Variance-based sensitivity analysis applied to hydrogen migration and redistribution model in Bison. Part I: Simulation of historical experiments

In a light water nuclear reactor environment, zirconium-alloy cladding undergoes corrosion that introduces hydrogen into the cladding. Hydrogen being slightly soluble in zirconium means that the terminal solid solubility of hydrogen is quickly reached. This leads to hydrogen precipitation into crystals of zirconium hydride. These hydrides decrease the ductility of the cladding, which increases the chance of cladding failure, especially during dry storage or transportation events, depending on the temperature and stress conditions. In this work, we tested the existing hydrogen migration and redistribution model implemented in the Bison fuel performance code by comparing Bison predictions of hydrogen distribution in various temperature profiles against historical published experimental data. Using 1D simulations of hydrogen evolution in zirconium-alloy, we demonstrated the accuracy of the existing model in the case of a sharp temperature gradient profile. We also performed a sensitivity analysis to investigate the impact of the model parameters on Bison predictions of hydrogen migration and redistribution to understand the discrepancies in the case of a low temperature gradient profile. In both cases, the Bison prediction of hydrogen distribution is accurate in the regions where the temperatures are higher. In these areas, the relative error on average does not exceed 1%. On the other hand, the error grows increasingly higher in the cold regions where hydrides are more likely to form and a steep gradient in the hydrogen concentration profile develops. The potential for the uncertainties in the model parameters to explain this behavior is investigated through the sensitivity analysis, including the impact of a clamping factor that artificially reduces the maximum allowable volume fraction of hydrides. This study also provides a literature review on hydrogen migration and redistribution experiments and a systematic comparison of the relative experimental data. (C) 2019 Elsevier B.V. All rights reserved.