• Energy Research

Subsurface reservoirs are targeted formations for geologic carbon dioxide (CO2) storage. Even if proper management of injection pressures minimizes the risks of induced seismicity, high pressure CO2 can interact with brine-saturated host rock and cause microstructural changes that lead to alterations in poromechanical properties of the rock. The effect is well pronounced in carbonate-rich rock, but observations on silica-rich reservoirs are ambiguous. In this study, we report a broad range of experiments performed on Berea sandstone, consisting mainly of quartz (∼90%), in three different states: pristine, thermally damaged, and thermally damaged then treated with liquid CO2. Drained and undrained poromechanical tests, ultrasonic velocity measurements, acoustic emission (AE), X-ray diffraction (XRD), and petrographic analyses are conducted. The tests reveal that thermal damage alone does not significantly affect poromechanical properties. However, CO2 injection does affect strength (10–15 % decrease), permeability (up to 100% increase), porosity (10% increase), and elastic creep rate (more than twice) ; corresponding icrostructural changes were observed from XRD test results. At the same time, the poroelastic moduli measured in triaxial compression experiments and load- induced fracture processes, as interpreted through acoustic emission data collected in uniaxial compression tests, were affected insignificantly. These experimental observations provide better understanding of the mechanical behavior of low-carbonate reservoir rocks that are subjected to high pressure CO2 injection.

CO2 injection in extensive saline aquifers that present no faults is unlikely to damage the caprock sealing capacity. In contrast, CO2 injection in closed reservoirs will induce a large pressure buildup that may reactivate the low-permeable faults that bound the reservoir. However, the vast majority of CO2 storage formations will be extensive saline aquifers bounded by a limited number of low-permeable faults. Such storage formations have received little attention and are the focus of this study. We model an extensive aquifer bounded by a heterogeneous low-permeable fault on one side and having open boundaries on the other sides. Simulation results show that the storage formation pressurizes between the injection well and the low-permeable fault, causing total stress changes and effective stress reduction around the fault. These changes lead to yielding of the fault core that is next to the lower half of the storage formation when injecting in the hanging wall. The yield of the fault core would induce a sequence of microseismic events with accumulated seismic moment equivalent to an earthquake of magnitude 1.7, which would not be felt on the ground surface and would not enhance permeability of the ductile clay-rich fault. © 2017 The Authors. V.V. acknowledges support from the ‘EPFL Fellows’ fellowship programme co-funded by Marie Curie, FP7 Grant agreement no. 291771. R.M. activities are sponsored by SCCER-SoE (Switzerland) grant KTI.2013.288 and Swiss Federal Office of Energy (SFOE) project CAPROCK #810008154. Peer reviewed

CO2 leakage is a major concern for geologic carbon storage. To assess the caprock sealing capacity and the strength of faults, we test in the laboratory the rock types involved in CO2 storage at representative in-situ conditions. We use the measured parameters as input data to a numerical model that simulates CO2 injection in a deep saline aquifer bounded by a low-permeable fault. We find that the caprock sealing capacity is maintained and that, even if a fault undergoes a series of microseismic events or aseismic slip, leakage is unlikely to occur through ductile clay-rich faults. © 2017 The Authors. Published by Elsevier Ltd. V.V. acknowledges financial support from the “TRUST" project (European Community's Seventh Framework Programme FP7/2007-2013 under grant agreement n. 309607) and from “FracRisk" project (European Community's Horizon 2020 Framework Programme H2020-EU.3.3.2.3 under grant agreement n. 640979). R.M. acknowledges partial support from the Center for Geologic Storage of CO2, an EFRC funded by the U.S. DOE, Office of Science, BES, under Award DE-SC0C12504. Peer reviewed

