ML20029C886

Proposed Tech Specs Re EDG Slow Start Surveillance Testing

ML20029C886
Person / Time
Site:	Comanche Peak
Issue date:	04/25/1994
From:	TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC)
To:
Shared Package
ML20029C883	List: ML20029C882 ML20029C886
References
TXX-94118, NUDOCS 9405020286
Download: ML20029C886 (20)

Text

as .

'i -

l o- ,

s  ;

./,

-ATTACHMENT 3 TO TXX-94118 AFFECTED TECHNICAL SPECIFICATION PAGES (NUREG-1468)

? .I

?

.i i

-- l -

9405020286 940425 1 PDR ADOCK 05000445 .l P- PDR

Attachment 3 to TXX-94118 Page 2 of 2 ELECTRICAL POWER SYSTEMS LIMITING CONDITION FOR OPERATION (Continued)

ACTION (Continued offsite source restored, restore at least two offsite circuits to GPERABLE status within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> from time of initial loss or be in at least HOT STANDBY within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in COLD SHUTDOWN within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />.

f. With two of the above required diesel generators inoperable, demon-strate the OPERABILITY of two offsite A.C. circuits by performing Surveillance Requirement 4.8.1.1.la, within I hour and at least once per 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereafter; restore at least one of the inoperable diesel generators to OPERABLE status within '2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> or be in at least HOT STANDBY within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in COLD SHUTDOWN within the fol-lowing 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />. Restore at least two diesel generators to OPERABLE status within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> from time of initial loss or be in at least HOT STAN0BY within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in COLD SHUTDOWN within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />.

SVRVEILLANCE REOUIREMENTS 4.8.1.1.1 Each of the above required independent circuits between the offsite transmission network and the Onsite Class IE Distribution System shall be:

a. Determined OPERABLE at least once per 7 days by verifying correct breaker alignments, indicated power availability, and

b. Demonstrated OPERABLE at least once per 18 months during shutdown by transferring (manually and automatically) the 6.9 kV safeguards bus power supply from the preferred offsite source to the alternate offsite source.

4.8.1.1.2 Each diesel generator shall be demonstrated OPERABLE:

a. In accordance with the frequency specified in Table 4.8-1 on a STAGGERED TEST BASIS by:

1) Verifying the fuel level in the day fuel tank,

2) Verifying the fuel level in the fuel storage tank,

3) Verifying the fuel transfer pump starts and transfers fuel from  !

the storage system to the day fuel tank, )

4) Verifying the diesel starts from ambient condition and ccel rates )

to_at least 441 rps in less than or equal to 10 seconds

• 4 i s.

M .1 % + d All oihu en3,nc Sinis fhr lNrformance o th6TJifsillqqq' d#ncUones3JA*8"nchil2X '@MgQd"" h

• All planned diesel engine starts for the purpose of this surveillance may be

@g The receded diewi by a prelube dtwear SM hoperiod in accordance (to se.conds) with vendor she be. vtWFe _tecommendati A at-lv<st once )

COMANCHE PEAX - UNITS 1 AND 2' 3/4 8-3= -

]

J J

a ENCLOSURE 1 TO TXX-94118 NUREG-1431 - TECHNICAL SPECIFICATIONS FOR WESTINGHOUSE PLANTS SEPTEMBER 1992 PAGES 3.8-6, 3.8-8 AND 3.8-17 i

s

'.j i

a l

I I

c, . ,;_, . ..:.- , , _ _ ., . _ _ _ .-

Enclosure 1 to TXX-94118 l Page 2 of 4 AC Sources--Operating

=3.8.1 SURVEILLANCE REQUI9EMENTS SURVEILLANCE FREQUENCY l

i l

SR 3.8.1.1 Verify correct breaker alignment and 7 days  !

indicated power availability for each

[ required] offsite circuit.

SR 3.8.1.2 -------------------NOTES-------------------

1. Performance of SR 3.8.1.7 satisfies this SR.

2. All DG starts may be preceded by an engine prelube period and followed by a warmup period prior to loading.

L 3. A modified DG start involving idling l and gradual acceleration to synchronous speed may be used for this SR as recommended by the manufacturer. When modified start procedures are not used, the time, voltage, and frequency tolerances of SR 3.8.1.7 must be met.

Verify each OG starts from standby As specified in conditions and achieves steady state Table 3.8.1-1 voltage t [3740] V and 5 [4580] V, and frequency a [58.8] Hz and s [61.2] Hz.

(continued) i WOG STS 3.8-6 Rev. 0,.09/28/92

Enclosure 1 to TXX-94118- I Page 3 of 4 AC Sources--Operating 1

l 3.8.1 l j

SURVEILLANCE REQUIREMENTS (continued) .;

SURVEILLANCE FREQUENCY SR 3.8.1.7 -------------------NOTE------------------_-

All DG starts may be preceded by an engine prelobe period.

Verify each OG starts from standby 184 days condition and achieves in s [10] seconds, voltage t [3740] / and 5 [4580] V, and frequency a [58.8] Hz and 5 [61.2] Hz.

SR 3.8.1.8 -------------------NOTES-------------------

1. This Surveillance shall not be performed in MODE 1 or 2.

2. Credit may be taken for unplanned events that satisfy this SR.

Verify [ automatic [and] manual] transfer [18 months]

of AC power sources from the normal offsite circuit to each alternate [ required]

offsite circuit.

(continued) .

WOG STS 3.8-8 Rev. O, 09/28/92

Enclosure 1 to TXX-94118 Page 4 of 4 AC Sources-0perating 3.8.1 Table 3.8.1-1 (page 1 of 1)

Diesel Generator Test Schedule NUMBER OF FAILURES IN LAST 25 VALID TESTS (a) FREQUENCY s3 31 days-a4 7 days (b)

(but no less than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />)

(a) Criteria for determining number of failures and valid tests shall be in accordance with Regulatory Position C.2.1 of Regulatory Guide 1.9, Revision 3, where the number of tests and failures is determined on a per DG basis.

(b) This test frequency shall be maintained until seven consecutive failure free starts from standby conditions and load and run tests have been performed. This is consistent with Regulatory Position [ ], of Regulatory Guide 1.9, Revision 3. If, subsequent to the 7 failure free tests, 1 or more additional failures occur, such that there are again 4 or more failures in the last 25 tests, the testing interval shall again be reduced as noted above and maintained until 7 consecutive failure free tests have been performed.

Nrte: If Revision 3 of Regulatory Guide 1.9 is not approved, the above table will be modified to be consistent with the existing version of Regulatory Guide 1.108, GL 84-15, or other approved version.

i WOG STS 3.8-17 Rev. O, 09/28/92 I

i

.I' e

,, .j

' t ENCLOSURE 2 TO TXX-94118 NUREG-1366, IMPROVEMENTS TO TECHNICAL SPECIFICATIONS SURVEILLANCE REQUIREMENTS DECEMBER 1992 i

PAGES 16 AND 53 THROUGH 59  ;

.l.

1

i i

l j

Enclosure 2 to TXX-94118 Page 2 of 9 3 General Findings water hammer could not occur), operators and health its best efficiency point which degrades the pump. .,

physicists who must climb ladders to the top of a tank farm discussed in more detail later.

must be dressed in protective clothing because the tank farm is a radiation area. Electrical and electronic equipment wears or breaks fror unplugging and removing equipment from cabinets fc At another PWR, the surveillance on the containment testing or from lifting leads and using jumpers.

area high radiatien monitor requires that a heavy (be-cause of shielding) high-level source be lowered to the ne use of valves for isolation or flowpath change caust monitor. leaks around the valve packing and other valve or vah The current industry effort on advanced reactor designs should include a study of how all required surveillance He testing of an emergency diesel generator in its eme:

testing will be performed in order to (1) minimize the gency mode induces thermal stresses and causes othc possibility of a transient caused by testing, (2) minimize . problems which are discussed later in this report.

the burden on phnt personnel who will have to perform these tests, and (3) minirnize the radiation exposure re. Rus, the importance of the test must be balanced again-ceived by people in performing the required testing. considerations of wear on equipmer.t as well as on othe considerations.

