.2E+000 IFROS.

Valves. Vacuum Breakers Falls to Operate /h 2 4.4E-004 2.5E-004 1.4E-003 5.6E+000 Y-PRA.

Falls to Operate /d 2 1.lE-002 6.3E-003 3.5E-002 5.5E+000 Y-FRA.

Valves, Operator Not Specif ied C

Falls to Operate /h 1 1.3E-004 1.2E-004 1.8E-004 1.5E+000 MN$t-PRA.

Fails to Operate /d 2 5.6E-004 3.?E-004 1.8E-003 5.6E+000 Y-PRA.

Fatis to Open /d 3 1.9E-002 1.0E 502 5.4E-002 6.2E+000 BRPI-PRA.

Falls to Close /d 1 8.6E-002 8.4E-D02 1.3E-001 1.5E+000 BRP1-PRA.

Spurious Open /h I 8.9E-007 7.6E-001 1.8E-006 2.3E+C00 BRPI-PRA.

(d Spurious Close /h 2 2.7E-007 1.6E-007 8.5E-007 5.4E+000 ERP1-PRA(1),

" 215-PRA(1).

Internal Leakage /h 2 8.00-005 4.6E-005 2.iO OO4 5.6E+000 Y-PRA.

a. Adjusted to equal 1.5 f f the NUCL ARR-calculated error factor is less than 1.5 (see Section 5.3).

b. Data snurce acronyms f rta Appendts A if < 3 sources are insolved; otherwise, the number of sources followed by " plants" for operational data f rom corrnercial nuclear power plants, " generic" for generic data f rom other sources, or "both."

c. Most of these ars control alves.

TABLE 4. GENERIC FAILURE RATES FCR ELECTRICAL /INSTCUMENTATICN COMPCNENTS No.

of Upper Error a b failure Mode U Rec. Mean Median Sound factor Socce Co%wt Type Automatic Transfer Switches Fatis to Transfer /d 1 2.0E-002 1.5E-002 4.9E-002 3.2E+000 IPS-PRA.

t. lect r ica lly Annunciators Falls to Operate /h I 9.0E-007 3.9E-007 3.3E-006 8.4E+000 IEEE-500.

Spurious Operation /h 1 3.0E-006 1.3E-006 1.1E-005 8.EE+000 IEEE-500.

Batteries Falls to Operate /5 10 1.lE-006 1.0E-006 2.lE-006 2.2E+000 6/ Plant Fails to Operate /d 2 1.2E-002 1.2E-002 2.2E-002 1.9E+000 SWD-HBK.

Charger. Battery Fatis to Operate /h !!  !.0E-005 7.5E-006 2.5E-005 J.3E+000 7/ Plant Ctreutt Breakers. Molded Ease Falls to Operate /h 1 2.1E-006 1.lE-006 6.9E-006 6.IE+000 IEEE-500.

Fails to Ocerate /d 7 4.9E-004 2.EE-004 1,7E-003 6.4E+000 3/Both Circuit Breakers. Po-er Fatis to Open /d 4 1.1E-003 5.3E 304 4.lE-003 7.BE+000 3/ Plant

/

Falls to Close 'd 3 1.EE-003 1.5E-003 2.3E-003 1.5E+000 3/ Plant 5purtous Operation /h 1 3.BE-007 3.3E-007 7.6E-007 2.3E*000 SVD-HBK..

Spurious Open th 6 3.1E-007 3.1E-007 4.6E-007 1.5E+000 a/Both Spurtous De-energtre /h 3 6.3E-007 4.9E-007 1.6E-006 3.2E-000 X-PE A.

/h 5 1.lE-007 8.2E-008 2. 7E -007 3.2E+000 4/ Plant y Conductors ratis to Operate Computation Modules fatis to Operate /h 6 3.7E-006 1.7E-006 1.3E-005 7.5E+000 IEEE-500.

Controllers / Regulators. Fails to Operate lh 2 3.4E-006 1.5E-006 1.2E-005 8.CC+000 IEEE-500.

Non-mec5an ica l Fuse Spurtous Operatten /n I 1.3E-007 1.2E-007 2.3E-007 f.9E+000 RAC-85.

Generator. with Diesel Engine Falls to Start /d 76 5.3E-003 4.0E-003 1.4E-002 3.4E+000 7/ Plant Detver 4.0E-003 2.6E-003 1.2E-002 4.;E+000 5/ Plant Fails to Pt.n /h 5 Fails to Run /d 70 1.0E-002 7.5E-003 2.6E-002 3.4E+000 NSAC-108.

Generator. with Hydro furbine Fails to Start /d 1 3.1E-003 3.0E-003 4.7E-G03 1.6E+000 NEE 3- PG A .

Driver Generator, with Mater Driver Fails to Run /h  ! ? 2E-005 2.IE-005 3.2E-005 1.5E+000 SWD-HBK.

Generator, with Gas Turbt,e fatis to Start /d 1 3.5E-002 3.4E-002 5.0E-002 1.5E+000 MMSI-PRA.

Oriver 2. DE-004 6.EE-004 3.2E+000 M451-PRA.

fails to Eun /5 1 2.EE-004

1 l

l TABLE 4. GENERIC FAILURE RATES FOR ELECTRICAL / INSTRUMENTATION COMPONENTS (CONTINUED) ho.

of Upper Errer a b U Rec. Mean Median Bound Facter Source Component Type F a ilur e Mode Heaters fails to Operate /h I 8.9E-006 7.8E-006 1.8E-005 2.3E+000 RAC-85.

Relays, Control Fatis to Operate /h 7 3.7E-007 3.EE-007 5.4E-007 1.5E+000 RAC-85.

Fails to Open /d 1 3,2E-006 3.lE-006 4.9E-006 1.EE+000 IEEE-500.

Fails to Close /d 1 4.lE-006 3.9E-006 6.2E-006 1.EE+000 IEEE-500.

Falls to Operate /h 7 1.IE-006 1.0E-006 1.9C-006 1. 9E+C3 RAC-85f7,,

Relays. Protective IEEE-500(1).

Fails to Open /d 1 1.4E-007 5.3E-007 2.CE-006 3.6E+000 IEEE-500.

Fails to Close /d 1 4.2E-006 3.0E-006 1.2E-005 3 9E+000 IEEE-500.

Transducers fails to Operate /h 2 1.5E-005 3.0E-006 5.6E-005 1.8E+001 IEEE-500.

(Det,c tors /E lement s/ Sen ser s )

Fatis to Operate /d 5 2.8E-004 2.1E-004 7.3E-004 3.4E+000 SWD-HBK.

Spurious Operatton /h 10 1.0E-006 7.9E-007 2.6E-006 3.3E+000 SWD-H6A.

Switches, Iristrumentat ion Fails to Operate /h 3 1.CE-006 4.lE-007 3.6E-006 8.9E+000 IEEE-500(2).

SWD-H3K(1).

w Falls to Operate /d 3 2.2E-005 7.3E-007 5.4E-005 7.4E+001 IEEE-500(2),

1. SWO-HSK(1).

Spurtous Operation /h 3 5.8E-007 2.4E-007 2.lE-006 9.3E+000 $WD-HBK(1),

IEEE-500(2).

Switches, Solid-state Falls to Operate /h 1 2.3E-006 8.2E-007 8.8E-006 1.lE+001 IEEE-500.

/d 3.7E-007 2.BE-007 9.4d-001 3.4E+000 IEEE-500. I Fatis to Operate 1 Spurtous Operation /h 1 5.6E-007 2.lE-001 2.2i-006 1.0E+001 IEEE-5GO.

Transmitters Falls to Operate /h 10 2.0E-006 1.3E-006 6.3E-006 5.0E+000 SWo +BK( 7) .

IEEE-500(3).

Spurtoos operation /h 3 1.lf-005 2.0E-006 4.0E-005 2.0E+001 IEEE-500.

Transformers, Control and f ails to Operate /h 5 1.lE-006 7.J-007 3.0E-006 3.9E+000 3/Soth Ir'st rument at ion Transformers Power fails to Operete /h 5 8.2E-007 6.4E-007 2.1E-006 3.2E+000 3/ Plant

a. Adjusted to equal 1.5 if the NUCE AWealculated error f actor is less than 1.5 (see Section 5.3).

b. Data source acrones f rom Appendta A tf < 3 sources are involved; otherwise the number of sources followed by " plants" f or coerat tonal data f rom corwrcial nuc lear pcwer plants. " generic" f or generic dat a from other sour ces, or **both."

~ .

TABLE 5. GENERIC FAILURE RATES FOa MECHANICAL COMPONENTS WITH VARIOUs DESIGNS AND NORMAL STATES 42.

kormal State cf Upper Error b c and Failure Mode, Rec. Mean Median Bouro Factor Source Co w t Type and Destgn U Air Conditioning Units / Chillers Air Conditiening Units Running /Fatis to Run /h 2 1.4E-005 1.lE-005 3.6E-005 3.2E+000 REE3-PRA.

Alt./ Fails to Start /d I 1.5E-002 1.3E-002 2.9E-002 2.3f+000 NEE 3-PSA.

Bicwers/Ctrpressors Corprassors Star:dby/ Fails to Run /h 2 6.IE-C05 5.9E-005 8.9E-005 1.5E+000 UFRA.

Unsp./ Falls to Run /h 1 2.21-004 2.2E-004 3.3E-004 1.5E+000 RAC-85.

Standby /Fatis to Start /d 2 1.1E-C01 1.1E-001 1.EE-C01 F.5E+000 1-PRA.

Control Rods Control pods Urso./ Falls to Operate /h 1 4.5E-006 3.5E-006 1.lf-0C1 3 fE+000 SWD-MSK.

! Control Rod Drives Control Rod Dr tves tmsp./Fa 61s to Operate /h 2 5.2E-004 4.7E-004 1.0E-003 2. !E +000 Si/D-HOK .

Damers. Euclusive of Operators l

Geeeral (preaggregated) Closed /Fatts te Operate ih 1 1.9E-006 1.4E-006 5.1E-006 3.7E+000 IE E E -50'1.

l w Darnpers. Operator mot Specified

• Design Not Specified Alt.! Fails to Operate /d 1 3.0E-003 2.?E-003 6.!E-003 2.3E+000 ZIS-PRA.

Fan Cooler ibits Fan Cocler lhtts Alt./ Fails to Run /h 2 8.EE-006 7.2E-006 2.1E-005 2.9E+c00 ZIS-PEA (1).

IPS- W 't).

Alt./Fatis to Start /d 2 1.9E-003 1.8E-003 2.1E-003 1.5E+000IPS-?RA,t). ZIS-PL :).

Fittings Elbow d Running /Esternal /n i S.5E-004 1.9E-005 1.9E-003 1.0E +002 IEEE-500.

Leakage / Rupture Fans. Venttlators Faas Alt./ Falls to Ran /h 1 1.9E-006 1.1E-006 6.2E-006' 5.EE*000 MPI-FR A .

Unso./ Falls te Run /h 7 5.5E-006 1.6E-006 2.1E-005 1.2E+00; DAC-85.

Alt./ Falls to Start /d 1 1.5E-003 8.EE-004 4.tE-003 5.EE+000 h%PI-FEA.

Ventilstors Runnirg/Fatis to aun /h 1 2.EE-006 2.5E-006 4.5E -006 1.8E+000 IEEt-500.

n

k. ..

w v. x ._ .

TABLE 5. GENERIC FAILURE RATES FOR MECHANICAL COMPONENTS l WITH VARIOU3 DESIGNS AND NonMAL STATES (CONTINUED)

No.

Normal Staw o' Upper Error b c and failure Mode, U Re, Hean Median Bound Factor Source Component Type ind Design Hest Exchangers Steam Generator Running / Internal Leakage /h  ! 5.6E-005 4.9E-005 1.2t 004 2.4E+000 IEEE-500.

Design Not Specified Open/ Plugged /h 4 1.EE-006 9.2E-007 5.lE-006 5.6E+000 ZIS-PRA(2),

IPS-PRA(21 Running / Internal Leakas- /h 2 ?.3F-006 7.5E-007 4.2E-006 5.6E+000 ZIS-PRA(1),

IPS-PRA(1).

Untp./' r nel Leakege /h 4 2.7E-005 2.4E-005 5 SE-005 2.3E+000 IEEE-500.

Miscellaneous Piping items Nonle (Note d) External /h 1 9.2E-004 1.8t-005 1. BE-tM3 1.0E+002 IEEE-500.

i- sge/ Rupture Pipe' Pipe, 7- ' Inches, ID Running / External /h 13 1.3E-001 :.lE-007 3. lE- 007 2.8E+000 PBS.

Leakage / Rupture Standby / External /h 8 1.4E-007 1.3E-007 2.8E-007 2.2E+000 PBS.

Leakage / Rupture Ptpe, < 3 Inches, ID Running /Enternal /h f4 1.8E-007 1.0E-001 6 3E-007 6.6E+000 PBS.

y Leakage / Rupture Standby /Enternal /h 7 6.3E-003 4.92-008 1.6E-007 3.2E+000 PBS.

Leakage / Rupture Pumps. Daesel Ortwen Centrifgst Standby / Falls to Start /d 5 1.1E-002 3.4E-003 4.3E-002 1.3E4001 IPRDS.

Design het Specified Standby /Fa tis to Run /h 1 7.6E-002 6.4E-002 1.7E-0J1 2.6E+000 ZIS-PRA.

Standby /Fatis to Start /d 4 8.2E-003 4.3E-003 2.BE-002 6.5E+G00 4/ Plant Pumps. Mator Priven Centrifugal Alt./ Falls to Run in 25 2.6E-005 1.9E-005 1.2E-005 3.8E+000 IPRDS.

Running / Fails to Run i" 16 2.4E-005 1.1E-005 8.7E-005 8.2E+000 '/i Plant

• Standby / Falls to Run n. 4 1.2E-005 9.2E-006 3. 0E-005 3.2E+0F0 NEE 3-PRA(3),

Z15-PRA(1).

Alt./Fatis to Start /J 11 6.2E-C13 5 9E-003 1.0E-002 1. E.E + 000 IPRDS.

Standby / Falls to Start /d 16 6.4E-003 4.6E-003 1.7E 002 3.8E+000 NEE 3-PRA(4), 1 i

IPRDS(12).

Unsp./ Falls to Start /d 1 9.4E-003 8.SE-003 1.6E-002 1.8E+000 IPS-PRA.

I TABLE 5. GENERIC FAILURE RATES FOR H cHANICAL COMPONENTS I wITH VARIOUS DESIGi4S AND NORMAL STATES (CONTINUED)

No.

Nermal State of Upper Error b c Rec. Mean Median Bound Factor Source Ccepon ' it Type and Design and Failure Mode

• U Pumps, Motor Driven (cont *d)

Design Not Specified Alt./ Fails to Run /h 12 3.5E-005 2.lE-005 1.lE-004 5.3F+000 5/P lant Running / Falls to Run /h 19 4.5E-005 2.9E-005 1.3C-004 4.6E+000 4/P lant Standby / Fails to Run /h 20 6.3E-005 7 21-006 2.2E-004 3.lE+007 8/P lant Unsp./ Falls to Run /h 1 5.2E-005 5.CE-005 7.5E-005 1.5E+000 EUR-PRA.

Alt./Fa 6ls to Start /d 11 4.0E-003 1.8E-003 1.4E-002 7.SE+000 4/ Plant Running / Fails to Start /d 6 9.2E-003 6.lE-003 2.1E-002 4.4E+000 X-PRA.

Standby / Falls to Start /d So 2.9E-003 2.0E-003 8.4E-003 4.2E+000 8/ Plant 9umps. Steam Turbine Driven Centrifugal Running / Fails to Run /h 2 1 OE-004 1.02-004 1.5E-004 1.5E+000 IFRDS.

Standby / Falls to Start /d 2 2.6E-002 2.5E-002 4.lE-002 1.EE+000 IPROS.

Design Not Specifiec Standby / Fails to Run /h 5 2.EE-003 1.3E-003 8.9E-003 7.0E+000 5/ Plant Unsp./ Fails to Run .h 2 7.7E-005 7.5E-005 1.lE-004 1. 5E + 000 EUR-PRA(1),

X-PRA(1).

Standby / Fails to Start /d 6 1.8E-002 1.3E-002 4.7E-002 3.5E+000 6/ Plant Unsp./ Fails to Start /d 1 7.lE-002 5.7E-002 1.7E-001 3.IC+000 EUR-PRA.

Pumps. Driver Not Specified

$ Design Not Specified Running / Fails to Run /5 2 3.lE-006 1.8E-006 1.0E-005 5.EE+000 ZIS-PRA.

5. 6E + 000 Z I S-PR A ( 2 ) ,.

Standby / Falls to Run /h 4 1. 5E -005 8.8E-006 5.0E-005 IPS-PRA(2).

halves, Check Manual Op-i or (mechanical Unsp./ Falls to Operate /d 2 1.6L-003 1.0E-003 5.7E-003 5.6E+000 X-PRA.

ha ndwee 1)

Unsp./ Internal teskage /h 3 6.3E-007 3.6E-007 2.0E-006 5.6E+00G X-PRA.

Motor Operator Unsp./ Falls to Operate /h 1 7.4E-006 4.3E-006 2.4E-005 5.6E+000 X-PRA.

Unsp./ Fails to Operate /d 1 7.0E-003 6.7E-003 1.lE-002 1.7E+000 SVD-HBK(5),

X-PRA(2),

No Operator Unsp./ Falls to Close /d 1 1.2E-303 6.8E-004 3.8E-003 5.6E+000 .TE3-PRA.

bnsp./ Fails to Operate /u 70 2 BE-005 2.2E-005 7.lE-005 3.2E+005 Y-PRA(il),

X-PPA (9).

Unsp./ Falls to Open /d 1 1.2E-003 6,6E-004 3.8E-003 5.EE+000 NEE 3-PRA.

Unsp./ Internal teakage /h I? 2.2E-005 9.3E-007 7.9E-006 8.5E+000 Y-PRA(2),

X4R A(10) .

Open/Sputous Close /h 1 3.4E-006 2.0E-006 1.lE-005 5.Ef f '.0 hE E 3- PR A .

4

- n -

TABLE 5. GENERIC FAILURE RATES FOR MECHANICAL COMPONENT 3 WITH VARIuuS DESIGNS AND NORMAL STATES (CONTINUED)

No.

Normal State of Upper Error b c and failure Mode, O Rec. bean Median Bound Factor Sotrce Component Type and Design l Valves, check (cont'd)

No Operator (tilting disk Unsp./ Falls to Close /d 1 5.5E-004 4.3E-004 1. E-003 3.2E+000 NEE 3-PRA.

checkvalve)

Unsp./ Fails to Operate /d 7 3.7E-004 2.9E-004 9.3E-004 3.2E+000 X-PRA.

Unsp./ Falls to Open /d 1 1.8E-001 1.lE-004 6.0E-004 5.6E+000 NEE 3-PRA.

Unsp./ Internal Leakage /h 8 1.4E-006 1.2E-006 3.0E-006 2.6E+000 X-PRA.

Open/Spurtous Close /h 1 2.lE-006 1.2E-006 6.8E-006 5.6E+000 HEE 3-PRA.

Pne mutic Operator Unsp./ Falls to Ope < ate /d 7 3.8E-003 2.0E-003 1.3E-002 6.3E+000 SWD-HBK(1),

Y-PRA(6).

Unsp./ Internal Leakage /h 6 1.9E-004 1.6E-004 4.lE-004 2.EE+000 Y-PRA.

Operator Not Specified Unsp./ Falls to Close /d 3 1.8E-003 1.6E-003 3.5E-003 2.2E+000 NEE 3-PRA(1),

MNSI-PRA(2).

Unsp./ Falls to Operate /d 1 6.9E-004 4.M -004 2.2E-003 5.6E+000 Y-PRA.

Unsp./ falls to Open /d 7 3.8E-005 2 2E-UO5 1.2E-004 5.6E+000 5/ Plant Unsp./ Internal teakage /h 5 3.4E-006 1.9E-007 1.0E-005 5.3E+001 4/ Plant Open/Spurtous Close /h 1 1.5E-006 8.4E-007 4.7E-006 5.6E+000 NEE 3-PRA.

Other Operator Unsp./ Fails to Operate /h 1 1.9E-005 1.lE-006 6.0E-006 5.6E+000 X-PRA.

Unsp./ falls to Operate /d 2 2.0E-003 1.2E-003 6.5E-003 5.4E+000 SWO-HSK(1),

X-PRA(1).

y Valves, Hydraulic Operator Design E t Specified Open/ Fails to Operate /d i 4.1E-003 3.9E-003 6. 4E- 003 1.6E+000 NEE 3-FRA.

Unsp./ Fails to Operate /h 2 1.9E-00" 1.1E-006 6.0E-006 5.EE+000 X-PRA(1),

Y-PRA(1).

Ursp./ Falls to Operate /d 3 6.4E-004 5.0E-004 1.6E-003 3.2E+000 X-PRA(1),

Y-PRA(2).

Open/ Spurious Close /h 1 1.8E-006 1.lf-006 5.9E-006 5.6E+000 NEE 3-PRA.

Valves, Motor Operator Angle Unsp./ Fails to Operate /d 1 3.0E-003 2.3E-00.1 7.4E-003 3.2E+000 IPR 05.

%tterf ly Unsp./ Fails to Operate /d 2 5.6E-003 5.3E-003 8.4E-003 1.6E+000 g re.5, Unsp./ Spurious Operation /h 2 1.0E-006 9.4E-007 1.9E-006 2.1E+000 h 35.

Glot>e Unsp./ Fails to Operate /d 1 9.5E-003 9.2E-003 1.4E-002 1.5E+000 IPRDS.

Unsp./ Spurious Operation /h 1 2.0E-006 1.7E-006 4.3E-006 2.6E+000 IFR05.

Gate Unsp./ Tails to Operate /d 2 1.lE-003 1.0E-003 1.6E-003 1. 5E +00') IPR 05.

Un.p./Spurtous Operetton /h 2 1.9E-006 1.8E-006 2.8E-UO6 !.5E+000 IPR 05.

TADLE 5. GENERIC FAILURE RATES FOR MECHANICAL COMPONENTS WITH VARIOUS DESIGNS AND NORMAL STATES (L.NTINUED)

No.

Normal State af Upper Error C Comorent Type and Design and Failure Mode

• O Rec. Mean Medt n Bound Factor Source Vsl.es, Motor Operator (cont'd)

Design hot Specified Open/ Falls to Close /d 5 5.CE-003 3.9E-003 1.3E-002 3 2E+000 3/F lant Closed / Falls to Operate /d 2 5.lE-003 4.9E-003 1.4E-003 1.5E+000 215-P9A.

Cpen/Fatis to Operate /d 2 7.0E-003 4.lE-003 2.2E-002 5.4E+000 HEE 3-FRA.

Unsp./ Falls to Operate /h 10 4.8F-006 1.9E-006 1.8E-005 9.6E+000 K-PRA(9).

Y-PRAll).

Unsp./ Fails to Operate /d 12 1.9E-003 5.5E-004 7.3E-003 1,3E+001 4/Tlant Closed /Fatis to Open /d 4 6.2E-003 5.9E-003 9.2E-003 1.5E+000 BRPI-PRA(2),

C l-PRA(2).

- Open/ Spurious Close /h 3 2.6E-008 1.5E-008 8.5E-008 5.6E+000 3/ Plant Closed /Spurtous Open /h 3 7.2E-007 4.lE-00T 2.3E-006 5.6E+000 215-PRA(1),

IPS-PRA(2).

Unsp./ Spurious Open /h 1 1.2E-006 9.3E-007 3.0E-006 3.2E+000 BRP1-PRA.

Other Unsp./ Falls to Operate /d 2 4. 7E-003 - 4.5E-003 6.8E-003 1.5E+000 SWD-eEst.

Valves, Manual Operator (handwheel)

Globe Unsp./ Falls to operate /d 1 4,2E-003 3.5E-003 9.lE-003 2.6E+000 IPRDS.

W Gate Unsp./ Falls to Operate /d 1 2.9E-003 2.6E-003 5.9E-003 2.3E+000 IPRDS.

• I Orso./ Spurious Operation /h 1.7E-006 1.3E-006 4.3E-006 3.2E+000 IPRDS.

Desigi kot Spectfted Unsp / Falls to operate /h 14 3.7E-008 3.lE-008 8.2i-008 2.6E+000 X-PRA.

Unsp./ Falls to Operate /d 16 4.3E-004 1.0E-004 1.7E-001 1.6E+001 3/ Plant Open/ Spurious Close /n 3 1.2E-007 1.0E-007 3.lE-007 3.2E+000 3/ Plant Valves, Pneumatic Operater Butterf ly Vasp./ Falls to Operate /d 2 2.8E-003 2.JE-003 4.0E-Oc3 1.5E+000 1*RDS.

Unsp./ Internal Leakage /h 1 4.9E-007 4.lE-007 1.1E-006 2.6E+000 IPRDS.

Unsp./Spurtous operation /h 2 6.4E-007 4.4E-001 1.8E-006 4.1E+000 IPRDS.

Ball Unsp./ Internal Leakage /h 1 1.9E-006 1.7E-CC6 3.8E-006 2.3E+000 IFRDS.

Globe unsp./ Falls to Operate /d 2 3.4E-003 3.3E-003 5.2E-003 1.6E+000 IPEDS.

Unsp./ Internal teakage /h 1 4.8E-007 4.!E-007 1.1E-006 2.6E+000 IPitDS.

Unsp./ Spurious Operation /h 1 1.3E-006 1.2E-006 2.2E-006 1.9E+000 IPRDS.

Gate Unsp./Fatis to Operate /d i 1.0E-002 1.0E-002 . SE-002 1.5E+000 IFRDS.

Unsp./ Internal teakage /h 1 8.2E-007 6.3E-007 2.? -006 3.2E+000 IPRDS.

