Uncertainty and sensitivity analyses for fuel temperature evaluations of U-Mo/Al plate-type dispersion fuel

Abstract

U-Mo/Al plate-type dispersion fuel is a promising candidate for the conversion of research reactor fuels from highly enriched to low-enriched uranium due to its high uranium density. The fuel temperature is a very important parameter, as it affects the performance of the fuel through various aspects, such as the formation of an interaction layer (IL) between the fuel particles and the matrix, swelling, and the release of fission gas. For these reasons, the fuel temperature as a function of the fission density was calculated for two representative heat flux profiles using best-estimate values and Monte Carlo simulations. Uncertainty and sensitivity analyses which utilized the uncertainties of the critical parameters were then conducted to determine the upper (maximum) and lower (minimum) bounds of the fuel temperature for the selected heat flux profiles. The uncertainty analysis used common uncertainty propagation approaches and a probabilistic sensitivity analysis (Monte Carlo simulation), randomly sampling numbers following a Gaussian distribution. Lastly, the Pearson correlation coefficient was used to identify the input uncertainties which influence the fuel temperature most in the sensitivity analysis. These analyses contribute to safety analyses and to the licensing process, as they are used in best-estimate approaches that apply realistic assumptions complemented with uncertainty analyses, such as the Best Estimate Plus Uncertainty (BEPU) approach.