3.10 Surveillance Testing and Power Reductions 3.12 Surveillance Testing on a Staggered Test Basis Some surveillance tests in both PWRs and BWRs require power reductions in order to prevent a transient that can Staggered testing is the scheduling of tests for the subsys tnp the reactor. In a PWR, a power reducuon is necessary tems or trains of a system in which the survettlance tet for strokmg turbine valves. In a BWR, there are three interval is divided into a subinterval for each subsystem o tests that typically require a reduction in power: MSIV train, testing, control rod movement testing, and turbine valve stroking.nerefore, another incentive for elimmating un. The advantage to testing on a staggered test basit .a -

necessa.'y testing is the increase in capacity factor if such the chances of a common-mode failure and equipmen -

testing were done at a reduced frequency. unavailability are reduced. A staggered test basis can hay disadvantages.

3.11 Surveillance Testing and One resident inspector stated that, at his plant, this typ Equipment Wear of testing requires additionallicensed operators and over -

time for operators. It also requires more individue Equipment is sometimes operated in a different way for entries into protection cabinets which causes schedulin survet!!ance testing than the way it would be used per. problems for licensees and may increase the chance of forming its design function. A simple example is an injec- reactor trip. It can also extend the time required to per tion pump which, when tested, recirculates water back to form surveillance tests by requiring initial setup time f(

a tank through a line that is smaller in diameter than the test equipment to be repeated for each test rather thr normal injection line, thus making the pump operate off setting up just once.

NUREG-1366 16

Encloeure 2 to TXX-94118 Page 3 of 9 10 ELECTRIC POWER 10.1 Emergency Diesel Generator Research done by NRC and the industry has shown that krveillance Re9uirements (PWR' s me f the assuptions in the analysts of the LOCA, 11WR) required by 10 CFR 50.46 and Appendix K to 10 CFR Pan 50, are very conservative. In addition to the conserv.

ative nature of the regulations, other conservatisms have Corresponding to their importance to safety, emergency diesel generators (EDGs) have the most detailed Techni- been included in the vendors' LOCA models.

SECY-83-472 provides a method to eliminate those con-cal Specifications surveillance requirements of any piece servatisms not specifically required by the regulations.

of mechanical or electrical equipment in a nuclear power plant. Surveillance requirements for EDGs are currently Under the sponsorship of the Electric Power Research based on Regulatory Guides 1.108 and 1.9. Institute (EPRI), calculations were performed, usmg the methods given in SECY-83-472, that show margin ts The safety function of the diesel generators is to supply ac available to the criteria of 10 CFR 50.46 that could be electncal power to plant safety systems whenever the used to extend the EDO start and load times. Studies preferred ac power supply is unavailable. nrough sur- done for a typical four-loop Westinghouse PWR veillance requirements, the ability of the EDGs to meet (NSAC-130) show that the diesel start and load time their load and timing requirements is tested and the qual- could be increased to 45 seconds from 10 seconds. De i:y of the fuel and the availability of the fuel supply are 45.second stan time is limited by environmental qualifi-monitored. cation considerations of equipment in containment. De calculated peak cladding temperature was below 2200 *F.

As part of the resolution of Unresolved Safety Issue (USI)

A-44, " Station Blackout,, the NRC staff has prepared A similar calculation fora typical BWR 4 showed that the Regulatory Guide (RG) 1.155 to provide guidance on diesel generator stan and load time could increase to 118 EDG reliability levels. RG 1.155 also specifies that the seconds (NSAC-%), and still be within acceptable limits.

rehable operation of onsite emergency ac power sources should be ensured by a program designed to maintain and However, for the purpose of evaluating the effects of monitor the reliabtlity level of each power source over surveillance testing, start and load times should be ad-dressed separately.

ttme to ensure that the selected reliabilitylevels af e being achieved.

A fast start (i.e., start and acceleration to synchronous speed at full fuel rack position) has the potential to accel- ,

Genenc Safety Issue (GSI) B-56, " Diesel Reliability," erate the degradation of the diesel generator if conducted l

was established to develop guidelines for an EDG reli- with ut the benefit of a prelube period. However, I abtlity program. In addition to these efforts, the Office of Prelubricating diesel generators is now common practice, Nuclear Regulatory Resean:h (RES)is conducting the and any remaining negative effects of fast starts are mim.

Nuclear Plant Aging Research (NPAR) Program, which is mal. Nevertheless, fast starts can be eliminated on some intended to resolve techmall safety issues related to the 6esel genento@ changg h gwernor cMgwa@n, l aging degradation of equipment important to reactor but only at the cost of reducmg diesel generator reliabil-safety. An irnportant part of this program is the study of ny, by clinunating a redundant perspeed protecticn fea.

the agm.g of emergency diesel generators. ture, that is, the backup mechamcal governor. In this case, the gain associated with slow starts does not offset t he loss of the backup overspeed protection.

He results of these programs were reviewed as part of this study to determine bow these programs will affect Fast loading (i.e., zero to full load in 120 seconds or less) surveillance requirements for EDGs in the Technical during surveillance tJing is, on the other hand, the most Specifications.

significant cause of accelerated degradation of diesel gen-erators. It can cause rapid piston ring and cylinder liner The current performance requirements are stnngent. wear (up to 40 times greater than normal wear) and The EDG mun start on any of several signals (e.g, man- should be elimmated in favor of loading in aa:ordance ual actuation, safety injection, or loss of normal power to w th the manufacturer's recommendations, except for the an emergency bus), increase to rated speed in a short time 18-month loss of offsite power (LOOP) test. Manufactur-(e.g.,10 seconds), and pick up its emergency load in blocks ers' recommendations for diesel genentor loading can be at programmed times (load sequencing).These times are 30 minutes or more to reach full load.

relatively short and are set by the requirements of the In an actual emergency, loads will be sequenced onto a large-break loss-of-coolant accident (LOCA).

diesel generator in approximately 60 seconds. This 53 NUREG-1366

Enclosure 2 to TXX-94118 i Page 4 of.9 10 Electric Power constitutes fast landmg of the diesel generator regan11ess By a letter dated December 7,1987 (Murphy,198i,. ..ie !

of whether the sequencing started at 10 seconds or at licensee for Vermont Yankee submitted a request to either 45 or 118 seconds (as suggested by the studies revise this survedlance/ alternate testing requirement, referenced above) after the diesel generator starts. The NRC sinff reviewed this proposed change to the !

Hence, design changes for slower diesel generator start. Technal Specifications and requested a more quantita- l ing and acceleration would not significantly reduce the tive analysis than had been originally supplied. In re '

degradation of diesel generators which is inherent with sponse to this request, the Vermont Yankee N tr rapid loading that is necessary to meet safety analysis Power Corp. submitted an analysts dated July 15, +988 1 requirements. (Capstick,1988), using reliability methods. De NRC 1 staff is reviewing that submittal. I ne NRC staff recommends that all testing of the diesel l generators, with the exception of the LOOP tests which De analysis quantified the unavailabilities of the systems l are performed with and without an ESF signal once each when required to perforrn their intended function upon !

refueling, be performed by gradual loading in accordance demand, both with and without alternate testing. Two J with the manufacturer's recommendations. systems were chosen for detailed analysis: the core spray .

EDG tests were typically started with the EDG initially at ambient conditions with no prelubrication or warmup De pros and cons of testing were quantified, that is, time. Generic Letter (OL) 84-15 changed this, stating (1) tne decreased potential for an undetected failure due that "[t]icensees are encouraged to submit changes to to the alternate testing and (2) the increased unavailabil-their Technical SpectTication(s) to accomplish a reduction ity due to (a) the alternate testing and (b) repair of ;

in the number of [ cold) fast starts." A typical technical demand-related and test related failures. Other disad-specification was included in GL 84-15 which required a vantages to alternate testing which were not quantified in start from ambient conditions every 184 days rather than this study are:

every month.

(1) reduced reliability due to equipment degradation Some nonstandard Technical Specifications require that, from emve testing -

i with an inoperable EDG, not only the remaining operable l

diesel generator (s) must be tested at a higher frequency (2) potential for unnecessary shutdowns that resu.. in '

than normally required but, in addition, other emergency plant transients and challenges to safety systems i equipment such as the emerlency core cooling system (ECCS), safety related cooling water pumps (e.g., service O) potential for plant transients initiated during sur.1 water), and other power supplies also must be demon.

veillance tests strated operable. Dis testing must commence "immedi-atety" upon discovering that a diesel generator is inoper. (4) diversion of operating personnel time and attention able.

