To safely store carbon dioxide in enhanced oil recovery/ CO2 sequestration projects it is important to ensure the integrity of the caprock during and after production and injection. A change in fluid pressure and temperature within a porous reservoir will generally induce stress changes within the reservoir and the rocks that surround it. Amongst the potential hazards resulting from these induced stress changes is the reactivation of existing faults or fractures and inducing new fractures, which may breach the hydraulic integrity of the caprock that bounds the reservoir.<p> The theories of inclusions and inhomogeneities have been used in this research to derive semi-analytical and closed-form solutions for induced stress change during pore pressure change within a reservoir and in the surrounding rock, under plane strain and axisymmetric conditions. Methods have been developed to assess fault reactivation and induced fracturing during injection or production within a reservoir. The failure stress change concept for a Coulomb failure criterion has been used to study the likelihood of fault reactivation and induced fracturing within the reservoir. Formulations have been adopted to calculate the critical pressure change for fault reactivation and induced fracturing within the reservoir and in the surrounding rock during injection and production. Sensitivity analysis has been performed to study the effects of different parameters such as initial in-situ stress, reservoir geometry, reservoir depth, reservoir tilt or dip , material property contrast between the reservoir and surrounding rock, fault geometry, fault strength, and intact rock strength. General patterns of induced stress change, in-situ stress evolution, fault reactivation, and induced fracturing have been identified.<p> The developed methodologies have been applied to six different case studies: fault reactivation analysis in the entire field for a synthetic case study; induced fracturing analysis in the entire field in a synthetic case study; fault reactivation and induced stress change analysis within the Ekofisk oil reservoir in North Sea; fault reactivation analysis in the Lacq gas reservoir in France; the Weyburn-Midale EOR/CO2 Storage project in southeast Saskatchewan; and acid gas injection in Zama oil field, Alberta. The results of these case studies show good consistency with field observation, and physical and numerical models.<p> The generality, simplicity, and straightforwardness of the developed methodologies, along with their flexibility to model different plausible scenarios and their ease of implementation for systematic sensitivity analyses makes them suitable for decision-making and uncertainty management, specifically in early stages of reservoir development or site assessment for geological sequestration of carbon dioxide.

To safely store carbon dioxide in enhanced oil recovery/ CO2 sequestration projects it is important to ensure the integrity of the caprock during and after production and injection. A change in fluid pressure and temperature within a porous reservoir will generally induce stress changes within the reservoir and the rocks that surround it. Amongst the potential hazards resulting from these induced stress changes is the reactivation of existing faults or fractures and inducing new fractures, which may breach the hydraulic integrity of the caprock that bounds the reservoir.<p> The theories of inclusions and inhomogeneities have been used in this research to derive semi-analytical and closed-form solutions for induced stress change during pore pressure change within a reservoir and in the surrounding rock, under plane strain and axisymmetric conditions. Methods have been developed to assess fault reactivation and induced fracturing during injection or production within a reservoir. The failure stress change concept for a Coulomb failure criterion has been used to study the likelihood of fault reactivation and induced fracturing within the reservoir. Formulations have been adopted to calculate the critical pressure change for fault reactivation and induced fracturing within the reservoir and in the surrounding rock during injection and production. Sensitivity analysis has been performed to study the effects of different parameters such as initial in-situ stress, reservoir geometry, reservoir depth, reservoir tilt or dip , material property contrast between the reservoir and surrounding rock, fault geometry, fault strength, and intact rock strength. General patterns of induced stress change, in-situ stress evolution, fault reactivation, and induced fracturing have been identified.<p> The developed methodologies have been applied to six different case studies: fault reactivation analysis in the entire field for a synthetic case study; induced fracturing analysis in the entire field in a synthetic case study; fault reactivation and induced stress change analysis within the Ekofisk oil reservoir in North Sea; fault reactivation analysis in the Lacq gas reservoir in France; the Weyburn-Midale EOR/CO2 Storage project in southeast Saskatchewan; and acid gas injection in Zama oil field, Alberta. The results of these case studies show good consistency with field observation, and physical and numerical models.<p> The generality, simplicity, and straightforwardness of the developed methodologies, along with their flexibility to model different plausible scenarios and their ease of implementation for systematic sensitivity analyses makes them suitable for decision-making and uncertainty management, specifically in early stages of reservoir development or site assessment for geological sequestration of carbon dioxide.

