ML13085A246

DC Cook Table V-1 & V-2 Record of Review March 22 2013

ML13085A246
Person / Time
Site:	Cook
Issue date:	04/22/2013
From:	NRC/NRR/DRA/APLA
To:	NRC/NRR/DRA/APLA
Snodderly M R
References
Download: ML13085A246 (50)

Text

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 1

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings PP-B2-1 Attachment 2 of the Plant Boundary Definition and Partitioning Report indicates credit given for an Appendix R partitioning feature between AA2/AA16. The boundary consists of steel plates coated with 2-inch Pyrocrete and a technical evaluation was performed to evaluate deviations to the 3-hour rating of the fireproofing material. Since fire coatings are typically not rated for direct flame impingement, Pyrocrete could not be installed on some portions, and small through- barrier openings exist, it is not clear from the documentation that this boundary meets this requirement. Basis for Significance: Documentation enhancement Documentation has been enhanced to demonstrate acceptability of fire coatings credited for fire barriers. For this particular example, Engineering Equivalency Evaluation 11.39, which gives reasonable assurance that the boundary is acceptable, has been referenced in the CNP Plant Boundary Definition and Partitioning Report.

(From peer review: Met at CC-II/III Suggestion)

RAI not needed.

Staff sees that the acceptability of the use of Pyrocrete Fireproofing on steel walls, ceiling, and structural steel components of the Unit 1 and 2 Auxiliary Feedwater Pump Enclosures is addressed in Engineering Equivalency Evaluation 11.39. Small portions of the structural steel that are not fully protected and minor gaps that exist in the 3-hr rated roll-up door and blow-out panel will not lead to spread of fire or impair the ability of the plant to reach a safe and stable condition.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PP-B2-2 In page 25 of Plant Boundary Definition and Partitioning Report Rev.1, (R1900-004-001), boundary identification AA2/AA2C has eight undampered ventilation penetrations between fire zones 110 and 114, & between 111 and 115. But this is not supported by a technical evaluation. Subsequent to the identific ation of this issue, the plant personnel generated a technical evaluation to address this issue. Therefore, this F&O now is a documentation issue. The technical evaluation needs to be referred to in the plant partitioning report. Basis for Significance: Documentation issue only, does not impact results. The CNP Plant Boundary Definition and Partitioning Report has been updated to reference Engineering Equivalency Evaluation 11.56. (From peer review: Met at CC-II/III Suggestion)

RAI not needed.

Staff that the Plant Boundary Definition and Partitioning Report (R1900-0041-001) was updated to reference Engineering Equivalency Evaluation 11.56 (see page 35 of Attachment 2). Engineering Equivalency Evaluation 11.56 documents the acceptability of eight has eight undampered ventilation penetrations between fire zones 110 and Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 2

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings 114, & between 111 and 115.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PP-B3-2 (From V-2 Table) (There is no F&O associated with this SR. However, Table B-6 of the fire PRA WOG Peer Review SR Assessment Table states reads that: "the plant portioning analysis at CNP does not credit spatial separation as a portioning element in defining physical analysis units in Pant Boundary Definition and Partitioning Report Rev. 1, (R1900-004-001)." Updates made to the Fire PRA since the peer review incorporated spatial separation (specifically for the YD compartment) and justification for such. This SR is now considered to meet Cat II/III. (From peer review: Met at CC-I Not an F&O)

RAI not needed.

Staff found that the Pant Boundary Definition and Partitioning Report Rev. 1, (R1900-004-001) did credit spatial separation in the case of buildings excluded and Yard partitioning (see page 6 and 7). However, Section 4.3 still contains the statement that "plant partitioning at CNP does not credit spatial separation".

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PP-B5-1 Capability Category I of PP-B5 precludes crediting active fire barrier elements outside of those included in the regulatory fire protection program.

The CNP PRA meets this r equirement. Category II/III of PP-B5 requires explicit bases and criteria when crediting active fire barrier elements. No explicit basis or criterion has been developed for normally open fire doors with fusible links, beyond meeting the specific design codes required by the fire protection program. Basis for Significance: Specific recommendations to A review of credited active fire barrier elements (fire dampers, held-open fire doors, water curtains) was expanded beyond the requirements of specific design codes, by identifying hazards at CNP that affect these elements, such as high energy arcing faults (HEAFs), as well as hydrogen and transformer explosions. These hazards were evaluated to determine thei impact on active fire barrier elements in order to demonstrate compliance with Category II/III criteria. The methodology and results of the review are documented in (From peer review: Met at CC-I Suggestion)

RAI submitted as RAI 34(i)

During the audit it was explained that further assessment of active fire barriers was expanded to include high hazard events (i.e., HEAFs, transformer explosions, and hydrogen explosions) and is documented in the Plant Boundary Definition and Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 3

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings meet CC-II/III the CNP Plant Boundary Definition and Partitioning Report.

Partitioning Report (R1900-0041-001), we ask an RAI to describe how the multi-compartment analysis was revised as a result. In the response to RAI 34(i) the licensee cited the provided a description of assessment of active fire barriers was expanded to include consideration of high hazard events (i.e., HEAFs, transformer explosions, and hydrogen explosions) and is documented in the Plant Boundary Definition and Partitioning Report (R1900-0041-001). Staff sees that Section 4.5 (on pages 9 and 10) discusses crediting active fire barriers. All active fire barriers (i.e., fire dampers, held-open fire doors, water curtains) are identified in Attachment 2 of the Plant Boundary Definition and Partitioning Report .

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PP-B7-1 Walkdown notes collected for confirming conditions and characteristics of credited partitioning elements are not referenced in the CNP Plant Boundary Definition and Partitioning Report. Basis for Significance: Documentation enhancement A reference to the confirmatory walkdown notes has been added to the CNP Plant Boundary Definition and Partitioning Report. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that a description of confirmatory walkdowns was added as Section 4.7 of the Plant Boundary Definition and Partitioning Report (R1900-0041-001). The walkdowns were performed in accordance with EPM procedure, EPM-DP-FP-002 (Performance of Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 4

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings Field Walkdowns) and are documented under EPM File Access Number 200100120-012. Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PP-C2-1 Exclusion of buildings within the licensee-controlled area from the global analysis boundary is documented and appears to be reasonable; however, more detailed documentation is recommended. Basis for Significance: Documentation issue The CNP Plant Boundary Definition and Partitioning Report has been updated to include more detailed bases for excluded buildings. In addition, drawings have be added to enhance graphical illustration of excluded locations. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that the Plant Boundary Definition and Partitioning Report (R1900-0041-001) contains a discussion and detailed listing of excluded locations. Attachment 3 provide s the plant partition boundary drawings.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

ES-A1-1 SR ES-A1 has specific definitions of which fire initiating events need to be included, defined in terms of initiators that cause automatic scrams, manual scrams per procedure, or LCO shutdowns under certain conditions. The Fire PRA Component Selection report (PRA-FIRE-17663-0002) simply references the internal events PRA initiating events analysis, without saying how (or even if) the criteria used there to identify internal initiators are consistent with Fire PRA SR ES-A1. Basis for Significance: If the internal events PRA initiating events analysis satisfies the requirements The description provided in the CNP Fire PRA Component Selection report (calculation PRA-FIRE-17663-002-LAR), Section 3.4.1, has been updated to show how the internal events PRA initiating event analysis meets the requirements of the combined PRA Standard SR ES-A1 (Chapter 4, Fire). (From peer review: Met at CC-I-II Suggestion)

RAI not needed.

Staff sees that the DC Cook Fire PRA Component Selection report (PRA-FIRE-17663-002-LAR) was updated to show how internal events initiators were considered in equipment selection. Tables 3.4.1-1 through 3.4.1-5 indicates explicitly how internal event initiators were dispositioned in the Fire PRA Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 5

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings for HL-IE-A, then it should be consistent with Fire PRA SR ES-A1. including which events were not included.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

ES-B1-1 Internal events systems were excluded from inclusion in the FPRA model based on sensitivity results generated from a scoping fire PRA (17663-002 calculation, Table 3.4.2-1). This sensitivity has not yet been repeated with the full scope fire PRA. Basis for Significance: Confirmation of conclusions produced from initial scoping sensitivity The sensitivity studies performed at the start of the fire PRA were performed again just prior to submittal of this LAR using the new full scope fire PRA. These sensitivity studies are documented in the Uncertainty/Sensitivity calculation (PRA- FIRE-17663-015-LAR) and were performed to confirm conclusions produced from the scoping fire PRA. (From peer review:

Met at CC-II Suggestion)

RAI not needed.

The sensitivity studies performed again just prior to submittal of the LAR using the full scope fire PRA and is documented in DC Cook Fire PRA Uncertainty and Sensitivity Analysis (PRA- FIRE-17663-015-LAR).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

ES-B2-1 MSO scenarios were generally well defined, but there were three issues identified: 1. Loss of the running charging pump on spurious closure of one of two VCT outlet valves is identified as a fire-induced spurious failure for loss of RCP seal injection. However, the failed status of the pump is not transferred to the HHI fault tree if the scenario results in RCP Seal LOCA. 2. Failure to trip the RCPs is modeled for impact on RCP seal leakage following loss of RCP seal cooling. However, it does not appear that spurious re-start of the RCPs was considered. 3. Spurious closure of both the charging pump discharge valve (QRV-251) and the miniflow valve on the running pump (QMO-225 or QMO-226) could result in pump failure. This is identified as PWR The three specific issues noted in the fire PRA peer review Finding were addressed with fault tree changes, and are documented in the Fire-Induced Risk Model calculation (PRA-FIRE-17663 LAR). Additionally, two general issues associated with this Finding were addressed. The first was to review the Fire PRA modeling to ensure that equipment failures leading to consequential events such as Loss of RCP Seal Cooling and Loss of Offsite Power were properly reflected in the fault trees of the CNP fire-induced risk model. The second general issue was to review the current industry guidance from NEI regarding identification of MSO scenarios, specifically to (From peer review:

Met at CC-II Finding) RAI not needed.

During the audit the licensee stated that the three issues specially noted in the peer review Finding were addressed with fault tree changes, and are documented in the DC Cook Fire-Induced Risk Model report (PRA-FIRE-17663 LAR). Staff found MSO consideration of spurious closing of the charging pump discharge valve, QRV-251, and mini- recirculation valves, QMO-225 and Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 6

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings generic scenario 51 in NEI 00-01, Rev. 2, Appendix G, but does not seem to have been addressed in the Cook MSO Expert Panel. Basis for Significance: The identified issues are not expected to significantly affect the results because of the availability of four pumps to fulfill the HHI function and the expectation that the cables associated with spurious re-start of the RCPs are also associated with failure to trip. However, this is identified as a finding because it could not be determined if additional cases of the same issues exist that could be more significant. ensure that subtle variations between scenarios have been covered in the CNP fire PRA. The CNP MSO Expert Panel report (calculation PRA- FIRE-17663-002-LAR), has been updated to reflect this review. 226, on pages 142, 144, 249 and 250.

