In Situ Gas Content Prediction of Deep Coal Seams Based on Logging in the Central Linxing–Shenfu Block Authors: Yuyan Che, Guangui Zou, Suping Peng, Jiasheng She, Xiaodong Wang, Hu Zeng, Ke Ren Published: 2026 ABSTRACT Deep coalbed methane, a key strategic resource for unconventional natural gas supply, is abundant in the Ordos Basin, China. However, limited systematic research on the state of gas occurrence and critical depth, combined with sparse gas content and in situ stress measurements, has severely restricted the optimization of developmental strategies. Focusing on the 8+9# coal seam of the Taiyuan Formation in the central Linxing–Shenfu Block, Ordos Basin, an improved gas content prediction model is established, which comprehensively accounts for matrix compression, thermal expansion, adsorption-induced swelling, and the pore volume occupied by adsorbed gas. The adsorbed gas content trend was exhibited by logging data. Continuous in situ stress profiles were predicted using logging data combined with the Eaton method and Huang Rongzun’s in situ stress model. Critical depth was determined using the gas content evolution and in situ stress regime transition. The minimum horizontal stress of the 8+9# coal seam exhibited a “higher in the south and lower in the north” distribution pattern. The gas content critical depth is 1076–1704 m, the in situ stress critical depth is 1138–1727 m, and the integrated critical depth is 1076–1727 m, showing a “deeper in the south and shallower in the north” distribution. Based on the three-stage variation pattern of gas content and stress regime, a layered development strategy was proposed. This study supports location optimization and drilling and fracturing parameter design, contributing to safe, efficient, and economical deep coalbed methane development under similar geological conditions. Introduction There have been four stages to the research and industrial development of deep coalbed methane (CBM) in China, namely, the initial exploration stage (1994-2005), the slow development stage (2006-2015), the steady progress stage (2016-2020), and the rapid development stage (since 2021). , Driven by advancements in CBM exploration and development technologies, the sector has progressively evolved in two directions: the expansion from high-rank coal to medium-low rank coal, and the gradual advancement from shallow to medium and deep burial depths. − In 2023, the PetroChina Research Institute of Petroleum Exploration & Development estimated that deep CBM resources (with burial depths exceeding 1500 m) across the Ordos Basin, Junggar Basin, Sichuan Basin, and Turpan-Hami Basin total 50.7 × 1012 m3, among which the Ordos Basin accounts for 18.3 × 1012 m3. In 2022, the horizontal well cluster in the Linxing Block yielded 5.7 × 104 m3/d; meanwhile, in 2023, the well cluster in the Shenfu Block sustained production of 0.42 × 104 m3/d. Together, these results suggest that the deep coal seam in the eastern Ordos Basin possesses several advantageous attributes, including substantial resource reserves, high gas content, the presence of free gas, and elevated initial production rates. − However, the absence of a standardized classification for deep coal seams has impeded the development of relevant theoretical and technological research. The definition of critical depth has been extensively discussed, with three primary approaches identified in the literature. , One type defines the depth boundary based on the reservoir permeability, specifically, the depth corresponding to a permeability of 0.1 × 10–3 μm2, with the criterion being whether large-scale stimulation measures are required to achieve economic production. The second approach defines critical depth using a saturated gas content model that incorporates three essential factors: coal rank, temperature, and stress. This involves constructing a critical depth conversion index that reflects the interrelationships among maximum vitrinite reflectance (R o,max), pressure, temperature, and burial depth. The third method defines critical depth using the lateral pressure coefficient (G S = σ H /σ v ), where σ H represents the maximum horizontal stress and σ v denotes the vertical principal stress. Deep coal seams exhibit reduced porosity and permeability, a higher proportion of free-phase gas, and significantly greater challenges for hydraulic fracturing stimulation. These attributes distinguish deep coal seams markedly from shallow CBM reservoirs. Thus, critical depth is closely associated with permeability, gas saturation, and in situ stress distribution. Clarifying the definition of critical depth is important for advancing theoretical research on deep CBM exploration and guiding engineering deployment. Xi et al. (2024) calculated a critical depth of 1200 m for the Ningwu Basin by integrating drilling, logging, and well testing data with isothermal adsorption results and gas content distribution trends. In the Enhong Area of eastern Yunnan, Ju et al. (2020) determined the critical depth to be approximately 700 m based on the in situ stress distribution. Lv et al. (2025) estimated a critical depth of 800 m using in situ stress and gas content models, whereas Shen et al. (2014) compared three distinct methods and concluded that the critical depth of the Qinshui Basin is 750 m. Zheng et al. (2019) employed a back propagation neural network model to determine the critical depth under the parameter system of permeability, porosity, gas content, in situ stress, and temperature, ultimately defining coal seams with a burial depth exceeding 1000 m in the Qinshui Basin as deep coal seams. Limited research has been conducted on the critical burial depths of deep coalbed methane reservoirs within the Linxing–Shenfu Block. Mi et al. (2024) analyzed the mechanical properties and stress distribution characteristics of coal seams and their roofs and floors in the Shenfu block using logging data. Additionally, Ju et al. (2017) discussed the principal stress distribution of the 4+5# coal seams in the Shanxi Formation but did not determine the critical depth or address the 8+9# coal seams in the Taiyuan Formation. Nevertheless, while the 8+9# coal seams represent the primary exploration target of the Linxing–Shenfu block, systematic research on these seams remains limited. This study shows that the free gas proportion in deep coal seams exceeds that in the shallow coal seams. Predicting pore pressure in CBM reservoirs is crucial for identifying the gas distribution of the CBM. However, one prominent challenge stems from the sparse distribution of measured in situ stress and gas content data sets, making it difficult to reliably characterize regional in situ stress distribution trends and gas accumulation regularities. Conventional methods for calculating gas content typically use a constant porosity measured in the laboratory, neglecting the effects of in situ stress and temperature, which leads to systematic biases in the evaluation of deep reservoirs. Although recent studies have introduced a framework for porosity evolution, they still fail to account for the pore space occupied by adsorbed gas films, resulting in a persistent overestimation of the free gas content. To address these shortcomings, this study established an improved gas content prediction model that couples matrix compression, thermal expansion, and adsorption-swelling effects and further introduced a correction term for the pore volume occupied by adsorbed gas to reduce calculation errors in free gas estimates. Combined with well-logging-based in situ stress prediction methods, this study quantified the vertical variation laws of gas content and in situ stress, elucidated their distribution characteristics, and defined the comprehensive critical depth for the study area, thereby providing theoretical support for the optimization of deep CBM development plans. Regional Geological Background The Ordos basin is in the western part of the North China Craton. Its southern, northern, eastern, and western boundaries correspond to the Qinling Orogenic Belt, Yanshan-Yinshan Orogenic Belt, Helan Mountain Range, and Lvliang Mountain Range, respectively. The basin exhibits an overall north-southward upward tilt: its western flank dips steeply, and its eastern flank gently transitions toward the basin margin. During the Late Indosinian, Middle Yanshanian, and Late Himalayan periods, the Ordos Basin underwent distinct tectonic processes, including thrust nappe-front subsidence in the southwestern part, differential uplift-coupled strike-slip thrusting in the northeastern part, and differential tilting of the eastern margin slope. The central region of the Linxing–Shenfu Block is situated within the eastern Ordos Basin, with a regional structural dip trending southwestward as a whole. Within this block, the coal-bearing strata comprise the Shanxi Formation (hosting the 4+5# coal seams) and the Carboniferous-Permian Taiyuan Formation (hosting the 8+9# coal seams). Based on well logging data, the burial depth of the 8+9# coal seams ranges from 1668 to 2530 m. In the research region, coal thickness is greater in the northern part (> 15 m) and thinner in the southern part (approximately 5–10 m), and functions as the main target layer for current deep CBM exploration (Fig. ). (a) Simplified geological map showing the major tectonic units in the North China Craton (modified from) and (b) coal thickness of the Taiyuan Formation in the central parts of Linxing–Shenfu Block. Methodology Total Gas Content Calculation Model Coalbed methane occurs in the pores of coal seams, making porosity data crucial for the evaluation of the resource potential. Under different burial depths and geological conditions, porosity is jointly controlled by formation pressure and temperature, showing distinct variation trends. Therefore, clarifying the variation law of porosity under different burial depths helps improve the prediction accuracy of the deep coalbed methane content. While the gas content of deep CBM reservoirs is governed by porosity, reservoir temperature, and pressure conditions at different burial depths also dominate the coal adsorption characteristics, thereby further influencing the occurrence features of coalbed methane. On the basis of previous studies, Li et al. (2024) comprehensively considered the matrix compression effect, thermal expansion effect, and coal adsorption-swelling effect, established a porosity evolution model with burial depth, and further calculated the coal seam gas content. Nevertheless, the model fails to consider the pore volume occupied by adsorbed gas films, which induces an overprediction of free gas and further biases the evaluation of the total gas content. Therefore, this paper comprehensively analyzes the combined influences of formation factors on porosity. In addition, the impact of adsorbed gas-occupied pore volume on free gas content is fully considered in the total gas content calculation, thereby establishing an innovative prediction approach for the coal seam gas content. Compared with existing research, the core innovations of this model are reflected in two aspects: (1) by integrating the effects of matrix compression, thermal expansion, and coal adsorption-expansion, the model reconstructs in situ porosity under deep geological conditions, thereby eliminating errors caused by the constant porosity assumed in laboratory settings; (2) by quantifying the pore volume occupied by adsorbed gas and correcting the effective pore space for free gas, the model significantly improves the calculation accuracy of free gas content in deep coal seams. Coal Seam Porosity Prediction Model Coal seam porosity, which exerts a direct influence on reservoir gas content, must be accurately determined under in situ formation conditions. In the formation environment, coal seam porosity is governed by the combined effects of coupling geological factors, such as formation pressure, formation temperature, pore pressure, and the adsorption-swelling effect of coal on methane (Fig. ). Ignoring formation factors will cause the calculated porosity to deviate from the actual situation, thereby affecting the estimation of the gas content in coalbed methane reservoirs. Schematic illustration of deformation of the coal matrix and pore volume under in situ stratigraphic conditions: (a) Initial state of the solid coal matrix and pore space under surface laboratory conditions, and (b) deformation state resulting from the coupled effects of confining pressure compression, thermal expansion, and methane adsorption swelling. Porosity can be derived as follows 1 φ=Vp0+ΔVpVb0+ΔVb=(Vb0−Vs0)+(ΔVb−ΔVs)Vb0+ΔVb=1−Vs0(1+ΔVsVs0)Vb0(1+ΔVbVb0)=1−(1−φ0)1+εV(1+ΔVsVs0) where φ is porosity, %; V p0 is the initial pore volume, m3; ΔV p is the variation of pore volume, m3; V b0 is the initial total volume of coal, m3; ΔV b is the variation of total coal volume, m3; V s0 is the initial volume of coal solid matrix, m3; ΔV s is the volume variation of coal solid matrix, m3; φ0 is the initial porosity, %; and ε V is coal matrix volume variation. It is assumed that elastic deformation occurs in coal solid particles under stress. The volume change of solid particles