(5) increased radiation exposure to operating personnel Some nonstandard Technical Specifications also require '

that if a train or subsystem of certam safety systems other than the diesel generatprs (for example, a low-head De analysis showed that, for the core spray system, alter-safety-tnjection purnp of the ECCS)is declared inoper- nate testag (which is required daily by the Verrnont Yan-able, not only the other train of the particular system but kee Technscal Specifications) prcduced unavailabilities at also other equipment of the emergency core cooling sys- least a factor of 4 greater than monthly testing. For the -

tems and the essel generators must be tested. Thus, a diesel generators, this factor was about 3. l failed train is one safety system can cause a great deal of l testing of appassusly unrelated systems. His type of test. Considering this analysis and similar conclusions in j ing is called "sherante testing." NUREG-1024, the staff recommends that alternate test 1 ing requirements be deleted from the Technical Specifi-An example of this in matris form is shown in Table 10.1 cations for all plants so that the failure of a train or (from a letter from Vermont Yankee Nuclear Power subsystem of a safety-related system other than an emer- l Corp., July 15,1988 [Capstick,1988]) which is based on gency diesel generator would not require testmg 'of the !

the Vermont Yankee Technical Specifications. diesel generators or any other equipment.

NUREG-1366 54

- o ot

= :s

• Sk O

wa

.c

-O at me i

c f4

, n Table it.1 Allernate testing requirements

~4 X

operasima W honey RHR Altermate Standtsy y',

e AmayoreMe emapememas Equt0 Case LPCI Desmet Nh ser%ce Serwke . coating gas e-seheyseman Ways) eastui sprey * ,C gamerators coelleg weser water eener ADS Rt'IC HPCI aresamsms [

2.

Sinedby inquad 7 t/D monat Core spray 7 1/D I I L?CI puunp 7 I IID* I I LPCI

, -- 7 I/D I/D 1/D I/D

- Diesel generasen 7 1/D I/D 1/D I/D Contaiesment mahag 30 1/D RHRSW

$ poemp 30 I/D RHR service weser 7 I/D I!D Service mener 15 1/D I/D Alteranse eamer 7 I/D I/D I/D ADS 7 I RCIC 7 IID HFCI 7 I. I I/D 1/D samedby saa nremessent 7 I/D UPS*

Nose:I-IW D-Daily a-Rh e oudy 5 b-See 17C1 ^ ,' . core spray, and diesel generator alternate testing requirements.

- C-

n

.m

^

.O ag

.' j

M a to TXA-94118 Page 6 of 9 10 Electnc Power ne NRC staff recommends that the requirements to test hours. During the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, the diesel is to o. .c the remaining diesel generator (s) when one diesel gen. with its 2 hour-rated load and for the last 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> it is to ~

erator is inoperable due to any cause other than operate at its contmuous rated load. ne Standard Tech-preplanned preventive mamtenance or testing be limited nical Specifications require that, within 5 minutes after to those situations where the cause for inoperabtlity has completing this 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test, the emergency buses must not been conclusively demonstrated to preclude the po- be deenergized and loads shed with a subsequent fast .

tential for a common mode failure. However, when such start and full load acceptance.

testmg is required it should be performed within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> of having determmed that the diesel generator is moper- Duke Power Co., by letter dated February 15,1988, on able. the Catawba Units I and 2 dockets (Tucker.1988) pro-posed to separate the 24-hour test from the 5-minute test.

De NPAR Program found that regulatory survedlance The NRC staff approved Duke's proposal in a letter to -

requirements are not the only contnbutor to EDG degra- Duke dated July 28,1988 (Jabbour,1988).

dation. NUREG/CR-4590, Volume 1, identified four categones of stressors that contnbuted to emergency die- ne reason for requesting this change is that separating sel generator agmg: vibration, infenor quality of compo- these two required tests gives plant operators added flex 2-nents, adverse environment, and human error, bdity and prevents critical path complications dunng the outages.

The NPAR Program did not spectfy the fraction of prob.

tems found with emergency diesel generators which are Duke stated that it has been necessary to shut down the -

due to testmg. A study done for EPRI(NP-4264. Vol. 2) diesel generator faster than recommended by the diesel '

looked specifim!!y at failures of emergency dieselgenera- generator shutdown procedure in order to perform the :

tors that result from surveillance testing. ne data for this hot restart test within 5 minutes of the 24-hour test run.

study constst of LERs from January 1979 through early Another problem with performing these tests in quick 1983, a period of just over 4 years. Note that this period succession is their potential for causing critim! path com-preceded the issuance of Generic !.etter 84-15 so that, plications and delays during an outage. Engineered safety ;

hopefully, the situation now would be somewhat better. features (ESF) actuation testing is performed at tHe-A total of $85 failures of 136 diesel generators were ginning of refueling outages. Block tagouts are d d.

found. Of these 585 failures,70 (12%) were determined untilcompletion of ESF testing. As a result of the teamg -

to be related to surveillance testing. He components sequence currently dictated by Technical Specifications, a that had the highest numbers of surveillance-test related minimum of 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> of cntical path time is spent each '

problems were: turbocharger, power assembly and bear- refueling outage runnmg the two dice! generators. By ings, starting system, cooling system, tube oil system, gov- revising the surveillance requirements as requested, the ernor and exciter, and regulator. However, no specific two 24-hour runs could be completed later in the outage fadures were widespread enough to be considered ge- or at some other convenient time.

nene. Generic failures with diesel generators have oc-curred in the past, but solutions to these problems are Duke proposed to substitute a diesel generator run at -

avadable and, in most cases, have been implemented. continuous-rated load for I hour or until the operating temperature had statn ^ ed, followed within 5 minutes by ;

Emergency diesel generator testing appears to be an area a diesel engine start. To ensure that operatmg tempera-that would benefit from a reliability-based testing pro- tures have stabilized, the NRC staff concludes that 2 gram (as discussed in Section 3.8 of this report). ne NRC hours is a more appropriate time limit.

staff is evaluating reliability-centered concepts for the resolution of GI B-56 that may further reduce unneces- ne hot restart test is performed to verify that the diesel saiy testing. NUREO/CR-5078 desenbes an approach to generator does not have, in any way, impaired perform-a reliability-based testing program for emergency diesel ance following operation at full load or equilibnum :

generators. As part of this reliabdity based approach, a temperature.

detailed root cause analysts procedure and a good pre-ventive maintenance program (also reliability based) Failure to restart when hot, or extended delay in restart-should be included. Detailed monitoring and trending are ing, is typically only experienced with small forced-air-unportant to assure good performance.

cooled diesel engines which, upon being tripped undergo j a temperature rise transient. The large diesel generaton Diesel generator surveillance requirements could also be are typimlly water cooled and do not experience an) 1 improved in another area. significant temperature rise transients dunng oper6on '

or after shutdown. In addition, diesel generatc 'e i The Standard Technical Specifications contain a require- normally maintained at hot standby conditions (hwed ;

ment to operate each emergency diesel generator for 24 coohng water and lubricating oil). I 1

NUREG-1366 56

Enclosure 2 to TXX-94118 Page 7 of 9 10 Electric Power The NRC staff, therefore recommends that other utilities tems not associated with an inoperable tram or sub-be permitted to change their Technical Specifications to system (other than an inoperable EDG).

separate the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test and the hot startup test if they propose doing so.

10.2 Battery Surveillance Requirements Findings (PWR, BWR)

• EDGs are very unportant to safety. Industry guidance for testing large lead storage batteries of the land used in nuclear power plants ts found in

• EDGs are tested too often because: Standard 450-1980 of the Institute of Electrical and Elec-tronics Engineers (IEEE). Regulatory Guide 1.129, (1) Technical Specifications at some plants requtre Revision 1 (February 1978) endorses an earlier version of testing tf other safety-related equipment ts this standard (IEEE 4W1975). The Standard Technical inoperable. Specifications follow this standard to some extent but are more conservative in some requtrements and less conser-(2) Technical Spectfications at some plants require vative in others. Table 10.2 compares IEEE 450-1980 not just one start to venfy operability but starts with the Westinghouse Standard Technical Spectfica-

"immediately," or within I hour, and every 8 tions, Version 4A.

hours thereafter.

Note that IEEE 4S1980 requires more visual inspec.

• Studies show that testing too frequently is counter- tions of the condition of the batteries (e.g., cleanliness, productive to safety m terms of equipment evidence of corrosion, cracks and leakage of electrolyte) availability, than the Westinghouse Standard Technical Specifica-tions (STS). On the other hand, the Westinghouse STS

• Raptd loading is a major cause of diesel generator are more conservative with respect to the frequency of degradation. measurements of battery charger output, pilot cell condi-tions, and total terminal battery voltage. (The West.