Evidence of the other two changes could not be identified from this report as it does not provide the fault tree logic. The licensee also points out that fire modeling was reviewed to be sure it was appropriately reflected in the fault trees and industry guidance on MSO was reviewed. Review of industry guidance of MSOs is evidenced in Attachment B of the DC Cook MSO Expert Panel report (PRA- FIRE-17663-002b-LAR). Even though staff could not find evidence of all updates asserted, given the low significance of the issues and the allowance that there may still be minor inconsistencies across supporting documents we found no need for additional information.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

ES-D1-1 While documentation in Calculation 17663-002 was sufficient to support peer review, the following minor issues were identified: 1. Treatment of interlocks and power supplies is not specifically mentioned in 17663-002. For MOVs and pumps, these appear to have been treated as within the primary component boundary and picked up in the cable selection. 2. Tables C-1 and C-2 list the applicable WinNUPRA basic event name is entered as "see 1- RCP-ALL" or "see 2-RCP-ALL." The fault tree models incorporate discrete basic events for each RCP breaker (e.g., 1ABCB----- 1B9FAF, 1AACB-----1C2FAF). 3. Tables C-1 and C-2 do not contain disposition remarks for 1-XSO-315, 1- XSO-The updated CNP fire PRA Component Selection report (calculation PRAFIRE- 17663-002-LAR) was reviewed to ensure it reflects changes that were made during updates to related fire PRA tasks such as development of the Plant Response Model (calculation PRAFIRE- 17663-005-LAR), the multiple spurious operation expert panel, and detailed HRA (calculation PRAFIRE- 17663-012-LAR). Additionally, a description of the treatment of interlocks and power supplies considered within the boundary of the primary component was added to The updated CNP fire PRA Component Selection report (calculation PRA-(From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that DC Cook PRA Component Selection report (PRAFIRE- 17663-002-LAR) was updated to include treatment of power supplies and explains treatment of interlocks (see Section 4.0). The license asserts that the PRA Component Selection report was reviewed to make sure it contained changes resulting from updates to other fire PRA Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 7

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings 325, 2-XSO-315, and 2- XSO-325. These dispositions were found in the UNIT-1-2-SSEL+PRA COMPS.xls file provided during the review. 4. The correlation between HEPs identified in Table 3.4.5-1 and the instrumentation listed in Table 3.4.5-2 could be improved. As currently fo rmatted, it is hard to determine which instruments were added to support each HEP. 5. The process described in 17663-002 for identification of components to address multiple spurious operation concerns does not match the expert panel process used. There is also no reference to the Multiple Spurious Operations Expert Panel Final Report. Basis for Significance: These issues do not affect the analysis results. FIRE-17663-002-LAR). updates. However, in checking for update of issue #3 of the peer review comment (we were told during the Nov. 2011 audit that disposition remarks for 1-XSO-315 and X-XSO-325 had been added), we could not find that this update had occurred. None-the-less, given the insignificance of this issue and the allowance that there may still be minor inconsistencies across supporting documents staff found no need for additional information.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. CS-A2-1 New MSO scenarios produced from an update of the MSO Expert Panel have not yet been incorporated into the cable selection analysis. Basis for Significance: Potentially significant impact to the FPRA risk profile. The update to the MSO Expert panel has been incorporated in the detailed circuit analysis task effort/evaluation. The evaluation addresses multiple cable failure mode to support the new MSO scenario, and also addresses multiple concurrent cable failures. (From peer review:

Met at CC-II Finding) RAI not needed. (But related RAI against FSS-A2- 1 was written to address multiple cable failure issue) During the audit the license stated that results of an update of the MSO Expert Panel were incorporated into the cable selection analysis. Sta ff sees that DC Cook PRA Component Selection report (PRAFIRE- 17663-002-LAR) dated August 29, 2011 explains on page 9 how the results of the Expert Panel were incorporated and references the DC Cook MSO Expert Panel report (PRA- FIRE-17663-002b-LAR) dated Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 8

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings August 20, 2011 Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. CS-A9-1 No specific discussion is provided in the Fire PRA Cable Selection/Cable Routing report about how proper polarity hot shorts on ungrounded DC circuits are considered by the model. Basis for Significance: Documentation enhancement Fire PRA Cable Selection and routing tasks are performed using guidelines from NEI 00-01, NUREG 6850 and a procedure developed to provide specific guidelines in performing these tasks, the procedure developed discusses how proper polarity hot shorts on ungrounded DC circuits are considered in the model. Additional to the guidelines in these references, the detailed circuit analysis evaluation also discusses the proper polarity on ungrounded DC circuits. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

The licensee points out that guidance documents, NEI-00-01 and NUREG-6850, provide guidance on proper polarity for ungrounded shorts. Staff sees also that guidance on proper polarity for ungrounded shorts is provided in Detailed Circuit Analysis (TE R1900-0049-0001) on page 7.

Given the additional information, the NRC staff finds the disposition to the F&O to be acceptable. CS-B1-1 Any additional circuits and cables associated with the identified potential associated circuit's issues have not been identified. Basis for Significance: FPRA plant response model does not address associated circuits issues Technical Evaluation 12.5 identifies power supplies that could have potential associated circuit issues (common power supply and common enclosure). These power supplied are being addressed under new technical evaluation to show that they will not have potential associated circuits issues. (From peer review:

Not Met at CC-I-III Finding) RAI submitted as SSD RAI-15(a) and 15(b)

Technical Evaluation 12.5 (not published Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 9

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings until September 26, 2011) is specifically devoted to associated circuit issues by common power supply and common enclosure. It identifies deficiencies in protection DC power supply cables, associated with 4KV switchgear and concludes (in Attachments 4 and 5) that secondary fires cannot be precluded. Also, Technical Evaluation AEPDCC-11-001 addresses the coordination of certain panels not addressed in Technical Evaluation 12.5. A follow up report, Secondary Fire Evaluation (Technical Evaluation 12.5.1), addresses the secondary fire issue for 125 DC circuits and concludes that the location of postulated secondary fires will not affect post fire safe shutdown.

However, it is not clear that Technical Evaluation AEPDCC-11-001 addresses the secondary fire issues for several 250 DC circuits identified in Attachment 4 of Technical Evaluation 12.5, pages 65 through

68. The responses provided to RAI-15(a) in a letter dated April 27 th 2012 stated, based on review, that despite failures that could lead to secondary fires that fires in credited NSCA equipment or cables would not occur. The response provided to RAI-15b in a letter dated August 9 th 2012. Determined that fuses for a number of 250 V DC circuits should be replaced with fuse of a lower current rating to provide assurance that Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 10

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings secondary fires would not occur in these circuits and has implemented compensatory measures that will remain in place until these fuses are replaced and has also committed to perform applicable modifications if a potential for secondary fires in other types of circuits is identified (see SE Section 2.8, Table 2.8-1: Modification S-2.3). Based on the licensees commitment to resolve the secondary fire concerns with modifications the NRC staff finds that not specifically meeting SR CS-B1 is acceptable for this application. Once the modifications are in place, the finding will be resolved. CS-C1-1 The Fire PRA Cable Selection/Location provides a table of location data on a cable-by-cable basis. To facilitate the understanding of fire impacts on a compartment basis, consider adding a table of compartment-by-compartment cable locations. Basis for Significance: Documentation enhancement Technical Evaluation R1900-0043-0001 -"Fire PRA Cable Selection and Routing" has been updated to show the cable compartment location(s) for cables. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that Table 1 in Fire PRA Cable Selection and Routing (Technical Evaluation R1900-0043-0001) shows all Fire Zones all the cable routing for a given cable.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. CS-C3-1 Section 4.0, Uncertainty, of R1900-004-003 R0 Fire PRA Cable Selection/Location identifies issues on the routing of trays and conduits due to drawing legibility issues. In some cases previous drawing revisions were available to review and provided insight, if not positive identification. For the remaining cases, the engineer assigned to t routing There were no instances where the engineer(s) routing a cable were not able to determine any fire zone for the cable's raceways, i.e. there were no instances where the engineers used an engineering judgment in the entire cable route. In cases where the engineer preparing the cable route or the engineer reviewing the (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

The licensee clarified during the audit and in Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 11

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings task applied their judgment assigning a fire zone location to the cable. This engineering judgment was based on the last known raceway or conduit location positively identified on a layout drawing. Basis for Significance: The discussion in Section 4.0, Uncertainty, implies that the routing of some population of trays and conduits could not be verified; however, discussion with those responsible for determination of the routings expressed a high level of confidence in the accuracy of the routings determined. route applied judgment it was discussed with CNP design personnel to verify the required fire zone information. the F&O disposition that the fire zones cables were routed through were in all cases verified with design personnel.

Based on this explanation, the NRC staff finds the disposition to the F&O to be acceptable.

CS-C4-1 Circuits/cable related to associate circuits concerns have not been included in the Associated Circuits by Common Power Supply and by Common Enclosure, Rev. 0 reports. Also, no reference is provided to this report from the Fire PRA Cable Selection/Location document. Basis for Significance: Documentation enhancement Technical Evaluation 12.5 identifies power supplies that could have potential associated circuit issues (common power supply and common enclosure). These power supplied are being addressed under new technical evaluation to show that they will not have potential associated circuits issues. (From peer review: Met at CC-I-III , although Table V-1 Table shows this as "Open" Suggestion)

RAI not needed.

Staff sees that Technical Evaluation 12.5 (not published until September 26, 2011) is specifically devoted to associated circuit issues by common power supply and common enclosure. It identifies deficiencies in protection DC power supply cables, associated with 4KV switchgear and concludes (in Attachments 4 and 5) that secondary fires cannot be precluded. Also, Technical Evaluation AEPDCC-11-001 addresses the coordination of certain panels not addressed in Technical Evaluation 12.5. A follow up report, Secondary Fire Evaluation (Technical Evaluation 12.5.1), addresses the secondary fire issue but concludes that the location of postulated secondary fires will not Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 12

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings affect post fire safe shutdown.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. CS-A11-1 Section 4.0, Uncertainty, of R1900-004-003 R0 Fire PRA Cable Selection/Location identifies issues on the routing of trays and conduits due to drawing legibility issues. In some cases, previous drawing revisions were available to review and provided insight, if not positive identification. For the remaining cases, the engineer assigned to the routing task applied his judgment in assigning a fire zone location to the cable. This engineering judgment was based on the last known raceway or conduit location positively identified on a layout drawing. Basis for Significance: The discussion in Section 4.0, Uncertainty, implies that the routing of some population of trays and conduits could not be verified; however, discussion with those responsible for determination of the routings expressed a high level of confidence in the accuracy of the routings determined. There were no instances where the engineer(s) routing a cable were not able to determine any fire zone for the cable's raceways, i.e. there were no instances where the engineers used an engineering judgment in the entire cable route. In cases where the engineer preparing the cable route or the engineer reviewing the route applied judgment it was discussed with CNP design personnel to verify the required fire zone information. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

During the audit the licensee clarified that the fire zones cables were routed through were in all cases verified with design personnel.

Based on this explanation, the NRC staff finds the disposition to the F&O to be acceptable.

