Among polymers used as encapsulant in photovoltaic (PV) modules, poly(ethylene-co-vinyl acetate), or EVA, is the most widely used, for its low cost and acceptable performances. When exposed to weather conditions, EVA undergoes degradation that affects overall PV performances. Durability prediction of EVA, and thus of the module, is a hot topic in PV process industry. To date, the literature lacks of long-term predictive computational models to study EVA aging. To fill this gap, a computational framework, based on the finite element method, is proposed to simulate chemical reactions and diffusion processes occurring in EVA. The developed computational framework is valid in either case of environmental or accelerated aging. The proposed framework enables the identification of a correspondence between induced degradation in accelerated tests and actual exposure in weathering conditions. The developed tool is useful for the prediction of the spatio-temporal evoltition of the chemical species in EVA, affecting its optical properties. The obtained predictions, related to degradation kinetics and discoloration, show a very good correlation with experimental data taken from the literature, confirming the validity of the proposed formulation and computational approach. The framework has the potential to provide quantitative comparisons of degradation resulting from any environmental condition to that gained from accelerated aging tests, also providing a guideline to design new testing protocols tailored for specific climatic zones.