text

LER-1994-017, on 940902,EDG 1-1 Degraded Load Carrying Capability Identified.Caused by Inadequate Design Basis, Failure to Effectively Monitor for Engine Performance Degradation & Failure to Control Mods
Event date:
Report date:
Reporting criterion:	10 CFR 50.73(a)(2)(ii)(B), Unanalyzed Condition
	2551994017R00 - NRC Website
	v • d • e

:on 940902,EDG 1-1 Degraded Load Carrying Capability Identified.Caused by Inadequate Design Basis, Failure to Effectively Monitor for Engine Performance Degradation & Failure to Control Mods
	ML18064A484
Person / Time
Site:	Palisades
Issue date:	11/22/1994
From:	Gire P; CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	;
Shared Package
ML18064A483	List: 05000255/LER-1994-017, :on 940902,EDG 1-1 Degraded Load Carrying Capability Identified.Caused by Inadequate Design Basis, Failure to Effectively Monitor for Engine Performance Degradation & Failure to Control Mods;
References
	LER-94-017, LER-94-17, NUDOCS 9412020137
	Download: ML18064A484 (48)
	v • d • e

". *-

1.

NRC Form 388 U.S. NUCLEAR REGULATORY COMMISSION (9*931 APPROVED OMB NO. 3160-0104 EXPIRES: B/31/86 LICENSEE EVENT REPORT !LERI FACILITY NAME l1l DOCKET NUMBER 121 PAGE 131 PALISADES PLANT olslololol2lsls 1 I OF 17 Tine 141 Licensee Event Report 94-017 Emergency Diesel Generator, EOG 1-1, Degraded Load Carrying Capability - Supplemental Report EVENT DATE 161 LER NUMBER 181 REPORT DATE 181 OTHER FACILITIES INVOLVED (BJ SEQUENTIAL REVISION FACILITY NAMES MONTH DAY YEAR YEAR NUMBER NUMBER MONTH DAY YEAR N/A o I s I o I o I o I I ol9 012 914

--, 0 I 1 I 1 -

0 I 1 ii 1 I

914 o I s I o 1 o 1 o I I 9 4 N/A THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10 CFR I: ICIH>ck °"" °' "'°"' of rM fol/owing} 1111 OPERATING N

MODE (91 20.402(b) 20.406(cl 60.73(a)(2Jlivl 73.7 llbl i3.711cl POWER 20.4061all1llil 60.38(cll1J 60.731all21M LEVEL I

I (10) 1 0

0 20.406(a)(1 )(iii 60.38(c)(2J 60.731aJl2J(Yiil OTHER (Specify in Abatroct 20.4061*111 Jliiil 60.73(a)(2Jlil

60. 7 31oll2Jlviii11Al below and. in Text,

- 60.73(oJl2Jlviiii1Bl 20.406(a)(1 )(ivl x

60.73(a)(2Jliil -

20.406(0)(1 J(v) 60.731*11211iii) 60.731aH21bd LICENSEE CONTACT FOR THIS LER 1121 NAME TELEPHONE NUMBER*

Paul J Gire, Staff Licensing Engineer 6ARIEA1CI°~ I 7 I 6 I 4 I I 8 I 9 I 1 I 3 COMPLETE ONE LINE FOR EACH COMPONENT FAILURE DESCRIBED IN THIS REPORT 1131 MANUFAC*

REPORTABLE MANUFAC*

REPORTABLE

CAUSE

SYSTEM COMPONENT TUR ER TO NPRDS

CAUSE

SYSTEM COMPONENT TURER *.

TO NPRDS I

I I I I

I I

I I I I I I I I

I I I I I I l*********i*l*

SUPPLEMENTAL REPORT EXPECTED (141 MONTH DAY YEAR rx, NO EXPECTED I

YES Uf yo.,,,,,,,.,,,.,. EXPECTED SUBMISSION.DATE)

SUBMISSION I

I I

DATE (161 ABSTRACT UJmlt ro 1400 _.:H, i.e., llfJPIOJC/mate/y f'fftee11 aing/e-_,,e typewritten lines) 1161 On September 2, 1994, while the plant was operating at 100% power, and during maintenance testing, it was discovered that the Emergency Diesel Generator 1-1 (EDG 1-1) maximum output was below the maximum required design basis load. A misadjusted engine governor output linkage and engine performance degradation limited the EDG 1-1 output. The engine problems were corrected and an electrical load test was completed on September 4, 1994 to verify design basis maximum load carrying capability. A similar load test and minor linkage adjustments were completed for the alternate* EDG.1-2 on September 7, 1994. The as-found load carrying capability of the EOG 1-2 was found slightly below the maximum analyzed design basis load.

The causes for this event were inadequate design basis testing, failure to effectively monitor for engine performance degradation, and failure to adequately control modifications made to engine governors in the past. Corrective actions for this event include periodic peak load testing for EDGs, and enhanced EOG performance monitoring, to include engine governor performance.

an engineering evaluation has determined that the safety consequences of the reduced load carrying capability are minor. The engine would have remained running during the peak load period, and the safety functions of the associated electrical loads would have been fulfilled.

9412020137 941122 PDR ADOCK 0500025~

S PDR

Also, I

NRC 'Form 388A (8-83)

U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMB NO. 3160-0104 EXPIRES: 8/31 /86 FACILITY NAME (1)

Palisades Plant

Event Description

LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMBER 121 YEAR LER NUMBER 131 SEQUENTIAL NUMBER REVISION

.NUMBER 0

5 0

0 0

2 5 59 4 -o 1 7 -o

. PAGE (41 2

OF 7

On September 2, 1994, while the plant was operating at 100% power, and during maintenance testing, it was determined that the load carrying capability of Emergency Diesel Generator 1-1 (EOG 1-1 ),[EKJ; was degraded. A misadjusted engine governor output linkage ~nd minor engine performance degradation combined to limit the EOG 1-1 output to below the maximum analyzed load, which is elevated for the first half hour period after a large break loss of coolant accident, LBLOCA. An engine tune-up and fuel linkage adjustments were completed and followed by a design basis load test on September 4, 1994. This test and two subsequent monthly peak load tests have verified design basis maximum load carrying capability. Also, an engineering evaluation has determined that the reduced load carrying capability would not have adversely affected core cooling or containment cooling following a postulated LBLOCA.

The engine and its safety loads would have remained running, but at reduced frequencies, during the peak load period. The safety functions of the associated electrical loads would still have been fulfilled at the reduced frequency.

A similar load test and minor linkage adjustments were completed for the alternate EOG 1-2 on September 7, 1994. The as-found load carrying capability of the EOG 1.;2 was equal to the maximum analyzed design basis load. However, when instrument inaccuracies are factored into the comparison, the EOG 1-2 performance was 12 KW below the analyzed load demand. In the as-found condition, it is evident that EOG 1-2 was fully capable of performing its design function.

The diesel engines were restored to operable status within the Technical Specification allowed equipment outage period. The reactor remained at 100% power during the testing and repairs.

Due to the fact that EOG 1-1 had been incapable of performing its full design function for a significant period, this event was reportable in accordance with 10CFR50.73(a)(2)(ii)(B) as a condition outside the design basis. There have been times recently when EOG 1-2 had been taken out of service for maintenance and testing. Thus, during these times both emergency diesel generators* were concurrently inoperable.

Cause Of The*Event The causes of this event were; inadequate design basis testing, failure to effectively monitor for engine degradation, and failure to adequately control modifications made to the engine governors in the past. Beginning in 1986, the analyzed peak accident loads surpassed the continuous load rating of 2500 KW for both EDGs. During this period, implications of the peak loading values were evaluated. Load testing above the continuous rating of the engine was considered potentially damaging to the engines, and pre-operational testing in 1971 had verified load capabilities above 2700 KW, with additional fuel rack travel still available. Also, monthly testing at the continuous rated load of 2500 KW appeared to provide appropriate performance monitoring. This evaluation was not totally correct, and was not appropriately conservative to provide the ultimate assurance that the EDGs were capable of performing their design function. A current evaluation of peak

NRC Form 388A

~'

111-83)

U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMB NO. 3160-0104 EXPIRES: 8/31 /86 FACILITY NAME 111 Palisades Plant LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMBER 121 YEAR LER NUMBER 13l SEQUENTIAL NUMBER REVISION NUMBER o s o o o 2 s s s 4

o 1

1 o*

PAGE 14) 3 OF 7

load testing has determined that infrequent diesel operation at elevated loads is acceptable as well as necessary to verify EOG capabilities.

Analysis Of The Event

Accident Load Analysis. History Palisades has two independent and redundant emergency EDGs with continuous load ratings of 2500 KW and two hour maximum load ratings of 2750 KW. (Refer to Attachment 1 which Sl,Jmmarizes EOG loading information.) Technical Specification Surveillance Test requirements have always required monthly starting and load testing at 2400 KW plus or minus 100 KW.

Beginning in 1986, as a* result of plant modifications, the maximum analyzed accident loads required to be supplied by the EDGs after a design basis accident exceeded their continuous rating of 2500 KW. The accident loads are elevated above 2500 KW for the. first half hour period following a LBLOCA, and. then decrease below 2500 KW for the remainder of the event.

However, the monthly load test requirements for the EDGs remained at 2400 plus or minus 100 KW. (Refer to Attachment 2 for the historical depiction of maximum analyzed EOG loads.) At that time, peak load testing of the EDGs was considered potentially damaging to the engines, and also Plant Technical Specifications, Section 4. 7.1.d., restricted loading of the EDGs to below 2500 KW. The Plant Technical Specifications, Section 4;7.1.d., was revised in 1987 to clarify the diesel load restriction to 750 amps at 2400 volts.

Discovery Of The Problem The first indication of a EOG load capability problem occurred during routine monthly testing of EOG 1-1 on July 19, 1994. The maximum output of the engine was unexpectedly limited to 2340 KW, (acceptable range is 2300 to 2500 KW). Operations personnel attempted *to load the engine to the nominal test load of 2400 KW. However, the electrical governor Lipper limit prevented further loading. Operations personnel recognized the implications of the reduced load carrying capability and a prompt operability evaluation was completed at this time. EOG 1-1 was determined to be operable. The operability call was based on the fact that the design basis

. function of the engine is to provide emergency power to safety related components in the "Unit"

. mode. Monthly engine testing is only performed in the "Parallel" mode, which is with the engine paralleled to the electrical grid. The load carrying capability of the engine in the "Unit" mode is not affected by the electrical governor upper limit, which provides a control function solely in the "Parallel" mode. The engine performance data was also reviewed at this time and all parameters were normal. System Engineers and Operations personnel were convinced that the condition only affected the load carrying capability during testing, and that the upper limit simply needed adjustment. This would allow future monthly testing to be performed at the higher nominal test value of 2400 KW. A Condition Report was initiated to document the operability call and provide for further control and review of the electrical governor limit adjustment.

NRC Form 388A (9-B31 U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMB NO. 3160-0104 EXPIRES: 8/31 /86 FACILITY NAME 111 Palisades Plant LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMBER 121 YEAR LER NUMBER (31 SEQUENTIAL NUMBER REVISION NUMBER 0500025594-017 -o PAGE 141 4

OF On July 29, 1994, EOG 1-1 was removed from service to complete the adjustments to the electrical governor limits and to gather additional testing data. The upper limit could not be.

completely adjusted due to unexpected interference from the speed control setting of the mechanical governor. The partial adjustment did allow the EOG 1-1 to be loaded to an improved 7

. value of 2400 KW. Also, several other findings from the ongoing evaluation continued to support the initial co_nclusion that the reduced loading capabilities were associated solely with the "Parallel" mode *of operation. Specifically, the speed droop setting was incorrect for the EOG 1-1, which limited its output in the "Parallel" mode. Also, an additional 3/16" of engine fuel rack travel was measured at the 2400 KW loading and would supposedly be available in "Unit" mode for EOG 1-1 to provide the additional accident loading required. Further planning was required to make the final adjustments to all of the engine governor settings.

The troubleshooting on EOG 1-1 resulted in the discovery that fuel rack linkage configuration was improperly li_miting full rack travel and, thus, there was available engine output that could be gained by correcting this problem. Initial adjustments of the linkages were completed to allow the

engine to supply up to 2500 KW, and further adjustments would be needed during a planned design basis test that was under development. Also, it became apparent that the engine performance had degraded because it was not providing the expected power output at the measured fuel rack travel, with respect to the original pre-operational test in 1971. Inspections, adjustments, and replacement of suspected components provided minor improvements in engine performance, but no specific component degradation was found to explain the lower engine output.

