L-2012-354, Supplemental Information Regarding License Amendment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Generating Plants (2001 Edition)

Supplemental Information Regarding License Amendment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Generating Plants (2001 Edition)

ML12278A106
Person / Time
Site:	Turkey Point
Issue date:	09/19/2012
From:	Kiley M Florida Power & Light Co
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
L-2012-354
Download: ML12278A106 (132)

Text

0Florida 0 L FPL, TODAY.

Power & Light Company, 9760 S.W. 344 Street, Florida City, FL 33035 September10 CFR 19, 52012 0.90 POWERING EMPOWERING TOMORROW." L-2012-354 U.S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555-0001 Re: Turkey Point Nuclear Generating Station Units 3 and 4 Docket Nos. 50-250 and 50-251 Supplemental Information Regarding License Amendment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Generating Plants (2001 Edition)

By FPL letter L-2012-092 dated June 28, 2012, in accordance with the provisions of 10 CFR 50.90, "Application of License or Construction Permit," Florida Power and Light Company (FPL) requested an amendment to the Renewed Facility Operating License (RFOL) for Turkey Point Nuclear Generating Station Units 3 and 4. The license Amendment Request (LAR) will enable FPL to adopt a new fire protection licensing basis which complies with the requirements in 10 CFR 50.48(a) and (c) and the guidance in Revision I of Regulatory Guide (RG) 1.205.

On September 5, 2012, the NRC Staff requested supplemental information regarding the LAR.

The attachments to this letter provide the requested supplemental information.

The supplemental information does not impact the 10 CFR 50.92 evaluation of "No Significant Hazards Consideration" previously provided in FPL letter L-2012-092.

This letter makes no new commitments or changes to any existing commitments.

If you should have any questions regarding this application, please contact Robert Tomonto, Licensing Manager, at 305-246-7327.

an FPL Group company

Florida Power and Light Company L-2012-354 Turkey Point Nuclear Generating Station Page 2 of 2 Supplemental Information Regarding License Amendment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Generating Plants (2001 Edition)

I declare under penalty of perjury that the foregoing is true and correct.

Executed on September 19, 2012.

Michael Kiley Vice President Turkey Point Nuclear Generating Station Attachments cc: Regional Administrator, Region II, USNRC Senior Resident Inspector, USNRC, Turkey Point USNRC Project Manager for Turkey Point Mr. W. A. Passetti, Florida Department of Health

Attachment I to L-2012-354 Page 1 of 21 Attachment 1 Florida Power and Light Company Turkey Point Nuclear Generating Station Units 3 and 4 Supplemental Information Regarding License Amendment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Electric Generating Plants, 2001 Edition

Attachment I to L-2012-354 Page 2 of 21 NRC Question #1 Florida Light and Power Company's review against the criteria in NEI 00-01 Section 3.2.1.2, Fire Damage to Mechanical Components, states that brazed components were assumed not to fail as a result of a fire in this analysis. This is contrary to the NEI 00-01 guidance. Provide justification for this assumption by discussing the extent of use of brazed components at Turkey Point and the significance of this assumption to the analysis results.

Response

The only instance in which fire damage to brazed tubing was excluded was for the Atmospheric Dump Valve (ADV) controls. The ADV control is via digital controllers with I/P converters in the control room (fire area MM/106). These digital controllers are analyzed as part of the safe shutdown analysis. The output pneumatic signal from the converters is transmitted to the valve operators located at fire areas OD-1 15 [Unit 3] and OD- 114 [Unit 4] via copper tubing. The copper tubing from the control room to the valves is routed in fire areas OD-078, OD-079, OD-080, OD-083, OD-084, OD-105 and HH/098.

The failure of copper tubing would result in the inability to open the ADV from the control room. Any action required to mitigate this is a longer term action required to cool down the plant.

The above outdoor areas have no potential for a hot gas layer development. Additionally, the tubing is located at large distances from ignition sources. The zone of influence consideration for Turkey Point is based on thermoplastic cables. Potential failure of the copper tubing would be at a higher temperature than thermoset cables, which have higher failure temperature than the thermoplastic cables. The location of copper tubing is not within the zone of influence of any significant ignition sources. Therefore, the possibility of failed brazed joints is negligible. All above outdoor fire areas are transitioning as performance based and these areas are relatively low risk further minimizing any potential effect of a failed brazed joint. Further, the loss of ADV control capability, even when assumed to occur in all scenarios in these fire zones [OD-078, OD-079, OD-080, OD-083, OD-084 and OD-105], has a minimal impact on the associated scenario risk and will not impact meeting the acceptance criteria in Regulatory Guide 1.174.

Fire area HH/098 is the Cable Spreading Room, which is one of the alternate shutdown

[ASD] fire areas. Redundant control for the ADVs is provided at the Alternate Shutdown Panel [ASP]. This control is an electrical signal to local I/P converters at the ADVs. The pneumatic signals from the ASD fire areas are isolated from the ASP. Therefore, control of the ADVs is ensured independent of any potential fire impact on instrument air in the alternate shutdown fire areas.

Attachment I to L-2012-354 Page 3 of 21 The revalidated Appendix R SSA and the NFPA 805 NSCA analyses assumed instrument air would be lost in any area where an instrument air component or an end device (e.g., air operated valve or pressure regulator) is located. It is assumed that fire damage to the end device would cause a large enough air leakage to depressurize the entire air header for that unit if adequate makeup capacity is not available. Because brazing may only occur in the copper tubing in the instrument air system and not in piping (which has threaded connections), and because the instrument air system typically transitions from piping to tubing only in the vicinity of end devices, there would not be any brazed tubing in the areas where end devices are not located. In addition, the majority of the copper tubing is constructed with mechanical joints (compression fittings). Tubing cables are installed to transmit pneumatic signals between remote devices. The joining of ADV tubing cables, when required, is performed inside tubing junction boxes with compression fittings.

Brazing is not used to join two sections of tubing cables. It has, however, been assumed that the tubing cable vendor may have randomly brazed two individual sections of tubing within a tubing cable during manufacturing process. Since the non-existence of brazed tubes within control tubing cables can not be field verified, the above conservative treatment of the copper tubing was deemed appropriate.

The only systems at Turkey Point that allow brazed copper pipes are the potable water and the breathing air systems. One system that allows the replacement of galvanized piping with brazed copper piping is the service water system. None of these systems are required for safe shutdown after a fire. The only other potential for brazed joints would be in air conditioning systems. Since these systems are failed on location and the brazed joints are local to these systems there is no impact on the analyses supporting NFPA 805.

Based on the above, there is no significance to the use of the assumption that brazed connections will not fail due to the effects of a fire. Therefore, the exclusion of the potential for an exposure fire to damage mechanical components at PTN does not affect the ability of the plant to safely shutdown.

The above information has been added to the corrective action program and will be incorporated into the analysis. After incorporation the analysis will comply with NEI 00-01 Section 3.2.1.2

Attachment I to L-2012-354 Page 4 of 21 NRC Question #2 Attachment V describes a UAM regarding electrical panel factors and provides the results of a sensitivity analysis against a NUREG/CR-6850 method as being an increase in CDF/LERF of 15 to 35% for affected fire zones. The NRC's position on this method is explained in a June 21, 2012 letter to NEI (ML12171A583). Please use an NRC accepted method which removes the panel factors and provide revised results for affected aspects of the application (CDF, LERF, ACDF, ALERF).

Response

A simplified and bounding approach was used to prepare an initial sensitivity study as reported -n the LAR. That simplified approach consisted of removing the applied factors for the fire scenarios listed in Attachment W. That removal effectively changes the applied severity factor to 1.0. The results represent an upper bound. The results of this sensitivity study have already been provided in the LAR as a percentage increase for the baseline CDF and LERF for Units 3 and 4. As requested, the same sensitivity study results for the ACDF and ALERF are 20 to 90% higher for affected fire zones.

Incorporation of the above increases in risk and delta risk, in conjunction with a reduction in the credit taken for the risk offset associated with control room incipient detection credit, result in total risk (CDF and LERF) within the Regulatory Guide 1.174 acceptance criteria, with the exception of select LERF results which are overpredicted.

Attachment 1 to L-2012-354 Page 5 of 21 NRC Question #3 Several of the Findings from the 2010 peer review F&Os identified in Table V-3 identified deviations from NUREG/CR-6850 (e.g., Finding 1-28 regarding propagation of fires outside of electrical cabinets, Finding 3-10 regarding use of lower HRRs for transient fires, and Finding 10-22 regarding "qualitative" disposition of sensitive electronics). Table V-3 subsequently dispositions these items as "This F&O has been closed. The action taken to address this item was specifically included in the focused-scope Peer Review." This response does not provide any information about how the item is currently addressed in your fire PRA. Please describe how the item was dispositioned and, for those F&O's that identify a deviation from NRC accepted methods that has been retained, provide a sensitivity evaluation (total and ACDF/LERF) replacing the method with an accepted method.

Response

The Peer Review history for the Turkey Point Fire PRA (FPRA) included both a full-scope Peer Review in 2010 and a focused scope Peer Review in 2012. The scope and extent of the Peer Review in 2012 included all of the supporting requirements associated with Technical Elements FSS, HRA, and PRM. The results of the focused scope Peer Review found that changes to the Fire PRA that had been implemented were sufficient to satisfy many of the Supporting Requirements (SRs) and also sufficient to close related Facts and Observations (F&Os) from the 2010 Peer Review. In all cases but one, the methods that were used are consistent with NRC accepted methods and did not involve the use or credit of any Unreviewed Analysis Methods (UAMs). The only instance in the Turkey Point fire PRA where a method not accepted by the NRC has been used is that associated with NRC Question 2.

In some instances, the updated analysis was insufficient to fully satisfy the SR. In such cases the F&O from the 2010 Peer Review was superseded by a new F&O. It was the intent of Florida Power & Light to include in Table V-3, only those F&Os from the 2010 Peer Review that either remained open or were superseded by the 2012 Peer Review.

Inadvertently, a number of F&O from the 2010 Peer Review was included in Table V-3.

These entries were dispositioned using the phrase 'This F&O has been closed. The action taken to address this item was specifically included in the focused scope Peer Review'.

Attachment I to L-2012-354 Page 6 of 21 The disposition of each F&O in Table V-3 associated with a finding that was not superseded by a focused scope peer review finding provides the action taken to disposition the associated F&O.

A revised Table V-3 is attached to this response with changes which delete those F&Os closed by the focused scope peer review or which are superseded by the focused scope peer review. A marked-up version of Table V-3 is provided as Attachment 2. A revised version of Table V-3 with the F&Os deleted is provided as Attachment 3 and replaces the entire Table V-3 of LAR 216 submitted by FPL letter L-2012-092.

The table below provides the basis for deletion of each of these Table V-3 F&Os including the F&O number from the focused scope peer review which superseded the initial peer review F&O. The disposition of the closed/superseded F&Os from the Focused Scope peer review is also provided in this table.

A column is provided in the table below identifying those F&Os where the focused scope peer review closed an F&O without a revision to the associated documentation. Only two instances have been identified where an F&O was closed by the focused scope peer review without a documentation revision, these are associated with F&Os 1-28 and 6-16. In both of these cases the focused scope peer review determined that the existing documentation met the ASME/ANS standard requirement and/or an applicable FAQ. A discussion of these F&Os is provided in the Table below.

F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) 1-11 Closed. Can be removed The team reviewed PTN-FPER-1 1- N from Table V-3. 001, Revision 0, "Turkey Point Nuclear Power Plant Units 3 and 4 Information Notice (IN) 92-18 Review (NFPA 805)." This evaluation supports the availability of the components recovered via the altered events process and is

Attachment 1 to L-2012-354 Page 7 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) considered sufficient to disposition this F&O.

1-22 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

1-24 Closed. Can be removed Subsequent to the 2010 Peer Review, N from Table V-3. changes were made to set the altered events to nominal values rather than zero. This was reviewed and is considered sufficient to disposition this F&O.

1-26 Closed. Can be removed Subsequent to the 2010 Peer Review, N from Table V-3. changes were made to set the altered events to nominal values rather than zero. This was reviewed and is considered sufficient to disposition this F&O.

1-28 Closed. Can be removed The approach utilized is consistent Y (Focused from Table V-3. with Final approved FAQ 08-0042. Scope Peer As such this F&O is considered Review closed. confirmed the analysis was consistent with the issued FAQ).

1-29 Closed. Can be removed MCR evaluations no longer use a 0.1 from Table V-3. assumed CCDP. This was reviewed and is considered sufficient to disposition this F&O.

Attachment I to L-2012-354 Page 8 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) 1-30 Superseded by Focused N/A N/A Scope Peer Review F&O 10-1. Can be removed from the Table V-3.

1-31 Superseded by Focused N/A N/A Scope Peer Review F&O 10-11. Can be removed from the Table V-3.

1-32 Superseded by Focused N/A N/A Scope Peer Review F&O 10-11. Can be removed from the Table V-3.

1-33 Superseded by Focused N/A N/A Scope Peer Review F&O 10-9. Can be removed from the Table V-3.

1-35 Superseded by Focused N/A N/A Scope Peer Review F&O 9-4. Can be removed from the Table V-3.

1-36 Superseded by Focused N/A N/A Scope Peer Review F&O 7-8. Can be removed from the Table V-3.

1-4 Superseded by Focused N/A N/A Scope Peer Review F&O 7-1. Can be removed from the Table V-3.

Attachment 1 to L-2012-354 Page 9 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (YIN) 1-41 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

1-44 Superseded by Focused N/A N/A Scope Peer Review F&O 8-10. Can be removed from the Table V-3.

1-45 Closed. Can be removed Transients now only use floor area N from Table V-3. ratio. Transients no longer apply SF and NSP factors which may have resulted in non-conservative results.

This was reviewed and is considered sufficient to disposition this F&O.

1-46 Superseded by Focused N/A N/A Scope Peer Review F&O 10-12. Can be removed from the Table V-3.

1-5 Closed. Can be removed Subsequent to the 2010 peer review N from Table V-3. the model was changed to remove the events referenced in this F&O. This was reviewed and is considered sufficient to disposition this F&O.

1-6 Closed. Can be removed Subsequent to the 2010 peer review N from Table V-3. Table A-I was modified to indicate credited instrumentation in bold test.

This is considered sufficient to address the issues raised and disposition this F&O.

Attachment I to L-2012-354 Page 10 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (v/N) 1-9 Closed. Can be removed Since the 2010 peer review, valve N from Table V-3. 114A and emergency cooler V30A/B cable routing were incorporated via excluded events based on RFI 0274 routing data. No other exclusion/altered events representing other components were credited in the analysis and no additional issues were noted during this review. The resolution is considered sufficient to disposition this F&O.

2-10 Closed. Can be removed Subsequent to the 2010 peer review, N from Table V-3. the values specified in the recovery rules were reviewed and the evaluation of combination events was revised to use the HRA calculator to identify combination events values.

This is documented in internal events PRA documentation. No additional examples of inappropriate application of HEP values in the recovery rules were noted during this review. This is considered sufficient to disposition this F&O.

2-12 Closed. Can be removed The process used for adjusting HEP N from Table V-3. values carried over from the internal events HRA is documented in Report 0493060006.002, Revision 1. This was reviewed and considered sufficient to disposition this F&O.

2-14 Closed. Can be removed Subsequent to the 2010 Peer Review, N from Table V-3. top contributors have been reviewed

Attachment I to L-2012-354 Page 11 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) and the mapping of fire-induced failures refined. The specific issues noted in this F&O were not observed in the current review. This is considered sufficient to disposition this F&O.

2-15 Closed. Can be removed Changes made to the model to N from Table V-3. incorporate fire impacts are addressed in the PTN PSA Model Update Calculation, PTN-BFJR-00-001, Revision 9. The documentation provided is consistent with the process normally used for PSA model updates and is considered sufficient to disposition this F&O.

2-16 Closed. Can be removed Subsequent to the 2010 Peer Review, N from Table V-3. the PRA model was changed to incorporate modeling of power circuits required to support tripping of the RCPs.. This allows fire-impacts on the equipment required to support the operator action to be captured.

However, the changes were not made in the most logical location and could cause confusion during model review and maintenance. F&O 2-16 is considered closed, but a new suggestion F&O (7-2) was generated to recommend relocation of the added gates.

2-18 Superseded by Focused N/A N/A Scope Peer Review F&O 7-3. Can be removed from

Attachment I to L-2012-354 Page 12 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) the Table V-3.

2-19 Closed. Can be removed Subsequent to the 2010 peer review N from Table V-3. the updated LERF model has been completed and is documented in FPL Calculation PTN-BJFR-99-010, Revision I. This is the model used in the fire PRA. This was reviewed and is considered sufficient to disposition this F&O.

2-22 Superseded by Focused N/A N/A Scope Peer Review F&O 10-14. Can be removed from the Table V-3.

2-26 Superseded by Focused N/A N/A Scope Peer Review F&O 9-6. Can be removed from the Table V-3.

2-29 Superseded by Focused N/A N/A Scope Peer Review F&O 9-8 (Suggestion F&O, not included in Table V-3).

F&O 2-29 can be removed from the Table V-3.

2-31 Superseded by Focused N/A N/A Scope Peer Review F&O 8-8. Can be removed from the Table V-3.

2-37 Superseded by Focused N/A N/A Scope Peer Review F&O 7-3. Can be removed from

Attachment Ito L-2012-354 Page 13 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) the Table V-3.

2-38 Superseded by Focused N/A N/A Scope Peer Review F&O 7-3. Can be removed from the Table V-3.

2-39 Superseded by Focused N/A N/A Scope Peer Review F&O 7-3. Can be removed from the Table V-3.

2-4 Superseded by Focused N/A N/A Scope Peer Review F&O 8-8. Can be removed from the Table V-3.

2-40 Closed. Can be removed FPL Calculation PTN-BJFR-99-010, N from Table V-3. Revision 1 documents only 2 HEPs associated with the LERF model.

These are addressed in the TURKEY POINT UNITS 3 & 4 FIRE PROBABILISTIC RISK ASSESSMENT HUMAN FAILURE EVALUATION REPORT (Report 0493060006.002 Revision 1). This is considered sufficient to disposition this F&O.

2-42 Superseded by Focused N/A N/A Scope Peer Review F&O 7-3. Can be removed from the Table V-3.

2-43 Superseded by Focused N/A N/A Scope Peer Review F&O

Attachment I to L-2012-354 Page 14 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) 7-3. Can be removed from the Table V-3.

2-6 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

2-7 Closed. Can be removed Changes made to the model to N from Table V-3. incorporate fire impacts are addressed in the PTN PSA Model Update Calculation, PTN-BFJR-00-001, Revision 9. The documentation provided is consistent with the process normally used for PSA model updates and is considered sufficient to disposition this F&O.

3-10 Superseded by Focused N/A N/A Scope Peer Review F&O 10-10 (Suggestion F&O, not included in Table V-3). F&O 3-10 can be removed from the Table V-3.

3-11 Superseded by Focused N/A N/A Scope Peer Review F&O 9-5. Can be removed from the Table V-3.

3-13 Superseded by Focused N/A N/A Scope Peer Review F&O 10-16. Can be removed from the Table V-3.

Attachment I to L-2012-354 Page 15 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) 3-14 Superseded by Focused N/A N/A Scope Peer Review F&O 10-17. Can be removed from the Table V-3.

3-9 Superseded by Focused N/A N/A Scope Peer Review F&O 10-4. Can be removed from the Table V-3.

4-10 Superseded by Focused N/A N/A Scope Peer Review F&O 10-3. Can be removed from the Table V-3.

4-11 Closed. Can be removed Subsequent to the 2010 peer review, N from Table V-3. transient fire modeling and section 8.1 to Fire Scenario Report 0493060006.004 has been revised to remove the separate transient fire location factors, and the approach is now consistent with industry approved methods. Therefore this F&O is closed.

4-12 Closed. Can be removed Parametric uncertainty analysis has N from Table V-3. been performed for Unit 4 (to be performed on Unit 3 using same methodology). This was reviewed and is considered sufficient to disposition this F&O.

4-13 Closed. Can be removed Cable with previously unknown N from Table V-3. routing which are credited in the FPRA have been traced via RFI-0279.

Attachment 1 to L-2012-354 Page 16 of 21 Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (YIN)

This was reviewed and is considered sufficient to disposition this F&O.

Closed. Can be removed This F&O concerned completion of N from Table V-3. the MCA. MCA analysis was completed (with some new F&Os on the analysis.) This was reviewed and is considered sufficient to disposition this F&O.

Superseded by Focused N/A N/A Scope Peer Review F&O 10-6. Can be removed from the Table V-3.

Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

Attachment 1 to L-2012-354 Page 17 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) 4-4 Superseded by Focused N/A N/A Scope Peer Review F&O 8-3. Can be removed from the Table V-3.

4-5 Closed. Can be removed Changes made to the model to N from Table V-3. incorporate fire impacts are addressed in the PTN PSA Model Update Calculation, PTN-BFJR-00-001, Revision 9. The documentation provided is consistent with the process normally used for PSA model updates and is considered sufficient to disposition this F&O.

4-6 Closed. Can be removed Closed out based on evidence that N from Table V-3. circuit failure modes were used as evidenced the application of spurious operation probabilities via the altered events table.

4-8 Closed. Can be removed Subsequent to the 2010 peer review, N from Table V-3. additional fire modeling analysis for the rooms and scenarios identified has been performed and is documented in Fire Scenario Report 0493060006.004. The issue is considered closed.

4-9 Closed. Can be removed Subsequent to the 2010 peer review, N from Table V-3. Appendix A to Fire Scenario Report 0493060006.004 has been updated to reconcile inconsistent use of severity factors between the Fire Scenario Report and the FRANC model.

Attachment I to L-2012-354 Page 18 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N)

Appendix H has been added to the Fire Scenario Report to document basis for ignition frequencies. The issue is considered closed.

5-16 Superseded by Focused N/A N/A Scope Peer Review F&O 10-2. Can be removed from the Table V-3.

5-3 Closed. Can be removed Subsequent to the 2010 Peer Review, N from Table V-3. additional MSO Expert Panel sessions were conducted and are documented in the Expert Panel for Addressing Multiple Spurious Operations Report (0027-0003-003-001, Revision 1).

The document now includes a description of how the MSO impact is incorporated in the model. In addition, PTN PSA Model Update Calculation, PTN-BFJR-00-001, Revision 9 specifically references those changes that are related to an MSO scenario. This is considered sufficient to disposition this F&O.

5-5 Superseded by Focused N/A N/A Scope Peer Review F&O 7-5 (Suggestion F&O, not included in Table V-3).

F&O 5-5 can be removed from the Table V-3.

6-1 Superseded by Focused N/A N/A Scope Peer Review F&O 8-4 (Suggestion F&O, not

Attachment 1 to L-2012-354 Page 19 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) included in Table V-3).

F&O 6-1 can be removed from the Table V-3.

6-11 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

6-12 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

6-13 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

6-15 Superseded by Focused N/A N/A Scope Peer Review F&O 7-6. Can be removed from the Table V-3.

6-16 Closed. Can be removed The issues raised in this F&O are Y (Focused from Table V-3. based on good practices from Scope Peer NUREG-1792 which is not directly Review referenced in the ASME/ANS PRA determined Standard as the basis for any SR. The that the practice being employed by FPL for requirements the PTN fire PRA is consistent with of the that observed in recent internal events ASME/ANS peer reviews using the HRA standard were calculator. While there are still met and the dependency issues to be addressed as requirement documented in F&O 7-8, this F&O is of the

Attachment 1 to L-2012-354 Page 20 of 21 F&O Disposition by Focused F&O Dispositions Table 4-4 Closed Scope Peer Review Without a Document Revision (Y/N) considered to represent reviewer referenced opinion and not compliance with the NUREG is requirements of the standard. not a Therefore, F&O 6-3 is considered requirement closed, of the SR.

The reviewer also indicated that the PTN approach is consistent with the method applied in other recent internal events peer reviews) 6-3 Superseded by Focused N/A N/A Scope Peer Review F&O 7-3. Can be removed from the Table V-3.

6-4 Superseded by Focused N/A N/A Scope Peer Review F&O 8-1 (Suggestion F&O, not included in Table V-3).

F&O 6-4 can be removed from the Table V-3.

Attachment I to L-2012-354 Page 21 of 21 NRC Question #4 The impact on the PRA results of implementing incipient detection (very early warning fire detection system - VEWFDS) in the Main Control Room is presented in Attachment W as a risk reduction of about 7E-4/yr for CDF and 1E-04/yr for LERF, which implies current risk levels at or above these values. How was this risk reduction calculated? If the licensee determines that these results are correctly presented, please describe what additional compensatory actions have been implemented to lower the risks to acceptable ranges prior to the installation of the incipient detection system.

Response

The review of the calculation used to generate the risk reductions reported in the LAR was found to contain an error. That error resulted in the risk increase factor of 50 (reciprocal of 2E-2) being incorrectly applied to all of the main control room panels instead of only those for which the incipient credit was relied upon for a risk reduction. The calculation for this risk reduction has been corrected and the corrected values for the risk reductions are as follows:

1.29E-5/rx-yr for Unit 3 CDF 1.82E-5/rx-yr for Unit 4 CDF No credit is taken for incipient detection offset for LERF quantification.

The error associated with the control room incipient detection credit has been determined to be applicable only to the control room and therefore does not impact the risk quantification for any other fire areas.

ATTACHMENT 2 TO L-2012-354 MARKED-UP TABLE V-3 of L-2012-092 Florida Power and Light Company Turkey Point Nuclear Generating Station Units 3 and 4 License Amendment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Electric Generating Plants, 2001 Edition

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 1-1 In numerous significant scenarios, the CF-Al The overall Fire PRA results appear to be This F&O has been resolved.

spurious operation probability is assumed to greatly impacted by setting spurious At the time of the Peer Review, the FPRA be 1.0 (true) for any events where spurious operation probabilities to 1.0. Scenario had only a very limited credit for fire induced operation can occur. For example, in 79ALA, which is presently 8E-06 would be spurious actuation probability. The specific scenario 79ALA (one of the top 5 scenarios reduced by at least an order of magnitude instance identified in the F&O was updated.

in unit 3), three events are set to true by assigning spurious operation probabilities The resolution of this F&O also included a affecting the top cutsets; GMMOGE100 to several events. Similarly, with 79AKA, and review of significant fire initiating events and (MOVs 878A or B spuriously operate), 79AJA also at 8E-06. additional credit for hot short induced MAVK3CV303A, OHTX3CNTRL. Capability Perform Circuit Failure Probability Analysis spurious operation was applied in the Category I requires setting spurious for significant spurious operations events, analysis as appropriate. In all cases, the operation probabilities to industry accepted and modify the FRANC model to assign a application of the spurious actuation factor is values. It appears most of the events set to Perform Circuit Failure Probability Analysis consistent with the guidance in NUREG/CR-true would be either MOVs (0.33) or AOVs for significant spurious operations events, 6850 and FAQ 08-0047.

(0.62) or similar, and should not be set to and modify the FRANC model to assign a true for significant fire scenarios. Analysis probability for the event in the cutsets. In using the specific circuit configuration order to meet CCII, the spurious operation for each significant spurious operation would probability should be based on the specific be required for CC II, and may lead to circuit configuration for each significant different results than the generic values, spurious operation.

depending on the circuit design and cable affected.

1-10 Transient Fires are postulated in all fire IGN-A9 Systematic issue. Appears as if numerous This F&O has been resolved.

compartments, as listed in Appendix B and compartment transient frequencies were A sensitivity evaluation was performed that Table 3-6 of the Ignition Frequency Report. underestimated, while others would have involved increasing the weighting factor for All factors affecting the fire frequency were been slightly over estimated as a result, occupancy and storage from 'low' to assessed based upon a slightly modified Initial review was confirmed by walkdown of medium' for all instances where such a NUREG/CR-6850 approach. However, the 5 areas. The ranking on all 5 areas did not condition could reasonably be expected to rankings that were provided do not appear appear to match the walkdown teams occur. The results of this sensitivity found to be consistent with the methods in estimate for each area. that the impact on the calculated CDF for NUREG/CR-6850, result in an Re-assess the transient fire rankings per the each unit was less than 1E-7. Given this underestimate for fire frequencies in some Guidance in NUREG/CR-6850. Confirm the small impact, the existing analysis is areas, and an over estimate in other areas. rankings by walkdown of each area, taking adequate for the application.

One F&O is provided on this SR. In into account the actual condition.

particular: a) Areas were ranked as zero in maintenance, occupancy, or storage even though entrance to the areas is physically possible, b) Areas were ranked as 1, even though activities were not prohibited by plant procedure.

Revision 0 Page V-1 2

Florida Power & Liqht Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update In areas where the room is sealed during operation (roof plugs), transients could have been left in the room prior to sealing, so the ranking on this factor should not be zero -

per the 6850 guidance. During the walkdown, Compartments 70 and 71 both had permanently stored breaker grounding devices, with poly-covers, and 71 had a temporary transformer for the polar crane (operating). Both should be ranked as

'medium' for storage. Similarly, the cable room had storage of 3 temporary fans, cables and blankets and should be marked as medium for storage. This room also appears to include numerous components that will likely be worked on during power, (ranking moderate for non-hot work), and numerous people were present during our limited walkdown. Compartment 88, an open area in front of the switchgear room, had numerous combustibles stored and located, and should probably be marked as medium or high (presently marked as low). Both area 85 and 88 have frequent foot traffic, and should be marked as medium for occupancy. 85 appears as if it should be moderate for storage (no controls). Similarly; no controls appear to be in place for 116.

The above are samples of identified issues, based on our limited walkdown. It appears there will be similar issues with other areas in the plant. We looked at other areas adjacent to the areas we were in (compartments 87, 84, etc), and expect similar problems with the present rankings.

(This F&O originated from SR IGN-A9)

Ee*nts in the altercd eVents table are QR Al4l Affects multiple r*ecVries in numermus This F&O has been closed.

ren...r.edvwth a 0.1 failur.e pro.bability

.. ... Gs P F.men2! The action taken to address this item was-(generally) that include opcning Of ValVeS, Perform 02 18Reis oP;M9n All recovered specifically icudinthe focusoed scopo Revision 0 Page V-1 3

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update etc. However, these components have not HRA D2 cOMPoncnts inthe Altered Events Report. PeeF Revoev.-

been confirmed as availahl* and aFre G Conider also that some roever' events unaffected by spurious operation failing the may involve the manual operation of more valve (92 18oncern), than one component.

(This F&O oreignated from SR CS GA! 1) 1-17 Table 3-2 includes uncertainty values (EF) IGN-A10 Systematic Issue. This F&O has been resolved.

for prior and posterior values. However, QU-E3 Estimate EFs for significant fire The quantitative uncertainty analysis was Error Factors are not propagated to the UNC-Al compartments. ESTIMATE the uncertainty prepared subsequent to the peer review.

compartment specific ignition frequencies. interval of the CDF results. ESTIMATE the A parametric uncertainty evaluation that The other parameters, such as conditional UNC-A2 uncertainty intervals associated with considers fire ignition frequency as well as failure probabilities for circuit failures, do not parameter uncertainties (DA-D3, HR-D6, other variables was performed that uses a have uncertainty intervals. HR-G8, IE-C15), taking into account the Monte Carlo sampling process. The results The lack of uncertainty intervals would not state-of-knowledge correlation. of the analysis showed a mean that was generate meaningful uncertainty interval of s f e lote ais sha meancthat was the CDF/LERF results. (This F&O originated slighter higher than the calculated results from SR IGN-A10) which was expected.

1-18 During walkdowns, several key areas IGN-A7 Appears to be missing components in This F&O has been resolved.

appeared to have ignition sources not numerous areas, based on a limited The specific instances identified in the F&O included on the ISDS. For example, in the sampling during walkdown. were reviewed and the analysis updated cable spreading room, 2 transformers were Perform a re-verification of the ISDS for accordingly. In addition, the supplemental in the compartment (3X033 - 75KVA, 3X130 significant fire areas in the FPRA. Add walkdowns that were performed as part of

- 45KVA), both within the screening distance missing components to each ISDS, where ongoing analysis refinements efforts for the of targets. Also in the compartment is CP- applicable, significant fire areas did not identify any 600 spectralink cabinet, an open cabinet, other omissions.

the RCP Vibration Monitoring Cabinet, 4P21 and 4P09 instrument AC panel. Note; we did not do a 100% review of the CS room, so additional cabinets may be missing. See also F&O 1-19. (This F&O originated from SR IGN-A7) 1-19 It appears the Ignition Source Counting did IGN-A7 Appears to be a systematic issue in the This F&O has been resolved.

not count Lighting Panels or other similar FPRA. A re-assessment of the lighting panels was panels. For example, there were at least 8 Include unsealed lighting panels and similar performed. The re-assessment focused on lighting panels in the cable spreading room electrical cabinets in the ISDS as potential the need for treatment as a fire initiating that were not on the ISDS. Additional similar ignition sources. event. No effort was undertaken to alter the panels are located in most electrical rooms population of electrical cabinets considered we walked down, such as the switchgear in the fire frequency development.

Revision 0 Page V-14

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update rooms and other electrical rooms. Based on Therefore, the existing values potentially our walkdowns, many of the lighting panels have a conservative bias. The assessment should be included in the ISDS, based on did not identify any instances were explicit guidance in 6850 and the subsequent FAQ treatment as a fire initiating event was on sealed cabinets. A review of the generic needed.

guidance provided for ignition counting did list the screening of small, wall mounted cabinets (sealed). However, the lighting panels do not appear to meet the criteria listed in the procedure (not sealed, numerous switches/breakers), etc. Many of the cabinets are located close to cable trays or other intervening combustibles, so a small fire could result in a larger fire due to spreading. (This F&O originated from SR IGN-A7) 1-2 Section 4.1of the Component Selection AS-B1 The significance of not identifying This F&O has been resolved.

