ML17038A436
ML17038A436 | |
Person / Time | |
---|---|
Site: | Waterford ![]() |
Issue date: | 02/07/2017 |
From: | Chisum M R Entergy Operations |
To: | Document Control Desk, Office of Nuclear Reactor Regulation |
References | |
W3F1 -2017-0001 | |
Download: ML17038A436 (77) | |
Text
W3F1-2017-0001Page 2 of 2cc: 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 toW3F1-2017-0001SAMA RAI Responses Waterford 3 License Renewal Application toW3F1-2017-0001Page 1 of 74RAI SAMA 11. Provide the following information regarding the Level 1 Probabilistic Risk Assessment (PRA) orProbabilistic Safety Assessment (PSA) used for the Severe Accident Mitigation Alternative(SAMA) analysis. The basis for this request is as follows: Applicants for license renewal arerequired by 10 CFR 51.53(c)(3)(ii)(L) to consider SAMAs if not previously considered in anenvironmental impact assessment, related supplement, or environmental assessment for theplant. As part of its review of the Waterford Electric Station, Unit 3 (WF3) SAMA analysis, NRCstaff evaluates the applicant's treatment of internal events and calculation of core damagefrequency (CDF) in the Level 1 PRA model. The requested information is needed in order for theNRC staff to reach a conclusion on the sufficiency of the applicant's Level 1 PRA model forsupporting the SAMA evaluation.1.a. WF3 Environmental Report (ER) Section D.1.4 indicates that there is approximately afactor of 3 increase in CDF and a factor of 3 decrease in large early release frequency(LERF) from PSA 2009 R4 to 2015 PSA R5 used for the SAMA analysis. Discuss themajor reasons for these changes.Waterford 3 ResponseSeveral changes were made in the Revision 5 PSA model update as listed in ER SectionD.1.4.4.The most significant change causing the increase in CDF was the revision of the batterydepletion modeling to include procedural direction to strip batteries to allow for extendedbattery life. This change had little impact on LERF because the associated sequencesare not early sequences. The condensate storage pool (CSP) update, for emergencyfeedwater to utilize source water from the demineralized water storage tank (utilizinggravity fed supply) feeding the CSP, caused a decrease in CDF. However, that decreasewas overshadowed by the increase from the battery depletion modeling.The decrease in LERF was due to removal of conservatisms in the LERF model.Contributing changes include the following. Removal of dependency to refill nitrogen accumulators (extended creditedoperation time from 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />) Added containment cooling system fan coil isolation valves into model Revision of modeling associated with refill of the CSP to reflect current proceduralguidance Updated human failure events1.b. ER Section D.1.4.5 indicates that the 2009 peer review concluded that approximately 9%of the applicable PRA standard's supporting requirements (SRs) were met at CapabilityCategory I while 10% of the SRs were rated as not met. Discuss any findings from this toW3F1-2017-0001Page 2 of 74review that remain open in the PRA models used for the SAMA analysis and theirpotential impact on the SAMA analysis.Waterford 3 ResponseThree findings from the 2009 peer review, unrelated to internal flooding, remain open inthe PRA models used for the SAMA analysis. These findings and their potential impacton the SAMA analysis are discussed below. IE-C12-01:ISLOCA - low pressure LPSI and HPSI line contain two check valvesin series. The failure rate of the check valves need to be treated as conditional,rather than independent. Additionally need to address small ruptures in the LPSIMOVs. At present only large leakage is considered.This finding is associated with the inclusion of State of Knowledge Correlation(SOKC). The increase in probabilities due to SOKC would be minor per WCAP-17154-P. This is insignificant because the conditional failure rate of both checkvalves would be about 2E-7. Additionally, a review of NUREG/CR-6928 showsthat small ruptures are defined as 1 to 50 gpm. Leaks of this size are notconsidered sufficient to meet the classification for ISLOCA. ISLOCA is not asignificant contributor to risk at WF3, with a CDF contribution of 9.44E-10 /rx-yrand a LERF contribution of 4.26E-10 /rx-yr. Therefore, adjusting the failure ratesof LPSI and HPSI line check valves to address SOKC would have a negligibleimpact on the risk profile and no impact on the SAMA cost-benefit conclusions. SY-A18a-01:HPSI system has an installed spare that can be aligned to eithersystem. Coincident unavailability due to maintenance for redundant equipment ispossible (spare pump OOS for extended periods and could be OOS with anotherpump). Need to specifically address this possibility. This may also be true forcharging pumps.SR SY-A18 (changed to SY-A20 in latest version of the standard) states:INCLUDE events representing the simultaneous unavailability of redundantequipment when this is a result of planned activity. The Plant Specific Failure DataDevelopment analysis (PSA-WF3-01-DA-01) documents the inclusion of allplanned concurrent maintenance (including installed spares). This remains 'notaddressed' due to documentation. The coincident unavailability is included in themodel; however the documentation does not fully explain the process used toconsider/model events.This is a documentation finding, with no impact on the SAMA cost-benefit results. SY-C2-01:Need to add a discussion of what the criteria for CCF considerationsare (which types of components were looked at, were inter- and intra- systemCCFs considered, etc. If the component types were determined based off of a listfrom a Reference, provide this information and a pointer to the referencedocument/methodology. toW3F1-2017-0001Page 3 of 74This finding is a documentation issue only. During the disposition of the F&O forSY-B4, the criteria for CCF considerations were reviewed. Inter-system and Intra-system CCFs considered are documented in the CCF calculation, but not explicitlyin each system notebook. This finding has no quantitative impact on the internalevents model. This is a documentation finding, with no impact on the SAMA cost-benefit results.Eight findings from the 2009 peer review, related to internal flooding, remain open inthe PRA models used for the SAMA analysis. These eight findings are summarizedbelow. Some of the findings are documentation issues and resolution of others wouldtend to decrease or have no impact on the internal flooding CDF.The internal flooding model was not used to analyze individual SAMA candidates.Rather, the internal flooding CDF contribution of 2.48E-06/rx-yr was used, along withexternal events CDF values, to calculate the internal/external events multiplier for theSAMA analysis. The multiplier was utilized because the current internal floodingmodel hasn't been integrated with the current internal events model or the Level 2and 3 models. As noted in the response to question 1.e, the CDF for the 2015 (R5)PSA model used for the SAMA analysis is larger than prior models because thatmodel includes a revision of the battery depletion modeling to include proceduraldirection to strip batteries to allow for extended battery life. This modeling increasedthe CDF from sequences initiated by a loss of offsite power. Since the internalflooding CDF comes from sequences initiated by internal floods, it would not besignificantly impacted by this model change. Thus, there is no impact on the SAMAcost-benefit conclusions. IF-B2-01:Although required by this SR, no evaluation of individual componentfailure modes, human-induced mechanisms, or other events that could releasewater into the area were identified. The evaluation assumed that using aguillotine rupture was adequate to not require any specific failures or human-induced mechanisms. This does not meet the intent or specifics of thisrequirement. Other SRs (IF-B3 and IF-D6) are also potentially not met when onlythe use of a guillotine rupture is used.(IF-B3) Waterford 3 basically characterized all flood sources as catastrophicruptures but where there are potential spray targets they do evaluate sprayimpacts. Waterford characterizes the flood in terms of gpm for larger sources oras total flood capacity for smaller flood sources. Waterford does considerpressure of the flood source to a limited extent, primarily when evaluating thepotential for spray impacts. However, there is no evidence that Waterfordconsidered the temperature of the flood source beyond stating that HELB istreated elsewhere. Waterford should include some discussion of temperature inPRA-W3-01-002.(IF-D6) Section 2.0 of the Internal Flooding analysis specifically states "all causesof flooding were considered except plant-specific maintenance activities. No toW3F1-2017-0001Page 4 of 74mention of inclusion/exclusion of generic maintenance activities was found.While Waterford discusses operator error contributions to flooding at a very highlevel in section 3.1.2, basically the only floods considered were catastrophicfailures. The flood scenario frequencies were then quantified using generic piperupture data and plant-specific pipe length. The resulting low frequencies oftenlead to scenarios being subsumed. While the operator induced floods may beless severe, the frequencies will be higher, so they should be consideredexplicitly. IF-C3c-01:There do not appear to be any Engineering calculations available tosupport some of the statements or inherent assumptions made in the InternalFlooding Analysis. In particular, room dimensions and flood rates are notavailable to justify flood depths stated for various rooms, some zones credit "airtight" doors as being structurally sound up to a depth of 6 inches with nojustification of door integrity against a static water load of this depth, "air tight"doors appear to be treated as "flood doors" with no justification as to how this wasdetermined (normally air tight door seals are not designed to prevent waterintrusion or extrusion), timing related calculations (time for flood to reachsusceptible equipment, flood rates, etc.) were not included or referenced, etc. Ifthese calculations exist, they should be either provided in appendices to the reportor referenced in the appropriate sections of the report.If the calculations do not exist, they should be performed, and the statements andinherent assumptions in the analysis re-verified to ensure they reflect the resultsof the calculations.On page 89 of the Internal Flooding Report, within the 2nd paragraph, a statementis made that a particular door is assumed to open out, and that the floodpropagation pathway will go through that door. No discussion, or calculation, isprovided to justify why that particular door will open versus another of the doorsfrom the room (there are multiple doors associated with the room). If there is nobasis behind that particular door failing prior to the other doors, then an evaluationof the flooding impacts from other doors opening should be performed.On page 215, there is an un-supported assumption that drain failures have afailure probability of 0.1. Need to provide basis for this assumption. IF-C7-01: The Fire Water pump house has been excluded from evaluation on thebasis that the failure of the fire pumps will not precipitate a reactor trip and the fireprotection itself is not used to mitigate any accident scenario that might lead tocore damage other than those occasioned by fire. This exclusion needs to be re-visited to determine if an internal flood in the fire water pump house has thepotential to initiate a flood/spray event elsewhere in the plant due to spurious firewater valve actuations (e.g. look at potential for spray/submergence on a firewater control panel to determine if it could cause spurious signals to fire waterequipment in the plant resulting in a plant spray/flood event.), and if this toW3F1-2017-0001Page 5 of 74inadvertent actuation could result in the need for a plant shutdown. If this impacthas been evaluated, document it. IF-D5a-01: Although Waterford calculates the initiating event frequency for eachevaluated flood scenario using generic data, and the specific calculations arepresented in a footnote for each scenario, a reduction factor has beeninappropriately applied to component rupture failure rates. The analysis statesthat the generic component failure rates are obtained from EGG-SSRE-9639 (seeTable 3.2.1.2 in Flood report). However, these failure rates are then reduced byan additional factor to convert them from "spray" failures to "rupture" failures. (Theexample provided shows a "1/27th" reduction for a 1000 gpm valve failure) Theapplication of the reduction factor is inappropriate since the data are "rupture"rates, not "spray" rates, and the EGG-SSRE-9639 source document has alreadyapplied a 1/25 reduction factor to ensure that the rates are applicable as rupturerates. Need to use the "rupture" failure rates without applying the additionalreduction factor. IF-D7-01: The discussion for excluding the condensate polisher building fromconsideration based on the assumption that the operators would bypass thecondensate polisher system in the event of a rupture/leak within the building isinadequate. IF-D7-02: The Internal Flooding report is inconsistent / incorrect in its use of"subsume" versus "screen". For example, in Section 4.2.1.3, the report states thatscenarios are "subsumed" but the justification for subsuming the scenarios isbased on the justification for "screening" of scenarios (screening is defined in SRIF-D7). IF-E5a-01: For operator actions, only actions outside of the Control Room appearto have been reviewed. Also, no analysis could be found to determine if therewere any "unique" (i.e. not credited in the base PRA) operator actions that shouldbe added for internal flooding recoveries, or if the operator actions credited weremodified to account for the stress level/timing differences associated with internalflooding scenarios. Of the actions credited in the base PRA model, 4 of theoperator actions appear to be removed by a recovery rule file as inaccessible.However, no additional analysis was found to justify why these 4 actions weredetermined to be inappropriate for internal flooding recovery, or why no otherhuman actions were impacted by the internal flooding scenarios. IF-E6-01: In general, WSES3 used the standard quantification processes fromsection 4.5.8 of the standard. However, WSES3 did not propagate the numericaluncertainties as part of the quantification. WSES3 needs to redo the InternalFlooding Quantification and include the propagation of the numerical uncertaintiesand provide the mean and ERF factors for the resultant CDFs. toW3F1-2017-0001Page 6 of 741.c. Provide the "freeze date" or the date which corresponds to the WF3 design and operationincorporated into the WF3 PSA used for the SAMA analysis. Identify any design oroperational (including fuel cycle) changes that have or, are planned, since this freezedate that might impact the SAMA analysis.Waterford 3 ResponseThe PRA model used in the SAMA analysis has a freeze date of 11/1/2015. Since thistime some changes have been made to the design and operation of the plant. Threechanges that would potentially have an impact on the SAMA analysis are themodifications made for FLEX (diverse and flexible coping strategies, which do not affectthe design basis of the plant but primarily operational response to an extended loss of ACpower), the temporary emergency diesel, and the proceduralization of local manualcontrol of EFW pump turbine and flow control valves in a more prominent way. Thesechanges would serve to reduce the SBO contribution to core damage and releasecategories. No fuel cycle changes are planned that might impact the SAMA analysis.1.d.Confirm that no changes have been made to the WF3 model used in the SAMA analysissince the peer review that would constitute an upgrade as defined by the PRA StandardASME/ANS RA-Sa-2009, as endorsed by Regulatory Guide (RG) 1.200, Revision 2.Waterford 3 ResponseNo changes have been made to the WF3 model used in the SAMA analysis since thepeer review that would constitute an upgrade as defined by ASME/ANS RA-Sa-2009, asendorsed by RG 1.200, Revision 2.1.e.The revised Attachment W to the WF3 National Fire Protection Association (NFPA) 805License Amendment Request (LAR) gives the internal events CDF and LERF as 6.5E-06per reactor-year (rx-year) and 8.7E-08 per rx-year respectively. These values areapproximately 60% of the results given for the 2015 (R5) PSA used for the SAMAanalysis (internal events CDF and LERF as 1.05E-05 per rx-year and 1.36E-07 per rx-year respectively). Identify which of these values best represents WF3 for license renewalpurposes, discuss the reasons for these differences and the impact on the SAMAanalysis.Waterford 3 ResponseIn enclosure 2 of letter W3F1-2015-0025, "Responses to Request for AdditionalInformation Regarding Adoption of National Fire Protection Association Standard NFPA805 License Amendment Request (LAR) Waterford Steam Electric Station, Unit 3(Waterford 3)," dated May 14, 2015, the internal events CDF and LERF are given as6.5E-06/rx-yr and 8.7E-08/rx-yr, respectively. toW3F1-2017-0001Page 7 of 74The CDF value differs from that given for the 2015 (R5) PSA model used for the SAMAanalysis because it is from a prior interim model revision which did not include therevision of the battery depletion modeling to include procedural direction to strip batteriesto allow for extended battery life. The prior interim LERF model