L-15-235, Response to Request for Additional Information Regarding License Amendment to Adopt Technical Specification Task Force Traveler-425

Response to Request for Additional Information Regarding License Amendment to Adopt Technical Specification Task Force Traveler-425

ML15237A035
Person / Time
Site:	Perry
Issue date:	08/24/2015
From:	Harkness E FirstEnergy Nuclear Operating Co
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
L-15-235, TAC MF3720
Download: ML15237A035 (49)

Text

Perry Nuclear Power Plant PO Box 97 10 Center Road RrstEnergy Nuclear Operating Company Perry. Ohl° 440a 1 Ernest J. Harkness 440-280-5382 Vice President Fax 440-280-8029 August 24, 2015 10 CFR 50.90 L-15-235 ATTN: Document Control Desk U.S. Nuclear Regulatory Commission Washington, DC 20555-0001

SUBJECT:

Perry Nuclear Power Plant Docket Number 50-440, License Number NPF-58 Response to Request For Additional Information Regarding License Amendment to Adopt Technical Specification Task Force Traveler-425 (TAC No. MF3720)

By correspondence dated March 25, 2014 (Accession No. ML14084A165), as supplemented by letter dated October 7, 2014 (Accession No. ML14281A125),

FirstEnergy Nuclear Operating Company (FENOC) submitted a license amendment request for the Perry Nuclear Power Plant (PNPP). The proposed amendment would modify the PNPP Technical Specifications by relocating specific surveillance frequencies to a licensee controlled program with the implementation of Nuclear Energy Institute (NEI) 04-10, "Risk-Informed Technical Specifications Initiative 5b, Risk-Informed Method for Control of Surveillance Frequencies." The proposed amendment is consistent with Nuclear Regulatory Commission (NRC)-approved Technical Specification Task Force (TSTF) Traveler TSTF-425, Revision 3, "Relocate Surveillance Frequencies to Licensee Control - Risk Informed Technical Specification Task Force (RITSTF) Initiative 5b," with certain proposed deviations.

On June 26, 2015 via electronic correspondence (Accession No. ML15179A007), the NRC requested additional information to complete the staff's review. FENOC's response to this request is attached.

There are no regulatory commitments established in this submittal. If there are any questions or additional information is required, please contact Mr. Thomas A. Lentz, Manager - Fleet Licensing, at (330) 315-6810.

Attachment L-15-235 Page 2 of 2 I declare under penalty of perjury that the foregoing is true and correct. Executed on August %H , 2015.

Sincerely, Ernest J. Harkness

Attachment:

Response to June 26, 2015 Request for Additional Information cc: NRC Region III Administrator NRC Resident Inspector NRC Project Manager Executive Director, Ohio Emergency Management Agency, State of Ohio (NRC Liaison)

Utility Radiological Safety Board

Attachment L-15-235 Response to June 26, 2015 Request for Additional Information Page 1 of 47 By correspondence dated March 25, 2014, FirstEnergy Nuclear Operating Company (FENOC) submitted a license amendment request for the Perry Nuclear Power Plant (PNPP). The proposed amendment was supplemented by letter dated October 7, 2014, The proposed amendment would modify the PNPP Technical Specifications by relocating specific surveillance frequencies to a licensee controlled program with the implementation of Nuclear Energy Institute (NEI) 04-10, "Risk-Informed Technical Specifications Initiative 5b, Risk-Informed Method for Control of Surveillance Frequencies."

On June 26, 2015 via electronic correspondence, the Nuclear Regulatory Commission (NRC) staff requested additional information to complete their review. The request for additional information (RAI) is presented in bold type, followed by the FENOC response.

The letter dated October 7, 2014 provided the response to RAI 1.

RAI 2

Capability Category II of the endorsed American Society of Mechanical Engineers (ASME)/American Nuclear Society (ANS) PRA standard (i.e., ASME/ANS RA-Sa-2009) is the target capability level for supporting requirements for the internal events probabilistic risk assessment (PRA) for this application. In the response to RA11, dated October 7, 2014, FirstEnergy Nuclear Operating Company (FENOC or the licensee) provided the list of Facts and Observations (F&Os) findings and suggestions from the 2008 Gap Analysis and the 2011 and 2012 focused-scope peer reviews for large early release frequency (LERF) analysis and internal flooding analysis, respectively. The licensee also stated:

"The 1997 PSA [Probabilistic Safety Assessment] Peer Review Certification F&Os were not included as the follow-on reviews were a complete reevaluation to the PRA standard in effect and supersede this information. Therefore, the 1997 PSA Peer Review Certification F&Os are not considered relevant to the application."

In the application dated March 25, 2014, the licensee provided a summary of the PNPP PRA history. This summary stated that a model update and Computer Aided Fault Tree Analysis System (CAFTA) model conversion were performed in 2011. (The PNPP PRA originally used the WinNUPRA code, as explained in the letter dated October 7, 2014.) The NRC staff notes that many PRA model conversions also result in "PRA upgrades."

Attachment L-15-235 Page 2 of 47 Based on Section 1-5, "PRA Configuration Control," of ASME/ANS RA-Sa-2009, and RG 1.200, Revision 2, Regulatory Position 1.4, "PRA Development, Maintenance, and Upgrade," a PRA upgrade must be peer reviewed. The NRC staff reviewed the response to RA11, in the letter dated October 7, 2014, and notes that a number of revisions were made to the PNPP PRA in response to the 2008 Gap Analysis. The licensee appears to treat all model revisions as PRA updates (i.e., PRA maintenance) and not PRA upgrades, and therefore, the PRA model does not need a subsequent peer review. The ASME/ANS PRA Standard defines upgrades as, "incorporation into a PRA model of a new methodology or significant changes in scope or capability that impact the significant accident sequences or the significant accident progression sequences." Based on this definition, please address the following:

a) For F&O IE-A4a, the Status/Gap suggested using EPRI [Electric Power Research Institute Technical Report] TR-1013490 as guidance for support system initiator fault trees. (The NRC staff notes that EPRI TR-1013490, "Support System Initiating Events: Identification and Quantification Guideline," published in 2006, appears to have been superseded by its Technical Update, EPRI TR-1016741, published in December 2008. EPRI TR-1016741 was also sponsored by the NRC Office of Nuclear Regulatory Research, and focuses on the treatment of common cause failures (CCFs) when modeling support system initiating events.) Although the disposition of the F&O states that the initiating event fault trees were updated to include CCF events for applicable components, it is unclear if the disposition resulted in a change in the capability of the PNPP PRA that impacted the significant accident sequences (similar to Example 5 in Nonmandatory Appendix 1-A of ASME/ANS RA-Sa-2009) and could constitute a PRA upgrade. Please expand on the discussion of how this gap was resolved (i.e., discuss the methodology used and describe the changes to the model) and why it is not considered an upgrade.

b) For F&O DA-D5, the Status/Gap states: "The Alpha Factor method is used to conduct the CCF analysis...." The Perry Resolution then states: "In the PRA model update, the Multiple Greek Letter method was used to perform the Common Cause analysis for the model...." The Multiple Greek Letter and Alpha Factor Method are similar, but given the other changes to CCF modeling, this could constitute a PRA upgrade. Please explain why this model change is not considered an upgrade.

c) For F&O HR-G7 the Status/Gap states: "Although some dependencies are identified during the identification and definition process, all possible HFE

[human factors engineering] combinations and dependencies are not addressed." The Perry Resolution then states: "HRA [human reliability analysis]... for post-initiators, including HEP [human error probability]

dependencies, has been completely redone...." It is unclear if a different

Attachment L-15-235 Page 3 of 47 HRA approach to human error analysis and dependency analysis than what has previously been used has been applied. A new approach could also constitute a PRA upgrade. Please expand on the discussion for the resolution of this gap and why it is not considered an upgrade.