Report mentions: "Since the FPRA ES-Al components as causing initiating events is The FPRA assumes each postulated fire quantification calculates a fire CCDP and the basically that the assumed model impact is results in at least a reactor trip. Logic is initiating event frequency for each zone is ESA3 accurate by modeling a reactor trip with a included in the model so that appropriate based on the fire ignition frequency, the ES-A4 subsequent failure of the function, rather event tree is quantified if the fire induces a initiating event faults are not required to be FQ-A2 than modeling the initiating event itself. In different type of event (event tree). The used for FPRA quantification." Fault tree some cases, this impact is a matter of timing overall structure of the FPRA model was initiating events were not impacted by the for operator actions. In the case of this reviewed to address the specific item component mapping, and are therefore not FPRA, the HEPs have been conservatively identified in the F&O and to confirm changed by fire damage. As a result, set assuming a loss of MFW as a starting appropriateness of overall treatment. The equipment associated with Fault Tree point. However, the fault tree initiating only change that was required was related initiating events were not identified as events include loss of CCW, loss of HVAC to biasing the application of recovery actions components potentially causing a fire- and others. It is not clear that the present so that they were based on an assumed loss induced initiating event. 163 events are model accurately determines CDF/LERF of MFW.

screened in Table A of the Equipment results for systems impacted which may Selection Analysis based on being cause a complicated reactor trip (special associated a initiating event).

fault tree initiating event. Most are modeled Modify FPRA to model the fire impact to in other system models. However, Several Fault Tree Initiating Events, and analyze the were found to not be modeled in the rest of FPRA assuming a fault tree initiating event the model: CPSD3PC61 1, CPSD4PC611 for those areas where the initiating event and 2 related failures. A few others (Cooling can occur.

units) do not appear to be modeled Revision 0 Page V-1 5

Florida Power& Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update elsewhere.

(This F&O originated from SR ES-Al) 4-2-Ev ..en.ts rDAC,'F3ECC-B & DACF3E=C-A are NRA G! A review of the Altered Events Report This FrO has been closed.

anded 'under the containment coolerF ault indicates there are likely numer*ous The acti9o taken t addresc this item was Tree Logic. Each aFe applied with a GOmbiations ef .opping HEPs (0.1 in the specifically included in the focused scope screening value re over.of 0. r.esulting in a alter-,ed events repo"t. Pee Pew-co-mbin e rl Review the altered events repoit fo

• AA.K3 PR28!mnd 282 csmine 'under an

• fire areas, and determnine the

ignificant AND gate, resulting in a combined recovery

- cmie E where they occur under Ra of 0.01 The above are only examples, AND gate. Use a single recover, for these based on a ra ndomn review'A of tMo fire areas, events, Or set only one of the events to 0.1 and a few HEPs for each. Two events in in the altered events report.

about 15 were found to have the above problem. (This E-O originated from SR HRA-G4) 4-24 Event EREE32B6G3F is listdasF; 0.0 NRA-C4- Multiple discrepancies in the altered events This F&O has been closed, probability in the altered events report rtable. Pfo SIome may be d.ocumentation issus The. action taken to address this im was areas, to.....

even though the description says set que'oy of thiz A^,tpted F=.... Table-(incorrect description). Howvver, may specificallyn in the focused scope to uey 1.A f heAltere Evns al appear to be errors in the probability foi ReRviw comes. up with 146 entries, where the these ovents. Setting the events to 0.0 probability is set to 0.0, but the descriptien results in the events being screened from says, to set the event to 0.1 A question Was the PRA result. However, many.may be as asked on this issue, and the response was a result of setting other related events te 0.1.

that these were set to 0.0 and anoether event in this case, the documnentation associated was sAt to 0.1. However, there is no way to .ith the reason/description needs to be verify, tr*k or repeat these cettings. In updated.

reviewing a few events, there was no easy C~orrect Altered Events Table, where way to determin that

. there is a applicable. Additionally, when events are set corresponding 01eettarcorsto zero, provide a tracoable method to the failure. (This F&O oiginated f..ro SR determine that a corresponding event is set HRA-G-) to 0. 1 or another value.

1-25 There does not appear to be a review of FQ-E1 Requirement of QU-D5 as called for by FQ- This F&O has been resolved.

non-significant cutsets in the PRA QU-D5 El Review of non-significant cutsets performed documentation. Perform a review of non-significant cutsets and documented.

(This F&O originated from SR QU-D5) and accident sequences, as discussed in QU-D5 for the FPRA.

Revision 0 Page V-16

Florida Power & Light Attachment V - Fire PRA Quality Florida Power & Light Attachment V Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 4-26 By uing the altered events tablc, and PRM A2 G-iven the large number of events set to zero This F&Dhhas been closed.

setting recovered events. to 0.1 (seeping) SYA! inthe alter~ed- events report (27614), and the The action taken to address this item was and the other events in the fault. t.re- t zer, SA.. !arge number of basic events greater than specifically included in the foe-used scope random.failures.that.may.fail.the.HP PRdrndo k ..... that may fail the HER- are ar E 03 i the PRA, there are likely numerous Per ReJeq not icueinthe results. ifthe random events no~t included inthe model.

events are greater than 2 orders of Re'viw the values set to 0.0 in the altered magnitude belo'w the top or ontribu-te to 1% events report and modify the solution to of the system fa.ilue rate, they should be include the random failures in the results, inc'luded based On SY A! 5. (This F&O wefa*=,e.

originated from SR SY A!5 and PRM A2) 1-27 Significant fire compartment contributors to FQ-E1 Requirement of LE-F1, F3. This F&O has been resolved.

LERF are documented in Appendix C of the LE-F1 Document the contributors to LERF based Added LERF top cutsets and importances summary report. However, the contribution LE-F2 on the requirements of LE-F1 of the internal run as well as sensitivity analysis in from plant damage states is not provided or events section of the standard, as required Summary Report. Also performed and the contributors from LEB SRs. Sources of LE-F3 by FE-Q1. Document the Sources of documented the uncertainty evaluation for uncertainty, including sensitivity analysis UNC-A1 uncertainty, including sensitivity analysis LERF.

performed, are not evaluated for LERF. performed for CDF in Appendix D of the (This F&O originated from SR LE-F1) Summary Report.

It appears in the analysis that MCC fires for FSS-A4 See FAQ 012. Scenario 058C A showed a This F&O has been closed.

un, v-ented MCC fires are net considered to CDF of 1E 05 prior to screenig. The action taken to address thisitem was damage targets outside the selected hG. Include in the model large MCC fires specifically included iR the focused scope Sec-P SGenario 058C A.The EForigial draft of prpgtngotsde of sealeId MCCs; PAPF RePYPMI.w FAQ 42 included discussion OR MCC fires, and pFrvide a probability of the-fire propagating ou-tsid-e the ['.1C. This recmmen.ded approach Was net incl'uded in the final FAQ, and as a result, the treatment of MCC fires does net meet the guidance in the FAQ for a sealed cabinet. The end result is that the FPRA should consider MCCs as unsealed, due to the possibility of energetic fires resulting in the MP.CC door being opened, or as a result Of maintenance on the MCC being the cause of the fire (actual events) when thc doer is open at the start of the fire. It is ur understanding that the NRC Fire PRA folks also do not consider MCCs to Revision 0 Page V-17

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update be sealed cabinets. The proposed industry approach basically summarizes to the folloin*g : M . open at the top are considered unsealed, but MCGs that arc sealed At the top have around a 0.1 probability of propagation. The data analysis for this Val1 'ast noetvalidatcd, but was considered conse,-ative at the time of the proposal, Gince 6Peveral fires- where it was not clear if they camoe outside of the MCC were-assumed unknown (1/2 an event).

(This F&O originated fom SR F98 A4)

Thecotrl romabnd nmet.eai The 0.1 scOPing analysis isant based On This F&OQ hat; been closed.

106 A, uses6 a CDP f 0 1 (scrpig value) analysis of plant specific procedures or The actin taken to address this was However, the scenario is one f.the top Cn, analysis. The HEP' may be higher or lower, specifically inc;lue, i-;.n*A the focused,opo scenarios. (This F&, originated from SR depending On p*roedures. Additionally, the . peerv ew-F=99 92)circuit analysis may show the Remote Shutdodzwn dIanel may not fuJnction for some scenarios due to M*SOs.

Provided detailed analysis for MCR abandonment CCDP.

1-3 The internal events PRA model has AS-B1 As a result of assuming a reactor trip and This F&O has been resolved.

numerous locations in the model where the ES-Al not mapping components/equipment to The issues and concerns identified in the specific initiating event results in a model modeled internal initiating events; the risk F&O related to the fire-induced initiating impact. For example, under gate U3QT07; ES-A3 can be under-estimated. In this case, since events were reviewed. The review found initiating events that can cause a PORV or ES-A4 the general approach used is systematic, several instances where a change to the SRV to lift are ANDed with the failure to FQ-A2 this problem is difficult to determine without modeling was required to allow the existing reclose the PORV or SRV. In this case, significant effort to combine the impact of treatment methodology to be retained. The special initiator %ZZIP6U3 is identified as an each modeled impact. In most cases, the review did not identify any instances where initiating event that will cause a PORV lift, modeling results in non-conservatism in the specific fire initiating event logic beyond that along with %ZZT2U3. Equipment that can result. However, the fix for feed-and-bleed already in the model was needed.

cause each are not mapped or modeled in resulted in conservatism for most of the the Fire PRA. As a result of a previous scenarios where FW is not initially lost. In review, the modeling of Feed-and-Bleed was either case; whether modeled conservatively changed to assume a loss of feedwater (low or nonconservatively, the standard SG level) occurred. The shorter time results requirements in this area are to model the in a higher HEP for feed-and-bleed in all impact of the FPRA accurately.

scenarios, regardless of whether a loss of Revision 0 Page V-18

Florida Power & Light Aftachnnent V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update FW occurred. However, numerous other Map all identified internal events initiating modeling impacts can occur, that are not events to the specific components that can modeled. Under gate 162115, logic for cause the event, and modify the FPRA to HVAC unit 3S230 failure to start is included determine the CCDP based on the fire-when a Loss of offsite power would occur. induced initiating event that results.

This logic is applicable only for when a LOOP occurs, and not applicable for non-LOOP events. This type of logic is contained throughout the internal events PRA modeling. Another example is under gate El 104A, where loss of DC power results in lockout relay failures. There are many other examples throughout the PRA. Additionally, the identification of the specific initiating event for quantification was not performed per the requirements of FQ-A2. For quantification, the modeled initiating event is assumed to be a reactor trip in all cases.

This treatment does not meet the intent of SR FQ-A2, where the quantified model should encompass the risk contribution from all applicable initiating events.

4-W Fire modeling war. conducted via generic fire FSS-G! The p.resnt analysis prov der. a bounding This'.F', has been closed.

modeling fromn which ZonRes Of Influence FS G approach for fire severity in most Gases. Thin actionn taken to address this itmwas (ZO') for Specific initiator types WaS Pefor-Fm 2 po;int fire modeling, when specifically includ in the focused scope geRnFrated. The Z*ls were used to def-ne applicable, for significa*t fir i PewRepiei bounding fire characteristics for each fire scenario. Characteristics that are used te b9und potentially rick contributing fire eVents are identified in Attachment B of the FiFre Scena.ie Report, (Repont 0493060006.004).

BRased- on the- ueOf AbounMding approach this SR is judged to be met at CC 1.

Significant fire scenarios should be developed with 2 paint fire modeling. (This F&G eriginatod from SR FSS C!)

Page V-19 Revision 00 Revision Page V-1 9

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Fir scnaio yauatontoos eree F6S-G2 Significant scenarios are not developed ThsFOa becocd developed based o-n the Generic Fire conGiGGsidoring firc growth. The current The -actionta;ken to address this item was Modeling Treatments (Project Nu-mber approach is consevative. specifically included in the focused scope SPH02902.030). These lncl'udc fire growth times for significant fires, P-ee7rRehAe.

walkd.W./evaluati. n tools arc based on where grot4h i'me ..;isalable in bo9unding fires that are assumed to caus~e N1UR EG/C.R 68-50-or FAQ 052 for transient target damage at a height above the base ip fire with the firo burning at peak intensity.

Because these tools assu5me a firs burning at peak intensity this SR is considered met at CC I. (This F&O originated from SR FSS G2-)

4432 No)cvidcnce was idniidthat suggests FSS-GQ Significant scenarios dIo not consider decay. This F&,O has; been closed.

that fires were assumed to burnout over a SS Gi Include growth and decay for significant fire The action taken to address this item was period of.time.....u...n.c..n wa. scenarOcs in the FPRA. specifically included in the focused scope estimated uosi a peak heat releasr .eeF as Rvew.-

dictated in Attachment B of the Fire Scenario Repoet, (Repert 0493060006.004).

AGccrdingly ths SIR is considered nOt Me for CC!1/111. (This F&O originated from SR 4-13 Except forF the MrCR fire scenarmios, no other F99 P9 Systemaio This- PR-O has been clos6ed-.

fire.scenario.has used the Non Suppre.sion Apply .o. suppress.io fac*ors for significan The action taken to address this e Was Probability (NSP) in PTN fire model at this n whe applic time. (This F&O originated from SR P88 Peew-P8) 1-34 No evidence was found that supported FSS-G4 Systematic issue. This F&O has been resolved.

confirmation of conformance of fire rated Provide the documentation that supports The treatment of barriers in the MCA is barrier segments to applicable test confirmation of conformance of fire rated based on information in the Fire Hazards standards. Additionally, the effectiveness, barrier segments to applicable test Analysis and supplemented with walkdown reliability, and availability of any passive fire standards, and the barrier effectiveness, observations. The analysis documentation barrier feature credited does not appear to reliability and availability, was updated to provide this information.

be performed. (This F&O originated from The MCA was modified as needed to SR FSS-G4) incorporate the results of this effort.

Page V-20 Revision 00 Revision Page V-20

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 4-35 The mul-ti compadment anaysic assumes a FSS-G5 Syemati-issue This F&O has been c!osed.

bounding value of TE 3 af For any scenai eeted if the adjoining The act taken to address this item was actAive fire barrier elemnents. Acatual fire physical analysis units are separated by specifically icueinthe focused scope barrier elements are not considered instead active fie barrir ele-m-e.nts, ANTIFY the Pee-Review, the failure probability of a fire door isf~tvnPF6 rlaityand availability of a ssumned- for. active

.......barrier

...... ...elem~ent

.... failur~e efecivnsreliabF~ie the lty, a activ-e fire barrier element.

beca;user this failur~e probability represents the highest single probability of a barrier failure. Accordingly this analysis provides a qualitative bounding assessment fire barrier feature failure probability. (This F&0 originated from SR FSS G5) 4-36 HFE included in the altered events repor F4QG4 SystematiGIssue This F&O has been closed.

are not treated under dependency NRA-Gt Perform dependency analy'is for HFEs in The action taken to address this item was ev-aluations Een though the v-alu-es are the Alter Events Table, and provide a specifically in*dWed in theP focu scope screeRn4in value, the*,ocess dependency quantification that irporates the P ee-FR:*e'vew evaluation may result.in a higher HEP, neHHFE dependency.

especially if more than 2 events are in a single outset. The sensitivity study case 3 doc'umented in Appendix D of FPRA Summary Repo~t 0103060006.00-sow that doubling all non recovery HE'Ps using multp !lersgreater than 1 yields a delta CDFn inrGease of 6.60E 5, or 24.7,0 o~f the base fire C-DF=. Fudrhermore, if doubling the HEPs increased C-DF= by -25%, it stands to reason that halving the same set of HE~s woul-'-d decrease the CDF by a s...imilr Ra-mon. it is reasonable to assume that Adetailed analysi c ould reduce moest of thee sreeingHEPs by at least half, and in many eases by much more. (This F&O ogSR r F CA)

Page V-21 Revision 0 Page V-21

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 1-37 Significant contributors to Fire PRA results FQ-E1 Requirement of QU-D7 This F&O has been resolved.

are included in Section 4.3 and the QU-D7 Provide importance measures as required Importance measures for CDF and LERF appendices of the Summary Report. This by QU-D7 and FQ-El have been determined and added to the includes a list of operator actions that Summary Report.

contribute to CDF. However, no importance measures are provided for CDF or LERF.

(This F&O originated from SR QU-D7) 1-38 Results of the Fire PRA did not include the FQ-F1 Systematic Issue This F&O has been resolved.

following: (e) the total plant CDF and QU-F2 Provide required documentation per QU-F2 The documentation of the analysis results contributions from the different initiating UNC-A2 and FQ-Fl. has been expanded to include the events and accident classes (i) the information noted in the F&O. These results uncertainty distribution for the total CDF () were also reviewed for reasonableness and importance measure results (I)asymmetries no issues or concerns were identified.

in quantitative modeling to provide application users the necessary understanding of the reasons such asymmetries are present in the model (m) the process used to illustrate the computer code(s) used to perform the quantification will yield correct results process. Some of these issues are listed in other F&Os.

However, item e(accident classes), I (asymmetries) and m (validation of computer codes) is not covered elsewhere. (This F&O originated from SR QU-F2) 4-4 Fire Induced Spurious PR opening is E Systematic issue of ho.. the PoRV, other This F&- has b coe.

modeled -rnder Gate 3S2ORFIREEQUIV PRMA small small LOCA initiating events aRd sm6al The action taken to address this item was (2i6 gate). However, this.ma.l =OCA impact LOCA iEe.r modeled

. in the RA. picy inluded in the focu..d scope

' ......... '............... PRM-B 139 sp ..........

is only added to three locations in the PRA For any fire induced initiating events, Peer Review, location inwhoe smaFTAModel. For example, i 1including PQRV opening and other smaill 1G~a1Gn6in M~el teFOFexamleLOC.A CA.FTA. initiating events, ensure the newly thc fre loa iducc loic i smal notdeveloped logic is added to all lctosin included under gate GHLO1 .or 7 ther the PRA model to ensu-rt the PRA moedel loat4ions.; Fo~r small1 small4 ILOCA, the- PORV. ovscrety oeti & oPM Simiarly, small. smAll I OA is included in when evaluated. Also a#B*tS the equipment locati in..th-..e PRA mdel, while the Felection, only with regard to the

.impact of Inducemd PORV logic (and other small smalltheqpmnonheodlditaig logic) is only icuein3 locations. The Revision 0 Page V-22

Florida Power & Light Attachment V - Fire PRA Quali Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update above are examples for two of the revicwcd event.

added logic. The fire indu-ced initiator e,-ent faut trees hav-e been added into the PRA model. The consistency of the model changes have been r..i.wed.. One question a-bouit the c-hanges is that the fire risk coulid be potentially underestimated. For example, gate 3FIRES2 has been added to simulate the small LOCAs induced by fires. Under gate 392ORP FIREQWIV, gate 3FIRES2 and intemal event t-L-CA4iitiator 04776213 are ORed, Which secMc to be appropriate. However, %ZZ02/_37 i11undei" 11 parent gates, while 3FIRES2 only has orne paret gate. If the intended fire damage is smnal! LOCA, all system functions affected by small LOCAc should be affected. (This F&Q originated fromn S.R ES213)4 1-40 The quantification of significant basic FQ-F1 Requirement of QU-F6 and FQ-F1. This F&O has been resolved.

events, not cutsets provided. and accident Additionally, thesequences definitions is QU-F6 Provide the quantification of significant basic The Summary Report has been updated to used for significant basic event, significant UNC-A2 events, cutsets and accident sequences, provide the importance measures of the and the definition used for significant basic model basic events, top 90% of all plant outset, and significant accident sequence event, significant cutset, and significant cutsets, and a review of the scenarios are not provided. (This F&O originated from accident sequence contributing more than 1% of the total risk.

SR QU-F6) 4 The HRA does not look at the Fire Specific HRA-D2 The simplified factors included in the HEP This F&O has been closed.

factors affeG;tin (a) quality [typo (classrom H 2 modifier approach does not provide the The aGtion taken to address this item was or simulator) and frequencYJ Of the operator equivalent of a eHRA-. speifically the focused scope training or experience (b) quality of the Provide a detailed HRA for significant HEPs P-eer-Review&

written procedures and administrative in the FPRA results. Detailed HRA should

.ontrols (G) availability Of instrumentation account for the Fire Specific faGc9to as listed needed to take corre.tiv.e ac.tio (d) degree above. The present multiplier method deoes of c!ar!ty 9f cues/indications (e) hu*mn* not appear to meet the requirements of the machine interface (1 time available and time standard for detailed HRA.

required to Womplete the response (9) cemplexity of the required response (h) environment (e.g., lighting, heat, radiation) under which the operator is working (i)

Revision 0 Page V-23

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update accessibility of the equipment requiring manipulation (j) necessity, adequac~y, and availability of special tools, padts, clothing, et. As a resul, the HRA does not appear to meet the PRA ta quRent*s foFra det~ailed HRA..(hsFOoiiao from SR HR-H2)

Page V-24 Revision 00 Page V-24

Florida Power & Light Affachment V - Fire PRA Qualitv Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 4-44 The FPRA models6 0.1 HEMP ývalues for HRA-CQ! Appears to be a significant non conservative This F&O has been closed.

several recovery actions that are not in the impact to PRA results. Given the actions in PRM-8R4 The action taken to- address this item was SSA and arc no-t in the fire ca.;fe shutdown t~he altered events report are being added to speciicall included in the focused sccpe procedures (See F&O 6 11). These recovery the mode! as needed reco-eries in order to JPeeF Reyvew actins a*re input into the PRA model by ensure risk is low, and given the resulting adjusting the random independent failure Fecoeye~ actionRs d-o not show~ up inthe prorpaility li fr Fnre*overc*rGcmpui*ne to results in m cases.,

.t the*e appears to be a 0.1 and the logical true setting fo ulcuAARiRIec uciWeen ;nc CnnT,mDnn

..... ne .... ...

TflJT 1-...

a;mm'oi Il,2 ;CCOCFve

..M e . . . e .a F

10 W

actions to the procedures and the 0.0 This approach introdues several quantification of these actions in the FPRA.

issues. including the following: The use of it appears part of the disconnect is that the logic modeling, as moedified by the altered recovered eGuI' at*'Q thp r) !

events table, Fresults in the recovery Values recovery HEPs altogether in several being screened fro-m the rsls instances. For compartment 070 AB3, basic Revise the approach described for assigning event E0BR330303 is set to 0. for recovery recovery HE=Ps via the altered events table of alternate feed to load- -e~nter 3H, but this and ensurethquniiainpocste recovery feeds into AND gate E3013H. This inen edrsults.

AND gate appear-s to bheInull1ified, however as the seconRd input to the gate is false clue to the 0.0 values present in altered events (based on visualizatien of settings from the altered events file for 070 AB in CAFTA).A quantification of 07-0 AB prFodued 10,000 cutsets and not insrta ofince EC ouR330303 waS present. A.Fuher review af the 070r A quantification inRdica-te-d the following events aced ithe alte'rend events table as r*riResdonet appear in the final utets A.TPXPUMPA.A2ASTRT, EB2F33003H, ECB33A15,FAVC3 1606, MAVX3 311, MSVR3 311. Further review of ether areas

And- otherF recovery values (0.1 and 0.0 value iA;n the altered events reoped) appears to iniaesimilar problemns Will occur. The above are examples (potential issues),

which appear to be logic problems resulting from the use of 0.0Q a;ind 0.1 inputs in the altered events reoped. Additional problems are likely for other events. (This F=&O originated from SR HRA (A)

Page V-25 Revision 0 Page V-25

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 1-4. The method and calculatins for transient F-R-R W The Severity factors used are in sionificant This F&O har heen Airesid fire severity factors (SF) arc not clearly fire scenarios in the FPRA. .I.- .- __

juJ.Ies documented and several SF v-alu-es in DocumenRt t-he severity facto)rs useRd for e.ac~h specitlcally includec )GLused SG013e FRANC model are not consistent with the scenario including the basis. Revise the PeeF Reymew.

.nes listed in the FSS repert. For example, transient severity factors to remove double fire scenarios 079A J/K/L in Appendix A of counting of the area factor !ncl'uded in both FSS report have an SF value of 3.05E 2, the square footage of the comFpartment, and whiGh is used ' in the FR AN "Cmodel, but is the BE 02 calculated in Section 8.4 of the not consistent wieth the BE 2 value incl,u-ded F9-5 repao in FSS report sectiOn 8.1, GENERAL TRANSISENT SEVERITY FACTORS.

FPL'ERIN staff reviewed this issue and stated that the SF is calculated based on a floor are. facr-r (FSS report sect*io 8.3) timnes the 9FE 2 transient ignition frequency adjustment factor (F=S report section 8.*)*

FANr Q079.A JKI'L, this method re*ul"ts in a factor ef 1.52E 2. A factor of 3.05E 2 was used in FRANC. The above points to the following: a) The severity factors- us9ed- aFre not Well documented, traceable, or consistent wit.h what is provided in.the text o the FSS repeot. B) The tw:o severity factors basically double co.unt the area severity factor. For the last case (B3),the BE 02 already includes a conside.ratio, based 9n experience, that the fire that occurred is near a target (compo.ent). As a result, the 6 events listed just happened te no~t beP nePar any components, resulting in a low probability of damage. To put this another way, it can not be demonstrated that the BE 02 factor is duo to the small size of tho transient fires rather than being due to the fire being in a location REA near!a component. A reviewi of the control/1aux bid ina ,utrm uppruu*y riue ;iyer

,whichi tes. a fairly large ifie Ii ireuprey g8Un Size fuRe 10 Page V-26 Revision 00 Page V-26

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update raise the fire detectors up to 160+ degrees.

.A.ceGonRd- event is described as "A leaking regulator ignited leaking propane." This can obviously be a larger fire. Since the location/area of the originating fire is in the 9F 02 factor, and in the "are-a" factor, the double cnting resultsI in an underestimation of the likelihood of fire-damage for a transient fire. (This F&O originated from SR FSS D3) 46 The PTN EPRA methodology generally does F=99-QQ Requirement of FSS DO This F&O has been closed.

not ncl'ude postulation or evaluation of Perform a qualitative evaluation of smoke The action taken to addres s item was smogke damage. Additional review shows~ damage to FPRA equipment per the specifically icueint~he focused sGcope that the smoke issues do9 not affect the Fe_________of____ 9.________ew FPRA results significantly. However, the- r m fe w FPRA does not include a qualitative evaluation of smoke damage to FPRA equipment. (This F&O originated from SR FSS-q4) 4-5 -HEP EH1FICLR2 R3H. is icuein the F=PRA HRA-CA. Completenes isu i the douen.ietation of This FRO has been closed.

model, without specific analysiS in the task. the HRA The act;ion t e address this item was 12 report This is a modified HEP from the E-.nsurder d mentatin of all HEPs is specifically included in the focused scope internal events PRA. Additionall, provided in the HFE report. Peer-Review, MHFP3BAMT is included in the model as a screen!ng value ,withot documentation in Task 12. (This F&O originated from SR HRA-G) 4-6 many Of thes

,Although n affectinog Es-GSystematic issue. As a result of this, the Ths FRO ha operator aGction a.re inc*lud in Table ^A.1of pmn lt fo..r the. FPRA. (considered eRAqAui The action taken to address this item was the HRA report, and many are traced, none separate for this review) does not include specifically in t 'he'focused scope

-inlded of the instrum+ents reviewed (partial review) any of the identified ins. Table A41 Pee.Review appear to have been tagged as FPRA of the-NRA.

affecting. As a result, the instruments do not Include needed Instruments in the PRA appear to be in the FPRA equipment .... ... +l .f m;ln the selection process or equipment list. For impamt On the FPRA HEP l example, under GHFPS1 RCRC, Fl 3 93 isp t listed and icueinthe SSA. However, Revision 0 Page V-27

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update this s; listed as non PRA affecting as is no inthe TAGBE, UJNL table or other associated table.

(This F&O originated fromn SR ES G-1).

Additionally, when new equipment is added that are noAt in the SSE (for example; LOCAL CHARGING PUMP HYDRAULICA COUPLING TEMPERATURE Ti 3/4 6716/6717/6718), these do not appear to be traced or included in the PRA equipment list.

Most of the PRA (non SSE) instruments are net identified by number.

4-9 Recovery to the Charging Pump Suction CS A!! 1st item does net appear to be ignifigat. This F&'.o has been cAloe va*lve 1 4A and the associate components Q9.-Q2A verifies MOVA350 is available, but the ction taken to address this item was is provided through the excluded events EA doc~u~ment-ation is very difficult to fellow, specifically icueinthe focused scope tabe ad he ltredevetstabe.However, the MOV and the associate Ppeeeepy~w~

However, the recovery is actually crediting PRM--84* component do Rot sho9W up in the flow through MOV 3 350.* This flow path is Equipment List for the PRA, nor Will it show n.. included in the FPRA model. Similarly, UP in the impor+tance lists, rest, I e;tc. The theQ Normnal Con-tainment Coolers are not suroate event (0.1 applied to 114A) de mndelel in the FPRA, but recoepred in the not represent what is in the model.

altered events table througgh Gill 0.1 Additionally, since the flow path was Rot recovery. The ahbve are examples. The modeled, the PR-A .onsideratins may not entirety of the altered events table has not all be conside.red. 2nd item:l.This one may been reviewed for recoveries that ma" be be significant, since the conRtainR.ment cooer01PEF bringing in additional o*.mponens not in the ran bhecom.piated, and a likely mismatch FPRA or FPRP, Equipment List (This F&O betw-een the SSA and the modeled FPRA originated froam SR CS All) logic could result in significant differences.

Add the MOV 3 350 andI the associated flow'A path to the PRA model,'normal containRment coolers. Additionally, review the altered events repo"t to determine if additional credited flow paths are moedeled, where the actul, components and supped. logic is not 0n4he-FPRA.

2-1 The plant partitioning task does not include PP-B1 Section 3.11.5 of FHA states that man-hole This F&O has been resolved.

detailed discussion with respect to this SR PP-B7 covers are justified as three-hour fire Walkdowns of fire zone boundaries were PP-B7. The manholes are modeled as boundary although they need not to be Revision 0 Page V-28

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update separate fire compartments. However, no specifically rated as fire barrier. Therefore, performed and documented in support of a walkdown for these manholes has been the modeling of manhole as fire review of the Fire Hazards Analysis update.

performed. No justification for the modeling compartments is considered acceptable Additional discussion regarding the basis for approach has been provided except being although no walkdown has been performed the ignition frequency for the manholes was briefly mentioned in Section 2.2 of Report for the manholes. Other credited barriers are added to the documentation.

PTN-PSA-7.01 Revision 2. Walkdowns were discussed in PP-B2-4 above also not documented for spatial separation Consider adding justification for the or other boundaries that are not fire rated modeling of manholes according to the but was credited in the FPRA. (This F&O requirements in SR PP-B7. Consider originated from SR PP-B7) performing walkdown for manholes with significant risk contribution. Also, document walkdowns on all credited, nonrated barriers credited in the FPRA.

2-0A review of thc HRA. report and the reco~very P-RM 813-11 One of the ietfdisc generates Rnon This; FXQ has been closcd.

rule file u'-soed in the FRANG model shows conser.'ati,-e resulits and 1Ae other one A s The action taken to addre6ss thiS item was-that the majoertity of the HEPs ;nd HFE cOnseReatie. The second eXample is likely specifically included in the focused scope combinations were treated properly in the signifieant. Pew-Review PTN firc PRA mod.el.Howe.er,.i.ol.t.d Consider reviewing the recovery rule file for e.ases indi*ate the followng issues: 1. Some consistency against the HRA report HEPs were not applied (9ordocumented) doc'umentation. Also consider updating the properly. For example, EHFPDO-TXT, HPR, combination evaluation.

Failu-re to cress connect u-nit diesel oil storage tanks to extevnd availability of fuel for EDG, was supposed to be set to 1.0 per TableA, 2 in HRRA report page A 19.

