ML20059H788

Responds to Ltr Expressing Committee Concerns Re trigger-value Approach Proposed by Staff in Reg Guide 1.160.Comments Are Appreciated But Unnecessary to Revise RG 1.160 or Ask NUMARC to Revise Any NUMARC Documents

ML20059H788
Person / Time
Issue date:	10/29/1993
From:	Taylor J NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
To:	Wilkins J Advisory Committee on Reactor Safeguards
Shared Package
ML20059D953	List: ML20059D951 ML20059H788
References
REF-GTECI-B-56, REF-GTECI-EL, RTR-REGGD-01.160, RTR-REGGD-1.160, TASK-B-56, TASK-OR ACRS-R-1545, NUDOCS 9311100216
Download: ML20059H788 (2)

Text

,

f **%g h

!iie/

S UNITED STATES NUCLEAR REGULATORY COMMISSION

('

s,

/

W ASNNGTON, D.C. M55M@i October 29, 1993 J. Ernest Wilkins, Jr., Chairman Advisory Comittee on Reactor Safeguards U. S. Nuclear Regulatory Comission Washington, DC 20555 i M5

Dear Chairman Wilkins:

~

In your letter to the Chairman, dated September 22, 1993 you expressed the Comittee's concern regarding the " trigger-value" approac,h proposed by the staff in Regulatory Guide (RG) 1.160 to resolve Generic Issue B-56, " Diesel Generator Reliability."

You stated that it is categorically impossible to demonstrate the reliability of emergency diesel generators (EDGs) using this method and that footnote 3 added to RG 1.160 does not satisfactorily resolve the Comittee's concern.

You also stated that the generic letter should be revised to remove the inconsistency regarding " problem diesel."

The staff agrees with the ACRS that conformance of individual EDG's with trigger values cannot be taken to mean, in any statistical fashion, that the EDG has demonstrated achievement of high reliability values such as 0.950 or 0.975, which licensees may have comitted to when implementing the station blackout rule.

Footnote 3 was added to RG 1.160 to emphasize this fact.

licensees chcose to use trigger values, they can be used as alert levels to If indicate EDG performance problems that may need additional monitoring or corrective actions.

In addition to this prospective choice RG 1.160 indicated that licensees also could use other performance me,asures such as establishing down-time constraints to ensure high EDG availability.

and individual licensees retain considerable flexibility in this matter and NUMRC NUMRC maintenance guidance, 93-01 and RG 1.160. understan in the NUMRC guidance and RG 1.160 do not invoke any requirements forTrig accelerated testing.

We continue to believe that NUMRC and individual licensees understand the proper usage of the trigger values as set forth in the NUMRC guidance and RG 1.160.

To avoid any misconceptions on the statistical signific remove reference to trigger values and the associated footnote from the generic letter.

of EDG reliability and availability.The staff's goal is for the licensee to mai achieved through implementation of the maintenance rule, which includ performance of a detailed root-cause analysis of all ED failures, and implementation of an effective performance-based preventive CONTACT: Om Chopra, EELB/DE 504-3265 74 t/wVf4

~...

,,d-

e J. Ernest Wilkins, Jr. maintenance program. The generic letter will remain voluntary in nature for licensees to adopt and it is expected that those licensees most affected by 1

requirements for accelerated EDG testing (and an inordinate amount of EDG failure reporting) will avail themselves of the revised technical J

specifications offered by the generic letter.

However, the staff will retain the requirement that licensees must provide a comitment to comply with the provisions of the maintenance rule and the guidance of RG 1.160 for emergency diesel generators as a necessary condition to eliminate the existing technical specification (TS) for the accelerated EDG testing and reporting requirements. We will also revise the generic letter to clarify that the need for the current prescriptive requirement in the plant TS for accelerated testing of a " problem diesel" has been eliminated with implementation of the maintenance rule.