On September 3, 1994, the initial design basis load test was performed for EOG 1-1 and the maximum output was measured at 2685 KW. Based on the results of the test and observation of the fuel rack positioning, further rack adjustments and testing were completed until full rack travel resulted in an output of 2714 KW. This value includes an instrument inaccuracy penalty of 17 KW. The fuel rack linkage adjustments appear to have resulted in the majority of the total performance improvement (276 KW) that was gained during the troubleshooting and repairs.

Based upon review of the final test results, the EOG 1-1 was declared operable on September 5, 1994. Additional peak load tests for EOG 1-1 were completed on 10/6/94 and 11 /2/94 in conjunction with the normal monthly testing. The output of EOG 1-1 was observed to be 2722

NRC F0tm 388A 19*B31 U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMB NO. 3160-0104 EXPIR.ES! B/31 /B6 FACILITY NAME 11 I Palisades Plant LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMBER 121 YEAR LER NUMBER 131 SEQUENTIAL NUMBER.

REVISION NUMBER 0

5 0

0 0

2 5

5 9

4* -* 0 1

7 0

PAGE 141 5

OF KW and 2735 KW I respectively I using precision instrumentation and accounting for instrument.

. inaccuracies. Thus, all peak load testing results, afterthe adjustment of the governor linkage, exceed the peak load of 2688 KW that could exist during the first half hour period after a LBLOCA.

7 On September 7, 1994, a design basis load test was performed for the alternate. EOG 1-2 to verify its ability to supply maximum analyzed loads. The as-fol.Ind load carrying capability of the EOG 1-2 was 2651 KW, which includes a 15 KW instrument inaccuracy penalty. This as-found performance was 1 2 KW below the maximum analyzed design basis load of 2663 KW. The root cause for the EOG 1-2 reduced load carrying capability was an improperly positioned fuel rack travel hard-stop. The hard-stop was adjusted and a second peak load test was performed. An improved output of 2697 KW was recorded during the test, which includes an instrument inaccuracy penalty. During the second EOG 1-2 peak load test, the electrical grid was experiencing slight fluctuations in voltage and frequency, and thus the load output of the EOG 1-2 was* unsteady. The engine output was observed at times to be over 2750 KW. Thus, the actual maximum output for the EOG 1-2 is higher than the value recorded in this tesJ. The EOG 1-2 was declared operable on September 8, 1994, based on its ability to provide the maximum design basis load of 2663 KW. Additional peak load tests for EOG 1-2 were completed on 10/10/94 and 10/27/94 in conjunction with ttie normal monthly testing. The output of EOG 1-2 was observed to be 2794 KW and 2736 KW, respectively, using precision instrumentation and accounting for instrument inaccuracies. *Also, *for.EOG 1-2 there was additional fuel rack travel available at the loads recorded for these tests.

The present output of EOG 1-2, based on a comparison of fuel rack positions,* is approximately equivalent to the output recorded in the original pre-operational testing in 1971. Thus, engine degradation does not appear to exist on EOG 1-2, but further evaluations and testing* are planned.

Several maintenance and modification activities from the 1970s and 1980s are primarily*

responsible for the reduction in peak load carrying capability below the initial 2800 KW for both EDGs. The mechanical governor was replaced on EOG 1-1 in 1979 and again in 1982. The

. electrical governor was replaced on EOG 1-1 in 1989. Similarily, the mechanical governor was

replaced on the EOG 1-2 in the early 1980s. It is suspected that, during the governor modification work, output linkage misadjustments occurred that limited the engine output to a value below the original capability of 2800 KW, but still above the 2500 KW rating. The impact

.. upon the total load carrying capability of the engines remained undetected at that time due to the existing operability test criteria of 2400 KW, plus or minus 100 KW.

The EOG 1-1 engine performance degradation issue remains under investigation at this time. The present maximum output of EOG 1""1 is approximately.100 KW below the original pre-operational testing results from 1971. The comparison is somewhat difficult to fully evaluate due to the inability to determine the potential inaccuracy of the instruments used in the pre-operational test.

NRC F0rm 388A

\\.:

19-831 FACILITY NAME 111 Palisades Plant

Safety Significance

LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMBER 121 YEAR LEA NUMBER 131 SEQUENTIAL NUMBER U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMS NO. 3160-0104 EXPIRES: 8/31186 REVISION NUMBER PAGE 141 0

5 0

0 0

2 5 59 4 -o 1 7 -o 6

OF 7

An engineering evaluation of the impact of EOG 1-1 reduced load carrying capabiiity has been

. completed. The evaluation of the EOG 1-1 degraded performance has determined that the EOG would have remained running at a reduced frequency. The evaluation has determined that the individual safety related loads on the EOG supplied safety buses would also remain running and provide their design function at the anticipated reduced flow rates. This result is based on the fact that slightly reduced safety related pump and fan flow rates for the first half hour following a LBLOCA, would only lead to a negligible delay in recovery from the event. During a LBLOCA, peak fuel clad temperatures and peak containment pressures occur within the first minute of the event. The peak values that these safety limit parameters reach are not impacted significantly by the pump or fan flow rates assumed in the analysis. Thus, the slightly reduced flows caused by the 30 minute operation at reduced frequency have negligible impact. Attachments 4 and 5 provide the evaluation results which were also independently reviewed by Sargent and Lundy technical experts. This review included discussions with equipment vendors and industry specialists, and the final report will be retained within the final Palisades corrective action document package at the plant site.

The EOG design basis accident load profile is a depiction of the postulated electrical load that will exist at a certain time after a worst case design basis accident. The accident load profile is a combination of the automatically sequenced loads and any operator manually initiated loads allowed by our Emergency Operating Procedures,(EOPs). Attachment 3 provides the time dependent LBLOCA accident load profiles for both EDGs. The load profiles were recently revised on 10/5/94. The difference between the present versions and the past versions is a timing change for the. manual initiation of control room ventilation condensing units and containment hydrogen recombiners by control room operators. The past versions assumed that the two mentioned loads were initiated at thirty minutes after the start of the event. The present versions.

assume that the loads are initiated at sixty minutes from the start of the event. The two manually initiated loads were conservatively placed in the past load profiles well in advance of their need, and also well in advance of their realistic starting times during the initial stages of an accident.

Thus, the past versions of EOG loading profiles contained approximately 104 KW of conservatism at the critical thirty minute period after the start of the event. The difference between the present and past load profiles is considered conservative margin that exists to be considered in the total overview of this issue. The reduction in the actual loading was not factored into this evaluation of safety significance because past EOPs did not restrict the manual loading from occurring at the earlier time.

EOP changes have been completed and an increased margin is now reflected in the load profiles.

r

.. \\.

NRC Form 388A 19*83)

U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMB NO. 3160-0104 EXPIRE~: B/31 /86 FACILITY NAME 111 Palisades Plant

Corrective Action

LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMBER 121 YEAR LER NUMBER 131 SEQUENTIAL NUMBER REVISION NUMBER 0500025594-o1 7 -o Corrective action for this event includes the following actions:

PAGE 141

1.

Complete short-duration peak load tests for both EDGs during the next two monthly tests.

. The monthly testing will monitor engine performance and establish trend information, pertaining to fuel rack position relative to KW loading, at loads of 2300 KW and above.

7

2.

Upon the completion of Action 1, (monthly peak load testing), establish operability criteria in the monthly EOG test procedures to ensure that EOG load carry_ing capability for design basis loads is verified. At this time it is anticipated that available fuel rack travel at a given load, with a known correlation between available travel and load output, will provide an accurate method to monitor design basis capabilities.

3.

Establish periodic design basis testing for both EDGs. The anticipated testing interval is once each refueling cycle, and will be based on the results from the trending and testing from Actions 1 and 2.

4.

Determine the root cause for the apparent minor engine degradation that has occurred on EOG 1-1 with respect to the original 1969 engine testing. Enhance the present performance monitoring program for the EDGs to maintain the engines at or near peak performance.

5.

Establish preventative maintenance controls to periodically monitor engine governor performance and control setpoints to ensure governors perform as expected in both "Parallel" and "Unit" modes.

6.

Establish administrative controls to ensure that changes to the plant design basis that occur through analysis are properly controlled and evaluated for potential verification testing.

7.

Evaluate the design margin that exists for both EDGs with respect to the maximum accident required loads and determine the possible alternatives that exist to increase the margin.

8.

Perform a detailed assessment of safety system testing to ensure plant design basis requirements are being properly verified. The order in w~ich the safety systems are assessed will be in accordance with their PRA ranking.

9.

Perform further EOG dynamic load model analyses to determine; the capability of an EOG to support the largest electrical transient motor start that results from a sequencer malfunction, and the approximate maximum transient motor start that results in an EOG trip,(to provide further validation of the EOG load model).

ATTACHMENT I Consumers Power Company Palisades Plant Docket 50-255 EMERGENCY DIESEL GENERATORS KILOWATT TABLE NOVEMBER 22, 1994 I Page

ATTACHMENT 1 EMERGENCY DIESEL GENERATORS KILOWATT TABLE PARAMETER DIG 1-1 DIG 1-2 TWO HOUR ACCIDENT RATING 2750 KW 2750 KW CONTINUOUS RATING 2500 KW

. 2500 KW TECHNICAL SPECIFICATIONS 2300 TO 2500 KW 2300 TO 2500 KW MONTHLY LOAD TEST RANGE MAXIMUM ANALYZED POST OBA PRESENT - 2584 KW PRESENT

- 2559 KW ACCIDENT LOAD I NG PREVIOUS - 2688 KW PREVIOUS - 2663 KW MAXIMUM LOAD CARRYING AS FOUND

- 2438 KW AS FOUND

- 2651 KW CAPABILITY-SEPTEMBER 1994 (reference notes 1, 2, AS LEFT

- 2714 KW AS LEFT

- 2697 KW and 3}

INCLUDING

- 2722 KW INCLUDING

- 2794 KW RECENT TESTS - 2735 KW RECENT TESTS

- 2736 KW ORIGINAL VENDOR PEAK 2810 KW 2827 KW LOAD TESTING-1969..

(reference note 3}

PALISADES PRE-OPERATIONAL 2730 KW 2720 KW TESTING-1971 (reference note 3}

NOTES:

l~ The as-found and as-left data includes an instrument inaccuracy penalty of 0.5 % of the reading.

The original vendor and pre-operational testing results are not penalized for instrument inaccuracy.

2. The first as-left recorded value for the EDG 1-2 is the best estimate of steady state engine output.

The first peak load test was.performed during a period when the electrical grid was unsteady.

The additional EOG 1~2 peak load test results shown above are more indicative of its full load capability.

3. During the as-left testing for EOG 1-2, the original vendor testing, and the Palisades pre-operational testing there was additional fuel rack travel available.

For the as-left testing results for EOG 1-1, the engine was essentially at full fuel.

ATTACHMENT 2 Consumers Power Company Palisades Plant Docket 50-255 EOG HISTORICAL LOADINGS NOVEMBER 22, 1994 I Page

KW

EOG Historical Loadings (Max. OBA Analyzed Load) 2100--------------------~-----~-2~4-""2liJS-_____ _

2flP3

+---------.1-----+------+------+-----**** *-*

.. 2~1

~OO-l------+---r---+----r---1---~~-~~---.~

81 82 83 84 86 BJ 88 90

~J 94 95 Legend e edg 1-1

)( edg 1 -2 YEAR

ATTACHMENT 3 Consumers Power Company Palisades Plant Docket 50-255

. EDG STEADY STATE LOADINGS

. NOVEMBER 22, 1994 4 Pages

EOG 1-1 STEADY ST ATE LOADING KILOWATIS VS TIME llDG 1WO HOUR RATING (17'0 KW) 0 TIME ICIO 120 140 MINUTES

. ::ia 0

l>

c:

Vt

~

0 l>

G')

n I

c r-3:

0 m

0 z

ia (A) 0.,,

r-,,,

N

.EOG 1-1 STEADY STATE LOADING

2900.-------------------~----------------------------------

EOG TWO HOUR RA TING (1750 ltW)

J6llO 1-HL...--~*:---r lll41tW "4 I

2]00 '-

'*---.c.-7 JIOO i-211111D..... ______ *.._ ______

______..,......... ______

______ __.. ______ -"---------

o

TIME IO HD llO.

140 MINUTES

a ITI Vt ITI z -I ITI c

.G')

I.....

r-t

a 0

r-ITI

EOG 1-2 STEADY ST ATE LOADING KILOWATrSVSTIME iDO TWO HOlll llA11NO C2'150 ltW)

JMJltW-4 A en,..,Sit.. a............ a.-.

