Abstract
Computational codes are necessary tools for geochemical modeling of the alteration of minerals due to their ability to handle key mechanisms, such as dissolution, precipitation, diffusion, and convection at many temporal and spatial resolutions. Modeling glass corrosion specifically requires a description of the amorphous layer that forms on the surface of the glass and its effect on glass alteration kinetics. The objective of the GRAAL model (glass reactivity in allowance of the alteration layer) is both to provide a simple implementation of the passivation process in a reactive transport code and to provide data relative to the composition and the solubility of the amorphous layer. The size and properties of the protective amorphous layer drives the glass alteration rate, with regard to passivation; the greater the quantity of the protective amorphous layer, the lower the dissolution rate of the primary mineral. Here, concepts, equations, and implementation of GRAAL are reported. Simple glass alteration experiments are used to apply the model and measure parameters. The International Simple Glass used for nuclear glass long-term behavior studies is at the center of the glass compositions studied.