ML20153B606
| ML20153B606 | |
| Person / Time | |
|---|---|
| Issue date: | 03/09/1988 |
| From: | Stello V NRC OFFICE OF THE SECRETARY (SECY) |
| To: | |
| References | |
| TASK-PII, TASK-SE SECY-88-074, SECY-88-74, NUDOCS 8808310022 | |
| Download: ML20153B606 (11) | |
Text
> . .
l
- !g#*% s c q
\...../
POLICY ISSUE March 9, 1988 SECW88-74 (lnfOrmation)
For: The Commissioners From: Victor Stello, Jr.
Executive Director for Operations
Subject:
EVALUATION OF LICENSEE PROGRAMS FOR TESTING AND INSPECTING CHECK VALVES
Purpose:
To provide the Commission with a report on licensees' check valve progrLms and their responses to INP0's Significant Operating Cvent Report (SOER) 86-3 on check valves.
Backaround: On November 21, 1985, a loss of power and water hamer event at San Onofre Unit I raised questions regarding the design, testing, and maintenance of safety-related check valves. During a Commission meeting on January 29, 1986, I proposed that the staff perform a generic analysis of check valve problems. In response to this proposal, on February 7,1986, the Commission issued a staff requirements memorandum (SRM) directing the staff to perform such an andlysis and provide the results to the Comission. By memoranda dated March 24 and July 30, 1986, the staff told the Commission about the actions taken by the staff ano those planned by the industry.
As part of the industry's response to the check valve issue, the Institute of Nuclear Power Operations (IfdO) issued SOER 86-3, "Check Valve Failures or Cegradation," on October 15, 1986. This report recccrended that preventive maintenanco procedures (i.e.,
inspection, testing, and repair programs) be established and that, as a minimum, the main steam, nuclear service water, auxiliary feedwater, diesel air start, suppression peci support, chemical end volume centrol, main feedwater, residual heat removal, and low-pressure injecticn systems be includec in the programs. -In addition, the report recorrended that licensees perform a design review of check valves, as installed, for correct sizing, orien-tation, type, and locaticn with respect to sources of turbulence
, (i.e., elbcws, tees, and pump discharge outlets). To facilitate implementation of the second recenvendation, the four nuclear steam supply system (NSSS) owners groups developed application CONTACT:
E. T. Baker, DRIS/NRR 49-24783 8808310022 G80818 '
.c The Comissioners .
guidelines for check valves. These guidelir.es were issued in Septemh r 1987.
On December 9, 1986, a feedwater line at Surry Unit 2 ruptured.
Subsequently, a number of previously undetected check valve failures in the feedwater system were discovered. Although these failures did not contribute to the event, the failures did identify generic check valve problems. The staff briefed the Comission on the Surry event on February 25, 1987. On March 11, 1987, the Commission issued an SRM directing the staff to provide (1). information on scheduled completion dates for the generic issues on check valves by April 3. 1987, and (2) an update on the generic review and its resolution by June 22, 1987.
In response to the first requirement, a memorandum to the Cormission dated March 30, 1987, outlined the status of actions taken by the industry and the staff. This memorandum noted that the staff was planning inspecticns at several sites to determine the effectiveness of the industry's actions.
SECY-87-168, dated July 9, 1987, described the results of the first two site' inspections of the total of five inspections conducted.
This paper discusses the results of the five site inspections performed to specifically examine check valve issues. The five site inspections were performed at Millstone Unit 1 Trojan, Fort Calhoun, Zion, and Crystal River.
Discussion: During the site inspections,'the staff reviewed the following items: check valve maintenance history; testing progrars requiredunder10CFR50,AppgndixJ,andSectionXIofthe ASME Boiler and Pressure Code ; testing procedures; the location of check valves with regard to sources of turbulence for those valves that were physically accessible; and the licensees' actions in response to SOER 86-3. Additionally, the staff reviewed the guidance that INPO has developed for I
I Current regulatory testing requirements (Appendix J to 10 CFR 50 and Section XI of the ASME Boiler and Pressure Vessei Code) provide for appropriate testing of check valves in certain applications. Appendix J requires a local leak-rate test of containment isolation valves each refueling outage.Section XI requires that all check valves within the Code boundary be exercised quarterly or during cold shutdown if exercising the valves at power is not practical, with normally open valves being shcwn to close er cessation of flow and/or reverse flow and normally closed valves being shown to open. Additicnally, the Code requires leak testing of valves where seat leakage is limited to a certain amount in order to show that the valve can fulfill its closed-position function. However, each licensee categorizes the valves and determines whether valves are tested under these programs. NRC review of l these programs is discussed later in this paper.
The Connissioners .
evaluating responses to SOER 86-3, sumary of the results of a questionnaire sent to licensees by INP0 requesting activities planned in response to SOER 86-3, and a sumary of the results of evaluations conducted by INP0.
During the NRC inspections, the inspectors reviewed system diagrams and chose check valves to be examined in detail, including testing and maintenance procedures. A total of bout 400 check valves were reviewed at the five plants.
The check valves were chosen from safety-related and balance-of-plant systems and consisted of inline and pump discharge check valves. In general, licensees inspected were not performing preventive maintenance. In the cases where licensees were performing preventive maintenance, this was done only because previ.ous failures, intpections or testing results indicated a need, not as part of a general preventive maintenance program. Moreover, for the main steam, nuclear service water, auxiliary feedwater (PWR only), diesel air start, suppression pool support (BWR only), chemical and volume control (PWR only), main feedwater, residual heat removal and low-pressure injection systems, the SOER recommends an aggressive general preventive maintenance program be established; this was not planned by the licensees inspected.
Of the approximately 175 check valves physically examined during the plant walkdowns, almost all were located in turbulent areas generated by pump discharges, elbows, and tees. This placement of valves does not conform with general guidelines developed by the valve manufacturers or the Owners Groups Task Force for valve performance and reliability. These guidelines specify at least 10 pipe-diameters downstream and 5 pipe-diameters
, upstream of flow disturbances as the distances which result ir.
l the best performance and reliat,ility of check valves. The widespread practice in existing designs of locating check valves in areas of turbulence increases the importance of performing preventive maintenance for these valves.
The test procedures used for the 400 valves also.were reviewed.
The itcensees ger.erally comply with the requiremen+s of their j NRC-approved Appendix J and Section XI test programs which apply to safety-related components. The Code requires that i
normally open valves be tested to show they will close, and normally closed valves be tested to show they will open. The staff interprets this to mean that testing should be performed to verify the ability of the valve to perform its required i
. safety function (s), i.e., to open, close, or both. In general, the licensees inspected do not recognize all the safety functicns of their check valves. This results in only one safety function being verified in some cases and valves being entirely omitteo
.o The Commissioners !
from the program in other cases. Because of the narrow interpretation of the Code by licensees, the inspectors found that the licensees were not performing adequate testing.
Additionally, in some instances, the test procedures were very difficult to follow, and the test acceptance criteria did not specify quantitative values for flow rates when flow was used as an indication of the valve reaching the fully open position or leakage when used to determine leak tightness.
Under 10 CFR 50.55a, each licensee is required to submit to the NRC for approval an inservice testing (IST) program for pumps and valves. This program is to be performed in accordance with the ASME Code,Section XI. Until recently, the staff did not have sufficient resources to complete more than a small fraction of the IST procram reviews, and interim approvals were issued pending a more comprehensive review by the staff. This problem is exacerbated by utilities frequently revising the IST program. A complete program review consists of examining the system function to ensure that the appropriate valve code category is specified. The reviews also verify IST program commitments that valves in the program are tested in accordance with the ASME Code,Section XI, requirements.
The IST programs are not required to contain implementing test proc 2dures and therefore, procedures are not reviewed as part of IST program reviews. However, a recently issued inspection procedure covering this area does contain general guidance for the inspectors on verifying the adequacy of check valve testing by reviewing procedures, acceptance criteria, and test results. The staff plans to modify this procedure to provide n: ore explicit inspection guidance.
Based on the inadequacies discovered during the five site inspections, it appears that the staff and industry need to revise the way that adequate and effective check valve tests are developed and implemented. The staff has begun initiatives directed toward improving the host. of problems in the IST area.