Caprock formations are intended to prevent upwards carbon dioxide (CO2) migration to the surface during CO2 geological storage. Caprock interaction with CO2, as well as its potential consequences, requires to be predicted, and thus, need to be studied experimentally. Laboratory investigations of caprock behavior are complex due to its low permeability, and the scarcity of experimental studies involving high-pressure CO2 injection into caprock representatives puts this difficulty into manifest. In this study, we perform laboratory experiments in an oedometric cell on intact and remolded Opalinus clay (Jurassic shale), evaluating the breakthrough pressure and permeability for liquid and supercritical CO2. Intact and remolded shale specimens present intrinsic permeabilities of 10-21 m2 to 10-20 m2, respectively. Applied axial stress ranges from 27 MPa to 42 MPa and the pressure and temperature conditions are representative of a caprock at a depth of 800 m. We found that the microstructure of the caprock has a great effect on the material properties. The intrinsic permeability of a more tight material (intact Opalinus clay) is around two times lower than that of remolded shale, which has a more open microstructure. Additionally, the intact rock becomes 30 times less permeable to CO2 than the remolded shale, which implies that the CO2 relative permeability is 15 times smaller for intact rock than for remolded shale. On the other hand, CO2 breakthrough pressure for the tighter material is almost three times lower than for the more permeable remolded shale. Breakthrough pressure of the remolded shale ranges from 3.9 MPa to 5.0 MPa for liquid CO2 and from 2.8 MPa to 4.6 MPa for supercritical CO2. For the intact shale, breakthrough pressure is 0.9 MPa for liquid CO2 and 1.6 MPa for supercritical CO2. Thus, the breakthrough pressure cannot be correlated with the intrinsic permeability of the caprock. © 2017 The Authors. Opalinus clay cores were provided by Swisstopo in the framework of Mont Terri Project, CS-C experiment. R. Makhnenko activities are sponsored by SCCER-SoE (Switzerland) grant KTI.2013.288 and Swiss Federal Office of Energy (SFOE) project CAPROCK #810008154. V. Vilarrasa acknowledges support from the ‘EPFL Fellows’ fellowship programme co-funded by Marie Curie, FP7 Grant agreement no. 291771. Peer reviewed

Abstract Failure in brittle rock happens because micro-cracks in the crystal structure coalesce and form a localized fracture. The propagation of the fracture is in turn strongly influenced by dissipation in the fracture process zone. The classical theory of linear elastic fracture mechanics falls short in describing failure when the dissipation in the fracture process zone is non-negligible; thus, a non-linear theory should be employed instead. Here we present a study in which we explore the characteristics of the fracture process zone in granite. We have combined fracture tests on Adelaide black granite with acoustic emission detection and finite element analyses by using a non-local integral plastic-damage constitutive theory. We have further employed the theory of configurational mechanics to support our interpretation of the evolution of the fracture process zone with strong energy-based arguments. We demonstrate that the size of the fracture process zone is non-negligible and dissipative phenomena related to micro-cracking play an important role. Our results indicate this role should be assessed case by case, especially in laboratory-sized analyses, which mostly deflect from theories of both size-independent plasticity and linear elastic fracture mechanics. When strong non-linearities occur, we show that fracture energy can be correctly computed with the help of configurational mechanics and that complex numerical simulation techniques can substantially facilitate the interpretation of experiments designed to highlight the dominant physical mechanisms driving fracture.

Abstract In the light of growing concerns for the climate change, it is of particular interest for governments to encourage efficient capture and safe storage of large amounts of carbon dioxide in the subsurface. In this perspective and in order to accurately predict the short and long-term response of the reservoir, a precise characterization of the geomechanical properties has to be carried out. In addition to the classical poroelastic properties, time-dependent deformation, such as viscous creep should also be considered. Storage capacity of a caprock may be seriously affected by local creep deformation allowing fast vertical fluid flow through an a priori very impermeable formation. Within this study, we investigate the ease to creep of Opalinus clay (Jurassic shale) under shallow geological storage conditions and predict the propagation of high porosity channels at operational time scales. The effective poroviscoelastic parameters of rock are inferred from the novel laboratory experiments that allow evaluation of time-dependent deformation. The bulk viscosity of the shale is found to be ∼ 1014 - 1015 Pa·s and it decreases with rise of temperature and pore fluid pressure to total mean stress ratio. Furthermore, the propagation speed of high porosity channels (porosity waves) is calculated to be on the order of the centimeters per year.