ML17114A432
| ML17114A432 | |
| Person / Time | |
|---|---|
| Site: | Waterford  |
| Issue date: | 04/21/2017 |
| From: | Chisum M R Entergy Operations |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| W3F1 -2017-0030 | |
| Download: ML17114A432 (15) | |
Text
W3F1-2017-0030Page 2 of 2
Enclosures:
- 1. SAMA Round 2 RAI Responses - Waterford 3 License Renewal Applicationcc: Kriss KennedyRegional AdministratorU. S. Nuclear Regulatory CommissionRegion IV1600 E. Lamar Blvd.Arlington, TX 76011-4511RidsRgn4MailCenter@nrc.govNRC Senior Resident InspectorWaterford Steam Electric Station Unit 3P.O. Box 822Killona, LA 70066-0751Frances.Ramirez@nrc.govChris.Speer@nrc.govU. S. Nuclear Regulatory CommissionAttn: Elaine KeeganDivision of License RenewalWashington, DC 20555-0001Elaine.Keegan@nrc.govU. S. Nuclear Regulatory CommissionAttn: Phyllis ClarkDivision of License RenewalWashington, DC 20555-0001Phyllis.Clark@nrc.govU. S. Nuclear Regulatory CommissionAttn: Dr. April PulvirentiWashington, DC 20555-0001April.Pulvirenti@nrc.govLouisiana Department of EnvironmentalQualityOffice of Environmental ComplianceSurveillance DivisionP.O. Box 4312Baton Rouge, LA 70821-4312Ji.Wiley@LA.gov to W3F1-2017-0030SAMA Round 2 RAI Responses Waterford 3 License Renewal Application toW3F1-2017-0030Page 1 of 12RAI 1Regulatory Basis for RAI 1NEI 05-01 provides the following guidance on the Level 2 model information to be provided in theSAMA submittal:Provide a table or matrix describing the mapping of Level 1 accident sequences into Level2 release categories and a description of the representative release sequences.The ER did not provide a table or matrix describing the mapping of specific Level 1 accidentsequences into the Level 2 release categories or a description of the representative releasesequences. Entergy was therefore asked in RAI 2.e to provide a description of the sequencesused to characterize the source terms for each of the significant release categories, the basis forthis selection and its appropriateness for use in determining the benefit for the Phase II SAMAsevaluated. While the Entergy RAI response 2.e provided an expanded general discussion of thereviews made to select the representative sequence, no specific sequences were provided or adescription of the representative release sequences provided.
RequestProvide a description of the specific Level 1 and Level 2 accident sequences used to characterizethe significant release categories (H-E, H-I and M-I) and why the particular Level 1 and Level 2accident sequences were chosen to be representative for those release categories used indetermining the benefit of the Phase II SAMAs.
Waterford 3 ResponseRelease Category H-EThe H-E cutset results show a distribution of accident sequences as follows: transient (TQX_H)81%, bypass (all induced SGTR) 11%, bypass (SGTR) 5.5%, reactor vessel rupture 1.7%,containment isolation failures < 1%; and ISLOCAs < 1%.The MAAP scenario selected to represent the H-E release category is identified as TQX_H. Level1 sequence TQX is a transient initiating event followed by successful reactor trip and RCSpressure control. The RCP seals develop a leak, due to loss of seal cooling, resulting in a smallLOCA. HPSI is initially successful, but fails during recirculation after the RWSP inventory isexhausted. Level 2 sequence TQX_H is TQX with early failures of both containment sprays andcontainment fans. In TQX_H, containment failure occurs prior to core damage. Scenario TQX_Hwas selected to represent the H-E release category based on its dominant frequency within therelease category as well as large release fractions of CsI. In addition, this accident sequence isunique in that several failure events occur in quick progression. High steaming rates and loss ofcontainment safeguard systems lead to rapid containment over-pressurization occurring during thecore damage phase but before reaching the core damage temperature of 2200 °F. toW3F1-2017-0030Page 2 of 12Other accident sequences included in the H-E release category included