PNP 2014-063, Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors

Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors

ML14169A046
Person / Time
Site:	Palisades
Issue date:	06/17/2014
From:	Vitale A Entergy Nuclear Operations
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
PNP 2014-063
Download: ML14169A046 (21)

Text

Entergy Nuclear Operations, Inc.

Palisades Nuclear Plant 27780 Blue Star Memorial Highway Entergy 4:o953o Anthony J Vitale Site Vice President PNP 201 4-063 June 17, 2014 U. S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, DC 20555-0001

SUBJECT:

Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors Palisades Nuclear Plant Docket 50-255 License No. DPR-20

References:

1.

ENO letter, PNP 2012-106, License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated December 12, 2012 (ADAMS Accession Number ML12348A455) 2.

ENO letter, PNP 2013-013, Response to Clarification Request

License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated February 21, 2013 (ADAMS Accession Number ML13079A090) 3.

NRC electronic mail of August 8, 2013, Palisades

- Requests for Additional Information Regarding Transition to the Fire Protection Program to NFPA Standard 805 (TAC No. MF0382) (ADAMS Accession Number ML13220B131) 4.

ENO letter, PNP 2013-075, Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated September 30, 2013 (ADAMS Accession Number MLI 3273A469) 5.

ENO letter, PNP 2013-079, Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated October 24, 2013 (ADAMS Accession Number ML13298A044)

~ Entergy PNP 2014-063 June 17, 2014 U. S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, DC 20555-0001 Entergy Nuclear Operations, Inc.

Palisades Nuclear Plant 27780 Blue Star Memorial Highway Covert, MI 49043-9530 Tel 269 764 2000 Anthony J Vitale Site Vice President

SUBJECT:

Response to Request for Additional Information - License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors Palisades Nuclear Plant Docket 50-255 License No. DPR-20

References:

1. ENO letter, PNP 2012-106, "License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors," dated December 12, 2012 (ADAMS Accession Number ML12348A455)

2. ENO letter, PNP 2013-013, "Response to Clarification Request-License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors," dated February 21, 2013 (ADAMS Accession Number ML13079A090)

3. NRC electronic mail of August 8, 2013, "Palisades - Requests for Additional Information Regarding Transition to the Fire Protection Program to NFPA Standard 805 (TAC No. MF0382)" (ADAMS Accession Number ML13220B131 )

4. ENO letter, PNP 2013-075, "Response to Request for Additional Information - License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors", dated September 30,2013 (ADAMS Accession Number ML13273A469)

5. ENO letter, PNP 2013-079, "Response to Request for Additional Information - License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors", dated October 24,2013 (ADAMS Accession Number ML13298A044)

PNP 2014-063 Page 2 of 3 6.

END letter, PNP 2013-083, Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated December 2, 2013 (ADAMS Accession Number ML13336A649) 7.

NRC electronic mail of March 11, 2014, Requests for Additional Information Palisades NFPA 805 Project LAR

- MF0382 (ADAMS Accession Number ML14118A293) 8.

END letter, PNP 20 14-035, Revised Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated April 2, 2014 9.

END letter, PNP 2014-050, Response to Request for Additional Information License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors, dated May 7, 2014

10. NRC electronic mail of May 21, 2014, Requests for Additional Information PRA

- Palisades NFPA 805 LAR

- MF0382 (ADAMS Accession Number ML14142A104)

Dear Sir or Madam:

In Reference 1, Entergy Nuclear Operations, Inc. (END) submitted a license amendment request to adopt the NFPA 805 performance-based standard for fire protection for light water reactors. In Reference 2, ENO responded to a clarification request. In Reference 3, END received electronic mail Request for Additional Information (RAls).

In Reference 4, ENO submitted the 60-day RAI responses.

In Reference 5, END submitted the revised 90-day RAI responses.

In Reference 6, END submitted the 120-day RAI responses. In Reference 7, END received electronic mail RAts on Fire Modeling.

In Reference 8, END submitted the revised response to RAI SSA 07.

In Reference 9, END submitted responses to the Fire Modeling RAls.

In Reference 10, END received electronic mail RAts on Fire PRA. Per discussion with the NRC, the RAI response schedule for the RAls in Reference 10 is as follows:

PRA RAls due in 30 days (no later than June 20, 2014):

PRA 01.e.01, PRA 01.f.01, PRA 01.h.01, PRA 01.h.02, PRA 01.k.01, PRA 01.mm.01, PRA 01.q.01, PRA 01.r.01, PRA 01.y.01, PRA 12.01, PRA 31 PNP 2014-063 Page 2 of 3

6. ENO letter, PNP 2013-083, "Response to Request for Additional Information - License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors", dated December 2, 2013 (ADAMS Accession Number ML13336A649)

7. NRC electronic mail of March 11,2014, "Requests for Additional Information - Palisades - NFPA 805 Project LAR - MF0382" (ADAMS Accession Number ML14118A293)

8. ENO letter, PNP 2014-035, "Revised Response to Request for Additional Information - License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors", dated April 2, 2014

9. ENO letter, PNP 2014-050, "Response to Request for Additional Information - License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors", dated May 7,2014

10. NRC electronic mail of May 21, 2014, "Requests for Additional Information - PRA - Palisades - NFPA 805 LAR - MF0382" (ADAMS Accession Number ML14142A104)

Dear Sir or Madam:

In Reference 1, Entergy Nuclear Operations, Inc. (ENO) submitted a license amendment request to adopt the NFPA 805 performance-based standard for fire protection for light water reactors. In Reference 2, ENO responded to a clarification request. In Reference 3, ENO received electronic mail Request for Additional Information (RAls). In Reference 4, ENO submitted the 60-day RAI responses. In Reference 5, ENO submitted the revised 90-day RAI responses. In Reference 6, ENO submitted the 120-day RAI responses. In Reference 7, ENO received electronic mail RAls on Fire Modeling. In Reference 8, ENO submitted the revised response to RAI SSA 07. In Reference 9, END submitted responses to the Fire Modeling RAls. In Reference 10, ENO received electronic mail RAls on Fire PRA. Per discussion with the NRC, the RAI response schedule for the RAls in Reference 10 is as follows:

PRA RAls due in 30 days (no later than June 20, 2014):

PRA 01.e.01, PRA 01.f.01, PRA 01.h.01, PRA 01.h.02, PRA 01.k.01,

PRA 01.mm.01, PRA 01.q.01, PRA 01.r.01, PRA 01.y.01, PRA 12.01, PRA 31

PNP 2014-063 Page 3 of 3 PRA RAIs due in 90 days (no later than August 19, 2014):

PRAO1.j.01, PRAO1.LO1, PRA 17.b.01, PRA2O.01, PRA23.01, PRA 23.a.01, PRA 23.c.01, PRA 28.a.01, PRA 30 In Attachment 1, ENO is providing 30-day responses to the RAIs noted above.

A copy of this response has been provided to the designated representative of the State of Michigan.

This letter contains no new commitments and no revisions to existing commitments.

I declare under penalty of perjury that the foregoing is true and correct. Executed on June 17, 2014.

Sincerely,

Attachment:

1. Response to Request for Additional Information Regarding License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors cc:

Administrator, Region Ill, USNRC Project Manager, Palisades, USN RC Resident Inspector, Palisades, USNRC State of Michigan ajv/jpm PNP 2014-063 Page 3 of 3 PRA RAls due in 90 days (no later than August 19, 2014):

PRA 01.j.01, PRA 01.1.01, PRA 17.b.01, PRA 20.01, PRA 23.01, PRA 23.a.01, PRA 23.c.01, PRA 28.a.01, PRA 30 In Attachment 1, END is providing 30-day responses to the RAls noted above.

A copy of this response has been provided to the designated representative of the State of Michigan.

This letter contains no new commitments and no revisions to existing commitments.

I declare under penalty of perjury that the foregoing is true and correct. Executed on June 17, 2014.

Sincerely, ajv/jpm

Attachment:

1. Response to Request for Additional Information Regarding License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactors cc:

Administrator, Region III, USNRC Project Manager, Palisades, USNRC Resident Inspector, Palisades, USNRC State of Michigan

ATTACHMENT 1 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REGARDING LICENSE AMENDMENT REQUEST TO ADOPT NFPA 805 PERFORMANCE-BASED STANDARD FOR FIRE PROTECTION FOR LIGHT WATER REACTORS NRC REQUEST PRA RAI O1.e.O1 The response to PRA RAI 01.e, in the letter dated December 2, 2013, Agencywide Documents Access and Management System (ADAMS) Accession No. ML13336A649, stated that the primary coolant pump (PCP) seal failure model used the methodology presented in WCAP-15749-P, Revision 1, Guidance for the Implementation of the Combustion Engineering Owners Group (CEOG) Model for Failure of Reactor Coolant Pump Seals Given Loss of Seal Cooling (Task 2083), December 2008. This topical has not been endorsed by the NRC.

Describe whether the PCP seal failure is the same for both the compliant and the post-transition PRA models such that the impact of this model on the change in risk estimates is minimal. If the PCP seal model differs between the compliant and post-transition PRA models, or if the model has a substantive impact on the change in risk estimates, provide a summary of the method and the quantitative results that are used in the PRA.

