ML061720173

Response to NRC Request for Information Proposed Change Number (PCN)-548 Battery and DC Sources Upgrades and Cross-Tie

ML061720173
Person / Time
Site:	San Onofre
Issue date:	06/19/2006
From:	Scherer A Southern California Edison Co
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
PCN 548, TAC MD0300, TAC MD0301
Download: ML061720173 (11)

Text

ISOUTIERN CALIFORNIA A. Edward Scherer EDISON' Manager of Nuclear Regulatory Affairs An EDISON INTEIRNATIO\ALI Company June 19, 2006 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555-0001

SUBJECT:

Docket Nos. 50-361 and 50-362 Response to NRC Request for Information Proposed Change Number (PCN)-548 Battery and DC Sources Upgrades and Cross-Tie San Onofre Nuclear Generating Station, Units 2 and 3

References:

(1) Letter from Brian Katz (SCE) to Document Control Desk, dated February 28, 2006,

Subject:

"Proposed Change Number (PCN)-548, Revision 1, Battery and DC Sources Upgrades and Cross-Tie" (2) Letter from N. Kalyanam (USNRC) to Richard M. Rosenblum (SCE), dated May 16, 2006,

Subject:

"REQUEST FOR ADDITIONAL INFORMATION (RAI) ON THE PROPOSED TECHNICAL SPECIFICATION AMENDMENT REQUEST TO UPGRADE BATTERY AND DC SOURCES INCLUDING A BATTERY CROSS-TIE CAPABILITY WITH AN EXTENDED 30-DAY BATTERY COMPLETION TIME (TAC NOS. MD0300 AND MD0301)

Dear Sir or Madam:

This letter responds to the Request for Additional Information (RAI) from the NRC staff reviewers (Reference 2) regarding the subject license amendment request, PCN-548, originally submitted as Reference 1.

The responses are provided in Enclosure 1.

P.O. Box 128

,N~of San Clemente, CA 92672 949-368-7501 Fax 949-368-7575

Document Control Desk June 19, 2006 Ifyou have any questions or require additional information, please contact Jack Rainsberry at (949) 368-7420.

Sincerely,

Enclosure:

1. Responses to NRC Questions cc: B. S. Malleft, Regional Administrator, NRC Region IV C. C. Osterholtz, NRC Senior Resident Inspector, San Onofre Units 2 & 3 N. Kalyanam, NRC Project Manager, San Onofre Units 2 and 3

Enclosure 1 Response to NRC Questions

REQUEST FOR ADDITIONAL INFORMATION SAN ONOFRE NUCLEAR GENERATING STATION, UNITS 2 AND 3 AMENDMENT REQUEST TO UPGRADE BATTERY AND DC SOURCES INCLUDING A BATTERY CROSSTIE CAPABILITY WITH AN EXTENDED BATTERY COMPLETION TIME (TAC NOS. MD0300 AND MD0301)

Page references are from the license amendment request (LAR) application dated February 28, 2006, from Southern California Edison (SCE, the licensee) for San Onofre Nuclear Generating Station, Units 2 and 3 (SONGS 2 and 3).

1. The licensee's preventive maintenance risk estimates for the 30-day battery completion time (CT) for change in core damage frequency (ACDF), change in large early release frequency (ALERF), incremental conditional core damage probability (ICCDP), and incremental conditional large early release probability (ICLERP) are within the Regulatory Guide (RG) 1.174 and RG 1.177 acceptance guidelines for a very small change. However, risk estimates for ICCDP (1.87E-6) and ICLERP (4.OE-7) during corrective maintenance do not meet the RG 1.177 acceptance guidelines of 5E-7 and 5E-8, respectively, and require further justification for an extended corrective maintenance 30-day battery CT. Please include a discussion on common-cause evaluation and operator actions required for corrective maintenance and performance of the battery cross-tie.