caused by the variation of pore fluid pressure ΔV s1 is expressed as follows 2 ΔVs1=−KYΔP·Vs0 where K Y is the volume compression coefficient of coal skeleton, MPa–1; ΔP = P – P 0 represents the variation of fluid pressure, MPa; P is the in situ pore pressure, MPa; and P 0 is the fluid pressure during laboratory porosity testing, MPa. The volume change of coal solid particles caused by thermoelastic expansion, ΔV S2, is expressed as follows 3 ΔVS2=βSΔT·VS0 where β s is the volumetric thermal expansion coefficient of coal, generally β s = 2.4 × 10–5 m 3/m 3·K; ΔT = T – T 0 is the variation of absolute temperature, K; T is the in situ temperature, K; and T 0 is the temperature during laboratory porosity measurement, K. The volume variation of coal solid particles induced by gas adsorption swelling, ΔV S3, can be determined as follows 4 ΔVS3=εp1−φ0·VS0 where ε p is the adsorption-swelling strain per unit volume of coal, which can be determined by eq () 5 εp=CswVads C sw denotes the expansion constant of methane, t/m3. Based on previous experimental data, C sw is taken as 0.0007 t/m3; and V ads represents the adsorbed gas content of coal seams, m3/t. Combining the above three factors, the porosity calculation formula can be further expressed as 6 φ=1−(1−φ0)1+εV(1+ΔVSVS0)=1−(1−φ0)1+εV(1−KYΔP+βSΔT+CSWVads1−φ0) For the total volumetric strain of coal induced by confining pressure, it can be derived from the definition of the volume compression coefficient 7 1+εV=1+ΔVbVb0=1+Vcoal−Vb0Vb0=VcoalVb0 where V coal = V b0exp (−K Y Δσ′) is the total volume of coal after the change in confining pressure; then, 8 1+εV=exp⁡(−KYΔσ′) Δσ′ is the variation of effective stress acting on coal, MPa. According to the definitions of the thermal expansion coefficient, volume compression coefficient, and elastic mechanics, the relationship between effective stress variation and volumetric strain of methane-bearing coal per unit volume can be expressed by thermodynamic parameters 9 Δσ′=−1KY(2−4−2βSΔT) The porosity can be further expressed as 10 φ=1−(1−φ0)exp⁡(2−4−2βSΔT)(1−KYΔP+βSΔT+CSWVads1−φ0) The fluids in the pore space of coal seams include water and coalbed methane. Coalbed methane consists of adsorbed gas, free gas, and dissolved gas. Among them, dissolved gas exists in a dissolved state in formation water; adsorbed gas adheres to coal matrix surfaces; and free gas is distributed in the residual pore space in a free state. Under reservoir conditions, adsorbed methane occupies part of the pore space, and the remaining pore space serves as the occurrence site for water and free gas (Fig. ). For methane in coal, the majority exists in an adsorbed state. Therefore, when calculating the space occupied by free gas, it is necessary to consider not only the volume occupied by water but also the pore space occupied by the adsorbed layer. Failure to consider this factor biases free gas content calculations, which further degrades the quantitative accuracy of the total gas content assessment. Since coalbed methane in coal seams is generated during sedimentary and geological evolution, a large amount of adsorbed gas occurs within the nanoscale pores of the coal matrix. These pores are not connected to the external fracture network, which is consistent with the pore connectivity characteristics of deep coal reservoirs in the Ordos Basin. Consequently, the pore volume of this isolated pore fraction cannot be measured in laboratory porosity tests by using bulk coal samples. Diagram of pore fluids distribution in coal. The porosity occupied by adsorbed gas per unit volume of coal is defined as the adsorbed gas porosity φ ads , which can be calculated by eq () 11 φads=Vads_rVb=Vads·ρg·ρb10·ρads where V ads_r is the volume of adsorbed gas under in situ conditions, m3; V b is the volume of coal, m3; ρ g represents the standard-state density of CBM, t/m3; ρ b is the bulk density of coal, t/m3; and ρ ads is the density of adsorbed gas, t/m3. Referring to the research method proposed by Shi et al. (2024), it is assumed that 90% of the adsorbed gas occurs in isolated dead pores, while the remaining 10% occupies the pore volume measured by laboratory tests. This hypothesis is consistent with the characteristics of coalbed methane occurrence: the vast majority of adsorbed gas is stored in the form of microporous filling within the nanoscale micropores of the coal matrix; conventional laboratory porosity tests can only detect pores connected to the fracture network and cannot identify isolated micropores within the coal matrix. The Linxing–Shenfu block is located on the eastern margin of the Ordos Basin and shares the same low- to medium-rank coal attributes and pore structure characteristics as the study area described by Shi et al. (2024), with micropores accounting for more than 70% of the total pore volume; therefore, this assumption is geologically applicable. To assess the uncertainty of this parameter, a sensitivity analysis was conducted by varying the proportion of isolated pores containing adsorbed gas within the range 80% to 99%. The results are shown in Table : within this reasonable geological range, the relative error in the predicted total gas volume is less than 3.8%. Changes in this proportion do not alter the three-stage evolution trend of the gas volume or the determination of the critical depth, indicating that the model exhibits good stability. 1Predicted Total Gas Volume for Well LS-7 Based on Different Coefficients α of Adsorbed Gas Occupying Measurable Effective Pore VolumeαTotal Gas VolumeError0.0122.26-0.0522.090.79%0.1021.871.78%0.1521.652.77%0.2021.423.77% Predicted Total Gas Volume for Well LS-7 Based on Different Coefficients α of Adsorbed Gas Occupying Measurable Effective Pore Volume On the basis of a unit volume of coal: The total pore volume fraction (total porosity) φ is the ratio of the total pore volume to the coal volume, the adsorbed gas volume fraction φads (within interconnected pores) is the ratio of the adsorbed gas volume to the coal volume, and the proportion of formation water volume to the total pore volume is S w . Therefore, the ratio of the water volume to the coal volume is φ·S w . The effective porosity available for free gas is the portion of total porosity remaining after subtracting the porosity occupied by adsorbed gas and water 12 φfree=φ(1−Sw)−φads where S w is the water saturation, %. This derivation is fully consistent with the logic behind the calculation of free gas content in eq (); all volume parameters are based on the same unit of coal mass. Total Gas Content Calculation Model The total gas content in coal