THE MODELLING OF FUEL VOLATILISATION IN ACCIDENT CONDITIONS

THE MODELLING OF FUEL VOLATILISATION IN ACCIDENT CONDITIONS H. Manenc1, P.K. Mason2, M.P. Kissane1 1Institut de Protection et de Sûreté Nucléaire, Département de Recherches en Sécurité, CEA/Cadarache, France 2AEA Technology, Winfrith Technology Centre, UK (under contract to IPSN)

For oxidising conditions, at high temperatures, the pressure of uranium vapour species at the fuel surface is predicted to be high. These vapour species can be transported away from the fuel surface, giving rise to significant amounts of volatilised fuel, as has been observed during small-scale experiments [1], [2], [3], [4] and taken into account in different models [5], [6]. Hence, fuel volatilisation must be taken into account in the conduct of a simulated severe accident such as the Phebus FPT-4 experiment, a large-scale in-pile test designed to investigate the release of fission products and actinides from irradiated UO2 fuel in a debris bed and molten pool configuration. Best estimate predictions for fuel volatilisation were performed before the test. This analysis was used to assess the maximum possible loading of filters collecting emissions and the consequences for the filter-change schedule. Following the successful completion of the experiment, blind post-test analysis is being performed; the boundary conditions for the calculations are based on preliminary post-test analysis with the core degradation code ICARE2 [7]. The general modelling approach is presented here and then illustrated by the analysis of fuel volatilisation in Phebus FPT4 (for which results are not yet available). Effort was made to reduce uncertainties in the calculations by improving the understanding of the controlling physical processes and by using critically assessed thermodynamic data to determine uranium vapour pressures. The analysis presented here constitutes a preliminary, blind, post-test estimate of fuel volatilised during the test.