containment bypass,containment isolation failures, ISLOCA and reactor vessel ruptures.As described in the response to RAI 2.d 1, these scenarios are classified as early releases becausethe initiating event failure leads to an immediate release pathway from the containment structure.In addition, these scenarios are classified as high severity because the containment releasepathway precludes the mitigation or retention of fission products due to scrubbing, retention, ordeposition mechanisms that occur within the containment structure. MAAP analyses were notperformed for these H-E sequences, due to the high uncertainties associated with evaluation ofthese types of sequences in the MAAP code, so the actual CsI release fractions were not calculated. More rigorous assessment of these scenarios using MAAP would result in thedistribution of these classes of accidents across release categories of lower severity including M-E,L-E and LL-E. Individually, these accident sequences (I-SGTR, SGTR, V, CI, and ISLOCA)contribute no more than 11% to the H-E release category and combined contribute < 20% to theH-E release category. It is judged that the existing conservatisms in the individual SAMA caseanalyses more than compensate for the potential for higher CsI release fractions from thesesequences. For example, if the objective of the SAMA was to reduce the likelihood of a certainfailure mode, the failure mode was completely removed from the model to estimate the benefit,even though the SAMA would not be expected to be 100% effective in eliminating the failure. Inaddition, as shown in the revised Table D.2-2 provided in the earlier RAI response, Phase IISAMAs 61 and 71, which were evaluated to reduce the risk from SGTR, are potentially cost-beneficial. However, the other SAMAs related to these accident sequences are far from beingpotentially cost-beneficial in the revised Table D.2-2. [Phase II SAMAs 54, 56, 57, 58, 59, 60 wereevaluated to reduce the risk from I-SGTR and SGTR, and Phase II SAMA 55 was evaluated toreduce the risk from containment isolation failure]. Therefore, scenario TQX_H was selected asrepresentative of the H-E release category.A summary of the dominant H-E cutset results is provided below.Dominant H-E Release Category Accident SequencesSequence IDContributionCsI %Description TQX_H~81%35%Transient followed by successful reactor tripand RCS pressure control. The RCP sealsdevelop a leak due to loss of seal coolingresulting in a small LOCA. HPSI is initiallysuccessful, but fails during recirculation afterthe RWSP inventory is exhausted.Containment fans and sprays fail early;containment fails due to over-pressurizationduring core uncovery, prior to core damage(>2200 °F).
1 Entergy Letter, W3F1-2017-0001 (ADAMS Accession Number ML17038A436) toW3F1-2017-0030Page 3 of 12Dominant H-E Release Category Accident SequencesSequence IDContributionCsI %DescriptionI-SGTR~11%not calculated Pressure and thermally induced SteamGenerator Tube RupturesSGTR~5.5 %not calculated Steam Generator Tube Ruptures RB, <1%This sequence represents a SGTR followed byreactor trip and loss of RCS and core heatremoval. This results in RCS pressurizationabove the HPSI shutoff head and loss of RCSinventory without adequate makeup from HPSIor the charging system. The result is early coredamage at high pressure.
RX, ~5%This sequence represents the case when aSGTR occurs followed by reactor trip andsuccessful primary to secondary heat removal.However, since the RCS is not depressurized,it remains at high pressure and inventory is lostthrough the steam generator and out theMSSVs. RCS inventory control will be lostwhen the RWSP is depleted. This event resultsin late core damage at medium pressure.
RU, <1%This sequence represents a SGTR followed byreactor trip with successful RCS heat removaland depressurization. However, RCS inventorycontrol is lost due to the failure of the HPSI andcharging system to make-up inventory lost outof the ruptured, unisolated steam generator.This results in early core damage at medium pressure.
