---',<cczr.za~rzD Drsraraurrox Der ror Sra~rroz SvSrz~wREGULATO INFORMATION DISTRIBUTIONSTEM (RIDS)

"ACCESSION NBR: 9209210173 DOC ~ DATE: 92/09/11 NOTARIZED: NO DOCKETFACIL:50-275 Diablo Canyon Nuclear Power Plant, Unit 1, Pacific Ga 05000275

AUTH.NAME AUTHOR AFFILIATIONSOKOLSKY,D. Pacific Gas & Electric Co.RUEGERFG.M. Pacific Gas & Electric Co.

RECIP.NAME RECIPIENT AFFILIATION

SUBJECT: LER 92-015-00:on 901112,inadequate maint of Hosgri reptcommitments. Caused by thorough design reviews & 10CFR50.59SE not performed for revs to certain design documents. Designdocuments & certain seismic files revised.W/920911 ltr.

DISTRIBUTION CODE: ZE22T COPIES RECEIVED:LTR I ENCL ~ SIZE:TITLE: 50.73/50.9 Licensee Event Report (LER), Incident Rpt, etc.NOTES:

RECIPIENTID CODE/NAME

PD5 LAROODFH

INTERNAL: ACNWAEOD/DOAAEOD/ROAB/DSPNRR/DLPQ/LHFB10NRR/DOEA/OEABNRR/DST/SELB SDNRR/DST/SPLBSD1

G~IIRGN5 FILE 01

EXTERNAL EG&G'RYCEFJ HNRC PDRNSIC POOREFW

COPIESLTTR ENCL

1 11 1

2 21 12 21 11 11 11 1l. 11 1

2 21 11 1

RECIPIENTID CODE/NAME

PD5 PD

ACRSAEOD/DSP/TPABNRR/DET/EMEB 7ENRR/DLPQ/LPEB10NRR/DREP/PRPB11NRR/DST/SICB8H3NRR/DST/SRXB 8ERES/DSIR/EIB

L ST LOBBY WARDNSIC MURPHYFG.ANUDOCS FULL TXT

COPIESLTTR ENCL

1 1

2 21 11 11 12 21 11 11 1

1 11 11 1

NOTE TO ALL "RIDS RECIPIENTS:

PLEASE HELP US TO REDUCE WASTE! CONTACT THF DOCUMENT CONTROL DESK.ROOM Pl-S7 (EXT. 504-2065) TO ELIMINATEYOUR NAME FROM D!STRIBUTIONLISTS FOR DOCUMENTS YOU DON'T NEED!

FULL TEXT CONVERSION REQUIREDTOTAL NUMBER C" COPIES REQUIRED: LTTR )1 ENCL 31

Pacific Gas and Electric Company 77 Beale Street

San Francisco, CA 94106415/973-4684

Gregory M. RuegerSenior Vice President andGeneral ManagerNuclear Power Generation

September 11, 1992

PG&E Letter No. DCL-92-198

U.S. Nuclear Regulatory CommissionATTN: Document Control DeskWashington, D.C. 20555

Re: Docket No. 50-275, OL-DPR-80Docket No. 50-323, OL-DPR-82Diablo Canyon Units 1 and 2Licensee Event Report 1-92-015-00Inadequate Maintenance of Hosgri Report Commitments

Gentlemen:

PG&E is submitting the enclosed voluntary Licensee Event Report (LER)concerning inadequate maintenance of Hosgri Report commitments. Thisreport is submitted for information purposes only as described in Item19 of Supplement 1 to NUREG-1022.

PG&E met with NRC Region V personnel on June 22, 1992, to discuss HosgriReport commitments related to seismic qualification of equipment atDiablo Canyon Power Plant. At this meeting, PG&E stated that it wouldsubmit the enclosed voluntary LER.

The events discussed in this LER did not affect the health and safety ofthe public.

