Performing Uncertainty Analysis of IIST Facility SBLOCA by TRACE and DAKOTA. International Agreement Report.

摘要

Nowadays, the increasing computing speed realizes the quantification of propagation of input uncertainties to output data with Monte Carlo simulation or modified simulation methods. The best estimate plus uncertainty (BEPU) methods have been proposed to be used instead of typical conservative methodologies. Based on the CAMP activity, this project demonstrates the capability of SNAP-TRACE-DAKOTA for 2 % small break LOCA (SBLOCA) of IIST experiment. The number of samples was determined by Wilks formula to generate the upper bound of peak cladding temperature (PCT) with 95/95 confidence level and probability. The PCTs by IIST experiment and best-estimate calculation are 804 K and 861 K respectively. The mean value and standard deviation of the 59 trial by SNAP-TRACE-DAKOTA are 938.7 K and 63.6 K respectively, and the maximum value of PCT is 1054 K.