Determination methodologies for input data including loads considered for reliability evaluation of fast reactor components

Yokoi, Shinobu*; Kamishima, Yoshio*; Sadahiro, Daisuke*; Takaya, Shigeru  

Many efforts have been made to implement the System Based Code concept aiming at optimizing margins dispersed in existing codes and standards. Failure probability calculated based on statistical information such as a type of probability distribution, mean (or median) and variance (or standard deviation) for random variables is expected to be a promising quantitative index for margin optimization. Statistical information on material strength, which is also required to calculate the failure probability, has been already reported in JAEA-Data/Code 2015-002 "Structural Properties of Material Strength for Reliability Evaluation of Components of Fast Reactors -Austenitic Stainless Steels-" whereas others have not been identified yet. This report provides methodologies and basic ideas to determine statistical parameters of other random variables, especially variable loads, necessary for reliability evaluation.