Incorporating reliability analysis into the design of passive cooling systems with an application to a gas-cooled reactor

Abstract

A time-dependent reliability evaluation of a two-loop passive decay heat removal (DHR) system was performed as part of the iterative design process for a helium-cooled fast reactor. The system was modeled using RELAP5-3D. The uncertainties in input parameters were assessed and were propagated through the model using Latin hypercube sampling. An important finding was the discovery that the smaller pressure loss through the DHR heat exchanger than through the core would make the flow to bypass the core through one DHR loop, if two loops operated in parallel. This finding is a warning against modeling only one lumped DHR loop and assuming that n of them will remove n times the decay power. Sensitivity analyses revealed that there are values of some input parameters for which failures are very unlikely. The calculated conditional (i.e., given the LOCA) failure probability was deemed to be too high leading to the identification of several design changes to improve system reliability. This study is an example of the kinds of insights that can be obtained by including a reliability assessment in the design process. It is different from the usual use of PSA in design, which compares different system configurations, because it focuses on the thermal–hydraulic performance of a safety function.