ML23006A026

NRR E-mail Capture - Audit Plan - Turkey Point Fire Protection Changes to Support RCP Seal Replacement LAR

ML23006A026
Person / Time
Site:	Turkey Point
Issue date:	01/05/2023
From:	Michael Mahoney NRC/NRR/DORL/LPL2-2
To:	Mack J Florida Power & Light Co
References
L-2022-LLA-0128
Download: ML23006A026 (11)

Text

From: Michael Mahoney Sent: Thursday, January 5, 2023 9:06 AM To: Mack, Jarrett

Subject:

Audit Plan - Turkey Point Fire Protection Changes To Support RCP Seal Replacement LAR (EPID L-2022-LLA-0128)

Attachments: Audit Plan_Turkey Point Fire Protection Program LAR.docx

Jarrett, By application dated August 26, 2022 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML22243A161), Florida Power and Light Company (FPL, the licensee) submitted a license amendment request (LAR) for the Turkey Point Nuclear Generating Station, Units 3 and 4.

The proposed LAR will revise the operating license, paragraph 3.D, Fire Protection, for fire protection program changes that may be made without prior U.S. Nuclear Regulatory Commission (NRC) approval. One of the criteria for such a change is that the risk increase resulting from the change is less than 1x10-7/year (yr) for Core Damage Frequency (CDF) and less than 1 x 10-8/yr for Large Early Release Frequency (LERF). The change is to support replacement of the currently installed reactor coolant pump (RCP) seals with the Framatome RCP hydrostatic seal package equipped with the Passive Shutdown Seal (PSDS).

To improve the efficiency of the NRC staffs review, FPLs representatives and the NRC staff have discussed the performance of an NRC staff audit using an online reference portal that would allow the NRC staff limited, read-only access to the information identified in Section 5.0 of the attached audit plan. The NRC staff plans to conduct a desk audit to review the documentation provided on the portal. The online reference portal would allow the NRC staff to audit internal licensee information to confirm that the information support statements were made in the LAR and to determine whether the information included in the documents is necessary to reach a safety conclusion on the application. Any audit information that the NRC staff determines to be necessary to support the development of the NRC staffs safety evaluation will be requested to be formally submitted on the docket. The audit may also include interactions (e.g., teleconferences or webinars) on a mutually agreeable schedule sufficient to understand or resolve issues associated with the information made available on the online reference portal.

Use of the online reference portal is acceptable, as long as the following conditions are met:

• The online reference portal will be password-protected, and passwords will be assigned to those directly involved in the review on a need-to-know basis.

• The online reference portal will be sufficiently secure to prevent NRC staff from printing, saving, or downloading any documents; and

• Conditions of the use of the online reference portal will be displayed on the login screen and with concurrence by each user.

These conditions associated with the online reference portal must be maintained throughout the duration of the audit process.

The NRC staff would like to request that the portal be populated with the information identified in section 5.0 of attached audit plan. The NRC staff may request additional documents during the review, which will be transmitted to you via email.

If you have any questions, please contact me.

Thanks Mike Mahoney Project Manager, LPL2-2 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation Desk: (301)-415-3867 Mobile: (301)-250-0450 Email: Michael.Mahoney@nrc.gov

Hearing Identifier: NRR_DRMA Email Number: 1887 Mail Envelope Properties (SA1PR09MB9486042AE64B5551A659A608E5FA9)

Subject:

Audit Plan - Turkey Point Fire Protection Changes To Support RCP Seal Replacement LAR (EPID L-2022-LLA-0128)

Sent Date: 1/5/2023 9:05:52 AM Received Date: 1/5/2023 9:05:56 AM From: Michael Mahoney Created By: Michael.Mahoney@nrc.gov Recipients:

