Uncertainty Quantification of SB LOCA With CSAU Using Optimal Statistical Estimator

Abstract

When performing best estimate calculations uncertainty needs to be quantified. An optimal statistical estimator algorithm was adapted, extended and used for response surface generation. The objective of the study was to demonstrate optimal statistical estimator for uncertainty evaluation of single value or time dependent parameters when Code Scaling, Applicability and Uncertainty (CSAU) method is used for uncertainty quantification. The scenario selected was small-break loss-of-coolant accident with break in cold leg of a two-loop pressurized water reactor Westinghouse type. The uncertainty was quantified for RELAP5/MOD3.2 thermalhydraulic computer code. The results showed that peak cladding temperature selected as one of primary safety criteria with added uncertainty does not exceed the safety limit. Uncertainty was additionally evaluated for nine time dependent parameters. This finding indicate that CSAU method can be applied to transients other than loss-of-coolant accident.