ENO RESPONSE The primary coolant pump seal failure model is based on the topical report generated by the owners group and endorsed by the NRC (WCAP-16175-P-A).

As part of a model update the revised topical report WCAP-15749-P, was reviewed for impact on the implementation of the seal model. WCAP-1 5749-P provides guidance on implementation of the seal model as developed per WCAP-1 6175-P-A. The review of WCAP-15749-P documented that no changes to the existing seal model were required and none were made.

Therefore, the existing seal model remains consistent with the consensus model as endorsed by the NRC as documented in WCAP-16175-P-A.

The seal model incorporated into the PRA model consists of two principal elements.

The first element is development and incorporation of seal failure probabilities into the PRA model. The second element includes the plant specific elements with respect to maintaining seal cooling, instrument and control related to primary coolant pump operation and the human error probability for failure to trip the primary coolant pumps.

Page 1 of 18 ATTACHMENT 1 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REGARDING LICENSE AMENDMENT REQUEST TO ADOPT NFPA 805 PERFORMANCE-BASED STANDARD FOR FIRE PROTECTION FOR LIGHT WATER REACTORS NRC REQUEST PRA RAI01.e.01 The response to PRA RAJ 01.e, in the letter dated December 2, 2013, Agencywide Documents Access and Management System (ADAMS) Accession No. ML13336A649, stated that the primary coolant pump (PCP) seal failure model used the methodology presented in WCAP-15749-P, Revision 1, IIGuidance for the Implementation of the Combustion Engineering Owners Group (CEOG) Model for Failure of Reactor Coolant Pump Seals Given Loss of Seal Cooling (Task 2083)': December 2008. This topical has not been endorsed by the NRC.

Describe whether the PCP seal failure is the same for both the compliant and the post-transition PRA models such that the impact of this model on the change in risk estimates is minimal. If the PCP seal model differs between the compliant and post-transition PRA models, or if the model has a substantive impact on the change in risk estimates, provide a summary of the method and the quantitative results that are used in the PRA.

ENO RESPONSE The primary coolant pump seal failure model is based on the topical report generated by the owners group and endorsed by the NRC (WCAP-16175-P-A).

As part of a model update the revised topical report WCAP-157 49-P, was reviewed for impact on the implementation of the seal model. WCAP-15749-P provides guidance on implementation of the seal model as developed per WCAP-16175-P-A. The review of WCAP-15749-P documented that no changes to the existing seal model were required and none were made.

Therefore, the existing seal model remains consistent with the consensus model as endorsed by the NRC as documented in WCAP-16175-P-A.

The seal model incorporated into the PRA model consists of two principal elements.

The first element is development and incorporation of seal failure probabilities into the PRA model. The second element includes the plant specific elements with respect to maintaining seal cooling, instrument and control related to primary coolant pump operation and the human error probability for failure to trip the primary coolant pumps.

Page 1 of 18

The seal failure probabilities were developed per and remain consistent with the criteria of WCAP-1 6175-P-A. The probability of seal failure based on the seal model is the same for both the compliant and post-transition plant. The probability of seal failure was not altered in the post transition plant results.

The probability of failure of support systems required for seal cooling and instrument and control necessary to trip the pumps is a plant specific input to the PRA model logic and is not governed by the consensus model. This element of the model is based on plant specific features with one exception. The human error probability for tripping the primary coolant pumps is based on the time available to accomplish the action defined by WCAP-1 6175-P-A.

Modification S2-5 (Provide Alternate Method of Tripping Primary Coolant Pumps during Fire Event) as described in Attachment S Table S-2 of the original PNP LAR is being implemented as part of transition to NFPA 805. This modification will provide an alternate capability to trip the primary coolant pumps from the control room.

Implementation of the modification impacts the plant specific inputs to the seal model.

Therefore, the difference between the variant and post-transition plant in the PRA model with respect to primary coolant pump seals is in the instrument and control logic associated with pump operation. The variant plant represents the existing plant (no modification). The post transition plant model includes the alternative capability to trip the pumps from the control room. The post-transition plant is compliant with respect to the requirement to ensure primary coolant pumps can be tripped from the control room following a fire. Consequently there is no difference between the compliant and post transition plant.

The modification reduces the risk associated with the existing pump control circuits which may preclude the ability to trip the pumps due to fire affects. Logic associated with the proposed modification is the only difference between the variant and post-transition plant with respect to the pump seal model.

A summary of the method and the quantitative results that are used in the PRA are not required because the difference in the seal model is:

in the plant specific element of the model, related to a modification to improve plant capability, and NOT related to the probability that the seal will fail on loss of cooling

REFERENCES:

1. WCAP-1 6175-P-A (Formerly CE NPSD 1199 P, Revision 1), Model for Failure of RCP Seals Given Loss of Seal Cooling in CE NSSS Plants, March 2007.

2. WCAP-1 5749-P, Guidance for the Implementation of the CEOG Model for Failure of RCP Seals Given Loss of Seal Cooling (Task 2083), Revision 1, December 2008.

Page 2 of 18 The seal failure probabilities were developed per and remain consistent with the criteria of WCAP-16175-P-A. The probability of seal failure based on the seal model is the same for both the compliant and post-transition plant. The probability of seal failure was not altered in the post transition plant results.

The probability of failure of support systems required for seal cooling and instrument and control necessary to trip the pumps is a plant specific input to the PRA model logic and is not governed by the consensus model. This element of the model is based on plant specific features with one exception. The human error probability for tripping the primary coolant pumps is based on the time available to accomplish the action defined by WCAP-16175-P-A.

Modification S2-5 (Provide Alternate Method of Tripping Primary Coolant Pumps during Fire Event) as described in Attachment STable S-2 of the original PNP LAR is being implemented as part of transition to NFPA 805. This modification will provide an alternate capability to trip the primary coolant pumps from the control room.

Implementation of the modification impacts the plant specific inputs to the seal model.

Therefore, the difference between the variant and post-transition plant in the PRA model with respect to primary coolant pump seals is in the instrument and control logic associated with pump operation. The variant plant represents the existing plant (no modification). The post transition plant model includes the altemative capability to trip the pumps from the control room. The post-transition plant is compliant with respect to the requirement to ensure primary coolant pumps can be tripped from the control room following a fire. Consequently there is no difference between the 'compliant' and 'post-transition' plant.

The modification reduces the risk associated with the existing pump control circuits which may preclude the ability to trip the pumps due to fire affects. Logic associated with the proposed modification is the only difference between the variant and post-transition plant with respect to the pump seal model.

A summary of the method and the quantitative results that are used in the PRA are not required because the difference in the seal model is:

in the plant specific element of the model, related to a modification to improve plant capability, and NOT related to the probability that the seal will fail on loss of cooling

REFERENCES:

1. WCAP-16175-P-A (Formerly CE NPSD 1199 P, Revision 1), Model for Failure of RCP Seals Given Loss of Seal Cooling in CE NSSS Plants, March 2007.

2. WCAP-15749-P, Guidance for the Implementation of the CEOG Model for Failure of RCP Seals Given Loss of Seal Cooling (Task 2083), Revision 1, December 2008.

Page 2 of 18

NRC REQUEST PRA RAIO1.f.O1 The response to PRA RAI 01.f in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649 indicates that the circuit analysis of identified instrumentation for dominant operator actions has been completed and will be incorporated into the transition fire PRA risk results, which is to be provided in response to PRA RAI 30.

a. Discuss what is meant by dominant relative to AG 1.200s, An Approach For Determining The Technical Adequacy Of Probabilistic Risk Assessment Results For Risk-Informed Activities, definition of a significant basic event and whether these non-dominant actions are assumed to be failed in the fire PRA.

b.

If not assumed to be failed, justify this treatment by discussing the risk significance of the credited non-dominant operator actions on the transition risk results.

ENO RESPONSE a.

Dominant operator actions in the context of the discussion provided in the original response to 01.f was related to a set of operator actions which would be required to be maintained as detailed human error probabilities (HEPs) to offset increases in core damage frequency (ODE) resulting from the assignment of screening or scoping HEPs to other human failure events (HEEs).

In addition, the discussion does not mean that other (non dominant) actions did not already have instrumentation supporting the operator action included in the model. The discussion was only meant to convey that some actions in the dominant set did not have instrumentation available at that time.

b.

It is not the case that all non dominant operator actions are assumed to be failed in the fire PRA. The group of non-dominant operator actions includes two subsets comprised of HFEs assigned either scoping or screening values. HFEs assigned a screening value (1.0), are assumed failed in the fire PRA. Events assigned scoping values are analyzed in the same manner as the dominant HFEs to the extent that instrumentation is included in the model, fire induced impacts are considered; access to the area where the action is to be completed is required, operator ability to complete the action is required and instrumentation availability impacts are considered.

Scoping HFEs without supporting instrumentation included in the model or those for which the fire fails the instrumentation would be failed in the fire PRA. Revised risk results reflecting the implementation of the above process for incorporation of operator actions will be provided in response to RAI 30.