However, it s set to 2.3= 2 ORn the recovery r'ue (in both the HRA. repo.t and actual rule file). A Feview' by FPL'ýERIN Staff showed that the example HEP was included in the LIL table (failed for"ascenarios), however, it was also icueinthe Excluded Events table for multiple fire scenarios. Since this HFE ha_ no cue in the mFan cntrol room, i was intended to fail this HEP for all fire scepn~arios. Therefore, the Exclu1ded Emvents table should be updated. 2. A review of the top cutsets in fire sequence 096A show the HFE cembination CHFPSTPRCP and GHFPN!JVLVS has not been considered in Revision 0 Page V-29

Florida Power & Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update the HRA evaluation. FPL'ERIN staff concurred that com~e HE=P combinations may be mis6sed, which render consenyative results. Because fire scenarios for fire comRpaFmnft 0Q96 wJfill be re*fined fr realism, the HEP combinations are expected to be re visited. (This F&Q originated from SR PRM 1311) 2-42 A compariGon of basi-c eavnt between the PRM-B-12 Isolated errOr. Hoqweve, the erro May be Thi; F&OQ has been closed.

i.nternal event (Rev. 7) and fire models ha.s PRM -83 S.g.ifia.. hphe action taken to address this item was been perfgmed.. Fr ....

the basic events Consider providing bases for the modeled specifically incude in the focused scope (about 650 basic events are identified) in fire new basic ev-ents in the fire PRA model. Peer-Review, model, the majority is set to either 0 or 1, Correct AHFPAFWFLO probability in which simu!ates the fire impact in the fire scenarios. The ones; wPith. other values are checked and found to be lacking sufficient basis as requ.red by SR PRM 8313. Fo9Frthe moPd-ified HE=Ps, the majority rseams to be ok by updating with a morEe bounding value ot 1.0. However, the following event is an out*ier, which m result in earl" truncation

.ay of the cutscts with hios AHFPAF=WFLO*=

OPER)ATOR FAILS TO THROTTLE UP AFW F=LOWV, I 17E 4 For the deleted basis even~ts in fire model, all the changes have been traced in the PRM report. (This F&O originated from SR PRM B13) 2-14 Fire compartment 096 is the top conr~ibutor PPRM- Cur rent analysis for top fire scenario 096A isThisF F&O- has been closed.

for U3 CPF. TraGcing the failed basic events / conse,"ative. This issue applies to other fire The action taken to address this item was components /ables shows that the scenarios in the FPRA. specifically incl[ude...d in,thep focused ecope sequeRGce failures seem. to contibute to loss Consider updating the top fire scenarios to Pee Review.,

of reunanie.. Since the cables travel.to remove ctnser-atism related to sequencer rooms housing redundant tFraiR, the modeling and failures.

assumed failure of cables fail the sequencer, which in tur fails the SThgear. Feo example, crcGuit analysis for 3X03 NPO 3MA states, "ADDED CABLES IN SEQUENCER THAT CAN' PREVEiNT SUT FROM Revision 0 Page V-30

Florida Power & Liqht Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update POWER!NG SWGR THRU 3,05 BREAKER." A discussion with FPLERIN staff indicated that the analysis for fire comnpartmnent 96 is not yet complete as docuHmented in the suimmary report. it i anticipated that the approaches and refinements uised for other plant locations will resu-lt n more realistic risk reslits for the room. On the other hand, FPL= staff also identified that there-is-an-open itern associated wAth fire zone 096 (SSA 3GG 13). (This F&O originated from SR PRM 139) 245 Most new events are adtded to the model in PRM 133 Systematic- ssue. This F&O has beeR closed.

order to assess spur'ius operation, and Co.-sider,. -ad4dingto the documentatien Th{e_{rac.tin taken to address this item was other Fire PRA effec*to..Ho .e.ver,there is n whether any events Added to the fire PRA specifically in4c in the 1n focused scope d.ocumentaion supporting the events, and Table D) 1, 2, or 3 of the component Peer-Review.

as such, they do not meet the DA selection report are new to the PRA. If new, requirements As refefrenced in PRM B13. add the details of the event to the This SR lists an exception (DEVELOP a . per the ass.oiated standard

{documentatio defined basis tcim Of non equirements or provide justification to the applicability of ayoftes euieensi PR.A douettonf non applicability ot Section 2), which is not provided. (This F&O originated from SR PRM B13)3 2- RCP Spurious start logico under PRM-19 Isolated iss-e Risk impact is not kno'-wn This F&O has been closed.

3FIRRCPPUR apparsto e inorrct.Consider refining the model for the spuriou The action taken to address this item was, First, the spurious start of an RCP typically start RC. sddrs the f.ire pat speifially included in the focused s.o.pe takes 2 spurious op.rati.. s, iRncuding sta*t of the operator aGtion to trip the RCPs. Pew Reytew.

of the lift pump, and then start of the main pump. Second, the logic, does not inc'lude any consideration of whether the operator already tripped the pumps, and they restarted or whether the RCP received a spuroussigaland then operator trip of the pump. is not possible(may be possible with a single spurious depending OR the design).

Finally, if the RCP restarts (as modeled),

tripping of the RCPs rmay not be possible Revision 0 Page V-31

Florida Power & Light Attachment V - Fire PRA Quality Florida Power & Light Attachment V Fire PRA Quality Table V-3 DISPOSITIC)N OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update depending on the operator action4S performed.

included T-hiS consderation needS to be in the determination of the event for operator trips the RCPs. Under gate 3FIRERCPS6PUR1, 3FIRERCPSPU21R1 (RCP Sea Fail'.re Following Spurious Pump Start Due to Fire) is AND'ed YAth the HEP CHF=PSTPRCP, which is not failed by any fire scenario. This event is evaluated in the HRA report, but the fire impast for an rpeFatorto trip the RC"Ps is not evaluated, (This F&O originated from SR PRM B9) 2-48 PTN LERF model development is Systematic ssu.. This F&O has been closed.

documented in the Component and Cable Identify and documnent any new accident The action taken to address this itmwas Sel*etion RepoIt 0493060006.01. Rev. 1, nr ion no~'onrl t~ nn~t "t '--~ro , GG;,*9;9

G&

Section 4.5, LARGE EARLY RELEASE damage that would be applicable to the Fire speG GllyIFIe U ... 1.;16 ;

FREQUENCY (LERF). This section PRM 91 PRA that were not add reseed for LERF described the excl-udd LERF sequences On PRM 144 estimation in the Intern internsa events model. No dicc-usio* is dent aroare identifi*d on Any new accidnt progressio sequence. resulting in a !arge early Fela. e beyond the onset Of core damage that would to determine if repair of equipment can be be applicable to the Fire PRA that were not credited. JUSTIFY credit given for repai addressed for LERF estimation in the (i.e., ensure that plant conditions do not Internal Events

.......... 6 PRA.

- 11=... D As. a result,

. . . significant

,1 ,,.;n fc n preclude repair and actuarial data exists a*niden progresotn sequences re sultintg in from which to- estimorate the repair failure a large early relearse have- not been probability [see SY A24, DA G15, and DA reviewed to deter* m if repair of equipment

,'e P9]). INCLU1-DE acci~denpt sequence can be credited for the FPRA LERF models dependencies in the accident progression

[LE G3 GCC , requirement]. In additi9R, sequence ina ma nsistent with the accident sequence dependencies in the applicable requiFrements of 2 2.2, as-accident progression sequences have not incl-ded in a manner consistent with the analysis [LE C, requirement].

applicable requiremnents of 2 2.2, as appropriate for the level of detail of the analysis [LE C8 requremnent]. (This F&C-originated from SR PRM 611)

Page V-32 Revision 0 Page V-32

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 249 The PTN firs LERF model is built upon a LE Draft interna event LERF madel is used. This F&O has been closed.

draft !nternal oen-R,'i ts LERF model (rev. 8), -A2- Changes te the draft model vwhen finalized The action taken to address this item was wAhich needFs to be Updated when that model coul1 be significan.

-epefially s. included in the focused scope isfinRalized. (This F&O originated from SR Update the fire LERF mopdel when the Pe&.Rey.ew.

RM34) internal event LERF model s finalized.

2-22 Beyond the Ceneic Fire Modeling FSS-G6 Without detailed fire mo.deling for significant This F&O- has been closed.

Treatm'ents, the 'Fi r PRPA'"did net ide F'2-D2 fire sGenarios, the resu,ls arc cnservati e. The action tak* n to address this itemn was maddtionadetaile fTire m&odeling oaote fire FSS-G4 Consider performing add~itinal detailed fire spcfclyicued in the focused scope copatmnt. oignaedfrmmodeling ThsF& for target dlamage tim~ing when -tho Pe&_pReAsy0 eq SR FSS C6) exposure environment exceeds the damage threshold.

2-26 The system unavailability records for the FSS-P7 Systematic issue. The intent for Capability Th-s FAO has ** p.n closed.

plant have net been reviewed in crediting CRategory ....is to additionally

. require a review The action taken to address this item was fire. detection, a suppression syste.. .ofplant records to dPt*fr*mie if the generic, specifically includd in th. focused scope (This F&O originated from SR FSS D7) unavailabilitycredit is consistent with actu.al PeeFiew system un'availability. Outlier .experienc

.ou.d be any experience indiating that actual system is unavailable more frequently than woud* be indicated by the generic values.

Consider performing and documnent the review of plant records to deteFrmine if the generi un~availability crFedit is consistent with actual system unavailability. Outi!er experience would be any experience inRdicating that actulal system is unav~ailable mor~e frequently than would be indicated by the generic values.

2-PTiN FS repo,"t 0493060006.004 ReyiFon FSS-4-0 Documentation iSsue. However, the This FR,(O has been clod.-

1, App. A documents the SCENARIO RSS- i...nadequacy of the walkdown documentation The action taken to address this item was

SUMMARY

REPORT, which includes the F98.... cannot provide detailed infor n f speci.f.ical included in the focused scope combinations, of fire sources and target sets. -s-enRario development or detailed fire Re e'w However, the walkdewn documentation is, medeling 7 lacking. Currently, the walkdOWn results are Con~sider enhancing the process and documFented directly in the FRANC documnentation of the source target-data Revision 0 Page V-33

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update database. Consider a walkdown comllectfion w~alkdowns. See recommnended documentation package whiGh would include.. alkdown NUREG/'CR 6850.

elements such as a data collectein procedure, docu.cn.tation of who perform.ed what walkdowns on what dates,-

dec'Amentation that review of the collected sourc~e target data was performed, etc. (This F&O originated from SR FSS DlO) 2-34 The PTN fire PRA model hac, not Gcompleted F.95 H6 Quantitative resu ls for the i e F a the quantitative result for any s significant fire se.narios in the multi The action taken to address this item was analyzed quantttvl 1 ina maneA that copdet analysis should be generated specifically included in the foeu~sed sco~pe facilitates FiFe PRA applications, upgrades, an docuMente d. Peer Review.

and peer review. (This F&O originated from Quantify the identified significant fire SR FS9 H8) scenarios in the multi comparment analysis and Gemplete dIasAumentfatio.ý-

The fire PRA has not identified how the A Requirement not met This F&O has been closed.

physical characteFimsrtF (unique to fire PRM Identi o w.the physical characteristics The action taken to address this item was scenarios,.or.affected by fire scenarios) (unique to fir, affected by fire spciialy.nluded in the focussed hscpe identified in LE A! and the accident ...... ___ idp_

`_ _ In AI- Rd -eet--Re,-

6eque~e Ga~a~eris'Gs E A2accident sequence characteristic~s iden tified are addressed in the LERF analysis. FPRA in L A ar addressed in the LERF im~pact can affect accident sequences foraalss LE=R.F such. AS failing containment ioain affecting containmriet coling or instrument air to con.tainment, opening pressurizer PORVs (pressurizing containment), etc.

(This F&O originated from SR LE A3) 2-38 The fire PRA did not re visited the plant LE-A5 Requirement not met. This F&O has been closed.

damage states defined in the internal events PRM-8 Re define the plant damage states in the The action taken to address this item wars LERF model. (This F&O originated from SR internal events*LERF model to account for specifically included in the focused scope any fieseifccaaceitc PeeF Reyew-2-39 The MSO Review List reviewed by the LE4 ytemati essue, This F&O has been closed.

Turkey Po.int Expe. Panel should have PRM131 Use a system.atic process and document the The anction taken to add.rs this item was.

captured meet of the requirement in this identification of the fire specific LERE specifically included in the focused scope 3R.

However, a cystematis process and contributors froM the set identified inAMSE Peer-Review,-

Revision 0 Page V-34

Florida Power & Light Aftachment V -Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update documnentation are not available with respect standard Table 2 2.6 9..

to the identificatio-n o-f LIERF conRtributors from thc spt id*nt ifiqd in AMSRE standard Table. 2 2.8 9. (This F&O originated from SR LE= 91) 2-4 The FRANCImodel itself generate

.annot.RM A3 This finding mainly focuses On This F4O has been cloe*,d total CDF , LERF= and risk ifane dumentatin. However, the unavailability The actin taken to address this was m~easures. The moqdels show th of the- cu,_tset files6 and ris~k importance seiial nlddi h eue cp contributions from eac~h quantified fire reports prevents the detailed analysis in perRPlwv sequences. As shown in the summary repor other tasksVI such asN ad circuit 0193060006, Rcv 1, the Unit 3 DFPtop aealysietc-.

sutsets arc listed. However, the UItI4 CDP, Document outset files for U312-L-ERP and* U4 Unit I CDF R, LERF, the risk impertance CD/LERF and risk impor"tance reports for values. for cash barsi events arc t eano.

inc-luded. (This F&O originated from SR PRM-A3) 2-40 A separate FPRA LERF pakage is not L G2 LE C-2 CC-- requirement This F&O has been closed.

available. SR LE C2 is assigned as not met PRM 1115 INCLUDE REALISTIC "treatmentof feasible The action taken to address this item w;as sin~e the LERF specific operator actions is operator actions for LERF Specific HEPs (it specifically inclu-ded in the focused scope not evident in the ,RA repert and the screen any) foloWiRg the onset o.f core damage Peer-Reyoeyw values; are used for num.eros operator CONSISTENT WITH APPLICABLE recovery aGcons fo.r the fire.indced PROCEDURES.

component failures. (This F&O originated from SR LE C24 2 FPRA specific significant accien LE-Q4 SR LE 0:1/5/6 Requirements This F&O has been closed.

progression sequences resulting in a large I F-05 For FPRA LEFRF m~odel, evaluate the fire- The ;action taken to address this. itemR Was early PPR.^ release LEr-RF=have not ch@nges been are developed.

di;ect~y Thespcfcapcsfrteflongsciclyicuddnthfcsdsoe LE G6" .................... .... GG~ed G~

he

... FP.RA R chanes q.arciretly requirements in SRs LE C4 through Peer-Re-iew, incorporated4 in the quantification.faut~t~r~eeC-6:CLUDE mnodel logic necessary to FPRA specific realisti generic or plant avdaFelrtGetnfe oth seific analyses for systemn m success - criteria poiearaitc specic., .. s ssignificant

- - accident siainoh progression sequences for the significant accident progression resulting in a large early release. INCLUDE sequences have not been developed. The mitigating actioGnr by operating staff, effect of FPP. LERF changes arc directl, fission product scrubbing on radionuJclide incorporated inthe quantification fault tree, release, and expecated benAefic~ial failures in FPRA specific system models that support significant t progression sequence.

the accident progression analysis have not PROVIDE technical Revision 0 Page V-35

Florida Power& Light Attachment V - Fire PRA Quality Flrd oe ih AtahetV-Fr R ult Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update been developed. The FPRA LER Fc hanges specific or applicable generic Ga*lulations arc diFeGtly

• ncorporated in the q uariifGatir-_ deenrG ating the feasibility of the ationi 5T fault tFre. (This F&O originated fr em-SR LEk scrubbing mnechanisms, or beneaficial f G&) raiiues) supporting tne i ot any o

4 11FP%;

genr8ic Or plant specifi analycee for success criteria for the significant ace progression sequences. USE conser:

Or*a co*mbnto . of .- l-Rative and realistic rsystem success criteria for ni He, K significant accident progressioneq

-'V-LL. syste* .m.aes tat suppo.t iRe accident progression analysis inamn Gconsistent With the applicable rcquiremnERtS for 2 2.1, as appropriate for the level At detail of the analysis.

2-43 The significa*t accident pr*g*rge*si LbE-Gig LE C! 0/12 CC 11requirement This F&O has been closed.

sequences resulting ina !arge early release LE G12 E*i;l" Ga i aGi.i*i~ien p ; ;mwess.,lo;; The ction taken n toaddressi th is itemn was have not been review=-ed to determine it 6e specifically included in the fiDOWSLed GGepe engineering analyses Gan suppedt continued PRM 1316 in a iarge eany release I- if 8Rnq ceorCin analyses can lper Repew, equipment operation or eperator actions support continued equipment operation during accident progression that could operator actiens during accident progre aR reduce LERF. The significant accident that could reduce LERF. USE consern:

pFgqFess 9R 69%Re-ear 1 I recsave not seen 1wpn40 rnon significant iete rmine it engineering a t÷ seouensers. RFý-EVIEW C

signifia ntaccident progresinseqec operator actions after containment failure resulting in a large early release to that coulwd redu ce LERF. (This F&O -IT- -AnIA~vsr-ARP originated from SR LE C--) l" l - ,

zontnuc equipmentH RMOR 0 actionsg after ongtainn 4k.-..! .-. .2.4-..4A...-.i1nr iio A r-A, - M - -r-A Wevaiiws Ot GonseRwati rell ;aqt accident progression sequences 2-44 Uncertainty Evaluations (Sensitivity studies) QU-E4 QU-E4 requirements. This F&O has been resolved.

should be performed for both CDF and UNC-A1 Perform sensitivity studies should be Parametric uncertainty and sensitivity has LERF model for Units 3 and 4 since the Revision 0 Page V-36

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update model uncertainties may have different UNC-A2 performed for both CDF and LERF model for been performed for CDF and LERF for both impact to specific model due to differences Units 3 and 4. Units. The results do not indicate any in plant designs, FPRA model details, and change in the selection of parameters or etc. (This F&O originated from SR QU-E4) assumptions are necessary.

r, f rnr + ~r iivr + Iriol, Tki, C 9fnk-kaa,,i .Ad for risk Significant human action, including HRA-G! sngifiea*tHFF=F 4-Hit JL A. IjRKi both thc new fire specific cafe shutdown moedetail, as specified in HR/k C;.Th sp G~alredu in the focused scooe HRA-02 actions identified and the non fire actions PeeF Review-that exist in the internal evcnts PRA and HRA E1 risk significant, new fire related operator included in the Fire PRA. it is noted that the actions should use an approach that HR G!

HEPc derived by the mu~ltiplier appreach forF addresses both failure in cogAition as Well the non fire actions are still considered as as faluE)t execute.

the screening values until a detailed analysis HR-G2 is performed. Update the HRA by performed detailed HRA There are numerous HR 11 for all risk significant HEPs, icungthe examples of signifi*c*t H"Ps in the results, HEPs analyzed us'ilng the multplie method as Well as signififant crieening Hr-Ps set to in the H/, as Well as the screening HE'Ps 0.1 in the altered eventc repot. The PRM-911 inthe altered event report. The HR/k.

estimation of the HEPs for the new, fire QU-04i approach should be revised to treat the related humnan actions did not accounIt for simplified approacw h for the non fire actions the fire scenario specific performance as a screenRng tool, a*d provide detailed shaping factors (PSF~s), including tm HRA_for significant HEPs. Additionally, even complete action), etC. The formal avammMm to_ilable the screeRing resus should be reviewed for HRA Of significant HEMPIS includes- eahfire scenario where the HEP is applied refererncinG and considering pr.cedures to including cOnsideratio Of ti~ming, lost performn actions, assessment of the impacts indfications, spurious operations ionthe instrumentation needed for cues and scenario) and other effects on the tiFming for execu1tion and resulting impacts to the HPE, feasibility, timin'g of the event, performance shaping factors, evaluation Of cognitive and executiRn error probabilities, scenario specific equipment imnpacts that mnay affect the timing of the human interaction, as well as considerations Of workload (for input to the dependency impact evaluation). Overall, it appears as if the Fire PRA treats the screening results of the method used for the non-fire hu-man actions as detailed results.

However, as iniae nthe diagram shown 9a iG a t6urupiiurcu&

Page V-37 Revision 00 Page V-37

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Fethod, and does not meet the requirements of a detailed HRA per the standard. Altheugh the original HEPs in the internal' events PRA. included all of the re-levant HRA factors, the fire specific HEPs woul,-,d have to consider the imnpact Of fire On these farctors n developing the HEP results.

The e-nd rslisan HRA that provides very general results, without specific application to a fire compa*tment or scenario, and -a ack of-detailed HRA for significant fire areas.

(This F&O originated from SR HR,\ CG) 2--7-- PTN systemn mo.del changes in the fire PRA PRM-B9 Systematic iSsue ytemn models were n This F&O") has clased.

.een models are sum.marized in Tables D 1, D 2 SY A!! updated according to the SY A and SY B The action taken to address this itemn was and D 3 of the Compoent and able. specifially in n the focus.ed scop Selection Report 0493060006.001, Revision Se(A;4P-Uifedet&~ h ytP AAR19ad ppR

'1.I TII*r 1, Appendix D, MULTIPLE Apenix*,r

,1*1 el*.IDirl SPURIOUS .. ,e S---A1&their SYA!5 associated Constier docu-me updating ntationmodel the syste according and to OPERATIONS EXPERT PANEL EVIEv 8-Y--4-Y Avan SY BvSR requirement..

AND DISPOSITION OF OPEN ITEMS. SY-A2 However, no additional documentation of the changes for PTIN fire PRA has been SY A23 provided. Requirements under SY A and SY-A3 SY B are not met as a;esult. In particular, S-Y--A4 the requirements ofSY A2, 3, '6,, 11,14 15, 17, and 23, as well as SY B5 through B., do not appear to be met based on a SY-Big reviewmp of the documentation in the tables.

Since the level Of analysis andSY12 doc-mentation for system models for Fire PRA is expected to be similar to that S-Y-4 performed for internal events, the 1Y--44 documentation in the Appendix D tables Y131 does not mneet what is expected for this requirement. For example: on Table D 3, page ID57, there appears to be a number of SY-E86 changes to 'correct' l0oG. The original logic SY137 is included in the system notebooks (e.g.,

Dr- power notebook). However, since this is just corrected here, the internal eventS PRA , ,

Revision 0 Page V-38

Florida Power & Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update is not updated. Additionally, without proper development, it is impossible to determine i the change isacuae In the DCG power ease, a reference to the power drawing is neeAdedi to-oneuroF the logic is now correct.

On Tablo D 2. itemp 23, there arc -alot o-f logic changes. However, there r.sno reason as to why the lo9gi changes are needed or why the changes are mnade as written. In T-ables -D2 and D 3, numerous references have-been made to the comments/

recommendations from site engineers, which add insights but shudnot be9 censidered as-the sole modeling basis. (This F&Q originated from SR PRM Be) 3-10 The PTN Fire PRA ugsed much lo'w-er HRR F-99-4 Transient fire ePvaluations conducted as This F&O has been closed.

for the evaluatien of transient fires. The HRR FS-D§ described in the Fire Scenario Repot result The acticn taken to address this item was for transient fires is based on the in screening fire damage to targets due to specifically included in the focused scope NUREG!CR 6550 HRR for eectric motor FSS-G4 the lower HRR which is believed to be non fires (See F&O 3 10). The Fire Scenari coGnservative for developed fires Vinvolvig Report (Report 0493060006.004) indicates ordinary comnbustible fuel packages such as that a tFranient NPP tFraRnsient fire is bette a trash cRn or trash bag.

represented by a temporary cabl Use the NURFG=G'CR 6850 HRR for transie:n installation,~~~~~

nldsa giinfres whc Or provide alternate justification for an seurce. Based en this the Fire Scenario area specific HRR based on the limiting fire Report indicates that the electric m*o*t HRR that co.uld occur ithin the area.

is used to describe transient fires.

A review of the EPRI Fire D13 Of transient fires indicatedA the followin~g: a) events in the DR3 indicated that the fire Was either above 75 kw or could have bheen. above 75 kw, if not suppressed. For example, one transient fire reule iMn an automatic suppression systemR actuation, which was likely above 75 kw due to the sprinkler head being above 160 F=as a reulof the firo, b) a r8eet event At one ot the peer review team members plant was above 75 kw. As a result of this review, and discussionF;F;0-.

ýamongst the peer review tea~m Revision 0 Page V-39

Florida Power & Light Attachment V - Fire PRA Qualitv Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update members, the HRR for transient fires does not appear to be substantlated for the PTN FPRA. (This F&O originated from SR FSS9 D6) 3-1-1 The screening criteria is defined inthe FSS G2 NU-RE-G 650 Section 11.5.4.3 Step 3.c.1 This F&O hoc been cosged.

TuJrkey Point Hot Gas Layer and Multi recommends development of a conse,'--ative The action taken to address this item was Compartment Analysis, (Report HRR based OR a com~bination Of ignition specifically included in the focused scope H0Q-4,3060006.006) methedology. coure and secondary combustibles that Peer-Review-Compartments that don't screen are retained produce the highest HRR. This for further analysis. A soncern identified with recommendation is provided to ensure thata the screening criteria involves the use ef a conser.ative/realistic- HRR is u4sed to standar~d fire scenario for each analysis- determine the potential for HGL formnation.

rather thin determining the most challenging Iseof A no.n cns.. .tive HRR may lead to fire scenario inherent to the analyzed underestimation of the potential for HGL-c~ompartm~ent. This appro~ach potentially formation and accordingly sprea.d- Of masks the potential forF forming an HGL i potentially damaging hot gas to adjacent the exposing compartment. eempartmeR&.

(This F&O originated from SR FSS C2) A realistic HRR should be developed for each fire compartment to ensure that the potential for formation of an HGL is appropriately assessed on a compartnmet by compartment basis.

3,43 Reviw of fire modeling in enclosed FSS-C1 Systematic application of generic fire This F&O has been closed.

compartments does not appear to accurately FS& G!4 modeling results to define target damage The action taken to address this item was nsider the addition of HRR from. when fire spread should be postulated. speihfically iuen thde fosted sope secondary combu-stibles. For instance a In sitatinons te*re the generic firo modeling

, P-ReYiew.

.. ithge. cubicle.fire.locted.in.fire.zonetreatments demonstrate fire spread to 71 is estimated to darmage trgets above the s.,onday combustibles, the scenario cubile4' horizontally and 7' vertically. The should ass-u-me full area damage. These sihgea ubicle are vented at the top so seaisaecniae o ealdfr a fire in these cabinets can be expected to modelng that woulA'd be uswed to show-fre spread to the cable trays abeoe. The cable damolge thPRA twodets aud the tshoe tf trays would be expected to ignite within the damage to.PRg Fretd t tim*- ro plume of the cabinet which is e-stima;t*d to damac be at least 3' wide. Gi-en that An initial width of 3' is reasonable, fire spread and additional HRR due to the resulting cable tray fire only assumes 1' Of fire spread along Revision 0 Page V-40

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update the cable tray. Comnpared to the NUREG/CR 6850 gulidance for flame spread along PVC cable (flame spread - 0.9 tmmeece) the estirmation of HRR for these typi!al scerioase c nono nSRSative.

Realistic estimation of the scearioF; HRR i nccessary to ensure the full iFmpact of the fire en exposed targets is presented and that the effects of a damaging HOL may also be estowaed.

(This F&O-oFiginated ferom SR FSSC) 3-44 According to the generic limitatn FSSG Systematic applicatin*' f generic fire This F&O has beenR losed.

Eontinred in Attachment t B of the Fire modeling results to define target damage The tarken t address this i was Senarioe Repor (Repod 0493060006.004) when fire spread should be postulated, specifically included nthe foused scope the generic fire modeling teatments do not The geric teatments used in relatively PeeF-Review, acco-Aunt fo4-rthe effects Of hot gas layer small roomsp shouldi be scArutinized to enpsure (HGL) on the correlations presented. The that an" HGL itrconoiscnsid-ered-and liFmit1atio indicates that because H4GL is not accounted foWf ondt be sigefisant GOA idered that these corltinhould not be usred in enclosed areas with small vo1rlumeswhere a significant HOLthickness may form. Bca3ue'J this relationship is net considered plume temperatures may be underestimnated because it is assumned- that amnbient temperFature air is being entrained inRto the plume, resulting incooler plumee temperatures, rather than heated air from the into hot gas layer. EnRtrainment of heated air tho fire plumIe resultsIn higher damage heights because the plume remains hotter at higher elevations. (This F&O originated from SR FSS G!)

3-2 Credit for fire compartment separation via PP-B1 As noted in the description non-fire rated This F&O has been resolved.

non-rated construction was commonly PP-82 construction is credited for separation of fire The configuration and construction of non-noted, e.g., according to the FHA the walls compartments, however no Fire PRA fire rated barriers was confirmed using a of fire compartment 034 are not fire rated specific justification for the validity of the fire combination of information in the Fire and they provide separation from fire compartment is provided. This is considered Hazards Analysis and supplemental plant compartments 036, 035, & 058. Separation a systematic issue for the FPRA. walkdowns. The anlaysis and related Revision 0 Page V-41

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update of FC 034 from the surrounding FCs is one Provide FPRA specific justification of the documentation was updated to provide this of many examples where non-fire rated construction separating fire compartments, information.

construction is credited for separation. Use where the barriers will substantially contain of this level of separation is acceptable the fire.

provided the separation is justified.

However, the justification does not appear to be provided for the FPRA. (This F&O originated from SR PP-B2) 3-3 A few cases of special separation are PP-B1 Two instances were identified where spatial This F&O has been resolved.

credited in the PB&P. Most notable are PP-B3 separation is credited for the separation of Openings between fire zones were separation of Fire Compartments 058 and fire compartments. No justification is addressed with respect to targets on the 037 and 004 and 010. The FHA notes in the provided for this separation. other side of an opening which are within the write-up for fire zone 004: 'There is a partial Provide justification for the use of spatial zone of influence of an ignition source.

height concrete wall on the South side of separation in the FPRA. If not justified, Targets were evaluated for fire damage this room with a full height opening to Fire combine the compartments in the FPRA. regardless of the zone in which they were Zone 10'. No justification is provided for this located.

separation, hence it is not clear that the The multi-compartment analysis considered credited separation may be expected to the mumpatent anacent zons contain the effects of a fire. Accordingly the the volume associated with adjacent zones effect of a fire beyond the identified fire with openings between the zones in compartment boundary may occur. While evaluating the potential for hot gas layer this effect would be expected to be identified formation.

through performance of the multicompartment analysis the level of documentation provided in support of the PB&P does not satisfy the standard requirements. (This F&O originated from SR PP-83) 3-4 The PTN self assessment points out that the PP-B1 As discussed in the description This F&O has been resolved.

FHA documents the use of active fire barrier PP-B5 justification/discussion is not provided for The walkdowns that were performed did not features as necessary for fire zone crediting active fire protection features in observe any open fire doors (active separation. However in cases where fire barriers that are identified as non-fire rated features). The documentation for the fire compartment separation is provided by structures. It is not clear if active features scenario development process was updated unrated barriers there may be active such as fire dampers exist in these barrier to provide the criteria and methodology that features that are not identified by the FHA segments because the FHA does not rely on were used.

but credited by the Fire PRA. In such cases them for separation. Documentation should active fire barrier features may be be provided that clearly establishes what unknowingly credited for separation but not features are credited in such barrier Revision 0 Page V-42

Florida Power & Light Aftachment V - Fire PRA Quality Flrdoe iqtAtcmn FiePA - ult Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update adequately maintained by the fire protection segments and why makes them acceptable.

program. Because these elements were not Given the large number of barriers credited purposely identified within the development in the FPRA that are discussed in the FHA, of the Fire PRA it is unknown if the Fire but without discussion of active elements, Protection Program identifies all of the there are likely a number of undocumented necessary features. Because the Fire PRA active elements in these barriers.

does not formally define and justify these Determine the active fire barriers on barriers features this element is judged not met. credited in the FHA (not SSA), and provide (This F&O originated from SR PP-B5) justification for any active elements credited in the FPRA.

3-5 According to the Section 3.13 of the PTN SF-Al As discussed in the description no This F&O has been resolved.

FPRA Summary Report the effect of an discussion was found that specifically The low seismic spectra applicable to the earthquake on ignition source scenarios is addresses fire ignition source scenarios that Turkey Point site have been validated via discussed in the IPEEE and Potential Fire may arise from an earthquake. Also, since the IPEEE with respect to the potential for Related Vulnerabilities self assessment, these scenarios are not identified a causing unique fire scenarios. Their Review of the Potential Fire Related qualitative assessment of their risk potential for causing damage to pipes or Vulnerabilities self assessment did not significance is not included, tanks containing combustible gases or reveal an analysis that specifically The analysis provided in the Potential Fire liquids or to initiation of electrical fires is addresses generation of fire ignition source Related Vulnerabilities self assessment considered negligible.

scenarios which could result from an should be expanded to look for unique earthquake, nor does this assessment ignition source scenarios that may arise from address the potential risk significance of an earthquake and a discussion of the risk these scenarios. This assessment does significance of these scenarios should be identify fire vulnerabilities in terms of fuels, qualitatively assessed.

ignition sources, and oxidizers however these discussions are not specific to seismic events nor do they include evaluation of special ignition scenarios that may arise from an earthquake. (This F&O originated from SR SF-Al) 3-7 According to report PTN-PSA-7.01 The IGN-Al As discussed in the description the revised This F&O has been resolved.

generic fire ignition frequencies provided in IGN-B4 generic fire frequencies contained in FAQ The guidance provided in FAQ 08-0048 NUREG/CR-6850 were used to establish the 08-048 are not incorporated into the PTN requires the use of the original fire ignition frequencies for PTN. While the fire frequencies nor is there justification for NUREG/CR_6850 fire frequency values as a use of these values is not entirely incorrect, their exclusion. This SR requires use of the sensitivity study. Rather than perform two this SR requires the use of'current nuclear current nuclear power industry event history analyses, the PTN analysis was developed power industry event history that includes or justification for data exclusion. Because using those original values for the NFPA 805 power plants of similar type, characteristics, the fire ignition frequency methodology does Revision 0 Page V-43

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update and vintage.' Accordingly this requirement not address the data contained in FAQ 08- application.

requires use of the EPRI revised generic fire 048 this SR is considered not met. Use of The application of the non-segregated bus frequency values included in FAQ 08-048 or the NUREG/CR-6850 values results in a duct information from FAQ 07-0035 is not justification for its exclusion. Also, it appears conservative estimate of CDF/LERF. FAQ applicable as the plant does not use non-that FAQs 07-35 (bus ducts) and 08-44 35 can have significant impact on fires in the segregated bus duct. The connections to (MFW pump fires) were not incorporated into area of bus ducts. However, it is not the station transformers are made using the FPRA. (This F&O originated from SR apparent if this is important for Turkey Point. cables. FAQ 08-0044 was also not needed IGN-A1) FAQ 44 can result in a lower MFW large fire and the conservatism associated with frequency. original method did not adversely affect the The fire ignition frequency information results.

contained in FAQ 08-048 should be incorporated into the PTN fire ignition frequencies. Additional FAQs should also be incorporated into the FPRA.

