CNL-22-069, Response to Request for Additional Information Regarding License Amendment Request to Revise Technical Specifications to Adopt Risk-Informed Completion Times TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times - Ritstf.

Response to Request for Additional Information Regarding License Amendment Request to Revise Technical Specifications to Adopt Risk-Informed Completion Times TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times - Ritstf.

ML22182A390
Person / Time
Site:	Sequoyah
Issue date:	07/01/2022
From:	Jim Barstow Tennessee Valley Authority
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
CNL-22-069, EPID L-2021-LLA-0145
Download: ML22182A390 (6)

Text

1101 Market Street, Chattanooga, Tennessee 37402 CNL-22-069 July 1, 2022 10 CFR 50.90 ATTN: Document Control Desk U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001 Sequoyah Nuclear Plant, Units 1 and 2 Renewed Facility Operating License Nos. DPR-77 and DPR-79 NRC Docket Nos. 50-327 and 50-328

Subject:

Response to Request for Additional Information Regarding License Amendment Request to Revise Technical Specifications to Adopt Risk-Informed Completion Times TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times - RITSTF Initiative 4b (SQN-TS-20-03) EPID L-2021-LLA-0145

References:

1. TVA Letter to NRC, CNL-21-026, License Amendment Request to Revise Technical Specifications to Adopt Risk-Informed Completion Times TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times---RITSTF Initiative 4b (SQN-TS-20-03), dated August 5, 2021 (ML21217A174)

2. TVA Letter to NRC, CNL-22-023, Supplement to License Amendment Request to Revise Technical Specifications to Adopt Risk-Informed Completion Times TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times - RITSTF Initiative 4b (SQN-TS-20-03)

EPID L-2021-LLA-0145, dated April 28, 2022 (ML22118A496)

3. TVA Letter to NRC, CNL-22-034, Second Supplement to License Amendment Request to Revise Technical Specifications to Adopt Risk-Informed Completion Times TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times---RITSTF Initiative 4b (SQN-TS-20-03), dated May 13, 2022 (ML22133A238)

4. NRC electronic mail to TVA, Request for Additional Information Related to Sequoyah Nuclear Plant's LAR to adopt TSTF-505 (EPID L-2021-LLA-0145),

dated May 26, 2022 (ML22146A332)

By letter dated August 5, 2021 (Reference 1), Tennessee Valley Authority (TVA) requested an amendment to the Sequoyah Nuclear Plant (SQN), Units 1 and 2 Renewed Facility Operating License. The proposed amendment would modify Technical Specifications (TS) requirements to permit the use of Risk-Informed Completion Times (RICT) in accordance with Technical

U.S. Nuclear Regulatory Commission CNL-22-069 Page 2 July 1, 2022 Specifications Task Force (TSTF)-505-A, Revision 2, Provide Risk-Informed Extended Completion Times - RITSTF [Risk-Informed TSTF] Initiative 4b. In Reference 2, TVA submitted a supplement to the Reference 1 license amendment request (LAR). In Reference 3, TVA submitted a second supplement to the Reference 1 LAR.

In Reference 4, the Nuclear Regulatory Commission (NRC) issued a request for additional information (RAI) and requested TVA respond by July 11, 2022.

The enclosure to this submittal provides the TVA response to the RAI.

TVA has reviewed the information supporting the no significant hazards consideration and the environmental consideration that was previously provided to the NRC in the referenced LAR.

The additional information provided in this RAI response does not impact the conclusion that the proposed license amendment does not involve a significant hazards consideration. The additional information also does not impact the conclusion that there is no need for an environmental assessment to be prepared in support of the proposed amendment.

There are no new regulatory commitments contained in this submittal.

In accordance with Title 10 of the Code of Federal Regulations 50.91, Notice for Public Comment; State Consultation, a copy of this supplement is being provided to the Tennessee Department of Environment and Conservation.

Please address any questions regarding this submittal to Stuart L. Rymer, Senior Manager, Fleet Licensing, at slrymer@tva.gov.

I declare under penalty of perjury that the foregoing is true and correct. Executed on this 1st day of July 2022.

Respectfully, Digitally signed by Rearden, Pamela S Date: 2022.07.01 09:10:38 -04'00' James Barstow Vice President, Nuclear Regulatory Affairs and Support Services Enclosure Response to Request for Additional Information cc (w/ Enclosure):

NRC Regional Administrator - Region II NRC Senior Resident Inspector - Sequoyah Nuclear Plant NRC Project Manager - Sequoyah Nuclear Plant Director, Division of Radiological Health - Tennessee State Department of Environment and Conservation

Enclosure Response to Request for Additional Information NRC Introduction

Background

On August 5, 2021, the Tennessee Valley Authority (the licensee) submitted a license amendment request (LAR) to revise technical specifications to adopt risk-informed completion times in accordance with TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times - RITSTF Initiative 4b (ADAMS Accession No. ML21217A174). As part of the NRC staff evaluation of this LAR, a virtual Regulatory Audit was conducted from September 2021 through March 2022 (ADAMS Accession No. ML22108A282). The LAR was supplemented on April 28, 2022 (ADAMS Accession No. ML22118A496), and May 13, 2022 (ADAMS Accession No. ML22133A238).