PRM-B2-1 Specific open F&Os from the 2001 peer review were identified that are judged to potentially impact the FPRA plant response model: 1) Observation TH-5 (Sub-Element TH-4) -CEQ fans are excluded from the hydrogen control evaluation for LERF estimation with no clear basis. 2) Observation SY-17 (Sub-element SY-17) -Cross-tie for AFW from Unit 2 does not consider the need for AFW at Unit 2 (e.g., prior or concurrent trip of the opposite unit), and 3) No apparent documented basis The internal events PRA has developed and documented resolutions for the C-level F&Os that could adversely affect the development of the Fire PRA plant response model. Resolution of SY-17 addresses all credited cross-ties. This review has been documented in calculation file PRA-FIRE-17663-005-LAR. (From peer review:

Not Met at CC-I-III Finding) RAI submitted as RAI-34(a)

LAR Table U-1 shows several Level-C F&O to still be open, including TH-4 and SY-17, (with proposed resolutions). RAI-34(a) asks to describe how the resolutions of open Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 13

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings for sufficient bottles air supply inventory for PORV operation for the 24-hour mission time. Basis for Significance: Potential impact to the FPRA risk profile F&Os from the 2001 internal events peer review were incorporated into the fire PRA. Include in the response specifically how the three cited open F&Os in PRM-B2-1, judged to have significant potential impact to the FPRA risk profile, were addressed.

The response submitted for RAI-34(a) in a letter dated August 9 th 2012 specifically provides dispositions for these open F&Os. In one case the PRA model was modified in response. In the other two cases adequate justification for not changing the model was provided.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PRM-B9-1 A review of calculation 17663-0005 Appendix B1, Table B-1 revealed that some items noted as included in fault tree modifications were not, in fact, incorporated. These include: 1. For valves 1-IMO-262 and 263, it is stated in the "How component is included in FPRA" column that basic events 1FAMV--IMO262CSF and 1FAMV--IMO263CSF were added to the 1HPI fault tree. The MSO Expert Panel Report noted that "Spurious opening/ or closing of minimum flow recirc lines for AFW, SI, CHP, RHR is modeled in fire PRA as in the internal events PRA, with the same success criteria." Spurious closure of these valves was modeled in the internal events PRA because control power is removed from the valves by a switch in the control room. However, it is possible that a fire- induced circuit failure could result in a hot short bypassing The fire-induced risk model was reviewed. Calculation PRA-FIRE-17663-005-LAR was updated to ensure that the changes identified were correctly incorporated into the FPRA and adequate justifications were developed for any fire-induced impacts that were not modeled. Also, the FPRA development team reviewed the FPRA to ensure that MSO Expert Panel was properly modeled and the modeling is consistent with that in the internal events PRA. (From peer review: Met at CC-I-III Finding) RAI not needed.

During audit the licensee confirmed that the fire induced risk model and the DC Cook Fire PRA Fire-Induced Risk Model report (PRA-FIRE-17663-005-LAR) were reviewed to ensure all updates were appropriately and consistently incorporated. Staff checked Appendix B, page 121 and can see that 1-IMO-262 and 1-IMO-262 are shown as in the fire PRA WinPRA model. With regard to disposition of MSOs: we found that the disposition of each MSO is provided (added)

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 14

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings this switch. Spurious closure of these valves in combination with spurious start of the associated SI pump could result in pump failure if RCS pressure is high. 2. For 1-TCSE, it is stated in the "How component is included in FPRA" column says "added CFF failure to 1T11A and 1T11D where appropriate." However, these events were not added to the applicable fault trees. The justification for not including the cooling fan failures was that the fans have an alternate power supply from the supported transformer. Therefore, the cooling is considered within the transformer boundary. Basis for Significance: Omission of failure modes identified in the equipment selection and MSO expert panel without adequate justification may impact the results. Although the identified items are not expected to significantly impact the results, there may be additional items not identified by the review team that could be significant. in the Section 6.3 evaluation tables as the "Status of MSO Evaluation in PRA Task 2" entry. Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

PRM-C1-1 Several documentation issues were identified in Calculation 17663-0005: 1. In Table B-1 the basic event IDs associated with component 12- CRV-51 is given as 1D0AV-12CRV51OSF. It appears that the correct ID should be 0D0AV-12CRV51OSF. 2. In Table B-1, the remarks for component 1- QCM-250 note that "per RCB event no longer considered to fail TBC." The review team was told that this disposition has been formally documented. The remarks should be updated to provide this more formal reference. 3. In Table B-2 the basic event associated with component 12-CRV-51 is given as 2A0CB----21BDCSF. It appears that the correct ID should be 0D0AV-12CRV51OSF. 4. Appendix A Tables 4-2, 4-4, and 4-8 lists the mission The fire PRA development team corrected the identified documentation issues and reviewed Tables B-1 and B-2 to ensure that the modeling notes reflect the current fire PRA resolution. The fire PRA plant response model is documented in calculation file PRA-FIRE-17663-005-LAR. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff checked the typographical errors and documentation deficiencies in the DC Cook Fire PRA Fire-Induced Risk Model report (PRA-FIRE-17663-005-LAR) and they appear to be corrected as indicated in the disposition. For example the basic event name associated with 12-CRV-51 has been corrected to "OD0AV-12CRV51OSF" in Table B-1 and Table B-2, and the remarks Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 15

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings time for Bleed and Feed as 0.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. It appears that the systems to support Bleed and Feed are modeled for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and that this is the correct mission time. If 0.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> is considered the correct mission time, the basis needs to be provided. 5. The list of items included in the small LOCA initiator in Appendix A Section 4.2.1 includes "RCP seal LOCA caused by failure of RCP cooling without reactor trip." This is not listed in Section 4.2.3 in the top event description for Small LOCA. 6. The description of top event SBO in Appendix A Section 4.1.4 discusses the fact that availability of the turbine-driven auxiliary feedwater pump impacts the time to perform manual cross- ties. However, it is not clear from the wording that the turbine-driven AFW pump and cross-ties are modeled within the SBO fault tree. 7. Calculation 17663-0005 Appendix A Section 4.1.4 says in the description of top event RCP-1 that the "Westinghouse seal LOCA model identifies three distinct seal LOCA sizes, 21 gpm/pump, 182 gpm/pump, and 480 gpm/pump." WCAP-16141 actually identifies four seal leakage rates of 21 gpm/pump, 76 gpm/pump, 182 gpm/pump, and 480 gpm/pump. It appears that the probability of the 76 gpm/pump leakage rate (.01) has been combined with the 21 gpm/pump leakage probability. Since there is no difference in the mitigation requirements for all leakage rates between 21 gpm/pump and 182 gpm/pump, this is considered to be only a documentation issue. Basis for Significance: The modeling for all of the identified items appears to be correct. Therefore, these issues are not expected to affect the results. associated with 1-QCM-250 in Table B-1 had been improved. Also, the list of Small LOCA contributors presented in Sections 4.2.3 of Appendix A was been updated to include RCP seal LOCA and is now consistent with Section 4.2.1. On the other hand the mission time of 0.5 for Feed and Bleed presented in Appendix A, Tables 4.2, 4.4, and 4.8 has not been changed or justified (that we could find). However, given that the peer reviewer states that the modeling related with the seven cited issues appears to be correct, staff judges the significance of not addressing all seven documentation issues as not important enough to pursue with an RAI. Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 16

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings PRM-B11-1 No comprehensive review has been made of all sequence-specific internal events HEPs to see whether these variations need to be reflected in the FPRA model. Basis for Significance: No significant impact to risk profile is anticipated. This is judged to be primarily a documentation issue. A comprehensive review was conducted to identify sequence-specific internal events human failure events and has been documented in calculation file PRA- FIRE-17663-012-LAR (HRA). (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

A comprehensive review of internal events human failure events that should be included in the Fire PRA appears to be part of the process to identify and incorporate human errors into the Fire PRA and is described in DC Cook Fire PRA Human Reliability Analysis report (PRA-FIRE-17663-0012-LAR). Section 3.4 of that report specifically identifies human error events from the internal events PRA were used as input to the Fire PRA HRA process. Staff notes the late completion date of 17663-0012-LAR: October 31, 2011.

Given the additional information, the NRC staff finds the disposition to the F&O to be acceptable.

FSS-A2-1 The assessed capability category for this item is "Not Met". The methodology that is described indicates that all cables whose fireinduced failure could adversely affect a credited component are identified. This is true for both functional failure concerns as well as spurious actuation (operation) concerns. The issue arises due to the process that is used whereby the spurious threats are propagated into the Fire PRA. The process as implemented results in spurious events being evaluated based on a single cable failure. Instances where the concurrent failure of more than one cable is required are not The treatment of fire induced spurious equipment operations was upgraded to address the potential cases where multiple concurrent cable failures could occur and cause undesired equipment operations. (From peer review:

Not Met at CC-I-III Finding) RAI submitted as RAI 34(b)

This F&O was written as being closely associated to SR CS-A2 which requires that fire-induced spurious operations from TWO cables be considered. During the trip the licensee explained that after the peer review comment an effort was undertaken to Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 17

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings currently addressed in the Fire PRA. This includes instances where the set of cables co-exist in the same raceway. The treatment of this SR is closely related to CS-A02. In order to meet FSS-A2, the propagation of the failure combinations must be consistent. The treatment needs to include in the grouping of riskrelevant targets the cables supporting any identified circuits where hot shorts impacting up to and including two cables (including both intracable and intercable hot shorts) could lead to spurious operation of selected equipment. Basis for Significance: A review of the information provided in Table 3-6 of FIRE-PRA-17663-0010 coupled with discussions with the utility team concluded that the determination of whether a spurious event occurs relies on the entry in columns for those events. Unfortun ately, there are other instances where the concurrent failure of multiple cables is required to cause the upset state. These instances are identified via a discussion in another column in the table. Consequently, these discussions were not considered in the development of the damage set and therefore, the damage set effectively is limited to only those spurious events that result from a single cable failure. Higher order cable failures are implicitly treated as not being credible. Examples include 1-CMO-419, 1-CMO-429, 1- ICM-305, and 1-ICM-306. It is anticipated that additional cases may arise as a result of the update of the related report to incorporate changes as noted in other F&Os (MSO Expert Panel). explicitly consider and document the impact from TWO cables. The licensee showed how extra columns had been added to Table 3-7 of PRA-FIRE-17663-010. Staff asks an RAI to "Describe how the treatment of fire-induced spurious equipment operations was upgraded in response to FSS-A2-1 to address potential multiple concurrent cable failures and cause undesired equipment operations. [And also ask to] Specifically address in the response how two cables that if affected and could cause spurious operations were considered."

In the response to RAI 34(b) in a letter dated August 9 th 2012, the licensee describes the upgrade performed in response to the F&O. The licensee explained that first each cable was evaluated one at a time to determine if it could spuriously actuate, then cables were examined in pairs to determine if a pair could spuriously actuate. As a result of the F&O the licensee reviewed their assessment and improved the documentation. A column was added to the cable analysis table in the Detailed Circuit Failure Analysis report (R1900-049-0001) title "Spurious Multiple Concurrent Failures to further clarify whether a cable was multiple concurrent failure contributors. Example entries are provided in the response.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 18

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings FSS-A2-2 PRA-FIRE-17663-0011a, CNP Fire PRA detailed Fire Quantification of Individual Areas. The report identifies those scenarios where burnup was acceptable and, where not acceptable, and it provides a correlation to the Detailed Fire Models conducted as part of Task 4.11. As a sample, Fire Modeling Report AA3, (R1900-003-AA3), was reviewed to verify the correlation. There was a good correlation between the two, except for an instance where Fire Zones were listed as acceptable with full room burnup and with detailed scenarios conducted. Basis for Significance: This appears to be a minor issue since the CCDP and CDF values were identical between the burnup and fire modeling scenarios. The correlation between two documents "PRA-FIRE-17663-0011a, CNP Fire PRA detailed Fire Quantification of Individual Areas" and "Task 4.11, Detailed Fire Models" was established for the few discrepancies and review conducted for other detailed fire models to ensure that this is an anomaly. (From peer review:

Not Met at CC-I-III Suggestion)

RAI not needed.

