Variations in the Pore Structure and Fractal Characteristics of No. 5 and No. 8 Coals in the Yichuan Area, Daning–Jixian Block, Ordos Basin

No. 5 and No. 8 coal seams in the Yichuan area, Ordos Basin, are primary targets for coalbed methane (CBM) development, yet their pore structure differences and controlling factors remain poorly understood. This study integrates CO2 adsorption, low-temperature N2 adsorption, and mercury intrusion porosimetry to characterize the pore structure across the full pore-size range, with fractal dimensions quantified by V-S, FHH, and J-function models. The results show that the No. 8 coal has greater CO2-accessible specific surface area and micropore volume, with micropores (<2 nm) accounting for 58–81%. In contrast, the No. 5 coal exhibits higher mesopore (2–50 nm: 1–4%) and macropore (>50 nm: 31–48%) proportions, along with higher porosity and permeability, and better pore openness and connectivity. Fractal analysis indicates that the No. 5 coal has higher V-S and FHH fractal dimensions, suggesting greater micropore–mesopore structural complexity, while the No. 8 coal exhibits a higher J-function fractal dimension, reflecting stronger pore throat heterogeneity. Correlation analysis further indicates that micropore development is closely associated with coalification degree and vitrinite content; mesopore characteristics are associated with mineral matter and show a non-monotonic relationship with ash yield; and macropore–fracture systems are related to depositional setting and burial depth. The delta-plain-sourced No. 5 coal, with shallower burial and greater lithological heterogeneity, appears to retain better-developed seepage pathways, whereas the marine–continental transitional No. 8 coal, buried deeper and more compacted, is characterized by a dominance of adsorption space with limited connectivity. Accordingly, the pore structure of the No. 5 coal appears more favorable for gas seepage, while that of the No. 8 coal appears more favorable for gas adsorption. These pore-structure-based interpretations may provide a basis for differentiated CBM development strategies in the study area, although they should be regarded as preliminary pending validation with production data.