had a slightly lower valuedue to the update in the SGTR sequences which are binned as LERF which saw anincrease due to the change in values of thermal-induced SGTR and pressure-inducedSGTR failure probabilities.The 2015 (R5) PSA model used for the SAMA analysis best represents WF3 for licenserenewal purposes. Additional changes to the revision 5 PSA model update can be seen inSection D.1.4.4 of the ER.1.f.Briefly describe the process and procedures in place to assure the technical adequacy ofchanges made to the WF3 PSA since the 2009 peer review.Waterford 3 ResponseThe PSA Maintenance and Update procedure describes the process for maintaining thePSA models current with the as-built and as-operated plants. It describes the modelchange request (MCR) database used to track plant changes, procedure revisions,nuclear licensing revisions, and model improvements that impact the PSA models. Thisprocedure is in place for PSA model maintenance in order to ensure that the modelremains current with the as-built, as-operated plant and to ensure that industry standards,experience, and technology are appropriately incorporated into the models. Thisprocedure gives specific instructions for identifying model change requests, documentingthose requests, and incorporating those requests into the PSA model.PSA engineers review all plant modifications and procedure revisions with the potential toimpact the PSA model and enter them into the MCR database. MCRs that are importantand necessary to assure the technical adequacy of the PRA or quality of the PRA resultin a condition report within the corrective action system and an interim model update.Those that have a minor impact, but still necessary to address may be deferred until thenext PSA update.The PSA engineers performing model updates are experienced, trained professionalsand each change is reviewed by a second, experienced, trained PSA engineer. Inaddition, expert panel reviews are used to enhance the technical quality of the PSAupdates. Changes from the expert panel review for an update are immediatelyincorporated into that update of the model. Therefore, the WF3 PSA is of sufficienttechnical quality for use in the SAMA analysis. toW3F1-2017-0001Page 8 of 741.g. ER Section D.1.1 (p. D-26) states that the CDF uncertainty factor of 1.99 is based on theratio of the 95th percentile CDF to the mean CDF. Since the PSA results used in the cost-benefit analysis are based on point estimates, the uncertainty factor should be based onthe ratio of the 95th percentile CDF to the point estimate CDF. Describe the uncertaintyanalysis and provide the 95th, mean and point estimate results from this analysis.Discuss the impact of the revised uncertainty factor on the results of the SAMA analysis.Waterford 3 ResponseThe parametric uncertainty analysis was performed by defining a probability distributionon the value of each parameter and then propagating this parameter value uncertaintythrough the result (CDF) using Monte Carlo sampling performed by the UNCERTsoftware. The parameters of interest are those of the probability models for the basicevents of the logic model and include failure rates, component unavailabilities, initiatingevent frequencies and human error probabilities. The95th, mean, and point estimate areshown in the table below.Point EstimateMean95%CDF (/rx-yr)1.05E-051.086E-052.164E-05The ratio of the 95th percentile CDF to the point estimate CDF is 2.06. The "Internal andExternal Benefit with Uncert" values in revised Table D.2-2 use an uncertainty factor of2.06.1.h. Discuss the scope of the 2009 WF3 internal events peer review and if all applicableelements of the ASME PRA standard were assessed in this review. Discuss the potentialimpact on the SAMA analysis of any elements that were not assessed.Waterford 3 ResponseThe scope of this review was a full scope review of the WF3 PRA with the exception ofthe Configuration Control requirements and the following High Level Requirements(HLR): HLR-IE-A, HLR-IE-B, HLR-HR-A, HLR-HR-B, HLR-HR-C, HLR-HR-D, HLR-HR-E,and HLR-HR-F. The Configuration Control requirements were not reviewed becauseEntergy uses a fleet-wide configuration control procedure for all of their plants and thisprocedure was reviewed as part of the Arkansas Nuclear One Unit 2 (ANO-2) peerreview. The Initiating Event (IE) HLRs and the Human Reliability Analysis (HRA) HLRslisted above were covered by an earlier review and did not need to be revisited. The teamdid do a confirmatory review of HLR-IE-A and HLR-IE-B. The results of the confirmatoryreview of HLR-IE-A and HLR-IE-B and results of the earlier peer review for the otherlisted HLRs are provided below. toW3F1-2017-0001Page 9 of 74HLR-IE-A: The confirmatory review concluded,"WSES-3 PRA has considered initiatingevents based on a number of sources such as the past PRAs, generic nuclear plantoperating data, and WSES-3 specific experience. Therefore, the requirements of thisHLR are satisfied."HLR-IE-B: The confirmatory review concluded,"The initiating events have been groupedproperly and the review team found these groupings to be consistent with other plantPRAs. Therefore, the requirements of this HLR are satisfied."HLR-HR-A: The earlier peer review concluded,"Pre-initiators were identified by a definedapproach. This included system/component alignment issues and instrumentmiscalibrations. This was completed in the system analyses. One issue that was initiallyidentified related to the lack of the documentation of the review of plant specificinformation. This issue has been resolved."HLR-HR-B: The earlier peer review concluded,"The screening was completed in anacceptable manner. A process for screening pre-initiators was defined and followed.Activities that could impact multiple trains were not screened. One issue was identifiedrelated to the screening out of the pre-initiator Human Failure Events (HFEs) of therunning systems."HLR-HR-C: The earlier peer review concluded,"This was completed appropriately.Human failure events were defined and added to the PRA model appropriately."HLR-HR-D: The earlier peer review concluded, "A systematic process was defined andfollowed for calculating human error probabilities. A detailed assessment was applied toall HFEs, as opposed to using screening values."HLR-HR-E: The earlier peer review concluded,"A systematic approach was followed toidentify the operator responses for each accident sequence. Operators were used viatalk-throughs to provide the information necessary to fully understand and assess theactions required. (Note that these talk-throughs did not include measurement ofsequence timing.) The operator input was well done and documented."HLR-HR-F: The earlier peer review concluded,"HFEs were defined and incorporated inthe PRA model appropriately. The definition of the HFEs was completed and inputinformation was well documented in individual operator action worksheets."Findings (one each for HLR-IE-A, HLR-IE-B, HLR-HR-A, HLR-HR-B, HLR-HR-D, andHLR-HR-F and none for HLR-HR-C or HLR-HR-E) from the earlier peer review werecarried over into the list of findings from the 2009 peer review. These findings have beenclosed, and no findings related to these HLRs remain open in the PRA models used forthe SAMA analysis. Thus, there is no impact on results of the SAMA analysis from thiscarry-over review. toW3F1-2017-0001Page 10 of 74RAI SAMA 22. Provide the following information relative to the Level 2 PRA or PSA analysis. The basis for thisrequest is as follows: Applicants for license renewal are required by 10 CFR 51.53(c)(3)(ii)(L) toconsider SAMAs if not previously considered in an environmental impact assessment, relatedsupplement, or environmental assessment for the plant. As part of its review of the WF-3 SAMAanalysis, NRC staff evaluates the applicant's treatment of accident propagation and radionucliderelease in the Level 2 PRA model. The requested information is needed in order for the NRCstaff to reach a conclusion on the adequacy of the applicant's Level 2 PRA model for supportingthe SAMA evaluation.2.a. The table in ER Section D.1.4 gives LERF for the 2015 (R5) PSA as 1.36E-06 per rx-year,while Section D.1.2.1 (p. D-27) and Table D.1-12 gives 1.88E-06 per rx -year. Explain thedifference.Waterford 3 ResponseThe LERF value given in Section D.1.4 for the 2015 (R5) PSA (1.36E-07/rx-yr) is thesimplified internal events LERF model for WF3. For the SAMA analysis, a detailed Level 2model was created which includes LERF and provides a value of 1.88E-06/rx-yr as showin Table D.1-12. Additional details of the full level 2 model and its development aredescribed in the response to 2.b.2.b.1 ER Section D.1.4.4 indicates that a full level 2 model was created for the 2015 (R5) PSAbased on the 2015 internal events model. Describe the full level 2 model in comparisonwith the prior LERF only model reviewed in the 2009 peer review, the changes made to itto obtain the 2015 (R5) level 2 model and the steps taken to insure the technical adequacyof the full Level 2 model.Waterford 3 ResponseThe previous simplified LERF-only model following the WCAP-16341-P methodology wasused as the starting point for the detailed WF3 Level 2 model. The conversion of thesimplified LERF model into a Level 2 analysis for WF3 included the following: Restructuring the event trees for addition and consolidation of nodes; Execution and incorporation of plant-specific MAAP calculations in thedetermination the event tree outcomes; Development of 12 release categories beyond the LERF, SERF, LATE andINTACT end states; Incorporation of the WF3 Emergency Action Levels, evacuation estimates, andMAAP 4.0.6 accident sequence timing; Utilization of FP results derived from MAAP analyses in the binning of therelease categories; and Development and incorporation of detailed ultimate containment capacity intothe Level 2 analysis. toW3F1-2017-0001Page 11 of 74The WF3 model is a Level 2 analysis capable of meeting the Category II requirements ofRegulatory Guide 1.200 and the ASME PRA Standard. The updated Level 2 analysis usesavailable technical work from the previous WF3 PRA analyses where appropriate, butapplies the most recent accident progression research, current industry practices, andrealistic plant-specific analyses. The level 2 analysis was performed by Jensen-Hughesand received in-depth technical reviews within Jensen-Hughes and by a representative ofEntergy with level 2 experience and all comments were resolved. Also, an expert panelcutset review of the significant and non-significant cutsets for the level 2 model wasperformed and all issues addressed.2.b.2 ER Sections D.1.2.1, "Containment Performance Analysis," and D.1.2.2.6, "Mapping ofLevel 1 Results into the Various Release Categories," both provide discussions regardingthe transfer of Level 1 core damage results to the Level 2 fission product release analyses.The ER states:For the WF3 Level 2 analysis, no grouping into [Plant Damage States] PDS wasperformed to group accident sequences with similar safety features and containmentfailure responses. A more rigorous approach was taken where each Level 2 accidentsequence was assessed individually based on the accident-specific containmentresponse.The WF3 Level 2 accident sequences were named using the two or three letteridentification for the CD sequences from the Level 1 core damage event trees (i.e.,AX, MU, SB, TQX, TKQ, and RB) and combined with a one-letter code to representcore melt sequences (core damage with containment safeguard systems).Provide additional information on this process including a description of the Level 1 andLevel 2 sequence naming nomenclature and how the Level 2 sequences or ContainmentEvent Tree (CET) endpoints were assigned to the Level 2 release categories.Waterford 3 ResponseThe Level 2 accident sequences are derived directly from the WF3 Level 1 core damagesequences. All Level 1 core damage sequences were evaluated following the modificationof the Level 2 scenarios for further analysis and incorporation into the Level 2 model. Themodifications of the Level 2 scenarios are based on Level 1 core damage sequences thathave been extended for a 36-hour duration, 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> beyond the 24-hour core damageperiod used to establish success criteria. In addition, the Level 2 scenarios are evaluatedfor containment integrity response based on the operational configurations of the WF3containment safeguard systems. The MAAP results from the analyses performed on theLevel 2 scenarios are used to develop the Level 2 model. toW3F1-2017-0001Page 12 of 74As part of the development of the Level 2 model, a detailed evaluation of the WF3containment capacity was performed to assess the integrity response during Level 2accident progression. Based on the containment response timings and FP releasefractions obtained from the MAAP analyses, the Level 2 scenarios are assigned to theappropriate Release Categories (RC). The parameters used for the RC classificationinclude the timing of occurrence of any containment failure and the maximum FP fractionover the 36-hour MAAP analysis period.The Level 2 accident sequences are identified and named in accordance the WF3 coredamage sequences as establish as part of the Level 1 success criteria. The Level 1 coredamage sequence names are based on the performance of the safety function systems incombination with the sequence initiating event. For example, the RC H-E sequence isidentified as TQX_H. This scenario is classified as a having a transient initiating event"%T", with a failure of the event "Q" to maintain the RCS pressure boundary and a failureof the event "X" to maintain long-term RCS control during the recirculation phase. Thisnomenclature is consistent with that used for the WF3 IPE and the Level 1 PRA. Inaddition, the containment safeguard system configurations are denoted in a mannerconsistent with the IPE designations for combinations of containment fan and sprayfailures. A "TQX_H" sequence has no functioning containment fans or sprays. Othercontainment safeguard combinations of the TQX sequence include: "TQX_B" with bothcontainment fans and sprays operation, "TQX_D" with containment spray failed, and"TQX_F" with containment fan coolers failed.The WF3 Level 2 accident sequences allow each core damage sequence to be evaluatedfour times, once for each safeguard configuration, and assigned to a RC uniquely basedon both the initial accident sequence and its corresponding containment response.Because the assignment of Level 2 accident sequences is directly assessed for eachaccident scenario, the establishment of PDS based on core melt conditions as waspreviously used in the WF3 IPE was not required.2.c. ER Section D.1.2.1 states that 4 CETs were used to model the core melt progression andradioactive releases. Four trees, Trees B, D, F and H, representing four combinations ofcontainment heat removal, are subsequently discussed. Confirm that these are the fourCETs used and describe the use of the four trees considering that the two containmentheat removal systems are explicitly represented by CET nodes.Waterford 3 ResponseYes, these are the four CETs used. To provide a more accurate determination of the Level2 sequence response, different configurations of Containment Heat Removal (CHR)system performance were applied to each Level 1 sequence and evaluated independently.Each of the four trees represents a different configuration of containment heat removalsystem performance. toW3F1-2017-0001Page 13 of 74CET-B: Both Containment Sprays and Containment Cooling Fans are available(CHR-B)CET-D: Only Containment Cooling Fans are available (CHR-D)CET-F: Only Containment Sprays are available (CHR-F)CET-H: No Containment Safeguards are available (CHR-H)ER Sections D.1.2.1.1 and D.1.2.1.2 discuss the CFC (containment fans) and CS(containment sprays) top events, but these are not actual nodes in the CETs. Rather, theydefine the entry points for each of the trees. Each Level 1 sequence was evaluated usingeach of the four trees.2.d. ER Section D.1.2.2.7 indicates that for: Containment Bypass Sequences, ContainmentIsolation Sequences, Reactor Vessel Rupture Events and Interfacing System Loss ofCoolant Accident (LOCA) Events; there was no consideration of fission product (FP)scrubbing, retention, or deposition and all were assigned to the High-Early (H-E) releasecategory (RC). Clarify this statement since with no scrubbing, retention or deposition, 100percent release of volatile FPs would be expected.Waterford 3 ResponseThe discussion in D.1.2.2.7 addressed the characterization of the FP releases for thepurpose of binning these release scenarios based on severity. Containment BypassSequences, Containment Isolation Sequences, Reactor Vessel Rupture Events, andInterfacing System Loss of Coolant Accidents were classified as High-Early releases.These scenarios are classified as early releases because the initiating event failure leadsto an immediate release pathway from the containment structure. In addition, thesescenarios are classified as high severity because the containment release pathwayprecludes the mitigation or retention of fission products due to scrubbing, retention, ordeposition mechanisms that occur within the containment structure. MAAP analyses werenot performed for these High-Early sequences.2.e. ER Section D.1.5.2.9 states:The representative accident sequences selected for each release categoryrepresented both the dominant accident class based on the Level 2 results and themaximum release of fission products