ML18158A514

Grand Gulf, Unit 1, Supplement to Application for Technical Specification Change Regarding Risk-Informed Justification for the Relocation of Specific Surveillance Frequency Requirements to a Licensee Controlled Program (TSTF-425)

ML18158A514
Person / Time
Site:	Grand Gulf
Issue date:	06/07/2018
From:	Larson E A Entergy Operations
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
EPID L-2018-LLA-0106, GNRO-2018/00029
Download: ML18158A514 (39)

Text

• e:=u Enterpv . b.J Entergy Operations, Inc. P.O. Box756 GNR0-2018/00029 June 7, 2018 Port Gibson, Mississippi 39150 Tel 601-437-7500 Eric A. Larson Site Vice President Grand Gulf Nuclear Station 10 CFR 50.90 ATTN: Document Control Desk U. S. Nuclear Regulatory Commission Washington, D. C. 20555

REFERENCES:

.

Dear Sir or Madam:

Supplement to Application for Technical Specification Change Regarding Risk-Informed Justification for the Relocation of Specific Surveillance Frequency Requirements to a Licensee Controlled Program (TSTF-425)

Grand Gulf Nuclear Station, Unit 1 Docket No. 50-416 License No. NPF-29 1. Letter from E.A. Larson (Entergy) to Document Control Desk (NRC) dated April 12, 2018, "Application for Technical Specification Change Regarding Risk..;lnformed Justification for the Relocation of-Specific Surveillance Frequency Requirements to a Licensee Controlled Program (TSTF-425)" (NRC Agencywide Documents Access and Management System (ADAMS) Accession No. ML 181028445)

2. Letter from L. M. Regner (NRC) to Vice President, Operations (Entergy) dated May 23, 2018, "Grand Gulf Nuclear Station, Unit 1 -Supplemental Information Needed for Acceptance of Requested Ucensing Action HE: Adoption ofTechnical Specifications Task-Force (TSTF) Traveler TSTF-425, Revision 3 (EPI D L-2018-LLA-0106)" (NRC ADAMS Accession No. ML 18138A468)

By letter dated April 12, 2018 (Reference 1), Entergy Operations, Inc. (Entergy) requested a *license amendment to modify the Grand Gulf Nuclear *station, *Unit 1 technical specifications

• by relocating specific surveillance frequencies to a licensee-controlled program with implementation of Nuclear Energy1 Institute 04-10, "Risk-Informed Technical Specification Initiative 5b, Risk-Informed Method for Control of Surveillance Frequencies." By letter dated May 23, 2018, (Reference 2), the NRC provided a formal request for supplemental information to support acceptance of the TSTF-425 license amendment request. This letter provides the requested supplemental information that includes the finding-level facts and observations (F&Os) from the 2015 full-scope industry probabilistic risk assessment peer review with the associated resolutions and conclusions.

The supplement includes a description GNR0-2018100029 Page 2 of 3 of whether each F&O is a maintenance update or an upgrade and how it meets the capability Category II of the American Society of Mechanical Engineers probabilistic risk assessment standard.

The information provided herein does not change the intent or the justification for the requested license amendment (Reference

1) and does not alter the conclusion that the proposed license amendment does not involve a significant hazards consideration as published in Federal Register 74 FR 32000 dated July 6, 2009. No new regulatory commitments are made in this submittal.

In accordance with 10 CFR 50.91, "Notice for public comment; State consultation," a copy of this application, with attachments, is being provided to designated State of Mississippi Official.

If you should have any questions regarding this submittal, please contact Douglas A. Neve, Manager, Regulatory Assurance, at 601.437.2103.

I declare under penalty of perjury that the foregoing is true and correct. Executed on th day of June 2018. Sincerely, Eric A. Larson EAURN/gwe

Attachment:

Supplemental Information Supporting License Amendment Request to* Relocate Specific Surveillance Frequency Requirements cc: with Attachment Ms. Lisa Regner U.S. Nuclear Regulatory Commission Mail Stop OWFN 8 81 Rockville, MD 20852-2738 Mr. Siva Lingam U.S. Nuclear Regulatory Commission Mail Stop OWFN 8 81 Rockville, MD 20852-2738 Dr. Mary Currier, M.D., M.P.H State Health Officer Mississippi Department of Health P.O. 8ox-1700 Jackson, MS 39215-1700 Email: mary.currier@msdh.ms.gov GNR0-2018/00029 Page 3 of 3 cc: without Attachment Mr. Kriss Kennedy Regional Administrator, Region IV U.S. Nuclear Regulatory Commission 1600 East Lamar Boulevard Arlington, TX 76011-4511 NRC Resident Inspector Grand Gulf Nuclear Station Port Gibson, MS 39150

\ ATTACHMENT to -G-NR0..;20t8/00029 Supplemental Information Supporting License Amendment Request to Relocate Specific Surveillance Frequency Requirements Attachment GNRO 2018/00029 Page 1 of 2 1. Summary Description By letter dated April 12, 2018 (Reference 1), Entergy Operations, Inc. (Entergy) requested a license amendment to modify the Grand Gulf Nuclear Station, Unit.1 (GGNS) technical specifications by relocating specific surveillance frequencies to a licensee-controlled program with the implementation of Nuclear Energy Institute 04-10, "Risk-Informed . Technical Specification Initiative Sb, Risk-Informed Method for Control of Surveillance Frequencies.".

By letter dated May 23, 2018, (Reference 2), the NRC provided a formal request for supplemental information to support acceptance of.the Technical Specifications Task Force (TSTF) traveler TSTF-425 license amendment request (LAR). 2. NRC Staff Request for Supplemental Information In the letter dated May 23, 2018 (Reference 2), the NRC staff states: In the letter dated May 3, 2017, the NRC staff states, in part, that "in order for the NRC to consider the F&Os closed so that they need not be provided in submissions of future risk-informed licensing applications, the licensee should adhere to the guidance in Appendix X in its entirety.

Following the guidance in Appendix X will reinforce the NRC staffs confidence in the F&O closure process and potentially obviate the need for a more in-depth review." The NRC observed GGNS's process for review and closure of F&Os on August 23-25, 2017, at Jackson, MS. The observations were limited to the onsite review; however, as stated in the Appendix X guidance, the onsite review and associated "consensus process, as described in the body of this document, should be followed during which the full team present on the day of the associated consensus session considers and reaches consensus on adequacy of closure of each finding." The guidance in Appendix X permits remote reviews (i.e., via web and teleconference connection to the onsite team), but only for a limited number of findings.

