ML17083C172

Summary of 890216 Meeting W/Util & Consultants in San Francisco,Ca to Discuss PRA,890217 Audit of Relay Chatter Analysis & Facility long-term Seismic Program.List of Attendees Encl

ML17083C172
Person / Time
Site:	Diablo Canyon
Issue date:	03/29/1989
From:	Rood H Office of Nuclear Reactor Regulation
To:	Office of Nuclear Reactor Regulation
Shared Package
ML16341F078	List: ML17083C172
References
TAC-55305, TAC-68049, NUDOCS 8904060327
Download: ML17083C172 (98)

Text

I gP,R REGG, (4

+4

~

0 Cy I

0O 4y w+n +**++

DOCKET NOS.:

50-275 and 50-323 I

/

I P

I t'

, UNITEDSTATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555 March 29, 1989 APPLICANT:

FACILITY:

PACIFIC GAS AND ELECTRIC COMPANY (PG&E)

DIABLO CANYON NUCLEAR POWER PLANT, UNITS 1

AND 2

SUBJECT:

Attendees at the meeting are given in Enclosure 1.

The agenda for February 16, 1989 is given in Enclosure 2.

Slides presented at the meeting by PG&E on the subject of "High Pressure Function of the ECCS" are given in Enclosures 3.

PG&E slides on "Low Pressure Function of the ECCS" are given in Enclosure 4.

After the meeting, PG&E performed additional calculations to address the importance of the residual heat removal system crosstie valve.

The results of these calculations were given to the staff on February 17, 1989, and are included in Enclosure 5.

Based on PG&E's submittal dated February 10, 1989, and the information presented in Enclosures 3, 4, and 5, the staff considers the issues raised in the two above-referenced BNL reports to be resolved.

However, it should be noted that the review of these two systems in ongoing, and if additional concerns arise, further discussions with PG&E may be necessary to resolve them.

SUMMARY

OF FEBRUARY 16, 1989 PUBLIC MEETING TO DISCUSS PROBABILISTIC RISK ASSESSMENT (PRA),

AND FEBRUARY 17, 1989 AUDIT OF RELAY CHATTER ANALYSIS, DIABLO CANYON LONG TERM SEISMIC PROGRAM (LTSP)

(TAC NOS.

55305 AND 68049)

On February 16, 1989 the NRC staff and its consultants met with PG&E in San Francisco, California to discuss PG&E's response to two reports by NRC staff consultants from Brookhaven National Laboratory (BNL).

The reports were transmitted to PG&E by NRC "DCL-89-008, Forwards Public Version of Revised Emergency Plan Implementing Procedures,Including Rev 10 to EP G-1,Rev 4 to EP RB-15:B,Rev 3 to RB-15:G & Rev 0 to [[procedure" contains a listed "[" character as part of the property label and has therefore been classified as invalid. RB-15:O.W/undated Release Memo|letter dated January 10, 1989]].

The BNL reports evaluated selected aspects of the Probabilistic Risk Assessment (PRA) that was included in PG&E's July 31, 1988 final report on the Diablo Canyon Long Term Seismic Program.

The BNL reports are identified as (1) Letter Report-02/Rev.l, "A REVIEW OF SYSTEMS ANALYSIS IN THE DCPRA:

HIGH PRESSURE FUNCTIONS OF THE EMERGENCY CORE COOLING SYSTEM," December 1988; and (2) Letter Report-03, "A

REVIEW OF SYSTEMS ANALYSIS IN THE DCPRA:

LOW PRESSURE FUNCTIONS OF THE EMERGENCY CORE COOLING SYSTEM," December 1988.

89p4p60 p~ppp275 327 S9p329 P'DR pOOCK PDC A~o P ZF<<N I Iy

V g 'a

'I

March 29, 1989 On February 17, 1989, at the PGKE offices in San Francisco, California, the NRC staff conducted an audit of the relay chatter analysis performed by PGSE as part of the Diablo Canyon PRA.

During the audit, PG8E gave the staff a summary of the methodology used in the relay chatter analysis (see ).

Two specific calculations were selected to be audited, the effects of relay chatter on the auxiliary salt water system, and the bus transfer scheme.

Comments by the staff consultants from BNL about the February 16, 1989 meeting and the February 17, 1989 audit are given in Enclosure 7.

Harry Roo

, Senior Project Manager Project Directorate V

Division of Reactor Projects - III, IV, V and Special Projects

Enclosures:

1.

Meeting Attendees 2.

Meeting Agenda 3.

PGEE Slides on High Pressure Function of the ECCS 4.

PGSE Slides on Low Pressure Function of the ECCS 5.

PGSE Slides on RHR Crosstie Valve Importance 6.

PGSE Summary of Relay Chatter Analysis Methodology 7.

BNL Letter Commenting on Meeting and Audit cc: w/enclosures - see next page

, ~

c

'I y L,

( March 29, 1989 On February 17, 1989, at the PG&E offices in San Francisco, California, the NRC staff conducted an audit of the relay chatter analysis performed by PG&E as part of the Diablo Canyon PRA.

During the audit, PG&E gave the staff a summary of the methodology used in the relay chatter analysis (see Enclosure 6).

Two specific calculations were selected to be audited, the effects of relay chatter on the auxiliary salt water system, and the bus transfer scheme.

Comments by the staff consultants from BNL about the February 16, 1989 meeting and the February 17, 1989 audit are given in Enclosure 7.

original signed by Harry Rood, Senior Project Manager Project Directorate V

Division of Reactor Projects - III, IV, V and Special Projects

Enclosures:

1.

Meeting Attendees 2.

Meeting Agenda

~

~

~

3.

PG&E Slides on High Pressure Function of the ECCS 4.

PG&E Slides on Low Pressure Function of the ECCS 5.

PG&E Slides on RHR Crosstie Valve Importance 6.

PG&E Summary of Relay Chatter Analysis Methodology 7.