Page 3 of 18 NRC REQUEST PRA RAI01.f.01 The response to PRA RAI 01.f in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649 indicates that the circuit analysis of identified instrumentation for Iidominant" operator actions has been completed and will be incorporated into the transition fire PRA risk results, which is to be provided in response to PRA RAI 30.

a. Discuss what is meant by Iidominant" relative to RG 1.200's, I~n Approach For Determining The Technical Adequacy Of Probabilistic Risk Assessment Results For Risk-Informed Activities", definition of a significant basic event and whether these non-11dominant" actions are assumed to be failed in the fire PRA.

b. If not assumed to be failed, justify this treatment by discussing the risk significance of the credited non-dominant operator actions on the transition risk results.

ENO RESPONSE

a. 'Dominant' operator actions in the context of the discussion provided in the original response to 01.f was related to a set of operator actions which would be required to be maintained as detailed human error probabilities (HEPs) to offset increases in core damage frequency (CDF) resulting from the assignment of screening or scoping HEPs to other human failure events (HFEs). In addition, the discussion does not mean that other (non 'dominant') actions did not already have instrumentation supporting the operator action included in the model. The discussion was only meant to convey that some actions in the dominant set did not have instrumentation available at that time.

b. It is not the case that all non 'dominant' operator actions are assumed to be failed in the fire PRA. The group of non-'dominant' operator actions includes two subsets comprised of HFEs assigned either scoping or screening values. HFEs assigned a screening value (1.0), are assumed failed in the fire PRA. Events assigned scoping values are analyzed in the same manner as the 'dominant' HFEs to the extent that instrumentation is included in the model, fire induced impacts are considered; access to the area where the action is to be completed is required, operator ability to complete the action is required and instrumentation availability impacts are considered. Scoping HFEs without supporting instrumentation included in the model or those for which the fire fails the instrumentation would be failed in the fire PRA. Revised risk results reflecting the implementation of the above process for incorporation of operator actions will be provided in response to RAI 30.

Page 3 of 18

NRC REQUEST PRA RAI O1.h.O1 In the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, the response to PRA RAIO1.h, subsection 3) Justifications forAssumptions Identified as Non-Conseivative in the licensees analysis describes that the treatment of location in the dependency analysis differs from the guidance in NUREG-1921, EPRI/NRC-RES Fire Human Reliability Analysis Guidelines, Draft Report for Comment NUREG-1921 guidance does not negate the possibility of success of all subsequent actions after failure of an action in the main control room as stated in the RAI response but does state that there would be high dependence between all actions. Simply stating that the approach is not realistic is not sufficientjustification to deviate from the NUREG. It also appears that the timing decision branch of Figure 6-1 of NUREG-1921 is not utilized by the dependency analysis for sequential actions due to this deviation.

Provide a time and distance justification for each set of control room actions considered to be in different locations or conform to the accepted method. Identify the final approach used in the response to PRA RAI 30.

ENO RESPONSE Palisades Nuclear Plant (PNP) will follow the NUREG-1 921 guidance and treat all actions taken in the control room as taking place within a single (same) location. The impact of these changes will be reflected in the quantification results documented in response to PRA RAI 30.

NRC REQUEST PRA RAI O1.h.02 The dependency analysis described in response to PRA RAI 01.h does not indicate that a minimum value was utilized for the joint probability of multiple human failure events (HFE) and the response. The statement, e.g., for zero dependence, the conditional human error probabilities (HEP) is equal to the independent HEP implies thatjoint HEPs may take on any value. Section 6.2 of NUREG 1921 addresses the need to consider a minimum (floor) value for the joint probability of multiple HFEs. Each value less than the floor value should be individuallyjustified.

Considering this guidance, describe andjustify thatjoint HEP values that appear in fire PRA cutsets including any values less than the floor value, If a HEP floor for cutsets was not used consistent with NUREG-1921 (i.e., 1 E-5 with justifications for lower values), provide updated risk results as part of the aggregate change-in-risk analysis requested in PRA RAI 30, which is consistent with NUREG-1921 guidance.

Page 4 of 18 NRC REQUEST PRA RAI01.h.01 In the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, the response to PRA RAI 01.h, subsection 3) 'ljustifications for Assumptions Identified as Non-Conservative in the licensee's analysis" describes that the treatment of location in the dependency analysis differs from the guidance in NUREG-1921, "EPRIINRC-RES Fire Human Reliability Analysis Guidelines, Draft Report for Comment". NUREG-1921 guidance does not "negate the possibility of success of all subsequent actions" after failure of an action in the main control room as stated in the RAI response but does state that there would be high dependence between all actions. Simply stating that the approach is not realistic is not sufficient justification to deviate from the NUREG. It also appears that the timing decision branch of Figure 6-1 of NUREG-1921 is not utilized by the dependency analysis for sequential actions due to this deviation.

Provide a time and distance justification for each set of control room actions considered to be in different locations or conform to the accepted method. Identify the final approach used in the response to PRA RAI 30.

ENO RESPONSE Palisades Nuclear Plant (PNP) will follow the NUREG-1921 guidance and treat all actions taken in the control room as taking place within a single (same) location. The impact of these changes will be reflected in the quantification results documented in response to PRA RAI 30.

NRC REQUEST PRA RAI01.h.02 The dependency analysis described in response to PRA RAI 01.h does not indicate that a minimum value was utilized for the joint probability of multiple human failure events (HFE) and the response. The statement, "e.g., for zero dependence, the conditional human error probabilities (HEP) is equal to the independent HEP" implies that joint HEPs may take on any value. Section 6.2 of NUREG 1921 addresses the need to consider a minimum ("f1oor'? value for the joint probability of multiple HFEs. Each value less than the floor value should be individually justified.

Considering this guidance, describe and justify that joint HEP values that appear in fire PRA cutsets including any values less than the floor value. If a HEP floor for cutsets was not used consistent with NUREG-1921 (i.e., 1E-5 with justifications for lower values), provide updated risk results as part of the aggregate change-in-risk analysis requested in PRA RA130, which is consistent with NUREG-1921 guidance.

Page 4 of 18

ENO RESPONSE PNP will follow the guidance of NUREG-1 921 and utilize a floor value of 1 E-5 for all conditional joint HEPs. The impact of these changes will be reflected in the quantification results documented in response to PRA RAI 30.

NRC REQUEST PRA RAIOLk.O1 The response to PRA RAI 01.k, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, indicates that main control room (MCR) abandonment is only postulated for those fires resulting in a loss of MCR habitability; however, the response to PRA RAI 03, in the letter mentioned above, states that the RAI Response Fire PRA Model will include additional scenarios that model MCR abandonment due to equipment damage, with control being transferred to other locations, such as the alternate shutdown panel If the intent is to credit MCR abandonment due to loss of control, provide a description of the method and its technicaljustification. Include an explanation of the supporting analysis, work performed, and process followed in the technicaljustification.

ENO RESPONSE The response to PRA RAI 01.k was intended to indicate that control room abandonment due to loss of control or function is not explicitly modeled in the Fire PRA. That is, specific identification of those fire events which lead to loss of control or function is not part of the fire scenario development and initial quantification process. Only scenarios that result in control room abandonment due to loss of habitability are explicitly identified as control room abandonment scenarios.

However, the Fire PRA model does include credit for operator deployment for local actions (including local actions at the alternate shutdown panel) as potential success paths in the accident sequence development. Use of these alternate success paths is not limited to control room abandonment scenarios due to loss of habitability.

The response to PRA RAI 03 for FSS-B1-01 was intended to indicate that additional control room scenarios are being added to the RAI Response Fire PRA model. These additional scenarios also credit operator deployment for local actions including local actions at the alternate shutdown panel. The intent is not to explicitly identify and credit control room abandonment due to loss of control.

Page 5 of 18 ENO RESPONSE PNP will follow the guidance of NUREG-1921 and utilize a floor value of 1 E-5 for all conditional joint HEPs. The impact of these changes will be reflected in the quantification results documented in response to PRA RAI 30.

NRC REQUEST PRA RAI01.k.01 The response to PRA RAI 01.k, in the letter dated December 2,2013, ADAMS Accession No. ML13336A649, indicates that main control room (MeR) abandonment is only postulated for those fires resulting in a loss of MeR habitability; however, the response to PRA RAI 03, in the letter mentioned above, states that lithe RAI Response Fire PRA Model will include additional scenarios that model MeR abandonment due to equipment damage, with control being transferred to other locations, such as the alternate shutdown panel".

If the intent is to credit MeR abandonment due to loss of control, provide a description of the method and its technical justification. Include an explanation of the supporting analysis, work performed, and process followed in the technical justification.

ENO RESPONSE The response to PRA RAI 01.k was intended to indicate that control room abandonment due to loss of control or function is not explicitly modeled in the Fire PRA. That is, specific identification of those fire events which lead to loss of control or function is not part of the fire scenario development and initial quantification process. Only scenarios that result in control room abandonment due to loss of habitability are explicitly identified as control room abandonment scenarios.

However, the Fire PRA model does include credit for operator deployment for local actions (including local actions at the alternate shutdown panel) as potential success paths in the accident sequence development. Use of these alternate success paths is not limited to control room abandonment scenarios due to loss of habitability.

The response to PRA RAI 03 for FSS-B1-01 was intended to indicate that additional control room scenarios are being added to the RAI Response Fire PRA model. These additional scenarios also credit operator deployment for local actions including local actions at the alternate shutdown panel. The intent is not to explicitly identify and credit control room abandonment due to loss of control.