RESPONSE: The risk estimatesfor ICCDPand ICLERP during correctivemaintenance were performed based on potentialcommon-causefailure of the remainingavailable batterieswith no credit givenfor inspection of the extent of condition of the remaining availablebatteries. In lieu of the higher incurredrisk ofpotentialcommon causefailure (CCF)of the other batteries,SONGS is committed to perform an operabilityassessment of the remainingavailablebatteriesto verify that a common-causefailure condition does not exist when any JE battery is rendered inoperabledue tofailure. By performing such an inspection and assessmentfor CCFand confirming CCFis not present,the corrective maintenancerisk would reduce to the preventive maintenance risk which meets the RG 1.177 acceptance guidelines. If CCFis presentfor other batteries,then SONGS would enter the appropriateTSfor multiplefailed batteries.

Perplantprocedures,maintenance electricianswill confirmfunctionality of other batteries(i.e., similarfailure has not occurredon other batteries). The results of the evaluation will be verified in an operabilityassessment to evaluate the extent of the condition. This evaluation will assess whether the other similar components are operable or inoperablefor (or susceptible to) the same common-cause.

Common-cause evaluation and operatoractions requiredforperformance of the battery cross-tie will be performed in accordance with plantproceduresand operating instructions. These proceduralsteps will include confirmation thatpriorto completing Page 1 of 8

the cross-tie that the subsystem to be cross-tiedto is available to support both subsystems.

2. Cumulative Risk Please provide a discussion on cumulative risk impact of previous changes and/or current risk-informed requests. The discussion should confirm SONGS' risk-informed changes are incorporated into the risk evaluation for the proposed extended battery completion time and that the SONGS' probabilistic risk analysis (PRA) includes the cumulative impact of these changes.

RESPONSE: All previous changes that have been implemented in the plant have been incorporatedin the PRA model used to assess this application. This includespreviously approvedrisk-informed allowed outage time extensions and the risk-informed in-service testingprogram. The actualrisk impacts ofprevious changes are included in the periodic PRA data update. There are no other risk-informedrequests that are currently awaitingNRC approval.

3. External Events The LAR does not provide an evaluation with respect to the potential risk impact of external events (seismic, fires, high winds, floods and other (HFO)). The evaluation as provided by the licensee in Reference 2 is limited to a discussion of external event peer review results for seismic and fire events. HFO events were not specifically addressed by the licensee in the submittal or the RAI. Please provide an assessment (qualitative or quantitative) as to the risk impact of external events (seismic, fire and HFO) on the proposed extended 30-day battery CT.

RESPONSE: All risk resultsprovided in the LAR quantitativelyinclude the impact of external events includingseismic andfire. It is SONGS continuingpractice to include the contributionsof internaland external events risk in all past andfutureapplicationsof PRA (includingall license submittals and configurationrisk managementpractices).

High winds, floods, and other external events were screenedwhen evaluatedfor the Individual PlantExaminationof External Events (IPEEE)[

Reference:

1PEEEFor SONGS 2 & 3, December 15, 1995]. In the examination, high winds andfloods were screened due to the high wind andflood capacity of thefacilities in relationto the low frequency of capacity challenginghigh winds andfloods. The scope of the high winds analysis includedhurricanesand tornados. The most likely event is a wind induced loss of offsite power, which is explicitly included in the loss of offsite power initiatingevent frequency. For more significant damage, it was concludedthat the only tornado capable of damaging safety relatedequipment and structuresis an F5 Fujitascale tornado,which has a returnfrequency of 8E-8/year. PerASME PRA Standard(ASME RA-Sb-2005, December 30, 2005) Requirement IE-C4, a screeningcriteria of JE-7/yearis used to eliminate initiatingeventsfromfurther evaluation. Therefore, high winds can contine to be screened. Additionally, a high wind sufficient to damage the structure housing a DC Page 2 of 8

battery/bus would also damage the adjacentroom housing the same train battery/bus whether it was cross-tied or not. Therefore, high winds risk does not impact this application.