Unsp./ Spurious Operation /h 1 3.0E-006 2,7E-006 5.4E-006 2.0E+000 IPRDS.

l l

l l

l

2 .

TABLE 5. GENERIC FAILURE RATES FOR MECHANICAL COMPONENTS WITH VARIOUS DESIGNS AND NORMAt. STATES (CONTINUED)

No.

Normal State of Upper Error b c and failure Mode, U Rec. Mean Hedlan Sound Factor Source Component Type and Design l

l Valves, Pr.eumatic Operator (cont'd) i l Design Not Specified Closed / Fails to Operate /d 1 2.6E-003 2.3E-003 5.2E-003 2.3E+000 NEE 3-PRA.

Unsp./ Fails to Operate /h 9 5.2E-007 4.4E-007 1.lE-006 2.6E+000 X-PPA.

Unsp./Fal's to Operate /d 13 1.4E-003 2.3E-004 5.2E-003 2.2E+001 5/P' ant Unsp./ Internal leakage /h 1 1.6E-006 1.5E-006 2.6E-006 1.7E+000 Y-PRA.

Open/ Spurious Closa /h 4 4.3E-001 3.4E-007 1.1E-006 3.2E+000 4/ Plant Unsp./Spuricus Operation /h 1 4.9E-006 4.5E-006 8.8E-006 2.0E+000 E UR -PRA.

Closed / Spurious Open /h 1 1.4E-005 7.8E-006 4.4E-005 5.6E+000 IPS-PRA.

Valves. Relief Indtrect Acting. Pilot Open/ Fails to Close /d 2 3.6E-002 2.8E-002 8.8E-002 3.2E+000 NEE 3-PRA.

Operated Unsp./ Fails to Close /d 2 1.4E-003 1.4E-003 2.lE-003 1.5E+000 SVD-HBK.

Closed / Falls to Open /d 2 1.2E-00? 7.0E-003 3.8E-002 5.5E+000 NEE 3-PRA.

Unsp./ Fails to Open /d 2 3.8E-003 2.9E-003 9.4E-003 3.2E+000 SVD-HBK.

Closed / Spurious Doen /h 1 4.0E-006 3.4E-006 8.9E-006 2.6E+000 ZIS-PRA.

Indirect Acting. Power Closed / Falls to Operate /d 1 2.9E-003 1.7E-003 9.3E-003 5. 6E +000 X-PRA.

3' Operated C' Closed / Spurious Open /h 1 3.7E-006 2.lE-006 1.2E-005 5.6E+000 X-PRA.

Indirect Act ing. Operater Closed / Spurious Open /h 1 5.0E-006 3.9E-006 1.3E-005 3.2E+000 IPS-PR A.

Not Specif ied Design Not Specified Closed / Fails to Close /d 1 5.7E-004 5.2E-004 1.1E-003 2.lE+G00 IPRDS.

Closed / Falls to Operate /d 5 8.lE-004 7.9E-004 1.:E-003 1.5E+000 Y-PRA(4).

IPROS(l).

Unsp./ Fails to Operate /h 13 8.5E-008 4.9E-008 2.8E-007 5.EE+000 X-PRA.

Unsp./ Fails to Operate /d 14 2.5E-003 2.4E-003 4.0E-003 1.7E+000 X-PRA.

Closed / Fails to Open /d 1 5.9E-003 3.4E-003 1.9E-002 5.6E+000 E UR-PR A .

Closed / Internal teakage /h 9 1.8E-007 1.4E-007 4.5E-007 3.2E+000 3/ Plant Closed / Spurious /h 1 7.0E-007 5.4E-007 1.8E-006 3.2E+000 IPROS.

Operation Closed / Spurious Open /h 1 9.4E-006 5.4E-006 3.0E-005 5.6E+000 E UR -PR A .

Unsp./ Spurious Open /h 1 1.9E-006 1.7E-006 3.!E-006 1.8E+000 SVD-H6K.

TABLE 5. GENERIC FAILURE RATES FCR MECHANICAL COMPONENTS WITH VARIOUS DESIGNS AND NORMAL STATES (CONTINUED)

No.

Norr.a1 State of Upper Error b c and failure b k , U Rec. Mean *edian

. Sound Factor Source Component Type and Design Valves, Solenoid Operator G lobe Unsp./Falle to Operate /d 1 2. l E- 002 1.aE-002 4.5E-002 2.6t+000 IFRDS.

Unsp./Spurtous Operation /h 1 1.7E-005 1.3E-005 4.3E-005 3.2E+000 IPRDS.

Unsp./ Fails to Cperate /h 9 1.0E-006 5.5E-007 3.1E-005 5.EE+000 Y-FRAf4),

Design hot Specified X-FRA(5)-

i Unsp./ Fails to Operate /d 8 1.2E-004 6.1E-004 1 CE-003 2.6E+000 1-PGA(4),

X-PRA.8).

Valves. Vacuum Breakers Y-FRA.

valves, vacuum Breakers Closed / Falls to Operate /h 2 4.aE-004 2.5E-004 1.4E-003 5.6E+000 Closed / Fails to Operate /d 2 1.lE-002 6 J.-003 3.5E-002 5.5E4000 Y-FRA.

Valves. 0;erater Not Specified Design hot Specified (hote f) Open/Fatis to Close /d 1 8.6E-002 8 4E-002 1.3E-0Cl 1.5E+000 82PI-PRA.

Closed / Fails to Operate /h 1 1.3E-004 1.2E-004 1.8E-004 1.5E+000 M 51-PRA.

Unsp./ Fails to Operate /d 2 5.EE-004 3.2f-004 1.BE-003 5.6E+000 Y-PRA.

Closed / Fails to Open /d 3 1,9E-002 1.0E-002 6.4E-002 6.2E+000 BRPI-PRA.

Unsp./ Internal teakage /h 2 8.0E-005 4.EE-005 2.6E-004 5.EE+000 Y-PRA.

A Open/Spurtous Close /h 1 4.5E-008 2.6E-008 1.5E-007 5.6E+000 215-PDA.

~ Ursp./Sourious Close /h 1 1.lE-006 1.0E-006 2.2E-006 2.IE+000 BRP1-FRA.

Unsp./ Spurious Or.en /h I 8.9E-007 7.8E-007 1.8E-006 2.3E+000 BAP1-PRA.

a. Alt. normal state is alternating; Unsp. Unsrecifted normal state.

b. Adjusted to equal 1.5 if the NUCL ARR-calculated error f actor is less than 1.5 (see section 5.3).

c. Data source acronyms f rom Appeadix A if < 3 sources are involved; otherwise, the number of sources followed by " plants" for operational data f rom comercial nuclear power plants. " generic" for generic data from other sources, or "both."

d. A large uncertainty is associated with the f ailure rate for this compenent. The error factor has been truncated at 100 and the upper bound adjusted to equal the mediai times 100. More cats f or this comorent are needed to reduce this uncer-tainty.

Pictng f ailure rates are per hour per nuclear power plant system Also, a "rur>ning" normal state for passive comporents e.

inctcates that the compocent is maintaining a pressure bou<dary, f Most ci these are ccetrol valves. _

col TABLE 6. ' GENERIC FAILURE RATES FOR ELECTRICAL / INSTRUMENTATION COMPONENTS WITH VARIOUS DESIGNS AND NORMAL STATES No.

Normal State b of Upper Error C d a

and Failure Mode U Rec. Mean Median Bound Factor Source Component Type and Design Automatic Transfer Switches 1.5E-002 t.9E-002 3.2E+000 IPS-PRA.

Design Not Specified Standby / Fails to /d 1 2.0E-C Transfer Electrically Annunclators Visual Oe-energized / Falls to /h 1 9.0E-007 3.9E-007 3.3E-006 8.4E+000 IEEE-500.

Operate Oe-energized / Spurious /h 1 3.0E-006 1.3E-006 1.1E-005 8.6E+000 !EEE-500.

Operation Batteries 1.lE-006 1.0E-006 2.lE-006 2.2E+000 5/P lant Vet Cell (e.g., lead-acid) Standby / Falls to Operate /h 9 Standby / Fails to Operate /d 1 8.2E-003 4.8E-003 2.6E-002 5.5E+000 SWD-HBK.

Unsp./ Falls to Operate /h 1 2.5E-006 1.4E-006 8.0E-006 5.6E+000 X-PRA.

Design Not Specified Standby / Falls to Operate /d 1 1.4E-002 1.3E-002 2.4E-002 1.8E+000 SWD-HBK.

Charger, Battery Charger, Battery Energized / Fails to /h 11 1.0E-005 1.5E-006 2.5E-005 3.3E+000 7/ Plant Operate y

Circuit Breakers, Molded Case G*neral (preaggregated) 'Jnsp./ Falls to Operate /h 1 2.!E-006 1.IE-006 6.9E-006 6.lE+000 IEEE-500.

Circuit Sreakers, Power Design Not Specified Open/ Fails to Close /c 3 1.6E-003 1. 5E -003 2.3E-003 1.5E+000 3/ Plant Closed / Fail * . Operate /d 3 2.6E-004 1.0E-004 1.0E-003 1.0E+001 MNSI-FRA(2).

SWD-HBK(1).

Unsp,/ Falls te Operate /d 4 8.8E-004 fi . 4E-004 2.4E-003 3.7E+00-0 X-PRA(2),

SVD-HBK(2).

Closed / Fat a to Open /d 4 1.lE-003 5.3E-004 4.1E-003 7.8E+000 3/ Plant Energized /5purious /h  ? 8.8E-007 6.8E-007 2.2E-006 3.2E+000 X-PRA.

De-energize l Unsp./ Spurious /h 1 7.3E-007 4.2E-007 2.4E-006 5.6E+000 X-PRA.

De-energize Unsp./ Spurious Operation /h 1 3.8E-007 3.3E-007 7.EE-001 2.3E+000 SVD-HBK.

Closed / Spurious Open /h 6 3.lE-007 3.lE-007 4.6E-007 1.5E+000 4/Both

- 1 TABLE 6. GENERIC FAILURE RATES F@R ELECTRICAL /INSTRUMENTATI@N COMPONENTS WITH VARIOUS DESIGNS AND NORMAL STATES (CONTINUED) ha.

Normal State b I Rec. Mean Median DP

Bound I" "e Factor Source d

Component Type and Design, and rallure Mode U Conductsrs Bus Er.ergtred/ Falls to /h 5 1.lE-007 8.2E-008 2.7E-007 3.2E+000 4/ Plant Operate Computation Modules Av;ragers Unsp./Fa tis to Operate /h I !.1E-005 1.CE-005 2.2E-005 2 2E+000 IEEE-500.

Converters Unsp./f alls to Operate /h 2 1.2E-006 8.7E-007 3.4E-006 3.9E+000 IEEE-500.

9tfferentiators Unsp./Fatis to Operate /h 1 2.0E-006 1.3E-006 6.lE-005 4.9E+000 IEEE-500.

Function Generators Unsp./.~ t t is to Operate /h 1 4.7E-006 3.3E-005 1.3E-005 3.9E+000 IEEE-500.

Sursrers Unsp /Fatis to Operate /h 1 1.2E-006 8.3E-007 3.2E-006 3.9E+000 IEEE-500.

l Cont rollers /Regu lators.

Non-mechanica l 1 E lectronic. Analog unso./ Fails to Operate /h 1 2.4E-006 1.2E-006 8.3E-006 7.0E+000 IEEE-500.

Design Not Specified Energtred/ Falls to /h 1 4. 6E -006 2.0E-006 1.7E-005 8.3E+000 IEEE-500.

Operate Electrical Function Items Energtred/Fatis to /h 1 1.3E-006 7.5E-007 4.2E-006 5.6E+000 NEE 3-PRA.

y Panel Operate Electrical Piece Parts 3 9E-006 3.0E-006 1.0E-005 3.2E+000 NEE 3-PRA.

Diode /Dxttfter Energized / Fails to /n 1 Operate Fuse Fuse. Power Unsp,/ Spurious Operation ih 1 1.3E-007 1.2E-007 2.3E-007 1.9E+000 RAC-85.

Generator, with Diesel Engine Drtver Generator Standby /Fatls to Run /h 5 4.0E-003 2.6E-003 1.2E-0-02 4.EE+000 5/ Plant Stenddy/ Fails to Run /d 70 1.0E-002 7.5E-003 2.6E-002 3.4E+000 MSAC-108.

Standby / Fails to Start /d 7% 5.lE-003 3.7E-003 1.4E-002 3.7E+000 t/ Plant Destg, Not Spectfted Standby / Falls to Start /d I 7.9E-003 7.7E-003 1.lE-002 1.5E+000 SVD-HBK.

4 __ __

TABLE 6. GENERIC FAILURE RATES FOR ELECTRICAL / INSTRUMENTATION COMPONENTS WITH VARIOUS DESIGNS AND NORMAL STATES (CONTINUED)

No.

Normal State b of Upper Error c d Component Type and Design, and Failure Mode U Rec. Mean Median Bound Factor Source Generator, with Hydro Turbine Driver Generator Standby / Fails to Start /d 1 3.lE-003 3.0E-003 4.7E-003 1.6E+000 NEE 3-PRA.

Generator. with Motcr Driver Convtrtor Unsp./ Fails to Run /h 1 2.2E-005 2.1E-005 3.2E-005 1.5E+000 SWD-HBK.

( Gererator, with Gas Turbine Driver Generator Standby / Falls to Run /h 1 2. 6E - O')4 2.0E-004 6.6E-004 3.2E+000 MNSI-PRA.

Stanct.y/ Fails to Start /d t 3.5E-002 3.4E-002 5.0E-002 1.5E+000 MNSI-PRA.

Heaters Design Not Specified Unsp./ Fails to Operate /h I 8.9E-006 7.8E-006 1.8E-005 2.3E+000 RAC-85.

Power Electronics (Solid-state)

Frequency Converter Energtred/ Falls to /h 1 4. 3 E- 006 4.1E-006 7.lE-006 1.7E+000 SVD-HBK.

Operate Energized / Fat h to /h 13 2.9E-005 6.2E-006 1.lE-004 1.8E+001 8/rsth g Inverter (not tnotor ctrir.)

Operate Energized / Falls to /d 5 7.8E-003 4.5E-003 2.5E-002 5.5E+000 X-PRA.

Operate Rectifter,Ctrid Oper./ Falls to Operate /h 1 1.5E-006 1.4E-006 2.EE-006 1.9E+000 SWD-HBK.

w/SCRs;encapt exciters I. Mir Design Not Srecified Energfred/ Falls to /h 1 3.lE-006 2.9E-006 5.7E-006 2.0E+000 IEEE-500.

Operate Relays, Control E lect romechanical Unsp./ Falls to Operate /h 2 3.5E-bO7 3.4E-007 5.lE-007 1.5E+000 RAC-85.

Reed Unsp./ Falls to Cperate /h 2 1.4E-006 1.4E-006 2.3E-006 1.7E+000 RAC-85.

Design hot Specified Open/ Falls to Close /d 1 4.lE-006 3.9E-006 6.2E-006 1.6E+000 IEEE-500.

Unsp./ Falls to Operate /h 3 1.5E-Ou6 1.0E-006 4.7E-006 d.7E+000 RAC-85.

Closed / Falls to Open /d 1 3.2E-006 3.lE-006 4.9E-006 1.6E+000 IEEE-500.

Nc n cal Unsp./ Falls to Operate /h 3 1.lE 006 1.0E-006 1.9E-006 1.9E+000 RAC-85.

I

' .

*  ;' ,. , j ,b a ! !

)

d 1 .

e .() . .

0 0

0. 0 0

0. 0

0. .

c 0340 0 . .

r 00 0 0 K. x. K K K0 0 0 0 0 0 0 0 u 5555 5 B i 8 B B5 5 5 5 5 5 5 5 o - 8 - - H H H H H - - - - - - - -

S EE - E E - - - - E E E E E E E E EECE E D D D O DE E E E E E E E _

EEAE E V V V V VE E E E E E E E I!RI I S S S S SI I I I  ! I I I ro "r 00 00 00 0

0 0

1 0

0 0

0 0

0 0

0 0

0 0

0 0

0 01 00 00 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

1 0

0 0

0 0

ot + + + + + + + + ++ + + + + + + +

N rc EE E E E E E E EE E E E E E E E ra 96 8 8 4 3

.8 E. 8. l. 4 8 7 9 7 1 4 O ) EF 31 3 1 3 3 1 5 12 4 6 3 3 3 1 3

- I D

- T E 57 6 5 4 6 6 4 56 7 7 1 1 7 6 7 A u rd 00 00 . 0 0 0 0

0 0

0 0

0 0

0 0

00 00 0

0 0

0 0

0 0

0 0

0 0

0 0

0 T h n - - - - - - - - - - - - - - - - -

N I u EE .E E E E E E EE E E E E E E E o 29 0 3 6 3 55 2 7 4 7 5 4 E T %B 18 2 E.

5 1 2

. 15 1 18 6 4 7 4 8 B.

8 9 MN U O 67 7 6 4 7 7 5 77 7 8 7 7 7 7 7 R C n 00 00 0

0 0

0 0

0 0

0 0

0 0

0 00 00 0

0 0

0 0

0 0

0 0

0 0

0 0

0 T( ia - - - - - - - - - - - - - - -

- S d e

EE 07 E

3 E

0 E

1 E

9 6

6 E

3 EE 60 E

4 E

0 E

0 E

2 E

3 E

2 E

- N S M S.

35 5 3 2 7 8 2 84 1 7 2 1 2 8 2 I E

/ T 67 7 5 4 6 7 5 76 1 7 7 1 7 6 7 L A .na 00 00 0

0 0

0 0

0 0

0 0

0 0

0 00 00 0

0 0

0 0

0 0

0 0

0 0

0 0

0 A T e EE E E

E E

E E

EE E

E E

E E

E E

C S M 29 4 5 8 0 2 0 22 1 4 7 7 2 3 7 I 45 7 1 2 1 9 4 92 2 1 2 1 3 2 3 R L T A .

C M .

of e c 14 1 2 5 0 1

1 1 11 1 1 1 1 1 1 1 E R N R L O E N dh d h d h h d hh d h n . d h h d U // / / / / / / /t / / / ./ / / -/

. R D n O N o e i F A b e

t a n t

a d er e r S S eo se op O p o o s u

o o t eo pt t o s u

o ot s u

o o tM t t t t t t

. E N a l O o O s s io s s io s s T G t e C o s ls i s s r .

sr o t l r i l sl l l 1 r l i A I u ot i u t l ui i u i 1 u l t s a la p a a oa a p a a p a a R S ll ai ma t 's l F

/ / / o F S n r F tr/

F F S nF

/o/

F

/ / o S n F F E rF lsll l

a dededi

/ /

dedeudededidededi

/ / /

dede E D o l a F et et et et et pet et et et et et et et Nd aF / zarara zazaSrarazazazaza zaza R n Fnp /. d rt rt r igegege i r i r /. gegegegegege t

rt riririri s i rt r gege U S a / e rprprp gege rpr prr prprcr prprp r pr p s

L U es o eOeOeO eOeOeeOeOeOeOeOeO eOeO pn l n n r v - n pn n n n n n n n I O OU C E E E E E OE E E E E~ E E E A I F R A

• n C V g )

I i ) s s d r R H e o n E T D 'tned s n i o

N I d oi e t n cf S a e EW a (i c

/

s t

n t

a G e ee t e t S p y iS vp n e w s-T T t m r e id

. N ct e t r t eo l s u l oe 6 E n th E n e t N e o s/s I r a Sl rn rsr u r ,b E O P g ero ,t s e a L P 3 M C v ,is se yD a

cos ut n dce se5 hi si em cL P s

e r

T p

m e

st es hi cB A O l nt ae i t

i t

T C R e'

T(

rD S

w S

w

~

_ .' :4 .' l { _!j  ; ' ]ii' i~ i' :j ,4 ;i4 '

I l

TABLE 6. GENERIC FAILURE RATES Foa ELECTRICAL / INSTRUMENTATION COMPONENTS 14ITH VARIOUS DESIGNS AND NORMAL STATES (CONTINUED)

No.

hormal State b f Uppu Error e d Component Type and Design, and failure Mode U R ec . Mean K dian Bound Factor' Source Switches, Solid-state (cont'd)

Energized /Spurtous /h 1 5.8E-001 2.lE-007 2.2E-006 1.0E+001 IEEE-500.

Operation Transmitters fransmitters Energired/ Falls to /h 3 5.4E-007 3.6E-007 1.6E-006 ?.4E+000 IEEE-500.

Operate Oper./ Fails to Operate /h 7 2.2E-006 2.2C-006 3.2E-006 1.5E+000 SVD-HBK.

Energeted/ Spurious /h 3 1.lE-005 2.0E-006 4.0E-005 2.0E+001 IEEE-500.

Operation Trcasformers, Control and Instrumentation Otfferenttal/ Regulating Energized / Fails to /h . 1.0E-005 6.GE-006 3.4E-005 5.6E+000 NEE 3-PRA.

Operate voltage-regulating Oper./ Falls to Operate /h 3 1.lE-006 1.0E-006 I.0E-006 2.lE+000 SWD-98K.

Design Not Specified Energized / Falls to /b 1 6.6E-007 3.lE-007 2.3E-006 7.5E+000 IEEE-500.

Operate 4

as Transformers, Powe Energized / Falls to /h 4 9.lE-007 7.1E-007 2.3E-006 3.2E+000 N"i3-PRA(3),

Design Not Specified Z15-FRA(1).

Operate Othat Energsted/ Falls to /h 1 2.7E-006 1 EE-006 8.8E-006 5.6E+000 IPi-PRA.

Operate

a. " General (preaggregated)" in the component design column indicates that the corresponding data were classified as pre-aggregated; i.e., as reflecting a mixture of the component designs. NUCLARR records from sources such as IEEE-500 that supply data combined across component designs contribute to these entries.

b. Alt., normal state is alternating; Unsp., Unspecified normal state; Oper., operating or running normal state.

c. Adjusted to equal 1.5 If the NUCLARR-calculated error factor is less than 1.5 (see Section 5.3).

d. Data source acronyms from Appendix A if

• 3 sources are involved; otherwise, the number of sources followed by " plants" f or operational data f rom corrercial Tuclear power plants, " generic" for generic data f rom other sources, or "both "

l

7. QUALITY OF DATA IN THE HANDBOOK Although collected in 1987, the data in this handbook may be useful as a source of generic component failuru rates for components that are typically found in risk assessments. These data offer some advantages over other generic failure data sources.

Past generic failure data sets developed for use in USNRC-sponsored risk assessments have been based largely on expert judgement. For example, the National Reliability Evaluation Program (NREP) developed a generic data set by assembling data experts from across the nuclear powcr plant industry and obtaining their concensus on values to be used for screening purposes in ,.

PRAs.5 More recently, the Risk Management Integration and Evaluation Program (RMIEP) has developed a data base for the La Salle PRA by using judgement in selecting from past generic data sources.6 The same is true for much of the generic data used in the NUREG-ll50 Reactor Risk Document studies;7 a major source of generic data for this series of risk evaluations has been the Interim Reliability Evaluation Program (IREP) 8 Procedures Guide and it contains the NREP data.

This handbook differs from generic component failure data sources used in past USNRC-sponsored risk assessments because this represents the first USNRC-sponsored effort to collect in one place the existing published compo-nent failure rates that are based on U.S. commercial nuclear power plant operational data. Thus, the data herein are based on U.S. commercial nuclear power plant experience. This plant experience serves as a baseline, thus reducing any dependence on expert judgement to obtain generic component failure rates (of course, expert judgement is always required in the process-ing and interpreting of operational events and other plant information to obtain failure rates).

The data tables show the number of data points and data sources contri-buting to each failure rate estimate. This information helps the user assess 47

i the credibility of the data and shows where additional data collection ef-forts would be warranted. Because of the care taken not to duplicate data, rat.. ti.at are backed by many sources represent experience across the nuclear power plant industry and are not likely to be biased by plant-specific data from possibly atypical plants.

Although this handbook is based mostly on plant experience data, other sources have been used. They are only to fill gaps, and all such instances are clearly identified in the enclosed data tables. Furthermore, all such uses are fully traceable. Page numbers are provided in NUCLARR for every data point, and an appendix discusses the general features of each source.

Another feature of the data base is its quality assurance. Automated checks ensure the validity and reasonableness of every entry. In addition, over half of the records for the component failure data handbook have been compared with the original sources after entry into NUCLARR.

Finally, when data from different sources are combined, the aggregation routines use standard techniques to capture the population variability.

Plots allow the user to further assess the compatibility of different sources.

The existence of the NUCLARR data base itself is another asset; one can acquire NUCLARR and do one's own tailoring of the data for specific applica-tions and sensitivity studies.

48

8. REFERENCES

1. O. V. Hester, C. D. Gentillon, and J. R. Fragola, Evaluation of the NRC Ooerational Data Reouirements, EGG-REQ-7361, August 1986.

2a. D. I. Gertman, W. E. Gilmore, W. J. Galyean, M. R. Groh, C. D.

Gentillon, B. G. Gilbert, Egclear Computerized Library for Assessina -

Reactor Reliability (NUCLARR) Volume 1: Summary DescriotigJl, NUREG/CR 4639 (EGG-2458), February 1988.

2b. r.. Beers, 0. J. Call, Nuclear Lomputerized library foi Assessina Reactor Reliability (NUCLARR) Volume 2: Prearammer's Guide, NUREG/CR-4639 (EGG-2458), August 1988.

2c. W. E. Gilmore, D. 1. Gertman, B. G. Gilbert, W. J. Galyean, Nuclear Computerized Library for Assessino Reactor Reliability (NUCLARR)

Volume 3: Data Base Manaaement Guide for Processina Data aEL Rey _isina the Data Manual, NUREG/CR-4639 (EGG-2458), November 1988, 2d. W. E Gilmore, C. D. Gentillon, D. 1. Gertman, G. H. Beers, W. J.

