U-601715, Provides Addl Info Re Proposed Amend to Plant Tech Spec 3/4.8.1.1, AC Sources - Operating, Per .Amend Will Allow Testing Frequency of Diesel Generator to Be Returned to Monthly

Provides Addl Info Re Proposed Amend to Plant Tech Spec 3/4.8.1.1, AC Sources - Operating, Per .Amend Will Allow Testing Frequency of Diesel Generator to Be Returned to Monthly

ML20058N096
Person / Time
Site:	Clinton
Issue date:	08/02/1990
From:	Jamila Perry ILLINOIS POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
JPS-0632-90, JPS-632-90, U-601715, NUDOCS 9008130212
Download: ML20058N096 (5)

Text

~~ ' ' ^

~ * ' ' ~ ~ ~ ~ ~ "' '~ ~ ~ ~ "^~ ' ~ ^

, ,7 ,

1 1 7) i . U-601715 D , 'h ; e '

L47-90 (08 - 02) -LP _

gq '

y =.. _8E 100a

, ILLINDIS POWER 00MPANYl cDNTON POWER STATION, iO. BOX 678. CUNTO(ILLINO;S 61727

.JPS-0632-90

,' 1 August 2, 1990

+

A L x -10CFR50.90 M

{t Nuclear" Regulatory Commission Document control. Desk

, Washington, D.C. 20555

= Subject Additional'.Infornation Regarding

, Proposed Amendment ~of Clinton Power Station Facility Ooeratina= License-No.'NPF-62 i

Dear Sir:

L By ' letteridated . ' July 11, 1990' (reference - U-601703) , Illinois' Power

"(IP): requestedi-amendment _ of Facility Operating License - No. NPF-62,:

Appendix - A _ ' Technical - Specifications, .for Clinton : Power Station (CPS).' . IP' requested, -in part, ' amendment of Technical.' Specification ' i

,3/4.8.1.1, "AC Sources-- Operating," to allow the testing frequency off r a Ldiesel generator to"be returned from - weekly to ' monthly if _ that .

diesel ~ generator'has completed seven consecutive failure-free demands

, Land the number of failures in the last 20 valid tests has'been reduced

- ton one or. L less,. : regardless of the number of - failures that diesel

' generator'has experienced in the last 100 valid tests. Additionally,

by,' letter : dated July 9, :1990 (reference U-601694), IP requested ' a

% ' waiver of compliancenin order to resume.a monthly testing frequency of

!thef Division I _ diesel generator (DG1A) a_t CPS . following: successful 1  ; completionc of . - testing at-a graduated, ' incremental increase , in - its

testing interval. Based upon discussions with the NRC staff and IP's ~ j

,^ . ,, understanding; that the identified Technical Specification amendment

t request is being : reviewed by the NRC on an expedited: basis, IP is o submitting.this letter to request ~ withdrawal of the; submitted-request b for L waiver of . compliance. This letter. also provides ~ additional 'or  !

> clarifying ;information regarding the two previous letters identified

'above. ,

i The current CPS Technical Specifications allow a diesel generator's

, test frequency .to be returned from weekly to monthly when that diesel 4 generator has completed seven consecutive failure-free demands and the number of failures in'the last 20 valid tests has'been reduced'to one ,

c .or'less.'

-However, the current CPS _ Technical Specifications require .i

r ,thatja weekly. testing. frequency be maintained if that diesel generator .!

has. experienced five.or more failures in the last 100 valid tests. As

, stated" above, IP's ' amendment request would -remove this restriction.

As identified in IP's amendment request, this proposed change has f previously been approved on the Perry plant docket.

f 8k 9008130212 900802

{DR ADOCK 05000461 I h PDC (

4 As of July 9, 1990, DG1A has experienced nine failures in the 98 valid

~

. tests-conducted since receipt of the CPS Operating License. OG1A has experienced six failures in the 56 valid tests conducted siace January 1, 1989. These failures (and resulting corrective actions) were