The guidance in Appendix X also permits evaluating and crediting post onsite closure work by the licensee but then requires a re-review of the licensee resolution and associated documentation and separate consensus session. As detailed in the observation report (ADAMS Accession No. ML 17356A055), the NRC observers could not conclude that the licensee fully adhered to the endorsed guidance in conducting the F&O closure audit. Therefore, as specified in the letter dated May 3, 2017, in order for the NRC to review *the technical adequacy of GGNS PRA with regard to Risk..:lnformed Technical Specification Initiative Sb, the LAR must be supplemented by the F&Os and associated resolutions and conclusion

[emphasis added], or the LAR must be supplemented by the following information demonstrating that the licensee adhered to the guidance in Appendix X in its entirety ... Entergy Supplemental Information:

Entergy herein provides the requested listing of the 39 finding-level facts and observations (F&Os) from the 2015 full-scope industry probabilistic risk assessment peer review that were closed by an independent assessment conducted August 23-31, 2017. The listing includes the resolutions and conclusions of the F&Os. In addition, the listing documents the written Attachment GNRO 2018/00029 Page 2 of 2 basis for each F&O to validate whether the F&O constituted a PRA upgrade, maintenance update, or other; and documents the results from the independent assessment team review of the supporting requirements to ensure that Capability Category II of the American Society ofMechariical-Engineers probabilistic risk assessment standard was met for the F&Os. 3.

References:

1. Letter from E.A. Larson (Entergy) to Document Control Desk (NRC) dated April 12, 2018, "Application for Technical Specification Change Regarding Risk-Informed Justification for the Relocation of Specific Surveillance Erequency Requirements to a 'Licensee Controlled Program {TSTF-425)" {NRC Agencywide Documents Access and Management System (ADAMS) Accession No. ML 181028445)

2. Letter from L. M. Regner (NRC) to Vice President, Operations (Entergy) dated May 23, 2018, "Grand Gulf Nuclear Station, Unit 1 -Supplemental Information Needed for Acceptance of Requested Licensing Action RE: Adoption of Technical Specifications Task Force (TSTF) Traveler TSTF-425, Revision 3 (EPID L-2018-LLA-0106)" (NRC ADAMS Accession No. ML 18138A468)

List of Finding Facts and Observations (F&Os) on the GGNS Internal Events PRA Model * (32 Pages)

List of ,Finding Facts and Observations

{F&Os) on the GGNS Internal Events PRA Model 1-3 QU-04 This was not addressed A comparison to similar (all Resolved, QU-04: Cat Maintenance This finding was as a comparison of Mark 111/BWR-6) plants was Closed 2-3 MET update. Added closed and therefore results to similar plants added to the Summary Report per documentation does has no impact on was not conducted.

associated with the Rev 4 independent not impact the surveillance test Possible Resolution PRA model. Causes for assessment methodologies used interval (STI) change Provide a comparison significant differences in of finding or change the PRA evaluations performed results between the plants closure. scope or capability.

in accordance with similar to the comparison were identified.

It is a comparison of the surveillance provided for some of the results from the frequency control other technical elements.

analysis.

program (SFCP). 1-4 QU-F2 The documentation does A detailed discussion of the Resolved, QU-F2: Cat Maintenance This finding was QU-F3 not describe significant significant accident Closed 1-3 MET update. Added closed and therefore accident sequences in sequences for both CDF and per documentation does has no impact on STI sufficient detail. LERF was added to the Model BWROG not impact the change evaluations Possible Resolution Integration and Quantification peer review methodologies or performed in Describe the accident Report for the internal events QU-F3: Cat change the PRA accordance with the sequences in detail model 2-3 MET scope or capability.

SFCP. Similarly, a detailed per It is a discussion of discussion of the significant independent results from the accident sequences for both assessment analysis.

CDF and LERF was added to of finding the Internal Flood Report for closure. the flood scenarios.

1-6 QU-F2 The documentation does A detailed discussion of the Resolved, QU-F2: Cat Maintenance This finding was not describe significant significant accident Closed 1-3 MET update. Added closed and therefore accident sequences in sequences for both CDF and per documentation does has no impact on STI sufficient detail. A LERF was added to the Model BWROG not impact the change evaluations sensitivity study on Integration and Quantification peer review methodologies used performed in LOOP recovery may be Report for the internal events or change the PRA accordance with the appropriate as the base model scope or capability.

SFCP. case. Similarly, a detailed It is a discussion of The key sequences use discussion of the significant results from the a battery lifetime of 4 accident sequences for both analysis.

hours. Appears division II CDF and LERF was added to The sensitivity study battery lifetime is 1 O the Internal Flood Report. used the same underlying Page 1 of 32 1-7 QU-F6 hours. Possible Resolution Refine analyses and upgrade documentation.

A quantitative definition of significant is not provided.

Possible Resolution Address During the Peer Review, a sensitivity study on the loss of offsite power (LOOP) initiators was performed, and based on the sensitivity study, the Recovery Rule files were changed from using the normal weather recovery probabilities to using the average weather recovery probabilities.

As documented in the LOOP timing analysis, RCIC operation is limited by suppression pool heat-up and/or Division I battery depletion while RCS depressurization is limited by Division II battery depletion.

The time assumed for power recovery prior to RCIC loss is 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (based on minimum design battery life), and the time assumed for power recovery prior to loss of RCS depressurization capability is 1 O hours (based on calculated battery depletion time without load shed). The quantitative definition of "significant" was added to the GGNS PRA Summary Report. Resolved, Closed Page 2 of 32 QU-F6: Cat 1-3 MET per independent assessment of finding closure methodology.

The battery lifetimes used in the analysis are division-specific.

Maintenance update. Although the definition of significant was not included in the GGNS documentation, the evaluation of the results was based on the definition of significant provided This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

1-8 QU-A2 RSC 14-15 (PRA Summary Report) provides results. Fault tree linking is used. Significant is not defined but sequences are rank ordered and provide a high percentage of the CDF results. [There are errors in the initiating event frequencies, wrong version of database was used, such that a new quantification is needed. In addition, there are conservatisms in significant accident sequences which should be addressed.

Primarily in the DC lifetime for division II which could impact the top two sequences and several others.] Possible Resolution Correct the model input and re-quantify.

The quantitative definition of "significant" was added to the GGNS PRA Summary Report, and the GGNS PRA Uncertainty and Sensitivity Report. These reports also identify the risk significant accident sequences based on this definition.

The Initiating Events Report was updated to ensure that it specified which column of values should be used in the CAFTA .RR file, and the .RR file used for quantification was updated to ensure it contains the values from the correct column. As documented in the LOOP timing analysis, RCIC operation is limited by suppression pool heat-up and/or Division I battery depletion while RCS depressurization is limited by Division II battery depletion.

The time assumed for power recovery prior to RCIC loss is 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (based on minimum Resolved, Closed Page 3 of 32 QU-A2: Cat 1-3 MET per independent assessment of finding closure in the Standard.

Added documentation does not impact the methodologies used or change the PRA scope or capability.

Maintenance update. Although the definition of significant was not included in the GGNS documentation, the evaluation of the results was based on the definition of significant provided in the Standard.

Added documentation does not impact the methodologies or change the PRA scope or capability.

The initiating event issue was a translation error when the values were transferred from the IE notebook to the RR file. The underlying methodology to calculate the initiating event This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

1-12 LE-E4 The LERF is quantified using the same general process as the CDF and is documented in the au notebook.

The review of the LE quantification against the requirements of Tables 2-2.7-2(a), (b) and (c) is essentially identical to the CDF reviews documented under the au High Level Requirement.

Direct linking of the Level 1 sequences with the CET provides assurance that all system dependencies are captured, etc. A LERF truncation sensitivity was performed, but does not meet the criterion identified in the au notebook.

