A survey of dynamic methodologies for probabilistic safety assessment of nuclear power plants

Dynamic methodologies for probabilistic safety assessment (PSA) are defined as those which use a time-dependent phenomenological model of system evolution along with its stochastic behavior to account for possible dependencies between failure events. Over the past 30 years, numerous concerns have been raised in the literature regarding the capability of the traditional static modeling approaches such as the event-tree/fault-tree methodology to adequately account for the impact of process/hardware/software/firmware/human interactions on the stochastic system behavior. A survey of the types of dynamic PSA methodologies proposed to date is presented, as well as a brief summary of an example application for the PSA modeling of a digital feedwater control system of an operating pressurized water reactor. The use of dynamic methodologies for PSA modeling of passive components and phenomenological uncertainties are also discussed.

► Dynamic methodologies for probabilistic safety assessment (PSA) are surveyed.
► These methodologies overcome the limitations of the traditional approach to PSA.
► They are suitable for PSA using a best estimate plus uncertainty approach.
► They are highly computation intensive and produce very large number of scenarios.
► Use of scenario clustering can assist the analysis of the results.