Page 5 of 18

NRC REQUEST PRA RAIO1.mm.O1 The response to PRA RAI 01.mm, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, indicates that key assumptions and sources of uncertainty were identified. Provide a table that describes these key assumptions and sources of uncertainty that assesses their impact on the NFPA 805 application.

ENO RESPONSE In the development of each Fire PRA report, a section was included that identified assumptions related to each of the associated Fire PRA tasks included in that specific notebook. For each of the identified assumptions, a qualitative assessment was documented regarding the potential quantitative impact as it applies to the base fire PRA model which serves as part of the characterization of the assumptions. In the PNP Fire PRA Quantification and Summary Notebook [1], these assumptions were reviewed to develop a table that identified sources of uncertainty by each NUREG/CR-6850 task and assessed the sensitivity of their impact on the NFPA 805 application. A modified version of this table is provided below, It has been updated to account for the status of the RAI Response Fire PRA model and updated to specifically identify the potential key assumptions associated with the sources of uncertainty.

Page 6 of 18 NRC REQUEST PRA RAI01.mm.01 The response to PRA RAI 01.mm, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, indicates that key assumptions and sources of uncertainty were identified. Provide a table that describes these key assumptions and sources of uncertainty that assesses their impact on the NFPA 805 application.

ENO RESPONSE In the development of each Fire PRA report, a section was included that identified assumptions related to each of the associated Fire PRA tasks included in that specific notebook. For each of the identified assumptions, a qualitative assessment was documented regarding the potential quantitative impact as it applies to the base fire PRA model which serves as part of the characterization of the assumptions. In the PNP Fire PRA Quantification and Summary Notebook [1], these assumptions were reviewed to develop a table that identified sources of uncertainty by each NUREG/CR-6850 task and assessed the sensitivity of their impact on the NFPA 805 application. A modified version of this table is provided below. It has been updated to account for the status of the RAI Response Fire PRA model and updated to specifically identify the potential key assumptions associated with the sources of uncertainty.

Page 6 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK ASSUMPTIONS RESULTS TO THE NO.

TASK TITLE DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Plant Boundary The fire PRA analysis This task posed a limited During scenario development, D

d boundary was opportunity for the the zone of influence was not e ml ion an determined, and the plant identification of potentially limited to the physical analysis Partitioning was partitioned into key assumptions and related unit boundary for most discrete physical analysis sources of uncertainty compartment scenarios.

If the units (PAUs) based on beyond the credit taken for zone of influence included the physical the physical presence of targets in adjacent fire characteristics of the boundaries and partitions.

areas/zones, these targets were various areas.

also included, regardless of their fire area/zone location. In addition, a multi-compartment analysis further reduced uncertainty by addressing the potential impact of failure of partition elements on quantification.

2 Fire PRA The fire PRA components This task posed perhaps the The potential for uncertainty was C

were selected by highest potential for error if reduced as a result of multiple omponen reviewing the not uncertainty. The overlapping tasks including the Selection components in the FPIE mapping of basic events to MSO expert panel process PRA model and the components required not combined with reviews of equipment included in the only the consideration of screening initiating events, deterministic Nuclear failure modes (active versus screened containment Safety Capability passive) but an penetrations, and screened Assessment (NSCA) understanding of the ISLOCA scenarios. Additional analysis. The data were Appendix RJNSCA functions internal reviews and the change analyzed with respect to not previously considered evaluation process provided the their suitability to be risk significant in the FPIE opportunity to further reduce included in the fire PRA model.

uncertainty in this task.

model. Additional considerations, including the potential effects of Multiple Spurious Operations (MSO5), were used to evaluate the need to include additional cornponents.

Page 7 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESUL TS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY 1

Plant Boundary The fire PRA analysis This task posed a limited During scenario development, Definition and boundary was opportunity for the the zone of influence was not determined, and the plant identification of potentially limited to the physical analysis Partitioning was partitioned into key assumptions and related unit boundary for most discrete physical analysis sources of uncertainty compartment scenarios. If the units (PAUs) based on beyond the credit taken for zone of influence included the physical the physical presence of targets in adjacent fire characteristics of the boundaries and partitions.

areas/zones, these targets were various areas.

also included, regardless of their fire area/zone location. In addition, a multi-compartment analysis further reduced uncertainty by addressing the potential impact of failure of partition elements on quantification.

2 Fire PRA The fire PRA components This task posed perhaps the The potential for uncertainty was Component were selected by highest potential for error if reduced as a result of multiple reviewing the not uncertainty. The over1apping tasks including the Selection components in the FPIE mapping of basic events to MSO expert panel process PRA model and the components required not combined with reviews of equipment included in the only the consideration of screening initiating events, deterministic Nuclear failure modes (active versus screened containment Safety Capability passive) but an penetrations, and screened Assessment (NSCA) understanding of the ISLOCA scenarios. Additional analYSis. The data were Appendix R1NSCA functions internal reviews and the change analyzed with respect to not previously considered evaluation process provided the their suitability to be risk significant in the FPIE opportunity to further reduce included in the fire PRA model.

uncertainty in this task.

model. Additional considerations, including the potential effects of Multiple Spurious Operations (MSOs), were used to evaluate the need to include additional components.

Page 7 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK K

TIE TASK ASSUMPTIONS RESULTS TO THE NO.

TAS TI DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Fire PRA Cable Cables were assigned to No treatment of uncertainty is The cable selection approach S

I the components based typically required for this task was based on the failure fault e ec IOfl on existing Fire Safe beyond the understanding of consequences identified for each Shutdown cable the cable selection approach cable relative to the operation of evaluations and for the various iterations of the associated component.

additional cable cable identification tasks.

These fault consequences were identification.. Tasks 2 Additionally, PRA credited identified in the original Appendix and 3 were performed components for which cable R data. A seperate effort was iteratively with the Plant routing information was not performed to review this data in Fire Induced Risk Model provided (credit by exclusion) light of current practices to (Task 5).

represents a potential key assure its fidelity. Since assumption and source of Palisades has undergone an uncertainty. Recognizing extensive effort to identify cables that the potential exists to for components beyond those improperly credit these addressed in Appendix R, components where their uncertainty associated with cables are located (non-unknown cable locations (UNL conservative), it can be components) has been greatly assumed that these reduced.

In order to eliminate components are failed excessive conservatism, UNL unnecessarily (conservative),

components were credited by exclusion either explicitly or based on assumed cable routing.

In any event, the assumed cable routing is identified as a potential key source of uncertainty.

Qualitative A small number of plant Structures from the global No structure with credited PRA S

areas met all of the analysis boundary, and components was excluded. This creening criteria necessary for ignition sources deemed to exclusion criterion is not subject qualitative screening.

have no impact on the FPRA, to uncertainty. In the event that were excluded from the a structure which could lead to a quantification based on plant trip was excluded qualitative screening criteria.

incorrectly, its contribution to The only assumptions CDF would be small (with a subject to uncertainty are the CCDP commensurate with base judgments regarding the risk) and would likely be more potential for plant trip used than offset by inclusion of the as part of the screening additional ignition sources and process.

the subsequent reduction of other scenario frequencies. A similar argument can be made for ignition sources for which scenario refinement was deemed unnecessary.

Page 8 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESULTS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY 3

Fire PRA Cable Cables were assigned to No treatment of uncertainty is The cable selection approach Selection the components based typically required for this task was based on the failure fault on existing Fire Safe beyond the understanding of consequences identified for each Shutdown cable the cable selection approach cable relative to the operation of evaluations and for the various iterations of the associated component.

additional cable cable identification tasks.

These fault consequences were identification.. Tasks 2 Additionally, PRA credited identified in the original Appendix and 3 were performed components for which cable R data. A seperate effort was iteratively with the Plant routing information was not performed to review this data in Fire Induced Risk Model provided (credit by exclusion) light of current practices to (Task 5).

represents a potential key assure its fidelity. Since assumption and source of Palisades has undergone an uncertainty. Recognizing extensive effort to identify cables that the potential exists to for components beyond those improperly credit these addressed in Appendix R, components where their uncertainty associated with cables are located (non-unknown cable locations (UNL conservative), it can be components) has been greatly assumed that these reduced. In order to eliminate components are failed excessive conservatism, UNL unnecessarily (conservative). components were credited by exclusion - either explicitly or based on assumed cable routing.

In any event, the assumed cable routing is identified as a potential key source of uncertainty.

4 Qualitative A small number of plant Structures from the global No structure with credited PRA Screening areas met all of the analysis boundary, and components was excluded. This criteria necessary for ignition sources deemed to exclusion criterion is not subject qualitative screening.

have no impact on the FPRA, to uncertainty. In the event that were excluded from the a structure which could lead to a quantification based on plant trip was excluded qualitative screening criteria.

incorrectly, its contribution to The only assumptions CDF would be small (with a subject to uncertainty are the CCDP commensurate with base judgments regarding the risk) and would likely be more potential for plant trip used than offset by inclusion of the as part of the screening additional ignition sources and process.

the subsequent reduction of other scenario frequencies. A similar argument can be made for ignition sources for which scenario refinement was deemed unnecessary.

