oa Coupling Unified Ensemble Modelling with Invasion Percolation Simulation for Early Appreciation of Risk in CO2 Containment

During the initiation stages of CO2 storage projects, predicting the ultimate plume distribution presents a significant challenge due to the presence of substantial uncertainties in subsurface conditions. These uncertainties, when predicting plume distribution, need to be handled in a consistent way to better assess risks related to containment and to ultimately determine a storage site as having a low risk profile. This paper summarizes a method that utilizes ensemble-based approaches to create a range of realizations that represent uncertainty from the outset, to characterize uncertainties in static geomodels. A dynamic simulation is performed on each of these geomodel realizations using a modified invasion percolation simulator to improve model runtimes and deliver accurate representations of plume migration. The simulation outcomes are compared to known plume distributions to assess conceptual leakage risks. What we discovered was that even in a highly studied storage site with relatively low heterogeneity, our predictions remain highly sensitive to uncertainties. This underscores the need for robust assessment methods when evaluating new targets, regardless of how much we may think we already know.