• There is no safety reason for performing a startup of inghouse Su require these every 7 days while IEEE a diesel wtthin 5 minutes of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test run as is 450-1980 requires these surveillances only monthly.)

required by Technical Specifications.

It is apparent from this compartson that the Westing-Recommendations house STS are most concerned with measurements of the operability of the batteries and not as concerned with

• When an EDO itself is inoperable (not including a mechanisms that degrade the batteries.

support system or independently testable compo-nent), the other EDG(s) should be tested only once Perhaps the most significant surveillance not included in (not every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />) and within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> unless the the Westinghouse STS is the surveillance for ambient absence of any potential common mode failure can room temperature. IEEE 450-1980 requires a monthly be demonstrated. surveillance. The Westinghouse STS do not. The West.

Uighouse STS do require a quarterly surveillance of elec.

• trolyte temperature in a representative number of cells, EDGs should be loaded in accordance with the ven.

dor recommmhtmas for all test purposes other but the requirement is that the temperature be greater than the refueling outage LOOP tests, than a minimum value, an operability requirement. There is no maxunum temperature specified.

• The hot start test fouowmg the 24-hour EDG test should be a simple EDO start test. If the hot start A limit on maxarnum ambient temperature would protect test is not performed within the required 5 mmutes the batteries from degradation mechanisms.

following the 24-hour EDG test, it should not be necessary to repeat the 24-hour EDO test. The only NUREG/CR-4457, which studied the aging of Class 1E requirement should be that the hot-start test is per- batteries for the NPAR Program, states that " thermal iormed within 5 minutes of operating the diesel gen- stresses, whether caused by internal sources. . .or by the crator at its continuous rating for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> or untt! room temperature, are probably the most detrimental operating temperatures have stabilized, with respect to accelerating the aging of batteries." As an example, the report cites a major battery manufacturer as

• Delete the requirement for alternative testing that stating that an increase in ambient temperature from requires testing of EDGs and other unrelated sys. 77'F to 95'F reduces the life of the battery by 50%

$7 NUREG-1366 to TXX-94118 Page 8 of 910 Elecirte Power Table 10 2 Comparison of requirements ofIEEE Standard 450-1980 with requirements of Westinghouse STS i 1

Requirement IEEE Westinghouse 450-1980 j

STS '

l 1.

General appearance and cleanliness of battery M -

and battery area 2.

Evidence of corrosion on terminals or connectors. M Q

3. Cracks in cells and leakage of electrolyte M -

4. Individual cell condition Y R

5. Tightness of bolted connections Y R

6. Integrity of battery rack Y R

7. Condition of venulation equiprnent M -

9. Electrolyte levels, each cell M Q

10. Ambient temperature M -

11. Voltage, specific gravity, each cell Q Q

12. Electrolyte temperature, representative cells Q Q

13. Total termmal battery voltage O Every 7 days

14. Pilot cell electrolyte level See item 9 Every 7 days

15. Float voltage M Every 7 days

16. Specific gravity M Every 7 days

17. Electrolyte temperature M -

Note: M = nmothly. Q = quanerty, Y = yearty, R = not to cacced 18 months.

The NRC staff therefore recommends a studyof the need EEE Standard 535-1986 requires that batteries that for a maximum (and minimum) allowable ambient tem- have been aged to their end-of life service be given a perature for batteries, pre seismic capacity test, a capacity test during a sirnu-There are other important phenornena discussed in lated seismic test, and a post seismic test. IEEE 535-1986 NUREG/CR-4457 thst are not covered by either the also requires seismic qualification of the battery rack.

Standard Technral Specifications or EEE 450-1980. herefore, batteries tested to IEEE 535-1986 should be These are the seasrme vulnerabdity of the batteries and acceptaHy qualified for seismic events, recogmzmg that excessive harmome fluctuations in the battery charger this is not a Techrucal Specifications issue and seismic voltage, called ac "rtpple." testing should not appear m the Technkal Specifications, According to NUREO/CR-4457, excessive harmonic The Standard Techrucal Specifications require several fluctuations in voltage frrAn the battery charger cause surveillances more often than called for by IEEE stresses at the battery plate sunilar to overcharging, accel

• 450-1980. These are the electrolyse ievel, fioat voitage, erste corrosion, and produce excessive internal tempera. specificgravity of the pdot cell (Category A items of Table '

tures. The NRC Office of Nuclear Regulatory Research 4.8.2), and the total terminal battery voltage. Several of should continue to study these to determine if this situ- the PWR ham identified these as burdensome sur.

ation is really a problern at nuclear power planta. vedlance intervals. However, as shown in Table 10.3 (taken from NUREO/CR-4457), some of the most com-mon causes of battery failure are associated with items The seismic event is the design-basis event for the me-charucal integrity of batteries. Seismic vulnenbility is covered by these surs,mm Note that the lea ^g caused by physical degradation of the structure of the cause of battery inoperability is low specific gravity 1 ficient charge and low electrolyte levels are also signus,-

battery. Dere are no good tests to detect this aging, but cant causes of battery failure.

NUREG-1366 58

. Enclosure'2 to TXX-94118 9 9 10 ElGctnc PowGr Table 10.3 Battery failure events reported in LERs Fallare cause No, 1.nw specific gravity 67 27 Personnel (operation, maintenance, testing) 52 21 Insufficient charge 27 11 Defective / weak cells 22 9

Low electrolyte solution level 14 6

Faulty connections 13 Defective procedures 11 4 Charger malfunction 9 4 Design, fabrication, construction 8 3 High electrolyte solution level 8

3 Unknown causes 5 2 Corrosion 4

2 Short circuit 4

2 Normal wear / natural end of life 3

1 Extreme environment 1

<1 Total 248 100 Source: NtJREG/CR-4457 Note also that testing (grouped together with operation Findings and maintenance) is the second largest contributor to battery failures. However, these 7. day surveillances e should not be significant contributors to testing failures. Operability surveillances of batteries required by Technical Specifications are performed more often -

than the industry standard recommends.

In addition, one utility representative told the NRC staff e

during a site visit that in addition to the Technical Specifi- There is no Technical Specifications requirement for monitoring or controlling battery room cations requirement, it was company policy to do these checks every 7 days. temperature.

e Seismic qualification isan important consideration The NRC staff therdore recomrnends that the battery for Class 1E batteriesand battery racks. All Class survettlance requirements remain as they are. 1E batteries and battery racks should be qualified to IEEE Standard 535-1986. This is not a Technical Specifications issue.

As noted earlier, many factors specified in IEEE e

450-1980 are important for degradation of batteries that Alternating current (ac) ripple from battery charg.

are not covered by Technical Specifications. This is prob- ers may be a degradation concern.

ably appropriate, if the purpose of the Technical Specifi. Recommendation ,

cations is limited to operability concerns. However, the H e

staff recommends that these factors be included in any The NRC should consider the above findings and i

preventive maintenance program, determine whether any additional action is war-ranted. l 59 NUR EG-1366 l

l

)

P 4

ENCLOSURE 3 TO TXX-94118 GENERIC LETTER 84-15, PROPOSED STAFF ACTIONS 10 IMPROVE AND MAINTAIN DIESEL GENERATOR RELIABILITY (PAGES 1, 2 AND ENCLOSURE 1) k t

f k

5 I

-l

. :. _ _ . _ _ --_:= _ --

-Enclosure 3 to TXX-94118 Page 2 of 6

' /ps% 9  %,

UNITED STATES

! - NUCLEAR REGULATORY COMMISSION WASHINGTON 0, C. 20555 g ,l-l

/ July 2, 1984 TO ALL LICENSEES OF OPERATING REACTORS, APPLICANTS FOR AN OPERATING LICENSE, AND-HOLDERS OF CONSTRUCTION PERMITS Gentlemen:

[

SUBJECT:

PROPOSED STAFF ACTIONS TO IMPROVE AND MAINTAIN

, DIESEL GENERATOR RELIABILITY (Generic Letter 84-15)

As part of the proposed technical evaluation of Unresolved Safety Issue (USI)

A-44, Station Blackout, the staff is considering new requirements that would reduce the risk of core damage from station blackout events. The reliability of diesel generators has been identified as being one of the main factors affecting the risk from station blackout. Thus, attaining and maintaining high reliability of diesel generators is a necessary input to the resolution of USI A-44 Plants licensed since 1978 have been required to meet the reliability goals of Regulatory Guide 1.108 for their diesel generators. However, the staff has determined that many operating plants do not have reliability goals in place for their diesel generators. Considering the critical role diesel generators play in mitigating various transients and postulated events following a loss of offsite power, the staff has determined that there is an important need to assure that the reliability of diesel generators at operating plants is maintained at an acceptable level. The staff has determined that the risk from station blackout is such that early actions to improve diesel generator reliability would have a significant safety benefit. Toward this objective, we have developed the following approach to assess and enhance, where necessary the reliability of diesel generators at all operating plants.