from the MAAP analyses.Provide a more detailed discussion of this process including a description of the Level 2sequences used to characterize the source terms for each of the significant releasecategories, the basis for this selection and its appropriateness for use in determining thebenefit for the Phase II SAMAs evaluated. Note that using the dominate sequence in eachRC to characterize the releases for that category may not necessarily lead to the correctbenefit for the individual SAMA cost-benefit analyses. toW3F1-2017-0001Page 14 of 74Waterford 3 ResponseThe process of selecting a representative accident sequence included a review of the riskimportance of the accident sequences in combination with the timing and severity of therelease.A cutset review of the Level 2 model quantification results was used to identifypredominant sequences. This review process included an evaluation to identify dominantaccident sequences that contribute either to the top cutsets or based on total cutsetfrequency within each release category.A review of the MAAP analyses of the accident sequences was conducted to identifysequences based on timing and release magnitude. A review of fission product sourceterms identified accident sequences with the highest source term based on CsI within theirassigned release category. Similarly, the initiation of containment release times wasreviewed to identify scenarios with the earliest release timing for their assigned releasecategory. The purpose of the review was to screen the sequences and capture potentialaccident sequences for additional review for each RC.Following the process of identifying and screening of potential accident sequences fromboth the cutset review and the MAAP analysis, an additional review of the candidatesequences was used to select an accident sequence for each release category that is bothconservative and representative of WF3.2.f. The start of release times given in ER Table D.1-10 are not consistent with the RCdefinitions in Table D.1-8 for a number of release categories. For example: for RC H-E(start of release less than 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> after general emergency declaration), the time of thestart of release (plume 1) is 13.4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> while the time of declaration of a generalemergency is 15 minutes; and for RC High - Intermediate (H-I) (start of release is greaterthan 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> after general emergency declaration), the time of release is 2.0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />.Provide a discussion of the reasons for these differences and the impact on the results ofthe SAMA analysis.Waterford 3 ResponseER Table D.1-8 provides the grouping of the release categories: early, intermediate andlate releases. These releases represent the time duration between the establishment ofWF3-specific conditions associated with a General Emergency (GE) and the time that theactual containment failure occurs.Time to General Emergency Conditions - The MAAP results are used to establish the timeto plant conditions used in the recognition and classification of emergencies. Plant criteriaused by WF3 to establish plant conditions for a General Emergencies are based on the toW3F1-2017-0001Page 15 of 74WF3 emergency response procedures. Using these plant criteria, the timing to reach theseconditions was extracted from the MAAP data to establish a scenario specific timing forGeneral Emergency conditions. The maximum time to meet these conditions defines thetime at which the WF3 plant and personnel recognize the plant-specific state of GeneralEmergency.Plume Release Time - The initiation time for a plume release is established using theMAAP results. Plume release times are determined as the time at which the containmenthas failed.The characterization of timings used to determine the Release Categories is a relativemeasure of time. The duration of time between the declaration of a General Emergencybased on plant criteria to the time of the plume release represents the characterizationtiming. Early releases are classified as having a containment release within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> ofplant General Emergency declaration.The plant-specific General Emergency timings used to classify the Level 2 scenarios arealso incorporated in the Level 3 model to support the timing of emergency response actionin context with the progress on the accident and plume release. The Level 3 modelparameter RDOALARM defines the declaration of GE by Waterford 3 to emergencyresponders and agencies. The RDOALARM used in the Level 3 model represented thetime for the plant to evolve to General Emergency conditions and also included a 15-minute assessment time for Waterford 3. The RDOALARM parameter was incorrectlyused to this assessment time in relation to the plant GE conditions rather than in context ofthe plant scram time. Values of the RDOALARM have been modified as shown in revisedTable D.1-10 and reflect the time between the recognition of GE conditions plus the 15-minute assessment time in relation to the initiating event (scram time).RevisedTable D.1-10 provides the timings associated with the release plumes used in the Level 3 analysis.The plume release times represent the time at which a release pathway opens(containment failure) and FP releases begin. Plume release times are determined inrelation to the start of the initiating event (reactor scram).The Level 3 model was updated using the correct RDOALARM times. Also, 48hr FPreleases were used as described in the response to 2.g and alternate accident sequenceswere selected to represent the M-I and L-I RC as described in the response to 4.b. Therevised Table D.1-12 is shown below. The updated mean values of PDR and OECR forWF3 are 17.1 person-rem/yr and $162,682/yr. toW3F1-2017-0001Page 16 of 74Table D.1-12Characteristics of Release ScenarioResults - Year 2010MRelease IDFrequency (per year)Population Dose1(person-sv)Offsite EconomicCost ($)Intact3.68E-061.28E+031.25E+08H-E1.88E-063.20E+042.88E+10H-I4.75E-062.20E+042.25E+10M-E2.74E-081.75E+041.56E+10M-I1.34E-078.94E+034.57E+09M-L1.84E-081.12E+048.54E+09L-I2.42E-095.70E+031.75E+09L-L5.56E-104.15E+037.86E+08LL-L3.85E-106.47E+032.74E+09Totals1.71E+01person-rem/yr1.63E+05$/yr1 Conversion Factor: 1 sv = 100 rem2.g. ER Section D.1.2.2.6 indicates that level 2 accident sequences were evaluateddeterministically using the Modular Accident Analysis Program (MAAP) 4.0.6 code and a36-hour accident time period, and that this time period was selected to ensure thatsufficient time was allotted to allow for late failures and to capture the peak steady-stateFP release concentrations. Provide support that the 36 hour4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> accident time period yieldsthe peak FP release over the 48 hour5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> time period beginning at the time of declaration of ageneral emergency. If the peak FP release does not occur using the 36 hour4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> accident timeperiod, discuss the impact on the SAMA analysis if the analysis is extended to 48 hour5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br />safter the declaration of a general emergency.Waterford 3 ResponseEach RC representative accident sequence was re-evaluated using the MAAP code over atime period extending for 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> following the declaration of the WF3 GeneralEmergency. This re-evaluation was performed to conservatively establish peak FPfractions. The Level 3 model was updated using the extended 48-hour MAAP FP fractions.Also, the RDOALARM times were corrected as described in the response to RAI SAMA 2.fand alternate accident sequences were selected to represent the M-I and L-I RC asdescribed in the response to RAI SAMA 4.b. The revised Table D.1-12 is shown above.The updated mean values of PDR and OECR for WF3 are 17.1 person-rem/yr and$162,682/yr. toW3F1-2017-0001Page 17 of 742.h. ER Table D.1-9 states that the frequency of the "intact" RC is obtained from the differencebetween the base CDF and the total of the other release categories. Provide the results forthe "intact" RC from the sum of the no containment failure containment event tree endstates. Discuss the impact of cut set truncation on the CDF and RC frequencies and thevalidity of the approach taken to determining the RC frequencies.Waterford 3 ResponseThe CDF and RC frequencies were both quantified at a truncation of 1E-11 andconvergence studies were performed on both the level 1 and level 2 model results. Thelevel 1 demonstrates CDF convergence (defined as a change of less than 5% per decade)at 1E-11. Also, the level 2 demonstrates a change of less than 5% at a 1E-11 truncation.The highest frequency release categories (H-E and H-I) also demonstrate a change of lessthan 5% at 1E-11 truncation. Thus no significant change in SAMAs would be expected byproviding the results for the "intact" RC from the sum of the no containment failurecontainment event tree end states vs taking the difference between the base CDF and thetotal of the other release categories. toW3F1-2017-0001Page 18 of 74RAI SAMA 33. Provide the following information with regard to the treatment and inclusion of external events inthe SAMA analysis. The basis for this request is as follows: Applicants for license renewal arerequired by 10 CFR 51.53(c)(3)(ii)(L) to consider SAMAs if not previously considered in anenvironmental impact assessment, related supplement, or environmental assessment for theplant. As part of its review of the WF3 SAMA analysis, NRC staff evaluates the applicant'streatment of external events in the PRA models. The requested information is needed in orderfor the NRC staff to reach a conclusion on the sufficiency of the applicant's PRA models forsupporting the SAMA evaluation.3.a. In response to NRC requests following the accident at the Fukushima Daiichi NuclearPower Plant, new seismic hazard curves have been developed for each nuclear powerplant site. The Entergy response to NRC staff RAIs on the WF3 NFPA 805 transition LAR(Agencywide Documents Access Management System (ADAMs) ML14162A506) providedan assessment of the seismic CDF that is different from that given in the integrated leakrate testing interval extension LAR used in the SAMA analysis. Provide an updated WF3seismic CDF using the approach of the NFPA 805 assessment but based on the new postFukushima hazard curves and discuss the impact of using this seismic CDF on the WF3SAMA analysis.Waterford 3 ResponseIn response to NRC staff RAIs on the WF3 NFPA 805 transition LAR (Enclosure 2 toLetter W3F1-2014-0025, dated June 11, 2014) Entergy provided an assessment of theseismic CDF that is different from that used in the SAMA analysis. This seismic CDF,9.02E-07/rx-yr, was based on the EPRI seismic hazard curves.Using the new post-Fukushima hazard curves to assess the seismic CDF in the samemanner as the NFPA 805 RAI response, would result in a seismic CDF of 6.48E-06/rx-yr.This would change the internal/external events multiplier from 3.02 to 3.57. The "Internaland External Benefit" values in revised Table D.2-2 use a multiplier of 3.57.3.b. As stated above, the revised Attachment W to the WF3 NFPA 805 LAR gives the internalevents CDF and LERF as 6.5E-06 per rx-year and 8.7E-08 per rx-year respectively. Thesevalues are approximately 60% of the results given for the 2015 (R5) PSA used for theSAMA analysis. If the 2015 (R5) value is the most appropriate for use in the licenserenewal applications (LRA), provide an assessment of the impact of this more recentinternal events model on the results of the fire PSA used in the SAMA analysis and theresulting impact on the SAMA analysis. toW3F1-2017-0001Page 19 of 74Waterford 3 ResponseAs explained in the response to 1.e, the CDF value differs from that given for the 2015(R5) PSA model used for the SAMA analysis because it is from a prior interim modelrevision which did not include the revision of the battery depletion modeling to includeprocedural direction to strip batteries to allow for extended battery life.The prior interimLERF model had a slightly lower value due to the update in the SGTR sequences whichare binned as LERF which saw an increase due to the change in values of thermal-induced SGTR and pressure-induced SGTR failure probabilities.As noted in Section D.1.3.2 of the ER, a fire CDF of 1.80E-05/rx-yr was used in calculatingthe SAMA internal/external events multiplier discussed in Section 4.15.1.4.4. This is thefire CDF reported in the revised Attachment W to the WF3 NFPA 805 LAR mentioned inthe question and was the most recent value at the time the SAMA analysis wasperformed. As noted in the response to question 1.e, the CDF for the 2015 (R5) PSAmodel used for the SAMA analysis is larger because that model includes the revision ofthe battery depletion modeling to includet procedural direction to strip batteries to allow forextended battery life. This change and the induced SGTR change for LERF in the internalevents model would not have such a significant impact on the Fire PRA model resultssince those model results are driven by fire-specific factors. Also, with the conservatismsincluded in both the fire PRA and the SAMA analysis, it is judged that the multiplierprovided in ER Section 4.15.1.4.4 is appropriate for the SAMA analysis.3.c. ER Section D.1.3.4 indicates that internal flooding is not included in the 2015 internalevents PSA used for the SAMA analysis. It is also stated that changes were made tointernal flooding analysis that allowed the internal flooding analysis to satisfy therequirements in the ASME Standard and RG 1.200. Provide further information on thisanalysis including consistency with the system modeling in the 2015 (R5) PSA, the impactof any differences on the internal flood CDF and the SAMA analysis and the process usedto insure the technical adequacy of the internal flooding analysis.Waterford 3 ResponseThe internal flooding analysis has not been updated since the peer review. The internalflooding analysis is documented in report PRA-W3-01-002, "W3 Internal FloodingAnalysis." The last major revision of this report, Revision 2, was completed in April 2008.In April 2009, editorial Revision 3 was completed to update the Appendix B walkdownnotes. Revision 3 of PRA-W3-01-002 was included in the 2009 peer review of the WF3internal events model and calculates a total CDF contribution of 2.48E-06/rx-yr frominternal floods. This value was used, along with external events CDF values, to calculatethe internal/external events multiplier for the SAMA analysis. The multiplier was utilizedbecause the current internal flooding model hasn't been integrated with the current internalevents model or the Level 2 and 3 models. toW3F1-2017-0001Page 20 of 74The contribution the flood scenarios make to the core damage frequency was calculatedby manipulating event trees and data prepared in quantifying other accident scenarios inthe 2003 (R3) PSA model. The sequence probabilities were then combined with initiatingevent (flooding) frequencies to determine the contribution of internal flooding to the coredamage frequency. The differences between the R3 PSA model and the R5 PSA modelare described in Sections D.1.4.3 and D.1.4.4 of the Environmental Report. The R3 CDFof 6.75E-06/rx-yr was increased to 1.05E-05/rx-yr in R5. This increase was predominantlydue to a revision of the battery depletion modeling to include procedural direction to stripbatteries to allow for extended battery life. This modeling increased the CDF fromsequences initiated by a loss of offsite power. Since the internal flooding CDF comes fromsequences initiated by internal floods, it would not be significantly impacted by this modelchange. Thus, there is no expected impact on the SAMA analysis.The PSA analyst performing the internal flooding analysis was an experienced, trainedprofessional and the analysis was reviewed by a second, experienced, trained PSAanalyst. The internal flooding analysis was performed in accordance with guidancedocuments existing at the time (ASME PRA standard ASME RA-Sb-2005, NRCRegulatory Guide 1.200 for Trial Use, April 2004, Draft Regulatory Guide DG-1161,September 2006, and draft EPRI guidance document, "Guidelines for Performance ofInternal Flooding Probabilistic Risk Assessment (IFPRA)", September 2006.) However,the internal flooding model hasn't been updated since the peer review and hasn't beenintegrated with the current internal events model or the Level 2 and 3 models. Therefore,the WF3 internal flooding model was not deemed to be technically adequate for analyzingindividual SAMA candidates. Thus, the internal flooding CDF was included with theexternal event CDF values to calculate the internal/external events multiplier for the SAMAanalysis.3.d. As discussed in the NRC staffs "Interim Staff Response to Reevaluated Flood Hazards" atWF3 dated April 12, 2016, there are a number of reevaluated flood hazards that exceedthe current design-basis. Provide a discussion of the current status of the WF3 MitigationStrategy Assessment (MSA) and integrated assessment or focused evaluation, and adiscussion of the impact of flood hazards on the WF3 risk. Provide support for the ER'sconclusion that flood hazards are negligible and need not be included in the externalevents multiplier.Waterford 3 ResponseThe WF3 Mitigation Strategy Assessment has been completed