can be expressed as the sum of adsorbed gas, free gas, and dissolved gas, which can be mathematically expressed as follows 13 Vtotal=Vads+Vfree+Vdis where V total is the total gas content of coal seams, m3/t; V free is the free gas content, m3/t; and V dis is the dissolved gas content, m3/t. Coal seams are well developed with abundant micropores, and the surface of coal molecules possesses a strong physical adsorption capacity for gas molecules. In coalbed methane, adsorbed gas generally accounts for 70% to 95% of the total coalbed methane volume. The adsorbed gas content can be estimated with the Langmuir isothermal adsorption equation, as follows 14 Vads=VL·PPL+P where V L is the Langmuir volume, m3/t, and P L is the Langmuir pressure, MPa. The results of high-pressure isothermal adsorption experiments conducted on 58 coal samples from the Linxing–Shenfu Block, Ordos Basin, indicate that both V L and P L are positively correlated with coal rank R o,max. Specifically, V L and P L gradually increase with increases in R o,max (Fig. a,b). V L and P L exhibit a positive U-shaped relationship with temperature (T): as temperature increases, V L and P L first decrease and then increase, with the transition temperature at approximately 54 °C (Fig. c,d). This pattern is consistent with the conclusions of a study on the adsorption characteristics of deep-seated medium- and low-rank coals in the Ordos Basin. , Scatter plots for different parameters. (a) and (b) cross-plots of Lambert volume (VL ) and Lambert pressure (VP ) against maximum vitrinite reflectance (R o,max); (c) and (d) cross-plots of VL and VP versus temperature (T). Based on the above correlation analysis, nonlinear regression analysis was adopted to establish prediction models for Langmuir isothermal adsorption constants V L and P L for coal in the study area on a dry ash-free (DAF) basis, which are expressed as follows 15 VL=(1.946Ro,max2−2.919Ro,max+2.575)(0.055T2−5.881T+166.632) 16 PL=(0.02Ro,max0.272)(1.276T2−141.630T+4041.828) Dissolved gas refers to the gas that exists in a dissolved state within pore water in coal seams. These gas molecules do not form an independent gas phase in the fluid but are dissolved in the liquid phase. The proportion of dissolved gas is generally low, accounting for <5% of the total coalbed methane volume. Nevertheless, its occurrence exerts certain influences on the occurrence state and production process of CBM. The dissolved gas content can be calculated by eq () 17 Vdis=φ·Sw·C where C represents the dissolved gas content per unit volume of fluid, m3/m3, which can be determined according to Henry’s Law, as follows 18 C=18.133PT−237.76 Free gas refers to the gas occurring as an independent phase (gaseous or supercritical state) within coal seam pores, which is not adsorbed on the coal matrix surface nor dissolved in pore water. Free gas can flow freely when stored in relatively large pores and fractures. Adsorbed gas dominates the CBM occurrence in shallow coal seams, while free gas constitutes fewer than 5% of the total gas volume. In deep coal seams, however, the adverse effect of temperature facilitates the release of some adsorbed gas back into the free gas state, leading to a remarkable increase in the proportion of free gas, which can reach 40%. Therefore, the free gas content must be accounted for to accurately estimate the total gas content. The free gas content calculation requires the reasonable determination of the effective pore volume for free gas occurrence in coal. However, previous free gas content calculations (eq ) have generally ignored pore volume variation under in situ formation conditions and neglected the occupation of pore space by adsorbed gas, resulting in overestimations of the free gas content. 19 Vfree=ρgrφ0(1−Sw)ρgρb where ρ gr is the density of free gas under in situ conditions, t/m3. In coal, free gas occupies the residual part of the total pore volume after deducting the space occupied by adsorbed gas and pore water (Fig. ). The porosity calculation method proposed in this study comprehensively considers the influences of formation conditions on porosity, as well as the occupation of pore space by adsorbed gas on free gas porosity. The corrected free gas content can be determined as follows 20 Vfree_new=ρgr(φ(1−Sw)−φads)ρgρb where V free_new is the corrected free gas content, m3/t. In summary, the total gas content without considering porosity variation caused by formation factors is calculated by eq () 21 Vtotal=Vads+φ0·Sw·C+Vfree The total gas content considering porosity changes caused by formation factors is calculated by eq () 22 Vtotal=Vads+φ·Sw·C+ρgrφ(1−Sw)ρgρb The total gas content considering porosity variation and pore space occupation by adsorbed gas is calculated by eq () 23 Vtotal=Vads+φ·Sw·C+Vfree_new Division Method of In Situ Stress Field Transition Depth For normal compaction conditions, the sediment compaction degree rises as burial depth increases, and the acoustic interval transit time of the sediments exhibits regular, depth-dependent variation. This relationship can be characterized by the establishment of a normal compaction trend line. In the presence of an anomalous pore pressure, sediment compaction was inhibited, resulting in deviations in the acoustic travel time from the normal compaction trendline. The Eaton method quantifies the pore pressure magnitude by leveraging this deviation. Using acoustic interval transit time logging data in the research region, a log–linear relationship was applied within the normal compaction interval to derive the expression for the normal compaction curve, and the acoustic interval transit time at the surface t 0 was obtained as follows 24 ln⁡t=−kZ+ln⁡t0 where t is the acoustic interval transit time at depth Z and k is a constant coefficient. Eaton proposed an empirical formula for predicting formation pore pressure in 1972 25 P0=σV−(σV−Pn)(tt0)m where σ V is the vertical principal stress, P n is the hydrostatic pressure, and m is the Eaton coefficient. Vertical principal stress σ V term denotes the vertical principal stress, which primarily originates from the gravitational load of overlying strata. The calculation formula is as follows 26 σV=∫0Zρ(Z)gdz where ρ(Z) is the formation density at depth Z and g is the acceleration due to gravity. Hydrostatic pressure is the pressure generated by the weight of fluids in the pores of the overlying strata. When the pore connectivity is good and the fluids are in a static state, the pressure increases uniformly with depth as follows 27 Pn=ρwgZ where ρ is the formation water density, which is generally taken as 1.03 t/m3. The Eaton coefficient is obtained from the measured pore pressure 28 m=ln[(σV−P0)/(σV−ρw)]ln(tt0) The principal horizontal stresses (maximum horizontal stress σ H and minimum horizontal stress σ h ) were predicted using the model proposed by Professor Huang Rongzun from China University of Petroleum. It is assumed that the in situ stress in underground rock formations is primarily generated by the overburden pressure and tectonic stress in the horizontal direction. 