The items covered by this letter fall into the following three areas:

1. Reduction in Number of Cold Fast Start Surveillance Tests for Diesel Generators This item is directed towards reducing the number of cold fast start surveillance tests for diesel generators which the staff has determined results in premature diesel engine degradation. The details relating to this subject are provided in Enclosure 1. Licensees are requested to describe their current programs to avoid cold fast start surveillance testing or their intended actions to reduce cold fast start surveillance testing for diesel generators.

2. Diesel Generator Reliability Data This item requests licensees to furnish the current reliability of each diesel generator at their plant (s), based on surveillance test data.

Licensees are requested to provide the information requested in Enclosure 2.

-84070esa06 1 y,

l to TXX-94118 I Pagn 3 of 6 '

1

3. Diesel Generator Reliability Licensees are requested to describe their program, if any, for attaining  !

and maintaining a reliability goal for their diesel generators. An example of a performance Technical Specification to support a desired diesel generator reliability goal has been provided by the staff in Enclosure 3.

Licensees are requested to comment on, and compare their existing program or any proposed program with the example performance specification.

Accordingly, pursuant to 10 CFR 50.54(f), operating reactor licensees are requested to furnish, under oath or affirmation, no later than 90 days from the date of this letter, the information requested in Items 1 through 3 above.

Applicants for op 2 rating licenses and holders of construction permits are not required to respond.

Licensees nay request an extension of time for submittals of the required information. Such a reouest must set forth a proposed schedule and justification for the delay. Such a request shall be directed to the Director, Division of Licensing, NRR. Any such request must be submitted no later than 45 days from the date of this letter.

This request for information has been approved by the Office of Management and Budget under Clearance Number 3150-0011, which expires April 30, 1985.

Sincerely,

\ '

f arre sbn u , r Division of icensing

Enclosures:

1. Reduction in Number of Cold Fast Starts for Diesel Generators

2. Diesel Generator Reliability Data

3. Diesel Generator Reliability 1

' Enclosure 3 to TXX-94118 Pagq 4 of 6 Ef' CLOSURE 1 REDUCTION IN N'JMBER OF COLD FAST START SURVEILLANCE TESTS FOR DIESEL GEtlERATORS Fast Start Testing The staff has for sometime had under review and assessment methods of diesel generator testing. The staff has determined that many licensees use'a method

. of testing which does not take into consideration those manufacturer recommended preparatory actions such as prelubrication of all moving' parts and warmup procedures which are necessary to reduce engine wear, extend life and improve

- availability. The existing Standard Technical Specifications reouire fast starts from ambient conditions for all surveillance testing which in many engine designs and operating practices subject the diesel engine to undue wear and stress on engine parts. Concerns were expressed by ACRS regarding the imposition of severe mechanical stress and wear on the diesel engine due to frequent cold fast starts.

Nuclear Industry related groups (INP0 and American Nuclear Insurer) have also expressed concern based on operating experience that cold fast start testing results in incremental degradation of diesel engines and that, if proper procedures covering warmup prelubrication, loading / unloading etc., were taken, an improvement in reliability and availability would be gained. Similar views have been identified by the nuclear power industry and the regulatory authority in Sweden. The authority in Sweden has taken corrective actions to reduce the frequency of fast starts.

It is the staff's technical judgement that an overall improvement in diesel engine reliability and availability can be gained by performing diesel generator starts for surveillance testing using engine prelube and other manufacturer recommended procedures to reduce engine stress and wear. The staff has also determined that the demonstration of a fast start test capability for emergency diesel generators from ambient :onditions cannot be totally eliminated because the design basis for the plant, i.e., large LOCA _ coincident with loss of offsite power, requires such a capability.

In view of the above, the staff has concluded that the frequency of fast start tests from ambient conditions of diesel generators should be reduced. An example of an acceptable Technical Specification to accomplish this goal _is provided in the attachment to this enclosure. Licensees _are requested to describe their current programs to avoid cold fast starts or their intended action to reduce the number diesel of cold fast start surveillance tests from ambient conditions for generators.

Licensees are encouraged to submit changes to their Technical-

. Specification to accomplish a reduction in the number of such fast starts.

Other Testing Also, the staff is concerned regarding a number of additicral diesel generator' tests that are currently being required by Technical Specifications for some of-the earlier licensed operating plants. For example, when subsystems of the emergency core cooling system on some plants are declared inoperable, the diesel  ;

generators are required to be tested. The staff has concluded that excessive i testing results in degradation of diesel engines. In order to make those few l plants consistent with the majority of the plants, it is_the staff's position that the requirements for testing diesel generaters while emergency core cooling equipment is inoperable, be deleted from the Technical Specifications for such plants.

The affected licensees are encouraged to propose Technical Specifications to make such changes.

l

Encionure 3 to TXX-94118 Page 5 of 6 ATTACTGT TO DiCLOSURE 1 TYPICAL TECHNICAL SPECIFICATION SURVEILLANCE REQUIREMENTS 4.8.1.1.1 Each of the above required independent circuits between the offsite transmission system shall be: network and the onsite Class 1E distribution a.

Determined OPERABLE at least once per 7 days by serifying correct breaker alignment, indicated power availat.ility, and b.

Demonstrated OPERABLE at least once per 18 months during '

shutdown by transferring (manually and automatically) unit  !

power supply from the normal circuit to the alternate circuit.

4.8.1.1.2 Each diesel generator shall be demonstrated OPERABLE:

a.

In accordance with the frequency specified in Table 4.8-1 on a STAGGERED TEST BASIS by:

1.

Verifying the fuel level in the day and engine-mounted fuel tank, '

2. Verifying ~the fuel level b the fuel storage tank.

3.

Verifying the fue'l transfer pump starts and transfers fuel from the storage sy, stem t,o the day and engine-mounted tank, 4

Verifying the diesel starts from ambient condition and accelerates to at least (900) rpm in less than or equal to 10 seconds.* The generator voltage and frequency shall be  ;

(4160) s'(420) volts and (60) (1.2) Hz within (10)* seconds after the start signal. The diesel generator shall be started for this test by using. one of the following signals:

a) Manual b) Simulated loss of offsite power by itself.

• The diesel generator sta'rt (10 sec) from ambient conditions shall be performed at least once per 184 days in these surveillance tests. All I other engine starts for the purpose of this surveillance testing may be preceded by an engine prelube period and/or other warmup procedures '

recomended by the manufacturer so-that mechanical stress and wear on the diesel engine is minimized.

NOTE:

Bars in the margin show changes made to the Standard Technical Specifications.

1 i

l

'Encionure 3 to TXX-94118 ,

Page 6 of 6 l

. 1' SURVEILLANCE REQUIREMENTS (Continued) c) Simulated loss of offsite power in conjunction with an ESF actuation test signal. -

d) An ESF actuation test signal by itself.

5. Verifying the generator is synchronized, loaded to greater than or equal to (continuous rating) in less than or equal to

( ) seconds,* and operates with a load greater than or equal to (continuous rating) for at least 60 minutes,

6. Verifying the diesel generator is aligned to provide standby power to the associated emer'gency busses,

b. At least once per 31 days and after each operation of the diesel where the period of operation was greater than or equal to I hour by checking for and removing accumulated water from the day and engine-mounted fuel tanks,

c. At least once per 92 days and from new fuel oil prior to additional to the storage tanks by verifying that a sample obtained in accordance with ASTH-0270-1975 has a water and sediment content of less than or equal to ,05 volude percent and a kinematic viscosity 0 40'C of greater than or equaT to 1.9 but less than or equal to 4.1 when tested in accord'ance with ASTM-0975-77, and an impurity level of less than 2 mg. of insolubles per 100 ml. when tested in accordance with ASTM-D-2274-70'.

d. At least once per 18 months, during shutdown by:

1. Subjecting the diesel to an inspection in accordance with procedures prepared in conjunction with its manufacturer's recomendations for this class of standby service.

