ML20150E090
| ML20150E090 | |
| Person / Time | |
|---|---|
| Site: | 05000447, 05000531, 05000550 |
| Issue date: | 12/04/1978 |
| From: | Brooks B, Vezey E GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20150E087 | List: |
| References | |
| NEDO-23978, NUDOCS 7812110116 | |
| Download: ML20150E090 (35) | |
Text
--
lg
""=:::
! oc10#2 l
! ADS SOLENOID VALVE l RELIABILITY DEMONSTRATION l AVAILABILITY ENGINEERING REPORT NO. 50 l
l
"' e"l e"."vezS 78121101161
NED0-23978 78NED289 Class I October 1978 AVAILABILITY ENGINEERING REPORT NO. 50 ADS SOLENOID VALVE RELIABILITY DEMONSTRATION
/
't / $-
B. P. Brooks / E. I. Ve Approved:
b Approved: "
C. B. nson, Manager R. J. McCandless, Manager Valves and Auxiliary Availability Engineering Equipment Design NUCLE AR ENERGY PROJECTS DIVISION . GENERAL ELECTRIC COMPANY SAN JOSE, CALIFORNIA 95125 GENER AL h ELECTRIC
DISCLAIMER OF RESPONSIBILITY This document was prepared by or for the General Electric Company. Netther the General Electric Company nor any of the contnbutors to this document:
A. Makes any warranty or representation, express or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this docu-ment. or that the use of any Information disclosed in this document may not infringe privately owned rights; or B. Assumes any responsibility for liabthty or damage of any kind which may result from the use of any Information disclosed on this 00cument.
NEDO-23978 TABLE OF CONTENTS .
Page ABSTRACT v/vi
- 1. INTRODUCTION 1
- 2. CONCLUSIONS 2 3
SUMMARY
3
- 4. DATA 4 5 DATA ANALYSIS 5
- 6. HARDWARE COMPARISON ANALYSIS 6
- 7. EVALUATION OF APPLICABILITY OF FAILURE RATE 6
- 8. REFERENCES 8 n
APPENDICES A. ADS SOLENOID VALVE RELIABILITY DEMONSTRATION PR00 RAM A-1 B. AUTOMATIC DEPRESSURIZATION SYSTEM RELIABILITI DEMONSTRATION B-1 PROGRAM FOR GESSAR APPLICATIONS C. COMPARISON OF ADS (AUTODEPRESSURIZATION SYSTEM) SOLENOID VALVES C-1 iil/1v ,
NEDO-23978 ABSTRACT A reliability analysis of the ADS solenoid valve arrangement was submitted to the Nuclear Regulatory Commission, showing that the current, redundant valve system with 21-month surveillance testing is more reliable than a proposed sys-tem with additional solenoid valves. The additional valves would permit sur-veillance testing at more frequent intervals.
A failure rate for ADS solenoid valves in the fail-to-operate mode of 5.6 x 10-6 per hour was used in the analysis. This is an industrial generic failure rete.
General Electric was asked by the Nuclear Regulatory Commission to demonstrate this failure rate. Field data on similar valves were acquired, and an actual field failure rate was determined. The field failure rate is 8.2 x 10~ per hour for fail-to-operate mode and 3.3 x 10
-6 per hour for all failure modes. This demonstrated failure rate, therefore, confirms the conservatism of the failure rate used in the system reliability analysis.
v/vi
- 1. INTRODUCTION The automatic depressurization system (ADS) solenoid valves control the flow l of air to the safety / relief valve (S/RV) air operators and cannot be tested without opening the S/RV, which would result in an undesirable pressure reduc-tion in the BWR nuclear boiler system. All other system elements are testable or operable during plant operation. Several schemes have been proposed to add solenoid valves to isolate the ADS solenoid valves during test. A comparison of the reliability of the existing design and other designs which would isolate and allow testing of solenoid valves showed that the alternate designs degraded the ADS reliability, and that it is preferable to use the existing design and conduct surveillance testing of the solenoid valves at outage intervals (a maxi- '
mum of 21 months). In this analysis, which was submitted to the Nuclear Regu-
+
latory Commission (NRC), a failure rate of 5.6 x 10 -6 per hour was used for the solenoid valves in a fail-to-operate mode. This was a generic failure rate taken from the literature. In return, the NRC requested General Electric to demonstrate that the solenoid valve failure rate used in the above referenced analysis is valid.
A proposed solenoid valve reliability demonstration program was submitted to the NRC for approval.(l) The proposed program consisted of a search of field data l
to determine the field failure rate of solenoid valves in applications similar to those proposed for use in BWR/6 application where the systems analysis applies.
The prcposed program is included in this report as Appendix A.
Subsequently, the NRC approved the proposed program for implementation.
