Experimental investigation of injection pressure effects on fault reactivation for CO2 storage

Abstract

Laboratory tests were conducted in a triaxial load frame with acoustic emission and transmission capability to investigate mechanisms that might be initiating the microseismicity experienced in CO2 injection operations. Although often related to reactivation of mapped faults or local fracturing due to reduced injectivity, the case of the Illinois Basin – Decatur Project is used here to illustrate the need for better understanding of what triggers microseismic events in relatively large permeability, good reservoir candidates. There, microseismicity has occurred in the CO2 storage target formation, the Mt. Simon sandstone, as well as in the underlying Precambrian basement. The microseismicity in the Mt. Simon sandstone occurred ahead of CO2 plume arrival and at relatively low injection pressure conditions, well below the fracturing pressure at the injection well. A hypothesis is suggested for the occurrence of such events in the field, whereby critically stressed planes are activated by the passage of the pressure front at injection start; these faults are small and thus not visible in the seismic survey. In order to test this hypothesis, sandstone plugs were prepared by two different methods to incorporate a fracture plane, which we attempted to reactivate by pore pressure pulses. The reactivation was successful at low pressure for a fracture created in the laboratory at reservoir conditions but was unsuccessful except at a much higher pore pressure in a saw-cut artificial fracture. The results suggest that tortuous, rough stress-induced fractures may be easier to reactivate because of the higher probability that sections are already favorably oriented with respect to critical shear stress at a low pore pressure increase. Saw-cut fractures may close completely under isotropic stress loading and may be difficult to activate unless exactly oriented with respect to critical shear stress at a low pore pressure increase. Acoustic emission accompanying fracture reactivation was also recorded and analyzed. This revealed a different event distribution energy between creating and reactivating the fracture.