2. Verifying the generator capability to reject a load of greater than or equal to (largest single emergency load) kw while maintaining voltage at (4160) : (420) volts and frequency at (60) : (1.2) Hz (less than or equal to 75% of the difference between nominal speed and the overspeed- trip setpoint, or 15% above nominal whichever is less).

3. Verifying the generator capability to reject a load of *

(continuous rating) kw without tripping. The generator voltage shall not exceed (4784) volts during and following the load rejection.

'See footnote on previous page

l l

-i ENCLOSURE 4 TO TXX-94118' SAFETY EVALUATIONS, INSPECTION REQUIREMENTS FOR TDI, DIESEL GENERATORS (TAC NO. M85325)

MARCH 17, 1994 PAGES 1, 2, SER 1, 6, 7 9, 10 AND 11

?

I b

i

-4ee- a In- n m y m y y m y Y

Enclosuce 6 to TXX-94118 Page 2 of-9

,,,,Q

. RECE.yED umrEc STATES

~'. : VUCLEAR REGULATORY COMMISSION pg g j gggg "l .

ase.arcu a: aww , d TS ya jg ,

Mr. R. C. Day # " 0 Ouke Engineering i Services. Inc.

TDI Diesel Generators Owners' Group Cow I A '

Clearingnouse

c. o a I + '

230 Soutn ', nn Street  ; ni t

E 4 l

?. O. Box ;s i,.QCM W f TlddM I c.hariotte. North Carolina 28201-1004 j p;g j

SUBJECT:

SAFETY EVALUATION, INSPECTION REQUIREMENTS FOR TRANSAMERICA DELAV INC. DIESEL GENERATORS (TAC NO. M85325)

The Transamerica submitted Delaval proposals on (TDI) diesel generators Owners' Group (0wners' Group)

Novemoer 30,1992 (Reference 1 in the enclosed Safety Evaluation) and December 7, 1993 (Reference 2), recommending removal of licensing conditions imposed as part of a technical resolution to address concerns regarding the reliability of the TDI emergency diesel generators "

(EDGs) following the crankshaft failure at Shoreham in August 1983. The technical as resolution identified involved implementation of Phase I and Phase II programs in NUREG-l'  ! Reference 3). The Phase I program focused on the resolution of known ene mponent problems that had potential generic implications, while tb 11 program focused on the design review of a large set of important manufacturing standpoin, :omponents to ensure their adequacy from a dell as operational performance. At that time, the staff concluded that these components merited special emphasis in the area of load restrictions and/or maintenance and surveillance. The 16 major components which were identified included connecting rods, crankshafts, cylinder blocks, cylinder heads, piston skirts, and turbochargers. Engine load restrictions were addressed in the plant Technical Specifications, license conditions, engine operating procedures and operator training, as appropriate, for five of these components. The most critical periodic maintenance / surveillance actions for these components were incorporated as license conditions.

On the basis of substantial operational data and inspection results the Owners' Group provided information in Rsfuences 2 and 3 to demonstrate that the special concerns of NUREG-1216 are no longer warranted. The Owners' Group stated that the TOI EDGs should be treated on a par with other EDGs within the-nuclear industry and subjected to the same standard regulations, without the special requirements of NUREG-1216.

In addition, the Owners' Group stated that this action will improve availability of the engines for service, .

especially during outages, while maintaining current reliability levels.

The NRCa staff completed reviewand itsoperational of the consultants dataatandPacific Northwest inspection Laboratories'(PNL results contained in the Owners' addition, Group submittal reports relative to the individual components. In independent opinions were obtained from three leading diesel engine experts regarding these inspection requirements.

Enclocure 4 to TXX-94118

.Page-3 of 9 R. C. Dat 2 On the basis of its review, the staff has concluded that there is adequate justification for removing the present component-based licensing conditions.

The staff's evaluation of the Owners' Group's submittal reports is in the attacned safety evaluation (SE).

It is intended that the attached SE be referenced by affected licensees in proposals for changes to facility licenses to the extent specified and under the limitations delineated in the licensee submittals and the associated NRC evaluation. The evaluation defines the basis for the approval of the reports and is applicable to the eight Owners' Group licensees: Texas Utilities for Comanche Peak; Entergy Operations for Grand Gulf; Duke Power for Catawba; Carolina Power for Shearon Harris; Georgia Power for Vogtle; Cleveland Electric Illuminating for Perry; Grand Gulf Utilities for River Bend; and Tennessee Valley Authority for Bellefonte.

In accordance with procedures established in NUREG-0390, the TDI Owners' Group is requested to publish approved versions of the Owners Group reports as generic topical reports within three months of receipt of this staff approval.

The accepted version should incorporate this approval letter and the enclosed evaluation between the title page and the abstract. The approved version -

shall include an -A (designating approved) following the report identification symbol.

Tne staff does not intend to repeat its review of the approved matters described in the approved generic topical reports when the reports appear as references in license applications except to assure that the material presented is applicable to the specific p:lant involved. The staff's approval applies only to the matters described in the reports.

Should the staff's criteria or regulations change so that the staff's conclusions as to the acceptability of the reports are invalidated, the Owners' Group and/or the licensees referencing the reports will be expected to revise and resubmit their respective documentation, or submit justification for the continued effectiv6 applicability of the reports without revisions of their respective documentation.

Sincerely, 2m 6f. A i

f James A. Norberg, Chief Mechanical Engineering Branch Division of Engineering Office of Nuclear Reactor Rcgulation

Enclosure:

Safety Evaluation

Enclosure 4 to IXX-94116 Page 4 of 9 SAFE 7 En.u:*:CN By 'uE Cr W E ;r NUC S JEACTCR arge gilc,y C ? E S.iB l U T Y 30 2ELIA5il!!f IEVIEW CF E"EDGENCY DIE $El GENE 8ATCRS

• 2NUFACTG EO 3Y W NS3EDICA DELAVAL. INC.

:NT:CCUCTICN Our mg tme '.d:s, any utilttles ordered diesel generators from Transamerica

elaval. Inc. (TOI) 'or installation at nuclear plants in the United States.

The first of these engines to become operational in nuclear service were these it San Onofre Unit 1 in 1977. Hewever, nuclear plant operating experience

,ith TCI uiergency :tesel gar,erators (EDGs) remained very limited until preoperational test programs were started at Shoreham and Grand Gulf Unit ! :n tne early 1980s.

Concerns about the reliability of large-bore, medium-speed diesel generators manufactured by TOI for application at domesti: nuclear plants were first promoted by a cranKsnaf t f ailure at Shoreham in August 1983. However,'a broad pattern of deficiencies in critical engine components subsequently became evident at Shorenam and at other nuclear and non-nuclear f acilities employing TOI diesel generators. These deficiencies stemed from inadequactes in design, manufacture. and quality assurance / quality control by TOI. .

In response to these problems, 11 (now 8) U.S. nuclear utility owners' formed a TDI Diesel Generator S ners' Group to address operational and regulatory issues relative to diesel generator sets used for standby emergercy power. On March 2. 1984, the Sners' Group submitted a proposed program (**::

Cwners' Group Program Plan") to the NRC that was intended to provide an in-depth assessment of the adequacy of the respective utilities' TDI engines to perform their safety-related function through a combination of design revie=s.

quality revalidations, engine tests, and component in'spections.

The Owners' Group program addressed three major elements concerning the manuf acture, inspection, and operation of TDI diesel engines:

(1) Phase 1: Resolution of known generic engine component problems to serve as a basis for licensing plants during the period before completion cf Phase !! of the Owners' Group program.

(2) Phase 11: A Design Review / Quality Revalidation (OR/QR), of a large se:

of important engine components to ensure that their design and manufacture, including specifications, quality control and quality assurance, and operational surveillance and maintenance, are ade:uate.

(3) Expanded engine tests and inspections as needed to support Phase i are

I pregran s.

' Carolina P:wer and Light Co. (Shearon Harris). Cleveland Electric Illuminating Co. (perry), Duke Power Co. (Catawba), Georgia Power Co.

(Vogtle), Gulf States utilities (River Bend) Entergy Operations. Inc.  !

i (Grano Gulf Units 1 1 2), TVA (Sellef:nta). Texas Uti'aities (C ranche Peak).