and concluded that theWF3 FLEX design basis flood is not affected by the results of the Mitigating Strategy FloodHazard Information (MSFHI). Specifically, the following flood mechanisms, which boundthe reevaluated flood hazards that exceed the current design basis, do not impact the siteFLEX strategies. toW3F1-2017-0001Page 21 of 74local intense precipitation eventprobable maximum flood on the Mississippi Riverprobable maximum flood combined with an hypothetical dam break within theMississippi River and levee failure at WF3combined event H.3 (alternative 3): 25-year flood in the Mississippi River,probable maximum surge including antecedent water level, levee failure, andcoincident wind-generated wavesTherefore, the current FLEX strategies can be fully deployed with no additional operatoractions. The letter documenting these results was sent to the NRC on 11/14/2016(ML16319A089).The focused evaluation has not yet commenced at WF3. No appreciable impact on risk isexpected due to interim actions taken as part of the Flood Hazard Re-evaluation to beginsump pump actuation promptly and utilize weather warning time triggers. toW3F1-2017-0001Page 22 of 74RAI SAMA 44.Please provide the following information regarding the Level 3 PRA used in the SAMA analysis.The basis for this request is as follows: Applicants for license renewal are required by 10 CFR51.53(c)(3)(ii)(L) to consider SAMAs, if not previously considered, in an environmental impactassessment, related supplement, or environmental assessment for the plant. As part of its reviewof the WF3 SAMA analyses, NRC staff evaluates the applicant's analysis of accidentconsequences in the Level 3 PRA. The requested information is needed in order for the NRCstaff to reach a conclusion on the sufficiency of the applicant's Level 3 PRA model for supportingthe SAMA evaluations.4.a.ER Table D.1-10 includes a time to declaration of general emergency (GE) and a warningtime that is said to include a 15 minute GE declaration. The GE declaration time is 15minutes for all release categories while the warning time ranges from 15 minutes to 9hours. Discuss the use of these times in the consequence analysis and how they weredetermined. The GE declaration time would be expected to be sequence specific andbased on site procedures.Waterford 3 ResponseRDOALARM is utilized in the WinMACCS code to represent the initiation of the movementof the cohorts. Within the WinMACCS code, the value of RDOALARM provides therequired context with both the timing of the accident scenarios and the progression ofemergency response measure.WF3 procedures facilitate the determination of general emergency (GE) conditions. Usingthe WF3 procedures, a determination of the timing for GE conditions is evaluated basedon degraded fission product barriers. To establish GE conditions, an assessment ofconditions that lead to the loss of two barriers with a potential loss of the third barrier isrequired for declaration of a WF3 General Emergency.MAAP results were used to establish the timing of the loss or potential loss of the fissionproduct barriers (GE condition) during accident progression. Timings were establishedfrom MAAP results for the loss or potential loss of each barrier. The maximum time toachieve a plant GE condition was used to represent the scenario-specific GE condition foreach release scenario.WF3 recognizes a 15-minute "assessment" period following the indication of plantconditions to a General Emergency and in advance of notification to offsite responders orpopulations. The start of this assessment period is coincident with the timing of the WF3plant GE conditions. The emergency director is expected to make the emergencydeclaration promptly without waiting for the 15-minute period to elapse once theemergency action level is recognized as being exceeded. However, since the declarationof a GE could take up to 15 minutes beyond the time that the necessary plant conditionsare reached, the sum of the time to achieve plant GE conditions and the 15-minuteassessment period is conservatively defined as the earliest time that offsite response toW3F1-2017-0001Page 23 of 74actions associated with sheltering or evacuation may occur. This is the timing value of theWinMACCS parameter RDOALARM.Updates to the plume characteristics and accident progress timing have been made toinclude this 15-minute "assessment" period in context with the onset of plant GeneralEmergency conditions with the WinMACCS parameter RDOALARM. The revised TableD.1-10 provides updates to the WinMACCS parameter RDOALARM.4.b. The NRC staff notes in ER Table D.1-12 that while the population dose for the Low-Intermediate (L-I) RC is greater than that for the High-Early (H-E) RC, the cesium andIodine release fractions given in Table D.1-10 are about 4 percent of those for the H-E RC.Similarly, while the population dose for the Moderate-Intermediate RC is higher than thatfor the H-E RC, the cesium and iodine release fractions are about 33 percent and 12percent of those for the H-E RC, respectively. Explain the reason for this unexpectedresult and the impact on the SAMA cost-benefit analysis. As part of this explanation and tothe extent applicable, summarize the treatment of relevant release characteristics (e.g.,energy of release, source term, etc.) used to define each RC.Waterford 3 ResponseA screening and review of the accident scenarios associated with each RC was performedfor the purpose of selecting both a conservative and representative sequence to bestrepresent each RC. Two Level 3 RC scenarios, M-I and L-I, represent a conservativeaccident scenario based onenergy of release and source terms derived from the MAAPanalyses. However, these scenarios generate population doses that are greater thanscenarios with higher or earlier release scenarios (H-E).The scenarios selected for the RC M-I and L-I represent early phase in-vessel core meltconditions under high RCS pressure (> 200 psi) and partially recovered/maintained reactorpressure vessel (RPV) water levels. Under these conditions, increased production ofsteam and hydrogen enhances fission product releases from the fuel rods and other corematerials. These in-vessel conditions facilitate the release of the alkaline and rare earth(non-volatile) isotopes from fuel fracturing or powdering. A review of the FP fractions forboth the M-I and L-I accident sequences show these sequences to be outliers with regardto the ratio of barium to iodide in comparison to the other accident scenarios. In the caseof the L-I scenario, barium exceeded iodide at a ratio of 6.61:1.Alternate accident sequences were selected to represent the M-I and L-I RC and wereused in the Level 3 model. Results obtained from the Level 3 model using scenarios withrepresentative ratios of alkaline and rare earth isotopes are shown in the revised TableD.1-12 in the response to RAI SAMA 2.f. By replacing the outlier accident sequences, theM-I and L-I RC have PDR less than earlier RC such as H-E.Since the outlier sequences are not being used in the revised SAMA cost-benefit analysis,a discussion of their impact on the SAMA cost-benefit analysis is not provided. toW3F1-2017-0001Page 24 of 744.c. ER Table D.1-11 provides the estimated core inventory input to the Level 3 analysis;however, there is no description regarding how this input was developed. Clarify that thecore inventory estimates applied in support of the Level 3 analysis are specific to WF3.Additionally, clarify whether additional adjustments of the core inventory values arenecessary to account for differences between fuel cycles expected during the period ofextended operation and the fuel cycle upon which the Level 3 analysis is based (e.g., toaccount for any changes in future fuel management practices or fuel design).Waterford 3 ResponseThe reactor core inventory used in the Level 3 analysis is based on long-term operation ata core thermal power level of 3,735 MWt (100.5% of the extended full power uprate of3,716 MWt). This core inventory is the LOCA Alternative Source Term in the WF3 FSAR[FSAR, Table 12.2-12] and is specific to WF3.The core inventory represents a bounding core design based on five power uprate fuelmanagement cycles. The bounding core inventory was determined using ORIGEN2 andbased on a cycle length, rated thermal power, assembly loading, average core enrichmentand feed batch size. Based on sensitivity analyses, contributors to the highest coreactivities included low enrichment, longest cycle length and highest quantity of onceburned fuel. Entergy believes that the core inventory values account for expected fuelmanagement practices and fuel design during the period of extended operation.4.d. Regarding ER Section D.1.5.3, the NRC staff notes that the consequence analysisassumed site-specific meteorological data from year 2010, given that it generated thehighest population dose and offsite economic cost. However, Section D.1.5.2.6 indicatesthat certain meteorological data, including that for year 2010, was not available and wasaddressed, at least in part, by using "data from approved data substitution methods asneeded." Quantify the amount of missing meteorological data, which were estimated usingdata substitution, and clarify the methods used.Waterford 3 ResponseData substitution methods were used to augment missing meteorological data for 2010.Missing data for 2010 include the following: 153 hours0.00177 days <br />0.0425 hours <br />2.529762e-4 weeks <br />5.82165e-5 months <br /> of wind speed, wind direction and temperature difference data (1.7% ofannual data); and 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> of precipitation data (0.2% of annual data)Data substitution methods were used to augment missing meteorological data for 2011.Missing data for 2011 include the following: toW3F1-2017-0001Page 25 of 74 64 hours7.407407e-4 days <br />0.0178 hours <br />1.058201e-4 weeks <br />2.4352e-5 months <br /> of wind speed, wind direction and temperature difference data (<0.1% ofannual data); and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of precipitation data (<0.1% of annual data)Precipitation data was missing for a few scattered hours (<0.1% of annual data) in 2005,2008, and 2009. The data for 2013 was missing all the hourly precipitation data (100% ofannual data). The missing precipitation data was zero-filled as discussed below.Missing data was completed using approved substitution methods (Atkinson and Lee1992). Data substitution methods used to complete the missing wind speed, wind directionand temperature difference data included: Replacement of primary tower data with secondary tower data (method utilized for6% of data replacement); and Substitution with data from previous year that is representative (method utilized for94% of data replacement)Data substitution methods used to complete the missing precipitation data included: Zero filling precipitation data. Zero filling the precipitation data results in aconservative result from the WinMACCS model (method utilized for 100% of datareplacement)4.e. ER Section D.1.5.2.1 discusses population data. Explain why the population distributionused in the analysis is appropriate, and justify the method used for populationextrapolation. In doing so, describe how those parishes with declining populationprojections were addressed (if applicable). Additionally, clarify whether transient andspecial facility populations were included, and if not, justify their exclusion.Waterford 3 ResponseThe US Census 2010 data was used as the basis for the population projections. Parishlevel population projection estimates were obtained from Woods & Poole Economics, Inc.This data provided population projection estimates for every year from 2011 through 2040.A cohort component method with an employment driven migration model was used byWoods & Poole Economics, Inc. for Louisiana to establish population projections byparish. Excel was used to establish the growth trends in known parish populationprojections. Linear and 2nd order polynomial regression equations were used, asappropriate, to extend the trend past the given projection dates to create projectedpopulation data to the half decade year (2045) following the end of the WF3 period ofextended operation (2044). This method is appropriate, and conservative, since populationprojections beyond the period of extended operation were used for the SAMA analysis,while NEI-05-01A recommends projecting the population to a time within the second halfof the period of extended operation. toW3F1-2017-0001Page 26 of 74For parishes with a projected negative population growth, the highest estimated populationvalue was held constant for the remaining portion of the projection period. One Louisiana parish, Orleans Parish, was shown to have a declining populationtrend. To produce conservative population estimates for this parish throughout theprojection period, the population value for 2011 was held constant. The graphs for Iberville, Plaquemines, St. Helena, St. James, and St. Maryparishes illustrated a growth tendency in the beginning of the projection, then adecline in growth rate and, in some cases, a decline in population. Linearregression did not fit the data for these three parishes and the R squared valueswere low. Second order polynomial regression equations fit the data better andhad better R squared values. However, when projected to the end of the timeperiod, the population tended to decline. For these five parishes, a composite linewas used. This line used the second order polynomial regression and, followingthe highest year of projected population, that year's data was held constant.Because the composite curve was not allowed to decline below the highestprojected population data point, the results are conservative.Transient populations were included. Transient population data for Louisiana wasobtained from official state sources (The University of New Orleans, Hospitality ResearchCenter), 2014. "Louisiana Tourism Forecast 2014 - 2017", prepared for LouisianaDepartment of Culture, Recreation and Tourism. Retrieved from<http://www.crt.state.la.us/tourism/louisiana-research/latestresearch/index>, (May 14,2014). Using corresponding years, a ratio of the transient population value and the parishpermanent population value was produced for each parish. A transient ratio wascalculated for 2013, 2014, 2105, 2016, and 2017, but the 2013 ratio was used as it wasthe largest. This ratio was assumed to be constant, and was used to project the transientpopulation to 2045 values, because it is an economic value which is proportional topopulation changes.Special facility populations were included as part of the permanent population counted inthe US Census 2010 data.4.f. ER Section D.1.5.2 describes the assumptions used for many of the parameters applied insupport of the Level 3 analysis, but significant gaps exist in the information provided.Specifically, the guidance in Section 3.4.2 of NEI 05-01, "SAMA Analysis GuidanceDocument," identifies several economic parameters utilized in the WinMACCS model thatare not discussed (e.g., cost of evacuation, cost of temporary relocation, cost of landdecontamination, etc.). Describe how each of these cost parameters were developed, andprovide the values and technical basis for any inflation/escalation factors utilized. toW3F1-2017-0001Page 27 of 74Waterford 3 ResponseThe current parameter values in the following table for CHEVACST, CHRELCST,CHCDFRM, CHCDNFRM, CHDLBCST, and CHPOPCST, are based on recommendedcost values in 1987 dollars and are adjusted to represent the increase in values to presentpricing values. These adjustments were calculated using average U.S. consumer pricesindices. A proportional factor of 2.08 was developed using the annual average for 1987CPI (113.6) and the annual average for 2014 CPI (236.736). This CPI factor was appliedto the 1987 values of the following parameters to represent current values as shownbelow.The 1987 values are based on the parameters described in NUREG-1150, "SevereAccident Risks: An Assessment for Five U.S. Nuclear Power Plants" and NUREG/CR-3673, "Economic Risks of Nuclear Power Reactor Accidents" and are consistent with theinputs used in MACCS2 Sample Problem A. These values have been subject to extensivepeer review since the late 1980s and continue to be used in licensee PRA and SAMAanalyses and state-of-the-art severe accident analyses. In these analyses, the values arecommonly escalated to account for inflation. The base values are considered applicable tothe region around WF3 because five nuclear power plants were analyzed in NUREG-1150, some of which have comparable or larger population densities than the regionaround WF3. For example, the Zion plant was part of NUREG-1150, is just north ofChicago, and has a large population density. Since economic parameters tend to increasewith population density, and the base values take into account sites with larger populationdensity, the values are considered applicable to the region around WF3.MACCSParameterDescription1987ParameterValueCurrentParameterValue(Updated to12/2014)CHEVACST001 Cost of evacuation2756.27CHRELCST001 Cost for temporary relocation2756.27CHCDFRM0001Farmland decontamination cost (dose reduction factorof 3)562.51,172Farmland decontamination cost (dose reduction factorof 15)1,2502,605CHCDNFRM001Non-farmland decontamination cost (dose reductionfactor of 3)3,0006,252Non-farmland decontamination cost (dose reductionfactor of 15)8,00016,672CHDLBCST001 Labor cost of a decontamination worker35,00072,938CHPOPCST001 