Coal demineralization evaluation using a novel 3D computational method based on pore-scale 3D morphological modeling

Coal permeability is crucial for the efficiency of coal seam gas (CSG) recovery. The mineralized natural fracture reduces the coal permeability and restricts the transportation of CSG. A novel demineralization evaluation method was developed based on the pore-scale 3D morphological modeling algorithm to improve the coal permeability, and the process of in situ demineralization in the coal was performed. The performance of the novel demineralization evaluation method is verified against the previous method on coal. Furthermore, nine coal models considering three different structures generated by the quartet structure generation set (QSGS) method were developed to assess the potential of the established algorithm. Additionally, the evolution of the microstructure, permeability, and fluid flow in coal models subjected to demineralization stimulation was investigated. The results suggest that the novel method reproduced the dynamic evolution of the connected structures and caught the promoted characteristic of pore development in the poorly-connected regions when compared with the previous method, with the connectivity and permeability increased by 18.77% and 57.14% after demineralization, respectively. The coal demineralization is most affected by porosity, sequentially followed by core distribution probability and the anisotropy degree in structure. The permeability increased with the demineralization extent, following an exponential law. In addition, demineralization effectively eliminates the tortuosity of flow pathways and increases the number of streamlines in coal models, revealing an enhancement of the transport capacity of coal. Furthermore, the relationship between the permeability and porosity in coal models has a high agreement with the existing permeability models, confirming that the generated coal models and the performance evaluation of this novel demineralization on these coal models are reliable.