Investigating poromechanical causes for hydraulic fracture complexity using a 3D coupled hydro-mechanical model

Hydraulic fracturing is widely used to increase permeability of tight deep geological formations for improving oil and gas production and enhance geothermal energy extraction. Prior studies often predicted simple planar or near planar hydraulic fractures even though these simple fractures do not adequately explain the measured data. Instead, it is likely that complex fracture networks are created. The phenomenon of hydraulic fracture branching that gives rise to complex fracture networks is poorly understood. In this study, we develop a numerical modeling tool, based on sequential coupling of solid solver HOSS and fluid solver PFLOTRAN, to investigate the mechanisms for hydraulic fracture branching. The spherocylindrical microplane constitutive model is implemented to model fracture growth in anisotropic rocks. We verify our coupled model using the KGD analytical solution. Using a set of simulations, we demonstrate that a hydraulic fracture can branch into lateral directions for certain in situ stress conditions if there are pre-existing permeable weak layers whose initial Biot effective stress coefficient is greater than that of the matrix. In addition, we investigate the effect of three-dimensional pre-existing geological discontinuities on the creation of complex fracture systems. Our results demonstrate that branched hydraulic fractures can be predicted if we account for (1) damage-dependent Biot effective stress coefficients and (2) pre-existing geologic discontinuities. This represents a 3D poromechanics mechanism for the creation of branched fracture networks where multiple fractures can propagate simultaneously in a dense parallel swarm. © 2022