Hierarchical Bayes model to quantify fracture limit uncertainty of high-burnup advanced fuel cladding tubes under LOCA conditions

成川 隆文; 濱口 修輔*; 高田 孝*; 宇田川 豊

Proceedings of Asian Symposium on Risk Assessment and Management 2022 (ASRAM 2022) (Internet), 11 Pages, 2022/12

To realize a more reliable safety evaluation of loss-of-coolant accidents (LOCAs) in light-water-reactors, we developed a quantification method of the fracture limit uncertainty of high-burnup advanced fuel cladding tubes using a hierarchical Bayes model that can quantify uncertainty even when experimental data are limited. The fracture limit uncertainty was quantified as a probability using the amount of oxidation and the initial hydrogen concentration (the hydrogen concentration in fuel cladding tubes before the LOCA-simulated tests) as explanatory variables. The hierarchical Bayes model was developed by dividing the regression coefficients into a hierarchical structure with an overall average term common to all types of fuel cladding tubes and a term representing differences between types of fuel cladding tubes. Using the developed model, we showed that the fracture limits of the high-burnup advanced fuel cladding tubes tended to be on average equal to or higher than that of an unirradiated conventional fuel cladding tube. Further, we proposed a method to reduce the fracture limit uncertainty by using non-binary data depending on the condition of the fuel cladding tube specimens after the LOCA-simulated test instead of the binary data, thereby increasing the amount of information in each data.