The present paper is dedicated to preliminary studies of the transient behavior of the ASTRID (Advanced Sodium Technological Reactor for Industrial Demonstration) demonstrator developed in France by CEA and its industrial partners. ASTRID is foreseen to demonstrate the progress made in Sodium Fast Reactor (SFR) technologies at an industrial scale by qualifying innovative options, some of which still remain open in the areas requiring improvements, especially safety and operability. Among the innovative options, a gas Power Conversion System (PCS) based on the Brayton thermodynamical cycle is currently considered. The main objective of such a PCS consists in physically avoiding the possibility of a sodium/water reaction with the secondary circuit.

To assess the transient behavior of such a PCS when facing incident/accident sequences, previous calculations were carried out using the CATHARE 2 thermal-hydraulics code, which considers by default the working gas as an ideal gas in its Equations Of States (EOS). However, this approximation is no longer valid for the high pressure levels of this Brayton cycle. This paper thus describes new calculations performed considering real gas EOS that are now available in CATHARE 3. The simulation of the nominal PCS working point is shown to be much more accurate than in previous CATHARE 2 calculations as the discrepancy regarding the theoretical working point is less than 1°C for the gas temperature and less than 1 % for all the components power levels (compressors, heat exchangers and turbines). The impact of this new real gas hypothesis in CATHARE 3 on an unprotected transient simulation has also been investigated on a loss of power supply case. For short time scales, the impact of such an hypothesis is demonstrated to be very low. However, an improvement of the heat extraction with the real gas option should enhance the natural convection in the primary circuit to the longer term.