ML25073A190

Audit Plan for MRP-480 (EPRI TR 3002023895)

ML25073A190
Person / Time
Site:	Electric Power Research Institute
Issue date:	03/20/2025
From:	Delosreyes J Licensing Processes Branch
To:	Electric Power Research Institute
References
EPRI TR 3002023895
Download: ML25073A190 (8)

Text

Attachment REGULATORY AUDIT PLAN BY THE OFFICE OF NUCLEAR REACTOR REGULATION IN SUPPORT OF THE REVIEW OF TECHNICAL REPORT 3002023895, MATERIALS RELIABILITY PROGRAM: XLPR ESTIMATION OF PWR LOSS-OF-COOLANT ACCIDENT FREQUENCIES (MRP-480), FEBRUARY 2024 ELECTRIC POWER RESEARCH INSTITUTE, INC (EPRI).

DOCKET NO. 99902021

1.0 BACKGROUND

By application dated April 26, 2024 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML24121A204), Electric Power Research Institute, Inc.

(EPRI), submitted topical report (TR) EPRI Report 3002023895, Materials Reliability Program:

xLPR [Extremely Low Probability of Rupture] Estimation of PWR Loss-of-Coolant Accident Frequencies (MRP-480), (ADAMS) Accession No. ML24123A223) to the U.S. Nuclear Regulatory Commission (NRC) for review and approval. The primary objective of MRP-480 is to demonstrate that analytically determined loss-of-coolant accident (LOCA) frequencies in pressurized water reactors (PWRs) using the probabilistic fracture mechanics (PFM) code, Extremely Low Probability of Rupture (xLPR), are similar to those presented in NUREG-1829, Estimating Loss-of-Coolant Accident (LOCA) Frequencies Through the Elicitation Process, Vol. 1 (ML080630013).

The NRC staff has performed an initial review of the submitted topical report and determined that a regulatory audit of the items in the Information Requests and Audit Topics sections of this audit plan would facilitate the timely completion of our review. The NRC staff is continuing to review other aspects of the topical report and may identify the need to audit additional subjects by separate correspondence or during audit discussions.

2.0.

REGULATORY AUDIT BASES The NRC staff determined that an audit is the most efficient approach toward a timely resolution of questions associated with this review. The audit will provide the NRC staff an opportunity to minimize the potential for requests for additional information (RAIs) by ensuring a clear mutual understanding of MRP-480 methodology and clarifying potential review questions with focused communication. As appropriate to facilitate efficient audit interactions, the NRC staff is

2 requesting that EPRI provide access to requested audit documentation through an online reference portal before the face-to-face and/or virtual portion of the audit.

Upon completion of this audit, the NRC staff is expected to achieve the following:

1.

Confirm internal EPRI information that supports statements made in MRP-480.

2.

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

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

3.0 REGULATORY AUDIT SCOPE The purpose of this audit is to gain a more detailed understanding of the analyses contained in MRP-480.

The areas of focus for the regulatory audit are the information contained in the submitted topical report, the enclosed audit information needs, and all associated and relevant supporting documentation (e.g., methodology, process information, calculations, etc.) identified below. The audit will be performed consistent with NRC Office of Nuclear Reactor Regulation Office Instruction LIC-111, Revision 2, Regulatory Audits, dated December 30, 2024 (ADAMS Accession No. ML24309A281).

4.0 INFORMATION REQUESTS Please make the following information for the NRC staff to audit via an online reference portal to support an efficient face-to-face audit:

1.

Responses to the audit topics in Section 9.0 Audit Topics of this audit plan at least one week prior to the audit discussions.

2.

xLPR files (e.g., input decks, output plots, etc.) associated with any of the audit topics in Section 9.0 of this audit plan that may be requested by the staff during the audit discussions.

In addition, to support the NRC staffs understanding of MRP-480 and to help identify any additional information needed to support its review, the staff requests that EPRI be prepared to discuss the audit topics identified in Section 9.0 of this audit plan.

5.0 REVIEW TEAM Members of the NRC staff in the audit team are provided below. Key EPRI personnel involved in the development of the topical report should be made available for interactions on a mutually agreeable schedule to respond to any questions from the NRC staff.

