open access

https://hdl.handle.net/2115/100425

Correlation Sensitivity Analysis of Reliability of Passive System of small reactor

英語

Gaussian Copula

2026-07-01

Proceedings of the 10th International Symposium on Reliability Engineering and Risk Management (ISRERM 2026)

The Emergency Heat Removal System (EHRS) is a critical passive safety system of the IRIS (International Reactor Innovative and Secure) reactor, and its reliability directly determines the decay heat removal capability following accidents. This paper employs the Active learning Metamodel-based Importance Sampling (AM-SIS) method to calculate and analyze the failure probability of the IRIS EHRS. To address the lack of information regarding correlations among input parameters, Gaussian Copula functions are introduced to impose artificially assumed correlation structures, systematically investigating the sensitivity of failure probability to pairwise correlations among eight key input parameters. The results indicate that among all parameter combinations, the negative correlation between in-containment refueling water storage tank (IRWST) water temperature and automatic depressurization system (ADS) discharge coefficient has the most significant impact on failure probability, with the maximum failure probability increasing by 163% compared to the baseline case. In contrast, the positive correlation between core power and decay heat power, which is of practical engineering concern, has a limited effect on failure probability, with an increase of only 18%. The study confirms that while certain parameter couplings may significantly amplify system risk, the correlation between core power and decay heat power does not pose a serious threat to the reliability of the IRIS EHRS. This work reveals the nonlinear impact of parameter correlations on the reliability assessment of passive safety systems and provides a theoretical basis for subsequent uncertainty analysis and model simplification.

会議発表論文

Version of Record