BNL Letter Commenting on, Meeting and Audit cc: w/enclosures

- see next page DISTRIBUTION f ~n NRC

& LPDRs JSniezek PD5 Plant File GKnighton HRood OGC EJordan BGrimes ACRS (10)

MJohnson NChokshi RRothman DRSP/PDS+

D

.D5 HRood G

ton 03~/89 0 /~PI/89

14 if I '

ll'

~

P kllk hfl I I'4 fl H

~

I h

H4'

~ 4 I M

P H

II

~ I' II c

f I' HII V

4 I

IIA l3 li

>y I.'1V V

f

'I

,I I

1 ) "II'I 3

f If I

I II h fk C.

h f

.r II h

h II,I

I J

Ih II,.>r 4

~

4I V

=H 1

H h,f g,',

Hfl h

Ht lh h

Mr. J.

D. Shiffer Pacific Gas and Electric Company Diablo Canyon CC:

'Richard F. Locke, Esq.

Pacific Gas 5 Electric Company Post Office Box 7442 San Francisco, California 94120 Janice E. Kerr, Esq.

California Public Utilities Commission 350 HcAllister Street San Francisco, California 94102 Ms. Sandra A. Silver 660 Granite Creek Road Santa Cruz, California 95065 Mr. W. C. Gangloff Westinghouse Electric Corporation P. 0.

Box 355 Pittsburg, Pennsylvania 15230 Managing Editor San Luis Obis o Count Tele ram Trl une 13'~o nson Avenue P. 0.

Box 112 San Luis Obispo, California 93406 Mr. Leland M. Gustafson, Manager Federal Relations Pacific Gas and Electric Company 1726 M Street, N.

W.

Washington, DC 20036-4502 Dian M. Grueneich Marcia Preston Law Office of Dian H. Grueneich 380 Hayes Street, Suite 4

San Francisco, California 94102 NRC Resident Inspector Diablo Canyon Nuclear Power Plant c/o U.S. Nuclear Regulatory Commission P. 0.

Box 369 Avila Beach, California 93424 Mr. Dick Blakenburg Editor 5 Co-Publisher South County Publishing Company P. 0.

Box 460 Arroyo Grande, California 93420 Bruce Norton, Esq.

c/o Richard F. Locke, Esq.

Pacific Gas and Electric Company Post Office Box 7442 San Francisco, California 94120 Dr. R. B. Ferguson Sierra Club - Santa Lucia Chapter Roc'anyon Star Route Creston, California 93432 Chairman San Luis Obispo County Board of Supervisors Room 270 County Government Center San Luis Obispo, California 93408 Director Energy Facilities Siting Division Energy Resources Conservation and Development Coamission 1516 9th Street Sacramento, Califor nia 95814 Hs. Jacquelyn Wheeler 3033 Barranca Court San Luis Obispo, California 93401

" Pacific Gas 5 Electric Company Diablo Canyon Ms. Laurie McDermott, Coordinator Consumers Organized for Defense of Environmental Safety 731 Pacific Street, Suite 42 San Luis Obispo, California 93401 Mr. Paul Szalinski, Chief Radiological Health Branch State Department of Health Services 714 P Street, Office Building 88 Sacramento, California 95814 Regional Administrator, Region V

U.S. Nuclear Regulatory Commission 1450 Maria Lane Suite 210 Walnut Creek, California 94596 Ms. Nancy Culver 192 Luneta Street San Luis Obispo, California 93401 President California Public Utilities Commission California State Building 350 McAllister Street San Francisco, California 94102 Michael M. Strumwasser, Esq.

Special Assistant Attorney General State of California Department of Justice 3580 Wilshire Boulevard, Room 800 Los Angeles, California 90010

'i

Pacific Gas and Electric Company Long Term Seismic Program Diablo Canyon CC:

Dr. Keiiti Aki Department of Geological Sciences University Park University of Southern California Los Angeles, California 90089-0741 Dr. Ralph J. Archuleta Department of Geological Sciences University of Califotnia Santa Barbara Santa Barbara, California 93106 Dr. Robert D. Brown, Jr.

U.S. Geological Survey Mail Stop 977 345 Middlefield Road Menlo Park, California 94025 Dr. David B. Slemmons Center for Neotectonic Studies MacKay School of Mines University of Nevada-Reno

~

~

~

Reno, Nevada 89557-0047 Dr. Robert Fitzpatrick Building 130 Brookhaven National Laboratory

Upton, New York 11973 Dr. C. J. Costantino Building 129 Brookhaven National Laboratory

Upton, New York 11973 Lorig Term Seismic Program Dr. Steven M. Day Department of Geological Science San Diego State University San Diego, California 92182 Dr. George Gazetas Dept. of Civil Engineering 212 Ketter Hall SUNY-Buffalo Buffalo, New York 14260 Dr. Jean Savy Mail Code L-196 Lawrence Livermore National Laboratory P. 0.

Box 808 Livermore, Califor nia 94550 Dr. Anestis S. Veletsos 5211 Paisley Avenue

Houston, Texas 77096 Dr. Ken Campbell U.S. Geological Survey P. 0.

Box 25046, Mail Stop 966 Denver Federal Center Denver, Colorado 80225 CC Dr. Michael Bohn Sandia Lab. - Organization 6412 Post Office Box 5800 Albuquerque, New Mexico 87185 Dr. J. Johnson E(E 595 Market Street - 18th Floor San Francisco, California 94105 Dr. M. K. Ravindra EgE 3150 Bristol Street, Suite 350 Costa Mesa, California 92626

ENCLOSURE 1

ATTENDEES Public Meeting on Probabilistic Risk Assessment Diablo Canyon Long Term Seismic Program

Thursday, February 16, 1989 NAME S. Abek Michael P.