29 {σH=(μ1−μ+ξ1)×(σV−αP0)+αP0σh=(μ1−μ+ξ2)×(σV−αP0)+αP0 where μ is Poisson’s ratio; ξ 1 and ξ 2 are the tectonic stress coefficients in the regional horizontal direction. These coefficients have constant values in a specified area and can be derived from the measured fracture pressure, where α is the effective-pressure coefficient. Results and Discussion Controlling Factors of the Gas Content Prediction Model Analysis of the coalbed gas content prediction model can help elucidate the control mechanisms such as coal metamorphism degree, reservoir physical properties, and temperature-pressure conditions on the adsorption-desorption, occurrence, and enrichment of coalbed methane. The model clarifies the action mechanisms, primary-secondary relationships, and coupling effects of each factor, provides theoretical support for establishing a refined gas content prediction model considering multiparameter collaborative constraints, and enriches the theoretical system of coalbed methane reservoir evaluation and resource prediction. Taking Well LS-7 as an example, based on the normal compaction curve shown in Fig. a (fitting parameters are listed in Table ), the Eaton method was employed to obtain the pore pressure prediction curve in Fig. b. Furthermore, according to the porosity model proposed in this paper, which takes formation factors, including matrix compression effect, thermal expansion effect and coal adsorption-swelling effect, into account, the variation curve of porosity with burial depth in Fig. c was acquired. Combined with the variation characteristics of free methane density with burial depth (temperature and pressure) obtained from the NIST database (Fig. d), the above parameters were substituted into the gas content prediction model established in this paper. Ultimately, we derived the curve depicting the evolution of the gas content with burial depth (Fig. e). Schematic diagram of gas content prediction: (a) Fitting of normal compaction curve; (b) pore pressure prediction curve; (c) porosity prediction curve; (d) free methane density curve; and (e) gas content prediction curve. 2Fitting Parameters of the Normal Compaction Curve (ln t = −kZ + ln t 0)Well NumberLS-1LS-2LS-3LS-4LS-5LS-6LS-7 k 1.9483e-041.6189e-048.3613e-056.1754e-051.3853e-043.2386e-056.7988e-05 t 0/(μs/ft)77.699778.980475.692271.683877.611168.162069.8485Well NumberLS-8LS-9LS-10LS-11LS-12LS-13LS-14 k 3.1703e-045.866e-057.3215e-058.6103e-051.5562e-042.0377e-042.2387e-04 t 0/(μs/ft)90.503668.136966.981375.480681.861577.369981.0359 Fitting Parameters of the Normal Compaction Curve (ln t = −kZ + ln t 0) The dissolved gas content in coal seams is extremely low and exhibits an insignificant variation with burial depth. As the burial depth deepens, the volume of free gas grows steadily, with values climbing from 0.69 m3/t to 5.26 m3/t. Within shallow strata, the coal porosity decreases progressively with burial depth. The concurrent rise in formation pressure drastically elevates free gas density, leading to a continuous growth in total free gas volume. In deep strata, the porosity undergoes a slow upward trend with deeper burial. Moreover, elevated in situ temperatures at depth transform part of the adsorbed methane into free gas, providing an additional contribution to free gas accumulation. The adsorbed gas content presents a segmented variation trend, increasing initially and subsequently decreasing with greater burial depth. In the first stage, the adsorption capacity increases rapidly with pressure and shows an approximately linear growth. At this stage, the formation pressure rises significantly as the burial depth increases. The increase in pressure remarkably enhances the methane adsorption capacity of the coal seams. In addition, the formation temperature in shallow strata is relatively low, which exerts no obvious inhibitory effect on the adsorption. In the second stage, the growth of the adsorption capacity slows. After entering this stage, the formation pressure continues to rise, while the contribution of pressure to methane adsorption gradually weakens, and the growth rate of the adsorbed gas content begins to decline. Meanwhile, coal pores are gradually compacted under high pressure, leading to the reduction of pore volume and limiting the further growth space of the adsorbed gas content. The increase of burial depth raises the formation temperature, which partially offsets the promoting effect of pressure on adsorption. Consequently, the adsorption capacity still increases, but at a gradually slowing rate. In the third stage, the adsorption capacity decreased progressively. With the further increase of temperature, the inhibitory effect of temperature on adsorption is significantly strengthened, which gradually counteracts the promotion effect of high pressure on adsorption. Moreover, coal pores are further compressed under overburden pressure, and the reduction of pore space and specific surface area limits the storage space for adsorbed gas, eventually resulting in a gradual decline of the adsorbed gas content. To analyze the influences of initial porosity, water saturation, vitrinite reflectance, and temperature-pressure conditions on the total gas content, four scenarios were calculated separately by the prediction model based on the measured data of the 8+9# coal seam in well LS-7. First, the water saturation is fixed at 11.05% and vitrinite reflectance is set to 1.22%, with temperature and pore pressure conditions determined by measured well data; the initial porosity increases from 2% to 14% (Fig. a). Second, the initial porosity is fixed at 6% and vitrinite reflectance remains 1.22%, with temperature and pore pressure based on measured well data; the water saturation rises from 10% to 70% (Fig. b). Third, the water saturation is fixed at 11.05%, and the initial porosity is set to 6%, with temperature and pore pressure provided by measured well data; vitrinite reflectance increases from 0.8 to 1.4 (Fig. c). Fourth, water saturation is fixed at 11.05%, vitrinite reflectance at 1.22%, and initial porosity at 6%; the influence of temperature is superimposed