Estimates of CO₂ leakage along abandoned wells constrained by new data

The viability of carbon capture and geological storage (CCS) projects depends in part on the risk that injected CO₂ or displaced pore fluid will leak out of the target formation into surrounding formations or to the surface. Abandoned oil and gas wells, of which millions exist both throughout the United States and globally, form a potential conduit for this leakage. Recently, specific field measurements have been made to quantify the range of effective permeabilities that can be expected in abandoned wells, enabling us to, for the first time, combine field-scale numerical simulations of CO₂ sequestration in deep saline aquifers with real data on effective permeabilities of leaky wells. Using a previously developed semi-analytical reservoir simulator that can accommodate an arbitrary sequence of alternating aquifers and aquicludes, as well as an arbitrary number of leaky wells, we investigated how the amount of CO₂ that leaks out of the target formation depends on the spatial density of nearby abandoned wells and their effective permeability. Furthermore, we assess the influence that variations in pressure and temperature found between geological targets have on this dependency. We find that the observed differences in leakage between geological targets are controlled almost exclusively by differences in density of CO₂ at the local subsurface conditions, causing the CO₂ plume to contact a different number of wells when injecting at the same constant mass rate. We quantitatively assess the results obtained from our numerical experiments by combining them with the permeability data that have recently become available, typical spatial densities of abandoned wells, and performance requirements put forward in the literature. Our results indicate that leakage of CO₂ through abandoned wells is unlikely to be a major limitation in storage security of CCS projects.