Nuclear Technology / Volume 160 / Number 3 / December 2007 / Pages 337-345
Technical Paper / Materials for Nuclear Systems / dx.doi.org/10.13182/NT07-A3904
A combined transport and radiolysis model is proposed in this paper to predict the oxidation/dissolution of UO2 under alpha radiolysis of water. The UO2-water interface is divided into an arbitrary number of layers. The Chemsimul kinetic code is used for radiolysis calculations in each layer, and the modeling of transport between the layers is based on Fick's law. The calculation proceeds in an iterative way, and an alpha dose rate profile is taken into account as input data. To limit the calculation time, which depends on the computer capacity and the duration of the leaching experiment described, a compromise between the thickness and the number of cells has to be found. At present, simulations of leaching experiments lasting several days cannot be carried out because of the very long calculation time. However, the calculation has been compared with experimental results obtained under irradiation at high flux levels of a UO2-water interface subjected to a beam of He2+ particles generated by a cyclotron. Owing to computer time limitation, calculations are carried out by considering 200 layers, each 10 m thick, to simulate 1-h experiments. In the experimental geometry (monoenergetic linear alpha beam), the alpha dose rate profile is well described by a summation of Bragg curves. The comparison relates to experiments performed in aerated and deaerated media at a high flux of 3.3 × 1010 cm-2s-1 and 3.3 × 1011 cm-2s-1. The calculated uranium content in solution is three times lower than the experimental value, and the hydrogen peroxide concentration is ten times lower in aerated media. In deaerated media, however, the comparison is quite good. Finally, a calculation was carried out with a large imposed dissolved hydrogen concentration in solution to check the inhibition of matrix dissolution. The release of uranium in solution is relatively high despite the hydrogen concentration in solution because of the primary formation of hydrogen peroxide. This is probably because of a lack of knowledge concerning the inhibitor mechanism under alpha radiolysis (influence of the surface under alpha irradiation, hydrogen activation, validity of primary radiolytic yield in presence of H2, etc.), which is not taken into account in our calculations.