Regulatory Basis Section 2.3 of Regulatory Guide (RG) 1.174, Revision 3, An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis (ADAMS Accession No. ML17317A256), describes acceptability of a probabilistic risk assessment (PRA). The PRA analysis used to support an application is measured in terms of its appropriateness with respect to scope, level of detail, conformance with the technical elements, and plant representation. RG 1.200, Revision 2 (ADAMS Accession No. ML090410014),

provides detailed guidance on technical acceptability of a PRA used to support risk-informed applications.

Request for Additional Information APLC-RAI-1 In its April 28, 2022, response to APLA Audit Question 07, the licensee explained that a revised state-of-knowledge correlation (SOKC) analysis is unnecessary because the point estimate values are within 2 percent of the calculated mean values from the current one top multi-hazard model (OTMHM) seismic PRA (SPRA) model. The licensee provided a comparison of seismic core damage frequency (SCDF) and seismic large early release frequency (SLERF) between the mean and point estimate values in the SPRA submittal made in its October 18, 2019, response to the NRCs post-Fukushima actions (ADAMS Accession No. ML19291A003) and current OTMHM SPRA. The results show that the mean and point estimate values are within 2 percent of each other in the OTMHM SPRA model, while these values are different by more than 2 times in the SPRA submittal in response to the NRCs post-Fukushima actions. The licensee explained that this substantial change was due to the correction of an error in the quantification of the SPRA stating that the error was made by not correctly setting a single parameter used by the UNCERT computer code (@POINTCALC) in calculating mean values, and that the error significantly affected how the mean values were calculated. However, the licensee did not provide any details on the error, the basis for its substantial impact on the quantification results, and why such an impact is expected due to correction of the error.

In addition, the NRC staffs review and discussion during the regulatory audit for the LAR appears to indicate that the OTMHM model does not support uncertainty analysis. Further, in Section 5 of Enclosure 2 to the LAR, the licensee stated that the SQN SPRA Facts and Observations (F&O) Independent Assessment and Focused-Scope Peer Review was performed from February 4 - 8, 2019. Since the licensee submitted its SPRA in response to NRCs post CNL-22-069 E1 of 4

Enclosure Fukushima actions on October 18, 2019, it appears that the change made to correct the quantification error was not part of the SPRA that was peer reviewed.

The NRC staffs experience reviewing SPRAs is that the SPRA mean values and point estimates are appreciably different, primarily due to the uncertainty in the hazard curve and, in certain cases, select fragilities. For example, Case numbers 18 and 19 in Table 5-7 of the licensees October 18, 2019, response to the NRCs post-Fukushima actions demonstrates the large difference in SCDF and SLERF is due to the uncertainty in the seismic hazard. An additional example is a recent peer reviewed SPRA supporting a 10 CFR 50.69 application, which demonstrated that the uncertainty mean is about 30 percent higher than the point estimate for both SCDF and SLERF.

Based on above observations, please address the following:

a) Provide an explanation of the error that was corrected which resulted in the changes to the mean and point estimate SCDF and SLERF values in the OTMHM SPRA model compared to the SPRA submitted in response to the NRCs post-Fukushima actions.

The explanation should discuss why, given the uncertainty in the seismic hazard, the correction of the error resulted in a large change in the mean value and why such a change is to be expected.

b) Clarify whether the OTMHM model supports quantification of parametric uncertainty for the seismic PRA, including the uncertainty in the seismic hazard curve as well as the fragility groups. If it does, explain the process for parametric uncertainty analysis, including the uncertainty in the seismic hazard curve, employed for the OTMHM model compared to the SPRA submitted in response to the NRCs post-Fukushima actions. If it does not, provide an explanation of how the mean and point estimates can be compared using the OTMHM SPRA model.

c) Justify why the change does not constitute a PRA upgrade per the definition in the 2009 ASME/ANS PRA Standard, as endorsed in RG 1.200, Revision 2, and therefore, require a focused-scope peer review.

TVA Response Response to Part a)

The error in the seismic Probabilistic Risk Assessment (SPRA) submittal relates to the configuration of the model database file that the uncertainty software uses to perform the parametric uncertainty analysis. There is a variable in the database that must be changed to allow the uncertainty software to sample the seismic hazard curve and propagate the sample to the discrete seismic hazard bin initiating event basic events. Similarly, the variable must be properly set in order to sample the structures, systems, and components (SSC) fragilities and propagate the sample to the discrete seismic hazard bin fragility basic events. In addition to this change, model refinements reduced the seismic cored damage frequency (CDF) and the seismic large early release frequency (LERF).

