ML20137Z517

Forwards RAI Re Plants Response to NRR Draft Rept on Plant Emergency Power Sys

ML20137Z517
Person / Time
Site:	Oconee
Issue date:	04/22/1997
From:	Labarge D NRC (Affiliation Not Assigned)
To:	Hampton J DUKE POWER CO.
References
TAC-M93550, NUDOCS 9704240212
Download: ML20137Z517 (8)

Text

. ._ ._. __ _ __

p < r. .

,y +

" April 22,.1997

.4r. J. W. Hampton Vice President, Oconee Site:

Duke Power Company

)

P. O. Box 1439 ' '

-Seneca, SC 29679

SUBJECT:

REQUEST FOR ADDITIONAL INFORMATION RELATED TO DUKE POWER COMPANY'S RESPONSE TO THE NRR DRAFT REPORT ON THE OCONEE EMER6ENCY POWER SYSTEM (TAC NO. M93550)

Dear Mr. Hampton:

By letter dited October 31,19%, you responded to the NRR draft report

)

on the Oconee Emerg:ncy Power system. As a result of our review of this information, we have determined that additional information and clarificatier are needed to close out open issues and complete our report on the Oconee/Keowee emergency electrical distribution system. We request that you respond to the attached questions within 30 days.

Sincerely, ORIGINAL SIGNED BY:

David E. LaBarge, Senior Project Manager  !

Project Directorate 11-2

. Division of Reactor Projects - I/II Office of Nuclear Reactor Regulation Docket Nos. 50-269, 50-270, and 50-287 i

Enclosure:

As stated l cc w/ enc 1: See next page DISTRIBUTION w/ attachment:

Docket File- J. Johnson, RII PUBLIC

R...Crlenjak, RII PD 11-2 Rdg. ACRS l S. Varga ' -

)

J. Zwolinski '

I r

D. LaBarge fMh N N f F. Ashe D. Nguyen Qy;O \ , g J. Lazevnick 40019 OGC To receive a copy of..this document, indicate in the box: "C" = Copy without attachment /enclosu(e- "E" = Copf with at,tachment/ enclosure "N" = No copy 0FFICE PM:POII-2 /#f] LA:PDIb2jp DfFhiV2 l 1 NAME DLABARGE:en LBliRRY N 4 HK40k DATE 4 /2 2. /97 (J/ NV /97 4 / s t/97 / /97 / /97 / /97 unurgau c.gur neanucn omcFICI AL RECORD COPY PDR ADOCK 0000269 P PDR

e # MZy p **,

UNITED STATES g

j 2

NUCLEAR REEULATORY COMMISSION WASHINGTON, D.C. 30666 4 001

% ,,, April 22,1997 Mr. J. W. Hampton Vice President, Oconec Site Duke Power Company 1 P. O. Box 1439 Seneca, SC 29679 l

SUBJECT:

REQUEST FOR ADDITIONAL INFORMATION RELATED TO DUKE POWER COMPANY'S RESPONSE TO THE NRR DRAFT REPORl ON THE OCONEE EMERGENCY POWER SYSTEM (TAC NO. M93550)

Dear Mr. Hampton:

By letter dated October 31, 1996, yn9 responded to the NRR draft report on the Oconee Emergency Power Syste.9. As a result of our review of this information, we have determined the additional information and clarification are needed to close out open issues and complete our report on the i Oconee/Keowee emergene.y electrical distribution system. We request that you respond to the attached questions within 30 days.

Sincerely, avid E. LaBarge, Senior Project Manager Project Directorate II-2 Division of Reactor Projects - I/II Office of Nuclear Reactor Regulation

~

Docket Nos. 50-269, 50-270, and 50-287

Enclosure:

As stated l cc w/ encl: See next page c

._. ___ . __ _ _ ._ _ _ _ . _ .~._ _ _ _ _ _ _ _ _ ____ _ ._. . _ . . _ _

( ,

i

! I Oconee Nuclear Station I Units 1, 2, and 3 l _

5 cc:

! Mr. Paul R. Newt 9n Mr. J. E. Burchfield i

Legal Department (PB05E) . Compliance Manager Duke Power Company Duke Power Company l 422 South Church Street Oconee Nuclear Sitr.

