Image-based fracture pipe network modelling for prediction of coal permeability

Abstract

In coal seam gas (CSG) reservoirs, the underlying fracture networks play a critical role in controlling flow pathways of methane and water phases. Thus, study of flow in fracture networks via different numerical modelling methods is of importance. Compared to direct flow simulation, the Fracture Pipe Network Model (FPNM) is an effective means of modelling fluid flow in a fracture network due to its computational efficiency. However, constructing FPNM for authentic samples is challenging due to the limited fracture data at the core scale. In addition, application of FPNM for coal is challenging due to its complex internal fracture network. With X-ray micro-computed tomography (micro-CT) imaging, internal fracture network of coal samples can be visualised non-destructively, providing deterministic input data for FPNM. However, analysing fractures individually and characterising network connectivity based on micro-CT images are difficult, due to complex interactions, irregular fracture geometry, resolution limitation and image noises. In this paper, we develop a framework for constructing image-based FPNM for real fractured coal cores for fast prediction of permeability. Compared with conventional FPNM, each pipe element of our improved FPNM has the identical properties (orientations, length, aperture and roughness) with its corresponding data from the real sample. By comparing permeability values obtained from FPNMs, micro-CT images and voxelised DFNs, we conclude that FPNM can effectively estimate the permeability of original fracture networks while requiring significantly lower computational cost that make it a suitable framework for expensive multi-phase multiphysics flow simulations.