Modelling of time resolved and long contact time dissolution studies of spent nuclear fuel in 10Â mM carbonate solution - A comparison between two different models and experimental data

Using two different models, radiation induced dissolution of spent UO2 fuel has been simulated. One of the models is conventional homogeneous radiolysis simulations where two different geometrical constraints were used and the second model is the recently developed steady-state model. The results of the simulations are compared to each other and to experimental results from spent fuel leaching experiments performed in carbonate containing aqueous solution under Ar-atmosphere. The influence of radiolytically produced H2 is incorporated (on the basis of a recently suggested mechanism) in both models and this reproduces the experimentally observed inhibition of spent fuel dissolution fairly well. The conventional radiolysis model reproduces the experimental concentrations of the radiolysis products H2 and O2 very well while it fails to reproduce the experimental H2O2 concentration. The reasons for this are discussed. The general trend in uranium concentration as a function of time is reproduced by both the conventional radiolysis model and the steady-state model. The conventional radiolysis model (in which the radiation dose is homogeneously distributed in the whole liquid volume) underestimates the uranium concentration while the steady-state model, which represents the worst case scenario, overestimates the concentrations to some extent. When applying the conventional radiolysis model, assuming that all the radiation energy is deposited within 40 μm from the fuel surface, the uranium concentrations during the initial part of the experiments are reproduced quantitatively. The differences between the models and the applicability of the models are discussed in some detail.