ML16203A404

Westinghouse Best-Estimate Loss-Of-Coolant Accident Analysis Methodology

ML16203A404
Person / Time
Issue date:	02/23/1996
From:	Banks M Advisory Committee on Reactor Safeguards
To:
References
Download: ML16203A404 (3)

Text

February 23, 1996 The Honorable Shirley Ann Jackson Chairman U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001

SUBJECT:

WESTINGHOUSE BEST-ESTIMATE LOSS-OF-COOLANT ACCIDENT ANALYSIS METHODOLOGY During the 428th meeting of the Advisory Committee on Reactor Safeguards, February 8-10, 1996, we reviewed the best-estimate, large-break, loss-of-coolant accident (LBLOCA) analysis methodology developed by the Westinghouse Electric Corporation. During this review, we had the benefit of discussions with representatives of the NRC staff, Westinghouse, Idaho National Engineering Laboratory, and several nuclear power plant licensees. Our Subcommittee on Thermal Hydraulic Phenomena has held a number of meetings on this matter as far back as 1991. The last meeting of the Subcommittee concerning this issue was held on January 18-19, 1996. We also had the benefit of the referenced documents.

Westinghouse has developed an improved method to evaluate the performance of emergency core cooling systems (ECCS) for the case of a LBLOCA in three-and four-loop pressurized-water reactors (PWRs) of Westinghouse design. Westinghouse has proposed that this improved method, based on the use of the WCOBRA/TRAC code, be accepted for routine use in demonstrating that the cores in these plants meet NRC licensing requirements pursuant to the revised ECCS Rule (10 CFR 50.46). The NRC staff has reviewed this proposal and has concluded that the new methodology can be used for licensing calculations. We concur with the staff; however, some improvements in the uncertainty analysis are desirable.

The improved method of analysis takes advantage of data and the understanding of thermal-hydraulic behavior developed during the past two decades. This method will reduce the conservative margins in the calculated peak cladding temperature that result from the use of current methods based on Appendix K. This will permit licensees of Westinghouse three-and four-loop PWRs to have greater flexibility in the operation of their plant reactor cores and in associated fuel management practices. We also believe that, when properly documented, the improved method will provide a straight-forward and understandable assessment of the performance of an important safety system.

The improved LOCA evaluation method makes use of realistic values for inputs and correlations rather than the conservatively biased

values used in the past. To meet licensing requirements, empiri-cally based uncertainty distributions for each of the important inputs and correlations are used and propagated through the solution algorithm, WCOBRA/TRAC, to obtain estimates of uncertainty distributions for the peak cladding temperature. A nominal 95 percent probability of nonexceedance is required for licensing purposes. Questionable models or correlations are adjusted to ensure that their predictions are conservative. Westinghouse expanded the Code Scaling, Applicability, and Uncertainty (CSAU) evaluation methodology outlined in NRC Regulatory Guide 1.157, by including additional parameters not considered during the earlier CSAU exercise conducted by the NRC staff.

We have some concerns about the Westinghouse best-estimate LBLOCA evaluation methodology. The method used by Westinghouse to obtain the heat transfer coefficient uncertainty distribution resulted in some high values that are nonphysical. Westinghouse should reevaluate the heat transfer uncertainty distribution with appropriate consideration of the dependencies on physical parame-ters such as reflood rate. The Westinghouse treatment of the minimum wetting (or rewetting) temperature is not satisfactory because the correlation ignores important phenomena and could lead to nonconservative results. The existence of compensating errors in WCOBRA/TRAC may be a reason for the skewed distribution in heat transfer coefficients. We believe that these concerns should be addressed.

Obtaining adequate documentation in a timely manner has been a problem from the outset of this review. This has unnecessarily complicated the reviews by both the NRC staff and the ACRS.

Westinghouse has committed to provide documentation that will clearly lay out its LBLOCA methodology. We believe that the staff should review this final documentation prior to approving use of the improved methodology. The staff should also prepare guidelines for documentation of future best-estimate LOCA submittals before the lessons learned from this review are forgotten.

It is important to realize that the deficiencies seen in codes like TRAC and RELAP may preclude their extension to the evaluation of best-estimate ECCS performance under small-break LOCA conditions or to passive plant designs. The use of WCOBRA/TRAC is acceptable for LBLOCA calculations because of the extensive test data available for code validation and the associated analytical expertise developed over the past 20 years. A comparable database does not exist for many other applications.

We commend the staff and Westinghouse for completing an important task. If the above concerns are adequately addressed, the result will be a much improved best-estimate method for the prediction of LBLOCA behavior in light-water reactors.

ACRS Member George Apostolakis did not participate in the Commit-tee's deliberation of this matter.

Sincerely,

/s/

T. S. Kress Chairman

References:

1. Westinghouse Topical Report, "Code Qualification Document for Best Estimate LOCA Analysis," WCAP-12945-P, Revision 1, Volumes 1-5, June 1992 (Proprietary)

2. U. S. Nuclear Regulatory Commission, "Safety Evaluation by the Office of Nuclear Reactor Regulation Related to Acceptability of the Topical Report WCAP-12945(P) 'Westinghouse Code Qualification Document for Best-Estimate Loss of Coolant Accident Analysis' for Referencing in PWR Licensing Applica-tions, Westinghouse Electric Corporation" (Draft) and "Draft Technical Evaluation Report, Westinghouse Code Qualification Document for Best Estimate Loss of Coolant Accident Analyses, WCAP-12945-P" (Proprietary), transmitted by P. Boehnert, ACRS staff, to the ACRS Thermal Hydraulic Phenomena Subcommittee, by memorandum dated January 4, 1996

3. Memorandum, dated November 3, 1995, from P. Boehnert, ACRS staff, to I. Catton, Chairman, ACRS Thermal Hydraulic Phenome-na Subcommittee,

Subject:

NRC/NRR-Westinghouse Meeting, October 23-24, 1995 - Westinghouse Best-Estimate ECCS Evaluation Model Code, WCOBRA/TRAC", including W memorandum, dated October 13, 1995, transmitting "Revisions to the W Best-Estimate Uncertainty Methodology" (Proprietary)

4. Memorandum dated January 5, 1996 from M. Nissley, Westing-house, to Members and Consultants of the ACRS Thermal Hydrau-lic Phenomena Subcommittee, transmitting the following reports:

. NTD-NRC-95-4505 - Roadmap Comparison with CSAU Methodology

. NTD-NRC-95-4575 - Revised Uncertainty Methodology Report (Proprietary)

. NTD-NRC-95-4586 - Assessment of Compensating Errors (Proprietary)

. NTD-NRC-95-4588 - Non-Proprietary Executive Summary

. NTD-NRC-96-4618 - Responses to Several Issues Identified in INEL's Review of NTD-NRC-95-4575 (Proprietary)

5. U.S. Nuclear Regulatory Commission, Regulatory Guide 1.157, "Best-Estimate Calculations of Emergency Core Cooling System Performance," May 1989