ML20154N522

Summary of 880510 Meeting W/Util Re Single Failure Problem W/Charging Pumps Sys.List of Attendees Encl

ML20154N522
Person / Time
Site:	Haddam Neck File:Connecticut Yankee Atomic Power Co icon.png
Issue date:	05/24/1988
From:	Wang A Office of Nuclear Reactor Regulation
To:	Office of Nuclear Reactor Regulation
References
NUDOCS 8806020328
Download: ML20154N522 (9)

Text

.

o, UNITED STATES NUCLEAR REGULATORY COMMISSION

[ ' g/f ,g g j WASHINGTON, D. C. 20$$5 k**u4/ *** NRf 14 1M8 Docket No. 50-213 LICEi;SEE: Connecticut Yankee Atomic Power Company FACILITY: HA0 DAM NECK PLANT

SUMMARY

OF MAY 10, 1988 MEETING REGARDING VOLUME CONTROL TANK OUTLET VALVE FAILURE On May 10, 1988, the staff met with Connecticut Yankee Atomic Power Company (CYAPCO) to discuss a single failure problem with the charging pumps system.

During routine inservice testing of motor operated valves (MOVs), MC"-257, volume control tank (VCT) outlet valve failed to operate. In perfoming tne root cause analysis, CYAPCO detemined that the charging system is susceptible to a single failure by either tne VCT outlet or RWST suction to charging POV failing.

CYAPC0 stated this single failure was missed because in an early single failure analysis (1970) this single failure was noted, but since no credit was taken for the charging flow it was not considered. As result of various reanalysis in 1986, CYAPC0 detemined that a break in the core deluge line was susceptible to single failure and detemined charging is needed in the recirculation phase. CYAPCO performed a modes and effects analysis to ensure the plant had no other single failure vulnerabilities. However, this analysis only provided a review for single fa'iure vulnerabilities after the injection phase. Therefore, this failure was not identified.

CYAPC0 has reviewed several options. They believe the best interim solution to eliminate this single failure vulaerability is to trip the charging pumps onasafetyinjectionsignal(SIS). CYAPCO reiterated that no credit is taken for the charging pumps except for the case of a core deluge line break during the recirculation phase. In the recirculation phase, the charging pumps are not susceptible to a single failure because the charging suction is supplied from the RHR pumps discharge, which provides sufficient back pressure to close the on-line check valve. If the check valves were to feil, this would only divert flow to the VCT and the charging pump integrity is maintained.

CYAPC0 stated they will perform a single failure review and any open items will be resolved before entering MODE 4. Additionally CYAPC0 will repair the VCT single failure problem to reestablish charging flow during the injection pnase before start-up from the 1989 refueling outrage.

8806020320 080524 PDR ADOCK 05000213 P PDR

4 2-The staff agreed that the actions taken by CYAPC0 are prudent. However, to assure ourselves of the scope and adequacy of the review the staff needs to

-review the single failure analysis. If the staff can agree that the single failure review is adequate and CYAPC0 will cormiit to repairing the charging pumps single failure problem Sy the end of the 1989 refueling outage the staff would be able to support an emergency license amendment to support restart.

Additionally, the staff asked if CYAPC0 could trip on some other logic which might start the charging purrps but still assure integrity of the charging pumps (for example tripping for VCT level). CYAPC0 agreed to look for other possi-bilities and call the staff on May 11, 1988 to provide a status of their work and the possible amendment.

Enclosed is the attendance list and an initial list of options examined by CYAPC0 to resolve the single failure vulnerability, iM $259 Y Alan B. Wang, Project Manager Project Directorate I-4 Division of Reactor Projects I/II

Enclosure:

As stated LA:PD yl- PM:PDI-4 PD DI-4 SNobis AWandtbV JS 544i/88 5/)Y/88 5/4 /88

' @ 21.)

MEETING

SUMMARY

DISTRIBUTION

-Docket File

NRC & Local PDRs Gray File AWang OGC-WF SNorris EJordan JPartlow l

NRC Participants Tom Shediosky NRC Region I Y Gene Hsii

( John F. Stolz Bob Jones Wayne Hodges Rudy Karsch HBClayton 1

1

Mr. Edward J. Mroczka y Connecticut Yankee Atomic Power Company Haddam Neck Plant CC: -

Gerald Garfield, Esquire R. M. Kacich, Manager Day, Berry and Howard Generation Facilities Licensing Counselors at Law Northeast Utilities Service Company City Place Post Office Box 270 Hartford, Connecticut 06103-3499 Hartford, Connecticut 06141-0270 W. D. Romberg, Vice President D. O. Nordquist Nuclear Operations Manager of Quality Assurance Northeast Utilities Service Company Northeast Nuclear Energy Company Post Office Box 270 Post Office Box 270 Hartford, Connecticut 06141-0270 Hartford, Connecticut 06141-0270 Kevin McCarthy, Director Regional Administrator Radiation Control L'rit Region I Department of Envirofmental Protection U. S. Nuclear Regulatory Commission State Office Building 475 Allendale P,oad Hartford, Connecticut 06106 King of Prussia, Pennsylvania 19406 Bradford S. Chase, Under Secretary Board of Selectmen Energy Division Town Hall Office of Policy and Management Haddam, Connecticut 06103 80 Washington Street Hartford, Connecticut 06106 J. T. Shedlosky, Resident Inspector Haddam Neck Plant D. B. Miller, Station Superintendent c/o V. S. Nuclear Regulatory Comission i