I

Enclosure 4 to TXX-94118

-Page 5 of 9 4

6

roposes to use tnt s generic diesel 9anagement program in lieu of *.ne current maintenance / surveill anca recu t rement s .

n Se basis of the substantial operational experience of the 70! EDGs ace ulated s:nce '.985 and the inspection results of the EDG components, the C.ne s' Occuo ,as :roviced information in its submittal reports of November 30, ;992, and Cecemoer 7,1993 (References 2 and 3) to demonstrate that the special concerns of NUREG-1216 are no longer warranted. The Owners' Group nas recomended removing the license conditions related to EDG component inscact1ons involving teardcwns and surq111ance requirements.

The Owners' Group has analyzed the need for engine overhauls in accordance with the current OR/0R requirements. Their analysis and conclusions are cased on an understanding of the historical concerns for each component af fected by the overhaul and the results of extensive inspections performed by the licensees who make up the TOI Owners' Group. The information in its submittal reports includes comoonent description, component identification number er the OR/QR Appenoix II. # Preventive Maintenance (PM) Task Description," the manuf acturer's replacement / overhaul recomendations, the number of engine hours run between Inspections or cumulative engine hours, number of engine starts, inspection findings, and the percentage of all components in s'ervice covered by the inspections. The results of the inspections compiled by the Owners' Group in its submittti reports (References 2 and 3) indicate that most teardowns have shown little or no wear on internal engine components.

However, with continuing operation, it is possible that problems could occur with specific components which could require inspection or overhaul of affected components. The Owners' Group is proposing that such actions ce determined on a case-by-case basis, and that inspections or overhauls be performed so that engine reliability and availability are maximized. The Owners' Group contends that the primary purpose of EDG 10-year teardown inspections is to document the condition of the specific components, not to replace components, since most components being inspected show little or no wear. However, as a matter of good maintenance practice, these components 4re generally replaced after a teardown inspection, regardless of conditton.

These teardowns can result in reassembly errors or entry of foreign materials resulting in increased wear or decr6ased engine reliability.

The Owners' Group believes that an overhaul will be needed during the life of-these engines as they are currently operated. However, due to the limited number of run hours and the availability of periods to perform major teardnwns

- the licensees need the flexibility to determine when an overhaul is recutred and how an overhaul is conducted.

ihs Owners' Group contends that some of the early concerns with EDG comoonents aere caused by the deleterious effects of the fast starts and loading of Els in nuclear service. The Owners' Group notes that the life expectancy of most engine components in cx.merc111 service, which are not subject to' fist ittris.

is f ar greater than the estimated 1ife of EDG components in nuclear service based on early data.

All licensees have the authority to delete fast-start and loading recu1rtments ,

on the basis of Generic Letter (GL) 84-15, and are committed to doing so.  !

l I

Enclosure 4 to TXX-94118-Page 6 of 9 7

twever, some licensees have not taken this step for a numcer of reasons.

Dest, 9any engines have control systems whicn will not allow a slow start.

'e necessary char.ges in sucn control systems.are currently being implemented.

Iec:nd. 3:me of tne TOI licensees want to consolidate all changes for a art'c A ar tecnnical scecification (TS) to lessen the impact en the licensee ano :ne VC .orticad resulting from a TS change recuest. The staff is

arrently preparing a GL addressing the requirements for accelerated testing

f emergency diesels. Most licensees are waiting for this GL to be issued tefore requesting a change to their TSs which would include a request for the

eletion of the fast starts. Once the slow start option is imp'.ecentad and accelerated testing is eliminated, engines at nuclear plants will be operated similarly to those in commercial service, and the expected life of components in engines at nuclear plants should compare favorably with commercial engine components. The data from engines in nuclear service which have implemented the slow-start option supports this contention. Since the manufacturer's recommendations for commercial operation of TOI/EDG components prior to

vernaul indicate that there are substantial safety margins available, appropriate changes can be made in M/S requirements based on realistic estimates of component life expectancy, and flexibility can be achieved in the frequency of performing teardown inspect'1ons. -

The Owners' Group, in its submittal reports, has also discussed the need for flexibility in schcauiing teardown inspections from the standpoint of shutdown risk management (SRM). According to the Owners' Group, the "available windows" of outage time of sufficient length to allow engine teardowns and/or overhauls are being shortened because of SRM requirements. The "available window" during which a diesel can be removed from service for maintenance depends in a numcer of factors, including plant design, availability of alternate power sources, fuel handling schemes, and other operational, maintenance, or inspectfen requirements. These factors cause the "available window" to vary from outage to outage. Typically, the "available window' is Oetween 10 and 21 days; however, SRM programs have compressed this

• window" by as mucn as 20%. As a result of this shortening of "available windows,' all plants need naximum flexibility in scheduling EOG maintenanca activities (i .e., schedule major diesel work during times when longer " windows' are available without impacting overall outage length). Time-airected tearcowns/ overhauls do net allow this flexibility. The Owners' Group is procesing a generic diesel management program which combines predictive aintenance, surveillance, and inspection. The Owners' Group contends that

. with this program, considerable flexibility can be achieved in the frequency of performing teardowns and/or overhauls without sacrificing engine reliability.

' /cical components that are inspected or replaced or both during an engine avernaul are turbochargers, main bearing caps / studs, cylinder clocks, ccnnecting rods / bear 1rgs/ bushings, cylinder heads, push rods, lower cylinder liner seals, base assemblies, crank shaf ts, cylinder liners, pistenserings, 6el injection tubing, and rocker arm capscrews/ drive studs, problems witn these components resulting from the intrusive inspections could certainly limit or preclude the engine's acceptable power output. Disas:embly of these cceconents can result in the acc1 dental introduction of dirt ar.d other ' Ore':-

taterials that may harm the engine. In addition, these components are

P ' Enclosure 4 to ?XX-94118 Paje 7 o f 9 9

• 3ecluse soecific surveillances/inscections were imcosed by regulati:n ::

ensure that acceptable engine conditions were being maintained, the nspection results snould not identify unacceptacle findings.

. he C ners' Group should have an alternative diesel management progen ita 2 1 9 ents that are judged by the regulate y staff to be easenacif v e:ualij effective compared to current license requirements in maintaining diesel reliability.

• The underlying source or technical basis for the proposed regulatory cnange snould be justified by authorities and expertise equal to that anicn determined the current regulatory requirements.

As discussed in the following paragraphs, all five criteria have been satisfied. The current TOI engine reliability was found to be equal to or better than the industry average. In the period between January 1990 and December 1992, the median reliability of TOI diesels was found to be 0.9906.

This is about 1% better than the nuclear industry average, and well above NRC's highest goal of 0.975.

Specific surveillances/ inspections were imposed by NRC regulations to erlsure that acceptable TOI engine conditions were being maintained. A review of the operational database and the inspection results for the key components, as discussed in Appendix A, show no unacceptable findings. In fact, nost -

inspections did not uncover any signs of wear or degradation that need to be 1ddressed.

NRC-sponsored research (Reference 8) has indicated the potentially negative consequences of intrusive inspections on components and engine reliability as a result of current practices. In a study of failures related to aging, a failure curve, sometimes called the " bathtub

• curve, correlates the change in failure rate with age. The beginning segment of the curve represents a " wear-in portion, with a higher failure rate associated with many pieces of new d

equipment. Once the machinery is broken in, the failure rate is at its lowest and remains constant for a period of time. As the machinery wears and reatnes the end of its lifetime, the failure rate increases. The challenge is to determine the time scale for these regions for each place of equi; ment. On the basis of these studies, it is generally believed that the diesel engine's reliability is considerably lower during the " wear-in" period, and some engines may be on the lower end of the acceptable range of reliacility, curr9 the wear-in" period of operation.

Some of the early concerns with EDG components' were due to the deleterious affects of fast start and loading of EDGs in nuclear service. Comoonent 1 'e

=xcectancy in comercial T01 engines which are not subject to fast starts :s far greater than life expectancy for TOI engine components in nuclear arv :e.

although the f ast-start requirements have been relaxed on the basis of GL U-15 not all licensees have implemented the changes in the EDG control sista-to permit slow starts. All members of the Owners' Group are committed to implementing those changes in the near future. The staff is alsJ 3dcr! ssp 9 the issues related to accelerated testing in a generic letter to te issued shortly. Cace the sicw start option has been implomnted and acce!erated L

Page 7 of 9-e s

]

9 3ecause sDecific surveillances/1nscections were imposed by regulation.t:

ensure that acceptable engine conditions were being maintained, the inspection results should not identify unacceptacle findings.

