ML20052F811

Discusses Acceptability of Results in Tests Performed on Capability of Diesel Generator to Start & Run Motor Driven Auxiliary Feedwater Pump.Safety Evaluation Encl

ML20052F811
Person / Time
Site:	Trojan File:Portland General Electric icon.png
Issue date:	05/05/1982
From:	Clark R Office of Nuclear Reactor Regulation
To:	Withers B PORTLAND GENERAL ELECTRIC CO.
References
TAC-46552, NUDOCS 8205130676
Download: ML20052F811 (7)

Text

r.p-f f.

~

DISTRIBUTION MAY 5 BSE l 06'cket File Local PDR ORB Rdg tot o

., d

~'Wyp' D.Eisenhut

, JHeltemes Docket No. 50-344 e

'n RAClark c/.:

1/ A 9 Kreutzer

-i

,AELD f

' 5, l.'hSIC Mr. Bart D. Withers

~i 9

/.; I&E (1)

%Q ACRS (10)

'i y

Vice President Nuclear Portland General Electric Company

') b - a. g CTrammell is ! I c Gray File 121 S.W. Salmon Street hKjg t

]

Portland, Oregon 97204 JWermiel

SUBJECT:

TESTING OF MOTOR-DRIVEN AUXILIARY FEEDWATER PUMP-TRGilAN NUCLEAR PLANT.

Our letter to you of April 16, 1981 requested that you conduct a test of the capability of adiesel-generator to start and run the new motor-driven auxiliary feedwater pump (AFP).

You responded to this request in letters dated May 15 and' November 12, 1981, and January 15 and April 30, 1982. Based on the information you provided, including manufacturer's test results, and our review of the matter, we have concluded that the Trojan Nuclear Plant diesel generators have already been tested to a capacity equivalent to that of the new AFP load and that further in-situ testing to assure adequate capability of the diesel generator to start the new AFP is, therefore, not required.

The basis for this conclusion is contained in the enclosed, Safety Evaluation.

Sincerely,

' 'i 09: v; nod bv Robcn A. Clark '

?

.A Robort A. Clark, Chief Operating Reactors Branch #3 J-Division of Licensing

Enclosure:

Safety Evaluation cc w/ enclosure:

See next page 8205130676 820505 PDR ADOCK 05000344 P

PDR a/

orrice )

. 3...

gTrhent.dd dD $ntA.,,,,,

Rgg3

~

ee-,

Slb!82 ynga,,

om>

sac rosu m oo* Nacu mo OFFICIAL RECORD COPY uso m um m..

~

~

i...

_; T,.

~

Portland General Electric Company cc: Multnomah County Library Social Science and Science Department 801 S.W. 10th Avenue Portland, Oregon 97205 Michael Malmros, Resident Inspector U. S. Nuclear Regulatory Commission Trojan Nuclear Plant 8

P. O. Box 0 Rainier, Oregon 97048 Robert M. Hunt, Chairman Board of County Commissioners Columbia County St. Helens, Oregon 97501 Donald W. Godard, Supervisor Siting and Regulation Oregon Departnent of Energy Labor and Industries Building Room 111 Salem, Oregon 97310 I

Regional Administrator Nuclear Regulatory Commiss, ion, Region V Office of Executive Director for Operations 1450 Maria Lane, Suite 210 Walnut Creek, California 94596 I

e o

4

-~

~^

~

ENCLOSURE SAFETY EVALU TION TROJAN NUCLEAR POWER PLANT DOCKET NO. 50-344 TESTING OF NEW MOTOR-DRIVEN AFW PUMP ON DIESEL GENERATOR PGE has modified the auxiliary feedwater system for Trojan plant by adding a new motor-driven auxiliary feedwater pump. These modifications potentially I

place additional loading on the diesel generators' by allowing a new non-Class lE motor-drivdn auxiliary feedwater pump to be powered from either emergency diesel generator. This pump is connected to a non-Class lE bus and would only receive power from the diesel generators under conditions when i

no other auxiliary feedwater pumo is available and there has been a loss of all other power sources to the affected non-Class 1E bus. To start the motor-driven auxiliary feedwater pump (1250 HP), both the non-Class lE bus and the diesel generator bus would have to be manually unloaded and dedicated to this pump in order to accommodate the starting transient inrush current.

.1 After starting the auxiliary feedwater pump, ESF loads may then be manually restarted as diesel generator capacity permits (the diesel generator has a continuous rating of 4418 kw and a 30 minutes rating of 5003 kw).

. In a letter dated April 16, 1981, we requested that PGE conduct a test to demonstrate the capability to start and run the motor-driven auxiliary n

feedwater pump on an emergency diesel generator and to verify that this 4

mode of operation would not degrade the Class lE power system below an acceptable level.

In response, PGE stated that since it is theoretically i

possible (PGE has analysed the diesel generator capabilities and determined g

that it can start and carry a load as large as the new AcW pump) to l

start the motor driven AFV pump on an unloaded emergency diesel generator, this is a step load in excess of the original design requirements N

of the diesel generator units. The design specification for the diesel generator units required the ability to start a 900 hp motor as the. largest step demand, whereas the driver on the motor driven auxiliary 'feedwater

~

pump is a 1250 hp motor. Hence, starting current of the motor-driven AFW-

. pump on an emergency diesel generator would approach the 3D minutes rating of the diesel generator. PGE does not consider this good er.gineering practice to operate the diesel generatorin this manner without a dire need to do so.