Galye.n, B. G. Gilbert, Euclear Computerized Library for Assessina Reactoi Reliability (NUCLARR) Volume 4: User's Guide, NUREG/CR-4639 (EGG-2418), June 1988.

2e. D. I. Gertman, B. G. Gilbert, W. E. Gilmore, W. J. Galyean, Nuclear Computerized Library for Assessino Reactor Reliability (NUCLARR)

Volume 5: Data Manual, NUREG/CR-4639 (EGG-2458), June 1988.

3. O. V. Hester, S. R. Brown, and C. D. Gentillon, AnnotateL Biblicaraohv of Reliability and Risk Data Sources, NUREG/CR-5050 (EGG-Rrg.7827), March, 1988.

4. J. R. Fragola, P. Appignani, M. Studzke, Component Reliability Parameter System Data Sheets, Science Applications International Corporation, September 1987 (unpublished).

5. A. J. Oswald, C. D. Gentillon, S. D. Matthews, and T. R. Meachum, Generic Data Base for Data and Models Chapter of the National Reliability Evaluation Proaram (NREP) Guide, EGG-EA-5887, June 1982.

6. A. C. Payne, Jr., February 3,1987 correspondence from Sandia Laboratories to Oren Hester at EG&G Idaho, Inc.

7. M. T. Drouin, F. T. Harper, A. L. Camp, Analysis of Core Damaag_

Freauency from Internal Events: Methodoloav Guidelines, Volume 1, NUREG/CR-4550 (SAND 86-2084), September, 1987.

8. D. D. Carlson, Interim Reliability F.v_aluation Program (IREP)

Procedures Guide, NUREG/CR-2728, SAND 82-1100, January 1983.

49

________----,m__.___ - - . _ _ _ _ - _ - - - - - - - _ . - _ . - - _ , , - _ _ . _ . - _ . - _ _ _ - _ . _ _ - - . . , - - - - - - _ _ - - - - - . _ _ _ - - _ . . . _ _ _ _ _ _ - _ - - - - , _ _ - - - - - - - - _ _ - - - - - - . , . ,

APPENDIX A COMPONENT FAILURE DATA SOURCES l

I l

CONTENTS A.l. PRA SOURCES ..................................................... A-3 A.1.1. Big Rock Point 1 (BRPl-PRA) ............................. A-3 A.1.2. Haddam Neck Plant (Conneticut Yankee) (HNPl-PRA) . . . . . . . . A- 4 A.l.3. Mil l s t on e 1 (MN S I - P RA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-5 A.I.4. Indian Point 2 (IPS2-PRA) ............................... A-5 A.1.5. Zion Station (ZIS-PRA) .................................. A-7 A.l.6. Oconee 3 (NEE 3-PRA)...................................... A-8 A.I.7. P l a n t " X " ( X - P RA ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A - 10 A.1.8. P l a n t " Y " ( Y - P RA ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A - 12 A.I.9. European Pl ant ( EUR-PRA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A- 13 A.2. ADDITIONAL DATA SOURCES .................................. ...... A-14 A.2.1. IEEE-Standard 500-1984 (IEEE-500) ....................... A-14 A.2.2. In-Plant Reliability Data System (IPRDS) . . . . . . . . . . . . . . . . A-18 A.2.3. Swedish Reliability Data Book (SWD-HBK) ................. A-20 A.2.4. Reli ability Analysis Center (RAC) . . . . . . . . . . . . . . . . . . . . . . . A-22 A.2.5. Reliability of Emergency Diesel Generators (NSAC-108) ... A-24 A.2.6. Pipe Break Frequency (PBS) .............................. A-26 A.3. REFERENCES ............................. ..... ................. A-28 A.3.1. Gene ral Re f e rence s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A- 28 A.3.2. PRA Sources ........................ ..... . ............ A-28 A.3.3. Additional Sources ..... ........ ....................... A-29 TABLES A-1. Selected Fields in NUCLARR Component Failure Data Base .... .... A-2 A-2. Supplemental References For IEEE-500 Data . . . . . . . . . . . . . . . . . . . . . . A-17 l

A-ii i

i' APPENDIX A COMPONENT FAILURE DATA SOURCES This appendix contains a brief description of each source contributing data to the component failure data (CFD) handbook. The probabilistic risk assessments (PRAs) used as sources are presented first, followed by other types of data sources. Each PRA source description prcvides tabular infor-mation such as the sponsor of the study, the number of records in the cr?c-nent failure handbook aggregates from the source, and an overview of its '

plant-specific data treatment. This is followed by a brief discussion of the_ associated data that addresses their quality and the major source-specific assumptions (if any) made in entering the data into the Nuclear Computerized Library for Assessing Reactor Risk'(NUCLARR), from which the handbook rates are derived. In general, these data sets tre all suitable

- for PRA applications as they. were developed.and structured for that

_ purpose.

Similar information is t,upplied for the additional data sources.

Table A-1 provides a brief description of the major non-numeric ' data fields.used in NUCLARR to describe the component failure data. All the fields are explained fully in Volume 4, Part 3 of the NUCLARR NUREG series.^4 They are defined just briefly here for reference because some of the source descriptions refer to them in explaining how the data were interpreted and entered.into the data bare. Handbook users need an understanding of the

-assumptions made in processing data from each source in order to make in-formed judgements on the use of the data for particular applications.

The title of each subsection below contains the " Study ID" acronym used in the main text data tables. The Integrated Risk Assessment Data .

Acquisition Program (IRADAP) bibliography nf data bases A-2 contains further information on most of these sources.

A-1

l TABLE A-1. SELECTED FIELDS IN NUCLARR COMPONENT FAILURE DATA BASE

, Fie_lt Descriotion Eyent-Definina Attributes Component Type 3-char. component code '(mechanical or electrical /instrum.)

Design 2-char. component design code Failure Mode Code for the undesired change or lack of change of state Normal State Code for normal state of a component during operation Additional Informai. ion Application As many as three attributes for particular component types (such as internal environment for valves)

Failure Origin Origin for failure data (in-depth plant records,. expert judgement, laboratory test reports, other)

Fail . Rec. Type Specific types of records that were sources for the failure information.

Exposure Orig. Origin for exposure data (in-depth plant records, expert judgement, calendar hours, critical hours, other)

Exp. Rec. Typ. Specific types of records that were sources for the denomi-nators of the failure rates Data Coll. Per. Starting and ending years for data collection Incl. Circuit Whether control circuits are included in the component boundary for assessing failures Nuclear Orig. Whether the data come f.om nuclear power plant experience Bayesian Upd. Whether Baysian methods were used to calculate a rate from data and prior experience IRADAP Suit. Whether the data record is to be included in this handbook Safety Grade Whether the data describe safety grade components Primary Fail. Whether only primary failures are included in tha numerator of the failure rate (rather than both primary and secondary failures)

Origin. Fac. Nuclear power plants or groups of nuclear power plants whose experience is aggregated to form a failure rate Severity Whether the failures counted for the failure rate are catestrophic, degraded, or just incipient Degree Code for details on the nature of degraded failures include System Whether the system field lists included systems or excluded systems System Nuclear power plant system (s) associated with the failures Subsys. Train Text field (36 char.) for subsystem or train information Reference Information Document No. A- 3-digit document code followed by the year of publication Detail Ref. A page number or table number Supplemental Ref Up to 20 char. of acronyms for original sources cited by the reference document Additional Comments Comments Additional information that cannot be conveyed in the coded fields A-2

A.1. PRA SOURCES PRA and reliability analysis sources using plant-specific data are described below.

A.1.1. Bio Rock POINT 1 (BRP1-PRA)

Sponsor of study: Consumers Power Company l

l Performing organization: Wood-leaver Associates PRA completion date: - March 1981 NUCLARR qualified record count: 30 Reference number: AP-1 Facility attributes: U. S., commercial, General Electric, BWR.

CFD attributes: Raw data for hourly and per demand rates with confidence limits for the diesel generators ano certain pumps and valves.

Most components characterized by system.

The BRP plant is one of the oldest in the United States, having start-ed commercial operation in December, 1965. Because of its extensive oper-ating experience and the uniqueness of the BRP design, BRP plant-specific data were used whenever possible in the PRA. Plant-specific data sources included plant maintenance orders, control room log books, surveillance tests, LERs, event reports, deviation reports, plant review committee meet-

-ing minutes, and USNRC correspondence. The plant-specific data used spanned the period from 1970 to 1979; data before 1970 do not incluoe maintenance orders or surveillance tests and therefore were excluded.

All BRP plant-specific component failure data were entered into NUCLARR except for a per-demand rate for containment isolation valves fail-ing leak tests. The failure mode for this event wou'd be " failure to A-3

\

m4 _ . - . . _ _ . . . . _ _ . _ _ . _ . -._ . . _ _ _ . . __ _

J close" (i.e., remain closed) but_th'e event was judged not to be catastroph-

= ic. [All " leakage" events are per hour in the NUCLARR since one would -

- never request a component to leak.)

A.1.2. HAnoAM NECK PLANT (CONNECTICUT YANKEE) (HNP1-PRA)

Sponsor of study: Northeast Utilities (NU)

Performing organization: Northeast Utilities .

PRA-completion date: December 1985 NUCLARR qualified record count: 21 Reference number: AP-2 Facility attributes: U. S., commercial, Westinghouse,-PWR.

CFD attributes: Hourly and per demand rates from raw data and means and variances obtained_.through i Bayesian updates using ccnjugate prior .

distributions.

The Connecticut Yankee PRA describes the process used to gather compo-nent failure history, demand history, and run time experience over a 10-year period beginning_in_1974. Included in the process was the retriev-al of' event-specific data from the Baseline Events Analysis Reliability Data Sys4em (BEARDS), a proprietary NU data base that includes failure rd maintena'nce reports. All plant-specific component failure data 'were en-tered into NUCLARR. Data are_given_for pumps -(by system), containment air-recirculation fans, batteries, chargers, inverters, and diesels. The PRA contains-no plant-specific valve data.

The PRA report also presents updated component failure data. The results reflect a Bayesian update of WASH-1400, IEEE-500, and Westinghouse Nuclear Technology Division means and variances, using plant-specific expe-rience where available. Failure-on-demand rates were modeled using beta--

distributed priors. Hourly failure rates were modeled using gamma-distrib-uted priors. Although these updated means and variances are stored in .

NUCLARR, the aggregation methods use just the raw plant-specific opera-

. tional data.

A-4 2

, .-w------m , , , - . , - , - - -

zi 1

1 A.1.3. MILLSTONE 1 (MNSI-PRA)

'Sponssr ofl study: . Northeast Utilities Performing organization: Northeast Utilities >

PRA completion date: July, 1985 NUCLARR-qualified record count: 37-Reference number: AP-3 Facility attributes: U. S., commercial, General Electric, BWR. ,

'CFD attributes: Hourly and per demand rates based on raw data with means and variances obtained through Bayesian updates using conjugate prior distributions.

-The Millstone Unit 1.PRA was. quantified using plant-specific system

-

- and component. failure history,--demand history, and run time experience over

! '3-yoar period of_ commercial operation. The data collection process was

~

tim 1:ar to that of the-HNP1_.PRA, including extracting data from the Base-f

'line Events Analysis Reliability Data System (BEARDS). -Data are present-

<for valves,. diesel and gas turbine generators, breakers, and battery char-

gers. In addition, raw nlant-specific data are provided for pumps from 13 systems for failure to start and from 8 systems for failure to run.

Updated component _ failure data reflect a Bayesian update of-WASH-1400

) means and variances using the plant-specific experience. Rates for Failure I

'on demands were modeled using beta-distributed priors. ~ Hourly failure 1- rates were modeled using gamma-distributed priors. These data are pretent_-

I -in NUCLARRL for archival and display; they are not.used in the aggregation-calculations.

A.1.4. INDIAN-POINT 2 (IPS2-PRA)

Sponsor.of study: Consolidated Edison Company A-5

i e

Performing organizations: Pickard, Lowe & Garrick, Inc.;

Westinghouse Electric Corporation; and Fauske & Associates PRA completion date: Harch, 1982 1

NUCLARR qualified record count: 39 Reference number: AP-4 Facility attributes: U. S., commercial, Westinghouse, PWR.

CFD attributes: Raw data for hourly and per demand rates, with means and variances obtained through Bayesian updates using discrete prior distributions.

The site-specific and generic component failure and service hour data sections of the Indian Point Unit 2 PRA were available for entry of site--

s. specific component failure data. In the PRA, the Licensee Event Reports (LERs) generated after the plant became critical (from May 23, 1973 to December 31,1979) were used as the primary source of component failure data. This was supplemented by operating logs, and component maintenance and testing records, and significant occurrence reports to clarify the LER narratives in some cast s. Command fault; and failures frca problems out-side the component be Jaries were excluded; thus, the " primary failures only" indicator i .. for these data. In addition, the documentation ex-plains that contrul circuits are included within the component boundary for pumps.

System drawings, unit operating statut summaries, and test plans were ,

used to establish the exposure times or number of demands. Exposure times for running equipment were in most cases based on the sum of plant hours in operation and in hot shutdown.

The resulting raw failure informati a is documented in a set of 39 data summaries that describe the types of failures that occurred and the assumptions made for each estimate. For some components, specification of the system was included because of its impact on the data values. For example, motor-driven pumps are distinguished by system for the failure to run failure-mode but not for the failure to start failure mode. Data are A-6

f L

present for pumps, valves, heat exchangers, diesels, fan cooling units, batteries, chargers, breakers, transformers, inverters, automatic transfer devices, and buses. For the spurious close failure mode, manual and motor-operated valves were combined.

Generic data from WASH-1400, IEEE Std 500, and the LER data summaries on valves, pumps, and diesels were combined with plant-specific failure data using discrete prior distributions. Updated failure information based on the resulting discrete posterior distributions is stored in NUCLARR but is not used in the aggregations.

LERs have been criticized as a source of failure data because report-ing criteria differ among plants. However, the analysts for the Indian Point Station argue that the LERs for this station are good sources because the technical specifications required LERs for all significant events that directly affect plant safety or those systems designed to maintain the plant in a safe condition, and the PRA is directly concerned with such events. Furthermore, the LERs at this station are "a source of relatively consistent, reviewed, and generally complete event descriptions" (Reference AP-4, page 1.5-2).

A.1.5. ZION STATION (ZIS-PRA)

Sponsor of study: Commonwealth Edison Company Performing organizations: Pickard, Lowe & Garrick, Inc.;

Westinghouse Electric Corporation; and Fauske & Associates PRA completion date: September, 1981 NUCLARR qualified record count: 40 Reference number: AP-5 Facility attributes: U. S., commercial, Westinghouse, PWR.

CFD attributes: Raw data for hourly and per demand rates, '

with means and variances obtained through Bayesian updates using discrete prior distributions.

A-7

The Zion PRA data base is very similar in form to that of Indian Point. Both PRAs were performed by Picard, Lowe and Garrick, and the ap-proach used for each PRA's data was the same. LERs were the primary fail-ure data source, for the same reasons as stated above for Indian Point 2's PRA. These were supplemented by control room log books, plant maintenance logs, plant maintenance test records, and deviation reports. Demands were in most cases obtained from a study of test procedures and plant drawings, while operating hours were computed from the monthly operating status re-ports. Most of the component failure rates were applicable to all systems, '

but exceptions are noted in some cases. Data are presented for pumps, valves, containment fan cooling units, reactor building fan cooler dampers, heat exchangers, diesel generators, bus feeder and reactor protection sys-tem breakers, metal-enclosed busses, transformers, battery chargers, batter-ies, and inverters. As with the IPS2 PRA, motor-driven pump data were combined across system 'or the " fail to start" failure mode but not for the

" failure to run" mode.

The documentation format for the plant-specific failure data and their use to form updated distributions with means and variances are identical to that of the IPS2 PRA.

x f

A.l.6. Oc0 NEE 3 (NEE 3-PRA)

Sponsor of study: EPRI and Duke Power Company Performing organization: EPRI's Nuclear Safoty Analysis Center (NSAC)

PRA completion date: May, 1981 NUCLARR qualified record count: 54 Reference number: AP-6 Facility attributes: U. S., commercial, Babcock & Wilcox, PWR.

CFD attributes: Raw data for hourly and per demand rates, with means and variances obtained through Bayesian updates.

A-8 1

l

The Oconee Unit 3 PRA centains plant-specific raw data for valves; pumps; reactor building cooling units; isolating diode assemblies; instru-ment inverters; four sizes of transformers; panel boards; low and high voltage busses; buswork; DC, low voltage, and high voltage circuit break-ers; batteries; battery chargers; and hydro-driven generators. Nearly all the data are system-specific except for the valve data. For motor-operated valves that fail to operate, data for the condenser circulating water sys-tem are separately listed. The Oconee data are unique in its detailed treatment of check valves; data for swing check, stop check, and tilting disk check valves are distinguished.

The failure data for these rates are obtained from maintenance work requests supplemented by incidence reports and Licensee Event Reports from the 1975-1980 time period. The work requests provide a complete history of all repairs performed at Oconee. They are not restricted to safety-related systems, they are written during all modes of unit operation, and they are not produced in response to licensing-based criteria. Thus, their use provides completeness, traceability, compatibility with success data, at resulting reduction in data-base uncertainty, Periodic test reports, control room operating logs, and piping and instrumentation diagrams were used to compute the numbers of demands.

Operating hours were based on run hour loas for the motor-driven pumps and cooling units; a review of normal plant operating procedures, system line-ups, and periodic test records provided component service hours for other component / failure mode combinations. An appendix to the PRA contains a data summary table for each failure rate estimate describing its basis.

t For the PRA, the plant-specific data were updated using discrete log-normal prior distributions that were derived from generic data. The updat-ed median, mean, and upper and lower tolerance bounds are in the data tables. These are stored in NUCLARR but only the raw plant-specific data are used in the aggregations.

A-9

1 A.1.7. PLANT "X" (X-PRA)

Sponsor of study: (Proprietary information)

Performing organization: Science Applications International Corp.

(SAIC)

PRA completion date: Data received in 1987 NUCLARR qualified record count: Approximately 130 Reference number: AP-7 Facility attributes: U. S., commercial, Babcock & Wilcox, PWR.

CFD attributes: System-specific raw data for hourly and per demand rates for pumps, valves, the instrument air compressors, and electri-cal equipment Science Applications International Corp. (SAIC) provided handwritten ,

tables describing proprietary PRA data for a U.S. commercial Babcock and Wilcox power plant which it denotes " Plant X." The data are based on a review of the plant's daily operating history covering a time period from 1978 to 1984. Plant work authorizations, LERs, NPRDS records, and NRC Monthly Operating reports were all used as plant-specific input to the data base. Numbers of demands and operating hours were based on a SAIC question-naire filled out by plant personnel for each system, supplemented by plant drawings and tt.e Final Safety Analysis Report (FSAR). Interfacing demands caused by testing and maintenance were considered in the construction of component demand histories.

Tables were received for valves, pumps, instrument air compressors, battery chargers, batteries, inverters, diesel generators, and breakers.

The valve tables provided information on air operated, motor operated, handhweel, hydraulic, pressure relief / safety, and check valves for several systems at Plant X. The check valves are further classified as swing, tilting disk, hand weilded stop check valves, and motor operated stop check valves. Failure, demand, and operating hour counts are provided on a per pump basis for the pumps in severai plant systems. Tables for the air A-10

compressors and the electrical equipment other than breakers are also on a component-specific basis. Breakers are grouped in to 4160 VAC pump break-ers, 480 VAC valve breakers, and VDC breakers. For NUCLARR, data for spe- k cific components having the same normal operating state were pooled for each system.

The tables do not provide detailed failure mode information. Each failure is classified as Jemand-related or time-related; counts of each are given. Beccuse the data were collected for use in a PRA, all failures were classed as failures that could disable a train of a system so that it could not perform its intended function. That is, the failures were classed in NUCLARR as having a "castestrophic" severity. Unless otherwise known (see the next paragraph), failure modes were assigned as follows:

Time-Related Demand-Related Comnonent Failure Failure Check valves without IL FT0 operators Other valves, break- FTO(/h) FTO(/d) ers, inverters Pumps, compressors FTR FTS Batteries, chargers FT0 --

Recall that the fail to operate hourly failure mode (FT0/h) refers to general operational failure and is not a detailed failure mode.

Diesel generator data were omitted in order to avoid duplication with Reference AA-8, described in Section A.2.5, which has plant-specific diesel generator data from all plants.

Data sheets from SAIC's Component Reliability Parameter System (CRPS) were also provided to the INEL; some of these contained data from Plant X.

The CRPS is a data base system developed by SAIC in 1987 to allow the dis-play and aggregation of component failure data. Fuur Plant X entries in NUCLARR came from these sheets. These entries reflect combined data over the entire plant, including systems that were not included in the handwrit-ten sheets. For these data, NUCLARR's Detailed Reference field includes A-ll I

- - - _ - - 1

=

f the' component and Lfailure mode codes that identify the individual data:

-sheets in the CRPS. ~ The CRPS sheets were also used to clarify the. failure mode descriptions for-over 60 of the NUCLARR records from Plant X.-

1 A.1.8. PLAaT "Y" (Y-PRA)  ;

Sponsor of. study: (Proprietary information)

Performing organization: Science Applications International Corr,.

(SAIC)

IPRA completion date: Data received in 1987 NUCLARR qualified record. count: Approximately 70 f Refere oce number: AP-8 Facility attributes: A U. S. commercial BWR CFDl attributes: Raw data for hourly and per demand rates for pumps and valves, by system Handwritten data tables summarizing a study of pump and valve records '

for a-U.S. commercial General Electric plant having two units were-received from SAIC. -The sheets are based on plant records, equivalent to mainte-nance work authorizations, for the period from 1980 to 1985. Although the

. data -are proprietary when identified- by the plant name, they can be re-

-leased for use in a-gene *ic data base with the plant ideatified simply as-

" Plant Y."

The data tables had the same format as that described for Plant X, above, and were treated in the same manner. . The tables describe the emer-gency service ' water, high pressure service water, residual heat removal, reactor core isolation cooling,'and core spray systems. Approximately half  ;

of the data were also described in CRPS sheets; these provided additional clarification for failure modes.

A-12

A.1.9. EUROPEt,W PLANT (EUR-PRA)

Sponsor of stedy: (Proprietary information)

Performing organization: Science Applications international Corp.

(SAIC)

PRA completion date: Data received in 1987 NUCLARR qualified record count: 13 Reference number: AP 9 Facility attributes: A European commercial nuclear power plant CFD attributes: Raw data for hourly and per demand rates for valves and pumps SAIC has performed a number uf plent-specific operatic,nal data assess-ments supporting PRAs. The data analysis .esults are proprietary if specif-ic plant names are provided. However, such data are valuable for use in generic data bases even if the. plant name is not provided. Such is the case for data from a European commercial nuclear power plant.

The data, for pumps and valves, were received at the INEL in the form of entries in CRPS data sheets. The CRPS is a data base system devel:yed by SAIC in 1987 to allow the display and aggragation of component failure daa This transmittal was inclucled in an August 31, 1987 first draft of the Component Failure Data Handbook fer NRC/AE0D under the IRADAP. The page numbers in tne d&ta set for NUCLARR refer to the page numbers that were applied to th(. SAIC data sheets in that draft. NUCLARR's Detailed Rcference field also includes the component and failure mode codes that f aentify individual clata sheets in the CRPS. Although only selected f ail-ure rates were available, these were entered into NUCLARR for use in thi-handbook because plant-specific maintenance records, the equivalent of work authorization forms, were processed to derive them.

A-13

l A.2. ADCITIONAL DATA SOURCES "

Additional data sources are listed below. Although th< a sources were not developed for specific risk assessments, use in risk assessments was a primary purpose for developing most of them.

A.2.1. IEEE-STANDARD 500-1984 (IEEE-500)

Sponsor of study: The Institute of Electrical and Electrca-ic Engineers, Inc. (IEEE)

Perforn.ing organization: Fluor Engineers, Inc.

Completion date: 1983 NUCLARR qualified record count: Approximately 70 Reference number: AA-1 Facility attributes: Primarily V. S., commercial, nuclear.

CFD attributes: Hourly and per demand rates with upper and lower bounds for various failure modes for electrical, instrumentation, and mechanical components in nuclear power plants. Where available, repair times are also included.

The IEEE Standard 500 document was originally issued in 1977 as a manual of reliability data on electrical, electronic, and sensing compo-nents. Since that time, data have been gathered on mechanical equipment reliability, making the 1984 version of IEEE Std 500 a more comprehensive j

, work that includes data on over 1,000 events. The primary source for the reliability data presented in IEEE Std 500 1977 was the judgment of nearly 200 experts nn component reliability in the nuclear power field. The Del-l phi method was used to compile and refine the expert opinion data. Tbi t information is included in IEEE Std 500-1984 (IEEE-500) for the nonmechani-cal components, but it is supplemented by data from 24 additional sources.

A-14 I

The Std 500 1984 reliability data set for electrical and electronic components is useful for PRA applications because this is one of the few sources that exists for many of the components it covers. However, users should be aware of the original sources used and the f act that some of them may overlap (a single failure may be reflected both in LERs and IPRDS, for example) . There is also overlap between design-based groupings and application based groupings of the data. Ior example, the pump data are primarily structured at a high level by the pump's internal method of opera-tion (centrifugal, positive displacement) rather than by type of driver (turbine, diesel), but data sheets with pump and driver both are included and some data sheets have this information at a system specific level. The amount of actual operational data supporting the rate estimates should be evaluated from a knowledge of the sources used by IEEE. For this handbook, the catestrophic mode failure data from lEEE-500 are used as a source only if there is a lack of several plant-specific data sources.

