Multi-scale modelling of multi-physics flow in coal seams

Abstract

The understanding and quantitative analysis of multi-physics flow in coal seam gas reservoirs are of importance for estimating potential gas production or assessing the feasibility of carbon dioxide sequestration and hydrogen storage projects. Thus, such phenomena as two-phase flow, gas and rock compressibility, as well as sorption and diffusion and their relative contributions to flow characteristics, should be quantitatively analysed. At the pore-scale, direct numerical simulations and pore-network modelling are used for micro-CT based flow simulations. However, there is a lack of modelling tools available for modelling fluid flow on micro-CT images of coal. Therefore, we develop an appropriate hybrid numerical solution for image-based flow simulations in coal, where the two-phase Darcy–Brinkman flow model is employed as a basis. This model tends towards the Navier–Stokes solution in the resolved regions of the image and towards the continuum Darcy flow in the sub-resolved zones. The gas phase is considered compressible, while the other phase represents water. The sorption processes are introduced by the Langmuir isotherm, while the transient nature of sorption is captured by the implicit combination of the Langmuir and pressure equations in the Darcy–Brinkman model. The surface diffusion is explicitly embedded in the proposed hybrid model by using the second Fick’s law. The impact of matrix swelling-shrinkage phenomenon on the coal porosity and absolute permeability is accounted for by using the Palmer–Masoori analytical model. The proposed hybrid numerical solution is implemented by using the OpenFoam open-source software. Moreover, several illustrative synthetic and realistic simulations are presented to demonstrate the capability of the developed solver to capture complex coal multi-physics. Furthermore, the employment of numerical irregular meshes for saving computational resources is also exemplified.