Coupled hydromechanical modelling of fault zones in clay-rich rock: towards management of fault-related risks in gigatonne-scale CO2 storage

Achieving climate neutrality by 2050, a target set by the EU, requires significant scale-up of CO₂ storage. One economically attractive option is CO₂ storage clusters where multiple operators inject CO₂  into the same aquifer that may reach hundreds of kilometers in size. Injecting CO₂ from multiple locations in the same aquifer introduces challenges, such as pressure interactions, even in the far-field beyond the storage project. Pressure build-up caused by neighboring fields could result in induced seismicity, potentially opening pathways in the fault damage zone (FDZ) within the caprock through which CO₂ could escape. Such risks must be identified early to avoid injecting CO₂ into problematic areas within a storage region.

For early-stage screening of high-risk locations, reduced-complexity methods, such as vertical equilibrium models, are suitable since the computational demand of detailed reservoir simulations is prohibitively high. However, currently these models do not account for the effect of pressure change on flow behaviour through the FDZ and the risk of fault reactivation since few constitutive relations exist, limiting their ability to reliably screen risks.

Developing such relations requires a detailed, project-scale understanding of key parameters controlling the risk of fault reactivation and flow behaviour through the FDZ, which includes multiscale fractures, ranging from grains to the thickness of the caprock. Those fractures cause stress perturbations, leading to heterogeneous permeability evolution that needs to be accounted for in models to reliably quantify upscaled flow properties under stress, as well as the risk of fault reactivation, in a realistic yet computationally feasible way. 

We investigate the hydromechanical behaviour of the main fault and surrounding FDZ, using project-scale, sequentially-coupled simulations that include up to thousands of multiscale, planar fractures, following observations of realistic FDZ architectures. Stress, shear- and normal-displacement relationships for single mudrock fractures, derived from experimental data, are used to model intrinsic fracture permeability evolution and resultant upscaled FDZ permeability, while a semi-analytical method is employed to simulate aseismic shear slip on the main, planar fault surrounded by the FDZ, during the nucleation phase, and therefore to assess the risk of fault reactivation. The classical crack tensor theory is used for elastic geomechanical simulations, while the embedded discrete fracture model (EDFM) is employed for single-phase fluid flow simulations.

Preliminary results indicate significant stress perturbations, particularly in regions with higher fracture density (i.e., close to the fault core), which enhanced upscaled FDZ permeability under realistic stress boundary conditions. Initial simulations suggest that stress boundary conditions, and orientation and frictional properties of fractures and the main fault play an important role in controlling the magnitude of pore pressure increase required for fault reactivation and significant increase in upscaled FDZ permeability. Work is ongoing concerning systematic sensitivity analyses, with the aim of identifying key parameters controlling flow behaviour and the risk of fault reactivation. The results are expected to inform the development of constitutive relations required for screening of the risks in reduced-complexity models.