ML22292A272
| ML22292A272 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 10/19/2022 |
| From: | Kimberly Green Plant Licensing Branch II |
| To: | Wells R Tennessee Valley Authority |
| References | |
| L-2022-LLA-0049 | |
| Download: ML22292A272 (10) | |
Text
From:
Kimberly Green Sent:
Wednesday, October 19, 2022 12:07 PM To:
Wells, Russell Douglas Cc:
Victor, William Ross; Eckermann, Jeremy Beau
Subject:
Request for Additional Information Related to TVA's Request to Adopt TSTF-505 and TSTF-439 for Browns Ferry Nuclear Plant, Units 1, 2, and 3 (EPID L-2022-LLA-0049)
Attachments:
Browns Ferry TSTF 505 Final RAI.pdf
Dear Mr. Wells,
By application dated March 31, 2022 (Agencywide Documents and Access Management System Accession No. ML22090A287), the Tennessee Valley Authority (TVA) submitted a license amendment request for Browns Ferry Nuclear Plant (Browns Ferry), Units 1, 2, and 3.
Specifically, TVA requested to revise certain technical specifications (TSs) to permit the use of a risk-informed completion time (RICT) for actions to be taken when a limiting condition for operation (LCO) is not met, based on Technical Specifications Task Force (TSTF) Traveler, TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times. In addition, TVA requested to implement Traveler TSTF-439, Eliminate Second Completion Times Limiting Time From Discovery of Failure to Meet an LCO.
By letter dated June 14, 2022 (ML22147A137), the U. S. Nuclear Regulatory Commission (NRC) staff transmitted an audit plan to TVA stating its plan to conduct a regulatory audit to examine TVAs non-docketed information with the intent to gain understanding, to verify information, and to identify information that will require docketing to support the basis of the licensing or regulatory decision.
Based on discussions during the audit meeting, the NRC staff determined that additional information would need to be placed on the docket to support the staffs regulatory decision regarding the application. A draft request for additional information (RAI) was previously transmitted to Mr. Eckermann on October 13, 2022. At TVAs request, a clarification call was held on October 19, 2022, to clarify the NRC staffs draft RAI. As a result of the clarification call, no changes were needed to the RAI.
A response to the attached RAI is requested no later than 45 days from the date of this email.
The NRC staff considers that timely responses to RAIs help ensure sufficient time is available for staff review and contribute toward the NRCs goal of efficient and effective use of staff resources. If circumstances result in the need to revise the requested response date, please contact me at (301) 415-1627 or via email at Kimberly.Green@nrc.gov.
Sincerely, Kimberly Green, Sr. Project Manager Plant Licensing Branch II-2 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation
Hearing Identifier:
NRR_DRMA Email Number:
1814 Mail Envelope Properties (DM6PR09MB5462AFEF2B8E00A4407B5A138F2B9)
Subject:
Request for Additional Information Related to TVA's Request to Adopt TSTF-505 and TSTF-439 for Browns Ferry Nuclear Plant, Units 1, 2, and 3 (EPID L-2022-LLA-0049)
Sent Date:
10/19/2022 12:07:15 PM Received Date:
10/19/2022 12:07:00 PM From:
Kimberly Green Created By:
Kimberly.Green@nrc.gov Recipients:
"Victor, William Ross" <wrvictor@tva.gov>
Tracking Status: None "Eckermann, Jeremy Beau" <jbeckermann@tva.gov>
Tracking Status: None "Wells, Russell Douglas" <rdwells0@tva.gov>
Tracking Status: None Post Office:
DM6PR09MB5462.namprd09.prod.outlook.com Files Size Date & Time MESSAGE 2342 10/19/2022 12:07:00 PM Browns Ferry TSTF 505 Final RAI.pdf 130802 Options Priority:
Normal Return Notification:
No Reply Requested:
No Sensitivity:
Normal Expiration Date:
LICENSE AMENDMENT REQUEST TO ADOPT TSTF-505, REVISION 2 REGARDING RISK-INFORMED COMPLETION TIMES AND TSTF-439 REGARDING ELIMINATION OF SECOND COMPLETION TIMES REQUEST FOR ADDITIONAL INFORMATION BROWNS FERRY NUCLEAR PLANT UNITS 1, 2, AND 3 DOCKET NOS. 50-259, 50-260, 50-296 By application dated March 31, 2022 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML22090A287), the Tennessee Valley Authority (TVA) submitted a license amendment request (LAR) for Browns Ferry Nuclear Plant (Browns Ferry or BFN), Units 1, 2, and 3. The proposed amendments would revise certain technical specifications (TSs) to permit the use of a risk-informed completion time (RICT) for actions to be taken when a limiting condition for operation (LCO) is not met. The proposed changes to implement RICT are based on a Technical Specifications Task Force (TSTF) Traveler, TSTF-505, Revision 2, Provide Risk-Informed Extended Completion Times (ML18183A493).