During the audit the licensee explained that the current version of the DC Cook Fire PRA Detailed Fire Quantification of Individual Areas PRA-FIRE-17663-0011a,) carries all Fire Area scenarios forward to quantification even if they were very small contributors to the CDF. Therefore, inconsistency between PRA-FIRE-17663-0011a, and Fire Modeling Report AA3, (R1900-003-AA3) no longer existed. Staff notes that, in fact, no quantification screening was performed at all. Staff confirmed that scenario CDFs for all Fire Areas, including AA3, are presented in table 3-3 of PRA-FIRE-17663-0011a (see pages 22 thru 27 for AA3).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 19

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings FSS-A6-1 Fire Scenarios for the Unit 2 MCR are not complete. Basis for Significance: Methodology is established. However, specific risk for Unit 2 MCR is not identified. The Unit 2 MCR fire quantification was completed using the methodology used for Unit 1. This is described in calculation PRA-FIRE-17663-0010b-LAR. (From peer review:

Met at CC-I/II Finding RAI not needed.

The MCR fire quantification approach and CDF and LERF results are presented in the DC Cook Fire PRA Main Control Analyses report (Unit 1 and Unit 2] (PRA-FIRE-17663-001b-LAR). For example CDF and LERF for Unit 2 is presented in Tables 28 and 19.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-A6-2 The Fire PRA treatment of the Main Control Room basically divides the analysis into a limited number of cases. In general, there are two specific considerations that do not appear to have been specifically included or addressed. One involves a postulated fire in the panel section that contains the Main Control Room ventilation system controls. The analysis of this panel section does not address the consequential fa ilure of the ventilation system. The treatment applies a node probability of 0.10 for the random failure of the system. The other consideration that does not appear to have been addressed is whether the postulated progression of the fire event at a particular panel section would result in an event progression that could not be mitigated using features and controls available via OHP-4025-001. The current treatment assumes that a screening HEP of 0.10 can be applied to all abandonment cases. The potential that a fire at a The MCR analysis was revised, and the two cases identified during the peer review were evaluated. The results were incorporated into calculation file PRA- FIRE-17663-011b-LAR. (From peer review:

Met at CC-I/II Suggestion)

RAI submitted as RAI-29 In response to RAI 29.01, the quantification of main control room scenarios was expanded to include modeling of human error probabilities to respond to scenario-specific equipment damage. The HEPs for all ex-control room actions were re-evaluated using the most recent feasibility analysis performed as an input. The operator actions were evaluated using the detailed human reliability analysis quantification, per Section 5.3 of NUREG-1921. The results of the re-evaluation showed an increase in delta risk greater than the acceptance guidelines of Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 20

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings panel could put the plant into a configuration that would require functions and controls not available via OHP-4025 should be evaluated using an HEP of 1.0. Basis for Significance: A cursory review of the MCR results indicates that incorporation of the issues noted herein would not results in a significant impact to the quantification totals for the MCR. RG 1.205. The licensee will include the modification to the reactor coolant pump seals as a committed modification in Table S-2 of the LAR to reduce the delta risk to within the RG 1.205 guidelines. Based on the licensee's response to RAI 29.01 the staff and the crediting of the reactor coolant pump seal modification, the NRC staff finds the disposition of this F&O to be acceptable.

FSS-C2-1 In order to meet CII/III, CHARACTERIZE ignition source intensity using a realistic time- dependent fire growth profile (i.e., a time-dependent heat release rate) for significant contributors as appropriate to the ignition source. There is an inconsistency in the time to reach peak HRR for transient fires (6 minutes per the procedure vs. 8 minutes per the modeling reports). There is also insufficient justification for the duration of both transient and unconfined oil spill fires. Basis for Significance: There is an inconsistency in the identification of duration to reach peak HRR for transient fires in EPM-DP-FP-001 R1 (6 minutes) and the detailed fire modeling reports (8 minutes). Eight minute duration is consistent with FAQ 08-0052. The 40-minute transient fire duration for transient fires in the detailed modeling reports has not been justified relative to the up to 60-minute duration in FAQ 08-0052. The duration of unconfined oil spill fires (20 minutes) in the detailed fire modeling reports is very conservative when compared against typical durations from NUREG 1805 spreadsheet applications. The Detailed Fire Modeling Procedure, EPM-DP-FP-001, has been updated to reflect the correct duration used to reach peak HRR for transient fires. Transient fires were extended to 60 minutes per FAQ-0052. The duration of unconfined oil fires, while conservative, did not result in any oil fire scenario becoming a significant risk contributor, and therefore, no changes were made. (From peer review:

Met at CC-I/II Suggestion)

RAI not needed.

A spot check during the audit revealed that a in a number of cases the duration used to reach peak HRR for transient fires was extended to 60 minutes per FAQ 0052 as asserted by the licensee (e.g., DC Cook Fire Risk Evaluation - Fire Area AA50, page 41, Attachment 6), confirming the F&O disposition.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-C3 (From V-2 Table) Updates made to the Fire PRA in response to F&O FSS-C2-01 have incorporated appropriate (From peer review: Met at CC-I Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 21

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings (There is no F&O associated with this SR. However, Table B-6 of the fire PRA WOG Peer Review SR Assessment Table reads that: "While a more realistic treatment of fire growth behavior is used with fire decay times identified, there is a lack of justification for the times used. CC-I was identified since sufficient deficiencies were found with the times used (40 minutes for transients when FAQ-0052 indicated durations of up to 60 minutes, and a 20-minute fire duration for an unconfined oil fire where such fires would have a duration of not more than one minute) that the key.") justification for heat release rate profile stages and the SR is now consid ered to meet Cat II/III.Not an F&O)

RAI not needed.

Staff found this issue to be similar to this F&O against FSS-C2. A spot check during the audit revealed that a in a number of cases the duration used to reach peak HRR for transient fires was extended to 60 minutes per FAQ 0052 as asserted by the licensee (e.g., DC Cook Fire Risk Evaluation - Fire Area AA50, page 41, Attachment 6), confirming the F&O disposition.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-C5-1 Damage thresholds are based on thermoset cable thresholds. It is known that thermoplastic cables are installed, and the impact is being reviewed. Also, solid state components are documented in the detailed fire modeling reports at the higher thermoset damage thresholds. Basis for Significance: Use of a higher damage threshold can result in misidentification of failures. However, the issues are understood and being pursued. In the case of solid-state components, NUREG 1805 spreadsheets (FTDs) have been run to determine if damage does occur at the lower thresholds. The Fire PRA has been updated and revised to include impact of thermoplastic cable installations and to address solid- state components. (From peer review:

Met at CC-I/II Finding) RAI not needed.

Cable damage was completely reevaluated in Technical Evaluation 11.64, Rev 0, "Cable Material Impact Review", dated Nov. 2010. This effort re-identified the cables types and quantity of thermoplastic cable and address solid state components. During the trip the licensee explained that the original analysis was performed in 2009 before the Sandia results showing greater damage for thermoplastic cable.

Staff checked a number of fire modeling Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 22

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings reports (e.g., CNP, Detailed Fire Modeling Report - Fire Compartment: AA50, R1900-0411-AA50) and can see this re-evaluation referenced in Section 5.5.2, and see cable classifications with respect to thermoset and thermoplastic cables listed in Attachment 5. In a related RAI we ask why temperatures for thermoset are used when thermoplastic is identified in Attachment 5 for Fire Area AA57A. Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-C8-1 FSS-C08 requires confirmation that credited wraps are not subject to mechanical damage or direct flame impingement from high hazard sources (unless qualified or tested under these conditions). Basis for Significance: The detailed fire modeling report for AA36/42 indicates that two credited raceway wraps are not expected to be subject to mechanical damage or flame impingement from high hazards sources, inclusive of HEAF. Walkdown data has not been located to confirm lack of damage or impact. Should such damage or impact occur, the cables protected within the raceways could be subject to the environment of postulated fire scenarios in the location. Technical Evaluation 11.57 has been updated to provide objective evidence that credited wraps are not subject to damage from high hazard sources. (From peer review:

Not Met at CC-I-III Finding) RAI submitted as RAI-34(c)

During the trip the licensee explained that Technical Evaluation 11.57 had been redone to provide objective evidence that credited wraps are not subject to damage from high hazard sources including HEAFs. However, only the original Technical Evaluation 11.57, dated Dec 20019, could be located on the Viewer or produced during the trip. Staff asked an RAI to describe how the Technical Evaluation was updated and to specifically explain how HEAFs were considered.

In response to RAI 34(c) in a letter dated August 9 th 2012, the license explains that Technical Evaluation 11.57 was updated to Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 23

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings include objective evidence from walkdowns that credited wraps are not subject to mechanical or heat related damage from hazard sources. This included determining whether a catastrophic transformer or hydrogen explosions could physically impact the fire wrap.

B Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-D4-1 Detailed fire modeling inputs for cable trays which have tray covers credit a 20-minute delay for damage and ignition. During walkdowns small holes were noted in some trays bottoms which had been credited with the 20-minute delay. The impact of the tray bottom design with these holes should be evaluated and documented. Basis for Significance: Delays in ignition and damage times could be overestimated. The Fire PRA has been updated and revised to remove credit for cable tray covers in locations where cable tray bottoms have small open holes. (From peer review: Met at CC-I-III Finding) RAI not needed.

During the audit the license explained that the credit for the tray bottom covers with holes in them had been removed in the few cases where they existed as cited Fire Area AA43 as an example. Staff checked the CNP Detailed Fire Modeling Report - Fire Compartment AA43 (R1900-007-AA43) and found this exception for a handful of cases but in other cases (presumably with no holes in the cover the 20 minute credit is taken). See pages 9, 10, and 16 concerning riser 2EM-C217.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-D7-1 CNP fire detection and suppression analysis has been completed using the approach and Cat 1 is acceptable for the application. CNP considers the estimates of unavailability used (From peer review: Met at CC-I Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 24

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings guidance of NUREG/CR-6850 Appendix P. Plant-specific values for "unavailability" have been estimated and an evaluation for outlier behavior based on plant information from maintenance activities is being conducted and tracked via a project open item. This evaluation needs to be completed and documented to achieve a CC II rating. Basis for Significance: Meets Category I in the analysis to be appropriate. Suggestion)

RAI submitted as RAI-34(d).