Consideration is being given to changing 10 CFR 50.55a to strengthen the.IST requirements and to clarify the scope of pumps and valves covered by IST. The revised rule would restructure the manner in which IST prograns are developed and implen'ented. More emphasis would be given towards better technical guidelines leading to less program review before program implementation, and more inspection of actual tests and procedures. The staff has also become r.cre a,tively involved in presenting NRC concerns to the A5ME c w ittee responsible for developing the testing standards. Staff views on the need to strengthen these standards also are being a brought to the attention of high level ASME ccanittees.
The Commissioners .
The SOER responses that were reviewed during the five site inspections indict.ted a general lack of aggressiveness by industry concerning the implementation of the SOER recomendations. (These SOER responses are internal licen:ee documents which are not forwarded to INP0. They are, however, available to INP0 site evaluation teams and NRCinspectors.) For the most part, the licensees were performing analyses and planning check valve programs at the time of the inspections. Howeve', actions outlined in the programs had not been implemented. A more detailed description of each of the SOER responses and the site inspection results is provided in Enclosure 1 to this paper.
The NRC and INP0 have repeatedly stressed the importance of the check valve issue. On February 27, 1987, the Director, Office of Inspection and Enforcement, wrote to the Chairman of the Owners Groups Task Force, urging that the task force reemphasize to. licensees the importance of evaluating check valve test programs and implementing the 50ER recomendations in a timely manner. In response to this letter, INP0 reemphasized the importance of check valves in a letter to utility executives dated March 25, 1987, that forwarded SOER 87-3, "Pipe Erosion." A copy of the INP0 letter is provided as Enclosure 2.
Shortly after that letters INP0 issued SOER evaluation guidelines to be used by the INP0 plant evaluation teams. The guicelines, which were made available to the staff, address each of the recommendations in the 50ER, offer information on the bases, and establish a consistent interpretation of the 50ER. Checklists were also developed by INP0 and made available to the staff.
The checklists are to be used by the INP0 plant evaluation teams i evaluating the acceptability of utility actions regarding the 50ER. Generally speaking, the staff feels that the guidelines and checklists will allow the INP0 plant evaluation teams to review utility actions in a thorough manner.
In addition, an INP0 letter of June 16, 1987 asked utility executives to comolete a survey on irplementation of the recomendations of SOER 86-3 and to return the information by July 17, 1987. As of September 2, 1987, out of 54 utilities, 41 responded to the survey with the remaining utilities committed to respond in the "near future." The initial results of the survey were made available to the staff and, although INP0 declined to release plant specific results to the staff, the results indicated the following: an average of 9 safety-related and balance of plant systems are being evaluated or inspected; i an average of 40 check valves per plant have been identified for inspection; emphasis is on valves greater than 2 inches in
The Comissioners .
diameter; some check valves have already been inspected either as part of existing inservice inspection programs or other programs initiated to detect check valve problems; as of September 2,1987, the inspections have revealed, according to INP0, a 19-percent check valve failure rate. With one exception, all the utilities that responded to the survey plan to expand the scope of their check valve inspection and testing programs. In addition, many of the utilities plan to and have conducted check valve inspection programs during outages.
INP0 has conducted plant evaluations that included review of licensee internal responses to SOER 86-3. However, the results made available to the NRC staff were not related to any specific plant; rather the results were a summary of INP0's findings.
INP0 found that the licensees' programs were either in progress or submitted to managerrent for approval. The INP0 sumary did not present any definite conclusions from these results and the NRC staff found the characterizations in the summary to be .too general to draw any conclusiens.
The Owners Groups Task Force issued "Application Guidelines for Check Valves in Nuclear Power Plants" in Septenher 1987.
The guidelines discuss such issues as application, performance characteristics, rrethodology for evaluation of check valves, methods to detect degradation, and problem areas associated with check valves. These guidelines shculd help utilities in developing a program in response to the SOER and ircorporating that program into maintenance activities.
Cerclusions: Based on the five site inspections conducted between April and October 1987 and our reviews of internal utility SCER respcnses, substantial additional efforts are still needed by the industry in respcnding to SOER 86-3, particularly with regard to preventive maintenance. The staff found that virtually ro praventive maintenance was being performed cr check valves. However, the Corrmir.sion will note that the PRC inspect 1 ens were conducted within a few trorths after INFO's letter emphasi: irs the importance of this matter (Enclosure 2). In addition, tbr Gwners Groups Task Force application guidelines were net issued until after three of the five inspections had been conducted.
