Study on best estimate method applied to nuclear criticality safety analysis

In Best Estimate method, multiple parameters are randomly sampled at the same time according to their probability distribution. Best Estimate method aims at simulating true condition and the number of code runs is related to the tolerance percentile and confidence level. Best Estimate method is applied to the criticality analysis of spent fuel rack and fuel transport package, using nonparametric sampling approach which means that multiple parameters are sampled at the same time. The sorted sequence of parameter sensitivity is got from parameter sensitivity analysis. The statistical analysis of calculation results demonstrates that the result of Best Estimate method is closer to true-value and the method of analyzing every single parameter sensitivity is conservative. The sequence of parameter sensitivity is stable when analyzing specific condition, which mostly depends on parameter's own sensitivity. The range and distribution of parameter have little effect on the sequence. The sensitive parameters should be paid close attention to in the design and manufacturing process of relevant equipment. (authors)