ML12100A103
| ML12100A103 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 04/11/2012 |
| From: | Ulses A nrc/nrr/dss/srxb |
| To: | Doug Broaddus Plant Licensing Branch II |
| Parks, Benjamin 415-6472 | |
| Shared Package | |
| ML12100A133 | List: |
| References | |
| TAC ME3775 | |
| Download: ML12100A103 (10) | |
Text
April 11, 2012 MEMORANDUM TO: Douglas A. Broaddus, Chief Plant Licensing Branch II-2 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation FROM: Anthony P. Ulses, Chief /RA/
Reactor Systems Branch Division of Safety Systems Office of Nuclear Reactor Regulation
SUBJECT:
AUDIT REPORT REGARDING TENNESSEE VALLEY AUTHORITY BROWNS FERRY, UNIT 1 AREVA FUEL TRANSITION EMERGENCY CORE COOLING SYSTEM EVALUATION MODEL APPLICATION (TAC NO. ME3775)
By letter dated April 16, 2010, the Tennessee Valley Authority, the licensee for the Browns Ferry, Unit 1, requested to amend the Browns Ferry facility operating license by amending the Technical Specifications as necessary to transition to AREVA fuel. An earlier version of this report is dated August 16, 2011 (Agencywide Document Access and Management System (ADAMS) Accession No. ML112230009). This submittal was revised to correct issues identified by the fuel vendor. A proprietary version of the audit report is also being made available.
As part of this review, Benjamin Parks and Jennifer Gall of the Reactor Systems Branch staff conducted an audit of the modified emergency core cooling system evaluation model and supporting analysis supporting the Browns Ferry Unit 1 fuel transition request at the AREVA Office in Richland, Washington on July 19-21, 2011.
The audit was conducted as detailed in an audit plan issued by memorandum dated July 6, 2010 at ADAMS Accession No. ML11181A019. The plan stated that the NRC staff would prepare a detailed audit report documenting the information reviewed during the audit and any open items identified as a result of the audit. Enclosure 1 to this transmittal provides the audit report. Enclosure 2 contains a request for additional information for conveyance to the licensee.
Enclosure:
- 1. Audit Report
- 2. Request for Additional Information CONTACT: Benjamin T. Parks, NRR/DSS/SRXB 301-415-6472
ML11181A019. The plan stated that the NRC staff would prepare a detailed audit report documenting the information reviewed during the audit and any open items identified as a result of the audit. Enclosure 1 to this transmittal provides the audit report. Enclosure 2 contains a request for additional information for conveyance to the licensee.
Enclosure:
- 1. Audit Report
- 2. Request for Additional Information CONTACT: Benjamin T. Parks, NRR/DSS/SRXB 301-415-6472 DISTRIBUTION:
Nonpublic SRXB r/f HCruz, NRR EBrown, NRR AMendiola, NRR BParks, NRR AUlses, NRR JDavis, NRR JGall, NRR ADAMS ACCESSION No. ML12100A103 OFFICE NRR/DSS/SRXB NRR/DSS/SRXB NRR/DSS/SRXB/BC NAME JGall BParks AUlses DATE 4/9/12 4/9/12 4/11/12 AUDIT REPORT TENNESSEE VALLEY AUTHORITY BROWNS FERRY UNIT 1 AREVA FUEL TRANSITION EMERGENCY CORE COOLING SYSTEM EVALUATION MODEL APPLICATION
1.0 INTRODUCTION
By letter dated April 16, 2010, the Tennessee Valley Authority (TVA), the licensee for the Browns Ferry, Unit 1, requested to amend the Browns Ferry facility operating license by amending the Technical Specifications as necessary to transition to AREVA fuel. To support this request, the Reactor Systems Branch (SRXB) is reviewing the application of the emergency core cooling system (ECCS) evaluation model (EM) to Browns Ferry.
This audit was conducted to address issues identified during the staff review of the AREVA fuel transition request. During the Nuclear Regulatory Commission (NRC) staff review, ((
)) was identified in the most severe loss-of-coolant accidents analyzed by AREVA. The NRC staff was unable to conclude that ((
)) was based on adequate models.
To address the NRC staff concerns, TVA prepared a modified ECCS evaluation and intends to augment the Browns Ferry ECCS evaluation with analyses that address the staff concerns
(( )). The staff reviewed the modified ECCS evaluation and supporting analysis for the Browns Ferry Unit 1 fuel transition.