"Mack, Jarrett" <Jarrett.Mack@fpl.com>

Tracking Status: None Post Office: SA1PR09MB9486.namprd09.prod.outlook.com Files Size Date & Time MESSAGE 3212 1/5/2023 9:05:56 AM Audit Plan_Turkey Point Fire Protection Program LAR.docx 137305 Options Priority: Normal Return Notification: No Reply Requested: No Sensitivity: Normal Expiration Date:

AUDIT PLAN REGARDING LICENSE AMENDMENT REQUEST TO REVISE THE FIRE PROTECTION PROGRAM IN SUPPORT OF REACTOR COOLANT PUMP SEAL REPLACEMENT PROJECT FLORIDA POWER AND LIGHT COMPANY TURKEY POINT NUCLEAR GENERATING STATION, UNITS 3 AND 4 DOCKET NOS. 50-250 AND 50-251 EPID NO. L-2022-LLA-0128

1.0 BACKGROUND

By application dated August 26, 2022 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML22243A161), Florida Power and Light Company (FPL, the licensee) submitted a license amendment request (LAR) for the Turkey Point Nuclear Generating Station, Units 3 and 4 (Turkey Point).

The proposed LAR will revise the operating license, paragraph 3.D, Fire Protection, for fire protection program changes that may be made without prior U.S. Nuclear Regulatory Commission (NRC) approval. One of the criteria for such a change is that the risk increase resulting from the change is less than 1x10-7/year (yr) for Core Damage Frequency (CDF) and less than 1 x 10-8/yr for Large Early Release Frequency (LERF). The change is to support replacement of the currently installed reactor coolant pump (RCP) seals with the Framatome RCP hydrostatic seal package equipped with the Passive Shutdown Seal (PSDS).

The NRC staffs review of the LAR has commenced in accordance with the Office of Nuclear Reactor Regulation (NRR) Office Instruction LIC-101, License Amendment Review Procedures. The NRC staff has determined that a regulatory audit should be conducted in accordance with the NRR Office Instruction LIC-111, Revision 1, Regulatory Audits, dated October 31, 2019 (ML19226A274), for the NRC staff to gain a more detailed understanding of the licensees proposed license amendment.

A regulatory audit is a planned, license-related or regulation-related activity that includes the examination and evaluation of primarily non-docketed information. A regulatory audit is conducted with the intent to gain understanding, to verify information, and/or to identify information that will require docketing to support the basis of the licensing or regulatory decision. Performing a regulatory audit of the licensees information is expected to assist the NRC staff in efficiently conducting its review or gain insights on the licensees processes or procedures. Information that the NRC staff relies upon to make the safety determination must be submitted on the docket. However, there may be supporting information retained as records under Title 10 of the Code of Federal Regulations (10 CFR) 50.71, Maintenance of records, making of reports, and/or 10 CFR 54.37, Additional records and record-keeping requirements, which although not required to be submitted as part of the licensing action, would help the NRC staff better understand the licensees submitted information.

Enclosure

2.0 REGULATORY AUDIT BASIS An audit was determined to be the most efficient approach toward a timely resolution of questions associated with this LAR review, because the NRC staff will have an opportunity to minimize the potential for further rounds of requests for additional information (RAIs) and ensure no unnecessary burden will be imposed by requiring the licensee to address issues that are no longer necessary to make a safety determination. The NRC staff is requesting an initial set of internal licensee information to be reviewed by the staff using an online reference portal. Upon completion of this audit, the NRC staff is expected to achieve the following.

1. Confirm licensee information which supports statements made in the LAR.

2. Determine whether the information included in the documents is necessary to be submitted to support a safety conclusion.

The audit information that the NRC staff determines to be necessary to support the development of the NRC staffs safety evaluation will be requested to be submitted on the docket.

3.0 REGULATORY AUDIT SCOPE OR METHOD The purpose of the remote audit is to gain a more detailed understanding of licensees proposed LAR. The areas of focus for the regulatory audit are the information contained in the licensees August 26, 2022, submittal, and the proposed audit questions in section 5.0 of this audit plan.