The LAR also proposed to implement TSTF-439, Eliminate Second Completion Times Limiting Time From Discovery of Failure to Meet an LCO (ML21272A357).
By letter dated June 14, 2022 (ML22147A137), the U. S. Nuclear Regulatory Commission (NRC) staff transmitted an audit plan to TVA stating its plan to conduct a regulatory audit to examine TVAs non-docketed information with the intent to gain understanding, to verify information, and to identify information that will require docketing to support the basis of the licensing or regulatory decision.
By emails August 19, 2022 (ML22238A065), August 24, 2022 (ML22238A067), September 13, 2022 (ML22259A037), and September 14, 2022 (ML22259A038 and ML22259A039), the NRC staff transmitted audit questions in advance of and during the audit meeting that was conducted the week of September 12, 2022. Based on discussions during the audit meeting, the NRC staff determined that additional information would need to be placed on the docket to support the staffs regulatory decision regarding the application. Below are the requests for the additional information.
RAI 1 (APLA/APLC) - PRA Model Uncertainty Analysis for Updated PRA Models (Audit Question 3)
Section 2 of Enclosure 5 to the LAR identifies the model of record (MOR) for each probabilistic risk assessment (PRA) used in the application. For the internal events PRA it is MOR 10, for the fire PRA it is MOR 6, and for the seismic PRA it is MOR 1. The NRC staff reviewed the portal documents provided and it appears that some uncertainty analyses were conducted using earlier models. The NRC staff notes that an MOR update can significantly affect the risk results.
It may raise the importance of a previously analyzed source of uncertainty that was not significant to the point where it becomes a key source of uncertainty. A model update can also introduce a new source of uncertainty.
2
- a.
Clarify how the uncertainty analyses that the staff has audited reflect the MORs used to support this application.
- b.
For any PRA hazard model for which the MOR used for this application is different from the model subjected to the last uncertainty review, justify the conclusion that the difference introduces no new key source of uncertainty relevant to this application.
Similarly, confirm the conclusion provided in Section 5 of Enclosure 9 of the LAR, based on any updated MOR uncertainty review, that the adjustments to the MOR to create the OTMHM introduce no key source of uncertainty relevant to RICTs.
RAI 2 (APLA) - In-Scope LCOs and Corresponding PRA Modeling (Audit Question 4)
The NRCs safety evaluation for NEI 06-09 specifies that the LAR should provide a comparison of the TS functions to the PRA modeled functions to show that the PRA modelling is consistent with the licensing basis assumptions or to provide a basis for when there is a difference. Table E1-1 in Section 1 of Enclosure 1 to the LAR identifies each LCO in the TSs proposed for inclusion in the RICT program. The table also describes whether the systems and components covered by the LCO are modeled in the PRA and, if so, presents both the design success criteria (DSC) and PRA success criteria. For certain LCOs, the table explains that the associated structures, systems, and components (SSCs) are not modeled in the PRAs but will be represented using a surrogate event that fails the function performed by the SSC. For some LCOs, the LAR did not provide an adequate description for the NRC staff to conclude that the PRA modeling will be sufficient.
- a.
Regarding TS 3.3.5.1.B, Table E1-1 states that the suppression pool valves will be used as surrogates for the Reactor Vessel Level Low - Level 0 instrument signal (Function 2.e). It is unclear to the NRC staff which systems are affected by the Function 2.e signal.