V~1.7 %not calculated Reactor Vessel Rupture CI<1 %not calculated Containment Isolation FailuresISLOCA<1%not calculated Isolating System LOCA toW3F1-2017-0030Page 4 of 12Release Category H-IThe H-I cutset results show a distribution of accident sequences as follows: station blackout(SBO_E) 66%, TB_H 28%, TQU_H 5%.The MAAP scenario selected to represent the H-I release category is identified as SBO_E. Level 1sequence SBO is a loss of offsite power followed by failure of both emergency diesel generatorsand failure of the turbine-driven EFW pump to start or run. This results in early core damage athigh pressure. Level 2 sequence SBO_E is the Level 1 SBO sequence with the early loss of bothcontainment sprays and containment fans due to loss of power. The SBO_E scenario was selectedto represent the H-I release category based on its dominant frequency in the release category aswell as the largest release fractions of CsI. This sequence results in containment failure prior to vessel breach.Other accident sequences included in the H-I release category included TB_H and TQU_H.Level 1 sequence TB is a transient initiating event followed by successful reactor trip, RCSpressure control, and RCS pressure boundary integrity. Decay heat is not removed from the steamgenerators by means of the MFW, EFW, or any backup systems. Sequences that result in lifting ofthe pressurizer SRV (RCS pressure control fails), but successful closure of the SRV are includedin this sequence. Failure to recover primary to secondary heat removal results in RCS inventoryloss through the pressurizer safety relief valves and early core damage at high pressure. Level 2sequence TB_H is TB with early failures of both containment sprays and containment fans. Thissequence results in a late containment failure (> 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />) after vessel breach.Level 1 sequence TQU is a transient initiating event followed by successful reactor trip and RCSpressure control. However, the RCP seals develop a leak due to the loss of seal cooling resultingin a small LOCA. High pressure safety injection fails to inject sufficient water from the RWSP. Thisevent leads to early core damage at medium pressure. Level 2 sequence TQU_H is TQU withearly failures of both containment sprays and containment fans. This sequence results in latecontainment failure (> 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />) after vessel breach.A summary of the dominant H-I cutset results is provided below. toW3F1-2017-0030Page 5 of 12Dominant H-I Release Category Accident SequencesSequence ID ContributionCsI %Description SBO_E~66 %32%Loss of offsite power followed by failure ofboth emergency diesel generators and failureof the turbine-driven EFW pump to start/run.Containment fans and sprays not availabledue to loss of power. Containment failureoccurs prior to vessel breach.
TB_H~28 %25%Transient with successful RCS pressurecontrol and boundary integrity with loss ofdecay heat removal and failure to recoverRCS inventory; early failure of containmentfans and sprays; vessel breach with late (> 4hours) containment failure.
TQU_H~5 %23%Transient with seal LOCA and successful RCSpressure control but failure of inventorycontrol; early failure of containment fans andsprays; vessel breach with late (> 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />)containment failure.Release Category M-IThe M-I cutset results show a distribution of accident sequences as follows: SU_H 77%, TB_F 12%, TB_B 10%.The MAAP scenario selected to represent the M-I release category is identified as TB_B. Level 1sequence TB is a transient initiating event followed by successful reactor trip, RCS pressurecontrol, and RCS pressure boundary integrity. Decay heat is not removed from the steamgenerators by means of the MFW, EFW, or any backup systems. Sequences that result in lifting ofthe pressurizer SRV (RCS pressure control fails), but successful closure of the SRV, are includedin this sequence. Failure to recover primary to secondary heat removal results in RCS inventoryloss through the pressurizer safety relief valves and early core damage at high pressure. Level 2sequence TB_B is TB with both containment sprays and containment fans available for operation.In this sequence, the vessel remains intact with containment failure due to hydrogen burnoccurring at about 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br />. Scenario TB_B was selected to represent the M-I release categorybased on its applicability to the SAMAs being evaluated and earlier containment failure time.Other accident sequences included in the M-I release category included small LOCA (SU_H) and