The audit focused on a review of analytic results to aid the staff in determining whether the EXEM BWR-2000 ECCS evaluation model has been applied in a manner that demonstrates (1) that the Browns Ferry-specific ECCS performance evaluation conforms to the required and acceptable features of an ECCS evaluation model set forth in Appendix K to 10 CFR 50, and (2) that the Browns Ferry ECCS evaluation appropriately demonstrates compliance with the requirements of 10 CFR 50.46.
This review activity was conducted in an audit setting for three reasons:
(1) The majority of the information that the NRC staff examined during the audit was proprietary in nature.
(2) The NRC staff requested the availability of cognizant technical analysts to discuss the information contained in the documents that the NRC staff reviewed.
Much like the information that the staff examined, these discussions were proprietary in nature.
Enclosure 1
(3) As the staff also reviewed analyses that were preliminary and unverified, these analyses and any revisions or sensitivity studies will be provided to the NRC in a single transmittal, in final, verified and quality assured format.
This audit activity enhances the review process because it facilitated timely communication without compromising documentation requirements. The audit report, and the licensees responses to additional NRC staff requests for additional information, will provide documentation of all items reviewed and discussed during the audit.
1.1 Regulatory Basis 10 CFR 50.46 requires that each pressurized light water nuclear power reactor be provided with an ECCS that must be designed so that its calculated cooling performance following postulated loss-of-coolant accidents conforms to, among other acceptance criteria, a peak cladding temperature that does not exceed 2200°F, calculated in accordance with an acceptable evaluation model. The emergency core cooling performance must be calculated for a number of postulated loss-of-coolant accidents of different sizes, locations, and other properties sufficient to provide assurance that the most severe postulated loss-of-coolant accidents are calculated.
The staffs review of the modified evaluation was conducted in accordance with NUREG 0800, Standard Review Plan (SRP) for the Review of Safety Analysis Reports for Nuclear Power Plants: LWR Edition, Section 15.0.2, Review of Transient and Accident Analysis Methods.
The SRP directs the staff to review the complete code documentation including, but not limited to: (a) the evaluation model, (b) the accident scenario identification process, (c) the code assessment, (d) the uncertainty analysis, (e) a theory manual, (f) a user manual, and (g) the quality assurance program.
SRP 15.0.2 Section III.2.D:
The reviewers should ensure that [1] all code closure relationships based in part on experimental data or more detailed calculations have been assessed over the full range of conditions encountered in the accident scenario [2] integral test assessments properly demonstrate physical and code model interactions that are determined to be important for full size plant accident scenarios [and 3] the documentation contains comparisons of all important experimental measurements with the code predictions in order to expose possible cases of compensating errors.
SRP 15.0.2 Section III.2.F:
The reviewers should confirm that the evaluation model is maintained under a quality assurance program that meets the requirements of Appendix B to 10 CFR 50. As a minimum, the program must address design control, document control, software configuration control and testing, and corrective actions. The reviewers should confirm that the quality assurance program documentation includes procedures that address all of these areas. The reviewers may conduct an audit of the implementation of the code developers quality assurance program.
2.0 AUDIT RESULTS The audit began with several presentations prepared and delivered by AREVA staff. AREVA provided a background overview of the EXEM-BWR 2000 ECCS Evaluation Model, followed by two presentations, one that summarized the issues identified in the staff audit plan, and another that described evaluation model changes and reanalyses that AREVA developed to respond to items identified in the audit plan.
((
)).
Open Item: Please provide a revised ECCS Evaluation Summary that provides a detailed description of the most severe loss of coolant accident analysis, along with a description of other break sizes, locations and other properties that were evaluated to support the determination that the most severe postulated loss of coolant accident has been calculated.
Open Item: Please provide a detailed description of the model changes made to address the staffs concern with the (( )).
2.1 Summary of Staff Activities During Audit The staff reviewed the analysis results performed according to the new method. The results indicated that the limiting recirculation line break was a 0.21 ft2 split break in the pump discharge with a battery board A failure and a top-peaked axial power shape. The limiting PCT was 1891°F. This compares to other recent results as follows:
% OLTP ADS Condition Single Failure Limiting Break Size PCT 105 Operable HPCI/LPCI 0.21 ft2 1891°F 105 Degraded HPCI/LPCI 0.5 ft2 1809°F 105 Degraded LPCI/ADS 0.25 ft2 1973°F 120 Operable HPCI/LPCI 0.5 ft2 1998°F The staff reviewed the results and in most cases did not identify any significant issues with the results. The staff identified minor issues and requested follow-up from the licensee as discussed later in this report.
