Experimental Reactivation of Shear‐fractured Berea and Boise Sandstones by Brine or Liquid CO2 Injection at Depth

Abstract

AbstractInjection‐driven reactivation of fractured/faulted reservoirs is a key concern in the design and safe operation of water/CO2 injection projects. Here, a new laboratory testing protocol is devised by which the reactivation conditions are quantified, providing pivotal data for effective risk assessment/management. We simulate a field injection operation in the laboratory on prefaulted Berea and Boise sandstones subjected to stress/pressure conditions prevalent at 2 km depth. The protocol consists of two key stages: (i) drained normal faulting: triaxial shear failure induced by increase of the overburden stress; and (ii) injection‐driven reactivation of the faulted rock by pore pressure increase. Injection is conducted with either brine or liquid CO2. The data show that at 2 km depth, (i) shear fracturing and subsequent faulting occurs when the differential stress reaches 36–46 MPa for Boise, or 96–98 MPa for Berea, and (ii) subsequent reactivation triggers when the pore fluid over‐pressure reaches 4–5 MPa, regardless of the injected fluid or specific sandstone tested. Spatiotemporal monitoring of the concomitant microseismic activity proves effective in time‐lapse imaging the structural changes leading to the triaxial faulting of the intact rock, or to the reactivation of the prefaulted rock. Microseismic imaging suggests that (i) initial shear faulting results in a single slip surface, consistent with brittle (Berea) or semibrittle (Boise) failure; (ii) upon injection‐induced reactivation at 2 km depth, the preexisting fault ruptures and slips first, before a new and steeper fracture nucleates, progressively propagates, then slips; and (iii) dynamic slip transfer occurs from the preexisting to the newly formed fault.