(A) study on the advanced statistical core thermal design methodology

Abstract

A statistical core thermal design methodology for generating the limit DNBR and the nominal DNBR is proposed and used in assessing the best-estimate thermal margin in a reactor core. Firstly, the Latin Hypercube Sampling Method instead of the conventional Experimental Design Technique is utilized as an input sampling method for a regression analysis to evaluate its sampling efficiency. Secondly and as a main topic, the Modified Latin Hypercube Sampling and the Hypothesis Test Statistics method is proposed as a substitute for the current statistical core thermal design method. This new methodology adopts "a Modified Latin Hypercube Sampling Method" which uses the mean values of each interval of input variables instead of random values to avoid the extreme cases that arise in the tail areas of some parameters. Next, the independency between the input variables is verified through "Correlation Coefficient Test" for statistical treatment of their uncertainties. And the distribution type of DNBR response is determined though "Goodness of Fit Test". Finally, the limit DNBR with one-sided 95% probability and 95% confidence level, $DNBR_{95/95``}$ is estimated. The advantage of this methodology over the conventional statistical method using Response Surface and Monte Carlo simulation technique lies in its simplicity of the analysis procedure, while maintaining the same level of confidence in the limit DNBR result. This methodology is applied to the two cases of DNBR margin calculation. The first case is the application to the determination of the limit DNBR where the DNBR margin is determined by the difference between the nominal DNBR and the limit DNBR. The second case is the application to the determination of the nominal DNBR where the DNBR margin is determined by the difference between the lower limit value of the nominal DNBR and the CHF correlation limit being used. From this study, it is deduced that the proposed methodology gives a good agreement in the DNBR r...