Page 8 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX The PNP fire PRA model was developed using applicable portions of the FPIE model. The model was expanded as necessary to include additional sequences associated with fire events. Cables were linked with basic events in the model and associated to plant locations allowing evaluation of fire-induced circuit failures on a per scenario basis.

The construction of the FPRA plant response model itself is a source of uncertainty. The same sources of uncertainty/sensitivity that are applicable to the base model are applicable to the FPRA. However, these are judged to be minor in the context of the overall Fire PRA model development process in the context of the NFPA 805 application.

Some 9,000+ failure modes (random and fire) are included in the FPRA plant response model. This includes a highly detailed representation of potential failures (e.g., down to the contact pair level) and fully developed common cause failure modeling. Several thousand cables are mapped to the associated basic events.

The bookkeeping challenge of managing this amount of data introduces potential error.

FPIE and FPRA peer reviews (including the F&0 resolution process and the subsequent RAI resolution process), internal assessments, and the change evaluation process are useful in exercising the model and identifying weaknesses. In addition, the FPRA model changes are incorporated into the FPIE model. This assures that these sequences are exercised and reviewed continually not just for fire PRA applications.

The potential for managing this amount of data was addressed by employing different industry codes that were used to validate the quantified results. By employing different codes, problems with input are better captured as each code provides different reports, different diagnostic capabilities, etc.

The detailed modeling employed in the Palisades analyses ensures better rigor, insights, and reduces errors, and reduces the epistemic uncertainty.

Moreover, such detailed modeling results in conservative numerical results as failures are double counted; however, this increases the aleatory uncertainty.

It is considered that the importance of reducing the epistemic uncertainty at the expense of increasing the aleatory uncertainty greatly benefits the development of additional risk insights.

5 POTENTIAL KEY SENSITIVITY OF THE TASK TASK ASSUMPTIONS RESULTS TO THE NO.

TASK TITLE DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Plant Fire Induced Risk Model Page 9 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESUL TS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY 5

Plant Fire The PNP fire PRA model The construction of the FPIE and FPRA peer reviews Induced Risk was developed using FPRA plant response model (including the F&O resolution applicable portions of the itself is a source of process and the subsequent RAI Model FPIE model. The model uncertainty. The same resolution process), intemal was expanded as sources of assessments, and the change necessary to include uncertainty/sensitivity that evaluation process are useful in additional sequences are applicable to the base exercising the model and associated with fire model are applicable to the identifying weaknesses. In events. Cables were FPRA. However, these are addition, the FPRA model linked with basic events judged to be minor in the changes are incorporated into in the model and context of the overall Fire the FPIE mode/. This assures associated to plant PRA model development that these sequences are locations allowing process in the context of the exercised and reviewed evaluation of fire-induced NFPA 805 application.

continually - not just for fire PRA circuit failures on a per Some 9,000+ failure modes applications.

scenario basis.

(random and fire) are The potential for managing this included in the FPRA plant amount of data was addressed response mode/. This by employing different industry includes a highly detailed codes that were used to validate representation of potential the quantified results. By failures (e.g., down to the employing different codes, contact pair level) and fully problems with input are better developed common cause captured as each code provides failure modeling. Several different reports, different thousand cables are mapped diagnostic capabilities, etc.

to the associated basic The detailed modeling employed events.

in the Palisades analyses The bookkeeping challenge ensures better rigor, insights, of managing this amount of and reduces errors, and reduces data introduces potential the epistemic uncertainty.

error.

Moreover, such detailed modeling results in conservative numerical results as failures are double counted; however, this increases the aleatory uncertainty. It is considered that the importance of reducing the epistemic uncertainty at the expense of increasing the aleatory uncertainty greatly benefits the development of additional risk insights.

Page 9 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK ASSUMPTIONS RESULTS TO THE TASK TITLE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY 6

Fire Ignition A fire ignition frequency The frequency values from A Bayesian update process for was estimated for each NUREG/CR-6850 and EPRI PNP events after 2000 was Frequency plant compartment based Report 1016735 include applied to the generic on fixed sources and uncertainty to account for frequencies taken from transient factors. The variability among plants NUREG/CR-6850 and the EPRI frequencies were along with some significant 1016735 data.

ultimately applied on a conservatism in defining the scenario basis. The frequencies, and their The applicabilIty of the ignition approtionment of the fire associated heat release frequency data is identified as a frequency was done in rates, based on limited potential key source of uncertainty.

accordance with detailed data.

NUREG/CR-6850 guidance and associated A potential key assumption is FAQ5.

that the fire ignition frequency data is applicable and provides an accepted estimate of the fire frequency for PNP.

Quantitative An initial quantification of Other than the conservative Quantitative screening was the fire PRA model was treatment asscoiated with limited to refraining from further Screening performed to identify the retaining all scenarios, there scenario refinement of those relative risk contribution is no uncertainty from this scenarios with a resulting CDF /

of each physical analysis task on the FPRA results.

LERF below the screening unit (PAU). No actual threshold. All of the results were screening was performed retained in the cumulative CDF /

as all PAUs were LERF.

retained in the quantification. This step was used to identify compartments where detailed analyses would be appropriate.

8 Scoping Fire Scoping fire modeling is a This task by itself does not The employment of generic fire coarse approach used to contribute to uncertainty, modeling solutions did not Modeling bound the fire effects of However, the approach taken introduce any significant certain ignition sources.

for this task included: 1) conservatism. Detailed fire A more refined approach, generic fire modeling modeling was performed on generic modeling, was treatments used in lieu of those scenarios which otherwise employed at PNP. A conservative scoping would have been notable risk detailed analysis was analysis techniques and 2) contributorsand applied where performed for typical limited detailed fire modeling the reduction in conservatism ignition sources based on performed to refine the was likely to have a measurable their physical properties scenarios developed using impact.

and prescribed heat the generic fire modeling release rates. This solutions. The primary The NUREG/CR-6850 heat analysis yielded a conservatism introduced by release rates introduce guideline for the this task is associated with significant conservatism given evaluation of fire damage the heat release rates the limited fire test data available effects for the various specified in NUREG/CR-to define the heat release rates ignition sources. This 6850.

and the associated fire enabled the development development timeline. However, of a basic scenario for alternative treatments are not many sources that could currently accepted.

be treated as bounding.

Page 10 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESUL TS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY 6

Fire Ignition A fire ignition frequency The frequency values from A Bayesian update process for Frequency was estimated for each NUREG/CR-6850 and EPRI PNP events after 2000 was plant compartment based Report 1016735 include applied to the generic on fixed sources and uncertainty to account for frequencies taken from transient factors. The variability among plants NUREG/CR-6850 and the EPRI frequencies were along with some significant 1016735 data.

ultimately applied on a conservatism in defining the The applicablilty of the ignition scenario basis. The frequencies, and their approtionment of the fire associated heat release frequency data is identified as a frequency was done in rates, based on limited potential key source of accordance with detailed data.

uncertainty.

NUREG/CR-6850 A potential key assumption is guidance and associated FAQs.

that the fire ignition frequency data is applicable and provides an* accepted estimate of the fire frequency for PNP.

7 Quantitative An initial quantification of Other than the conservative Quantitative screening was Screening the fire PRA model was treatment asscoiated with limited to refraining from further performed to identify the retaining all scenarios, there scenario refinement of those relative risk contribution is no uncertainty from this scenarios with a resulting CDF /

of each physical analysis task on the FPRA results.

LERF below the screening unit (PAU). No actual threshold. All of the results were screening was performed retained in the cumulative CDF /

as all PAUs were LERF.

retained in the quantification. This step was used to identify compartments where detailed analyses would be appropriate.

8 Scoping Fire Scoping fire modeling is a This task by itself does not The employment of generic fire Modeling coarse approach used to contribute to uncertainty.

modeling solutions did not bound the fire effects of However, the approach taken introduce any significant certain ignition sources.

for this task included: 1) conservatism. Detailed fire A more refined approach, generic fire modeling modeling was performed on generic modeling, was treatments used in lieu of those scenarios which otherwise employed at PNP. A conservative scoping would have been notable risk detailed analysis was analysis techniques and 2) contributorsand applied where performed for typical limited detailed fire modeling the reduction in conservatism ignition sources based on performed to refine the was likely to have a measurable their physical properties scenarios developed using impact.

and prescribed heat the generic fire modeling The NUREG/CR-6850 heat release rates. This solutions. The primary analYSis yielded a conservatism introduced by release rates introduce guideline for the this task is associated with Significant conservatism given evaluation of fire damage the heat release rates the limited fire test data available effects for the various specified in NUREGlCR-to define the heat release rates ignition sources. This 6850.

and the associated fire enabled the development development timeline. However, of a basic scenario for alternative treatments are not many sources that could currently accepted.

be treated as bounding.

Page 10 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK ASSUMPTIONS RESULTS TO THE NO.

