Effect of matching relation of multi-scale, randomly distributed pores on geometric distribution of induced cracks in hydraulic fracturing

Abstract

Reservoir rock contains many multi-scale, unevenly distributed pores, and the pore structures of shale in different reservoirs and geological environments vary greatly. Because the seepage velocity and pressure field are related to the pore spatial variations, the inhomogeneity of the seepage is superimposed on the anisotropy of the rock’s physical properties, which will affect the distribution of the induced cracks. A method for calculating the pore size in the bonded particle model, based on Delaunay triangulation, is proposed. A modeling approach capable of simulating the multi-scale pore distribution of actual rock is presented based on the proposed method. To understand how microcracks connect micropores in the process of fracturing, several bonded particle model samples with different pore structures were established, and numerical experiments were conducted based on the coupling calculation of the discrete seepage algorithm and discrete element method. The focus of this study was on the interactions between the distribution characteristics of multi-scale pores, the specific physical properties of the fracturing fluid, and the distribution differences of the induced cracks caused by the special seepage characteristics when using different fracturing fluids. The numerical results showed that the advantages of supercritical CO2 fracturing are maximized in deep reservoirs (high in-situ stress) and that a suitable in-situ stress condition is required (i.e. a stress ratio close to 1).