Externalfloods causedby the probable maximum flood (PMF)and the probable maximum precipitation(PMP)events were assessed in the IPEEE. In the analysis,floods were deemed to be risk insignificantbased on the location ofpotential externalflood sources and the designedgradingof the land to preclude the PMP.

Other external events include transportationaccidentsfrom Interstate 5 and the adjacent railroadtracks. These risks are re-evaluatedon a tri-annualbasis based on the changingtransportfrequencies, shipment sizes, cargoesand accidentprobabilities. The most recent study [

Reference:

SONGS Units 2 & 3, 2005 Offsite Hazards Update, October 2005] confirmed that the asphyxiant, toxic hazard,explosive/flammable hazard frequencies are each less than JE-6/year. Because the DC buses, switches, and batteries are located in the auxiliarycontrol building with many interveningstructures (including the containment building)between the DC system and the highway andrailroad,low frequency transportationaccidents would not have a risk impact on the DC system cross-tie configuration.

In summary, seismic andfire risk estimates were quantifiedand explicitly included in the originalLAR andsubsequent RAI response. The risk of HFO in the 1PEEEwas reviewed and reconfirmedto be risk insignificantas it applies to this application.

4. Tier 2 Evaluation A specific Tier 2 analysis is not referenced in the LAR although a Tier 2 discussion with respect to compensatory measures is referenced in Reference 2. The licensee's Tier 2 evaluation concluded that no risk significant configurations exist based solely on the low risk increase of the proposed extended 30-day battery cross-tie CT. The Nuclear Regulatory Commission staff is concerned that the licensee's Tier 2 evaluation is not consistent with the guidance of RG 1.177, Section 2.3, "Evaluation of Risk Impact,"

Subsection, "Tier 2," which states that reasonable assurance must be provided that risk-significant plant equipment outage configurations will not occur when specific plant equipment is out of service. RG 1.177 states that one appropriate method to evaluate such combinations of equipment out of service is to use the Tier 1 ICCDP and ICLERP acceptance guidelines to identify risk-significant configurations associated with the proposed change. This information is required to ensure that the Tier 2 evaluation in Reference 2 meets the guidance of RG 1.177.

RESPONSE: SONGS performed a Tier 2 evaluationto identify potentialrisk significant configurationsgiven a DC bus subsystem is cross-tiedto a same trainDC bus subsystem.

The Safety Monitor software used by SONGS is able to assess and rank the importance of all remaining availablecomponents given a specific plant configuration. The ranking shows the riskfor each availablecomponent ifthat component were removedfrom service while in the given configuration. The lists of importantavailable componentsfor Page 3 of 8

the cross-tiedcase and the base case (no cross-tie with all batteriesavailable)were comparedto determine whether there are high risk configurationsunique to the cross-tied case. The comparison identified only one unique high risk configuration (i.e., > IE-3/year CDF)for the cross-tiedcase: removal of an opposite train battery chargerwithout creditedalignment of an installedand equivalentswing battery. Such a configuration is limited by the Technical Specification to 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> andper the SONGS MaintenanceRule Risk ManagementProgramwould not be entered into voluntarily (see response to RAI Question #6 below).

5. Reference 2 states that the seismic PRA requires no modeling changes for replacement batteries B009 and B010. However, batteries B007 and B008 are also being upgraded, which will also replace the existing battery racks. The licensee stated that the seismic PRA will be reviewed again when the upgrade is designed and implemented. Please identify if the additional seismic review is a licensee commitment per the LAR.

RESPONSE: SONGS commits to verify the seismic capacity of the replacement battery racks and to incorporatethe results into the SONGS PRA as appropriate.

6. The LAR discusses the use of the SONGS' risk monitor to avoid high-risk configurations but does not provide information on applicable acceptance guidance (Tier 3) used to identify a risk-significant configuration. Please provide this information to complete the review.