Response

The 2008 Gap Assessment, performed to the ASME PRA Standard Addendum "b",

endorsed by Regulatory Guide 1.200 (RG 1.200) Rev. 1, was considered by FENOC to be equivalent to an industry Peer Review. The reason for not formally declaring it a Peer Review at the time was due to the fact there were no utility peers on the review team. This gap assessment brought in a team of highly qualified outside contractors with the direction to lead and perform the review, document the results per the industry peer review guidance of NEI 05-04, "Process for Performing Internal Events PRA Peer Reviews Using the ASME/ANS PRA Standard" and to identify where the PRA model did not fully address the Supporting Requirements (SRs) in the ASME/ANS PRA Standard. This gap assessment also reevaluated the identified Findings and Observations (F&Os) from the previous 1997 peer review, as it was felt based on preliminary review that they had not been adequately addressed, and then expand upon those insights additional findings and observations as needed based on the current state of knowledge and methods. Furthermore, this assessment also determined the amount of effort that would be required to close the identified F&O's (gaps), achieve Capability Category (CC) II for SRs in the standard, and develop a gap closure project plan, schedule, and cost estimate. During the review it was determined that some areas were deficient such that it would be more cost-effective to completely re-perform the analysis. High level requirements and supporting requirements for Section 2-2.8, "LERF Analysis (LE)" and Section 3-2, "Internal Flood Analysis (IF)" of the ASME/ANS PRA Standard therefore were not reviewed in the 2008 gap assessment. Following close-out of the identified gaps in the Level 1 internal events portions of the model, new analysis then commenced for LE and IF.

Some members of the gap assessment team were retained following the assessment to support FENOC staff in addressing the F&O's and review the changes to ensure the responses were adequate. FENOC did not consider the F&O's from the gap assessment closed at this point, since the reviewer was directly involved in the solution and therefore no longer independent. The new modeling and methodologies applied in LE and IF did constitute an upgrade and were the subject of focused scope peer reviews.

The original WinNUPRA model was a linked fault tree model at the time of the review, and the conversion of a WinNUPRA linked fault tree to a CAFTA linked fault tree is not considered an upgrade as both models employ the same general methodology (Example 11 of Appendix 1-A of the ASME/ANS PRA Standard).

Before proceeding with F&O resolution, sensitivity studies were performed at the time of conversion and no significant changes in results were identified, as documented in analysis/assessment PRA-PY1 003. Following conversion,

Attachment L-15-235 Page 4 of 47 cosmetic and structural improvements were made to address F&O's as directed in the recommendations. In many cases, to accomplish this it was found to be easier to re-create the fault trees in CAFTA rather than re-use the original WinNUPRA fault trees. For example, the WinNUPRA model often considered a single train to be running with the alternate in standby. The revised CAFTA fault trees included the probability that a given train was in standby vs. run. The new CAFTA fault trees and resulting cutsets and importance measures were then compared to the original fault trees and results to ensure no systems, structures, and components (SSCs) were omitted and that no logic errors were introduced. This was considered to be PRA maintenance as no fault tree logic was altered such that it significantly impacted the model results, except by the correction of noted errors. There were significant editorial changes made to the model as well, such as consistent use of basic event names via type code and type set capability in CAFTA, the use of the common cause module in CAFTA, and other cosmetic changes such as consistent modeling philosophy applied (for example, not collapsing OR Gates) so a PRA analyst could transition from one fault tree to the next and see the same general structure used.

Therefore, it is FENOC's interpretation that none of these changes constituted a "new methodology or significant changes in scope or capability."

Similar to the fault trees, the event trees were also reviewed and restructured to ensure consistency in modeling philosophy, making it easier to understand and review. Some unique independent event trees were removed from the model and logically relocated or subsumed in the event trees following industry best practices, such as loss of feedwater and station blackout (SBO). The loss of feedwater initiator was re-mapped to use the general transient event tree, which has an identical operator response with the exception of the availability of feedwater, which was captured in the underlying fault trees. Station blackout is not an initiating event, but rather a possible condition following a loss of offsite power (LOOP). Thus it is now captured in the LOOP event tree, which includes in the underlying logic the potential that either or both divisional diesels would fail. This change removed the need for additional mutually exclusive logic to prevent failures to both diesels from appearing in the LOOP tree. The operator response is otherwise the same for both SBO and LOOP, with the exception of what systems may be available to mitigate the event, which is again captured in the underlying fault tree logic. Additionally, the anticipated transient without scram (ATWS) event trees were combined into a single event tree. After changes were made, sensitivity studies between the two models were performed to ensure the resultant insights and importance were understood. It was found that a Data Update, including updated initiating event frequencies (for example, Finding and Observation (F&O) DA-D1), had a more pronounced effect on the results than any other changes made. Therefore, it is FENOC's determination that none of these changes constituted a "new methodology or significant changes in scope or capability."

Attachment L-15-235 Page 5 of 47 The PRA model documentation was comprehensively revised and updated following the 2008 gap assessment to address F&O's. The original documentation was found to be inconsistent in format and content and not sufficient to meet the PRA standard requirements. The implementation of new fleet processes for PRA model maintenance, update, and configuration control to address other F&O's caused the PRA documentation to be reclassified in the form of PRA notebooks, with new records retention requirements, naming conventions, and new notebook format requirements. The original WinNUPRA model documentation was reviewed to ensure no significant information was omitted in the new model notebooks. The changes were reviewed by a gap assessment team member to ensure the F&O's were addressed appropriately. Therefore, none of these changes constituted a "new methodology or significant changes in scope or capability."

Following receipt of the request for additional information (RAI) discussed here, a review of model changes made following the 2008 Gap Assessment has been performed and is provided as Table 1, with a disposition as to whether each identified change was regarded as a "PRA update" or a "PRA upgrade." One item was determined to be a PRA model upgrade: offsite power recovery modeling. This was judged to be a PRA Upgrade based on Example 13 of Appendix 1-A of the ASME/ANS PRA Standard due to a change in methodology from a point value estimation to the convolution method. This was considered to be adequately covered under the Section 2-2.8, "LERF Analysis" (LE) focused scope peer review.

As a contributor to LOOP and SBO sequences, impacts from offsite power recovery would be important to the Level 2 model results and thus concerns or discrepancies in the application of the method would be identified during the peer review, via SRs LE-C8, which refers back to Section 2-2.2, "Accident Sequence Analysis (AS)," and through the results review under SRs LE-E4 and LE-F. However, upon further review of the LE peer review documentation, a review of offsite power recovery is not explicitly documented. Furthermore, some aspects of the offsite power recovery may not have been thoroughly reviewed during the LE peer review. Therefore, FENOC conducted a focused scope peer review of the SRs pertinent to offsite power recovery. These SRs include DA-A1, DA-C1, DA-C16, AS-B7, QU-A1, QU-A2, QU-A5, and QU-F2. This peer review was performed July 24 - July 31, 2015, and found that the methodology was traceable and applied in a consistent manner, yielding reasonable results. No findings on the application of the methodology were identified. Some suggestions/observations to improve model documentation were provided. The F&O's from the peer review are included in the RAI 2 response Attachment. The suggested documentation enhancements will be incorporated into the model documentation during the next model update, currently scheduled for 2018, as the enhancements have no impact on model quantification or implementation.

The methodology employed for the Internal Events analysis was also consistently applied for the Level 2, and Internal Flooding, PRA models, which were re-assessed under focused scope peer reviews. The Level 2 and Internal Flooding models are

Attachment L-15-235 Page 6 of 47 linked to the underlying Level 1 model and are not stand-alone. Any issues in the underlying model would propagate into the Level 2 and Internal Flooding (IF) models and thus be identified during the focus scope peer reviews. Additionally, various SRs in the Section 2-2.8, "LERF Analysis (LE)" and Section 3-2, "Internal Flood Analysis (IF)" requirements point back to other SRs in the internal events portion of the ASME/ANS PRA Standard, including Section 2-2.2, "Accident Sequence Analysis (AS)," Section 2-2.6, "Data Analysis (DA)," Section 2-2.4, "Systems Analysis (SY)," and Section 2-2.7, "Quantification (QU)." If there were challenges in the underlying logic modelthey would have been identified through these SRs, therefore it is concluded that the above methodology changes were appropriately assessed during the focused scope peer reviews. To demonstrate this, examples in Table 2 identify those SRs from LE and IF that reference back to other SRs.

In conclusion, while a number of changes were made to the model during this update process, only one item was determined to be a PRA model upgrade: offsite power recovery modeling, and a focused scope peer review was pursued for this item. Other than the offsite power recovery modeling, these changes did not appreciably impact the model results (Section B, page 33 of the ASME/ANS PRA Standard) and therefore, it was not necessary to perform an additional peer review on these items per Appendix 1-A of the ASME/ANS PRA Standard.