0

~

t-t 0 c

~

~*

l>

c

~

G') n

~

24llO

z::

~

lb

()Q I

3:

11>

N w

r-m 2JOO 0

z c

- 4 (A) nDD

0 0..,,

t-t r-za*

am IJO 140 TIME MINUTES

EOG 1.. 2 STEADY ST ATE LOADING wrm PROPOSED EOP OIANGES 2'00.-------------------------------------------------------.

EDO 1WO HOWl RATING (J'HO ICW)

Ja -

ta -

~ *

a I

JSBICWrC

... -Lmd......_/

)>

Vt

'" ::i UIO z

~

)>

~

JG1KW'\\

IC n 2400 -

~

"ti

J:

lb OQ I

i:

ID N

~

llOO -

, __.,,,.7 m

r-0 z

):II -t IC 2llOO -

~ w

a 0..,,

ZICIO -

r-

JlllllD I

I I

I 0

40 eo IO, ICD IJO 140 TIME MINUTES

ATTACHMENT 4 Consumers Power Company Palisades Plant Docket 50-255 _

EOG 1-1 DEGRADED PERFORMANCE EVALUATION NOVEMBER 22, 1994 26 Pages

EDG 1-1 OPERATION DURING STEADY STATE AND TRANSIENT CONDITIONS Willi 2438 KW OUTPUT CAPABILITY GENERAL OBSERVATIONS EDG 1-1 steady state capability was determined, based on a precision transducer, to be 2438 KW on 8/30/94. Automatically sequenced loads are 2556 KW prior.to BOP manually actuated loads. Total peak automatic + manually added loads are estimated to be 2688 KW from 30-32 minutes after the design basis event. This loading drops to 2397 KW at 32 minutes due to RAS.

The significance of the 2438 KW maximum capability of EDG 1-1 resides in (1) the resulting reduced frequency steady state operation during accident conditions and (2) the reduced engine

short time response capability during induction motor starting transients. These two items are discussed in this evaluation.

STEADY STATE CONDITIONS As shown in Figure I, EDG 1-1 load demand will be 2556 KW immediately following OBA sequencing, rise to 2573 KW from 2 to 20 minutes, fall to 2537 KW between 20and 30 minutes, *and then briefly rise to 2688 KW for a 2 minute duration. Load demand will drop to.

2397 KW at 32 minutes and eventually rise to 2457 KW for the remainder of the first two hours of the accident. Load will remain below 2500 KW for the remainder of the accident.

Since the EDG engine cannot fully supply these steady state KW loads, the engine will reduce in speed which in tum results in a reduction in system operating frequency for the connected 2400 and 480 volt loads. A reduction in reduced operating frequency will directly affect motor and connected pump and fan speeds. Pump flows and fan outputs will be reduced which in tum will result in lower KW loading on the EDG. The exact reduced* KW loading is difficult to determine since the connected loads will not all react the same to the reduction in system frequency. Existing inhouse models have limited engine and pump models to accurately simulate such a steady state event.

Induction motors can operate successfully with frequency variations of 5 % if voltages are maintained at 100% (ANSI C50.41-1982). Thus, assuming the static excitation system can sustain I 00 % voltage during reduced frequency operation, steady state conditions at 57 hertz would be assumed to be acceptable to connected induction motor loads. While motor currents will slightly rise on large induction motors due to reduced frequency operation, the increase in current will be offset by the reduction in connected pump or fan load. No adverse affects on equipment or tripping of individual loads are expected. Tables I and IT, attached, summarize the effects of small frequency changes on equipment and the present EDG trips.

It is estimated and verified by modeling that EDG 1-1 frequency will be approximately 58 Page 1

'Y' I

hertz following automatic sequencing of DBA loads and will remain at 58 hertz until 30 minutes when loading increases to 2688 KW. At this time, EDG 1-1 frequency will most likely decay to approximately 56 hertz for the 2 minute loading period and then return to 60 hertz at 32 minutes when EDG load demand drops to 2392 KW (which is below the EDG 1-1 2438 KW capability). Voltages are expected to remain at 2400 volts due to adequate excitation system response.

No adverse motor currents or EDG response is expected since 56 hertz is only slightly below the -5 % frequency rating of the connected induction motors.

Pump flows and thus connected loads will be slightly reduced. Operation at 56 hertz will occur for only a 2 minute duration after which frequency will return to 60 hertz since load demand drops to 2392 KW at 32 minutes into the accident.

In summary, electrically EDG 1-1 is expected to be fully operable during the reduced steady state frequency operation at 58 hertz and over the 2 minute period when frequency may decay to 56 hertz. However, the effects of reduced pump flows on accident mitigation must be determined. What are the effects of a 2-4% reduction in pump speeds (operation at 58 hertz) and subsequent reductions in pump flows during the first 30 minutes of an accident?

Similarly, what are the effects of a 4-6% reduction in pump speeds (operation at 56 hertz) and subsequent reductions in pump flows for a 2 minute duration? Answers to these questions will determine the safety significance of the reduced 2438 KW steady state output capability of EDG 1-1.

TRANSIENT CONDITIONS The short time output torque of the engine at rated speed is only as great as engine fueling will permit. If momentarily overloaded, the engine will lose speed. It will slow if instantaneous fuel supply is not sufficient to drive instantaneous load. Some stored energy is available due*

to angular velocity and mass of the engine and generator which will aid in engine response to an applied transient load. The static excitation system used on the EDGs *is relatively fast and sized to handle the large starting transient currents demanded by the induction motors during starting.

It is estimated that the worst case transient step load could occur when EDG 1-1 is loaded to

2334 KW and Auxiliary Feedwater P8A starts due to delayed AFAS. This results in an initial momentary transient load demand of 397 KW which is 293 KW above the EDG maXimum output capability of 2438 KW. Two things may occur due to this short time transient overload demand:

(I) The excitation system is able to overcome the voltage reduction due to loss in engine speed and the motor will start and, once the starting transient is over, the engine

. will recover in speed at a reduced steady state system operating frequency as previously discussed.

(2) The excitation system cannot sustain adequate voltage during the motor start.

Engine speed and voltage continues to decay resulting in "voltage collapse."

Individual loads and the EDG trip on overcurrent.

Based on engineering judgement it is felt that EDG 1-1 would be capable of starting and accelerating all DBA loads (including P8A) with a 2438 KW output capability (scenario (1) above). This conclusion is based on the following observations and later verified by dynamic simulations of the event as summarized in the next section of this evaluation:

(1) Per ALCO infonnation, the "engine is capable of accepting a 1000 KW step load."

(2) Typically, diesel engines are sized to accept a step load of 50 % rated (1250 KW)

(3) The maximum step load in this case in 397 KW I 2500 KW or 16% of continuous EDG output rating (4) Measurements of frequency during motor sequencing in 1987 (Attachment A) indicated less than 1,3 deviations indicating the present engine/ governor is very capable of handling the KW demand of the induction motors being started during DBA sequencing.

(5) The static excitation system has been sized to handle the large inrush current demanded by the induction motors which is primarily reactive load during starting. It would most likely be capable of sustaining adequate voltage during reduced engine speeds.

DYNAfv.llC SIMULATIONS OF WORST CASE STARTING TRANSIENT Computer models have been developed of the Palisades emergency diesel generators and their

, associated emergency core cooling system induction motors during sequencing and results compared with field tests. Attachment Aof this evaluation documents the IEEE Paper which was published in September, 1993 documenting the inhouse models. In order to examine the frequency excursions experienced by EDG 1-1 while starting Auxiliary Feedwater Pump P8A, computer simulations were completed starting the motor during each of the loading plateaus as shown in Figure 1. In each case the prestart loading prior to starting P8A was assumed to be 354 KW lower than the given loading plateau. Table I summarizes the results of the dynamic simulations. Included in Table 1 are the results of a dynamic simulation assuming an engine output capability of 2714 KW - the final maximum output capability following corrective actions. As can be seen in Table 1, with EDG 1-1 output capability of 2438 KW, P8A is successfully started in all cases with a subsequent reduction in EDG steady state frequency.

The steady state frequencies from the computer simulations are very close to the predicted frequencies previously discussed i.e. 58 hertz for the first 30 minutes declining to a minimum of 56 hertz for only a 2 minute period. Motor speeds (pumps and fans) decay 2-6% below nonnal operating values during the first 32 minutes of EDG 1-1 loading.

Figures 2-5 summarize the response of the engine mechanical power, generator excitation system, and engine speed/frequency when starting P8A at the various loading plateaus. As can be seen in these plots, the generator excitation system responds and maintains voltages at 2400 volts during the motor starting transient. Subsequent engine transient response and resulting system steady operating speed/frequency is dependent on the preload prior to starting the motor. In each case, the generator and engine, as predicted, maintain the connected. loads at minimal reductions in operating frequencies (reductions of only 2-4 hertz). No problems were identified as summarized in Tables I and II.

Figures 6-9 document the ability of the EDG to start and load P8A with the maximum as left EDG 1-1 output capability of 2714 KW. As can be seen in these plots, the EDG is capable of starting P8A during all loading plateaus. Steady state frequencies return to 60 hertz following the P8A motor starting tlansient. No operating problems were found.

S&L REVIEW OF EDG RESPONSE DUE TO REDUCED ENGINE OUTPUT CAPABILITY The following summarizes the scope of an independent S&L review of CPCo efforts to determine the EDG capabilities to perform their design functions during accident conditions.

It will include a review of past operability (steady state and ttansient) when EDG 1-1 maximum output capability was 2438 KW as well.as its present capability of 2714 KW and when EOG 1-2 maximum output capability was 2651 KW as-well as its present capability of 2697 KW.

1.

Review industry experience at other Nuclear Plants such as Salem. Determine if a similar problem has been experienced and what was done to resolve it. Include a.

search of industry publications such as IEEE on such a topic. (Paul Gire, CPCo Licensing has submitted a search of other Nuclear Plants regarding the subject and*

would be a good.first contact.)

2.

_Review the completed EDG dynamic simulations of OBA sequencing which were

perfo~ed using the inhouse CPCo PSS/E dynamic model of the EOG from a qualitative perspective (do not repeat computer modeling or simulations). Are the results adequately predicting engine and system response? Do they appear reasonable?

(KE Yeager is the contact for this area of review.)

3.

Review the effects of reduced motor speeds on motor pump flows which are being determined by the Palisades Reactor Engineering Dept from a qualitative perspective (Do not repeat computer modeling or simulation,s). Are the predictions reasonable?

(TC Duffy is the contact for this aspect of the system response.)

4.

Contact industry vendor experts and review the expected engine response for the steady state and transient conditions. Does it agree with the conclusions drawn from the CPCo dynamic simulations of EDG response? Are the predictions of pump flows Page 4

acceptable?

5.

Review the effects of reduced frequency operation on the associated induction motors connected to the safety buses.

6.

Develop a summary report of the S&L review of the EDG response due to reduce engine output capability for items 1-5 above. Draw conclusions on EOG and system operability based on the infonnation reviewed which can be presented to the NRC in upcoming discussions (Enforcement Conference).

KE YEAGER 10/27/94 Page 5

Table I Summary of the Effects of Reduced EDG Operating Frequency on Connected System Components Componet 2300 and 460 volt Induction Motors 480 volt AC Contactors Motor overcurrerit relays, differential relays, undervolta~e relays Safety Related MOVs Safety related relays/controls Battery Chargers effect Overall reduction in Runing I due tb reduced pump/fan flows Increase in No Load current (estimated to be 3-8%)

reference PSS/E Simulations, discussions with PTI, S&L GE Bulletin Increase in Locked GE Bulletin Rotor Current (estimated to be 3-8%)

Slight increase in current. AC Contactors are dual frequency rated; 50/60 hz Reduction in relay sensitivity -

pickup Norie. Started in first 60 seconds of DBA at 60 hertz None. Fed by preferred AC or Class lE DC None. SCR controlled, no effect from inp"tit frequency variations of 3-7%

S&L Review, and Cutler*

Hammer Bulletin Generating Station Protection -

GE*info DBA sequenc;ing logic*

P&ID

Drawings, schematics JD Slinkard, System Engineer

EDG Trip Engine Overspeed Lbw Lube Oil Pressure Overcrank Field shutdown timer Overcurrent relaying Generator differential relaying Table II Surrunary *of EDG Trips *

Description

At 10% erigine overspeed a single sensor actuates the oversp~ed relay which energizes the SDR Enabled on both engine start circuits after jacket water pressure switch is closed (at lOpsig and 35 sec delay). If two oil pressure sensors see

<40 psig (approx 1/2 engine speed -

450 rpm) the SDR will be actuated When starting ijOU have 35 sec to obtain:

start ckt A: >10 psig or 120rpm start ckt B: >10 psig If the tach pack senses engine rpm less than 600 RPM, the field shutdown timer will trip the output breaker and shutdown the generator field within 120 seconds unless engine speed recovers above 600 rpm x phase -

824 amp pickup y phase -

824 amp pickup z phase -

752 amp alarm only differential relay settings Page 7

D>

()Q ro 00 7Aa1.~ -t l!DG 1-1 RKDOCSD SNGINB.CAPABILITY 2714 kw max 2 438 kw max c..Af.<\\ll/L IT~

2688 kw load (1) 2556 kw load(2) 257j,kw load(3) 2537 kw loa (4) 0-2 min load 2-20 min load EguiEment

.!...fil seieu!