However, the truncation was as low as could be achieved, and the lack of convergence does not significantly affect the results. Also when uncertainty is considered LERF mean value is calculated to design battery life), and the time assumed for power recovery prior to loss of RCS _ depressurization capability is 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> (based on calculated battery depletion time without load shed). The updated quantification of the Internal Events PRA and the Internal Flood PRA both now show convergence for both the pre-recovery and the post-recovery cases. The revised quantitative uncertainty analysis yields a mean value of CDF that is greater than the mean value of LERF. Resolved, Closed Page 4 of 32 LE-E4: Cat 1-3 MET per BWROG peer review values and for quantification were not affected by correcting the values The battery lifetimes used in the analysis are division-specific.

Maintenance update. A review of the LERF model identified an error in the LERF model where a gate that was supposed to be an AND gate was inadvertently modeled as an OR gate. With this error corrected, convergence was obtained, and the LERF was calculated to be a decade lower than CDF, as expected.

Correction of the modeling error to achieve convergence did not impact the methodologies used or change the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

1-13 IFPP-A5 exceed mean CDF value. This is not possible.

Possible Resolution Consider the reasons and address. Walkdowns are documented in RSC 13-20 Internal Flooding Walk down Documentation.

In general, this information was found to substantiate the flood zone definition discussions in Section 4.0 of RSC 13-37, Revision O (Internal Flooding Analysis).

Flooding scenarios associated with Control Building area OC125, which contribute to approximately 5% CDF may be overly conservative.

Based on discussions with GG PRA consultants, these scenarios were dominant due to the presence of DC equipment in this room, as documented in the GG equipment database.

However, this critical equipment is not Correction made to the equipment location mapping (OC215 replaced OC125) in the flooding analysis software (TIFA) and FRANX database.

Room OC215 has no flood sources, so no new scenarios were introduced by the correction of this mapping. Resolved, Closed Page 5 of 32 IFPP-A5: Cat 1-3 MET per BWROG Peer Review Maintenance update. The DC equipment was incorrectly mapped to room OC 125 rather than OC215. Correction of the mapping error did not result in new scenarios and did not impact the methodology used or change the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

2-1 IE-C15 located in this area. Possible Resolution Reevaluate the subject scenarios and equipment locations.

The mean values provided in the IE Notebook were not used in the quantification of the PRA results. The values from Table 9 in the IE Notebook were not correctly used in the CAFTAmodel.

Possible Resolution Update the CAFTA database to reflect the updated initiating event analysis mean value frequencies.

The Initiating Events Report was updated to ensure that it specified which column of values should be used in the CAFTA .RR file, and the .RR file used for quantification was updated to ensure it contains the values from the correct column. Resolved, Closed Page 6 of 32 IE-C15: Cat 1-3 MET per independent assessment of finding closure Maintenance update. This was a translation error when the values were transferred from the IE notebook to the RR file. The underlying methodology to calculate the initiating event values and for quantification were not affected by correcting the values, nor was there a change to the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

3-1 IE-C12 IE-C4 Table 6 of the initiating events report shows data used for Bayesian updating of plant specific initiating events. In some cases it appears that the plant experience would imply a substantially higher frequency than the prior data. For example for % T2 the prior is 1.12E-2 /yr whereas the plant specific experience is -0.3/yr. Also for % TSTT1 the prior is 8.80E-3 /yr whereas the plant experience is -0.13/yr.

These differences are large enough that the prior may not be appropriate for Bayesian updating.

Some explanation of this difference is warranted especially with regard to the Bayesian process. Also since the experience timeframe covers a period of much earlier GGNS operation, it is possible that more recent data is better because of plant fixes. Possible Resolution Provide justification for this deviance or consider alternate methods for calculating the IE frequencies.

Based on a review of the Plant-specific Data Analysis and Initiating Events Reports, there was a typo in the Prior Frequency Mean value for % T2 (LOCHS) and corresponding spreadsheet.

The value should be 1.12E-1 instead of 1.12E-2. The values were updated in the Initiating Events Report, the associated spreadsheet, and the CAFT A. RR file. The analysis for % TSTT1 incorrectly included the Loss of Switchyard Power Lines in*

• both the LOOP and the %TSTT1 IE frequencies, rather than the LOOP frequency only. Correction of the analysis reduced the frequency for % TSTT1 to be comparable to the generic estimate.

The current value is 9.19E-3/yr.

Resolved, Closed Page 7 of 32 IE-C4: Cat 1-3 MET per BWROG Peer Review IE-C12: Cat 1-3 MET per BWROG Peer Review Maintenance update. The issue associated with % T2 was a typo, for which the resolution did not impact the methodology or change the PRA scope or capability.

The issue associated with % TSTT1 was inclusion of applicable data when evaluating the IE frequency for transformer ST11. The underlying methodology or the PRA scope or capability were not changed, but the classification of the events was corrected to apply only to the LOOP frequency.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

3-2 IE-C15 Table 9 of the Initiating The Initiating Events Report Resolved, IE-C2: Cat Maintenance This finding was IE-C2 Events Notebook was updated to ensure that it Closed 1-3 MET update. This was a closed and therefore includes a summary of specified which column of per translation error has no impact on STI the Initiating Events values should be used in the BWROG when the values change evaluations Frequencies derived CAFTA .RR file, and the .RR Peer were transferred performed in from the updated IE file used for quantification was Review from the notebook to accordance with the analysis.

The Frequency updated to ensure it contains IE-C15: Cat the RR file. The SFCP. / per reactor year (the the values from the correct 1-3 MET underlying fourth column from the column. per methodology to left) shows the final independent calculate the updated number that assessment initiating event should be used for of finding values or for quantification.

However, closure quantification were the IE frequencies used not affected by for quantification have correcting the values, come from other columns nor was there a that do not represent the change to the PRA most recent data. scope or capability.

Possible Resolution Correct this transposition error. 4~4 HR-F1 There are multiple A review of the HFEs in the Resolved, HR-F1: Cat Maintenance This finding was human failure events GGNS model of record (MOR) Closed 1-2 MET update. The closed and therefore (HFEs) for performing was performed to identify per methodology used has no impact on STI the same action, only on those for performing the same BWROG for the calculation of change evaluations a different piece of action on different pieces of Peer the HEP values, and performed in equipment.

For example, equipment within the same Review the PRA scope or accordance with the there are three different system. A total of 16 capability, are not SFCP. HFEs for failing to start individual HFEs were replaced impacted by this The pending standby air compressors.

with 6 common HFEs for the change. documentation update If an operator fails to actions. The common HFEs to the System start a compressor, they have the same value of the Analysis Report to likely fail to start any/ individual HFEs they replaced, reflect the new HFE compressor, not just one which is effectively a names also has no in particular.

There dependency of 1.0, instead of impact on STI change should be only one assigning a dependency of evaluations.

failure for the operator to 1.0 during the dependency start a compressor that analysis.

fails the action for all air These changes were made in compressors.

Otherwise Page 8 of 32 4-5 HR-F2 HR-H2 there are failed and unfailed actions in the model to start the compressor.

Possible Resolution Group similar operator actions into one action. The timing of cues is not explicitly documented in the HRA calculator.

The time delay to the cue is set to zero in every instance.