Reference 2 is attached as Appendix B. I l
In parallel with the data search, a comparison review was conducted on the in-service ADS solenoid valves and the solenoid valves which will be installed in BWR/6 plants. This study is included in this report as Appendix C.
This report presents the data accumulated on the program, the conclusion reached, and supporting discussion.
1
- 2. CONCLUSIONS
- 1. The actual ADS solenoid valve failure rate derived from field applica-tion data is lower than the industrial generic failure rate used in the systems analysis.
(a) The failure rate used in the previously submitted reliability analysis was 5.6 x 10 -6 per hour in the failure-to-operate mode.
~
(b) The field failure rate was found to be 8.2 x 10 per hour for f ailure-co-operate modes.
(c) The field failure rate was found to be 3.3 x 10~ per hour for all failure modes.
- 2. The Licensee Event Reports (LER), the General Electric Component Infor-cation Retrieval system (CIR), plus on-site investigation of repair records, were adequate to establish valid failure rates for the ADS solenoid valves.
3 The solenoid valves =anufactured by Automatic valve Company (AVCO),
Farmington, Michigan, are utilized in the in-service application and are essentially of the same configuration as the valves which will be used on the Crosby and Dikkers S/RVs in BWR/6 applications. The failure rate based on the field experience of the AVC0 valve is applicable to the solenoid valves used on Crosby and Dikkers S/RVs.
4 Solenoid valves =anufactured by Automatic Switch Company ( ASCO), Florham Park, New Jersey, are also installed in in-service applications (in addition to those manufactured by AVCO); however, these valves are a somewhat different design than the solenoid valves used on Crosby and Dikkers S/RVs, and data from these valves are therefore not useable and are not included in the evaluation.
2
SUMMARY
There are three kinds of ADS solenoid valves currently in service. Of these, the valves made by AVC0 are equivalent to the ADS solenoid valves to be used on Crosby and Dikkers S/RVs. The other two (ASCO and the new valves on the two-stage Target Rock S/RVs), are sufficiently different from Crosby and Dikkers valves that the data are not used in this demonstration program.
Eleven domestic plants have AVC0 ADS solenoid valves. Several other plants have a mixture of AVC0 and ASCO valves; however, since it was not considered feasible to separate the operating time between AVC0 and ASCO valves, the data from these plants are not included in this study. Data and operating hours from one plant are not included because of the difficulty of obtaining data from that plant; thus, there are 10 plants involved in the study.
The total accumulated valve service time (i.e. , plant operating hours times number of S/RVs) on the 10 AVC0 equipped plants included in the study was 1,231,195 hours0.00226 days <br />0.0542 hours <br />3.224206e-4 weeks <br />7.41975e-5 months <br /> as of June 30, 1978.
The ADS solenoid valve anomaly history on the AVC0 equipped plants is shown in the Chart of ADS Scienoid Events (Figure 1). The chart is largely self-explana-tory. A limitation in a study of this type is that initial event reports are of ten erroneous or =isleading, since they are based on gross sy=ptoms. La ter failure analysis is not of ten correlated with the initial report. However, all reports were treated as valid, and on-site investigations were made to clarify and verify the actual event. The only failure which resulted in a failure of the solenoid valve to operate occurred at Plant D on June 14, 1976. This results in a calculated failure rate for the failure-to-operate = ode, as follows:
1 -7 A 8.2 x 10 per hour closed (1,231,195 - 6760) 3
NEDO-23973 8760 hours0.101 days <br />2.433 hours <br />0.0145 weeks <br />0.00333 months <br /> (one year) are subtracted from the total operating hours because the valve is presumed to have _ failed at the beginning of the cycle. There were three other failures of solenoids. In two cases, they failed open during shut-down and were self-announcing on startup. The third failed open af ter manual operation. The . combined failure rate of all modes is:
-6 per hour A = = 3 3 x 10 open (1,231,195 - 8760 x 4)*
Although only the failure-to-operate mode is applicable, it is interesting to note that both of these failure rates are lower than the failure rate of
-6 5.6 x 10 per hour used in the reliability analysis.
1
- 4. DATA The Chart of ADS Solenoid Events presents all of the events which were found as a result of Licensee Event Report (LER), Component Information Retrieval (CIR) operated by General Electric, and site data searches. In each case, the source is given on the chart. General Electric has an availability engineer on site at each of the plants involved, except one. For this reason, this plant was omitted from the study, as were the plants that have a mixture of AVC0 and ASCO valves. Site investigations included interviews with personnel involved, a study of maintenance work requests to determine work actually done, and a search for events not otherwise reported.
'The three fail-to-close mode failures occurred during a startup but may have been in a failed mode since prior manual operation - as much as one year before.
4
NEDO-23978 Since the study of field data was carried out in parallel with the study of ADS valve similarity, the failures of ASCO as well as AVC0 valves were also reviewed.
There was a total of nine field failures on AVC0 and ASCO valves during !