We appreciate your comments and recomendations; however, we believe that it is unnecessary to revise RG 1.160 or ask NUMARC to revise any of the NUMARC documents to further clarify the issue of trigger values, which we believe is fully understood not to be a measure of reliability.

Sincerely,

/

apes M. Jh [ lor

[

fecutive Director for Operations cc:

The Chairman Comissioner Rogers Comissioner Remick Comissioner de Planque SECY 0GC OCA OPA i

i

[p.a now%,

UNITED STATES

' 'g NUCLEAR REGULATORY COMMISSION g

{,

,y ADVISORY COMMITTEE ON REACTOR SAFEGUARDS o

.7 W ASHINGTON, D. C. 20555 h

• sse e September 22, 1993 The Honorable Ivan Selin Chairman U.S.

Nuclear Regulatory Commission Washington, D.C.

20555

SUBJECT:

PROPOSED GENERIC LETTER REGARDING REMOVAL OF ACCELERATED TESTING AND SPECIAL REPORTING REQUIREMENTS FOR EMERGENCY DIESEL GENERATORS FROM PLANT TECHNICAL SPECIFICATIONS During the 401st meeting of the Advisory Committee on Reactor Safeguards, September 9-10, 1993, we reviewed the subject generic letter (GL).

During this

meeting, we had the benefit of discussions with representatives of the NRC staff and NUMARC.

We also had the benefit of the documents referenced.

The staff has informed us that this version of the proposed GL reflects consideration of the comments made by the Committee to Review Generic Requirements (CRGR).

The proposed GL has been issued for public comment in accordance with our agreement that this could be done prior to our review.

The proposed GL would alivw licensees to request removal of the Technical Specification (TS) provirions for accelerated testing and special reporting requirements for the emergency diesel generators (EDGs).

When requesting this license amendment, licensees must, however, commit to implement a maintenance program for monitoring and maintaining EDG performance consistent with 10 CFR 50.65,

" Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants,"

and of Regulatory Guide (RG) 1.160,

" Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," that was developed by the staff to provide guidance for complying with the provisions of the Maintenance Rule, 10 CFR 50.65.

6 In our April 26, 1993 (revised June 24, 1993) report on the draft version of RG 1.160, we noted that:

On many occasions, we have provided comments on the trigger-value approach proposed by the staff to resolve Generic Issue B-56, " Diesel Generator Reliability." The proposed regulatory guide for implementing the Maintenance Rule explicitly endorses the trigger-value

f. <1 % ~ G O S ( G '

D.

jf

k, i

The Honorable Ivan Selin 2

i procedure for " monitoring emergency. diesel generator (EDG) performance against EDG target reliability levels."

l It is categorically impossible.to demonstrate the reliability of EDGs using this method.

We remain i

strongly opposed to its use for this purpose and continue to recommend that the staff's implementation guidance for i

the Station Blackout Rule, 10 CFR 50.63, be revised to l

deal with this issue.

When this is done, the regulatory l

guide should be appropriately revised.

I The staff's response was to include a footnote in RG 1.160 which i

states:

The triggers-are intended to indicate when emergency diesel gener ator performance problems exist such that additional monitoring or corrective action is necessary.

It is recognized that it is not practical to demonstrate by statistical analysis that conformance to the trigger values will ensure the attainment of high reliability, with a reasonable degree of confidence, of individual EDG units.

We do not believe that this footnote satisfactorily resolves our concern.

Regulatory Guide 1.160 endorses Section 12.2.4 of NUMARC 93-01,

" Industry Guidelines for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," which, in turn, references Appendix D of NUMARC 87-00, Revision 1,

" Guidelines and Technical Bases for NUMARC Initiatives Addressing Station Blackout at LWRs."

Each of these documents clearly implies that use of the " trigger values and monitoring methods" (as described in Appendix D of NUMARC 87-00, Revision 1) provides an acceptable means of monitoring EDG target reliabilities of 0.95 or 0.975 in accordance with the intent of 10 CFR 50.63 for coping with station blackouts.

