Abstract Investigation of coal microstructure based on 3D reconstruction is of great significance to the development of coalbed methane and the spatial migration of fluids in coal. In this paper, high-resolution X-ray micro-computed tomography (micro-CT) is used to perform CT scanning of the coking coal sample from Shaqu Coal Mine. Then, the quantitative identifying the pores, coal matrix and mineral in coal by using 3D visualization software AVIZO. An equivalent pore network model (PNM) for statistical pore size distribution (PSD) is established based on the “Volume Fraction module” in AVIZO. Finally, the variation of pore pressure, seepage velocity and flow path during the process of methane seepage in pore space are simulated. The results show that the distribution of pores, mineral, and coal matrix are highly heterogeneous, accounting for 8%, 3%, and 89% of the total volume, respectively. PSD statistical analysis showed that most of the pores in the sample space were distributed in the range of 500–1400 nm. With the increase of throat radius and throat length, the number of throat increased first and then decreased. Most of throat equivalent radius is less than 700 nm. The equivalent length of the throat is mainly distributed in the range of 1500–4500 nm. Under the same pressure gradient, the distribution of the pore pressure, seepage velocity and flow path in the three directions are different, which shows the high heterogeneity of the pore structure. During the methane seepage process, the pore pressure gradually decreases, and the smaller the pore radius, the more obvious the pressure change. As the pressure gradient increases, the seepage velocity of methane gradually increases, and the curve presents the obvious nonlinear relationship. In the pore structure, the sudden decrease of the pore radius in some areas causes the seepage velocity increases sharply.

中文翻译：

---

基于 CT 的 3D 重建煤微观结构的定量表征和大孔隙的可视化渗流

摘要 基于三维重建的煤微观结构研究对煤层气的开发和煤中流体的空间运移具有重要意义。本文采用高分辨率X射线显微计算机断层扫描（micro-CT）对沙曲煤矿焦煤样品进行CT扫描。然后，利用3D可视化软件AVIZO对煤中的孔隙、煤基质和矿物进行定量识别。基于AVIZO中的“体积分数模块”建立了用于统计孔径分布（PSD）的等效孔隙网络模型（PNM）。最后，模拟了甲烷在孔隙空间渗流过程中孔隙压力、渗流速度和流动路径的变化。结果表明，孔隙、矿物和煤基质的分布高度非均质，占8%，分别占总体积的 3% 和 89%。PSD 统计分析表明，样品空间中的大部分孔隙分布在 500-1400 nm 范围内。随着喉道半径和喉道长度的增加，喉道数量先增加后减少。大多数喉部等效半径小于700 nm。喉道等效长度主要分布在1500～4500 nm范围内。在相同的压力梯度下，三个方向的孔隙压力、渗流速度和流道分布不同，显示出孔隙结构的高度非均质性。甲烷渗流过程中，孔隙压力逐渐降低，孔隙半径越小，压力变化越明显。随着压力梯度的增加，甲烷的渗流速度逐渐增加，曲线呈现明显的非线性关系。在孔隙结构中，部分区域孔隙半径突然减小，导致渗流速度急剧增加。