Bohn George Bozoki D. A. Brand Ni lesh Chokshi Lloyd S. Cluff Charles Coffer Allin Cornell Robert G. Fitzpatrick David W. Ogden Harry Rood Bimal Sarkar George Sarkisian

~

~

~

~

Bruce D. Smith Raymond Thierry Michael R. Tresler

. George Wu ORGANIZATION BNL (NRC)

Sandia National Laboratory (NRC)

BNL (NRC)

PG&E NRC/RES/PRAB PG&E PG&E Consultant to PG&E BNL (NRC)

PG&E

" NRC/NRR/PDV Bechtel (PG&E)

PG&E PG&E PG&E PG&E PG&E

ENCLOSURE 2 DIABLO CANYON LONG TERM SEISMIC PROGRAM NRC/PG&E MEETING ON PROBABILISTIC RISK ASSESSMENT FEBRUARY 16, 1989 ROOM 1753, 77 BEALE STREET SAN FRANCISCO, CALIFORNIA 8:30 8:40 INTRODUCTION MEETING ORGANIZATION PG&E - LLOYD CLUFF NRC

- HARRY ROOD BRUCE SMITH 8:45 ll:00 11:30 12:00 1:00 3:00 3'30 4:00 NRC CAUCUS NRC COMMENTS LUNCH PG&E RESPONSE TO REPORT 3

NRC CAUCUS NRC COMMENTS ADJOURN NRC/BNL HARRY ROOD/NILESH CHOKSHI RAYMOND THIERRY NRC/BNL HARRY ROOD/NILESH CHOKSHI PG&E RESPONSE TO REPORT 2

RAYMOND THIERRY

I I

ENCLOSURE 3 PGSE SLIDES ON HIGH PRESSURE FUNCTION OF THE ECCS

0

HIGH PRESSURE FUNCTIONS OF THE EMERGENCY CORE COOLING SYSTEM REVIEW COMMENTS AND REQUEST FOR ADDITIONALINFORMATION

I

1

SUMMARY

F BNL MMENT 1 According to the boundary condition for split fraction HRA (4kv Buses F and H failed) BNL believes the split fraction value for HRA should be i.0.

SUMMARY

F PGJkE RE P.

NSE The split.fraction value used by PGEcE is appropriate.

DISCUSS lON o

High Pressure Recirculation is Not Required to the Safety injection Pumps for this Boundary Condition o

Valves Modeled as a Conservatism for This Boundary Condition o

Procedures instruct the Operators to Close Valves 8974A and 89?4B Either Remotely or Locally o

Sl Recirculation Flow 30 gpm, RHR Diverted Flow Much Less {Approximately 1/3) With a Total Diversion of 4%

of RWST in 24 Hours

TOP EVENTS SUPER COMPONENTS RCS LOOP I RCS LOOP 2

~

OOA I

~ 004

~

1 b

I OA RECKN HKaT ElCNANCER IIOT H

~ 100 H VOLIIHE CONTROL TANd H LCV Iltb H LCV lllC

~Ido H

CONTAINHENT SPRAT INST.

ACC I

I HIIOIA RC5 LOOP S RCS LOOP ~

49000 Idl 0 RCS LOOP I

~

~

144

~ lta

~ ZZA 49OOC 441 C

IZO 4 Olb dd0 4 Ido I

~ dta ~ 4744

~ 944 CNI I 4<<te 4 698 ~ 4768 43948

~

4 4415

~ 17 4 IATIA 8

4924 4

0 I I ~

H J G

I 0 4 741 I 564 ACC I-I ACC 1-2 HII048 HIIOID RCS LOOP 0

~ 460 RCS LOOP 2

~ 444

~4190 4210

~

140 I 144

~

194 IZZI RCS NOT LEOS Ido i 45 49124 I ZIA 49194 44114 49104 418 CH I?

Sl I-I I 14

~ 0 4

~ IOT K

91 t25 916

~ Sdb ACC I

HIIOIC RCS LOOP 3 44 Itc IZZC 4014C H

RCS HOT LEOS Si ~

~

0 I CaSo HE40ER 3m 4 49tlb It208 49198 I Zlb ~ 12b SI 1-2 91 IIOlb4 Co5o HEADER t.d RCS HOT I

I F 04 RCS HOT

~ ~ ~

10 PRT H

44564 404 ITI6AHCv 634 LEIDovN NEAT KTCHANCKR 47168

~134a

~ 1344 424 sl V dali scv 364 HOR 8 4~4 1304

~7154 RN

~ 1154 SCV 56$

CCN NPR 4 IOA RCS LOOP 0 HOT I.KC 701 10 RECIRC I 0 b PRT Hcv 637 44568 T

TOP EVENT ~

H sCv 64 lb

~ 148

~ 1 4 0012

~ 0 I PRT

~ TO SCREENS

~

~ 4 H

444 (Idden ISO)441957oK99 II td II IAR.

PlAQRAH E.4-3

I

480V AC BUSES 1FAle 16 480VAC BUS IF RI% PUHP TRAINA Ale 480VAG BUS fQ ISOtATE AWST FROM CCPa MOVa8805AIB A

5OLATERWST FROM SIPs 8974h ISOUtTE RWST FROM SIPs CV 8977

%%A TO CHAASN3 MOY8804A fSOtATE AWST FROM CCPs CVf624 ISOLATERWST FROM SIPs 88T48 ISOUtTE RWST FROM SIPa MOV8916

%%870 Sl PUMPS MOV8804 b Sf SUCTION VALVES MOVa8028NS I

480V AG BUS fH 480VAG BUS IH fFSUPPORT NRAVAlhKE'MOVs8805 NBWOULDNOT HAVEOPENED INW1EGTION PHASE ANDDONOT NEED TOBE CLOSED FOR SUCCESS OF IIR.

f5% PUhP TRhlN 8 480V AG BUS fH FIGtIRE E.4-9.

