Optimization of uncertainty parameters screening in the physical process failures of the secondary-side passive residual heat removal system

BackgroundThe study of physical process failures is a core aspect of research on passive systems. Given that passive systems play an indispensable role in ensuring the safety and reliability of advanced nuclear power systems, precisely identifying key parameters that influence the failures of passive systems holds immense significance.PurposeThis study aims to devise an efficient and reliable parameter optimization method to identify high-impact parameters, thereby enhancing the reliability assessment of passive systems.MethodsThis study focused on the Secondary-side Passive Residual Heat Removal System (PRS) of HPR1000. Firstly, a small-sample uncertainty quantification screening mechanism that integrated Failure Mode and Effects Analysis (FMEA) with parameter correlation analysis was innovatively constructed, and the FMEA method was employed to analyze potential failures during the operation of PRS. Secondly, key uncertainty factors related to failure modes in physical processes were identified. Then, based on accident scenarios, a parameter list was determined and probability distributions and variation ranges were assigned to each parameter. Subsequently, based on probability density functions, Latin Hypercube Sampling (LHS) was applied to generate input sample combinations. The ratio of instantaneous heat removal power to core decay power was taken as the key performance indicator, and the indicator value at the moment when the system entered a stable operation state during the accident process was selected as the evaluation basis. Through the best-estimate program, transient simulation calculations were conducted to obtain the output performance indicator result set corresponding to the input sample set. Finally, the Spearman's rank correlation coefficient was chosen to measure the correlation between input parameters and output indicators, completing the preliminary screening of parameters. Thereafter the ASYST was employed to conduct uncertainty quantification and transfer on a small sample group (Group A, n=100) generated through LHS after 24 uncertainty input parameters were determined through FMEA.ResultsBy combining parameter correlation analysis, 11 high-impact parameters, whose consistency and validity are verified through analysis results from a large sample group (Group B, n=10 000), are identified.ConclusionsResults of this study reveal that the small-sample screening strategy proposed in this study exhibits both efficiency and reliability in the identification of uncertainty parameters for physical process failures in passive safety systems. The findings of this study offer solid theoretical support and practical technical references for uncertainty quantification and reliability design optimization in passive systems, contributing to enhanced safety and operational efficiency in advanced nuclear power applications.