In encoding selected data sheets from IEEE-500, informatior on IEEE-500's sources has been made readily available. The " Supplemental Reference" field is provided it capture the original sources of referenced data. Table A-2 shows the codes that have been used for supplemental refer-ences for IEEE-500 data.

All the data in IEEE-500 are in the form of rates; no raw data (num-bers of components and f ailures) are supplied. Furthermore, the upper and lower bounds included in the document are weighted averages of upper and lower bounds in the original source documents and as such have no specified coverage probability. In ratering the data for NUCLARR, the " recommended rates" have been treated as medians and bounds have been conservatively treated as 20% and 80% tolerance bounds.

Since the IEEE-500 data have been combined from many sources, the assignment of values for fields in NUCLARR such as data collection period and nuclear origin requires further explanation. The data collection A-15

l TABLE A-2. SUPPLEMENTAL REFERENCES ron IEEE-500 Ident i fier. Reference 500 ANSI /IEEE Std 500-1977, IEEE Guide to the Collec. ion and Pre-sentation of Electrical, Electronic, and Sensing Component Reliability Data for Nuclear-Power Generating Stations.

493 ANSI /IEEE Std 493-1980, IEEE Recommended Practice for Design of Reliable Industrial and Commercial Power Systems (Gold Book).

JdB The following five references:

Interim EGG-EA-5816, April 1982, Data Summaries of Licensee Event Reports of Valves at US Commercial Nuclear Power Plants.

NUREG/CR 1362, March 1980, Data Summaries of Licensee Event Reports of Diesel Generators at US Commercial Nuclear Power Plants.

NUREG/CR 1331, EGG EA 5079 Feb 1980, Data Summaries of Licensee Event Reports of Control Rods and Drive Mechanisms at US Commercial Nuclear Power Plants.

NUREG/CR-1730, EGG-EA-5188 Sept 1980, Data Summaries of Licensee Event Reports of Primary Containment Penetrations at US Commercial Nuclear Power Plants.

NUREG/CR-1205, Jan 1980, Data Summaries of Licensee Event Reports of Pumps at US Commercial Nuclear Power Plants.

RMDB Corps of Engineers (HND) R/M Data Base, Ground Stationary Equipment, Rep No 16,04/12/73.

RAC Nonelectric Parts Reliability Data (NPRD-2) Summer 1981, Reli-ability Analysis Center, Rome Air Development Center, Griffiss Air Force Base, NY 13441.

N06666 NUREG-06666, April 1981, A Probabilistic Safety Analysis of DC Power Supply Requirements for Nuclear Power Plants.

NERS80-02 Failure Data Estimates for Diesel Generators in US Commercial Nuclear Power Plants. NERS 80-02 (CRBR Project) July 1980.

E46 Nuclear Unit Productivity Analysis (Special EPRI Report No 46)

Aug 1976 (Nonredundant Equipment Only).

EA 2205 EPRI Rep No AP 2205, Feb 1982, Component Failure and Repair Data; Gasification-Combined-Cycle Power Generation Units.

A-16 I

TABLE A-2. SUPPLEMENTAL REFERENCES FOR IEEE-500, c0NTINUED Identifier Reference EA 2071 EPRI Rep No AP 2071, Oct 1981, Component failure and Repair Data for Coal Fired Power Units.

EA 2321 EPRI Rep No AP 2321, March 1982. High Reliability Gas Turbine tombined-Cycle Development Program: Phase 11.

EJ R&D Status Report-Advanced Power Systems Division, EPRI Jour-nal, March 1982, p 37, NPRD NUREG/CR-2232, Nuclear Plant Reliability Data System (NPRDS) 1980 Annual Reports of Cumulative System and Component Reli-ability.

IPRD NUREG/CR 2886, May 1982, in-Plant Reliability Data System (IPRDS) Interim Draft Report on the Reliability Characteris-tics of Selected Pumps in Four Nuclear Plants.

CIGRE The First International Enquiry on Circuit Breakers Failure and Defects in Service. Extract from Electra No 79 (CIGRE)

(1974-1977).

SHIM SHIHlZU, Use of Experience Data of Nuclear Power Plant Diesel-Generator Emergency Power Systems in Reliability Analy-sis. Presented at the Annual West Coast Reliability Sympo-stum, Feb 26, 1977.

UA Utility "A" AEA Architectural Engineering Company "A" AEB Architectural Engineering Company "B" PWR10 Pressurized Water Reactor - Plant No 10 GTA Gas Turbine Manufacturer "A" period for a data sheet was taken to be from the earliest date present among the sources to the latest date. In some cases, the ending date was based on the source's publication date. For the nuclear origin field, a value (yes or no) was assigned if it was known to be applicable for all the sources contributing to the aggregated value; wnen the analyst was in doubt,'the field was left blank. A similar situation applies for the data A-17

origin field; for example, a " plant" designation was used for data from IPRDS or from the "PWR 10" reference. When the same failure origin type did not apply to all the sources, "other" was encoded in this field and "See Sup. Ref" was included in the comment field to refer the analyst to the supplemental reference. For the failure record type field, the record types known by the INEL analysis team to be used in the associated root sources were assigned. For example, LERs are a failure record type for failure rates from the LER Summary series of NUREG reports. Where the INEL encoding team lacked such detailed knowledge of the sources, the fields were left blank. The exposure origin and record type are blank for all the IEEE 500 records because no raw exposure data are provided.

A.2.7 IN-PLANT RELIABILITY DATA SYSTEM (IPRDS)

Sponsor of study: U. S. NRC Office of Nuclear Regulatory Research Performing organization: Oak Ridge National Laboratory Completion date: March, 1982 NUCLARR qualified record count: Approximately 120 Reference number: AA-2 through AA-5 Facility attributes: U. S., commercial, nuclear power plants; five PWR units and four BWR units.

CFD attributes: Plant and system-specific hourly and per  ;

demand rates for pumps, valves, and main electrical distribution system components based on raw data, with confidence bounds based on the binomial and Poisson distri-butions. Repair action records also exist in some cases.

The main objective of the In-Plant Reliability Data System (IPRDS) is to develop a comprehensive and component-specific data base for PRA and other component reliability-related statistical analysis.AA-r To accomplish this objective, in the 1978-1981 time period data base personnel visited selected plaats and copied all the plant maintenance work A-18 l

I requests. They also gathered plant equipment lists and plant drawings and in some cases interviewed plant personnel for information on component popu-lations and duty cycles. An IEEE committee participated in selecting the specific plants; three PWR stations and 90 BWR stations were visited. Be-cause some plants participated on the condition that their identity would not be divulged, the plants are denoted by numbers.

The maintenance records describe preventive .s well as corrective maintenance. IPRDS personnel screened the maintenance records for selected components to separate out the cases of corrective maintenance occurring since the start of commercial operatio1. These records were reviewed to determine such things as failure modes, severity, and, if possible, failure cause. The data from these reports we e encoded into a computerized data base. By having a team of data base enperts process all the data, problems with differences in the interpretation of reporting requirements were mini-mized. Usir9 in-plant records providet assurance that even disabling incip-ient hardware problems are included. lhus, the resulting data base has consistency and depth.

The IPRDS data for NUCLARR and the component failure data handbook have been entered from the three reports that have been issued containing IPRDS failure data. These reports describe pumps,"'3 valves, " and major electrical components."'S The electrical component report pre-sents data for diesel generators, batteries, battery chargers, and inver-ters in nuclear power plant essential ac power distribution systems. The table below shows the time span and numliers of components and failure records that were analyzed to derive the failure rates:

Unit Plant Numbers Paulation failuret Years Pumps 1 through 4 1468 3998 27 Valves 1 and 4 24825 5712 24 Electrical 1 through 5 213 698 33 A-19

(

l The IPRDS data that were described as catestrophic in severity were encoded for NUCLARR.* Because the primary data source is maintenance {

work requests, the " primary failures only" indicator was set to show both primary and secondary failures; all of these would result in corrective maintenance. The component boundaries are described in the IPRDS reports and all include control circuits. The data time frames werc estimated for each of the five stations (plants) submitting data from the known years of commercial operation per component and estimated data collection dates as indicated in the following table.

Plant No. of Years of Commercial Data Data I Plant No. fl0 lyne_ Units. Operation (per uniti S, art Date End Date 1 IPl PWR 1 5.1 74 79 2 IP2 PWR 1 1.3 79 80 3 IP3 BWR 3 3.1 77 80 4 IP4 BWR 1 6.2 75 80 5 IPS PWR 5 3.8 77 81 The IPRDS data base is very helpful for corrective maintenance frequencies and repair times. However, further information on component duty cycles and use would improve the demand-based failure rate estimates contained in the reports, in addition, because the data base was developed from relatively few nuclear power stations, caution should be used for other than generic applications. For the component failure data handbook, these data am combined with other plant-specific data to cbtain failure rates.

A.2.3 SWEDISH RELIABILITY DATA BOOK (SWD-HBK)

Sponsor of study: Nuclear Safety Board of the Swedish Utilties (RKS) and the Swedish Nuclear Power inspectorate (SKI)

a. Diesel generator cata were reitted in ordsr to avoid ducittation with other sources (such as Ref erence AA-8, described in Section A.2.5, which has plant-specif ic data f rra all plants). Also, to avoid high f ailure rate estimates based on little data, a rule was adopted for entry of f ailure to stirt data for pumps. Whenever 0 failures were observed, the data were not entered unless the number of demands enceeded 303.

A-20

(

l l

Performing organization: ASEA-ATOM and Studsvik Energiteknik AB Completion date: 1985 NUCLARR qualified record count: Approximately 75 Reference number: AA-6 ,

Facility attributes: 7 Swedish PWR units (from 4 stations) and 1 Swedish BWR unit CFD attributes: Numbers of failures and either demands or operating hours for pumps, valves, con-trol rods and control rod drives, electri-cal equipment, and instruments, with means and upper bounds based on fitting gamma distributions The Swedish Reliability Data Book was developed to provide improved failure data for reliability calculations. It is based primarily on evalua-tions of failure reports in the Swedish Thermal Power Reliability Data Sys- ,

tem (ATV), a data collection system jointly established by the Swedish util-ities and maintained and managed by the Swedish State Power Board at Stockholm, Sweden. LER reported to the Swedish Nuclear Power Inspectorate and Information provided by the operation and mnaintenance staff of each plant also contributed to the results presented in the Data Book.

The data cover operating periods starting as early as 1974 and con-tinuing through to the end of 1982 for the BWR units and through the middle of 1981 for the BWR unit, Ringhals 2. The initial startup periods for cach reactor _are not included. The data pertain primarily to safety related systems, since these systems are tested regularly in accordance with the technical specifications and have the most clear reporting requirements.

Demand counts were obtained from a knowledge of testing requirements; the impact of events such as scrams was considered in some cases. Time clocks have been installed on pumps to record actual exposure time. For other components, the operational profiles of the plants were used to calculate exposure times.

A-21

Data for pumps, valves, control rods and driver, sensors, and electri-cal equipment were entered in NUCLARR. No records from the Data Book were pooled; instead, the application field was used (where applicable) to cap-

, ture additional differentiating attributes for the data such as voltage levels for electrical components. The Data Book presents data combined across systems and, for the BWR plants, across units. The mean values estimated from the gamma distribution fit to the data were entered into NUCLARR, although they are not used in the calculations for this handbook.

The Data Book contains a discussion of failure modes for each type of component. It distinguishes " critical failures" and " degraded / incipient failures." Only data for critical failures (catestrophic failures) were entered into NUCLARR. The Data Book also contains a discussion of the failure events that are common for each cc por. nt type and schematics that show typical component boundaries.

A.2.4 RELIABILITY ANALYSIS CENTER (RAC)

Spensor of study: U. S. Department of Defense Performing organization: IIT Research Institute, contractor for the Reliability Analysis Center managed by Rome Air Development Center, Griffiss AFB, NY 13441-5700 Completion date: Fall, 1985 n

NUCLARR qualified record count: Approximately 20 Reference number: AA-7 Facility attributes: U. S. commercial and military installations CFD attributes: Numbers of failures and numbers of thou-sands of operating hours for more than 380 major nonelectronic part types, cate-gorized by component design, environment, and whether the source was military or commercial,

, A-22 1

l The Reliability Analysis Center (RAC) is charged with the collection, analysis, and dissemination of reliebility information pertaining primarily to parts used in electronic systems; however, data are also collected for selected nonelectronic parts used in military, space, and commercial appli-cations. A special system is in place for the collection and analysis of data on operating and planned military systems and equipment. Data are also collected on a continuous basis from such sources as testing labora-tories, device or equipment manufacturers, government laboratories, and equiprant users. Automatic distribution lists, volunatary data submittal, and field failure reporting systems supplement an intensive data solicita-tion program. A three-year period of strong emphasis on data acquisition l provided more data for the 1985 edition of Nonelectronic Parts Reliability l QLt3.*

To support nuclear power plant risk assessment, data for fans, compres-sors, heaters, relays, and fuses were selected for entry into NUCLARR.

. These are areas for which relatively little specific nuclear power plant operational data exist. In all cases, the entered data were from ordinary commercial applications with a " ground fixed" environment or from military applications with a " ship sheltered" or " submarine" environment. These terms are defined as follows:

Ground fixed--conditions less than ideal to include installation in permanent racks with adequate cooling air, maintenance by military personnel and possible installation in unheated buildings.

Ship sheltered--surface conditions similar to Ground Fixed but subject to occasional high shock and vibration.

a, This coreent in the report was the bas ts for assessing the data collection pariod as f rom 19E2 t, 1984. Some older data may also have been used. The report does not state specific data sources i- A-23

(

Submarine--conditions normal to operation aboard a submerged vessel.

Temperature and humidity controlled.

The RAC failure rates are not differentiated with regard to failure modes. A separate section of the report provides a failure mode breakdown for most of the components. Factors (percentages) from these tables could be used to break the overall failure rates into rates for particular fail-ure modes. However, the environments for those failures are not shown, and thus they may not be applicable for particular sets of data. These factors were not used in assessing the data for this Handbook. Thus, failure mode designations for data from RAC must be interpreted broadly.

A.2.5. RELIABILITY OF EMERGENCY DIESEL GENERATORS (NSAC-108)

Sponsor of study: Electric Power Research Institute (EPRI)

Performing organization: Nuclear Safety Analysis Center (NSAC)

Completion date: September, 1986 NUCLARR qualified record cot.nt: Approximately 250 Reference number: AA-8 Facility attributes: U. S., commercial, LWRs from 52 sites.

CFD attributes: Numbers of failures and demands for emergency diesel generator failure to start and for failure to load and run for the required duration.

EPRI's NSAC surveyed the U. S. nuclear power plant industry in order to determine emergency diesel generator (DG) reliability for the years from 1983 to 1985. For each of 154 diesel generators, reliability data are pro-vided for each of these three years. Both testing and unplanned demands and associated failures were included. However, the unplanned demands re-present only about 2 percent of the total and have very few failures (for the 75 units and three years in the study, there were 431 start demands and 223 load-run demands with only 2 start failures and 4 load-run failures).

Because of this sparsity, only the total demands and associated failures were entered into the data base. These are present on a unit-specific A-24

(

_m . _ _ - _ _ . _ _ _ . _ . _ . . . ._ .- _ . - . _ . -

)

basis. Note that starts and load-runs that were conducted as an aid to i trouble-shooting and maintentoce were not included in the NGAC study since they do not provide an indication of the DG's ability to fulfill its mission in a real emergency. 1 Two failure modes were considered in the NSAC study. The first is l failure to start; in addition to u.iplanned demands this includes 00th fast start tests (less than 10 seconds) and slow start tests (where '.he require-ment to start allows approximately 5 minutes from the first sthrt attempt to stable performance at rated frequency and soltage). Since fast start .

testing causes stress and wear on the diesel and fast starts are needed only for a simultaneous blackout and loss of coolant accident (o ich is expected to occur only very rarely if ever), fast start testing is done very infrequently. There were no observed failures during the few tests that were classified this way, and the report authors reconnend combining these tests into a single category.

The failure to run mode includes ali failures occurring from the time when load was applied to the DG until the diesel is no longer needed (for unplanned demads) or until the end of the runnir.g duration required by technical specifications (for a test). Although failure to run is often represented by a per hour failure rate, the focus of the NSAC study was on unreliability; i.e. whether the DG could run long enough to supply the power needed for a mission. For plants with one hour testing durations, these estimates will coincide. More generally, any intention to operate the DG at greater than 50% of plant ESF load rating for one hour or longer was counted as a potential load-run demand.

l

For both failure modes, terminations caused by conditions other than the DG and its immediate support sy

tems were not counted (as demands or as failures). Conditions that invalidated tests or demands fur this study include any operating errors that would not have prevented the DG from being restarted (possibly manually, but frt..n the control room) and brought to load in a few minutes without corrective maintenance; incorrect trip signals that would not have been operative in the emergency mode; and minor A-25 I

water or oil leaks that would not have precluded operation of the DG in an emergency. Also, cases of start or load run failures that occurred during successive retry attempts, and that were caused by the same malfunction  ;

before it was realized that an underlying problem existed, were considered to be one demand and one failure (unless the DG was maintained, declared operable, and returned to service between attempts).

In summary, data from this source are highly recommended for use in reliability and risk assessment studies. EPRI made an unremitting effort to ensure that the data were both comprehensive and reported / evaluated on a consistent basis. Data from this source were entered at a plant unit level, unless the diesels are shared in which case station specific records were presented.

A.2.6 PIPE BREAK FREQUENCY (PBS)

Sponsor of study: U. S. NRC Office of Nuclear Regulatory Research Performing organization: Idaho National Engineering Laboratory (EG&G Idaho, Inc.)

Completion date: May, 1987 NUCLARR qualified record count: 44 Reference number: AA-9 Facility attributes: 81 U. S. commercial nuclear power plants.

CFD attributes: Nineteen occurrences of pipe failures (breaks), supplemented by expert-opinion estimates of frequencies of safety significant pipe breaks in commercial

. U.S. nuclear power plants lhis study empirically developed frequencies and bounds for safety-significant pipe failures in commercial nuclear power plants. its purpose is to update the pipe break frequencies reported in the Reactor Safety Study (WASH-1400),^~3 which are used in many risk analyses. The study involved reviewing various data sources for actual piping failure A-26

events of significant magnitude. More specifically, all breaks in U.S.

commercial nuclear power plant piping (81 plants) during the time frame from commercial operation through December, 1984 having a leak rate of at least 1 gpm for pipes at least 2 inches in diameter and all leak rates of 50 gpm or more regardless of pipe size were recorded. The primary data )

sources were LERs and the Nuclear Power Experience notebooks published and updated by S. M. Stoller Corporation. To the extent possible, information on conditional factors such as the system in which the failure occurred, the operational mode of the plant, and the size of the pipe involved was obtained for each event to permit estimation of conditional pipe break _

frequencies useful to risk analysts. The study also includes a survey of the amount of piping and number of welds typically found in major nuclear power plant systems.

Because there have been tew significant pipe failures, the sparse real data were supplemented with expert opinion data. The report presents the results of combining the real and subjective data through Bayesian statis-tical methods. In addition, an analysis of variance on the combined output to provide a model for the effect of system, pipe size, and the operational mode of the plant was performed.

Due to wide variations in the expert-ooinion data, the results of the analyses that used that data were inconclusive. Therefore, the component -

failure data set contains, at a system level, just the raw data provided by this study.

A-27

\

A.3. REFERENCES A.3.1. GENERAL REFERENCES A 1. W. E. Gilmore, C. D. Gentillon, D. 1. Gertman, G. H. Beers, W. J.

Galyean, B. G. Gilbert, !Lyclear Computerized library for Assessing _

Rgactor Reliability (NUCLARR) Volume 4: User's Guide, NUREG/CR-4639 (EGG-2458), June 1988.

A-2. O. V. Hester, S. R. Brown, and C. D. Gentillon, annotated Biblicarachy of Reliability and Risk Data Sources, EGG-REQ-7827, September 1987.

A-3. Reactor Safety Study, " Appendix lll- Failure Data," WASH-1400, NUREG-75/014, U. S. Nuclear Regulatory Commission, 1975.

A.3.2. PRA SOURCES AP-1. Consumers Power Company, Bio Rock Point Probabilistic Risk Assessment, Harr.h 1981.

AP-2. Northeast Utilities, (nnnecticut Yankee Probabilistic Safety 5ludy.

NUSCo-149, February 1986.

AP 3. Northeast Utilities, M.il] stone 1 Probabilistic Safety Study, ,

NUSCo-147, July 1985..

AP-4. Pickard, Lowe and Garrick, Inc., Westinghouse Electric Corporation, and Fauske & Associates, Inc., Indian Point Probabilistic Safety S19dy, internal document prepared for Commonwealth Edison Company of New York, Inc. and the New York Power Authority, March, 1982.

AP-5. Pickard, Lowe and Garrick, Inc , Westinghouse Electric Corporation, and Fauske & Associates, Inc., Zion Probabilistic Safety Study, prepared for Commonwealth Edison Company, September 1981.

AP 6. Electric Power Research Institute (Nuclear Safety Analysis Center),

OCONEF-3 PRA A Probabilistic Risk Assessment of Oconee Unit 3, NSAC-60, Vols. 1-4, June 1984.

AP-7. J. R; Fragola, P. Appignani, et. al., Unpublished plant-cpecific PRA data for a commercial V. S. Babcock and Wilcox nuclear power plant (" Plant X") [ Handwritten tables and Component Reliability Parameter System (CRPS) sheets], Science Applications International Corp., received in August, 1987.

A-28

(

AP-8. J. R. Fragola, P. Appignani, et al., Unpublished plant-specific PRA data for a commercial V. S. General Electric nuclear power plant (" Plant Y") (Handwritten tables and Coponent Reliability Parameter System (CRPS) sheets], Science Applications International Corp., 1987 (handwritten data received in January, 1988).

AP 9. J. R. Fragola, P. Appignani, et. al., Unpublished plant-specific PRA data for a European commercial nuclear power plant (Component Reliability Parameter System (CRPS) sheets], Science Applications International Corp., 1987 (handwritten data received in January, 1988).

A.3.3. ADDITIONAL SOURCES AA-1. IEEE Guide to the Collection and Presentation of Electrical.

Electronic. Sensina Component and Mechanical Ecuipment Reliabil_i.b _ '

Data for Nuclear Power Generatina Stations, IEEE Std-500-1984, Nuclear Power Engineering Committee, IEEE Power Engineering Society, 1983.

/A-2. J. P. Drago, R. J. Borkowski, D. H. Pike, and F. F. Goldberg, The In-Plant Reliability Data Base for Nuclear Power Plant Components:

Data Cpilection and Methodoloav Report, NUREG/CR-2641, ORNL/TH-8271, July 1982.

AA-3. J. P. Drago, R. J. Borkowski, J. R. Fragola, and J. W. Johnson, The__

In-Plant Reliability Data Base for Nuclear Plant Components:

Interim Data Report--The Pumo Comoonent, NUREG/CR-2886, ORNL/TM-8465, December 1982.

AA-4. R. J. Borkowski et ai., The in-Plant Reliability Data Base for Nuclear Plant Comoonents: Interim Data Reoort--The Valve Component, NUREG/CR-3154, ORNL/TM 8647, December 1983.

AA-5. W. K. Kahl and R. J. Borkowski, The In-Plant Reliability Data Base -

for Nuclear Plant Components: Interim Report--Diesel Generators.

Batteries. Charaers, and Inverters, NUREG/CR-3831, ORNL/TM-9216, January 1985.  ;

AA-6. J-P. Bento, et al., Reliability Data Book for Components in Swedish_

Nuclear Power Plants, RKS 85-25, 1985.

AA-7. Reliability Analysis Center, Nonelectronic Parts Reliability Data, NPRD-3, Rome Air Development Center, Griffiss AFB, New York,13441, Summer 1985.

AA-8. H. Wyckof', The Reliability of Emeroency Diesel Generators at U.Sm Nuclear Power Plants, NSAC-108, September, 1986.

AA-9. R. E. Wright, J. A. Steverson, and W. F. Zuroff, Pioe Break Freauency Estimation for ,9 c, 3 lear Power Plants, NUREG/CR-4407, EGG-2421,.May 1987.

A-29

.~ _ __. _.._ _ _ _ -

/

1 APPENDIX B l l

COMPONENT FAILURE DATA PLOTS i

I i

s b

i l

i -

I

(

r k

7. .

FIGURES B-1. Failure rates and bounds for mechanical components ..... ....... B-3 B-2. Failure rates and bounds for electrical and instrumentation components ..................................................... B-20 B-3. Failure rates and bounds for mechanical components with various designs and normal states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-28 B-4. Failure rates and bounds for electrical and instrumentation components with various designs and normal states . . . . . . . . . . . . . . B-45 B-ii

APPENDIX B COMPONENT FAILURE DATA PLOTS

\ Each row in Tables 3-6 in the body of this report contains a record count; plots in this apperidix display the data set corresponding to each row having a record count equalling or exceeding four. The plots are organized in four multi-paged figures, one set of pages for each table, as follows: _

lable Number ivoe of Componenti Format fiaure Number 3 Mechanical Simple B-1 4 Electrical Simple B-2 5 Mechanical Detailed B-3 6 Electrical Detailed B-4 The electrical components listing contains both electrical and electronic components. For the detailed format, components are' listed by their de-sign and normal state as well as their failure mode. Plots for compo-nent/ failure mode combinations- appear in the same order as the respective entries in Tables 3 through 6. The units (per hour or per demand) for each plot show in the y axis label.

In each figure, there are three frames of plots per page, Each frame can accomodate a maximum of 20 h ta points. These points are estimates of median failure rates or probabilities of failure on demand, with asso-ciated bounds (if available). Each record in the Nuclear Computerized Library for Assessing Reactor Risk (NUCLARR) that is included in this handbook and is in.a group with at least four records gives rise to one wch point.