3,160,212 valve hours. The produces a failure rate of: l l
= -6 A = 2 9 x 10 per hour gross (3,160,212 - 8760 x 9)*
This shows that the generic failure rate for all types of solenoid valves used
-6 on S/RVs in the reactor environment is less than the value of 5.6 x 10 per hour derived from generic industry data.
- 5. DATA ANALYSIS Data analysis results are shown on the Chart of Solenoid Events. The sources of failure analyses information are shown on the chart along with conclusions which affect solenoid valves. Many of the failure analyses covered other prob-lems which occurred and, in some cases, were inconclusive (e.g., the 7/15/77 event at Plant A). In such cases, the solenoid valve was conservatively charged with a failure.
l The determination of failure rates was covered in Section 3 The failure rates were determined by simple point estimates, as proposed in Appendix A. The failure rates for fail-to-open, and for all other modes, were lower than the generic failure rate used in prior analyses. This result justifies the use of the generic failure rate.
l
' Assumes that all nine failures occurred at beginning of one year fuel cycle i
but were not discovered until end of cycle.
5 1
- - . ________ ___ _______ _ __ _ ___ _ ___ _ __________________ _______________.------_______a
NEDO-23978- .
- 6. HARDWARE COMPARISON ANALYSIS The hardware comparison analysis was conducted to gain assurance that the data l from solenoid valves in the field would be meaningful when extrapolated to the solenoid valves being used on the Crosby and Dikkers S/RVs in BWR/6 application.
This study was completed as a separate, parallel task, and the report of that study is included as Appendix C.
An early conclusion of the study was that valves manufactured by Automatic Switch Company and the valves being used by Target Rock on the new two-stage S/RVs were not of the same basic configuration as the ones being used on Crosby and Dikkers S/RVs. For that reason, the study concentrated on the valves manufactured by Automatic Valve Company (AVCO). The conclusion of the study is that the AVC0 valves are sufficiently similar to the Crosby and Dikkers solenoid valves to justify direct correlation of AVC0 field data with anticipated performance of Crosby and Dikkers solenoid valves.
7 EVALUATION OF APPLICABILITY OF FAILURE RATE It is concluded that the failure rate determined for the AVC0 solenoid valves is valid for use as a nuclear power system failure rate in reliability analysis.
The AVC0 failure rate (as well as the combined AVCO/ASCO failure rate) was lower than the generic failure rate takea from published tables.
The in-service AVC0 failure rate data are exhibited by equipnent which has been subject to the same quality control, handling, environment and maintenance as future ADS solenoid valves will experience and covers experience in 10 different plants. The failure rates of 8.2 x 10 per hour for fail-to-operate and 3.3 x
-6 per hour for "all failure modes" are, therefore, submitted as " nuclear 10 generic" failure rates for use in f uture ADS solenoid valve reliability analyses.
6
~
The failure rates as calculated in this report are " point estimates" or " maximum likelihood estimates".' The method of calculation is appropriate because the
)
valves are maintained at periodic intervals and are replaced when failure occurs. The confidence level associated with failure rates calculated by this ]
l is in the order of 50% but is difficult to evaluate because so few failures have l l
occurred. However, this is characteristic of " generic" failure rates for high reliability hardware. I l
l i
l l
J l
I l
l l
l l
i l
l l
i l
l l
1 1
I l
\
i "I. Bazovsky, Reliability Theory and Practice,1961 Edition, Prentiss Hall, pp. 212 et seq.
- B. Epstein, Truncated Life Tests in the Exponential Case, Annals of Mathema-tical Statistics, 25 3, September 1954, pp. 555-564.
7
- 8. REFERENCES i
- 1. G. G. Sherwood, Letter to U. S. Nuclear Regulatory Comission, dated January 16, 1978.
Subject:
ADS Solenoid Valve Reliability Demonstration Program Docket Numbers - STN 50-531 and STN 50-550.
- 2. U. S. Nuclear Regulatory Comission, Letter to G. G. Sherwood, dated April 26, 1978.
Subject:
Automatic Depressurization System Reliability Demonstration Program for GESSAR Applications.
8
FAILt487 I II a
Plant Peport Date Event Reported E $" EG Failure Analysis Remarks A 02-27-75 All ADS valves remained open Sneak circuit .hrough relay indicator lights No No Not solenoid valve failure, after actuation during PREOP provided enou d current to hold solenoids test. (Source: LER Files) open in actuated position. (Source: Inves-tigation on site.)
A 04-29-75 S/RV "B" failed open at 105 Solenoid valve disassembled. No cause found. No No Failed safe for ADS function -
power. (Source: 1.ER Files) Failure charged to crud it steam pilot valve. not charged to solenoid valve.
Solenoid valve remained in service. (Source:
AOR Files.)
