Safety objectives of SEPIA systems for sodium-cooled fast reactor

The Sodium-cooled Fast Reactor safety strategy for preventing core disruptive accident is based, in particular, on a plant design capable of ensuring that the combination of any initiating event with the complete failure of all automatic shutdown systems (in particular the two redundant and diverse scram systems) cannot initiate a core disruptive accident. For mitigating such combinations, only diverse measures (i.e., diverse with regard to the automatic shutdown systems), not damaged by the initiating event (or by the hazard which could initiate it) can be postulated to be efficient. Analyses performed on the previous Liquid Metal Fast Reactors (e.g., European Fast Reactor (EFR)) show that for large cores, with uranium-plutonium oxide fuel, short term management of the reactivity control could be achieved by improvement of the inherent reactivity feedback effects, mainly those resulting from thermal structure expansion (in particular the control rod thermal expansion and the dia-grid thermal expansion). At a longer term, this needs to be completed by measures capable of decreasing the core reactivity. This is in particular required for ensuring a safe state. For future SFR, among some of the solutions that are being investigated, additional passive and diversified devices allowing core reactivity decrease, called SEPIA systems (French acronym for 'SEntinelle Passive d'Insertion d'Anti-reactivite' which means Passive Sentinel for Anti-reactivity Insertion) could cope with the short and long term management of unprotected transients (i.e., with complete failure of all automatic shutdown systems). Some designs of SEPIA systems which are being studied consist of dedicated neutron-absorbing devices (fully diverse compared to those connected to the automatic shutdown systems) which are actuated and released into the core in case of significant core temperature increase. For instance, this could be achieved by implementation of B4C pebbles inside some fuel subassemblies. Calculations performed with relevant system codes confirm the potential efficiency of such SEPIA systems to cope with unprotected transients, in particular unprotected loss of flow and unprotected loss of heat sink transients. The paper presents the motivations for implementing SEPIA systems, the specifications for such systems and a summary of the transients' calculations performed in support of the definition of these specifications. R and D actions are in progress to identify the most promising design solutions, and feasibility studies on these are on-going. (authors)