Within each set of data, points are sorted from small to large. The bars show the upper 95% and lower 5% bounds of the corresponding fitted ,

B-1 4., - ,. -

lognormal distribution for each individual failure rate or probability; the points marked with dots are the medians. Horizontal dotted lines show the 5% and 95% bounds of the fitted lognormal distribution for the aggre-gate of the data set as a whole. Each bar is marked on the abscissa with the study-id of its data source.

For component / failure mode combinations (or, for the detailed tabits, also component designs and normal states) for which more than 20 "IRADAP Suitable" records exist, several successive frames are used to capture the information. Regardless of whether one frame or several frames are used to plot each group of data, plots for successive groups are separated by dashed lines. The cases with multiple frames of p'ots are denoted by a "Part n of m" indication below the corresponding frame. Note that the scale for the vertical axis is not necessarily the same for each plot within one of thess groups; the scale for each individual frame is ,

selected based on the specific data to be plotted.

s B-2

,-we~ ,-e-., , , . . -

,.,e-, - - , , ,-m-e,- ~ -me,m,

l l

Fons, Ventilater s-Fc;is to Ran I DE - 004 r

N e v 10E-005 -, -

o I CC $

10E-000 . . -

v V

4

'S 10E-007 4 L.

10E -008 s e 9~1, 9., v.,

9, 9,,

- . 'l 4 , G- -p y *1 , .m 9lL,,.W d s -

Vp 1

W-s s_t x

&J s&v s

Av ,&s J J  %

1 -

Heat Exchanges-Inter noi Leakage 10E-oo3 N

g 1OC-004 _

o . .

10E-005 . _ _ . .

v

_3

'610E-CC6 _ , ,

u.

1DE-007 u.

q. vp, F;,,. r, p

c c c ss'kt 't"I J, 4

Dm '.$

Q D,,

m d,

m i

• h 1OE-004 r

N

• 3y 4, ,

• L1. ~ t)r, m _

b<

t).

v

~.' 006 _

O L.

10E-C07

&2,, r,

> s, x-J

", 4

/ 9 1,  :

- - = = p -== -- - .----..-.---_ . --- --_-

h a '1 eU B-3

Pipe-Externol LeoLoge/ Rupture 10E -006 _

r N

• iot-oo?

o

_.__ __ _ .__ .,_,,_,_m_ ,,_

e , , , , , < < ,. <, , , , < ,

9

~ ""

310E-005

' ~ ~

_~ _

'5 a

10E-009 aq o 4aoe%4Aa o$g Vy ug'V3s Q 4 4Up 4 g&f(&}

G3 Vo. 4,y 3g O g v3' &e, Cry p &g vg Q} G}

Poi t 1 of 3 (42 Records Totcl)

Pipe-Externo; Leckage/ Rupture 10E -005 x

N 310E-006 o _

o , ,

1 -- l , o

) 10E-007 _ _

~S L

~ "~'~~~" '~""

10E-00S . .

%g%3 %gRy%{by %g'by %f043  % %y%g%3%g%y%{&%g%p Port 2 of 3 (42 Records Total)

Pipe-Externo! Leckage/Ruoture 1CE-005 _,

r N

m O

O 10E-006 "

E v

)1OE-007 _

'5 a

q q v3 \r3 Port 3 of 3 (4 2 Records Totan Fiqure B- 1. Foilure rotes and bounds f or rnec ncrocal comoor.en t s -

(cdntir,ued).

B-4

Pumps, Diesel Driven-Falls to Stort 10E +000

\ .OE-001 1

f v 5

[010E-002 ,

{10E-003 L ,

3 u.1,0E- 004 y

5 10E-005 . . . -J

• %q *1n, *eb, R. %yL *% %q 1.3 "e4 1 4 At -

Pumps, Motor Driven-Fails to Stort toe-001 u

N 10E-002 .,

o -

c' toe-oo3 _ o , ,

o y

6

'6 toe-004 I _

u.

1.C E -005 . _ . _ . _ _ _ _

p I

5 Chs, &, , *, %,13l, k '<! %c,A '< Is,, t, U 5,k,k h Y,%  %  % tf t f 'Q Port 1 of 5 (51 Records Total)

Pumps, Motor Driven-Fcils to Start 10E-001 7

N 3o 10E-002 .

CC '

o g,

_,, , _ o-,_,, ,

) 10E-003 - ,

g .. ._ _ _ ..._

10E-004 s 4. 4 J- '-e Eof eo, % ' o Sea s

6 .L u -t, r, sY 4 r n S. '

la 1$ Ag sg%% a.,dys .tc %s% 'o% T'4,.q's'v% % c r g/r> -

s,%,4 4 Yg'Q Aj% ,. $%; 4% % %y%'e 4 V 's,%

Port 2 of 5 (51 Recor cs Totat)

Fiqure B- 1. Foilure rates and bounds f c4 rrec hanical  : o corents (continued).

B-S

___m_-.-.____ _ _ _ _ _ _ _ _ . _ - . - _ _ _ _ - _ . _ . - - - _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ . . _ _ _ _ _ _ _ _

Pumps, Motor Driven-Fails to Stort I CE-001 V

N

-y 10E-032 -

, o ,

cr; ' " '

)

, , i.<

v jtot-co3 . _

g -... . . . _ . _ . . .._ . . . - . . .

10E-004 . .

'% / k'* [  % h *4 Port 3 of 5 (81 Records Totol)

Purnos, Motor Dr;ven-Fai!s to Stort 1oE+000 m

N

, 10E-001 .

o o

{

,4 i<

" 1CE-002 o' ' D

        ^O1OE-003

u. ...... .. . . . . _ . ~ _ ._ _._..;

1 OE-004 , ,  %'S ^% ' Port 4 of 5 (81 Records Total) _ Pumps, Motor Driven-Foils to Start 1.0E + 000 t N y 10E-001 , _. d Cr 10E -002 _ y - _5

         '6 10E -003 L                                                            . _ . _ _ _ _ .            _

10E-004

                                                                                                                                       + s o

21 Port 5 of 5 (81 Pe;ords Totci) Figure B--1. Fa: lure rates and bounds f or rnechankCl Components (continued). i B-6 e

Nmos Motor Driven-Fails to Run 10E-oo3 r N

        ,10E-004        _                                                                                                                                                         _

E CE 10E-005 - o , ,

       )                ,
                                                     ,, ,               ,           i i         '
       '6 10E-006 _<                                                                                                                                                              ,

u. ,

I r toe-007 _ _ _ . . . _ 3 Yi 3 h#%s _ Ds '3 '* Q  %}fb Port 1 of 5 (97 Records Total) Pumps, Motor Driven-Fails to Run 1.0E -003 r N

         ,10E-004         -

l o 10E-005 ._

                                  , _,       _.._i        _<         _,          __' _"._."_

b

        '6 10E-006        ,

w , -. - . . . . - - 10E-007 _ _ _

                                    %              hYb A.fhhh%, k% 3, R                                                                          o, t             'l'             *r
                                                      ~*k%y            4                                               %,               *%%      ,. %,                                   ~%

Port- 2 of 5 (97 Records Total) _ Purnps, Motor Jriven-Fails to Run 10E -003 k.

          ,10E-004         _

g g , , < < o < < < " {, o o f '"tf 'f_ 1CE-005 _ _ y

• b

         '6 10E-006         .                                                                                                                                                          _

L . . . _ _ . . _ .- . . . _ . . _ . . . 10E-007 . . . _ _ . D4 f 'I #

                                                                                                    '%,y @

I 3, E?7 ,,, a 44

                                                                %                                                       w Dg         k. vxC                       y.

4 M, c, ,

                                                                                                                         %,           4    *cl7 Part 3 of 5 (97 Records Toto!)

Figur Failure rates and bounds for mechanica! cornconents (cont,einued). B- 1. B-7 L...-

Purr , Motor Driven-Fails to Run tCE-oo3 N. '} ,, toe-co4 {, tf u {- { , O CC 10E-005 .- - - - o b

                   ~6 10E-006        _

u. . . . . . . . . . _ . . . . . . . _ . . . . . . . . . . . _.. . . _ . . . . . . . . . . . . ,

10E-007 -. . . _

                                                          *o % !                     '

O J it Port 4 of 5 (97 Records Totol) Pumps, Motor Driven-Fails to Run 10>000 (10E-oot . _ o10F-002 _ t ' f_ r2LOE-003 _ 0 10E-004 4 _ kY' _ j10E-005 10E-006 __ . . . - . _ . - . . . _ _ _ . . . . . . . . , . _ . . _ . . _ . . . . _ . . . _ . . . . . _ . ._ 10E-007

                                                                                                                                                                                                 'v +

QQQQ+qr+ ++d,%+ Ap .k 94/ . 't'y '*g'*g'*g,p. *g G '* Q Part 5 of 5 (97 Records Totcl) Pumps, Steam Turb;ne Driven-Fals to Stort 10E+000 0 N

                      ,10E-001            -                                                                                                                                                                         ,

o o f u f LOE-002 . . _ . . . _ . . - ~~ _ ...._.. ,.- o - . _ . - ii

                     ~5loE-oo3 L.

i _ _ 10E-OC4 . (1g 4%. 'n'J %'x4

                                                         +g                                    %,

g '%'/ s v $ $ P 4 O 'N Fiaure B-1 Failure rates and bounds f or mechanico! comoonents ' (co'ntinued). u B-8 1 _ _ _ _ _ _ - - - _ ._ ~

Pumps Steam Turbine Driven-Foils to Run 10E4000 (10E-001 _ -

                                                                                                                                                           }

1o 1OE-002 } { _ 10E-003 - - o , ,,

   ) 10E-004          _

4

                                                             <         f                                                                            -

O 4

   ' 10E-005          -                                                                                                                             -

TOE-006 -- -~ ~ s y

                                                 ~

s >- +

                                  +%               r'                  e$$                                e%'             %

17 a'4p # 3'Sp, ,

                                                                                                                             % 4k, 94           #

Pumps. Driver Not Specified-Fails to Run 1.0E-001 r N 10E-002 ~. _ . 3o 10E-.003 - Z 10E-001 _ _ v ' 31.0C-005 .  ! 5

l. ' 10E-006 _ _

         'l DE-007                       .-                .

l ds d& %y *es. 'bs ds s y

                                                                                                             '$4             4              '

9, 4 < b,7 Vo!ves, Check-Foits to Operate 1 10E-002 t

   -N
      ,a  10E-003      -                                                                                                                            -

3 g o , < , m m o e a -~ 10E-004 ,. o , e.

                                                .,-o.           .                           ~ . .         .    ..   ~.           ..               -
    ^610E-005          ,_

L toe-Oc6 _+ 5 5 > ;- +e# -+A A A & .r wTjsw Ty kw'th  %,w'w+ww

                                                      ?$ i 'd;      f T7 l'7    wbysw   s
                                                                                               't'yw+vs:7w+'gw+co+ng
                                                                                                            'oy 'h t t,g ' C7 h7 'N y T/ w Port 1 of 3 (46 Records Total)

Fioure B- 1. Follure rates and counds for meencnicci cornoonents (c o'ntinuad). B-9 L

_. . _= _ _ _ . . Valves, Check-Fails to Operate 1.0E -001 V-N 1 3 '10E-002 I ' ff. ' o o .,.I I x ' ' ', o 10E-003 ' 4 < o u , l

    =

k o10E .004 - -. 10E-005 .

                                          .++

k + + R>. '- S + S 3 r 3 M ' T*4'c SJ Tps,

                              + %*      +J4           .,

fi if'c %+h 3 ep+*g' gfN4, 't.7

                                                                               %                  yT/ %              i/po         \     %+%.pO
                                                                                                                                        $T
                                                                                                                                   / / 7 '4 ,9,                    N D ,a S/          ;C,, 3D 4, 4-4                      "r 4 Part 2 of 3 (46 Records Total)

Valves, Check-Fails to Operate 1.0E-001 V 1 N I 1 3 0E-002 t i o - o Z 1.M -003 - . o  :

    ~D 6 10E-004           -                                                                                                                                                                -

10E-005 Sp V k l-O +'*g D +Ng, %g M, g g 'i4 Pcrt 3 of 3 (46 R cords Total)

v. '

C hec'< - Fails to Open 10E-00' 10E-003 _. , .- o , 1 13E-004 -. o g . D

    ~5 1.0E-005         y                                                                                                                                                                -

u. 10E-006 hgv. hg  %, 4 4 O %,s 't 4p "?p

                                         $s                                                                              9                    v            4. ,,               4, s.
                                                             ~%                     g%                   %                       'c,            '#
                                                                                                                                                 'p+.          .~ %_                  k tf                     '

t7 17 t;. tf Figure B-1 Fa+re rates and bounds f or machanico! componer.ts (continued). B-10 k

1 1 Volves, Check-f ails to Close 10E-001 . r N

       ,10E-002          -                                                                                                                                                                                                                                       -.

o CC 10E-003 - ~ v o L q 10E-005 . t s y v'C I,

                                                                                $;s-                                                           %%                                Qv                            j g
                                                                                                                                                                                                                                      'h w

k k D; T1 _ Volves, Check-Internal Leckage , 10E-OO4

I N u.,c nndr * -

                                                                                                                                                                                                                                               ,     ,          ,m Q                                                                                                                                                                                                                 ,   
                            .                                                                                                                                                     ,I   i 0 10E-006 s.v o                                                                                                               -,

q. "'

                   '07 m,
       .. ' (,        OS      ,                                                                                                                                                                                                                                     -

s . . _ .

             ,.c. 39                                                                                         .                                                         .                                      .

d - A- c t r

                                  +%+%+%                                                                   A+%+,                                                                                              4
                                        %;%#9%Mi                                                                                             2; w+w+%r%+w
                                                                                                                                                    % % m; 9; $7w+% Ac . +w w+s w+. w+'cq                   Q,'r L; m; %s; %; % y
                                                                                                                      <t 1 of 2 (34 Records Toto!)                                                                                                                                                       _

Volves, Check-Internal Lecxage 10E-001 r P f s^10E-002 10E-Of ' t {_ c) E 10t- - 10E, k i ' " o _3_ 10E W'. _

        ~610E-007               _                                                                                                                                                                                                                                     -.

L 10E-008 _ _.

             $ 0E--009
                                     + 4                                                       4                                 + >-                   i. +                         9         w7 +                   1        a                  5          .c k w% $ %                                                                                                's %W                               4.,                3    WW                't                   u           .

f YJ f l 'l 'l ' bj f $,9 .<l %4 %l Part 2 of 2 (34 Records TotcQ

                                                                                                 .$                                                                                        a                                                                                       +

(c5ntinued). , B-11

Valves, Hydraulic Operator-Fails to Operate 1.0E-001

                   'o N-g     10E-002        -                                                                                                                 ,

o k e 10E-003 _. _ . . _ . . . - . . . _.. 2 4 h

                   '6 10E-004            _

L 10E -OO5 _ . s

• A

                                                                          %              +%                     $f                       %
                                                                                               -Q                    'R#                  Q Q

Valves, Motor Operator-Fails to Operate 1CF-002

                    .c N 10E-003             _                                                                                                                   _
                     *o 10E-004           _                                                                                                                   _

Z 1Cf-005 - - o 310E-006 -

                    '5                             . . . <. . . . ,. . . . . . . '

w - . . . - . . . . . . . . - . 1.0E-008

                                                         +                                                                        k      +    +%
                                                                   +%. +% +% +% +% +%.                                              %
                                                                             %       % k % % %                                       N                %

Valves, Motor Oc= itor-Fails to Operate r N g 10E-002 -. g _ s x ..

                                                                                   ,  ,,            ((+tif' e

10E-003 - ( 1 I { --- I I 2 , L 51CE-004 - 10E-005

• a

                                                     +'4
                                                                                                                   -- 6
                                                                                                                                  # -t 9,'%, g g'f]% + pgp" egf':'sl*+,'T s s + +3'!y *g Port 1 of 2 (24 Records Totai)

Figure 8-1. Foilure rctes and bounds f or mechanicci ccmoonents (c6ntinued). B-12 1

I l l l Voives. Motor Oper ator-Fa;!s to Operate 10E + 000 t N 1 I I a o 1OE-001 i o 1 1OE-00' i o _3_

                     '5 10E-CO3                                                                                                                                                                                                                      -

u. toe-oo4

                                                                                                                                                                                               +                       %
                                                                                                                                   +\c.
                                                                                                                                          ^d7
                                                                                                                                                   "_  Sc
                                                                                                                                                                                                  'c,
                                                                                                                                                                                                     %f
                                                                                                                                                                                                                          * ($,,

xy Part 2 of 2 (24 Records Total) Volves, Motor Operator-Fails to Open 1.0E-001 t N o o C:' 10E-00' O

                                                                    ~

o 5 h

                       ~5 L.

1OE-003 4, d  %&

                                                                                                                                         *s,            %                                       e%,
                                                                                                                                            '%g             %,                                        '%y                               e7 Valves, Motor Operator-Fans to Cbse 10E+003 TJ N

3o 1CE-00t - o 1 ii 2 1.0E-002 I _,

                         ~5 u-                                                                                                                                                                                             ,

o 10E-003 G,. 4  % k. Wm

                                                                                                                                   'y            *f,               'J,                                        "3 N                                                                                     T j                                                a,                                                 _,
                                                              - - = - - - - - - . - . .                                                                     ,    .                                      - - . -           - - - - - - - - - - - -

g P' y p P (cdotinued). B-13

Volves, Motor Operator-Spurious Operation 10E-005

                                 .c N

2O e e 10E-006 _

                                                                                                                                                                           ._.....~..._:
                                 =

5 o u 10E-007 q q \ q og Valves, Motor Operator-Spurious Open 10E-002 r N 10E-003 _ _ 3O 10E-004 _ e 10E-005 _ _ o

                                  .$10E-006 o
                                                                                                                          }'                                                                            _

a 10E-007 _

                                                                                                                     --"~                                            " ~ ~ ' ~ ~      '"-' ' " "

10E-003 6  %

                                                                                                       %'g                 %g                       %,                    %'g, l

Valves, Manual Operator (mechanical handwheel)-Fails to Operate 10E-005 , z $

                                                                                =.
                                   \

g 10E-C06 O-C 10E-007 - , e " g ,,

                                     =                                              .

0 10E-008 <" v - - L 1CE-009

                                                                                                                    +                     +
                                                                                                                                              %_ +%+g +'g. +g+% +'.g .
                                                                                        +'v3  k4+g+g+%A   W         1 g+g+'g,/

W W

• 4 "I 4 NiW 4 Ul Fiqure B-t FcHure rctes and bounds ear mechanical cornoonents '

(c6ntinued). B-14 p

Vanes, ' Manual Operator (mechanical nandwheel)-Fa;is to Operate 10E-001 v N 10E-002 l_ ) f, e

                    ~

m m Io 0 ' c'10E-003 ,

   =

{10E-004 ,_ , 0 u.10E-005 ...., ..< . 10E-006

                                                                                                                  . xx +++ . +%

J-

                         +Ns+w,+

gy +w+w%w% $ w w w g <+- c%www w+w+w+kg+w%w w w+w% ,., g Vanes, Pneumatic Operator-Fails to Operate 10E-004

   .C N

g 10E-005 -

   ~                                                                                                                                          .

c Lc 1DE-006 -. . o D " "~~ '~

                            ~ '                                                                     ~                                     ~

3 10E-0C7 - _ L f

                    ~

10E--008 . . . . . . .

                                                                       +
                                                                                    +w                  +
                                                                                                                   +w
                                +w% +w%w$    .
                                                                                             %. w$                       42,,  +w% +w               %                  w%

VoWes, Pneumatic Operator-Fails to Ope ate 1DE+000 n N 10E-001 _ ff,

   ~

v f g1OE-002 , , t

  'ytoE-003                                  ,, ,                f                                                                                                            _

m ~~~ ~ o i u.10E-004 r < 1.0E-005

                         + %J. % ,, k e7 s                             Y .r r n               + m. + n                                       a +

e

                                                                                 %+% k h,+A W                             QL3.Tg+%+'o.e                                    4.

Nk9 $,yP i / ' NN" k ^ l l ~57 l 1 Figure B- 1. Failure rates and bound _ f .' e ine:hanko! :crrounents (continued). 8-15

Volves, F.'eamatic Operator-Spurious Oper ation 10E-005 _. r i 11 0

            -o10E-006          -                                                                                                                                               _

C . . . _ . . , . . . - . . . _ . . . ~ . . ._,. , . . . - . . . v

            .Dg 10E-007 u

10E-008 64 t . c.>. h q & s.Ry 't$ *R ti Vcives, Pneumatic Operator-Spurious Close 10E -004 r N 10E-005 - _

            ,v                                                                                                                                            o O

Z 10E-006 - _ v _h

            '5 toe-oo7         _ . . _ _ . _ - . . -                  . . _ . . .       . _                                                                  -            ..._

L 10E-003 0 Nt #fr; G9 X, 9  %, i)4g Valves, Pneumatic Operator-internal Leckage 10E-005 N v C , g o 10E-006 _ v - u o 2 . _. _. _ . , . . . . ,

            ~6 L

1CE-007 es., c. .i- n, W - x'p ' .g,  %@A Figure B-1. Failure rates and bounds f or mecnonical components (continued). B-16

                                                                                                                                                                                                 ~ .

Va!ves, ReGef-Faib to Operate 10E-004 < r I

   \ .OE-005
                      ~

1 ,. l ., a E0 10F-005

                      -                                                                       <             4 o

007 , , , 908 . ~ . . ~ . . _ _ . . _ . . . _ . _-. . . _ _ 10E-009 . .

                                + +
                                                                                             +                              +                  +V +
                                                            +% +% +%                                       +% +% ' % +% _ I, % +%
                                           .                           4 Ty fy                                     tf %            g t;                        17 %

Volves, ReGef-Fails to Operate 10E+000

   'o N _0E-001 1            L                                                                                                                                                        ,

e., .> a0 10E-002 ' o

                                                                 ""'             -[

1_0E-003 - q> _g

                                                 . ~ . .                                          -

w 1.0E-004 10E-005 . .

                                                                 *                                                                  * * .                       *+
                                                           %,n.        %+V+ 4   %g%+
                            +%A     ?4         +%+3   y                 4            t; -                 w% t; 2, 2; 94c.4 ti 1M
                                                                                                            +      t  +k+%+w+%                          4     %;%+%+%ty                  w_

Valves. Relief-Foas to Open 1.0E+ 000 f V - N g 10E-001 _ o fv . <>

   ^ 1.0E-002        -                                                                                                         i                                              __,

e f

   .b
   '5 10E-G03 w                    _ - . . . . . . . . ,                     -.                                                                                                        ..-.

1.0E-004 . _ _ _

                                                 $t3                          k9                                               $-.#

s% %Y f. I;,f* x t y. t.f'q'

                                                                't                                                    %                   Y                     M7 n               '              "

g (continued).g B-17

Volves, Relief--Fails to Close 10E+000 V N

    ,10E-001       -                                                                                                                                      .,

o e y o t0E-002 - , e o

   .o 10E-003      -

o

u. .

10E-004

                                             %               Q                                                 J'9p                 J'9
                                                                  %                  '%                                                      4 Voives, Relief-Spurious Open 10E-004
   .c N

3O 10E-005 _ _ E o

    *                                                     }
   .510E-006                            -
                                                         'l"                                                                                             _
   'O L.

10E-007

• de 'D #

                                                         #n" M

r4 J'g 0p's Valves, Relief-Internct Leckage 10E-001__ Q10E-002 - - e 10E-003 o m o - 1310E-OO4 _ l _ m 10E-005 _ _ o o

   -g 10E-006      _                                                                                                                                     _
   '6 u.

10E-007 _ _ 10E-008 _ - 10E-009

                                   +%,       e.       1.             A          ,-        A                    A           A              A
                                              ~%g+      %               %          %        'c ,.                 M          '4.             %
                                        %                 t,                 %      %           %                       S;      %:             9 Figure B-1.                     Failure rates and bounds for mechanical cornponents (continued).

B-18

Nfolves, Solenoid Operator -FoHs to Operate ', 10E-001 (10E-002 - _ v 10E-003 . . < _

            "o                                                                                                                                                                      '

e 10E-004 _ _ o 10E-005 _ _ h '

             ~6 10E-005              -

tu 1OE-007 .. _ . - . . _ . - . . . _ . . . _ .- . _ _ 10E- 008 -- _J t- >- >- n -

                                                                        %                                       +%                      +%                                    +%              %

ty

                                                                                                                                                                                                         %        +%         %         %_
                                                                                  %                                        'B7                          %                                17                  %         t7       T7        ty Vanes, Solenoid Operator-Foits to Ope ate 10E-001                                                                                                                                                                                                     _
                                                                                                                                                                                                                                                     ~
             'O N                                                                                                                                                                                                                       "

g 10E-002 <

             ~

h il X 10E-003 - _ o n

             ._h
             '5 10E -004               n ..

tu P-10E-005

                                                                                                                                                                                                        +           +
                                                                                                                        >-                   >-                                                                              n      e
                                                                  +%                                                      %                        %                              +% +%y                    %         %cy       %       'hT
                                                                                       $                                    'R,                             my                            t7                 $                   v7 Fiaure B-1.                                                  Faure rates and bounds f or mechanica! components                                                                                                     '

(cdntinued). l B-19

Batteries-Fails to Operate 10E-004 r N

              ,10E-005        _                                                                                                                                                         ,
             ~

o 2 10E-006 o < o  : y._. y . . - . . _ . . . . . _ . . . . . . . - . . . . . .

             '5 10E-007        _                                                                                                                                                                                                      _

L 10E-003

                                                                                                                                       ~
                                                             *c4                                      x                                st                b                 :               3              'q                     ?

