A review: passive system reliability analysis – accomplishments and unresolved issues

Reliability assessment of passive safety systems is one of the important issues, since safety of advanced nuclear reactors rely on several passive features. In this context, a few methodologies such as Reliability Evaluation of Passive Safety System (REPAS), Reliability Methods for Passive Safety Functions (RMPS) and Analysis of Passive Systems ReliAbility (APSRA) have been developed in the past. These methodologies have been used to assess reliability of various passive safety systems. While these methodologies have certain features in common, but they differ in considering certain issues; for example, treatment of model uncertainties, deviation of geometric and process parameters from their nominal values, etc. This paper presents the state of the art on passive system reliability assessment methodologies, the accomplishments and remaining issues. In this review three critical issues pertaining to passive systems performance and reliability have been identified. The first issue is, applicability of best estimate codes and model uncertainty. The best estimate codes based phenomenological simulations of natural convection passive systems could have significant amount of uncertainties, these uncertainties must be incorporated in appropriate manner in the performance and reliability analysis of such systems. The second issue is the treatment of dynamic failure characteristics of components of passive systems. REPAS, RMPS and APSRA methodologies do not consider dynamic failures of components or process, which may have strong influence on the failure of passive systems. The influence of dynamic failure characteristics of components on system failure probability is presented with the help of a dynamic reliability methodology based on Monte Carlo simulation. The analysis of a benchmark problem of Hold-up tank shows the error in failure probability estimation by not considering the dynamism of components. It is thus suggested that dynamic reliability methodologies must be