Charlotte, North Carolina 28242-0001 P. O. Box 1439

Seneca, South Carolina 29679 I

! J. Michael McGarry, III, Esquire

! Winston and Strawn Ms. Karen E. Long

1400 L Street, NW. Assistant Attorney General

! Washington, DC 20005 North Carolina Department of  !

i Justice l Mr. Robert B. Borsus P. O. Box 629 l j Framatome Technologies Raleigh, North Carolina 27602 i Suite 525  ;

i 1700 Rockville Pike Mr. G. A. Copp i i Rockville, Maryland 20852-1631 Licensing - EC050 l

. Duke Power Company I

Manager, LIS 526 South Church Street

NUS Corporation Charlotte, North Carolina 28242-0001 l l 2650 McCormick Drive, 3rd Floor l Clearwater, Florida 34619-1035

.l Senior Resident Inspector i U. S. Nuclear Regulatory Commission

!. Route 2, Box 610 l Seneca, South Carolina 29678  ;

i Regional Administrator, Regich II 4

U. S. Nuclear Regulatory Co'aission

101 Marietta Street, NW. FJite 2900 l

Atlanta, Georgir. 30323 '

i

! Max Batavia, Chief

, Bureau of Radiological Health l South Carolina Department of Health

and Environmental Control l 2600 Bull Street j Columbia, South Carolina 29201 l County Supervisor of Oconee County

Walhalla, South Carolina 29621 i Mr. Dayne H. Brown, Director i Division of Radiation Protection i North Carolina Department of j Environment, Health and 4 Natural Resources I P. O. Box 27687

i. Raleigh, North Carolina 27611-7687 i

- i i

REQUEST FOR ADDITIONAL INFORMATION I

DUKE POWER COMPANY i  !

OCONEE NUCLEAR STATION ELECTRICAL DISTRIBUTION SYSTEM

1. The response to open issue No. 7 states that operating Oconee unit shutdoc loads (approximately 2 MW) are block loaded on the Lee combustion turbines, and additional loads are then manually started until approximately 5 MW is obtained. The shutdown loads, therefore, are ap)arently deenergized prior to initiation of the test in order to suasequently block load them on the Lee combustion turbine. This conflicts with the response to question A8 in Duke Power Company's  ;

(DPC's) responses to staff questions dated January 31, 1996. That response states that during this test, the required startup equipment for the Oconee unit is not lost since the loads are transferred to the Lee

CTG without a loss of power. Please clarify which of these responses is accurate.

If the shutdown loads are indeed briefly deenergized during the Lee (a) test, then readdress the staff's original question relative to your

,. January 31, 1996, response. The question was: "If this degree of loading can be obtained on the Lee combustion turbines during l startup, why can't the same test be performed on the Keowee units?"

2. The response provided to open issue No. 9 appears to contradict the information in calculation KC-UNIT 1-2-0098 (Keowee Governor fiechanical Single Failure Analysis) dated September 29,. 1993, and calculation KC UNIT l-2-0106, Rev. 1 (Keowee Power Operating festrictions for NSM-52966) dated May 4, 1995. The information in the referenced calculations indicates that, when the partial shutdown solenoid is deenergized, the )

wicket gates are limited to 25 percent open, and when the solenoid is

energized there is no limit on the wicket gates. The calculations also indicate that on an emergency start from standby there is initially no limit on gate position, then Keowee speed switch 14/1 operates at 52 rpm to limit the gate position to 25 percent, and speed switch 13/1 operates

, at 122 rpm to remove the gate limit and return control to the governor.

The response to open issue No. 9, however, indicates that on an emergency start the partial shutdown solenoid energizes initially to allow the gates to open to 50 percent and the gate limit is set at 50 percent.

Please clarify.

(a) It also appears from this response that the monthly test performed to meet technical specifications is actually a modified normal start, since the auto synchronizer is turned off. Are there any other differences between that start and the normal start?

(b) Is the voltage and frequency the only acceptance criteria specified for the monthly test? What is the specified voltage and frequency acceptance criteria? Is the start time to specified voltage and frequency monitored, and what is its acceptance criteria?

