Numerical Modeling of Capillary Hysteresis and Coupled Elastoplasticity for Geological Carbon Storage

<p>When CO₂ is injected into the saline aquifer or depleted reservoir for geological carbon storage, physical processes are tightly coupled, affecting CO₂flooding and its trapping mechanisms.</p> <p>For example, injection induces pore pressure build-up and dilation of pore space, which can uplift the ground surface or tensile/shear failure of the caprock which may result in the leakage of CO₂. Thus, rigorous analyses of coupled flow and geomechanics are necessary to predict the long-term security of geological carbon storage. In this study, we focus on two irreversible (path-dependent) processes that are coupled through flow and geomechanics: hysteretic capillary pressure in flow and elastoplasticity in geomechanics. Hysteresis in capillary pressure during drainage and imbibition processes can be seen as mechanical energy dissipation. We employ our previously proposed numerical model based on the 1D elastoplasticity algorithm for constitutive relation of the hysteretic capillary pressure in two-phase flow, i.e., capillary pressure and irreducible water saturation. In particular, we model the irreducible (plastic) water saturation being attributed to the part from the hysteretic capillary pressure, which yields a mathematically well-posed problem. We implement the irreversible flow and geomechanics simulation, calculating the residual saturations and plastic strain from each iteration of flow and geomechanics, as we employ the fixed-stress sequential method solving coupled flow and geomechanics.</p> <p>From the numerical experiments, we find robust computations of the coupled processes, highlighting the coupled effects of capillary hysteresis and elastoplasticity. As residual/capillary and structural trappings are major trapping mechanisms for CO₂ geological storage, the proposed constitutive relation and algorithm for coupled path-dependent processes can predict flooding and trapping of CO₂ underground more accurately. &#160;</p>