During the audit the licensee argued that the individual RAW values for these unavailabilities to be "unimportant" (i.e. no higher than 1.32). It is not clear what the aggregate unavailability might be. Staff asked an RAI to provide rationale for why outlier behavior based on plant information being conducted and tracked through an open project item was not used and why using generic unavailability is acceptable.

The response to RAI 34(d) in a letter dated August 9 th 2012, the licensee explained that the evaluation of plant-specific outlier behavior had been completed and that the evaluation shows that the actual unavailable time for fire detection and suppression systems is greater than the generic unavailability value used in the CNP Fire PRA. The licensee provided a sensitivity analysis in Section 3.4.7 using the actual unavailability times indicating the correct values do not significantly impact the risk results in the LAR. Based on staff finds that the PRA is technically adequate with regard to SR FSS-D7 to support the fire risk evaluations and other risk calculations required for the NFPA 805 application. The self-approval acceptance guidelines are much smaller than the transition acceptance guidelines and the NRC Staff concludes that Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 25

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings the correct plant specific fire protection system unavailabilities should be incorporated into the licensee PRA before the quantitative results may be used to support future self-approval. The licensee included updating the PRA to use the plant specific values as implementation item S-3.19 FSS-D8-1 The detailed fire modeling reports, in which detection and suppression is credited, should be updated to include a specific section addressing system effectiveness. Currently some data is included in Section 5.2, some is in Attachment 1 and some is in Attachment 9. NFPA Code evaluations are complete but not specifically referenced. Basis for Significance: Enhancements A discussion of fire protection system effectiveness has been incorporated into all fire modeling reports. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff checked a few detailed fire modeling reports and found a section (Section 5.8.1.1, "Detection Effectiveness Evaluation" under Section 5.8.1 "Detection Analysis" that specifically addresses this concern.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-E3-1 A sentence in Section 3.3.4.2 of Scientech report PRA-FIRE-17663-0015 stops mid-sentence: "Results of sensitivity studies will be documented in" Basis for Significance: This editorial error should be fixed. Section 3.3.4.2 of Scientech report PRA-FIRE-17663-0015-LAR was updated to state the location of the results of the sensitivity studies. (From peer review: Met at CC-I Suggestion)

RAI not needed More than the Suggestion (i.e., related to the typo) the more important issue here is why the SR is met only at CC-I. Page B-39 of the Fire Peer Review Report (LTR-RAM-II-10-041) indicates that the detailed fire modeling reports present only a qualitative Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 26

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings discussion of uncertainty. In RAI 31 the NRC Staff requested that the licensee perform a statistical propagation of parametric uncertainty and assess the impact on the risk results presented in the LAR. Finding UNC-A1-1 on SR UNC-A1 noted that the licensee had not performed a propagation of parametric uncertainty through the PRA model. In response to the RAI (Reference 10) the licensee provided a sensitivity analysis in which PRA model parameters having distributions were statistically propagated in the CNP Fire PRA model. The sensitivity analysis results show that propagating the parametric uncertainties through the PRA model could potentially significantly impact the results of decisions because of the little margin available before the proposed CNP RI/PB FPP potentially exceeds the risk acceptance guidelines for Region II (small change) in RG 1.174. See Section 3.4.7 of this SE for the NRC staff's evaluation of this sensitivity study. Consequently, the NRC staff finds that the PRA is technically adequate with regard to SR UNC-A1, and associated SR QU-E3, to support the fire risk evaluations and other risk calculations required for the NFPA 805 application, and the licensee has demonstrated the capability to a perform the propagation of parametric uncertainty through the PRA model as needed to support future self-approval.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 27

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings FSS-G1-1 Table 3 of Scientech calculation file 17663-011c, Multi-Compartment Analysis, identifies adjacent areas that did not screen at Steps 1, 2 or 3 at a CDF of 1E-7 or less. The non screened areas are supposed to go onto the next stage. Area AA2 to AA25 has an identified CDF of 1.30E-7 in Table 3, which exceed 1.0E-7. However, these are not addressed in the subsequent stage. Basis for Significance: The significance of this multicompartment is low given that Table 3 identifies the CDF of this MCA as 1.30E-7. All other MCAs in Table 3 did progress to the next stage. This appears to be a single instance of incomplete documentation of the results of the next stage. The Multi-Compartment Analysis for Areas AA2 to AA25 with CDF 1.30e-7 was addressed in the subsequent stages of the Screening, with the final results provided in a table on page 36 of 56. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff found Scenario AA2-to-AA25 to be Dispositioned in Table 6 on page 29 of DC Cook Multi-Compartment Analysis (PRA-FIRE-17663-0011cLAR).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-G2-1 DEFINE screening criteria for multicompartment fire scenarios that provide reasonable assurance that the contribution of the screened physical analysis unit combinations are of low risk significance. The screening criterion, when based on a fire zone to fire zone analysis, is predicated on the results within each zone, and not necessarily fire spread between zones, especially when barriers between zones are credited. Basis for Significance: When a fire scenario was limited to a zone-basis, the approach was to field-verify the ability of the barriers to contain the fire. Appendix A fire barriers, which have been incorporated into the fire protection program as Appendix R barriers, form the zone boundaries. The approach is not clearly documented. For any zone where "whole room burnout" was employed rather than "detailed fire modeling" a discussion of the fire barriers was provided to justify that a fire can be expected to be contained within the single fire zone. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff found discussions of barriers in Table 6 and in cases of whole zone burnout. In DC Cook Multi-Compartment Analysis (PRA-FIRE-17663-0011cLAR)

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 28

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings FSS-G5 (From V-2 Table) (There is no F&O associated with this SR. However, Table B-6 of the fire PRA WOG Peer Review SR Assessment Table: "There is no evaluation of normally open fire doors with fusable links. A MCA was performed on AA2 to AA25, which are separated by a fusible link held open fire. Refer to FSS-PP-B5 relative to the evaluation of fusible link held open fire doors in compartment boundaries. Plant Boundary Definition and Partitioning Report Rev. 1, (R1900-004-001), Section 4.3, reads: "the plant portioning analysis at CNP does not credit spatial separation as a portioning element in defining physical analysis units.") The quantification of scenarios in the multi-compartment analysis has been reviewed and revised to take credit for active fire barriers, consistent with the changes made in SR PP-B5 above. This documented in calculation file PRA-FIRE-17663-0011c. (From peer review: Met at CC-I Not an F&O)

RAI submitted as RAI34(i)

There is no finding or suggestion for FSS-G5 but because PP-B5 was considered CC-I, FSS-G5 also assigned to CC-I.

During the audit it was explained that further assessment of active fire barriers was expanded to include high hazard events (i.e., HEAFs, transformer explosions, and hydrogen explosions) and is documented in the Plant Boundary Definition and Partitioning Report (R1900-0041-001), we ask an RAI to describe how the multi-compartment analysis was revised as a result. In the response to RAI 34(i) in a letter dated August 9 th 2012, the licensee cited the provided a description of assessment of active fire barriers was expanded to include consideration of high hazard events (i.e., HEAFs, transformer explosions, and hydrogen explosions) and is documented in the Plant Boundary Definition and Partitioning Report (R1900-0041-001). Staff sees that Section 4.5 (on pages 9 and 10) discusses crediting active fire barriers. All active fire barriers (i.e., fire dampers, held-open fire doors, water curtains) are identified in Attachment 2 of the Plant Boundary Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 29

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings Definition and Partitioning Report.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-G6-1 In order to meet CII/III, it is necessary to quantify the risk contribution of any selected multicompartment fire scenarios consistent with FQ requirements. Basis for Significance: The SR is met. The F&O only provides guidance for meeting CC-II/III. All the Table 5 MCAs were quantified for the risk contribution of any selected multicompartment fire scenarios consistent with FQ requirements. (From peer review: Met at CC-I Suggestion)

RAI not needed.

During the audit the licensee explained and staff sees that DC Cook Multi-Compartment Analysis (PRA-FIRE-17663-0011c-LAR) page 48 explains that the MCA was quantified but that the total sum of all MCA scenarios was 4.7E-8/yr which is 0.2% of the total CDF and therefore it's contribution was not carried forward into quantification of the total fire CDF.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

FSS-H5 (From V-2 Table) (There is no F&O associated with this SR. However, Table B-6 of the fire PRA WOG Peer Review SR Assessment Table reads: "No documentation of parametric uncertainty analysis. The SR is judged to be met at CC-I.") Results Documentation. The fire PRA results documentation has been updated to reflect recent model changes and also to provide additional information regarding the quantification process, to better support Fire PRA applications, updates, upgrades and future peer reviews.

Parametric Data Uncertainty. Because of the wide range in modeling choices, such as the t-squared growth model provided for in (From peer review: Met at CC-I Not an F&O)

RAI submitted as RAI 31 Staff found that the DC Cook Fire PRA Uncertainty and Sensitivity Analysis (PRA-FIRE-17663-015) provides a detailed description of which elements of the Fire PRA that parametric uncertainty was Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 30

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings NUREG/CR-6850, modeling uncertainties drive the fire PRA results insights. The development of the parametric data uncertainty characterization provides information that does not provide meaningful insights into the decision-making process. Thus, the parametric data uncertainty is not applicable to the NFPA-805 LAR submittal. The parametric data uncertainty will be accomplished in a future update to the fire PRA. In the fire PRA, the endpoints of the fire damage state trees represent mean values for each of the fire scenarios. The uncertainty distributions for these fire damage states include the parametric data uncertainty associated with the ignition frequencies as well as the other branch points on the fire damage state tree. The results of this development will be provided in an update to calculation PRAFIRE- 17663-015-LAR. propagated and accompanying rationale and why in some cases only a qualitative treatment was performed. However, Staff notes that a statistical propagation of parametric uncertainty was not completed and so a number of F&Os related to the performance of uncertainty analysis on different elements of the PRA are open. Staff asks an RAI to provide an estimate of the difference between the mean and the point estimate and the impact this may have on risk acceptance.

In RAI 31, the NRC Staff requested that the licensee perform a statistical propagation of parametric uncertainty and assess the impact on the risk results presented in the LAR. Finding UNC-A1-1 on SR UNC-A1 noted that the licensee had not performed a propagation of parametric uncertainty through the PRA model. In response to the RAI (Reference 10) the licensee provided a sensitivity analysis in which PRA model parameters having distributions were statistically propagated in the CNP Fire PRA model. The sensitivity analysis results show that propagating the parametric uncertainties through the PRA model could potentially significantly impact the results of decisions because of the little margin available before the proposed CNP RI/PB FPP potentially exceeds the risk acceptance guidelines for Region II (small change) in RG 1.174. See Section 3.4.7 of this SE for the NRC staff's evaluation of this sensitivity study.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 31

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings Consequently, the NRC staff finds that the PRA is technically adequate with regard to SR UNC-A1, and associated SR QU-E3, to support the fire risk evaluations and other risk calculations required for the NFPA 805 application, and the licensee has demonstrated the capability to a perform the propagation of parametric uncertainty through the PRA model as needed to support future self-approval.