In view of this, the small number of plants inspected, and the variance in the responses, the staff intends to conduct further inspections to evaluate the industry's centinuing effort in this area. These inspections will be carried out as a part of the inspections cf the maintenance area per the SRM on SECY 87-014
, In particular, future inspections will review whether licensees l have implemented the additional mai'itenance and inspection i
l
The Commissioners '
activities that are recommended as a result of the analyses performed. The staff understands that INPO plans to increase their emphasis on check valve programs during plant evaluations.
The staff will provide a subsequent report to the Commission on this subject by the end of the fiscal year.
(A-
[dM ctor Stello, J .
l Executive Director for Operations ;
Enclosures:
- 1. Inspection Results
- 2. INPO Letter of 3/25/87 DISTRIBUTION: -
Conunissioners OGC (H Street) ,
OIA GPA EDO '
ACBS SECY O
A 9
INSPECTION RESULTS The Millstone Unit 1 site inspection was performed April 20-24, 1987. At the time of the inspection, the Millstone licensee had not completed a formal internal response to the SOER; however, discussions with licensee representatives gave the NRC inspectors the impression that the licensee considered the ) resent testing programs sufficient and did not intend to make any significant c1anges.
The inspection report (50-245/87-09), however, details several weaknesses in the licensee's testing programs. These include questionable test acceptance criteria, test requirements that do not ensure valve functionality, and inservice test programs that do not clearly state which valves are subject to or exempt from testing.
The Trojan licensee had developed formal internal recommendations to the SOER.
These were reviewed during the May 18-21, 1987 inspection. The recommendations proposed to management that additional testing and inspections, as well as a check valve reliability study, be performed. The study would determine which check valves had the greatest effect on plant safety and reliability and propose. tests and inspections to increase the reliability of the valves. In addition to identifying new check valves to be tested, the program also would include a review of existing procedures to ensure that the test can detect check valve degradation or inoperability. Discussions with licensee personnel indicate the licensee is developing an aggressive program which should reduce the likelihood of failed check valves going undetected.
The Fort Calhoun site was inspected August 17-21, 1987. At that time, the licensee had only outlined a program of planndd actions in response to the 50ER. The licensee had started, but not completed, a review of volve locatien, valve maintenance history, and valve size. This is scheduled to be ccmplete ir. February 1988, thereby allowing the licensee tc determine what check valves should be included in a preventive maintenance progran. The inspection report, 50-285/87-22, cutlines weaknesses in Fort Calhoun's current testing and maintenance programs as well. Th9se include an instance of referenced testing procedures that did not cover the subject valves; an instance of maintenance orders that were written and cancelled because of lack of parts, with no evidence of any work having been performed at a later date; and valves not being tested in a manner that would ensure the valve would perform all -
required safety functions. In acdition, several unresolved items were identified concerning the type of testing performed on check valves.
The Zion sits inspection was performed September 14-18, 1987. The Commonwealth Edison Company (Ceco) corporate staff had developed the internal response and were performing analyses in response to the 50ER. This response, developed by CECO, was a comprehensive program which would evaluate check valve reliability and performance. At the time of the inspection, Zion plant personnel were implementing seme of the actions outlined in the program. These actions included measuring actual flow and reviewing valve locations. Some of the testing being currently performed at Zion on check valves, however, was found to have problems. For example, the Event V pressure isolation valves (P!V)
Enclosure 1
were being tested in series rather than individually and some valve tests were not valid because of the equalizing pressure on both sides of the valve.
(A proposed civil penalty has been issued in this matter.) To alleviate some of the weaknesses identified, Ceco is taking actions to clarify and upgrade Zion's program for measuring leakage through PIYs by increased monitoring and more stringent acceptance criteria. However, the failure of Zion's plant engineering staff to develop adequate implementing procedures for present check valve testing requirements warrants future inspections by NRC staf f to ascertain whether program improvements recommended by CECO corporate staff are adequately implemented by the Zion staff.
The content of the Crystal River licensee's internal response was good from the standpoint that it required analyses of critical flow, review of failure history, and inspection of valve location. The response to the 50ER was started late, but the October 5-9, 1987 inspection revealed that the licensee was now aggressively pursuing the issues. It was noted during the review, however, that not all check valves in the systems discussed in the 50ER were included in the internal response. Currently, testing is not being performed on all the safety functions of the valves, and the present implementing test procedures were difficult to follow and lacked quantitative acceptance criteria for flow rates when used as indication of the valve reaching the fully open position.