'.!..1&

SElEU!

1..1&

SElEU!

CCW P52A 67 0.99 60.6 0.96 59.8 0.96 SW P7B 89 0.99 80.5 0.96 7.9. 5 0.96 AF P8A 104 0.99 95.7 0.96 94.4 0.96 CCW P52C 67 0.99 60.6 0;96 59.8 0.96 LPSI P67B 99 0.98 90.0 0.96 88.8 0.95 CSP P54B 50 1.0 45.7 0.97 45.1 0.96 HPSI P66B 91 0.98 83.0 0.96

81. 9 0.95 CSP P54C 50 1.0 45.7 0.97 45.1 0.96 CHG P55C 102 0.99 97.8 0.96 97.3 0.96 BA P56B 39 0.99 36.3 0.96 35.9 0.96 CCF V4A 46 1.0 44.3 0.97 44.1 0.96 EDG 1-1 718 1.0 656 0.97 654 0.97 60 Hz 58.4 Hz 58.1 Hz Prnech*2688 kw Pmech*2371 kw Pmech*2361 kw

References:

1)

D2714kw.out,d2714kw.prn,d2714comp.sav.

2) dl4.out, dl4.prn, d14comp.sav, dl4a.out, d14a.prn, dl4acomp.sav.

3) d13.out, d13.prn, dl3comp. sav, dl3a. out, d13a.prn, dl3acomp. sav..

4) dl5.out, dl5.prn, d15comp.sav, dl5a.out, dl5a.prn, dl5acomp.sav.

5) dll.out, dll.prn, dllcomp.sav.

GJBrock 9/30/94

~

1ojI/f:/

20-30 min load

.!..1&

Seleu!

61. 6 0.97 81.8 0.97 97.1 0.97

61. 6 0.97

91. 4 0.96 46.4 0.97 84.2 0.96 46.4 0.97 98.4 0.97 36.8 0.97 44.5 0.97 659 0.98 58.6 Hz Pmech*2381 kw 2688 kw load(5) 30-32 min load 1..1&

Seleu!

54 0.93 72 0.93

. 86 0.93 54 0.93 81 0.93 41 0.94 75 0.93 41 0.94 94 0.93 33 0.93 42 0.94 623 0.94 56.4 Hz Pmech-2293 kw

~

.. ~...

EDG 1-1 STEADY STATE LOADING KJLOWATIS VS TIME*

2900..--------------------------------------------------------------------------------.

2800 2700 2(00

.7s-n Js~r.:.

2500 2400 2300 2200 2100 2000 0

As Left 9194 *Max Load Carrying Capability (2714 KW) 2688 KW--<.r--- Automatically Sequenced Load Ph11 Manual Load ---

0>S3 20 40 60

.TIME-:MIN 80 100 120 140

I

\\,

I 59.000 2163 KW 2400 v 900 RPM 60.HZ 60.500

.. -..;R--,'

EOG 1-1 2438 KW ENGINE MAX MECH POWER OUTPUT 2556 KW LOAD @ 60 HZ START AUX FEED P8A 2371 KW

'-----+--------------------+--------------------*--------------------

62.000 65.000 68.000 63:500 66.500 69'.500 TIME(SECONDS) 71.000 2400 v 875 RPM 58.4 HZ 74.000

12. 500 2438 (11 0

0 0

0 ::c

~

P.I 0

Cil I.....

i 0

(11 0

0 0

0 KW 0

0 0

t I

l I

0 0

0 ::c

~

w P.I 0

Cil I.....

(/)

"O 0

0 0

I I

I

0 0

. THU; 0 ::c

~

Cl>

P.I 0

Cil I.....

"O

~

0 ::c SEP 29 MAX/2556 3: trl g~

H :x:i ITJ G) 1-( trl z

~Q trl ITJ 0 II H

. trl 00 (/) e w trl wt-'

w G)

ITJ trl oz

x:i trl ITJ

x:i H

w )>I t-'

x:i 1-3 trl trl 0

~ :x:i

0.

trl.....

x:i I

()

0 t-'

c:.

0 rt

(/)

c

~*

~

0 c e

(/)

trl z

trl

~

()

0 z trl

()

1994 14:24 KW LOAD

I-

., \\:

~*

I

~ 2180 KW 2400 v l

900 RPM J

60 HZ 59.000 60.500 EOG 1-1

2438 KW ENGINE MAX MECH POWER OUTPUT 2573 KW LOAD @ 60 HZ START AUX FEED P8A 2362 KW\\

'-----*--------------------+--------------------+--------------------

2400 v \\

871 RPM 58.1 HZ\\

-*-A-62.000 65.000 68.000 71. 000 74.ooo*

63.500 66.500 69.. 500 72.500 TIME(SECONDS) 2438

(.11 0

0 0

0

i::

z:

I:"..

l"l 0

G'>

I

i 0

(.11 0

0 0

0 KW 0

0 0

I I

l I

0 0

0 :c

~

w l"l 0

G'>

I en "O

0 0

0

FRI, 0 :c

~..

Cl) l"l 0

G'>

I "O

~

0 :c SEP 30 MAX/2573 3: t11 03:

0 t11 H ::tJ "1 Gl i< t11 z

~R t11 "10 II H t11 00(/) e w t11 wt-<

w Gl "1 t11 oz

0 t11 "1

0 H

ttl )>I t-<

0.,

t11 t11 0

~ ::0

0.

t11.._.

0 I w

()

0 t-<

c 0

rt

(/)

c

0 t11 g;

0 c e

(/)

t11 z

t11

~

n 0 z t11

()

1994 08:45 KW LOAD

N

~.

'-- 2144 KW 2400 v r

900 RPM 60 HZ*

59.000 60.500 EOG 1-1 2438 KW ENGINE MAX MECH POWER OUTPUT 2537 KW LOAD @ 60 HZ START AUX FEED P8A hr; u/l. e </

, 2381 KW-....._

'--- - - -+ -- - ----- -- - - - -- - - - - -+-- - -- --- - - - - ---- --_--+-- -- -- - --- -- - - - - -- - -

62.000 63.500 65.000 68:000 66.500 TIME(SECONDS) 69.500 2400 v~.

~ --

71.000 72.500 2438 U1 0

0 0

0 U1 0

0 0

0 0 =

~..

PJ 0

G)

I.....

i 0

0 0 *o 0

f I

l I

0 0

0 =

z t-<

w PJ 0

G)

I.....

(/)

"Cl 0

0 0

0 =

~..

CJ)

PJ 0

G)

I.....

"Cl

~

0 =

THU, SEP 29 1994 KW MAX/2537KW t'lj H

t-<

tr!

p.

01 0 c rt 3: tr!

03:

0 tr!

H )::I t'lj G) 1-( tr! z

aQ tr! '

t'lj 0 II H tr!

<X> (/) --

W trl wt-<

w G) t'lj tr!

oz

)::I tr!

)::I 01 )>I

)::I.,

tr! 0

~ )::I tr!.....

)::I I.....

()

t-<

0

(/)

c

~

0 c

(/)

tr!

z tr!

~

()

0 z tr!

()

15:06 LOAD

"C lb OQ ti>

w EOG 1-1 24 38 KW ENGINE MAX MECHANICAL POWE.R OUT 2688 KW LOAD @ 60 HZ START AUX FEED P8A p:::"', f,_,,A e S--

2294 KW

~t-------

r-2293 KW

'-- - - - -+- - - - - - - -.- -- - -- - - - - - - * - - - - - - - - - - - - -- --- -.-- ~ - -- -- - - - - - - - - - --- - - -

. 2400 v f 900 RPM 60 HZ 59.000 62.000 60.500 2400 v~

846 RPH

'56.4 HZ

~~-------~-------

63.500.

.65.000 68.000 66.500 TIME(SECONDS) 69.500 71.000 74.0QO 72.500 2438 I-'

(.11 0

0 0

0 :c z t"'..

I-'

P:I 0

(;')

I-'

I I-'

=i 0

(.11 0

0 0

0 K;W 0

0 0

f I

l I

0 I-'

0 0

0 0 :c z t"'

w P:I 0

(;')

I-'

I I-'

(/)

0 0

0 I

I I

0 0

WED, 0 :c

~..

(I)

P:I 0

(;')

I-'

I I-'

~

0 :c SEP 28 MAX/2688 3:M g~

H~

":! G) t-< M z

§Q M

":JO II H

. M e oo en WM Wt-<

w G).

":JM oz

~M

~

H w )>I t-<

~11 M

MO

~~

0.

Ml->

~I

()

0 t-<

i::

0 rt en c::

~

M.

)>I z 0

c::

{/)

M z

M

~

()

0 z M

()

1994 14:37 KW LOAD I

'-o ll>

OQ (1)

I-

- t-

---.---~---.--.-----.-----:-i-1.

~I 1,,

I, I

I I

'\\__ 2163 KW 2400 v 60 HZ r 900 RPM 59.000 62.000 60.500 63.500 EOG 1-1 2714 KW ENGINE MAX MECH POWER OUTPUT 2556 KW LOAD @ 60 HZ START A,UX FEED PBA

~'i...,.,c.e r;,

2556 KW~

2400 v 60 HZ 900 RPM\\

' 65.000 68.000 71.000 74.000 66.500 69.500 72.500 TIME(SECONDS) 2714 I-'

°'

0 0

°'

0 0

0 :c

~..

I-'

('IJ 0

Ci)

I-'

I I-'

'.j KW 0

I-'

0 0

t I

l I

0 I-'

0 0

i::

~

w

('IJ 0

Ci)

I-'

I I-'

(I>

'O 0

0 0

I I

I

0 0

THU, 0

i::

~-..

Q)

('IJ 0

Ci)

I-'

I I-'

'O

~

0

i::

SEP 29 MAX/2556 3: tr:1 g~

H ::0 "1 G) 1-( tr:1 z

~Q tr:1 "1 0 II H M

00(/)e Wtr:I wt-<

w G)

"1 tr:1 o*z "1

0 tr:1 H

IJj ~

t-<

0.,

tr:1 tr:IO

0.

~ ::0 t:r:I....

0 I A

Pl

()

0 t-<

i::

0 en rt c::

0 tr:1

)"

z 0

c:: e en tr:1 z

tr:1

~

()

0 z tr:1

()

1994 14:12 KW LOAD

"C Ill (IQ l'D

\\JI 2180 KW I

I I

1'

~ 1

I 2400 v 1--...

~

I I

60 HZ r 900 RPM

/

\\,

59.000 62.000 60.500 63.500 EOG 1-1 2714 KW ENGINE MAX MECH POWER OUTPUT 2573 KW LOAD @ 60 HZ START AUX FEED PBA F19vA.c 7 2573 KW 1 65.000 68.000 66.500 TIME(SECONDS) 71.000 69.500 2400 v 900 RPM 60 HZ

74. 000 72.500 2714 (JI 0

0 0

0 :c

~..

P:I 0

G'l I

i 0

(JI 0

0 0

0 KW 0

0 0

f

.1 I

J.

I 0

0 0

0 0 :c

z:

t"'

w P:I 0

G'l I

(/).,,

0 0

0 I

I I

I I.*

0 0

THU, 0 :c

~..

Ill P:I 0

G'l I

~

0 :c SEP 29 MAX/2573

--111..

~iii

- R

~

3: tr!

g~

H~

"l G) 1-( tr! z

1Q tr!

"10 II H

tr!

oocne Wtrl Wt;-t w

G)

"l tr!

oz "l

~ tr!

H tJj ~

t;-t

~ >-:3 tr!

tr! 0

0.

~~

tr!........

~I w

Ill

()

0 t-'

0 c:

(/)

rt c:

~

0 c: e

(/)

tr!

z tr!

~

()

0 z tr!

()

1994 13:41 KW LOAD

>ti lb

()Q CD

~.