Per capita removal cost for temporary or permanentrelocation of population and businesses in a regionrendered uninhabitable during the long-term phasetime period.5,00010,420 toW3F1-2017-0001Page 28 of 744.g. NUREG-1530, Revision 1, Reassessment of NRC's Dollar per Person-Rem ConversionFactor Policy (DRAFT) is publicly available in ADAMs at ML15049A114. SinceCommission approval of the NUREG is expected by the middle of 2017, this would be newand significant information that would need to be evaluated before the WF3 licenserenewal is issued. WF3 used the old value of $2000 per person-rem in the current SAMAanalysis. In anticipation of this change, please provide a sensitivity analysis using theanticipated new value of $5,200 per person-rem.Waterford 3 ResponseA sensitivity analysis was performed that utilizes the value of $5,200 per person-remrather than a value of $2000 per person-rem. This resulted in one additional potentiallycost beneficial SAMA (SAMA 9 to add a new backup source of diesel cooling). The resultsof this sensitivity can be seen in Table D.2-4.4.h. On May 4, 2016, the Commission issued a decision (CLI 16-07) in the Indian Point licenserenewal proceeding, in which it directed the Staff to supplement the Indian Point SAMAanalysis with sensitivity analyses. Specifically, the Commission held that documentationwas lacking for two inputs (TIMDEC and CDNFRM) used in the MACCS computeranalyses, and that uncertainties in those input values could potentially affect the SAMAanalysis cost-benefit conclusions. The Commission therefore directed the Staff to performadditional sensitivity analyses.The two inputs (TIMDEC and CDNFRM) are commonly used in the SAMA analysesperformed for LRAs. These two input values were generally based on the values providedin NUREG 1150, "Severe Accident Risks: An Assessment for Five U.S. Nuclear PowerPlants" and NUREG/CR 3673, "Economic Risks of Nuclear Power Reactor Accidents."The TIMDEC input value defines the time required for completing decontamination to aspecified degree. The CDNFRM input parameter defines the cost (on a per person basis)of decontaminating non-farmland by a specified decontamination factor. The CDNFRMvalues used in NUREG 1150 ($3,000/person for decontamination factor of 3 and$8,000/person for decontamination factor of 15) stem from decontamination costestimates provided in NUREG/CR 3673, the same 1984 economic risk study referenced insupport of the decontamination time inputs. These decontamination cost inputs arecommonly escalated to account for inflation.The NRC Staff believes the Commission's decision in CLI 16-07 may be applicable to theSAMA analysis performed for WF3, inasmuch as that analysis may have also relied uponthe NUREG 1150 values for TIMDEC and CDNFRM. We therefore request that Entergyeither justify why CLI 16-07 does not apply to the SAMA analysis performed for WF3 orsupplement the SAMA analysis with sensitivity analyses for the CDNFRM and TIMDECvalues. Entergy is requested to review the input values specified in CLI 16-07 for the toW3F1-2017-0001Page 29 of 74Indian Point LRA, and (1) to apply the maximum values specified by the Commission (oneyear (365 days) for TIMDEC and $100,000 for the CDNFRM values for thedecontamination factor of 15) or, in the alternative, (2) to explain, with sufficientjustification, its rationale for choosing any other value(s) for its sensitivity analyses. In anyevent, Entergy should execute sensitivity analyses for the release categories modeled thatexceed 1015 Becquerels of Cs 137 released. Entergy is requested to evaluate how thesesensitivity analyses may affect its identification of potentially cost-beneficial SAMAs.Finally, upon completing its sensitivity analysis, Entergy is requested to submit thespreadsheet (or equivalent table if another method is used) that conveys the populationdose and off-site economic cost for each release category and integrates the results into aPopulation Dose Risk and an Offsite Economic Cost Risk for WF3.Waterford 3 ResponseTo address this, Entergy has chosen to use Option 1, and a new WinMACCS TIMDECand CDNFRM sensitivity case was developed with the following changes (as compared tothe WF3 WinMACCS base case documented in the Environmental Report(ML16088A335)):TIMDEC was escalated to one year (365 days) for decontamination factor (DF) =15CDNFRM was escalated to $100,000/person for DF=15These changes were applied to all release categories (even those with total releasesof Cs-137 below 1015 Becquerels).Both the conditional and frequency weighted WinMACCS results of this sensitivity case foroffsite dose and economic cost are presented in the table below for each releasecategory, as the audit question requests. The frequencies used to weight the results arethose from the Environmental Report (ER) Table D.1-12. For the specified TIMDEC andCDNFRM input changes, the WinMACCS Offsite Economic Cost Risk (OECR) increasedapproximately 98% and the Population Dose Risk (PDR) increased approximately 31% ascompared to the WinMACCS base case results of $1.63E+5/yr and 17.1 person-rem/yr(presented in updated Table D.1-12 in the response to RAI SAMA 2.f). The increasesseen in the OECR and PDR for this TIMDEC and CDNFRM sensitivity case are boundedby the 95th percentile uncertainty factor of 2.06, which was included as part of the SAMAcandidate cost-benefit evaluation. Therefore, no new SAMA candidates are identified aspotentially cost-beneficial based on this new TIMDEC and CDNFRM sensitivity case.There are no changes to the conclusions of the SAMA analysis based on the TIMDEC andCDNFRM sensitivity case. toW3F1-2017-0001Page 30 of 74IDFrequency(per year)Baseline(person-sv)ECON(person-sv)Baseline($)ECON($)Intact3.68E-061.28E+031.28E+031.25E+086.29E+08H-E1.88E-063.20E+044.08E+042.88E+105.10E+10H-I4.75E-062.20E+042.96E+042.25E+104.61E+10M-E2.74E-081.75E+042.30E+041.56E+104.01E+10M-I1.34E-078.94E+039.85E+034.57E+091.82E+10M-L1.84E-081.12E+041.31E+048.54E+093.06E+10L-I2.42E-095.70E+035.96E+031.75E+097.33E+09L-L5.56E-104.15E+034.38E+037.86E+083.35E+09LL-L3.85E-106.47E+036.91E+032.74E+091.09E+10Totals1.71E+012.24E+011.63E+053.21E+05% ChangeNA30.92%NA97.50%Unitsperson-rem/yrperson-rem/yr$/yr$/yr1Conversion Factor: 1 sv = 100 remCharacteristics ofRelease ModePopulation Dose(person-sv)1 Offsite EconomicCost ($) toW3F1-2017-0001Page 31 of 74RAI SAMA 55. Provide the following information with regard to the selection and screening of Phase I SAMAcandidates. The basis for this request is as follows: Applicants for license renewal are requiredby 10 CFR 51.53(c)(3)(ii)(L) to consider SAMAs if not previously considered in an environmentalimpact assessment, related supplement, or environmental assessment for the plant. As part ofits review of the WF3 SAMA analysis, NRC staff evaluates the applicant's basis for the selectionand screening Phase I SAMA candidates. The requested information is needed in order for theNRC staff to reach a conclusion on the adequacy of the applicant's Phase I SAMA selection andscreening process for the SAMA evaluation.5.a. Based on the review of importance analysis in ER Tables D.1-2:5.a.i. The risk reduction worth (RRW) for event %TAC3, "Loss of 4.16Kv Bus 3A3-S"(1.0914), is considerably less than that for %TAC4, "Loss of 4.16Kv Bus 3B3-S"(1.318). Explain the reasons for this difference and consider a potential SAMAthat addresses the cause of this difference.Waterford 3 ResponseThis difference is attributed to an asymmetry related to component cooling water(CCW). If a safety injection actuation signal (SIAS) occurs, the CCW systemautomatically splits into two independent trains by closing the CCW pump suctionand discharge cross-connect valves and the train B supply and return valves tothe AB header. Flow to the AB header (i.e., reactor coolant pumps (RCPs))continues to be supplied from train A. Thus, following an SIAS, CCW train Bdoes not provide flow to the RCPs. Thus, when initiator %TAC4 (Loss of 4.16kVBus 3B3-S) occurs and the operators fail to align CCW train AB and fail to trip theRCPs, it leads to a failure of the RCP seals. Phase II SAMA 5 was evaluated toaddress this asymmetry. Phase II SAMA 77 which was proposed, evaluated, andfound potentially cost-beneficial as presented in RAI SAMA 7.b (to provide diversebackup auto-start signals for the standby CCW trains on loss of the running train)also mitigates this failure.5.a.ii. Event ZDHFBAT_LSP, "Failure to shed loads on the A or B battery," is failure of ahuman action and is addressed by several hardware related SAMAs. Discuss thepotential for SAMAs relating to improvements in procedures and training to reducethe impact of this human error and other human error events (e.g. Events ZHF-C2-011).Waterford 3 ResponseThe RRW tables (ER Tables D.1-2, D.1-4, and D.1-5) are intended to show thePhase II SAMAs which were evaluated in the cost-benefit analyses, which wouldmitigate each of the important events. Many enhancements to procedures andadditional training to reduce the impact of human errors were also considered for toW3F1-2017-0001Page 32 of 74human actions in the RRW tables, but were screened out during Phase I and are,therefore, not listed in the RRW tables.ZDHFBAT_LSP, Failure to shed loads on the A or B battery, is a good example.NEI-05-01A recommends a SAMA candidate to "Improve DC bus loadshedding." During the Phase I screening analysis, procedure OP-902-005,"Station Blackout Recovery Procedure," was reviewed to determine ifimprovements could be made. Since no improvements were identified, thiscandidate SAMA was screened as "already installed."ZHF-C2-011, is a combination of two human actions, failure to align CCW train ABto replace lost train A or B and failure to trip RCPs after loss of seal cooling, and isanother good example.For the first human action, failure to align CCW, NEI-05-01A recommends aSAMA candidate to "Enhance procedural guidance for use of cross- tiedcomponent cooling or service water pumps." During the Phase I screeninganalysis, procedure OP-901-510, "Component Cooling Water SystemMalfunction," was reviewed to determine if improvements could be made. Noimprovements were identified, so the CCW portion of the candidate SAMAwas considered "already installed." (Since this SAMA candidate addressestwo systems, it was retained to evaluate adding the ability to cross-tie theACCW pumps and was found to not be cost-beneficial (Phase II SAMA 21).For the second human action, failure to trip RCPs after loss of seal cooling,NEI-05-01A recommends a SAMA candidate, "On loss of essential rawcooling water, proceduralize shedding component cooling water loads toextend the component cooling water heat-up time." During the Phase Iscreening analysis, procedure OP-901-510, "Component Cooling WaterSystem Malfunction," was reviewed to determine if improvements could bemade. No improvements were identified, so the CCW portion of the candidateSAMA was considered "already installed."In addition, for the second human action, failure to trip RCPs after loss of sealcooling, another SAMA candidate, "Human actions to automatically trip theRCP on loss of CCW" was found in the Sequoyah SAMA analysis. During thePhase I screening analysis, procedure OP-901-510, "Component CoolingWater System Malfunction," was reviewed to determine if improvements couldbe made. No improvements were identified, so the candidate SAMA wasconsidered "already installed."Phase I SAMA candidates related to training were also investigated todetermine if additional training would mitigate high RRW events. Examplesinclude the following. toW3F1-2017-0001Page 33 of 74 Increase training on response to loss of two 120V AC buses whichcauses inadvertent actuation signals. In training, emphasize steps in recovery of offsite power after an SBO. Emphasize timely recirculation alignment in operator training. Additional training on loss of component cooling water. Improve operator training on ISLOCA coping. Increase training and operating experience feedback to improveoperator response. Institute simulator training for severe accident scenarios.During the Phase I screening analysis, procedure EN-TQ-114, "LicensedOperator Requalification Training Program Description," was reviewed todetermine if significant improvements could be made. The operators arerepeatedly trained on risk-significant actions. Classroom exercises andsimulator training are provided on these actions as well as on implementationof the severe accident guidelines. Severe accident scenarios are alsodeveloped for emergency planning exercises. The need for improvements inthis area was not identified.5.b. ER Section D.1.3.4 indicates that, while the internal flooding analysis is not integrated withthe internal events analysis, changes were made to the internal flooding analysis thatallowed the internal flooding analysis to satisfy the requirements in the ASME Standardand RG 1.200. Two SAMAs, SAMA 67, "Improve internal flooding response proceduresand training to improve the response to internal flooding events," and SAMA 68, "Installflood doors to prevent water propagation in the electric board room," were included in thein the Phase II evaluation. Provide a discussion of the identification of additionalcandidate SAMAs for mitigating internal flooding risk based on review of importantcontributors to the internal flooding CDF.Waterford 3 ResponseIn addition to phase II SAMAs 67 and 68, a number of phase I candidate SAMAs related tointernal flooding, from NEI-05-01A, were considered and found to be non-applicable oralready installed. Phase I SAMA candidates from other plants, listed below, were alsoconsidered. These SAMA candidates, and the two that were retained for evaluation, werecompared with the internal flooding scenarios to determine if the candidates wouldsignificantly mitigate the internal flooding CDF. The SAMA candidates were consideredglobally, rather than specifically. Phase II SAMAs 67 and 68 were found to be potentiallysignificant and were retained for evaluation, but the others were not. toW3F1-2017-0001Page 34 of 74 Install spray protection on motor-driven AFW pumps and space coolers. Install a globe valve or flow limiting orifice upstream in the fire protection system. Protect important equipment in the turbine building from internal flooding. Install spray protection on component cooling pumps and space coolers. Modifications to lessen impact of internal flooding path through Control Buildingdumbwaiter. Replace mercury switches in fire protection system to decrease probability ofspurious fire suppression system actuation.In addition to considering these phase I SAMAs, the internal flooding analysis wasreviewed to identify significant unique vulnerabilities that WF3 has to internal flooding. Aflood in the Reactor Auxiliary Building which propagates between Electrical Switchgearrooms A, B, and AB has the largest scenario contribution to the WF3 internal flooding andwas identified as a vulnerability. SAMA 68 to "Install flood doors to prevent waterpropagation in the electric board room" was evaluated to address this vulnerability andwas not found to be cost-beneficial.5.c. The ER indicates that the WF3 fire PRA was utilized to identify potential SAMAs. Threefire related SAMAs (74, 75 and 76) are included in the SAMA analysis based on theirbeing commitments in the WF3 NFPA 805 LAR. The WF3 fire PRA model gives a CDF forinternal fires that is 1.7 times higher than the internal events CDF after crediting thesecommitments. Provide a discussion of the identification of other candidate SAMAs formitigating internal fire risk based on review of important contributors to the internal fireCDF.Waterford 3 ResponseThe three fire-related SAMAs included in the SAMA analysis because they werecommitments in the WF3 NFPA-805 LAR have been changed to the following based onthe NFPA 805 LAR supplement letter dated January 18, 2016. These SAMAs havealready been implemented. Update six fire area heat detectors that have incorrect trip set points. Remove personnel offices and other combustible materials in Fire Area RAB 27(RAB+7). In Fire Area RAB 6 install a 1-hour fire resistance rating ERFBS fire wrap barrierfrom fire damage.Similar to the review for internal flooding, described in the previous response, a number ofphase I candidate SAMAs related to internal fires, from NEI-05-01A and from other plants,were considered and found to be non-applicable or already installed. Some examples arelisted below. These SAMA candidates were considered globally, rather than specifically. toW3F1-2017-0001Page 35 of 74 Upgrade fire compartment barriers. Install additional transfer and isolation switches. Enhance control of combustibles and ignition sources. Install lower amperage fuses for various 14 American wire gauge (AWG) controlcircuits in the main control room (MCR). Modify quick response sprinkler heads in cable chases A-11, C-30, and C-31 Install redundant fuses and isolation switches for MCR evacuation procedure Add fire wrap to the B Chilled Water cables in the vicinity of the A Chiller.In addition to considering these phase I SAMAs, the significant fire scenarios werereviewed for significant unique vulnerabilities, but no additional SAMAs were identified tomitigate the fire risk at WF3. No fire-related SAMAs were retained for cost-benefitevaluation.5.d. The disposition of Individual Plant Examination (IPE) and Individual Plant Examination ofExternal Events (IPEEE) insights is given in Table D.2-1.5.d.i. Phase I SAMA 184, "Install a portable generator to charge the AB battery," isscreened