AP1000 BEST ESTIMATE LARGE BREAK LOCA ANALYSIS PERFORMEDWITH THE WESTINGHOUSE AUTOMATED STATISTICAL TREATMENT OFUNCERTAINTY METHOD (ASTRUM)

摘要

The AP1000 is an advanced passive plant derived from theAP600. The AP1000 and AP600 reactor vessel geometries arebasically the same and similar to a standard PWR. The AP1000is rated for a thermal power of 3400 MWt. A unique feature ofthese passive plants is that the ECCS is designed to operatethrough passive means where pumps and emergency diesels arereplaced by natural circulation and gravity drain tanks.A realistic Large Break (LB) Loss of Coolant Accident(LOCA) analysis was performed using the WCOBRA/TRACcomputer code and the results are presented herein. The analysisincludes the use of the Westinghouse Automated StatisticalTreatment of Uncertainty Method (ASTRUM) which wasrecently licensed by the NRC. The methodology relies on theorder statistics for the determination of 95th percentile of PeakClad Temperature (PCT), Local Maximum Oxidation (LMO),and Core-Wide Oxidation (CWO) at a 95% confidence level.The method requires the execution of 124 LBLOCA transientcalculations, where the uncertainty contributors are sampledusing a direct Monte Carlo method.This paper discusses the results of the realistic LOCA analysisof AP1000 using the new methodology. The technique includesthe sampling of break type, break size, power distribution, andtime in the cycle as uncertainty contributors. Results indicatethat for the postulated LBLOCA scenario, the AP1000 has significant margin against the three traditional 10 CFR 50.46criteria (PCT < 2200 F, LMO < 17% and CWO < 1%).The analysis includes two sensitivity studies to assess theimpact of the off-site power availability assumption and theassumption on the burnup for the average and low power regionof the core.