Comparison of Revised Analytic Results to Prior Results The prior, degraded ADS analyses indicated that the fuel cladding would exhibit a single heatup due to the lack of steam cooling associated with the prompt depressurization and flashing
brought about by timely ADS initiation.1 However, the ADS-operable analyses indicated ((
)).
Overall, the modified analysis resulted in a change of the limiting single failure case, and a small adjustment to the limiting break size. Due to the modeled hardware modification, a PCT benefit was realized in the overall break spectrum.
Break Spectra Discrepancies Previous review activities associated with Browns Ferry requests for extended power uprates revealed that there was a discrepancy in the limiting break size among LOCA break spectra calculated by the staff, AREVA, and General Electric (Reference TVA response to staff RAI).
Because 50.46(a)(1)(i) requires licensees to analyze a number of postulated loss of coolant accidents of different sizes, locations and other properties sufficient to provide assurance that the most severe postulated loss of coolant accidents are calculated, the discrepancy caused the staff to question whether the AREVA break spectrum, the results for which were different from the staffs and General Electrics, provided the requisite assurance that the most severe postulated LOCA had been calculated.
While 105% OLTP, degraded ADS analyses had indicated reasonable agreement among the various break spectra, the analyzed restoration of the ADS appeared to re-introduce this discrepancy. Therefore, during this audit, the staff investigated this discrepancy to determine its apparent causes and safety and regulatory significance. The most readily identifiable difference between staff/GE analyses, which were consistent, and the AREVA analysis, which was discrepant, was a difference in the limiting break size. AREVA had been predicting a limiting break size of (( )), and currently predicts a limiting break size of 0.21 ft2, whereas GE and staff predicted much smaller limiting break sizes. The staff determined that a significant difference in the AREVA model was ((
)). In a consequence-driven event, the main steamlines do not isolate until the level one setpoint is reached. Until this time, the pressure regulator attempts to control reactor pressure by throttling the turbine steam admission valve closed gradually, allowing coolant to continue leaving through the main steam lines and reducing the overall reactor vessel inventory.
1 ADS initiation causes a prompt depressurization, which, for small break BWR loss of coolant accidents, can be an initiator for lower plenum flashing.
This modeling difference appears to cause the various evaluation models to differ in their predictions of the smaller break sizes. The staff requested that AREVA provide sensitivity studies of some of the break spectrum cases, including both the most severe break and a very small break to illustrate the differences. The staffs goal was to establish that, ((
)), a comparison among various break sizes would illustrate that the AREVA method was conservative - predicted a higher PCT - when analyzing a given break size, which would lead the model to provide break spectrum results that are conservative enough to provide assurance that the most severe loss of coolant accidents had been calculated, consistent with the requirements of 50.36(a)(1)(i).
Open Item: Please provide the results of a sensitivity study demonstrating the effect of the EXEM-BWR 2000 pressure control assumptions on the break spectrum. Please include results for the limiting break size as determined using the modified EXEM-BWR 2000 analyses (0.21 ft2) and for a smaller break size that would result in a delayed pressurization following a level-driven main streamline isolation.
2.2 Staff Review of Modified ECCS Evaluation Results Although the staff did not identify any significant issues with the revised evaluation results, the staff made several observations:
(1) The break spectrum for mid-peaked power shaped, split discharge breaks appeared to have abrupt temperature changes when plotted as a function of break size.
(2) For the subset of break sizes from (( )), a smaller, intermediate temperature transient was observed to occur between the heatup following core uncovery and that following lower plenum flashing.
Abrupt Temperature Changes With Break Size The staff plotted PCT vs break size for the mid-peaked pump discharge split breaks on the recirculation discharge line and observed a drop in PCT of approximately 200°F between two small breaks. This drop appeared, based on the remaining break spectrum results, to be rather abrupt. The staff requested AREVA to examine the break spectrum in this area and determine the cause of this abrupt drop in peak cladding temperature.
Open Item: Please provide a summary of the break spectrum results that includes sufficient detail to compare the break spectra for each combination of power shape, core flow statepoint, single failure, and break geometry.
Open Item: Please explain why the break spectrum results exhibit slightly discontinuous behavior in the range of break sizes from (( )). If the behavior is attributable to the use of varied closure relationships or constitutive models based on differing thermal-hydraulic behavior, please identify the specific models that are being
activated or inactivated, explain why, and provide an estimate or description of the impact on the evaluation.