out as "already installed". The stated disposition indicates that the intentof this SAMA is met by the ability to manually control the turbine-drivenemergency feedwater pump after loss of direct current (DC). Provide theimportance of this human action and the potential for a SAMA involving the use ofa portable generator.Waterford 3 ResponseThe operator action to manually control the turbine-driven EFW pump is notcredited in the version of the PRA model utilized for the SAMA analysis. The intentof this SAMA is to use a portable generator that can continue to supply DC powerto the EFW turbine driven pump controls (and necessary monitoringinstrumentation) to decrease the likelihood of core melt before AC power isrestored. The intent of the SAMA is already installed by implementation of theFLEX strategy to manually control the turbine-driven emergency feedwater pump.Phase II SAMA 7 evaluated a similar modification to install a gas turbine generatorwhich was retained as cost beneficial.5.d.ii. Phase I SAMA 185, "Add guidance for aligning the low pressure safety injection(LPSI) pump for containment spray," is screened out because it is "alreadyinstalled." The procedure implemented is stated to address use of LPSI pumps forcontainment spray only for Large LOCAs. Discuss the benefit of this SAMA forother LOCAs or transients. toW3F1-2017-0001Page 36 of 74Waterford 3 ResponseThe procedural guidance to align the LPSI pump for containment spray is astandard appendix to the emergency operating procedures (EOPs). Standardappendices are used for evolutions that are called-out by several different EOPswhen conditions warrant. Thus, this guidance can be utilized if both containmentspray pumps are not available and high containment pressure exists, not onlyduring large LOCAs. The disposition for phase I SAMA 185 has been updated toavoid confusion to state, "Using LPSI to replace containment spray isproceduralized in OP-902-009, Attachment 28."5.e. Identify the number of Phase I SAMA candidates identified from the various sources (i.e.NEI 05-01 Generic List, other industry documents of PWR SAMAs, the WF3 IPE andIPEEE, plant specific internal events importance analysis and other sources). If the totalnumber of Phase 1 SAMA candidates is different than the 201 identified in Section D.2.1of the ER, then provide an explanation for this difference.Waterford 3 ResponseThe total number of Phase I SAMA candidates identified from various sources (i.e. NEI-05-01A Generic List, other industry documents of PWR SAMAs, the WF3 IPE and IPEEE,plant specific internal events importance analysis and other sources) is 202. Thebreakdown by source is as follows. NEI-05-01A Generic List153(Some of the NEI-05-01A SAMA candidates were also potentially cost-beneficial in otherPWR SAMAs, but are not counted in the next bullet.) Other PWR SAMAs32 Plant-Specific Fire Risk Analysis 3 Plant-Specific IPE8 Plant-Specific IPEEE5 NRC RAI SAMA 7.b1 toW3F1-2017-0001Page 37 of 745.f. Section D.1.2.1 states that Table D.1-5 provides the correlation between all level 2 releasestates RRW risk significant events down to 1.005 identified from the WF3 PRA Level 2model and the SAMAs evaluated in Section D.2. Clarify specifically which releasecategories are included in the importance analysis: all release categories, all except theintact RC, or all except intact and high-early release categories?Waterford 3 ResponseTable D.1-5 includes all release categories except the intact release category. Basicevents that are correlated in Tables D.1-2 (based on CDF) and D.1-4 (based on LERF)are not listed again in Table D.1-5.5.g. It is noted that the Phase II candidate SAMAs did not include adding an emergency dieselgenerator (EDG). Discuss why the cost-benefit of adding an EDG was not performed orprovide such an evaluation.Waterford 3 ResponseA phase I SAMA candidate to add an emergency diesel generator was evaluated anddetermined to be already installed. WF3 has two emergency diesel generators (EDGs)with each Diesel oil feed tank containing a minimum of 339 gallons of fuel, a separatediesel generator fuel oil storage tank, and a separate fuel transfer pump.WF3 also has "temporary" diesel generators (TEDs) that are staged on-site prior toremoving a permanent plant Emergency Diesel Generator (EDG) from service forextended preplanned maintenance work or prior to exceeding the 72-hour AOT forextended unplanned corrective maintenance work. When the TEDs are installed in placeof an out of service EDG, the TEDs are aligned in the event of a loss of offsite power andfailure of the operable EDG and can be started and ready to load within 25 minutes.In addition, WF3 has two FLEX diesel generators capable of supplying 400kW. One ispre-staged in an enclosure situated on the reactor auxiliary building (RAB) +41' el. roofand placed into service within 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of the onset of a beyond design basis externalevent, which is 30 minutes before the batteries deplete with the extended load shedstrategy. The other, N+1 FLEX diesel generator, is stored in the "N+1" storage building(south of the nuclear plant island structure) and can be swapped out with the FLEX dieselgenerator should the FLEX diesel generator become unavailable. The N+1 generator maybe pre-staged within the RAB due to hurricane or flood warning.Therefore, WF3 has many sources of power already installed and the cost-benefit ofadding another EDG was not evaluated. toW3F1-2017-0001Page 38 of 74RAI SAMA 66. Provide the following information with regard to the Phase II cost-benefit evaluations. The basisfor this request is as follows: Applicants for license renewal are required by 10 CFR51.53(c)(3)(ii)(L) to consider SAMAs if not previously considered in an environmental impactassessment, related supplement, or environmental assessment for the plant. As part of itsreview of the WF3 SAMA analysis, NRC staff evaluates the applicant's cost-benefit analysis ofPhase II SAMAs. The requested information is needed in order for the NRC staff to reach aconclusion on the acceptability of the applicant's cost estimations for individual SAMAs andcost-benefit evaluation.6.a. The benefit of SAMA 31, "Install a digital feedwater upgrade," is addressed by Case 2,"Improve Feedwater Reliability." Case 2 was evaluated by eliminating the loss offeedwater initiating event. Discuss the added benefit that might occur if the upgrade wouldincrease the availability of feedwater subsequent to other initiating events.Waterford 3 ResponseThe added benefit that might occur if the upgrade would increase the availability offeedwater subsequent to other initiating events may be seen by looking at analysis Case17, Main Feedwater System Reliability. Case 17 analyzed the benefit of increasing theavailability of the feedwater system for phase II SAMA 33. The following discussion isprovided for Case 17."This analysis case was used to evaluate the change in plant risk from installing amotor-driven feedwater pump. A bounding analysis was performed by setting loss ofmain feedwater to zero in the PSA model. Initiator %T4 was set to zero and gateBT02 was deleted, which resulted in an internal and external benefit (with uncertainty)of approximately $3,572,561. This analysis case was used to model the benefit ofphase II SAMA 33."By removing gate BT02, analysis Case 17 included the benefit of increasing theavailability of feedwater subsequent to other initiating events. Analysis Case 17 resultedin an internal and external benefit (with uncertainty) of $3,572,561. SAMA 31, with a costof $6,100,000, remains not cost-beneficial when compared with the Analysis Case 17benefit.6.b. The assumption for Case 7, "Reduced Frequency of Loss of Auxiliary Component CoolingWater (ACCW)," given in ER Table D.2-2 is the elimination of failure of ACCW. SectionD.2.3 indicates that the model was changed by adding the ability to cross-tie the ACCW.Provide further information on the modeling to clarify this apparent difference.Waterford 3 ResponseCase 7 was performed to evaluate the risk reduction from adding the ability to cross-tie theACCW trains. Instead of modeling the cross-tie of the ACCW trains, a bounding case wasmodeled by eliminating the ACCW system failure, which will provide a conservative benefit toW3F1-2017-0001Page 39 of 74compared to adding the cross-tie capability. This was done by removing the gates listed inthe Case 7 description in Section D.2.3 of the ER.6.c. SAMA 19, "Add redundant DC control power for SW pumps," is evaluated in Case 12,"Increase Availability of ACCW," by eliminating the DC control power gates to the ACCWpumps. While this SAMA is from the generic PWR list in NEI 05-01 and does notnecessarily represent an important failure mode at WF3, discuss the benefit associatedwith eliminating DC control power failures for the component cooling water (CCW) pumps,in addition to the ACCW pumps.Waterford 3 ResponseA sensitivity was performed for Case 12 in which the DC control power to the CCW pumpswas removed via gates D100A (under gate SB4R3), D200A (under gate SA4R3), andD300AW (under gates SAB9R3 and SAB9R3W) in addition to the DC power gates thatwere removed previously in the Case 12 analysis. This resulted in a relatively smallincrease in the total benefit. With this change incorporated, the benefit (internal andexternal events with uncertainty) is $38,944. Adding redundant DC control power to morepumps would also increase the implementation cost. However, with the existing costestimate of $100,000, the conclusion that the SAMA is not cost-beneficial is unchanged.6.d. Provide more details on the WF3 specific cost estimate for SAMA 35, "Provide aredundant train or means of ventilation." It is not clear if the scope of the cost estimate isconsistent with the assumed elimination of failure of emergency diesel generator (EDG)room 3A cooling for Case 23, "Increased availability of [Heating, ventilation and airconditioning] HVAC used to assess the benefit of SAMA 35."Waterford 3 ResponseThe cost estimate is consistent with providing a redundant train of EDG room ventilationfor EDG 3A. The cost estimate assumes that the train would include instruments, anexhaust fan, and exhaust damper controls. A redundant power source is not neededbecause the EDG ventilation system is designed to maintain room temperature wheneverthe EDGs are in operation. Therefore, the existing EDG ventilation system is powered bythe EDG, through safety-related bus and MCC, and the cost estimate assumes the newtrain would also be powered by the EDG.Since a new train of ventilation for a battery room or the main control room would need aredundant source of power, the implementation cost for such a modification would belarger for those rooms. toW3F1-2017-0001Page 40 of 746.e. Clarify that the scope of SAMA 36, Implement procedures for temporary HVAC, isapplicable to rooms other than EDG room 3A. Analysis of this SAMA only assumedelimination of failure of EDG room 3A cooling (Case 23). Based on the benefit results forCase 23, it appears likely that the implementation of temporary HVAC for the other roomslisted in SAMA 36 may also be potentially cost-beneficial.Waterford 3 ResponseThe scope of SAMA 36 is to implement procedures for temporary HVAC for the MCR,EDG rooms, and battery rooms. Analysis Case 23 included three cases, each eliminatingone system, MCR HVAC, EDG Room 3A cooling, or EDG Room 3B cooling. The casewith EDG Room 3A cooling removed provided the greatest benefit; therefore, it was usedto represent the bounding benefit for Case 23. Since SAMA 36 was determined to bepotentially cost-beneficial, it is potentially cost-beneficial to implement procedures fortemporary HVAC for the battery, EDG, and main control rooms.6.f. Case 24, "Debris Coolability and Core Concrete Interaction," was evaluated by eliminatingfailure of debris coolability and core concrete interaction to determine the benefitassociated with the relatively low cost SAMAs; 38, 47, 72 and 73. These low cost SAMAsprovide water to the cavity or otherwise improve core coolability or reduce core concreteinteraction. Case 28, "Increase Cooling and Containment of Molten Core Debris," wasevaluated by eliminating containment core melt propagation and was used to determinethe benefit associated with relatively high cost SAMAs 44, 45, and 46. The benefitassociated with Case 28 is approximately $6,900,000 compared to that for Case 24 of$61, 000. It appears that the SAMAs evaluated by Case 24 would achieve much of thebenefit associated with Case 28. Discuss the reasons for this significant difference and thepotential for SAMAs 38, 47, 72 and 73, or some combination of them, to be cost-beneficial.Waterford 3 ResponsePhase II SAMAs 44, 45, and 46 were conservatively evaluated using Case 28. Case 28removed failure to maintain the cavity at lower pressure via the containment cooling fans,which removes all possibility of basemat failure. Case 28 is a very conservative SAMAcase which did not need to be further refined due to the high cost of phase II SAMAs 44,45, and 46.Phase II SAMAs 38, 47, 72 and 73 were evaluated using Case 24, which is lessconservative than Case 28, but still bounds the achievable benefit of the SAMAs. Case 24removed failure to cool debris and core-concrete interaction, but did not remove failure tomaintain the cavity at lower pressure. This modeling bounds the achievable benefit fromthe SAMAs which would introduce water to the cavity or otherwise cool the external lowervessel head. toW3F1-2017-0001Page 41 of 746.g. Case 43, "Gagging Device to Close a Stuck Open Safety Valve," is evaluated byeliminating failure events for stuck open relief valves and was used to estimate the benefitof SAMA 71, "Manufacture a Gagging Device for a Steam Generator Safety Valve anddevelop a procedure or work order for closing a stuck-open valve."6.g.i.Provide a more detailed description of the failure events listed and theirrelevance to limiting release following a steam generator tube rupture (SGTR)event.Waterford 3 ResponseThe description for Case 43 lists 13 basic events that were set to zero:PRYMS106BT, PRYMS112BT, PRYMS108BT, PRYMS113BT, PRYMS110BT,PRYMS114BT, PRYMS106AT, PRYMS112AT, PRYMS108AT, PRYMS113AT,PRYMS110AT, PRYMS114AT, and OHFMSSGAGR. Basic eventOHFMSSGAGR represents the failure to gag a leaking main steam safety valvefollowing a SGTR. The other 12 basic events represent Steam Generator safetyvalves failing to close (e.g., PRYMS106BT is safety valve MS106B fails toclose). Gagging or closing the safety valves will limit the loss of inventory fromthe steam generator. SAMA 71 has been changed to conservatively use thesame benefit as SAMA 61 and isnow retained as potentially cost-beneficial.6.g.ii.The benefit of SAMA 61, "Direct steam generator flooding after a SGTR," priorto core damage as assessed in Case 33, "Reduce Consequences of SteamGenerator Tube Ruptures," is approximately $100,000 whereas the benefit ofSAMA 71 is only $76. Both of these SAMAs are intended to reduce the releasesresulting from a SGTR. The very large difference between assessed benefit isnot expected. Explain the reasons for this difference or revise the assessmentsas appropriate.Waterford 3 Response SAMA 71 has been changed to use the same benefit as SAMA 61 and isnowretained as potentially cost-beneficial.6.h. Case 41, "Improve Internal Flooding Response Procedures and Training," and Case 42,"Water Tight Doors for the Largest Contributor to Internal Flooding," were evaluated byassuming that the reduction in risk was proportional to the reduction in internal floodingCDF. SAMAs evaluated by these cases were SAMA 67, "Improve internal floodingresponse procedures and training to improve the response to internal flooding events,"and SAMA 68, "Install flood doors to prevent water propagation in the electric boardroom." An examination of the reductions in risk given in ER Table D.2-2 for other casesindicates that this assumption may be non-conservative depending on the failures toW3F1-2017-0001Page 42 of 74resulting from the specific flooding events mitigated. Describe the system failures involvedin the internal flood events mitigated by these SAMAs and select evaluation cases thatwould be more representative for these specific internal flooding SAMAs.Waterford 3 ResponseIt is acknowledged that the assumption that the reduction in risk is proportional to thereduction in internal flooding CDF could be non-conservative. However, even if one wereto multiply the benefit produced by Cases 41 and 42 by three, SAMAs 67 and 68 wouldremain not cost beneficial. Multiplying the benefit by three is considered conservative asthe largest contributor causes loss of power and the PDR or OECR reduction would not bethree times greater than the CDF reduction. In addition, removing all of the internalflooding contribution for the identified scenarios in each analysis case is conservative.6.i. The cost for SAMA 68, "Install flood doors to prevent water propagation in the electricboard room," is given as $4,695,000 and stated to be from the Sequoyah cost estimate.The Sequoyah LRA ER indicates that this is the cost for both Sequoyah units. Further, thecost of such a modification would appear to be strongly dependent on a specific plantlayout. Provide a cost that is valid for the WF3 plant configuration. Also discuss ifsomething less than a full flood door, such as a flood barrier, might achieve the same riskreduction benefit.Waterford 3 ResponseA plant specific WF3 cost estimate was developed to modify doors D16 and