BLOCK DIAGRAM fOR TOP EVENT HR:

COLD LEG RECIRCULATION MITH RHR DISCUARGE ALIGMLD TO SI PUMP OR CIIARGING PUMP (TOP EVENTS SI ANO Cll SUCCLSSfUL)

~dent tC Qfk5

~tIte I

~LAAtfttttST Ttett CCte QOn eetttAOI

~eev JC QQtt

~TS tINSf tlS SPe NTIA t$1$ITT ttttST TAGES SW cvet tt tl~ t tLNt~

tItlhttI tellffNttI tO CttItrttO ItoveeotI 4lev Ic ttttS tC st sLtc feet ViLVL5 LtOvtetftdd t

QOVIC

~LJS tO

~10 St CletSSTE ttoe NotA CdtdteCI 10 St CKTSST%

etov eeotd VIL~

~%AT% ttWST 1teste ccte CYtete XATK ttteST tt&vStte Nttd 0

TTLSAtfttttST fttOtt $%

etOT elle tete ttttttd 10 St ttov eeetd St SLtCTOtt VALVT5 ttOVteeffetI I

OJS ttt

~tev IC LIJS III 0 tONOIT01%%~V$00TefAVALAdLL7l4 VRV5$TtOLAOtef ttee$ OtfI%0 et~ eLACT~ttdt$$5 %OK%0 teST CLTSTKfCH ItCC$$$ Cd 1&'EVE ttfttt tltttL~

ttteetd

~eeV IC dttf Itt FIQJRK K.4-10.

BLOCK DIAGRAM FOR TOP KVENT HR:

COLO LEG RECIRCULATION MITH RHR DlSCHARGE ALlGNEO TQ AN OPERATlNG HlGH PRESSURE PlNP (TOP EVENT CH ASSUMED FAlLEO)

l I

UMMARY F SNL OIVIIVIENT2

~

~

~

~

Split Fractions HRA and HR5 are Numerically Very Close, this Appears Inconsistent Since the HRA Boundary Condition is More Severe Than That for HR5.

SUIVIMARY F PG E RESPON E

The Split Fraction Values Should Be Numerically Close Since Both Cases Have Similar Cutsets.

DISCUSSION o

Each Split Fraction Has Similar Dominant Cutsets o

HR5 8974B Failure to Close on Demand or, in IVIaintenance (90%)

o HRA 8804A Failure to Open on Demand or, in Maintenance (89%)

~~,

DlS CUSS lON CONTINUED o

HRA Hardware and Maintenance Contribution HW = HWl+ HWD HWI = HRBKA1~HRBKB+ HRBKC~HRBKD

+ HRBKE + ~HRSK HWD = D2VMOO MN = ZMVBOF~ZMVBOD o

HR5 Hardware and Maintenance Contribution HWl = HRBKA1~HRBKB+ HRBKD

+ HRBKE~HRBKF + HRBKG~(HRBKH+HRBKI)

HWD = D2VMOO + 02VMOC MN = ZMVBOF~ZMVBOD+small

I

UMMARYOF BNL COMMENT The Value Calculated by BNL for Split Fraction HRC is in Disagreement with the PGEcE Calculated Value.

In Addition, the PGEcE Split Fractions for HRC and HR& are Numerically Similar Even Though the Boundary Condition for HR& is More Severe, This Seems Inconsistent.

MIVIARY F PG E RE P

E (1) The Value Calculated by PGEcE for Split Fraction HRC is Appropriate.

(2) The Numerical Values for HRC and HR& Should be Very Similar.

,DISCUSSION (1) o A Review of the BNL Cutsets for HRC Identified Three Missing Cutsets:

o HRBKK - MOV 8807B Fails to Open or Transfers Closed (1.556E-3) o HRBKL Manual Valve 8925 Transfers Closed (2.189E-4) o D2VIVlOO-Common Cause Failure of MOVs 8807A and 8807B to Open on Demand (8.430E-5) o Adding These to the BNL Value:

1.&59E-3 Total of Missing Cutsets 4.536E-BNL Value 6.395 E-3 6.430E-3 PGEcE Value

I y'

e I

1

DISCUSSION (2) o The Numerical Values for HR8 and HRC Should be Similar Since Their Boundary Conditions are Functionally Similar.

o HR8 4kv Bus F Unavailable, Top Event CH or Sl Failed, and Top Event LA or LB Failed o

HRC - 4kv Bus 6 Unavailable and Top Event CH or Sl Failed o

Each Split Fraction Has Similar Dominant Cutsets:

HR8 MOV 8974B Failure to Close on Demand and MOV 8804A Failure to Opened on Demand.

HRC MOV 8804B and 8807B Failure to Open on Demand

I I

SUMMARY

OF BNL COIVlMENT4 The Unavailability Calculations for RC1 and RC2 Seem to be Inconsistent with the Failure Rate Values Provided to BNL by PGcE.

SUMMARY

OF PGEcE RE P

MSE The Split Fraction Values for RC1 and RC2 are Consistent with the Failure Rate Data.

The Difference is Due to Monte Carlo Quantification Versus Point Estimate Quantification.

DISCUSS lON o

LTSP Chapter 6 Results MONTE CARLO (Table 6-3?)

RC1 4.43E-5 RC2 1.1&E-3 POINT ESTIMATE (Table 6-46)

RC1 4.389E-5 RC2 1.175E-3 o

BNL RESULTS POlNT ESTIMATE (Table 2.3)

RC1 4.3&QE-5 RC2 1.176E-3 o

The Mean of the Square of a Random Variable is Greater Than the Square of the Mean of the Random Variable o

Driving Cutset RCBKA~RCBKA

L

IVllVIARY F BNL RFI 1

Does the Quarterly Safety Injection Pump Operability Test Contribute to System Unavailability?

UMIVlARYOF PG4.E RESP N

E No, the Quarterly Safety Injection Pump Operability Test Does Not Contribute to System Unavailability.

DISCUSSION o

Test Alignment is the Same as Normal Alignment o

Flow Recirculation to RWST o

Not a Full Flow Test (30 gpm) o System Never Made Unavailable During Test

I

)

IVIIVIARY F BNL RFI 2

J What Constitutes the Monthly Test of MOVs in the ECCS Which Verity Operable Pathways?

