Publisher Summary One of the key issues surrounding the underground storage of carbon dioxide is the prediction of the long-term fate of the injected gas, as a means of assessing whether a potential injection site is likely to contain the carbon dioxide for very long time periods. This involves the study of a variety of factors such as seal capacity and geomechanical integrity, but it also depends on the physical process by which carbon dioxide will dissolve in the formation water. After injection ceases, the lesser density of the free carbon dioxide compared to the formation brine causes a vertical migration of the gas, until the gas is distributed between a mobile layer underneath the caprock, and gas trapped at the residual gas saturation throughout the injection zone. For injection into a structural trap, such as might occur in a depleted gas field, the gas cap could be quite thick. However, in a regionally extensive saline formation, such as at Sleipner, the gas can easily spread out laterally and migrate beneath the top seal. In this latter case, the thickness of the gas layer is initially only limited by capillary pressure, and there is a competition between migration and dissolution of the gas in the formation water. The balance between these effects determines the ultimate distance to which the carbon dioxide will migrate from the injection site. The saturation of underlying brine with dissolved CO2 creates a density instability, and on long time scales, this causes convective mixing, greatly increasing the overall rate of dissolution compared to a purely diffusive mechanism.