Coal structures exhibit considerable differences in rock properties and adsorption capacities. The physical properties of coal rocks are fundamental to understanding oil and gas reservoirs, while adsorption capacity directly impacts the gas content in coal seams. The accurate recognition of coal structures is essential for evaluating productivity and guiding coalbed methane well development. This study examines coal rocks of Benxi Formation in Ordos Basin. Using core photographs and logging curves, we classified the coal structures into undeformed coal, cataclastic coal, and granulated-mylonitized coal. AC, DEN, CAL, GR, and CN15 logging curves were selected to build a coal structure recognition model utilizing a long short-term memory (LSTM) neural network. This approach addresses the gradient vanishing and exploding issues often encountered in traditional neural networks, enhancing the model’s capacity to handle nonlinear relationships. After numerous iterations of learning and parameter adjustments, the model achieved a recognition accuracy of over 85%, with 32 hidden units, a minimum batch size of 28, and up to 150 iterations. Validation with independent well data not involved in the model building process confirmed the model’s effectiveness, meeting the practical needs of the study area. The results suggest that the study area is predominantly characterized by undeformed coal, with cataclastic coal and granulated-mylonitized coal more developed along fault trends.

The development of tight sandstone gas reservoirs in the XUJIAHE Formation of western Sichuan faces critical challenges due to extreme reservoir conditions, including burial depths exceeding 4500 m, abnormal pore pressures (pressure gradient > 1.8 MPa/100 m), high breakdown pressures (> 90 MPa), elevated temperatures (> 120 °C), and ultralow permeability (< 0.1 mD). Conventional fracturing technologies exhibit low success rates (< 40%) and inadequate proppant placement efficiency (< 50%) under these conditions, severely limiting commercial gas production. This study presents an integrated fracturing strategy combining three novel approaches to address these operational and geological constraints. First, a high-density potassium formate brine system (1.5–1.6 g/cm³) reduced the bottomhole fracture pressure gradient to 0.0326 MPa/m through hydrostatic compensation, achieving a 21.83 MPa reduction in surface pump pressure compared to conventional fluids. Second, a hybrid fracturing design combining low-viscosity slickwater (3–5 mPa·s) and high-concentration cross-linked gel (80–120 mPa·s) simultaneously enhanced fracture complexity (Fracture Complexity Index > 2.5) and conductivity (> 100 Darcy), achieving a fivefold improvement in proppant transport capacity (600 kg/m³). Third, gelled acid pre-treatment (15% HCl system) dissolved calcareous interlayers (CaCO₃ >70%), reducing injection pressure by 14.83 MPa and increasing proppant migration efficiency by 30% in challenging Xu 4 intervals. Field implementation demonstrated breakthrough performance: fracturing success rates improved to 82%, with average post-fracture gas production reaching 2.7 × 10⁴ m³/d—a 210% increase over conventional methods. These innovations tight gas stimulation by synergistically addressing high-stress environments, fracture network complexity, and lithological heterogeneity, providing a replicable framework for deep unconventional reservoirs.

Efficient coalbed methane (CBM) recovery combined with carbon dioxide (CO₂) sequestration is a promising strategy for sustainable energy production and greenhouse gas mitigation. However, the molecular mechanisms controlling pressure-dependent CH₄ displacement by CO₂ in coal nanopores remain insufficiently understood. In this study, molecular dynamics simulations were conducted to investigate CO₂-driven CH₄ recovery in a slit-pore coal model under driving pressures of 15, 20, and 25 Mpa. The simulations quantitatively captured the competitive adsorption, diffusion, and migration behaviors of CH₄, CO₂, and water, providing insights into how pressure influences enhanced coalbed methane (ECBM) recovery at the nanoscale. The results show that as the pressure increases from 15 to 25 Mpa, the mean residence time of CH₄ on the coal surface decreases from 0.0104 ns to 0.0087 ns (a 16% reduction), reflecting accelerated molecular mobility. The CH₄–CO₂ radial distribution function peak height rises from 2.20 to 3.67, indicating strengthened competitive adsorption and interaction between the two gases. Correspondingly, the number of CO₂ molecules entering the CH₄ region grows from 214 to 268, demonstrating higher invasion efficiency at elevated pressures. These quantitative findings illustrate a clear shift from capillary-controlled desorption at low pressure to pressure-driven convection at higher pressures. The results provide molecular-level evidence for optimizing CO₂ injection pressure to improve CBM recovery efficiency and CO₂ storage capacity.

Coal exhibits heterogeneous pore networks and chemically diverse surfaces, resulting in complex competitive adsorption among CH₄, CO₂, and H₂O. The underlying molecular mechanisms remain unclear. In this work, molecular simulation methods were applied to investigate the adsorption behavior and interaction characteristics of CH₄/CO₂/H₂O mixtures on two typical coal components (inertinite and vitrinite) under different CO₂ enrichment levels, corresponding to gas-phase CO₂ mole fractions of 4.8%, 9.1%, and 16.7%. The results demonstrate that CH₄ dominates surface occupation in all cases, maintaining 30–70 adsorbed molecules, whereas CO₂ adsorption is significantly weaker, remaining below 5 at low loading and increasing to only 10–17 at high loading. This indicates a limited competitive capability of CO₂ for adsorption sites. From an interaction perspective, water governs the electrostatic environment, with surface–water Coulombic energies consistently distributed around − 600 to − 750 kJ mol⁻¹. In contrast, CO₂–water interactions decrease from − 220 to − 360 kJ mol⁻¹ to − 100 to − 170 kJ mol⁻¹ as CO₂ loading increases, reflecting a pronounced screening effect. Meanwhile, direct CO₂–surface interactions remain weak (typically − 5 to − 15 kJ mol⁻¹). Overall, CH₄ adsorption is primarily controlled by dispersion interactions, while CO₂ is constrained by weak surface affinity and reduced hydration strength, resulting in a secondary role in multicomponent competitive adsorption within coal systems.