Team Member Division Area of Responsibility James Delosreyes and Lois James NRR/DORL/LLPB Project Management David Dijamco NRR/DNRL/NVIB Technical Reviewer (lead)

Seung Min NRR/DNRL/NPHP Technical Reviewer

3 Eric Palmer NRR/DNRL/NVIB Technical Reviewer David Rudland NRR/DNRL Senior Technical Advisor 6.0 LOGISTICS The audit will be conducted from March 31, 2025 to April 30, 2025, through an online portal (also known as electronic portal, ePortal, or electronic reading room) established by EPRI.

The audit team will conduct a TEAMS-based entrance meeting with the vendor on March 31, 2025, for the purposes of introducing the team, discussing the scope of the audit, and describing the information to be made available on the portal. Through the audit period between March 31, 2025 to April 30, 2025, the NRC staff will hold breakout sessions with representatives of EPRI to answer audit team questions and to have technical discussions. An exit meeting/call will be held at the conclusion of the audit on April 30, 2025.

The NRC staff does not foresee the need for an onsite visit or in-person discussions between the NRC and vendor staff to discuss information to be provided on the portal at this time.

However, if the need for a such a meeting is identified in the future, the audit plan will be revised, and the schedule for the audit will be adjusted accordingly. The NRC project manager (PM) will coordinate any changes to the audit schedule and location with the vendor.

7.0 SPECIAL REQUESTS The NRC staff would like access to the documents listed above in Section 4.0 through an online portal that allows the NRC staff to access documents via the internet. The following conditions associated with the online portal must be maintained throughout the duration that the NRC staff have access to the online portal:

The online portal will be password-protected, and separate passwords will be assigned to the NRC staff who are participating in the audit.

The online portal will be sufficiently secure to prevent the NRC staff from printing, saving, or downloading any information on the online portal.

Conditions of use of the online portal will be displayed on the login screen and will require acknowledgement by each user.

Username and password information should be provided directly to the NRC staff. The NRC PM will provide to EPRI the names and contact information of the NRC staff who will be participating in the audit. All other communications should be coordinated through the NRC PM.

8.0 DELIVERABLES An audit summary report 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 NRC staff will issue RAI(s) to the vendor.

9.0 AUDIT TOPICS

1. Section 1.1 of MRP-480, last paragraph, states: While the ALS is an immediate driver for the investigation into NUREG-1829 LOCA frequency results and time between

4 detectable leakage and LOCA, it should be noted that the results herein are intended to be generic and of use to other projects. The staff noted that the overall alternative licensing strategy (ALS) leverages the approved leak-before-break (LBB) analyses for the reactor coolant system main loop piping of pressurized water reactors, and that the xLPR analyses in MRP-480 are used to confirm and/or supplement the approved LBB analyses to show that the likelihood of a LOCA is low. The specific issue addressed in the ALS is the potential for fuel fragmentation, relocation, and dispersal (FFRD) in high-burnup fuel during design basis accidents. Since MRP-480 was submitted for review and approval under the overall umbrella of ALS, the staffs review of MRP-480 is based on the reports use for supporting ALS. Generally speaking, any topical report and the staffs review of it are based on a clear tie to a specific technical and/or regulatory issue that the report is addressing. Uses of MRP-480 outside the scope of ALS (i.e., outside the evaluation of potential for FFRD and its impact) would need to be reviewed and approved by the staff on a case-by-case basis. Discuss planned and/or expected, or examples of uses of the results of MRP-480 other than those associated with ALS.

2. MRP-480 relies heavily in two reports, the xLPR Piping System Analysis report (Ref. 8 of MRP-480, ML21217A088) and xLPR Generalization Study report (Ref. 9 of MRP-480, ML22088A006), that contain detailed xLPR analyses in the context of LBB-behavior in Alloy 82/182 dissimilar butt welds (in which an active degradation mechanism of primary water stress corrosion cracking [PWSCC] is present) in PWR piping systems. The overarching objective of the two reports was to demonstrate that the probability of rupture in these piping systems with the active degradation mechanism remains low, consistent with the requirement in 10 CFR 50, Appendix A, General Design Criteria 4.

The analyses in the two reports are based on PFM analyses. In the context of risk-informed decision making (RIDM), PFM analyses are integrated with other RIDM principles, particularly performance monitoring, when used as the basis in regulatory and safety decisions. Clarify/summarize the performance monitoring programs (e.g., ongoing inspection programs) in the PWR piping systems within the scope of MRP-480 and discuss if the inspections of the piping systems considered in this effort are sufficient for performance monitoring.