- i.
Describe the systems that are affected by the Function 2.e signal. In this description, address each configuration of the systems and how they are affected.
ii. Explain how the suppression pool surrogate bounds these effects.
- b.
Regarding TS 3.3.6.1.A, Table E1-1 states for Function 3 (HPCI isolations), Function 4 (RCIC isolations), Function 5 (RWCU isolations), and Function 6 (SDC isolations) that one of the modeled pathways will be used as a surrogate. It is unclear to the NRC staff which pathways will be used for each of these functions.
- i.
Clarify which modeled pathways will be used as a surrogate for each of the system isolation functions.
ii. Explain how the surrogate bounds each of the isolation functions.
- c.
Regarding TS 3.6.1.2.C, Table E1-1 states that one of the modeled pathways will be used as a surrogate when one of the primary containment airlocks is inoperable. Indicate the impact of this surrogate on large early release calculations compared to the airlock.
- i.
Briefly describe the effect of the failure of early containment isolation (i.e., plant response to the failure of the modeled pathway).
ii. Explain how this bounds the effect of an inoperable containment airlock door.
3
- d.
Regarding TS 3.6.1.3.A, Table E1-1 states that, for the valves not modeled, a pathway that is modeled will be used as a surrogate. It is unclear to the NRC staff which pathways will be used for each affected function.
- i.
Clarify which modeled pathways will be used as a surrogate for each of the system isolation functions affected.
ii. Provide justification that the surrogate bounds each of the isolation functions.
RAI 3 (APLA/APLC) - Update of Fire and Seismic PRAs with the resolution of Internal Event F&Os (Audit Question 5a)
Regulatory Guide 1.200, An Approach for Determining the Technical Adequacy of Probabilistic Risk Assessment Results for Risk-Informed Activities, Revision 2 (ML090410014) recommends a peer review of the PRA model and its results. The primary results of peer review are the facts and observations (F&Os) recorded by the peer review team and the subsequent resolution of these F&Os using the closure process documented in NEI 05-04/07-12/12-06 Appendix X, Close-out of Facts and Observations (F&Os) (ML17086A431), that the NRC has accepted for use.
The NRC staff notes that implementation of risk-informed categorization of SSCs under 10 CFR 50.69 previously has been authorized by license amendments issued to Browns Ferry (ML21173A177). A prerequisite to implementation of these amendments is that the resolutions to the internal events findings with the potential to impact the FPRA [fire PRA] and SPRA
[seismic PRA] modeling will be incorporated into the FPRA and SPRA. Enclosure 2 to the LAR does not discuss the resolution of internal events F&Os in either the fire or seismic PRA models.
It is not clear to the NRC staff whether the PRA models to be used in RICT calculations have incorporated all relevant F&O resolutions.
Confirm that all internal events PRA modeling updates performed to resolve F&Os that could impact fire or seismic risk were incorporated into the fire and seismic PRA models.
RAI 4 (APLA) - Risk Calculation Based on the Model of Record (Audit Question 6a)
Regulatory Guide 1.174, Revision 3, An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis (ML17317A256),
provides risk acceptance criteria for total core damage frequency (CDF 1E-04/year) and large early release frequency (LERF 1E-05/year). Section 6.4 of NUREG-1855, Revision 1, Guidance on the Treatment of Uncertainties Associated with PRAs in Risk-Informed Decision Making (ML17062A466), states that for a capability category II risk evaluation, the mean values of the risk metrics (total risk and change in risk) are evaluated against these guidelines. More specifically, these metrics are the means of the probability distributions that result from the propagation of input parameter uncertainties and model uncertainties explicitly reflected in the PRA models.
In general, the point estimates of core damage frequency (CDF) and large early release frequency (LERF) obtained by quantification of the cutset probabilities using mean values for each basic event probability do not represent true mean values of the CDF and LERF. Under certain circumstances, a formal propagation of uncertainty may not be required if it can be demonstrated that the state-of-knowledge correlation (SOKC) is unimportant (i.e., the risk results are well below the acceptance guidelines).
