ABSTRACT
Probabilistic Safety Analysis (PSA) is a systematic probabilistic method for assessment of reliability and safety of the complex systems including nuclear power plants. The event tree and the fault tree are two basic methods used in PSA. The basic events are the ultimate parts of the fault tree, representing the undesired events. Component unavailability is defined as the probability of being in a failed state when required. The constant unavailability model is most frequently used basic event reliability model in the PSA of the nuclear power plants. The constant unavailability model is commonly used when failure probability per demand is needed, typically when components or systems are activated or change state as it is case for the most of safety systems components. The increased and extended application of the PSA requires appropriate consideration of uncertainties in analyses and interpretation of the results. Inadequate treatment of uncertainties may lead to poorly supported or even wrong conclusions whose final consequence is a loss of adequate level of safety. Epistemic uncertainty results from the imperfect knowledge or incomplete information regarding values of parameters of the underlying model. It is also called state-of-knowledge uncertainty. Epistemic uncertainty is considered in the models by probability distributions associated with uncertain parameters. Probability distributions associated with uncertain parameters represent the state of knowledge about the right values of the parameters and are therefore very often derived from expert judgment. This paper presents the results of the analysis of the introduction of probability distributions associated with component unavailability parameters, on the overall unavailability of the analyzed system. Implications on introduction of different probability distributions for different number and sets of components are investigated. The analysis is done on reference nuclear power plant safety system model. The main findings of analysis and corresponding conclusions are given.