In response to the specific items raised:

In general the Support System Initiating Event (SSIE) fault trees (FTs) should mirror the post-initiator fault trees, with the notable exception of the exposure time. However, the previous model SSIE FTs did not capture all failure modes, and only included common cause failures (CCF) in a limited way. The SSIE fault trees were revised to include all the failure modes modeled in the post-initiator fault trees, including comprehensive common cause failure modeling. Although F&O IE-A4a states that the "the initiating event fault trees do not contain common cause failure events," in reality, CCF terms were included in the fault trees in a limited manner, for example, the basic event IACMCC modeled the CCF of the four instrument and service air compressors. Also, as stated in F&O SY-B4, "Many CCF events are modeled as single events and not expanded to capture multiple failure combinations." This comment was applicable to both the support system fault trees and support system initiating event fault trees. Therefore, as part of the disposition of both the IE-A4a and SY-B4 F&Os, the system modeling was reviewed from a common cause perspective and CCF terms were added or removed per EPRI 1016741 and the superseded EPRI EPRI TR-1013490, "Support System Initiating Events: Identification and Quantification Guideline," as directed in the F&O recommendation. Example 25 of Appendix 1-A of the ASME/ANS PRA Standard states that adding additional CCF terms using existing methodology is considered PRA maintenance. Therefore it is FENOC's determination that this was simply a correction to ensure the methodology was applied consistently within the existing initiating event fault trees, and is therefore

Attachment L-15-235 Page 7 of 47 not deemed an upgrade. Sensitivity studies were performed to demonstrate there were no appreciable changes and that data update rather than structural changes had the largest impacts to the results.

a) As noted in RAI 2b, the Alpha Factor method and Multiple Greek Letter method are viewed to be similar and this was considered part of the data update and thus PRA maintenance. The two parameter estimations for deriving common cause factors are similar enough that this would not constitute a PRA upgrade, based on the fact that the Alpha Factor and Multiple Greek Letter methods are directly related through simple mathematical equations as described in Appendix A of NUREG/CR-5485, "Guidelines on Modeling Common-Cause Failures in Probabilistic Risk Assessment," (Equation A-25 and Table A-2). The methodology used to apply either parameter estimation in a PRA model is the same. Other changes in CCF were implemented due to other modeling requirements as discussed previously. Common Cause modeling improvements were done to enhance the model flexibility and match current industry practices but did not change the results appreciably. Therefore this did not constitute a "new methodology or significant changes in scope or capability that impacted significant accident sequences or accident progression sequences."

Sensitivity studies were performed to demonstrate there were no appreciable changes and that data updates rather than structural changes had the largest impacts to the results.

b) Human Reliability Analysis (HRA) was fully updated (completely redone) using the same methodology as previously utilized (EPRI HRA Calculator) and modeled in the PRA. The purpose of updating the HRA analysis was to ensure consistency between Human Error Probabilities (HEPs) and to subsequently enhance the documentation. Therefore, the update to the HRA analysis is considered a data update. New HEPs were added as necessary to add further resolution and detail to the model (PRA maintenance per Example 20 of Appendix 1-A of the ASME/ANS PRA Standard). However, the process and methodology of identifying and assessing HEPs using the HRA Calculator software did not change. Improvements were also made to the individual HEPs to better track timing for the purposes of dependency analysis, this was to ensure the software would correctly identify the sequence of HEPs during the dependency analysis.

The HRA dependency analysis was updated in an effort to capture as many dependencies as possible. This involved taking advantage of new software capabilities in the HRA Calculator to more robustly apply the existing dependency methodology. This was in direct response to F&O HR-G7 indicating that while some treatment of dependency analysis was performed, a more thorough and detailed approach was necessary. Therefore this update was performed in direct response to an F&O. The methodology

Attachment L-15-235 Page 8 of 47 employed for the HRA analysis, including the dependency analysis, was also consistently applied for the Level 2 and Internal Flooding models. While this was a PRA update, it had the potential to have a more significant impact on the model results and was subsequently reviewed. For example, during the Level 2 peer review, a Finding was received (LE-C7-01) for not fully considering the dependencies between Level 1 and Level 2 events.

However, no methodology issues were noted. As another example, the peer review team stated for IFQU-A5: "the dependency analyses for both Level 1 and LERF analyses are performed using the same methodologies as in the internal events models." Table 2 gives further examples and identifies those SRs from LE and IF that require the reviewer to go back to previous sections of the ASME/ANS Standard, including the SRs from HR. No issues with the human reliability analysis were noted in the Internal Flooding Peer Review, including the additional HEPs added to address failures to isolate flood sources and the methodology itself was found to be acceptable. Therefore the enhancements that were made to the dependency analysis were reviewed under the two focused scope peer reviews in sufficient detail to verify that it was performed correctly.

Attachment L-15-235 Page 9 of 47 Table 1: Disposition of Model Changes Update or SR Item Description of Change Justification Upgrade?

Some Event Trees were revised to take credit for an alternate means of high The overall process, including event tree pressure injection: Standby Liquid modeling, fault tree modeling, and success Control (SLC) and Control Rod Drive criteria analysis, remains unchanged, and no (CRD) injection. This revision was new methodologies were introduced. Example prompted by Operator Interviews which 10 of Appendix 1-A of the ASME/ANS PRA Crediting revealed that in the event Feedwater Standard identifies such a change as an update.

AS-A3 Injection Update (FDW), Reactor Core Isolation Cooling This change was prompted when addressing System (RCIC), and High Pressure Core Spray F&O HR-E3 and to meet SRs SC-A6 and HR-E3 (HPCS) were unavailable, operations to interview operations and training personnel to would attempt to maintain level with CRD confirm the success criteria modeled is and SLC injection before depressurizing consistent with plant operations and operating the vessel and utilizing low pressure philosophy.

injection sources.

The methodology for developing the Event Trees was unchanged. Event Trees were combined to avoid unnecessary duplication of logic. Example 13 of Appendix 1-A describes a The Station Blackout Event Tree was scenario in which the Station Blackout modeling subsumed into the Loss of Offsite Power is revised and the Loss of Offsite Power Event Event Tree. The Loss of Feedwater and Tree is incorporated into the Transient Event Inadvertent Stuck Open Relief Valve Tree, and considers this change to be an AS-10 Event Trees were subsumed with the General Update upgrade. However, incorporating the Station Transient Event Tree. The Anticipated Blackout Event Tree into the Loss of Offsite Transient Without Scram (ATWS) Event Power Event Tree is straightforward as they Trees were combined into a single Event share similar operator response, mitigating Tree.

systems, and success criteria. Station Blackout is a subset of the Loss of Offsite Power Condition. Operators would be working from the same flowcharts and procedures for a Loss of

Attachment L-15-235 Page 10 of 47 Table 1: Disposition of Model Changes Update or SR Item Description of Change Justification Upgrade?

Offsite Power or Station Blackout, but would be in a different set of flow charts and procedures in a general transient. Incorporating the Station Blackout and Loss of Offsite Power Event Trees into the Transient Event Tree would involve much more complex changes to the model logic.

Incorporating the Station Blackout Event Tree into the Loss of Offsite Power Event Tree is a minor change and thus constitutes a PRA update rather than a PRA upgrade.

WinNUPRA used Fault Trees with point This was judged to be a PRA Upgrade based on estimate basic events to model the Example 13 of Appendix 1-A of the ASME/ANS recovery of offsite power. Due to issues PRA Standard due to change in methodology Offsite power importing this tree and the corresponding from point value estimation to the convolution AS-B7 Upgrade recovery House Events into CAFTA model, Offsite method. This item is discussed in detail in the power recovery was changed to the response to RAI 2.

Convolution Method implemented by Recovery Rules.

The two parameter estimations for deriving common cause factors are directly related through simple mathematical equations as described in NUREG/CR-5485 Appendix A Common Revision from Alpha Factor method to (Equation A-25 and Table A-2). Therefore, this DA-D5 Update Cause Multiple Greek Letter (MGL) method. revision would not constitute a PRA upgrade.

The process methodology used to apply either numerical data set in a PRA model is the same under industry consensus common cause modeling practices.

Attachment L-15-235 Page 11 of 47 Table 1: Disposition of Model Changes Update or SR Item Description of Change Justification Upgrade?

Some dependency analysis was present in the WinNUPRA model. The Gap Assessment Update, but produced an F&O to extend the scope of HRA significant Dependency Analysis to capture the possible enough to combinations. This update was in response to impact the F&O and followed the recommendation HR-D5, Re-performed HRA dependency analysis results. exactly.