4 Table E5-1 of Enclosure 5, Section 2 of Attachment 1 to the LAR presents point estimates of CDF and LERF based on the optimized one-top multihazard model. For each of the three units, values for total risk are listed along with risk contributions from internal events (which includes internal flooding), internal fire, and seismic events. These reported values did not match the point estimates developed from the model of record.
When quantification explicitly addresses the SOKC, point estimates are likely to be lower than the mean risk values. The total LERF is reported as 9.08E-06/year, 8.63E-06/year, and 8.53E-06/year for Browns Ferry, Units 1, 2, and 3, respectively. When both model updates to resolve F&Os and the SOKC are considered, the total risk could approach the criteria from Regulatory Guide (RG) 1.174.
Demonstrate that the total risk for Units 1, 2, and 3 meet RG 1.174 risk acceptance guidelines for CDF and LERF. Mean values based on the model of record should be appropriately combined to include risk from internal events, seismic hazards, and fire as well as accounting for the SOKC. In this demonstration, account for changes in risk due to PRA model updates needed in response to NRC staff information requests or previous commitments and include identification of the parameters that are assumed to be correlated in the parametric uncertainty analysis of fire and seismic events.
RAI 5 (APLA) - Consistency of the Results Produced by the CRM Tool With the PRA Model Of Record (Audit Question 7a)
Regulatory Guide 1.174, Revision 3, recommends that the level of detail in the PRA should be sufficient to model the impact of the proposed licensing basis change. The characterization of the problem should include establishing a cause-effect relationship to identify portions of the PRA affected by the issue being evaluated. Full-scale applications of the PRA should reflect this cause-effect relationship when quantifying the effect of the proposed licensing-basis change on the PRA elements.
Section 4.2 of NEI 06-09-A, Risk-Managed Technical Specifications Guidelines, (ML12286A322) describes attributes of the tool used for configuration risk management (CRM).
The PRA models are normally transformed to create a real-time risk (RTR) model suitable for use with the CRM tool. Item 3 of Section 2.3.5 of the NEI guidance document states, The following specific CRM tool attributes are required for [Risk Managed Technical Specifications]
RMTS implementation: Appropriate benchmarking of results from the CRM tool against the PRA model shall be performed to demonstrate consistency.
The staff reviewed how the one-top multihazard model (OTMHM) was integrated with the Phoenix model and how example evaluations were used to check the Phoenix model. However, it is not clear how results using the CRM tool were directly benchmarked against results from the PRA MOR. The NRC staff noted that benchmarking was performed between the individual PRA hazard models on one hand and the OTMHM on the other, with each hazard quantified separately. Total CDF and total LERF were quantified for each unit, using the OTMHM.
Clarify how the benchmarking activities performed can confirm consistency of the RTR model results with the results of each PRA model of record.
RAI 6 (STSB) - Justification for ECCS Instrumentation Variation (Audit Question 11)
5 Attachments 2.1, 2.2, and 2.3 to the LAR contain the proposed TS markups for Browns Ferry, Units 1, 2, and 3, respectively. In proposed TS 3.3.5.1, ECCS Instrumentation, each of the Required Actions proposed for the RICT program includes a loss of function in Conditions B, C, E, F, and G. Per TSTF-505, Revision 2, loss of function is a prohibited condition. Provide a justification for this variation.
RAI 7 (STSB) - Apparent Loss of Function for TS Conditions (Audit Question 12a-d)
Table E1-1 of Enclosure 1 to the LAR includes descriptions of the DSC of TSs proposed for inclusion in the RICT program. Given the inoperable equipment in the TS condition, the DSC represents the minimum set of remaining credited equipment sufficient to perform TS safety function while in an RICT. The following TS conditions appear to include a loss of function.
Clarify the associated DSC in Table E1-1:
a) TS 3.3.5.1.E (One or more channels inoperable for Function 3.f, HPCI Pump Discharge Flow - Low (Bypass))
b) TS 3.5.1.C (HPCI System inoperable) c) TS 3.5.1.D (HPCI System inoperable and Condition A entered) d) TS 3.5.3.A (RCIC System inoperable)
RAI 8 (EEEB) - Other Comments (Audit Question 13)
Table E1-1 of Enclosure 1 to the LAR lists the TS LCO Conditions to which the RICT Program is proposed to be applied and documents the certain information regarding the TSs with the associated safety analyses, the analogous PRA functions, and the results of the comparison.