Charger, Battery-Fcils to Operate 10E-004 r N o o o 3

              -o O

1.0E-005 _ Z o .._ < _._ . _ _ . _ _ . _ . . . _ _ _ . . _ . _ .

              ,)10E-006          _

o L 10E-007 . . _ r  %  % .c. +

                                                                                                                          %q                      'O            ik 'De%M4 c.,            b,                                ^,:.

Circuit Breckers. Powec-Fails to Operate 10E-001 - 7

              \ 10E-002 L                                                                                                                                                                                $'

v

                                                                                                                                                                                           +
               "M10E-003           -                                                                                                                                 f                                                                    _

k

                &10E-004 L                                                    j
               .---                                                   .. L.                                        . . . _ . .                                _                                               _                ..._.'

w 10E-005 ~,_ , 10E-006

                                                      +.m,                  4,                                                         +          b.                34                     3.,                     3 t'N                                               'oc t                     O                 b 's, .                   b i                                                      T/                                                f3                                T      'N                         / .,
                                                                                                                                                                                                                              'Q
                                                                                                                  ?4                                      5;                      &                     '&                        $

t ______________.____________._____.___________ Iigure B-2. FOiiure rGtes God b._' inds f Or e:ectrical Cnd instru-mentation components. B-20 l

                                                                                                                                                                                                                                                )

1-Circuit Breckers, Power-Faih to Open 10m-CC1 V N l g 10E-002 - 1 -

    ~                                                                                                                                                                                                                  l U

C 1.0E-003 - o ,

    -D                                                                       '
    'O u.,

1.0E-004 . 1 OE-oc5 ee e e, ams J e%'r' 3' g,

                                                                                      %                         4                                                                                                              1 4

Circuit Breckers, Power-Spurious Open 1GE-005 I

       .c
     \
       -o o 1.0E-006                            _

e o w v - 2 10E-007 _

       ~6 u.

10E-008 vp dy b'J  % igs  %., st. g1 '99 4

d.  %,.

M) Mp E. Conductors-Feils to Operate 10E-005

          .c N
             ,1.0E-006                               _                                                                                                                                                                                         _

o C " 10E-007 _. y , ._

                                                           . . . . . .                                      ..s
          'O 10E-003                                                                                                                                                                                                                           _

L_ 1.0E-009 e,  % J. c, 4-o tc. g' 2 J' q 'J ~. beu  %)

                                                                                                                            'n G  u                  %,.$,                                                                                 My                                    'F o

Ficure B-2. Failure rates and bounds fcr i : : t r c. a ~1 H stru - rnEntotion cor.1ponents (continued). B-21

Computation Modules-Fails to Operate 1OE-004

                                                         .=

N jt0E-005 ,. , _ m o 2 10E-006 _ o , E 10E-007 _ tp t.ie P. t~q., hc  % p Q e ~4

                                                                                                                                                        %                       S             %           %                    C              %

Generator, with Diesel EngMe Driver-Fails to Start 10E-001 _ 1- g N

                                                           ,10E-002                                  -                                                                                                                                           ,-

o i* t0E-003 _. - m , , , e - o

                                                          '5 10E-004                                    -                                                                          1                                                                  _

10E-005t_ . . . . . C t TOI c e It I TOt / e

                                                                                                                                            %D%D4[  e              e              g bgg 44g6,b,4b,b,bg'gg                                   g s

Por i 1 of 4 (75 Records Total) - Generator, with Diese! Engine Dr:ver-Fails to Start 10E-001 o N

                                                           ~$10E-002 U

Z

                                                                                                                                                                                                          "               ,   ,   , .,,-           i-e

t. , o o , o o , , "' ' '

                                                                                                                                                                                                                                       -[
                                                            @10E-003                                                                                               .

O L

,                                                               1CE-004
                                                                                                                                        % % % 4w h h de 4g 4w ? 4 -

C Y[t[t['O t #O t %(3 % 6 % 6#d4g4 'O h  % Y [ b% 4 % %,% % [' s%%  %[b[%%%%s%[% 9

                                                                                                                                                                                                                                 % @s %(                   i 9 Part 2 of 4 (76 Records Toto!)

Fiqure B-2. Failure rates and councs for eiec trica and instru-mentation cornponents (continuea). B-22 i I

f Generator, with Diese! Engine Driver-Fdis to Star t 10E-OO1 v N 3 1.0E-OO2 _ _

          -o                                                                                                                                                                                                                                                             ,     ,            ,

Q' m_m__o__" < " " " ' ' " { v 310E-003 _- -

          '5 a

10E-004 . . . . . . .

                                                                         % 4w 4p 4 .                                                           e.g         14 A4 4A1 4 7%3                                                                                               # 1w A .3         g%s s4y,ic                                       n S,%'Mc &c,p<f,                                                                      5 Q'p,y           4l,   4c Q +'%,O ' y
                                                                           '%'f4$s s          cs c4 so'A                                              s s                                                          s s

s s s v s b, Qp b, sty b, tg b,.3Ss723 4 6, q,stpQ, g Qp,ty 4,4 99 Port 3 of 4 (76 Records Totc!) Generator, with Diesel Engine Driver-Fails to Start 10E-001 t i l N e 1, 1 - o o y < n 10E-002 _ qi < <

           .* ' ^b o

L 10E-003 -- - b0be 4,c#$441 4 y s s 3 s 4 4

                                                                                                                    %, c c b 0l %'O         s        s     s pc s

44, l.g 14 Qe %:ip'c%7 s 'cs c cW s s f Os D$ gD G D ,",I D9 h kp b9 kp Dg b@ Dy 9  ; . bp kp Part 4 of 4 (76 Records Totc0 Generator, with Diesel Engine Driver-Fai!s to Pun 10E-001 N 3o 1CE-002 _. _ I.f n 0 310E-003 ._

              '5                                                                                             .__.

L 1CE-004 b, ' > .ru

                                                                                          <>,s                                                 $2, N.

d

                                                                                                                                                                                                                                                                                 l, f
                                                                                                             ,                                     'e n                                 4                                                                        9,                    4
                                                                                                                                                                                              'dj
                                                                                                                                                       'dj                                                                                                        #

k' Fiaure B-2. Fai:ure rates and boands W eiec t r o 2 cd atru - rndntction cornponents (continuea) B-23

Generator, with Diesel Engine Driver-Fans to Run 10E-001 m N

      ,10E-002                                                 -                                                             i                                                                                                      _

o , , g _. . . _ .. . ., 10E-003 _ , i e < 5

     'O 10E-004                                                _                                                                                                                                                                    ,

u. 10E-075

                                                                        - 4                  '
                                                                                                               .1                     1           4 % 48                    Ai.4                                  %4I h'iIg1$

s n s s s 4 s !1-9 ,4,f4 '3,1 s s s 8 9s s '4m'4 ' p 4.m(w s 9 1 ig<f s s s9 4m 4 s%sb eee9es9 b b C b b 4 %s bee ( b % %s%sb e e %sb e %b % Port 1 of 4 (70 Records Total) Generator, with Diesei Engine Driver-Fails to Run 10E-001 7 I

      ~*

o 10E-002 _ o__" ' ' o , ,_ _.__o_m__. Z < o ,, < o . . . . . .. . . . .

      ) toe-003                                                  _                                                                                                                                                                    _
                                                                                                                                                                                                                                                   )

o L 1OE-004 400 4 0 4 0 4 40

                                                                                                                                                                                                                                      'q f4 "Yf*o \*                         ' 1                                      40 ^1 D, g 6, bg94b gbg g4ggb,gC ggggg                                                                         3 Part 2 of 4 (70 Records Totcl)

Generator, with Diesel Engine Driver-Fails to Run tot-nol n N O , , , g , , 4 o m , o '

       " 10E-002                                                   _<                    '       '

o - No L 10E-CC3 . . 1 h%hN 4 h- 4- h 4w h h 4 A4h 'S' % {,%M.ltw$,h*1,%. 4. (s "4ys {s4 W(s% s s (s$m 's s 4(s' s s {s s 's s s . s s s D3 Dg D;9 D;g g D y 9D gD h % D 9D,9 g D9Ug, D g 99D D tg Ug Part 3 of 4 (70 Records Total) Figure B-2. Failure rates and bouncs for etectrical and instru-rnentation components (continued). B-24 l i 1

               - - - _ _ _ _ _ _ _ _ _ - - - _ _ . _ _ _ - _ _                                                                                                                                                                                      l

Generator, with Diesel Eng;ne Driver-Fails to Run 10E+000 , t N 4v o 10E-001 _ e " I

                                                                                                                                                                                                                                                 }

u

                                ._0                    i CE-002                                                                                                                                                                                                                                -
                                .mo L

10E-003 _ _ 4,g 1 4 dy Q id

                                                                                                                                                                                                                        4c 1

4,o 1 4cs  !.g ig, .3,s, es se. cs , s

                                                                                                                                                                                                                                                                     ,e s
                                                                                                                                                                                                                                                                                ,c         ys h                  .USS DS                                                                          DW                    b h h                                      N9      k9          k9 Part 4 of a (70 Records Totai)

Reicys, Control-Faits to Opercie ICE-004 N

                                       ,,,10E-005                          _

{

                                    ~

o C * ' 1 CE -006 - e o o y , _ g . 4 u, 0 10E-007 _ 10E-008

                                                                                               '9q                        %,

O

i. C,

                                                                                                                                                                                                                                               yO                 S7           %

O t'g 'g 'g i

                                                                                                                                                                                          'O'd                  3                                       '63                 '63      'O'g Reicys, Protective-Faas to Operate 10E-005
                                          .c N
                                             ,, 10E-006                                                                                                                                                                                                   ,                ,          .
                                                                                                                                                                                                                  - e-7,:;                                                                           _

o v t 1CE-007 - _ e _=- h

                                          ~ 5 1.0E-009                        _                                                                                                                                                                                                                    _

u. 1.CE -003

                                                                                              %,,                      t,.                                    %.                                                                          .              5.n,              t,,       4,

o. g C,..'d b .g s i w g$ 6- r de 6a 6.

3 9>- u v u 3 Fiqure B-2. Fahure rates and bc unas f cr e'ectrkt cr -J instru - r-Enta tion cornccnents (continued). B-25 1

' i Transoucers (Detectors / Elements / Sensors)-Fcils to Operate 10E-C01 t N

                                                                                                                      ,10E-002                                                                                                                                                                   -

o tr " 1CE-003 i - y _ _b. i

                                                                                                                     '6 10E-OO4                                   <                                                                                                                              ,

u. -

10E-005 . . . Q3 J' g '

                                                                                                                                                                                                                              %g           %g                           %g M                                                          4                     %g
                                                                                                                                                                                                                                                                                     '$4 Transducers (Detactors/E:ernents/ Sensor s)-Spurious Operation LOE-005
                                                                                                                      .c                                                                                                                                                                  g
                                                                                                                      \                                                                                                                                                  ,
                                                                                                                     --oe10E-006                                                                                         1 Oc                                                             <
                                                                                                                                                                                               -1 T

3 f

                                                                                                                                            ,._.+.,I                                              . . ~ . . . . . . - . . . _ . . . _

Q) ,.., . . . . . _ . . - . . . - -. .< _h 10E-OO7 _ _

                                                                                                                                                                                                                                                                                                             /
                                                                                                                      '6 u

10E-008 __ k %c k k %g %#% % % %c l 4 44 4 '44 4 % 4 4 44 % %g 44 Transn-itters-Fails to Operate 10E-005 P - . . Il 'I 3o 10E-006 ._ _ A T , u j10E-007 _ _ o L 10E-00S kgg k %w k.k k k,k k.k so s s s s s g v ag 'g 14 '4 s%4 'T4 '94 44 % , Ficure B-2. Failure rates and bounds for electricci cnd instru-m5ntction cornoonents (continued). B-26

1 i l- Transformers, Centrol and !nstrumentation-Feits to Operate 10E-004 fN g,10E-005 - -

                 ~

o 1 10E-006 - _ v b_. . . . . . . . -.. ._. - - . _ . _ . . _ . . o 10E-007 , L. IfE-008 . . Qy Q Q, 0p Ary

                                                                     .,J,                     Q                        g e.

v.. , 9 4 .4 v,g,

                                                                               -y                  -4. _

Ironsformers, Power-Fails to Operate 10E-004 N 10E-005 _ - e 10E-006 - o

                  'O ICE-007          -                                                                                                                                                 ._

L_ 10E-003 - . _ .

  .                                                            4__

f p e,., p^ e6

                                                                             #'o              C,'g                        '+4                  V         'hg
                                                                                                                                                                                             ~

Figure B-2. Failure rates and bounds fer electrica! an" instro-mentation components (continued).- B-27

Fans, Ventilators (Fons)-Unsp./Fcils to Run 10E-004 c l 1 N . 10E-005 ' ' ~ e 1 i ) 10E-006 T ,, o 1

                                                                                                                                                            ~

r b I 5 10E-007 _ ,

                                      ~ ~ ~ ' " "                                       ~ - ' ~ ~          ~
                                                                                                                                           ""^-"- "";

u , 10E-008 k'g k'g k'g k'g k'g k'g k'g Heat Exchangers (Design Not Specified)-Open/ Plugged 10E-004 I

.c                                                                                                                                                                   i N                                                                                                                                                                     '

3o 10E-005 _ _ Z o g <

,)10E-006            _                                                                                                                                   _

o L. LOE-007 Y'% k'% &'% *P'% Heat Exchangers (Design Not Specified)-Unso./Internc! Leckage 10E-004 r N o E m . . O h 5 o . u. 10E-005-Q ?b G.m-Yps g p'% 7'% g '% (g Figure B-3. Fciiure rctes and bounds for mechanico! components with various designs and ncrma! states. B-28

                                                                                                                                                                   /

Pipe (Pipe, > = 3 inches. O)-Running /E xter nc! Leal age /Puc ture 1CE-005 N e,1.OE -006 _ o O t 10E-007 - , u , e o , o ,

                                                                                                                                                                                                                                                             ~
                                                       ' mea.--5 5 10E-005                                                                                                                                                                                          -

u. 10E-009 .

q. ,s c, a(j n c n , c, s n su r sn es wv 8

c.,s 9 v o

                                                                                                                                                                        .w. V vn Q

s.v sJ s.s v .w. s

                                                                                                                                                                                                                                                    %J l                                                                      Pipe (Pipe, >= 3 inches. O)-Standby / External Leckcae/'R up t ur e 10E-005 r

N 1 e .0E-006 _ o I o m 10E-007 - , ,' CJ .i, , _b

                                                         '510E-005                 _                                                                                                                                                                          -

u 1 CE-00 3 a <, c,  %  % <x nj's (% (/ , (fes f[.f3 (M,$ 'fn _s#IYa, o J q v v J w a Pipe (Pipe, ( 3 inches. O)-Running / Externa! Leakage /Ruotur-tCE-005, z ( N

                                                           , ICE-006 o

f_ Z 1.CE -00 7 _ _o u_.u_..g , o , , , , o

                                                           .5
                                                          ~ 5 10E-003 '_                                                                                                                                                                                        _

t 1.CE -009 s .s A. scye. c. .. .a u c. . C.% %u jm % e ~fm -. *r- -yn ..'y-- -~ ~ r u J J J w a u i e e a -

                                                                           --- ---- ---_.---                                                                        ~~ - - - - - - - - - - - - - - - - - _ _ ~

e 1 I - - With vari.3us des;gns cra nc-rm ~i s t at-2 ; (c - M) B-29 k____ _ ___ __ ____ ___ ___ _ _ _ _ _ _ _ _ _ _ ___ _ _ ___

4

                             ' Pipe (Pipe, < 3 inches, ID)-Standby / External Leakage /Rupttre 10E-006
                   .=              ,h.

y {1.0E-007 - cc < 8 v

                                                                                                                                                               ..[.....--         . . . - -
                   '6 L

10E-009 4 Pumps, Diesel Driven -(Centrifugal)-Standby /FaHs to Start IDE+000

                   \ 10E-001                                                                                                                                               t L                                                                                                                                                              -
                   .e y10E-002 '-                                                '

l {t0E-003 -

g -.' . - .-

a 10E-004 - l LOE-005

                                                          %                      %                       %                                                  %                q
                         - Pumps, Diesel Driven (Design Not SpecMied)-Stanaby/FaBs to Stcrt 10E-G01 o

N f o g 0E-002 1 _. - g o o 310E-003 - -

                     '6                                       - - - - - - - - - - - - - - -                                                                 --                              - - - - - -

L 10E-004 k

                                                           %a                            %                                   '~';                                         h*7 A,*P                           4,.,                                  ' N, g

4 M;x  ?,, Fiqure B-3. Failure ro:es and bounds fer mechanical components wIth various designs and normcl states (continued). B-30

r Pumps, Motor Driven (Centrif ugaQ-Mt./ Fails to Run 1.0E-033 r N g 10E-004 _ ., o , o o , t _o_o_o_,,_o 10E-005 - __l__.-_..__.. ~ O , , _b <

              '5 10E-006          _                                                                                                                                                       ,

u. 10E-007 _ . . e a st e a g c,., e n . . e g <v,. e, n e.. e <

                                        %   ~ ,,rggw; 4 w-up g g up , 9 g gy@w9 w                         3                   v3 w,
                                                                                                                       -; 4y 1
                                                                                                                                                  . ww1 ws
                                                                                                                               - 9 v-y -4 q q vp '.y s

Part 1 of 2 (25 Pecorcs Total) Pumps, Motor Driven (Centrif ugal)-Alt./Fals to Run 10E-003 _ N I

               -eo10E-004           _

l E o o I e

                $ 1CE-005           _                                                                                                                                                    _
               '5                                        ,                                           . - . .

10E-oos _

                                                             <3,                 c.                            e                    e                          c
                                                                %v-
                                                                                     %v                          hv                    hv                          t, s

Part 2 of 2 (25 Pe:cres TctcQ Pumos Motor Driven (Centrif uga!)-Running / Fa:!s te Run 10E-00 3 N ., [ o j,,, 10E -00 4 o

                                                                                                                                   '          '       f                                    _

r<

                 ~ 10E-005            -                                        ,-+_             o-o_"                                                                                ,

e 2'

                 '5 ICE-GO6           _

L. 10E-007 n.N nN %n

• h

                                                   %nn N.  %

i c>

                                                                           %s Wb A%

b r, . n k M .'mr w# 'e s #D W q. e w' w

                                                                                                                                                                                .d
                                                                                                                                                                                   ~

n ' a 9 4 , 7 ,

                                                                                                                                 -~     nt e1 with various cesgis anc ncrrra sta r(

B-31 L _ ______________ _____ _

1 Pumps, Motor Driven (Centrif ugal)-Standby / Fails to Run 10E-001 a N 10E-002 _ _ 3O 10E-003 _ _

                       - Q: 1 C2-004_                                                                                                     o                                          _

o

                               )10E-C05 0                                      .
                               ' lOE-006             .                                                                                                                                ,

10EiOO7 Us

                                                                                                          %J                              %e                         %M gT                                     P 4                                4 Pumps, Motor Driven (Centrif ugal)-Alt./ Foils to Start 10E-001 o                                                                                                                                          "
                          .N                                                                                                                                     "

o

                                    ,10E-002                ,
                                                                                                                                        ,             u
                                   ~                                                          o g                     - . _

tCE-003 _

                                   '6 t0E-004                                                                                                                                           .,
                                   "                                                                                                                                                               I t0E-005                                                 _
                                                                 \k                              %                           %%%                                   L}            %
                                               - Pumps, Motor Driven (Centrifugal)-Standoy/ Fails to Start 10E+000 7
                                   \ 10E-001
                                                       ~

2 "4 1

                                                                                                                                                                           }

o jt0E-002 i [ o o e I - j10E-003 L- - - . - - -- w 5 ' L 10E-004 , _ 10E-005 . J J lJ kk $$ ,r h

                                                                                    **w                                                                                    %            i Figure B-3. Failure rates and councs f or rnechanicci components -

with various designs cnd normal states (continued). B-32

                                                                                                                                                                                                                           <t f

i Pumps, Motor Driven (Des;gn Ibt ip(:ified)- Att. /FoA to Run 10E-003 r N g

                     ,10E-OO4                                                                                                                                                         f                      _
                    ~

o __ a y _. 10E-005 - o _, e _g ,

                    ^6 tee-006         _                                                                                                                                                                     _

u. 10E-C07

                                                                                                                                                                                               &4,__
                                                    -q;                             -+-                                     .tn         1.m        0,     p,         +v,              v.
                                                                                                  + ;o, +g
                                                                 *,:le;                                                                                      gs,                        %
                                                                            ,~'

c. g, y

                                                                                                              't        Y
                                                                                                                                                                 \}                         'A
                                                                                                                                     'A                   y
                                                                                                                                                                                                          ~Q Pumps, Motor Driven (Design Not Specified)-Running /Faas to Rt.r.

1 OE-003 _

                     .c N                                                                                                                                                                                  f 10E-004 g                  -

hh

                                                                                                                                                                                        }f                      ,

a _ __ _ _ . . ___ _o_o___ . 10E-005 - _. o . . .o 3

                     ~6 10E-006                 <

u. _

1CE-007 n :n + +_ G 1 <o. 9 .f fy,+%+s's 1 , Q, gs &")<% ',x

                                                                                                                               +m, %g tW, 'g W,d.,RH3s,tA %. '.{Q++A 1 7 g.gt; 'Sf -f g 3'y4y%t r T7 4 7                                            v%,,q     P4 9     ' g -;

• y9 x; q My %7 Pumps, Motor Driven (Design Not Specified)-Standby'F/ ods to Run 10E+000 (1OE-001 v 1CE-002 _

e t >f_ c, , , , cd 10E-003 _

• 10E -004 _

{t f _.

                       =                                                                          t
                      =o 1CE-005             -.

r L 10E-006 _' _ 10E-007' . . . . g 4 *p laf c, &,, g g g, .

                                                                                                                                     ,Y                                                  .

[

                                                                                                                                         '?

g Wl ,n s.x'< A ..s vn , c. ,,

                                                              /L 'I %.l'.G. t[]
                                                                                                                           */ El                      I I       I
                                                                                                                                                                          ;p Q, l              "
                                                                                                                                                                                               'l             cf'I m;                                                                          > <                                -

x; v4 x, ,,  :, d y A k ' with various designs cod ner rnd stat - (c atin g:d j ,, B-33 i...... .. _ _ _ _

Pumps, Motor Driven (Design Not Specified)-Alt./ Fails to Start 10E-001 . t- 1 N I 10E-002 2 -f _. 2o e

                ,10E-003            _

o . l- _D_ ..,._a . . _ . ... . . _ . . . . . . _ . . . . ~ . . . . . . . . . . - - ~ . .

               '510E-004             _                                                                                                                                                                                4 L-                                                                                                                                                                                                        !

10E-005 g %* +'% Gg 4.4,

                                                                                                                            +

o 4  % M, 4,3, +M i g3,'% % 5% g ') *3;'% '%, '%, i'%, '*s4 Pumps, Motor Driven (Design Not Specified)-Running /Foi.s to Stcrt 10E+000 ,

                'O N
                   , 10E-001          _                                                                                                                                                             ,                  .,
                ~

o Z 10E-002 - o - o

                 ~g                                                                 i,
                '5 10E-003             -
                                                             " - " " - "                                                             ~~-
                                                                                                                                                                                                                  ~";-

( 10E-004

                                                          +%                       +%                                           +%                   +%                   +M'
                                                                                                                                                                                                  +%

r Pumps, Motor Driven (Design Not Specified)-Standby / Fails to Stcrt

                        'OE-001 t

N g 10E-002 o l o o , g " ' a_.o o 10E-003 - , o m o 0 u_. ....o ..o_ . "_ . . m . . _ ._ . . . . . . _ . . . . . . . . . _ . _ . . . - 2

                   '6 10E-004            _                                                                                                                                                                                 _

u_ 10E-005 . . 6p 6, q A -t A6 A 4

                                                            'N                    &+++Yp%*%+*4
                                                                                     '?n                                           t
                                                                                                                                                                                             :           h Port 1 of 2 (36 Records Toto!)

Figure B-3. Failure rates and bounds for mechcnical components . with various designs and normal states (continuec). B-34

Pumps, Motor Driven (Design Not Specified)-Stanscy/Ms to Star t

              ~ 1.0E + 000 V

N g 10E-001 - - 0 \ 10E-002 - , o o wa

          '5 TOE-003                                        _                                                                                                                                                                                     -

u. . _ . _ . _

10E-004 _ . . - r, 5 w,Io,>w , 1 u 9 4,,+x +v p s . q 2t,ly,L, .s r,4 n , sr - a to, o g xg ,g w.v tr w s,9 4 % tf 9,

                                                                                                                                                                 -f       4v 4                       's            1 7 y 4 % 44 v
                                                                                                                                                                                                                 %+        g     ,g Port 2 of 2 (36 Records Total)

Purnps. Steam Tur0ine Driven (Design Not Specified)-Standby /FaHs to Pun 10E+000

           \ 10E-001
                                                                ~

o ., 1 0 10E-002 y o ,

             *10E-003                                             :

2 2 "

            ~6 u ICE-004                                             -                                                                                                                                                                                  -

TOE-005 "y

                                                                                                                                         %                      (%                         '!c,_

9 t 3.

