July 25, 2001
The Honorable Richard A. Meserve
U.S. Nuclear Regulatory Commission
Washington, D.C. 20555-0001
Dear Chairman Meserve:
|SUBJECT:||FEASIBILITY STUDY ON RISK-INFORMING THE TECHNICAL REQUIREMENTS OF 10 CFR 50.46 FOR EMERGENCY CORE COOLING SYSTEMS|
During the 484th meeting of the Advisory Committee on Reactor Safeguards, July 11-13, 2001, we met with representatives of the NRC staff and the industry to discuss the status of staff and industry initiatives to risk inform the technical requirements of 10 CFR 50.46, "Acceptance criteria for emergency core cooling systems for light-water nuclear power reactors." Our Subcommittees on Materials and Metallurgy, Thermal-Hydraulic Phenomena, and Reliability and Probabilistic Risk Assessment discussed this matter with representatives of the NRC staff, the Nuclear Energy Institute, the Westinghouse Owners Group, and the Boiling Water Reactor Owners Group on July 9, 2001. We also had the benefit of the documents referenced.
|1.||We recommend that the Commission approve the staff's request to proceed with rulemaking to modify the existing 10 CFR 50.46 to replace the prescriptive emergency core cooling system (ECCS) acceptance criteria with a performance-based requirement and to modify the 10 CFR Part 50, Appendix K evaluation model.|
|2.||We recommend that the Commission approve the staff's request to proceed with the development of a voluntary risk-informed alternative to 10 CFR 50.46, Appendix K, and General Design Criterion (GDC) 35 of 10 CFR Part 50, Appendix A.|
|3.||The staff should continue to develop the technical bases and requirements for redefining the large-break loss-of-coolant accident (LBLOCA).|
The ECCS requirements codified in 10 CFR 50.46, Appendix K, and GDC 35 are intended to ensure that plants can safely cope with a LBLOCA. The ECCS has been designed to accommodate pipe breaks up to and including a double-ended guillotine break of the largest pipe in the reactor coolant system. GDC 35 requires that the ECCS be capable of providing sufficient core cooling for a full spectrum of postulated LOCAs using either offsite power or onsite power. To comply with this requirement, ECCS evaluations generally assume that pipe breaks are coincident with a loss of offsite power (LOOP). In addition, the system must have sufficient diversity and redundancy to accomplish its safety function assuming a single failure.
Because LBLOCAs are rare, the current requirements for ECCS performance may have a detrimental effect on safety. These requirements focus attention and resources on events that are extremely unlikely to happen rather than on events which can have a larger contribution to risk. For example, the postulated occurrence of a LOOP coincident with a LBLOCA leads to requirements for rapid emergency diesel generator (EDG) start times and load sequencing. Such requirements could reduce the reliability of the EDGs and diminish the capability of the system to deal with the more likely small and medium break LOCAs.
The industry has proposed a revision of 10 CFR 50.46 that is based on a redefinition of the LBLOCA. Instead of dealing with a full spectrum of break sizes up to and including the double-ended guillotine break of the largest pipe in the reactor coolant system, the industry proposes to define a new maximum LBLOCA size based on leak-before-break (LBB) methodology and probabilistic assessments of the frequency and consequences of the new LBLOCA size.
The staff has accepted LBB methodology for the analysis of dynamic effects of pipe failure for pipe sizes down to 8-inches in some cases. The NRC pioneered the application of probabilistic fracture mechanics to piping through the development of the PRAISE code. The staff argues, however, that the prediction of leak rates for all sizes of cracks in all locations in piping systems is technically much more demanding than predicting whether a detectable leak will occur before failure. The staff also argues that a more rigorous assessment of uncertainties is needed to justify the redefinition of the LBLOCA for ECCS requirements. Thus, the staff believes this is a longer-term activity that will require a substantial technical effort.
We agree that the effort to define a new LBLOCA size requires an extension of current LBB and probabilistic fracture mechanics methodology. We believe that it is technically feasible, but the justification of the new LBLOCA size will become increasingly difficult as the proposed maximum break size is decreased. The industry has stated that it is willing to invest substantial resources to accomplish this objective. The staff should continue to develop the technical basis and requirements for the redefinition of LBLOCA.
In its Feasibility Study, the staff has investigated a number of options for revising 10 CFR 50.46 that it believes can be implemented on a shorter time scale and will provide safety benefits and some reduction in unnecessary conservatism and associated regulatory burden.
One of these options would make changes in the Appendix K evaluation model and would replace the current prescriptive ECCS acceptance criteria with a performance-based requirement. This would permit licensees to use cladding materials other than zircaloy or ZIRLO without having to seek an exemption. The current criteria, such as the 2200°F peak clad temperature and 17% oxidation limit, would be relegated to a regulatory guide as acceptance criteria for zircaloy and ZIRLO. We support the proposed development of the new performance-based acceptance requirement.
Possible changes in the evaluation models suggested in the staff Feasibility Study include replacing the current 1971 American Nuclear Society (ANS) decay heat curve with the 1994 ANS standard, replacing the current decay heat multiplier of 1.2 with an uncertainty estimate, and replacing the Baker-Just oxidation model with the Cathcart-Pawel oxidation model for heat generation. The intent of these changes is to use improved technical understanding to remove excessive conservatism from Appendix K models.
