The internal surface roughness of coals determines, to an extent, the quantity of natural gas, specifically methane, that they can readily adsorb/store and the ease with which that gas can flow and be produced from coal formations. Coals’ fractal dimensions (D) provide a quantitative measure of the degree of roughness of their internal surfaces (pores and microfractures). Low pressure nitrogen adsorption isotherms are commonly used to derive D values for coal applying the Frenkel–Halsey–Hill (FHH) method. However, the isotherm data points for many coals do not conform to the theoretically expected FHH trends. In this study, two other fractal determination methods, those of Neimark (NMK) and Wang & Li (W&L), are evaluated, together with the FHH method for 26 published coal isotherms. The NMK method is considered unsuitable as it generates unreasonably high D values. For many coals W&L and FHH methods are observed to derive consistently close D values. Detailed graphical analysis of the isotherm data points and their trends can explain the discrepancies that exist in the D values derived from the whole isotherms of some coals. Discrepancies are also observed in the fractal dimensions derived by the different methods for the low (D1) and high (D2) relative pressure portions of the coal isotherms. More realistic D1 values are derived by the W&L method than the FHH method. A novel workflow integrating detailed graphical analysis with optimized curve fitting is proposed to better quantify and verify the derivation of coal fractal dimensions from nitrogen adsorption isotherms.

煤的内部表面粗糙度在一定程度上决定了天然气的数量，特别是甲烷，它们可以很容易地吸附/储存，以及天然气可以流动和从煤层中生产的难易程度。煤的分形维数 ( D ) 提供了对其内表面（孔隙和微裂缝）粗糙度程度的定量测量。低压氮气吸附等温线通常用于推导D使用 Frenkel-Halsey-Hill (FHH) 方法计算的煤值。然而，许多煤的等温线数据点不符合理论上预期的 FHH 趋势。在这项研究中，评估了另外两种分形确定方法，即 Neimark (NMK) 和 Wang & Li (W&L)，以及 FHH 方法对 26 条已发表的煤等温线。NMK 方法被认为是不合适的，因为它会产生不合理的高D值。对于许多煤，观察到 W&L 和 FHH 方法可以得出始终接近的D值。等温线数据点及其趋势的详细图形分析可以解释D中存在的差异从某些煤的整个等温线得出的值。对于煤等温线的低 ( D1 ) 和高 ( D2 ) 相对压力部分，通过不同方法得出的分形维数也存在差异。与 FHH 方法相比，W&L 方法推导出更实际的D1值。提出了一种将详细图形分析与优化曲线拟合相结合的新工作流程，以更好地量化和验证从氮吸附等温线推导出煤分形维数。