4.0 AUDIT TEAM The audit will be conducted by NRC staff from the NRR Division of Risk Assessment (DRA),

Probabilistic Risk Assessment (PRA) Licensing Branches A and B, and the PRA Oversight Branch (APLA, APLB, and APOB, respectively), who are knowledgeable in probabilistic risk assessment, safe shutdown and circuit analysis, fire modeling, and fire protection engineering.

NRR staff in the Division of Engineering and External Hazards (DEX), Mechanical Engineering and Inservice Testing Branch (EMIB) who are knowledgeable in mechanical components will also participate in the audit. The audit will be led by staff from the NRR Division of Operating Reactor Licensing (DORL). NRC staff from other organizations may be assigned to the team as appropriate and others may participate as observers. Observers at the audit may include NRR Project Managers and various Regional staff.

The following are members of the NRC audit team:

Team Member Title Organization Mike Mahoney Project Manager NRR/DORL/LPLII-2 JS Hyslop Senior Reliability and Risk Analyst NRR/DRA/APLB Thinh Dinh Fire Protection Engineer NRR/DRA/APLB Ching Ng Reliability and Risk Analyst NRR/DRA/APOB Jeff Circle Senior Reliability and Risk Analyst NRR/DRA/APLA Thomas Scarbrough Senior Mechanical Engineer NRR/DEX/EMIB Gurjendra Bedi Mechanical Engineer NRR/DEX/EMIB Summer Sun Senior Nuclear Engineer NRR/DSS/SNSB

5.0 PROPOSED AUDIT QUESTIONS AND INFORMATION REQUEST Audit Questions

1. In the LAR submittal dated August 26, 2022, in the first paragraph of section 2.1, System Description and Operation, the licensee states that Westinghouse Model 93 Reactor Coolant Pumps (RCPs) are installed at Turkey Point. The licensee indicates that an RCP seal prevents reactor coolant from exiting the RCP body along the pumps impeller shaft.

Turkey Point currently has a three-stage RCP seal design installed in Unit 3 and Unit 4.

These seal types were first installed during the Unit 3 fall 2015 and Unit 4 spring 2016 refueling outages. The current RCP seal has three identical hydrodynamic mechanical stages in series, plus an abeyance seal designed to be installed on the Westinghouse RCPs. Please describe the performance of the currently installed RCP seals and any lessons learned from that performance applicable to the new RCP seal design.

2. In the LAR submittal dated August 26, 2022, in the fourth paragraph of section 2.2, Framatome RCP Hydrostatic Seal with PSDS, the licensee states that the April 26, 2022, pre-submittal meeting included a discussion of the NRC staffs audit in calendar year 2015 of the Framatome RCP hydrostatic seals with the PSDS device. In the LAR, the licensee states that Framatome has reviewed the audit questions and responses to determine if any changes were made to the PSDS design that would impact the Turkey Point installation. The licensee states that Framatome determined that the majority of the audit items were specific to the design under review, and that none of the audit items affect the current Framatome RCP hydrostatic seal package with PSDS to be installed at Turkey Point. Please describe these audit questions and the Framatome resolution.

3. In the LAR submittal dated August 26, 2022, in section 3.6, Safety Margin, the licensee states that NRC Regulatory Guide (RG) 1.174, section 2.1.2, Safety Margins, lists two specific criteria that should be addressed when considering the impact of plant changes on safety margins: (1) the codes and standards or their alternatives accepted for use by the NRC are met; and (2) safety analysis acceptance criteria in the licensing basis (e.g.,

final safety analysis report, supporting analyses) are met or provide sufficient margin to account for analysis and data uncertainty. Please describe the applicable codes and standards that have been met for Framatome RCP seal design change.

4. The LAR submittal dated August 26, 2022, in Attachment 1, provides proprietary information regarding the replacement RCP seals at Turkey Point. Please describe the RCP seal design changes and their improvement over the current Turkey Point RCP seals.