The time delay is an important step because it can limit the amount of time in the scenario to recover from the action. The only timing listed in the time window is the median response and execution time. Operator recovery is based on the remaining time available, but without the time delay to the cue included, more time is allowed to recover than is actually available.

Possible Resolution Use the identified delay times in the HRA calculator to accurately reflect the timing of the actions in the scenario the Rev 4a MOR, HRA Report, Quantification Report, and Summary Report. The affected System Analysis Report Appendices will be updated per a Model Change Request (MCR). Time delays were added into the HRA Calculator, and the dependency analysis was updated using the new information.

Resolved, Closed Page 9 of 32 HR-F2: Cat2 MET per independent assessment of finding __ closure HR-H2: Cat 1-3 MET per BWROG Peer Review Maintenance update. The HRA Calculator is used for the calculation of the human error probabilities (HEPs). Inclusion of the timing of cues does not impact the calculated HEP but could impact the "order" of the HFEs in the dependency analysis.

The methodology used for the dependency analysis, and the scope and capability of the PRA are not changed by including the timing of cues. The inclusion of the cues helps to ensure correct ordering of the HFEs in a combination during the dependency analysis.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

4-6 HR-F2 HR-G4 and recovery actions. Scenario timeframes are included in the evaluation of the HFE. However, there are no references to where the scenario timeframes are calculated.

There was some indication that MAAP had been used in the past to develop the scenarios, but nothing could be found to support the times used. Following plant uprate a scaling evaluation of the increased power was performed to revise the scenario times. Additional MAAP cases were performed following the uprate, but these have not been incorporated into the HFE analysis.

Possible Resolution Determine the reference for each scenario timeframe and document the link between the HFE As documented in the Success Criteria Report, new MAAP thermal/hydraulic analyses were done after the extended power uprate to support the Rev. 4 PRA update. Scenario time frames were reviewed and addressed by adding the delay times for the HEP cues. The bases for the HFE timing were updated based on the new MAAP analyses and documented in the timing notes in the HRA Calculator database and in a detailed table on HFE timing in the HRA Report. Resolved, Closed Page 10 of 32 HR-F2: Cat 2 MET per independent assessment of finding closure HR-G4: Cat 2 MET per BWROG Peer Review Maintenance update. The methodology used by the HRA Calculator to calculate HEPs, and the PRA scope and capability were not impacted by documenting the updated bases for the HFE time frames. This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

and the reference.

4-7 HR-G2 All operator actions The updated HRA evaluation Resolved, HR-G2: Cat Maintenance This finding was include an estimation of no longer sets the execution Closed 1-3 MET update. For HFEs closed and therefore the failure in cognition.

probability to zero and instead per where no execution has no impact on STI However, a number of is based on the maximum BWROG contribution was change evaluations operator actions had the combined value for the Peer included, the performed in execution failure CBDTH/HCR approach.

Review execution actions accordance with the probability set to zero were added using SFCP. stating that the action is the same methods memorized and practiced as for all other HFEs. routinely.

These actions The underlying HRA are in the first few methodology, and minutes following an the PRA scope and initiating event and capability were not based on the time impacted by adding available may have high additional detail for HEPs. some of the HF Es. Possible Resolution Include execution failure probabilities for all operator actions. 4-10 HR-H3 The independent Time delays were added into Resolved, HR-H3: Cat Maintenance This finding was HR-G7 evaluation of HFEs did the HRA Calculator, and the Closed 1-3 MET update. The HRA closed and therefore not include any delay dependency analysis was per Calculator is used for has no impact on STI time to the cue. This updated using the new BWROG the calculation of the change evaluations carried forward into the information.

Peer HEPs. Inclusion of performed in dependency analysis In addition, the most Review the timing of cues accordance with the where all HFEs were significant HFE combinations HR-G7: Cat and consideration of SFCP. evaluated to have the were reviewed as part of the 1-3 MET intervening same delay time of zero. dependency analysis for per successes does not This paired events that separation of events and independent impact the calculated should be separated in intervening successes.

When assessment HEP but could the accident sequence identified, the default of finding impact the "order" of by hours together dependency was adjusted in closure the HFEs in the resulting in dependent the HRA Calculator software.

dependency combinations that should analysis.

The not exist or have a lower methodology used dependency.

With all of for the dependency Page 11 of 32 4-13 DA-C3 the actions having the same delay time, complete dependence was calculated resulting in much higher

• dependent failure probabilities than actually exist. The HRA calculator software has overrides available to offset delay times or reduce dependence, but these were not used. There also does not appear to be any evaluation of intervening successes which would remove the dependence between actions. Possible Resolution Perform the dependency analysis using accurate delay times. Include review for intervening successes.

A number of component types were excluded from the evaluation including motor operated valves, air operated valves, and temperature switches in PSA-GGNS-01-DA-01.

These component types were not reviewed for plant specific failures to determine if Bayesian updating of the generic failure data should be Additional plant-specific data was obtained for various valves and air compressors which were previously not included in the PRA. The new data includes number and type of failures, demand data, and exposure data per component and type code, and this data was analyzed consistent with the established data analysis Bayesian update methodology.

The new data was compiled into the Resolved, Closed Page 12 of 32 DA-C3: Cat 1-3 MET per independent assessment of finding closure' analysis, and the scope and capability of the PRA are not changed by including the timing of cues and consideration of intervening successes.

Maintenance update. Using Bayesian updating to evaluate reliability data for additional component types did not result in a change in methodology or in the scope or capability of the PRA. This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

4-14 DA-C3 4-15 DA-C13 performed.

Possible Resolution Include evaluation of these component types and subtypes so that plant specific data can be evaluated for inclusion to the generic failure rates. The failures removed from consideration do not have adequate justification for disregarding previous plant failures.

Many failures were removed in previous model revisions, but there is no documentation as to why the failures were no longer applicable.

Possible Resolution Develop bases for failure inclusion and exclusion and document the failures using th~ bases. One discrepancy was identified for battery charger unavailability.

In the notebook unavailability was calculated for the L51 battery chargers based on past history. However, the reliability database had zero unavailability for each of the battery chargers.

Possible Resolution spreadsheets used for the data analysis, and all changes and additions were documented in the Specific Data and CCF Report. The bases for failure inclusion and exclusion are established in the Plant-Specific Data and CCF Report, where it is now documented that all failures included in the PRA must have occurred during the time frame for the PRA update (September 1, 2006 through August 31, 2012) and must meet the definition of a PRA functional failure. The unavailability data for the 125V DC battery chargers was updated as documented in the Plant-Specific Data and CCF Report. All unavailability data was reviewed for similar concerns, and data for the following were also updated: radial well pumps, air compressors, AC circuit breakers, and switchyards.

Resolved, Closed Resolved, Closed Page 13 of 32 DA-C3: Cat 1-3 MET per independent assessment of finding closure DA-C13: Cat 2-3 MET per BWROG Peer Review Maintenance update. Enhancement of the documentation to describe the bases for excluding some equipment failure data does not impact the methodology used or change the PRA scope or capability.

Maintenance update. Update of unavailability data for several components to be consistent with plant operating history does not involve a change in methodology or change the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP. This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

4-17 DA-C14 4-19 DA~E1 Update the model unavailabilities for 125V DC battery chargers.

