Uncertainty quantification through reduced modelling of CO2 injection sites

The project involves building on other work to model how to create physically meaningful models of the long-thin flows of CO2 (gravity currents) that naturally occur when storing CO2. This will involve grappling with the inherent heterogeneities within geological formations, with processes like dissolution which happen on a variety of length scales, and with the complex multi-phase flow of CO2 which can lead to trapping by surface tension. Simplified models will be used to identify the dominant physical processes. This PhD project looks to both understand those reduced physics models and then to utilise this computational efficiency to understand the sensitivity to the unknowns in
the problem, chiefly the geological reservoir itself. An additional aim is to understand how far CO2 might propagate from the injection point, which is sensitively dependent on the underlying geology. Viewing this more in a probabilistic sense, how far might we expect the CO2 to progress can be addressed using a simplified, gravity current model for the CO2 while using a more statistical representation of the underlying geology to give predictive maps of the risk of CO2 propagating a certain distance.