'*1 '*

I I

I

', ___ ~

'--- 2144 KW 2400 v r

900 RPM 60 HZ 59.000 62.000 60.500 63.500 EOG 1-1 2714 KW ENGINE HAX MECH 2537 KW LOAD @ 60 HZ START AUX FEED P8A h'fvAC S1.

~

........ ~

POWER OUTPUT 2537 KW ~

~---------------------*--------~----------~

2400 v 60 HZ

oo RPM\\

65.000 68.000 71.000 74.000

. 66.500 69.500 72.500 TIME(SECONDS) 2714 U1 0

0 0

0 ::c

~..

ts:!

0 Ii>

I

)

0 U1 0

0 0

0 KW 0

0 0

.0

c

.~

f I

l I

0.

0 0

0 w

ls:I 0

Ii>

I (I>.,,

0 0

0

THU, 0 ::c

~..

a>

ls:I 0

Ii>

I

'ti 3:.

ts:!

0 ::c SEP 29 MAX/2537 1994 KW

0.

l11 DJ 0 c*

rt i4:56 LOAD

"ti Pl (IQ 11>

I-'

I EOG 1-1 2714 KW ENGINE MAX MECHANICAL 2688 KW LOAD @ 60 HZ START AUX FEED P8A POWER OUT h9 vtte '1 2688 KW""\\

+

1*',*

~

/,

'-- 2294 KW 2400 v 900 RPM r

60 HZ 59.000 60.500 2400 v 60 HZ 900 RPM:\\

62.000 65.000 68.000 71.000 74.000 63.500 66.500 69.500 72.500 TIME(SECONDS) 2714 I-'

UI 0

0 0

0 :c

~..

I-'

. Pl 0

t;)

I-'

I I-'

j 0

UI 0

0 0

0 KW 0

I-'

0 0

0 :c

~

w Pl 0

t;)

I-'

I I-'

(I>.,,

o f

I I

0 I-'

0 0

0 I-'

0 0

THU, 0 :c

~..

CX>..

Pl 0

t;)

I-'

I I-'.,,

~

0 :c SEP 29 MAX/2688 1994 KW 3: t':I g~

H

x:J

":! G) to< t':I z

3Q t': I

":! tJ II H t':I CD(/) e Wt':!

wt-'

w G)

":! t':I oz H

x:i t':I t-'

x:i t':I tJj )>I

x:i >-,3 t':IO n.

~ :x:i N

-..J t':l 1-'

x:J I

~

()

~

t-'

0

_Q

(/)

c c

~

tJ c e

(/)

t':I z

t':I

~

().

0 z t':I n 11:23 LOAD

Urr4-cllh'J,Vr /I

. IEEE Traruactioru on Energy ConvezJll VoJ.. 8. No.~. Seplember 1993

, a

' MODEL"'8 OF EMERGENCY DIESEL d!'NERATORS

- 433 IN AN 800 MEGAWATT NUCLEAR POWER PLANT K.E. Yeager, Member IEEE Consumers Power Company Jackson, H1ch1gan Abstract -

Comc>uter models have been developed of emergency d1esel generators and the1r assoc1ated emergency core cooling system 1nduct1on motors during sequenc1ng and results compared w1th f1eld tests.

Models requ1red to perform stud1es of emergency diesel generators 1n a nuclear *plant are presented.

F1eld measurements 1nd1cat1ng d1fferent response of two seemingly identical generator exc1tation systems are discussed.

Results of 480 volt ac contactor dropout testing are prov1ded for*determin1ng voltage 11m1ts 1n the 480 volt syste111 during motor start1ng trans1ents.

Keywords -

01esel Generators~ Induct1on Motors.

INTBOOUCTION Design requirements of nuclear plants include backup power supplies provided by emergency diesel.

generators (EOGS).

These generators provide the sole source of power to large emergency core cool1ng system (ECCS) loads during a loss-of-coolant accident (LOCA)

coincident with a loss of off-site power. via the util1ty transmission network.

Digital computer* software and hardware technology make it possible for a ut111ty to develop and maintain accurate models of these critical power supplies to assess their adequacy due to electrical lllO!l1fications and load additions over the 11fe of the plant.

The models also make 1t possible to simulate LOCA scenarios and verify the capabi11ty of the EDGS to provide a safe and reliable source of power to cr1tical loads during an accident.

This paper sulllllllrizes the efforts of one utility to develop and maintain computer models of the EDGS and their associated ECCS loads in an 800 megawatt nuclear power plant. The models are compared to field test data frOlll routine testing of the EDGS following a refueling outage.

Additional discussion is provided concerning voltage limits in 480 volt systems through field testing of ac contactors. FinallJ, differences in response of two supposedly identical EOG excitation control systl!llS are presented.

THE NEED fORG£lilf !I ihfmGENCY DIESEL The need for develoPing and uintaining acC\\lrate models of EDGS and their associated syst1111 1n nuclear power plants can be sulllll!rized as follows:

1.

To verify that all ECCS induction llOtors will start and accelerate during a LOCA when feel frOCll the EDGS,.ttidl are weaker po.er supplies than the off-site power supply via the utility transmission nett110rtc.

2.

To verify that ECCS 110tors already running and.

loaded to LOCA conditions will continue to operate (e.g., not slow down and possibly 92 SK 561*1 !C A paper raco... ndad and approved by the IE!! Electric Kachinary co.. ittea of th*

IEE! Power Engineering Society for preaantation at the I!EE/P!S 1992 SWiiier Keating, Seattle, WA, J~ly 12*16, 1992.

K&nuacript aubmittad October 7, 1990; 11&de available for printing

..*ril '16, 1'992.

J.R. Willis, Helllber IEEE Power Technologies, Inc.

Schenectady, New York tr1p) during remaining motor start1ng transients.

3.

To ver1fy that 480 volt ac contactors w111 have adequate voltage to p1ckup when required and w111 not drop out during motor starting trans1ents.

4.

To assess whether electr1cal mod1f1cat1ons and load add1t1ons to the orig1nal EDGS and associated ECCS systems will exceed di ese 1

generator capabilities.

5.

To verify that ECCS induct1on motor sequenc1ng steps do not *overlap*, result1ng 1n excessive motor starting trans1ents and potential tripping of critical loads required during a LOCA.

6.

To provide s1rm.ilations of the EDG

and associated ECCS system loads during postulated LOCA cond1tions that cannot be duplicated through f1eld tests.

EMERGENCY DIESEL GENERATOR SYSTEMS Two independent EOG systems, each capable of

supply1ng power to ECCS motors for safely shutting down the reactor following a LOCA, are provided for the 800 megawatt nuclear plant presented in this paper.

The EOGS and associated ECCS induct1on motors used 1n the studies are su11111arized in Figures 1 and 2.

Each generator is rated 2400 volts, 3.125 HVA, and uses automatic sequencers to start the ECCS induction motor loads *. The tab le provided in each figure su11111ar1 zes the starting times of the *ECCS induct1on motor loads during f1eld tests* associated with required EOG testing as part of a recent plant refueling outage. As can be seen, ECCS motor loads such as containment spray are not started during routine EOG testing.

Such loads are included,

however, in the

final computer simulat1ons representative of LOCA conditions.

INQ!JCTION !!)!OB l<<)QELS The electrical and mechanical characteristics of the ECCS induction motors for each EOG system sunnarized in ffgur9S 1 and 2 are presented in Table I.

Manufacturers' speed versus.torque and current curves, as well as motor and load inertias, were available.'

These IKltor starting characteristics are critical when developing the llOdeTs of the overall EC~ systems fed by the EDGS.

figure 3 su-rizes the lllOdel chosen by the authors to represent the induction S>tors su111111rized in Table I and figures 1 and 2. It is a double-cage representation used successfully to represent deeii-bar rotor effects 1 n squirrel-cage inductfon motors during starting [1,2].

The model parameters were determined through trial and error by us 1ng the 1111nufacturers' data and a software package to interactively select parameters until the model reasonably duplicated the motor torque, current and power factor during start1ng. Table II summarizes the equivalent circuit parameters found for the ECCS induction motors using this method. Figures 4 and 5 present a c0111Parison of the manufacturers' data and model results for the 400 horsepower LPSI and HPSI motors. As can be seen in the figures, the double cage induction motor model dupl1cates the !llllnufacturers' data quite closely. Similar results were obta1ned for thl:!

remaining ECCS

fnduction.ator models used 1n the studies.

088,*8%9193$03.00 C 1992 IEEE Page 18

434

"'6 I'~

n:~.*U r!e~m

... I *IO YOLI

! ~~=~(...,

I 16 CO*' Cl..C t.1', I

~!21

CC~*~

1 1~c6 j 0110~

Alllt ".. OLllllG

'2'6 IO~IC OCIO **H Flpn 1. EDI 1-1 Syst..

2"100 t'OI.. T IUS *10 J-S(AY.,..

.,. -6 S(ll'IW !fT*

7C t51 q) [OC 1-1 srr:, 1 1 r 1111 1uc1

.,.... 20

10 YOI. I LOllO

([NT[~ 20 o.oo*

2."

b.bO

~.*o IJ.00 1~.u u...

2*. H 121 I

t.--.

6 CO"it Ct.C

52C ou1 *no 121

IC CO*I IOMY

to YO&. I LOllO C!*TI~ 12

~,,,

~**

"l.luU"\\IM

- - 12

l*o IO~IC ACIO **u 10 bO. at.

Or I fllllflD ~::' CM-CC'

- H YIA 121 CC' YJa Flpn *2. Ell 1-2 S,..

Flpn 3. l1Ul1 Cap Ila.till llltlr.....

Page 19 TaMtl s_, tf li*llll lmr Chncurtstln Motor Horse Sync ICVA H(sec) pa.er,,.

. ~ -.-

ED6 1-1 Misc 118 1800 118 0.5 EOG 1-2 Misc 205 1800 205 0.5 8oric Acid 30 3600 29.5 1.0 Charaina Pumgs 75 1800 72.4 0.256 Chargin~ Pump 100 1800

99. l 0.44 PS A Cont Cool i nq 75 1800 69.3 1.16 Cont Sgray 250 3600 219 1.18 HPSI 400 3600 343 1.18 LPSI 400 1800 367 0.426 Service Water 350 1200 330 0.73 Conmonent Cla 300 1800 267 0.385 Auxiliary Feed 450 3600 402 0.6%

P8A Auxi l ital Feed 400 3600 343 1.18 Air Handling 25 1800 20 0.5 V95 96 Air Handling 20 1800 21 0.5

V26A, 8 Tablt II Saawy 1f l1~1etl11 lltlr D11blt Cage Eqalval11t Clrcllt Para*ttn (per nit 11 *otor base) llOtor IA LA UI 11 L 1

2 c2 BIG 1*1 0.02 0.015 2.7 0.02 0.15 0:015 o:o5 BIG 1*Z, 0.02 0.015 Z.7 0.02

0. 15 0.015 0.05

ltc lorlc 0.01 o.cm Z.7 o. 16 O.OI 0.009

0. lO kid Cllerl 0.009 0.02 2.7 O.lO 0.007 0.011 0.015 Cllerl 0.02 o.ozs 2.7 O. IS 0.04 0.015

0. I]

I'll '5SA c-O.OOI 0.04 2.7 0.11 0.05 0.011.

0. 14 eoa11-c-

0.015 0.068 J.,

0.06 O.OJS 0.007 O.il4 0 01J 0.05 J.O

0. 1 0.09 0.018 0.-JCS l"I 0.015 o.oss J.O 0.065 0.09 0.02

0. :2 Service 0.02 0.029 J.O 0.06 0.09Z 0.014 O.CI *

~

o.oz o.os J.7 0.06 O.OI 0.012 J. :a Cll AMII O.OOI O.QI J.6 O.Ql 0.07 0.015 o.oa AMII,...

0.015 0.061 J.1 0.06 0.035 O.J07 J.04 PIC Atr lld 'I a.oz 0.015 2.7 0.02 0.15 0.015

0. J5 9! 96 Air lld o.oz 0.015 2.7 0.02 0.15 0.015 O.J5 VZ6A,I i

I I

I I t i I i

i t I i

/

TOtqUe (MOdel)

*-****-******--***--L*~=~.,

C!*C!

_..~_..-_,_ _ _.. _ _._ _ _. oO o.o SPUD(pu) 1.0 Flsn 4. lulf11tww Vena l1dll1 Cap lmr l*MI fw LPSI lmr t -~~

I rr.:.-. :-.:.-. ::-. :-.:.-. ::-. :.-. :-:-. :-. : - - - -..

_/ - **----- -

~

I r-

~

(118nUf.)

r-I Torque (lhnuf.)