L -

i' i

j l l

) (c) What are i.he reasons and difficulties associated with performing an  ;

! emergency start test on a monthly basis instead of a modified normal l

', start? The modified normal start does not test emergency start gate  !

limit operation, immediate closure of the field circuit breaker, or  !

emergency start relay contacts. '

(d; The response to open issue No. 9 indicates that the combination of i existing and proposed technical specifications results in three emergency start tests from a dead stop, and four from a running Keowee unit every 2 years. Are these numbers consistent with those j J

provided in the response to open issue No. 447 That response  ;

indicates that the Keowee units are emergency started (both star.

1 and hot started) three times per 18 months and a total of 13 times

! over a 3-year period.

' (e) The response to issue No. 9 also indicates that under current technical specifications there is an 18-month emergency power switching logic (EPSL) functional test that is performed with a '

" start from dead stop on an ONS unit." What is the relationship of this test to the EPSL functional test provided to the staff in the January 31, 1996, DPC letter? The test provided to the staff was not a Keowee start from dead stop but rather an emergency start from <

generating to the grid. How many EPSL functional tests are from a '

dead stop and how many from generating to the grid? Are any of the

, Keowee starts from dead stop performed as a black start?  !

3. Are any of the Oconee Technical Specification electrical tests referenced l in the most recent October 31, 1996, DPC letter going to be eliminated in the Technical Specification 3.7 rewrite? If so, which ones and why?

i

4. In the response to open issue No. 28 it is indicated that analysis conducted to evaluate the past operability of the electrical system concluded that all required safety functions would have been accomplished  ;

with the voltage at 11.9 kV. In the same response it is stated: 1 Due to different system loading, it is expected that the voltages required at the terminals of the Keowee generator to assure proper operation of safety loads will vary, depending on which failure scenario is being analyzed. The minimum generator voltage that bounds all cases and scenarios for the Keowee analysis is 13.5 kV.

It is not clear what the differences are that would lead to the conclusion that "all . required safety functions would have been accomplished" with the Keowee generator at 11.9 kV, and also conclude that "[t]he minimum generator voltage that bounds all cases and scenarios for the Keowee analysis is 13.5 kV." Please clarify. Also, please list l the full complement of failure scenarios that are referred to in the above quotation.

i 1-

1

5. It is not clear in the response to open issue No. 29 what DPC's final

position is. The response states that procedure PT/0/A/0620/09, Keowee

[ Hydro Operation, does not put the Keowee unit through a load run; but it j doesn't indicate why it is not a load run. It references a one-time load i run test that was run in August 1996, and states that the test data l demonstrates that monthly load runs of Keowee are not necessary in light j of the fact that Keowee is frequently operated to the grid. However, the 4

final paragraph in the response states that normal operation of Keowee i verifies the first two surveillance requirements (synchronization and i load acceptance) in standard technical specifications, and Duke plans to

. trend data monthly on Keowee operation to the grid in order to verify  :

1 proper heat exchanger operation. Please indicate why PT/0/A/0620/09 does i

not meet the criteria of a load run, and what combination of existing

Keowee operation and additional verification will be used to demonstrate j the load run criteria is met by the Keowee hydro units.

6. The response to open issue No. 30 indicates that the LOCA signal is verified in step 12.35 of the EPSL functional test that was provided to

the staff in a January 31, 1996, letter from DPC. It further states

that subsequent steps 12.36, 12.37, 12.38, and 12.39 demonstrate that

! this LOCA signal is providing the emergency start to the Keowee units.

l In fact, it appears from the procedure that at step 12.35 both Keowee l units are already operating. Please verify that the LOCA signal of step 4 12.35 does not actually start the Keowee units but rather is used to

verify logic actuation; contact closure, etc., necessary for the

emergency start of the Keowee units. If this is not accurate, i specifically explain what steps 12.35 through 12.39 are verifying.

a l 7. The item 8 comment in attachment 2 notes that a single breaker failure

! will not cause the lockout of both Keowee units during periods of dual Keowee. unit grid generation and a simultaneous ground fault. It states that both ACB 1 and 2 would need to fail in order to lockout both Keowee units in the postulated scenario. These statements are made with regard to a statement made in the NRR report that "[a] subsequent single failure of a safety-related breaker to clear a fault on the overhead emergency power path could potentially cause the lockout of both Keowee units if '

they were generating to the grid."  !

