Modeling brine leakage to shallow aquifer through an open wellbore using T2WELL/ECO2N

The efficiency and sustainability of CO2 storage in deep saline formations depends on the integrity of the overlying caprocks. Existing oil and gas wells, which penetrate the formations, are potential leakage pathways. In this paper, T2WELL/ECO2N, a coupled wellbore-reservoir flow simulator of carbon dioxide and brine, is employed to analyze the effects of salinity and excess pore pressure (EPP) on brine leakage from an injection reservoir to a shallow fresh aquifer. The effectiveness of the Equivalent Porous Media (EPM) approach, a widely used approximation of wellbore flow in wellbore-reservoir simulations, is also evaluated. A hypothetical model is built to create test cases with EPP of 0.1 MPa, 0.5 MPa and 1.5 MPa under low-, middle-, and high-salinity. The results show that a quasi-steady-state leakage rate will be quickly established if the EPP caused by CO2 injection is higher than the threshold EPP. However, the brine salinity has an important impact on the brine leakage rate including the threshold EPP below which no leakage occurs. The leakage rate decreases with the increase of the salinity whereas the threshold EPP increases with the increase of the salinity. For the same brine salinity, the quasi-steady-state leakage rate increases with the EPP nonlinearly. The leakage rate calculated by the EPM approach is very sensitive to the equivalent permeability of the wellbore, which however, is not an intrinsic parameter of a given wellbore-aquifer system. Although the EPM approach could reproduce the dynamic leakage rate reasonably well with a fitted permeability, it could over- or underestimates the leakage rate by orders in magnitude for the same system depending on whether the true leakage rate itself is smaller or larger than that of the fitting case due to changes in the salinity and/or the EPP through an abandoned well. The dependence of the equivalent permeability on the flow rate makes the EPM approach not to be a viable method for predicting wellbore leakage except for the cases with very small leakage rates. (C) 2012 Elsevier Ltd. All rights reserved.