We are generally supportive of this effort, but note that in dealing with a mix of models in which some elements are conservative and some elements are nonconservative, removing "excessive" conservatism without a real understanding of the uncertainties in the overall model can lead to unsatisfactory results. For example, although the Cathcart-Pawel model gives a more accurate description of the oxidation behavior of unirradiated zircaloy tubing in laboratory studies, the more conservative Baker-Just model was deliberately chosen in an attempt to ensure that the effects of variables such as irradiation and behavior such as spalling of the oxide film that were not explicitly included in the models would not lead to nonconservative results. In addition, although the staff is developing performance-based acceptance criteria to permit use of other cladding materials, both the Baker-Just and Cathcart-Pawel models build "zircaloy behavior" into the evaluation model. The staff should consider a performance-based requirement for a heat generation model that includes the effects of cladding oxidation, irradiation, and the potential for cladding spallation rather than a prescriptive requirement. Acceptable heat generation models for different cladding materials could then be discussed in a regulatory guide. If implementation of the Appendix K option proves to be more challenging than anticipated, then the staff should proceed with a rulemaking that includes only the update of the decay heat curve to the 1994 ANS standard.
The second shorter-term option recommended by the staff is a voluntary risk-informed alternative to 10 CFR 50.46 that would replace the current requirements intended to ensure ECCS reliability (i.e., the coincident LOOP and the single-failure criterion) with more risk-informed approaches that reflect the lower frequencies of LBLOCAs. Licensees could choose either generic deterministic reliability requirements developed by the NRC (e.g, a requirement that a coincident LOOP be postulated only for smaller, more frequent LOCAs) or show that they can meet an acceptable threshold value for the core damage frequency (CDF) and large, early release frequency (LERF) associated with the LOCA initiators with appropriate consideration of uncertainties. ECCS reliability evaluations could reflect plant-specific features and operational data. Frequencies of LOCAs with different break sizes could be determined using the analysis provided in NUREG/CR-5750, updated to reflect more recent operating experience. Alternatively, probabilistic fracture mechanics together with a review of service history data could be used, but the technical work to support this would be similar in magnitude to that required to define the new LBLOCA size. We believe the approach outlined by the staff in this option would provide a much more realistic and risk-informed approach for ECCS requirements. The staff should proceed with the technical work and the rulemaking for this option.
We look forward to reviewing the technical work and regulatory guidance needed to support these rulemaking efforts as they evolve.
George E. Apostolakis
- Draft memorandum received June 3, 2001, from William D. Travers, Executive Director for Operations, to The Commissioners, Subject: Status Report on Study of Risk-Informed Changes to the Technical Requirements of 10 CFR part 50 (Option 3) and Recommendations on Risk-Informed Changes to 10 CFR 50.46 (ECCS Acceptance Criteria), and attached Feasibility Study report.
- Memorandum dated January 19, 2001, from Annette L. Vietti-Cook, Secretary, NRC, to William D. Travers, Executive Director for Operations, NRC, Subject: Staff Requirements - SECY-00-0198 - Status Report on Study of Risk-Informed Changes to the Technical Requirements of 10 CFR Part 50 (Option 3) and Recommendations on Risk-Informed Changes 10 CFR 50.44 (Combustible Gas Control).
- Memorandum dated February 3, 2000, from Annette Vietti-Cook, Secretary, NRC, to William D. Travers, Executive Director for Operations, NRC, Subject: Staff Requirements - SECY-99-264 - Proposed Staff Plan for Risk-Informing Technical Requirements in 10 CFR Part 50.
- Letter dated February 8, 2001, from Anthony R. Pietrangelo, Nuclear Energy Institute, to Thomas L. King, Office of Nuclear Regulatory Research, NRC, Subject: Preliminary Industry Response to NRC Questions on Redefinition of Large-Break Loss-of-Coolant Accident.
- Letter dated October 17, 2000, from Robert H. Bryan, Westinghouse Owners Group to Thomas L. King, Office of Nuclear Regulatory Research, NRC, Subject: "WOG Large Break Loss of Coolant Accident (LBLOCA) Redefinition Discussion of Benefits."
- Letter dated January 8, 2001, from Adrian Heymer, Nuclear Energy Institute, to Mary T. Drouin, Office of Nuclear Regulatory Research, Subject: "Draft Large Break LOCA Redefinition Program, Project Summary."
- Letter dated January 19, 2000, from Joe F. Colvin, Nuclear Energy Institute, to Richard A. Meserve, Chairman, NRC, Subject: SECY-99-264, Proposed Staff Plan for Risk-Informing Technical Requirements in 10 CFR Part 50.
- American Nuclear Society, ANSI/ANS-5.1-1994, American National Standard for Removing Decay Heat Power in Light Water Reactors, dated August 23, 1994.
- U.S. Nuclear Regulatory Commission, NUREG/CR-5750, Rates of Initiating Events at U.S. Nuclear Power Plants: 1987-1995, February 1999.