Best estimate plus uncertainty analysis of metal-water reaction transient experiment

Abstract

Uncertainty analysis is applied in the licensing process for nuclear installations to complement best estimate analysis and to verify that the upper bound value is less than the threshold corresponding to the safety parameter of interest. Metal-water reaction is a critical safety phenomenon of water-cooled nuclear reactors at accident conditions, e.g. Loss-Of-Coolant Accidents (LOCA). AISI 348 cladding is able to increase the accident tolerance comparing to Zr-based alloys and differently from other accident tolerant fuel cladding options, there is operational experience of nuclear power plants with stainless steel. In this study, a transient oxidation experiment of AISI 348 by steam was conducted and the major sources of uncertainty were addressed. An evaluation model was developed to calculate the evolution of mass gain during the experiment. Meanwhile, uncertainty propagation of experimental data was performed. The results show that the mass gain predicted by the transient metal-water reaction model lays within the experimental data uncertainty band. Furthermore, the selection of the oxidation kinetics model seems to be important whether the analysis wills to provide conservative results.