Speaker: Saied Mighani; Massachusetts Institute of Technology
Enhanced reservoir connectivity generally requires maximizing the intersection between hydraulic fracture (HF) and preexisting natural fractures (NF), while having the hydraulic fracture cross the natural fractures (and not arrest). Extensive theoretical and experimental developments have been put forward to analyze the quasi-static aspects of this interaction. To understand the dynamics of this interaction, I designed and performed a suite of triaxial experiments (Pconf =5 MPa) in Solnhofen limestone. In the experiments, a fluid-driven tensile fracture initiates from a pressurized borehole and travels towards a saw-cut fault (used as a proxy for a natural fracture). Acoustic emissions (AE), fault slip, stress drop, and pore pressure were recorded at a 5 MHz sampling rate. By varying the fault surface roughness and fluid viscosity, I was able to modify the fluid diffusivity (fault transmissivity/fluid viscosity) of the fault and study its influence on this interaction. A threshold value for fluid diffusivity was observed, below which the HF was able to cross the fault. Otherwise, the fluid-infiltrated fault activated; this resulted in a finite fault-parallel slip, 0.1-4 MPa stress drop, and a burst in AE signals. Moreover, the slip induced an enhanced fluid diffusivity of ~>4 orders of magnitude on the fault. The calibrated AE sensors recorded seismic moment magnitudes of ~-5 for slip events vs. ~-7 magnitudes for HF initiation events. The experiments imply that the diffusion of fluid, driven by the fault slip, majorly controls the cross/arrest conditions. Hence, this variable has to be considered in the interaction problem. Also, the low tensile/slip events’ seismic moment ratios might explain the paucity of tensile events during microseismic recordings of reservoir stimulations.
In addition to the presented mechanical pore fluid effects, a major aspect in subsurface deformations is the chemo-mechanical processes upon the natural or anthropogenic injection of reactive fluids into the subsurface, such as CO2 in hydrothermal rock systems. Observations suggest that the resulting alterations may modify the physical properties such as porosity-permeability-velocity relationships, as well as the frame stiffness, of the matrix. Building upon these currently studied signatures, I would like to add the mechanical aspects such as the generated AEs and strength evolution upon matrix alterations. This will provide a unique venue for knowledge transfer between seismology, rock mechanics, and rock physics. Therefore, we might get closer to solve a crucial question in such geological settings, which is to determine whether, how, and when the deformation under these conditions will lead to the final rupture. Time-lapse seismic monitoring and microseismic acquisitions during fluid injection projects, as well as volcanic activities, would immediately benefit from the implications of these observations.