HRA HR-G7, with an effort to capture the possible Dependency QU-C2 HEP combinations Methodology However, the dependency analysis can have a reviewed as significant impact to the model results. As such, part of LE it was reviewed during the LE and IF focused and IF peer scope peer reviews. It is FENOC's reviews. interpretation that these peer reviews reviewed the update in sufficient detail to verify it was performed correctly.

The methodology used for both developing IE point estimates and developing support system initiating event fault trees was previously used in During the model update process new the model. No new methodologies were initiating events were identified and introduced by this update. This was also added to the model. For many of these performed to make the model more in-line with IE-A1, initiating events, new fault tree logic was Addition of new industry best practices. New initiating events IE-C1, developed to calculate the IE frequency. Update lEs were identified through the resolution of F&Os IE-C8 New initiating events included a loss of such as IE-A1 directing the performance of a nuclear closed cooling (NCC) and a loss failure modes and effects analysis, and IE-B1 of numerous alternating current (AC) or and IE-C6 to re-consider improperly grouped direct current (DC) buses.

events such as a loss of NCC or a loss of bus events. Example 1 of Appendix A-1 supports the view that this is an update.

Attachment L-15-235 Page 12 of 47 Table 1: Disposition of Model Changes Update or SR Item Description of Change Justification Upgrade?

The loss of feedwater support system Numerous other initiating events utilized the initiating event was modeled using a point estimate method, and thus no new Developed developed fault tree for the WinNUPRA methodology was introduced by the change.

Fault Trees for model. This was changed to a point IE-C1, Support estimate for the CAFTA model, based on Update This change was also identified during IE-C8 System a recommendation from industry peers benchmarking and comparison with other plant Initiating Events that a developed fault tree for a loss of IE frequencies and results, as required by the feedwater tends to over-estimate the ASME/ANS PRA Standard. It was desirable to actual frequency. be more in line with current industry practices.

WinNUPRA quantified specific fault trees (DAM) to obtain cutset files containing Overall process of developing fault trees to Mutually the mutually exclusive logic. CAFTA QU-B8 Update model the mutually exclusive logic remained the Exclusive Logic utilizes NOT logic to apply the mutually same.

exclusive logic contained in the DAM fault trees.

Attachment L-15-235 Page 13 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s)

The methodology for modeling systems only credited in Level 2, such as Hydrogen Igniters and Containment Isolation, is Met CC l/ll/lll with Suggestion (LE-C6-01) identical to that for the Level System models affecting the accident progression 1 system modeling.

(such as hydrogen igniters and containment sprays) Systems significant in Level LE-C6 2-2.4 SRs in SY are developed in the system notebooks consistent 2 modeling, such as with the applicable requirements of the ASME Residual Heat Removal Standard section 2-2.4. (RHR), in the Containment Spray mode, were reviewed as part of the focused scope peer review. The F&O's were addressed as recommended.

Not Met with Finding (LE-C7-01) The methodology for The LE analysis credits operator actions that are developing Level 1-specific contained with the procedural guidance, including the HEPs was identical to that Emergency Procedure Guidelines (EPGs). The used for developing the actions developed for the Perry Level 2 were Level 2-specific HEPs.

evaluated using the HRA calculator, with the Furthermore, the information presented in Appendix E of the L2 methodology for performing LE-C7 2-2.5 SRs in HR notebook. The analyses included an appropriate level the Dependency Analysis of detail in terms of examining procedural guidance, was also the same.

training and time windows. However, the HEP However, a Finding was dependency analysis only considered dependencies received as the Level 2 between the Level 2 HEPs and not between level 1 dependency did not account and Level 2 events. The justification for this is not for Level 1 to Level 2 documented in the analysis, but the PRA staff dependencies. The F&O's

Attachment L-15-235 Page 14 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s) indicated that since the Level 2 actions are initiated by were addressed as EPGs by emergency response personnel, there recommended.

should not be a cognitive dependency. If true, then there is still the potential for dependency between the events if they would occur in a similar time window as the Level 1 HEP(s), if there are not adequate resources, or if there is a common manipulation error.

The combinations of significant Level 1 and Level 2 HEPs should be examined for such dependency.

Attachment L-15-235 Page 15 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s)

Met CC l/ll/lll with Suggestion (LE-C8-01)

Inter-system dependencies are captured throughout the accident sequence because of the linked Level 1/Level 2 approach developed in the CAFTA model.

Functional dependencies, such as failure to depressurize after core damage, given early success or failure are generally accounted for by use of the same basic events pre and post core damage.

The methodology for However, the approach of having a small number of modeling accident sequence Plant Damage States (PDSs) and questioning system dependencies is identical logic in the Containment Event Trees (CETs) has the LE-C8 2-2.2 SRs in AS from the Level 1 model to potential to lose some functional information about the the Level 2 model and the core damage sequence and must be performed F&O's were addressed as carefully. For example, the Level 2 CET event that recommended.

questions injection uses gate LATEINJ that is low pressure Emergency Core Cooling System (ECCS)

(level 1 logic) and Level 2 HEP CPHISAG1-INJLATE.

For Level 1 sequences with successful injection, this should not be considered a failure. It is acceptable to have the conservative modeling if it is documented and noted to be insignificant, but this was not noted in the L2 notebook.

Met CC l/ll/lll 2-2.5 SRs in HR Equipment survivability is addressed in Appendix F. Methodologies used to HRA probabilities are calculated consistent with that determine parameter for Level 1 with consideration for conditions present. estimates for equipment and LE-E1 The fault tree models developed for the Level 2 operator actions was 2-2.6 SRs in DA analysis utilize the same data and HRA methods as identical for the Level 1 and are used in the Level 1 system models, in accordance Level 2 models.

with the DA and HR assessments.

Attachment L-15-235 Page 16 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s)

Between LE-E4 and LE-F3, Met CC l/ll/lll with 1 Finding (LE-C7-01) and 1 2-2.7-2(a) QU-A QU SRs except F Suggestion (LE-E4-01)

(Documentation) are Should be The LERF quantification process uses the same covered. The methodology 2-2.7-2(b) 2-2.7.3(b) - process as the CDF quantification process, meeting to quantify the model, QU-B the QU requirements. The only requirement from including determining Tables 2-2.7-4(a), (b) and (c) that is not met is QU-C2 LE-E4 importance rankings, (assessing the degree of dependency between HFEs) performing truncation for the Level 1/Level 2 HEP combinations. The HFE Should be analysis, and uncertainty dependency issue refers back to F&O LE-C7-01. In 2-2.7-2(c) 2-2.7.4(c) - analysis, is the same for the addition, the documentation for QU-B3 (establishing QU-C Level 1 and Level 2 models.

acceptable truncation limits) should be clarified.

The F&O's were addressed Otherwise, this SR is met.

as recommended.

Met CC l/ll/lll with Suggestion (LE-F3-01) Between LE-E4 and LE-F3, Of the items listed in Table 2-2.7-2(d) and (e) of the QU SRs except F Standard, the following are noted for LE: Reviews of (Documentation) are the results were performed in Sections 3.3.1 and covered. The methodology Should be 2- 3.3.2. A comparison of the results to Clinton was to quantify the model, LE-F3 2-2.7-2(d) 2.7.5(d) - QU-D presented in Section 3.6.4. Significant contributors to including determining LERF are identified in Section 3.3.3. Sources of importance rankings, modeling uncertainty and generic modeling performing truncation uncertainty are considered in Table H-49, although the analysis, and uncertainty disposition of the treatment of ex-vessel cooling and analysis, is the same for the

Attachment L-15-235 Page 17 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s) hydrogen combustion are not clear as to why they are Level 1 and Level 2 models.

not a source of uncertainty at Perry. Table H-50 The F&O's were addressed presents a list of plant-specific assumptions that could as recommended.

affect LERF. Section 4.3 compiles the list of modeling uncertainties that could have a noticeable impact on LERF. QU-E4 calls for identifying how the PRA model is affected which has been qualitatively addressed (see also HLR-QU-E which states the potential impact Should be of the modeling uncertainties and assumptions can 2-2.7-2(e) 2-2.7.6(e) -

have on the LERF results is to be understood. This QU-E SR is considered to be met because the bulk of the uncertainty characterization has been performed, but a suggestion F&O is presented to further evaluate the impact of these on the LERF results quantitatively.

The following are also noted: In Table H-49, #20 is missing information. #21 discusses MCCI but not debris contact with containment.