What is the Duration of These Tests and Their Contribution to System Unavailability?

SUMMARY

OF PG&cE RESP N

E o

Once Every 31 Days the Flow Path From the RWST to RCS Via the Charging Pumps is Verified by Checking Each Valve Position o

Every Quarter the Charging Pumps are Tested in Normal Alignment System Response is Unchanged Pump Need Not Start, Already Running No Contribution to System Unavailability o

ECCS Valve Stroke Tests Are Performed Quarterly Short Duration (I to 15 Minutes).

Single Train Unavailability Due to Test is 3.8E-5 o

A Coincident Failure Would be Required to Fail the Opposite Train CHI 6.2E-4 (Two Trains)

CH2 1.4E-2 (One Train)

CH Actuation Signal Operational Sli 3.3e-3 (Two Trains)

SI2 1.6E-2 (One Train)

HR1 Valves Must Change Position RCi Valves Must Change Position

1 l

UIVIIVIARY F BNL RFI 3 lf unscheduled maintenance of IVIOVs in the ECCS requires valve isolation, would the plant shutdown immediately or only after the allowed outage time?

SUIVIIVIARYOF PGJkE RESP M

E lfit is Clear That the Required Maintenance Cannot be Accomplished Within the Allowed Outage Time, Then the Unit Would Begin Shutdown immediately.

DlSCUSSION o

MOV Maintenance included o

One Hour Tech. Spec. Valves Not Included

ENCLOSURE 4

PGEE SLIDES ON LOW PRESSURE FUNCTION OF THE ECCS

LOW PRESSURE FUNCTIONS OF THE EMERGENCY CORE COOLING SYSTEM REVIEW COMMENTS AND REQUEST FOR ADDITIONALINFORMATION

UMMARY F BNL OMMENT 1 For LOCA Scenarios, the Assumption That the Rupture Occurs in Cold Leg 1 is Not as Accurate (for Top Events LI and AC ) as the Assumption That the Rupture May Occur in Any of the Four Cold Legs.

SUMMARY

OF PG&E RESPONSE The Assumption That the Break Occurs in Cold Leg 1 is Valid.

D IS CU SSI ON o

RHR Cross-tie Line Normally Open (8716A and 8716B) o Injection Independent of RHR Train o

BNL Cutsets Four Times PG&,E Cutsets Factor of 1/4 Required o

PG&,E and BNL Values Agree

MMARY F BNL MMENT 2 The Use of a More Conservative Large LOCA Success Criteria (i.e., Two Out of Three) Would Result in an Increase of the LI1 Split Fraction. It Would Also Require Restructuring of the Large LOCA Event Tree.

SUMMARY

OF PG&cE RE P

NSE (1) The Safety Evaluation of The RHR Cross-Tie Line Isolation Prepared by Westinghouse Provides the Justification for the One-Out-Of-Three Injection Line Success Criteria.

In Light of This Analysis, the Use of a More Conservative Success Criteria is Not Necessary.

(2) The Split Fraction LI1 is Not Used in the Large LOCA Event Tree; Therefore, an Increase in LI1 Does Not Affect the Frequency of Core Damage.

(3) If a Success Criteria of Two Out of Three Were Used, it Would Not Require a Restructuring of the I.arge LOCA Event Tree.

DISCUSSION (1) o Indian Point PRA (FSAR Conservative) o No Impact on Plant Risk o

Use AllAvailable Information (Plant Specific)

e

DISCUSSION (2) o LI1 Not Used For Large LOCA Scenarios o

LI2 Used For Large LOCA - Injection After Melt o

AC Minor Impacts Dominated by 3/3 Accumulator Success DISCUSSION (3) o Change Success Criteria Requantify Top Events o

Large LOCA Event Tree Unchanged

S

UMIVIARYOF BNL COIVIMENT Small Discrepancies Were identified Between PGEcE and BNL Values in the Quantification of Certain HWI Values.

SUMMARY

OF PGJkE RESP NSE The Discrepancies Have Two Causes:

(1) Differences Between Monte Carlo Quantification and Point Estimate Quantification and (2) The Method Used to Quantify Conditional Failure Frequency.

The Discrepancies are'Not Significant Enough to Affect the Overall Accuracy in the Quantification.

DISCUSSION (1) o Monte Carlo Versus Point Estimate o

20 Bin Histogram Approximation of Log-Normal Distribution DISCUSSlON (2) o IVIonte Carlo and Division o

Obtain Means and Divide o

Fully Develop Equations o

One Step Versus Two Step Quantification o

Overall Accuracy Adequate

MMARYOF BNL COIVIMENT4 i 1 A Small Discrepancy Between the PGEcE and BNL Value for Split Fraction LI2 Was Noted.

UMIVIARYOF PG&,E RESPONSE The Source of This Discrepancy is Probably Due to Monte Carlo Quantification.

DISCUSSION o

PG&,E Re-evaluated Point Estimate LI2 5.25E-4 o

BNL Point Estimate LI2 5.2TE-4

SUMMARY

OF BNL RFl 1

What is the Duration of the RHR Pump Test and MOV Stroke Tests?

IVIMARY F PG E RE P

N E

The RHR Pump Test Has a Duration of Approximately One Hour.

Valve Stroke Tests Are of Short Duration Lasting From One to Fifteen Minutes With an Average of Approximately 5 Minutes.

DISCUSSION o

Pump Operates in Miniflow o

RCS Pressure Drops, MiniflowValves Close o

Test Unavailability 7.6E-7 1 hr/2160 hr 1.65E-3 lVIOV Failure to Close o

IVIOVStroke Test (Single Train Only) o Off-Normal Alignment 3.8E-5 o

Disables One Train Only o

Easily Recoverable

l

MMARY F BNL RFI 2

Could estimates for the Test Contribution to the Unavailability Be Provided Demonstrating That Test Contributions are Indeed Negligible.

