Status of severe accident studies at the end of the conceptual design of ASTRID: Feedback on mitigation features

•The status of ASTRID severe accident studies at the end of 2015 is presented.•The good natural behavior of the CFV core facing a severe accident is shown.•Severe accident mitigation strategy and associated studies are presented.•Main families of accident initiators and core degraded states are studied.•Dedicated core material relocation ducts lower the energy release of one order of magnitude. The ASTRID reactor developed by the CEA with its industrial partners, will be used for demonstration of the safety and operability, at the industrial scale, of sodium fast reactors of the 4th generation. Among the goals assigned to ASTRID, one is to improve the safety and the reliability of such reactor (compared to previous built sodium-cooled fast reactors). Regarding the innovations promoted in the ASTRID design, a low sodium void worth core concept (CFV core) has been developed. By means of various design provisions enhancing the neutron leakage in case of sodium draining, the overall sodium void effect of the ASTRID core is near zero and could even be negative. Additionally, mitigation devices should be implemented into the core in order to limit the thermal energy released in the fuel during a severe accident. This paper deals with a synthesis of severe accident studies performed during the second period of the pre-conceptual design stage of the ASTRID project (2013–2015). The main insights of the studies in term of mitigation strategy and of mitigation device design are highlighted in the paper. The CFV core transient behavior has been investigated in case of generalized core melting situations initiated by postulated reactivity insertion ramps (UTOP) and unprotected loss of flow (ULOF). In case of UTOP transients, according to our calculations, the mechanical energy released by molten fuel vapor expansion does not exceed several tenths of megajoule. Simulated ULOF transients do not lead to energetic power excursions thanks to the mitigation provisions and to the core design. Regarding ULOF transients, early boiling phase leads to core power decrease and the primary phase of the accident is not governed by a power excursion. The paper deals with the approach and the presentation of preliminary findings regarding mitigation provisions. Those provisions are investigated by considering a postulated core degraded state representative of the end of the transition phase. The possible scenario evolutions from this degraded state provide the following parameters: mas