An improved digital-rock random construction method and flow simulation considering pore and fracture coupling of shale gas reservoirs

Abstract

Random construction of digital rocks based on scanning electron microscope (SEM) images is an effective method to investigate shale gas flow. Limited researches have been conducted to take into account the practical shape of organic matter or the coupling of fractures and pores during the reconstruction of digital rock samples, which would lead to the flow simulation results deviating from the real situations. In this work, the traditional segmentation method of SEM image using threshold algorithm was modified to calculate the morphological parameters of multi-components shale structures. Then, the classical four-parameter random growth algorithm was improved to construct the shale digital rock. Finally, we conducted flow simulation on the digital rocks through pore network model (PNM). Some complex structures (e.g., rough fracture and pyrite framboid) in shale can still be properly generated by controlling the growth parameters in the algorithm. Compared to the previous reconstruction methods, we incorporated the diagenetic sequence of shale components and more realistically described the morphology of organic matter and inorganic matter. After considering the complex structures, the structure characterization parameters such as pore size distribution and porosity distribution among 2D slices are consistent with the real sample, thus improved the representativeness of reconstructed rock samples. The coupling of fracture significantly improved the connectivity of PNM, resulting in an increase in absolute permeability from 3.492 × 10−9μm2 to 7.019 × 10−5μm2 and lower left movement of the capillary pressure curve. For the gas-water relative permeability curve, the addition of fracture extends the two-phase flow region, and decreases the bound water saturation from 0.63 to 0.19 during the drainage process, which facilitates the flow of the gas-water two phases. This work highlights the importance of considering organic matter morphology and fractures in shale digital rock, which provide fundamental insights into reservoir characterization and simulation.