LAUSR

To obtain the real physical model characterizing the pore structure of coal rocks and further explore the rules governing the flow of coalbed methane (CBM) using numerical simulation, we proposed a sequential method from real coal model to computer-aided designed (CAD) coal model to finite element coal model. First, six different coal samples were scanned using $$\upmu $$μCT225kVFCB CT system. The obtained CT data were subject to threshold segmentation using the compensated digital terrain model and a coal model in STL format was established. Then, we elaborated a reverse engineering method to convert the coal model into a CAD coal model. Lastly, we obtained the finite element coal model after setting the boundary conditions with CFX software, numerically simulated CBM flow at 30 different pressure gradients based on the finite element coal model and analyzed the impacts of effective porosity on permeability and seepage velocity of CBM. The results showed that (1) at the microscale ($${<}100\,\upmu \hbox {m}$$<100μm), the seepage velocity and pressure gradient were in agreement with the Forchheimer law of high-speed nonlinear seepage; (2) permeability of CBM shows a fluctuating, but not absolute increase with effective porosity increasing; (3) at the same pressure gradient, the overall seepage velocity of CBM increases with the effective porosity increasing, but declines with non-Darcy flow coefficient increasing; (4) the non-Darcy flow coefficient reduces with both effective porosity and permeability increasing. Fitting to power function analysis shows that compared with permeability, the effective porosity has more significant impact on the non-Darcy flow coefficient.