SIMULATION OF THE OECD MAIN STEAM LINE BENCHMARKUSING THE WESTINGHOUSE RAVE? METHODOLOGY

摘要

In order to determine the safety of a reactor withrnrespect to reactor system failures, a set of postulatedrnevents is analyzed, and the results are presented inrnChapter 14 or 15 of the plant Final Safety AnalysisrnReport (FSAR). In the analysis of events that are notrninitiated by a Loss of Coolant Accidents (non-LOCArnevents), the typical approach has been to makernconservative and bounding analysis assumptions,rneither because of analysis expediency, or because ofrnthe simplified modeling assumptions. In some cases,rnthis has resulted in combination of assumptions thatrncannot occur in reality. The consistency of the analysisrnassumptions can be improved by externally linking thernreactor coolant system thermal-hydraulic calculationrnmodel to a more realistic 3-dimensional corernneutronics and heat transfer model, that replaces thernsimple point kinetics model typically used to representrnthe core neutronics in the system code.rnOver the past few years, Westinghouse hasrndeveloped the RAVE? methodology for thernapplication of three-dimensional core neutron kineticsrnto the analysis of non-LOCA FSAR events. Thisrnmethodology uses the NRC-approved core neutronrnkinetics code SPNOVA and the NRC-approvedrnWestinghouse version of the core thermal hydraulics code VIPRE-01, in conjunction with the NRC-approvedrnWestinghouse version of the reactor coolantrnsystem thermal hydraulic code RETRAN-02. ThernWestinghouse methodology has been submitted to thernNRC for approval in April, 2004 and the NRC SafetyrnEvaluation Report (SER) is expected to be issuedrnbefore this paper will be presented.rnAs part of the development and licensing of thernRAVE? methodology, Westinghouse has performedrnan analysis of the OECD Main Steam Line Breakrn(MSLB) benchmark. This benchmark problem hadrnbeen defined in a cooperative program sponsored byrnthe OECD, the NRC, and the Pennsylvania StaternUniversity, in order to simulate the core response andrnthe reactor coolant system response to a relativelyrnsevere steamline break accident condition. A PWRrnMSLB is characterized by significant space-timerneffects in the core caused by the asymmetric coolingrnand an assumed stuck-out rod during reactor trip. Thisrnproblem was therefore considered appropriate to testrnthe incorporation of a full three-dimensional modelingrnof the reactor core into the system transient code tornallow simulation of interactions between reactor corernbehavior and plant dynamics.rnThis paper documents the results of the applicationrnof RAVE? methodology to the OEDM MSLBrnbenchmark problem.