}-

\\_ Tcwque (Model)

I,

C! 0 o*

Flsn s. lulfutss,.,. llBll Cap -

fw IPSI lltlr GENERATOR PQJEL The generato" are modeled us1ng a sa11ent pole representation and 1nclude the effects of the amortisseur w1nd1ngs.

F1gure 6 s~r1zes the block diagru of the model. Generator equivalent circuit and saturation data were obtained from the *nufacturer and are presented in Table III. A aift01' reduction of th*

subtransient and transient reecUnces provided a better fit between th* model and f1eld... sUNMnts of the motor starting transients.

The reduction of these paraaete" 1111 1nd1cat* add1t1onal aach1ne saturation dur1ng th* large motor starts not r911rnented 1n the.

or1g1nal 111chine data.

435

..... _____,d Flgn I. El11tnmpltl1.... If* S1H11t Piii Cnerater T1M1 Ill S..uy If IHntlr Pan.ten Inductances Nach1ne (per un1t) con~~ts Saturation Data Ld* 1.56 La* 1.06 Tdo'* 3. 7 S(l.0)* 0.1724 Ld I *.296 La** 0.177 Tdo** 0.05 S(l.2)* 0.6034 Ld** 0.177 Tao** 0.05 Ll* 0.088 H* 1.0716

.. Note:

Ld* and Lq* adjusted to 0.15 and Ld' to 0.26 based on f1eld tests. All parameters 1n per- '

unit on 1111ch1ne base of 3.125 MVA, 2400 volts.

EXCITATION SYSTEM !)DEL A stat1c exc1tat1on syste11 (si111lar to the IEEE t:ti>e ST2 a:ldel [3]), hav1ng both generator current and voltage as 1nputs 1s *used. for the excitation control systee for each EOG.

No aanufacturer's data, however, was available to develop such a model. A control system block d1agr111 of the excitation system could have b~n develOJ)ed by the. 111nufacturer by remving the excitation system and perfort11ng bench tests. Th1s option did not

.at plant aporoval.

A suitable exc1tat1on control syst* model, h011Mver, had to be found which would approx1*t* the generator teMl1nal voltage response dur1ng the motor starting transients.

F1gure 7 SUllmllr1zes the exc1tat1on control system chosen by the author1, one wh1ch has been used successfully by other1 (C]. The t1* constants and exciter gain were adjusted unt11 a reasonable 1111tch was obtained betlllffn the model and f1eld tests for generator tera1nal voltage response during motor starting cond1t1ons for eech ED6. (The aotor starting conditions provided the res~* of the ED6 w1th both voltage and current feedback.) Table IV sumar1zes the final values used in the studies.

EMAX p.u.

K I

gr*..

1 *TA' ----*

1

T8 1 Other SIQnaJe Flan 7. Ellltltlll SJlll9....,

EMIN Page 20

436 e

Talllcll

- s_, " Excltatltl srn. P1r181tn G9rler1tor TA Tl c

Tl lRllma lfllllfn C-1

(.eel INCi EDG l*I J.O 6.0 77.5 2.0 6.0 o.o EOG 1*Z J.O 6.0 77.5 1.0 6.0 0.0 Detennining the ti1111 constants and gains in Figure 7 (through tr1al and error) to match response of the generator dur1ng *motor start1ng was t1me consuming.

Wh11e the model matched the motor start1ng trans1ents add1tional voltage oscillations were founcf in the lllOdei following the motor starts that were not evident 1n the f1eld measurements.

It 1s of the authors' opinion that further investigation is needed in the area of developing f1eld testing methods for excitation syst811S wh1ch use both current and voltage as feedback.

Such field testing

would enable the engineer to deten1ine 90del parameters for 1 more complete and accurate excitation syst111 representat1on for EDGS such as the IEEE type ST2 model.

It would also avo1d removal of the excitat1on system for bench tests wh1ch 1s usually unacceptable to plant operat1ons.

GOVERNOR MOQEL The governor model chosen bl the authors was also used successfully -in reference (4J. F1gure 8 su111111rizes the contro 1 syste111 and assoc1ated p1r11111ters used 1 n the studies.

The ga 1 ns and t 1* constants used in the governor model were adjusted until a reasonable match was obtained between the lllOde 1 and the resu 1 ts from

. field measurements of generator speed during 1110tor starting.

I

Cl*T381 s~

-, -. -T.-a""".-T-. T-S-2 I I

I z I

"----....)

~

Electrle Control Boa T MIN

~

Actlaa*

T1 *.01 I2 *.02 l3 *. 2 T4 *.25 Ts *.009 T6 *.0384 Flsnl.,,,.......

IC

40

!o *. 024

!max

1.1 T*in

0

,

lloeH C0111Puter s1*1at1ons of the ECCS mtOr' s~enc1ng on each EOG, as s~riztd 1n Figures 1 and 2, were coqileted and cQllplred to f1eld 911Suraents of EOG ten11nal voltage and changes 1n speed.

Figures 9 through 12 su-rize the generatOr' t81'9inal voltage response of EOG 1-1 while starting the HPSI, Service Water, LPSI, COllPOnent Coo 1 i ng

Water and Aux111 ary Feech11ter mtOr'S.

As can bt seen 1n tht f1gures, the model is qu1te close to field test... surtments during the amtor starting transients and is sa..hat conservative in te1'11S of initial voltage drop Ind recovery ti* above 1.0 per-unit operating voltage.

Table V suillarizes the mtor acceleration tims 911sured during the field tests and frcm tht c~uter si*lations. Reductions in several mtOr' inertias wre

- requ1red to obtain acceleration tims near those measured 1n the field.

However, the original motor manufacturer's inertia constants, llhich generally result Page 21 1n acctlerat1ng t1mts longer tha.n those from field tests and thus g1v1 coM1rvatfv1 results, are used 1n the f1na1 studies for each EOG when reiiresent 1 ng LOCA

cond1t1ons.

1~

1 i

r-

.Model\\

i

=

-~

- -"w

. r-... c

~...

. ~

o l~~_._-"""'._._ __ 1.._ __ *.*~*--'----l----...L_ __..J._~_~_:~*;_*.:__::

s.o TIMI (MC) 10.0 Flsn I. Ell 1-1 T.... al lllta1* Rnpn11 Dut11 IPSI ** ~ Stnlce Wam llur Start111

.___...___,,,...__..._ ___...._ _ _._ _ _,_ __ _._~_._-_.;........... o 12.0 TIMI (MC) 18.0 Flpn 11. Ell 1-1 T...... llltap ln,_11 Dlri11 LPSI llltlr Stlrtlll

L l

L ID

~---~...._---~_._~...._~i-........1..~...1...~0 Flgn 11.

TIMI (MO)

M.O EDI 1-1 Tn1aJ V11ta11 R11pt1n Dll'fi1

. CCW liter Start1' I.

r:--

~-------~-----------2 41.0 Fl&n 12.

Tl*<*

Ell 1-1 TM'llllll ltllap 111,1111 Ina AnDl8Y Fllftltlr lllW S...

Figures 13 through 15 sum11riz1 the speed fluctuations found in the actual field measu,...nts and through CQlllf1uter si.,lations during the starti!'V of the LPSI, COllPQnent Cooling Water, ancf Aux11iary Feedwater Pu~s on E06 1-1. As can be seen in these figures, the coaiputer mode 1 111tches th* field measuremnts reasonab 1 y we 11 and is s 1ight1 y conservat 1Y1.

The speed fluctuations are very 11ini1111 (less than 1 percent) incHcating the E06 governor essentially 111intains syst*

base frequency during the motor starting transients.

Figures 115 and 17 sumerizt th* generator te,..inal voltage response of E06 1-2.tine starting the HPSI, Service Water, and Auxiliary Feedwlter PUllPS. As can be seen in the Figures, the model is quite close to field test 1111surements during the motor starting transients Ind, si1111ar to ED6 1-1, is sOllll!lllhat conservative in tel'llS of initial voltage drop and recovery ti* above 1 _

per-unit aperating voltage.

Table V sumarizes the 1110tor acceleration ti*s 811Sured during the f1e1d tests versus the si.,1at1ons.

Page 22 t

L I

i

[_

I

!----*~ ------

.0 l

~.

1-

.2

- j -

r*~

z Ill

~

Ill a:...

~

0,.;

':'=-":-'-~....__._~...___._~...1..-_;_~...:.,____;.~e tZ.O nMI (MC) 19.0 FIP'I 13. Ell 1-1 Spm Fl1Ct11dH1 Dni11 LPSI liter Start1,

~ I r I

Fl&n 14. Ell 1-1 SllllM Fl1Ct11d.U Dst11 CCW llttlr StartQ

O

~

' C!...

I*

IO n.o

!O i2 I

1

~..

'----'-~"'--'-~-'-_._~.......___._~_.___.~e Fl&n 15. Ell 1-1 SllllM Fl1Ct11d*1 ln1 Aalllary Fllftater llttlr StartlJ 437

1 r~

I I f

---~-------~---------~.___---=

U T1.(...

1U Ftsn 11. HI 1-z r.-.. r11ta11111,1111 Ina IPSI IM Stnla l1t1r._ ~

I I r t

~ -

'~

""" 17. Ell 1-2,..............,,,........

lalllryf11ta1*---

TIMll I

'J If -

l111h ldll 1'11111..... 114 rlll a..I......

Page 23 DCIIAUQI mlQ UV91.a TM UC1tlt10ft 111t-llMd llr t"e ho E06S.,..

1dlflt1cal '" Y1ntlfl, 11Ze Ind Held sett1nqs.

i.o dllftlll 1\\1¥9 b-'.... 11..:e or1911111 l"sta11et*on.

F1tld t*t1,.,.....1e1t, llGl,__,_l lldl Sla.r gtfterttcr tt,..1na1 voltqe,..,..... Oft UM 1-1 tl\\ln ED; 1-Z llf'len st1rt1,. s1*11*r tnduet1on mtor 10ldt. T'll1s d1ffe,..nc*~

'" l'tSPOftH cu Des* DJ Ulm1n1"f F1"'"9 18 e1'd 19, lllfl1dt c.-re '1t1d..U~I flt t"'91nal voltage of lldt E1)Q 11Mn start1.. tN.00 Ito HPSI Ind 400 110 LPS!

~ors. No exo11nat1on could De found 11 to ""Y tilt rts.oonsn..,.. d1ffe,..ni.

F1t1d adJust91ftts of

~~*

voltage r19ul1tor en EOi 1-1 d1d. not 1111Drove

~~*

resoonse. 1h11 1s offered es a caution to '.lt~ers wf'o 11111 usu* tne rtsDO"H of tne uc1tat~o" s1stelll9 :,,

1de"t1ca1 EDGs 1s tne,... after test1"9 o"ly :ne ~:G.

Tht exc1tat1on syst* on ~

ECG ~st be

~esteci 6 '.lr resoonse to dete,..1nt 1U c11aracttrlst1cs for :ierfoM11.,9

~or start1"9 stud1n.

- L I

IDG 1*1 \\;

i

~, __ _._~..........;..~ 0 mmc-.

LO Ft... IL DI 1-1,._ DI 1-2 T....., Ylftlp

1111 lll'tlaftlllW Swtlll r-t*I f I'

\\

~

I I

~o..~..__.. __.....

~.._,..__,,_C __

....... __.___.~~~

.. o F1&n 1i.

1-1 r.. DI 1-2 T..... YllU11 II. UI lll'tJaLnl-S~

LIMITS IMP0$ED By.yPLT SYmMS In order to deten11 ne 11 ia1 ts 1 mposed by the 480 volt syste111, f1eld test1ng of ac contactors was perfonlled to deter111n1 p1ckup and d~t voltages.

Table VI SU111114r1zes the results of f1eld tests performed on all contactors fed by 480 volt Motor control Centers 1 and 2.

As can be seen 1n the Table* the h1gh~st ac contactor dropout voltage was 0.625 per un1t (480 volt base).

T1b11 VI Rtslftl If IC CNtactw 'Fiii~ Tnts

418 Vtlt l1tll' C11tnl Cnten 1 u~ 2 Cont actor M1n1mu11 P1ckup fl4ax1mum Dropout S1ze Volts Ccer un1t)

Volts (per un1t) 1

.725

.425 1

.692

.375 l

.692

.400 1

.692

.367 1

.692

.400 1

.692

.450 l

.725

.383 l

. 725

.367 l

.708

.383 l

.708

383 1

.650

.. 358 1

.717

.358 1

.700

358 2

.808

.567 2

.733

.625 2

.775

.567 2

.725

.525 3

.750

525 3

.675

.508 a

439 Based on 'e resu 1 ts, an ac contact or dropout voltage of 0.65 per un1t (for conservat1s*) was assumed as the 11*1t 1n the stUdies. F1gure 20 SUlllllr1zes the lowest voltage trans1ent seen by ~ volt motor control center 2 fed by EOG 1-2 during the start1ng of the HPSI and Serv1ce Water Motors. As can be seen 1n the f1gure, the m1n1mu11 voltage 1s 0.78 per un1t, wti1ch 1s well above the contactor dropout 11111t of 0.65 per un1t.