TASK TITLE DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Detailed Circuit Circuit failures were Uncertainty considerations Circuit analysis was performed evaluated on a failure are limited to errors in circuit as part of the Fire Safe Failure Analysis mode basis using the failure analysis where a Shutdown / NSCA analysis and data provided in the cable was deemed incapable supplemental circuit selection original Appendix R of causing loss of a particular efforts. Refinements in the analysis and additional function credited in the application of the circuit analysis cable data selection FPRA. Similar to Task 2 results to the fire PRA were efforts. In many cases (with the exception of the performed on a case by case additional circuit reviews MSO process), this task has basis where the scenario risk were necessary to no associated uncertainty if quantification was large enough determine the specific performed correctly.

to warrant further analysis.

failure consequences of cables on individual equipment.

10 Circuit Failure Circuit failures based off The uncertainty associated Circuit failure mode likelihood M d L ih d

the failure mode were with the applied conditional analysis was generally limited to 0 e i,e I 00 evaluated in Task 9. In failure probabilities posed those components where Analysis some cases, additional competing considerations.

spurious operation could not be circuit failure likelihood On the one hand, a failure caused by the generation of a analysis was needed. If probability for spurious spurious signal. This approach applicable, failure operation could be applied limited the introduction of non-probabilities were applied based solely on cable scope conservative uncertainties.

to specific cable failure without consideration of less Additional refinement to this modes.

direct fire effects (e.g., a approach was performed on risk failure likelihood applied to significant scenarios. Given this the spurious operation of an treatment, the application of MOV without consideration of circuit failure probabilities is not the fire-induced generation of considered to be a potential key spurious signal to close or source of uncertainty.

open the MOV). On the other hand, a failure probability for spurious operation could be applied despite the absence of cables capable of causing spurious operation in that location.

Page 11 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESUL TS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY 9

Detailed Circuit Circuit failures were Uncertainty considerations Circuit analysis was performed Failure Analysis evaluated on a failure are limited to errors in circuit as part of the Fire Safe mode basis using the failure analysis where a Shutdown / NSCA analysis and data provided in the cable was deemed incapable supplemental circuit selection original Appendix R of causing loss of a particular efforts. Refinements in the analYSis and additional function credited in the application of the circuit analysis cable data selection FPRA. Similar to Task 2 results to the fire PRA were efforts. In many cases (with the exception of the performed on a case by case additional circuit reviews MSO process), this task has basis where the scenario risk were necessary to no associated uncertainty if quantification was large enough determine the specific performed correctly.

to warrant further analysis.

failure consequences of cables on individual equipment.

10 Circuit Failure Circuit failures based off The uncertainty associated Circuit failure mode likelihood Mode Likelihood the failure mode were with the applied conditional analysis was generally limited to evaluated in Task 9. In failure probabilities posed those components where Analysis some cases, additional competing considerations.

spurious operation could not be circuit failure likelihood On the one hand, a failure caused by the generation of a analysis was needed. If probability for spurious spurious signal. This approach applicable, failure operation could be applied limited the introduction of non-probabilities were applied based solely on cable scope conservative uncertainties.

to specific cable failure without consideration of less Additional refinement to this modes.

direct fire effects (e.g., a approach was performed on risk failure likelihood applied to significant scenarios. Given this the spurious operation of an treatment, the application of MOV without consideration of circuit failure probabilities is not the fire-induced generation of considered to be a potential key spurious signal to close or source of uncertainty.

open the MOV). On the other hand, a failure probability for spurious operation could be applied despite the absence of cables capable of causing spurious operation in that location.

Page 11 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK ASSUMPTIONS RESULTS TO THE TASK TITLE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Detailed Fire 1 1 Modeling The application of Utlimately, the treatment of Detailed fire modeling was detailed fire modeling these issues has evolved performed only on those was limited to the Main through the various RAIs and scenarios which otherwise would Control Room (MCR) subsequent model have been notable risk abandonment scenario, refinements to reduce the contributors and only where and a few risk significant number of potential key removal of conservatism in the areas (e.g., in the 1 C and assumptions.

generic fire modeling solution 1 D swithcgear rooms).

was likely to provide benefit The majority of the other The analysis methodology either via a smaller zone of scenarios were analyzed conservatism is primarily influence or to credit automatic using the generic fire associated with conservatism suppression.

modeling treatments, in the heat release rates specified in NUREG/CR-Additional refinement of the fire This task also includes 6850.

scenarios was pursued using the devleopment of a multi-point analysis of the heat multi-compartment The primary potential key release rates as opposed to the analysis and structural assumption and related use of a bounding fire for most steel analysis.

source of uncertainty in this scenarios. Additional fire task is in the area of the time modeling was pursued in areas delay associated with cable of high risk, notably the damage that resulted in switchgear rooms.

several different related RAIs.

The time delay associated with cable damage that was incorporated into the fire modeling is identified as a potential key source of uncertainty.

Post-Fire Human 12 Reliability The post-fire HRA was Human error probabilities Detailed fire HEP values have Analysis (HRA) performed by developing represent a potentially large not been developed in all cases, a post-fire human error uncertainty for the FPRA and screening or scoping HEP probability (HEP) for each given the importance of values have been applied to credited action. For human actions in the base some of the less risk significant cases where detailed model. A potential key HEPs. This approach should post-fire HEPs were not assumption is that the HRA help reduce the impacts of developed, screening or methods utilized for PNP uncertainty associated with this scoping values were provide representative HEP issue.

used consistent with the values in the analysis guidance provided in commensurate with their In any event, the human error NUREG-1 921.

importance.

probabilities used in the Fire PRA model are identifed as a potential key source of uncertainty.

Seismic Fire 13 Interactions A qualitative seismic-fire Since this is a qualitative Seismic-fire interaction has no review was performed evaluation, there is no impact on fire risk quantification.

and documented.

quantitative impact with respect to the uncertainty of this task.

Page 12 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESUL TS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Detailed Fire 11 Modeling The application of Utlimately, the treatment of Detailed fire modeling was detailed fire modeling these issues has evolved performed only on those was limited to the Main through the various RAls and scenarios which otherwise would Control Room (MCR) subsequent model have been notable risk abandonment scenario, refinements to reduce the contributors and only where and a few risk significant number of potential key removal of conservatism in the areas (e.g., in the 1C and assumptions.

generic fire modeling solution 1 D swithcgear rooms).

The analysis methodology was likely to provide benefit The majority of the other either via a smaller zone of scenarios were analyzed conservatism is primarily influence or to credit automatic using the generic fire associated with conservatism suppression.

modeling treatments.

in the heat release rates specified in NUREGlCR-Additional refinement of the fire This task also includes 6850.

scenarios was pursued using the devleopment of a The primary potential key multi-point analysis of the heat multi-compartment release rates as opposed to the analysis and structural assumption and related use of a bounding fire for most steel analysis.

source of uncertainty in this scenarios. Additional fire task is in the area of the time modeling was pursued in areas delay associated with cable of high risk, notably the damage that resulted in switchgear rooms.

several different related RAls.

The time delay associated with cable damage that was incorporated into the fire modeling is identified as a potential key source of uncertainty.

Post-Fire Human 12 Reliability The post-fire HRA was Human error probabilities Detailed fire HEP values have Analysis (HRA) performed by developing represent a potentially large not been developed in all cases, a post-fire human error uncertainty for the FPRA and screening or seoping HEP probability (HEP) for each given the importance of values have been applied to credited action. For human actions in the base some of the less risk significant cases where detailed model. A potential key HEPs. This approach should post-fire HEPs were not assumption is that the HRA help reduce the impacts of developed, screening or methods utilized for PNP uncertainty associated with this scoping values were provide representative HEP issue.

used consistent with the values in the analysis In any event, the human error guidance provided in commensurate with their NUREG-1921.

importance.

probabilities used in the Fire PRA model are identifed as a potential key source of uncertainty.

Seismic Fire 13 Interactions A qualitative seismic-fire Since this is a qualitative Seismic-fire interaction has no review was performed evaluation, there is no impact on fire risk quantification.

and documented.

quantitative impact with respect to the uncertainty of this task.

Page 12 of 18

FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK IT TASK ASSUMPTIONS RESULTS TO THE NO.

T LE DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Fire Risk 14 Quantification The fire PRA was As the culmination of other Since the fire PRA solves for quantified using the tasks, most of the uncertainty CCDP (prior to the application of FRANC analysis tool.

associated with quantification frequency) at a truncation limit of The quantitative results has already been addressed.

1.OE-09 for CDF and 1.OE-1O for are summarized in the One source of uncertainty is LERF, there should not be a Fire PRA Quantification the selection of the truncation significant truncation and Summary Notebook.

limit, contribution. These truncation limits are several orders of magnitude below the typical values calculated. Additionally, the final truncation values utilized in the integrated one-top model are compared to the PRA standard requirement of less than 5% change per decade of truncation and further discussed in the Fire PRA Quantification and Summary Notebook. As such, the truncation values utilized are not identified as a potential key source of uncertainty.

Uncertainty and 15 Sensitivity Uncertainty and This task does not introduce N/A Analysis Sensitivity are discussed any new uncertainties but is in the Fire PRA intended to address how Quantification and uncertainties may impact the Summary Notebook, fire risk.

Fire PRA 16 Documentation The FPRA is documented This task does not introduce The documentation task in a series of reports.

any new uncertainties to the compiles the results of the other fire risk. Uncertainty tasks. See specific technical considerations should be tasks above for a discussion of documented in a manner that their associated uncertainty and facilitates FPRA applications, sensitivity.

upgrades, and peer review.