TB_F. toW3F1-2017-0030Page 6 of 12Level 1 sequence SU is a small break LOCA with reactor trip, but failure of the HPSI to injectsufficient water from the RWSP. This event leads to early core damage at medium pressure. Level2 sequence SU_H is SU with early failures of both containment sprays and containment fans. Thissequence results in containment failure due to over-pressurization. SU_H is the dominantcontributor to the M-I release category with a CsI fraction of 7.7%. As described in the priorresponse to RAI 4.b 2, SU_H was identified as an outlier based on the ratio of barium to iodide incomparison to the other accident scenarios. It is acceptable to exclude SU_H because the M-Irelease category is < 2% of the total level 2 release frequency and the CsI release fraction issimilar to the selected scenario. Also, as shown in the revised Table D.2-2 provided in the earlier RAI response 2, Phase II SAMAs 13 and 18, which were evaluated to reduce the frequency of coremelt from a small LOCA, are far from being potentially cost-beneficial.As described above for TB_B, Level 1 sequence TB is a transient initiating event followed bysuccessful reactor trip, RCS pressure control, and RCS pressure boundary integrity. Decay heat isnot removed from the steam generators by means of the MFW, EFW, or any backup systems.Sequences that result in lifting of the pressurizer SRV (RCS pressure control fails), but successfulclosure of the SRV, are included in this sequence. Failure to recover primary to secondary heatremoval results in RCS inventory loss through the pressurizer safety relief valves and early coredamage at high pressure. Level 2 sequence TB_F is TB with containment sprays available, butcontainment fan coolers failed. The reactor vessel remains intact, but the containment fails late (19hours) due to containment over-pressurization.A summary of the dominant M-I cutset results is provided below.Dominant M-I Release Category Accident SequencesSequence IDContributionCsI %Description TB_B~10 %6.3 %Transient with successful RCS pressurecontrol and boundary integrity with loss ofdecay heat removal and failure to recover RCSinventory. The vessel remains intact withcontainment failure occurring at 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br /> dueto hydrogen burn. Both containment fans and sprays are available.
SU_H~77 %7.7%Small LOCA with containment failure due toover-pressurization; failure of containment fans and sprays.
2 Entergy Letter, W3F1-2017-0001 (ADAMS Accession Number ML17038A436) toW3F1-2017-0030Page 7 of 12Dominant M-I Release Category Accident SequencesSequence IDContributionCsI %Description TB_F~12 %7.9 %Transient with successful RCS pressurecontrol and boundary integrity with loss ofdecay heat removal and failure to recover RCSinventory. The vessel remains intact withcontainment failure due to over-pressurizationoccurring at 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />. Containment spray isavailable, but containment fans are failed. toW3F1-2017-0030Page 8 of 12RAI 2Regulatory Basis for RAI 2NEI 05-01 provides the following guidance for the identification of SAMAs:SAMAs may be hardware changes, procedure changes, or enhancements to programs,including training and surveillance programs. Hardware changes should not be limited topermanent changes involving addition of new, safety-grade equipment, but should alsoinclude lower cost alternatives, such as temporary connections using commercial gradeequipment (e.g., portable generators and temporary cross-ties).NEI 05-01 further states:If a SAMA was not evaluated for a dominant risk contributor, justify why SAMAs to furtherreduce the contributor would not be cost-beneficial.The dominant contributor to internal flooding risk is water propagation in the electric board roomand was evaluated in SAMA 68. In SAMA 68, Entergy proposed to install permanent flood doorsto prevent water propagation to the electric board room. The WF3 plant specific cost estimate forthe flood doors was determined to be approximately 1.3 million dollar. In comparison, the benefitfor removing this internal flooding risk is approximately three hundred thousand dollars. Therefore,Entergy found SAMA 68 to not be cost-beneficial.The staff requested in RAI 6.i for Entergy to consider a lower cost alternative than a permanentflood door, such as a flood barrier.
RequestConsider a lower cost alternative to the flood doors, such as a flood barrier, for SAMA 68 orexplain why a lower cost alternative was not considered or necessary. Provide a cost-benefitcomparison or justify why SAMAs to further reduce the water propagation in the electric boardroom would not be cost-beneficial.