D9 to be flooddoors to prevent water propagation to the other electric board rooms. The WF3 plantspecific cost estimate is $1,268,119, which is substantially lower than the Sequoyah costestimate of $4,695,000. A flood barrier is not expected to achieve the same risk benefit asa flood door. SAMA 68 remains not cost beneficial.6.j. The cost for SAMA 8, "Use fire water system as a backup source for diesel cooling," isgiven as $2,000,000 and stated to be from the Seabrook cost estimate. Implementation ofa similar SAMA for the Grand Gulf plant (SAMA 9) was estimated to cost $1,344,000. Thisis very near the assessed benefit at WF3 of $1,338,000. Provide a WF3-specificjustification for the cost estimate for SAMA 8.Waterford 3 ResponseThe implementation cost estimate for the Grand Gulf plant was a conceptual estimateperformed using 2012 dollars. A recent PWR implementation estimate was consideredmore applicable than the Grand Gulf estimate. Escalating the Grand Gulf 2012 estimate tocurrent dollars using a ratio of the consumer price indices would increase the estimate tojust over $1.4 million. SAMA 8 is now retained as potentially cost-beneficial. toW3F1-2017-0001Page 43 of 746.k. In the evaluation of the benefit of SAMA 61, "Direct steam generator flooding after aSGTR," prior to core damage in Case 33, ER Table D.2-2 states that the SGTR CDFcontribution was assigned from the H-E RC to the L-I RC. However, the NRC staff notesthat the population dose for the L-I RC is greater than that for the H-E RC. Justify theapproach used to evaluate the benefit of SAMA 61 in Case 33.Waterford 3 ResponseBased on the updated Level 3 results and approach presented in RAI SAMA 4.b thepopulation dose results for the L-I RC are no longer greater than that for the H-E RC. Theupdated benefit for SAMA 61 in Case 33 is $557,676 and this SAMA is now retained aspotentially cost beneficial. toW3F1-2017-0001Page 44 of 74RAI SAMA 77. For certain SAMAs considered in the WF3 ER, there may be lower cost or more effectivealternatives that could achieve much of the risk reduction. In this regard, provide an evaluationof the following SAMA. The basis for this request is as follows: Applicants for license renewalare required by 10 CFR 51.53(c)(3)(ii)(L) to consider SAMAs if not previously considered in anenvironmental impact assessment, related supplement, or environmental assessment for theplant. As part of its review of the WF3 SAMA analysis, NRC staff considers additional SAMAsthat may be more effective or have lower implementation costs than the other SAMAs evaluatedby the applicant. The requested information is needed in order for the NRC staff to reach aconclusion on the adequacy of the applicant's determination of cost-beneficial SAMAs.7.a. SAMA 27, "Install an additional component cooling water pump," is evaluated as a meansto increase cooling water availability. Consider a potentially lower cost modification ofreplacing one of the pumps with a diverse design that would lower the common causepump failure.Waterford 3 ResponseThe common cause failure of the CCW pumps to fail to run (SCCMDPNRUN) is 8.51E-07/yr and 9.03E-11/yr for fail to start (SCCMDPSTRT) in the PRA model utilized for theSAMA. A sensitivity was performed in which basic events SCCMDPNRUN andSCCMDPSTRT were set to zero. This resulted in a benefit (internal and external eventswith uncertainty) of $7,257. Replacing a CCW pump would cost significantly more thanthis amount; therefore, this lower cost modification would not be cost-beneficial.During evaluation of this response, an error was found in the calculation of the CCW typecodes which affects the CCW independent and Common Cause Failure (CCF) basic eventprobabilities. A Model Change Request and condition report have been initiated to fix theissue during the next model update. To evaluate the impact of the error, a sensitivity wasperformed on the current WF3 internal events model in which the CCW type codes werecorrected and which resulted in an insignificant (less than 1%) change in the modelresults. Thus, the impact to the WF3 SAMA analysis is negligible.7.b. Also, regarding SAMA 27, Table D.1-2 indicates a portion of this benefit is due toeliminating the operator failure to align CCW train AB to replace lost Train A or B. Providean assessment of a potentially lower cost SAMA candidate to provide diverse backupauto-start signals for the standby CCW trains on loss of the running train.Waterford 3 Response:Installation of diverse backup auto-start signals would improve the reliability of the CCWsystem and lower the importance of the operator action to align the standby CCW train. Abounding analysis was performed in which the operator action SHFABCCWRP (Failure to toW3F1-2017-0001Page 45 of 74align CCW train AB to replace lost train A or B) was set to zero, and this led to a benefit of$4,524,670. The plant-specific estimated cost for this modification is $1,091,240, so thisSAMA is retained as potentially cost-beneficial. toW3F1-2017-0001Page 46 of 74Table D.1-10WF3 Release Category Source TermsTime (sec)BasisTPQU_BI00Plume #117280005429432003777165.381.28E+03Plume #248.00432001296008640065.382.07E+03TQX_H49984.47< 0Plume #1222784.4705849984499268.080.00E+00Plume #261.8849984129600796168.083.52E+06SBO80629.9868892.98Plume #1253429.98191173780630688938.080.00E+00Plume #270.4080630129600489708.089.10E+05------------------SX_B39185.4119673.41Plume #1211985.4055.5391854320040158.080.00E+00Plume #258.8843200129600864008.086.46E+04TB_B53997.6748852.11Plume #1226797.6713.6514653998488528.080.00E+00Plume #263.0053998129600756028.083.98E+04TB_F7255269353.03Plume #1245351.61324319972552693538.080.00E+00Plume #268.1572552129600570488.085.52E+06------------------TKC_B73785.9943337.63Plume #1246585.991123044873786433388.080.00E+00Plume #268.5073786129600558148.082.87E+04TQX_B11291587573.07Plume #1285714.78624.3325342112915864008.080.00E+00Plume #279.3772112915129600166858.087.79E+04------------------------------------AX_D123785117007.42Plume #1296585.3853337385123785864008.080.00E+00Plume #282.3812378512960058158.082.40E+06NOTES:1234RDPDELAY - Start ofPlume release -from scram time(sec)MAAP TimingRelease Ends(sec)RDPLUDUR -Duration ofRelease (sec)RDPLHITE - Height ofplume release -centerline of EscapeHatch (26.5 ft)PLHEAT - Energyof ReleaseEREL(6), WReleaseCategoryReleaseFrequency(per year)Level 2 MAAP Run IDRelease Time -Time fromscram (sec)Elasped Time -GE to Release(sec) / (hr)H-I4.75E-06INTACT3.68E-06Loss of Fuel Clad(FCB3)Loss of Fuel Clad(FCB3)2H-E1.88E-06M-I1.34E-075145.56H-LNA1NAM-E2.74E-08L-L5.56E-10M-L1.84E-08L-ENA1NAL-I2.42E-09Loss of Fuel Clad(FCB3)LL-ENA1LL-L3.85E-10These Release Categories were included as part of the Level 2 PRA Model, but were not present in the Level 2 cutset results. As a result, release scenarios were not developed as part ofthe Level 3 analysis.LL-INA1NABasis for OALARM is Loss of Fuel Clad Barrier FCB3 [EP-001-001, Rev. 030] as defined by core exit thermocouple reading >= 1200 deg F. This is conservatively defined by MAAP Event code690.Basis for OALARM is Potential Loss of Containment Barrier CNB2 [EP-001-001, Rev. 030] as defined by containment pressure >= 17.7 psia. This is defined by MAAP parameter PRB(4) as foundin MAAP data file ".d85".Basis for OALARM is Potential Loss of Fuel Clad Barrier FCB3 [EP-001-001, Rev. 030] as defined by core exit thermocouple reading >= 700 deg F. This is defined by MAAP parameter TCRHOTas found in MAAP data file ".d85".RDOALARM - Time to declare GE, sec5429.3355827.9511737.0019512.00Loss of Fuel Clad(FCB3)NAPotential Loss ofFuel Clad (FCB3)4Loss of Fuel Clad(FCB3)Loss of Fuel Clad(FCB3)Potential Loss ofContainmentPressure (CNB2)3Loss of Fuel Clad(FCB3)3198.5830448.3725341.716777.97 toW3F1-2017-0001Page 47 of 74Table D.1-10 (cont'd)NobleGasesICsTeSrRuLaCeBaTPQU_BIPlume #12.79E-035.31E-051.33E-049.36E-051.97E-078.72E-073.07E-097.17E-083.86E-07Plume #21.36E-023.24E-041.92E-041.18E-042.15E-079.28E-073.30E-097.80E-084.81E-07TQX_HPlume #10.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.99980.34760.21270.24029.90E-041.36E-024.43E-051.02E-046.49E-03SBOPlume #10.00E+000.00E+000.00000.00000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.99860.31521.57E-011.77E-011.80E-039.04E-042.99E-051.11E-031.03E-03---------------------------SX_BPlume #10.00001.82E-071.40E-072.92E-081.13E-102.10E-093.03E-121.73E-123.50E-09Plume #20.66610.09347.17E-021.58E-013.33E-043.59E-031.72E-054.19E-052.14E-03TB_BPlume #10.00000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.55526.27E-029.97E-034.84E-031.75E-083.86E-071.03E-092.14E-095.66E-06TB_FPlume #10.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.99967.87E-022.10E-021.16E-022.84E-085.76E-071.87E-094.88E-095.62E-06---------------------------TKC_BPlume #10.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.50396.59E-034.20E-035.44E-031.67E-051.59E-053.98E-071.55E-063.39E-05TQX_BPlume #10.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.51081.07E-021.57E-033.58E-024.54E-078.65E-062.42E-081.61E-074.69E-06------------------------------------------------------AX_DPlume #10.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+000.00E+00Plume #20.61515.80E-046.22E-033.39E-032.22E-065.66E-066.34E-081.71E-071.11E-05NOTES:1LL-L3.85E-10LL-INA1LL-ENA1L-L5.56E-10M-L1.84E-08L-ENA1L-IM-E2.74E-08M-I1.34E-07H-LNA1H-E1.88E-06H-I4.75E-06INTACT3.68E-06ReleaseCategoryThese Release Categories were included as part of the Level 2 PRA Model, but were not present in the Level 2 cutset results. As aresult, release scenarios were not developed as part of the Level 3 analysis.ReleaseFrequency(per year)Level 2 MAAP Run IDRDRELFRC001 - Release FractionsNANA2.42E-09NANA toW3F1-2017-0001Page 48 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion1. Case SBO ReductionEliminated SBOcontribution.34.4%44.0%46.8%$3,791,951$7,811,4191. Provide additional DCbattery capacity.WF3 plant specific cost$3,172,695Retain2. Replace lead-acidbatteries with fuel cells. WF3 plant specific cost$6,185,319Retain7. Install a gas turbinegenerator.Davis-Besse costestimate$2,000,000Retain2. Case ImproveFeedwater ReliabilityEliminated failure offeedwater.0.9%0.2%0.2%$21,596$44,48831. Install a digital feedwater upgrade.SeaBrook Cost estimate$6,100,000Not costeffective toW3F1-2017-0001Page 49 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion3. Case Add DC SystemCross-tiesChanged gates torequire multiple DC busfailures.20.8%31.1%29.5%$2,430,355$5,006,5323. Provide DC bus cross-ties.WF3 plant specific cost$1,449,686Retain4. Case IncreaseAvailability of On-Site ACPowerChanged gates torequire multiple AC busfailures.22.2%32.1%30.5%$2,515,301$5,181,5205. Improve 4.16-kV buscross-tie ability.WF3 plant specific cost$1,554,988RetainCase 5. Reduce Loss ofOff-Site PowerReduce the frequency ofthe LOOP initiator byremoving severeweather contributionaffecting OSP lines.11.3%11.8%12.4%$1,022,379$2,106,10010. Bury off-site powerlines.SeaBrook Cost estimate$3,000,000Not costeffective6. Install an additional,buried off-site powersource.SeaBrook Cost estimate$3,000,000Not costeffective toW3F1-2017-0001Page 50 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)ConclusionCase 6. Provide BackupEDG CoolingEliminated failure ofCCW cooling to theEDGs.4.4%10.0%10.8%$847,186$1,745,2038. Use fire water systemas a backup source fordiesel cooling.Grand Gulf Costestimate$1,400,000Retain9. Add a new backupsource of diesel cooling. SeaBrook Cost estimate$2,000,000Not costeffectiveCase 7. ReducedFrequency of Loss ofAuxiliary ComponentCooling Water3Eliminated failure ofACCW.6.4%0.9%0.4%$92,762$191,08921. Enhance proceduralguidance for use of cross-tied component coolingor service water pumps.WF3 is adding the abilityto cross-tie ACCW andnot procedure change.WF3 plant specific cost.$6,528,828Not costeffective22. Add a service waterpump.Sequoyah cost estimate$1,043,000Not costeffectiveCase 8. Increasedavailability of feedwaterEliminated DWST failureto supply the CSP.1.2%0.3%0.2%$27,688$57,038 toW3F1-2017-0001Page 51 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion32. Create ability foremergency connection ofexisting or new watersources to feedwater andcondensate systems.WF3 plant specific cost$885,760Not costeffectiveCase 9. High PressureInjection SystemEliminated failure ofHPSI.8.4%3.1%2.5%$267,690$551,44117. Replace two of thefour electric safetyinjection pumps withdiesel-powered pumps.Callaway cost estimate$1,500,000Not costeffective13. Install anindependent active orpassive high pressureinjection system.Callaway cost estimate$1,500,000Not costeffectiveCase 10. Extend ReactorWater Storage PoolCapacityReduced failure fromoperator actions andtank rupture.1.8%0.2%0.1%$22,100$45,52716. Throttle low pressureinjection pumps earlier inmedium or large-breakLOCAs to maintainreactor water storagetank inventory.SeaBrook Cost estimate$3,000,000Not costeffective toW3F1-2017-0001Page 52 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion29. RWST fillfromfirewater duringcontainment injection-Modify 6 inch RWST flushflange to have a 21/2-inchfemale fire hose adapterwith isolation valve.WF3 plant specific cost$747,640Not costeffective30. High-volume makeupto the refueling waterstorage tank.Sequoyah cost estimate$565,000Not costeffective49. Install automaticcontainment spray pumpheader throttle valves.ANO-2 cost estimate$2,500,000Not costeffectiveCase 11. Eliminate ECCSDependency onComponent CoolingWater SystemEliminated failure ofECCS motor cooling.0.7%2.9%3.2%$246,497$507,78520. Replace ECCS pumpmotors with air-cooledmotors.SeaBrook Cost estimate$6,000,000Not costeffectiveCase 12. IncreaseAvailability of ACCWEliminated the DCcontrol power gates tothe ACCW pumps.0.2%0.1%0.1%$12,454$25,655 toW3F1-2017-0001Page 53 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion19. Add redundant DCcontrol power for SWpumps.Callaway cost estimate$100,000Not costeffectiveCase 13. Low PressureSafety Injection SystemEliminated failure of theLow Pressure SafetyInjection system.0.0%0.0%0.0%$8$1614. Add a diverse lowpressure injectionsystem.Callaway cost estimate$1,000,000Not costeffective15. Provide capability foralternate injection viadiesel-driven fire pump.Davis-Besse costestimate$6,500,000Not costeffectiveCase 14. IncreaseComponent CoolingWater AvailabilityEliminated failure ofCCW pump failures andCCFs.13.5%28.5%27.5%$2,210,419$4,553,46427. Install an additionalcomponent cooling waterpump.SeaBrook Cost estimate$6,000,000Not costeffectiveCase 15. DecreasedCharging Pump FailureEliminated the normalcharging pump powergates.0.4%0.8%0.7%$59,385$122,332 toW3F1-2017-0001Page 54 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion12. Install modification topower the normalcharging pump from anexisting spare breakerfrom the alternateemergency powersystem.Callaway cost estimate$350,000Not costeffectiveCase 16. Reactor CoolantPump SealsEliminated RCP SealLOCA.16.0%31.9%30.7%$2,475,719$5,099,98124. Install anindependent reactorcoolant pump sealinjection system, withdedicated diesel.Sequoyah Cost estimate$8,233,000Not costeffective25. Install anindependent reactorcoolant pump sealinjection system, withoutdedicated diesel.Sequoyah Cost estimate$8,233,000Not costeffective26. Install improvedreactor coolant pumpseals.SeaBrook Cost estimate$2,000,000RetainCase 17. Main FeedwaterSystem ReliabilitySet loss of mainfeedwater to zero.33.3%19.4%19.2%$1,734,253$3,572,561 toW3F1-2017-0001Page 55 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion33. Add a motor-drivenfeedwater pump.Sequoyah cost estimate$10,000,000Not costeffectiveCase 18. EDG Fuel OilSet the failure of fuel oilpumps to zero.17.1%21.1%22.3%$1,815,628$3,740,19411. Install a large volumeEDG fuel oil tank at anelevation greater thanthe EDG fuel oil daytanks.WF3 plant specific cost$26,300,000Not costeffectiveCase 20. Create a reactorcoolant depressurizationsystemEliminated small LOCAevents.14.5%2.1%0.9%$203,552$419,31818. Create a reactorcoolant depressurizationsystem.SeaBrook cost estimate$1,000,000Not costeffectiveCase 21. SteamGenerator InventoryReduced the frequencyof turbine-driven AFWpump failure duringSBO.67.3%63.2%65.8%$5,511,067$11,352,79834. Use fire water systemas a backup for steamgenerator inventory.Cost from Indian point(IP2)$3,073,130Retain toW3F1-2017-0001Page 56 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)ConclusionCase 22. Instrument AirReliabilityEliminated the loss ofInstrument Air.0.1%0.0%0.0%$2,783$5,73437. Replace service andinstrument aircompressors with morereliable compressorswhich have self-contained air cooling byshaft driven fans.Callaway cost estimate$500,000Not costeffectiveCase 23. IncreasedAvailability of HVACEliminated failure ofEDG room 3A cooling.9.4%12.0%12.7%$1,030,922$2,123,69935. Provide a redundanttrain or means ofventilation.WF3 plant specific cost$3,574,481Not costeffective36. Implementprocedures fortemporary HVAC.Callaway cost estimate$100,000RetainCase 24. Debriscoolability and coreconcrete interactionEliminated failure ofdebris coolability andcore concreteinteraction.0.0%0.5%0.6%$42,490$87,530 toW3F1-2017-0001Page 57 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion38. Create a reactorcavity flooding system.Cost from Indian Point(IP2)$1,741,724Not costeffective47. Provide a reactorvessel exterior coolingsystem.Cost from ANO-2$2,500,000Not costeffective72. Provide water fromthe fire protectionsystem to thecontainment sump.WF3 plant specific cost$715,918Not costeffective73. Enhancecommunication betweensump and cavity.WF3 plant specific cost$702,551Not costeffectiveCase 