Sincerely,

Gregor H. Rueger

cc: Ann P. HodgdonJohn B. HartinPhilip J. HorrillHarry RoodCPUCDiablo DistributionINPO

DCO-90-EN-N027

Enclosure

5853S/85K/NWZ/2246

9209210173 920911PDR ADDCK 050002755 PDR

ICENSEE EVENT REPORT (LE+IFACIUTYNAME (11

DIABLO CANYON UNIT 1

TITLE(4I INADEQUATE MAINTENANCE OF HOSGRI REPORT COMMITMENTS

DOCKET NUMB(R 2 PACE 3

0 5 0 0 0 2 7 5 1" 12

EVENT DATE (Sl LER NUMBER (Sl REPORT DATE (TI OTHER FACILITIES INVOLVED (B)MON DAY SEOUENTIAL

NUMBERREVISIONNUMBER

MON DAYFACIUTYNAMES

DIABLO CANYON UNIT 2

DOCKEt NIAVIBIR(Sl

0 5 0 0 0 3 2 3

11 12 90 92 — 0 1 5 0 0 09 11 92 0 5 0 0 0OPERATINGMODE (9)

THIS REPORT IS SUBMITTED PURSUANT To THE REQUIREMENTS OF 10 CFRI (11)

POWERLEVEL

1 0 010 CFR

X OTHER VOLUNTARY

(Specify in Abstract below and in text, NRC Form 366A)

UCENSEE CONTACT FOR THIS LIR (12I

DAVID SOKOLSKY, SENIOR NUCLEAR GENERATION ENGINEER

COMPLETE ONE UNE FOR EACH COMPONEN't FAILURE DESCRIBED IN TIES REPORT (13l

AREA CODE

415 973-9717

T PHON NUM R

CAUSE SYSTEM CONPOHENT MANUFAC-TURER

REPORTABLETo NPRDS

CAUSE SYSTEH COMPONENt HAHUFAC-TVRER

REPORTABLETo NPROS

?;gM."'";>;.?'I>,".\II

SUPPLEMENTAL REPORT EXPEC'TED (14)

YES (if yes, complete EXPECTED SUBNISSION DATE) i X ( NO

ABSTRACT (16)

EXPECTEDSUBNISSIONDATE (15)

HONTH DAY YEAR

This voluntary LER is submitted for information purposes only as described in Item19 of Supplement 1 to NUREG-1022.

On November 12, 1990, PG&E determined that the boric acid tank level transmitters inplace at that time did not have a seismic qualification file as required by the HosgriReport.

The Hosgri Report and equipment seismic qualification files were reviewed to determineif other equipment files were similarly deficient. Discrepancies were found in someactive valve, instrument, and electrical qualification files, in the methodology usedfor passive valve qualification, and with battery operated light (BOL) qualification.PG8E has performed analysis or testing which verifies that the identified components arecapable of performing their required functions following a Hosgri earthquake.

The root causes of these findings were determined to be: (1) the classification systemthat defines component design requirements did not in all cases convey Hosgri Reportrequirements; and (2) thorough design reviews and 10 CFR 50.59 safety evaluations werenot performed for revisions to certain design documents.

Corrective actions include: (1) revising certain seismic qualification files; (2)implementing a program to control special qualification requirements of Design Class IIequipment; and (3) revising design documents and the UFSAR to reflect Hosgri Reportseismic qualification requirements.

5853S/85K

LICENSEIVENT REPORT (LER) TEXT COINUATION

FACILITY NAHE (I)

TEXT (17)

DIABLO CANYON UNIT 1

DOCKET NUHBER (E)

0 5 0 0 0 2 7 5

YTAR

92

LER NIIHBER

SIOVCNklNUMBOI

015

REVSIO N

0 0

PAGE 3

2 " 12

I. Plant Conditions

Units 1 and 2 have been in various modes and at various power levels withthe conditions described below.

II. Descri tion of Event

A.

B.

Summary:

On November 12, 1990, PG&E determined that the boric acid tank(BAT)(CA)(TK) level transmitters (CA)(LT) in place at that time didnot have a documented seismic qualification as required by PG&Elicensing commitments. Further review identified additionalcomponents lacking seismic qualification documentation.

This event and subsequent findings were not considered to bereportable because engineering assessments indicated that seismicqualification files could be prepared for all components without theneed for plant modifications.

Background:

The original design basis for Diablo Canyon required consideration ofa design earthquake (DE) and double design earthquake (DDE), whichcorrespond to the operating basis earthquake (OBE) and safe shutdownearthquake (SSE) as defined in 10 CFR 100, Appendix A.