'he Cwners' Group should have an alternative diesel management Drogram

.ith al cents that are judged by the regulatory staff to be -easonaci f inc e:ually effective compared to current license requirements in maintaining diesel reliability.

The underlying source or technical basis for the proposed regulatory cnange snould be justified by authorities and expertise ecual to that whicn determined the current regulatory requirements.

As discussed in the following paragraphs, all five criteria have been satisfied. The current TO! engine reliability was found to be equal to or tetter than the industry average. In the period between January 1990 and Cecember 1992, the median reliability of T0! diesels was found to be 0.9906.

This is about 1% better than the nuclear industry average, and well acove NRC's highest goal of 0.975.

Specific surveillances/ inspections were imposed by NRC regulations to ensure that acceptable TOI engine conditions were being maintained. A review o'f the operational database and the inspection results for the key components, as discussed in Appendix A, show no unacceptable findings. In fact, most -

inspections did not uncover any signs of wear or degradation that need to be addressed.

NRC-sponsored research (Reference 8) has indicated the potentially negative consequences of intrusive inspections on components and engine reliability as a result of current practices. In a study of failures related to aging, a failure curve, sometimes called the " bathtub' curve, correlates the change in failure rate with age. The beginning segment of the curve represents a

• wear-in* portion, with a higher failure rate associated with many pieces of new equipment. Once the machinery is broken in, the failure rate is at its lowest and remains constant for a period of time. As the machinery wears and reatnes the end of its lifctime, the failure rate increases. Tiis challenge is to determine the time scale for these regions for each piece of equi; ment. On the basis of these studies, it is generally celleved that the diesel engine's reliability is considerably lower during the " wear-in" period, and some engines may be on the lower end of the acceptable range of reliacility, dur g l the

• wear-in' period of operation.

Some of the early concerns with EOG componenti were due to the deleterious ef fects of fast start and loading of EDGs in nuclear service. Comoonent 1 'e l svoectar.cy in commercial TOI engines which are not subject to fast starts :s i far greater than life expectancy for TDI engine components in nuclear scrc:e. '

Althougn the fast-start requirements have been relaxed on the basis of GL U- ,

15. not all licensees have implemented the changes in the EDG c= trol .jsti- i to permit slow starts. All members of the Owners' Group are committed to implementing these changes in the near future. The staff is also adcr! ssp; the issues related to accelerated testing in a generic letter to te issued shortij. C.xe the s';w start option has been implo ented and accelerated I

Enclosure 4 to TXX-94118 Page 8 of 9 D

m t'rg has Oeen eliminated, nuclear service engine Operation will more

' :sely match that of engines in c: v ertial service and the expected c:mconent

' 'fe for T0! engines in nuclear service should ccmoare f avorably with

.: rercial engine c:mponent life. The data from engines in nuclear service

.nicn ra<e mole ented the slow-start o,) tion succorts this :ntent':t 4 eview of P.e manuf acturer's recomendat1ons for ecmercial operat1:n of DI/EDG remponents before overnaul indicates that there are substantial safety margins available for most components in nuclear service. The staff concurs with the Owners' Group recomendation that by ccmbining predictive maintenance, surveillance, and inspections, as in the proxsed gerer'c Pasel 9anagement program, considerable flexibility can be achieved in the frecuency of performing engine teardowns and/or overnauls .<ithout sacrificing engine reliability.

The Owners' Group contends that the *available windcws" of outage time of sufficient length to allow engine teardowns and/or overhauls are being snortened due to SRM reouirements. As a result of this snortening of available windows, all plants neea maximum flexibility in scheduling EDG maintenance activities. The adoption of a predictive maintenance program for EDGs as proposed, in lieu of the current time-directed teardown /overnaul requirements would give the licensee this flexibility without jeopardizing engine reliability.

The Owners' Group has requested the removal of inspection requirements from the license conditions. The Owners' Group proposes to continue appropriate inspections; however, scope, ins 9ection schedules, and especially the amount of intrusive inspections involving disassembly would be changed to maximize EDG availability and reliability. Inspections would be planned to respond to monitoring and trending results and where other maintenance activities make the component accessible, such as in response to failures of nearby components

• or where monitoring is indicating an end of component life conditions, The Cwners' Group will continue appropriate inspections, especially those not involving engine disassembly. Inspections will be defined and included as part of a well-managed engine program currently under preparation. Elments of terrect engine management have been reported previously to the NRC and industry (References 8 and 9). Key features of an EDG management pecgra.n.

acceptable to the staff (see Appendix C of this safety evaluation) nave teen discussed and provided to the Owners' Group. The Owners' Group agrees that each member would adopt the group's proposed generic management program, resolution, or mitigating actions, and that all actions are intended to te 1cceptable to the manufacturer.

'inally, the underlying source or technical basis for the proposed regulatory

? 4993 1s equal in excertise to that which was re900nsible for reccrending the current regulatory requirements. The TOI Owners' Group, with support tr:,

'he nnuf acturer as instrumental in preparing the technical basis for the original regulatory conditions n NUREG-1216. j V. CVERALL CONCLUSICNS The staff, with assistance from its consultants and recogni43d d esal N '! l experts, concluded that the regulatory requirements on TOI engines may :e l l

L I

1 Enclosure 4 to TXX-94118 Pag,e 9 of 9 11 i mc:,5 dered at thts time. This conclusion is based on a review of 'he nrqat reliioliity cata of t e TDI engines, the Owners' Group inspect:ns :f I e last several years, ano the opinion of experts .ho have experten:e in t9e

es';9 trd 3:erits:n of large diesel engines. The staff telieves that Me 7::

Nrers' ho, ' : se an < et',cr Owners group, must address *ia un'aue ,

at tenance needs for its s:ecific engine to keep the reliability f actor acce:t:cle. with a current medt an reliability of 0.9906, tne IDI Cwners' 3rcuo and its individual cwners, seem to fully understand the maintenance needs of this engine, the staf f further believes that there is suf ficie.a*

'nformation in the Caners' 3roup submittal reports to ccnclude :'at CI ;1 e

peration at authorized loads is acceptable under normal NRC regulatory overstght procedures for EDGs, The staff and its consultants, in their review of the TDI submittal reports and the operational datab.se, did not uncover any aew concerns or issues. [ndividual reports from recognized experts endorse many of the TDI engine management practices, inspections, or precautions. The

? ners' Group intends to incorporate most of the inspections and precautions

' rem the current M/S requirements in its generic diesel management program ir.o tocropriately supplement tnese inspections with alternate condition monitoreg

rocecures. All members of the Owners' Group are cemitted to implement this 31esel management program.

The key features of a maintenance program which the stsff finds acceptable are delineated in Appendix C of this safety evaluation. The staff has reviewed the preliminary version of the diesel management program, which the Owners' Group is proposing in lieu of the current M/S requirements. The stiff finds the principal elements of this program acceptable. The proposed maintenance oregram is in conformance with the requirements in Regulatory Guide 1.150,

' Monitoring the Effectiveness of Maintenance at Nuclear Power Plants,' dated June 1993, which endorses NUMARC 93-01 dated May 1993, "!ndustry Guide for Monitoring the Effectiveness of Maintenance at .iuclear Power Plants.*

Acc:rdingly, the staff has concluded that the license conditions related to the periodic M/S program (see Appendix 0 of this safety evaluation) for certain components (see Appendix E of this safety evaluation) which were

' ened on the licensees based on the recomendations in NUREG-1216, ce retoved at this time. Therefore, the detailed steps of the preventive M/S programs will not be subject to NRC staff review and approval. However, the staff believes that future revisions of the M/S program would be subject u tne provisions of 10 CFR 50.59 (Code of Federal Reoulations) in vtew of ite

. moortance of the M/S program in ensuring the opera 0111ty and reliability ;f

• ne engines. The staff will recuire that the owners of each plant c:mmit u me current M/S program in the interim period preceding the implementatt n :'

• ne generic diesel management program currently under development in ass:ct at'on and agreement ,ith the mant facturer. The transitic9 fecm Se

arrent M/S orogram to the generic diesel management program could be 1:camoli: ed ;nder the promions of 10 CFR 50.59. The TS recuirements of suojecting the diesel to an inspection in accordance with procedures precare:

'n conjunction with its 311ruf acturer's recemmendations for the class of stanoby service would continue to remain in effect, similar to the IS r*Cutrements on other EDG manufacturers.