under variable pore pressure conditions (Fig. ). Curve of total gas content: (a) Effect of initial porosity on total gas content; (b) effect of water saturation on total gas content; and (c) effect of vitrinite reflectance on total gas content. Variation curves of total gas content with temperature and pore pressure. It is found that the initial porosity, vitrinite reflectance, and water saturation exert distinctly differential controlling effects on the total gas content of coal seams. With the continuous increase in initial porosity, the pore and fracture development of coal reservoirs is enhanced, which not only provides more storage space for free methane gas but also improves the pore connectivity. This facilitates gas accumulation and preservation, leading to a continuous upward trend in the total gas content. An increase in vitrinite reflectance indicates a higher degree of coal metamorphism, accompanied by a growth in the specific surface area and adsorption sites of micropores within the coal. Correspondingly, the methane adsorption capacity is significantly strengthened, which further promotes a gradual increase in the total gas content. On the contrary, higher water saturation means that more pore and fracture space in coal seams is occupied by formation water. A large number of adsorption sites are covered by water films, which inhibit the methane adsorption and storage capacity. Meanwhile, formation water occupies the storage space of free gas, resulting in a continuously decreasing trend of total gas content as water saturation increases. With the variation of burial depth, the total gas content of coal seams is synergistically governed by pore pressure and formation temperature, showing an obvious regular variation trend. As pore pressure continuously rises, the compression degree of methane in coal reservoirs increases, and the storage capacity of free gas is significantly improved. Meanwhile, higher formation pressure can enhance the adsorption capacity of coal matrix for methane and promote the simultaneous increase of adsorbed gas content. Consequently, the total gas content presents a continuous increasing trend with the rise of pore pressure. In contrast, temperature exerts an inhibitory effect on gas content. Methane adsorption represents a typical exothermic reaction. Increasing in situ temperature reduces the binding strength between methane molecules and coal matrices, which promotes the desorption and escape of adsorbed gas. In addition, elevated temperatures accelerate the thermal movement of gas molecules and degrade the reservoir gas-sealing capability. Coupled temperature–pressure interactions induce a progressive reduction in total gas content as formation temperature rises. Characteristics of In Situ Stress Distribution Taking Well LS-12 as an example, based on the normal compaction curve shown in Fig. a (with the fitting parameters listed in Table ), the pore pressure prediction curve shown in Fig. b was obtained using the Eaton method. Flowchart of prediction: (a) Normal compaction curve fitting; (b) Pore pressure prediction; (c) Horizontal principal stress estimation; and (d) critical depth determination. Conventional in situ stress prediction relies on fracturing operation data or in situ stress testing. Obtaining such data is notably costly, and in situ stress data remain scarce in the Linxing–Shenfu Block. In contrast, well logging curves enable the acquisition of continuous vertical variation patterns across the entire wellbore. In the absence of fracturing data, reconstructing a continuous in situ stress profile and achieving reliable critical depth prediction via well logging curves not only mitigates the heavy reliance of traditional stress models on fracturing data and fills gaps in exploration data sets but also provides theoretical support for optimizing “sweet spot” target zones in deep CBM development. Using existing results from three injection-falloff well tests and in situ stress tests conducted in the central of the Linxing–Shenfu Block (Table ) and integrating the in situ stress model proposed by Professor Huang Rongzun (eq ()), the maximum and minimum horizontal stresses of the 8+9# coal seam in the research region were estimated (Table ). With these stress values serving as constraint conditions, the additional regional stress coefficients in the model were inversely calculated, yielding the average values of ξ 1 = 0.6019 and ξ 2 = 0.3020. From this analytical process, the single-well in situ stress profile presented in Fig. c can be generated. 3Measured In Situ Stress Data of the Central Parts in Linxing–Shenfu BlockWell NumberAverage Burial DepthPore Pressure P 0 Closure Pressure P c Breakdown Pressure P b Minimum Horizontal Stress σ h Maximum Horizontal Stress σ H LS-52260.521.6242.2856.8942.2851.52LS-62523.324.4748.1663.1048.1658.92LS-82313.924.5443.3555.1543.3552.55 Measured In Situ Stress Data of the Central Parts in Linxing–Shenfu Block 4Comparison between Injection/Falloff Method and Logging Calculation Results of In Situ Stress for 8+9# Coal Seam in the Central Parts of Linxing–Shenfu Block Maximum Horizontal Stress σ H /MPaMinimum Horizontal Stress σ h /MPaWell NumberDepth/mMeasured ValueCalculated ValueRelative Error/%Measured ValueCalculated ValueRelative Error/%LS-12070.2–2076.8–57.88––48.03–LS-21865.7–1880.3–44.28––37.04–LS-31933.1–1948.7–48.95––40.48–LS-42414.1–2429.3–59.93––48.29–LS-52255.4–2265.651.5253.494.0342.2843.993.84LS-62516.7–2529.958.9261.756.2748.1651.184.79LS-71975.0–1983.6–46.68––38.10–LS-82309.5–2318.352.5558.0612.9043.3548.9410.49LS-91979.1–1984.5–49.94––40.54–LS-102159.8–2166.1–50.93––41.99–LS-112092.9∼2107.7–55.18––46.58–LS-121861.9–1873.5–50.96––42.60–LS-131668.0–1682.9–42.35––34.96–LS-141696.0–1710.6–45.49––37.88– Comparison between Injection/Falloff Method and Logging Calculation Results of In Situ Stress for 8+9# Coal Seam in the Central Parts of Linxing–Shenfu Block Calculations derived from single-well logging data indicate that the in situ stress within the central of the Linxing–Shenfu Block gradually increases with increasing burial depth. By applying the same analytical approach to the existing logging data of 14 individual wells, a planar distribution map of the in situ stress in the roof of the 8+9# coal seam was generated using Surfer software and ordinary Kriging interpolation (Fig. ), with a grid cell size of 1 × 2 m2. This method is commonly used for planar mapping of geological reservoir parameters in coalfields and can effectively reflect regional trends in continuous distribution even when well data are limited. The