A 05-06-75 S/RV "B" failed open at 905 =0" ring dislodged; burr on solenoid valve Yes No Failed safe for ADS function, power. (Source: LER Files) core plunger; possible ground in solenoid 1ead.
A 07-15-77 S/RV "F" failed open after Grounded wiring in conduit bor in drywell. Yes No Combination of failures, manual actuation. (Source: Solenoid valve coils had cracked and chafed including solenoid valve.
LER Files) insulation. S/RV steam pilot valve leaked. Failed safe for ADS function.
Replaced valve top works and solenoid coll. h, (Source: Work Request and investigation.)
B 09-11-77 On post outage hydro S/RV Solenoid valve from S/RV "It" was full of l
"11" was open. Removed Yes No Solenoid valve exposed to abnor- N water. Soleroid valve had bee left mal moisture. However count as $
solenoid valves feca S/RVs hanging in drywell with valve failure because damage could "H" and "C". (Source: unprotected while S/RV out '
. sing
-# 'enance.
y Site Work Request, inves- have developed after startup.
Source of water unknown. 5 - Ive Failed safe for ADS function.
tigation) plunger frozen. Solenoic n <+. n 'C" OK. (Source: investigation + >> <e.)
B 09-16-77 S/RV "H" opened during Did not clear problem on S/RV "H". No No Problem repeated at next test.
vessel filling. Replaced (Source: Investigation on site.) See next item, solenoid valve as part of trouble-shooting procedure.
(Source: Site Work Request; investigation.)
B 09-24-77 Electrical ground in S/RV "H" Found "Amphenol" connector full of water.
solenoid circuit. (Source:
No No Not a solenoid valve failure.
Dumped water out of connector; ground Failure associated with power Site Work Request; investi- cleared. "Drywell was like Turkish bath." cf.rcuit.
gation.) (Source: Site Work Request; investigation on site.)
C 03-05-75 S/RV stuck open. Lapped Found broken airline fitting to solenoid pilot valves. Repaired No No No solenoid valve involvement.
valve. No solenoid valve failure. (Source:
solenoid valve. (Source: Vendor Report)
CIR Files) 71gure 1. Chart of ADS Solenoid Events - AVCO Solenoid-Valves
FA!!fRE?
I 0" da 25 EG Remarks Report Date Event Reported Failure Analysis Plant These 3/DVs are new 2 stage Target Rock No No Ifot AVC0 or similar type valve.
C 06-09-78 3/RVs "C" and "L" would not valves. Solenoid valves are not AVCO nor Therefore, data is not included operate on manual on startup test following outage. similar type. Solenoid valves had not been in the study.
(Source: LER Files) assembled properly. (Source: Vendor Field Report)
, Steno *d had been leproparly assembtw; So Yes Quality Control problem. Did not D 06-14-76 3/RV "E" solenoid valve in-operative. (Source: CIR resulting in stripped threads in solenoid fail safe for ADS function.
Report.) Modified per site valve body, allowing plunger assembly to ,
failure report: it was loosen to extent the valve was rendered 3/RV "A". Inoperable. (Source: Site Failure analysts.)
6 woNE y
Plants g
I w
- w o W M
cn Figure 1. Chart of ADS Solenoid Eventa - AVC0 Solenoid Valves (Continued)
NEDO-23978 APPENDIX A A.1 ADS SOLENOID VALVE RELIABILITY DEMONSTRATION PROGRAM A.1.1 Purpose The purpose of the BWR/6 ADS solenoid valve reliability demonstration program l l
is to verify the failure rate used for solenoid valves in the previously sub-mitted ADS reliability analysis. The failure rate used was 5.6 x 10-6 for j failure-to-operate mode of failure.
A.1.2 Proposed Demonstration Method The proposed method for demonstrating the reliability of the BWR/6 ADS solenoid valves is to determine the field failure rate for S/RV solenoids on operating domestic BWRs. Many plants have S/RVs with solenoids similar to those used on BWR/6 valves; thus, a large body of data exists from which a highly applicable failure rate can be developed for verifying ADS solenoid valve reliability.
In addition to the hardware similarity, the drywell environment, duty cycles,
- installation, maintenance, quality control and user variability will be essen-tially identical in the current operating plants and in the BWR/6 plants, so that the failure rate will be directly applicable without applying modifiers.
Failures will be identified by researching the two existing failure reporting systems. The Licensee Event Reports (LER), which the utilities are required to sutruit to the NRC, will be the primary source of information relating to the l S/RV solenoid failures. The Component Information Retrieval (CIR) system which is operated by GE has similar data and will also be utilized.
A preliminary look at the field failure data and a review of data at one operat-ing site indicates that the approach is entirely feasible, and that a valid failure rate can be established.
A-1
NEDO-23978 A.1 3 Demonstration Program The ADS solenoid valve demonstration program will have three phases:
(1) a review of failure report data in the LERs and CIRs; (2) review of records at several plant sites to gather "in-depth" infor-mation on the reported failures and reporting system; and (3) analysis of hardware and data.