SUMMARY

OF PGEcE RESP N

E The RHR Pump Test Has a Negligible Contribution to RHR System Unavailability. Valve Stroke Tests In the RHR System are Also Minor Contributors.

"ra~

l ~

UMMARY F BNL RFI Top Event AC Does Not Consider Common Cause Failures Between the Accumulator Outlet Check Valves. What is the Plant Specific Failure Rate for Check Valves Exposed to Boric Acid Deposition and corrosion for Both Independent and Common Cause Failures?

UMIVIARYOF PG E RESP MSE lt is Not Necessary to Model Common Cause Failures of the Accumulator Outlet Check Valves. The Three-Out-Of-Three Accumulator Success Criteria Makes This Unnecessary.

There is No Plant Specific Evidence that Would Justify Separate Failure Rates for Check Valves Exposed to Boric Acid.

DISCUSS lON o

Three-Out-Of-Three Success Criteria o

Total Failure Rates o

Conservative - Double Counts Common Cause o

4300 Check Valve Demands, Zero Failures o

Inspection Reports Do Not Indicate Problems o

No Basis

TOP EVENT LI PER COMPONENTS VOLVHE CotlTROL TAtN RC5 RCS L~ 2 89008 BSIOB SCOt4A REGEH HEAT EXCH4tIGER BI07 tlP-SISS H

H LCV I I2B H LCV I I2C 8840 COHTAIHHEHT SPRAY RCS s ooc seioc 820 SBot8 8 884 8475 CH I-3 8477A 84794 47 8924 88054 ACC I-I t488084 89564 esooD RCS RCS LOOP I

89484 BSIOD E

~~

88 ISA 8 I94 8224 880 4

880 8 83944 8 S94 84784 CH I. I 84798 63948 8 s98 84788 CH

-2 880 8 880 IA 74l ACC l-2 ACC 4

HBSOBD 8 560 RCS LOOP 4 8948D 8819D 822D SSISD BSIBB F RCS ti gl HOT LEG5~

I.2

~ceo 4

H 8922A 892 IA 89 I 94 ee2IA 89204 SI I-I Seois 892 4

tl H

8925 H88088 89568 ACC l-3 HSBOBC RCS LON'

'RCS LOOP 3 8948~

8 l98 8228 SSISC 822C 88 I SC 88 H

RCS 89ISB HOT LEGS~~

3, ~

88028 S92ls 89228 SI I-2 89238 88048 8976 e

56G H C, PRT 88564 Ce5.

%40CR 3,4 FG1 365 CCII HDR 4

~ceSe HEADER 2.4 RCS HOT 8 404 RCS HOT 8 408 H

COHT SltttP sceTEIIS n 0 HCV 638, EAT xCHCR ST I6A LETOOVII H eries 89824 tl 0

PRT Hcv 637

'8568 87344 87348 T

FCV 364 87248 87284 8730A RHR

-I 7 2es 873os RHR l-2 8 004 87008 Slot dNt 4940 RCS LOOP 4 HOT LEG 870t 8 F 02 It+0, I30)882624.ESS II-2S.88 HAIL 69828 plAI)RADII F, S-I SIIFF r 4

OF S

CV88I8A CV8848A E

I

%Q TRAlNA OfSCHAAGE 8809A CV8948b II8IHXt OSCHAROE IICVar b

IIIRTRNNb OISCHARGE 88098 0

0 (I) ASS~ NO LMI IS RUtTUREO. TIIEREFORE ONE OF THE REMlWINO THREE COI.O LEGS REOUlREO FOR SUCCESS OF I. I FlGURE E.5-22.

BLOCK OlAGRAH FOR TOP EVENT Ll

'h 4

V Boundary condition designator:

LI1 Hardvare unavailability cutsets due to indcpcndent failures; HWI =

Case A

Success Criterion:

1/3 Rupture in Cold Leg 1

C p

Case B

Success Criterion:

1/3 Rupture in any of the Four Cold Legs cocci L

":.K L

p o

1 ~

.1

=

Ci L

KC

~ LI

'<v

+

1 ~ 4~i-=-i" LIEK<

P LTEKc

+

h

~l 47t/C-0p LT"Yn LI"KF

+

CP'"

+5>

p L

c (

LTgKQ

~ o7 W<<Ilp C l D

L c

p OPh L cYD P

L aKJ

+ ct~

Y4<<K

+@5

~

~

~ ~

+ c(3

+

p ~.4Q7 7 Ll<<YQ ~ LI<<KF + gi 14 c ~

+ ct.z.

1.

cw4 16 6 t49>2=

CP 17

&.~ 49" 2= - 6'.

9K L FKI +~ S LI":Q + LT":KG +c.A LIDK":

LIP<I

+ CI+

18

& ~ 4~7 2:

C0 LIQKt

+

LTC KY

+ cA 'f 9

f '447 2h a Cap 0

be4c 2:-0<

LIBK";.

LI;

~ "J p col LI--Y>

~

LTB (L

+ c.~)

TOP EVENT A UPER COMPONENTS VOLIIHE <OHTROL TaHF.