RESPONSE: The Tier 3 requirementfor a configuration risk managementprogram to assess risk when a risk-informed TS is appliedhas been incorporatedinto the SONGS Maintenance Rule Risk Management Program(MRRMP). This program is administered throughplantprocedure S0123-XX-10, "MaintenanceRule Risk Management Program Implementation." In the procedure,the following guidelines areprovided which control the instantaneousand cumulative incrementalrisk:

Page 4 of 8

(The following in an excerptfrom plantprocedureS0123-X"-10, "MaintenanceRule Risk Management ProgramImplementation," page 13 of 41:)

6.5 Action Thresholds Based on Quantitative Considerations 6.5.1 The established quantitative risk management actions thresholds consider temporary risk increases associated with the configuration-specific CDF, as well as the Incremental Core Damage Probability (ICDP) and Incremental Large Early Release Probability (ILERP).

6.5.2 Maintenance configurations with a configuration-specific CDF in excess of 10 3/year should be carefully considered before voluntarily entering such conditions. If such conditions are entered, it should be for very short periods of time and only with a clear detailed understanding of which events cause the risk level.

6.5.3 ICDP and ILERP, for a specific planned configuration, with respect to establishing risk management actions, should be considered as follows:

ICDP ILERP

> 10. - configuration should not > 101 normally be entered voluntarily 10-_ 10- - assess non quantifiable i0-7_ 10-1 factors

- establish risk management actions

< 10. - normal work controls < 10.1 In the MRRMP program,plant configurationswith an instantaneouscore damage frequency of greaterthan 1E-3/year and/ora large early releasefrequency of JE-4/year are high risk configurationsthat are not entered into voluntarily. In addition, the MRRMP procedureensures that the cumulative incrementalrisk ofplanned configurationsare managed to acceptable levels (as shown in the table of the procedural steps above).

7. Table 3, Line 7, on page 13 of Attachment I of Reference 1 estimates the single allowed outage time risk ICCDP as 1.87E-6. The risk estimate appears to be inconsistent with the guidance of RG 1.177 in that the estimate appears to use 15 days instead of the proposed CT of 30 days when estimating ICCDP. Note c for Table 3 states that battery Page 5 of 8

replacement and testing requires 30 days, which is given as the mean duration in Table 3, Line 10. Please explain the discrepancy.

RESPONSE: This comment is correct,the ICCDPshould be for 30 days with an ICCDP of 3.76E-6.

8. Reference 2 discusses additional technical specification (TS) changes including CT and surveillance test intervals (STIs) not related to the proposed risk-informed 30 day battery cross-tie CT. The RAI response qualitatively accepts the additional risk based on conformance to either TSTF-360 Revision 1, IEEE 450-2002, or with a specific risk assessment statement. Please discuss whether these additional STI, CT and TS changes were addressed in the modified SONGS risk estimate of the proposed 30 day battery crosstie CT.

RESPONSE. The additionalSTI, CT and TS changes were not quantitativelyaddressed in the modified SONGS risk estimate of the proposed30 day battery crosstie CT.

The majority of changes involve editorial,clarification,improved test methods and/or revisedformatchanges. However, there are eight changes that may have risk impact and they are consideredhere.

Three surveillances (i.e: monitoringfloat voltage,float currentandpilot cell voltage) are proposedto be performedevery 31 days instead ofevery 7 days as are presently performedper the current TS. The acceptabilityof the longer interval is based on the vast accumulatedexperience by the nuclear industry at large andSONGS in particular.We have observed that the weekly datagatheredduring thepast 20 years has remained virtually unchanged.

Figure1 below demonstratesthat the DC busfloat voltage measureddoes not significantly change over time (refer to the 10 month sample periodshown in Figure 1).

The float voltage values measuredweekly at the battery terminalswere essentially unchanged.This observation is typicalfor all 8 Class 1E batteries installedat SONGS.