The first phase will consist of a search of the LERs and CIR reports, to find all reported incidents of S/RV solenoid failure. Safety / relief valve failures will also be reviewed for any reported failures which may have been due to solenoid valve malfunction but not reported as such. The definition of failure will be from IEEE-500: "Tennination of the ability of an item to perform its required function." Both "announcedr and " unannounced" failures will be included (i.e. , not only actual failures on demand, but also conditions found by inspection that would preclude operation will be classified as failures).
All S/RV solenoid failures will be included in the determination of failure rate except:
(1) Failures of control circuitry, wiring and piping external to the valve assembly are not applicable to the component failure rate and there-fore will not be included.
(2) Failures which are clearly the result of out-of-spec conditions i= posed on the valve are failures of external systems ar.d are not applicable to the component failure rate and also will not be included.
All identified S/R solenoid failures will be documented. Those failures not considered as part of the component failure rate will be analyzed and clearly demonstrated as to why they are not included.
A-2
. NED0-23978 The second phase will consist of a review of plant records to verify the com- ,
pleteness of the LER and CIR reports. The reports in both systems are usually l written before the failure analysis is complete and updating of reports of diffi-cult to coordinate. There is also a time lag in the system, and recent reports may not be available in the printouts. Plant maintenance records are typically more detailed as to failure mode and cause and wil3 e valuable in understanding reported events. An initial data review identified reported solenoid valve failures at several plants. A visit has alreadv been made to one plant, and visits to the other plants with reported failures would be of primary interest.
If the data at these plants substantiate the LER and CIR systems reports, no further plant visits would be planned.- Telephone inquiries should suffice to confirm absence of failures at the other plants. Visits to-the above plants have not been coordinated with the utilities at this time.
The final phase of the demonstration program includes calculation of the field failure rate of ADS solenoid valves, a design comparison of Target Rock, Crosby and Dikkers solenoid valves and an evaluation of the applicability of the field failure rate to the BWR/6 solenoid valves. Solenoid valves furnished with the Target Rock valves and the Crosby valves are manufactured by the same supplier; Automatic Valve Company. The solenoid valves furnished with the Dikkers valves are manufactured by Seits Company. All of the above solenoid valves are of similar type and construction and use similar materials.
The failure rate will be calculated as a point estimate:
y, nu nber of applicable failures numoer of applicable plant hours i
Failures-per-plant-hour is the significant failure rate for this equipment. The total number of valve operations in a normal valve maintenance period (5 years) is relatively very small. A failure rate per demand is not a relevant statistic in a system of this type because it is not the demands that cause the failure; the demand reveals a failure which has already occurred due to a standby or failure-per-hour mode. The ADS solenoid valves are in a normally closed, daenergized state; and the standby, failure-to-open failure rate is the rate of concern.
A-3
NCC-23073 If nultiple, independent failures have occu-red en a valve duri .g the standb7 peried between actuatices, they will te Ocutted as nultiple faibres. A failed valve vill be censidered to have failed in=ediately after the last actuatien, and none of the standby tine innediately preceding the discovery Of falh e vill be included in cunulative valve operating tine.
0:neurrent with the above tasks, a desig2 e cpa-iscn between the ~arge ?.cc solenoid valves (two different de. signs) and tnese en Crosty and Dikkers S/RTs will be :enducted. Target Ecek valves are en nest currently cpe s*ing D? plants.
A prelininary cec;arison stevs equivalence in qualification test prog ans, in CA/OC acw ;tance luirenents and, for one of the Ta get *-cek solencid valve designs, design and Ocestruction epivalence, yart of this task will be to dete&
nine the installation history and c;e-ating tine en the Target Tock solenoid valves.
~he final task of this pregan vill be an evaluation of the applicability of the ae~n:alated failure rate data te the n?/6 valve design. Tnis vill consist ,
pr'-mally of resolving the effect of any detail design differences in lig:t of I the failure nodes reported in phase tv0 of the prog an. It is not anticipated that any substantial obstacles to direct applicability will exist.
I t
A.2 ALTEREATI MlTEDS FOR RELIABILC ODC1STRATICN T.ro other a;;roactes to eliability dener.stration c- failure rate de .e-nira*ic:
vere also considered:
(1) 1.ateratory testing of valves, and (2) ceservation of S:lenoid valves en the first n?/6 plants to Ottain ea-17 indleations of weaknesses and *: detemine the fail:re -ste.
The prinary reas.:n f r rejecting these tv nethods is that the selected netted will give better results nere quickly than either Of the Other two. A brief discussion of teth of the alte nate nethods of reliatility den cst-ati:n f:11 vs.