-summarized in IP's Special Report dated January 29, 1990 (reference U-601599) and a subsequent Special Report contained in Licensee Event Report (LER)90-011 (reference U-601688 dated June 14, 1990). As described in these reports, three of these test failures were the result of isolated failures of specific equipment. Specifically, the October 19, 1989 failure was the result of air entrainment in the fuel oil supply piping due to tha fuel oil day tank level being drained below the fuel oil supply suction during a . surveillance test; the

l. December 30, 1989 failure was the result of failure of the 2D-2E O . contact-pair of the K19. cc,ntrol set-up relay; and the May 15, 1990 failure was the result of Service Water system supply valves to the diesel generator jacket water heat exchangers being insufficiently open as a result of an error during system tagout restoration. l In contrast to the three failures just described, the failures which l occurred on November 20, 1989; December 11, 1989; and December . 27, 1989 were the result of slow starts. These slow starts were the subject of an Action Plan which was developed in order to identify the

, root cause of the slow start problem and resolve.those factors which l ' may - impact the ability of DG1A to routinely meet its starting requirements. This Action Plan was developed and implemented in a systematic manner in order to evaluate all potential contributors to the slow start problem of DG1A. This Action Plan is more . fully described-in IP's Special Reports dated January 11, 1990 (reference U- l 601589); January 29, 1990 (reference U-601599); March 29, 1990 -l (reference U-601632); and May 30, 1990 (reference U-601678).

Based on a more detailed review of the starting history of DC1A, IP Nuclear Station Engineering Department (NSED) personnel, with the l concurrence of.the diesel generator unit manufacturer, replaced the  !

governor on the 12-cylinder engine on December 27, 1989. (The l Division I and II diesel generators are each 12-cylinder and 16-cylinder tandem engine -

generator units.) The decision to replace the governor on the 12-cylinder engine centered on the fact that the preceding slow starts were characterized by a marked dacrease in engine acceleration midway up the acceleration curve. This distinct feature began. occurring following the replacement of the governor on the 12-cylinder engine in February 1989. (See Attachment 3 to IP's request for waiver of compliance letter U-601694 dated July 9, 1990.)

I' Although the governors of the tandem engines were of the same type p (Woodward EGB13P), the governor installed in February 1989 was a 1 different model (Model 9903-266) than that of the 16-cylinder engine l (Model 9903-265). The governor manufacturer stated that these two l models are interchangeable. Notwithstanding, IP replaced the governor on the 12-cylinder engine on December 27, 1989, such that matched governor models (Model 9903-265) currently exist on DG1A.

Actions taken by IP in response to failures of DG1A la November and December 1989 (in addition to implementation of the DG1A Action Plan) also included corrective actions for the K19 relay failure on December 30, 1989. These actions involved replacement of Potter & Brumfield

' type-MDR 5095 relays and an assessment of past performance of these relays at CPS. The assessment included an evaluation of the service

' life of these relays that considered both the number of anticipated relay _actuations and previous failure history. This assessment led;to the establishment of a periodic replacement interval for these relays.

=at CPS. The established replacement interval provides assurance of relay replacement prior to any anticipated failure, and is compatible-with the scheduling of other periodic diesel generator maintenance.

As a result of' implementation of the DG1A Action Plan, IP considers the root _ cause' of the slow- start problem of DG1A to be, pending any clarification from the Woodward Governor Company, misapplication of a Woodward EGB13P Model 9903-266 governor when used in conjunction with a Woodward EGB13P Model 9903-265 governor on tandem engines. Since replacement of the 12-cylinder engine Model 9903-266 governor with a Model 9903-265 governor on December 27, 1989, the start times of DG1A have been very stable. (Again see Attachment 3 to U-601694.)

Following replacement of the governor on the 12-cylinder engine, DG1A has experienced only two failures in the 31 valid tests performed.

(As previously noted, the failure on December 30, 1989 was the result of the failure of the K19 relay, and the failure on May 15, 1990 was the result of the Service Water supply valves to the jacket water heat exchangers being insufficiently open. These problems were not specifically related to the slow start problem and corrective actions

-have been taken to correct and prevent .-their recurrence.)

Additionally, DG1A has completed nine consecutive failure-free demands and has experienced only one failure in the last 20 valid tests (due  !

'to the aforementioned service water valve problem). As a result, IP believes that the slow start problem of DG1A has been resolved.