25. Decay HeatRemoval CapabilityEliminated latecontainment failure dueto over-pressurization.0.0%20.3%22.7%$1,690,200$3,481,81241. Install an unfiltered,hardened containmentvent.WF3 plant specific cost$15,083,162Not costeffective42. Install a filteredcontainment vent toremove decay heatOption 1: Gravel BedFilterSeaBrook cost estimate$20,000,000Not costeffective toW3F1-2017-0001Page 58 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)ConclusionOption 2: MultipleVenturi ScrubberCase 26. ImproveContainment SprayCapabilityReduced failure ofcontainment spray.5.8%44.6%52.3%$3,908,776$8,052,07839. Install a passivecontainment spraysystem.SeaBrook cost estimate$10,000,000Not costeffective50. Install a redundantcontainment spraysystem.SeaBrook cost estimate$10,000,000Not costeffective40. Use the fire watersystem as a backupsource for thecontainment spraysystem.WF3 plant specific cost$2,455,808RetainCase 27. ReduceHydrogen IgnitionEliminated hydrogendetonation.0.0%0.1%0.0%$3,434$7,07543. Provide post-accidentcontainment inertingcapability.Callaway cost estimate$100,000Not costeffective toW3F1-2017-0001Page 59 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion51. Install anindependent powersupply to the hydrogencontrol system usingeither new batteries, anon-safety gradeportable generator,existing station batteries,or existing AC/DCindependent powersupplies, such as thesecurity system diesel.Callaway cost estimate$100,000Not costeffective52. Install a passivehydrogen control system. SeaBrook cost estimate$100,000Not costeffectiveCase 30. ReduceProbability ofContainment FailureEliminated containmentfailure.0.0%85.8%93.0%$6,947,477$14,311,80348. Construct a buildingto be connected toprimary/secondarycontainment andmaintained at a vacuum. SeaBrook cost estimate$56,700,000Not costeffectiveCase 31. ContainmentIsolationEliminated containmentisolation failure.0.0%0.1%0.1%$9,153$18,855 toW3F1-2017-0001Page 60 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion55. Add redundant anddiverse limit switches toeach containmentisolation valve.Sequoyah cost estimate$692,000Not costeffectiveCase 32. ReduceFrequency of SteamGenerator TubeRupturesEliminated steamgenerator tuberuptures.1.0%5.7%5.5%$430,225$886,26356. Institute amaintenance practice toperform a 100%inspection of steamgenerator tubes duringeach refueling outage.Callaway cost estimate$3,000,000Not costeffective57. Increase the pressurecapacity of the secondaryside so that a steamgenerator tube rupturewould not cause therelief valves to lift.Callaway cost estimate$10,000,000Not costeffective58. Install a redundantspray system todepressurize the primarysystem during a steamgenerator tube ruptureCallaway cost estimate$10,000,000Not costeffective59. Route the dischargefrom the main steamsafety valves through astructure where a waterspray would condenseCallaway cost estimate$10,000,000Not costeffective toW3F1-2017-0001Page 61 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusionthe steam and removemost of the fissionproducts.60. Install a highlyreliable (closed loop)steam generator shell-side heat removal systemthat relies on naturalcirculation and storedwater sourcesCallaway cost estimate$10,000,000Not costeffectiveCase 33. ReduceConsequences of SteamGenerator TubeRuptures5Reassigned the SGTRCDF contribution fromH-E release category torelease category L-I.0.0%3.3%3.6%$270,716$557,67661. Direct steamgenerator flooding after asteam generator tuberupture, prior to coredamage.Generic cost estimatefor procedural changewith engineering andtesting/trainingrequired.$200,000Retain71. Manufacture agagging device for asteam generator safetyvalve and developing aprocedure or work orderfor closing a stuck-openvalve.Cost from Indian Point(IP2)$453,745Retain toW3F1-2017-0001Page 62 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)ConclusionCase 34. Reduce ATWSFrequencyEliminated ATWScontribution.1.4%0.2%0.1%$21,060$43,38363. Add an independentboron injection system.SeaBrook cost estimate$500,000Not costeffective64. Add a system of reliefvalves to preventequipment damage frompressure spikes during anATWS.SeaBrook cost estimate$500,000Not costeffective65. Install motorgenerator set tripbreakers in control room. Sequoyah cost estimate$100,000Not costeffective66. Provide capability toremove power from thebus powering the controlrods.Sequoyah cost estimate$100,000Not costeffectiveCase 37. ReduceProbability of a LargeLOCAEliminated the initiatorsfor a Large LOCA and amedium LOCA.0.4%0.2%0.2%$17,568$36,19069. Install digital largebreak LOCA protectionsystem.SeaBrook cost estimate$500,000Not costeffective toW3F1-2017-0001Page 63 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)ConclusionCase 38. PreventSecondary SideDepressurizationEliminated the initiatorfor a steam line breakoutside containmentand for inadvertentclosure of MSIVs.0.3%0.1%0.0%$6,385$13,15370. Install secondary sideguard pipes up to themain steam isolationvalves.SeaBrook cost estimate$500,000Not costeffectiveCase 39. EliminateThermally Induced TubeRuptures Following CoreDamageEliminated thermalinduced steamgenerator tube rupture.0.0%0.2%0.2%$18,357$37,81654. Modify proceduressuch that the water loopseals in the reactorcooling system (RCS) coldlegs are not clearedfollowing core damage.South Texas costestimate$100,000Not costeffectiveCase 40. ReplaceCARMVAAA201-B with afail closed AOVEliminated motivepower dependency fromMOV CARMVAAA201-B.0.0%0.0%0.0%$0$062. Hardware change toeliminate MOV CS-V-17AC power dependency.SeaBrook cost estimate$300,000Not costeffective toW3F1-2017-0001Page 64 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)ConclusionCase 41. ImproveInternal FloodingResponse Proceduresand Training1Eliminated thecontribution to internalflooding CDF fromfloods in the TurbineGenerator Building +15elevation and ReactorAuxiliary Building +46elevation.N/AN/AN/A$15,744$32,43367. Improve internalflooding responseprocedures and trainingto improve the responseto internal floodingevents.Sequoyah cost estimate$400,000Not costeffectiveCase 42. Water tightdoors for the largestcontributor to internalflooding1Eliminated thecontribution to internalflooding CDF fromfloods in flood zoneRAB21-212/225B.N/AN/AN/A$161,278$332,23368. Install flood doors toprevent waterpropagation in theelectric board room.WF3 plant specific cost$1,268,119Not costeffectiveCase 44. CCW BackupAuto-Start SignalOperator actionSHFABCCWRP (Failure toalign CCW train AB toreplace lost train A or B)was set to zero.13.5%28.3%27.4%$2,196,442$4,524,670 toW3F1-2017-0001Page 65 of 74TableD.2-2Summary of Phase II SAMA Candidates Considered in Cost-Benefit EvaluationAnalysis Case (bold)SAMA Number andTitle2,4AssumptionsCDFReduction(%)PDRReduction(%)OECRReduction(%)Internal andExternalBenefit ($)Internal andExternalBenefit withUncert ($)WF3 CostEstimate ($)Conclusion77. Provide a diversebackup auto-start signalfor the standby CCWtrains on loss of therunning train.WF3 plant specific cost$1,091,240Retain(1)These analysis cases only impact internal flooding and have been evaluated as described in Section D.2.3.(2)The WF3 NFPA 805 SAMAs (Phase II SAMAs 74, 75, and 76) have been updated based on the latest NFPA 805 RAIs. Based on the latest NFPA 805 RAIs all of the modifications thatimpact the fire PRA have been installed.(3)Phase II SAMA 23 (analysis case 7) to proceduralize shedding component cooling water loads to extend the component cooling water heat-up time upon loss of essential raw coolingwater has been changed from retain to N/A. This has been changed due to the fact that if an SIAS is not generated before RCP seal failure, CCW can operate with only the dry coolingtowers (without ACCW), so shedding CCW loads on loss of raw cooling water (ACCW) is not necessary. In those limiting cases when an SIAS is generated, the CCW trains automaticallysplit and the non-safety CCW header is isolated. Since the SIAS automatically sheds non-essential CCW loads, a procedure is not necessary.(4)Phase II SAMAs 44, 45, 46, and 53 (analysis cases 28 and 29) have been changed from retain to N/A as these SAMAs are for a new plant and it's not practical to back fit thesemodifications into a plant which is already built, and operating. The cost of implementation of these SAMAs would likely exceed the maximum benefit.(5)Based on RAI response 6.g.ii Phase II SAMA 71 was moved to utilize the benefit of analysis case 33 rather than analysis case 43. toW3F1-2017-0001Page 66 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)Case 1. SBO Reduction$7,811,419$9,738,7371. Provide additional DC battery capacity.$3,172,6952. Replace lead-acid batteries with fuel cells.$6,185,3197. Install a gas turbine generator.$2,000,000Case 2. Improve Feedwater Reliability$44,488$54,79631. Install a digital feed water upgrade.$6,100,000Case 3. Add DC System Cross-ties$5,006,532$6,364,7463. Provide DC bus cross-ties.$1,449,686Case 4. Increase Availability of On-Site AC Power$5,181,520$6,584,2845. Improve 4.16-kV bus cross-tie ability.$1,554,988Case 5. Reduce Loss of Off-Site Power$2,106,100$2,624,22510. Bury off-site power lines.$3,000,0006. Install an additional, buried off-site power source.$3,000,000Case 6. Provide Backup EDG Cooling$1,745,203$2,183,1468. Use fire water system as a backup source for dieselcooling.$1,400,0009. Add a new backup source of diesel cooling.$2,000,000Case 7. Reduced Frequency of Loss of AuxiliaryComponent Cooling Water$191,089$232,91521. Enhance procedural guidance for use of cross-tiedcomponent cooling or service water pumps.$6,528,82822. Add a service water pump.$1,043,000Case 8. Increased availability of feedwater$57,038$70,17632. Create ability for emergency connection of existingor new water sources to feedwater and condensatesystems.$885,760Case 9. High Pressure Injection System$551,441$688,98317. Replace two of the four electric safety injectionpumps with diesel-powered pumps.$1,500,00013. Install an independent active or passive highpressure injection system.$1,500,000Case 10. Extend Reactor Water Storage Pool Capacity$45,527$54,845 toW3F1-2017-0001Page 67 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)16. Throttle low pressure injection pumps earlier inmedium or large-break LOCAs to maintain reactorwater storage tank inventory.$3,000,00029. RWST fill from firewater during containmentinjection-Modify 6 inch RWST flush flange to have a21/2-inch female fire hose adapter with isolation valve.$747,64030. High-volume makeup to the refueling waterstorage tank.$565,00049. Install automatic containment spray pump headerthrottle valves.$2,500,000Case 11. Eliminate ECCS Dependency on ComponentCooling Water System$507,785$634,26820. Replace ECCS pump motors with air-cooledmotors.$6,000,000Case 12. Increase Availability of ACCW$25,655$31,86719. Add redundant DC control power for SW pumps.$100,000Case 13. Low Pressure Safety Injection System$16$2214. Add a diverse low pressure injection system.$1,000,00015. Provide capability for alternate injection via diesel-driven fire pump.$6,500,000 toW3F1-2017-0001Page 68 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)Case 14. Increase Component Cooling WaterAvailability$4,553,464$5,795,36727. Install an additional component cooling waterpump.$6,000,000Case 15. Decreased Charging Pump Failure$122,332$156,13612. Install modification to power the normal chargingpump from an existing spare breaker from thealternate emergency power system.$350,000Case 16. Reactor Coolant Pump Seals$5,099,981$6,492,65624. Install an independent reactor coolant pump sealinjection system, with dedicated diesel.$8,233,00025. Install an independent reactor coolant pump sealinjection system, without dedicated diesel.$8,233,00026. Install improved reactor coolant pump seals.$2,000,000Case 17. Main Feedwater System Reliability$3,572,561$4,429,75233. Add a motor-driven feedwater pump.$10,000,000Case 18. EDG Fuel Oil$3,740,194$4,663,364 toW3F1-2017-0001Page 69 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)11. Install a large volume EDG fuel oil tank at anelevation greater than the EDG fuel oil day tanks.$26,300,000Case 20. Create a reactor coolant depressurizationsystem$419,318$516,27718. Create a reactor coolant depressurization system.$1,000,000Case 21. Steam Generator Inventory$11,352,798$14,129,89734. Use fire water system as a backup for steamgenerator inventory.$3,073,130Case 22. Instrument Air Reliability$5,734$7,04337. Replace service and instrument air compressorswith more reliable compressors which have self-contained air cooling by shaft driven fans.$500,000Case 23. Increased Availability of HVAC$2,123,699$2,649,04235. Provide a redundant train or means of ventilation.$3,574,48136. Implement procedures for temporary HVAC.$100,000Case 24. Debris coolability and core concreteinteraction$87,530$109,415 toW3F1-2017-0001Page 70 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)38. Create a reactor cavity flooding system.$1,741,72447. Provide a reactor vessel exterior cooling system.$2,500,00072. Provide water from the fire protection system tothe containment sump.$715,91873. Enhance communication between sump andcavity.$702,551Case 25. Decay Heat Removal Capability$3,481,812$4,364,66341. Install an unfiltered, hardened containment vent.$15,083,16242. Install a filtered containment vent to removedecay heatOption 1: Gravel Bed FilterOption 2: Multiple Venturi Scrubber$20,000,000Case 26. Improve Containment Spray Capability$8,052,078$9,991,50539. Install a passive containment spray system.$10,000,00050. Install a redundant containment spray system.$10,000,00040. Use the fire water system as a backup source forthe containment spray system.$2,455,808 toW3F1-2017-0001Page 71 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)Case 27. Reduce Hydrogen Ignition$7,075$9,95743. Provide post-accident containment inertingcapability.$100,00051. Install an independent power supply to thehydrogen control system using either new batteries, anon-safety grade portable generator, existing stationbatteries, or existing AC/DC independent powersupplies, such as the security system diesel.$100,00052. Install a passive hydrogen control system.$100,000Case 30. Reduce Probability of Containment Failure$14,311,803$18,038,74548. Construct a building to be connected toprimary/secondary containment and maintained at avacuum.$56,700,000Case 31. Containment Isolation$18,855$24,18055. Add redundant and diverse limit switches to eachcontainment isolation valve.$692,000Case 32. Reduce Frequency of Steam Generator TubeRuptures$886,263$1,134,57556. Institute a maintenance practice to perform a100% inspection of steam generator tubes during eachrefueling outage.$3,000,00057. Increase the pressure capacity of the secondaryside so that a steam generator tube rupture would notcause the relief valves to lift.$10,000,000 toW3F1-2017-0001Page 72 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)58. Install a redundant spray system to depressurizethe primary system during a steam generator tuberupture$10,000,00059. Route the discharge from the main steam safetyvalves through a structure where a water spray wouldcondense the steam and remove most of the fissionproducts.$10,000,00060. Install a highly reliable (closed loop) steamgenerator shell-side heat removal system that relieson natural circulation and stored water sources$10,000,000Case 33. Reduce Consequences of Steam GeneratorTube Ruptures$557,676$702,76961. Direct steam generator flooding after a steamgenerator tube rupture, prior to core damage.$200,00071. Manufacture a gagging device for a steamgenerator safety valve and developing a procedure orwork order for closing a stuck-open valve.$453,745Case 34. Reduce ATWS Frequency$43,383$53,20163. Add an independent boron injection system.$500,00064. Add a system of relief valves to prevent equipmentdamage from pressure spikes during an ATWS.$500,00065. Install motor generator set trip breakers in controlroom.$100,00066. Provide capability to remove power from the buspowering the control rods.$100,000 toW3F1-2017-0001Page 73 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)Case 37. Reduce Probability of a LOCA$36,190$45,00969. Install digital large break LOCA protection system.$500,000Case 38. Prevent Secondary Side Depressurization$13,153$16,07070. Install secondary side guard pipes up to the mainsteam isolation valves.$500,000Case 39. Eliminate Thermally Induced Tube RupturesFollowing Core Damage$37,816$48,49554. Modify procedures such that the water loop sealsin the reactor cooling system (RCS) cold legs are notcleared following core damage.$100,000Case 40. Replace CARMVAAA201-B with a fail closedAOV$0$062. Hardware change to eliminate MOV CS-V-17 ACpower dependency.$300,000Case 41. Improve Internal Flooding ResponseProcedures and Training1N/AN/A67. Improve internal flooding response proceduresand training to improve the response to internalflooding events.$400,000Case 42. Water tight doors for the largest contributorto internal flooding1N/AN/A68. Install flood doors to prevent water propagation inthe electric board room.$1,268,119 toW3F1-2017-0001Page 74 of 74Table D.2-4 Sensitivity Analysis ResultsAnalysis Case (bold)SAMA Number and TitleBaseline Internaland ExternalBenefit withUncert. ($)Sensitivity with$5,200 per person-rem ($)WF3 Cost Estimate($)Case 44. CCW Backup Auto-Start Signal$4,524,670$5,759,24477. Provide a diverse backup auto-start signal for thestandby CCW trains on loss of the running train.$1,091,240(1)These analysis cases only impact internal flooding and have been evaluated as described in Section D.2.3.