Subsequently, upon further discussions with PGE, we concluded that the theoretical analysis to start the motor-driven AFW pump on an unloaded diesel generator in lieu of an actual test would be acceptable provided PGE could showthat(1)theTrojandieselgeneratorshavebeentestedtoacapacity 4

1 equivalent to that of AFW pump load during diesel generator qualification k

testing, (2) the loading of AFW pump will not cause spurious operation of the a

Class IE bus undervoltage relays and load shedding devices, (3) this loading would not actuate overcurrent devices provided for AFW pump motor / feeder and (4) operating procedures and restrictions will be implemented to ensurr that AFW pump is started as a last restort and both the non-Class lE bus and the diesel generator bus are unloaded and dedicated to this pump.

By letters of January 15, 1982 and April 30, 1982, PGE submitted the above information for our review. Our evaluation of this information is as follows:

'i Ji The test reports for the manufacturer's testing conducted on Trojan diesel generators included test data extracted from the manufacturer's Quality Assurance Documentation and Test Manual for Trojan's emergency diesel 1

2 generators. This test data is separated into two sections:

(1) a continuous

- load test and (2) a transient load test. The largest continuous load test consisted of loading the diesel generator to 4920 kw at a 0.8 power factor up to a maximum of 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />. The transient load test consisted of sudden application of 4920 kw to the diesel generator at 0.8 power factor. Ti$e maximum voltage dip due to this transient loading was 22%. PGE states that this test is more severe than starting of the AFW pump since the mechanical load change on the shaft of the diesel generator was equivalent to 6595 hp, whereas the actual kilowatt load on the diesel generator during starting of AFW pump would be approximately 933 kw or 1250 hp. We concur with PGE with this conclusion and find this to be acceptable.

The maximum calculated voltage dip due to starting the motor-driven AFW pump is less than 24%, with the voltage significantly recovering within 0.5 seconds.

I

'This voltage dip would not affect the undervoltage relays since,they operate when voltage falls to 60% of full value after a one second time dalay. The

,{

~

degraded grid voltage relays would not be affected since they are set at 92.5% of full voltage with a 55 second time delay. This satisfies our 1

above requirements 1 and 2 and is, therefore, acceptable.

aj The following discussion described why the start of the AFW pump motor would not actuate the overcurrent devices provided for the AFW pump motor / feeder protection. The instantaneous overcurrent trip setpoint of the overcurrent device is set at 2000 amps which is more than double the motor locked rotor 1

current of 862 amps.

Comparing inverse time characteristic curve for this 1j breaker and the locked-rotor amps of 862 amps, a long-time overcurrent trip would occur in 5.4 seconds at 100% voltage and in 7 seconds at 75% voltage, however, recovery time for the overcurrent condition due to starting the

s j

}

. 'N y

I AFW oump is less thcn two seconds and so will not result in an overcurrent x

trip.

This satisfies our requirement #3 ar.d is, therefore, acceptable.

~

s PGE will develop procedures for starting of the AFW pump to ensure that' the loads connected to the ncn-Class 1E bus as well as the diesel generator bus are first disconnected and then starting of the AFW pump in this fashion'is resorted to only in the extreme unlikely event when all other methods of feeding the steam generator are not available.

Implementation of these s

procedures satisfies our above requirement #4 and hence we find this to be acceptable.

The integrity of the circuit including the breakers from the diesel generator i

to the non-1E bus feeding the AFW pump is periodically tested every'18 months as part of the testing to assure the ability to supply power to the 4

l pressurizer heaters from the diesel generator bus.

It is our understanding i

that the pump is used on plant startup to provide low end feedwater flow during which time the pump and its supply breaker are also exercised. This provides an overlapping test wherein all the circuit components are periodically tested and therefore provides some assurance of system 3

operability.

i Based on our review of the infonnation provided by PGE in response to our request as discussed above, we conclude that the Trojan Nuclear Plant diesel j

generators have previously been tested to a capacity equivalent-to.that the.

l AFW pump load and, therefore, further in situ testing to assure adequate

.i capacity of the diesel generators to handle the AFW pump load is not required,

~

j We further conclude that the starting of the AFW pump with the above

4

- ~-

+

\\

\\'

s i

s g

l Q

A Y g.

~

~

s,,

') -

1 5-(

4 N.

e stipu?,ated restrictions will not degrade the Class lE system below an -

acceptable, level and we find this design to be acceptable.

i e -,

' ' Principal Contributor:

0.'Chopra w

c r.

g g

s s

f

~,

s q

~ L J s

-c s

,4

,, y.

t-9

'g k

4 e

4 t

\\.

(

4 f

1 l

\\

'\\'

j t -

f i

)

f.

n

%W N ',.

i a

1 4

~

m

(

f n

'/

s e

~.

p g

ke 4

4 I

1 g

T

\\

i.

k

\\

t 7

i t