3-8 Review of the plant-specific fire events for IGN-A4 As discussed in the description review of the This F&O has been resolved.

outlier experience indicates that some fires identified in Appendix A reveals fires events may have been considered outliers that may have become challenging had they The scope of plant specific fire events were or unknown if the selection criteria had not been extinguished early. The selection Te scope of an spe d gre of considered treatment of fires that are criteria for challenging fires contained in re-assessed with an expanded group of extinguished prior to full development as Appendix A is based on section C.3.3.1 of plant personnel with particular focus on the potentially challenging. Several cases NUREG/CR 6850, however the criteria subjective criteria from C.3.3.2. The results identified in Appendix A of the Fire Ignition contained in C.3.3.2 is not included; had the of the re-assessment affirmed the previous Frequency Development Report, PTN-PSA- criteria of C.3.3.2 been included more fires dispositions.

7.01 may have developed into challenging may have been selected as challenging or fires had they not been discovered and identified as unknown.

extinguished early in their development.

Fires 7, 8, 9, 21, 22, 27, 30, 31 appear to be The criteria for selecting challenging fires in potentially challenging fires (or unknown). Appendix A of the Fire Ignition Frequency See also the previous assessment from 9- Development Report, PTN-PSA-7.01 should

09. (This F&O originated from SR IGN-A4 be revised to include the criteria contained in C.3.3.2 of CR/NUREG 6850 and the fire events should be revisited to determine if additional fires should be selected.

One Gituatien was identified for which credit F-99G8 This finding it based on identification of This F&O has been closed.

of a firc wrap is propesed. The FRAN'C FSS-G credit for a wrp in the FRANC Exclu'ded The action taken to address this item was Evxchluded Evnts Table, Aftachment F of the Evcns Table, Alchmcnt F of the Firp ire Scenario Report, (Repert Scena r a Report, (Report 04.93060006.004) 049306900096.004) indicates that a fire wrap Revision 0 Page V-44

Florida Power & Light Aftachment V - Fire PRA Quality Florida Power & Light Attachment V Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Will be credited in Fire Zone 071. This fire Credit fo^.r the proposed fire wrap should be wrap proecL-ts a-n MCCG 23B c-able and is bein*g addresse in thc wrap integrity should be credited to exclude basic ev-ent 3B0n*. N e-stablshed, with respect to fire resistance, technical basic for the fire resista ee-rating mpechahnical p*rtection, and potential fire of this wrap was found in the FPFRA nO related e)xposur.....eto "i;sh the wrap may be fhnr- i ",+ifirn+in #nr r'rar0fon +k, r f flC  ; mp ; N!g-=N +

I A assumng mehanical damage, direct flame etG+/-

impingement oF HEAF. Accordingly this SR is;consfideredý not met. During the walkdcwn, Thermo lag seen throughout the plant.

.as Thermo lag has had problems inthe past, and the rating would need to be justified prior to credit. (This F&O originated from SR FSS G8) 4--0 The HRR uscd for fire modeling of the zone FSS G4 Severity factor ir l-red extenni'vel" in thk F4re This F&O has been closed.

of influence is based on mnotor fires, which is PRA~ The action takep toIn addre.. t~his 4i4 was-substantially lwrthan the NUREG/OR Use the NUREG!CR 6850 HRR fo-r tran-sient specifically icl-uded in the esused 6Gepe 6850 recommended HRR of 317 k'P. As a F22-D2 fires, or develop an accepted industry HRR PeerReywew:,

result, tho use of severity factor could, approach (presently being discussed by pteRti*ally be double ciontng the loweAred EPRI). Develop transient fire severity factors HRR for transient fires (note that even when based on the likely HRR and locnation of eoverhead cables or lcoat!on of equipment.

Ieee nent factor for transient fires, Fo exml, if cable is 7 feet overhead, the it is dependent on the HRR in terms of the severity factor would be based On the zone of influence), if the severity factor minimum HRR that would damage the cable develOp*en*t is beoin the NRlIrEGIOR -1. 11-.4 J:- . . A - 1*:I:----I. $1....

- I**:--

taRGe

• L:I* '4[G',*;:.;I L:;T;',.¢ 6850 HRR for transient fires. The severity can be -sed in determining non suppression factor for transient fires discussed in Section time for generic cases, based on the latest 8.4 of the fire scenario repor FAQ 52. Finally, it is recommended that the (049306* 06.00n) doeS not provide current conditional probabilities- in 7.1.2 ta-;b'le sufficient justification for a generic transient for electrical cabinets should not be used in fiFe severity fa*tOr. In add4iton, the severity the-FPRA4 factor deriv-ed from an aRnalyss of the number of fire events includes non suppression results, and would therefore not be independent of any non suppression probabilities applied later. FSS C4 requires severt f eos t be independent of othe factors. It is noted, however, that the Page V-45 Revision 0 Page V-45

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update implementation of the severity faGcEt in the PTN FRANC model did not involve the suppression factor in the same scenario.

Firo severity factor as disu,-ssed in Section 7.1.2 for lwvoltage electrcal c.abinets, i not developed or applied consistently with the NUREG/CR 69850 methods. This is developed from a supplemental repodt (ERIN report, Supplemerntal Fire PRA Methods).

Additionally, fi e propagation ou.tsid of the, electrical cabinets is.also dependent on the non suppression probability. Therefre, some dependency exists in this data if Used i owitha nn suppr.essin factor.

The nm-bers listed in 7.1.2 for electrical caeines were de.rived using the to tal number of cabimnet fires inthe denominator, rather than the number of fires of the specific DaRel tvye. Due to this incorrect derivatien of the conditiona;l probabilities for fire

+

e conditional pro-babilities8 thusR developed (and applied in the FRA NC modlel 01*;4 for low voltage cabinets) could potentially* prpgf be low by an order of m~agnitude a ;A(non conservative). For

+*

both the transienAt h firesF- and- low-voltage

..+;

cabinet. . fires, the severity factors are basically developed usinig fire events data from the E=PRI ropedt. Given the fire data duration and damage is a result of multiple factors (growlh, suppression, severity, location, ete), and given the fire data often does Rot have sufficient infonrmation to make aR reason-nable deterrmination of either the fire size or whether a fire propagated outside the cabinet, the peerM reviw teamR deterFmined tha;t the -useof firee events da;ta-for developing the above severity factors is not allm"49te-Revision 0 Page V-46

Florida Power & Light Affachment V -Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update (This F&O originated from SR FSS C4) 4-1r The aregment in Sti 8.1-of the fire F22 Q4 The l0o4atio/everity factor applied should Thic FR&O has ben cd.

scenario report about the application of the F99D3 have already accounted for the prbability The action taken to address this item was probability of an ignition source b that the target located within the impact specifically included inthe focused scope located within an area around the target may area around the transient fuel aFnd an PeeF-Reyiew.

not be co.rect. This is becauso the FSS G4 ignition source is located within an.are in of-a2loain atr oh application-- the vicinity to ignite the transient fuel.

transient fuel to the fire ignition frequency Revisep Section 8.1 and do not apply has already accounted for the probability of .. prat ato* n fac*t*rs for transient fu.el the target being within the influence zone e.soreof and ignition simultaneouslyin any a-fire. As such, ignition isa given condition. fr ~nFe With the apportioned frequ.ency, the target must be lcated within the impact area areund the transient fuel and an ignition source must be lcoated within an area in the vicinity to ignite the transient fuel. (This F&Q originated from SR FSS C4) 124 Exep orte O fire scenarios, no6 FSS3 Mean values and uncertainty itrasforF This; FRO has been clo9sed.

,scenario specific fire modeling has been F-8844 the parameters us... ed for mo.d.eling the The action taken to address this item was, pcdowmed to provide a mean value of, and N:A significant firo scenarios have not been specifically included inthe focGused-scope statistical representation Of, the uncertafinty provided- PeeF Review.

inqter-als for the parameters u-sed for Consider developing mean vale and modeling the compatments with significant u.ncertainty intevals for the parameters used fi.re ri.sk.onriuto. Therefore.. for modeling the significant fire scenarios.

compartments ether than MCR, only the results Of onservative, generic fi mo ,ndeling developed in the Generic Fire Modeling Treatments repeot were applied to t#e a.nalysis of fire scenarios. (This F&O originated from SR FSS E3) 4-1-3 Uncertainties associated with cases where F98-E4 The required uncertainty has not been This F&O has been closed.

Gable routing has been assumed (e.g., the F. evaluated performed. The action taken to address this item was E-XCLUDEDEVENTS table has assumedn the uncertainies specifically iclu1ded in th* f6ocued scope tteaclaeooeto associated with cases with assulmed cable PeeF Reyiew, selected areas based on ':agkdow:n or engineering judgment) have not been ivestigated w-th a decumented basis. (This Revision 0 Page V-47

Florida Power & Liqht Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update F&O originated from SR FSS E4) 444 T Ht.hG@6 . Layer and Multi CompartmPet F22G2 Multi

- .pa.t..net fiyc Setnarioc' rick ThOS F& Q hau been clodd.

Analysis report (H0493060006.006, F-99-G6 ,g.ifi.an. is not yet evaluated. Thepactin taken to address this item was Revision 0) perfor.ed a Screening Complete the detailed multi compartment p y inluded in the focused scope evaluation of the need for hot gas layer and analysis and add the discussion on multi Peer Review multi compartment analysis, and identified compartment fire sRenarios.rik

• on.trtion.

scenarais/zones that warrant further evaluations. No detailed multi compartment analysis is campleted (still in progress) in thisrepert, and ne disc'Ussion onmu-lti compartment fire cnais risk contributionR is pFrEvided. (This F&Q originated from SR FSS-G6) 4 15 Treatment for transient fire damage to FSS-PD6 Significant modeling isses eulting in This F=ROC has; bheen closed.

targets is measu-red from the compartment F=99-G! many transient fire scenarios being The acti0O taken te address this item was flor rFather than the height of the tFransient screened r dui detailaspe~iy uedi the focnrused scope fuel package that is typically codered. The reslt M;.f the F=PA are theref.re.

PT-N FSS report sectionR 8.2 states, "Cable potentially non canServativc for the analyze trays (Eorthe lowest tray wthin a stack oE detailed scenarios. Tr*an.sient fire evaluatin trays) that were at leacst 54.8' off the floor conducted as described in the Fire Scenario were considered beyonRd the zonRe of Report result in screening fire damage to infl.uen.e o-f the transient fire for nongualified targets that are ocGated , 5.8' above the floor:

cables." This apparently was based on the which is believed to be non conse,'-atfve for lowered HRR values (See F&,O 3 10) used develeped fires involving rdinaFy*

for the trasient fires, as well as the transient cmutbefuel pack~ages such -asai tr-as-h fire being located at the floGr. Thar may can or tFash bag. In.. r.espo to thi concrn..

result in the reperformance of transient fire itwas pointed out that the the~rmal plumne walkdown if the tase fire HRR valu .ompon.nt relies on empirical relationships need to be updated. Discussion with FP&L between the source strength and the followig the onsite review provided some distance betweeR the virtual origin of the fire basfis, for the damage height (indicating that and the target. The fire plume begins to transient fiFre Aabve the floor will1 have; Ran entrain Air At the lowest point Of bu1rning, overall l average .suface HRR). Whi.h defines the base.of.the fie normall However, the supplemental dm6Gu6siOR was at the floorF. However this argument ignores still considered inconsistent with past events the potential that a fire could begin burning and existing guidance on analysis ot at the top of a fuel pac~kage thus elevating its, transient fires. (This F&Cn originated from SR bas. At a minimum, duing the initial period.

ofSH6) burning, damage temperatures generated Revision 0 Page V-48

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update by the fire would likewise be elevated. OVe time the base Of the fire may change due to collapse of the fuel package Or burning away of the fuel, however the empirical rndel presented did no develop these8 ideas as a reason for assuming that the base -ofthe fire is at the floor for its 8ntire duration.

The transient fires should be coensid-ered to be above the floor level inthe analysis.

4-17 Per summary report, Task 9 is fulfilled with CF-B1 Incomplete evaluation and document for This F&O has been resolved.

the NISYS SSD database, "PTN NFPA 805 circuit failure. The NISYS DB can include Circuit failure probability was considered for Database.mdb". This database has been identification of when spurious operation high risk scenarios and only in cases where significantly expanded for the NFPA 805 may occur, but does not provide the circuit doing so would result in a reduction in total tasks. A sample circuit analysis worksheet analysis or circuit failure probability analysis risk. Additional details with respect to circuit (e.g., for component 20ASB/G3) has needed to support the FPRA. configuration and raceway type have been signatures at the bottom, which were not Provide a documented basis, and detailed added to the altered events table.

populated yet. The NISYS circuit analysis is circuit analysis for any spurious operation an Appendix R type circuit analysis and probability used in the FPRA per Tasks 9 does not identify the circuit failure modes and 10 of NUREG/CR-6850 (or equivalent).

and address likelihood of failure. Failures of the required cables identified are assumed to have a probability of 1.0 unless specifically modified in the ALTEREDEVENTS table of the FRANC model. The treatment of the circuit analysis seems to be bounding (i.e., the likelihood was not part of the analysis). Although Appendix D of the fire scenario report states the bases for the altered FRANC event probabilities, it seems that there is no linking between the altered probabilities and the circuit analysis package. The majority of the altered events are based on operator manual actions while some based on the simple spurious actuation probabilities from NUREG/CR-6850, which were based on specific evaluation (with no basis provided in the FRANC database), but do not directly linked to any specific circuit analysis Revision 0 Page V-49

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update worksheet. Since the "basis" column of the Altered event table in the FSS report does not appear to include sufficient documentation to allow review/peer review of the results and the NISYS database does not include the analysis, the analysis (not the results) has not been documented. The evaluation and documentation of the review of the fire-induced circuit failure modes and the assignment of the appropriate industry-wide generic values to their conditional failure probabilities for risk-significant contributors based on the specific circuit configuration under consideration should be included in the circuit failure report for Tasks 9 and 10. (This F&O originated from SR CF-B1) 4--1 The new fire specific safe shutdown actions HRA-B12 Risk significant human actions should be This PRO has been closed.

identified and incorporated into the PTN Fire HRA-13 defined in accordance with SR HR F2. The action. taken to address thisie wa-s PRA have not been defined (even for the HR F2 dentif;y the risk significant new fire specific speifiall inluddMi the focused sco~pe risk 6ignificant actions) by Spccif'ing (a) safe shutdoGwn actions and define these Ppar RPV*ew, accident ccquence specific timing Of cues, actions in accordance with SR HR F2 and and time window far su-ccessful completion HRA-133.

(b) accident .equence specific procedural guidance (e.g., Afes, andetFhe ) (c) !he availability of cues and other indications foi detectfion and evaluation errors (d)the specific high level tasks (e.g., train level) required to achieve the goal of the response, Or the com~plexity of the response. (This F&Q originated fromn SR HR F2) 4-24 The time available and time required to MA-Q Systematic issue. The evaluation of the time This F=O has been closed.

complete actions. were net evaluated for the HRG available and time required to complete the I he action taken to address this item was, new, risk SignificantPTN fire EPR^

related'RO cafe report risk__________________________________

risk i c action is r specfica luded in the fo.used sope shutdown actons

-N Fee~tEvalluate FRA HA tirme available to comAplete the Frik PeeF Review.

Tables A 1 and A 2 inc-lude evaluation of the significant fife related safe shutdown time available to 'cmplete actiors. However, the point in time at which operators are Revision 0 Page V-50

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update expected to receive relevant i ndications are not evaluated. (This F&O originated fromR SR 4-24 It appears that the reasonableness of rick FQ-G4 The reasonableness of risk signifiant, pest This F&O has been closed.

significant, post initiator HEPs relative to HRA CnitiatoFr HEPs relative to each othcr should The act'io taken to address this item was each other was not yet reviewed in the be review:ed and checked in the scenario specifically included in the focused scope scenario context, plant history, procedures, N-G6 context, plant histo;g, procedures, Pee&Re..'-

oporational pract!ies, and experience.. (This Q4J-04 operational practices, and epIence .

F&O originated from SR HR GE) Review thc reasonableneec of risk significant, post initiator HEPc relative to each o~ther in the scenario context, plant history, procedures, operational practices, a~dexperieRnee 4-25 Uncertainty characterization of the HEPs NRA-CG HEP u-ncertainty characterization is needed This F&O has been closed.

developed for the Fire PR. Was not NR-G8 for the evaluation of uncerteinty in the The oction taken te address this item was QpJ-Ed4 o ris results. specifically incl,-ded in the focused scope (This F&O)originated from SR HR G8) Develop-uncertainty characterization of the Peee-Review.

UNG A! HEPs used in the Fire PP.A (especially for thos16e risk significt HFEs).

4-4 There does not appear to be a dcument PRM-12 SR PRM B2 and NEI 07 12 require proper This F&O has been closed.

rocording the proper resolution of disposition of the deficiencies from the The action taken to address this item was deficiencies fr*m the previous peer review of previous peer review thal m adversely specifically included in the focused scop the nera events moedel. in addition, a peer affecnt the, accuracy Of the FirePR model PerRevyieA-revewor apasserssment of the major COF has been significantly reduced (mnore chage sicethe previous peer review of than an order of mnagnitude) since the last the internal events model doe ne appear to internal events peer review. Additionally, exist (agap database was provided, bu itnethodology changes have occurred, no supporting documentation). Finally, a gap inclui,4ng us.e of the HRA calculator and use assessment of the PRA standard changes of a new- CCF model.

fromn RO 1.200 Rev. 1 to Rev. 2 does not Resolve all signific-ant d~efici.enci_.esk_

from thep appear to exist. (This F&O) originated fromp o peer eview that may adverely SRWRM 2) affect the accuracy of the Fire PRA meodel results, if not already completed, and document these dispositions. Provide a gap assessment or new peer review en the in.teral e.vets PRA, latest Fevision.

Revision 0 Page V-51

Florida Power & Light Attachment V - Fire PRA Quality Florida Power & Light Attachment V Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 4-5 Documentation of the changes in the PRM CA The level of documentation provided in This F&D has been closed.

Internal E1vcntc PRA model to develop the Section . and Appendix D of the Fire PR* The action taken to address this item was Fire PR.A model aFr prim*arhily provided in Component and- Cnable S'el-ecttion Report specifically included in the focused scope Section 4 and Appendix D of the Fire PRA (Report 0493060006.001) does not meet the Peer Reymew.

mCmpenent and Cable Selectien Report SRs for IF 0, AS-C- andSYC (Report 0493060906.091). However;*

• Inclu-de addition)al dcmnainto the level additienal, detailed docu-mentation of the satisfying the SRs for IF D, AS C and SY C.

ch~anges is provide. Based

• on a rev.ie... t, the documentation in the tables in these sectiotns (e.g., Tables .1 2, P 1, 1) 2, D 3, etej), SRs for IF D, AS C and SY C, etc. are not met. Since the level of analysis and documentation for Firo PR-A model is expected to be similar to that for internal events, the doc'umentation in the above sections and tables does not meet what is expected for this requirement. Let's look at 6ý an L.-" IJ; there appears to be: An'-mber of eha logic. The Or Fcorrect' in the system notebo oks (e.g., DC power A. H since this is nil iust correction

. F8-.- .e here, the intrna events...PR,* is not updated. Additionally, without proper development, it is diffic.ut to determine if the change is aecurate. (This F&O originated from SR PRM Cl) 4-6 The Fire Scenario Report (0493060006.n00) F-S-A2 Circit failure mo.des are necessary for the This F&O" has b-een cle*Ad.

Appendices A,D, and F=,and the FRANC assessment Of circui failure (e.g., hot short) Th-cintke.oades this item was model document the equipment failure probabilities and required for meeting SR Speciially ithe focused scope modes for each fere scenario However, F99 A2. The mothod ued in the PTN FPRA. PeeAew-circuit failurWe modes.associat.ed with failurescn significantly over estimate the likeliho of the required sables were not identified or of the c.iri . c;,ausing such par*t*ular failure documented. Relevant circu.t failu*re modes mod as spurious operation.

are necessary for the assessment of circu-it Document circuit faire modes for the failure (e.g., hot short) probabilities. For required sables for each fire scen..ario most components, there is no differentiation i the FPRA betw-een failure modes that can e-sult due to failuire of each cable/circuit. As Revision 0 Page V-52

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update a resul1t, the fire scenarios assume each.

fail-re mode wNo-uld occur from damage to al cables identified in the SSA. In order to refine the fire scenario) under F tasks, the circuit failure would A Peed. to be prvided for

,n, oach risk relevant circuit. (This F&Q originated fromFSR FSS A2) 4-- The process of defining fire scenarios with PSS-A5 Issue with potentially significant impact. This F&O has been closed.

the source/target combinations and its F=orisk significant fir com.parFt8mts, The act t ta to 44

.a .ddress this item was F mRANCrimp.ementation process are such develop addit ioA fi .re scenarios ouch tha specificlly included in the focused scope thattheris of cah cotriutio rsk pecific targets are determined based on the Peee-Review, relevant ignition suce was cha.rate.ri.zed, ln-catmon of each taget (e.g., affected tray)

The evaluation and results are doc'umented relative to the igntio sources in the room.

inthe fire scenario report and the FRANIC mogdel. However, the analysis doss not appear to always differentiate between targets (e.g., sables in different cable trays).

Foanu-mber of the top 10 scenarios (i.e.,

fUll zone burn out scenarios), it appears all fires dam~age all equipment and all cablers in all trays (without specific knowledge ef where each of the targets are located, for example, in the trays). Basically, t1h process for developing detailed scenarios for all significant fire com~partments has not been completed in the FPR'A'fo 'rCDF and LERF.

For example, se scnro in 096, 019 and 020 (base case CDF scenarios). (This F=&O originated from SR FSS A5) 4h4. it appears that the application Of severity FSS G4 This appears to be a systemoati issue. This F&O has been closed.

facor etwenth i icositet FAN FS.1.HoweverYP, this may also be just a Th cintknt drs hsitem was-model and- those listed in AppenRdix A of the cofgrto oto n~rdcmnain specifically inc'luded in the focused scope Fire Scenario Report (04930600069.004). In issue. eveity factor should be applied to all Pe eiw the FRANC mo~del, severity faster is only applicable scenarios to derive realistic result.

used for transient fires and as the split Reconcile the differences betw'een the fraction between severe and .on severe FRANC model and the fire scenario report.

MCC firs.. For transient fires, it appears that.Apply the sever ty facGt* or non suppression the severity factor is used in the sense of a ro al applicable scenarios in a manner Revisio 0o se- Page V-53al Revision 0 Page V-53

Florida Power & Liqht Aftachment V -Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update lccation factor associated with te* osistct With thc meth*o*legy and data placement of the transient fuel,.In Appendix discussed in the fire scenario report.

.A.of the fire scenario report, severity faGctr ie also used fr goil fire, PUMP fire, and electrical cabinet fire, in addition to tr`ansient firec and MCC fires;. Further, the values of tho severity factor -sed in the FRANC model and Appendix A of the fire scenario report do not match. The bases for neither: were docu-,--mented for eacGh. in~dividual scenario.

Alsotthe-sevehiy factor values used for the oil and pUmp fires do not appear to be consistent w-ith the tab-lated values given in Table 5 1 of the fire scenario report. It appeaFrs that ceverity factors Or non suppression fa~tors can be applied to many more detailed scenarios in the FRANC rmo"del1 to mrlake the estimate of the detailed scnririk moeralisic In addition, the S.eeare-ignition frequencies listed in Appendix A do not appear to be consi.tent with those uJsed in the FRANC mdel. (This F&O originated from SR FSS CG) 5-11 Review of Turkey Point NISYS NFPA 805 ES-B1 The deviation between the ESFAS This F&O has been resolved.

Compliance Assessment Database within components actuation and Control Room SI The circuit analysis process used for the the Cable Routing and Respective components actuation should be disposition project has been confirmed to be consistent Equipment table, it was noticed that the and reconcile, to ensure Fire Safe Shutdown with the latest industry guidance (NEI 00-Spurious ESFAS signal /Appendix R equipment are appropriately 01). In addition, the asymmetry was "Spurious/ESFAS/Lacks/Analysis" have total credited in the Fire PRA. discussed with plant staff and confirmed to of 56 respective components impacted. Unit Reconcile the FPRA component list with the be reflective of the actual plant design and 3 Train A SI signal from the Control Room SSA component list for equipment impacted configuration.

"3MRASI/3CO6/3QR43/006" have total of 29 by an Sl signal respective components impacted, Unit 3 Train B SI signal from the Control Room "3MRBSI/3CO6/3QR45/006" have total of 28 respective components impacted, Unit 4 also have similar components impacted. The concern is the potential mismatch between FPRA and the SSA component lists. (This Revision 0 Page V-54

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update F&O originated from SR ES-B1) 5-13 Turkey Point FPRA Summary Report FQ-A3 It appears that there is inconsistent basic This F&O has been resolved.

NUREG/CR-6850 Task 16 Report No. event mapping between the database files. The identified data differences were 049306006.005 Rev. 1 Tables A-i, A-2, B-1 A sensitivity run was performed by copying reviewed and confirmed to be reflective of and B-2 documented the Units 3 & 4 Fire the U4 events to the U3 tables, and re- the design and layout of the units.

PRA quantification Results for both CDF and evaluated U3 CDF. The results are the top Additional comparison of the quantification LERF for all fire scenarios that were scenario in 96 dropped from 4.5E-05 to 1E- results between the two units was also quantified. Scenario 096-A was randomly 06. Based on this, the error appears to be performed to ensure that any significant picked review for both Units 3 &4. The significant. differences in results are consistent with the CDF/LERF results are consistent between Need to ensure that the altered events table actual unit differences. Various the Summary Report and Zone Scenarios in is correctly developed for both U3 and U4 for asymmetries in the plant layout were database files, Unit 3 CDF the CDF and LERF quantification. identified.

"PTNFIRE W LERF MH ESF.mdb", Unit 3 LERF"PTNFIREW_LERFMHESF.mdb",

Unit 4 CDF "U4PTNFIRE W LERF MH ESF.mdb",

and Unit 4 LERF U4PTNFIREWLERFMHESF.mdb".

However, reviewing the Altered Events table in each database files shows inconsistent basic events impacted between Unit 3 and

4. Unit 3 have no basic event impacted, while Unit 4 have 9 basic events listed. (This F&O originated from SR FQ-A3)

Review of PTN Tarsks 8 And I! Report PQ A3 The discssion between the .we repats aFre This F&Q has been cosed.

9493060006.00'4, Rev. i. Page P23Section F9A inconriecn read ng oms hydrogen fre The actfion ta-kPn to address this itemn war.

7.5.2, states "no hydrogen fires other than Incorrect apportioning the fire frequency and specifically include inPCt'he'i focused scope turbine/gnereater have been po.tulated. The defiRe appropriate fire sc.enarios . Go .d hav P..r.Reviw-basis appears to be that they use excess significant impact to the ODE and LERF flow check valves to limit H2 release. results.

Question was asked during the review, theMiscellaneous H2 fires in the response said "The sm~all 9antif o. .PRA,identifying other c..ompartments hydrogen d*wnstream of the check valves __________ ___l and its potential leakage Will result in smallcotnnghdgepin.

acc-mulations of hydrogen and are unlikely to result in cembustible concentrations of hydfrogen in any area of the plant". H5wever, furtiher rev~pie of PTNh NEPA 80 Fir-ie Ignitio Revision 0 Page V-55

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Frequency Report PTN PSA 7.01, Rev. 2, Fire Compartment Ignition Source Data Sheet (ISDS) for com~partment 082 "Unit 4 Auxiliary Transfo,,er Area" indicate that 19 Mlsc. Hydrogen F;i-e

,IN was identified in this area. Far example, H2 feed to the VCT should be looked at. (This F&O originated from SR FSS Al) 63~ Based on a review of the Turkey P*oit FPRA ES-P-i The report identifies mlp cases wih Thiws F&O has hbeen clotad mo.n.one.t and Cable ge89 lcie Reor Rep..t G9f"P . . investigation requirede Th___this _was 049306UR006O1 Rev. 1, the documentation is

- ----- P-RM R33cnimo netgto eurd oeo h cintknt drs hsie a the reselutienOs were documented as spcfclyicue pee . nthe focused scope su.fficien to SUPport the supporting requirements. P-RM- 4 addressed, but there some cases with no . ..-

The do-montaton issu-es identified in the F&Os docum.entatin

.. of resolution Or justification generally include rsuggestons6 to clrG _____________________

information in the calculation to a-ccurately reflect can be found. The PRA documentation does the process followedCin the PRA.. ES-Dl 01 not clearly chow the resoIlution of the MSO concerns a finding related te the MSO aftashment. items as required by the standard. This is This attachment needs to be updated to clearly considered a documentateron concern docu-menit how each MSO wacos addressed in the because the resolution of these items can model in accordance with the Actions from the be fou-n.d. in the PRA madel. Suggestion t0

.. M9O expert panel. Ano.thezr example: scenaro 45 update Appendix A with a clear resol,utin of involving (diesel overload) has a note to complete .

the on -going evaluation; scenario 31 (l Of 9Cpen inv~entor') has a note to verify total flow rates; Provide documnentation on open items fromn enRsure adequate Fire PRA documnentation. the MSO expert panel. Include modeling in Another example is scenario 7, 'Normal letdown the FPRA of any know-n scenarios that are tfais to isnlte aRnd inventor; is lost to the found to be an issue.

pressurizer relief tank (PRTT)', which has a open action to rwesit PRA* mondl strucntre for letdown solation. The FPRA doulmFentation doesn't discuss the disposition for this scenario and the FPRA doesn't model it. (This F&O originated from SR Es 1)1)

-64 Turkey Point FPRA Human Failure E-=A2 Any alarm p that does not require This FRO has bee.n closed Ev'aluatin Report 0493060006.002, page 6 ES aternate confirmation will need to be The action taken to address this item was states that a simulator review... was pefoermed HRA review..ed ad either disp..itioRed or specifically included in the focused scope to idniyisrmnainthat should be assessed for equipment to be added to the PeeF evoew-explicitly modeled in +nc!uding the FPRA, NRA-R4 FPR-Dcomponent list. There does not "Identify any alarms Or indications that would appear to be an extensive reviewo. of these lead operators to take immnediate control alarm response procedures for either actions wA-ithoeut fu-rther -verification." identification of instrumentation failures Revision 0 Page V-56

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Appendix C of the same report, page C 3 leading to a trip or causing an operator to has a Section. a.king whih annunciator tiles shutdown plant equipment.

cause an operator to take immediate action. Reie alarm re.pon.e proc.dres t The re.poRse . is Operators Will cfirm the. address this isu'e and document signal Mth an alternate indicator before appropriately. This aligns with the guidance takin. anY acti." This an.wer ic not in section 2.5.5 of NUREG/CR 6850.

specific, and may Rot always be procedur.ally cr9r~ect. in a typical NPP, each annUnciatoi sends the operators to some alarm response procedure. The alarm response proedure will typically require confirmnation using an alternate indicator; however, this is not always the case. (This F&O originated from SR ES A2) 6-4 No documentation is provided of a ES-A3 Systemrc issue This F&Q has been closed.

com.pFehensive

• rev iof fire impacts to FS;-A4 PeGr a reviewOf - -fireccenArio equipment The agction taken to ,address.this itern was plant equipment for unique initiating events. ES i

... to identify fire specific initiating specifically included in the focused S.. pe The MS9 ist incldes combinatos o evnt Examine gups omponents tha P-spurious operation components. However, PRM- ei

.an be disabled by a singole fire and include this review does not inclu-de consideration oe P-REM-4 the potential for a single spurious event.

c.ombtinations; of fire induced fa^.iluens that can F=Xaine these equipment impacts in terms lead to an initiating event. Additionallye,there ti success crit.ria.

did net appear to be a review of screened and the effects on the operability and initiating PDA, thier events than from the inteMal ISl O'CA pathways.,e'ents (Thispdrac performance oof operators prtr _ admtgtn and m__tgat_

.....A, ...... t .. .......

pathwas-. ..

(This systems. For eac....,,h scnario, identify a new F&O originated from SR ES A3) fire specific i*itiating event if no eXistiRg Mnita*ng event bounds or adequately represents the equipment impacts.

6-10 The altered events table in the FSS report FQ-Al This approach is not consistent with the level This F&O has been resolved.

includes several instances where a single FQ-A4 of detail modeled elsewhere in the PRA. The methodology and the analysis has been basic event combines a hot short spurious HRA-E1 Translate specific failure modes into basic updated to eliminate the use of this operation likelihood with an HEP to recover events and avoid combining disparate failure approach. The use of altered events for the spurious operation. For example, HR-I1 modes into combined basic events, spurious probability is used only as required ORZR30455C represents a combination of HR-12 and a singular value.

spurious opening of a PORV and operator QU-A3 human error probability to close the PORV.

Supporting requirement FQ-Al addresses the need to translate specific failure modes Revision 0 Page V-57

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update into basic events. Embedding an HEP with a spurious operation likelihood bypasses this requirement, and this approach is not consistent with the level of detail modeled elsewhere in the PRA. Also, the approach prevents the ability to address the state of knowledge correlation. (This F&O originated from SR FQ-Al) 6-t- The FPRA models through the Altered HRA-A4 The operator recovery actions were based This F&O has been closed.

evnts table sevpral 0.1 values for rcoovery HRAA on proposed new procedures instead of the The action taken to address this itmwas actions that are net in the safe shutdown HRAQ- exiSting ones, don't refe*et the as built as specifically iRnud4d in the focused

. sope analysis a*d are not in the fire cafe operated plant and are not *-*.ofirmed to e Peer ev..

shutdown procedures. There are about 198 HRA-D2 feasibej .

unique instances of suc.h recovery actins. HA-E= Ensure that all FPRA human failure events For example, e.ve.nt AW1.TRNAMAN,

.HR- in reflect the. as. b.ilt as operated plant, and area 63, is not incFluded. in operations t prcedalized. Verify all procedure 0 I9 P 1.0 This; humn

, . ............ HRt thtHh"2r hedited actions, including those modeled in interaction modeling doesn't reflect the as HR-14 the HRA and those included in the altered built as operated plant, and no eval.ation oA MR--12 events report, are included in the plant feasibility is d9cumented for these actions, operational procedures.

