On the CSAU employment during ULOF accident initiating phase for sodium-cooled fast reactors

Unprotected accidents are considered highly unlikely events in nuclear power plants (NPP) safety evaluation. Nevertheless their assessment is a primary concern for the safety analysis of fast breeder reactors (FBR) because of the severe consequences they might cause. System safety analysis codes such as ARGO (Tachibana et al., 1992) are usually employed for the assessment. tem safety analysis codes the complexity of a NPP is inevitably reduced through the geometrical discretization and the physical modelization, which lead to several approximations, assumptions and compromises. This set of models should be considered carefully and quantified for licensing purposes given the inherent coupled and non-linear nature of the phenomena involved. esent work starts with a brief description of the accident focusing on the core behavior. The transient is divided into phases in order to highlight the important phenomena on the safety criteria. Subsequently the definition of the significant phenomena modelization in the code provides the basis for the identification of the relevant and sensitive input parameters and a discussion on their uncertainties. Eventually, through the use of the tolerance limit for the uncertainty combination, the paper evaluates the uncertainty propagation through the code. aper is intended to demonstrate the applicability of the Code Scaling, Applicability, and Uncertainty (CSAU) methodology for unprotected transients in an FBR, showing the core characteristics at the end of the evaluated transient and focusing on the effect that the modelization of relevant phenomena has on the uncertainties evaluation.