IGN-A7-1 Observation: Calculation 17663-0006 in section 3.4.7.2 assigns a storage weighting factor of 0.1 for plant areas where no storage locations are to be applied and a maintenance weighting factor of 0.05 for plant areas where hot work will be prohibited at power. It is recognized that establishing enhanced administrative controls will reduce the likelihood for the storage of transient materials or hot work in these areas. However, the specific bases for the reductions used are not documented in a manor that related to CNP specific plant experience. Basis for Significance: Impacts robustness of the update process. Bases should be known so this assumption can be included in some type of monitoring program to ensure changes to plant operating experience specific to hot work and transient combustible controls are evaluated. The bases for 0.1 storage and 0.05 maintenance weighting factors were added as they relate to CNP operating experience to section 3.4.7.2 of calculation file PRAFIRE- 17663-006-LAR. (From peer review: Met at CC-I-III Finding) RAI submitted as RAI 34(e)

The licensee used a special weighting factor for apportioning fire frequency for transient fire that deviates from the guidance in NUREG-6850. An RAI was asked to perform a sensitivity analysis on this deviation.

In RAI 34(e), the NRC Staff requested that the licensee provide sensitivity analyses in which the maintenance, occupancy, and storage influence factors were each assigned values consistent with the guidelines in NRC-endorsed FAQ 12-0064. In response to RAIs 34e the licensee provided sensitivity analyses in which the maintenance, occupancy, and storage influence factors were each assigned values consistent with the guidelines in NRC-endorsed FAQ 12-0064. See Section 3.4.7 of this SE for the NRC staff's evaluation of these sensitivity Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 32

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings studies. The results of these sensitivity studies by the licensee demonstrate that the risk calculations do not change significantly for this LAR. However, the NRC staff does not find the licensee's proposed method to be acceptable because it provided excessive flexibility to distribute transient fire frequency among different plant locations with no technical justification for modifying the acceptable method. In response to RAI 61, the licensee provided an integrated analysis which provided the risk results after changing this, and several other unacceptable proposed methods, to acceptable methods. Accordingly, the NRC staff concludes that the licensee has demonstrated that the CNP Fire PRA is technically adequate with regard to SR IGN-A7 to support the fire risk evaluations and other risk calculations required for the NFPA 805 application. However, because the method employed in apportioning transient fire frequency is not acceptable to the NRC, the NRC staff concludes that the NRC-endorsed method in FAQ 12-0064 should be incorporated into the PRA before the quantitative results can be used to support self-approval. The licensee included modify storage and hot work procedures and reflecting these modifications in the PRA in implementation items S-3.3, S-3.4. and 5-3.19 IGN-A7-2 For plant areas where hot work is to be prohibited during power operations, a maintenance weighting factor of 0.05 was assigned. Administrative controls would reduce the probability of the transient Plant area specific maintenance factors were assigned for bins 7, 25, and 37 as described in calculation file PRA-FIRE-17663-006-LAR. (From peer review: Met at CC-I-III Finding)

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 33

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings hot-work-related bins. However, the reduction has also been applied to Bins 7, 25, and 37 which are maintenance related transients. Therefore the reduction would not apply. Basis for Significance: Transient-related ignition frequencies are too low for the affected plant areas. The licensee used a special weighting factor for apportioning fire frequency for transient fire that deviates from the guidance in NUREG-6850. An RAI was asked to perform a sensitivity analysis on this deviation.

The response to RAI 34(e) provides a sensitivity study that shows the comparison between three variations of assigning transient fire weighting factors. However, none of the variations match the guidance in draft FAQ 0012-0064.

In response to follow-up RAI 34.01, the licensee provided a sensitivity analysis in which the maintenance, occupancy, and storage influence factors were each assigned consistent with the guidelines in NRC-endorsed FAQ 12-0064. The results of the analysis showed the total CDF and LERF for both CNP Units 1 and 2 remain well below 1E-4/year for CDF and 1E-5/year for LERF. In addition, the delta CDF and delta LERF for both CNP Units 1 and 2 remain below the risk acceptance guidelines of 1E-5/year for CDF and 1E-6/year for LERF for Region II (small change) in RG 1.174, which is acceptable to the NRC staff.

IGN-A7-3 Fire Influence weighting factor of "0" is assigned to outside compartments Unit 1 Yard, Unit 2 Yard, and 146 while assessing that transient fires will not impact risk important equipment or circuits. However, the basis for this treatment is not explicitly provided in the calculation. Transient fires are Weighting factors were assigned following the guidance of Table 3-5 of PRA-FIRE-17663-006-LAR for affected compartments and spatial separation to subdivide the yard was applied. (From peer review: Met at CC-I-III Finding) The licensee used a special weighting factor for apportioning fire frequency for transient Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 34

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings possible in these outside areas and are not precluded by design. Basis for Significance: The transient fire risk in the affected compartments may be underestimated. fire that deviates from the guidance in NUREG-6850. An RAI was asked to perform a sensitivity analysis on this deviation.

The response to RAI 34e provides a sensitivity study that shows the comparison between three variations of assigning transient fire weighting factors. However, none of the variations match the guidance in draft FAQ 0012-0064.

In response to follow-up RAI 34.01, the licensee provided a sensitivity analysis in which the maintenance, occupancy, and storage influence factors were each assigned consistent with the guidelines in NRC-endorsed FAQ 12-0064. The results of the analysis showed the total CDF and LERF for both CNP Units 1 and 2 remain well below 1E-4/year for CDF and 1E-5/year for LERF. In addition, the delta CDF and delta LERF for both CNP Units 1 and 2 remain below the risk acceptance guidelines of 1E-5/year for CDF and 1E-6/year for LERF for Region II (small change) in RG 1.174, which is acceptable to the NRC staff.

IGN-A7-4 Bins 13, 17, and 29 were excluded from consideration at CNP. The bases for exclusion should be added to the bin descriptions in 17663-006 Section 3.4.6. Basis for Significance: No impact to risk results. The bases for exclusion to the bin descriptions has been added, and described in calculation PRA-FIRE-17663-006-LAR. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that the basis for excluding Bins 13, 17, and 29 was added to Table 5-1 of DC Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 35

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings Cook Fire Ignition Frequency Calculation (PRA-FIRE-17663-006-LAR).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

IGN-A7-5 Bin 19 was excluded from consideration at CNP. During a plant walkdown of zones 110 and 111, H2 in 15% concentration cylinders with piping attached to a plant system was observed. These should be addressed in bin 19 or the bases for exclusion should be added to the bin description in 17663-006 Section 3.4.6. Basis for Significance: Low level risk impact Hydrogen systems in fire zones 110 and 111 were evaluated and their treatment documented in calculation PRA-FIRE-17663-006-LAR. (From peer review: Met at CC-I-III Finding) RAI not needed.

Staff sees that Bin 19 was included in DC Cook Fire Ignition Frequency Calculation (PRA-FIRE-17663-006-LAR) by attributing 50% of the bin frequency to Room 110 and 50% to Room 111.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

IGN-B1-1 The following issues were noted in Calculation 17663-0006 document: 1) FAQ Table 2-1 does not reflect current status and should be updated to reflect all FAQ's are approved. 2) The spread sheets for Attachments A and B were not attached. 3) Page 18 discussion of weighting factors states the low category is used for no hot work locations when a plant-specific 0.05 value was actually used. This is inconsistent with the page 16 discussion. 4) Numerous locations in the calculation contained an error message relating to a reference; example: pg 11. 5) Bin 10 discussion "They are" should be "They are", 6) Section 3.4.7.2 under transients "hto" should Editorial issues as needed were corrected in calculation PRA-FIRE-17663-006-LAR. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees editorial suggestions incorporated into DC Cook Fire Ignition Frequency Calculation (PRA-FIRE-17663-006-LAR) report. Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 36

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings be "hot", Basis for Significance: Does not impact risk results.

IGN-B1-2 Calculation 10766-0006 contains numerous assumptions such as the 5 located at the end of Attachment B spreadsheet that are not included as assumptions in Section 3.2. If these are not located in the "assumptions" section, they could be missed in the consideration of uncertainty analysis. Basis for Significance: No direct risk impact Calculation for embedded assumptions was reviewed and added to Section 3.2 in calculation PRA-FIRE-17663-006-LAR. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that the cited embedded assumptions were added to section 3.4.7.2 and referred to in Section 3.2. of into DC Cook Fire Ignition Frequency Calculation (PRA-FIRE-17663-006-LAR) report.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. CF-A1-1 The appropriate industry-wide conditional failure probabilities for fire-induced circuit failures have not been selected for the specific circuit configurations under consideration. Basis for Significance: Potentially significant impact to CNP FPRA risk profile All of the circuit failure likelihood probabilities in the CNP fire PRA were reviewed and, if appropriate, updated based on data located in NUREG/CR-6850 Appendix K. Additionally, valves with double-break design were re-analyzed to develop best-estimate circuit failure likelihood probabilities accounting for this plant-specific feature. The updated circuit failure probabilities are documented in calculation PRA-FIRE-176663-010-LAR. (From peer review:

Not Met at CC-I-III Finding) RAI submitted as RAI 34(f)

This F&O as it is presented in WOG FPRA Peer Review (LTR-RAM-II-10-041) provides eleven specific suggestions for correcting circuit failure probabilities. Staff asks the license in an RAI to address these eleven suggestions.

In response to RAI 34(f) in a letter dated August 9 th 2012, the licensee provides Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 37

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings explicit explanation of how the recommendation was followed. It most case the recommendation was followed as written. In two cases (#5 and #7) the recommendations were shown not to apply. In one case (#9) a more conservative value was used and in another case explanation for not following the recommendation was provided. Staff found all explanations to be reasonable.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. CF-A2-1 Observation: Parametric uncertainties of applied hot short probabilities have not been incorporated into the model. Basis for Significance: Technical adequacy of the PRA Appropriate error factors were applied to the hot short probabilities and were incorporated into the PRA model. The circuit failure probability error factors are documented in calculation PRA-FIRE-176663-010-LAR. (From peer review:

Not Met at CC-I-III Finding) RAI not needed (but related to RAI asked on uncertainty).

Staff found that the DC Cook Fire PRA Uncertainty and Sensitivity Analysis (PRA-FIRE-17763-0015-LAR) provides discussion of the quantitative treatme nt of circuit failure uncertainty and that in fact parametric data uncertainty for circuit failure was propagated using error factors (see page 7, 9, 10, and 22. Based on F&Os from other PRA elements related to uncertainty we ask a general RAI on propagation of parametric data uncertainty because for some elements of the PRA this was not performed quantitatively.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 38

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings Given the additional information and the fact that the more important underlying issue (i.e., completing the parametric propagation of uncertainty) is addressed in RAI 31, the NRC staff finds the disposition to this F&O to be acceptable.