DiscussiOns vith the FPRA developmeRt team indicated that the post fire operating procedures will be updated to incorporate the new recover", actions and feasibility will be epvalu-ated at that time. The FPRA will need to be updated, as necessary to reflect the outcome of feasibility evaluations. (This F&O originated from SR HR El)

Based on a review., of the FPPRA HRA HRA- A! Step ot peFfermed This F&O has been closed.

Rep..., no talk thr.Ughor.re FHRAA4. appear Pform talk th-oughs or reviews of The action taken to address this im was and training persennel of proh ,u r.e..san RAE procedures and sequences of events with specifically the focused scope a~dt~aRig PFBG~e ~dplant 9fPFGedes operations and training personel to Pewr-Review.-

sequences of events to confirm that H con that

+Afirminterpretatin Gofthe prcGedures interpretation of the procedures by FPPA HR is consistent with plant 9bse'-ations and developers is consistent with plant HR 12 obser'-ations and training. The report indicates that a simulator review was Revision 0 Page V-58

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update performned to identify instrumentation that should be explicitly modeled in the fire PR.A as generally required to shutdown the plan or to perform credited operator aGtions, and a review agend-a is provided in Appendix C, b-t no documentation of such a revLew is prvided. (This F&O originated from SR HR E4) 6-1-3 NoIsirmulator cbser.ations or talk throughs HRA-A4 Step not pereFormed This F&O has been closed.

wvfit-operators have been performed to HR 1Perform4 silmuwlator observations or talk The action ta;ken to address this itmwas, scnarfios the leSPGse med( s for fire d M =4 throughs with operators to confirm the specifically included in the focused scope rsmodeled.

(This F& rigited rpse models f4or fire scenarios modeled.

..... Peer-Review.

firom SR HR E4) HR-I1-HR-12 645 The documentation of credited recoveries in NRA-El Systemetio-6sue This F&O has been closed.

the altered events table in man" instances is - Perform analyses to verify operatoi vague (e.g., "Mababil ty of re ore st bus," HR 12 reco.veries and. repairs cedited in the The action taken to address this itemn was and.ailue.prbabiity.f.ne .insde.cnto Altered events Table in Appendix P are specifically iRnclude in the foGused scope room HEP, required in less than 60 minutes RM 13B feasible, prior to crediting an" recovery via Pe....e..e..

(M..A, D 205)"). The effo. .to evaluate manual operation of the equipment.

these action for incluiolrin the SSA and feasibility would be facilitated by moree detailed de-scriptions of the actOnS. PTN system mRodel changes in the fire PRA models are summarized in Tables D 1, D 2 and D 3 of the Component aRnd Cable Selection Roped1 0193060006.001, Ravicien

1. However, no additional doc~umentation ot the changes is previded. Requirements unlder- S-Y A.and- S-Y BR are not m~et. Repair of components that are spuriously operated or fire damaged is modeled,using the altered events table of Attashment 0), FRANC Altered Events Table, of the Firo Scenario Analysis Repope 093060006.004, Rev 1.

However, the substitution does not include a

-erififation that the actions are possible or feasible. (This F&O originated from SR Revision 0 Page V-59

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update HRA-E-)

6- 19 The HEP dependnyeauto prouco ~Q1 UnreasonRably low values accigned t9 Thic; F&OC h-ac been closod.

.eveal* depcndent HEPs on the order of !E -RA dependent HEP co.bination. The action taken to address this item 'A-as 11 to 1 E 13 (for example, cases 62, 93, 110

__________________________2 AsSign a floor for dependlent HEP specifically iinclde -in.th focused scope and 96). NUREG 17.92. "Good Practices foco Gmbinations usin a justified minimum Pee..Revi..

Implementing Human. Reliability Analysis,"

recommends that the total combined probability of all the HFEs in the came accident sequence/cut set should not be less than- a justified value. NUREG 1792 suggests that the value .not bhebelow.. -1 5, since it is typically hard to defend that other dependenRt failure m*ordAs that areRnt readily anticipated cannot occur. Howeverf .sme.

inRdustry PRAs are using a floor value of

-IF 06. (This F&O originated ftrm SR HRA G-4-)

6-20 The parametric uncertainty associated with CF-A2 Step not performed This F&O has been resolved.,

conditional circuit failure probabilities are not UNC-A2 Step not performed Parametric uncertainty has been performed evaluated and are not incorporated into the for CDF and LERF for each unit's FPRA.

model. (This F&O originated from SR CF-A2) 6-3 For the FPRA, no.. accident sequences were AS A4 Required step not perform.ed. This F&Q has been clesed.

identified beyond these modeled by the AS-Ag Perform a review of FPRA scenarios to The action taken to address this item was interaa event RA Th ER aciet. L  ! enur tha-t the existing event tree Structures spcdial incudd inthe focus-Fed scope seuec acdent progression, success aGG~wately P-Gedel the spocifi G PRA initiating 126er-Review, crite timing are therefore based- en

.and- e.ents, ic'luding considerations ef timing, the itra vents PRA. ConsiPderation RRM--1-14 plant response, and human interactions.

should be given, however, to...... sc15 . PRM criteria and timing specific to the FEPRA. Fore -R-example, ne evaluation is m.ade of the timing PRM-85 associated with RWST draindo-n. Also, PRM-87 RVVST draindown mayMrequire smp recirculatiOn, b hic is nop irntly represented in the- nOn LOCGA. transient event tree accident sequences. This step has not been performed and finding is mnad-e Revision 0 Page V-60

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update to include su-ch considerations in the FPPRA.

development and dcmnain Also, when any new FPPR-A. initiating events are identfiedas at of resolving ES F&Os, considerationý will need to be. mnade of tho accident coquoncoe accident progression and timing associated with any new accident sequences. (This F&O originated from SR PRM-137 6-4 NoiFPRA modeling appears to have been AS-Al Step not performed. Potentially significtant This FO has been closed.

Made to add.ress . the actions directed by the ASA10 impact On F.PR ".. A aRccident sequences and The action taken to address this it was fire safe shutdown procedures that deviate Fesuts. specifically included in the focused scope from the actions directed by the FOPs. (This Model ing of the following procedural P-eer-Review,.

F&O originated- fromn SR- PRM B)AS-A5 responses to a fire mnay be needed: 1)

NRA-A!- equipment is disabled to preclude spurieus I'RA-A2 actuations; 2) human actions to iso-late-HR E! unprotected equipment; 3) human acGt*on to HR....manually operate pro.tected equipment.

Some of the above hu-man actions could PRM-B5 alsE indu-ce new sequences not traditionealy PRM-16 co'vered in the internal Events PRA- hew sequences to account for these effeGts may.

also need-to be incorporated int the- Fire_

PRA-Model.

6-9 The parametric uncertainty analysis as FQ-A4 Step not performed. This F&O has been resolved.,

discussed in QU-E3 (estimate of uncertainty QU-A3 Perform the FPRA uncertainty analysis, Parametric uncertainty has been performed intervals, etc.) is not performed. Also, the including estimates of uncertainty bounds, for CDF and LERF for each unit's FPRA.

"state-of- knowledge" correlation between per the requirements of QU-A and QU-E.

fire-specific event probabilities (e.g., When performing parametric uncertainty suppression system unavailabilities, fire calculations, ensure uncertainty intervals for ignition frequencies, hot short conditional event probabilities utilized by the FPRA are probabilities, etc.) hasn't yet been applied, correlated when significant.

(This F&O originated from SR QU-A3) 7-1 A review of the quantification results for PRM-A3 The method in which the fire-induced This F&O has been resolved.

selected compartments involving fire- PRM-A4 spurious safety injection actuation and A review of the model was performed and induced safety injection actuation and fire- PRM-B5 spurious opening of atmospheric dump revision made to address and resolve the Revision 0 Page V-61

Florida Power & Light Attachment V - Fire PRA Quali Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update induced opening of atmospheric dump PRM-B9 valve initiating events were linked into the issue identified in the F&O. Additional valves was performed to verify that the fault tree produces conservative results that reviews were performed as part of the modeling was consistent with the internal could impact the determination of significant overall results and cutset reviews and no events PRA treatment of similar initiating contributors to fire-induced risk. additional instances were identified.

events. This review revealed that duplicate Review the quantification results for the fire-cutsets were being introduced by the induced initiating events to verify that the manner in which the new logic for capturing results are consistent with the comparable the fire-induced initiating events was linked internal events.

into the fault tree. For example, a review of Review the mapping of the fire-induced cutsets for zone 098-A showed that the top initiating event impacts to ensure that they two cutsets were identical except that one inistent ith tompare thal used bleed aandversion feed of the HFE based for alignment on reactor trip of are consistent events initiator, wit thate that comparable appropriate internal differences bleedrinndifeed baseleon ranctor orip rdue to the considerations of the fire PRA are occurring with SG low level and the other incorporated (e.g., application of bounding used a version of the HFE based on timing timing for HEPs to capture uncertainty in the associated with trip with nominal SG level.seuneofi-ndcdalrs)ortt Similar for zoneissues were identified in the cutsets 091-ETL. sequence of deviations in fire-induced the modelingfailures), or that are documented and justified.

As noted in the 2010 peer review in F&O 1- Review the application of the feed and bleed 4,

of there are also inconsistencies the fire-induced small LOCA when in modeling HFE H Eto t ensure a p o ra ettiming n u etthe eappropriate m n iss used ue compared to the internal events small LOCA during the fire quantification. Since the initiating event. PTN explainsthath was MFW pumps are assumed failed for all fire due to circular logic issues, and a sensitivity areas, the most appropriate value may be due to is lo be a non-significant the HEP based on timing assuming the trip case shows this to beannsgiiatoccurs with low level inthe SGs.

issue. However, it is not clear that the circular logic issue could not be resolved Review treatment of any additional HEPs and that all potential impacts of the modeling with event-specific timing assumptions to approach taken are understood, ensure that the appropriate values ate used (This F&O originated from SR PRM-A3) in the fire quantification.

7-3 The current model uses the LERF model for PRM-B14 It cannot be determined from the existing This F&O has been resolved.

the PTN revision 9 model (PTN-BJFR documentation that an assessment was A review of the mapping of Level I 010, Rev. 1) and maps appropriate performed to identify new accident sequences to the plant damage states in the equipment impacts into the system models progressions beyond the onset of core LERF model was reviewed. No new used to model LERF. No new accident damage that would be applicable to the Fire accident progressions that required progressions beyond the onset of core PRA that were not addressed for LERF modification of the LERF model were damage were identified for the fire PRA. estimation in the Internal Events PRA. identified.

However, there is no documentation that a Document an assessment to determine if specific review of the accident progressions Revision 0 Page V-62

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update leading to LERF was conducted to identify there are potential fire-induced LERF whether new considerations should be mechanisms not captured by the internal addressed in the fire PRA. events accident progression models. If none In addition, effects on PDS mapping due to are identified, document the basis of that fire-induced failures may not be conclusion.

appropriately captured. For example, RWST diversion of the RWST to the containment sump is modeled as a failure of HHSI which would normally go to a dry containment PDS. However, the actual PDS should be one for wet containment. While this is a late containment failure concern rather than a concern for LERF, there may be similar fire-induced failures that could affect the mapping of LERF accident progressions.

(This F&O originated from SR PRM-B 14) 7-6 The new fire-specific safe shutdown actions HR-El Final post-fire safe shutdown actions have This F&O has not been resolved.

which are credited in the final Fire PRA will HR-E2 not been defined and appropriate The FPRA includes various actions that are be proposed to be added to the plant fire HR-E3 procedures revised to include the actions to being included as required plant changes in response procedures. These human actions be credited in the Fire PRA. the NFPA 805 LAR. The development and are included in the ALTEREDEVENTS table HR-E4 Complete the identification of new fire- implementation of related procedures has of the FRANC model using component basic HR-H2 specific safe shutdown actions which are not yet been initiated as it is part of the events as surrogate. HR-l1 credited in the final Fire PRA and evaluate overall integrated process associated with However, the safe shutdown actions HR-12 and document the HEPs consistent with transition to an NFPA 805 license basis.

modeled in the FPRA are not currently processes used for internal events HEPs.

consistent with those specified in the plant HR-13 Include consideration of fire effects on the fire response procedures, there is no HRA-A2 operator action, availability of cues, documented assessment of the cues HRA-A4 availability of time to complete the action, required to initiate the actions, no training feasibility of the credited actions given a fire, has been provided to operators on the new HRA-B2 and potential 92-18 impacts for both fire-specific actions, no operator reviews or HRA-B3 screening values and detailed HEP talk-throughs of the credited actions has HRA-D2 development.

been documented, and the applicable HRA-E1 Also, complete operator reviews and/or talk-performance shaping factors have not been throughs when the procedure updates are considered, including time available for the PRM-B1 1 completed to ensure that the interpretation action. of the actions is consistent with the This F&O supersedes 2010 Peer Review operator's understanding and training.

F&Os 1-41, 2-6, 6-4 and 6-11. Finally, consider expanding the discussion Revision 0 Page V-63

Florida Power & Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update (This F&O originated from SR HRA-A2) of sources of model uncertainty related to the HRA to include consideration of the accuracy and completeness issues noted in NUREG-6850, Volume 2, Appendix V.

7-8 Dependency between multiple altered HR-H3 The dependency associated with operator This F&O has been resolved.

events representing new HFEs in the same HR-12 actions applied using the altered events The use of altered events as a surrogate for cutset and between action represented by HRA-D2 method has not been addressed, a recovery action has been significantly the altered events and other HFEs in the Address dependency between multiple reduced as noted previously. Those same cutset has not been assessed based PRM-B11 altered events representing new HFEs and remaining instances are addressed by on the assumption that the dependency between the altered events and other HEPs modifications to the recovery rule file so that effects are bounded by the application of in the same cutset. If detailed dependency only a single instance of this use would exist conservative screening values. However, analysis is not performed, provide a in any cutset. This eliminates the potential there is no documented assessment to justification supporting the assumption that for multiple surrogate recovery events to support this assumption. There are cases the values chosen for the altered events appear together in the same cutset.

where complete dependency between bounds dependency effects.

events may be appropriate. For example, cutsets 40 - 45 in the provided Aggregate CDF-aggregate.cut file contain altered events MAVC4200A_1.00E-01 and MAVC4460_1.OOE-02 in each cutset. The product of these two events is therefore 1.OOE-03. However, since both events involve failure to isolate the letdown line, it could be assumed that there is complete dependence between the events since they would share a common cue.

This F&O supersedes 2010 Peer Review F&O 6-16.

(This F&O originated from SR HR-H3) 8-3 Attachment U - Internal Events PRA Quality PRM-B2 The potential effect of internal events F&O This F&O has been resolved.

(DRAFT), document applicability of Internal disposition on development of the FIRE PRA The internal events PRA model F&Os that Events F&Os to internal events PRA, but not was not addressed, have not been resolved/closed have been to Fire PRA. There was no evidence that Review internal events F&Os and provide reviewed and found to have no negative the review of F&O disposition status documentation as to how disposition of impact on Fire PRA results or this addressed the question of whether the those F&Os may impact development of the application.

disposition that was taken would adversely Fire PRA.

affect the development of the fire PRA.

Revision 0 Page V-64

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update This F&O is derived from 2010 Fire PRA peer review F&O 4-4.

(This F&O originated from SR PRM-B2) 8-5 The Fire PRA model changes were HRA-B1 The standard requires a review of the fire- This F&O has been resolved.

constructed so as to allow credit for the HRA-B3 induced initiating events accident current internal events PRA model structure sequences, and success criteria included in The review of results and cutsets that were using existing accident sequence the internal events model, to identify new perrevid idents anumbetstanes progression, success criteria and timing. PRM-B7 accident sequence progressions or success performed did identify a number of instances The internal events HRAs areThe modifiedinernalF&O.

with criteria due to unique aspects of fires,* such asInthat all instances, specificallyit identified was determined in the that a screening modifier. Travel paths are This review will help assure that there are no the existing model structure was appropriate considered in the human failure evaluation revised actions where the screening and that opportunities for recovery actions report. multipliers are not appropriate, were limited either because of a lack of However, there is no indication that a review Conduct and document a review of the fire- appropriate cues or insufficient timing to gain was performed to identify accident induced initiating events accident any meaningful benefit via recovery. The sequences that may require modification sequences, and success criteria included in analysis documentation of the HFE based on unique aspects of the plant fire the internal events model, to identify new treatment was updated to address the response procedures. For example, RWST accident sequence progressions or success internal events PRA model human actions draindown may affect the evaluation of criteria due to unique aspects of fires. that are used in the FPRA. The timing for aligning sump recirculation, which documentation addresses the applicability, is not presently represented in the non- numerical adjustment, and availability of LOCA transient event tree accident necessary cues.

sequences used for the majority of the fire scenarios.

A review should be performed for possible changes to success criteria, particularly due to model changes from the MSO evaluation.

This F&O is derived from 2010 Fire peer review F&O 6.3.

(This F&O originated from SR PRM-B5) 8-8 Several portions of the analysis are not FSS-E3 Unit 3 results not fully documented, although This F&O has been resolved.

documented for Unit 3. Specific examples PRM-A3 they are available for inspection using the The U3 results have been added to the include: PRM-Cl quantification software. analysis documentation.

Unit 4 significant contributors are identified Document Unit 3 results consistent with the in 0493060006.005, Rev. 2. Unit 3 Unit 4 results.

significant contributors are available, but not fully documented.

Revision 0 Page V-65

Florida Power & Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Unit 4 Fire Scenario information is presented in 0493060006.004, Rev. 2, but the equivalent Unit 3 information (Attachment D and E) is not provided.

(This F&O originated from SR PRM-A3) 8-10 2010 FPRA peer review F&O 1-44 finds HRA-C1 This appears to have a significant non- This F&O has been resolved.

issues with masking fire effects by setting PRM-B1 1 conservative impact to PRA results. Given The use of '0'; has been eliminated in the basic events to 0 in the Altered Events table. the actions in the altered events report are Altered Events table. Instead, events are This has partially been addressed by setting being added to the model as needed set to nominal. In the case of the application these events to 'nominal' in the Altered, recoveries in order to ensure risk is low, and calculations for NFPA 805, the 'compliant' Events table and for reviewing cases where given the resulting recovery actions do not case is determined by using a '0' value the nominal value is on one side of an AND show up in the results in most cases, there which would under-estimate the compliance gate and the modified HEP value is on the appears to be a disconnect between the case risk and thereby provide a conservative other side. However, there are still cases addition of new actions to the procedures estimate of the risk increase for the where fire impacts are masked when the and the quantification of these actions in the application.

nominally adjusted event is on both sides of FPRA. It appears part of the disconnect is an AND gate or the HEP event is on one that the logic modeling, as modified by the side of an AND gate and nominally adjusted altered events table, results in the recovery events are on the other side. Scenarios 030 values being screened from the results.

PTB and 067E PTB are two examples. This Due to the complexity of this methodology, it old F&O is converted to a new F&O 8-10. seems a difficult task to review and address (This F&O originated from SR HRA-C1) for these masking issues. Perhaps a more systematic and comprehensive approach, with an independent review, could provide confidence that these nonconservatisms are addressed. Adding new HEP basic events, consistent with the approach used for internal events, would address this issue.

9-1 A general screening based on the ability to FSS-G2 No basis for the screening criteria is This F&O has been resolved.

form a damaging HGL in an exposing FSS-G3 described. Since no MCA scenarios are The existing HLG/MCA analysis includes a compartment was developed. If no developed, there is no way to determine if number of occurrences where the simplified damaging HGL could form in an exposing FSS-G6 the exceeded (yet applied) screening criteria screening approach was found to generate compartment then there was no possible are significant. over-conservative results. Incrementally associated MC scenario. Provide a basis for the 1 E-07/yr screening enhanced treatments were applied to When a damaging HGL could form, a criteria including additional information confirm that these locations had a very low second screening was performed whereby required when the criteria are included. The likelihood of creating or causing formation of the frequency of developing the HGL was basis for the screening criteria should HGL conditions and consequently a possible Revision 0 Page V-66

Florida Power & Liqht Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update determined. If the frequency was less than ensure the frequency is not too high, thereby multi-compartment scenario 1E-07/yr, then the scenario could be potentially masking significant MCA eliminated. However, there was no basis scenarios.

provided for the 1 E-07/yr criteria, nor was Evaluate impact of exceeded screening the criteria adhered to; in fact, the criteria criteria; for example, a qualitative analysis of was exceeded, yet still applied, in over 150 the expected CCDP based on known targets different scenarios. Some of the screened in the exposing and exposed compartments.

scenarios were slightly over the 1E-07/yr threshold, while others ranged as high as nearly 6E-07/yr.

The impact of exceeding the criteria cannot be determined as no specific MCA scenarios were ever developed; therefore, it is not known if the scenarios would be significant.

For example, if it is assumed that 10 of the scenarios with a frequency of 5E-07 had CCDPs of 1.0, this would result in an increase in total CDF of 5E-06 which is about 10% of the total fire CDF.

(This F&O originated from SR FSS-G2) 9-4 The multi-compartment analysis assumes a FSS-G4 Systematic generic assessment of active fire This F&O has been resolved.

bounding value of 7.4E-3 for evaluation of FSS-G5 barrier elements may lead to non- The update of the analysis to incorporate a active fire barrier elements. Actual fire conservative results, barrier failure probability that integrates all barrier elements are not considered; instead If a screening value is desired, NUREG/CR- possible barrier elements was found to result the failure probability of a fire door is 6850 Section 11.5.4.4 suggests using a in a value of approximately double the assumed for active barrier element failure screening value of 0.1 for active fire barrier current value. However, since the entire because this failure probability represents elements. This value is much more likely to analysis approach involves a screening the highest single probability of a single encompass multiple fire barrier elements. strategy, additional analysis refinements are barrier failure. This method ignores the For scenarios that do not screen out, actual possible. An assessment of the use of a potential for multiple fire barrier elements, fire barrier elements identified during higher barrier failure probability to account Per NUREG/CR-4840 (source document for walkdowns (or document review) can be for failure of multiple barrier elements found NUREG/CR-6850 Table 11-3, "Barrier used to develop a more realistic barrier that the overall conclusion that MCA Types and Their Failure Probabilities") the failure probability, scenarios are not risk significant and need total barrier failure rate is a union of the not be explicitly included in the FPRA was probabilities of the individual failure rates. confirmed. However, the analysis Therefore, a value of 7.4E-03 may be documentation has not yet been updated ot conservative or non-conservative. reflect these results and insights.

This is based on 2010 FPRA peer review Revision 0 Page V-67

Florida Power & Light Attachment V - Fire PRA Qualitv Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update F&O 1-35.

(This F&O originated from SR FSS-G5) 9-5 The screening criteria are defined in the FSS-G2 Use of a standard fire scenario may be non- This F&O has been resolved.

Turkey Point Hot Gas Layer and Multi- conservative for some zones. The use of the 5 minute delay to Compartment Analysis, (Report Review zones to ensure that the standard combustible cable ignition is considered H0493060006.006) methodology, fire scenario is actually the most challenging realistic. Other conservatisms in the Compartments that don't screen are retained scenario inherent to the analyzed analysis ensure the overall conservatism of for further analysis. compartment. For zones where the the MCA/HGL evaluation.

A concern identified with the screening standard fire scenario is not the most criteria involves the use of a standard fire challenging, determine the most challenging scenario for each analysis rather than scenario and evaluate accordingly.

determining the most challenging fire scenario inherent to the analyzed compartment. This approach potentially masks the potential for forming an HGL in the exposing compartment.

For example, in zones 67 and 68 the standard fire scenario is non-conservative due to the potential for HEAF in 4kV switchgear. The damage time of 5 minutes is non-conservative for HEAF scenarios (should use 0 minutes).

This F&O supersedes 2010 FPRA peer review F&O 3-11.

(This F&O originated from SR FSS-G2) 9-6 The system unavailability records for the FSS-D7 This is a systematic issue. The intent for This F&O has been resolved. The fire plant have not been reviewed in crediting Capability Category II is to additionally protection system availability data for PTN fire detection and suppression systems. require a review of plant records to has been reviewed and no outlier behavior This F&O supersedes 2010 FPRA peer determine if the generic unavailability credit has been identified.

review F&O 2-26 is consistent with actual system unavailability. Outlier experience would be (This F&O originated from SR FSS-D7) any experience indicating that actual system is unavailable more frequently than would be indicated by the generic values.

Consider performing and documenting the review of plant records to determine if the Revision 0 Page V-68

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update generic unavailability credit is consistent with actual system unavailability. Outlier experience would be any experience indicating that actual system is unavailable more frequently than would be indicated by the generic values.

9-10 Section 3.1 of the FSS Report FSS-A6 Assumption made that no cabinet/panel fires This F&O has been resolved.

(0493060006.004, Rev. 2) states: "For the FSS-H7 in the MCR will ever spread to an adjacent Panels with communication between electrical panel fires, the scenarios are cabinet even if the cabinets are open to one adjacent panels are to be provided with developed similar to scenarios involving another. This incorporates an implied incipient detection to ensure early electrical panel fires outside the Control assumption that every MCR panel/cabinet identification of fire to preclude spread Room and are adequately described in fire will be extinguished prior to spread. between panels.

Attachment A. Fire spread to adjacent Identify adjacent MCR cabinets/panels panels was determined for these scenarios which could result in fire spread given failure based on a walkdown of the control room of suppression. Apply NUREG/CR-6850 during which panels with potential barriers Appendix L, S or other relevant document to for spread of fire were opened to confirm the address the potential for fire spread.

existence of such barriers. For MCB fires, Ensure documentation is consistent with the the method from NUREG/CR-6850 proc ually us in the Appendix L is applied. NUREG/CR-6850 process actually used in the analysis.

Appendix L defines a non-suppression probability applicable to the MCB. From Figure L-1 of NUREG/CR-6850, for non-qualified cables, and for a bounding distance of 0 meters (assuming that the cables terminating at the individual MCB are in very close proximity), a non-suppression frequency of 8.30E-3 is used for the MCB."

However, based on discussion with FPL/ERIN staff, this was not done.

Essentially, no fire spread for any cabinet in the MCR was assumed. For panels with incipient detection, success of the detection results in no damage as it is assumed operators isolate the circuit prior to additional damage in the cabinet. If insipient detection fails, the MCB panel fails completely, but never spreads another cabinet.

Revision 0 Page V-69

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update For all other cabinets/MCB panels without incipient detection, full burnout of the cabinet/panel is assumed, but again, no spread to adjacent cabinets is assumed even if the cabinets are open to one another (e.g., walkthrough MCB).

(This F&O originated from SR FSS-A6) 9-11 Several entries in Table 3-1 of Report FSS-G2 It was not confirmed during the peer review This F&O has been resolved.

H0493060006.006 says "Walkdown required that this was just a documentation issue. It was confirmed that the lower damage to confirm no combustibles within the 383 Therefore, this is classified as a finding threshold associated with thermoplastic Zor"when the cables are not IEEE-383 because it could affect the analysis results. materials was used for the analysis. The qualified. Based on discussions with FPL Verify that the damage criteria used is typographical error has been corrected.

and contractors, it is believed that this is a consistent with non-383 cable damage and typo, and the correct damage criteria were revise the documentation as required. If it is actually applied. discovered that the incorrect damage criteria (This F&O originated from SR FSS-G2) were applied, update the analysis with the correct criteria.

10-1 The 2010 peer review identified that "Fire FSS-Cl The present analysis provides a bounding This F&O has been resolved.

modeling was conducted via generic fire FSS-G1 approach for fire severity in most cases, The recommended resolution action in the modeling from which Zones-Of-Influence since the 98th percentile fire heat release F&O was assessed in the context of the (ZOI) for specific initiator types was rate is used. dominant fire risk contributors. This generated. The ZOls were used to define However, use of the split fraction method is assessment concluded that further bounding fire characteristics for each fire based on industry events rather than site refinements such as that described in the scenario. Characteristics that are used to specific fire ignition sources and target F&O would not substantively change the bound potentially risk contributing fire events configurations. Therefore, this could result results of the analysis. The existing are identified in Attachment B of the Fire in non-conservative frequency estimates of treatment retains some conservatism which Scenario Report, (Report 0493060006.004). target damage. results in this SR meeting CC I. This is Based on the use of a bounding approach Perform 2-point fire modeling, when adequate for the NFPA 805 application, as this SR is judged to be met at 00 1. applicable, for risk significant fire scenarios, this conservative bias would tend to over-Significant fire scenarios should be estimate the risk metric that is used to judge developed with 2-point fire modeling." the acceptability of this application.

Since this review, FP&L has stated that "The The issue regarding the ERIN panel split use of a panel split fraction to differentiate fraction is addressed in the disposition for between fires impacting the panel and F&O 10-3.

components with cables terminating at the panel versus panel fires impacting cables outside of the panel provides an equivalent Revision 0 Page V-70

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update and more useful two point fire model."

The Panel Split fraction is developed from a supplemental report (ERIN report, Supplemental Fire PRA Methods, dated February 2010). This document was submitted to the EPRI Fire PRA Methods Review Panel. This review is not complete as of the date of this peer review.

Use of the split fraction method is based on industry events rather than site specific fire ignition sources and target configurations and therefore, could result in non-conservative frequency estimates of target damage.

(This F&O originated from S.R FSS-C1) 10-2 The 2010 review of PTN Tasks 8 and 11 FSS-A1 Including the fire frequency and associated This F&O has been resolved.

Report 0493060006.004, identified that 'no fire scenarios from hydrogen fires will have Miscellaneous hydrogen fires have been hydrogen fires other than turbine/generator impact to the CDF and LERF results. incorporated in the Fire PRA in the charging have been postulated.' (Previously F&O 5- Incorporate the hydrogen fire scenarios pump room fire areas where the hydrogen

16) being developed into the model, and update lines associated with VCT cover gas are Since this Finding was identified, FP&L has documentation as necessary. routed.

determined that 'Miscellaneous Hydrogen piping at PTN is limited to hydrogen supply to the VCT tanks. The associated piping is located in the charging pump rooms (Fire Zones 45 and 55). Fires in these fire zones are assumed to impact all components in the fire zone. The associated risk is low given the availability of thermal barrier cooling for RCP seals and HHSI pumps. Allocation of the IGF associated with miscellaneous hydrogen fires to these fire zones would result in an increase in the ignition frequency for these zones by less than a factor of 3.

Given the low risk significance of these zones this will have a negligible impact on overall plant risk and the charging pump rooms will remain low risk contribution fire Page V-71 Revision 0 Page V-71

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update zones.

Incorporation of this ignition frequency into the associated documentation will be incorporated in a future revision to the documentation.'

Hydrogen fires are also being developed for H2 piping and valves in Compartments 82 and 87 (scenarios 82-P and 87-P).

However, since these do not appear yet in the Fire Scenario Report, action is required.

This finding is currently being addressed and appears to be resolved once the new H2 fires are included in the model and documentation is updated.

(This F&O originated from SR FSS-Al) 10-3 FSS-C4 requires severity factors to be FSS-C4 Severity factor (panel split fraction) is used The FPRA quantification uses the panel independent of other factors. Fire severity FSS-D5 extensively in the Fire PRA. factors consistent with the latest guidance factor as discussed in Section 7.1.2 for Use the severity factor method described in from the EPRI Methods Review panel. A electrical cabinets is not developed or FSS-G1 NUREG/CR-6850, or develop an accepted sensitivity study has been performed to applied consistently with the NUREG/CR- industry approach (presently being address the impact of elimination of the 6850 methods. This is developed from a discussed by EPRI). Develop fire severity credit for the panel factors. The results of supplemental report (ERIN report, factors based on the likely HRR and location this evaluation indicate that the delta Supplemental Fire PRA Methods, dated of overhead cables or location of equipment. CDF/LERF would exceed the Reg Guide February 2010). This document was For example, if cable is 7 feet overhead, the 1.174 guidelines should these factors be submitted to the EPRI Fire PRA Methods severity factor would be based on the completely eliminated (the 1E-5/1 E-6 delta Review Panel. This review is not complete minimum HRR that would damage the cable CDF/delta LERF limits would be exceeded as of the date of this peer review. at that distance. Additionally, the growth but the conservatively calculated delta risk Using this method, fire propagation outside time can be used in determining non- would be less than 2E-5/2E-6). Further of the electrical cabinets is dependent on the suppression time. refinements of this sensitivity evaluation are nonsuppression probability. Therefore, possible to reduce the calculated delta risk.

some dependency exists in this data if used Credit for additional Defense In Depth in conjunction with a non-suppression factor. measures may be taken in areas of concern Due to this derivation of the conditional as necessary to compensate for the probabilities for fire propagation outside of increased delta risk.

the cabinets, the conditional probabilities thus developed (and applied in the FRANC model) could potentially be non-Page V-72 Revision 0 Page V-72

Florida Power & Light Aftachment V - Fire PRA Qualitv Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update conservative.

The severity factors are developed using generic fire events data from the EPRI fire events database. Given the fire data duration and damage is a result of multiple factors (growth, suppression, severity, location, etc), and given the fire data often does not have sufficient information to make a reasonable determination of either the fire size or whether a fire propagated outside the cabinet, the severity factor used (panel split fraction) may not necessarily bound the conditions of the specific fire scenarios under analysis.

(This F&O originated from SR FSS-C4) 10-4 One situation was identified for which credit FSS-C8 This finding is based on identification of This F&O has been resolved.

of fire wrap is taken in Compartment 96 for credit for a wrap in Attachment A of the Fire A qualitative assessment has been ignition source 3B04, which is a 480V load Scenario Report, (Report 0493060006.004). performed to assess the potential impact of center. This fire wrap protects PB3319, Any credited fire wrap should be addressed this F&O.