A subsequent reevaluation of the plant to a different magnitudeearthquake was documented in the Seismic Evaluation for Postulated7.5H Hosgri Earthquake (Hosgri Report), submitted as amendments to theDiablo Canyon operating license application during the period fromJune 1977 through June 1980. The Hosgri reevaluation was performedusing criteria and methodology that were different from the originalseismic design basis. As a result, variations exist between theDE/DDE analysis and the Hosgri analysis regarding the scope ofqualification, load combinations, and allowable stresses.

With respect to safe shutdown, the original seismic design basis forDiablo Canyon required capability to achieve hot shutdown conditions.During the Hosgri reevaluation, additional components were identifiedand seismically qualified to demonstrate the capability to achievecold shutdown and maintain the plant in a safe condition after apostulated Hosgri earthquake, assuming a single active failure andloss of offsite power. Some of the identified components qualified tothis new criteria, including the BAT level transmitters, were DesignClass II components that normally do not require seismicqualification.

5853S/85K

LICENSEVENT REPORT (LER) TEXT COSINUATION

FACILITY NANE (I) DOCKET NONSER (2) LER NUMBER

VTAR ' RECONNVMOOI

PAGE 3

TEX'I (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 3 " 12

C.

The NRC review of PG&E's Hosgri reevaluation is documented in DiabloCanyon Supplemental Safety Evaluation Reports (SSERs) 7, 8, and 9.

Event Description:

On November 12, 1990, PG&E determined that the files for the DesignClass II BAT level transmitters did not contain analyses or testresults to demonstrate the required seismic qualification for theHosgri earthquake.

Previously installed transmitters had been qualified for Hosgri loads,but replacement transmitters were installed by design changesperformed in 1987 (Unit 2) and 1988 (Unit 1) without confirmingqualification for Hosgri loads.

As a result, PG&E initiated Nonconformance Report DCO-90-EN-N027 andperformed investigative actions to determine if additional componentsexisted that required seismic qualification as a result of HosgriReport commitments, but whose qualification had not been maintained.This investigative action identified the following findings andissues.

Other Instrumentation

~ Seismic qualification files had not been maintained for thesteam generator pressure indicators (SB)(PI), charging pump(CB)(P) discharge header flow transmitter (CB)(FT), andfirewater storage tank (KP)(TK) level indicator (KP)(LI). Thisinstrumentation was listed as requiring qualification to theHosgri earthquake. Files have been revised to documentqualification for the Hosgri earthquake; no plant modificationswere required.

Batter 0 crated Li hts

~ The seismic qualification of battery operated lighting units(BOLs)(FH)(LF) was not maintained in accordance with HosgriReport commitments. Seismic testing has verified that theseBOLs will remain functional following the Hosgri earthquake.

Valve uglification~ Twenty-one active valves (V) per unit were identified for which

the Hosgri load case qualification had not been fullymaintained. The valves were modeled in the piping systemanalysis for the DE, DDE, and Hosgri earthquake load cases;however, valve qualification to active valve stress limits forthe Hosgri earthquake was not documented. The files have beenverified and revised where necessary to document qualification

5853S/85K

LlCENSE/ENT REPORT (LER) TEXT COIINUATION

FACTLITT NAHE (1)l

DOCKET NUN8ER (2) LER NUN8ER

SCDUOIRAL RtVlSKlN

PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 4 '" 12

to active valve stress limits; no plant modifications wererequired.

Eleven containment HVAC valves (VA)(V) per unit were identifiedfor which formal seismic qualification files had not beenmaintained. The files have been revised to documentqualification; no plant modifications were required.

Two reactor coolant system (RCS) sample valves (AB)(SMV) perunit were identified that had been designated as passive valves,but are required to be active to meet Hosgri Report commitments.The files have been revised to document qualification; no plantmodifications were required.

The Hosgri Report established stress criteria to be applied tothe extended upperstructures of passive valves. Valves inaccident mitigating systems that are passive for the Hosgriearthquake were qualified by modeling the valves in the pipinganalysis and demonstrating compliance to pipe stress criteria.This was consistent with industry practice at that time and wasbelieved to be sufficient to assure passive qualification.Additional analysis has been performed and files have beenrevised where necessary to document qualification to passivestress criteria for these 127 valves. No plant modificationswere required, which validates the original approach.