A.:
1
A.2.1 Laboratory Testing I i
Laboratory testing is particularly suitable for small parts which will have a relatively short duty cycle and relatively high cycles of operation. Large num-bers of parts can be put on test for significant time and, in many cases, the environment and duty cycles can be reasonably simulated. This type of testing yields valid results as to failure modes and cycles to failure for this kind of application. For the ADS solenoid valves the application is different. These valves remain passive up to 21 months and must actuate on demand. Failures for this type of operation will probably occur while the valve is in the passive state as opposed to failures caused by a large number of cycles of operation.
Consequently, an appropriate test would consist primarily of long periods of non-operation followed by a single " random" actuation. This would be an inordinately ,
long test with questionable results. Also, the BWR environment is difficult to I simulate, the valves are relatively large and awkward and the plant-to-plant variation in maintenance and quality control is not readily simulated. All these elements contribute to make the laboratory test not appropriate for the ADS solenoids.
A.2.2 Observation of BWR/6 Valves It is obvious that the most accurate failure rates would be obtained from the actual hardware of concerr in the actual use of concern. The obvious problem is the time delay of several years after startup of several plants before enough data are accumulated to establish a failure rate with reasonable confidence.
If no field data were available on similar hardware in the actual use environ-ment, then this would be the recommended method. However, in view of the estimated 5-8 years required to obtain limited data by this approach, it was obvious that utilizing the data bank provided by existing operating would be far better.
A-5 i
A.3 FINAL REPORT The final report will include all of the findings of the study both in-house and from the field. The outline proposed for this report is the following:
- 1. INTRODUCTION i
- 2. CONCLUSIONS AND RECOMMENDATIONS i l
1 t
3
SUMMARY
- 4. DATA Failure Report Study Field Containment Study Field Anomalies Found
- 5. DATA ANALYSIS Review of Each Failure Review of Operating Hours Calculation of Failure Rate
- 6. HARDWARE COMPARISON ANALYSIS
- 7. EVALUATION OF APPLICABILITY FAILURE RATE i
i A-6 l
\
[
APPENDIX B 4 UNITED ST ATES
.f *
. . NUCLEAR REGULATORY COMMISSION j,.g. wAssiNcToN o.c rosss -
as, ss. a<j An a es Docket Nos: STN 50-447 STN 50-531 and STN 50-550 General Electric Company ATTN: Dr. G. Shenvood, Manager P6'go ,
Safety & Licensing Operation V.N g 1976 175 Curtner Avenue !
San Jose, California 95125 g g nw000 l Gentlemen:
SUBJECT:
AUTOMATIC DEPRESSURIZATION SYSTEM RELIABILITY DEMONSTRATION PROGRAM FOR GESSAR APPLICATIONS During our review of the GESSAR-238 NSSS aoplications, the General Electric Corrany committed to develop a program to establish the relia-bility of pilot solenoid valves. The specific valves are those that will be used in the automatic depressurization system and the pressure relief system of plants utilizing one of the GESSAR designs. As part of that corr.itment the General Electric Company agreed to submit the program for staff review and approval prior to its implementation.
Subsequently, on November 17, 1977, representatives of the General Electric Company met with the staff to discuss a proposed program to satisfy that commitment. As a result of that meeting, we provided our coments on the proposed program in a summary of the meeting, dated November 22, 1977. These comments resulted in a revised program which you transmitted by letter, G. Sherwood to E. Case, dated February 8, 1978. We have reviewed the revised program, find it acceptable, and in satisfaction of our requirement that the program be reviewed and approved prior to its implementation.
In your letter you also noted your interest in the staff's intended l review schedule. We plan no furthur review of the program until the I
results are submitted. We will advise you of our review schedule at that time. In the interim we would suggest that you keep the staff i advised of the progress of the program. l B-1
A?R z f G76 1
If you require any clarification of the matters discussed in this letter please contact the staff's assigned licensing project manager. I 9
- h. .
4 d
.k.va/%8srge,g chief Light WateNReactors Branch No. 4 Divisien of Project Management i
B-2
NEDO-23978 General Electric Company 1
Cu:
Mr. W. Gilbert, Manager Safety and Standards General Electric Company 175 Curtner Avenue San Jose, California 95125 Mr. L. Gifford, Manager Regulatory Operations Unit General Electric Company 4720 Montgonery Lane I Bethesda, Maryland 20014 2 f
/
v i B-3/B-4 t_________. ______--_____ ___ _ ____ _ __ _ -- _ -_ _
NEDO-23978 APPENDIX C COMPARISON OF ADS (AUTODEPRESSURIZATION SYSTEM) SOLENOID VALVES C.1 PURPOSE OF REPORT The purpose of this report is to provide a comparison of solenoid valves, cur-rently in in-service use, with those which will be furnished for BWR/6 appli-cation in ADS service.