The methodology for Met CCI/II with Finding (1-7) grouping and subsuming SRs in IE; Basis: Based on a review of the information presented Internal Flooding events was grouping and in Appendices E and G, it does not appear that any IFEV-A3 2-2.1 similar to that for Internal subsuming of initiating event scenarios have been improperly Events. The F&O's were lEs grouped with existing internal event groups.

addressed as Therefore, this SR is considered met.

recommended.

Attachment L-15-235 Page 18 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s)

Internal Flooding utilized a few of the Internal Events Event Trees, namely General Transient (T3A),

Loss of Condenser (T2),

Loss of Instrument Air (TIA),

Met CC I/I I/I 11 with Suggestion 2-5 and Loss of Service Water or Basis: In most cases, internal flooding accident NCC (TSW). As the Internal sequences were modeled using the general transient Flooding analysis including event tree. Flooding events involving service water consequential events such pipe breaks were modeled using the loss of service as ATWS and LOOP, those IFQU-A1 2-2.2 SRs in AS water initiating event. This treatment was determined Event Trees were also to be the most conservative treatment of internal utilized for the Internal flooding accident sequences. While the association of Flooding analysis. The accident sequence appears to be appropriate, the methodology used to basis for the association is documented sparsely in develop these Event Trees Section 8.1.4.

was the same as that for the remaining Event Trees (primarily LOCAs) from the Internal Events analysis.

The F&O's were addressed as recommended.

Attachment L-15-235 Page 19 of 47 Table 2: SRs from the Focused Scope Peer Review that Refer Back to Other SRs in Detail Description of Referenced SR Referenced Focus Scope Peer Review Results PY notes SRs SR(s)

Met CC I/I I/I 11 Basis: The probability value for each new human The methodology to develop failure event (HFE) defined for the internal flooding the new HEPs is the same scenarios is calculated using the same methodology as that used for the Internal as in the internal events PRA. The values and Events. Similarly, the IFQU-A5 2-2.5 SRs in HR analyses reviewed appear reasonable based on the methodology to perform the documentation in Appendix K. Similarly, the dependency analysis was dependency analyses for both level 1 and LERF the same as that used for analyses are performed using the same the Internal Events.

methodologies as in the internal events models.

Therefore, this SR is considered met.

Attachment L-15-235 Page 20 of 47 RAI3 In the application dated March 25, 2014, as supplemented by letter dated October 7, 2014, FENOC indicated that portions of the PNPP internal events PRA model have been assessed against ASME RA-Sb-2005, "Addenda to ASME RA-S-2002 Standard for Probabilistic Risk Assessment for Nuclear Power Plant Applications," with the clarifications and qualifications in Regulatory Guide (RG) 1.200, Revision 1 (ADAMS Accession No. ML070240001). According to NRC Regulatory Issue Summary 2007-06, "Regulatory Guide 1.200 Implementation" (ADAMS Accession No. ML070650428), the NRC staff expects licensees to fully address all scope elements with Revision 2 of RG 1.200 (ADAMS Accession No. ML090410014) by the end of its implementation period (i.e., one year after the issuance of Revision 2). Regulatory Guide 1.200, Revision 2, endorses, with clarifications and qualifications, the use of the combined ASME/ANS PRA Standard, ASME/ANS RA-Sa-2009, "Addenda to ASME/ANS RA-S-2008 Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications."

Identify and address any gaps between the PNPP PRA model and ASME/ANS RA-Sa-2009, including the clarifications and qualifications in RG 1.200, Revision 2, that are relevant to this application, or explain why addressing the gaps would have no impact on this application. (Portions of the PRA model that have been peer reviewed to ASME/ANS RA-Sa-2009 and RG 1.200, Revision 2, would not require a gap assessment.)

Response

The following table (Table 3), presents the Gap Assessment between the ASME/ANS RA-Sa-2009, "Addenda to ASME/ANS RA-S-2008 Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications," from the previous assessment performed to the ASME/ANS RA-Sb-2005, "Addenda to ASME RA-S-2002 Standard for Probabilistic Risk Assessment for Nuclear Power Plant Applications." The table provides a comparison between the portions of ASME/ANS RA-Sb-2005 that Perry is assessed against to the related portion in ASME/ANS RA-Sa-2009. The comparison between ASME/ANS RA-Sb-2005 and ASME/ANS RA-Sa-2009 resulted in only minor changes made to the internal events portion of the standard. No gaps were identified between ASME/ANS RA-Sb-2005 and ASME/ANS RA-Sa-2009. The Focused Scope Peer Reviews on Internal Flood (IF) and Large Early Release Frequency (LE) were peer reviewed to ASME/ANS RA-Sa-2009; therefore a comparison is not provided in Table 3.

Attachment L-15-235 Page 21 of 47 Table 3 Keywords Keyword in "Change From 05 to 09" Column Description of Keyword Exact No change, not even to wording Same intent with slightly different wording (for example. Writing out Same acronyms, rearranging a sentence, and others)

Change Small change Different Big change N/A SR deleted New New to standard Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments Section 2 (internal events)

HLR-IE-A HLR-IE-A exact HLR-IE-B HLR-IE-B exact HLR-IE-C HLR-IE-C exact Change in the order of the wording, no change HLR-IE-D HLR-IE-D same wording to intent IE-A1 IE-A1 exact removes internal Meet CC I/I I/I 11 IE-A2 IE-A2 change flooding initiators Internal Flooding assessed in IF IE-A3 IE-A3 exact

Attachment L-15-235 Page 22 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments IE-A3a IE-A4 exact IE-A4 IE-A5 exact IE-A4a IE-A6 exact IE-A5 IE-A7 exact IE-A6 IE-A8 exact IE-A7 IE-A9 exact IE-A8-deleted n/a n/a SR deleted in ASME/ANS RA-Sb-2005 IE-A9-deleted n/a n/a SR deleted in ASME/ANS RA-Sb-2005 IE-A10 IE-A10 exact changed reference from Paragraph 4.5.2 to 2-2.2 and from text of the referenced Sections is essentially IE-B1 IE-B1 same 4.5.8 to 2-2.7 the same IE-B2 IE-B2 exact Capability Category 2 wording change from "AVOID subsuming event into a group unless ..." to "DO Intent of CC II unchanged. Initiating events are NOT SUBSUME not subsumed unless the impacts are scenarios into a comparable or less than those of the remaining IE-B3 IE-B3 same group unless..." events in that group IE-B4 IE-B4 exact

Attachment L-15-235 Page 23 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments IE-B5 IE-B5 exact IE-C1 IE-C1 exact IE-C1a IE-C2 exact reference to other supporting requirements changed in text of the referenced SRs is essentially the IE-C1b IE-C3 same numbers same No change to intent of SR, prior distribution Note 2 changed data obtained from NRC 2010 Parameter IE-C2 IE-C4 same to Reference 2-2 Estimation Update IE-C3 IE-C5 exact reference to Paragraphs 4.5.6 changed to Section 2-2.7, reference to Paragraph 4.5.8 changed to text of the referenced Sections is essentially IE-C4 IE-C6 same Section 2-2.7 the same added an acceptable method (NUREG/CR-IE-C5 IE-C7 same 6928) no change to intent of SR

Attachment L-15-235 Page 24 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments changed reference from Paragraph 4.5.4 to Section 2-2.4, and Paragraph 4.5.6 to Section text of the referenced Sections is essentially IE-C6 IE-C8 same 2-2.6 the same changed reference from Paragraph 4.5.4 text of the referenced Sections is essentially IE-C7 IE-C9 same to Section 2-2.4 the same IE-C8 IE-C10 exact changed reference from Paragraph 4.5.5 text of the referenced Sections is essentially IE-C9 IE-C11 same to Section 2-2.5 the same IE-C10 IE-C12 exact

Attachment L-15-235 Page 25 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments changed "Use generic data and INCLUDE plant-specific functions" to "Use generic IE frequencies, including LOCA frequencies, data and based on generic-BWR numbers from NRC INCLUDE plant- Parameter Estimate 2010. Interfacing Systems specific features LOCA and Break Outside Containment LOCAs to decide which were calculated based on plant-specific generic data are information. IE frequencies were compared IE-C11 IE-C13 same most applicable" with other BWR/6's to verify reasonableness.