During review of the Hosgri Report, certain ambiguities in theHosgri Report and the SSERs regarding qualification requirementsfor valves in accident mitigating systems were identified.Table 7-1 of the Hosgri Report defines active components as"Mechanical equipment which is needed to go from normal fullpower operation to cold shutdown following the earthquake andwhich must perform mechanical motions during the course ofaccomplishing its design function." Tables 7-7 and 7-7A of theHosgri Report list valves required to be qualified to activevalve criteria for cold shutdown of the plant. SSER 7 refers toTable 7-7 in describing those valves required to be qualified asactive valves. These specific statements are consistent withwhat was done in the PGEE, NSSS, and other contractor valvequalification files and calculations.

However, other parts of the Hosgri Report and SSERs make generalstatements regarding valve qualification. These generalstatements can, be construed as implying that remotely operatedaccident mitigating system valves were also qualified to activevalve criteria for the Hosgri earthquake, as was the case forthe DE (OBE) and DDE (SSE). This indicates that the specificcriteria established during the Hosgri reevaluation andreflected in associated SSERs, along with the engineering

5853S/85K

LICENSEIVENT REPORT (LER) TEXT COINUATION

FACILITY NAME (I) DOCKET NUMBER (2)YEAR

LER NUMBER

REYIQOII

PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 5 " 12

actions taken during that period, were not clearly reflected inthe licensing documentation.

To provide increased conservatism, qualification files for thesevalves will be revised to document qualification to activecriteria for the Hosgri earthquake (see Section V.C. 1.).

Other Issues

Three switches (JL)(IS), two fuses (JL)(FU), and a fuse block(JL)(FUB) that require seismic qualification for Class 1Ecircuit integrity were identified in the dedicated shutdownpanel (JL)(PL). The panel, its anchorage, and the electricaldevices had not been seismically analyzed. The necessaryqualification files for this equipment have been prepared and noplant modifications were required. The root cause analysis ofthis finding will be addressed by the Regulatory Guide (RG) 1.97Review Program presently being performed by PG&E as a part ofNonconformance DCO-91-EN-N005. This issue is riot discussedfurther in this LER as it is not related to Hosgri Reportcommitments.

Position switches (JL)(33) internal to various safety-relatedvalve motor-operators were inappropriately designated as DesignClass II. However, this finding has had no adverse impact ondesign, procurement, and maintenance of the devices since thevalves and operators are designated as Design Class I andinternal parts have been managed as safety-related components.Drawing changes are being initiated to reclassify thesecomponents. This issue is not discussed further in this LER asthis finding does not impact the seismic qualification of thesecomponents.

Remote valve operators on some Design Class I valves have beendesignated as Design Class II. These valves: (1) have apassive function; (2) are only relied upon for manual handwheeloperation in accident analyses; or (3) are in the designatedcold shutdown flowpath that was evaluated for the Hosgriearthquake. Appropriate seismic analyses had been performed forthese valves. However, the Class II designation for the valveoperators does not assure configuration control of plantmaintenance activities for this unique situation. Therefore,drawing changes are being initiated to reclassify the valveoperators to Design Class I. As part of the design changeprocess, PG&E will verify that appropriate maintenanceactivities have been performed on these valve operators. Thisissue is not discussed further in this LER as there is noevidence that maintenance practices have impacted the seismicqualification of these operators.

5853S/85K

LICENSEINVENTREPORT (LER) TEXT COIQINUATION

FACILITY RANE (1) OOCKET NUHSER (8) LER NIRISER 6YTAR;: SEOVOITNL

IAIMSOI

PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 6 " 12

D. Inoperable Structures, Components, or Systems that Contributed to theEvent:

None.

E. Dates and Approximate Times for Major Occurrences:

June 3, 1977: Hosgri Report issued as Amendment 50 to theoperating license application. The report wasamended numerous times, with the last amendmentissued in June 1980.

June 22, 1987: Event Date. Design change failed to maintainevidence of seismic qualification for BAT leveltransmitter in Unit 2.

January 14, 1988: Event date. Design change failed to maintainevidence of seismic qualification for BAT Leveltransmitter in Unit 1.

November 12, 1990: Discovery Date: PG&E determined that Units 1 and2 BAT level transmitter files did not includeevidence of qualification for the Hosgriearthquake. PG&E initiated Nonconformance DCO-90-EN-N027.

July 3, 1991:

Hay 26, 1992:

Review of Hosgri Report and equipmentqualification files identified additionalpotential inconsistencies.