critical depth map (Fig. ) was generated using exactly the same interpolation method and parameter settings to ensure consistency in the mapping approach. In the central Linxing–Shenfu Block, the in situ stress of the 8+9# coal seam roof is characterized by a south-high and north-low spatial distribution. Specifically, the vertical principal stress ranges from 41.31 to 59.90 MPa, with an average of 49.85 MPa; the maximum horizontal stress ranges from 42.35 to 61.75 MPa, with an average of 51.85 MPa; the minimum horizontal stress ranges from 34.96 to 51.18 MPa, with an average of 42.90 MPa; and the differential stress between the horizontal principal stresses ranges from 7.25 to 11.64 MPa, with an average of 8.95 MPa. In Situ Stress Distribution in the Roof of the 8+9# Coal Seam: (a) Vertical principal stress; (b) maximum horizontal stress; (c) minimum horizontal stress; and (d) difference between principal horizontal stresses. Contour map of critical depth: (a) Critical depth of gas content; (b) critical depth of in situ stress. Critical Depth Prediction The critical depth of gas content refers to the key depth threshold at which the total gas content (or adsorbed gas content) of coal seams changes from “continuous increase” to “slowing growth/decrease” with increasing burial depth. It is the boundary where the competitive effects of pressure (promoting adsorption and free gas enrichment) and temperature (inhibiting adsorption and promoting desorption) reach equilibrium, as well as the boundary at which the occurrence state and development conditions of CBM change significantly. The interval near and below the critical depth of gas content is a composite gas-bearing zone characterized by “saturated adsorbed gas + enriched free gas,” serving as the core target interval for high-yield “sweet spots” of deep CBM. As shown in Fig. d, the interval above the gas content critical depth belongs to the pressure-dominated stage. As burial depth increases sharply, formation pressure elevates rapidly, while the influence of temperature remains relatively insignificant. Pressure significantly enhances the methane adsorption capacity of the coal matrix and simultaneously increases the density of free gas; both adsorbed gas and free gas increase synchronously, and the total gas content presents an approximately linear rising trend. The vicinity of the gas content critical depth corresponds to the temperature–pressure equilibrium stage. As burial depth further increases, the inhibitory effect of rising temperature gradually strengthens and offsets the promotional effect of pressure on gas adsorption. The growth rate of adsorbed gas slows down obviously, whereas free gas continues to accumulate with increasing formation pressure, causing the total gas content to reach a peak or enter a high-value plateau stage. The interval below the gas content critical depth is dominated by temperature. The temperature effect becomes predominant, methane desorption is intensified, and the adsorbed gas content begins to decrease. Meanwhile, the coal matrix is compacted under high formation pressure, resulting in reduced pore space and restricted growth of free gas. Consequently, the total gas content changes from an increasing trend to a slowly decreasing trend. With the inflection point of the gas content curve adopted as the dividing standard, the distribution range of the gas content critical depth in the central Linxing–Shenfu block is determined to be 1076–1704 m (Fig. a). In geomechanics and rock mechanics, the critical depth for in situ stress state transformation denotes the crustal depth threshold at which the relative magnitude relationship between the vertical principal stress (σ V ) and principal horizontal stresses (σ H , σ h ) undergoes significant changes. This depth serves as a critical boundary where the rock mechanical behavior transitions from being dominated by shallow horizontal tectonic stress to deep vertical principal stress, and it holds a significant guiding value for deep CBM development. Shen et al. (2014) proposed an indicator for defining the critical depth based on the variation in the dominant direction of fracturing cracks induced by changes in the maximum principal stress field direction, denoted as G s = σ H /σ V . As illustrated in Fig. d, the shallow interval above the critical depth (particularly in tectonically active zones or areas adjacent to plate compressional boundaries) is dominated by the horizontal tectonic stress, as indicated by a G s value greater than 1. In this stress state, rocks typically develop dense horizontal joints or inclined faults, and engineering excavations are prone to sidewall sliding and horizontal shear failure. , With growing burial depth, the in situ stress field evolves to the critical burial depth, where vertical principal stress increases steadily and achieves equilibrium with horizontal stress, resulting in an G s value approaching 1. When the depth exceeds the critical depth, the gravitational effect of the overlying rock formations surpasses the tectonic compression, rendering the vertical principal stress dominant. The mechanical behavior of coal seams below the critical depth exhibits dual sensitivity: for high-permeability coal seams, vertical principal stress dominance accelerates pore fluid pressure and induces vertical compressive deformation; for low-permeability coal seams, residual horizontal stress may sustain fracture network stability, which facilitates the CBM reservoir stimulation. Using G s as an indicator, the critical depth in the central segment of the Linxing–Shenfu block was delineated to range from 1138 to 1727 m, characterized by a prevailing south-deep, north-shallow burial distribution (Fig. b). The Zijinshan pluton in the southern part of the study region provided a strong horizontal tectonic stress, leading to a relatively large critical depth. The composite critical depth was determined using the union principle: the lower limit is taken as the smaller of the lower limits of the two types of critical depths, and the upper limit is taken as the larger of the upper limits of the two types of critical depths. The critical depth range of the study area is determined to be 1076–1727 m. Within this range, both the gas occurrence conditions and the in situ stress field have entered the stage characteristic of deep coal seams, and this range can serve as a comprehensive boundary to guide stratified mining strategies. The calculation results for all 14 individual wells are summarized in Table . 