In reviewing the calculations performed to respond to APLA Audit Question 07, it was determined that a second error was made in preparing the computer-aided fault tree analysis (CAFTA) database for the UNCERT calculations. According to the FRANX 4.4 User Manual:

CNL-22-069 E2 of 4

Enclosure There are at least two potential approaches to using the @POINTCALC switch to prepare for parametric uncertainty sampling. The two obvious methods are mentioned below (each involves using UNCERT on pre-solved cutsets, a typical CAFTA parametric uncertainty process):

1. The first method is to begin by changing the value of @POINTCALC to 0 and then re-quantify the SPRA model, resulting in a different cutset file than the base point estimate quantification. UNCERT parametric uncertainty is then performed on the new cutset file.

2. In the second method you begin with the SPRA point estimate cutset file (i.e., the cutset file produced with @POINTCALC set to 1). You then adjust the value of the

@POINTCALC Type Code to 0 in the associated database (invoking the seismic interval initiator and fragility basic events sampling equations) and then use the CAFTA cutset editor to reload this database into the cutset file. UNCERT parametric uncertainty is then performed on the cutset file.

While the @POINTCALC variable was set to 0, the SPRA was not requantified with the updated database (method 1 above) and the CAFTA cutset editor was not used to reload this database into the cutset file (method 2 above). This led to the failure to sample the seismic fragility basic events and seismic bin (initiator) basic events. Since discovery of this error, mean seismic CDF and LERF for each unit have been recalculated and are summarized below:

Seismic PRA Point Estimate Mean Percent Change U1 CDF 4.04E-06 5.61E-06 39%

U1 LERF 3.11E-06 4.02E-06 29%

U2 CDF 3.84E-06 5.48E-06 42%

U2 LERF 2.90E-06 3.70E-06 27%

These results are consistent with what the NRC cited in the RAI as a recent peer reviewed SPRA supporting a 10 CFR 50.69 application, which demonstrated that the uncertainty mean is about 30 percent higher than the point estimate for both seismic CDF and seismic LERF.

The table below shows the updated comparison of the point estimate values to the mean values for the all-hazards (Internal events, internal flooding, internal fire, and seismic) PRA model. It also shows that the mean U1 and U2 CDF and LERF values for the all-hazards PRA model are still below the NRC Safety Goals. Given this, the SOKC is unimportant (i.e., the risk results are below the acceptance guidelines) and does not need to be incorporated into the calculations for the RICT calculations.

All-Hazards Point Percent NRC Safety Mean PRA Estimate Change Goal U1 CDF 7.11E-05 7.36E-05 3% 1.00E-04 U1 LERF 8.12E-06 9.13E-06 12% 1.00E-05 U2 CDF 7.54E-05 7.76E-05 3% 1.00E-04 U2 LERF 8.91E-06 9.74E-06 9% 1.00E-05 CNL-22-069 E3 of 4

Enclosure TVA Response to Part b)

The OTMHM model does support quantification of parametric uncertainty for the seismic PRA, including the uncertainty in the seismic hazard curve as well as the fragility groups.

The process is the same as for the SPRA model submitted in response to the NRCs post-Fukushima actions. This process is described in Section 5.6 of the response to the NRCs post-Fukushima actions (ADAMS Accession No. ML19291A003). The only difference germane to the parametric uncertainty analysis is which software tool was used to produce the electronic cutset file that the uncertainty software utilizes. Outside of that difference, the process to perform the parametric uncertainty analysis (including the need to properly configure the variable as described in the answer to part a) is the same.

Parameter uncertainty relates to the uncertainty in the computation of the input parameter values used to quantify the model (i.e., initiating event frequencies, component failure probabilities and Human Error Probabilities (HEPs)). These uncertainties can be characterized by probability distributions that relate to the degree of belief in their values. A formal propagation of uncertainty is the best way to correctly account for this, and the PRA software UNCERT has the capability to propagate these uncertainties. The uncertainty analysis was performed with UNCERT 10.0, using Monte Carlo sampling and ACUBE processing.

The UNCERT analysis included distributions for seismic bin frequencies. Those distributions were generated by the FRANX code assuming lognormal distributions and estimating error factors (EFs) from the various seismic hazard curves inputted to the code (16th, median, and 84th). Sampling of the individual seismic bin frequencies was performed using the correlated approach.

Seismic failure probability distributions are determined automatically by FRANX given the fragility parameter estimates (Am, R, and U). The seismic failure probability of all components within a given fragility group are correlated in this process. Distributions for HEPs and combination factors were calculated in the Human Reliability Analysis (HRA) Calculator Version 5.2. Distributions for internal events PRA (IEPRA) basic events were left unchanged from the IEPRA model. Component failure random probabilities are correlated though the use of type codes for similar components.

TVA Response to Part c)

This is not a PRA upgrade because no new methodology has been utilized in the parametric uncertainty analysis. The only change was to properly configure the software to allow sampling of the seismic hazard and fragilities and propagation to the associated discrete seismic hazard bin scenario basic events. This was a software configuration error.

CNL-22-069 E4 of 4