Development of Uncertainty and Sensitivity Approaches for the Analysis of Severe Accidents in a Small Modular Reactor

A research performed by the Nuclear Safety Department of CNEA on the analysis and comprehension of the influence of uncertainties on severe accidents in a small modular reactor is presented. The work included in the paper was accomplished due to the participation in the Coordinated Research Project (CRP) I31033 launched and supported by the IAEA entitled "Advancing the State-of-Practice in Uncertainty and Sensitivity Methodologies for Severe Accident Analysis in Water-Cooled Reactors". Insights on best practices and a possible methodology for uncertainty and sensitivity calculations associated with severe accident codes applications are included. The present study is applied to a CAREM-like plant model, based on a CAREM-25 simplified design to improve the knowledge related to the reactor severe accident management. MELCOR 1.8.6 was considered as the simulation tool to perform severe accident tests in both uncertainty and parametric analysis and Dakota 6.10 was chosen as the calculation tool for sampling and managing the interface with the plant code, performing pre and post processing jobs. After developing the MELCOR plant model and establishing the coupling between codes, an uncertainty studio was performed. A kind of Phenomena Identification and Ranking Table process (PIRT) was carried out to select relevant parameters and their probability distributions regarding severe accident key times proposed as figures of merit (FOM) of the analysis. Correlation tables, minimum and maximum values and probability distributions were obtained after sampling and execution. Four out of eleven parameters were obtained as relevant factors in the accident progression and their impact was measured by performing a centred parametric studio. Studio failure rates were calculated and parameter dependence patterns were searched. As a conclusion, after describing methodology stages, lessons learnt throughout the investigation regarding best engineering practices on uncertainty analysis are discussed and recommendations related to best engineering practices are issued.