Scoping-level Probabilistic Safety Assessment of a complex experimental facility: Challenges and first results from the application to a neutron source facility (MEGAPIE)

Abstract

This paper presents a scoping-level application of Probabilistic Safety Assessment (PSA) to selected systems of a complex experimental facility. In performing a PSA for this type of facility, a number of challenges arise, mainly due to the extensive use of electronic and programmable components and of one-of-a-kind components. The experimental facility is the Megawatt Pilot Target Experiment (MEGAPIE), which was hosted at the Paul Scherrer Institut (PSI). MEGAPIE demonstrated the feasibility of a liquid lead-bismuth target for spallation facilities at a proton beam power level of 1 MW. Given the challenges to estimate initiating event frequencies and failure event probabilities, emphasis is placed on the qualitative results obtainable from the PSA. Even though this does not allow a complete and appropriate characterization of the risk profile, some level of importance/significance evaluation was feasible, and practical and detailed recommendations on potential system improvements were derived. The second part of the work reports on a preliminary quantification of the facility risk. This provides more information on risk significance, which allows prioritizing the insights and recommendations obtained from the PSA. At the present stage, the limited knowledge on initiating and failure events is reflected in the uncertainties in their probabilities as well as in inputs quantified with bounding values. Detailed analyses to improve the quantification of these inputs, many of which turn out to be important contributors, were out of the scope of this study. Consequently, the reported results should be primarily considered as a demonstration of how quantification of the facility risk by a PSA can support risk-informed decisions, rather than precise figures of the facility risk.