IE-C12 IE-C14 exact IE-C13 IE-C15 exact IE-D1 IE-D1 exact IE-D2 IE-D2 exact added reference to QU-E1 & QU- Added reference to SRs on uncertainty and IE-D3 IE-D3 same E2 assumptions, no change to intent HLR-AS-A HLR-AS-A exact HLR-AS-B HLR-AS-B exact Change in the order of the wording, no change HLR-AS-C HLR-AS-C same wording to intent AS-A1 AS-A1 exact AS-A2 AS-A2 exact

Attachment L-15-235 Page 26 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments reference to another SR changed from SC-A4 to SC-A3 (this is the same text of the referenced SR is essentially the AS-A3 AS-A3 same SR) same reference to another SR changed from SC-A4 to SC-A3 (this is the same text of the referenced SR is essentially the AS-A4 AS-A4 same SR) same AS-A5 AS-A5 exact AS-A6 AS-A6 exact AS-A7 AS-A7 exact AS-A8 AS-A8 exact AS-A9 AS-A9 exact AS-A10 AS-A10 exact AS-A11 AS-A11 exact AS-B1 AS-B1 exact AS-B2 AS-B2 exact AS-B3 AS-B3 exact AS-B4 AS-B4 exact AS-B5 AS-B5 exact AS-B5a AS-B6 exact AS-B6 AS-B7 exact

Attachment L-15-235 Page 27 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments AS-C1 AS-C1 exact AS-C2 AS-C2 exact added reference to QU-E1 & QU- Added reference to SRs on uncertainty and AS-C3 AS-C3 same E2 assumptions, no change to intent HLR-SC-A HLR-SC-A exact HLR-SC-B HLR-SC-B exact Change in the order of the wording, no change HLR-SC-C HLR-SC-C same wording to intent SC-A1 SC-A1 exact SC-A2 SC-A2 exact SC-A3-deleted SR deleted in ASME/ANS RA-Sb-2005 SC-A4 SC-A3 exact SC-A4a SC-A4 exact SC-A5 SC-A5 exact SC-A6 SC-A6 exact SC-B1 SC-B1 exact SC-B2 SC-B2 exact SC-B3 SC-B3 exact SC-B4 SC-B4 exact SC-B5 SC-B5 exact SC-C1 SC-C1 exact SC-C2 SC-C2 exact

Attachment L-15-235 Page 28 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments added reference to QU-E1 & QU- Added reference to SRs on uncertainty and SC-C3 SC-C3 same E2 assumptions, no change to intent HLR-SY-A HLR-SY-A exact HLR-SY-B HLR-SY-B exact Change in the order of the wording, no change HLR-SY-C HLR-SY-C same wording to intent SY-A1 SY-A1 exact SY-A2 SY-A2 exact SY-A3 SY-A3 exact added This SR requires interviews with knowledgeable knowledgeable plant personnel. Interviews (before plant were performed with the responsible System SY-A4 SY-A4 same personnel) Engineer for each system.

SY-A5 SY-A5 exact SY-A6 SY-A6 exact SY-A7 SY-A7 exact SY-A8 SY-A8 exact SY-A9-deleted SR deleted in ASME/ANS RA-Sb-2005 SY-A10 SY-A9 exact SY-A11 SY-A10 exact reference to another supporting SY-A12 SY-A11 same requirement text of the referenced SR is the same

Attachment L-15-235 Page 29 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments changed SY-A14 changed to SY-A15 SY-A12a SY-A12 exact SY-A12b SY-A13 exact reference to another supporting requirement changed SY-A12 changed to SY-SY-A13 SY-A14 same A11 text of the referenced SR is the same Made supporting requirement applicable to all No change from previously incorporated SY-A14 SY-A15 same three categories corrections.

references paragraph 4.5.5 SY-A15 SY-A16 same changed to 2-2.5 references overall HR Section references paragraph 4.5.2 SY-A16 SY-A17 same changed to 2-2.2 references overall HR Section SY-A20 changed SY-A17 SY-A18 same to SY-A22 text of the referenced SR is the same SY-A18 SY-A19 same wording no affect

Attachment L-15-235 Page 30 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments DA-C13 changed SY-A18a SY-A20 same to DA-C14 text of the referenced SR is the same SY-A19 SY-A21 exact SY-A20 SY-A22 exact SY-A21 SY-A23 exact DA-C14 changed SY-A22 SY-A24 same to DA-C15 text of the referenced SR is the same note (1) changed SY-B1 SY-B1 same to reference [2-4] Referenced NUREG is the same SY-B2 SY-B2 exact SY-B3 SY-B3 exact SY-B4 SY-B4 exact SY-B5 SY-B5 exact SY-B6 SY-B6 exact SY-B7 SY-B7 exact SY-B8 SY-B8 exact SY-B9-deleted SR deleted in ASME/ANS RA-Sb-2005 SY-B10 SY-B9 exact SY-B11 SY-B10 exact SY-B12 SY-B11 exact SY-B13 SY-B12 exact SY-B14 SY-B13 exact SY-B15 SY-B14 exact SY-B16 SY-B15 exact

Attachment L-15-235 Page 31 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments SY-C1 SY-C1 exact SY-C2 SY-C2 exact wording, added reference to QU- Added reference to SRs on uncertainty and SY-C3 SY-C3 same E1 and QU-E2 assumptions, no change to intent HLR-HR-A HLR-HR-A exact HLR-HR-B HLR-HR-B exact HLR-HR-C HLR-HR-C exact HLR-HR-D HLR-HR-D exact HLR-HR-E HLR-HR-E exact HLR-HR-F HLR-HR-F exact HLR-HR-G HLR-HR-G exact HLR-HR-H HLR-HR-H exact Change in the order of the wording, no change HLR-HR-I HLR-HR-I same wording to intent HR-A1 HR-A1 exact HR-A2 HR-A2 exact HR-A3 HR-A3 exact HR-B1 HR-B1 exact HR-B2 HR-B2 exact HR-C1 HR-C1 exact HR-C2 HR-C2 exact HR-C3 HR-C3 exact

Attachment L-15-235 Page 32 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments note 1 and note 2 changed to references 2-5 HR-D1 HR-D1 same and 2-6 Referenced NUREGs are the same HR-D2 HR-D2 exact HR-D3 HR-D3 exact HR-D4 HR-D4 exact HR-D5 HR-D5 exact HR-D6 HR-D6 exact HR-D7 HR-D7 exact HR-E1 HR-E1 exact HR-E2 HR-E2 exact HR-E3 HR-E3 exact HR-E4 HR-E4 exact HR-F1 HR-F1 exact HR-F2 HR-F2 exact HR-G1 HR-G1 exact HR-G2 HR-G2 exact HR-G3 HR-G3 exact HR-G4 HR-G4 exact HR-G5 HR-G5 exact HR-G6 HR-G6 exact HR-G7 HR-G7 exact HR-G8-deleted SR deleted in from ASME/ANS RA-Sb-2005

Attachment L-15-235 Page 33 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments HR-G9 HR-G8 exact HR-H1 HR-H1 exact wording, added "if the following The intent and the criteria in this SR is the HR-H2 HR-H2 same occur" same HR-H3 HR-H3 exact HR-11 HR-11 exact HR-I2 HR-I2 exact wording, added reference to QU- Added reference to SRs on uncertainty and HR-I3 HR-I3 same E1 and QU-E2 assumptions, no change to intent HLR-DA-A HLR-DA-A exact HLR-DA-B HLR-DA-B exact Change in the order of the wording, no change HLR-DA-C HLR-DA-C same wording to intent HLR-DA-D HLR-DA-D exact Change in the order of the wording, no change HLR-DA-E HLR-DA-E same wording to intent DA-A1 DA-A1 exact wording, changed references to text of the referenced SRs is essentially the DA-A1a DA-A2 same other SRs same DA-A2 DA-A3 exact DA-A3 DA-A4 exact DA-B1 DA-B1 exact

Attachment L-15-235 Page 34 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments DA-B2 DA-B2 exact DA-A3 is now DA-A4, moved references from Notes to References list, added NUREG-1715 and NUREG/CR-6928 to part (a), added line to say "see NUREG/CR-6823

[2-1] for a listing text of the referenced SRs is essentially the of additional data same, NUREG/CR-6928 is the primary source DA-C1 DA-C1 same sources" for generic probability failures in the PY PRA DA-A2 and DA-A3 change to DA- text of the referenced SRs is essentially the DA-C2 DA-C2 same A3 and DA-A4 same DA-C3 DA-C3 exact DA-C4 DA-C4 exact DA-C5 DA-C5 exact DA-C6 DA-C6 exact DA-C7 DA-C7 exact DA-C8 DA-C8 exact DA-C9 DA-C9 exact DA-C10 DA-C10 exact DA-C11 DA-C11 exact

Attachment L-15-235 Page 35 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments DA-C11a DA-C12 exact This SR states that interviews should be conducted in the case that "reliable estimates or the start and finish times are not available."