Coincident unavailability was identified to occur in the data analysis timeframe (PSA-GGNS-01-DA-01 ). This unavailability is not included in the model so is therefore not included in the final results. Possibl~ Resolution Include coincident maintenance in the model where analysis has determined it exists. There are numerous conflicts between the two data analysis notebooks and the two common cause notebooks.

This is likely due to a two-year gap between publishing of the notebooks.

Information is not consistent between notebooks and even within the same notebook.

The final data rollup notebook appears to be accurate, but its information is based off the plant specific notebook which has information that is out of date, not used, and A thorough review of previous analyses and the current system notebooks determined that the previously modeled coincident unavailabilities did not meet the criteria for inclusion.

It was confirmed that no more than one related system is scheduled to be in maintenance at any given time. The Plant-Specific Data and CCF Report was updated to document this review and information.

All data documentation was aggregated into a single Plant-Specific Data and CCF Report, which directly incorporates the supporting calculation spreadsheets, and describes the formulas used in the spreadsheets.

Resolved, Closed Resolved, Closed Page 14 of 32 DA-C14: Cat 1-3 MET per independent assessment of finding closure DA-E1: Cat 1-3 MET per independent assessment of finding closure Maintenance update. Enhancement of the documentation to describe the plant practices and criteria for modeling coincident unavailability did not change the methodology for modeling maintenance unavailability, or the PRA scope or capability.

Maintenance update. Consolidation, consistency update, and enhancement of documentation did not change the underlying data analysis methodologies or change the scope or capability of the PRA. This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP. This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

5-4 SC-AS AS-87 results in contradictory information to the data development notebook.

The same is true of the common cause notebooks.

Much of the plant specific data (run and demand estimates, maintenance unavailability data) was not found in the notebooks, but in spreadsheets provided separately.

This information should be included in the notebook for ease in identification.

Possible Resolution Resolve conflicts between notebooks and include supplementary data into the notebooks.

DC battery life is presented as 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> in the SC notebook, but the Div 11 battery was credited to 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> per the LOSP notebook.

The documentation is not consistent, and it is not clear if an operator action for load shedding is required.

Possible Resolution Determine realistic battery life times with and without load shedding, and model with any necessary HEPs in the As documented in the LOOP timing analysis, RCIC operation is limited by suppression pool heat-up (MAAP runs that credit RCIC under SBO conditions) and/or Division I battery depletion while RCS depressurization is limited by Division II battery depletion.

The time assumed for power recovery prior to RCIC loss is 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (based on Div. I minimum design battery life), and the time assumed for power recovery prior to loss of RCS depressurization capability is Resolved, Closed Page 15 of 32 SC-AS: Cat 2-3 MET per BWROG peer review AS-87: Cat 1-3 MET per BWROG Peer Review Maintenance update. Enhancement of the documentation to clarify the specific battery depletion times used in the analysis does not change the analysis methodologies are used or change the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

5-6 AS-87 PRA. AC power recoveries are developed on a cutset level to account for timing in the LOSP notebook (report GGNS-01-IE-01 ). Spot checks of the Qrecover file compared to the notebook identified the following errors/ inconsistencies:

used the "average" recovery of 6.56E-1 ZHE-OSP-DLGO-NW

-was entered into the Qrecover file with a probability of 1.22E-2 instead of 1.22E-1. Approximately 1 O other events were spot checked and found to be 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> (based on Div. II calculated battery depletion time without load shed). The DC Power system analysis documentation was updated to reference the division-specific battery lifetime from the LOSP analysis and revise the battery depletion assumption.

The Success Criteria Report was also updated to document the division-specific battery depletion times. The recovery factor typos documented in the F&O were corrected.

A detailed review of the remaining AC power recovery rules found no additional issues. During the Peer Review, a sensitivity study on the loss of offsite power (LOOP) initiators was performed, and based on the sensitivity study, the Recovery Rule files were changed from using the normal weather recovery probabilities to using the average weather recovery probabilities.

The timing for long term scenarios was addressed by a sensitivity study documented in the peer reviewed Quantification Report which has now been included in the base model. Resolved, Closed Page 16 of 32 AS-87: Cat 1-3 MET per BWROG Peer Review Maintenance update. Correction of typographical errors in recovery factors and update to include more appropriate offsite power recovery factors for average weather and term scenarios does not change the underlying methodology used to calculate the recovery probabilities or change the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

entered properly.

Additionally, the normal weather offsite power recovery data were applied to all the LOSP initiating events. The weighted average of the offsite power recovery probabilities did not include the severe weather portion in the weighting.

This makes the application of the non-recovery probabilities conservative Finally, the GG PRA team self-identified that offsite power recoveries for failure of OHR sequences was overly conservative, more aligned with loss of makeup timing than loss of OHR timing. The GG team performed a sensitivity that significantly reduced COF and LERF. Possible Resolution Review the entire list of offsite power recovery events to confirm they are entered into the Qrecovery file properly.

Apply normalized offsite power non-recovery probabilities that include the severe weather Page 17 of 32 5-7 AS-A7 5-8 AS-81 SC-83 component.

Re-evaluate the offsite power non-recovery probabilities for loss of OHR sequences to consider realistic probabilities.

The very small LOCA (%53) was identified as an initiating event in the IE analysis.

In Table 1 of the AS notebook; it was listed as being treated as a transient.

However, no basis is given, and the %53 initiating event is not included in the CAFTAmodel.

Possible Resolution Either provide a defendable basis for excluding the very small LOCA, or develop it for analysis.

The small and medium LOCA A TWS scenarios do not appear to have considered the LOCA effects on system success criteria, such as SLC. Large LOCA A TWSs have not been addressed with either a valid qualitative argument or a The %53 initiating event was Resolved, added to the list of transient Closed events in the Accident Sequence Analysis Report because it can be mitigated by the same equipment as a transient initiating event. This is consistent with inclusion as a transient event in the PRA model logic. The Success Criteria Report Resolved, was updated to document that Closed any medium LOCA or large LOCA with failure of rod insertion is assumed to lead to core damage. A small LOCA would not impact the mechanical or electrical reactor protection system (RPS), or the ability to manually scram, perform Page 18 of 32 AS-A7: Cat Maintenance This finding was 1-2 MET update. Clarification closed and therefore per of the documentation has no impact on STI BWROG on the treatment of change evaluations Peer very small LOCA performed in Review does not change the accordance with the methodology for SFCP. identification and grouping of initiating events or change the PRA scope or capability.

AS-81: Cat Maintenance This finding was 1-3 MET update. closed and therefore per Enhancement of the has no impact on STI BWROG documentation on change evaluations Peer the success criteria performed in Review for LOCA A TWS accordance with the SC-83: Cat does not change the SFCP. 1-3 MET methodology to per define accident BWROG sequence Peer progression.

There is quantitative evaluation.