The Copper Creek thrust fault in the southern Appalachians places Cambrian over Ordovician sedimentary strata. The fault accommodated displacement of 15-20 km at 100-180 °C. Along the hanging wall-footwall contact, microstructures within a ~2 cm thick calcite and shale shear zone suggest that calcite, not shale, controlled the rheology of the shear zone rocks. While shale deformed brittley, plasticity-induced fracturing in calcite resulted in ultrafine-grained (<1.0 μm) fault rocks that deformed by grain boundary sliding (GBS) accommodated primarily by diffusion creep, suggesting low flow stresses. Optical and electron microscopy of samples from a transect across the footwall shale into the shear zone, shows the evolution of rheology within the shear zone. Sedimentary laminations 1 cm below the shear zone are cut by minor faults, stylolites, and fault-parallel and perpendicular calcite veins. At vein intersections, calcite grain size is reduced (to ~0.3 μm), and microstructures include inter-and-intragranular fractures, four-grain junctions, and interpenetrating boundaries. Porosity rises to 6 percent from <1 percent in coarse (25 μm) areas of calcite veins. In coarse-grained calcite, trails of voids follow twin boundaries, and voids occur at twin-twin and twin-grain boundary intersections. At the shear zone-footwall contact, a 350 μm thick calcite band contains coarseand ultrafine-grained layers. Ultrafine-grained (~0.34 μm) layers contain microstructures similar to those at vein intersections in the footwall and display no lattice-preferred orientation (LPO). Coarse-grained layers cross-cut grain-boundary alignments in the ultrafine-grained layers; coarse grains display twins and a strong LPO. Within the shear zone, ultrafine-grained calcite-aggregate clasts and shale clasts (5-350 μm) lie within an ultrafine-grained calcite (<0.31 μm) and shale matrix. Ultrafinegrained calcite (<0.31 μm) forms an interconnected network around the matrix shale. Calcite vein microstructures suggest veins continued to form during deformation. Fractures at twin-twin and twin-grain boundary intersections suggest grain size reduction by plasticity-induced fracturing, resulting in <1 μm grains. Interpenetrating boundaries, four-grain junctions, and no LPO indicate the ultrafine-grained calcite deformed by viscous grain boundary sliding. The evolution of the ultrafine-grain shear zone rocks by a combination of plastic and brittle processes and the deformation of the interconnected network of ultrafine-grained calcite by viscous GBS enabled a large displacement along a narrow fault zone.

As a safe long-term management of nuclear wastes, deep geological repositories (DGRs) have been proposed or currently being constructed in several countries. The host rocks in DGRs are saturated with water after the geological disposal facilities (GDFs) are closed and sealed. Significant gas can be generated due to several processes, e.g., the metal corrosion, water radiolysis or microbial reaction of organic materials, etc. The generated gas is anticipated to span throughout the long-term disposal of waste, which may jeopardize the stability of host rocks. Correspondingly, the performance of GDF will be affected since the host rocks provide a final impediment to the radionuclide transport. As gas migration in saturated host rocks is a highly coupled hydro-mechanical (HM) process, either gas-induced micro-fracturing or macro-fracturing may contribute to the development of preferential gas pathways, which needs to be concerned to ensure the feasibility and safety of geological disposal. Current numerical studies on the gas migration behavior devoted to explaining the experimental phenomena in the gas injection tests conducted on the rock materials, in which some behaviors still cannot be well represented, i.e., gas induced fracturing, volulme dilation, anisotropic radial deformation. Therefore, to better represent the actual physical process of preferential gas flow, two modelling frameworks, i.e., macroscopic HM framework and two-scale HM framework, are proposed in the PhD study. For the macroscopic HM framework, a double porosity model is firstly developed based on the dual continuum method, in which the volumetric strains of the porous continuum (PC) and fractured continuum (FC) are work-conjugated to the respective effective stress level. The treatment in two types of porosity allows us to capture that the opening/closure of the fractures is caused by the interaction between the dilation of the PC and the dilation of the FPM, which is beneficial to describe the gas induced fracturing in an implicit way. Then, an enriched embedded fracture model (EFM) is proposed to address the mechanical behavior of fractures. A hyperbolic relation of fracture deformability is incorporated into the rock matrix, as a result the fractured rock shows a nonlinear elastic behavior, which can capture the stiffness degradation due to fracture opening. The equivalent continuum method is provided to derive the effective compliance tensor, which includes the transverse isotropic matrix and two fracture sets. Using the enriched EFM with a three-dimensional (3D) geometry is able to capture the anisotropic radial deformation during gas migration. Although the macroscopic HM framework is able to capture the major HM behaviors related to preferential gas flow, the development of gas dilatant pathways is still represented in an implicit way. Therefore, a two-scale HM framework is developed to explicitly simulate the development of preferential gas pathways. Initiating from the periodically distributed microstructures with microcracks, the asymptotic homogenization method is used to derive the macroscopic governing equations coupled with the normalized damage variable. The time-dependent damage evolution law is obtained from the microscopic mechanical energy analysis for evolving microcracks. Both time effect and size effect are incorporated in the damage model that will affect the overall HM behavior of rocks. The developed two-scale HM framework with single gas flow can qualitatively capture important behaviors, such as the discrete pathways, localized gas flow, unstabilized fracture branching. More specifically, the simulated results demonstrates that the inter-connection of fractures from gas inlet to outlet is a prerequisite for gas breakthrough, accompanied by large amounts of gas flowing out of the sample and a rapid drop in gas injection pressure. Incorporating water flow in the two-scale framework allows the model to quantitatively reproduce the experimental phenomena observed in the laboratory air injection tests, such as gas pressure evolution and mechanical deformation. More importantly, the model exlpaines that the significant differences in controlling gas breakthrough and mechanical deformation are resulting from the arbitrary nature of microstructural heterogeneities. To account for the gas-water interaction in the two-scale HM framework, a fully coupled two-phase flow and elaso-damage model is developed to simulate the laboratory and in-situ gas injection experiments. The model can quantitatively capture the experimental behaviors, e.g., gas pressure evolution and non-desaturation phenomenon. Furthermore, model results show that the highly localized fracture pathways are the major places where gas and water interacts each other, and as a result the rock is still kept fully saturated. As a whole, the obtained numerical results are synthesized and analyzed, the pros and cons of the developed models are discussed. To better improve the model performance, some recommendations are proposed for the future studies.