Waterford 3 ResponseSummary of Internal Flooding Analysis for the B Switchgear / Electrical Board Room [FloodZone RAB21-212/225B]An 8" fire protection line runs vertically through flood zone RAB21-212/225B. The guillotine ruptureof this line will result in the release ~2700 gpm of water. It is assumed that 15 minutes will elapsebefore the release is terminated. In this time water will accumulate to a 7" depth in switchgear area"B" and the electrical penetration room. This will cause submergence damage to the reactor tripswitchgear, multiplexer RA-2104, 4.6-kV switchgear 3B-3S, several B train 480-V switchgear andmotor control centers, B train battery charger 3B2-S, and other B train electrical components. As a toW3F1-2017-0030Page 9 of 12result of submergence damage to multiplexer RA-2104, the reactor would trip. Simultaneously,damage to 4.6 kV switchgear 3B would result in the loss of the "B" trains of HPSI, LPSI, CS andEFW, the "B" EDG and the "B" component cooling and "B" ACCW pumps.Floodwater released into this flood zone will tend to accumulate as there is no rapid or directdrainage to lower elevations and the only door leading outside the reactor auxiliary building, doorD51 that opens into the cooling tower area, is tight. Floodwater will flow under or through doorsinto several rooms, most importantly into the motor-generator set room, switchgear room A/B, andswitchgear room A.Should the release following the guillotine rupture of the firewater line be terminated in 15 minutes,the level of floodwater will rise to only ~1" in switchgear room A/B and will not enter switchgearroom "A" (a 4" curb lies inside the door). This is flood scenarioRAB21-212-225B-15MIN.Should the release last for more than 15 minutes, damage to the A/B switchgear might occur. Thisis flood scenario RAB21-212-225B-15-45MIN.Should the release last for more than 45 minutes, damage to switchgear in switchgear room Amight also occur. This is flood scenario RAB21-212-225B-45MIN.If the release is still not terminated, once flood levels reach ~ 3', doors will burst open and waterwill flood down to the -4' and -35' elevations of the reactor auxiliary building. Submergencedamage at these lower levels will occur, resulting in submergence damage to the EFW pumps.Assuming that all water released descends to the -35' elevation, sufficient water will be dischargedin 80 minutes after a guillotine rupture to cause such damage. Should the flood persist for 2 to 3hours, sufficient water would enter the safeguard rooms through the drains if the drain sumppumps failed to cause submergence damage to the HPSI, LPSI and containment spray pumps inthese flood zones. This is flood scenario RAB21-212-225B-80MIN.
Flood Scenario CDF (/rx-yr)RAB21-212-225B-80MIN 7.19E-07RAB21-212-225B-45MIN 3.47E-10RAB21-212-225B-15MIN 8.40E-11RAB21-212-225B-15-45MIN 2.76E-11Total7.19E-07 toW3F1-2017-0030Page 10 of 12Analysis Case 42 for SAMA 68, "Water Tight Doors for the Largest Contributor to Internal Flooding"This analysis case was used to evaluate the change in plant risk from installing flood doors toprevent water propagation in the electric board room. The electrical equipment rooms at WF3 donot have water tight flood doors. Specifically this SAMA will evaluate water tight doors for thelargest contributor to internal flooding, which is flood zone RAB21-212/225B. This analysis casewas used to model the benefit of phase II SAMA 68.Since the internal event risk analysis does not include internal flooding, this internal flooding SAMAwould not mitigate internal event risk. A bounding analysis was performed by assuming the SAMAwould eliminate the contribution to internal flooding CDF from flood zone RAB21-212/225B. TheSAMA case 42 benefit with uncertainty was estimated to be $332,233.Analysis of Lower Cost Alternatives for SAMA 68As stated above,"Once flood levels reach ~ 3', doors will burst open and water will flood down tothe -4' and -35' elevations of the reactor auxiliary building. Submergence damage at these lowerlevels will occur resulting in submergence damage to the EFW pumps. Assuming that all waterreleased descends to the -35' elevation, sufficient water will be discharged in 80 minutes after aguillotine rupture to cause such damage." This flood is designated RAB21-212-225B-80MIN and isthe risk significant scenario. Since this flood scenario assumes that no actions are taken and theflood levels reach 3', the scenario can only be mitigated by a SAMA that would prevent the doorsfrom opening when the flood reaches the 3' level, or by extensive room drain modifications thatwould prevent water accumulation. It would not be mitigated by lower cost alternatives like a curbor flood barrier. Thus, water tight doors were postulated to mitigate this flood.Lower cost alternatives like a curb or flood barrier could, however, mitigate other flood scenarios inthis room. As described above, the floods that last more than 15 minutes, but less than 80 minutescould damage the A/B switchgear and the A switchgear. Therefore, a lower cost alternative wouldeliminate the risk from floods RAB21-212-225B-15-45MIN (2.76E-11/rx-yr) and RAB21-212-225B-45MIN (3.47E-10/rx-yr), for a total of 3.75E-10/rx-yr. This is approximately 0.02% of the totalinternal flooding CDF (2.48E-06/rx-yr).The benefit of eliminating this flood risk is calculated below, following the Case 42 method ofcalculating the benefit.