REGEN HEAT EHCHANGER H LCV II28 RV57 ACC I-I 88084 H

89564 RCS 89ooa esloa RC5 LOOP I

69464 86 I 94 822A RCS LPO 2

69008 68 I08 RC5 890OC BSI RCS sooo elp E

SS ISA 820 880 la eeoIB 860 a

880 8 SIOI H SIOB H 83884 S475 647 4

8479A 47 II3944 8 694 84784 CH I-I 84798 8 ~ 7 8 83948 8 698 84'ISB CH

-2 H LCV II2C 6840 86094 8924 860 8

86074 COHTAINHENT Sf PaV 6 04A 874I AEC P

4 H8808D IC5 L~

89480 BSISP 822D BSI 80 H

RCS HOT LEGS~

I ~ 2 860 A

892 IA 89 I 94 H

8922A Sl I-I S92 4

Bso re 6925 ACC I-2 esose H

8 56P RCS LOOP 2 89468 BSIBB F 86 I 98 8228 83 862IA 682IB 89204 89208 H

H 69748 9744 RVST 6977 6976 BS568 ACC 0

1-3 Heeoec 8 56C RC5 LOOP 3 6946C BSI9C 822C SSISC H

RCS HOT LEGS~

68028 Cr5.

HEAOEH 3,4 89ISB 69218 89228 sl I 222 H

8604B~ Cr Sr HEADER 2r 4 RCS HOT 8 404 RCS HOT 87408 870 CONT SIIHP PRT 88564 H

C 6 0 A

87 I6A LETOOVH H

" p'TIGB 880 8

PRT 68568 69824 H

HCV 636 EAT XCHGR HCV 637 87344 87348 III FCV 344 D.1M m NOR 8 87248 FCV 64IB CCII HDR A FCV 365 8 24A 8728A 67304 RHR I

7268 87308 RHR I

2 8 004 67008 asa I 4740 RC5 LOOP 4 HOT LEG 8 70 l 6702 RECIRC PPZA lol PILI(F TAHAR 69828 Ilao. I30]682624. E99 II-28-88 IAIL PlACRAM F. l-I SHFFT

~ OF S

ACCUKMTM

'Col I

~CW~

CV~

I' cow KClAMAll7I 1~7 ICWIIIII CV~

JCCLMAATl&

l~1 1'C nese o

NSl%00$ LNfONPNNVON JCCVNAAfOO TNNi KW~

eY 07ÃO

~

FIGURE E.5-19.

BLOCK OlAGRAH FOR TOP EVENT AC

ENCLOSURE 5 PG&E SLIDES ON RHR CROSSTIE VALVE IMPORTANCE

RHR CROSS-TIE VALVEIMPORTANCE o

LLOCA INITIATINGEVENT o

1/3 SUCCESS CRITERIA o

CORE DAMAGE FREQUENCY CHANGE LESS THAN 1E-8 o

2/3 SUCCESS CRITERIS o

CDF CHANGE LESS THAN 1E-7

0

i~*IN ~

~

~b I.AlE5 a6 21$

21 gm II' d

m

~

1I

(.1-O(i Za -8 2eO

,; I,m I

4 I

01O.-Ig I.'llg4 FA l.Ctp 4

= 0 Z52.

&)

S s 3

I d

dm d

&IVES l

g ( I/<)

4s, ce)

D<b [I ichor

'b 5PIiII(

Liiv YV 4 Li VALu 2()

oP 1teo 1'gG 5

IpH2 IO NfAT'L i+

(ie 2.

,Oq i -2.5(.te t 6-'t

~IWmm <4 lemm Cmm4.

tC M

P oAL be 5'E3

0 0

ENCLOSURE 6 1

PGSE

SUMMARY

OF RELAY CHATTER ANALYSIS METHODOLOGY

RELAY CHATTER Objective

'-During an earthquake, relay chatter could impact the availability of components required to maintain the reactor in a safe shutdown condition.'he objective of

~

this analysis is to';

Identify contacts that affect components required for safe shutdown.

2.

Determine which contacts are subject to seismic relay chatter.

3.

Determine the consequences, if any, of contact chatter.

4. If contact chatter is a problem, determine how the operator can diagnose the problem.

5.

When the problem is diagnosed, determine the means available for the operator to correct the problem.

~Sco e

The systems and components included in this analysis are limited to those identified by the DCPRA systems analysts as important to risk.

Specifically, the component list is based on DCPRA letter PLG-123 (Chron 8500944).

Human factors, such as recognition and diagnosis of relay chatter consequences, are not included in this analysis.

0846E/0029E-1

+%

'%i+8

Page 2

Methodolo III 1;

System and Components:

Although the systems and major components are preselected, the system piping, instrument, and electrical schematics were reviewed to identify support systems and components that are required to function.

2.

Electrical Schematics:

Based on the component list, the appropriate electrical schematics were selected.

The latest drawing revision in the key Electrical drawings binder was used.

3.

Component Function:

The schematics were reviewed to determine the initial component state and the state required to accomplish reactor shutdown or other safety function.

Where the component had multiple "safe" positions, the analysis considered all required states.

4.

Contact Identification:

The contact paths are identified that cause the component to assume an undesirable position, prevent operation on demand, or cause degraded safety function.

For example, if a valve is required to close to perform its safety function, only the "open" contact path will be evaluated.

The relay models and types for the contacts to be evaluated are determined.

Contacts not subject to chatter, such as rotary or disk-type relays

, are assumed to remain in their initial position.

All other 1

contacts are assumed to move to the position that causes the component to move to the non-safe condition.

All disk-type relays have a bale installed on the disk spring to prevent vertical movement of the spring that could short the disc contact to ground; 0846E/0029E-2

S ~

Page 3

5:

Conse uence Evaluation: If the contact chatter results in the component reaching a non-safe position, the circuit is analyzed to determine whether the condition is intermittent (i.e., the component returns to the initial state after the earthquake) or latched (i;e., requires automatic or operator intervention);

6.

Reset:

.If the component remains in an undesirable condition after the earthquake, the simplest means for reset is determined.

Reset may be performed automatically (i.e., Solid State Protection System) or by operator intervention (i.e., resetting lockout or protective relays).

The reset method and location are identified.

~Aproach Simplifying assumptions were made for ease of analysis.

1.

Plant Condition:

Prior to the earthquake, the plant is assumed to be operating at 10(C power in a normal configuration.

Standby systems are considered to be off (i.e.,

no testing).

2.

Seismic Trip:

Reactor trip is assumed to occur during the initial seismic acceleration.

Relay chatter cannot prevent reactor trip and is likely to guarantee that the breakers trip.

3.