(See, for example, the language of the first paragraph of the introduction to Appendix D of NUMARC 87-00, Revision 1.)

It can't be both ways!

We strongly recommend that the staff-and NUMARC collaborate in resolving this matter by appropriate revision of these documents.

We have had a longstanding concern that the EDGs at many nuclear power plants are being subjected to excessive and unnecessary surveillance testing and other testing es required by TS limiting conditions for operation, and that such testing may actually be degrading the reliability of these machines.

Data for the years 1988 to 1991, provided to us by NUMARC, show that some EDGs are subjected to start testing only 12 to 15 times each year, while i

other EDGs are tested over 100 times each year.

)

A h

The Honorable Ivan Selin 3

This disparity in testing frequencies results, in part, from the wide variation in relevant TS requirements that were negotiated with licensees over the years.

The fact that this situation has existed for so many years reflects badly on both the staff and licensees with respect to their effectiveness in dealing with an acknowledged problem having safety inplications.

We believe that this proposed GL is an important step in achieving a more rational testing program.

In addition to our recommendation that RG 1.160 and the NUMARC documents on which this proposed GL is based be revised to reflect statistical reality, we believe that the language of the proposed GL needs improvement.

The proposed GL quotes a statement from RG 1.160 that triggers and testing of " problem diesels" will be addressed separately by the NRC.

In the next paragraph of the GL, licensee commitments required for approval of the removal of accelerated testing and special reporting requirements from the TS are described, including the need for a commitment to RG 1.160.

A statement is then made that these actions are intended to close the issues of triggers and testing for " problem diesels."

The staff should clarify this apparent contradiction and state clearly that the former prescriptive requirement for accelerated testing has been eliminated by this proposed generic letter.

Sincerely, J.

Ernest Wilkins r.

Chairman I

References:

1.

Memorandum dnted August 13, 1993, from J. Larkins, ACRS, to B.

i

Grimes, Office of Nuclear Reactor Regulation,

Subject:

Proposed NRC Generic Letter " Removal of Accelerated Testing and Special Reporting Requirements for Emergency Diesel Generators from P'. ant Technical Specifications" 2.

Memorandum dated August 12, 1993, from G.

Marcus, Office of Nuclear Reactor Regulation, for J. Larkins, ACRS, forwarding proposed NRC Generic Letter Regarding Removal of Accelerated Testing and Special Reporting Requirements for Emergency Diesel Generators from Plant Technical Specifications 3.

U.S. NRC Regulatory Guide 1.160, " Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," June 1993 4.

NUMARC 93-01,

" Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," May 1993 5.

Appendix

.D.

"EDG Reliability Program" to NUMARC Report,

" Guidelines and Technical Bases for NUMARC Initiatives E

o

lN P

The Honorable Ivan Selin 4

Addressing Station Blackout at LWRs," NUMARC 87-00, Revision 1,

August 1991 6.

SECY-93-044 dated February 22, 1993, for the Commission from i

James M.

Taylor, NRC Executive Director for Operations,

Subject:

Resolution of Generic Safety Issue B-56,

" Diesel Generator Reliability" 7.

Letter dated April 26, 1993 (Revised June 24, 1993), from Paul Shewmon, ACRS Chairman, to Brian K. Grimes, Office of Nuclear Reactor Regulation,

Subject:

Implementation Guidance for the Maintenance Rule 8.

Letter dated August 7, 1992, from Alex Marion, NUMARC, to Paul

Boehnert, ACRS, providing Industrywide Emergency Diesel

[

Generator Reliability Data 9.

Letter dated December 18, 1992, from Raymond Fraley, ACRS, to Alex Marion, NUMARC,

Subject:

Industrywide Emergency Diesel f

Generator Reliability Performance Data 10.

Letter dated March 1,

1993, from Alex Marion, NUMARC, to Raymond

Fraley, ACRS, responding to questions regarding Industrywide Emergency Diesel

. Generator Reliability Performance Data l

l

_