Mechanisms of enhanced hydraulic fracture complexity through carbon dioxide pre-injection induced permeability evolution

Abstract

Sequential injection of fluids with contrasting mobility in porous media causes the leading fluid's pressure history to reshape the permeability field, which then governs the seepage and fracture propagation of the following fluid. Using a cross-scale model that couples seepage and discrete element fracturing and distinguishes the flow behaviors of CO2 and water, this study investigates CO2 pre-injection enhanced hydraulic fracturing in coal, capturing permeability evolution driven by CO2 pressure diffusion and subsequent hydraulic fracture propagation. The CO2 pressurization rate controls the time for pressure diffusion, thereby determining permeability uniformity; an intermediate rate creates a moderately heterogeneous field, maximizing fracture length, area, and fractal dimension, while lower and higher rates, respectively, over-homogenize and over-localize the field, reducing complexity. CO2 pre-injection pressure governs the magnitude and heterogeneity of permeability enhancement, transitioning fracture propagation from length-dominated to complexity-dominated. Constant pressure duration regulates the spatial extent of CO2 diffusion, longer durations homogenizing the field and favoring primary fracture extension over complexity. Water injection pressure governs the driving pressure gradient and the shift from branching networks to localized channel flow. Below this threshold, the gradient cannot fully penetrate the CO2-treated zone, limiting fracture complexity; above it, faster flow suppresses branching, widens fractures, and reduces the fractal dimension. Orthogonal analysis reveals a hierarchical control: water pressure has the greatest influence, followed by CO2 pressure, water pressurization rate, constant pressure duration, and CO2 pressurization rate. These findings link pore pressure diffusion and permeability evolution to fracture complexity, providing a theoretical basis for the optimization of process parameters.