5. In LAR section 3.1.5, Operator Action, the licensee states that the current required time for the operator to trip the RCPs following a LOSC event was established as 20 minutes to protect the RCP seals from damage. The licensee proposed that for the replacement Framatome RCP seals, the available operator time to trip RCPs is 16 minutes during a LOSC event. The change in the RCP trip time on a LOSC event would affect the Human Error Probability values, which are inputs to the Fire probabilistic risk assessment model.

In LAR section 2.1, the licensee states that the current operator action time of 20 minutes was previously approved by the NRC (Reference 5 of the LAR, ML19064A903).

The associated NRC safety evaluation report (SER) indicated that the approval was

based, in part, on the data of the RCP trip time in LOSC conditions applicable to the Flowserve RCP seals, which are similar to the current RCP seals at Turkey Point.

Specifically, The SER (page 10) quoted the following licensees statement as part of the basis for approval: the licensee further stated that the time available to trip the RCPs following a loss of seal cooling is 20 minutes, whereas the time available assumed in the Flowserve topical report (TR) is 60 minutes, and that assuming a shorter time for operators to recognize the need and trip the RCPs results in increases in the calculated failure probability, which in turn increases the calculated core damage frequency (CDF) and LERF values, and is conservative.

Provide a discussion to justify that the reduction in time to trip the RCPs during a LOSC from 20 minutes to 16 minutes, to be implemented for the replacement Framatome RCP seals, is acceptable for protecting the RCP seals from damage at TP.

6. Sections 3.1.5 and 3.5 (Operator Action and Defense-In-Depth) of the Turkey Point LAR dated August 26, 2022 (ML22243A162) discusses a required RCP trip time of 16 minutes but does not provide the bases for this assumption. Explain the bases for the RCP trip time of 16 minutes. Explain whether this operation is a recovery action or a required post-fire manual operator operation, and whether this action has been incorporated into appropriate procedures and demonstrated to be feasible and reliable. Discuss if there are any other operator actions which need to be performed along with an RCP trip and if so, what are their impact and how are they credited in the PRA models.

7. Section 3.3 (Risk Analysis Results) of the LAR reports the total core damage frequency (CDF) for full power internal events, internal flood, and fire probabilistic risk assessment (PRA) for U3 and U4 as 8.01E-5 and 7.82E-5, respectively. However, there is no estimate for seismic and other external hazards. Provide seismic and other external hazard contributions to CDF/LERF, or alternately provide a justification for why the Regulatory Guide 1.174, An Approach For Using Probabilistic Risk Assessment In Risk-Informed Decisions On Plant-Specific Changes To The Licensing Basis, guidelines are met without quantitative contributions from those hazards. Provide a discussion for how, after taking uncertainty into account for the addition of seismic and other external hazards, the total CDF and LERF will not exceed the RG 1.174 acceptance guidelines.

8. Section 3.1.4 (New and Modified Scenarios) of the LAR, indicates that high energy arcing fault (HEAF) scenarios for different sources in a room are merged into one HEAF scenario. However, Table 3-2, New Scenarios, of Attachment 1 appear to divide the scenarios for load centers into individual fire scenarios, rather than grouping them.

Address this apparent inconsistency.

9. Appendix A of Attachment 1 of the LAR uses the fire CDF and LERF to determine the internal events/flooding delta CDF and LERF. The LAR indicates that fire induced failures of systems to prevent the loss of the RCP seals are larger than random failures; thus, extrapolating internal events/flooding delta CDF and LERF from fire is conservative.

However, the NRC staff notes that other fire damage in general plays a role in these scenarios, and thus, comparing the fire induced failure of systems relevant to loss of the RCP seals is only one factor, and perhaps not the dominant factor. Therefore, the NRC staff does not have sufficient information to determine whether the delta risk from internal events/flooding being derived from fires only would be an acceptable approach. As a result, provide a calculation of internal events/flooding delta CDF/LERF which relies only on the internal events/flooding PRA model, and is not extrapolated from the fire PRA

model.

