Global warming attracts great social, economic, political and scientific attention. A major proportion of the carbon dioxide (CO2) emitted in the atmosphere is due to anthropogenic activities and represents one of the main causes of this global problem. A possible solution is represented by carbon sequestration: CO2 is captured from power plants and injected in underground geological formations, where it dissolves into the resident fluid (brine) and can be safely stored for hundreds of years. In this frame, the properties of the rocks play a key role: after injection, CO2 follows sinuous pathways among the rock grains and spreads in a complex manner, making predictions on the long-term dynamics hard to obtain. For this reason, the identification of suitable sequestration sites and the design of the injection process is still a challenging task. Moreover, injection of CO2 takes place at depths between 1 and 3 km beneath the earth surface, which makes in-situ measurements hard to obtain: simulations and lab-scale experiments are essential. To make practical and prudent decisions about the future energy production strategies, the European Union (EU) must be able to accurately identify its carbon storage capacity. This research work improves our understanding and design capabilities of carbon storage processes in geological formations. This study focused on the analysis and interpretation of experiments and simulations of convection in porous media. The results are used to develop models that describe the rock properties, contributing also to improve commercial reservoir simulators. I (1) examined the effect of dispersion via innovative experiments in bead packs, (2) quantified the effect of dispersion with state-of-art numerical pore-scale simulations, and (3) identified appropriate models of dispersion for large-scale simulations.
