Prediction of local diffusion coefficient based on images of fractured coal cores

Abstract

Diffusion is a transport phenomenon in coal that plays a significant role in the productivity of coal seam gas (CSG) reservoirs. The diffusion coefficient quantifies the extent of gas diffusion in a coal sample. Diffusion coefficient is commonly measured as a bulk value for a core; however, it can vary spatially in a core, and its local variation is commonly ignored. We present an automated method to calculate local diffusion coefficients of krypton in a coal core using 3D microcomputed X-ray tomography (micro-CT) images. Fractures are identified from greyscale micro-CT images automatically and compared with the conventional watershed segmentation method. Local diffusion coefficients within the matrix are calculated automatically considering different fluid states. As gas within a coal core may be in the free state or adsorbed state, our method calculates the total diffusion coefficient taking into account the state of the fluid. We observe a variation of diffusion coefficient with time suggesting that as the gas accesses different regions of the pore matrix the diffusion rate changes. 3D diffusion coefficient maps are generated from the automated calculation, which shows the heterogeneity of the coal sample. The fractions of free and adsorbed gases are calculated, which allows further analysis of their individual impact on transport. Additionally, the proposed method can be applied to other fractured materials for automated diffusion coefficient measurements.