For PY reliable start and finish times for the added significant maintenance activities are available knowledgeable through the history of the Online Risk Monitor (before plant and the Plant Narrative Logs. Therefore no DA-C12 DA-C13 same personnel) specific interviews have been conducted.

wording, added phrase "that is a result of a planned, repetitive activity",

added example of intersystem DA-C13 DA-C14 same unavailability intent of SR unchanged DA-C14 DA-C15 exact DA-C15 DA-C16 exact DA-D1 DA-D1 exact

Attachment L-15-235 Page 36 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments DA-D2 DA-D2 exact DA-D3 DA-D3 exact DA-D4 DA-D4 exact DA-D5 DA-D5 exact wording, forCC II added the phrase "in a manner" prior to "consistent with the component DA-D6 DA-D6 same boundaries" intent of SR unchanged DA-D6a DA-D7 exact DA-D7 DA-D8 exact DA-E1 DA-E1 exact DA-E2 DA-E2 exact added reference to QU-E1 and Added reference to SRs on uncertainty and DA-E3 DA-E3 same QU-E2 assumptions, no change to intent HLR-QU-A HLR-QU-A exact HLR-QU-B HLR-QU-B exact HLR-QU-C HLR-QU-C exact LERF added to the model following 2008 Self-Assessment and assessed via focused scope HLR-QU-D HLR-QU-D change added LERF peer review

Attachment L-15-235 Page 37 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments HLR-QU-E HLR-QU-E exact Change in the order of the wording, no change HLR-QU-F HLR-QU-F same wording to intent QU-A1 QU-A1 exact QU-A2a QU-A2 exact QU-A2b QU-A3 exact QU-A3 QU-A4 exact QU-A4 QU-A5 exact QU-B1 QU-B1 exact QU-B2 QU-B2 exact QU-B3 QU-B3 exact QU-B4 QU-B4 exact AVOID changed to DO NOT; Circular logic broken using the guidance from reference NUREG/CR-2728, with care to ensure no changed from unnecessary conservatisms or Note to nonconservatisms were introduced into the QU-B5 QU-B5 change References model logic QU-B6 QU-B6 exact QU-B7a QU-B7 exact QU-B7b QU-B8 exact QU-B8 QU-B9 exact QU-B9 QU-B10 exact

Attachment L-15-235 Page 38 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments QU-C1 QU-C1 exact QU-C2 QU-C2 exact QU-C3 QU-C3 exact QU-D1a QU-D1 exact QU-D1b QU-D2 exact QU-D1C QU-D3 exact QU-D2-deleted SR deleted in from ASME/ANS RA-Sb-2005 QU-D3 QU-D4 exact QU-D4 QU-D5 exact QU-D5a QU-D6 exact QU-D5b QU-D7 exact Efforts made to capture the sources of removed "key" uncertainty, and not just key sources, utilizing QU-E1 QU-E1 ' same from key sources NUREG 1855 and EPRI 1016737.

removed "key" The assumptions are documented in the QU-E2 QU-E2 same from key sources respective notebooks HR-G9 is now HR-G8, IE-C13 is text of the referenced SRs is essentially the QU-E3 QU-E3 same nowlE-C15 same

Attachment L-15-235 Page 39 of 47 Table 3 : Gap Assessment from ASME/ANS RA-Sb-2005 to ASME/ANS RA-Sa-2009 ASME/ANS RA-Sb- ASME/ANS RA-2005 Sa-2009 Change (Reg Guide 1.200 (Reg Guide 1.200 from Description of Rev1) Rev 2) 05 to 09 Change PY Comments changed from 3 The PY Quantification Notebook identifies distinct capability assumptions and sources of uncertainty and categories to a assesses how they impact the overall model QU-E4 QU-E4 different single CC l/ll/lll and results QU-F1 QU-F1 exact QU-F2 QU-F2 exact QU-F3 QU-F3 exact The intent to document the model assumptions wording (took out and sources of uncertainty is unchanged. The the examples PY documentation includes discussions on QU-F4 QU-F4 same "such as..) these items.

QU-F5 QU-F5 exact QU-F6 QU-F6 exact

Attachment L-15-235 Page 40 of 47 RAI 2 Response Attachment Offsite power recovery Focused Scope Peer Review Findings and Observations The offsite power recovery focused scope peer review did not result in any findings, but did identify a number of suggestions. Those suggestions were items to correct or enhance the documentation, and are not expected to have any impact on the results of the offsite power recovery analysis and implementation. For the sake of completeness, they are listed below along with the proposed edits to the documentation. These edits will be incorporated into the next model update, to the extent practical as other revisions to the model may make some items obsolete. These resolutions were reviewed by the peer review team and found to be acceptable and fully address the suggestions.

Attachment L-15-235 Page 41 of 47 RAI 2 Response Attachment Suggestion F&Os F&O Number F&O Details In the equation used to calculate the non-recovery factors, Delay is 1

defined as "Sum of delay times," this term should be "Max of delay times."

Associated SR(s) Basis for Significance QU-F2 The "max of delay times" was used in the analysis, as appropriate.

Possible Resolution Correct the documentation to use the correct implementation of the delay times Perry Response The draft documentation for the next model update has been corrected to identify the Delay as the "Max of delay times." Note that this is strictly a documentation error as the calculations were performed using the maximum of the delay times, which the Peer Review team agreed was correct and appropriate.

The term "mts" (mission time start) is defined in the equation for offsite power non-recovery but is not explained how it is defined and will be used in the equations.

Associated SR(s) Basis for Significance QU-F2 This is a suggestion to enhance the documentation.

Possible Resolution Add further discussion to the documentation as indicated.

Perry Response The "mts" (mission time start) is used to identify basic events that are associated with failures to run for over at least 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />, as it is common practice to collect diesel failure data assuming failures within the first hour are demand failures. The Perry PRA model identifies three failure modes for diesels and standby pumps: failure to start, failure to load/run less than 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />, and failure to run for greater than 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. The convolution calculation accounts for this by using a mission time start (mts) of 1.0 for the diesel failures, and a mission start time of 0 for other failures.

The draft documentation for the next model update has been updated to include this discussion.

Attachment L-15-235 Page 42 of 47 RAI 2 Response Attachment Suggestion F&Os (continued)

F&O Number F&O Details It is unclear for the reviewers the percentage of cutsets that contains LOSP 3

events are recovered.

Associated SR(s) Basis for Significance This does not impact the results of the analysis but can provide another QU-F2 means to review the results.

Possible Resolution Add the identified information to the documentation as indicated.

Perry Response The draft documentation for the next model update will be updated to include a review of pertinent cutsets and provide the percentage of cutsets that includes loss of station power (LOSP) events and are recovered. In the current effective model, 23 percent of LOOP cutsets include a recovery term; however, this number is expected to change slightly as other updates are incorporated during the normal model updated process.

4 The non-recovery curve parameters values were taken directly from NUREG/CR-6928 . It does not seem to be correct. It could be NUREG/CR-6890 because NUREG/CR-6928 does not seem to have the non-recovery curve parameters listed.

Associated SR(s) Basis for Significance This does not impact the results of the analysis but the appropriate reference QU-F2 is necessary for the repeatability of the work performed.

Possible Resolution Correct the reference in the documentation.

Perry Response The actual reference is an NRC 2004 online database:

http://nrcoe.inel.gov/resultsdb/publicdocs/LOSP/loop-summary-update-2004.pdf The draft documentation for the next model update has been revised to include the correct reference.

Attachment L-15-235 Page 43 of 47 RAI 2 Response Attachment Suggestion F&Os (continued)

F&O Number F&O Details What's the minimum time required to restore onsite power after offsite power is restored? For some situations, operator actions to restore onsite power following the restoration of offsite power become very high priority actions for Operations.

Associated SR(s) Basis for Significance This is not expected to impact the results of the analysis, but a discussion of QU-F2 these aspects should be included in the documentation.

