Sensitivity analysis of BISON model for characterization of impact of experimental parameters on hydrogen migration and redistribution in zirconium-based alloys

Abstract The formation of hydrides, a result from excess hydrogen uptake by the cladding during nuclear reactor operation, can significantly impact cladding integrity. We present a sensitivity and uncertainty analysis of a hydrogen predictive model in the BISON fuel performance code to characterize the key input parameters involved in the model. This includes identifying the key parameters necessary to simulate hydrogen behavior in the fuel cladding, revealing the impact of environmental conditions on hydrogen distribution, and informing the envelope of conditions for ongoing experimental work conducted by the University of Michigan. The Sobol sensitivity analysis reveals the quantitative impacts of environmental conditions on the predicted total hydrogen concentration, as well as the respective impact on their sensitivity with respect to the physical parameters. Overall, the precipitation of hydrogen that occurred at the cold end of the sample is the most important phenomenon in the prediction of hydrogen concentration. The optimization study using the results from sensitivity analysis indicates that the BISON simulations produce accurate hydrogen predictions when the sets of parametric ranges are shifted to enable more precipitation to occur at the cold end. Lastly, the sensitivity and uncertainty (S/U) analysis for the ongoing benchmarking experiments supports the focus of experiments that lies on the measurement of Soret effect of hydrogen driven by linear temperature gradients. The outputs are expected to better characterize the various parameters involved in the hydrogen transport model in the BISON code, and improve the understanding of the hydrogen transport behaviors in zirconium-based fuel cladding in a range of expected environmental conditions.