The float voltage stayed between 131 V-132V, equal or greaterthan 2 Volts above the minimum Operable Voltage of 129V Also, based on industry experience and specific SONGS observation,the cell internal resistancestays constant, thereforefloat currentof a battery andthe measuredpilot cell voltage, due to the constant appliedfloat voltage, would remain constant over weekly and monthly observationperiods. Therefore, increasinginterval between these surveillances from 7 days to 31 days will not likely affect the calculatedrisk.

Page 6 of 8

Figure 1 - Float Voltage at the DC Bus 2D4 over a 10 Month History Similarly, afourth change involves a change in battery service testingfrom a 24 month frequency to a 30 monthfrequency to coincide with the 60 month modifiedperformance test. In practice,the modifiedperformancetest, which is a combinedservice and performance test, will be performed every 30 months. A modifiedperformance test is comprisedof two parts,thefirst part tests a battery by subjecting it to the most severe service loadprofile, and the secondpart is a constant currentcapacity discharge test which continuesuntil the acceptance criteria(1. 75Volts per cell (Vpc)) is reached This test is more rigorousthan the service test and will be performed consistently every 30 months which will provide a more consistent data basefor trending due to the performance of the same test (the modifiedperformance test) to meet the service and performance test requirements.

A sensitivity analysis wasperformed to assess the change in baseline CDFandLERF that conservatively simulatesand bounds the abovefour changes by assuming that the batteryfailureprobabilityfor all batteriesincreaseslinearlyfrom a 7 day'frequency to a 31 dayfrequency (i.e., failureprobabilitieswere multipliedby 31 / 7 = 4.43). The results show insignificantincreasesin the baseline CDF (increaseof 2E-8/year) andthe baseline LERF (increaseof less than JE-9/year).

The lastfour (4) of the remainingchanges involve additionalbatteryrequirements, reduced surveillanceperiods,and a new LCS controlled "BatteryMonitoringand MaintenanceProgram." These changes are likely to increase battery reliability.

Page 7 of 8

Collectively, the changes are qualitativelyassessedto be risk neutral to risk beneficial.

9. In Reference 2, the licensee commits to control the battery/bus crosstie using a step-by-step procedure with local independent (second checker) verification and to complete this verification within the 2-hour CT. Please identify whether development of the battery crosstie procedure is a licensee commitment per the LAR.

RESPONSE: SONGS commits to develop a step-by-step battery crosstieprocedure.

10. The SONGS' individual plant examination of external events (IPEEE) references plant improvement measures in Section 7.1, "Plant Improvements, "for both seismic and fire initiators. Please confirm that the improvements for fire and seismic initiators in the IPEEE have been implemented and are reflected in the SONGS extended battery PRA evaluation.

RESPONSE: In Section 7.1, "PlantImprovements, "there iwere six (6) seismic related recommendationsand three (3) fire relatedrecommendations. All recommendations have been implemented and were included in the LAR calculations.

11. Attachment I, Table 2, "SONGS Conditional CDF and LERF Contributions for Preventive Maintenance (PM)," located on page 11 of Reference 1, states in note b, 6 tests

• 4 batteries/l0 years. Table 2 on page 16 of Reference 2 lists 7 tests

• 4 batteries/ 10 years. Please clarify why the number of service tests has been reduced in the revised submittal. See also revised Table 2, Attachment I included in Reference 2.

RESPONSE: The number of tests has been adjusted to conservatively reflect closer to actualtesting intervals. This number of tests is still conservative to what SONGS expects to perform. Table 2 is correct as submitted in Revision I in February.

12. Attachment I, Table 2, "SONGS Conditional CDF and LERF Contributions for Preventive Maintenance (PM)," located on page 11 of Reference 1, lists the proactive multiple jar replacements mean outage duration as 20 days. Table 1 on page 5 of Attachment I states that the duration for jar replacement is 21 days. Please clarify what is the correct jar replacement duration to be used in the estimate of mean duration for Table 2.

RESPONSE: The mean durationfor batteryjar replacementon Table 2, page 11 of Attachment 1of Reference I in Februaryis correct (20 days). The duration of21 days in Table I is a typographicalerror.

Page 8 of 8