PB3813, PB7022, and PB7521. The wrap and the wrap integrity should be established appears as being credited in a HEAF The hose stream test imposed on the fire with respect to fire resistance, mechanical barrier qualification subsequent to fire scenario. No justification for crediting this protection, and potential fire related wrap assuming mechanical damage and exposure is considered to provide a exposure to which the wrap may be exposed comparable level challenge to the thermolag direct flame impingement from the HEAF is (direct flame impingement, HEAF, etc.).

provided. Similar issue for 3B03 also in barrier as would the HEAF force applied at the onset of fire exposure.

Compartment 96.

Thermo-lag is also seen as credited in some scenarios, which would require justification due to issues with this particular type of cable barrier.

(This F&O originated from SR FSS-C8) 10-6 Treatment for transient fire damage to FSS-D6 Many transient fire scenarios have been This F&O has been resolved.

targets is measured from the compartment screened during detailed scenario analysis. Supplemental walkdowns were performed to floor rather than the height of the transient The results of the FPRA are therefore re-assess the treatment of transient fires.

fuel package that is typically considered. potentially non-conservative for the analyzed These walkdowns focused on two key Discussion with FP&L during the review detailed scenarios. attributes - the appropriateness of the provided some basis for the damage height The transient fires should be considered to selected HRR characterization and the Revision 0 Page V-73

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update (indicating that transient fires above the floor be above the floor level in the analysis. location of the postulated fire scenarios.

will have an overall lower average surface With respect to this specific F&O, the HRR). However, the supplemental placement (elevation) of the assumed fire discussion was still considered inconsistent was based on the physical features of the with past events and existing guidance on location. The fires were not artificially analysis of transient fires, and could lead to elevated in the absence of a physical non-conservative estimates of transient fire feature, damage to targets.

Transient fire evaluations conducted as described in the Fire Scenario Report result in screening fire damage to targets that are located > 7.3' above the floor which is believed to be non-conservative for developed fires involving ordinary combustible fuel packages such as a trash can or trash bag. In response to this concern it was pointed out that the thermal plume component relies on empirical relationships between the source strength and the distance between the virtual origin of the fire and the target. The fire plume begins to entrain air at the lowest point of burning, which defines the base of the fire; normally at the floor. However this argument ignores the potential that a fire could begin burning at the top of a fuel package thus elevating its base. At a minimum, during the initial period of burning, damage temperatures generated by the fire would likewise be elevated. Over time the base of the fire may change due to collapse of the fuel package or burning away of the fuel, however the empirical model presented did not present sufficient basis for assuming that the base of the fire is at the floor for its entire duration.

(This F&O originated from SR FSS-D6) 10-8 Ambient conditions are assumed in the FSS-D4 Underestimating the ambient conditions This F&O has been resolved.

Revision 0 Page V-74

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Generic Fire Modeling Treatment Report FSS-H4 could result in non-conservative estimations A qualitative assessment has been (prepared by Hughes). Ambient of zones of influence and targets considered performed to assess the potential impact of temperature is assumed to be 68°F for all to be fire damaged. this F&O.

calculations. No technical discussion or Assess areas where elevated ambient The sensitivity of the ZOI dimensions to the justification is provided in the Fire Scenario temperatures could be experienced and ambient temperature is relatively low as Report to substantiate that this is a justify the acceptability of the models used. described in the original Hughes Generic reasonable value for the compartments Otherwise, incorporate elevated ambient Fire Modeling treatments report, in particular where this was applied. temperatures into the zone of influence for IEEE-383 qualified/Thermoset cables. In (This F&O originated from SR FSS-D4) calculations. the case of an initial ambient temperature of 35°C, the expected affect on the ZOI dimensions is within the measurement uncertainty in the field.

10-9 The 2010 peer review identified that 'Except FSS-D8 The method currently employed could result A qualitative assessment has been for the MCR fire scenarios, no other fire in optimistic times for suppression activation performed to assess the potential impact of scenario has used the Non-Suppression or fire brigade response. this F&O.

Probability (NSP) in PTN fire model at this Assess and document the effectiveness of The HGL and MCA analyses credit both time.' Since this review, FP&L has taken suppression with respect to: automatic suppression system and fire credit for suppression (both automatic and brigade actions. In the context of the HGL manual) in the Multi-Compartment/Hot Gas codes and standards, and current fire and MCA, the fire brigade action of interest Layer evaluation. However, this evaluation protection engineering practice, is fire control as that would terminate the does not include an assessment of the fire . possibility for HGL formation. However, the protection system effectiveness. The time available to suppress the fire prior only readily available numeric credit is fire Of particular concern is that fire detection to target damage, suppression credit. To reduce the and/or suppression timing (i.e., thermal Specific features of physical analysis unit conservatism introduced into the analysis, response of the detector and/or sprinkler) and fire scenario under analysis (e.g., fire detection time is ignored for the HGL was not calculated and subtracted from the pocketing effects, blockages that might and MCA. The timeframe associated with time considered for manual suppression impact plume behaviors or the "visibility" of detection and suppression is significantly when using the FAQ-0050 process. In the fire to detection and suppression less than the timeframe required to reach a addition, fire detection reliabilities are not systems, and suppression system hot gas layer temperature which would included in the assessment. If the detection coverage), and impact the HGL analysis.

system does not function as intended, the Suitability of the installed system given the time to detection to initiate fire brigade nature of the fire source being analyzed.

response would be substantially longer.

(This F&O originated from SR FSS-D8) 10-11 The 2010 peer review identified that "fire FSS-C2 The present analysis provides a bounding This F&O has been resolved.

scenario evaluation tools were developed FSS-C3 approach in most cases, since the 98th The recommended resolution involves the based on the Generic Fire Modeling percentile fire heat release rate is used from crediting of growth and decay in the Revision 0 Page V-75

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Treatments. These walkdown/evaluation FSS-G1 fire initiation without growth and burnout, modeling of the postulated fire. The existing tools are based on bounding fires that are However, use of the split fraction method is analysis does not take credit for these assumed to cause target damage at a height based on industry events rather than site variables. A review of the dominant fire above the base fire with the fire burning at specific fire ignition sources and target scenarios found that the risk benefit that peak intensity and without burnout times, configurations. Therefore, this could result might be gained is minimal. Therefore, this Because these tools assume a fire burning in non-conservative frequency estimates of refinement was not performed. The at peak intensity and without burnout, this target damage. resulting categorization of the related SR is SR is considered met at CC I." Include fire growth and decay for risk CC 1. Since the conservatism approach results bigrtie nterslsin some Since the review, FP&L has stated that "The significant fire scenarios. thiserais being retained in the results, use of a panel split fraction to differentiate this CC is judged to be adequate for the between fires impacting the panel and NFPA 805 applications as the conservative components with cables terminating at the bias would tend to result in the over-panel versus panel fires impacting cables estimation of the risk metrics used for this outside of the panel provides an equivalent application.

and more useful two point fire model... The application of the two point treatment to individual fire scenarios is carried through to the MCA/HGL evaluation which addresses the impact of each scenario on MCA."

The Panel Split fraction is developed from a supplemental report (ERIN report, Supplemental Fire PRA Methods, dated February 2010). This document was submitted to the EPRI Fire PRA Methods Review Panel. This review is not complete as of the date of this peer review.

Use of the split fraction method is based on industry events rather than site specific fire ignition sources and target configurations and therefore, could result in non-conservative frequency estimates of target damage.

(This F&O originated from SR FSS-C2) 10-12 The 2010 peer review identified that "The FSS-D9 This appears to be a documentation issue, This F&O has been resolved.

PTN FPRA methodology generally does not but FP&L should confirm that smoke An analysis of the impact of smoke damage include postulation or evaluation of smoke damage has been considered and document has been completed and documented in the damage. Additional review shows that the accordingly. PTN FPRA Scenario Report.

smoke issues do not affect the FPRA results Confirm that smoke damage has been Revision 0 Page V-76

Florida Power& Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update significantly. However, the FPRA does not considered and document accordingly.

include a qualitative evaluation of smoke damage to FPRA equipment."

Since the 2010 review, FP&L stated, "Section 6.2 of the Scenario Report was added to address this concern." However, section 6.2 provides a high level discussion and methodology including the statement that "Exposure time plays a key role in the likelihood of failures from smoke. As a result, damage from short term smoke exposure will only result from severe conditions.... Instruments, control components and all high voltage powered components are exceptionally vulnerable to circuit bridging as a result of airborne smoke and deposited particulates."

However, there is no documented discussion of the smoke damage assessment results, and none of the targets in the scenarios indicated smoke damage as the failure mode. NUREG/CR-6850 recommends considering smoke damage to banks of interconnected panels, and this should be considered.

(This F&O originated from SR FSS-D9) 10-13 A credit for incipient detection is taken for FSS-A6 This appears to be a documentation issue, This F&O has been resolved.

MCB fires (non-suppression probability of FSS-D7 but since the system is not yet installed, The credit taken for incipient detection is 0.02). There is no documentation to justify FSS-H7 there could be an impact to the assumptions consistent with that specified in FAQ this value. Per discussion with FP&L the made within the Fire PRA. 0046.

approach appears to be in agreement with Document the basis for probability of non-FAQ-08-0046. The approach also does not suppression value assumed in analysis.

use the NUREG/CR-6850 Appendix L factor When the incipient system is installed, the for panels that credit incipient detection. FPRA should be reviewed and updated Secondly, the incipient detection system is accordingly.

not yet installed, and therefore, the Fire PRA should be reviewed and updated as needed to reflect any differences between the Revision 0 Page V-77

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update assumed and as-built conditions of the system.

(This F&O originated from SR FSS-A6) 10-14 Beyond the Generic Fire Modeling FSS-A5 For risk significant fire scenarios, detailed This F&O has been resolved.

Treatments, the Fire PRA did not include fire modeling should be performed to ensure The current analysis is consistent with a additional detailed fire modeling for most fire you are not masking the "true risk significant Capability Category I analysis. This compartments. fire areas". Without detailed fire modeling provides a degree of conservatism in the Note 4 (under FSS-A5 of the ASME for significant fire scenarios, the results are analysis which would also tend to over-Standard) states that "once a fire scenario conservative, estimate the change in risk which is reported has been 'selected,' this implies that the Consider performing additional detailed fire for the NFPA 805 application. A review of scenario will eventually be evaluated and/or modeling to provide "reasonable assurance the results of the application analyses quantified at a level of detail commensurate that the fire risk contribution of each indicates more rigorous analyses consistent with the risk significance of the scenario." unscreened physical analysis unit can be with CC IIor CC III would not alter the (This F&O originated from SR FSS-A5) characterized." conclusions of the analyses.

10-15 PTN credits multiple suppression paths for FSS-C7 Lack of dependency analysis could lead to This F&O is resolved.

MCA/HGL evaluation. However, the FSS-G1 an optimistic estimate of suppression A review of the credited suppression dependencies have not been evaluated and FSS-H7 probability. systems in the Multi-Compartment /Hot Gas modeled. For example, fixed suppression When multiple suppression paths are Layer analysis has confirmed that no and fire brigade response may both rely on credited, perform a review and address any dependency exists between the suppression a single detection system. dependencies between suppression and systems and detection systems. Detection (This F&O originated from SR FSS-C7) detection systems credited in the MCA/HGL in the zones with suppression systems is calculation. associated with an independent detection system.

10-16 Review of fire modeling in single FSS-C1 Discounting of secondary combustibles This F&O has been resolved.

compartments does not consider the FSS-D3 when considering localized fire damage Supplemental walkdowns have been addition of HRR from secondary FSS-G1 could lead to non-conservative results, performed to identify and address the combustibles. It is acknowledged that Include secondary combustibles in the heat potential for fire spread for scenarios where secondary combustibles were considered for release rates used for zone of influence the non-383 cables are not protected by the MCA/HGL evaluation. estimates. Flammastic material. The analysis has been Fire spread and additional HRR due to the updated to include these scenarios as resulting cable tray fire and adjacent appropriate.

cabinets would increase the total fire size and the subsequent zone of influence.

Compared to the NUREG/CR-6850 guidance for flame spread along PVC cable (flame spread = 0.9 mm/sec) the estimation Revision 0 Page V-78

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update of HRR for the applied scenarios is non-conservative. Realistic estimation of the scenario HRR is necessary to ensure the full impact of the fire on exposed targets is presented and that the effects of a damaging HGL may also be estimated.

FP&L has stated that walkdowns are in progress to include fire spread to cable trays and incorporate this into the fire scenarios.

(This F&O originated from SR FSS-Cl) 10-17 The 2010 peer review identified that FSS-D1 Modifying the zone of influence to account This F&O has been resolved.

Attachment B of the Fire Scenario Report for HGL effects could impact the defined The MCA/HGL evaluation has been modified (Report 0493060006.004) generic fire target damage set. to address the concern noted in the F&O.

modeling treatments do not account for the The generic treatments used in relatively The potential for a larger zone of influence is effects of hot gas layer (HGL) on the zones small rooms should be scrutinized to ensure addressed via new fire scenarios added to of influence. The limitation indicates that that any HGL interaction is considered and the fire PRA.

because HGL is not considered that these accounted for if found to be significant.

correlations should not be used in enclosed The selection of which generic fire modeling areas with small volumes where a significant treatment is used to define target damage HGL thickness may form. Because this for HGL effects on a scenario basis should relationship is not considered plume be documented in a clear manner to temperatures may be underestimated facilitate updates and peer reviews.

because it is assumed that ambient temperature air is being entrained into the plume, resulting in cooler plume temperatures, rather than heated air from the hot gas layer. Entrainment of heated air into the fire plume results in higher damage heights because the plume remains hotter at higher elevations.

Since this review, FP&L states that "The impact of a hot gas layer on the zone of influence is evaluated for all fire zones/scenarios in the MCA/HGL evaluation." A review of this evaluation confirms that HGL effects on ZOI were in fact considered for the generic treatments; however, there is not sufficient Revision 0 Page V-79

Florida Power & Light Aftachnnent V -Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update documentation in Attachment A to the Fire Scenario Report to determine which zone of influence was applied to which scenario, and whether it was applied correctly to consider the effects of HGL. The Generic treatments include several iterations and combinations of variables, including opening percentage of the compartment. The fire scenario documentation at the time of this review did not provide sufficient information on opening percentage to confirm that the ZOI was applicable to the compartment.

(This F&O originated from SR FSS-D1) 10-18 In at least two cases, transient fire scenarios FSS-Al The exclusion of transients in some This F&O has been resolved.

have not been included in the fire modeling compartments may lead to a non- Supplemental walkdowns were performed to for some compartments (e.g., fire conservative estimate of CDF and LERF. re-assess the treatment of transient fires.

compartments 67 and 68). Per discussion These walkdowns focused on two key with FP&L the transients may have been attributes - the appropriateness of the Include transient scenarios in all excluded based on the dominance of the selected HRR characterization and the frequency of fixed scenarios. However, compartments where fire modeling has been employed. location of the postulated fire scenarios.

transients should only be excluded when The postulated location for the treatment of precluded by design. Based on the size of transient fires was based on where a these rooms, and the presence of secondary transient ignition source might reasonably combustibles, transient fires could lead to occur. The results of these walkdowns were fire growth and eventually HGL, and incorporated into the FPRA analysis.

therefore should be analyzed.

(This F&O originated from SR FSS-A1) 10-19 For fire modeling analysis of transient fires, FSS-H1 Lack of documentation on transient fire This F&O has been resolved.

FP&L implements a floor area weighting locations and boundaries will present a The specific instance noted in the F&O was factor. However, the documentation does challenge for updates and peer reviews, corrected. In addition, supplemental not include a graphical representation of the Update documentation to include a graphical walkdowns were performed to re-assess the assumed transient locations and representation of transient fire locations and overall treatment of transient fires, These boundaries. It is therefore not possible to boundaries. walkdowns focused on two key attributes -

review (or update) transient fires. the appropriateness of the selected HRR Also during review of transient weighting characterization and the location of the factors it appears to have been double postulated fire scenarios. However, the counted in some compartments (e.g., documentation that was generated did not Revision 0 Page V-80

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update compartment 63). Based on discussion with specifically produce graphical FP&L this was due to an error in the Excel representations. Instead, the information based spreadsheet tool for transient was incrementally enhanced to provide a frequency quantification. This appears to be spatial reference to a location with in the an isolated case and will be corrected. space. The need for special depiction of (This F&O originated from SR FSS-H1) transient fire scenario locations will be addressed in conjunction with the development of procedures for post transition configuration control.

10-20 The fire modeling analysis of the Turbine FSS-A1 Lack of consideration of the catastrophic T/G This F&O has been resolved.

Generator (T/G) fires is performed in fire may lead to a non-conservative estimate The analysis documentation has been accordance with Appendix 0 to NUREG/CR- of CDF and LERF. updated to address catastrophic T/G fires 6850. However, there is no discussion Perform a review of the catastrophic T/G fire that may lead to building collapse or other regarding the lack of analysis of the in accordance with Appendix 0 to significant widespread damage. The results catastrophic T/G fire event, which should NUREG/CR-6850, or document the of this update did not identify any new risk consider blade ejection, oil line rupture, and justification for excluding this event at PTN. significant contributors or insights.

hydrogen explosion. Per discussion with FP&L, the catastrophic fire was discounted since the T/G is located outdoors. While this may not result in hot gas layer formation and structural collapse, a review of the guidance is warranted, and inclusion of this event frequency should as a minimum map to the loss of the T/G and if suppression fails, all equipment within the T/G structure.

(This F&O originated from SR FSS-Al)

Page V-81 Revision 00 Page V-81

Florida Power & Light Aftachnnent V - Fire PRA Qualltv Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 10-21 The supplemental generic Fire Model FSS-C3 The current approach results in many This F&O has been resolved.

Treatments: Transient Ignition Source FSS-G1 transient fire scenarios being screened Supplemental walkdowns were performed to Strength includes an assumption for FSS-H2 during detailed scenario analysis. The re-assess the treatment of transient fires.

transient burnout of 12 minutes. This results of the FPRA are therefore potentially These walkdowns did not identify any burnout time is based on an assumed fire non-conservative for the analyzed detailed instances where an altering of the transient loading and the 317kW heat release rate, scenarios, fire duration had any material impact on the and appears to be optimistic given the Provide additional justification for the applied HGL and MCA. The documentation has uncertainty in transient fire loading. The transient fire analysis as a screening also been updated to address the criteria burnout is then used to develop a zone of approach. Consider increasing the burnout used for selecting the characteristic transient influence for thermoplastic targets, based on time and using the NUREG/CR-6850 fire HRR. The approach is consistent with the thermal response tables in Appendix H recommended damage threshold to 205°C the recently issued guidance from the to NUREG/CR-6850 for thermoplastic cable to bound uncertainties in fuel loading for EPRI/NRC review panel. The results of at 260'C. Since this resultant vertical zone transient fires, these walkdowns were incorporated into the of influence is used to screen transient FPRA analysis.

scenarios from impacting secondary targets The twelve minute fire corresponds to the higher than 7.3 feet from the floor, additional 317 kW el minute only a nd s tet s justification is needed 12 minute fire, to demonstrate and subsequent use of that 260°0aC 317lbkW 35 of fuel Classpackage onlyAdditional A material. and represents damage threshold is appropriate for discussion is provided in Rev. 0 of screening purposes.

Supplement 3 of the Hughes Generic Fire Modeling treatments that examines the fire Also noted is that Attachment B to the Fire durations and test durations of all Scenario Report zone of influence does not NUREG/CR 6850 tests. It is shown that the reflect the same values recommended by method used to determine a 12 minute fire the Generic Fire Model Treatment. As an predicts or overestimates the fire duration in example, the differentiation between all cases and is therefore a sound approach.

transient Severe and Non-Severe categories is not based on a 317kW fire. This appears to be a documentation issue only.

(This F&O originated from SR FSS-C3)

Page V-82 Revision 0 Page V-82

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 10-22 Per NUREG/CR-6850, appendix H, FSS-C6 For smaller volume rooms, estimates of This F&O has been resolved.

temperature sensitive equipment should be FSS-G1 equipment damage may be non- The consideration of sensitive electronics considered to fail at 65°C. Supplemental conservative, was addressed in a qualitative fashion in the Generic Fire Model Treatments: Hot Gas Apply the appropriate hot gas layer and Scenario Report.

Layer Tables includes new zone of influence zone of influence for temperature sensitive and hot gas layer treatments for temperature equipment where applicable.

sensitive equipment. However, per discussion with FP&L these have not been implemented in the fire scenarios.

(This F&O originated from SR FSS-C6) 10-23 The PTN FSS report 0493060006.004, Rev FSS-C5 Equipment damaged by suppression This F&O has been resolved.

2, section 6 discusses the damage criteria activities may impact estimates of CDF and The specific issue raised in the F&O is for thermal, smoke, and sensitive LERF for some scenarios, beyond the scope of the associated SR. In equipment. However, suppression effects Perform an assessment of electrical addition, no known consensus method do not appear to have been considered for equipment that may be vulnerable to water exists for treatment. A qualitative the potential to damage equipment. intrusion from suppression activities (or assessment, based on other guidance for (This F&O originated from SR FSS-C5) thermal shock from gaseous systems), and evaluation of potential impact of suppression include any additional failed equipment, not effects was performed which indicated that already considered damaged by fire, in no specific change in the analysis is needed.

scenarios as appropriate.

Page V-83 Revision Revision 0 Page V-83

ATTACHMENT 3 TO L-2012-354 REVISED TABLE V-3 Replaces Table V-3 of L-2012-092 Florida Power and Light Company Turkey Point Nuclear Generating Station Units 3 and 4 License Amendnment Request No. 216 Transition to 10 CFR 50.48(c) - NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Electric Generating Plants, 2001 Edition

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 1-1 In numerous significant scenarios, the CF-Al The overall Fire PRA results appear to be This F&O has been resolved.

spurious operation probability is assumed to greatly impacted by setting spurious At the time of the Peer Review, the FPRA be 1.0 (true) for any events where spurious operation probabilities to 1.0. Scenario had only a very limited credit for fire induced operation can occur. For example, in 79ALA, which is presently 8E-06 would be spurious actuation probability. The specific scenario 79ALA (one of the top 5 scenarios reduced by at least an order of magnitude instance identified in the F&O was updated.

in unit 3), three events are set to true by assigning spurious operation probabilities The resolution of this F&O also included a affecting the top cutsets; GMMOGE100 to several events. Similarly, with 79AKA, and review of significant fire initiating events and (MOVs 878A or B spuriously operate), 79AJA also at 8E-06. additional credit for hot short induced MAVK3CV303A, OHTX3CNTRL. Capability Perform Circuit Failure Probability Analysis spurious operation was applied in the Category I requires setting spurious for significant spurious operations events, analysis as appropriate. In all cases, the operation probabilities to industry accepted and modify the FRANC model to assign a application of the spurious actuation factor is values. It appears most of the events set to Perform Circuit Failure Probability Analysis consistent with the guidance in NUREG/CR-true would be either MOVs (0.33) or AOVs for significant spurious operations events, 6850 and FAQ 08-0047.

(0.62) or similar, and should not be set to and modify the FRANC model to assign a true for significant fire scenarios. Analysis probability for the event in the cutsets. In using the specific circuit configuration order to meet CCII, the spurious operation for each significant spurious operation would probability should be based on the specific be required for CC II, and may lead to circuit configuration for each significant different results than the generic values, spurious operation.

depending on the circuit design and cable affected.

1-10 Transient Fires are postulated in all fire IGN-A9 Systematic issue. Appears as if numerous This F&O has been resolved.

compartments, as listed in Appendix B and compartment transient frequencies were A sensitivity evaluation was performed that Table 3-6 of the Ignition Frequency Report. underestimated, while others would have involved increasing the weighting factor for All factors affecting the fire frequency were been slightly over estimated as a result. occupancy and storage from 'low' to assessed based upon a slightly modified Initial review was confirmed by walkdown of 'medium' for all instances where such a NUREG/CR-6850 approach. However, the 5 areas. The ranking on all 5 areas did not condition could reasonably be expected to rankings that were provided do not appear appear to match the walkdown teams occur. The results of this sensitivity found to be consistent with the methods in estimate for each area. that the impact on the calculated CDF for NUREG/CR-6850, result in an Re-assess the transient fire rankings per the each unit was less than 1 E-7. Given this underestimate for fire frequencies in some Guidance in NUREG/CR-6850. Confirm the small impact, the existing analysis is areas, and an over estimate in other areas. rankings by walkdown of each area, taking adequate for the application.

One F&O is provided on this SR. In into account the actual condition.

particular: a) Areas were ranked as zero in maintenance, occupancy, or storage even though entrance to the areas is physically possible, b) Areas were ranked as 1, even though activities were not prohibited by plant procedure.

Revision 0 Page V-1 2

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update In areas where the room is sealed during operation (roof plugs), transients could have been left in the room prior to sealing, so the ranking on this factor should not be zero -

per the 6850 guidance. During the walkdown, Compartments 70 and 71 both had permanently stored breaker grounding devices, with poly-covers, and 71 had a temporary transformer for the polar crane (operating). Both should be ranked as

'medium' for storage. Similarly, the cable room had storage of 3 temporary fans, cables and blankets and should be marked as medium for storage. This room also appears to include numerous components that will likely be worked on during power, (ranking moderate for non-hot work), and numerous people were present during our limited walkdown. Compartment 88, an open area in front of the switchgear room, had numerous combustibles stored and located, and should probably be marked as medium or high (presently marked as low). Both area 85 and 88 have frequent foot traffic, and should be marked as medium for occupancy. 85 appears as if it should be moderate for storage (no controls). Similarly; no controls appear to be in place for 116.

The above are samples of identified issues, based on our limited walkdown. It appears there will be similar issues with other areas in the plant. We looked at other areas adjacent to the areas we were in (compartments 87, 84, etc), and expect similar problems with the present rankings.

(This F&O originated from SR IGN-A9) 1-17 Table 3-2 includes uncertainty values (EF) IGN-A10 Systematic Issue. This F&O has been resolved.

for prior and posterior values. However, QU-E3 Estimate EFs for significant fire The quantitative uncertainty analysis was Error Factors are not propagated to the compartments. ESTIMATE the uncertainty prepared subsequent to the peer review.

Revision 0 Page V-13

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update compartment specific ignition frequencies. UNC-A1 interval of the CDF results. ESTIMATE the A parametric uncertainty evaluation that The other parameters, such as conditional UNC-A2 uncertainty intervals associated with considers fire ignition frequency as well as failure probabilities for circuit failures, do not parameter uncertainties(DA-D3, HR-D6, other variables was performed that uses a have uncertainty intervals. HR-G8, IE-C15), taking into account the Monte Carlo sampling process. The results The lack of uncertainty intervals would not state-of-knowledge correlation. of the analysis showed a mean that was generate meaningful uncertainty interval of slighter higher than the calculated results the CDF/LERF results. (This F&O originated which was expected.

from SR IGN-A10) 1-18 During walkdowns, several key areas IGN-A7 Appears to be missing components in This F&O has been resolved.

appeared to have ignition sources not numerous areas, based on a limited The specific instances identified in the F&O included on the ISDS. For example, in the sampling during walkdown. were reviewed and the analysis updated cable spreading room, 2 transformers were Perform a re-verification of the ISDS for accordingly. In addition, the supplemental in the compartment (3X033 - 75KVA, 3X130 significant fire areas in the FPRA. Add walkdowns that were performed as part of

- 45KVA), both within the screening distance missing components to each ISDS, where ongoing analysis refinements efforts for the of targets. Also in the compartment is CP- applicable, significant fire areas did not identify any 600 spectralink cabinet, an open cabinet, other omissions.

the RCP Vibration Monitoring Cabinet, 4P21 and 4P09 instrument AC panel. Note; we did not do a 100% review of the CS room, so additional cabinets may be missing. See also F&O 1-19. (This F&O originated from SR IGN-A7) 1-19 It appears the Ignition Source Counting did IGN-A7 Appears to be a systematic issue in the This F&O has been resolved.

not count Lighting Panels or other similar FPRA. A re-assessment of the lighting panels was panels. For example, there were at least 8 Include unsealed lighting panels and similar performed. The re-assessment focused on lighting panels in the cable spreading room electrical cabinets in the ISDS as potential the need for treatment as a fire initiating that were not on the ISDS. Additional similar ignition sources, event. No effort was undertaken to alter the panels are located in most electrical rooms population of electrical cabinets considered we walked down, such as the switchgear in the fire frequency development.

rooms and other electrical rooms. Based on Therefore, the existing values potentially our walkdowns, many of the lighting panels have a conservative bias. The assessment should be included in the ISDS, based on did not identify any instances were explicit guidance in 6850 and the subsequent FAQ treatment as a fire initiating event was on sealed cabinets. A review of the generic needed.

guidance provided for ignition counting did list the screening of small, wall mounted cabinets (sealed). However, the lighting panels do not appear to meet the criteria Revision 0 Page V-14

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update listed in the procedure (not sealed, numerous switches/breakers), etc. Many of the cabinets are located close to cable trays or other intervening combustibles, so a small fire could result in a larger fire due to spreading. (This F&O originated from SR IGN-A7) 1-2 Section 4.1of the Component Selection AS-B1 The significance of not identifying This F&O has been resolved.

Report mentions: "Since the FPRA ES-Al components as causing initiating events is The FPRA assumes each postulated fire quantification calculates a fire CCDP and the basically that the assumed model impact is results in at least a reactor trip. Logic is initiating event frequency for each zone is ESA3 accurate by modeling a reactor trip with a included in the model so that appropriate based on the fire ignition frequency, the ES-A4 subsequent failure of the function, rather event tree is quantified if the fire induces a initiating event faults are not required to be FQ-A2 than modeling the initiating event itself. In different type of event (event tree). The used for FPRA quantification." Fault tree some cases, this impact is a matter of timing overall structure of the FPRA model was initiating events were not impacted by the for operator actions. In the case of this reviewed to address the specific item component mapping, and are therefore not FPRA, the HEPs have been conservatively identified in the F&O and to confirm changed by fire damage. As a result, set assuming a loss of MFW as a starting appropriateness of overall treatment. The equipment associated with Fault Tree point. However, the fault tree initiating only change that was required was related initiating events were not identified as events include loss of CCW, loss of HVAC to biasing the application of recovery actions components potentially causing a fire- and others. It is not clear that the present so that they were based on an assumed loss induced initiating event. 163 events are model accurately determines CDF/LERF of MFW.

screened in Table A of the Equipment results for systems impacted which may Selection Analysis based on being cause a complicated reactor trip (special associated a initiating event).

fault tree initiating event. Most are modeled Modify FPRA to model the fire impact to in other system models. However, Several Fault Tree Initiating Events, and analyze the were found to not be modeled in the rest of FPRA assuming a fault tree initiating event the model: CPSD3PC61 1, CPSD4PC611 for those areas where the initiating event and 2 related failures. A few others (Cooling can occur.

units) do not appear to be modeled elsewhere.

(This F&O originated from SR ES-Al) 1-25 There does not appear to be a review of FQ-E1 Requirement of QU-D5 as called for by FQ- This F&O has been resolved.

non-significant cutsets in the PRA QU-D5 El Review of non-significant cutsets performed documentation. Perform a review of non-significant cutsets and documented.

(This F&O originated from SR QU-D5) and accident sequences, as discussed in QU-D5 for the FPRA.

Revision 0 Page V-1 5

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 1-27 Significant fire compartment contributors to FQ-E1 Requirement of LE-F1, F3. This F&O has been resolved.

LERF are report.

summary documented in Appendix However, C of the the contribution LE-F1 Document the contributors to LERF based Added LERF top cutsets and importances from plant damage states is not provided or LE-F2 on the requirements of LE-F1 of the internal run as well as sensitivity analysis in events section of the standard, as required Summary Report. Also performed and the contributors from LEB SRs. Sources of LE-F3 by FE-QI. Document the Sources of documented the uncertainty evaluation for uncertainty, including sensitivity analysis UNC-Al uncertainty, including sensitivity analysis LERF.

performed, are not evaluated for LERF. performed for CDF in Appendix D of the (This F&O originated from SR LE-F1) Summary Report.

1-3 The internal events PRA model has AS-B1 As a result of assuming a reactor trip and This F&O has been resolved.

numerous locations in the model where the ES-Al not mapping components/equipment to The issues and concerns identified in the specific initiating event results in a model ESA3 modeled internal initiating events; the risk F&O related to the fire-induced initiating impact. For example, under gate U3QT07 - can be under-estimated. In this case, since events were reviewed. The review found initiating events that can cause a PORV or ES-A4 the general approach used is systematic, several instances where a change to the SRV to lift are ANDed with the failure to FQ-A2 this problem is difficult to determine without modeling was required to allow the existing reclose the PORV or SRV. In this case, significant effort to combine the impact of treatment methodology to be retained. The special initiator %ZZIP6U3 is identified as an each modeled impact. In most cases, the review did not identify any instances where initiating event that will cause a PORV lift, modeling results in non-conservatism in the specific fire initiating event logic beyond that along with %ZZT2U3. Equipment that can result. However, the fix for feed-and-bleed already in the model was needed.

cause each are not mapped or modeled in resulted in conservatism for most of the the Fire PRA. As a result of a previous scenarios where FW is not initially lost. In review, the modeling of Feed-and-Bleed was either case; whether modeled conservatively changed to assume a loss of feedwater (low or nonconservatively, the standard SG level) occurred. The shorter time results requirements in this area are to model the in a higher HEP for feed-and-bleed in all impact of the FPRA accurately.

scenarios, regardless of whether a loss of Map all identified internal events initiating FW occurred. However, numerous other events to the specific components that can modeling impacts can occur, that are not cause the event, and modify the FPRA to modeled. Under gate 162115, logic for determine the CCDP based on the fire-HVAC unit 3S230 failure to start is included induced initiating event that results.

when a Loss of offsite power would occur.

