MANAGEMENT AND UNCERTAINTIES OF SEVERE ACCIDENTS

ended

In the current state of maturity of severe accident codes in terms of phenomena addressed and extensive validation conducted, the time has come to foster BEPU,Best Estimate Plus Uncertainties, application in the severe accident (SA) domain, and accident management (AM). The advantages with respect to deterministic analysis are known: avoid adopting conservative assumptions in the model and allow identifying safety margins, quantify likelihood of reaching specific values and, through the distribution variance provide insights into dominating uncertain parameters.The overall objective of the Management and Uncertainties of Severe Accident (MUSA) project is to assess the capability of SA codes when modelling reactor and SFP (Spent Fuel Pool) accident scenarios of Gen II and III. To do so UQ (Uncertainty Quantification) methods are to be used, with emphasis on the effect of already-set and innovative accident management measures on accident unfolding, particularly those related to ST (Source Term) mitigation. Therefore, ST related Figures Of Merit (FOM) are to be used in the UQ application. The MUSA project proposes an innovative research agenda in order to move forward the predictive capability of SA analysis codes by combining them with the best available/improved UQ tools and embedding accident management as an intrinsic aspect of SA analyses.MUSA develops through key activities which also describe the main outcomes foreseen from the project: identification and quantification of uncertainty sources in SA analyses; review and adaptation of UQ methods; and testing such methods against reactor and SFP accident analyses, including AM. Given the focus of FOM on source term, the project will identify variables governing ST uncertainties that would be worth investigating further. All the ingredients necessary to conduct the project are already available: analytical tools, experimental data, postulated reactor and SFP scenarios and, technical and scientific competences.