3. Section 1.3 of MRP-480, second paragraph, states: By design, xLPR analyses consider one weld at a time. All cases modeled either an Alloy 82/182 dissimilar metal weld, or a genericized representative similar-metal weld. The most limiting configuration between these two options was chosen for each individual line modeled depending on the corresponding materials and relevant material degradation mechanisms. The NUREG-1829 results (as described in the executive summary of the report) appear to be results for the total system (i.e., entire piping system) fleetwide, not for an individual weld in the piping system nor for a piping system for a specific plant. Since the xLPR analyses are used for comparison with the NUREG-1829 analyses, explain/discuss the following:

a. It appears that most of the cases analyzed in MRP-480 (i.e., in Table 3-3) were for dissimilar metal welds, at least for the line sizes relevant to ALS, and thus, it appears to the staff that dissimilar metal welds were considered the limiting locations in the analyses. Confirm/clarify whether dissimilar metal welds are considered the limiting locations in the analyses in MRP-480.

b. A follow-up to part a. above, discuss how the probabilities of rupture/LOCA at the limiting location (i.e., the limiting weld location) are appropriate for representing probabilities of rupture/LOCA for the entire piping system fleetwide of which the limiting location is part.

5

c. A related question is in the last paragraph of Section 3.6 of MRP-480, where it states: Welds are assumed to behave independently from each other, providing an upper bound on system-level probabilities. Is this saying that a probability for each weld is calculated and then combined to those for a system probability that will be an upper bound?

d. Section 4.1.2 of MRP-480 states that the xLPR Generalization Study report (ML22088A006) also included an assessment of system-level failure frequencies.

Discuss this assessment of system-level failure frequencies in the xLPR Generalization Study report.

4. Section 4.1.5 of MRP-480 discusses the use of rupture frequency as an analogue for large break LOCA frequency. It is not clear to the staff how only two cases out of the many cases in MRP-480 are sufficient to show that rupture frequency is an approximate analogue for large break LOCA frequency. Also, the argument in this section is that every realization of rupture had a corresponding realization of large break LOCA. The staff noted that not every realization of large break LOCA would necessarily lead to rupture, in which case the large break LOCA frequency could actually be higher than rupture frequency.

5. Figure 2-2 of MRP-480 shows the Piping only LOCA frequency results from NUREG-1829 and cites the non-piping-to-piping contribution ratios in Table 7.2 of NUREG-1829.

Clarify how the Piping only results were determined from the ratios.

6. Figure 2-3 of MRP-480 shows NUREG-1829 results that include adjustment for expert panel overconfidence. This adjustment is described in Section 5.6.2 of NUREG-1829.

Section 2.1.3 of MRP-480 references this figure and states that subsequent figures in this report conservatively consider comparison using the base case 40-year LOCA frequencies. Clarify the following:

a. What is the relevance of overconfidence adjustment in the context of the comparison of the MRP-480 LOCA frequency results with those of NUREG-1829?

b. Regarding the phrase quoted above, how is it that the subsequent figures in this report, i.e., MRP-480 which considers 80-year LOCA frequencies, conservatively consider comparison with the 40-year LOCA frequencies? Why not use the 60-year results in Figure 2-3 of MRP-480?

7. Section 2.2.2 of MRP-480 states: The ALS investigates the impacts of LOCAs in different PWR line sizes on LOCA-induced FFRD. For smaller break sizes, there is no cladding rupture for high burnup fuel, and thus no dispersal evaluation is required. EPRI report 3002028673 (ML24121A207, ALS document report), Section 1.9 states: For LOCA sizes below RCS main loop piping, core cooling analysis is performed to determine if clad burst occurs. By justifying that HBU [high burnup] fuel will not undergo cladding burst, potential consequences of dispersal are avoided. EPRI report 3002028675 (ML24121A208), which contains the cladding rupture LOCA analyses, appears to be only for Westinghouse-designed 2-loop, 3-loop, and 4-loop PWRs. Based on this information, the staff notes that PFM analyses are used for the main loop piping sizes to justify a low likelihood of a LOCA and that cladding rupture analyses are used for piping that are less than the main loop piping sizes to justify that FFRD will not occur.

The specific report that applies to each of the three nuclear steam supply system (NSSS) designs appears to be as follows:

6 Main Loop Piping (PFM analyses)

Less than Main Loop Piping (cladding rupture LOCA analyses)

Westinghouse (W)

MRP-480 EPRI report 3002028674/75 Combustion Engineering (CE)

MRP-480

?