A success criteria basis could not be found for using RCIC to depressurize to allow SDC in transients or ATWS. In transient sequences with success of depressurization, SDC is credited to prevent core damage, which disagrees with the MAAP calculation CALMAP-2014-1202, which shows this sequence as core damage. Possible Resolution Document the success criteria for LOCA A TWS events. Document bases for use of RCIC and SDC to make a sequence a safe, stable end state. If this cannot be justified, should be considered core damage. Remove credit for depressurization/SDC to prevent core damage in transients.

alternate rod insertion, or trip the recirculation pumps. Therefore, the only system in question for a small LOCA A TWS is standby liquid control (SLC). Based on the system design criteria for SLC, and in accordance with GDC 4, SLC is designed to operate following a LOCA. Therefore, the leakage during a small LOCA is not large enough to render SLC ineffective regardless of the location of the leak. The Success Criteria Report also documents that RCIC is not credited as a method of depressurization in the transient or A TWS accident sequences.

Decay heat removal options with successful RCIC injection are limited to RHR in Suppression Pool Cooling (SPC) Mode and RHR in Containment Spray (CS) Mode. Decay heat removal via RHR in Shutdown Cooling (SDC) Mode is not credited as a viable option when RCIC is injecting for inventory control. A new MAAP calculation was performed for a transient with depressurization in which LPCI and SPC alternate based on RPV level. The plant reaches a safe stable state after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Based on the Page 19 of 32 Review also no change to the PRA scope or capability.

5-9 SC-81 SC-82 GGNS assumes that suppression pool makeup (SPMU) is required in combination with containment venting in order to avoid cavitation of ECCS pump suction in containment heat-up sequences.

The assumption that venting fails the ECCS pumps is conservative, which is noted in Topic 7 of Table 11 of the QU notebook.*

Regarding SPMU successfully facilitating pump operation, there is no analytical basis for this success criteria, but instead is based upon the expert judgment of the modeler. While this may be a reasonable assumption, it would be better to have an analytical basis or at least carry th is item as an additional source of modeling uncertainty.

Since these assumptions are a significant driver to similarity of flow between the LPCI/SPC alignment and SOC, this case also provides a basis to credit SOC. SPMU is only required for large and medium LOCAsthat experience failure of decay heat removal. For large LOCA, the ECCS pumps are assumed to fail due to loss of NPSH if containment venting or containment failure occurs. The above assumptions are based on MAAP analysis showing SP level drops to the SPMU limit shortly after the sequence mission time. The ECCS pumps can pump saturated water if SP level remains above SPMU limit, unless there is flashing of the SP water, steam entrapment, cavitation, or a pump trip when containment fails. However, injection of CST volume will increase level in the SP and the potential for trip was eliminated in most of cases. The use of HPCS after containment failure is now addressed in the Accident Sequence and Success " Criteria Reports. Resolved, Closed Page 20 of 32 SC-81: Cat 2 MET per BWROG Peer Review SC-82: Cat 2-3 MET per BWROG Peer Review Maintenance update. Enhancements to documentation to clarify and provide the analytical basis for analysis assumptions does not impact the methodology used to determine accident sequence progression or success criteria.

There is also no change to the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

5-10 LE-A2 the CDF and LERF, consideration should be given to attempt to refine the assumption.

At a minimum, sensitivity analyses should be performed to ensure the impact of these SC assumptions are fully understood for risk characterization.

Possible Resolution Since these assumptions are a significant driver to the CDF and LERF, consideration should be given to attempt to refine the assumptions.

At a minimum, sensitivity analyses should be performed to ensure the impact of this SC assumption are fully understood for risk characterization.

The characteristics identified as important in LE-A 1 are documented in Section 1 of the LE notebook (PSA-GGNS-01-LE). However, the LE notebook does not provide any bases for the binning of sequences (e.g., determination of which sequences are high pressure and which are low). Per the Grand Gulf PRA team, selection was based on The LERF Report was updated to clearly define the high to low pressure transition at 200 psig, based on MAAP analysis.

The updated report also clarifies that only the pressure at the time of RPV failure is relevant for this binning criterion.

This resulted in a change to the binning of small LOCA sequences with successful depressurization

-prior to RPV failure to low Resolved, Closed Page 21 of 32 LE-A2: Cat 1-3 MET per BWROG peer review Maintenance update. Enhancement of the documentation, assumptions and bases for sequence binning did not changes the methodology used for binning or result in a change to PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

information from MAAP pressure scenarios.

gathered from both success criteria and LE RF-specific assessments and the engineer's experience working on other BWR 6 designs. This SR is considered met because the binning appears reasonable in most cases, but documentation of more definitive bases is needed. Some examples of sequences for which the high/low pressure binning are not obvious are: P-009 {SORV, RCIC initially successful, but LPI fails and RX depressurization not questioned) is "Low" pressure, and all Small LOCAs { even with depressurization successful) are binned as high pressure.

Possible Resolution Document the bases for the binning of the characteristics.

Qualitative evaluation of many of the sequences is intuitive (e.g., large LOCAs are low pressure),*

but some detail should be provided for the binning of less Page 22 of 32 5-12 LE-C10 LE-C12 LE-F2 LE-C3 LE-G3 LE-G6 obvious sequence characteristics.

Clearly identify the criterion for high vs. low pressure binning (200 psi). There is no quantitative definition of significant accident progression sequences.

There are SRs that require documenting the quantitative definition, as well as review of the significant severe accident progression sequences for possible credit for repairs and engineering analyses to provide a more realistic analysis.

An example of the lack of reviews for excess conservatism is that the operator action for turning on the H2 igniters was set to 1.0 in the analysis, yet is very significant to the results. Possible Resolution Define severe accident progression sequence and review the results to remove significant conservatisms.

The quantitative definition of "significant accident progression" was added to the Model Integration and Quantification Report. The new LERF MAAP Analysis Report documents the basis of the accident sequence progression.

The Quantification Report was updated to discuss the review and relative contributions of the LERF sequences, and the Summary Report was updated to provide a comparison of initiating event and other relative contributions to LERF as well as a more detailed comparison of the relative sequence contributions to LERF. The operator action to start the igniters was updated and documented in the HRA Report. The updated HEP was incorporated in the model during the rule-based Resolved, Closed Page 23 of 32 LE-C3, C10, C12: Cat 2 MET per independent assessment of finding closure LE-F2: Cat 1-3 MET per independent assessment of finding closure LE-G3: Cat 2 MET per independent assessment of finding closure LE-G6: Cat 1-3 MET per independent assessment of finding closure Maintenance update. Update of the LERF model and results with the updated igniter HEP did not require a change in methodology or PRA scope or capability, but updated insights were obtained.

Additional review and enhanced documentation of the LERF results contributions, and documentation of definitions also did not affect methodology or PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

5-13 LE-C1 LE-C2 The approach to the LE analysis was the NUREG/CR-6595 analysis, with a more detailed evaluation of the loss of OHR sequences.

Since the Level 1 and LE results are dominated by loss of OHR, this SR was evaluated as met to Category 11. However, the following items are also noted from the peer review: -The Level 1 SR review identified many items that will change the COF risk results (incorrect IE frequencies utilized, incorrect offsite power recoveries applied, etc.). -The LE results are recovery process. The model was -re-quantified and reviewed after changes were made. The igniter operator action identified in the Finding was the only significant conservatism that required refinement.

Although the relative importance of hydrogen combustion was reduced, the updated LERF results continue to show that LERF is dominated by containment failures caused by loss of suppression pool cooling and failure of the hydrogen igniters.