10. Indicate whether any upgrades have been made to the PRA since the last peer review of either corresponding hazard.

11. Indicate whether the internal events/flooding model of record identified in Appendix A of Attachment 1 to the LAR is the latest PRA and includes the proposed RCP seal change described in the LAR. If not, indicate why the CDF and LERF of these hazards is acceptable, or otherwise provide quantitative risk values which incorporates the proposed seal changes.

12. In the uncertainty matrix, Table C-4 of Attachment 1 to the LAR, task no. 5, it is stated that a reactor trip alone is assumed as the initiator for all fires that are quantified in the fire PRA. However, the NRC staff acknowledges that other initiators may arise from fire (e.g., loss of offsite power (LOOP) that have the potential to be more severe. Provide an updated fire CDF/LERF which incorporates these other potential initiators, or a justification for excluding them.

13. The uncertainty matrix, Table C-4 of Attachment 1 to the LAR, task no.10, states that Circuit failure mode likelihood analysis was generally limited to those components where spurious operation could not be caused by the generation of a spurious signal. Explain why the likelihood analysis was not applied to those components where spurious operation could be caused by the generation of a spurious signal.

14. The NRC staff notes that the installation of the new RCP seals results in a risk increase.

Explain the basis for the increase in risk in the PRA model.

15. The second + on pg. E-3 of Attachment 1 to the LAR states that the interruptible fire introduces the concept of a fire that is not expected to grow to a point that would be capable of activating an automatic suppression system. The NRC staff considers that statement an oversimplification of what is documented in NUREG-2230, Methodology for Modeling Fire Growth and Suppression Response for Electrical Cabinet Fires in Nuclear Power Plants. NUREG-2230 states that, However, the FPRA must account for the possibility that the fire could grow and be capable of not only damaging targets, but also activating an automatic suppression system. Thus, NUREG-2230 explains that an interruptible fire can cause an automatic suppression system to actuate. Similar to a growing fire, the interruptible fire heat release rate (HRR) profile, as described in section 4.2.2 of NUREG-2230, should be used when estimating the activation time of an automatic heat detection or thermally activated automatic suppression system for an interruptible fire. Explain if the analysis allows for the actuation of an automatic suppression system from interruptible fires, and if not, then justify why it was not considered.

16. Section E.2.1.3 of Attachment 1 to the LAR, indicates that the MCR response is 100%

effective, yet there is still the possibility of having an alarm that is silenced by the MCR or not properly investigated, which is reflected in the MCR and plant personnel fault trees. In the application of the methodology presented, this affects the probability listed in Table E-4 for 2. The probability of personnel not present in the room should not be 0. The probability should be dictated by the MCR response failure probability. The MCR response is independent of the ability for the system to provide advance warning or prompt warning. Provide further justification for the choice of MCR response and how it

was determined that the probability of personnel present in the room is 100%. Alternately, adjust the credit in the table below. (The relevant portion of Table E 2 is provided below).

17. Page 22 of Attachment 3 to the LAR indicates that the failure to open of check valves is used as a conservative failure rate for the wave spring failure. Provide the justification for this assumption.

18. The NRC staff notes that (1) using the Jeffreys non-informative prior, the failure to actuate probability would be 4.95E-03, which would represent an approximately 3x the RCP Seal Failure Probability; (2) the c uncertainty parameter discussed on page 21 of Attachment 3 to the LAR, is assumed to be 0.15, which is at the lower range of the c. Explain the basis of using a value of 0.15 for c and, provide the justification why a sensitivity study of 2x RCP Seal Failure Probability as shown on page C-1 of Appendix C of Attachment 1, to the LAR, is sufficient to address aforementioned uncertainties.

19. Page 22 of Attachment 5 to the LAR references a Report of the aging and corrosion resistance tests of the polyether ether ketone (PEEK) polymer regarding corrosion resistance testing for the sealing ring. It further indicates that the results of the testing program show that failure of the PEEK material due to corrosion or aging is not significant.