The circuit breaker that the staff was referring to in the NRR report is I not ACB 1 or 2, but rather an OCB in the switchyard. The failure i postulated was a failure of a Switchyard OCB to isolate a ground fault in I the switchyard or on a transmission line, outside the Keowee differential zone of protection. Such an uncleared ground fault would be seen by the ground fault protection scheme (59G relay) of both Keowee hydro units when they were generating to the grid and cause both units to trip.

Please respond whether this scenario is accurate.

W - ,.w - .---e w --n-- e e w ,- w - ~-

1

?

! 8. The item 13 comment in attachment 2 indicates that the diesel generator j hot restart test is not applicable for the Keowee units because the concern associated with the diesel generator's ability to start at high

diesel temperatures does not exist at Keowee.

We recognize that the Keowee hydro units do not necessarily have the same vulnerabilities as diesel units, but are there other vulnerabilities that

'. might be peculiar to the hydros relative to hot starting? We note that 3

in the response to open issue No. 30, DPC has indicated that a hot

] restart test was performed as part of a load run test on August 22-23, l 1996.

i 9. The item 15 cosmient in attachment 2 indicates that as a result of NSM ON-52966, the reclosure timers for the Keowee ACBs are set at 8.2 to 8.4 seconds, and the acceptance criterion is that they be greater than 4

4 seconds. What is the basis for the acceptance criterion? Why is there no upper bound on the acceptance criterion?

10. The item 16 comment in attachment 2 states that during the standby bus source undervoltage sensing test, each logic chain is actuated to ensure i

i that the resulting retransfer to startup signal is achieved. It indicates that the two-of-three verification is not necessary for the i standby bus undervoltage logic since a single failure is necessary in j order to require actuation of this logic.

I It is not clear what portion of the undervoltage sensing logic this test verifies and does not verify. How many logic chains are there, and is

! the two-of-three undervoltage signal provided to each logic chain? What

types of failures would the test not capture and how is the test conducted?

11. Item 57 in attachment 2 of DPC's letter, indicates that item 18_in the i

NRR draft report (page 27) ir< correctly states that there is no similar 4

technical specification requirement at Oconee for the standard diesel ,

generator test of reaching rated voltage and frequency within a specified time. It comments that the annual emergency start test verifies that i Keowee can obtain the rated voltage and frequency within the test acceptance limits.

i

We note that the subject standard technical specification test referred i

to in the NRR draft report is a 10-year test of the simultaneous start l capability of both diesel generators. Is the annual emergency start test of the Keowee units referred to in the comment to item 57 a test of the

{ simultaneous start capability of both Keowee units? If both units are not emergency started simultaneously from a standby condition in that

{ test, it is not comparable to the subject 10-year test. Is a 4 simultaneous emergency start from standby periodically performed on both Keowee units?

4

)

h

• l

12. The final portion of the response to open issue No. I discusses the electrical loading that would occur if ECCS actuation signals were received during a three-unit Oconee LOOP. It indicates that if an l

ECCS actuation were to occur, only one additional HPI pump would be started in each unit, over and above the other HPI pumps and essential )

loads that were energized by other automatic features during the  ;

LOOP event. The response states that the load associated with these i additional HPI pumps is smaller than the Oconee LOOP or LOCA loads that  !

' are currently analyzed to be block loaded onto an overhead or underground Keowee unit. It concludes, therefore, that the emergency power system would perform its intended function and is bounded by the current j analyses. '

! It is not clear why the fact that the load of the HPI pumps is leis than

the entire LOOP or LOCA load, leads to the conclusion that the current .

. analyses bounds the subject scenario. Is this indicating that, because l

! .the loads are staggered during the event, the voltage transients seen are

! less than those analyzed? If so, how does the final steady-state load l during this scenario compare to the analyses in terms of voltage and CT4 i or CTS loading capability. Has a three-unit LOCA/ LOOP event been i analyzed, or does the conservatism used in the CYME analyses bound the j three-unit LOCA/ LOOP event?

I i,

4 i e n

i l -

i e

i i

i l

i i

i I

.