This logic is applicable only for when a LOOP occurs, and not applicable for non-LOOP events. This type of logic is contained throughout the internal events PRA modeling. Another example is under gate El 104A, where loss of DC power results in lockout relay failures. There are many other examples throughout the PRA. Additionally, the identification of the specific initiating Revision 0 Page V-1 6

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update event for quantification was not performed per the requirements of FQ-A2. For quantification, the modeled initiating event is assumed to be a reactor trip in all cases.

This treatment does not meet the intent of SR FQ-A2, where the quantified model should encompass the risk contribution from all applicable initiating events.

1-34 No evidence was found that supported FSS-G4 Systematic issue. This F&O has been resolved.

confirmation of conformance of fire rated Provide the documentation that supports The treatment of barriers in the MCA is barrier segments to applicable test confirmation of conformance of fire rated based on information in the Fire Hazards standards. Additionally, the effectiveness, barrier segments to applicable test Analysis and supplemented with walkdown reliability, and availability of any passive fire standards, and the barrier effectiveness, observations. The analysis documentation barrier feature credited does not appear to reliability and availability, was updated to provide this information.

be performed. (This F&O originated from The MCA was modified as needed to SR FSS-G4) incorporate the results of this effort.

1-37 Significant contributors to Fire PRA results FQ-E1 Requirement of QU-D7 This F&O has been resolved.

are included in Section 4.3 and the QU-D7 Provide importance measures as required Importance measures for CDF and LERF appendices of the Summary Report. This by QU-D7 and FQ-E1 have been determined and added to the includes a list of operator actions that Summary Report.

contribute to CDF. However, no importance measures are provided for CDF or LERF.

(This F&O originated from SR QU-D7) 1-38 Results of the Fire PRA did not include the FQ-F1 Systematic Issue This F&O has been resolved.

following: (e) the total plant CDF and QU-F2 Provide required documentation per QU-F2 The documentation of the analysis results contributions from the different initiating UNC-A2 and FQ-F1. has been expanded to include the events and accident classes (i) the information noted in the F&O. These results uncertainty distribution for the total CDF () were also reviewed for reasonableness and importance measure results (I) asymmetries no issues or concerns were identified.

in quantitative modeling to provide application users the necessary understanding of the reasons such asymmetries are present in the model (m) the process used to illustrate the computer code(s) used to perform the quantification will yield correct results process. Some of these issues are listed in other F&Os.

Revision 0 Page V-17

Florida Power & Light Aftachnnent V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update However, item e(accident classes), I (asymmetries) and m (validation of computer codes) is not covered elsewhere. (This F&O originated from SR QU-F2) 1-40 The quantification of significant basic FQ-F1 Requirement of QU-F6 and FQ-Fl. This F&O has been resolved.

events, cutsets and accident sequences is QU-F6 Provide the quantification of significant basic The Summary Report has been updated to not provided. Additionally, the definitions UNC-A2 events, cutsets and accident sequences, provide the importance measures of the used for significant basic event, significant and the definition used for significant basic model basic events, top 90% of all plant cutset, and significant accident sequence event, significant cutset, and significant cutsets, and a review of the scenarios are not provided. (This F&O originated from accident sequence contributing more than 1% of the total risk.

SR QU-F6) 2-1 The plant partitioning task does not include PP-B1 Section 3.11.5 of FHA states that man-hole This F&O has been resolved.

detailed discussion with respect to this SR PP-B7 covers are justified as three-hour fire Walkdowns of fire zone boundaries were PP-B7. The manholes are modeled as boundary although they need not to be performed and documented in support of a separate fire compartments. However, no specifically rated as fire barrier. Therefore, review of the Fire Hazards Analysis update.

walkdown for these manholes has been the modeling of manhole as fire Additional discussion regarding the basis for performed. No justification for the modeling compartments is considered acceptable the ignition frequency for the manholes was approach has been provided except being although no walkdown has been performed added to the documentation.

briefly mentioned in Section 2.2 of Report for the manholes. Other credited barriers are PTN-PSA-7.01 Revision 2. Walkdowns were discussed in PP-B2-4 above also not documented for spatial separation Consider adding justification for the or other boundaries that are not fire rated modeling of manholes according to the but was credited in the FPRA. (This F&O requirements in SR PP-B7. Consider originated from SR PP-87) performing walkdown for manholes with significant risk contribution. Also, document walkdowns on all credited, nonrated barriers credited in the FPRA.

2-44 Uncertainty Evaluations (Sensitivity studies) QU-E4 QU-E4 requirements. This F&O has been resolved.

should be performed for both CDF and UNC-A1 Perform sensitivity studies should be Parametric uncertainty and sensitivity has LERE model for Units 3 and 4 since the UNC-A2 performed for both CDF and LERF model for been performed for CDF and LERF for both model uncertainties may have different Units 3 and 4. Units. The results do not indicate any impact to specific model due to differences change in the selection of parameters or in plant designs, FPRA model details, and assumptions are necessary.

etc. (This F&O originated from SR QU-E4) 3-2 Credit for fire compartment separation via PP-B1 As noted in the description non-fire rated This F&O has been resolved.

non-rated construction was commonly construction is credited for separation of fire Revision 0 Page V-18

Florida Power & Liqht Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update noted, e.g., according to the FHA the walls PP-B2 compartments, however no Fire PRA The configuration and construction of non-of fire compartment 034 are not fire rated specific justification for the validity of the fire fire rated barriers was confirmed using a and they provide separation from fire compartment is provided. This is considered combination of information in the Fire compartments 036, 035, & 058. Separation a systematic issue for the FPRA. Hazards Analysis and supplemental plant of FC 034 from the surrounding FCs is one Provide FPRA specific justification of the walkdowns. The anlaysis and related of many examples where non-fire rated construction separating fire compartments, documentation was updated to provide this construction is credited for separation. Use where the barriers will substantially contain information.

of this level of separation is acceptable the fire.

provided the separation is justified.

However, the justification does not appear to be provided for the FPRA. (This F&O originated from SR PP-B2) 3-3 A few cases of special separation are PP-B1 Two instances were identified where spatial This F&O has been resolved.

credited in the PB&P. Most notable are PP-B3 separation is credited for the separation of Openings between fire zones were separation of Fire Compartments 058 and fire compartments. No justification is addressed with respect to targets on the 037 and 004 and 010. The FHA notes in the provided for this separation. other side of an opening which are within the write-up for fire zone 004: 'There is a partial Provide justification for the use of spatial zone of influence of an ignition source.

height concrete wall on the South side of separation in the FPRA. If not justified, Targets were evaluated for fire damage this room with a full height opening to Fire combine the compartments in the FPRA. regardless of the zone in which they were Zone 10'. No justification is provided for this located.

separation, credited hence itmay separation is notbeclear that the expected to The Temulti-compartment analysis considered credtedsepratin tothe my b expcte volume associated with adjacent zones contain the effects of a fire. Accordingly the th openings ciaten the zones effect of a fire beyond the identified fire with openings between the zones in compartment boundary may occur. While evaluating the potential for hot gas layer this effect would be expected to be identified formation.

through performance of the multicompartment analysis the level of documentation provided in support of the PB&P does not satisfy the standard requirements. (This F&O originated from SR PP-B3) 3-4 The PTN self assessment points out that the PP-R1 As discussed in the description This F&O has been resolved.

FHA documents the use of active fire barrier PP-1B5 justification/discussion is not provided for The walkdowns that were performed did not features as necessary for fire zone crediting active fire protection features in observe any open fire doors (active separation. However in cases where fire barriers that are identified as non-fire rated features). The documentation for the fire compartment separation is provided by structures. It is not clear if active features scenario development process was updated unrated barriers there may be active such as fire dampers exist in these barrier to provide the criteria and methodology that Revision 0 Page V-19

Florida Power& Light Attachment V - Fire PRA Quality FlriaPoe & ih tahetV-Fr R ult Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update features that are not identified by the FHA segments because the FHA does not rely on were used.

but credited by the Fire PRA. In such cases them for separation. Documentation should active fire barrier features may be be provided that clearly establishes what unknowingly credited for separation but not features are credited in such barrier adequately maintained by the fire protection segments and why makes them acceptable.

program. Because these elements were not Given the large number of barriers credited purposely identified within the development in the FPRA that are discussed in the FHA, of the Fire PRA it is unknown if the Fire but without discussion of active elements, Protection Program identifies all of the there are likely a number of undocumented necessary features. Because the Fire PRA active elements in these barriers.

does not formally define and justify these Determine the active fire barriers on barriers features this element is judged not met. credited in the FHA (not SSA), and provide (This F&O originated from SR PP-B5) justification for any active elements credited in the FPRA.

3-5 According to the Section 3.13 of the PTN SF-Al As discussed in the description no This F&O has been resolved.

FPRA Summary Report the effect of an discussion was found that specifically The low seismic spectra applicable to the earthquake on ignition source scenarios is addresses fire ignition source scenarios that Turkey Point site have been validated via discussed in the IPEEE and Potential Fire may arise from an earthquake. Also, since the IPEEE with respect to the potential for Related Vulnerabilities self assessment, these scenarios are not identified a causing unique fire scenarios. Their Review of the Potential Fire Related qualitative assessment of their risk potential for causing damage to pipes or Vulnerabilities self assessment did not significance is not included, tanks containing combustible gases or reveal an analysis that specifically The analysis provided in the Potential Fire liquids or to initiation of electrical fires is addresses generation of fire ignition source Related Vulnerabilities self assessment considered negligible.

scenarios which could result from an should be expanded to look for unique earthquake, nor does this assessment ignition source scenarios that may arise from address the potential risk significance of an earthquake and a discussion of the risk these scenarios. This assessment does significance of these scenarios should be identify fire vulnerabilities in terms of fuels, qualitatively assessed.

ignition sources, and oxidizers however these discussions are not specific to seismic events nor do they include evaluation of special ignition scenarios that may arise from an earthquake. (This F&O originated from SR SF-Al) 3-7 According to report PTN-PSA-7.01 The IGN-A1 As discussed in the description the revised This F&O has been resolved.

generic fire ignition frequencies provided in IGN-B4 generic fire frequencies contained in FAQ The guidance provided in FAQ 08-0048 NUREG/CR-6850 were used to establish the 08-048 are not incorporated into the PTN requires the use of the original fire ignition frequencies for PTN. While the fire frequencies nor is there justification for Revision 0 Page V-20

Florida Power & Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update use of these values is not entirely incorrect, their exclusion. This SR requires use of the NUREG/CR_6850 fire frequency values as a this SR requires the use of'current nuclear current nuclear power industry event history sensitivity study. Rather than perform two power industry event history that includes or justification for data exclusion. Because analyses, the PTN analysis was developed power plants of similar type, characteristics, the fire ignition frequency methodology does using those original values for the NFPA 805 and vintage.' Accordingly this requirement not address the data contained in FAQ 08- application.

requires use of the EPRI revised generic fire 048 this SR is considered not met. Use of The application of the non-segregated bus frequency values included in FAQ 08-048 or the NUREG/CR-6850 values results in a duct information from FAQ 07-0035 is not justification for its exclusion. Also, it appears conservative estimate of CDF/LERF. FAQ applicable as the plant does not use non-that FAQs 07-35 (bus ducts) and 08-44 35 can have significant impact on fires in the segregated bus duct. The connections to (MFW pump fires) were not incorporated into area of bus ducts. However, it is not the station transformers are made using the FPRA. (This F&O originated from SR apparent if this is important for Turkey Point. cables. FAQ 08-0044 was also not needed IGN-A1) FAQ 44 can result in a lower MFW large fire and the conservatism associated with frequency. original method did not adversely affect the The fire ignition frequency information results.

contained in FAQ 08-048 should be incorporated into the PTN fire ignition frequencies. Additional FAQs should also be incorporated into the FPRA.

3-8 Review of the plant-specific fire events for IGN-A4 As discussed in the description review of the This F&O has been resolved.

outlier experience indicates that some fires identified in Appendix A reveals fires events may have been considered outliers that may have become challenging had they or unknown if the selection criteria had not been extinguished early. The selection The scope of plant specific fire events were considered treatment of fires that are criteria for challenging fires contained in re-assessed with an expanded group of extinguished prior to full development as Appendix A is based on section C.3.3.1 of plant personnel with particular focus on the potentially challenging. Several cases NUREG/CR 6850, however the criteria subjective criteria from C.3.3.2. The results identified in Appendix A of the Fire Ignition contained in C.3.3.2 is not included; had the of the re-assessment affirmed the previous Frequency Development Report, PTN-PSA- criteria of C.3.3.2 been included more fires dispositions.

7.01 may have developed into challenging may have been selected as challenging or fires had they not been discovered and identified as unknown.

extinguished early in their development.

Fires 7, 8, 9, 21, 22, 27, 30, 31 appear to be The criteria for selecting challenging fires in potentially challenging fires (or unknown). Appendix A of the Fire Ignition Frequency See also the previous assessment from 9- Development Report, PTN-PSA-7.01 should

09. (This F&O originated from SR IGN-A4 be revised to include the criteria contained in C.3.3.2 of CR/NUREG 6850 and the fire events should be revisited to determine if additional fires should be selected.

Page V-21 Revision 0 Page V-21

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 4-17 Per summary report, Task 9 is fulfilled with CF-B1 Incomplete evaluation and document for This F&O has been resolved.

the NISYS SSD database, "PTN NFPA 805 circuit failure. The NISYS DB can include Circuit failure probability was considered for Database.mdb". This database has been identification of when spurious operation high risk scenarios and only in cases where significantly expanded for the NFPA 805 may occur, but does not provide the circuit doing so would result in a reduction in total tasks. A sample circuit analysis worksheet analysis or circuit failure probability analysis risk. Additional details with respect to circuit (e.g., for component 20ASB/G3) has needed to support the FPRA. configuration and raceway type have been signatures at the bottom, which were not Provide a documented basis, and detailed added to the altered events table.

populated yet. The NISYS circuit analysis is circuit analysis for any spurious operation an Appendix R type circuit analysis and probability used in the FPRA per Tasks 9 does not identify the circuit failure modes and 10 of NUREG/CR-6850 (or equivalent).

and address likelihood of failure. Failures of the required cables identified are assumed to have a probability of 1.0 unless specifically modified in the ALTEREDEVENTS table of the FRANC model. The treatment of the circuit analysis seems to be bounding (i.e., the likelihood was not part of the analysis). Although Appendix D of the fire scenario report states the bases for the altered FRANC event probabilities, it seems that there is no linking between the altered probabilities and the circuit analysis package. The majority of the altered events are based on operator manual actions while some based on the simple spurious actuation probabilities from NUREG/CR-6850, which were based on specific evaluation (with no basis provided in the FRANC database), but do not directly linked to any specific circuit analysis worksheet. Since the "basis" column of the Altered event table in the FSS report does not appear to include sufficient documentation to allow review/peer review of the results and the NISYS database does not include the analysis, the analysis (not the results) has not been documented. The evaluation and documentation of the review of the fire-induced circuit failure modes and the assignment of the appropriate industry-Page V-22 Revision 00 Page V-22

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update wide generic values to their conditional failure probabilities for risk-significant contributors based on the specific circuit configuration under consideration should be included in the circuit failure report for Tasks 9 and 10. (This F&O originated from SR CF-B1) 5-11 Review of Turkey Point NISYS NFPA 805 ES-B1 The deviation between the ESFAS This F&O has been resolved.

Compliance Assessment Database within components actuation and Control Room SI The circuit analysis process used for the the Cable Routing and Respective components actuation should be disposition project has been confirmed to be consistent Equipment table, it was noticed that the and reconcile, to ensure Fire Safe Shutdown with the latest industry guidance (NEI 00-Spurious ESFAS signal / Appendix R equipment are appropriately 01). In addition, the asymmetry was "Spurious/ESFAS/Lacks/Analysis" have total credited in the Fire PRA. discussed with plant staff and confirmed to of 56 respective components impacted. Unit Reconcile the FPRA component list with the be reflective of the actual plant design and 3 Train A SI signal from the Control Room SSA component list for equipment impacted configuration.

"3MRASI/3C06/3QR43/006" have total of 29 by an SI signal respective components impacted, Unit 3 Train B SI signal from the Control Room "3MRBSI/3C06/3QR45/006" have total of 28 respective components impacted, Unit 4 also have similar components impacted. The concern is the potential mismatch between FPRA and the SSA component lists. (This F&O originated from SR ES-B1) 5-13 Turkey Point FPRA Summary Report FQ-A3 It appears that there is inconsistent basic This F&O has been resolved.

NUREG/CR-6850 Task 16 Report No. event mapping between the database files. The identified data differences were 049306006.005 Rev. 1 Tables A-i, A-2, B-1 A sensitivity run was performed by copying reviewed and confirmed to be reflective of and B-2 documented the Units 3 & 4 Fire the U4 events to the U3 tables, and re- the design and layout of the units.

PRA quantification Results for both CDF and evaluated U3 CDF. The results are the top Additional comparison of the quantification LERF for all fire scenarios that were scenario in 96 dropped from 4.5E-05 to 1 E- results between the two units was also quantified. Scenario 096-A was randomly 06. Based on this, the error appears to be performed to ensure that any significant picked review for both Units 3 & 4. The significant. differences in results are consistent with the CDF/LERF results are consistent between Need to ensure that the altered events table actual unit differences. Various the Summary Report and Zone Scenarios in is correctly developed for both U3 and U4 for asymmetries in the plant layout were database files, Unit 3 CDF the CDF and LERF quantification. identified.

"PTNFIRE_W_LERF MH ESF.mdb", Unit 3 LERF"PTNFIRE_W_LERFMHESF.mdb",

Unit 4 CDF Revision 0 Page V-23

Florida Power& Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update "U4PTNFIRE W LERFMHESF.mdb",

and Unit 4 LERF U4PTNFIRE W LERF MH ESF.mdb".

However, reviewing the Altered Events table in each database files shows inconsistent basic events impacted between Unit 3 and

4. Unit 3 have no basic event impacted, while Unit 4 have 9 basic events listed. (This F&O originated from SR FQ-A3) 6-10 The altered events table in the FSS report FQ-A1 This approach is not consistent with the level This F&O has been resolved.

includes several instances where a single FQ-A4 of detail modeled elsewhere in the PRA. The methodology and the analysis has been basic event combines a hot short spurious HRA-E1 Translate specific failure modes into basic updated to eliminate the use of this operation likelihood with an HEP to recover events and avoid combining disparate failure approach. The use of altered.events for the spurious operation. For example, HR-I1 modes into combined basic events, spurious probability is used only as required ORZR30455C represents a combination of HR-12 and a singular value.

spurious opening of a PORV and operator QU-A3 human error probability to close the PORV.

Supporting requirement FQ-A1 addresses the need to translate specific failure modes into basic events. Embedding an HEP with a spurious operation likelihood bypasses this requirement, and this approach is not consistent with the level of detail modeled elsewhere in the PRA. Also, the approach prevents the ability to address the state of knowledge correlation. (This F&O originated from SR FQ-Al) 6-20 The parametric uncertainty associated with CF-A2 Step not performed This F&O has been resolved.,

conditional circuit failure probabilities are not UNC-A2 Step not performed Parametric uncertainty has been performed evaluated and are not incorporated into the for CDF and LERF for each unit's FPRA.

model. (This F&O originated from SR CF-A2) 6-9 The parametric uncertainty analysis as FQ-A4 Step not performed. This F&O has been resolved.,

discussed in QU-E3 (estimate of uncertainty QU-A3 Perform the FPRA uncertainty analysis, Parametric uncertainty has been performed intervals, etc.) is not performed. Also, the including estimates of uncertainty bounds, for CDF and LERF for each unit's FPRA.

"state-of- knowledge" correlation between Revision 0 Page V-24

Florida Power & Light Aftachnnent V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update fire-specific event probabilities (e.g., per the requirements of QU-A and QU-E.

suppression system unavailabilities, fire When performing parametric uncertainty ignition frequencies, hot short conditional calculations, ensure uncertainty intervals for probabilities, etc.) hasn't yet been applied, event probabilities utilized by the FPRA are (This F&O originated from SR QUoA3) correlated when significant.

7-1 A review of the quantification results for PRM-A3 The method in which the fire-induced This F&O has been resolved.

selected compartments involving fire- PRM-A4 spurious safety injection actuation and A review of the model was performed and induced safety injection actuation and fire- spurious opening of atmospheric dump revision made to address and resolve the induced opening of atmospheric dump RMB5 valve initiating events were linked into the issue identified in the F&O. Additional valves was performed to verify that the PRM-B9 fault tree produces conservative results that reviews were performed as part of the modeling was consistent with the internal could impact the determination of significant overall results and cutset reviews and no events PRA treatment of similar initiating contributors to fire-induced risk. additional instances were identified.

events. This review revealed that duplicate Review the quantification results for the fire-cutsets were being introduced by the induced initiating events to verify that the manner in which the new logic for capturing results are consistent with the comparable the fire-induced initiating events was linked internal events.

into the fault tree. For example, a review of Review the mapping of the fire-induced cutsets for zone 098-A showed that the top initiating event impacts to ensure that they two cutsets were identical except that one i ni stent i th t ompare th a l used bleed aand version feed of the HFE based for alignment on reactor trip of are eventsconsistent initiator,witthate that comparable appropriate internal differences bleedrinndifeedGbasedevnlrendtor orip rdue to the considerations of the fire PRA are occurring with SG low level and the other incorporated (e.g., application of bounding used a version of the HFE based on timing timing for HEPs to capture uncertainty in the associated with trip with nominal SG level.seuneofi-ndcdalrs)ortt Similar for zoneissues were identified in the cutsets 091-ETL. sequence deviations of in fire-induced the modelingfailures), or that are documented and justified.

As noted in the 2010 peer review in F&O 1- Review the application of the feed and bleed 4, there are also inconsistencies in modeling HFE to ensure the appropriate timing is used of the fire-induced small LOCA when during the fire quantification. Since the compared to the internal events small LOCA MEW pumps are assumed failed for all fire initiating event. PTN explains that this was MFW pum oste a te vale may be due to circular logic issues, and a sensitivity areas, the most appropriate value may be case shows this to be a non-significant the HEP based on timing assuming the trip issue. However, it is not clear that the occurs with low level in the SGs.

circular logic issue could not be resolved Review treatment of any additional HEPs and that all potential impacts of the modeling with event-specific timing assumptions to approach taken are understood, ensure that the appropriate values are used (This F&O originated from SR PRM-A3) in the fire quantification.

Revision 0 Page V-25

Florida Power & IIQht Attachment V - Fire PRA Qualitv Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 7-3 The current model uses the LERF model for PRM-B14 It cannot be determined from the existing This F&O has been resolved.

the PTN revision 9 model (PTN-BJFR documentation that an assessment was A review of the mapping of Level 1 010, Rev. 1) and maps appropriate performed to identify new accident sequences to the plant damage states in the equipment impacts into the system models progressions beyond the onset of core LERF model was reviewed. No new used to model LERF. No new accident damage that would be applicable to the Fire accident progressions that required progressions beyond the onset of core PRA that were not addressed for LERF modification of the LERF model were damage were identified for the fire PRA. estimation in the Internal Events PRA. identified.

However, there is no documentation that a Document an assessment to determine if specific review of the accident progressions there are potential fire-induced LERF leading to LERF was conducted to identify mechanisms not captured by the internal whether new considerations should be events accident progression models. If none addressed in the fire PRA. are identified, document the basis of that In addition, effects on PDS mapping due to conclusion.

fire-induced failures may not be appropriately captured. For example, RWST diversion of the RWST to the containment sump is modeled as a failure of HHSI which would normally go to a dry containment PDS. However, the actual PDS should be one for wet containment. While this is a late containment failure concern rather than a concern for LERF, there may be similar fire-induced failures that could affect the mapping of LERF accident progressions.

(This F&O originated from SR PRM-B14) 7-6 The new fire-specific safe shutdown actions HR-El Final post-fire safe shutdown actions have This F&O has not been resolved.

which are credited in the final Fire PRA will HR-E2 not been defined and appropriate The FPRA includes various actions that are be proposed to be added to the plant fire procedures revised to include the actions to being included as required plant changes in response procedures. These human actions HR-E3 be credited in the Fire PRA. the NFPA 805 LAR. The development and are included in the ALTEREDEVENTS table HR-E4 Complete the identification of new fire- implementation of related procedures has of the FRANC model using component basic HR-H2 specific safe shutdown actions which are not yet been initiated as it is part of the events as surrogate. HR-l1 credited in the final Fire PRA and evaluate overall integrated process associated with However, the safe shutdown actions HR-12 and document the HEPs consistent with transition to an NFPA 805 license basis.

modeled in the FPRA are not currently processes used for internal events HEPs.

consistent with those specified in the plant HR-13 Include consideration of fire effects on the fire response procedures, there is no HRA-A2 operator action, availability of cues, documented assessment of the cues HRA-A4 availability of time to complete the action, required to initiate the actions, no training feasibility of the credited actions given a fire, Revision 0 Page V-26

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update has been provided to operators on the new HRA-B2 and potential 92-18 impacts for both fire-specific actions, no operator reviews or HRA-63 screening values and detailed HEP talk-throughs of the credited actions has HRA-D2 development.

been documented, and the applicable Also, complete operator reviews and/or talk-performance shaping factors have not been HRA-E1 throughs when the procedure updates are considered, including time available for the PRM-B1 1 completed to ensure that the interpretation action. of the actions is consistent with the This F&O supersedes 2010 Peer Review operator's understanding and training.

F&Os 1-41, 2-6, 6-4 and 6-11. Finally, consider expanding the discussion (This F&O originated from SR HRA-A2) of sources of model uncertainty related to the HRA to include consideration of the accuracy and completeness issues noted in NUREG-6850, Volume 2, Appendix V.

7-8 Dependency between multiple altered HR-H3 The dependency associated with operator This F&O has been resolved.

events representing new HFEs in the same HR-12 actions applied using the altered events The use of altered events as a surrogate for cutset and between action represented by HRA-D2 method has not been addressed. a recovery action has been significantly the altered events and other HFEs in the Address dependency between multiple reduced as noted previously. Those same cutset has not been assessed based PRM-B1 1 altered events representing new HFEs and remaining instances are addressed by on the assumption that the dependency between the altered events and other HEPs modifications to the recovery rule file so that effects are bounded by the application of in the same cutset. If detailed dependency only a single instance of this use would exist conservative screening values. However, analysis is not performed, provide a in any cutset. This eliminates the potential there is no documented assessment to justification supporting the assumption that for multiple surrogate recovery events to support this assumption. There are cases the values chosen for the altered events appear together in the same cutset.

where complete dependency between bounds dependency effects.

events may be appropriate. For example, cutsets 40 - 45 in the provided Aggregate CDFaggregate.cut file contain altered events MAVC4200A_1.OOE-01 and MAVC4460 1.OOE-02 in each cutset. The product of these two events is therefore 1.OOE-03. However, since both events involve failure to isolate the letdown line, it could be assumed that there is complete dependence between the events since they would share a common cue.

This F&O supersedes 2010 Peer Review F&O 6-16.

(This F&O originated from SR HR-H3)

Revision 0 Page V-27

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 8-3 Attachment U - Internal Events PRA Quality PRM-B2 The potential effect of internal events F&O This F&O has been resolved.

(DRAFT), document applicability of Internal disposition on development of the FIRE PRA The internal events PRA model F&Os that Events F&Os to internal events PRA, but not was not addressed. have not been resolved/closed have been to Fire PRA. There was no evidence that Review internal events F&Os and provide reviewed and found to have no negative the review of F&O disposition status documentation as to how disposition of impact on Fire PRA results or this addressed the question of whether the those F&Os may impact development of the application.

disposition that was taken would adversely Fire PRA.

affect the development of the fire PRA.

This F&O is derived from 2010 Fire PRA peer review F&O 4-4.

(This F&O originated from SR PRM-B2) 8-5 The Fire PRA model changes were HRA-B1 The standard requires a review of the fire- This F&O has been resolved.

constructed so as to allow credit for the HRA-B3 induced initiating events accident current internal events PRA model structure sequences, and success criteria included in using existing accident sequence the internal events model, to identify new progression, success criteria and timing. PRM-B7 accident sequence progressions or success performed did identify a number of instances The internal events HRAs are modified with criteria due to unique aspects of fires. such as that specifically identified in the F&O. In all instances, it was determined that a screening modifier. Travel paths are This review will help assure that there are no the existing model structure was appropriate considered in the human failure evaluation revised actions where the screening and that opportunities for recovery actions report. multipliers are not appropriate, were limited either because of a lack of However, there is no indication that a review Conduct and document a review of the fire- appropriate cues or insufficient timing to gain was performed to identify accident induced initiating events accident any meaningful benefit via recovery. The sequences that may require modification sequences, and success criteria included in analysis documentation of the HFE based on unique aspects of the plant fire the internal events model, to identify new treatment was updated to address the response procedures. For example, RWST accident sequence progressions or success internal events PRA model human actions draindown may affect the evaluation of criteria due to unique aspects of fires, that are used in the FPRA. The timing for aligning sump recirculation, which documentation addresses the applicability, is not presently represented in the non- numerical adjustment, and availability of LOCA transient event tree accident necessary cues.

sequences used for the majority of the fire scenarios.

A review should be performed for possible changes to success criteria, particularly due to model changes from the MSO evaluation.

This F&O is derived from 2010 Fire peer review F&O 6.3.

(This F&O originated from SR PRM-B5)

Revision 0 Page V-28

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 8-8 Several portions of the analysis are not FSS-E3 Unit 3 results not fully documented, although This F&O has been resolved.

documented for Unit 3. Specific examples PRM-A3 they are available for inspection using the The U3 results have been added to the include: PRM-C 1 quantification software. analysis documentation.

Unit 4 significant contributors are identified Document Unit 3 results consistent with the in 0493060006.005, Rev. 2. Unit 3 Unit 4 results.

significant contributors are available, but not fully documented.

Unit 4 Fire Scenario information is presented in 0493060006.004, Rev. 2, but the equivalent Unit 3 information (Attachment D and E) is not provided.

(This F&O originated from SR PRM-A3) 8-10 2010 FPRA peer review F&O 1-44 finds HRA-C1 This appears to have a significant non- This F&O has been resolved.

issues with masking fire effects by setting PRM-B11 conservative impact to PRA results. Given The use of '0'; has been eliminated in the basic events to 0 in the Altered Events table. the actions in the altered events report are Altered Events table. Instead, events are This has partially been addressed by setting being added to the model as needed set to nominal. In the case of the application these events to 'nominal' in the Altered recoveries in order to ensure risk is low, and calculations for NFPA 805, the 'compliant' Events table and for reviewing cases where given the resulting recovery actions do not case is determined by using a '0' value the nominal value is on one side of an AND show up in the results in most cases, there which would under-estimate the compliance gate and the modified HEP value is on the appears to be a disconnect between the case risk and thereby provide a conservative other side. However, there are still cases addition of new actions to the procedures estimate of the risk increase for the where fire impacts are masked when the and the quantification of these actions in the application.

nominally adjusted event is on both sides of FPRA. It appears part of the disconnect is an AND gate or the HEP event is on one that the logic modeling, as modified by the side of an AND gate and nominally adjusted altered events table, results in the recovery events are on the other side. Scenarios 030 values being screened from the results.

PTB and 067E PTB are two examples. This Due to the complexity of this methodology, it old F&O is converted to a new F&O 8-10. seems a difficult task to review and address (This F&O originated from SR HRA-C1) for these masking issues. Perhaps a more systematic and comprehensive approach, with an independent review, could provide confidence that these nonconservatisms are addressed. Adding new HEP basic events, consistent with the approach used for internal events, would address this issue.

9-1 A general screening based on the ability to FSS-G2 No basis for the screening criteria is This F&O has been resolved.

form a damaging HGL in an exposing described. Since no MCA scenarios are Revision 0 Page V-29

Florida Power & Li-qht Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update compartment was developed. If no FSS-G3 developed, there is no way to determine if The existing HLG/MCA analysis includes a damaging HGL could form in an exposing FSS-G6 the exceeded (yet applied) screening criteria number of occurrences where the simplified compartment then there was no possible are significant. screening approach was found to generate associated MC scenario. Provide a basis for the 1E-07/yr screening over-conservative results. Incrementally When a damaging HGL could form, a criteria including additional information enhanced treatments were applied to second screening was performed whereby required when the criteria are included. The confirm that these locations had a very low the frequency of developing the HGL was basis for the screening criteria should likelihood of creating or causing formation of determined. Ifthe frequency was less than ensure the frequency is not too high, thereby HGL conditions and consequently a possible 1E-07/yr, then the scenario could be potentially masking significant MCA multi-compartment scenario eliminated. However, there was no basis scenarios.

provided for the 1E-07/yr criteria, nor was Evaluate impact of exceeded screening the criteria adhered to; in fact, the criteria criteria; for example, a qualitative analysis of was exceeded, yet still applied, in over 150 the expected CCDP based on known targets different scenarios. Some of the screened in the exposing and exposed compartments.

scenarios were slightly over the 1 E-07/yr threshold, while others ranged as high as nearly 6E-07/yr.

The impact of exceeding the criteria cannot be determined as no specific MCA scenarios were ever developed; therefore, it is not known if the scenarios would be significant.

For example, if it is assumed that 10 of the scenarios with a frequency of 5E-07 had CCDPs of 1.0, this would result in an increase in total CDF of 5E-06 which is about 10% of the total fire CDF.