Investigative actions completed and scope ofcorrective actions defined.

June 22, 1992: PG&E met with NRC Region V to discuss thenonconformance on Hosgri Report commitments. PG&Eagreed to submit a voluntary LER.

G.

Other Systems or Secondary Functions Affected:

None.

Hethod of Discovery:

During review of a design change to replace the BAT leveltransmitters, Hosgri Report seismic qualification requirements for thetransmitters were identified,

5853S/85K

LICENSEI/ENT REPORT (LER) TEXT COINUATION

FACILITY RANE (I) DOCKET NIGGER (E)YCAR

LER NURSER 6STGVUIRAL

PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 7 " 12

When the design change was reviewed for Seismically Induced SystemInteraction (SISI) considerations, it was identified that the SISIProgram considers the Design Class II BAT level transmitters astargets (which indicates that they are required for safe shutdownafter an earthquake). This led to a review of the Hosgri Report,which lists the BAT level transmitters as requiring seismicqualification for the Hosgri earthquake. The transmitters installedat that time did not have a seismic qualification file.Subsequent review of the Hosgri Report identified additional seismicqualification commitments that had not been maintained.

H. Operator Actions:

None.

I. Safety System Response:

None.

III. Cause of the Event

A. Immediate Cause:

The conditions that directly affect the seismic qualification ofequipment within the scope of this LER are the result of severalimmediate causes. The causes are listed by equipment type as follows.

BAT level transmitters: PG&E implemented design changes in 1987 and1988 to replace the original, seismically qualified BAT leveltransmitters. The Hosgri Report seismic qualification requirement forthe transmitters was not recognized because the requirement was notreflected in design documents.

Other instrumentation: PG&E implemented design changes to reclassifyinstrumentation to conform with RG 1.97 requirements that wereinconsistent with the seismic qualification requirements of the HosgriReport.

The firewater storage tank level indicator was seismically qualifiedto maintain system pressure boundary integrity; however, therequirement for functional qualification was not recognized becausethe requirement was not reflected in design documents.

Batter o crated li hts: The BOLs were installed as a result of thepre-10 CFR 50, Appendix R fire protection review and were seismicallyqualified as documented in the Hosgri Report in response to an NRCverbal request. When additional units were installed to comply with10 CFR 50, Appendix R fire protection requirements, the seismic

5853S/85K

LICENSEIVENT REPORT (LER) TEXT COINUATION

FACILITY NAHE (1) DOCKEt NVM8ER (E) LER NVH8ER PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 05 0 0 0 2 7 5 92 01 5 0 0 8 " 12

qualification requirement for the BOLs was deleted because Appendix R

does not require seismically qualified lighting units.

Valve ualifications: Requirements for Hosgri active qualification of21 valves per unit were eliminated in a design criteria revision. Thebasis for this revision could not be determined.

Passive valve qualifications for the Hosgri earthquake were notestablished based on extended upperstructure stress allowables asstated in the Hosgri Report. Instead, valves were modeled in thepiping system and accepted based on compliance with piping stressallowables. This methodology was consistent with industry practiceand was considered to be appropriate for passive valve qualification.

Eleven containment HVAC system valves per unit were not included inthe valve list for identifying valves requiring qualification. As aresult, qualification files verifying that actual valve accelerationswere less than allowables were not established for these valves.

Two RCS sample valves were improperly listed on the valve list aspassive instead of active.

B. Root Cause:

The following two root causes have been identified for these findings.

A specific configuration control system was not established toidentify Hosgri seismic qualification commitments for equipment whoseclassification would not normally require seismic qualification. Thisapplies to the BAT level transmitters and the other instrumentation.

Seismic qualification requirements for BOLs were deleted as the resultof a design document revision that did not receive a thorough designreview and 10 CFR 50.59 safety evaluation. Similarly, seismicqualification requirements for valves were incomplete andinappropriately modified as the result of design basis documents thatdid not receive thorough design reviews and 10 CFR 50.59 safetyevaluations. At that time, design basis documents were not consideredto be within the scope of 10 CFR 50.59.

IV. Anal sis of the Event

The event analysis for each equipment type is as follows.

