A mesoscopic chemical reaction kinetics model to predict the formation of zirconium oxide and hydride accumulation light-water reactor (LWR) fuel clad is presented. The model is designed to include thermodynamic information from ab initio electronic structure methods as well as parametric information in terms of diffusion coefficients, thermal conductivities and reaction constants. In contrast to approaches where the experimentally observed time exponents are captured by the models by design, our approach is designed to be predictive and to provide an improved understanding of the corrosion process. We calculate the time evolution of the oxide/metal interface and evaluate the order of the chemical reactions that are conducive to a t 1/3 dependence. We also show calculations of hydrogen cluster accumulation as a function of temperature and depth using spatially dependent cluster dynamics. Strategies to further cohesively integrate the different elements of the model are provided.