Possible Resolution Add further discussion to the documentation as indicated.

Perry Response The draft documentation for the next model update has been revised to include the following discussion:

Following recovery of offsite power sources, it is necessary to then restore impacted systems so that they can fulfill their functions to prevent core damage or containment failure. Restoration of these systems becomes a high priority for operations.

Following the Perry 2003 LOOP it took approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 10 minutes for the blackstart source (Eastlake line) to be available, restoring offsite power to the site. It took an additional 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 15 minutes to restore the first emergency bus (EH11) from that offsite source. However, during this event the emergency diesel generators were running, as well as RCIC, so while offsite power recovery was a priority it was not an urgency. In an SBO with RCIC running, offsite power recovery to an emergency bus will either initially allow the dumping of the upper pool (immediate action in flowchart following power recovery) and prolong the use of RCIC (increasing heat capacity temperature limit via cooler water addition to the pool) or in the case of the diesel fire pump, allow actions to be taken to recover ECCS Systems. It is expected that on the order of 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> would be required to fill and vent an ECCS System, should such actions be necessary. Recovery of an initial system would then allow a cascading recovery of the other systems.

Based on a review of NUREG/CR-6980, the generic data represents the probability of not recovering offsite power to a safety bus following the initiation of the LOOP. As such, the analysis then must consider the time required to restore systems once the electrical bus(es) have been re energized. Note that if the system was not initially de-energized, restoration is not required (that is, offsite power was recovered prior to the diesel generator failing to continue to run). Therefore, the time of recovery actions is only important for recoveries involving a delay time following diesel failures.

Attachment L-15-235 Page 44 of 47 RAI 2 Response Attachment Suggestion F&Os (continued)

F&O Number F&O Details A review of the time delay events credited in the offsite power recovery was 5 (continued) performed, and is presented below. Note that this review will be re-performed on the updated model during the next model update, to verify that the justification given here is still true and to identify if any new patterns emerge in the cutsets.

• CPHI0NISPID7-DFPF0 / CPHI-DFPFUELOIL, Operator failures to refill the diesel fire pump's fuel oil tank: Injection was being supplied by the diesel fire pump, until the pump's fuel oil tank was emptied. The diesel fire pump can run for 7.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> on a 1A tank of fuel. The offsite power recovery calculation only credited 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> for this delay event. Therefore there is sufficient margin for offsite power to be recovered and restoration actions on an ECCS system, including fill and vent, to be completed prior to loss of the diesel fire pump.

• CRHIARI-LOSSOFHVAC, Loss of switchgear HVAC: Cutsets that appear with this term and credit offsite power recovery include additional 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> delay flags. Thus this delay term provides no further value in this analysis.

This item will be considered for removal from the recovery rules and documentation.

• DCBTDP-U1/ DCBTDP-U2, Battery Depletion Flags: These flags indicate station batteries were available until the batteries were depleted. A cutset review identified the following patterns:

• Failure to maintain the reactor pressure vessel (RPV) depressurized: injection is successfully supplied by a low pressure injection source (typically the diesel fire pump). However, a loss of DC due to battery depletion resulted in an inability to maintain the vessel depressurized, resulting in SRVs closing, the vessel re-pressurizing and a subsequent loss of the low pressure injection source. Recovery of offsite power prior to battery depletion would allow for the battery chargers to be placed in service and maintain the vessel depressurized before a loss of injection would occur.

Should offsite power be recovered just as the batteries are depleted, it is judged that there is sufficient time to restore the battery chargers and re-open the SRVs before the vessel reaches a pressure at which injection would be lost.

• RCIC is supplying injection until a loss of DC occurs due to battery depletion. The loss of DC results in the RPV becoming overfilled, tripping RCIC on a Level 8 signal. Recovery of offsite power prior to battery depletion would allow for the battery chargers to be placed in service to maintain RCIC control or isolate/bypass RCIC trips as appropriate. Should offsite power be recovered just as the

Attachment L-15-235 Page 45 of 47 RAI 2 Response Attachment Suggestion F&Os (continued)

F&O Number F&O Details batteries are depleted, it is judged that there is sufficient time to 5 (continued) restore the battery chargers before level 8 is reached in the RPV, resulting in the undesired RCIC trip.

• SPFG-2HR/ SPFG-4HR, Suppression Pool Heat Up Flags: These flags indicate that RCIC was successfully supplying injection until suppression pool temperatures required the vessel to be depressurized, resulting in a subsequent loss of RCIC injection. If the two hours was credited, restoration of power will immediately allow a dump of the upper containment pools into the suppression pool, extending RCIC availability for another two hours while ECCS systems are restored, including any fill and vent actions. A review of cutsets containing the four-hour flag (indicating the Upper Pool dump was performed prior to the loss ofonsite power) was performed, and noted the vast majority of cutsets also included a CVOx term indicating a long term event (see CV01-CV05, below).

A small percentage of cutsets did not contain a CVOx term and instead followed the following pattern:

DDFG-1HR/DGFG-3HR. Loss of fuel oil transfer pumps: These basic events indicate that the diesel was running until the diesel fuel oil day tank emptied (1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> for division 1 and 2, and three hours for division 3).

Restoration of offsite power prior to emptying the day tank indicates that the safety-related bus(es) were never de-energized and additional restoration actions for ECCS systems is not required.

CV01 - CV05, Loss of Injection due to Containment Failure: These flags indicate that an injection source was available and supplying the RPV; however, no means of containment heat removal was available.

Containment failure due to no heat removal is a long term event, yet only 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> was credited in this analysis. Following recovery of offsite power, containment venting can be aligned and initiated in approximately 35 minutes. Alternatively, RHR in the suppression pool cooling mode can be restored and initiated, although this action is expected to take longer as it will likely involve a fill and vent. There is sufficient time available for these recovery actions following restoration of offsite power, prior to containment over-pressurization and subsequent injection failure.

Attachment L-15-235 Page 46 of 47 RAI 2 Response Attachment Suggestion F&Os (continued)

F&O Number F&O Details It is unclear for the reviewers if recovery of failed diesel generators (DGs) is credited in the analysis. Discussion with the utility indicated recovery of DGs was not credited in the analysis. The documentation should be updated to include such a statement.

Associated SR(s) Basis for Significance This does not impact the results of the analysis, but a statement should be included in the documentation to indicate the scope of the analysis.

QU-F2 Possible Resolution Add the identified information to the documentation.

Perry Response No recovery offaiied DGs is credited in this analysis or in the overall PRA model. A statement to this effect has been added to the draft documentation for the next model update.

Attachment L-15-235 Page 47 of 47 RAI 2 Response Attachment Suggestion F&Os (continued)

F&O Number F&O Details Some delay type events were identified as being used in the offsite power recovery rule file, but they are not listed as delay time events in the Associated documentation. Examples are: CV01, CV02, CV03, CV04 and CV05. Also SR(s) basic event "CRHIARI-LOSSOFHVAC" is missing from the table in Appendix D of the quantification notebook.

QU-F2 Basis for Significance This is does not impact the results of the analysis, but is a suggestion to enhance the documentation for consistency and completeness.

Possible Resolution Add the identified information to the documentation.

Perry Response The following discussion has been added to the draft documentation for the next model update. In addition, the CRHIARI-LOSSOFHVAC will either be added to the quantification notebook, or removed from the analysis based on the cutset review as indicated in F&O 5.

Time Delays Electrical Basic Event Description Hours Failure CV01 CORE VULNERABLE HPCS OPERATING 4 No CV02 CORE VULNERABLE LOW PRESS INJ OPER 4 No CV03 CORE VULNERABLE FEEDWATER OPERATING 4 No CV04 CORE VULNERABLE INJ OUTSIDE AB 4 No CV05 CORE VULNERABLE ANCHORAGE FAILURE 4 No The above basic events represent containment failures due to over-pressurization, resulting in damage to an injection line that was supplying the RPV. Appearance of one of these terms in the cutsets indicates that injection was successful but containment heat removal was not. There is a significant period of time until containment failure would occur; however, only 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> was credited in this analysis. Recovery of offsite power would allow operations to restore a mode of containment heat removal, most likely containment venting through the fuel pool cooling and cleanup path (there is a containment isolation valve inside containment, valve 1G41F0140, for which no credit is given in the PRA for manual action to open. Restoration of power would allow this valve to be remotely opened from the Control Room).