BAT level transmitters: The non-qualified level transmitters installed in1987 and 1988 have been evaluated and no failure modes were identified thatcould have prevented boration from occurring or adversely impacting anyother safety functions. The transmitters had no moving parts andseismically induced failure of the transmitters was unlikely. PG&E believes

5853S/85K

LICENSEINI/ENT REPORT (LER) TEXT COINUATION

FACILITY NAME (I) DOCKET NUMBER (2) LER NUMBER

BIOVIKIIALKVMBOI

IILYISOII„:I.."$ KVMSOI

PAGE 3

TEXT (IT)DIABLO CANYON UNIT 1 05 0 0 0 2 7 5 92 01 5 0 0 9 '" 12

that the transmitters could have been seismically qualified if necessary;however, other unrelated design considerations warranted their replacementin the fourth refueling outages for both units. The replacementtransmitters are seismically qualified.

Therefore, installation of the non-qualified transmitters did not result inany safety or operability issues.

Other instrumentation: The Hosgri Report required that additionalinstrumentation required for cold shutdown be seismically qualified.Seismic qualification file revisions to document the qualification of thesedevices have been completed; no plant modifications were required. Nosafety or operability concerns exist.

Batter o crated li hts: The commitment to seismically qualify the BOLsexceeds regulatory requirements for fire protection. The BOL unitspresently installed at Diablo Canyon are similar to the original units thatwere seismically tested in 1978, but utilize lead-acid batteries (FH)(BTRY)instead of the nickel-cadmium batteries (FH)(BTRY) installed in the originalunits. To comply with the commitment in the Hosgri Report, seismic testingof the BOL units was performed to qualify the lead-acid batteries. Ratherthan identifying and qualifying only those BOLs in specific areas requiringlighting for post-earthquake shutdown, PGEE chose to qualify all BOLslocated in the plant (the quantity of BOLs in the plant is substantiallygreater today than when the original test was performed). A seismic testspectra was developed that enveloped all plant locations with BOLs. Thisspectra resulted in a test condition substantially more severe than theoriginal qualification test. Two different mounting rack configurationswere used to envelope all BOL support conditions.

Because the BOL units were subjected to a more severe seismic test spectrathan in 1978, during the second front-to-back and vertical test, excessivebuckling deformation of the sheet metal mounting rack occurred. However,the BOLs remained functional during and after the test. In order tocomplete the test in accordance with IEEE-344 requirements, the BOL mountingracks were stiffened using aircraft cable and gusset plates. The seismictest was resumed and successfully completed. The seismic testing confirmedthat the lead-acid batteries are capable of withstanding the Hosgriearthquake. Upon completion of the seismic testing, functional tests forthe BOLs were successfully completed. Therefore, no safety or operabilityconcerns exist.

Mounting racks for the BOL units in those locations where the requiredresponse spectra exceeds the 1978 test spectra will be modified to conformto the tested configuration (see Section V.C.2).

Valve ualifications: PG8E has reestablished the Hosgri valve qualificationfor those valves considered to be active valves for the Hosgri earthquake.

5853S/85K

LICENSEIVENT REPORT (LER) TEXT COIQINUATION

FACTLITY KAME (1)1

DocKET KUM8ER (E) LER KOM8ER 6 PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 10 o" 12

PG&E has completed the analyses for the valves that require passivequalification for the Hosgri earthquake. No plant modifications wererequired. Therefore, no safety or operability concerns exist.

Additional evaluations, currently in progress, have the potential to affectthese valve qualifications. These evaluations are: (1) incorporation ofGeneric Letter 89-10 motor-operator thrust loads in valve qualificationanalyses; (2) evaluation of the effect of revised vendor supplied valveweight data; and (3) a reevaluation of valves that utilize cast iron in non-pressure retaining applications. These issues are being addressedindependently of the issues in this LER.