CONCLUSIONS Models of EDGS and the1r ECCS 1nduct1on lllOtor loads dur1ng sequenc1ng have been developed for an 800 megawatt nuclear plant. The models were developed us1ng d1g1ta 1 computer software and hardware. F 1e ld tests and correspond1ng s1mulat1ons using the models 1nd1cate a close correlat1on dur1ng motor sequenc1ng. Additional test1ng of ac contactors in 480 volt systems to es tab 11sh contact or p1ckup and dropout vo 1 tages have been presented to prov1de 11m1ts when perform1 ng EOG sequenc1ng stud1es.

F1eld tests have been presented wh1ch 1nd1cate.the response on two seem1ngll 1dent1cal EOG exc1tat1on systems can be d1fferent *. To avoid potential problems, separate field test1ng is reconnended when develop1ng models of mult1ple EDGS.

Add1t1onal work is needed 1n the area of excitation control system f1eld test1ng for systems emQloying both generator term1nal voltage and current feedback. Such test1ng methods could be used to develop more accurate control system models representat1ve of static exc1tat1on systems

REFERENCES (l] B.J. Chalmers, A.$.

Mulk1, *"Design Synthesis Doub 1 e-Cage Induct 1 on Motors,* eroc rEE, Vo 1. 117, No. 7, July 1970 *

(2] *s.s. Waters. R.D. W111oughb/* *Mode11ng Induction Motors for System Stud1es.

IEEE Transactions on

!ndystry App11cat11\\.

Vol.

IA-19, No.

s; eptember/Octo er, l (3]

IEEE Coan1ttee Report, *Excitat1on System Models for Power Syst111 Stab111ty Studies,*

li.E..E.

Trl"J8gt1ons i" ~c APRfijfYS And Systems, ~

PA -

, No.

ruary (4] L.N. Hannett, F.P. de Nellot G.H. Ty11nsk1, W.H.

Beckerz *va11dat1on of Nuc1ear Plant Aux111ary Power )Upply by Test,. IEEE va"iHWPS on Power

~oparotus f. System, Vo *

, No.

9,

eptember,

~

BIQGBAPtiIES I

i-----~------~-~-~----~-

fllft 20. 418 ftlt lltm* Vtltap fwsa AC Ctttacur Dn111n Vlltap I

I i

~en~et~ ~* f'tfia~r received the BS£E frOll '41ch1gan ec no o9 ca nversity 1n 1976.

He was employed by Goodyear Atoaic Corporat1on, Piketon, Oh1o fro111 1976 to 1978.

He then Joined ConsU111rs Power CQll'Clany and 1s currently responsible for the design and analys1s of nuclear and fossil po.ier plant aux111ary systetnS. lilr.

Yeager is a lllellber of IEEE and Registered Profess1onal Engineer in the.State of Michigan.

~obnny R. W1111s received the B.S.E. and M.S.E degrees roia the University of Alabw 1n Birm1nghu i!I 1972 and 1978. and E.E. degree frOll the Univers1ty of Michigan 1n 1985. He was 111Ployed by Rust Eng1neer1ng, Birm1ngham, A1. from 1974 to 1979 and Consu.. rs Power C~any from 1982 to 1987.

He then Jo1ned Power Technolog1es, Inc.

as a Sen1or Engineer wtiere he conducts studies related to PO'ller syst* dynam1cs.

Mr. W1111s 1s a member of IEEE and a Registered Ptofessional Engineer 1n Alabama and N~.York.

~t

440 Discussion W. G. Bloetbe. N. L Deeb and s~ S. Sbab (Sargent & Lundy, Chicago. Ill: The authors arc to be congratulated on their timely paper. While the NRC has not yet taken an official pcisition on the* subject. the question of the dynamic perfor-mance of the emergency diesel generators has been raised during some Electrical Distribution System Functional Inspec-tions ! EDSFl'sJ. Therefore. other stations may be required to pt:rform an analysis similar to that described in the paper.

We would appreciate the authors* comments on the following

- concerns which arise from our experience in performing such studies:

!. The authors state that their induction motor models are based on the manufacturer"s data. However. a recent paper 1 indicates that the performance of the motors in the field can be considerably different from that indicated by the manufacturer*s data. Could the authors comment on the significance of comparing the actual performance of the motors with the manufacturer's data used in the modeling?

The authors also indicate that the effective moment of inertia for the motors differ from the values iiven by the manufacturers. We agree with the authors* use of the more conservative values given by the manufacturers. Neverthe-less. we would be interested in any comments that the authors may have on what might cause the manufacturers' moments of inertia to vary from the values observed in the field tests.

2. The authors also stated that the induction machine param-

. eters were selected so that the model reasonably duplicates the motor torque. current. and power* factor during start-ing. However. the authors indicated previously that only the manufacturers' s*peed versus torque and speed versus current curves as weU as the motor and load moment of inertia were available for the study. The results of the motor modeling shown in Figures 4 and 5 for the LPSI and HPSI pump motors show the speed versus torque and speed versus current characteristics only. Our own experi-ence in induction motor modeling has shown the difficulty of matching all three characteristics of a motor. i.e., speed versus torque, speed versus current, and speed versus power

factor. The authors did not rcftect the speed versus power factor characteristics of their motor models in their paper.

We would appreciate the authors comments on the signifi*

cance of matching the motor speed versus power factor characteristics. lf speed versus power factor information is available, can the authors provide a comparison of the speed versus power factor characteristics of the LPSI and HPSI pump motor models with the actual motor character*

is tics?

3. Numerical techniques have been used to asmt in the matching of the motor _model parameters to the motor characteristics. Have the authors used any such techniques to reduce the amount of trial and error required to model an induction motor?

4. The authors state that the generator excitation system used both a potential and a current power supply. Could the authors provide additional information on the excitation systems? We have successfully used the IEEE type ST2 (former type 3) inodel for representing the excitation sys*

' Hassan. I. D.: Weronick. R.: Bucci. R. M.: and Busch. W. "Evaluating the Transient Performance of Standby Diesel-Generator Units by Simulation."

IEEE TfOIU4Cnon.s on EMr/lY ConuerJion. September. 1992. Vol. 7. No. 3. pp.

470-4n.

Page 25 e

tern of an emergency diesel generator with both potential and current power supplies. The effect of the current power supply is most important during the initial recovery from the voltage dip. This may explain why the initial recovery given by the authors' model tends to be slower than that shown in the test data. The addition of the current power supply to the model may allow the voltage regulator loop of the excitation system model to_ be revised to reduce or eliminate the oscillatory behavior during the latter part of the recovery described by the authors.

5. We agree with the authors that additional work is needed in modeling and determining the model paramete~ for emergency diesel generator excitation systems. In.iddiuon.

we feel that there is room for improving the modeling oi governor systems for the same reasons given by the au*

thors. Also. when we have had modeline: information from the governor manufacturer. the form o(the model did :iot match the standard models given by the IEEE commiuee.

6. Obtaining adequate test data in a power plant en\\'lronment is not a simple task. Could the authors describe lhrn testing program and the methods used to acquire the resr.

data? What techniques were used to minimize the <!'.'feces of noise during data collection?

"'fanuscript -received August 7. 1992.

.I. D. llaaau, R. WeroDict, &Del a. M. Bucci.

(Sbaaco **rvioe8 IDoorporatecl, *** Tort, ***

Yort):

Th*

author*

present a-dynamic

imulation which appear* to correlate with field teat*.

However, aom*

additional infol"llAtion would be helpful to us and, we believe, other intereated readers, to compare the author*'

approach with the approach described in (l].

Could the authors collllllent on the type o!

software that wa* uaed and cits accessibility to nuclear utilitie*?

Ha* the software been verified in accordance with nuclear quality aaauranc* atandarda?

In [l] it i* atated that th* EMTP proqram was uaed tor the die.. 1 9enerator aiaulationa, and the appropriate module* were verified !or nuclear applicationa.

Th* EMTP proqram.has wide availability and acc*Hibility to potential uaer*

(varioua veraion*

are available throuqb EPRI and !M'1'P uaer CJrOUps).

Alao, the user *can* llOd*l virtually any

lectroaechanical device repreaented in a trar\\*tu-function tor.

by u**

ot th*

appropriate !M'1'P llOdul***

We have tound that comaercially aarketed dynaaic proqraaa of the kind typically uaed tor tranaient atability analy*i* 9enerally have fixed control system modela.

Th*** require recoapilinq (which may require th* intervention of th* software developer) to develop cuatomized control

Y*t-arran9eaents that closely match the performance ot in service voltaqe requlators and en9ine qovernors.

Could the authors comaent on th* relative ease in applyinq their software to develop customized models?

Also, it would be helpful it the simulated EDGs' t*rainal voltaq* reapon** durinq the load application in Step 2 i* included in the

paper.

l l (l] I.e. lla**an, R. Weronick, R.K. Bucci and w.. luacb.

Evaluati"9 th*

Tran*i*nt Pertoraance ot Stand}?y Di***l-C.nerator Unit*

by Siaulation*, IEll Paper 92 WM 011-6 BC.

Manuacript received Auguat 14, 1992.

K.E. v.. ger, J.R. Wllll1: The authors appreciate tl'le efforts and com-mems of tl'le diScussers. We feel tl'lat by snaring experiences in tl'lis area.

all comnbutors can benefit. we lirst address tl'le pcintS raised by Messrs.

Bloetl'le. Deeb and Shall:

1) Anl'lougll tl'le manufacturers data has been used to devetop the motor models. experiences with motor starting s!Udies at other plants have indieated tl'lat manutadurer data is usually aca.irate in terms ot locked rotor currem and power factor. Actual acceleration times. l'lowever. are more diflia.ill to matcn. indicating inac:a.iraciaa in ( 1) 11'18 ma~acturers motor speed-torque charac:teristlcl. (2) motor anG'or.purnp inertias. or (3) the pump speed-10f1'Je cl'larac:teristica (flow 0011f9'rations cile to valving of piping systems during tesling).

It may ifT1)10ve motor modeling accuracy to record adUaJ motor performance when started lrom tl'le normal' power supply. This would provide data wl'liCh could be compared to motor starting simulatiOns of eacl'I individual motor and would guide adjustments of the motor model prior to using it in diesel S8(Jlenc:ing studies. Unfortunately. pertonnng additional indi'liQJal indudion motor starting tests and measurements 1n an operating nuclear plant can be a formidable task. In addition. sening up pump conditions wllich represent adual LOCA !low oonditions dunno motor starting may not be Possible.

The autl'lors chose to limit the testing configurations. measuremems.

and engineering analysis to sequencing conditiOns via the emergency diesel generators and used the results to adjust tl'le motor inertias to matcl'I motor acceleration times. It is believed tl'lat tne changes to the Plant sysrem due to variOus valving con191i-anons. etc.. affected the acceleration times of the motors cilring testing (the motors were started unloaoed or at some other pump condition). Adjustments in the inertias were made instead of pump speed-torque cl'laracteristica only as a maner of cl'loiCe. Use of worst case inertias and pu~ speed-torque conditions during LOCA sequercing studies assures col'ISefVatlve resulla.

2) Lodted rotor power factors were available from the mar&lfadlJrer tor a few of the motors. If not. a value of.25 was assumed. During the deve!opmert of the motor modell. adjuStmetts in the motor Slator

~ivalenl drt:uit parameters were made tor eacn motor until a relatively accurate matcn was obtained between tl'le given or assumed (.25) IOdcad rotor power factor. Once the IOdced rotor power lac:tOr was matcned. tl'le general shape ot the speed - pcwer factor curve was ~ndenl on the

~ivalent cira.iit of 11'18 motor and its respcnse during SlaJting. The autl'lors experience indlcat" that this tec:hn~e will provide an adequale model ot motor power factor tor motor starting studies.

3) The authors nave noted technical publicallonS tor W1inO ll.lmetical method to aid to matching itQic:lton mocor. ctwKteftlllcl. and haw contributed to this effort (A).

For ll'lil lllUdy, l"IOweWr, trial and error techniques were used.

l The autnora plan to irnpiove the exeitldon IYll9"' ITIDdel using an IEEE type ST2 (tanner type 3).

5) No comment.