                                                                                                                                            %,?f "I                                        v UI uf Pumps, Steam Turbine Driven (Design Not Specified)-Stancby/ Fails to St:rt 10E4 000 t

N

               ,., 1 DE-001                                                                                                                                                                                                                            _
             &                                                                                                                                                            f 1GE-002                                            -                                                                                                                                                                                _

o 4

              .=1                                                     .                                                    ,

o 1CE-003 .- _ u. 10E-004

                                                                                                                         +

M, in e

                                                                                                                                                                           ". G s

J. b l--

                                                                                                                                %7 g                 '<                                                       s                    -4,
                                                                                                                                                                 -jD'J                                                 I        ')
                                                                                             = _

with various des,igr.s and norrnci states /cy t r -1). B-35 L____ _ _ _ _. _ _ _ - . _ _ _ _ _ _ _ _ - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _____ _ _

1-Pumps. Driver Not Specified (Design Not Specified)-Standby / Fails to Run 10E+000 ( 10E-001 _-- t etcE-002 _ _

                   -e 10E-003                      _-

0 10E-004 - _

                    ~h
                   -:gtoE-005                       _                                                                                                                                                                                              _

L TOE-006 10E-G07 > As%  %'%  %  %'%

                                            .Volves, Check (Motor Operator)-Unsp./ Fails to Operate 10E-00t t

N 3 10E-002 -- ,- 1- -t -. g _ _ . _ . . _ . . - . , . _ . . . . _ . . .

                                                                                                                                                                               . . . . _ .. j .

x 4 10E-OO3 _

                                                                                                                                                          <                                                                                                      4 o                                                                                                                                                                                                                                         i 2
                     '6 10E-004                         _

u. _

10E-005s . .

                                                                                                                +                                                           J g             #g              #g        #g             #ro
                                                                                                                                                         +3 4               '             'e              %         '4               N i
                                                              -Valves, Check (No Operator)-Unsp./ Fails to Operate LCE-001
                      'o
                     \ 10E-002                             _                                                                                                                                                                               , -,_

y10E-003 ,,

• 1.0E-004 ' , . . I g ,

                                                                                                                                                                                                              ~

O10E-005 L. _. . . . . _ .. . . . . t0E-006 k k k k + i- + k k & + k + +' k ++r+

                                                                                      '%'%'%'%'%'%%'%'%'%+'%h;.'%                                                           7                                   7 7 S 'h'%'%'%'%'%

i Figure B-3. Feliure rates and bounds f or mechanical components with various designs and normo; states (continued). B-36

I Vo!ves. Check (No Operator)-Unsp./interrd Leckage 1OE-CO3 _ r

          \10E-004                                                                                                                                                                                                                   ,

o Z 0 10E-005 ~,_ o

• 10E-On6 l

2 1

          '6                                                                 <.

L 10E-007 -- - - - 10E-00S

                                                                                                               +
                                                                                                                                                               +                       +          b       k             k
                                                                           +%s+'o      _ s +% %'9 +%'0 %,
                                                                                            ~
                                                                                                                                                                              +% 's                  %%M                       s +%

T7 ty k 7 7 .'7 ^d7 I7 My ty % '97 Valves, Check (No Operator (tilting disk check vaive))-Unso./Fa!!s to Operate 10E-001 TJ N g 1.CE-002 - , o r.r ' 1.0E-003 - , o m o o _b.

           '6 1.GE-004 - _ . _ . '                                                             . . _ _                   _                                                                                                               .. ,

L 10E-005 4 +'o , -

                                                                                                                                                                                         +               +                  +
                                                                                  +%s                        '4                                                         W                                   %                   'c.

U7 %7 '& %7 'o&

                                                                                                                                                                                                ~7             k                    &

Volves, Check (No Operator (tiiting disk check vc!ve))-Unso finterna! Leckage 10E-OO2 N 10E-003 _ o

             -o1CE-004                                           _                                                                                                                                                                            _

Z 10E-005 _ _ e u o 910E-005 _ o , _ . . . .

             ' 10E-007                                            _                                                                                                                                                                            _

10E-008 _ _

                                                                                 +                  +%,                            +v                          +                 r               +g            +                +

w

                                                                                         '07 w                 s._                          w                 w'd; Yl                                       '*ij           %*                hj               <              My

e. " + 4 b

                                                                                                                                                                                                                    .                    ')

With /GnOUS ue Sigr'G Cnd nOrmai States (: Or,tir d B-37

Volves, Check (Pneumatic Oprotar)-Unse./ Fos to Operate toc-001 -. o 1 -- N ' I 4

        ,tuE-002         -

q 15 y ', . _ . <, --, 10E-003 y , e u i

       '5 1.0C-004        _.                                                                                                                                             .,

a  : 10E-005 h i + 6' + &

                                                                                           %,                                                            %,        'v _

w% 4 w% y,  % $ Volves, Check (Pneumatic Operator)-Unsp./ internal ceokage 10E-001 ,

m. ,

N 1 OE-0% , - e

        .W 2C 10C-003
        $10C~004F.: .                                                                         .

__ 9

        ~5                 '

u 10E-005 ,

4. .

10E-006

                                                                            +                 +                        +           +              +            +

$. w% w% w% w% w% w% i- } . Voh'es, Check (Operator Not Specified)-Unsp./ Fails to Open 1.0E-002 t . N

          ,10E-003                                                                                                                                                  ,       -

Y LOE-004 , . o

         )                                                               '

5 '6 10E-005 , u. 10E-006 , N31 # 8's ly I k D $17 'i+  %  % Fiaure B-3. Failure rates and bounds f or mechanical components 5 with various designu cad nor:nal states (continsed). B-38 l

                   -           __-______.___ma-----_m. _ _ _ _ _ - - _ _ _ _ _ . _ . . -

l Valves, Check (Cp:t oter Not c'pe:ifiedbUnrp /!nterni Le's age C 10E-002 t q 1.0E -003 . _ S

       .v10E-004                                   -                                                                                                                                                    _

i o tt 10E -005 _ _ o w 100-006 . f _

       $ 10E-007           _                                                                                                                                                                            -

LL. 10E-003 - _ 10E-009 . _ . s s. , c 1. MR v,. x;R 4 4,, ty 3 ",' w t.

                                                            ~1                                                                 %;

J i' Valves, Motor Oper at or (Design Not Specified)-Cpen/'Fa% to Cbse 10E 4000 7 N o o 10E -001 _ e E ii o

         .D10E-002           _                                                                                                                                                                             _

h

u.  ; _-

10E-003 . . Q .. t<<, Q~, t?, 1. m, 9_ g, ty, g, "9

                                                                                                      '9-
                                                                                                                              'v                                      %

. Valves, Motor Oper ator (Des;gn Not h. rcified)-Unsp./F ai!s to Operate 1.0E -00' N 1CC-003 _ _ 3 10E-004 o _ _ CY A k 10E-005 _ i

p 2 1 CC-006 .

                                                                                 <                      o-                                                                                                   _

y , o

          ' 10E-007                                                                                                                                                                                          _

1.CE-008

                                        +              +         +              +N        +           +             +           ",,     +                                             4
                                           %                                                'n_

vg ' <i .., 4 ' .<, ,

                                              ~i              $f                                  $j                        Lt                                    A,
                                              'f                         $f          '$f                       ,e                  'C'j                                                                  t Figur e B-3. Failtre r ates cm) bot..ncja f or tr:e n.am: y : oreponent:

wIth various desi;;ns and ncrrnal st at es (c or= tins-d) . B-39

Volves, Motor Oper ater (Design Not Specified)-Unsp./ Foils to Operate 10E4000 V N 1 R-001 '_ t q e f I t E 10E-002

             'c'
                                                                                                            -I          }                                              9

{ 4 ( ,

             .{ 10E-003                                  t
             '5 w 10E-004        -                                                                                                                                         ,

10E-005' . k e  % 4 +

                                           +w                %A R 4 +w 4+% 1+wk+% +w& +w'k . %.N w,N N1                f                                                1 g                       -------.--                                       -----------------.--_.---------

Volves, Motor Operator (Design Not Specified)-Closed /Folls to Open 10E-001 7 N a 10E-002 _ _ o o Ng . . . .. $ . . . - . . . . . _ . . . . - . . . . . . . . L l 10E-003 f,, G G 4. 9,  %, 'ny v t, '%, R,'%  % Valses, Mcnual Operator (mechanical hand 6 heel) (Design Not Specified)-Unsp./ Fails to Opera 10E-005 r N

                , tot-006        _                                                                                                                                         _
               ^o 2

10E-007 . , y ,

                                                                                  ,           . o
               '6 10E-003 L.

A' i -' - - 1.0E-009

                                                                   +
                                                                                                                                     + ++
                                           +'q +% % +% +% +% +% +% +% +% +'q % 'q                                                                                   %

Figure B-3. 7 ailure rates and counas for mechonkol components with various CeSigns and norrnsi states (continued). B-40 l

I Va'ves, Manual Operator (mechanica; hande,heet) (D-shn N c,t Er+:;f'eJ)-Ung 6 4: t o 0; ' at e 10E -OD I V N 10E-002 , m

                                                                                                                                                                                       }     }     ,

o o O10E-003 cc l , 1

c. o I

                     . OE -00 4                        _ _ _      ___        __
               ~6 u.10E-DOS          - ,              ,

1 GL -006 .

                                         + + +. + +
                                                                                            "           p.                  F + + .+ + + + ++

o . ' o , es i e, _ ',, 'o

                                          'o.
                                                                  '4 'o&j          '4 k tj y, 't, '2^%; 'o.Ty' ' 'dy
                                                                                             % '$,                                .r s     ,
                                             '97            r,Qr        'C'7                                                                                        ?.?j 'My        'S; 87 My
                                                                                                                                -t Vc!ves, Pneumatic Operator (Design tbt Specified) -Unsp / Fails to Operate 10E-004
                .c N
                  ,10E-005         _                                                                                                                                                                -

o 10E-DO6 , I v i g _ 0 10E -007 _ tt 10E-008 . . . .

                                                                                                       +                            4             +           +             +Y,.         +
    }                                         +'4                   +'4.           +W                            %                       %           'et         '4                         '4
                                                             %              Mj            My                              Mj               %;          'd;          'Sy            %;          '$j y+3

t. e Valves, Pneumatic Operator (Design Not Specified)-Untp f Fa;!3 to Operate 10E+000 m

N 10E-001 + f , o t f

                  ]10E-002                                                                                                                    '

I ' k {10E-003 _ , I ,, 4

                  .-o u 10E-OO4            ,

o 10E-005

                                           +,                 t           %o    d.s
                                                                                              -                           -           +       +          +       +          +         '+

o_ o. My 6' j S , v,

                                                                                         ' ?V ~

g

                                                                                                                   <            .,hy 1

c.

                                                                                                                                                              %l %

a. -

                                                                                                                                                                                  'Lj u
                                                                                                                                                                                           ~~dy Y;

Fiqure B-3. ;c;!ure rctes and b:urc 6 mecnarcta u c ercone is wiin varicus designs and normal states 'cor tinuen B-41

Volves, Pneumatic Operator (Design Not Specified)-Open/ Spurious Cbse 10E-004 r N

              ,10E-005        _                                                                                                                                                                 .
             ~

o 10E- 006 - - o ' b

             '510E-007          . . . . . . . - . .              -. . a     .

L 10E-008 5 Ye

                                                                       %                                  Lp              'h.
                                                                               '%                           'g               t.,                                                4 9,g Valves, Relief (Design Not Soecified)-Cbsed/Fa:!s to Operate t0E+000 t

N 10E-001 _ o a " 10E-002 . o b

              'O 10E-003          -                                 1-                                                                                                                            ,

L _ .. p i  ! 10E-004 6 & k 4 eg'%

                                                                                                      '%             %g             Mg,,                                            Mg Valves, Relief (Design Not Specified)-Unsp./ Fails to Operate 10E-004
               .c
               \1 C'E-005                                                                                                                                                                          ,

e , o j 10E-006 , , __{10E-007 ,

               '5 Li.1.0E-003 L_                      .. .. . _ .. .. ._ , _ _ , . . . . . . . _ . . .                                 . _ . . . . . . _ _ . . , _ . . , . . ,

1.OE-009

                                                                              +                             +                   +                                             +       +
                                              +4 +y                                                            % +%+% % ww
                                                                                                                                          -r w%y+w 49 2 a%+%          7                        'h %+% 4 'th %, x; %;                                                         a;       My

• Fiqure B-3. Failure rates and bound for mechanical components with various designs and normal sto-tes (continued).

B-42 ) 1

i Vah en. Rehef (D % o tbt E;u;tr 0-Unt, -f,; to 0;-, : at. 1OE4000 , V x

         ' 10E - 001         .                                                                                                                                                                               .,

v o 10E-00 ct m _ . _ . _ . _ _ _ ,..__o. _ . <,_ _ _ _ . _

                                                                                                                }              _._.          _ _ _      _ _ _ . _. _ . _                . _ _ _ _ _ _

gf 10E-003 < , n i 10E-004 ., 1 OE-005 - t

                                         '4   .

4-' 4-

                                                                           + + + '>+' +

4-e

                                                                                                                                                      +' +             si
                                                                                                                                                                             .t~            t +'

32 f tj Q % if ,f tf :f uf ~i:f . Volves, Pehef (Design Not 5peu. CbseJ/ int +:- N! Led 07? . 10E-001 f 10E-00: N q,10E-003 _ , m o _ E 10E-004 _ cr 1CE-005 ._ v

        %10E-006 x

o 10E-007 2 - a L 10E-00E _ 10E-003 4 x, n L .. e e u

                                                   'c                 ga               4                  s                   .>,                 ,             'o                      t Bf                              %                  tf                tf                   tf               t,,                 M;,           _.ty Vanes, Sciencid Operater (Docign hot Specified)-Untp./r8 to Operate 10E-001 c

(10E-OS2 _

         .v ,

1 OE -093 _ o c2 1OE-004 _ c'10E-005 ._ _ 3 " 610E-006 ,_ k 1OE-007 _ 10E -003

                                                +                   + A
                                                                                     +                .c!
                                                                                                                          -                     n             4                   -

d,_ Iy_ M  %

                                                                            }                 4                  'I               %*                                                          i             i
                                                                                                                                                                                                               ,-       i wit n vo'ious denigns cod nor m; state- i - - -r t '

B-43

Volves, Solenoid Operator (Design Not Specified)-Unco / Fails to Operate 10E-001 o N '

        ,10E-002     .                                                                            ,
       "o cc 10E-003    .

1 e

       .5              . . _ . . . . .
       'S 10E-004    .

L 9 i0C-005 _ _ . u u + + +

                                                       '*g       '4g +'9;4 'og +!;9 '9    '9
                                       +'%                                              4       4 Figure: B-3. Foilure rates and bounds for mechanical components vath various designs and normal states (continued).

i M s i B-44

B a t t er :e <. (Wet CM (e g,, leM- a ;10)- ' tan Odf

                                                                                                                                                        ~
                                                                                                                                                                                        ;              to 0; or ate ict.004 o
           ,,1DE-005               ..

o LL ' 1 DE -006 _ _ . . __.i . _ -_ - _ _ . _ _ _ _ . . - _ , v 6 10E -00 7 G. 1 CE -00$ . _

                                                 <                .,                  b,                    qs                   e                         ,           e r.           ".,                    s .,-
                                                       ,t.                                                                                                                   ,          v.

r*.)t . t g v:7 r.'; y b,y 1 Char ge , Battery (C har ger , Bat t er v )-E m r gc c CT ans to 0:+ rate 10E-004 N

                                                                                                                                                                               q                                        q;4 .

14 Circuit Breakers, Power (Design Not 5;:>rOf sd)--Unsp./F ahu to Cree ate 10E -00 ? n o }

• 1DE-003 , _ _ _ _ _ . _ . . .__ _ _ __ _ __._ __ _. . _ _ _ . . _ _ --

O cr

             ^ 10E-DO4                   ,                                                                      ,                                                                                                             ,

qi _5 ' 6 1 CE --005 - - u - 10E-006 4 t. ', a i, ""

                                                                                                                                                               %6' n
                                                                                                                                                                                                'Y
                                                                             ,,<?                                           -l
                                                  .             J   N,Y                                .       4               L-'      .
                                                                                                                                                 ,                                      m     .A           a                         'E~

mOntation components with veren & R,m ana - r - f e t ~ ' -f c . B-45

l-l i. l- Circuit Breakers. Power (Design Not Spccified)-Closed / Fails to Open , 1OE-00i _._ !- x t i-1OE-002 - _ i O i 10E -003 _ l e ~1 l ja . ! '5 10E-004 . _. j a . . . _ . - _ . . .._.. _ . . _ - . - . . . . . . . . . i l' 10E-005 e.J l 4p %ass e 1 a v J'4 I

                                                                                         'R,                                   N                                                    At l'

i Circuit Breckers. Power (Design Not Specified)-Closed / Spurious Open 10E-005 i r g ! 3o 10E-006 _ _ Z 4 0 <

                                                                                                                                   . ~ . .          . . . _ . _

i _D 10E-007 _ _

'5

a 10E-008 . . f

                                                                        %g'                                                                                        %ss              #*q
                                                                                  g      %'% b)g %,'%                                                                ,M 9          94 1                                        - - _ _ - - _ - _ - _ - - - _ - - - - - - - _ _ _ - - - - _ - . - - - - _ _ .

i i. i l Conductors (Bus)-Energized / Fails to Operate 10E-00$ r I l N g 10E-006 _ _ w o l ^ 10E-007 _ < , m . . ! b _ . . . . . ~ . . . . . . . l '5 10E-008 _ , i k l l TOE-009 by '%g  ?> Ye

                                                                                                  %p'8     '

u l-

                                                                                   %                               by                       %.

g l. rigure B-4. Follure rates and bouncs for electrical and instru-rnentation components with various desians and normal states

                                                                                                                                                 ~

(continued). i B-46 \ l

                                                                                                                                                                                                         ?

I' l^

Cer,erator, with Dieset Engine Driver (Ce ner at or )-S t andby / F alis to Pun 1CE-031 N v

                        - 10E-002           .                                                                                                                                                                           _

o Ct p --

                     )10E-00T                                         <                      <
                     ~5 L.L.

I CiE -004 eut. (9  % - 9, s, Q, 'o,  ;

                                                                                                                                                                                                  ' ~
                                                                                                     '/'                                                       * .                                    '4 4

4

                                                                                                                                                                      --.--_.-.---.'4                                   - . . -

Generatc<, with Diesel Engine Driver Wno ator )- St andby/F ok to Run 10E-001 7 N 10E-002 _ _

                      $o                                                                                                                                                   .
                                                                                                                                                                                                                      ..~

10E-003 _ , <

                                                                          <    <    '          o v                           ,

N 5 10E -004 . L. 10E-C05 _ _ _1 4s _ I K1 h % h h e. n h '4(0 1.1, ;I. !s : <. : h io

                                                      R,5(3,$

N N 4 N f s y4ygyny4 s s . s c'4y1s

                                                                                                                                                            -   p      s 9 0

4 s c

                                                                                                                                                                                 '4   y R,sQ s       s
                                                                                                                                                                                                  '4e Q'4p s          s
                                                                % % % % % % % ty % %s% % %'s% N9 hy %stt%                                                                                                          9 y Part 1 of 4 (70 Records Total)

Gener at or, with D:esel Eng;ne Driver (Generatcr)-Standby /Fals to Run 10E-001

                       'o N

2o I vE -002 - l

                                                                                                                                                                     ._o_"_" _ _"__"_             , ,,                , _
                                                                                            ._ g_m_.o-__m_o q      li     j       4 e

e J._ 10E -003 O L& 10E-004

                                                                                  . 16 'fi 4h4y4-;-Q"y                6 1- le fe                                     b in !E,W.'g    ic is A         l . !<

is 4t h 4 p d e.

                                                        gfR,s s

M(1-;'4(M s cp s s s <. s

                                                                                                                                                                   ;'q.;'q;'Q s    s      s    s     s s
                                                                                                                                                                                               ,*?yx-;'49s            s f ??

f +mp s.f.9 h sh 3 ' ip r')p f(9 h %f fh . p 'yh ,^-f.y Iv"'s,p m'g fr ~p h h'-@ h-gsh '-9sh ->p

                                                                                                                                                                                                                      .)s [

9 19 Por t 2 of 4 (70 Records Ict1) FIOure 8-4. f 0LT e r Otes On d bOuni for Oe,;trP.01 and iv u - rnhntation corrpOnents with vari?u5 deCiara c'nd nOr rpal state ' (continued) B-47

Gener a t er , with D'ese! Enyne Driver (Gener at c r ) -S tancr y/ F cE tc R e n 1OE-001 y N 1 o , o o , o Lt' , , 10E-002 < < o - v - _ -- - .... __

           =

5 o u. 10E-003 7 n A -1t Ir : , A 1., A :s A i, is i t hAAAL 7c ,uggm,9. p4g9t, w_,sqqq.qq s e, .s s s s s s s sq. sqqv..s1.7q<q,949,9., .,7.,

                                                                                                                     .st ..s .s s q .s t s .s s s..s b-g F      ' b h,q h h h bh b bhg y g (,9 [.9 F9 bbh
                                                         -g 9 y 4 e g 9                                                                                                   -e 9 9 (.9 -9            h Port 3 of 4 (70 Records Totoi)

Generator, with Diesel Engirie Driver (Generator)-Standby / Fails to Run 10E + 000 t N

            -O o

10E.-301 - _ Cr o "

                                                                                                                ,,                o o

w. 2 10E -002 _ t L. 10E-003

                                                 !           in              4                   y              h                 1,             'i ..      -i n           %            %
                                                  $(s hp        s 4c My M c s               s
                                                                                                                                                   " ,(        R.,

s Mc s Mp s b+ N(s.9 b ts ts te s^ v ,^9 h n9 be te Part 4 of 4 (70 Records Total) Gener at or , with Diesel Er.gine Driver (Generater)-Standbj/ Fails to St art ICE _001 p N

               ,10E     N~"

a _ _ _- _ ___ __. _ _. _ __ __ __ _ _ cr 10E-003 . o m o , , o

                                                                                                                                                             ,              4       -

o ,. , o < g

             .=                            <
             '610E-004 L.

1CE-005

                                          % h h in % in % % 1, in                                                               l~ l -. l-             iv l~ n %l                       !, lc 44+t.N.fqqvg94N44 t   ,ss      C s   s ., .s                s
                                                                                                      's v h, s
                                                                                                                  'q,-Jy.cwow-t,s",q,N,          , t                      s ,s    i cv, *g ;,s s         .

Y~. h b b vg '@ bp -p $ v.yb , k f'y. b..E;, fi h b , F b -.9 'y % ', h.y '* s. h$ 9 *f d *b b y ' f '- 9 $ PCr t 1 of 4 (75 Fecorc:, Tota ) , Eigure 8-4 IC Iuf t rotec Cnd b%nd fcr et - tri' Q! Cnc i istr U-I rnent a tion c o mpone nt s ,vi t n earh: as cesdna c 1 r-:rnici states ' (continuec) B A8

Generator, Mth D, i el Enjne [riwr (E ' f- c t o )- S i c"U , -T $ ^ t u. i t : .- t 1M -001 o N c' o

  - 10E-002 o                                                                                                                                                                                                                                                          _

1

                                                                                                                                                                                ,             ,,       ,,,i,-n-                                         .-

h ' q q q. l q l w < l

 ._:.e. 10E -00 3  _.                                                                                                                                                                                                                                         _

o a i 100-004 i, -:. h. --:. y i- :, % --i 1, : . 1 - l 1. , 1

                           4g'q ;,[nf,'4,y,*:jg-lr.

s s s s p  % s L.1_[<.% s s t .[, s

                                                                                                                                                                                                      ,y s

ad  : ,- b ' y * ' 4: p wp bb

                                      .h
                                                   -9 99c ~$. 9h s.                      rg h . 8y.                  sE    + $ p;9 4. p9 h9 p,y -9                                          n 1 1, h t.

_ y 3 , ,y . .s %.7b y Port 2 of 4 (75 Fecords Tota 9 i Ge ner at or, wit h DieEel Engine Drier (Gerrer a t o.-)-S t on@v / F ak to Sta t - 10E-001 -__ O N 3o 1OE-002 , , , , 1 " ' ' ' '

                        . . .. ._ o_o_< _< _<                                            o                                              '           '       '

9 A.

   .s10E-003        .                                                                                                                                                                                                                                        ._

o - w. 10E-004 . _ _ . _ _ _ 1 1_ , , tm : 1, % :. % y

                            'ggy%.y43       h, hLy4g4g.n           1y h. iy,3 3;:.,- .!,t v;:4y.y:4fp, ,93gg;> z.;y -.

s s s s s s u s s s s s s s s - s .s , s b-pb,9 t-@s h-g p<p F.y^ 6y p9sg F wp b b h ph b s-,9h b h

                                                                                                                                                                             ---p 9 -9                                 9 -g ' p 9 y rn or t s of 4 (7cs nec or cs t ota!)

Gener at or , with Diesel Engine Driver (Gewr at or)- S t andby/ Tan to S?'ri 10E-001 T2 il z"; ot _.00 _ , , , e , o " ' o < - 1 h 6 _s10E-003 _ ._ o L 1GE-OO4

                               .<          <                           s               ,                         <                 e              .           .           ,                 .

C..r 6O #4, 'i $ ih 6.,, a* -a* se .*

                                                                                         %t 'f"[q af',[3 e T.                                                                            e                                                                           e
                                  'Y S       'y q               V
                                                                                                                                                    Tg           '7 ,        #              '             '
                                                                                                                                                                                                                                                  'f . 'b7_

s s s s.- ,'7 s 5

                                                                                    .s                    s                  s<

x . u. s - f , _ f b .( b, [s

                                                                                          ")

hs3 sb.3 , .3

                                                                                                                                     -1

[

                                                ,9              ny                          $ 3                                     9,.             -3             3                                                                               ; . . $,            4 m,+

• c

                -_-----_.._____)_

a Y. t n v (s n ( s - _ _ _ ' e:>_

                                                                                                                                                          ,, +; u- ar .s
                                                                                                                                                                                                     ;g a                                        $                                                                                                   E rnent atiOn .OTOCnonts with car c;

• _ ; w c;r- 1 ny m _ :, t j t p e, (Continued).

