Prediction Methods for Gas-bearing Properties in Deep Coal Measures in the Ordos Basin

Abstract Deep coalbed methane(CBM)in the Ordos Basin, which provides a strategic replacement for the resources from the Changqing Oilfield, China, challenges the conventional understanding of petroleum geology and exhibits enormous resource potential. Coal-bearing strata are widely developed across the basin, and coal seam No. 8 is the primary exploration target. This seam has a thickness of 6–16 m (average: 7.8 m), and its seismic reservoir properties are critical for improving exploration and development efficiency. However, the quantitative prediction of gas-bearing properties has long been hampered by three critical limitations: (i) the direct application of conventional amplitude-versus-offset (AVO) methods developed for sandstone or carbonate reservoirs neglects the unique dual pore–fracture system of coal and the effects of non-coal intercalations, leading to severe overlap between fluid and lithological responses and considerable non-uniqueness; (ii) common prediction workflows lack systematic geological constraints, and petrophysical inversion and attribute analysis remain largely qualitative; and (iii) single-attribute approaches often fail in complex coal structures because of the inherent trade-off between gas sensitivity and robust lithological discrimination. To address these challenges, this study establishes a progressively constrained prediction framework guided by geology, bridged by petrophysics, and centered upon frequency-dependent seismic inversion. We propose a novel evaluation parameter which is characterized by both high gas sensitivity and robust lithological discrimination; this parameter integrates geological fairway delineation, Bayesian AVO P-wave dispersion inference, and fusion of the product of the Lamé constant and density with dispersion attributes. This parameter effectively decouples the response interference in AVO inversion caused by thin seam and lithological variations, fundamentally enhancing gas-bearing property assessments from qualitative detection to quantitative characterization. Practical application of this parameter within the Ordos Basin demonstrates a significant enhancement in prediction accuracy. It thus provides a transferable technical paradigm for coal-measure gas exploration in similar basins globally.