C.2 APPLICATION OF EQUIPMENT The solenoid valves are a component part of the dual function main steam safety /
relief valves and are utilized to control the flow of air to a pneumatic actuator which causes the safety / relief valve to open and reclose when the solenoid is energized and subsequently deenergized.
The solenoid valves currently in in-service usage or those which have accumu- j lated a significant time history of usage are/were installed in safety / relief valves in the 22 boiling water reactor plants:
Seven plants with solenoid valves other than as depicted in this analysis are also utilized. Per-formance data accounts for this variance.
These solenoid valves are produced by the Automatic Valve Company, Farmington, Michigan and are installed on the safety / relief valves produced by the Target Rock Corporation, E. Farmingdale, New York.
The solenoid valves, with which these in-service valves will be compared, will be furnished with safety / relief valves for BWR/6 application. Two suppliers are involved:
C-1
~ . . _ _ _ _ .
NEDO-23978 (1) Solenoid valves produced 'oy C. Seitz AG, Zurich, Swit::erland are installed on safety / relief valves produced by Dikkers Valve Co.,
Hengelo, Netherlands. These valves are scheduled to be installed in the following BWR/6 plants:
Clinton 1 I
Perry 1 and 2 Black Fox 1 and 2 Grand Gulf 1 and 2 3 (2) Solenoid valve cocponents (coil, plunger assembly and plunger tube asse:n-bly) are produced by the Peter Paul Co. , New Britain, Connecticut; the balance of parts are produced by Crosby Valve Co. , Wrentham, Massachu-setts, and are installed on safety / relief valves produced by Crosby Valve Co. These valves are scheduled to be installed in the following SWR /6 plants:
Correntes CNV 1, 2 Kuo Sheng 1, 2 River Bend 1, 2 TVA X17-22 incl.
l C.3 EQUIPMENT DESCRIPTION Figure C-1 depicts the in-service solenoid valves, Figure C-2 depicts the BWR/6 Dikkers solenoid valves and Figure C-3 depicts the WR/6 Crosby solenoid valves. Table C-1 lists the caterials/ construction used in each of the solenoid valves.
4 l
1 J
$ C-2
. .. ._ _]
NEDD-23978 C.4 DISCUSSION Construction, design, and function of the BWR/6 solenoid valves are equivalent to the in-service solenoid valves in the following areas:
(1) Electrical enclosure is NEMA class IV (water-tight and dust-tight) .
(2) Solenoid coil insulation is Class H (high temperature, moisture resistant).
(3) When deenergized: (a) spring force holds the plunger assembly and main poppet in the closed position blocking the passage of inlet air; (b) inlet air pressure acts in a direction tending to open the main poppet; and (c) the vent seat is open, venting any air leakage from the main seat to atmos-phere, f 1
i (4) When energized: (a) electromagnetic force and inlet ait pressure overcome the closure spring force and act to move the plunger assembly (includes the main poppet and vent poppet seats) to open the main poppet seat: (b) the vent seat is closed; (c) inlet air pressure is supplied to the pneumatic actuator; and (d) af ter deenergizing the main poppet recloses, the vent poppet reopens and exhausts air from the actuator to the ambient.
(5) The three-way valve function is accomplished by direct action of the plunger and related seat seals (no linkage, etc. is involved).
(6) Static air seals are produced of Viton material.
(7) Dynamic air seals (valve seat materials) are produced of Viton except for the main poppet seats of the Dikkers solenoid valves which are metal-to-metal.
C-3
NEDO-23978 Construction and design of the BWR/6 solenoid valves are different than the in-service solenoid valves in the following areas:
(1) The Crosby solenoid valves utilize a coil enclosure constructed of aluminum and stainless steel, whereas the in-service valves utilize an enclosure of steel.
(2) The Dikkers solenoid valves utilize a stainless steel ball and stain-less steel seat for the main poppet, whereas the in-service valves utilize a flat disc, constructed of Viton, and a stainless steel seat.
(3) The Dikkers and the Crosby solenoid valves utilize a valve body con-structed of aluminum, whereas the in-service valves utilize a valve body constructed of brass.
(4) The Dikkers solenoid valves utilize multiple closure springs, whereas the in-service valves and the Crosby valves utilize a single closure spring.
(5) The Dikkers and the Crosby solenoid valves utilize hermetically sealing )
electrical connectors, whereac the in-service valves utilize direct wiring within a threaded conduit connection.
Quality Control Action Quality control action applicable to the BWR/6 solenoid valves is equivalent to that conducted for the in-service solenoid valve. The basic program for each is as follows:
(1) Qualification tests to demonstrate operability during normal and emergency environmental conditions. The qualification program to which the BWR/6 solenoid valves have been (or will be) subjected is as follows:
C.u
NED0-23978 (a) reference frame tests to demonstrate leaktightness, response time and operability to specified limits; (b) radiation aging consisting of a cumulative radiation dose of 30 x 10 RADS l (carbon); j (c) thermal aging for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> at 200 0C; l
[ (d) mechanical aging, consisting of 1000 actuation cycles; l
(e) repeat of Item (a); i (f) exposure to emergency environmental conditions at 340 F, decreasing to 2500 F (operability is demonstrated during and after the exposure);
(g) repeat of Item (a); and (h) disassembly and inspection.