Babcock & Wilcox (B&W)

MRP-480

?

The staff is requesting the following clarifications:

a. Confirm that MRP-480 is intended for the main loop piping of all NSSS designs.

b. For the main loop piping, clarify if cases 3.1.0, 3.1.1, and 3.1.2 in Table 3-3 of MRP-480 are the only cases that cover CE and B&W.

c. Clarify which report (or reports), if any, covers the line sizes below the main loop piping for CE and B&W.

8. Table 3-3, Appendix B, and Appendix C of MRP-480 list and describe the cases performed in the report. Clarify the following:

a. Which cases are existing xLPR analyses, and which ones are new analyses performed specifically for MRP-480.

b. Section 4.3 of MRP-480 cites cases in the xLPR Piping System Analysis report (ML21217A088) and xLPR Generalization Study report (ML22088A006). Clarify whether the cases discussed in this section of the report are sensitivity cases or not.

9. Sections 3.1 and 3.2 of MRP-480 discuss seismic effects in both the xLPR Piping System Analysis report (ML21217A088) and xLPR Generalization Study report (ML22088A006).

a. Clarify how seismic effects, both operating basis earthquakes and safe shutdown earthquakes, were considered in both reports.

b. NUREG-1903 describes seismic considerations for the transition break size in reactor coolant primary loop piping. Since MRP-480 is under the umbrella of ALS, which ultimately is about addressing the potential piping line breaks larger than the transition break size, discuss how plant-specific seismic behavior is addressed in MRP-480. As part of the discussion, clarify whether the scope of MRP-480 addresses the applicability of NUREG-1903.

10. Section 3.4.4 of MRP-480 states that the calculation methodology for the time between detectable leakage and rupture output considers the leak rate of the largest circumferential flaw Does largest circumferential flaw mean the largest out of all the realizations?

11. Section 4.1.1 of MRP-480 describes the approach taken for estimating LOCA frequencies for cases when crack initiation models are used and for cases when initial flaws are modeled. The approach taken needs clarification, as well as the following:

a. For cases using crack initiation models, describe the conservatisms and uncertainties in the model and sensitivity studies performed. Describe which sensitivity studies were focused on addressing model uncertainty.

b. Discuss how the recent work on PWSCC initiation presented during the Materials Information Exchange public meetings (summarized in the slides in ML23152A170 and ML24173A153) is addressed in the crack initiation models used in the xLPR

7 results in MRP-480. It is indicated in the slides that one of the objectives of the work presented was to validate the xLPR Code.

c. For cases using initial flaws, describe the flaw distribution and/or flaw characteristics (size, depth, aspect ratio, etc.).

d. With regard to large break LOCA times, intuitively initial flaw cases would result in shorter large break LOCA times as compared to crack initiation cases since realizations would start with an initial flaw. Would there be any case where initial flaw cases would lead to longer large break LOCA times, which could then impact the evaluation in MRP-480 of time from detectable leakage to large break LOCA (or rupture)?

12. Clarifications needed on the plots of LOCA frequency results in Figures 4-1, 4-3, and 4-4 of MRP-480:

a. When applying inservice inspection (ISI), clarify the ISI interval period.

b. The staff was able to confirm from Appendix A of the xLPR Piping System Analysis report (ML21217A088) and xLPR Generalization Study report (ML22088A006) some of the xLPR results in Figure 4-1. However, there are cases the staff could not confirm, such as Case 1.1.1 in Table 3-3 of MRP-480, which appears to be a sensitivity case for starting with an initial flaw, results in a probability of rupture that is quite high: 7.68E-01 (per Appendix A of Ref. 8 of MRP-480) or 7.68E-01/80 years =

9.60E-3 per year. This result doesnt seem to be reflected in Figure 4-1 of MRP-480, but Section 4.1.2 of MRP-480 says that this figure is for all xLPR cases evaluated and listed in Table 3-3. Another case is Case 1.1.19 in Table 3-3 of MRP-480.