(This F&O originated from SR FSS-G2) 9-4 The multi-compartment analysis assumes a FSS-G4 Systematic generic assessment of active fire This F&O has been resolved.

bounding value of 7.4E-3 for evaluation of FSS-G5 barrier elements may lead to non- The update of the analysis to incorporate a active fire barrier elements. Actual fire conservative results. barrier failure probability that integrates all barrier elements are not considered; instead If a screening value is desired, NUREG/CR- possible barrier elements was found to result the failure probability of a fire door is 6850 Section 11.5.4.4 suggests using a in a value of approximately double the assumed for active barrier element failure screening value of 0.1 for active fire barrier current value. However, since the entire because this failure probability represents elements. This value is much more likely to analysis approach involves a screening the highest single probability of a single encompass multiple fire barrier elements, strategy, additional analysis refinements are barrier failure. This method ignores the For scenarios that do not screen out, actual possible. An assessment of the use of a potential for multiple fire barrier elements. fire barrier elements identified during higher barrier failure probability to account Per NUREG/CR-4840 (source document for walkdowns (or document review) can be for failure of multiple barrier elements found Revision 0 Page V-30

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update NUREG/CR-6850 Table 11-3, "Barrier used to develop a more realistic barrier that the overall conclusion that MCA Types and Their Failure Probabilities") the failure probability, scenarios are not risk significant and need total barrier failure rate is a union of the not be explicitly included in the FPRA was probabilities of the individual failure rates. confirmed. However, the analysis Therefore, a value of 7.4E-03 may be documentation has not yet been updated ot conservative or non-conservative, reflect these results and insights.

This is based on 2010 FPRA peer review F&O 1-35.

(This F&O originated from SR FSS-G5) 9-5 The screening criteria are defined in the FSS-G2 Use of a standard fire scenario may be non- This F&O has been resolved.

Turkey Point Hot Gas Layer and Multi- conservative for some zones. The use of the 5 minute delay to Compartment Analysis, (Report Review zones to ensure that the standard combustible cable ignition is considered H0493060006.006) methodology, fire scenario is actually the most challenging realistic. Other conservatisms in the Compartments that don't screen are retained scenario inherent to the analyzed analysis ensure the overall conservatism of for further analysis. compartment. For zones where the the MCA/HGL evaluation.

A concern identified with the screening standard fire scenario is not the most criteria involves the use of a standard fire challenging, determine the most challenging scenario for each analysis rather than scenario and evaluate accordingly.

determining the most challenging fire scenario inherent to the analyzed compartment. This approach potentially masks the potential for forming an HGL in the exposing compartment.

For example, in zones 67 and 68 the standard fire scenario is non-conservative due to the potential for HEAF in 4kV switchgear. The damage time of 5 minutes is non-conservative for HEAF scenarios (should use 0 minutes).

This F&O supersedes 2010 FPRA peer review F&O 3-11.

(This F&O originated from SR FSS-G2) 9-6 The system unavailability records for the FSS-D7 This is a systematic issue. The intent for This F&O has been resolved. The fire plant have not been reviewed in crediting Capability Category II is to additionally protection system availability data for PTN fire detection and suppression systems. require a review of plant records to has been reviewed and no outlier behavior This F&O supersedes 2010 FPRA peer determine if the generic unavailability credit has been identified.

Revision 0 Page V-31

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update review F&O 2-26 is consistent with actual system (This F&O originated from SR FSS-D7) unavailability. Outlier experience would be any experience indicating that actual system is unavailable more frequently than would be indicated by the generic values.

Consider performing and documenting the review of plant records to determine if the generic unavailability credit is consistent with actual system unavailability. Outlier experience would be any experience indicating that actual system is unavailable more frequently than would be indicated by the generic values.

9-10 Section 3.1 of the FSS Report FSS-A6 Assumption made that no cabinet/panel fires This F&O has been resolved.

(0493060006.004, Rev. 2) states: "For the FSS-H7 in the MCR will ever spread to an adjacent Panels with communication between electrical panel fires, the scenarios are cabinet even if the cabinets are open to one adjacent panels are to be provided with developed similar to scenarios involving another. This incorporates an implied incipient detection to ensure early electrical panel fires outside the Control assumption that every MCR panel/cabinet identification of fire to preclude spread Room and are adequately described in fire will be extinguished prior to spread, between panels.

Attachment A. Fire spread to adjacent Identify adjacent MCR cabinets/panels panels was determined for these scenarios which could result in fire spread given failure based on a walkdown of the control room of suppression. Apply NUREG/CR-6850 during which panels with potential barriers Appendix L, S or other relevant document to for spread of fire were opened to confirm the address the potential for fire spread.

existence of such barriers. For MCB fires, Ensure documentation is consistent with the the method from NUREG/CR-6850 proc ually us in the Appendix L is applied. NUREG/CR-6850 process actually used in the analysis.

Appendix L defines a non-suppression probability applicable to the MCB. From Figure L-1 of NUREG/CR-6850, for non-qualified cables, and for a bounding distance of 0 meters (assuming that the cables terminating at the individual MCB are in very close proximity), a non-suppression frequency of 8.30E-3 is used for the MCB."

However, based on discussion with FPL/ERIN staff, this was not done.

Essentially, no fire spread for any cabinet in the MCR was assumed. For panels with Revision 0 Page V-32

Florida Power & Light Affachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update incipient detection, success of the detection results in no damage as it is assumed operators isolate the circuit prior to additional damage in the cabinet. If insipient detection fails, the MCB panel fails completely, but never spreads another cabinet.

For all other cabinets/MCB panels without incipient detection, full burnout of the cabinet/panel is assumed, but again, no spread to adjacent cabinets is assumed even if the cabinets are open to one another (e.g., walkthrough MCB).

(This F&O originated from SR FSS-A6) 9-11 Several entries in Table 3-1 of Report FSS-G2 It was not confirmed during the peer review This F&O has been resolved.

H0493060006.006 says "Walkdown required that this was just a documentation issue. It was confirmed that the lower damage to confirm no combustibles within the 383 Therefore, this is classified as a finding threshold associated with thermoplastic Zor' when the cables are not IEEE-383 because it could affect the analysis results, materials was used for the analysis. The qualified. Based on discussions with FPL Verify that the damage criteria used is typographical error has been corrected.

and contractors, it is believed that this is a consistent with non-383 cable damage and typo, and the correct damage criteria were revise the documentation as required. If it is actually applied. discovered that the incorrect damage criteria (This F&O originated from SR FSS-G2) were applied, update the analysis with the correct criteria.

10-1 The 2010 peer review identified that "Fire FSS-C1 The present analysis provides a bounding This F&O has been resolved.

modeling was conducted via generic fire FSS-G1 approach for fire severity in most cases, The recommended resolution action in the modeling from which Zones-Of-Influence since the 98th percentile fire heat release F&O was assessed in the context of the (ZOI) for specific initiator types was rate is used. dominant fire risk contributors. This generated. The ZOIs were used to define However, use of the split fraction method is assessment concluded that further bounding fire characteristics for each fire based on industry events rather than site refinements such as that described in the scenario. Characteristics that are used to specific fire ignition sources and target F&O would not substantively change the bound potentially risk contributing fire events configurations. Therefore, this could result results of the analysis. The existing are identified in Attachment B of the Fire in non-conservative frequency estimates of treatment retains some conservatism which Scenario Report, (Report 0493060006.004). target damage. results in this SR meeting CC I. This is Based on the use of a bounding approach Perform 2-point fire modeling, when adequate for the NFPA 805 application, as Significant fire scenarios should be applicable, for risk significant fire scenarios, this conservative bias would tend to over-estimate the risk metric that is used to judge Revision 0 Page V-33

Florida Power & Light Affachment V -Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update developed with 2-point fire modeling." the acceptability of this application.

Since this review, FP&L has stated that "The The issue regarding the ERIN panel split use of a panel split fraction to differentiate fraction is addressed in the disposition for between fires impacting the panel and F&O 10-3.

components with cables terminating at the panel versus panel fires impacting cables outside of the panel provides an equivalent and more useful two point fire model."

The Panel Split fraction is developed from a supplemental report (ERIN report, Supplemental Fire PRA Methods, dated February 2010). This document was submitted to the EPRI Fire PRA Methods Review Panel. This review is not complete as of the date of this peer review.

Use of the split fraction method is based on industry events rather than site specific fire ignition sources and target configurations and therefore, could result in non-conservative frequency estimates of target damage.

(This F&O originated from SR FSS-C1) 10-2 The 2010 review of PTN Tasks 8 and 11 FSS-A1 Including the fire frequency and associated This F&O has been resolved.

Report 0493060006.004, identified that 'no fire scenarios from hydrogen fires will have Miscellaneous hydrogen fires have been hydrogen fires other than turbine/generator impact to the CDF and LERF results, incorporated in the Fire PRA in the charging have been postulated.' (Previously F&O 5- Incorporate the hydrogen fire scenarios pump room fire areas where the hydrogen

16) being developed into the model, and update lines associated with VCT cover gas are Since this Finding was identified, FP&L has documentation as necessary. routed.

determined that 'Miscellaneous Hydrogen piping at PTN is limited to hydrogen supply to the VCT tanks. The associated piping is located in the charging pump rooms (Fire Zones 45 and 55). Fires in these fire zones are assumed to impact all components in the fire zone. The associated risk is low given the availability of thermal barrier cooling for RCP seals and HHSI pumps. Allocation of the IGF associated with miscellaneous Revision 0 Page V-34

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update hydrogen fires to these fire zones would result in an increase in the ignition frequency for these zones by less than a factor of 3.

Given the low risk significance of these zones this will have a negligible impact on overall plant risk and the charging pump rooms will remain low risk contribution fire zones.

Incorporation of this ignition frequency into the associated documentation will be incorporated in a future revision to the documentation.'

Hydrogen fires are also being developed for H2 piping and valves in Compartments 82 and 87 (scenarios 82-P and 87-P).

However, since these do not appear yet in the Fire Scenario Report, action is required.

This finding is currently being addressed and appears to be resolved once the new H2 fires are included in the model and documentation is updated.

(This F&O originated from SR FSS-A1) 10-3 FSS-C4 requires severity factors to be FSS-C4 Severity factor (panel split fraction) is used The FPRA quantification uses the panel independent of other factors. Fire severity FSS-D5 extensively in the Fire PRA. factors consistent with the latest guidance factor as discussed in Section 7.1.2 for Use the severity factor method described in from the EPRI Methods Review panel. A electrical cabinets is not developed or FSS-G1 sensitivity study has been performed to NUREG/CR-6850, or develop an accepted applied consistently with the NUREG/CR- industry approach (presently being address the impact of elimination of the 6850 methods. This is developed from a discussed by EPRI). Develop fire severity credit for the panel factors. The results of supplemental report (ERIN report, factors based on the likely HRR and location this evaluation indicate that the delta Supplemental Fire PRA Methods, dated of overhead cables or location of equipment. CDF/LERF would exceed the Reg Guide February 2010). This document was For example, if cable is 7 feet overhead, the 1.174 guidelines should these factors be submitted to the EPRI Fire PRA Methods severity factor would be based on the completely eliminated (the 1E-5/1 E-6 delta Review Panel. This review is not complete minimum HRR that would damage the cable CDF/delta LERF limits would be exceeded as of the date of this peer review. at that distance. Additionally, the growth but the conservatively calculated delta risk Using this method, fire propagation outside time can be used in determining non- would be less than 2E-5/2E-6). Further of the electrical cabinets is dependent on the suppression time. refinements of this sensitivity evaluation are nonsuppression probability. Therefore, possible to reduce the calculated delta risk.

some dependency exists in this data if used Credit for additional Defense In Depth Page V-35 Revision Revision 0 Page V-35

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update in conjunction with a non-suppression factor. measures may be taken in areas of concern Due to this derivation of the conditional as necessary to compensate for the probabilities for fire propagation outside of increased delta risk.

the cabinets, the conditional probabilities thus developed (and applied in the FRANC model) could potentially be non-conservative.

The severity factors are developed using generic fire events data from the EPRI fire events database. Given the fire data duration and damage is a result of multiple factors (growth, suppression, severity, location, etc), and given the fire data often does not have sufficient information to make a reasonable determination of either the fire size or whether a fire propagated outside the cabinet, the severity factor used (panel split fraction) may not necessarily bound the conditions of the specific fire scenarios under analysis.

(This F&O originated from SR FSS-C4) 10-4 One situation was identified for which credit FSS-C8 This finding is based on identification of This F&O has been resolved.

of fire wrap is taken in Compartment 96 for credit for a wrap in Attachment A of the Fire A qualitative assessment has been ignition source 3B04, which is a 480V load Scenario Report, (Report 0493060006.004). performed to assess the potential impact of center. This fire wrap protects PB3319, Any credited fire wrap should be addressed this F&O.

PB3813, P87022, and PB7521. The wrap and the wrap integrity should be established The hose stream test imposed on the fire appears as being credited in a HEAF with respect to fire resistance, mechanical barrier qualification subsequent to fire scenario. No justification for crediting this protection, and potential fire related exposure is considered to provide a wrap assuming mechanical damage and exposure to which the wrap may be exposed comparable level challenge to the thermolag direct flame impingement from the HEAF is (direct flame impingement, HEAF, etc.). barrier as would the HEAF force applied at provided. Similar issue for 3603 also in Compartment 96. the onset of fire exposure.

Thermo-lag is also seen as credited in some scenarios, which would require justification due to issues with this particular type of cable barrier.

(This F&O originated from SR FSS-C8)

Revision 0 Page V-36

Florida Power & Liaht Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 10-6 Treatment for transient fire damage to FSS-D6 Many transient fire scenarios have been This F&O has been resolved.

targets is measured from the compartment screened during detailed scenario analysis. Supplemental walkdowns were performed to floor rather than the height of the transient The results of the FPRA are therefore re-assess the treatment of transient fires.

fuel package that is typically considered. potentially non-conservative for the analyzed These walkdowns focused on two key Discussion with FP&L during the review detailed scenarios. attributes - the appropriateness of the provided some basis for the damage height The transient fires should be considered to selected HRR characterization and the (indicating that transient fires above the floor be above the floor level in the analysis. location of the postulated fire scenarios.

will have an overall lower average surface With respect to this specific F&O, the HRR). However, the supplemental placement (elevation) of the assumed fire discussion was still considered inconsistent was based on the physical features of the with past events and existing guidance on location. The fires were not artificially analysis of transient fires, and could lead to elevated in the absence of a physical non-conservative estimates of transient fire feature.

damage to targets.

Transient fire evaluations conducted as described in the Fire Scenario Report result in screening fire damage to targets that are located > 7.3' above the floor which is believed to be non-conservative for developed fires involving ordinary combustible fuel packages such as a trash can or trash bag. In response to this concern it was pointed out that the thermal plume component relies on empirical relationships between the source strength and the distance between the virtual origin of the fire and the target. The fire plume begins to entrain air at the lowest point of burning, which defines the base of the fire; normally at the floor. However this argument ignores the potential that a fire could begin burning at the top of a fuel package thus elevating its base. At a minimum, during the initial period of burning, damage temperatures generated by the fire would likewise be elevated. Over time the base of the fire may change due to collapse of the fuel package or burning away of the fuel, however the empirical model Page V-37 Revision 0 Page V-37

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update presented did not present sufficient basis for assuming that the base of the fire is at the floor for its entire duration.

(This F&O originated from SR FSS-D6) 10-8 Ambient conditions are assumed in the FSS-D4 Underestimating the ambient conditions This F&O has been resolved.

Generic Fire Modeling Treatment Report FSS-H4 could result in non-conservative estimations A qualitative assessment has been (prepared by Hughes). Ambient of zones of influence and targets considered performed to assess the potential impact of temperature is assumed to be 68°F for all to be fire damaged. this F&O.

calculations. No technical discussion or Assess areas where elevated ambient The sensitivity of the ZOI dimensions to the justification is provided in the Fire Scenario temperatures could be experienced and ambient temperature is relatively low as Report to substantiate that this is a justify the acceptability of the models used. described in the original Hughes Generic reasonable value for the compartments Otherwise, incorporate elevated ambient Fire Modeling treatments report, in particular where this was applied. temperatures into the zone of influence for IEEE-383 qualified/Thermoset cables. In (This F&O originated from SR FSS-D4) calculations, the case of an initial ambient temperature of 35°C, the expected affect on the ZOI dimensions is within the measurement uncertainty in the field.

10-9 The 2010 peer review identified that 'Except FSS-D8 The method currently employed could result A qualitative assessment has been for the MCR fire scenarios, no other fire in optimistic times for suppression activation performed to assess the potential impact of scenario has used the Non-Suppression or fire brigade response. this F&O.

Probability (NSP) in PTN fire model at this Assess and document the effectiveness of The HGL and MCA analyses credit both time.' Since this review, FP&L has taken suppression with respect to: automatic suppression system and fire cred)it forthuppession (bothtomaetc Gand System design complies with applicable brigade actions. In the context of the HGL manual) in the Multi-Compartment/Hot Gas codes and standards, and current fire and MCA, the fire brigade action of interest Layer evaluation. However, this evaluation protection engineering practice, is fire control as that would terminate the does not include an assessment of the fire possibility for HGL formation. However, the protection system effectiveness. The time available to suppress the fire prior only readily available numeric credit is fire Of particular concern is that fire detection to target damage, suppression credit. To reduce the and/or suppression timing (i.e., thermal Specific features of physical analysis unit conservatism introduced into the analysis, response of the detector and/or sprinkler) and fire scenario under analysis (e.g., fire detection time is ignored for the HGL was not calculated and subtracted from the pocketing effects, blockages that might and MCA. The timeframe associated with time considered for manual suppression impact plume behaviors or the "visibility" of detection and suppression is significantly when using the FAQ-0050 process. In the fire to detection and suppression less than the timeframe required to reach a addition, fire detection reliabilities are not systems, and suppression system hot gas layer temperature which would included in the assessment. Ifthe detection coverage), and impact the HGL analysis.

system does not function as intended, the Suitability of the installed system given the time to detection to initiate fire brigade Revision 0 Page V-38

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update response would be substantially longer, nature of the fire source being analyzed.

(This F&O originated from SR FSS-D8) 10-11 The 2010 peer review identified that "fire FSS-C2 The present analysis provides a bounding This F&O has been resolved.

scenario evaluation tools were developed FSS-C3 approach in most cases, since the 98th The recommended resolution involves the based on the Generic Fire Modeling percentile fire heat release rate is used from crediting of growth and decay in the Treatments. These walkdown/evaluation FSS-G1 fire initiation without growth and burnout. modeling of the postulated fire. The existing tools are based on bounding fires that are However, use of the split fraction method is analysis does not take credit for these assumed to cause target damage at a height based on industry events rather than site variables. A review of the dominant fire above the base fire with the fire burning at specific fire ignition sources and target scenarios found that the risk benefit that peak intensity and without burnout times. configurations. Therefore, this could result might be gained is minimal. Therefore, this Because these tools assume a fire burning in non-conservative frequency estimates of refinement was not performed. The at peak intensity and without burnout, this target damage. resulting categorization of the related SR is SR is considered met at CC "ig CC 1. Since the approach results in some Include fire growth and decay for risk Since the review, FP&L has stated that The significant fire scenarios. conservatism being retained in the results, use of a panel split fraction to differentiate this CC is judged to be adequate for the between fires impacting the panel and NFPA 805 applications as the conservative components with cables terminating at the bias would tend to result in the over-panel versus panel fires impacting cables estimation of the risk metrics used for this outside of the panel provides an equivalent application.

and more useful two point fire model... The application of the two point treatment to individual fire scenarios is carried through to the MCA/HGL evaluation which addresses the impact of each scenario on MCA."

The Panel Split fraction is developed from a supplemental report (ERIN report, Supplemental Fire PRA Methods, dated February 2010). This document was submitted to the EPRI Fire PRA Methods Review Panel. This review is not complete as of the date of this peer review.

Use of the split fraction method is based on industry events rather than site specific fire ignition sources and target configurations and therefore, could result in non-conservative frequency estimates of target damage.

Revision 0 Page V-39

Florida Power & Light Aftachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update (This F&O originated from SR FSS-C2) 10-12 The 2010 peer review identified that "The FSS-D9 This appears to be a documentation issue, This F&O has been resolved.

PTN FPRA methodology generally does not but FP&L should confirm that smoke An analysis of the impact of smoke damage include postulation or evaluation of smoke damage has been considered and document has been completed and documented in the damage. Additional review shows that the accordingly. PTN FPRA Scenario Report.

smoke issues do not affect the FPRA results Confirm that smoke damage has been significantly. However, the FPRA does not considered and document accordingly.

include a qualitative evaluation of smoke damage to FPRA equipment."

Since the 2010 review, FP&L stated, "Section 6.2 of the Scenario Report was added to address this concern." However, section 6.2 provides a high level discussion and methodology including the statement that "Exposure time plays a key role in the likelihood of failures from smoke. As a result, damage from short term smoke exposure will only result from severe conditions.... Instruments, control components and all high voltage powered components are exceptionally vulnerable to circuit bridging as a result of airborne smoke and deposited particulates."

However, there is no documented discussion of the smoke damage assessment results, and none of the targets in the scenarios indicated smoke damage as the failure mode. NUREG/CR-6850 recommends considering smoke damage to banks of interconnected panels, and this should be considered.

(This F&O originated from SR FSS-D9) 10-13 A credit for incipient detection is taken for FSS-A6 This appears to be a documentation issue, This F&O has been resolved.

MCB fires (non-suppression probability of FSS-D7 but since the system is not yet installed, The credit taken for incipient detection is 0.02). There is no documentation to justify FSS-H7 there could be an impact to the assumptions consistent with that specified in FAQ this value. Per discussion with FP&L the made within the Fire PRA. 0046.

approach appears to be in agreement with Document the basis for probability of non-Revision 0 Page V-40

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update FAQ-08-0046. The approach also does not suppression value assumed in analysis.

use the NUREG/CR-6850 Appendix L factor When the incipient system is installed, the for panels that credit incipient detection. FPRA should be reviewed and updated Secondly, the incipient detection system is accordingly.

not yet installed, and therefore, the Fire PRA should be reviewed and updated as needed to reflect any differences between the assumed and as-built conditions of the system.

(This F&O originated from SR FSS-A6) 10-14 Beyond the Generic Fire Modeling FSS-A5 For risk significant fire scenarios, detailed This F&O has been resolved.

Treatments, the Fire PRA did not include fire modeling should be performed to ensure The current analysis is consistent with a additional detailed fire modeling for most fire you are not masking the "true risk significant Capability Category I analysis. This compartments. fire areas". Without detailed fire modeling provides a degree of conservatism in the Note 4 (under FSS-A5 of the ASME for significant fire scenarios, the results are analysis which would also tend to over-Standard) states that "once a fire scenario conservative, estimate the change in risk which is reported has been 'selected,' this implies that the Consider performing additional detailed fire for the NFPA 805 application. A review of scenario will eventually be evaluated and/or modeling to provide "reasonable assurance the results of the application analyses quantified at a level of detail commensurate that the fire risk contribution of each indicates more rigorous analyses consistent with the risk significance of the scenario." unscreened physical analysis unit can be with CC II or CC III would not alter the (This F&O originated from SR FSS-A5) characterized." conclusions of the analyses.

10-15 PTN credits multiple suppression paths for FSS-C7 Lack of dependency analysis could lead to This F&O is resolved.

MCA/HGL evaluation. However, the FSS-G1 an optimistic estimate of suppression A review of the credited suppression dependencies have not been evaluated and FSSH7 probability. systems in the Multi-Compartment/Hot Gas modeled. For example, fixed suppression When multiple suppression paths are Layer analysis has confirmed that no and fire brigade response may both rely on credited, perform a review and address any dependency exists between the suppression a single detection system. dependencies between suppression and systems and detection systems. Detection (This F&O originated from SR FSS-C7) detection systems credited in the MCA/HGL in the zones with suppression systems is calculation, associated with an independent detection system.

10-16 Review of fire modeling in single FSS-C1 Discounting of secondary combustibles This F&O has been resolved.

compartments does not consider the FSS-D3 when considering localized fire damage Supplemental walkdowns have been addition of HRR from secondary FSS-G1 could lead to non-conservative results. performed to identify and address the combustibles. It is acknowledged that Include secondary combustibles in the heat potential for fire spread for scenarios where secondary combustibles were considered for release rates used for zone of influence the non-383 cables are not protected by Revision 0 Page V-41

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update the MCA/HGL evaluation, estimates. Flammastic material. The analysis has been Fire spread and additional HRR due to the updated to include these scenarios as resulting cable tray fire and adjacent appropriate.

cabinets would increase the total fire size and the subsequent zone of influence.

Compared to the NUREG/CR-6850 guidance for flame spread along PVC cable (flame spread = 0.9 mm/sec) the estimation of HRR for the applied scenarios is non-conservative. Realistic estimation of the scenario HRR is necessary to ensure the full impact of the fire on exposed targets is presented and that the effects of a damaging HGL may also be estimated.

FP&L has stated that walkdowns are in progress to include fire spread to cable trays and incorporate this into the fire scenarios.

(This F&O originated from SR FSS-Cl) 10-17 The 2010 peer review identified that FSS-D1 Modifying the zone of influence to account This F&O has been resolved.

Attachment B of the Fire Scenario Report for HGL effects could impact the defined The MCA/HGL evaluation has been modified (Report 0493060006.004) generic fire target damage set. to address the concern noted in the F&O.

modeling treatments do not account for the The generic treatments used in relatively The potential for a larger zone of influence is effects of hot gas layer (HGL) on the zones small rooms should be scrutinized to ensure addressed via new fire scenarios added to of influence. The limitation indicates that that any HGL interaction is considered and the fire PRA.

because HGL is not considered that these accounted for if found to be significant.

correlations should not be used in enclosed The selection of which generic fire modeling areas with small volumes where a significant treatment is used to define target damage HGL thickness may form. Because this for HGL effects on a scenario basis should relationship is not considered plume be documented in a clear manner to temperatures may be underestimated facilitate updates and peer reviews.

because it is assumed that ambient temperature air is being entrained into the plume, resulting in cooler plume temperatures, rather than heated air from the hot gas layer. Entrainment of heated air into the fire plume results in higher damage heights because the plume remains hotter at higher elevations.

Revision 0 Page V-42

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update Since this review, FP&L states that "The impact of a hot gas layer on the zone of influence is evaluated for all fire zones/scenarios in the MCA/HGL evaluation." A review of this evaluation confirms that HGL effects on ZOI were in fact considered for the generic treatments; however, there is not sufficient documentation in Attachment A to the Fire Scenario Report to determine which zone of influence was applied to which scenario, and whether it was applied correctly to consider the effects of HGL. The Generic treatments include several iterations and combinations of variables, including opening percentage of the compartment. The fire scenario documentation at the time of this review did not provide sufficient information on opening percentage to confirm that the ZOI was applicable to the compartment.

(This F&O originated from SR FSS-D1) 10-18 In at least two cases, transient fire scenarios FSS-Al The exclusion of transients in some This F&O has been resolved.

have not been included in the fire modeling compartments may lead to a non- Supplemental walkdowns were performed to for some compartments (e.g., fire conservative estimate of CDF and LERF. re-assess the treatment of transient fires.

compartments 67 and 68). Per discussion These walkdowns focused on two key with FP&L the transients may have been attributes - the appropriateness of the Include transient scenarios in all excluded based on the dominance of the selected HRR characterization and the compartments where fire modeling has been frequency of fixed scenarios. However, location of the postulated fire scenarios.

employed.

transients should only be excluded when The postulated location for the treatment of precluded by design. Based on the size of transient fires was based on where a these rooms, and the presence of secondary transient ignition source might reasonably combustibles, transient fires could lead to occur. The results of these walkdowns were fire growth and eventually HGL, and incorporated into the FPRA analysis.

therefore should be analyzed.

(This F&O originated from SR FSS-A1) 10-19 For fire modeling analysis of transient fires, FSS-H1 Lack of documentation on transient fire This F&O has been resolved.

FP&L implements a floor area weighting locations and boundaries will present a The specific instance noted in the F&O was Revision 0 Page V-43

Florida Power & I-I-qht Attachment V - Fire PRA Quali Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update factor. However, the documentation does challenge for updates and peer reviews, corrected. In addition, supplemental not include a graphical representation of the Update documentation to include a graphical walkdowns were performed to re-assess the assumed transient locations and representation of transient fire locations and overall treatment of transient fires. These boundaries. It is therefore not possible to boundaries. walkdowns focused on two key attributes -

review (or update) transient fires, the appropriateness of the selected HRR Also during review of transient weighting characterization and the location of the factors it appears to have been double postulated fire scenarios. However, the counted in some compartments (e.g., documentation that was generated did not compartment 63). Based on discussion with specifically produce graphical FP&L this was due to an error in the Excel representations. Instead, the information based spreadsheet tool for transient was incrementally enhanced to provide a frequency quantification. This appears to be spatial reference to a location with in the an isolated case and will be corrected, space. The need for special depiction of (This F&O originated from SR FSS-H1) transient fire scenario locations will be addressed in conjunction with the development of procedures for post transition configuration control.

10-20 The fire modeling analysis of the Turbine FSS-Al Lack of consideration of the catastrophic T/G This F&O has been resolved.

Generator (T/G) fires is performed in fire may lead to a non-conservative estimate The analysis documentation has been accordance with Appendix 0 to NUREG/CR- of CDF and LERF. updated to address catastrophic T/G fires 6850. However, there is no discussion Perform a review of the catastrophic T/G fire that may lead to building collapse or other regarding the lack of analysis of the in accordance with Appendix 0 to significant widespread damage. The results catastrophic T/G fire event, which should NUREG/CR-6850, or document the of this update did not identify any new risk consider blade ejection, oil line rupture, and justification for excluding this event at PTN. significant contributors or insights.

hydrogen explosion. Per discussion with FP&L, the catastrophic fire was discounted since the T/G is located outdoors. While this may not result in hot gas layer formation and structural collapse, a review of the guidance is warranted, and inclusion of this event frequency should as a minimum map to the loss of the T/G and if suppression fails, all equipment within the T/G structure.

(This F&O originated from SR FSS-Al)

Page V-44 Revision 00 Page V-44

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 10-21 The supplemental generic Fire Model FSS-C3 The current approach results in many This F&O has been resolved.

Treatments: Transient Ignition Source FSS-G1 transient fire scenarios being screened Supplemental walkdowns were performed to Strength includes an assumption for FSS-H2 during detailed scenario analysis. The re-assess the treatment of transient fires.

transient burnout of 12 minutes. This results of the FPRA are therefore potentially These walkdowns did not identify any burnout time is based on an assumed fire non-conservative for the analyzed detailed instances where an altering of the transient loading and the 317kW heat release rate, scenarios, fire duration had any material impact on the and appears to be optimistic given the Provide additional justification for the applied HGL and MCA. The documentation has uncertainty in transient fire loading. The transient fire analysis as a screening also been updated to address the criteria burnout is then used to develop a zone of approach. Consider increasing the burnout used for selecting the characteristic transient influence for thermoplastic targets, based on time and using the NUREG/CR-6850 fire HRR. The approach is consistent with the thermal response tables in Appendix H recommended damage threshold to 205°C the recently issued guidance from the to NUREG/CR-6850 for thermoplastic cable to bound uncertainties in fuel loading for EPRI/NRC review panel. The results of at 260°C. Since this resultant vertical zone transient fires, these walkdowns were incorporated into the of influence is used to screen transient FPRA analysis.

scenarios from impacting secondary targets The twelve minute fire corresponds to the higher than 7.3 feet from the floor, additional 317 kW fuel package only and represents -

justification is needed to demonstrate that a 35 lb of Class A material. Additional 12 minute fire, and subsequent use of 2600C discussion is provided in Rev. 0 of damage screeningthreshold purposes.is appropriate for Sussion i 3 of thes Hughes Generic Supplement Genev. Fire Fire Asoreeninoe iupss.

thModeling treatments that examines the fire Also noted is that Attachment B to the Fire durations and test durations of all Scenario Report zone of influence does not NUREG/CR 6850 tests. It is shown that the reflect the same values recommended by method used to determine a 12 minute fire the Generic Fire Model Treatment. As an predicts or overestimates the fire duration in example, the differentiation between all cases and is therefore a sound approach.

transient Severe and Non-Severe categories is not based on a 317kW fire. This appears to be a documentation issue only.

(This F&O originated from SR FSS-C3)

Page V-45 Revision 00 Page V-45

Florida Power & Light Attachment V - Fire PRA Quality Table V-3 DISPOSITION OF 2010 PTN FIRE PRA PEER REVIEW 'FINDING' F&Os Finding F&O Discussion Fire PRA SR Basis and Recommendation Disposition in Fire PRA Update 10-22 Per NUREG/CR-6850, appendix H, FSS-C6 For smaller volume rooms, estimates of This F&O has been resolved.

temperature sensitive equipment should be FSS-G1 equipment damage may be non- The consideration of sensitive electronics considered to fail at 65°C. Supplemental conservative, was addressed in a qualitative fashion in the Generic Fire Model Treatments: Hot Gas Apply the appropriate hot gas layer and Scenario Report.

Layer Tables includes new zone of influence zone of influence for temperature sensitive and hot gas layer treatments for temperature equipment where applicable.

sensitive equipment. However, per discussion with FP&L these have not been implemented in the fire scenarios.

(This F&O originated from SR FSS-C6) 10-23 The PTN FSS report 0493060006.004, Rev FSS-C5 Equipment damaged by suppression This F&O has been resolved.

2, section 6 discusses the damage criteria activities may impact estimates of CDF and The specific issue raised in the F&O is for thermal, smoke, and sensitive LERF for some scenarios, beyond the scope of the associated SR. In equipment. However, suppression effects Perform an assessment of electrical addition, no known consensus method do not appear to have been considered for equipment that may be vulnerable to water exists for treatment. A qualitative the potential to damage equipment. intrusion from suppression activities (or assessment, based on other guidance for (This F&O originated from SR FSS-C5) thermal shock from gaseous systems), and evaluation of potential impact of suppression include any additional failed equipment, not effects was performed which indicated that already considered damaged by fire, in no specific change in the analysis is needed.

scenarios as appropriate.

Page V-46 Revision 00 Page V-46