In addition, it should be noted that CPS site procedures currently require' System Engineers to trend and evaluate various diesel

_ generator parameters, including diesel generator start tines.. As-a result, the performance of the. CPS diesel generators is c antinuously monitored by knowledgeable personnel. Adverse  : rends or characteristics identified by Engineering personnel from this data

. require initiation of corrective action, or further investigation, as appropriate. Particular attention is paid to diesel generators which >

have experienced a relatively large number of failures in the last 100 valid tests (such as DG1A) . This program provides further assurance that an abrupt decline in the reliability of a diesel generator will be detected and acted upon in a timely manner. As a result, this program provides additional confidence that the reliability of the

. diesel generators at CPS will be maintained at a high level.

As described above, DG1A has demonstrated a very high degree of reliability since replacement of the governor on the 12-cylinder engine on December 27, 1989 and the K19 relay on December 30, 1989.

However, testing of DG1A at the current increased frequency (weekly) is resulting in unnecessary wear of this diesel generator. IP believes the long-term solution to the problem of unnecessary testing of DG1A is to obtain NRC approval of the proposed Technical Specification change, but given the period of time normally reqaired

, 1 s

forithe NRC to process proposed Technical Specification changes, IP

-initially believed that immediate action was required to prevent-excessive - testing of DG1A. IP therefore prepared and submitted- the request for waiver of compliance dated July 9, 1990. NRC approval of the waiver request would have allowed IP to increase the testing

- interval of DG1A in a stepwise manner until the testing frequency-was restored to monthly. At the same time, the proposed test plan would have provided adequate. time for the NRC to process ' the proposed Technical Specification change on a routine basis.

- IP's' proposed plan to increase the testing interval of DG1A in a stepwise manner was based upon a perceived test-interval dependent characteristic of the f ailures of DG1A. (That is, it appeared that start times increased as the elapsed time between starts increased.) .

This percepcion was based upon the following data taken between January 26, 1989 and December 27, 1989 regarding the DG1A start times and the elapsed time since the previous start:-

Number of Starts Time Between Start .ime Starts (T) Total >12 SEC 10-12 SEC <10 SEC T $ 3 days 37 3 3 31 3 days > T $ 5 days 4 2 0 2 5 days > T $ 10 days 5 2 2 1 10 days > T.$ 15 days 0 0 0 0 15 days > T $ 20 days 4 1 2 1 e 20 days > T $ 25 days

. 1 0 0 1  !

T > 25 days- 5 4 1 0 l 1

As can be seen from the above data, slower start times may loosely be '

correlated to greater elapsed time periods between starts; however, a test-interval dependent factor is by no means conclusive. It should be noted that the start times of DG1A had been erntic during this time period (see Attachment 3 to U-601694) and that there were also ,

many successful starts of DG1A during this time period, even when the I diesel generator had~not been started for several weeks. Therefore, l the existence of a test-interval dependent factor for these slow I starts is not, a conclusive- determination. l A graduated increase in test intervals would allow examination of the impact' (if any) of dormant periods of varying duration on DG1A start times subsequent to implementation of the DG1A Action Plan.

In keeping with the systematic approach and the depth of investigation and corrective e ction taken by IP during the performance of the DG1A 'i

. Action Plan, IP believes that a systematic approach to resuming a monthly testing frequency must be followed in order to verify that all possible contributors to the DG1A slow start problem have been 1 eliminated, even though a test-interval dependent factor to the previous slow start problem has not been conclusively established.

Therefore, resuming monthly testing (at least in a gradual manner) would provide an opportunity to conclusively demonstrate that a test-interval dependent failure mode does not currently exist.

l l

l

,c ,

0; .' ,

t-

It' is ; now - ~ IP's , understanding . that- IP's proposed -Technical.

.Spacification change to Technical Specification 3/4.8.l'.1 will be.

= processed ~by_the NRC on an' expedited basis. As a result, the:need for

'a waiver ~of-compliance, as requested _in IPfs July 9, 1990 letter, has E been greatly reduced and - is no longer considered to be necessary.

Therefore,-IP requests that the request for waiver of. compliance dated' July 9, 1990 be - withdrawn from further consideration. We appreciate

-your. understanding:and the prompt consideration you have given to this matter so that unnecessary testing of DGlA can be minimiEed.

Sincerely yours, j

,o l

. S. Per e Presi ont j

• DAS/alh

t. cc 'NRC Clinton Licensing Project Manager NRC Resident Office l 3 Illinois _ Department of Nuclear Safety l Regional Adminstrator, Region III, USNRC l

p <

l i

1 l

1 l

e ,

I

• a