Based on the uncertainty and sensitivity review summarized above, potential key assumptions (i.e., those that could impact the NFPA 805 application) were identified to include: non-suppression probabilities associated with the cable damage time, human error probabilities, fire ignition bin frequencies (in addition to the sensitivity analysis required by the use of NUREG/CR-6850 Supplement 1 (EPRI) ignition frequencies for all bins), and assumed cable routings.

Sensitivity analysis are performed for each of the potential key sources of uncertainty identified above, and these sensitivity cases will be re-performed with the base PAl Response Fire PRA Model. The results of these sensitivity cases will be included in the updated revision to the Fire PRA Fire Risk Quantification and Summary Notebook for the RAI Response Fire PRA Model.

Page 13 of 18 FPRA UNCERTAINTY AND SENSITIVITY MATRIX POTENTIAL KEY SENSITIVITY OF THE TASK TASK TITLE TASK ASSUMPTIONS RESULTS TO THE NO.

DESCRIPTION AND SOURCES OF SOURCE(S) OF UNCERTAINTY UNCERTAINTY Fire Risk 14 Quantification The fire PRA was As the culmination of other Since the fire PRA solves for quantified using the tasks, most of the uncertainty CCDP (prior to the application of FRANC analysis tool.

associated with quantification frequency) at a truncation limit of The quantitative results has already been addressed. 1.0E-09 for CDF and 1.0E-1 0 for are summarized in the One source of uncertainty is LERF, there should not be a Fire PRA Quantification the selection of the truncation significant truncation and Summary Notebook.

limit.

contribution. These truncation limits are several orders of magnitude below the typical values calculated. Additionally, the final truncation values utilized in the integrated one-top model are compared to the PRA standard requirement of less than 5% change per decade of truncation and further discussed in the Fire PRA Quantification and Summary Notebook. As such, the truncation values utilized are not identified as a potential key source of uncertainty.

Uncertainty and 15 Sensitivity Uncertainty and This task does not introduce N/A Analysis Sensitivity are discussed any new uncertainties but is in the Fire PRA intended to address how Quantification and uncertainties may impact the Summary Notebook.

fire risk.

Fire PRA 16 Documentation The FPRA is documented This task does not introduce The documentation task in a series of reports.

any new uncertainties to the compiles the results of the other fire risk. Uncertainty tasks. See specific technical considerations should be tasks above for a discussion of documented in a manner that their associated uncertainty and facilitates FPRA applications, sensitiVity.

upgrades, and peer review.

Based on the uncertainty and sensitivity review summarized above, potential "key" assumptions (Le., those that could impact the NFPA 805 application) were identified to include: non-suppression probabilities associated with the cable damage time, human error probabilities, fire ignition bin frequencies (in addition to the sensitivity analysis required by the use of NUREG/CR-6850 Supplement 1 (EPRI) ignition frequencies for all bins), and assumed cable routings.

Sensitivity analysis are performed for each of the potential key sources of uncertainty identified above, and these sensitivity cases will be re-performed with the base RAI Response Fire PRA Model. The results of these sensitivity cases will be included in the updated revision to the Fire PRA Fire Risk Quantification and Summary Notebook for the RAI Response Fire PRA Model.

Page 13 of 18

REFERENCES:

1. Palisades Nuclear Plant Fire Probabilistic Risk Assessment Fire Risk Quantification and Summary, ERIN Report 0247-07-0005.01, Revision 1, November 2012.

NRC REQUEST PRA RAIO1.q.O1 The response to PRA RAI 01.q, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, states that the time delay method will replace the damage accrual method originally employed by the fire PRA. Note that in Section H. 1.5.2 of NUREG/CR-6850, the failure times reported in Table H-8 assume steady-state fire exposure conditions and are therefore, not applicable for use in calculating exposure conditions that evolve over time. Provide a technicaljustification for how the Wme delay method accounts for pre-heating of targets that occurs at heat fluxes prior to reaching the peak heat flux for the fire being analyzed including those below the target damage threshold, and those not already taken into account by Table H-8.

Provide updated risk results as part of the aggregate change-in-risk analysis requested in PRA RAI 30 that appropriately account for pre-heating or that conseivatively do not credit the time delay associated with the pre-heating period.

ENO RESPONSE Consistent with industry precedent (References 1, 2), PNP will revise the Fire PRA RAI Response Model to use the damage accrual method using elements of the Arrhenius methodology (Reference 3, 4). As such, technical justification of the time delay method is not provided. The updated risk results will be included in the response to RAI 30.

Due to the revised approach of using the damage accrual method, reference to the time delay method in the previously submitted responses for RAI FM 01.p and RAI FM 02.b is superseded.

REFERENCES:

1. Turkey Point NFPA 805 LAR RAI Responses 4-4-14

2. Turkey Point NFPA 805 LAR RAls 5-27-14 ML14132A081

3. User Need Request on the Acceptability of the Arrhenius Methodology for Environmental Qualification (EQ) for LOCA and POST-LOCA Environments, ML003701987, February 24, 2000 Page 14 of 18

REFERENCES:

1. Palisades Nuclear Plant Fire Probabilistic Risk Assessment Fire Risk Quantification and Summary, ERIN Report 0247-07-0005.01, Revision 1, November 2012.

NRC REQUEST PRA RAI 01.q.01 The response to PRA RAI 01.q, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, states that the "time delay method" will replace the "damage accrual method" originally employed by the fire PRA. Note that in Section H.1.S.2 of NUREGICR-68S0, the failure times reported in Table H-8 assume steady-state fire exposure conditions and are therefore, not applicable for use in calculating exposure conditions that evolve over time. Provide a technical justification for how the "time delay method" accounts for pre-heating of targets that occurs at heat fluxes prior to reaching the peak heat flux for the fire being analyzed including those below the target damage threshold, and those not already taken into account by Table H-8.

Provide updated risk results as part of the aggregate change-in-risk analysis requested in PRA RAI 30 that appropriately account for pre-heating or that conservatively do not credit the time delay associated with the pre-heating period.

ENO RESPONSE Consistent with industry precedent (References 1, 2), PNP will revise the Fire PRA RAI Response Model to use the 'damage accrual' method using elements of the Arrhenius methodology (Reference 3,4). As such, technical justification of the 'time delay' method is not provided. The updated risk results will be included in the response to RAI

30.

Due to the revised approach of using the 'damage accrual' method, reference to the

'time delay' method in the previously submitted responses for RAI FM 01.p and RAI FM 02.b is superseded.

REFERENCES:

1. Turkey Point - NFPA 805 LAR RAI Responses 4-4-14

2. Turkey Point - NFPA 805 LAR RAls 5-27-14 ML14132A081

3. User Need Request on the Acceptability of the Arrhenius Methodology for Environmental Qualification (EQ) for LOCA and POST -LOCA Environments, ML003701987, February 24, 2000 Page 14 of 18

4. PLP-RPT-00057, Attachment PRA-RAI-Ol.q.01 NRC REQUEST PRA RAI O1.r.O1 The response to PRA RAI Olr, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, states that a one-minute time delay will be applied for credited automatic detection systems.

a. How is the probability of failure of automatic detection included in the PRA?

b.

If the automatic detection fails, is manual detection then credited?

c.

When manual detection is credited after automatic detection fails, is the 15 minute delay used?

d.

If a logical scenario of detection failure, manual detection with 15 minute delay, and attempted manual suppression is not included in the PRA. Evaluate the impact on the results of not including this scenario or add it to the PRA.

ENO RESPONSE

a. The fire PRA model is being updated to include the failure probability of automatic detection systems credited in the calculation of manual non-suppression probabilities (NSPs). As stated in the response to PRA RAI 01.r, no automatic detection systems were credited in support of the activation of automatic suppression systems as the automatic suppression systems are all wet-pipe systems.

In order to account for the failure probability of automatic detection systems credited in support of manual suppression, two sets of manual non-suppression probabilities are being calculated for each applicable set of fire scenarios.;

1) The first set is calculated assuming the automatic detection system fails and the corresponding manual detection time is used (e.g. 15 minutes).

2) The second set is calculated assuming the automatic detection is successful and the corresponding time to detection is used (e.g. 1 minute).

These two sets of NSPs are pro-rated by the automatic detection system success/failure rates. The first set of NSPs are multiplied by the automatic detection system failure probability (e.g. 0.05) and the second set of NSPs are multiplied by the complement of the failure probability (e.g. 0.95). The pro-rated NSPs from each set are summed and applied to the appropriate fire scenarios.

Page 15 of 18

4. PLP-RPT-00057, Attachment PRA-RAI-01.q.01 NRC REQUEST PRA RAI01.r.01 The response to PRA RAI 01 r, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, states that a one-minute time delay will be applied for credited automatic detection systems.

a. How is the probability of failure of automatic detection included in the PRA?

b. If the automatic detection fails, is manual detection then credited?

c. When manual detection is credited after automatic detection fails, is the 15 minute delay used?

d. If a logical scenario of detection failure, manual detection with 15 minute delay, and attempted manual suppression is not included in the PRA. Evaluate the impact on the results of not including this scenario or add it to the PRA.