Intermediate Temperature Transient Based on the staffs evaluation of the results in this break size range, the staff also identified the intermediate temperature transient identified in Item 2, above. The intermediate temperature transient is best described as follows. In most of the results presented, two cladding heatups were identified. The first heatup reached a peak, trended to a trough, at which point the heatup resumed and reached a second peak, with a nearly linear heatup trend until the quench occurred. In the cases ranging from (( )), however, the second heatup exhibited a smooth peak, followed by a slight decrease, prior to the utilization of the HUXY Appendix K heat transfer coefficients, at which point the cladding heatup resumed. See Figure 1.
The staff attempted to analyze the results to identify the cause of this intermediate temperature transient, but the calculation notebook furnished for staff audit did not contain plots of enough parameters to definitively identify the cause of the intermediate temperature transients.
Technical discussions with AREVA revealed that the analysts had also observed the intermediate temperature transient and had been investigating its cause. The investigation had not concluded by the time the staff exited the audit.
Open Item: Please determine the cause of the intermediate temperature transient observable in the plots of peak cladding temperature vs. time for the (( ))
cases and provide a summary explanation. Justify the validity of the results, given the temperature trends depicted.
Licensing Implications Based on the use of a modified ECCS evaluation model, the staff will need to determine whether the revised evaluation has been performed in accordance with the methodology contained in EMF-2361 and its approving safety evaluation. Should the modified evaluation model be found not to conform to the approved evaluation model, TVA will be requested to revise its proposed TS 5.6.5 to reflect the use of plant-specific analytic methods. This may be accomplished by appending the citation to EMF-2361 so that it includes reference to the specific Break Spectrum Report and the SE approving its use, consistent with GL 88-16.
Observation: The use of the modified EM may require treatment as a plant-specific evaluation model per GL 88-16.
Adequacy of EM Modification The staff reviewed the modification to the evaluation model to determine whether its implementation addressed the staffs concerns with ((
)) - it appeared that the modified EM would be capable of modeling a distribution of liquid flow countercurrent to the
steam flow between the bypass and fuel channels of the core model. The staff requested that TVA specifically identify those cases that did not employ the EM revision in its break spectrum results summary.
This information was provided during the audit, and ((
)). For these breaks, however, a large amount of steam flow up through the core is expected at the time of core spray initiation, such that the overall dynamics would be countercurrent flow limited. In these cases, the staff does not believe that the use of the unmodified EM is significant, because it correctly predicts the countercurrent flow-limited behavior. In fact, ((
)).
Based on this information, the NRC staff did not establish that the EM, modified or unmodified, would provide a realistic prediction of the spray distribution between the bypass and fuel channels in the core, consistent with the phenomena exhibited by SSTF.
Observation: The adequacy of RELAX to predict a realistic distribution of liquid countercurrent flow between the core and bypass is not established, but is shown to be insignificant in the Browns Ferry ECCS Evaluation Results.
This observation was based on the fact that the cladding heatup occurred during a period of spray cooling predicted by HUXY using Appendix K spray heat transfer coefficients.
REQUEST FOR ADDITIONAL INFORMATION TENNESSEE VALLEY AUTHORITY BROWNS FERRY NUCLEAR PLANT UNIT 1 REQUEST TO TRANSITION TO AREVA FUEL AND SAFETY ANALYSIS METHODS
- 1. Please provide a revised ECCS Evaluation Summary that provides a detailed description of the most severe loss of coolant accident analysis, along with a description of other break sizes, locations and other properties that were evaluated to support the determination that the most severe postulated loss of coolant accident has been calculated.
- 2. Please provide a detailed description of the model changes made to address the staff's concern with (( )).
- 3. Please provide the results of a sensitivity study demonstrating the effect of the EXEM-BWR 2000 pressure control assumptions on the break spectrum. Please include results for the limiting break size as determined using the modified EXEM-BWR 2000 analyses (0.21 ft2) and for a smaller break size that would result in a delayed pressurization following a level-driven main streamline isolation.
- 4. Please provide a summary of the break spectrum results that includes sufficient detail to compare the break spectra for each combination of power shape, core flow statepoint, single failure, and break geometry.
- 5. Please explain why the break spectrum results exhibit slightly discontinuous behavior in the range of break sizes from (( )). Identify the significant model aspects that are causing the behavior and provide an estimate or description of the impact on the evaluation.
- 6. Please determine the cause of the intermediate temperature transient observable in the plots of peak cladding temperature vs. time for the (( )) cases and provide a summary explanation. Justify the validity of the results, given the temperature trends depicted.
Enclosure 2