Analytical estimation of risk due to pressure buildup during CO2 injection in deep saline aquifers

Abstract

When CO2 is injected in a deep saline aquifer on the scale of tens of millions of tonnes, pressure buildup in the aquifer during injection will be a critical issue. Fracturing, fault activation and leakage of brine along pathways such as abandoned wells occur at various threshold fluid pressures, so operators and regulators will be concerned with pressure elevation at considerable distances from the injection well. Thus a critical contour of overpressure (CoP) is a convenient proxy for risk. The location of this contour varies depending on the target aquifer properties (porosity, permeability etc.) and the geology (presence and conductivity of faults). The CoP location also depends on relative permeability, and we extend the three-region injection model [1,2] to derive analytical expressions for a specific CoP as a function of time. The risk of pressure-induced leakage from the aquifer can therefore be cast in terms of phase mobilities and speeds of saturation fronts. We consider two boundary conditions at the aquifer drainage radius, constant pressure or an infinite aquifer. The model provides a quick tool for estimating pressure profiles. Such tools are valuable for screening and ranking sequestration targets. Because pressure profiles are relatively insensitive to spatial variability in aquifer permeability, a simple model can provide as good an estimate of pressure buildup as a sophisticated simulation that requires much longer to set up and to run. Relative permeability curves measured on samples from seven potential storage formations [3] are used to illustrate the effect on the CoPs. The relative permeability curve with the largest two-phase region mobility (MBL) gives the smallest pressure buildup, so that a given CoP is nearest to the injector. All else being the same, decreasing the two-phase-region mobility increases the risk associated with pressure elevation during injection. Thus characterizing relative permeability should be included in the implementation of CO2 storage projects. In the case of a constant pressure boundary, the CoP for small overpressures is time-invariant and independent of relative permeability. This result significantly reduces the uncertainty in predicting risk associated with small overpressures. Depending on the relative values of overall mobilities of two-phase region and of brine region, the risk due to a critical CoP which lies in the two- phase region can either increase or decrease with time. In contrast, the risk due to a CoP in the drying region always decreases with time. This analysis helps set limits on the maximum possible radial extent of a desired CoP, thereby providing a basis for establishing an Area of Review (AoR) for the storage project monitoring. The assumption of constant pressure boundaries is optimistic in the sense that CoPs extend the least distance from the injection well. We extend the analytical model to infinite-acting aquifers to get a more widely applicable estimate of risk. An analytical expression for pressure profile is developed by adapting water influx models from traditional reservoir engineering to the “three-region” saturation distribution. For infinite-acting boundary condition, the CoP trends depend on same factors as in the constant pressure case, and also depend upon the rate of change of aquifer boundary pressure with time. Commercial reservoir simulators are used to verify the analytical model for the constant pressure boundary condition. The CoP trends from the analytical solution and simulation results show a good match.