B-49

P A e' E kf c tr c n:cs ($0 lid-state) (!rh erttr (rot mat c r c t r ir '1)-[rtr R e d JT = t- '

                                                                                                                                                                                            +" .

1 k -D33 r

        \ 10E -004                                                                                                                                                       ,

v 0 10E - DO S - R , ___o_ _. _- .-_ .

                                             ,          4 10E -006     -
        '6 a 10E-007                                                                                                                                                        ,

10E-003 * < , . .. - a . # 7.'s,%c- ,L., t. ';2 t 7 -; , 1 ,.- > e.5 m f y., g' ,

                                          ' 'Q       'D., ' ',f;4-1        ,,

Lf L/  !+ v s -

                                                                                         'r   ,

5, ,7 Pc Atr Electr on cs (SMd-state) (inv ..rt- (not motor c tr ir .))-E ner p:e 3 /r a> e te Oper at. 1 OE 4 000 t N 10E-001 .. _ o CE ' 1CE-022 ,. e __ -_ a

          '51DE-003 g                  ....

10E-004

                                                   +
                                                     't.,.
                                                                       +                                    +'A                      y b,
                                                                                                                                                      +

y 4 ty 'd; 'c:7 Fe!cvs. Pr atet tive (Duir,n Not Spedf Ld)-Unw /Ftc to Opr ;ta 1CE-005 E r N v g 10E -OC r: _ _ G. , e - 0 S 10E-007 _

           '6 L

1CE-00*

                                                     *,q                        ,

i 'v'g

                                                              ]                    '5                                h-ry                      r                1y-i ]Ur e * '- 4,               f CI!j e fCIT$                      Tj'>d Dv UndC i C * +fC'f'1C Y C "c j I"4S t f U -

m entatiOn cOmLcrants wit: v 7I: C Ne ign. p1 'm O' 'T o  ? JtC5 (Ccntinued). B-50

Transduce,3 (Detec tc r s/E 'ements /Ser>sor s'i (9nOr s)-E ner pre 1/Folc tc 0;4 t ate 10E-001 T) N g 10E-002 _ q o 1 10E-003 - v i

                             -.5                                                                 1                                 .
                             '6 10E -004           _                      ,

u. 10E-005 c e i

                                                                        'tby                    'picy                      %, ,            9.,                m;y
                                                                                                                                 ,                                   ~,

g. v4 14 t4 -:4 Transducers (Detectcrs/Ekmento/ Sensor s) (Sensor s)-Energked/Epur ice, Oper ation 10E-005 r i N j T q, I

                               - 10E-006                                                                              -          -

o o 1 { } _- a , T y , _ s

                               .?10E-007 o

tw 10E-005 s x & n s"q s"Or s'b i "q 3'b ~ 'O , j'b , y":>~"O .'.O t s. , i,4; 4 e4 g u4 g_ g .e4 rp 4 Transmitters (Trcosmitterc)-Cpe ./Fons tc< Oper ate 1DE-005 N v g , m < 10E-006 _ v < 4 o La 10E - 00 7 __ n .

                                                                                        ,                 n                   m          n           r
                                                                     *gg,                              v
                                                                                          'g3              c3                 %.,           ,          , , ,,           ,
                                                                            'A;,              ~,.                                                                            ~<

k 7 ' .g '%'.y J 7 ., Ji, P

                                                                                                                                                                                  =4 Figf e B-4                       FoHure rates and boundi or e ectric ci and instru-mentation components with various Osans and rcrnd states (continued).

3-51 (

Transformers, Power (Design Not Specified)-Energized / Faits to Deerote 100-004 r N g,10E-005 ,

          ~

o Q: q, 1.0E-006 - -

          )                        .
          '5 10E-007                 .                                                                               .,

u. 10E-008 .

                                                                            ^
                                                      'I                                        WeJ QJ                                                Q
                                                               %                                                O, r,

8g Figure B-4. Failure rates and bounds for electrical and instru-mentation components with various designs and normal states (continued). + A W 4 i B-52 'l t-

                                                    -~ - m - m
                                                     . -w ,,. .

APPENDIX C AGGREGATION METHODS

CONTENTS C-1. GENERAL APPROACH ............................................... C-2 C-2. AGGREGATION CATEGORIES ............................... ....... . C-4 C 3. AGGREGATION ALGORITHMS ......................................... C-7 C-4. EXAMPLE OF AGGREGATION METHOD .................................. C-10 C-5.

SUMMARY

OF AGGREGATION METHODS ................................. C-12 C-6. REFERENCES ..................................................... C-13 TABLES C-1. Data Set: for Aggregation Methods ............................. C-6 C-2. CFD Aggregation Algotithms .................................... C-8 C-3. Example Data for Battery Chargers ............................. C-9 FIGURE C-1. Battery Charger Failure Data .................................. C-9 C-li (

i APPENDIX C AGGREGATION HETHODS This appendix provides a brief description of the methods implemented in the NUCLARR program for treating the uncertainty in the failure data and for aggregating or combining data from different sources. Understanding how the data are aggregated and how the upper bounds are formed will help the user to better appreciate their use. Two main quantities result from the aggregation methods: a point esti-mate and an upper tolerance bound. The point estimate is a weighted aver-age of the component failure data contained in the set being aggregated. The upper bound is a percentile of a distribution that is intended to des-cribe the population of failure rates or donand probabilities that are being combined. it is not intended to be a confidence interval describing how well the mean or median of that distribution is known. Rather, the tol-erance bound is a number that the failure rates or probabilities are ex-pected to remain less than with a probability specified by the user. In this sensa the interval from 0 to the bound is intended to cover that spe-

cified s ' aunt of the distribution of the failure rates for an event. The specified probabi;ity is thus a coverage probability rather than a confi-

                                                                                                ^

dence. It is useful for predicting where future failure rates or probabi-lities sampl_d from the population variability distribution might lie. A further remark about the tolerance interval is warranted. A classi-cal statistical tolerance bound has two probabilities associated with it: a coverage probability and a confidence indicating how likely it is that the interval actually does cover a distribution as desired. These bounds consider the sample size and the fact that parameters influencing the tol-erance bound are not really known but are estimated from the data. Atwood -1 describes a method for correcting tolerance bounds to C account for the uncertainties in parameters estimated from the data and C-1

thus for increasing the confidence. However, in applications with diesel generator data, he found that the correction had little effect. Based on adoitional experience with the method, he concludes in Reference C-1 that, when the sample size is at least moderately large, the uncertainty due to lack of data is small compared with the inherent variability in each fail-ure rate population. That is, generally, the variation between records dominates over the variation within records and thus, these corrections are not needed. The following sections provide an overview of the aggregation methods, including the major formulas. For a more detailed presentation, see Refer- - - ence C-2. C-1. GENERAL APPROACH The component failure data aggregation algorithms are based primarily on the treatment of generic data described in "On Combining Data for Esti-mating the Frequency of low-Probability Events with Application to Sodium Valve Failure Rates" by H. F. Martz and M. C. Bryson.c-3 This arti-cle discusses the combining of various types of generic data to form prior distributions describing population variability. Although the article does not discuss the per demand type of rate, the methods for combining data therein are easily extended to this case, in nearly all cases, the variability distributions are taken to be lognormal. This distribution is discussed in the PRA Procedures Guide.C'8 Apostolakis et al .C'S make use of lognormal prior distributions, as did the Reactor Safety Study.c4 It is appropriate for quantities that vary by or- , . , ders of magnitude. Although it is not bounded, it is appropriate for proba- l[ bilities as well as rates if the probabilities are small. The aggregation f, methods include adjustments for the lognormal distributions that model failure on demand probabilities that approach one; the output legnormal C-2

I I distilbution is checked to ensure that at least 95% of its values are less than 1 and a truncated distribution is used if this test fails. The data treatment discussed in the Martz/Bryson article focuses on the type of data supplied by the various data sources. for generic data, there are two main possibilities. The data may be raw data; i.e., histori-cal observed frequency data on similar events in similar applications. Al-ternatively, the data may be reduced, in the form of point or interval esti-mates for frequencies of similar events. In the latter case, there are sev-eral possible ways to specify bounds for the rates. Tolerance bounds (for one or two-sided intervals) may be specified, or error factors may be - given. These methods have been extended to allow standard deviations or variances to be used to describe the variability. For all of these cases, the underlying hypotheses are that the informa-tion froni each separate source (i.e., individual data record) has a within-source variation and that the goal of the aggregation is to provide a point estimate and a tolerance bound that descrites tne between-source varia-tion. This motivates an empirical Bayes (FB) treatment of the data in which the within-source variation is modelled by a conditional distribu-tion, conditioned on the particular (but unknown) value of the occurrence rate for a particular source. The (unconditional) distribution that this value comes from is the desired population variability distribution. Tol- _ erance intervals for aggregated rates are thus based on this distribution. Further remarks on the nature of the assumed within-source variation are appropriate. For raw data, Martz and Bryson assume that a constant failure rate applies for each source and that the within-source variability is thus the variability that is characteristic of a Poisson distribution. For reduced data, likewise, the bounds or other measures of variation for euch source are assumed to describe the variation of data for that source. However, in this case the data may have already been aggregated by the data supplier and may already reflect population variability. This is particu-larly the case for genuic component failure data; some individual data C-3 (

4 records may describe aggregated data. The aggregation procedures consider this distinction and do not attempt to remove within-source variation in such cases. In all cases the point estimates recommended by Martz and Bryson are the geometric means t.7 the values from the individual sources. In some cases, the means I.T8 e19M ed inversely according to the variances of the individual sourcet. After combiring m iggregating various sources based on the form of data provided, sm erd garegates may result. Each of these is character-ized by a lognormd (mtribution with an associated point estimate (its median) and upper tour.mco bound (related in a simple manner to the vari-ance of its underlying unnal distribution). To form a single overall ag-gregate, a mixte e d btrthution is formed. This, in turn, is fitted to a lognormal distriktion by matching moments. The individual distributions are weighted by the number of individual records contributing to each aggregate. < C-2. AGGREGATION CATEGORIES Following the basic aporoach of Reference C-3, the set of data records to be aggregated is first split into sets based on the type of information supplied (whether a data record describes homogeneous data and whether it contains raw or reduced data). An aggregate is formed for ear.h set that is present among the data being considered, then the resulting aggregates are combined into a mixture distribution. A single data point may represent homogenous data or it may represent data that have been previously aggregated from several sources. Data points-from different plants and studies are not expected to be homogene-ous. The Bayt:s procedures described by Martz are not applicable unless C-4 (

individual records are homogeneous. A single rec s treated as homogene-ous if (a) it has at least component design detail and (b) it is from a single plant and/or was used as plant-specific data for a Bayesian update. Reference C-2 contains a discussion of the rationale for these rules. The remainder of this section is a discussion of how the form of data for a record is classified based on the minimal information that is pre-sent. If an entry is present for the number of failures and for either the total number of operating hours or demands or for both the average number , of operating hours or demands per component and the number of components whose experience is being combined, raw data are being provided. The aggregation methods recognize just two types of rate variation in-formation beyond the raw data: confidence intervals and tolerance inter-vals. Error factors and standard deviations or variances are assumed to be supplying tolerance interval information. Specifying any type of rate var-iation information except for standard deviations and variances requires a specification of the coverage probability or confidence as well as an upper bound. Lower bounds are optional; if they are not specified the bounds are assumed to be one-sided. For error factors, the " error factor type" that indicates whether the corresponding tolerance interval is one or two sided is required. If one of these sets of data is present, variation information is provided. There are several possible ways to describe tolerance intervals. The aggregation methods check first the upper tolerance bound and associated coverage probability; if this is complete, tolerance interval data are pro-vided and any remaining tolerance interval information will be ignored. Othen ise, the error factor information will be used; the variance or stan-dard deviation information will be used only if no other tolerance interval information is provided in the data record. These considerations lead to six possible categories of data for aggregation. These are described in Table C-1. C-5

1 TABLE C-1. DATA SETS FOR AGGREGATION METHODS , Set Aggregation Number Tvoe Form of Data Present* Aaarecation Method 1 Homogeneous Raw data a k-1 or all f, < 2 Noninformative prior b k>l and at least one Empirical 7 ayes Pro-f, > 1 cedure #a 2 Homogeneous No raw dato; confidence Raw data conversion interval only 3 Homogeneous No raw data and no variation information a k-1 No bound possible b k>l Fit rates to lognor-mal distribution 4 Homogeneous No raw data; tolerance interval a k=1 Use single point b k>l Empirical Bayes Pro-cedure #2 / 5 Aggregated Tolerance interval , a k-1 Use single point k>l 'ixture b No raw data (unweighted) c Raw d?ta (weighted) 6 Aggregated No tolerance incerval a k-1 No bound possible k>l Fit rates to lognor-mal distribution b No raw data (unveighted) c Raw data (weighted)

a. k is the number of points in the set; (f, for i 1 to k) is the set of failure counts.

C-6

C-3. AGGREGATION ALGORITHMS The third column of Table C-1 identifies the aggregation methods to be used. For most of the data sets, the methods are further subdivided based on the amount of data present. Two other data sets are subdivided because the presence of raw data can affect the weighting of the data. Table C-2 provides an overview of the actual aggregation algorithms. In Table C-2, " dealing with 0 rates" requires that a non zero estimate of the median be obtained. Using a non-informative gamma or beta prior 8 produces, respectively, 0.23/T for per hour rates and F/(2n - 1 + F), where F is the median of an F(1,2n+1) distrioution, for per demand probabili-ties. (Here, 1 is the exposure time and n is the number of demands). Both of these require that raw data be available; if this is not the case, the data are omitted. The f(llowing equations apply to compute the variance, S,2, and the mean, M,, of an underlying normal distribution given tolerance infor-mation. The normal distribution describes the logarithm of the correspond-ing failure rate or probability. In these equations, 7 3 is an upper bound from the standard normal distribution (e. g., the ninety-fifth percentile, 1.645, when o is 0.95). Given the median (r,) and upper bound (U,): M, - in r, (this holds in all cases) 5 2- [(in V, - in r,)/zal 3 Given the median and error factor (EF,): S ,2 - (in EF,/zo ]2 C-7 t

TABLE C4'. CFD AGGREGATION ALGORITHMS For Homcaeneous Data Records: Other Data Form Preferred Data Form: Raw Data. Tolerance Interval Neither Empirical Bayes (EB) Proc. #1: EB 4:.oc. #2: Fit rates to lognormal

 . Compute'                                                                .      Deal with 0 rates b      distribution (See below.)

R = ( I f,)/( I ,) . Find normal dist. R2

• I 2 ft (ft -1)/T,]/( I T,) stdv., S, (See note b)

 . Match moments. Output                                                     .      Comnute*

underlying normal distribution: X* = I in r, /k M = 2 in R - 0.5 in R, 5 ,2 - I (in r, - X*)2 /k 3 S2 = i n R, - 2 i n R S,y32, y 3,2 /k W, = (1/S,2) / I (1/Sj r)

  . If 52 < 0, compute separate                                               .      Output normal distribution:

rates and fit them to a lognormal distribution. S2 - S g ,2 3,y9 2 H= I W, in r,

   . Similar treatment for per demand data.                                                               . If S2 < 0, use Sr . 3 q,   r For Aaarecated Ditta Records:

Preferred Data Form: Tolerance Interval No Tolerance Interval Mixture: Fit rates to lognormal distribution: 1

   . Deal with 0 rates.b                                                       . Deal with 0 rates.b Find normal dist. stdv. (S,)                                              . Define weights (W,).b
    . Define weights (W,).b                                                    .      Output normal distribution:
    . Output normal distribution:                                                M=      I W, in r, H=     I W, in r,                                                             Sr - I W, (In r,)2 -M      2 S2- I W, [(in r,)2 + 3,2] - M                                  2

a. Here, f, is the number of failures in time T,.

b. See text for a description of the method.

c. Here, r is the median of the ith rate distribution and k is the number of re, cords being combined.

C-8 l t

         . Given the median and variance (V,):

Si r - In [(r, + / r,' + 4 V,) / (2r,)] Given the mean (X,*) and upper bound: Sr=(z i a

                                  / z o' - 2 in (U,/X,*) ]2 (If this is not possible,             X,* is taken to be the median).
         . Given the mean and error factor:*

Si r , [7 o ./Z' a - 2 in EF, ]r

         . Given the mean and variance:

Si r - in (V, / (X,*)r + 3), In the last three cases, where the median is not provided and the mean is used, the median is calculated as r i- X,* / exp(S,2/2). The weighting of the data is a subject for further research; current-

 -ly, in a group of records with raw data weights are based on the exposure (time or number of demands) associated with each record; when groups are combined, the weights depend on the number of records in the group.

In most of the cases in Table C-2, special provisions apply if there is only a minimal number of records in a group; for example, a noninfor-mative prior is used to provide R and R, in EB Procedure #1 if there are insufficient data, See Reference 6 for these details and for further insights on how the formulas are derived. The final result is a lognormal distribution describing the aggregated rates.

a. Here. the EF is treated as the upper bound divided by the mean. If it were U/r, the median r should be reported in the data source.

C-9

C-4. EXAMPLE OF AGGREGATION HETHOD For the " fails to operate" failure mode of battery chargers, most of the data shown in Table C-3 are available. For this handbook, the last two entries are not used; they are not needed since plant-specific data are available. The last point entered in Table C-3 is present for illustration only, so that there is more than one point among the data lacking counts of numbers of failures and operational hours. The first eleven entries in Table C-3 are used for the component failure data handbook aggregate probability and are plotted in Figure C 1. They provide plant-specific data which are regarded as homogeneous, while the last two represent aggregates that are in_;nded to reflect perforn'ance across the industry. For the first eleven entries, the boundt listed are the 5 and 95 percentage confidence limits calculated using a chi-square distribution. Figure C-1 shows the results of fitting each data point to a lognormal distribution. The lognormal distributions match the specified upper bounds and the rate estimates (number of failures divided by operating hours). The rate estimates for the plots are treated as medians. The resulting lognormal 5, 50 and 95 percentiles are shown. For the aggregation algorithms, the first eleven points fall into Table C-l's Set 1, while the last two are in Set 5. Thus, two aggregates are formed and then combined. For Set 1, using Empirical Bayes Procedure

                                             #1 as outlined in Table C-2, one finds that R is 9.7E-6 and R2 is 1.6E-10; thus; M is -11.8 and S[ is 0.52. Note that the upper and lower         3 bounds given in Table C-3 are not used in this calculation.

For the second aggregate, the mixture method at the bottom of Table C-2 applies. From medians and upper bounds, the underlying normal distribution variances are found to be, respectively,1,81 for the IEEE-500 data and 0.67 for the assumed data; the normal distribution means are sim-ply the logarithms of the point estimates of the rates. (Note that lower bounds are not needed for this calculation). In combining these two data C-10

TABLE C-3. EXAMPLE DATA FOR BATTERY CHARGERS Source' failures Operating Hours lower Bound ypoer Bound HNP1 PRA 6 175200 1.52E-5 6.76E-5 IPS2 PRA 0 95800 0 3.llE-5 MNSI PRA 5 229488 1.73E-5 3.55E-4 NEE 3 PRA 1 96426 5.29E-7 4.41E-5 ZlS-PRA 0 202000 0 9.28E-6 X-PRA _ 8 304128 1.31E-5 4.53E-5 IFRDS (Plant 1) 3 223375 3.67E-6 3.47E-5 !!RDS (Plant 2) 0 56950 0 5.23E-5 IPRDS (Plant 3) 1 733320 1.12E-6 1.06E-5 IPRDS (Plant 4) 3 162930 5.03E-6 4.76E-5 IPRDS (Plant 5) 5 1007400 1.95E-6 1.04E 5 IEEE 500 --(Rate: 6.20E-7)-- - 1.22E-5 _ ZZZ (Dummy data) --(Rate: 1E-6)-- 3 E-7 3 E-6

a. See Appendix A.

Chorger, Bottery-Fails to Operate 10E-004 k o { 10E-005 - ct m g; _ _ . e

   ,2 10E -006        -

u 10E-007 e,

                        %#h"%4 1 %s'% ' % i , #s, h k #b,%%
                                   *

• h , '%, n c:  % 9 Figure C-1. Battery charger failure data (cases with raw data).

C-Il

i N points, the weights are taken to be inversely proportional to the vari-ances. Thus, the weights are, respectively. 0.12 and 0.88. Substituting these data in the mixture equation at the bottom of Table C-2 yields -13.9 for M and 0.81 for Sg t, 2 1 The last step is to u:e the same mixture equation at the bottom of  ! l Table C-2 to combine the two overall aggregates, with weights Wi= 11/13 tnd W, = 2/13. The result of the calculation is M = -12.12 and 52 = 1.13. This is the mean and variance of a normal distribution; the corr.-sponding lognormal distribution for the rates has median, exp M; nnn, exp M

• exp (S2 /2); and error factor, exp (1.645 S). The 95%

upper bound is found by multiplying the median times the error factor. L For the battery chargers, the resulting distribution for the rate of failure to operate, per hour, has the following attributes: Number of records: 13 Mean: 9.6E-6 , Median: 5.4E-6 Upper bound: 3.lE-5 , Error Factor: 5.7. C-S.

SUMMARY

OF AGGREGATION METHODS The aggregation methods work with the data in whatever form the data are given and use a mixture distribution to combine the results. The me-thods are based on three attributes of each data record: a) whether the record represents homogenous data or is already an aggregate that reflects population variability; b) whether the record contains raw or reduced data; and c) whether the record describes an actual failure rate or a probability of failure on demand. The methods use lognormal distributions to capture - the population variability. The median, mean, upper 95% bound, and error factor of the aggregate lognormal distribution are output. Of course, C-12 l l t __ - - - - - - - - _ - - - -

failure rates per hour and rates or probabilities of failure per demand are never aggregated. Additional features of the aggregation method include the f act that k all aggregations are based on lowest level information (i.e., individu: records), and all standard confidence / tolerance leveh are accepted in r t input data (not just 95%). An assumption that underlies all the aggregation methods is that the data from the individual records being aggregated are independent. Before component failure data records are entered into the data base, they are checked as much as possible to ensure that they contain unique data. Fur-thermore, raw data are sought as the primary foim of failure data, and they are the primary data used in the aggregaticn methods, in order to avoid dependence on data that may have originated from common sources such as WASH-1400.10 s C-6. REFERENCES

   ,  C 1.                                    C. L. Atwood, " Approximate Tolerance Intt.rvals, Based on Maximum Likelihood Estimates," Journal of the American Statistical Association,J_2:386, June, 1984, p. 459-465.

C-2. C. D. Gentillon, Aaarecation Methods for ComnoneJt Failure Data in h the Nuclear Comouterized Library for Aisessina Reactor Reliability, Y EGG-REQ-7775 (draft), August 1897. C-3. H. F. Martz and M. C. Bryson, "On Combining Data for Estimating the Frequency of low-Probability Events with Application to Sodium Valve Failure Rates," Nuclear Science and Enaineerina, S3, 3 1983, pp. 267-280. C-4. PRA Procedures Guide, NUREG/CR-2300, Vols. 1 and 2, U. S. Nuclear Regulatory Commission, January 1983. C-5. G. Apostolakis, S. Kaplan, B. J. Garrick, and R. J. Duphily, " Data Specialization for Plant-Specific Risk Studies," Nuclear Enaineerino_ and Desian, 56, 1980, pp. 321-329. C-6. Reactor Safety Study, " Appendix Ill--Failure Data," WASH-1400, NUREG-75/014, U. S. Nuclear Regulatory Commission, 1975. C-13

a y1.NUCL8A:tR GULATO4 Y COMMisslON ' t. Na .Ceu 323

                                                                                                                                                                                           ~~~

M'M BIBLicGRAPHIC DATA SHEET

                                                                                                     <s .,- ,,c m .,.e     . ,..                                                          EGG-EAST-8563
 . riva as; sesnia 3        O a'i a g a;a- a,3, 3-g ;

Component Failure Data Handbook .~ ... April 1991 a .. ca: asts osia 06152

s. Tv es ; a vc a r

 .. .u r-caisi C.D. Gentillon                                                                                                                                                            Technical Evaluation n a ica :: e. e 2 ~ <.. :.....

a J"w e .*y.a es w ,c _ . w ,, e.- - . m . 5 PeeE.a

          .e e 5 0 .aou.e
                     *.Nu one.wm N G OomGAN g a t iam - N au t am o A C ; a t sa o, *.c .+= J+

EG&G Idaho, Inc. P.O. Box 1625 Idaho Falls, 10 83415

   , s,o sea,~G ca o ANu A ricN - Nau e a~o a cc a iss , .e ,                                                                    s. . .. ,,,-,- -                           ..e o--  e, . -        os             . ~~~ : ---

Office for Analysis and Evaluation of Operational Data q United States Nuclear Regulatory Commission Washington, D.C. 20555 _.

    ' 3. $UP.LEME NT A 8 V NcTE 5 us:a ac m                  ., . .

This report presents generic component failure rates that are used in reliability and risk studies of commercial nuclear power plants. The rates are computed using plant-specific data from published probabilistic risk assessments supplemented by selected other sources. Each data source is described. For rates with four or more separate estimates among the sources, plots show the data that are combined. The method for combining data from different sources is presented. 'he resulting aggregated rates are listed with upper bounds that reflect the variability observed in each rate across the nuclear power plant industry. Thus, the rates are generic. Both per hour and per demand rates are included. They may be used for screening in risk assessments or for forming distributions to be updated with plant-specific data. o +- c~ a<s .c a os. c e sca e r c a s s.- ~ . ~ o -- --. .. - , Unlimited generic component fail Jre data ..m- . . * . . nuclear power plant ri:k assessment ..... failure rates Unclassified aggregation of data .r...-, Uncl assi fied is wueea a .acas it_ .m eC1 1

       %#C 808tu 32112494

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _}}