DOI: 10.1002/gj.70486 ISSN: 0072-1050

Model‐Based Characterisation of High‐Pressure Transcritical CH 4 / CO

Zhang Kun, Wang Zhi, Liu Qimeng, Chang Yanhai, Ma Mengya, Liu Hewu, Liu Huihu, Xu Hongjie

ABSTRACT

In deep coal seams, the high‐pressure adsorption of CH 4 and CO 2 governs both methane recovery and CO 2 storage capacity. To clarify how pore structure controls transcritical adsorption–desorption and model applicability, two coal samples of similar rank from the Huainan–Huaibei coalfields—Liuzhuang (LZ) and Qidong (QD)—were characterised by low‐temperature CO 2 and N 2 adsorption and mercury intrusion porosimetry; gravimetric high‐pressure isotherms were corrected to absolute adsorption by the intercept method and fitted with the Langmuir, BET and S–DR models. LZ coal is enriched in micropores, whereas QD coal exhibits a more developed fracture network. For both gases, the excess adsorption rises, peaks and then declines with pressure, a decline reflecting increasing free‐phase density rather than a real loss of capacity. After correction, absolute adsorption keeps rising and levels off at high pressure, with QD coal showing higher absolute adsorption and larger capacity parameters than LZ coal. Model performance is stage‐dependent. All three models fit the low‐pressure stage well, with Langmuir marginally the best (mean R 2  = 0.9945); S–DR attains the highest mean R 2 in the transition stage (0.7526 vs. 0.5239 for Langmuir); and on the high‐pressure plateau, the three are comparable in mean R 2 but S–DR is the most stable across samples and temperatures. Langmuir therefore remains the more convenient full‐range empirical estimator of capacity, whereas S–DR better resolves stage‐specific CO 2 behaviour at mid‐to‐high pressure; BET is least stable there. Despite its lower micropore volume, QD coal adsorbs more CO 2 , which may be associated with its more developed fractures and improved pore‐throat connectivity, reflecting pore‐size‐selective occurrence. Evaluating CO 2 storage potential therefore requires absolute‐adsorption correction together with stage‐specific model behaviour and pore selectivity, rather than total pore volume, excess adsorption or a single model parameter in isolation.