In the context of the Generation IV initiative, the consequences of a severe-accident (SA) in a sodium-cooled fast reactor must be studied. A SFR (Sodium cooled Fast Reactor) severe accident involves the disruption of the core by super-criticality involving the destruction of a certain number of fuel assemblies. Subsequently the interaction between hot fuel and liquid sodium can lead to a vapor explosion which could create a breach in the primary system. Some contaminated liquid sodium would thus be ejected into the containment building. In this situation, the evaluation of potential releases to the environment (the source term) must forecast the quantity and the chemical speciation of the radiocontaminants likely to be released from the containment building.
One critical risk of a SA is the production of contaminated aerosols in the containment building by spray ejection of primary-system sodium. Being pyrophoric, the sodium droplets react with oxygen first oxidizing then burning, with significant heat of combustion. As well as evaluating the consequences of a pressure rise inside the containment, the evolution of the sodium must be assessed since not only is it activated and contaminated but, in oxide form, very toxic. Ultimately, the aerosols are the main radiological risk acting as the vector for radionuclide transport to the environment in the event of a problem with the confinement. These aerosols could evolve and interact with the FP (Fissile Products) and these interactions could modify the physical and chemical nature of the PF.
We model a large part of the events that occur during a SA inside a SFR from the sodium spray fire to the reaction between sodium aerosols and PF (iodine).
At first, we develop a numerical model (NATRAC) that simulates the sodium spray fire, calculates the temperature and the pressure inside the containment as well as the mass of aerosols produced during this kind of fire. The simulation has been validated with different experiments chosen in the literature. The mass of oxide aerosols produced by a sodium spray fire can involve more than 60% of the ejected sodium.
Then, we develop the numerical simulation STARK based on the Cooper model that model the physico-chemical transformations of the aerosols. However, this model has never been validated and the literature does not permit to do so. In these conditions, we designed and performed our own experiment to obtain the missing values of parameters that govern the Cooper model. The Cooper model has been improved with the results of this experimental study, ESSTIA, and we present a modified Cooper model that improves the accuracy of Cooper model to investigate the transformation of the sodium aerosols.
The last part of the manuscript deals with the interaction between sodium aerosols (hydroxide) and a key fission product (iodine). We use density functional theory numerical simulation (the VASP code) to discover the affinities that can be identified. The results will facilitate simulation of the source term evolution because the sodium aerosols will interact with this FP.
All the data and numerical simulations presented here will contribute to implementation of models in the future SFR SA numerical simulation of the IRSN, ASTEC-Na.