Passive safety systems are increasingly implemented in advanced nuclear power plant designs. These systems rely on intrinsic, physics-based driving forces of low magnitude, enabling fully passive operation without active components. However, the small magnitude of these forces makes system performance more sensitive to disturbances, boundary conditions, and uncertainties in the underlying physical phenomena, thereby complicating reliability assessment compared to active systems.
This paper illustrates the application of an approach for assessing the reliability of passive safety systems, based on the Reliability Methods for Passive Systems (RMPS) methodology and Bayesian inference, to the isolation condenser system (ICS) of the BWRX-300 small modular reactor (SMR).
Application of the approach yielded an extremely low estimated probability of failure for the natural circulation in the isolation condenser system. To account for residual uncertainty, a conservative failure probability of 1.E-7 was assigned based on engineering judgment. The results were subsequently incorporated into a simplified PSA model of the BWRX-300 to evaluate system-level unavailability and the overall Core Damage Frequency (CDF).