PG&E performed additional seismic evaluations and analyses that weredocumented in the Diablo Canyon Long Term Seismic Program (LTSP) FinalReport submitted to the NRC in July 1988. The NRC review of the LTSP isdocumented in SSER 34. The findings in this LER do not affect theconclusions from the LTSP. The Diablo Canyon LTSP established fragilityestimates for air and motor-operated valves from a sampling of valvesrequired for safe shutdown. The allowable acceleration level for thesevalves as identified in the design criteria document was taken to correspondto the acceptable level of stress of 1.8S ("S" is the ASHE Section IIImembrane stress allowable), which is the criterion for active valves. Onthe basis of this assumption, the high-confidence-of-low-probability-of-failure (HCLPF) capacity, when compared to the LTSP demand, provided aseismic margin of 2.21 (reference LTSP Final Report, Table 7-2). Valvesconsidered as passive components for the Hosgri earthquake have an allowablestress criteria of 2.4S. For the LTSP study, the HCLPF seismic margin forvalves qualified as passive instead of active, in those few situations whichmay approach the worst case, would be reduced to 1.65. This seismic marginis still in excess of the 1.4 margin established in the LTSP deterministicseismic margin assessment (PG&E letter DCL-90-226). Therefore, the findingsin this LER do not affect the conclusions from the LTSP.

V. Corrective Actions

A. Immediate Corrective Actions:

2.

3.

The BAT level transmitters were replaced with seismicallyqualified transmitters.

The Hosgri Report was reviewed to identify any other seismicqualification commitments that had not been maintained.

Files demonstrating qualification to the Hosgri earthquake havebeen established or updated as necessary to document the seismicqualification of the identified components, except for the BOLs.Seismic qualification files are expected to be completed for theBOLs by October 15, 1992.

5853S/85K

F.

LICENSEI/ENT REPORT (LER) TEXT COINUATION

FACILITY NAHE (1)

TEXT (17)

DIABLO CANYON UNIT 1

COCKET NVNSER (E)

05 0 0 0 2 7 5 92

LER RIBBER 6SKOVENTIAL

NVQI>OI >,Sx.

015

PAGE 3

0 0 11 " 12

B. Corrective Actions to Prevent Recurrence:

Specific controls are being implemented to ensure thatconfiguration control of Design Class II components that requireseismic qualification will be maintained. These include:

~ Design Class II components with seismic qualificationrequirements specific to the Hosgri earthquake will bereclassified to assure qualification requirements aremaintained.

2.

3.

~ Plant modification and replacement part evaluationprocedures have been revised to assure seismicconfiguration control.

Design documents are being revised to incorporate appropriateHosgri Report seismic qualification requirements. The designchange to revise these documents is expected to be issued byOctober 15, 1992.

The Updated Final Safety Analysis Report (UFSAR) will be revisedto reflect Hosgri Report seismic qualification requirements.Revisions will be included in the annual updates issued in 1992and 1993.

A review of the valve requalification effort will be performedby PG&E's guality Assurance Department to ensure compliance withdesign documents. The review is expected to be completed byOctober 15, 1992. Verification of all corrective actions isroutinely performed by the gA Department as part ofnonconformance closure.

5. Procedures governing design criteria document revision have beenrevised over the past several years to require thoroughtechnical review, design/safety review, independentverification, and 10 CFR 50.59 safety evaluation. Personnelperforming safety evaluations receive documented training basedon the content of NSAC 125, "Guidelines for 10 CFR 50.59 SafetyEvaluations."

C. Additional Actions:

Seismic qualification files will be revised to demonstrateactive qualification for the accident mitigating system valvescurrently designated as passive for the Hosgri earthquake. Thiseffort, which is outside the scope of the nonconformancecorrective action, will provide increased conservatism.Engineering analysis will be completed in 1993 and plantmodifications, if required, are anticipated to be completed

5853S/85K

LICENSElfENT REPORT (LER) TEXT CONUATION

t:ACILITY NAME (I) DOCKET NUMBER (2) LER NUMBER PAGE 3

TEXT (17)

DIABLO CANYON UNIT 1 0 5 0 0 0 2 7 5 92 01 5 0 0 12" 12

2.

3.

during the sixth refueling outage in each unit (scheduled in1994).

Mounting racks for the BOL units in those locations where therequired response spectra exceeds the 1978 test spectra will bemodified to conform to the tested configuration. Plantmodifications will be completed during the sixth refuelingoutage in each unit (scheduled in 1994).

Seismic qualification files have been prepared to document thequalification of the dedicated shutdown panel and the requiredelectrical devices within the panel. No plant modificationswere required.

Design changes will be initiated to reclassify (1) positionswitches internal to safety-related valve motor-operators and(2) valve operators on Design Class I valves that wereidentified as a result of the nonconformance investigativeaction.

VI. Additional Information

A. Failed Components:

None.

B. Previous LERs:

None.

5853S/85K