TS Condition 3.3.8.1.A The TS Condition states: One degraded voltage relay channel inoperable on one or more shutdown board(s). AND The loss of voltage relay channels on the affected shutdown board(s) are OPERABLE.
Under the column Other Comments it states, Not explicitly modeled. Undervoltage DG start relays are used as surrogates. This is conservative because one of the two diesel start signals are failed.
Provide an explanation of Other Comments, preferably with an example, considering that the TS Condition description applies to one or more shutdown board(s).
RAI 9 (EEEB) - Availability of Temporary DGs (Audit Question 14)
TS Condition 3.8.1.B (U1 & U2) and TS 3.8.1.B (U3)
The TS Conditions state, One required Unit 1 and 2 [diesel generator] DG inoperable, and One required Unit 3 DG inoperable.
The current Competition Time varies from 7 days to 14 days depending on the availability of Temporary/Supplemental diesel generators (TDG).
Clarify whether credit of availability of TDG(s) is or will be considered in the calculation of RICT of DGs.
6 RAIs 10-15 The following applies to RAIs 10 through 15 Table E1-1 of Enclosure 1 to the LAR includes descriptions of the design success criteria (DSC) of TSs proposed for inclusion in the RICT program. Column 2, TS Condition Description, lists the current TS Condition for which an RICT is being proposed, Column 3, SSCs Covered by TS LCO Condition, describes the SSCs addressed by each action requirement, and Column 6, Design Success Criteria, provides a summary of the success criteria from the design basis analyses. Regarding the information in these columns for certain TS Conditions, the NRC staff requests the following information.
RAI 10 (EEEB) - TS 3.8.1, Condition F (U1 & U2) and Condition F (U3) (Audit Question 18)
Column 2 of Table E1-1 for TS 3.8.1, Condition F, states One required offsite circuit inoperable.
AND One required Unit 1 and 2 [or Unit 3] DG inoperable. Column 6 states that the DSC is One offsite AC power source AND Three of four U1 and U2 [or U3] DGs. These two statements seem to be inconsistent with statements in Table E1-1 for TS 3.8.1 Conditions A and B, and TS 3.8.1 Required Actions (RAs) F.1 and F.2, which require either power source for safe shutdown Explain the minimum required alternating current (AC) sources needed to fulfill the required safety function.
RAI 11 (EEEB) - TS 3.8.4, Condition A (U1 & U2) and Condition A (U3) (Audit Question 19)
Column 2 for TS 3.8.4.A (U1 & U2) states One Unit DC electrical power subsystem inoperable.
OR One Unit 1 and 2 Shutdown Board DC electrical power subsystem inoperable, whereas column 3 lists the SSCs covered by the LCO as Three 250 VDC Unit subsystems AND Four 250 VDC shutdown board subsystems AND Four Unit 2 and Two Unit 3 DG DC subsystems.
These two statements appear to be inconsistent. Similarly, columns 2 and 3 of TS 3.8.4, Condition A for Unit 3 seem to have same inconsistency.
- a.
Explain why column 3 is inconsistent for SSCs listed in column 2. Inoperable DG DC subsystems are addressed by TS 3.8.4, Condition C (e.g., U1 TS).
- b.
Explain why column 6 refers to an inoperable direct current (DC) subsystem in reference to 250 VDC shutdown boards, which has no impact on 250 VDC shutdown boards since they are powered by their own batteries according to Browns Ferry Updated Final Safety Analysis Report Section 8.5.3.5, Distribution System (ML21286A411).
RAI 12 (EEEB) - TS 3.8.7, Condition A (U1 & U2) and Condition A (U3), and TS 3.8.7, Condition E (U1) and Condition F (U3) (Audit Question 20)
In the LAR, the TS Bases for TS 3.8.7, Action A.1 (page B 3.8-89) states, in part, that however, because another single failure in the remaining three 4.16 kV shutdown boards could result in the minimum required ESF functions not being supported. Therefore, the 4.16 kV shutdown board must be restored to OPERABLE status within 5 days. However, column 6 of Table E1-1 for TS Condition 3.8.7.A (U1 & U2) lists the DSC as One of two divisions.
