Abstract The aims of this research are to quantitatively evaluate the complexity of the pore structure in coal frozen with liquid nitrogen (LN2) and then study the influence of the modified pore system on coalbed methane (CBM) extraction. To do this, nuclear magnetic resonance (NMR) and fractal dimension theory were used to determine the properties of the coal's pore system after samples of low rank coal had been frozen and then thawed. The fractal dimensions of pores in frozen-thawed coal samples were divided into five types according to pore size and the state of the fluid in the coal pores. The results showed that the fractal dimension DA of adsorption pores was less than two, indicating that these pores did not exhibit fractal characteristics. The fractal dimensions Dir and DT representing closed pores and total pores presented low fitting precision, so the closed pores showed insignificant fractal characteristics. However, the fractal dimensions DF and DS representing open pores and seepage pores had high fitting precision, suggesting that open and gas seepage pores exhibited a favorable fractal characteristic. Correlation analysis revealed that DF and Ds were negatively correlated with LN2 freezing time and the number of freeze-thaw cycles. After being frozen and thawed, coal porosity and permeability showed a strong negative correlation with fractal dimension and this relationship allowed predictive models for permeability and fractal dimensions (DF and DS) to be constructed. The models showed that the smaller the fractal dimension, the more uniformly the pores were distributed and the higher their degree of connection. These properties favor the production of CBM. This study also showed that compared with single LN2 freezing events, repeated cyclic freezing with LN2 followed by thawing is more favorable for CBM production.

中文翻译：

---

基于核磁共振的液氮冻融低阶煤分形维数应用于煤层气回收

摘要 本研究的目的是定量评价液氮（LN2）冻结煤孔隙结构的复杂性，进而研究改良孔隙系统对煤层气（CBM）提取的影响。为此，使用核磁共振 (NMR) 和分形维数理论来确定低阶煤样品冷冻然后解冻后煤孔隙系统的特性。冻融煤样孔隙的分形维数根据孔隙大小和煤孔隙中流体的状态分为五种类型。结果表明，吸附孔的分形维数DA小于2，说明这些孔不表现出分形特征。代表闭孔和总孔的分形维数 Dir 和 DT 拟合精度较低，闭孔分形特征不显着。然而，代表开孔和渗流孔的分形维数DF和DS拟合精度较高，表明开孔和渗气孔表现出良好的分形特征。相关分析表明，DF和Ds与LN2冻结时间和冻融循环次数呈负相关。冻融后，煤的孔隙度和渗透率与分形维数呈强烈负相关，这种关系允许构建渗透率和分形维数（DF和DS）的预测模型。模型表明，分形维数越小，孔隙分布得越均匀，它们的连接程度就越高。这些特性有利于煤层气的生产。该研究还表明，与单次 LN2 冻结事件相比，使用 LN2 重复循环冻结然后解冻更有利于煤层气生产。