These are just two examples, but there may be other cases in Table 3-3 of MRP-480 that cannot be confirmed from the two source reports. Show how the results from Appendix A of the xLPR Piping System Analysis report and xLPR Generalization Study report are plotted in Figure 4-1 of MRP-480.

c. Staff couldnt confirm the results in Figure 4-3 from the xLPR Piping System Analysis report and xLPR Generalization Study report. Clarify how results in these two reports are plotted in Figure 4-3, and which specific analysis cases of the referenced source reports correspond to the data points plotted in Figure 4-3.

d. The one result (yellow dot) just over 10 inches effective break size in Figure 4-3 is very close to one of the 95% upper bound result, but much higher than the other 95% upper bound result for the same break size. Explain the difference between the two 95% upper bound results for the break size just over 10 inches.

e. The executive summary of MRP-480 states: Ruptures did not occur when evaluating in-service inspection (ISI) or leak rate detection (LRD) in all base cases and all but three sensitivity cases. Discuss these three sensitivity cases with respect to results in Figures 4-3 and 4-4 of MRP-480.

13. Section 4.1.2 of MRP-480 states that all cases in Figure 4-1 of the report were modeled with crack growth due to PWSCC, noting that cases with crack growth due to only fatigue showed no leaks or ruptures. The staff understands that running PWSCC crack growth and fatigue crack growth separately and getting rupture results in the PWSCC case imply that PWSCC is dominant. Discuss a sensitivity case, which compares the combined PWSCC and fatigue crack growth with the PWSCC only case, or similar evaluation in order to confirm the dominance of PWSCC.

14. Section 4.1.3 of MRP-480 investigates the 95% upper bound confidence intervals for the cases with zero occurrences of rupture with leak rate detection. Section 4.1.1 of

8 MRP-480 discusses how the 95% confidence interval was estimated, and based on the discussion in Section 4.1.2 of MRP-480, Figures 4-3 and 4-4 of MRP-480 (for results with leak rate detection and leak rate detection plus inservice inspection, respectively) show 95% upper bound confidence intervals. It is not clear to the staff the estimation approach discussed in Section 4.1.1 of MRP-480, and how that is related to the discussion of number of realizations in Section 4.1.3 of MRP-480. Clarify the discussions in Sections 4.1.1 and 4.1.3 of MRP-480. Clarify also why Figure 4-1 of MRP-480 (for the results for occurrence of rupture without crediting leak rate detection nor inservice inspection) does not have a 95% upper bound confidence interval, since some cases listed Table 3-3 of MRP-480 resulted in zero occurrences of rupture per the summary of results in Appendix A of the xLPR Piping System Analysis report (ML21217A088) and xLPR Generalization Study report (ML22088A006).

15. Section 4.1.4 of MRP-480 states that the probability of crack initiation was zero in the associated base cases for Cases 3.1.1 and 5.2.1. The staff noted in Appendix A of Ref.

9 of MRP-480 that the Probability of 1st Crack at 80 EFPY in the associated base cases are not zero. Clarify what is meant in Section 4.1.4 of MRP-480 by the statement that the probability of crack initiation was zero in the associated base cases for Cases 3.1.1 and 5.2.1.

16. Discuss the four cases in Table 4-6 of MRP-480 for the steam generator nozzles (Cases 4.1.1, 4.1.2, 4.1.3, and 4.1.4 of the xLPR Generalization Study report) relevant to ALS to help the staff understand the sensitivity runs performed that are associated with these cases.

17. There is little discussion of convergence in MRP-480. Table 4-10 of the report states that for convergence criteria (Table C-7 of Regulatory Guide 1.245), there was no direct acceptance criteria.

a. Even though there was no direct acceptance criterion, discuss how it was determined that convergence of the xLPR results was reached.

b. Table 4-4 and Appendix C of MRP-480 show comparison of Case 1.1.6 of the xLPR Piping System Analysis with 70,000 and 1,000,000 realizations. Were there other sensitivity cases like this to show convergence of results?

18. Section 5.1.5 of MRP-480 discusses reduction of fracture toughness of cast austenitic stainless steel (CASS) due to thermal aging as a potential degradation mechanism. The section states that low levels of delta ferrite present in well-controlled austenitic stainless steel welds (3% to 10%) are unlikely to lead to a significant reduction in fracture properties of CASS. To better understand what this statement means in terms of probabilities of rupture and/or LOCA, demonstrate through recent probabilistic fracture mechanics work on CASS (e.g., nonproprietary report MRP-479) in the PWR fleet that the probability of rupture and/or LOCA values in CASS piping/welds are in the same order of magnitude or lower as those shown in Figures 4-1, 4-2, and 4-3 of MRP-480.