An evaluation of diffusivity and resolution models for fission gas release in UO2

Gas atom behavior is of particular interest to nuclear reactor designers as it plays a key role in determining fission gas release and fuel swelling. Various models have been developed for both the diffusivity and resolution of fission gas in UO 2 for use in fuel performance codes, including both empirical models fitted to experimental data and mechanistic models that represent physical behaviors. This work establishes a consistent methodology for comparing empirical and mechanistic diffusivity and resolution models by evaluating their impact on calculated fission gas release in BISON for several integral light water reactor irradiation experiments. Model accuracies are assessed by comparing the BISON-calculated release with the final fission gas release from the irradiation experiments. Bland–Altman analyses of each combination of models indicate that the mechanistic combination of Matthews diffusivity and Setyawan resolution, with other input parameters fixed, has a lower bias of -0.028% FGR than the historically used empirical pairing, the Turnbull-D1–D2–D3 diffusivity with White resolution, which produces a bias of -1.938% FGR. These results provide a comparative basis for selecting diffusivity and resolution models and suggest that mechanistic models can improve agreement with integral irradiation experiments while offering a more physically based framework for predicting fission gas release beyond the existing experimental database.