Modeling of Gas Migration through Low-permeability Clay Using Information on Pressure and Deformation from Fast Air Injection Tests

The characterization of gas migration through low-permeability clay formations has been a focus of R&D programs for radioactive waste disposal, which is also of great importance for shale-gas exploration, and cap-rock behavior of hydrocarbon reservoirs and CO2 sequestration. Laboratory tests on Opalinus clay cores from a shallow borehole in the Mont Terri Underground Research Laboratory (URL) and from a deep borehole in northern Switzerland included specific water and air injections tests, as well as oedometer and isotropic compression tests. For tests under different confining stress conditions, the rock compressibility was determined and the measured deformation was used to estimate changes in void ratio and to derive a relationship between void ratio and stress, and the corresponding changes in permeability as a function of changes in porosity. For the shallow cores from Mont Terri, largely linear-elastic deformation associated with the gas injection test could be inferred and the change in void ratio was accounted for by the pore compressibility. The corresponding change in permeability was obtained from the results of the water tests, indicating a log-linear relation between permeability and porosity. The derived porosity change and corresponding change in permeability was implemented in the standard TOUGH2 code, which reproduced the measured gas test results using fitted water-retention data derived from laboratory measurements. Similar injection tests performed on Opalinus clay cores from the borehole at greater depth showed overall similar behaviour, but at lower permeabilities, lower pore compressibilities and lower changes in porosity. These cases indicated non-linear behaviour which was implemented using an effective stress-dependent porosity change and associated change in permeability. In addition, the anisotropy associated with the bedding of the clay formation was considered by assuming different properties for “soft” and “hard’ layers to account for storage capacity for the injected gas prior to gas breakthrough. The computed change in the overall porosity could be compared to the measured axial deformation during the gas injection test and was used for calibration of the parameters describing the relationship between the effective stress and porosity and the corresponding change in permeability and capillary pressure.