6 Ou* to long data gatheMg II"" (~

1 ~*l oseillograpna nave provided a depelldllble IOUl'Ce ol aoquS1n; test cs.a during diesel S8CJ,lercing mt hmve beef'I used a a baCkup to digital recorcllng inStNrnenll. We nave had protlleml wlltl noile on several~

r~.

HoweYer. PT1 has bMrl succeutul UPlg ~*PC-based Dynamie Syt1en1 Monitor tor '9COttSlng ~

csara IOI' the main generator a1 power p1arn IOI' tater u* in ITQ3el de111181ton. Thia instrumeni snould a1so be IR)licallle to dtesel ~

.,,...,,.menll *well.

Next we addrea tile Poiltl railed by Meun. Hauan. Weronick ard Bac:ci.

The software used IOI' the study wa the PSSIE ptagram. wl1ie11 is commercially availatlle to ~ar utilities. FOi' severa1 studies. tile software with accompanying rrodel data l'lal been ventied with nuclear

~ality assurance procedures by co~n ot simulated reSPonse versus field rneasuremems. as was the case 1n 11'18 diScusser5* relerenced woni.

Wrth regard to developing customized dynamie models tor tile PSSiE program. tile software design r~ires the user to develOp computer coae wllieh can be linked imo the main program. While tl'lis is more 1n1101veo ano r~ires a higher level ol program tamilianty of tile user than mooe1 building from transfer !unctions. it iS also more llexible. Eacn metnod nas its strengths and weaknesses.

The simulated terminal voltage respcnse of the EDGs during Step 2 load application is shown in Figure 21.

(A) B.K. Johnson, J.R. Willia, "Tailoring lnciletion Motor Analy1ical M0<1e1s to Fil Known Motor Pertonnanoe Cl'laradenstica ard Satisfy Panicu1ar Study NeedS.' IEEE Transactions on Power Sysrems. Vol. 6. No. 3.

~st 1991, pp. 959-965.

!_ EDG1*2 0 0 l.-.....L~....L...~.l.---"~-1-~"--_,L~_.,_~..___.

0 TIME (MC) 5 F"igure 21. Step 2 Dietel Genetator Voltages.

Manuacript received September 22, 1992.

~I

0 ATTACHMENT 5 Consumers Power Company Palisades Plant Docket 50-255 QUALITATIVE EVALUATION-SAFETY SI~NIFICANCE OF REDUCED PUMP FLOW

.NOVEMBER 22,

1994 4 Pages

v I

I

,\\

1'>1*

()

~*

QUALITATIVE EVALUATION. ~

SAFETY SIGNIFICANCE OF REDUCED PUMP FLOW A peak D/G load of 2573 KW is automatically sequenced onto the D/G following a.

DBA LOCA. An evaluation by KEYeager has shown that D/G 1-1 would not have been able to support the operation of these sequenced equipment at normal motor speeds of 60 hertz between July 1994 and September 1994.

Had this automatic

sequencing been required during the time in question it would have resulted in a 4% reduction in pump and fan speeds for the D/G loads over the first 30 minutes of the event.

This speed reduction would have increased to 9% between 30 and 32 minutes when b/G loading is increased to 2688KW due to the addition of several manual loads.

After 32 minutes the D/G loading is reduced and normal pump operation would have been restored for. the duration of the event.

Based on the pump affinity law, a pumps' flow will vary one for one with its' speed.

Therefore ~ny percentage reduction in pump speed will equate to a similar reduction in flow.

A fan will behave in~ similar manner.

The DBA sequenced loads that would have operated with a reduction in flow had EDG 1-1 been required for a DBA event prior.to August 1994 are :

P-548 & P-54C P-55c*

P-568.

P-668 V-4A..

P-7B

P-67B P-52A & P-52C P-8A.

V-95...

CONTAINMENT SPRAY PUMPS CHARGING PUMP BORIC ACID PUMP HPS I PUMP CONTAINMENT AIR COOLER FAN SERVICE WATER PUMP LPS I PUMP COMPONENT COOLING WATER PUMPS AUX FEEDWATER PUMP CONTROL ROOM VENTILATION FAN The Charging and Boric Acid pumps do not provide any mitigating effects in the first 32 minutes of the DBA LOCA event.

The containment air cooler associated with this fan is not required to operate for any safety related purpose and cooling w~ter is automatically isolated from the cooler on an SIS.

The control room ventilation fan maintains a positive pressure in the control room envelope and following an accident helps reduce.iodine isotopes entering the C.R. by drawing entering air through charcoal filters. A reduction in the speed of V-95 would have a positive effect on its required function following ahd accident.

One of the main sources of iodine to the control room is unfiltered air inleakage through the isolation dampers.

Reducing the fan speed will reduce the pressure drop across these dampers thus reducing the amount of unfiltered air that enters the control room.

The limiting Chapter 14 events that would have been impacted by the reduced equipment speeds are the LOCA fuel analysis, and the LOCA containment analysis. These are the only events that would require the D/G to support the maximum expected DBA loads, and they have the least amount of margin to peak clad temperature, containment pressure, and containment temperature safety limits~ Equipment operating during other Chapter 14 events, where the* load may have exceeded 2438 KW, would have experienced similar but less severe reductions in speed.

Since all other events have less restrictive load profiles ~nd more margin av~ilable to safety limits, the consequences of Page 1

.o

.-)

reduced pump and fan.eeds will be bounded by the evlation of their impact on the LBLOCA. analyses.

The following discussion summarizes the results of the present analyses and the safety limits that need to be protected.

It also provides a discussion of the conservatisms in the analyses that would serve to mitigate any detrimental effects of reduced flows, and a qualitative evaluation of their impact.

The 9% flow reduction between 30 and 32 minutes will have little or no impact on this evaluation due to its short duration and the fact that peak temperatures and pressures occur much earlier in the event.

LBLOCA FUEL ANALYSIS EMF-91-177 Supplement 1, calculated a Peak C)ad Temperattire (PCT) of 2095 °F.

This PCT occured 63.53 seconds after the initiation of the LBLOCA.

The Standard Review Plan (SRP) safety limit is 2200 °F.

There are several conservatisms associated with this analysis that could have mitigated any increase in PCT.

PCT is driven by three major factors, the blowdown rate of the PCS, the magnitude and duration of Safety Injection Tank (SIT) flow rate, and the magnitude and timing of LPSI pump flow.

The PCS blowdown rate has the greatest impact on PCT, therefore the LBLOCA Fuel analysis minimizes containment pressure to maximize the blowdown of the primary system.

The analysis accomplishes this by assuming that all containment heat removal equipment is operating at time zero of the accident.

The reduced speed scenario being evaluated here is only a credible event when EOG 1-2 is not available, which would reduce the available containment heat removal equipment for the reduced speed scenario by more than 50% from what was assumed in the LBLOCA analysis. Additionally it would cause a slight reduction in the capability of the equipment that is available. This would result in a PCT lower than the 2095 value calculated in EMF-91-177.

The PCT analysis did conservatively use a LPSI & HPSI pump allignment representative of the failure of a D/G, so the injection pump flow paths used in the ana1ysis are consistent with the expected available flow paths associated with the reduced speed scenario being evaluated.

The predicted PCT occurs at 63~53 seconds, just prior to the SITs emptying at 63.7 seconds.

A reduction in LPSI and HPSI flows will not be significant since the flow from the SITs' is more than adequate to keep the downcomer full

and PCT is reached prior to the end of SIT flow.

After PCT occurs the LPSI pu~ps will be pumping against a depressurized system and only need to remove decay heat.

Special test T-339 verified that the measured LPSI flow capability was significantly higher than that assumed in the analysis, especially at lower system pressures.

There is also conservatism built into the other pump flows used in the analysis (HPSI, AFW, and Cont. spray). These flows were all calculated using a degraded pump head flow curve.

The degradation was based on pump performance at the required action range for the quarterly ISi tech spec surveilence test~ The analysis flow rates also accounted for the maximum allowable instrument uncertainty in the measurement loop used for the ISi test. Additionally the HPSI and AFW pump flows have very little impact on the results of the analysis, and SW and CCW pumps are not included in the analysis for this event.

A 4% reduction in flow applied to the LPSI pump performance used in the present analysis may have caused the PCT to exceed 2095 °F by delaying the time to PCT beyond the SIT injection phase; however, the PCT would not have approached the SRP linit of 2200°F.

This assesment is support~d by an evaluation of the impact of reduced LPSI flow performed by our fuel vendor in Page 2

SI

** )~~

ll*

1986.

The evaluatio.as done for the analysis of re.d a:t the time and concluded that a 30% reduction in LPSI flow only increased the PCT by 10 degrees.

This general trend can be applied to the present LBLOCA analysis and further supports the argument that a 4% decrease in flow would not cause a considerable increase in PCT, and would not eliminate the entire 105 degrees of margin between the calculated value of 2095 and the limit of 2200.

Considering the discussion above and the inherent conservatisms in the Siemens Power Corporation LBLOCA methodology the safety significance of reduced flows for the LBLOCA fuel analysis would have been minimal.

LOCA CONTAINMENT The calculated peak pressure for the LOCA containment analysis is 53.46 psig, the SRP limit is 55 psig. The rapid blowdown of the PCS for this event results in mass and energy releases to containment ending just prior to the calculated containment building peak pressure at 13.5 seconds.

It is the termination of the energy input along with the heat removal capability of the passive heat sinks that maintain the peak pressure below its design limit.. Since this is prior to any loads being sequenced on the D/G a reduction in pump or fan speed will not have any impact on the calculated peak pressure.

The SRP also requires the pressure at 24 h~s to be reduced to 27.5 psig,(les~

than half the design pressure). The Analysis calculates the pressure at 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months to be 12.9 psig.

The event will be under control by the time 24 hrs is reached.

The pumps will have been returned to normal speed for over 23 hours2.662037e-4 days 0.00639 hours 3.80291e-5 weeks 8.7515e-6 months .

The reduction in speeds for the first 32 minutes would result in slightly higher Containment atmosphere temperatures than predicted by the analysis. The higher temperatures will tend to cause the heat removal equipment to operate more efficiently, and by 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months should have brought the containment temperature and pressure back to the values predicted by the analysis.

The temperature profile al~o must stay below the allowable EEQ env~lope determined in A-PAL-93-074. As di~cussed above a 4% reduction in the Containment Cooling flow rates assumed in the present LOCA analysis would cause the temperature profile to increase slightly in the first 32 minutes.

This increase is expected to be less than 5°F, which is the minimum margin between the EEQ profile and the calculated profile. This minimum margin occurs between 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months and 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

As discussed above the temperature pr6file for a reduced speed scenario will begin to migrate back toward the predicted analysis profile after the pump flbws are returned to normal at 32 minutes.

Even if *the atmosphere did exceed the EEQ envelope by a few degrees it would not result in any equipment failure.

There are several conservatisms in the analysis that would.very likely keep the temperature profile below the EEQ limit.

The pump flows used in the analysis (LPSI, HPSI, SW, CCW, Cont. spray) have the same conservatisms as discussed in the LBLOCA fuel evaluation above.

The water source for the containment sprays is an outdoor tank vented to the ~tmosphere with a normal.

. capacity of 250,000 gallons of water.

This tank was conservatively assumed to be at an equilibrium temperature of 100 F.

The Service Water system is the ultimate heat sink at Palisades.

It takes suction from Lake Michigan.

The Service Water was assumed to be a constant 85 F throughout the analysis. Historical data has shown that the lake only exceeds 80 degrees periodically and for short durations, and it has not exceeded 82 degrees over the reviewed operating history (1980 to present).

Based on past sensitvity calculations on the impact of source temperature it is judged that the reduction in flow would be offset by the actual source Page 3

, L:"' ~.temperatures that WO. have actually been present. clidering the discussion

~

above the s~fety significance for the LOCA contain~ent event would have been minimal.

P.-:~e 4

05000255/LER-1994-017, :on 940902,EDG 1-1 Degraded Load Carrying Capability Identified.Caused by Inadequate Design Basis, Failure to Effectively Monitor for Engine Performance Degradation & Failure to Control Mods

Contents

text

CAUSE

CAUSE

Event Description

Analysis Of The Event

Safety Significance

Corrective Action

Description

References:

Navigation menu

05000255/LER-1994-017, :on 940902,EDG 1-1 Degraded Load Carrying Capability Identified.Caused by Inadequate Design Basis, Failure to Effectively Monitor for Engine Performance Degradation & Failure to Control Mods

text

CAUSE

CAUSE

Event Description

Analysis Of The Event

Safety Significance

Corrective Action

Description

References:

Navigation menu

Search