HRA-A1-1 The assessed capability category for this item is "Met". The methodology that is described indicates that all credited HFEs are addressed for fire-related effects. The treatment includes consideration of cognitive and execution impacts. The treatment also includes consideration of the location of credited x-MCR actions and the location of the postulated fire events. A potential issue arises in that it is unclear whether or how the pathway for that action was addresses and resolved to have not been adversely affected. Basis for Significance: The occurrence of a large fire event could have consequential impacts related to fire brigade actions, smoke removal efforts, or other considerations could impact operator access to the location for the action. A confirmation of the pathways has been performed and is documented in PRA- FIRE-17663-012-LAR. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff notes that in DC Cook Fire PRA Human Reliability Analyses (FIRE-17663-012-LAR) that operator interviews and dependency analysis is specifically addressed (See staff evaluation of F&O FQ-C1-1).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

HRA-A2-1 This SR was assessed as Met. The SR involves the treatment of fire- specific safe shutdown actions in the Fire PRA. The treatment as described in FIREPRA-17663-0012 selectively applies the actions associated with 01/02-OHP-4025-001. However, this procedure is specific to Main Control Room Abandonment. The documentation includes a note in the operator interview discussion in Appendix B indicating that 'operators can refer to procedure for additional guidance'. However, there is no other reference, basis or justification provided. For example, it is unclear whether the 'credit' provided by Additional documentation has been provided to more clearly describe and justify the stated treatment of the MCR Abandonment procedures. This has been added to calculation PRA- FIRE-17663-012-LAR. (From peer review: Met at CC-I-III Suggestion)

RAI submitted as RAI 29 Staff notes that Section 3.5.4 of DC Cook Fire PRA Human Reliability Analyses (FIRE-17663-012-LAR) specifically describes use of 0.1 for MCR abandonment and provides criteria for its use. Staff asks the license in an RAI to explain why this is an appropriate Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 39

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings the treatment is consistent with operator training, whether the operator are trained in this specific usage of the procedure, or whether that usage is consistent with procedure itself. A concern was noted in that a review of 02-OHP-4025-001 found that the usage and credit of this procedure is inconsistent with the state purpose, entry conditions, and caution provided before Step 1 of the procedure. Basis for Significance: The analysis documentation and treatment are internally consistent in the credit assigned to the referenced procedure. Separate discussions and procedure reviews found that the treatment was consistent wi th the structure of the plant procedures but much of this discussion and evidence is not referenced or discussed in the documentation. Because the treatment was valid, this SR was judged to have been met, but that additional documentation should be provided to more clearly describe and justify the stated treatment. References to training material, guidelines, or other materials should be used and referenced as applicable. As it currently exists, a notable amount of discussion was necessary to examine this attribute. For example, OHI-4023, 4.6.9.f provides clarification of the phrase 'refer to'. 1-OHP-4023-ECA-0.0, step 12 provides a clear 'refer to' to 1- OHP-4025-LS-3 which provides evidence that CNP does in fact treat elements of OHP-4025 as individual recovery actions. value. In response to RAI 29.01, the quantification of main control room scenarios was expanded to include modeling of human error probabilities to respond to scenario-specific equipment damage. The HEPs for all ex-control room actions were re-evaluated using the most recent feasibility analysis performed as an input. The operator actions were evaluated using the detailed human reliability analysis quantification, per Section 5.3 of NUREG-1921. The results of the re-evaluation showed an increase in delta risk greater than the acceptance guidelines of RG 1.205. The licensee will include the modification to the reactor coolant pump seals as a committed modification in Table S-2 of the LAR to reduce the delta risk to within the RG 1.205 guidelines. Based on the licensee's response to RAI 29.01 the staff and the crediting of the reactor coolant pump seal modification, the NRC staff finds the disposition of this F&O to be acceptable.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 40

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings HRA-A3-1 This SR was assessed as Met and it involves the potential for fire-induced failures associated with instruments and alarms that could mislead operators such that they would perform an undesired action. The documents that were reviewed concluded that no undesired actions would occur. That conclusion is based in part on a statement that verification of the status is required before or after the action. A number of instances were identified wherein the action involves the tripping of a potentially critical pump (RHR, CCW, ESW, etc.). The information provided calls for the operator to trip the pump. No further discussion is provided to justify or credit a restart of the pump. As presented, it would appear that these tripping threats should have been included in the Fire PRA. Further discussions and reviews found that plant procedure OHI-4000, Section 3 includes explicit Expectations that the confirmation of the alarm condition is to occur before execution of the specific alarm response actions. The information provided in this procedure provides a much stronger basis and justification for the applied treatment and should be added to the documentation. Basis for Significance: The applied treatment and evaluated response of the plant operators to possible fire induced spurious alarms currently relies on interview notes. The plant procedure structure actually includes a much stronger justification and basis for the applied treatment and references to that procedure should be added to the documentation. The documentation has been enhanced to add references to, and discussion of, OHI-4000, Section 3. This procedure provides a concise and objective statement as to the expected response of the plant operators to alarms. This has been added to calculation file PRA-FIRE-17663-012- LAR. (From peer review:

Met at CC-II Suggestion)

RAI not needed.

Staff notes discussion in DC Cook Fire PRA Human Reliability Analyses (FIRE-17663-012-LAR) related to annunciators in Section 3.5.1, and related to staffing n Attachment A.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

HRA-C1-1 This SR was assessed as Met because only a single issue was noted. The documentation provided in FIRE-PRA-17663-0012 included a The cues associated with each human failure event included in the Fire PRA model was reviewed and confirmed, especially for those (From peer review: Met at CC-II Finding)

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 41

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings number of instances where the treatment was divided into two separate elements -one to specifically address the cognitive element and another to address the execution. An example was 1D- OPENHOORHE. This action involves only the execution of 'opening the door' for the motor driven AFW pump room. The cognitive element was addressed via ESW-20MINCOGHE. The indicated cues for this cognitive element involved only the status of the ESW pumps (trip, flow, pressure). A review of the implementation of this action found that 1D--- OPENDOORHE is credited for recovery for all failure modes of the cooler itself-not just loss of the cooling water dependency. In this instance, the scope of cues is inconsistent with credited recovery. Basis for Significance: The treatment of 1D---OPENDOORHE is such that credit was provided in the Fire PRA even for cases where no valid cue is provided to the operator. The cues are based solely on the availability of ESW to the cooler itself. A random or fire induced failure of the cooler itself would not be detected and hence the credited recovery would not be valid. human failure events where the cognitive contribution was modeled in a separate basic event. The cues modeled in the fire PRA have been documented in calculation file PRA- FIRE-17663-012-LAR.

RAI not needed.

Staff notes that the cited example (i.e., 1D-OPENHOORHE) occurs in only one fire area. During PRA audit the licensee explained that they assumed that the recognition in a fire event (fore for the cited HFE event) was considered to twice as unlikely as in an internal events event.

Based on this explanation, the NRC staff finds the disposition to this F&O to be acceptable.

FQ-A3-1 The quantification does not account for all scenario-specific quantification factors. Basis for Significance: Potential impact to FPRA risk profile. The fire PRA model has been modified to properly account for scenario-specific factors. This includes the following changes: 1) Impacts to HEPs based on equipment unavailability due to fire. For example, for fire zone 144, when the steam supply valves for the TD AFW pump spuriously close, but an HFE for the AFW cross-tie assumes the TD AFW pump is available for four hours. 2) Dual unit trip considerations have been added to account for dependency of mitigating systems needed for (From peer review: Met at CC-I-III Finding) RAI not needed.

Staff notes that the licensee appears to have modified the model per the F&O comment (e.g., the HEP for Zone 144 is removed). During the trip we were asked to note Table 4 which shows Fire Areas/Zones where local Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 42

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings opposite unit (e.g. AFW or charging crosstie if both units are tripped). actions are not credited.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. FQ-A4-1 The "state-of-knowledge" correlation between fire-specific event probabilities (e.g., suppression system unava ilability, fire ignition frequencies, hot short conditional probabilities, etc.) has not yet been applied. Basis for Significance: The "state-of-knowledge" correlation has not been fully addressed, but no significant impact on parametric uncertainty evaluation is expected. All applicable data was correlated. The initiating event basic events for "small pump fire" have a common match in the WinNUPRA prm file and "pre-action sprinkler suppression" systems have a common math in the WinNUPRA prm file. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

During the audit it was clarified that state-of-knowledge correlations (SOKC) was performed for like components (i.e., with the same type code). The F&O pertains to lack of SOKC for suppression system unavailability, fire ignition frequencies, hot short conditional probabilities, etc.

Given that a sensitivity study based on calculated mean values was performed in response to RAI 31 and the explanation here, the NRC staff finds the disposition to this F&O to be acceptable.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 43

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings FQ-B1-1 Quantification truncation limits were selected based on an iterat ive evaluation with a preliminary fire model. This iterative evaluation is recommended to be repeated for the full scope fire PRA. Basis for Significance: Current truncation limits appear to be appropriate. No likely impact to risk profile. Iterative calculations have been performed to ensure appropriate truncation limits. Convergence can considered sufficient when successive reductions in truncation value of one decade result in decreasing changes in CDF or LERF, and the final change is less than 5%. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that Section 5.7 of the DC Cook Fire PRA Integrated Fire Risk Analysis report specifically describes the truncation convergence method used (see FIRE-17663-014-LAR).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. FQ-B1-2 Observation: WinNUP RA utilizes the rare event approximation, which is not recommended when basic event probabilities are above 0.1. Basis for Significance: WinNUPRA is an industry-accepted code The documentation of modeling uncertainties includes a note that t he FPRA utilizes the rare event approximation, which introduces some conservatism into the results, and that for some applications this may be a consideration (particularly for cutsets in which several basic event probabilities are above 0.1). (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff acknowledges that WinNUPRA's use of the rare event approximation is somewhat conservative for cutsets with basic events with probabilities greater than 0.1. However, such approximation is state-of-art and winNUNPRA is an industry accepted code.

Based on this explanation, the NRC staff finds the disposition to the F&O to be acceptable. FQ-C1-1 The peer review team observed that the final confirmatory step to ensure that all significant HEP combinations with potential dependencies have A confirmatory analysis has been performed to ensure all significant HEP combinations with potential dependency have been identified. The global CDF and LERF equations (the (From peer review: Met at CC-I-III Finding)

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 44

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings not been identified. Basis for Significance: Potentially significant impact on the FPRA risk profile "global" equations combined together the cutsets from all significant FPRA sequences or all FPRA sequences) have been quantified with HEP values set to values that are sufficiently high so that the cutsets are not truncated. The additional HFE combinations that were identified were then evaluated and modified as appropriate to address the degree of dependency between the HFEs in the cutset or sequence. RAI not needed.

During the audit the licensee confirmed that this final step had been performed and that only a handful of combinations needed modification and refers to an updated report. The FPRA peer review report (LTR-RAM-11-10-041) dated July 2010 states on page B-61 "The HEP dependency evaluation has been performed for the FPRA as documented in Attachment D of the DC Cook Fire PRA Human Reliability Analysis. Identified HEP dependencies have been characterized and are modeled directly in the fault tree (rather than by cutset post-processing) which is judged to be a strength". Given the additional information, the NRC staff finds the disposition of the F&O to be acceptable. FQ-D1-1 Documentation of results for significant and non-significant sequences (cutset review) has not been performed for LERF. Review of importance values has not been performed for reasonableness. F&Os generated for other FQ SRs should be applied to the LERF model, as well as the CDF model. Basis for Significance: Potential impact to quantitative results A review of significant and non-significant LERF results was conducted. The results section of the fire PRA has been updated to provide importance values for LERF. This has been provided in calculation PRA-FIRE-17663-011-LAR. (From peer review: Met at CC-I-III Finding) RAI not needed.