I 1
t l
l i _ _ - _
PW() x Nuctar Pceer Operations t1oo carce 75 Parkway Suae1500 Marta, Geoga SCC 39 Tokpecoe 4o4 e63 36co March 25, 1987 Wr. R. Patrick Mcdonald Senior Vice President Nuclear Generation Alabama Power Company 600 North 18th Street '
P. O. Box 2641 Sirmirgham, Alabama 35291 0400
Dear Mr. Mcdonald:
The purpose of this letter is to provide information and recourwne action relating as follows: to tw areas in which the industry has experienced recent problees, o piping failures due to steam and water erosior/ corrosion o
check valves that fail to function as desigred This letter also transmits SOER 87-3.
Picing Failure Problems The failure of the main fethster system pipirg at Svery Unit ? from arosien/cerosict resulted in personnel irjuries, including four fatalities, and subscantial equipment damage. IMPOSignificantEventReport(SER)1-87 E*csion/Cor"osion - Induced Failure cf N P'oinc, issued January 7,1937, desertoes this event. Subsequently, virginia Power provided more detailed informatien to the industry in five simultanecus regional presentatiens on Feb*uary 10, 1987.
IMpo has new crepa-e4 a new co v erensive Significant Operatirs Expe-ience Report (50ER) 87-3. Pioe Failures ir Hich-erariv Systens Due to Ero s t en/Co-ve si en.
Lnisnee to acministrative points of contact with this letter.50ER Senior 87-3, along managemnt should De aware of the important reccanendations in tnis SCIR.
SOER 87-3 uses th0 infomation prowidad by Virginia Power, werk dore by EPRI, and builds on two earlier SEE IN products: SciR 8211. Erosion of 5 tem P90inc sad aesultire Failure, issued November 17, 1982; and SER 23-65, water Pi;e dall Eresien 0:wnstrem of Flow Restrictino Oevices. (50ER 87-3 incertiorates supersede SER 23 65 anc SER 187 as references ano can be considered to them.)
Enclosure 2 n ; . ._._. . ;. a .: : ;-.~.- .~ . A .~ ~*c~~~ ~ -~ ~
,- w.,
March 25,1987 Page Two' ,
Of special note in the reference package are copies of SOER 82 11 (Reference furniseed to administrative 1 of 50ER 87-3) pointsand SIR 23 85Checks of contact). (included for pipe in N the nning reference mat shoule have been done in respense to these d:cuments, but many uti Mies have limited eariter checks to steam systems or systens witn expected tve-phase flow. The Surry event, as well as others, shows that checks It for erosien in is recortsended high temperature, single phase (water) systems is needed.
that you review your utility's work in response to 50ER 8211 and SIR 23-85 to detemine if actions taken were adMutte and to ensure that these actions are integrated with yeur response to 50ER 87 3.
Checir Valve Prebtems failures. ThesoccedproblemthatreqWesmanaa$ementattentionischeckva Since snat time, addi-
- 15. 1966, covering check valve f airure or cegracation. Such failures have the potential tienal chMk valve f ailures have occurred. systeres. SOEe 86-3 45 a "yell:w for cefeating essential reactor core cec 1NWM!e we setieve this classificati:n tac" (as opposed to a "red t e") 50ER.
remains appropriate, we ccMicer it important that senior mar 49emert ensure that the recerseendations in tMs 50ER are being aggnssively accressed. For
- f rrformation in this regard, EPD.I .is working eith an tecustry grous, under The guidethe is auspices of MUMARC, to develo; a chtek valve application guide.
expected to be available late this year. In the interim and under the assviso-tion that nost check valves installed in plant systems are used in acceptabletse applicatichs, the two recedinandatiC Ps in 50ER 66 3 should be carried c.t t:
ma.xista ==actical extent. A cecy of $0!R 86-3 is o mvi(td as Attac h t 3.
Sincerely.
Zact T. Pate P re s ident LTP:No Attschnents (As stated cove)
__ - - . - _ _ _ , . , . _ , _ _ - _ -