CO2 leakage potential as a result of induced seismicity.

The feasibility of permanent geological storage of CO2 has been questioned due to the potential reactivation of faults, induced seismicity, and significant increase in fault permeability leading to the CO2 leakage (Zoback and Gorelick, 2012). At the same time, Vilarrasa and Carrera (2015) argued that the characteristics of CO2 injection make CO2 storage favorable for mitigating induced seismicity, in contrast to wastewater injection, which has induced numerous earthquakes. Mitigating induced earthquakes is crucial for the success of geologic carbon storage not only because of the concerns for potential CO2 leakage, but also because of the negative effect that this seismicity has on public perception. Here, we further investigate the potential of CO2 leakage by experimentally quantifying the permeability of intact and sheared clay-rich rock representative of caprock formations and by numerically assessing fault stability as a result of CO2 injection. On the one hand, we measure in the laboratory the hydromechanical properties of Berea sandstone with the permeability ~ 10-14 m2, as representative of the storage formation, and Opalinus clay (shaly facies), as representative of the caprock with the permeability ~ 10-21 m2. The properties are measured under in-situ conditions both for intact and restructured rock, where the latter one is assumed to represent the sheared or damaged material in the faults. Experimental measurements show that the reservoir rock is significantly stiffer than the caprock and the damaged material is softer and more permeable than the intact rock. For the assessment of CO2 leakage, the changes in permeability and entry pressure of the caprock are of particular importance. We find that for the damaged material the permeability increases by less than two orders of magnitude, still remaining below 10-18 m2, and that entry pressure is reduced by a factor of roughly two, but stays above several MPa (Vilarrasa and Makhnenko, 2017). Subsequently, we use the material properties measured in the laboratory and model a scenario in which the storage formation is bounded by one fault with an offset equal to half of the thickness of the storage formation. We consider that the storage formation is isolated from the crystalline basement by a low-permeable, clay-rich formation with similar properties to those of the caprock. The fault is characterized by a low-permeable core and a damage zone on both sides of the core. The properties of the damage zone vary along the fault and adopt the properties of the damaged material the fault is in contact with, i.e., fractured reservoir rock or damaged caprock. The low permeability of the fault induces pressure buildup on the side of the fault where CO2 is injected, reducing effective stresses and thus, affecting fault stability (Vilarrasa et al., 2016). In our simulation, CO2 plume remains far away from the fault, so the CO2 by itself would not leak despite of significantly enhancement of fault permeability and reduction in entry pressure. Nevertheless, considering gigatonne-scale storage projects, CO2 could potentially reach some large faults. In this case, given the low permeability of clay-rich geomaterials (both intact and sheared material), pressure diffusion along the fault is very limited. As a result, stress changes on the fault are confined to the region immediately surrounding the storage formation. Thus, the area that may become unstable is limited, so potential induced seismicity would be of magnitudes that could hardly be felt on surface. Furthermore, permeability enhancement as a result of shear slip would not occur across the whole caprock and CO2 flow along the fault will become negligible. Most importantly, since the entry pressure of sheared clay-rich material is of several MPa, CO2 could barely migrate upwards.

Peer reviewed