During the hydraulic fracturing process, the created rough fracture surface and fracturing fluids with high viscosity greatly challenge proppants placement in the thin aperture of fractures. Thus, it is essential to detailly investigate the effect of surface roughness on the proppant distribution. In addition, the multiphase flow in the rough nanoscale microfractures in the variety of orientations have not been cleared. Taking all of these into consideration; rock grain geometries, packing mechanisms, the presence of clay content, and in-situ stress field will be affected and will affect the presence of the microcracks, and consequently control the permeability and porosity of the sedimentary rock. In the failed rock after fracturing work, a processed zone where the pre-existing natural fractures get activated, and induced microcracks including intergranular and intragranular grain boundaries are brought to connect to the main fracture. Hence, the rock grain and pore size distributions at fracture processed zone are altered. This, in turn, controls the fluid transport in the rocks.
Our novel approach incorporates the image analysis software (ImageJ) by organizing desired image processing codes to study the critical features of the post-fracturing core sample, including main fracture roughness, mechanical rock properties, crack density, grain, and pore size distributions. Tennessee sandstone was undergone the hydraulic fracturing test and polished on a cross-section perpendicular to the main fracture. This cross-section was recorded by the high-resolution SEM images after ion-milling. Corresponding grain size and pore size distributions are studied at each representative location with respect to its distance to the main fracture to probe alterations of the fracturing process from the core sample original state. The results of grain size and pore size distributions are compared. The discussions of their alterations mechanisms and their effects on the rock porosity and permeability are analyzed.
We find that the roughness presence of fractures strongly increases conduits open to fluid flow. In addition, our developed image processing code perfectly captured the rock grains with the promising precision. Further, we are able to observe the grain size deduction due to the incremental intragranular grain boundaries while intergranular grain boundaries are still majorities outside the fracture processed zone (FPZ). Grain size renders a lognormal distribution at each representative location and coincides with the permeability distribution of most reservoir rocks. Grain size averages also match the literature values with reasonable uncertainties (20%). The pore size distribution and its average value vary spatially. Results from this study kindle the insights of the heterogeneity of the fractured formation with proper petrophysics parameters quantitatively. We also found that the aspect ratio from 2D image analysis does not reflect the significance in the mechanics.
This novel approach will commit to supporting the lab measurements, gives field preliminary hydraulic fracturing performance assessment and lower the cost needed for hydraulic fracturing design.