The staff reviewed DC Cook Fire PRA Integrated Fire Risk Analysis report (see FIRE-17663-014-LAR) specifically provides LERF importance values in Tables 5-7, 5-9, 5-11, and 5-13. These results were found to be reasonable.

Given the additional information the NRC staff finds this deficiency has been Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 45

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings appropriately resolved by the licensee. FQ-E1-1 It is not apparent that any of the items listed in QU-D and LE-F, as clarified in the SR, were considered for the fire quantification -for example, review of significant and non significant cutsets, identification of key contributors to CDF and LERF, and review of importance of components and basis events. Basis for Significance: This SR requires that CDF and LERF internal events' quantification SRs be addressed for fire. This portion of the work was not completed at the time of the peer review. The results section of the fire PRA has been updated to provide the data in the supporting requirements from the internal events portion of the combined PRA standard (Chapter 2) associated with high level requirements QU-D and LE-F for LERF. This has been provided in calculation PRA-FIRE-17663-011-LAR. (From peer review: Met at CC-I-III Finding RAI not needed.

It was pointed out during the audit and staff sees that DC Cook Fire PRA Integrated Fire Risk Analysis report (see FIRE-17663-014-LAR) specifically provides LERF top contributors in Tables 5-17 and 5-19, and importance values in Tables 5-7, 5-9, 5-11, and 5-13. These results are reviewed for reasonableness.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee. FQ-F1-1 The documentation of the fire quantification provides the CDF and LERF results, but not to the extent that internal events quantification would be done, e.g., tables of most significant fires, ignition sources, cutset reviews, etc. Basis for Significance: The typical types of quantification tables identified above are needed to facilitate reviews and applications. Tables typically included for internal events quantification have been added to the report. The tables have been added to calculation PRA-FIRE-17663-011-LAR. (From peer review: Met at CC-I-III Finding) RAI not needed.

It was pointed out during the audit and staff sees that DC Cook Fire PRA Integrated Fire Risk Analysis report (see FIRE-17663-014-LAR) specifically provides normal results tables in Tables 5-1 through 5-19.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 46

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings SF-A3-1 Discussions with utility indicate that there are existing strategies to cope with a complete loss of onsite fire suppression systems, e.g., city water and pumper truck capability. Basis for Significance: This SR has been met; however, the additional existing strategies discussed with the utility to cope with a complete loss of onsite fire suppression systems needs to be included in the documentation. Note that since SF is allowed to be handled qualitatively, an issue here would have no impact on any quantitative result. Report PRA-FIRE-17663-013-LAR has been updated to include information on coping strategies. (From peer review: Met at CC-I-III Finding) RAI not needed.

Staff notes a paragraph on "first responders" on page 11 of DC Cook Fire PRA Seismic-Fire Interactions (PAR-FIRE-17663-00) that was not previously included.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

SF-A4-1 Scientech report PRA- FIRE-17663-0013 does not reference a "seismic response procedure" that was used to fulfill this SR. The plant's seismic response procedure was requested and provided by the utility (1- OHP-4022-001-007, Rev. 10, "Earthquake"). This procedure was reviewed in accordance with this SR. The procedure did not mention the possibility of a seismically-induced fire or spurious operation of fire suppression systems that could compromise postearthquake plant response; however, additional discussions with utility indicate that there are other existing procedures for coping with large-scale events, such as seismically-induced fires, e.g., emergency plan, fire pre-plans for beyond design basis events. The later of these provides actions to cope with a complete loss of onsite fire suppression. Basis for Significance: This SR has been met; however, the additional existing procedures discussed with the utility to cope with large-scale Report PRA-FIRE-17663-013-LAR has been updated to include the additional information on the plant's seismic response procedure. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees where DC Cook Fire PRA Seismic-Fire Interactions report (PAR-FIRE-17663-00) had been updated to include general discussion on ARPs, ONPs, EOPs, and SAMGs that are germane to coping with large-scale events, such as seismically-induced fires.

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 47

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings events needs to be included in the documentation. Note that since SF is allowed to be handled qualitatively, an issue here would have no impact on any quantitative result.

SF-A5-1 Scientech report PRA- FIRE-17663-0013 does not reference "fire brigade training procedures" that were used to fulfill this SR. The plant's fire brigade training procedures were requested and the FP training program description was provided by the utility (TPD-600-FP). This document was reviewed in accordance with this SR. There is no training discussed in the procedures regarding the ability/inability to fight a fire after an earthquake (either in terms of responding to fire alarms and fires, storage and placement of equipment, and access routes); however, additional discussions with the utility indicate that there is extensive training on large scale events such as those included in the fire preplans for beyond design basis events-Training activities that involve offsite fire response personnel, equipment, and water sources. Basis for Significance: This SR has been met; however, the additional existing fire-related training discussed with the utility to cope with large-scale events needs to be included in the documentation. Note that since SF is allowed to be handled qualitatively, an issue here would have no impact on any quantitative result. Report PRA-FIRE-17663-013-LAR has been updated to provide the additional information. (From peer review: Met at CC-I-III Suggestion)

RAI not needed.

Staff sees that training and fire brigade training is generally described however, notes that TPD-600-FP is not discussed in the DC Cook Fire PRA Seismic-Fire Interactions report (PAR-FIRE-17663-00).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

SF-B1-1 The documentation of the seismic/fire interaction analysis relies exclusively on the work done for the IPEEE. Several unvalidated assumptions are made in Scientech report 17633-0012 that are intended to justify that the work done Report PRA-FIRE-17663-013-LAR has been updated to validate the assumptions and fix the typographical errors. (From peer review:

Not Met at CC-I-III Finding) RAI submitted as 34(g)

Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 48

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings for the IPEEE remains valid today. These unvalidated assumptions are: 1) All fire suppression features have been installed in accordance with a Standard and all installed equipment is maintained in accordance with a Standard, 2) CNP as no cast iron fire mains, and 3) No significant changes to the plant fire protection features have been made that would render the IPEEE assessment invalid. These assumptions are made without any validation.

Therefore, the adequacy of reliance on the IPEEE seismic/fire (SF) interaction analysis is in question. The IPEEE assessment of seismic/fire interaction (circa 1992) was reviewed. It appears that the issue of having (or not having) cast iron fire mains was raised at that time and appears to have never been resolved (at least in the documentation provided to me). The IPEEE discusses future walkdowns to be performed to address this question, but later walkdowns were silent on this issue. The plant contends that no cast iron fire mains are used, but no documentation has been provided to verify this. Scientech report PRA-FIRE 17663-0013 is misnumbered in the body of the report (it is numbered 17663-0012). Basis for Significance: The validity of the seismic/fire interaction report is dependent on three unvalidated assumptions. Note that since SF is allowed to be handled qualitatively, an issue here would have no impact on any quantitative result.

Staff could not determine that the validation of these assumptions had been added to the DC Cook Fire PRA Human Reliability Analyses report (PRA-FIRE-17663-013-LAR). Neither could the licensee point them out during the trip.

In response to RAI 34(g) in a letter dated August 9 th 2012, the licensee explicitly addressed the three assumptions cited in the F&O. For the first two assumptions the license identifies the engineering documents that demonstrate compliance to the appropriate standards. In the third the licensee lists the fire protection features credited in the IPEEE and confirms that the Fire Hazards Analysis identifies the same fire protection features currently credited. Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.

UNC-A1-1 Calculation 17663-0015, Fire PRA Uncertainty and Sensitivity Analyses, does a good job of characterizing the potential sources of uncertainty in the Fire PRA model. However, propagation of parametric uncertainty for the fire I&M performed parametric uncertainty analysis of the CDF and LERF results for the risk-significant fire zones/areas, and documented this update in calculation file PRA-FIRE-17663-(From peer review:

Not Met at CC-I-III Finding) RAI Submitted as RAI 31 Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 49

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings CDF and LERF to address component failure probabilities, fire ignition frequencies, detector response, and auto suppression availability has not been performed. The Standard expects that the parametric uncertainty will be propagated through the baseline PRA model independent of any planned risk- informed applications. Basis for Significance: This is a required element for meeting the supporting requirement. 015-LAR. Staff found that the DC Cook Fire PRA Uncertainty and Sensitivity Analysis (PRA-FIRE-17663-015) provides a detailed description of which elements of the Fire PRA that parametric uncertainty was propagated and accompanying rationale and why in some cases only a qualitative treatment was performed. However, we note that a statistical propagation of parametric uncertainty was not complet3ed and so a number of F&Os related to the performance of uncertainty analysis on different elements of the PRA are open. Staff asks an RAI to provide an estimate of the difference between the mean and the point estimate and the impact this may have on risk acceptance.

In RAI 31, the NRC Staff requested that the licensee perform a statistical propagation of parametric uncertainty and assess the impact on the risk results presented in the LAR. Finding UNC-A1-1 on SR UNC-A1 noted that the licensee had not performed a propagation of parametric uncertainty through the PRA model. In response to the RAI (Reference 10) the licensee provided a sensitivity analysis in which PRA model parameters having distributions were statistically propagated in the CNP Fire PRA model. The sensitivity analysis results show that propagating the parametric uncertainties through the PRA model could potentially significantly impact the results of decisions Record of Review D.C. Cook LAR Attachment V - Tables V-1 and V-2 Fire PRA Peer Review - Facts and Observations (F&Os) 21-Oct-11 50

\Facts and Observations (F&Os) Licensee Disposition NRC Staff Findings because of the little margin available before the proposed CNP RI/PB FPP potentially exceeds the risk acceptance guidelines for Region II (small change) in RG 1.174. See Section 3.4.7 of this SE for the NRC staff's evaluation of this sensitivity study. Consequently, the NRC staff finds that the PRA is technically adequate with regard to SR UNC-A1, and associated SR QU-E3, to support the fire risk evaluations and other risk calculations required for the NFPA 805 application, and the licensee has demonstrated the capability to a perform the propagation of parametric uncertainty through the PRA model as needed to support future self-approval.

MU-B3-1 PRA Model Update Procedure 12- EHP-9010-PRA-001 Section 3.3 describes the process used for periodic PRA updates and Section 3.4 describes cases where an interim update may be appropriate. However, there is no explicit direction to: 1) Perform updates in a manner that maintains consistency with the supporting requirements of the ASME/ANS PRA Standard, and 2) Determine if changes to the model incorporated during a periodic or interim update meet the definition of a PRA upgrade requiring a peer review. Basis for Significance: Discussion with AEP personnel revealed that the actual practice meets the requirement even though the procedure does not explicitly require it. Sections 3.1.1, 3.1.2, and 3.3.1 of the CNP PRA Model Update procedure were revised to ensure that updates are performed to Capability Category II of the PRA Standard, and that the PRA Supervisor will identify whether a model change is an update or an upgrade requiring follow-on peer review. (From peer review:

Not Met at CC-I-III Suggestion)

RAI not needed.

During the audit the license presented the PRA Model Update Procedure 12-EHP-9010-PRA-001 showing that the PRA supervisor will determine whether a model change is significant enough to require a follow-on peer review. See item Section 3.1.1 item (e).

Given the additional information the NRC staff finds this deficiency has been appropriately resolved by the licensee.