Given,Maximum internal benefit is $2,355,400Total internal flooding CDF = 2.48E-06/rx-yrInternal events CDF = 1.05E-05/rx-yrMaximum internal flooding benefit = Maximum internal benefit x Total internal floodingCDF/Internal events CDFMaximum internal flooding benefit = $2,355,400 x (2.48E-06/1.05E-05) = $556,323 toW3F1-2017-0030Page 11 of 12SAMA benefit = (averted internal flood CDF/Internal events CDF) x (Maximum internalflooding benefit) = (3.75E-10/2.48E-06) x $556,323SAMA benefit = $84Applying the uncertainty factor of 2.06,SAMA benefit with uncertainty = $84 x 2.06 = $173Since the minimum implementation estimate for a hardware modification is $100,000, thepostulated lower cost alternatives would not be potentially cost-beneficial.In response to RAI 6.h 3, Entergy indicated that even if one were to multiply the benefit produced byCases 41 and 42 by three, SAMAs 67 and 68 would remain not cost-beneficial. Similarly, if onewere to multiply the new benefit with uncertainty of $173 by three, the SAMA would remain notcost-beneficial.
3 Entergy Letter, W3F1-2017-0001 (ADAMS Accession Number ML17038A436) toW3F1-2017-0030Page 12 of 12RAI 3Regulatory Basis for RAI 3NEI 05-01 provides the following regarding potentially cost beneficial SAMAs:This analysis may not estimate all of the benefits or all of the costs of a SAMA. Forinstance, it may not consider increases or decreases in maintenance or operation costsfollowing SAMA implementation. Also, it may not consider the possible adverseconsequences of procedure changes, such as additional personnel dose. Since the SAMAanalysis is not a complete engineering project cost-benefit analysis, the SAMAs that arecost-beneficial after the Phase II analysis and sensitivity analyses are only potentially cost-beneficial.Thus, in the ER, Entergy stated:Although the above SAMA candidates do not relate to adequately managing the effects ofaging during the period of extended operation, they have been submitted for detailedengineering project cost-benefit analysis to further evaluate implementation of thesepotentially cost beneficial SAMAs.In the response to several RAIs, Entergy revised the calculated benefits for several of the Phase IISAMAs, performed additional sensitivity analyses, and evaluated potentially lower costalternatives. As a result of these analyses, 5 new potentially cost-beneficial SAMA candidateswere identified.RAI 3Will the potential cost beneficial SAMAs added as a result of the RAIs dated 11/22/2016 and03/28/2017 be submitted for detailed engineering project cost-benefit analysis to further evaluateimplementation of these potentially cost beneficial SAMAs?
Waterford 3 ResponseEngineering change requests have been initiated for detailed engineering project cost-benefitanalysis in accordance with the engineering change process for the potentially cost-beneficialSAMA candidates identified as a result of RAIs.