Additionally, TS Condition 3.8.1.B (U1 & U2) in Table E1-1, column 6 lists the DSC for one required Unit 1 and 2 DG (1 of 4 DGs) inoperable as Three of four U1 and U2 DGs. The DSC
7 for TS 3.8.7.A appears to be inconsistent with the information in the TS Bases and the DSC for TS 3.8.1.B. Similarly, TS 3.8.4, Condition A for Unit 3 seems to have the same inconsistency.
Explain why a minimum of three of four shutdown boards are not listed as the DSC (for safe shutdown) for TS Condition 3.8.7.A.
RAI 13 (EEEB) - TS 3.8.7, Condition B (U1, U2, U3) (Audit Question 21)
TS Condition 3.8.7.B in Table E1-1 states One Unit 1 480 V Shutdown Board inoperable. OR 480 V (reactor motor operated valve (RMOV)) Board 1A inoperable. However, column 3 for the entry does not list the RMOV board as an SSC cover by the LCO Condition.
Explain why column 3 does not list the RMOV board available per unit identified specifically in the LCO.
RAI 14 (EEEB) - TS 3.8.7, Condition D (U1) and TS 3.8.7, Condition E (U2 or U3) (Audit Question 22)
- a.
Column 3 for TS Condition 3.8.7.D in Table E1-1 states, in part, Five Shutdown Board 250 V DC electrical power distribution subsystems and Column 6 states, in part, Three of four Unit 1 & 2 Shutdown Boards. These statements appear to be inconsistent.
Additionally, The TS Bases on page B3.8-58 refers to four 250 V DC shutdown boards for Units 1 and 2, and TS 3.8.4, Condition A (U3) in Table E1-1 refers to four shutdown boards.
Clarify the correct number of 250 VDC shutdown boards needed for the DSC.
- b.
Column 2 for TS 3.8.7.E (U3) in Table E1-1 refers to the shutdown boards for Unit 3; however, column 6 for the TS condition states, in part, Three of four Unit 1 & 2 Shutdown Boards.
Explain why column 6 for TS 3.8.7.E (U3) includes only the U1 and U2 shutdown boards and not the U3 shutdown boards.
RAI 15 (EEEB) - TS 3.8.7, Condition F (U2) (Audit Question 23)
In Attachment 2.2 to the LAR, there is a markup for TS 3.8.7, Condition F for Unit 2.
Additionally, there is an entry for TS 3.8.7, Condition F in LAR Table E1-3. However, there is not a corresponding entry in LAR Table E1-1.
Explain why there is not a line item for Unit 2 in LAR Table E1-1.
RAI 16 (STSB) - (Audit Question 24)
In attachments 2.1, 2.2, and 2.3 of the LAR, TVA proposed to add an RICT to Browns Ferry, Units 1, 2, and 3, TS 3.8.1 Required Action B.5 for one inoperable DG. Currently, Required Action B.5 has Completion Times with nested logic related to the availability of the TDG. The addition of the RICT to this set of Completion Times, as proposed, has the potential to be a confusing requirement.
Discuss revising the nesting of the and logical connectors for clarity.
8 RAI 17 (APLA/APLC) - Difference Between CDF and LERF Mean and Point Estimate Values (Audit Question 25)
During the audit discussions, TVA noted that it updated the values for CDF and LERF from internal events, fire, and seismic PRA for all three units. Based on the updated values, the NRC staff noted that the mean values are greater than the point estimate of these quantities for all cases considered, and the differences are greater than 2 percent. Section 1 of Enclosure 9 to the LAR states, A comparison of CDF results from the normal model quantification and the parametric uncertainty evaluation shows that the difference between the point estimate and mean results is less than 2%. Therefore, SOKC is not a key source of uncertainty for the TSTF-505 application. The cited statement appears to contradict statements made by TVA in response to Audit Question 6.
Explain this apparent contradiction or correct the cited statements to make them consistent with the updated PRA results.