Uncertainty quantification analysis and statistical estimation for LBLOCA in a PWR using Monte-Carlo and alternative methods

Highlights: • LBLOCA of APR-1400 with 10% power uprate was analyzed using Monte-Calro method. • The PCT data were obtained from different sampling size and method. • The PCT95, CEP, and their 95% confidence intervals were estimated. • They were compared with results of Wilks' and Johnson transformation method. The direct Monte-Carlo method has recently emerged as an alternative uncertainty propagation and quantification method to calculate the 95-percentile peak cladding temperature with 95% confidence level (PCT95/95) or the conditional exceedance probability (CEP) as the computational capability improves. However, it requires a large sample size, and it is still being debated that how many samples are required to obtain the result with low uncertainty and high convergence. In this study, assuming the 10% power uprate of APR-1400, the uncertainty quantification analysis of LBLOCA was conducted by using the direct Monte-Carlo method. Based on the PCT data calculated from different sample sizes (100, 200, 500, 1000, 2000, 5000) and different sampling methods (simple random sampling and Latin hypercube sampling), the PCT95, the CEP, and their 95% confidence intervals were estimated. Then, they were compared with results of Wilks' method and with those of Johnson's normal distribution transformation method, respectively.