5Statistics on Critical Depths for Individual Wells in Coal Seams 8 + 9 in the Study AreaWell NumberGas Content Critical Depth/mStress Field Critical Depth/mComprehensive Critical Depth/mLS-11300.951392.401300.95–1392.40LS-21156.801312.001156.80–1312.00LS-31448.901501.401448.90–1501.40LS-41279.751203.201203.20–1279.75LS-51532.101525.901525.90–1532.10LS-61611.651677.851611.65–1677.85LS-71318.601381.651318.60–1381.65LS-81704.901638.251638.25–1704.90LS-91404.701363.051363.05–1404.70LS-101495.201463.801463.80–1495.20LS-111648.651727.151648.65–1727.15LS-121462.201372.601372.60–1462.20LS-131076.351138.551076.35–1138.55LS-141411.251384.501384.50–1411.25 Statistics on Critical Depths for Individual Wells in Coal Seams 8 + 9 in the Study Area The spatial distributions of the two critical depths are highly coupled, which reflects the controlling effect of the in situ stress field on the evolution of coal seam gas content. Guidance of Critical Depth on Exploitation Variations in coal seam gas content and dominant in situ stress follow a three-stage trend with increasing burial depth. This characteristic exerts a direct control over CBM occurrence patterns, production mechanisms, and development schemes at different depths, providing critical theoretical guidance for the exploration and exploitation of deep coalbed methane resources. Depths above the critical value correspond to the pressure-dominated stage, where the gas content exhibits an approximately linear increase with burial depth. The reservoir is predominantly adsorbed-gas-dominated and generally undersaturated. The core development approach relies on dewatering and pressure reduction, which enables the desorption of adsorbed gas by lowering the reservoir pressure and represents a typical development mode for conventional CBM. Well pattern deployment requires an infill well arrangement along the horizontal stress direction to improve pressure relief sweep efficiency and promote methane desorption. Reservoir modification can be implemented with a relative ease. Governed by horizontal stress predominance, fracture propagation aligns with the maximum horizontal stress direction, and oriented perforation serves as an effective measure for the reservoir connectivity improvement. The interval near the gas content critical depth corresponds to the temperature-pressure equilibrium stage. The growth rate of adsorbed gas slows down, while free gas begins to accumulate, the total gas content reaches its peak or enters a high-value plateau, and the reservoir approaches saturation. This interval acts as a composite sweet spot with combined contributions from adsorbed gas and free gas, possessing dual advantages of desorption gas production and rapid free gas response and thus serves as the optimal target for deep CBM development. Development practices need to balance dewatering-pressure reduction and reservoir stimulation: on the one hand, pressure reduction promotes the desorption of adsorbed gas; on the other hand, hydraulic fracturing establishes migration pathways for free gas. As the fracture morphology begins to transition from horizontal to vertical extension, fracturing parameters should be optimized to avoid reservoir damage caused by excessive fracture propagation. Below the critical depth is the temperature-dominated stage. The adsorbed gas content decreases due to desorption under high temperature, while free gas continues to accumulate under high formation pressure and tight caprock sealing, maintaining a relatively high total gas content with generally saturated reservoirs. The development mode shifts from desorption-dominated gas production to free gas dominance, featuring rapid gas breakthrough after production commissioning, high initial production, and a long stable production period. Hydraulic fracturing is required to reconstruct the fracture network, counteract the sealing effect of the vertical principal stress, and provide migration pathways for free gas. Reservoir stimulation should focus on preventing interlayer channeling damage induced by overextended vertical fractures and meanwhile optimize drilling fluid density to address wellbore stability risks under high vertical principal stress. Above the critical depth (e.g., 4+5# coal seam), most CBM reservoirs exhibit undersaturation and are predominantly influenced by adsorbed gas. Therefore, to facilitate gas production, it is essential to implement water drainage strategies to lower the pressure and promote gas desorption. Below the critical depth (e.g., 8+9# coal seam), the coal is saturated with adsorbed gas. Compared with shallow coal seams, the caprocks of deep coal seams are denser and exhibit stronger sealing capacities and superior preservation conditions, which are more conducive to the free gas accumulation. Consequently, the deep CBM exhibits the distinct characteristics of immediate gas production upon commissioning and high yield once gas flow is established. From a development strategy perspective, in tectonically active intervals above the critical depth, although dynamic horizontal stress may elevate development risks, dense natural fracture networks are conducive to enhancing the initial productivity of the unconventional gas. At stable vertical principal stress levels below the critical depth, even though higher cost reservoir stimulation measures are necessary, the stability of the stress environment reduces long-term geological hazard risks. This makes the interval suitable for large-scale, long-term development plans. Therefore, an accurate definition of the critical depth is essential to balance safety, efficiency, and cost in deep coal seam development. Conclusions Within the central Linxing–Shenfu Block, roof in situ stress of the No. 8+9 coal seam rises with increasing burial depth, exhibiting a consistent south-high and north-low distribution pattern. The improved gas content prediction model corrects for the overestimation of free gas content in traditional models. Coal seam gas content exhibits a three-stage variation with burial depth. Integrated determination using gas content prediction and in situ stress prediction methods gives a critical depth range of 1076–1727 m for the study area, with greater depths in the south and shallower depths in the north. The two types of critical depths show strong spatial correspondence, and the zone of greater critical depth in the south is the preferred target for large-scale development. A layered development strategy is proposed based on the three-stage evolution law, and accurately defining the critical depth is a necessary prerequisite for the safe, efficient, and economical development of deep coalbed methane.