The LERF model reviewed by the BWROG Peer Review team double-counted early and late hydrogen events in many of the cutsets which resulted in overestimation of LERF. The LERF model was updated to eliminate the counting, as verified by LERF cutset reviews. The Level 1 model/data

• issues identified in other F&Os were corrected as described in those F &Os listed above. A new "Basis for Value" column was added to the LERF basic events table in the LERF Analysis Report to explain the bases for the values used. Resolved, Closed Page 24 of 32 LE-C1: Cat 2 MET per BWROG peer review LE-C2: Cat 2 MET per independent assessment of finding closure Maintenance update. As stated in the finding, the LERF analysis and quantification methodology used conforms to Cat 2. Update of the LERF model and results to address issues in the Level 1 and 2 PRA models, and the updated igniter HEP dtd not require a change in methodology or PRA scope or capability, but updated insights were obtained.

This finding was closed and-therefore has ho impact on STI change evaluations performed in accordance with the SFCP.

5-14 LE-C2 LE-C4 LE-E1 dominated by containment failure at vessel breach, but the majority of this fraction is hydrogen-related failures of containment.

The HEP for turning on igniters was set to 1.0; the peer review team set it to 0.1 in the LERF cutset file (1 E-10/yr truncation), and the LERF dropped from 5.81 E-6/yr to 1.74E-6/yr. Because of the peer review team's uncertainty in how the risk profile will change with the identified errors, a finding is developed to ensure the LE analysis is examined.

If the LE risk profile changes significantly, then some conservative assessments may require more detailed analysis.

Possible Resolution After updating the Level 1 and igniter operation action probability, consider if a more detailed approach to LE is necessary.

No credit is given to operator actions in the LE analysis.

There is no documentation of a Updates to the igniter HEPs were performed in the HRA and the values were added to the HRA and LERF Analysis Reports. The model was re-quantified and reviewed after the above changes were made. Although the relative importance of hydrogen combustion was reduced, the updated LERF results continue to show that LERF is dominated by containment failures caused by loss of suppression pool cooling and failure of the hydrogen igniters.

The LERF Analysis Report was updated to document the review of procedures that had previously been performed to Resolved, Closed Page 25 of 32 LE-C2: Cat 2 MET per independent assessment Maintenance update. Enhanced documentation of the process used to This finding was closed and therefore has no impact on STI change evaluations review of Grand Gulf identify the operator actions of finding identify and model performed in procedures for severe that would be used to respond closure operator actions for accordance with the accident responses by to the LERF sequence in LE-C4: Cat LERF sequences did SFCP. operations.

There is a progress.

As part of this 2 MET per not change the HEP identified for turning review, the systems that could BWROG methodology used on hydrogen igniters, but be used by the operators to peer review for the LERF it is set to 1.0 in the respond to the scenario had LE-E1: Cat analysis or HRA, or model. been identified and modeled 1-3 MET the PRA scope or Possible Resolution appropriately.

The LERF-per capability.

Nor did Document a review of related operator actions were BWROG the update of the GG procedures for evaluated as part of the peer review HEP to actuate the potential operator actions Human Reliability Analysis.

igniters.

to reduce the LERF. Updates to the igniter HEPs were performed in.the HRA and the values were added to the HRA and LERF Analysis Reports. 5-16 LE-03 Per the containment The evaluation of containment Resolved, LE-03: Cat Maintenance This finding was capacity report (GGNS failure in the updated Success Closed 2 MET per update. Additional closed and therefore 92-0034), the failure Criteria Report considers the independent evaluation of has no impact on STI location with the lowest location of the containment assessment containment failure change evaluations mean pressure is the failure. The specific locations of finding locations did not performed in basemat (65 psid). The considered were based on the closure change the accordance with the containment failure same design basis

• methodology used SFCP. location was not Containment performance for the LERF *considered in the LE calculation used to identify the analysis, or the PRA analysis (all basemat as the weakest point. scope or capability.

overpressurization was The LERF analysis does not The lack of credit for considered a large credit any fission product a decontamination release, after the 0.5 scrubbing based on factor for releases scrubbing credit for the Containment Failure location has no numerical Auxiliary Building).

since no approve,c;I impact on LERF. Since basemat failures methodology for crediting could potentially result in scrubbing due to Containment underground releases to failure location currently allow significant exists. scrubbing, the approach There is some probability that taken is conservative.

the containment failure could It is noted that other be located such that it impacts Page 26 of 32 5-18 LE-C1 LE-E3 containment failure locations have mean failure pressures that are not much higher than the basemat failures, but some credit could be given to reduce the LERF. [Basis: SR LE-03 Category II states when containment failure location affects the LERF, define failure location using a realistic assessment.]

Possible Resolution Evaluate the potential for containment failures to result in underground releases to potentially remove conservatism from the LERF. The GG LE analysis does not provide a quantitative definition of 'Large' releases, and does not document the evaluation of sequences as resulting in a 'Early' release. Discussions with the GG PRA team identified that the 'Early' evaluations were based on comparison of predicted containment failure time for the dominant sequence (loss of OHR) with the time of declaration of a general HPCS operation and the probability of this occurrence was developed in the Calculation of Split Fraction for GGNS ECCS Equipment Given Containment Failure. Additional GOTHIC room heatup analyses were run to evaluate the environment that would be present in the HPCS and LPCS rooms following a

• Containment failure at a location other than at the base mat. The new LERF MAAP Analysis Report defines 'Large' and 'Early' releases, documents the results of the LERF MAAP analyses versus the defined criteria, and summarizes the MAAP runs that contribute to LERF. Resolved, Closed Page 27 of 32 LE-C1: Cat 2 MET per BWROG peer review LE-E3: Cat 2 MET per independent assessment of finding closure Maintenance update. Enhanced documentation of "large" and "early release definitions did not change the methodology used for the LERF analysis, or the PRA scope or capability.

• This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.

5-20 LE-F1 LE-F2 LE-G3 emergency.

This is acceptable, but the evaluation needs to be documented.

[Basis: The LERF contributors from the CSET are all evaluated and carried through the accident progression analysis to the CSET end states. The end states categorization of LERF or not LERF is not documented, but appears to be reasonable.]

Possible Resolution Document a definition of 'Large' releases, and document the evaluation of sequences as resulting in a 'Early' release. The evaluation of 'Large' was qualitative, but appears reasonable (e.g., ISLOCA, Containment Isolation, Containment rupture), but needs to be documented, and the bases should be tied to a quantitative definition of 'Large.' The Quantification notebook QU-01) presents the total LERF, the top 100 LERF cutsets, and some LERF importance analyses.

There is no presentation The Quantification Report was Resolved, updated to discuss the relative Closed contributions of the LERF sequences, and the Summary Report was updated to provide a comparison of initiating event and other Page 28 of 32 LE-F1, G3: Cat 2-3 MET per independent assessment of finding Maintenance update. Enhanced documentation of the LERF results contributions did not change the methodology used This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the of the relative relative contributions to LERF closure for the LERF SFCP. contribution to LERF as well as a more detailed LE-F2: Cat analysis or from various contributors comparison of the relative 1-3 MET quantification, or the other than the sequence contributions to per PRA scope or im~ortance analysis.