Haddam Neck Plant Post Office Box 116 Connecticut Yankee Atomic Power Company East Haddam Post Office RFD 1, Post Office Box 127E East Haddam, Connecticut 06423 East Hampton, Connecticut 06424 G. H. Bouchard, Unit Superintendent Haddam Neck Plant RFD #1 Post Office Box 127E East Hampton, Connecticut 06424 l

l l

l l

ATTENDANCE LIST MAY 10, 1988 NAME COMPANY PHONE Jack Stanford Connecticut Yankee 203-267-2556 x452 Marty VanHaltern NU 203-665-5321 Bill Herwig NU 203-665-3198 Gene krineik NU 203-665-5865 Don Katze NRR/ DEST /SRXB 301-492-0886 Tom Shediosky NRC Region I, Sr. Res. 203-267-2571 Y. Gene Hsii NRC/SRXB 301-492-0887 Alan Wang NRC/PM 301-492-1313 Steve Vick NU 203-665-5537 Richard Kacich NU 203-665-3298 Don B. Miller CYAPCO 203-267-2556 John F. Stolz NRC/NRR/PDI-4 301-492-1300 John A. Blaisdell NU 203-665-5869 Bob Jones NRC/NRR/ DEST 301-492-3232 Wayne Hodges NRC/NRR/ DEST 301-492-0895 Rudy Karsch NRC/NRR/00EA 301-492-1178 I

l l

t

{

l i

l

Attachment 1 i 812918/Page 1 May 9, 1988 Ootions With regards to the imediate corrective action to resolve'two single failure vulnerabilities (CH MOV 257 and CH MOV 373), CYAPC0 explored nine options and chose the safest, most prudent option. The nine options, discussed in order of hargare modifications involved from least to greatest, are discussed below.

o Ootion No. 1 Recuest a Temoorary Exemotion From the Recuirement This option, to request an exemption for Cycle 15 from the GDC No. 35--

single failure criterion, was cor sidered because the long-term ECCS modifications already scheduled for completion during the 1989 refueling outage would result in the charging system no longer being required to show protection in the event of design basis events.

The exemption option was rejected because it did not resolve the under-lying problem the single failure vulnerability of the charging systera.

o Ootion No. 2 Dedicated Valve Ooerator Stationing a person full time outside the "B" charging pump cubicle (where CH MOV 257 is located) was considered, though not seriously. This option would have an adverse impact on the Operations Department, would be dependent on the reliability of the communication between the control room and the operator, and would require almost immediate recognition and actions by the operators. For these reasons, this option was discarded.

o Ootion No. 3 Alian One Pumo to the RWST. One to the VCT In order to ensure that one charging pump is available in a LOCA event, one pump (the "A" charging pump) could be isolated from its normal suction (the VCT) by closing manual Valve CH V-269 and opening CH MOV-32, aligning the "A" charging pump to the RWST. Piping and valve configura-l tion prevents isolating the "B" charging pump from the VCT and aligning it to the RWST while still aligning the "A" charging pump to the VCT.

With this option, only the "B" charging puep is available f" normal charging from the VCT, if back leakage occurred through the ioie pump's

("A" charging pump) discharge check valve and through the now open (1) It should be noted that not every option would resolve bath single failure vulnerabilitier. Only Option Nos.1, 4, 6, 7, and 8 will resolve the single failure vulnerability of CH MOV 373.

e

) Attachment 1 812918/Page 2 May 9, 1988 CH MOV-32, the RWST level would be raised. Having only one charging pump available for normal operations is highly undesirable.

The idle ("A") pump in this alignment would inject highly borated water from the RWST in the event its standby feature were actuated (low discharge pressure). This negative reactivity addition would be undesir-able, especially toward end of life (EOL). However, the most unattrac-tive implication of this option is that, if in a LOCA with a loss of normal power (LNP) the B pump fails when it is restarted for recircula-tion, the operator would be required to open the manual valve (CH V 269)

. in the "A" charging pump suction to allow the "A" charging pump to take a suction from the RWST. The time to perform this action may not be sufficient to meet the times used in the accident analysis. For these reasons, this option was rejected.

o Ootion No. 4- Place One Pumo Switch in the Pull to-lock Position This option would ensure one charging pump is protected in the event of CH HOV 257's failure to close on an SIAS. It would allow the single failure criteria to be met and involve no hardware changes. However, the disabling of the automatic start of the standby pump in the case of low charging pump discharge pressure is deemed unacceptable from an oper-atcr's viewpoint. The option also poses questions concerning operability requirements of technical specifications. Loss of the running charging pump would mean a temporary loss of RCP seal injection an event that is undesirable from an operator's viewpoint. For these reasons, the pull-to lock option was rejected.

o Ootion No. 5 Limit VCT Pressure and level The option to limit VCT pressure and level such that its head relative to the RWST head in a LOCA and in the event of CH MOV 257 failure to close was considered. This option essentially eliminates the potential of failing both charging pumps by ensuring that in a LOCA event the head of

( the piping from the RWST would be greater than that from the VCT, thus preventing hydrogen gases from the VCT from entering the charging pump

suction lines. Several variations of this theme were investigated.