ENO RESPONSE

a. The fire PRA model is being updated to include the failure probability of automatic detection systems credited in the calculation of manual non-suppression probabilities (NSPs). As stated in the response to PRA RAI 01.r, no automatic detection systems were credited in support of the activation of automatic suppression systems as the automatic suppression systems are all wet-pipe systems. In order to account for the failure probability of automatic detection systems credited in support of manual suppression, two sets of manual non-suppression probabilities are being calculated for each applicable set of fire scenarios. ;

1) The first set is calculated assuming the automatic detection system fails and the corresponding manual detection time is used (e.g. 15 minutes).

2) The second set is calculated assuming the automatic detection is successful and the corresponding time to detection is used (e.g. 1 minute).

These two sets of NSPs are pro-rated by the automatic detection system success/failure rates. The first set of NSPs are multiplied by the automatic detection system failure probability (e.g. 0.05) and the second set of NSPs are multiplied by the complement of the failure probability (e.g. 0.95). The pro-rated NSPs from each set are summed and applied to the appropriate fire scenarios.

Page 15 of 18

b. Yes, manual detection is credited if automatic detection fails as discussed in the response to part a) above.

c.

Yes, as discussed in the response to PRA RAI 01.r, the application of a 15 minute manual detection time is applied when appropriate, If manual detection is not considered credible, manual suppression will not be credited when the automatic detection system is assumed to fail or is nonexistent.

d. As discussed in the response to part a) above, the fire PRA model is being updated so that the NSPs applied to fire scenarios crediting automatic detection also take into account the failure probabilities of these automatic detection systems, and the resulting impact on the detection times. An evaluation of the impact of not including these scenarios is therefore not required.

NRC REQUEST PRA RAI O1.yOl The response to PRA RAI O1.y, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, appears to indicate that the barrier failure probability is defined by the most limiting barrier (e.g., non-rated barrier, door, damper, or wall) and not the sum of the types of barriers present.

Demonstrate that the impact on the results is not significant or provide updated risk results as part of the aggregate change-in-risk analysis requested in PRA RAI 30, summing the barrier failure probabilities for each type of barrierpresent per NUREG/CR-6850.

ENO RESPONSE In response to this RAI, the multi-compartment barrier failure probability is being updated to sum the barrier failure probabilities for each type of barrier present per NUREG/CR-6850. The risk results provided with the response to PRA RAI 30 will reflect this change.

NRC REQUEST PRA RAI 12.01 The ASME PRA standard calls for a focused scope peer review for PRA upgrades, where PRA upgrade is defined in the standard as:

The incorporation into a PRA model of a new methodology or significant changes in scope or capability that impacts the significant accident sequences or the significant accident progression sequences.

Page 16 of 18

b. Yes, manual detection is credited if automatic detection fails as discussed in the response to part a) above.

c. Yes, as discussed in the response to PRA RAI 01.r, the application of a 15 minute manual detection time is applied when appropriate. If manual detection is not considered credible, manual suppression will not be credited when the automatic detection system is assumed to fail or is nonexistent.

d. As discussed in the response to part a) above, the fire PRA model is being updated so that the NSPs applied to fire scenarios crediting automatic detection also take into account the failure probabilities of these automatic detection systems, and the resulting impact on the detection times. An evaluation of the impact of not including these scenarios is therefore not required.

NRC REQUEST PRA RAJ 01.y.01 The response to PRA RAI 01.y, in the letter dated December 2, 2013, ADAMS Accession No. ML13336A649, appears to indicate that the barrier failure probability is defined by '1he most limiting barrier (e.g., non-rated barrier, door, damper, or wall)" and not the sum of the types of barriers present.

Demonstrate that the impact on the results is not significant or provide updated risk results as part of the aggregate change-in-risk analysis requested in PRA RAI 30, summing the barrier failure probabilities for each type of barrier present per NUREG/CR-6850.

ENO RESPONSE In response to this RAI, the multi-compartment barrier failure probability is being updated to sum the barrier failure probabilities for each type of barrier present per NUREG/CR-6850. The risk results provided with the response to PRA RAI 30 will reflect this change.

NRC REQUEST PRA RAJ 12.01 The ASME PRA standard calls for a focused scope peer review for PRA upgrades, where PRA upgrade is defined in the standard as:

'The incorporation into a PRA model of a new methodology or significant changes in scope or capability that impacts the significant accident sequences or the significant accident progression sequences. "

Page 16 of 18

The response to RAI 12 states, the detailed HEP methodology was reviewed by the peer review and has not been changed. As such, a focused scope review of the HEP analysis is also not warranted. The response to RA123.e states, the use of NUREG 1921 methods for screening, scoping and detailed HEP values constitutes data and methods not included in the fire PRA peer review. However, these data and methods are considered acceptable for use.

a.

Clarify these conflicting statements considering that using data and methods acceptable for use is unrelated to the need for a peer review.

b. Describe the method that will be used to ensure that any PRA upgrade will be peer reviewed.

ENO RESPONSE

a. The response to PRA RAI 12 should be clarified as:

the detailed HEP methodology was reviewed by the peer review and has not been changed. As such, a focused scope review of the detailed HEP methodology is also not warranted.

The response to PRA RAI 23.e should be clarified as:

the use of NUREG-1921 methods for scoping HEP values constitutes a method not included in the fire PRA peer review. Therefore, the new methods are considered to require a focused scope peer review.

A focused scope peer review on the use of NUREG-1921 scoping methods will be performed consistent with ASME/ANS RA-Sa-2009. Any findings and their resolution will be described in the response to PRA RAI 30.

b.

ENO PRA configuration control procedure EN-DC-151 ensures that any PRA upgrades receive appropriate peer reviews.

REFERENCES:

1.

NUREG-1 921, Fire Human Reliability Analysis Guidelines, Final Report, EPRI 1023001, EPRI/NRC-RES, July2012.

2. ASME/ANS RA-Sa2009, Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications, ASME/ANS RA-S Committee and ASME, February 2009.

3.

EN-DC-151, Revision 5, PSA Maintenance and Update, Nuclear Management Manual, November 2013.

Page 17 of 18 The response to RAI 12 states, "the detailed HEP methodology was reviewed by the peer review and has not been changed. As such, a focused scope review of the HEP analysis is also not warranted." The response to RA123.e states, 'Yhe use of NUREG-1921 methods for screening, scoping and detailed HEP values constitutes data and methods not included in the fire PRA peer review. However, these data and methods are considered acceptable for use. "

a. Clarify these conflicting statements considering that using data and methods acceptable for use is unrelated to the need for a peer review.

b. Describe the method that will be used to ensure that any PRA upgrade will be peer reviewed.

ENO RESPONSE

a. The response to PRA RAI 12 should be clarified as:

the detailed HEP methodology was reviewed by the peer review and has not been changed. As such, a focused scope review of the detailed HEP methodology is also not warranted."

The response to PRA RAI 23.e should be clarified as:

the use of NUREG-1921 methods for scoping HEP values constitutes a method not included in the fire PRA peer review. Therefore, the new methods are considered to require a focused scope peer review."

A focused scope peer review on the use of NUREG-1921 scoping methods will be performed consistent with ASMEIANS RA-Sa-2009. Any findings and their resolution will be described in the response to PRA RA130.

b. END PRA configuration control procedure EN-OC-151 ensures that any PRA upgrades receive appropriate peer reviews.

REFERENCES:

1. NUREG-1921, "Fire Human Reliability Analysis Guidelines", Final Report, EPRI 1023001, EPRI/NRC-RES, July 2012.

2. ASME/ANS RA-Sa-2009, "Standard for Level1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications", ASME/ANS RA-S Committee and ASME, February 2009.

3. EN-OC-151, Revision 5, "PSA Maintenance and Update", Nuclear Management Manual, November 2013.

Page 17 of 18

NRC REQUEST PRA RAI31 The responses to several PRA RAIs (e.g., 01.g, 01.cc, and 03) are contingent on the development of a new all-inclusive fire response procedure. Describe if there is an Implementation Item in table S-3 that addresses the development and implementation of this procedure. If not, describe the method that will be used to ensure development of the procedure.

ENO RESPONSE The completion of a new all-inclusive procedure is an implementation action.

Implementation item 1, in Table S-3 of the PNP NFPA 805 LAR, Attachment S, addresses the development and implementation of the new all-inclusive fire response procedure. Completion of this implementation item is controlled via the PNP Commitment Tracking Process, specifically under LO-LAR-201 3-00052.

Page 18 of 18 NRC REQUEST PRA RAI31 The responses to several PRA RAls (e.g., 01.g, 01.cc, and 03) are contingent on the development of a new '~II-inclusive" fire response procedure. Describe if there is an Implementation Item in table 5-3 that addresses the development and implementation of this procedure. If not, describe the method that will be used to ensure development of the procedure.

ENO RESPONSE The completion of a new 'all-inclusive' procedure is an implementation action.

Implementation item 1, in Table 5-3 of the PNP NFPA 805 LAR, Attachment 5, addresses the development and implementation of the new "all-inclusive" fire response procedure. Completion of this implementation item is controlled via the PNP Commitment Tracking Process, specifically under LO-LAR-2013-00052.

Page 18 of 18