LERF. independent capability.

Possible Resolution assessment Document the relative of finding contribution to LERF closure from plant damage states and significant LERF contributors from Table 2-2.8-3. 5-22 LE-GS Limitations in the The Limitations of the LERF Resolved, LE-GS: Cat Maintenance This finding was analysis have not been analysis were added to the Closed 1-3 MET update. Enhanced closed and therefore identified.

The LE LERF Analysis Report. per documentation of the has no impact on STI analysis should be independent LERF analysis change evaluations examined to identify how assessment limitations did not performed in any simplifying of finding change the accordance with the assumptions can impact closure methodology used SFCP. applications.

for the LERF Possible Resolution analysis, or the PRA Examine the LE analysis scope or capability.

to identify how any simplifying assumptions can impact applications.

7-1 IFPP-83 There is no apparent The uncertainty associated Resolved, IFEV-83, Maintenance This finding was IFS0-83 documentation of an with internal flood plant Closed IFPP-83, update. Enhanced closed and therefore IFSN-83 uncertainty analysis for partitioning; internal flood IFQU-83, documentation of the has no impact on STI any of the following:

sources; flood-induced IFSN-83, uncertainty change evaluations IFEV-83 internal flood plant initiating events; accident IFS0-83: associated with the performed in IFQU-83 partitioning; internal flood sequences and quantification Cat 1-3 IF analysis did not accordance with the sources; flood-induced were added to Internal MET per change the SFCP. initiating events; accident Flooding Analysis Report. The independent methodology used sequences and discussion addresses both assessment for the IF analysis, or quantification aleatory and epistemic of finding the PRA scope or Possible Resolution (modeling) uncertainty.

closure capability.

Document the uncertainties in Page 29 of 32 7-2 7-4 IFSN-A12 IFSN-A13 IFSO-A3 IFEV-A3 accordance with the standard The EDG building is screened from further analysis based upon .a statement in FSAR that "pipe cracks are not postulated inside the diesel generator building;" Possible Resolution Consider spray scenarios for the diesel generator buildings.

The flood induced initiating event defaults to loss of power conversion system plant initiator (T-2) is conservatiye.

Possible Resolution Run sensitivities to simple turbine trip for comparison to loss of power conversion system. The Internal Flooding Analysis Report was updated to provide a more robust basis for screening of the EDG building, based on the inability of a flood within the EDG . building to result in a reactor trip since offsite power would not be affected.

Review of the internal flooding (IF) FRANX database revealed that pipe breaks associated with a loss of a system that results in a reactor trip (e.g., plant service water, circulating water) are actually grouped with the internal events initiator associated with that failed system. Only IF initiators that do not result in a loss of a system that causes a plant trip are grouped with the loss of power conversion system initiating event. This is much more detailed than suggested by the peer review team. The Internal Flooding Analysis Report was updated to better describe the grouping of Resolved, Closed Resolved, Closed Page 30 of 32 IFSN-A12 IFSN-A13, IFSO-A3: Cat 1-3 MET per BWROG peer review IFEV-A3: Cat 1-2 MET per BWROG Peer Review Maintenance This finding was update. Enhanced closed and therefore documentation of the has no impact on STI EDG building change evaluations screening from the IF performed in analysis did not accordance with the change the SFCP. methodology used for the IF flood area screening, or the PRA scope or capability.

Maintenance This finding was update. Enhanced closed and therefore documentation of the has no impact on STI IF initiating event change evaluations grouping did not performed in change the accordance with the methodology used SFCP. for the IF quantification, or the PRA scope or capability.

7-5 IFQU-A1 It is not evident that accident sequences were performed and documented.

There is little evidence contained in RSC-CALKNX-2015-0803 Possible Resolution Perform and document quantification in accordance with H LR IFQU-A1 7-7 IFQU-A1 In Table 1, reference to IFQU-A2 Section 14.0 seems IFQU-A3 incorrect.

IFQU-A4 Possible Resolution Revise Table 1 as IFQU-A7 appropriate IFQU-A10 IFQU-B1 [IFQU-A7 Basis: it is not evident that the IFQU-B2 requirements of 2-2. 7 were satisfied.

For example, no evidence of convergence determination or uncertainty analysis.]

[IFQU-B2 Basis: While initiating events and document the detailed grouping in an appendix.

The internal flooding model is integrated with the internal events model, and the internal flood accident sequences are quantified using the same methodology as the internal events accident sequences.

The Internal Flooding Analysis Report was updated to document the IF accident sequences (formerly included in the PRA Summary Report). Table 1 (IF "roadmap" for the ASME/ANS PRA requirements) was revised to reference the correct sections and/or other reports as necessary.

The updated quantification of the Internal Events PRA and the Internal Flood PRA both now show convergence for both the pre-recovery and the post-recovery cases. A review of the Internal Flood (IF) analysis was performed and documented in the Internal Flooding Analysis Resolved, Closed Resolved, Closed Page 31 of 32 IFQU-A1: Maintenance This finding was Cat 1-3 update. Enhanced closed and therefore MET per documentation of the has no impact on STI BWROG IF accident change evaluations Peer sequences did not performed iri Review change the accordance with the methodology used SFCP. for the internal events or IF sequence quantification, or the PRA scope or capability.

IFQU-A1, Maintenance This finding was IFQU-A2, update. Corrections closed and therefore IFQU-A4: to the IF "roadmap" has no impact on STI Cat 1-3 for the ASME/ANS change evaluations MET per PRA requirements, performed in BWROG and enhanced accordance with the Peer documentation of the SFCP. Review IF quantification, IFQU-A3: convergence and Cat 1-2 results did not MET per change the BWROG methodology used Peer for the IF Review quantification, or the IFQU-A7, PRA scope or IFQU-A10 capability.

7-8 IFQU-A10 most are documentation

~lements satisfied, the is no evidence to support d, results of the IF analysis consistent with D] Although it is apparent that quantification of the flooding model was performed as documented in CALKNX-2015-0803, it is not evident that the LERF analysis was reviewed and documented.

Possible Resolution Document the LERF analysis in accordance with IFQU-A10 Report. This review included cutset reviews for the IF, a review and discussion of the significant IF accident sequences, a review and discussion of the significant IF cutsets, and identification of the top IF basic events, HFEs, Maintenance events, CCF events, and initiating events based on Fussell-Vesely and RAW. A review of the Internal Flood (IF) LERF analysis was performed and documented in the Internal Flooding Analysis Report. This review included cutset reviews for the IF LERF, a review and discussion of the significant IF LERF accident sequences, a review and discussion of the significant IF LERF cutsets, and identification of the top IF LERF basic events, HFEs, Maintenance events, CCF events, and initiating events based on Fussell-Vesely and RAW. Resolved, Closed Page 32 of 32 IFQU-81, IFQU-82: Cat 1-3 MET per independent assessment of finding closure IFQU-A10:

Cat 1-3 MET per independent assessment of finding closure Maintenance update. Enhanced documentation of the IF LERF results did not change the methodology used for the IF LERF quantification, or the PRA scope or capability.

This finding was closed and therefore has no impact on STI change evaluations performed in accordance with the SFCP.