Explain how the parameters in the radiation and chemical testing for the PEEK polymer represent the operating conditions of the plant.

20. Page 19 of Attachment 7 to the LAR indicates that to reasonably account for possible vibration and hydraulic loads on the fuse spacer, the mechanical stress of 435 psi was increased by a factor of 10%, which resulted in an estimated stress load of approximately 479 psi. The new mechanical stress probability distribution curve is included in Figure 6-4.

However, Figure 6-5, which shows the mechanical stress and strength probability distributions for the high temperature fuse spacer are plotted together, which seems to be inconsistent with Figure 6-4. Confirm that a value of 479 psi is used when calculating the failure probability of the fuse spacer due to mechanical stress/strength interaction.

21. Page 21 of Attachment 3 to the LAR, indicates that the expected accident temperature environment at the Fuse Spacer would be 547F. Provide the justification that the expected accident temperature environment at the plant would not exceed 547F.

22. Section 3.5 of Attachment 1 to the LAR presents failures to actuate, remain sealed, and spurious actuation of the PSDS as time dependent failure rate frequencies. However, the development and evaluation of the failure modes for the PSDS is presented in Attachment 3 to the LAR, as per demand failures. Explain the rationale and justification for the inconsistency of units.

23. Section C.1.3 of Attachment 1 to the LAR provides a sensitivity of fire risk results with a 2X RCP seal failure probability. The NRC staff notes that the change in fire CDF due to additional risk from the sensitivity study approaches the RG 1.174 risk threshold. Provide a discussion on why the sensitivity yields acceptable results.

24. Section 3.6.2.1 of Attachment 1 to the LAR indicates that the Application Specific Model (ASM) fault tree explicitly and individually models each of the six RCP seals (3 per unit).

The staff notes that common-cause failure (CCF) of the RCP seals is not modeled in the ASM fault tree provided on pgs. 9-12 of Attachment 1 to the LAR and no basis has been provided for not modeling CCF of the RCP seals. Provide CCF modeling of the RCP seals or alternately provide a justification for why CCF modeling of the RCP seals is not needed.

25. The NRC staff notes that in the RCP Seal Replacement Fault Tree figure shown on pgs.

9-12 of Attachment 1 of the LAR, a few basic events are assigned a value of zero for the RCP Spuriously Starts Due to Fire. Explain the basis for assigning a probability value of zero for these basic events.

Information Request Please make the following information available for the NRC staff to audit:

1. In the LAR submittal dated August 26, 2022, in the fifth paragraph of section 2.2, the licensee states in part that the replacement Framatome RCP hydrostatic seal package with PSDS has been evaluated for installation in Turkey Point under the criteria of 10 CFR 50.59 to determine if any attribute requires prior NRC approval. The licensee states that it considers that no attribute for the installation of the replacement RCP seals requires prior NRC approval at Turkey Point. Please make the 10 CFR 50.59 evaluation available for NRC staff audit review.

2. The LAR submittal dated August 26, 2022, Attachment 1, references testing conducted to demonstrate the performance of the Framatome RCP seals. Please make the referenced testing documents available for NRC staff audit review.

3. If a Topical Report (TR) for the Framatome RCP seal design was not submitted to the NRC for prior review, please make the TR available for NRC staff audit review. If a TR was submitted for NRC review, please provide the associated accession number(s).

6.0 LOGISTICS The audit will be started once an electronic reference portal is set up and the documentation is made available to the NRC staff. The initial desk audit will be conducted over several weeks.

The licensee will be kept informed on a regular basis during periodic discussions with the project manager regarding the progress. The audit may include interactions (e.g.,

teleconferences or webinars) on a mutually agreeable schedule sufficient to understand or resolve issues associated with the information made available.

7.0 DELIVERABLES An audit summary will be prepared within 90 days of the completion of the audit. If the NRC staff identifies information during the audit that is needed to support its regulatory decision, the staff will issue RAIs to the licensee.