Uncertainty quantification for severe-accident reactor modelling; Results and conclusions of the MUSA reactor applications work package

Brumm, S.*; Gabrielli, F.*; Sanchez Espinoza, V.*; Stakhanova, A.*; Groudev, P.*; Petrova, P.*; Vryashkova, P.*; Ou, P.*; Zhang, W.*; Malkhasyan, A.*; Herranz, L. E.*; Iglesias Ferrer, R.*; Angelucci, M.*; Berda, M.*; Mascari, F.*; Agnello, G.*; Sevbo, O.*; Iskra, A.*; Martinez-Quiroga, V.*; Nudi, M.*; Hoefer, A.*; Pauli, E.-M.*; Beck, S.*; Tiborcz, L.*; Coindreau, O.*; Clark, G.*; Lamont, I.*; Zheng, X. ; Kubo, Kotaro*; Lee, B.*; Valincius, M.*; Malicki, M.*; Lind, T.*; Vorobyov, Y.*; Kotsuba, O.*; Di Giuli, M.*; Ivanov, I.*; D'Onorio, M.*; Giannetti, F.*; Sevon, T.*

The completed Horizon-2020 project on "Management and Uncertainties of Severe Accidents (MUSA)" has reviewed uncertainty sources and Uncertainty Quantification methodology for the purpose of assessing Severe Accidents (SA). The key motivation of the project has been to bring the advantages of the Best Estimate Plus Uncertainty approach to the field of Severe Accident. The applications brought together a large group of participants that set out to apply uncertainty analysis (UA) within their field of SA modelling expertise, in particular reactor types, but also SA code used (ASTEC, MELCOR, etc.), uncertainty quantification tools used (DAKOTA, RAVEN, etc.), detailed accident scenarios, and in some cases SAM actions. This paper synthesizes the reactor-application work at the end of the project. Analyses of 23 partners are sorted into different categories, depending on whether their main goal is/are (i) uncertainty bands of simulation results; (ii) the understanding of dominating uncertainties in specific sub-models of the SA code; (iii) improving the understanding of specific accident scenarios, with or without the application of SAM actions; or, (iv) a demonstration of the tools used and developed, and of the capability to carry out an uncertainty analysis in the presence of the challenges faced. The partners' experiences made during the project have been evaluated and are presented as good practice recommendations. The paper ends with conclusions on the level of readiness of UA in SA modelling, on the determination of governing uncertainties, and on the analysis of SAM actions.