The qualification program to which the in-service solenoid valves have been sub-jected is equivalent to the above, except that thermal aging was not applied and mechanical aging consisting of 200 actuation cycles was performed.
(2) Dimensional control and verification of critical dimensions of production units.
(3) Materials identification and control (including lubricants).
(4) Workmanship control.
(5) Component production tests to assure operability.
(6) Extended cycle test of randomly selected valves.
(7) Performance test at assembly with the safety / relief valve.
C-5
NDO-23978 C.5 CONCLUSION Solenoid valves which are currently in use or those which have accumulated sipi-ficant time history of usage in boiling water reactor plants as listed in Section 2 of this report are evaluated to be equivalent to those which will be .
installed in WR/6 plants. This equivalence applies to the design, construction, and function of the solenoid valves and is based en close similarity of the several valves and negligible differences.
C-6
~
Table C-1 COMPARISON OF MATERIALS AND CONSTRUCTION In-Service Dikkers Crosby Solenoid Valve Solenoid Valve Solenoid Valve Part Name Material Material Material Coil Enclosure Steel Steel Aluminum / Stainless Steel Enclosure Type NEMA Class IV NEMA Class IV NEMA Class IV Coil Insulation Class H Class H Class H l
i l Plunger Tube Assembly Stainless Steel Stainless Steel Stainless Steel
?
Plunger Assembly Stainless Steel Stainless Steel Stainless Steel Main Seat Stainless Steel /Viton Stainless Steel / Stainless Steel /Viton Stainless Steel Back Seat Stainless Steel /Viton Stainless Steel /Viton Stainless Steel /Viton Seals Viton Viton Viton Valve Body Brass Aluminum Aluminum i
I L _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ._
NCC-23773 A n _ r. --
I i
J .e e .' \
\
\(\\\
S$'%?%Y-'N ! /s , ,3 \\
" j, " ( ${A*
- ..y..-
% i .. ,
rjg j ;.g f
I I'
1335V35' )MM\ wl f *( /
+ ~
( . 4= [fr'Ii + l s i 'N
~ l Y ,A ; A I \ / '. N l
' M pl{+
- D. % \ . . . . .
s.--r r
/
q/ .si N - .
t 7" .Na!"~ 1 t ,* * ; * * * -
mac
..,: :-.-.. s
- \s '
, a :ilev.'m* ::c A
- L-s i
' : c< N'
- ..'NSE:
- c:w:<
.\,o 9 it N
,g ,
.....c; . .
- . g.
%s \
6 g\ \ }
v:,
- y;- ---
- - . y,x ; y m 'y,
\
- s st
- -
.k,..
k i --w' ...,,"...-
t '"e a 'w 4w Eif;.*.? C-1 h e*7i0e Scle.:id Y117e ss
NEDo-23978 I ,
l l
l l
EXHAUST I
\
- \ V273 y $
CDIL ENCLOSURE E 7 0
h/
T'I"5 x:x'
\
s\g5 %
) SOLEN 0ID
'_ggh D '
b h / ls COIL AND 9 _
- I N S U L AT .". '
r-( g:,
/
, )
\
(
N PLUN3EC TUBE ASSEMSLv 4 _
brf;:
\ l -
Ml .
s 1 ///
vuu s
g s mm F lh d .t '
- N ,', T ,1_.
N SEALS
! - s \
L - - s- ,u E! '
VALVE
[' 'I 1
/ BO'M CLDSURE SDRI.NG r 1 /s
[/ ,
/
I 3
I
/J 9'_ N MAIN PLUN3 e..
ASSEMBLY SEA ~
TO ACTUATOR
% INLET u
Figure C-2. Dikkers Solenoid Valve C-9
- N SEAL g EXHAUST
/
05URE
/
/' h.
CLJ';GER TUSE ,_
l\NNN
/ si 1/
j 7 x_ x_x_Ni(
I - COIL AND SOLENOID ,
ASSE"SLv s INSULATION N _7 _ -
k M _ !u- {~
~
\ k -BACK SEAT j i !'-
&c- = %~Q 4'N u W ~~~u5 4
"( -
iy Q ~
\ fNNNN .
l' N.kNNNI AC ATOR
] / >
l
$\ # '
h SEAL .
N MAIN SEAT s
\ l CLOSURE SPRING PLUNGER ASSEMBLY N INLET
's
/
b
. - . ~ . - . . . . . . . \
C-10
)
GEN ER AL h ELECTRIC