Circuit Breakers and Contactors:

Circuit breaker, manual transfer switches, and contactor main contacts are designed to withstand short circuit magnetic forces.

Therefore, these contacts are considered immune to seismic chatter.

The auxiliary contacts, however, are considered susceptible to chatter.

0846E/0029E-3

a I

r w't

Page 4

4.

Solenoid Yalves:

Solenoid valves are designed to vent air to accomplish the safety function.

Therefore, solenoid valves were not analyzed since chatter would tend to vent air.

However, control circuits for the solenoid valves were analyzed.

5.

Operator

Response

Because of the short duration of an earthquake and the likelihood that the operator may be temporarily incapacitated, the analysis assumes that the operator will not respond to intermittent alarms or false indications during the earthquake.

After the earthquake, the analysis assumes that the Solid State Protection System (which is not subject to chatter since the output relays are MDR rotary relays) actuates the required equipment that is not in a lockout condition due to contact chatter.

6.

Contact Welding:

In certain circuit configurations with large reactive loads such as coils, contact arcing occurs.

This arcing occurs primarily in DC circuits since there is not a voltage zero interruption point as in AC circuits.

In worst case situations the arcing may melt contact material, causing the contacts to we)d.

However, contact welding is not considered significant because relay contacts bounce during normal operations.

Thus, contacts are designed to, withstand arcing.

In addition, designers include arc suppressors on coils to mitigate the effects of inductive transients on contacts.

7.

Arcing Overvoltages:

In rare cases on lightly-loaded control circuits, cable capacitance may be sufficient to generate overvoltages during contact chatter.

These overvoltages are unlikely to damage equipment; however, the possibility exists that the overvoltage may induce sufficient voltage in 0846E/0029E-4

0

Page 5

adjacent circuits to energize an unassociated relay.

This condition (if it occurs at all) is transient in nature and cannot electrically latch a relay.

This scenario could theoretically occur in control circuits, but has not been observed at Diablo Canyon during normal contact bounce conditions.

Because of the rarity of configuration, the transient nature of the effect, and its absence in normal operation, transient overvoltages in control circuits are considered unlikely and are not considered in this analysis.

8.

Transformer Cooling:

Forced air fan cooling is required to extend large transformer ratings beyond the self-cooled (natural oil circulation) ratings.

The cooling fan control circuits are not analyzed because shutdown loads are normally within the transformer's self-cooled rating.

In worst

case, exceeding the transformer's self-cooled rating without fans would not disable the transformer.

The only consequence would be a slight reduction in transformer life.

9.

Reversing Starters:

Safety-related starters at Diablo Canyon are designed with a mechanical and electrical interlock to prevent the main contacts sets from closing simultaneously.

Therefore, the control circuits cannot cause a

short circuit due to both contactor sets closing simultaneously.

A similar concern is the possibility of tripping the thermal overload due to frequent reversing of the motor (for those valves lacking a seal-in contact).

When a motor is started and then reversed, the starting current may exceed twice normal starting current because of the motor back EMF and rotor inertia.

During chatter, however, this scenario is not likely to occur.

This is because relay chatter is, by definition, a cyclical 0846E/0029E-5

Page 6

phenomenon.'ts cyclical nature ensures that the contact will be closed for only a portion of the cycle.'ven assuming 25% close time, the motor is de-energized for 75% of the chatter cycle.

This provides sufficient time for the back EMF to discharge and rotor inertia to decrease.

Thus reversing overload relay operation is not considered credible for this analysis.

10.

MOY Pressure Boundaries:

Motor operated valves (MOVs) are used to isolate some pressure boundaries at Diablo Canyon.'hese valves are electrically disconnected under normal plant operation.

If they were electrically connected, contact chatter is not expected to cause them to crack open.

Contact chatter should close the contact only during a portion of the electrical cycle.

Since these valve actuators are driven beyond the seal point during closing by a torque switch, the motor must run for approximately 1/10 second or more before the valve cracks open.

However, the chattering contact does not remain closed long enough to drive the motor to crack the valve.

At the other end of the chatter cycle, the motor is driven closed again by a greater amount.

Based on this argument, MOVs are not considered to breach the pressure boundary unless a seal-in contact drives the motor through a complete cycle.

11. Earthquake Duration:

In accordance with IEEE 344-1975, the earthquake duration is assumed to be 30 seconds.

TR.~Tii 21:

2 fg 2

dgy 1

2222, 2

f1g be de-energized for sufficient time to reset the timing mechanism.

There-fore, timing relays with settings greater than 1 second are not considered de-energized.

However, this does not preclude the timing relay's output 0846E/0029E-6

, \\

~

E1 i4 Page 7

contacts from chattering in response to the seismic input; 13; Rotary Relays and Control Switches:

Certain relays such as the Potter-Brumfield MDR (Solid State Protection System), the Westinghouse W-2 control switch, and most lockout relays, have a rotary actuator.'his type of actuator is immune to contact chatter because vertical or horizontal seismic motion cannot move the actuator in a polar motion.

Therefore, rotary relays and switches are not considered to chatter in this analysis.

14. Non-Relay Contacts:

Non-relay contacts are included in the analysis.

This category includes the following control devices:

(

QO Control Switches (non-rotary)

Pressure Switches Temperature Switches Flow Switches Circuit Breaker Auxiliary Contacts Contractor Auxiliary Contacts Overload Relay Trip Contacts

15. Auto Start:

Components with an "auto" position on the control switch are assumed to be in that position.

16. Control Switch Position:

Control switches may be designed to be either main-tained (switch remains in the position it is set) or non-maintained (a spring returns the switch to neutral).

Control switches are assumed to be in the neutral position for non-maintained (spring return) switches.

Maintained contacts are assumed to be in the position matching the component position.

0846E/0029E-7

I

( ~

4

ENCLOSURE 7

BNL LETTER COMMENTING ON MEETING AND AUDIT

,Li '"

w F

Ng lf /

1$