Automatic venting of the VCT with an SIAS, automatic closure of hydrogen supply to the VCT, and regulating VCT level limits to some higher level were all examined. Automatic venting of the VCT was rejected for radio-logical and operational reasons. Automatic closure of the hydrogen supply on an SIAS would involve procurement of new equipeent and rather extensive engineering to implement. Calculations showed that adjustment I of VCT level alone would not be sufficient to ensure hydrogen did not l enter the charging pumps. For these reasons, this option was rejected.

i 1

{

l l

J

Attachment 1

• B12918/Page 3 May 9, 1988 o Ootion No. 6 Protect a Desionated Pumo on in SIAS This option consists of designating one of the charging pumps (e.g., the "B" charging pump) as the pump to be protected in case CH MOV 257 fails to close with an SIAS. It involves minor hardware changes. While this option solves the single failure problem for both CH MOV 257 and CH MOV 373, it has negative aspects. For instance, the reliability of the charging pumps with this option is less than others that could be chosen. If the "protected" pump happened to be the pump in use for normal operations when an SIAS occurred, the running pump would trip and

, the "nonprotected" pump would start. This is less desirable than the option to protect the nonrunning pump (see Option No. 7).

The safety of this option was determint also to be less than that of other options. If the nonprotected pump is inoperable for maintenance or other reasons and ah SIAS occurred, the protected pump may not be able to be started, depending on the design of the logic circuitry, until the SI signal were reset.

However, more importantly from a safety standpoint is the issue of hydrogen accumulation in the suction piping. This was a concern for both Option Nos. 6 and 7. The concern is that though one pump is protected, the other pump is assumed to have failed due to hydrogen gas binding of the pump. Due to the piping configuration, some amount of hydrogen gas is assumed to be in the protected pump's suction line as well.. Starting the protected pump without venting the hydrogen may result in failure of the "protected" charging pump. Venting the hydrogen was determined to be too time consuming and posed personnel safety risks.

In addition, the revision of emergency operating procedures (EOPs) and resultant operator training requirements were determined to be too burdensome. For these reasons, the protecting a designated pump option was rejected.

o Ootion No. 7 -Protect the Nonrunnino Pumo This option differs from Option No. 6 only in that the nonrunning pump (or, in the rare case when both charging pumps are running, a designated pump) is protected. This option has the same positive aspects as Option No. 6, plus an increase in reliability. However, all the negative aspects of Option No. 6, including the hydrogen accumulation issue, are still there for this option. For this reason, this option was rejected.

l l

l l

l l

s

l

. Attachment 1 812918/Page 4 May 9, 1988 o Ootion No. 8 -Trio Both Pumos on an SIAS This option entails automatically tripping both charging pumps on an SIAS. The advantages of this option are that the single failure crite-rion is met, for both CH MOV 257 and CH MOV 373 the change has minimal impact on the E0Ps, this option eliminates the hydrogen accumulation in the suction line issue, and involves minimal hardware changes.

While this option is determined to be safe, CYAPCO recognizes that having both charging pumps trip on an SIAS, even when normal off-site power is available, is not ideal from an operations standpoint. However, this scenario is bounded by the plant LOCA response in which both charg pumps trip automatically in the event of a coincident LOCA and LNP.g Further, this operational concern is minimal in that the design allows the operator to restart the charging pumps 30 seconds after the SIAS, even before the UAS signal is reset. This minimizes the period seal injection is lost in the case of a spurious or inadvertent SIAS.

This is the option CYAPC0 has chosen as an interim measure. Details of the design of this option are contained in Attachment 2.

o Ootion No. 9 -install Redundant Motor Ocerated Valve This option, an obvious solution to the single failure vulnerability, is the likely long term solution for CH MOV 257 single failure. Coincident with this hardware modification would be a design change to include CH-MOV 32 as one of the valves that open automatically with an SIAS.

This would solve the single failure vulnerability of CH MOV-373. These modifications would allow the automatic tripping of the charging pumps on an SIAS with normal off site power available to be returned to the original configuration in which the charging pumps start or continue to run on an SIAS with normal off site power available.

The disadvantages of this option as an immediate solution is that the time to en procure, and install a new MOV is considerable.

Furthermore,gineer, even if the equipment (new valve) was available now, the l single failure vulnerability would remain until the completion of the new switchgear room to allow a diverse vital power supply. The new switch-i gear room is scheduled to be operational by the end of the 1999 refueling

! outage.

(2) See letters from W. G. Counsil (CYAPCO) to D. L. Ziemann (NRC), dated May 22 and September 22, 1978.

l l

l l

l r

_ _ _ _ . _ . , . . ..____.,..._..,m. __ - _ . _ _ , , , . _ . - _ . . . _ . , __