Mining-induced stress–gas permeability characteristics based on protective layer pressure-relief mining Authors: Kun Wang, Xiangrui Meng, Guangming Zhao, Xiang Cheng, Guanghui Han, Anhong Wang Published: 2025 ABSTRACT AbstractProtective layer pressure-relief mining is an effective strategy for preventing coal and gas outbursts in deep seams, while simultaneously facilitating safe and efficient gas extraction from low-permeability coal. This study employed the self-developed ZST 1500 test system to perform permeation tests on the axial pressure-relief zone (APRZ) and radial pressure-relief zone (RPRZ) of the protected layer under various gas pressure conditions. The primary objective was to elucidate the gas permeability mechanism of the protected layer under combined mining-induced stress and gas pressure during pressure-relief mining. The results indicated that the application of axial stress and seepage pressure during the initial seepage stage led to axial compression and circumferential expansion of the sample, thereby resulting in the formation of equivalent stable seepage channels. Comparative analysis revealed that the RPRZ exhibited higher gas permeability than the APRZ, while the gas permeability gradually rose as the sample’s circumferential stress decreased. Notably, the circumferential stress was identified as the predominant factor influencing the formation of equivalent stable seepage channels. The upper limit of gas permeability was determined by the number and extension of these equivalent stable seepage channels in the coal body. Furthermore, variations in seepage pressure significantly affected the degree to which the gas slippage effect enhanced sample gas permeability. A capillary-based permeability model for porous media effectively fitted the experimental permeability of the protected layer, thus validating the model’s feasibility. The findings provide a theoretical basis for delineating gas accumulation zones, optimizing design of gas drainage boreholes, and enabling safe and efficient co-extraction of coal and gas in underground mining operations.Keywords: Protective layer pressure-relief mining, Stress–strain path, Gas seepage, Sensitivity coefficient, Gas permeability enhancement rateSubject terms: Energy science and technology, Engineering Introduction China has abundant coal resources, which are projected to remain pivotal in safeguarding national energy security as a strategic “ballast”1 and serving as a stabilizer for structural adjustments and transitional development in the energy sector over the coming decades2. The secure extraction and efficient utilization of coal resources are critical prerequisites for ensuring national energy security and facilitating sustainable socioeconomic development3,4. As shallow coal resources are being exhausted, the development and utilization of deep resources have emerged as a new strategic focus5. However, the geological environment of deep coal seams, characterized by “five highs and two disturbances,” has led to a significant increase in gas content and in-situ gas pressure, thereby intensifying the risk of coal and gas outbursts6,7. Protective seam mining is an effective pressure-relief technique widely used to prevent coal and gas outbursts and to enable safe, efficient gas extraction from deep, low-permeability coal seams 8,9. The mining-induced stress redistribution in the protective seam propagates into adjacent protected seams to induce pressure relief and enhanced permeability. These changes significantly reduce gas pressure within the protected coal seams and therefore effectively mitigate the risk of coal and gas outbursts. Therefore, understanding the influence of the mining-induced stress field on the gas permeability of low-permeability coal seams is crucial for the safe and efficient co-mining of coal and gas in deep coal seams. In recent years, a considerable body of research has been conducted to investigating the characteristics of coalbed methane seepage10–14. Regarding the research on gas seepage theory and models, Terzaghi’s seminal contributions include the formulation of the effective stress equation for fluid–solid coupling in porous media and the establishment of a one-dimensional consolidation model15. Subsequently, he proposed a three-dimensional consolidation theory incorporating total stress and pore pressure as variables, thereby establishing the foundation for fluid–solid coupling16. Lubinski and Geertsma developed a theory of elasticity in porous media based on the Biot equation17,18 and established an analogy model between thermoelasticity and saturated porous elastic media21,22. Zienkiewicz introduced a theoretical framework that addresses the nonlinear characteristics of geometry and materials and formulated a generalized Biot equation based on three-dimensional consolidation theory23. Metigue established a thermos–fluid–solid coupled model that considers the effect of temperature on seepage based on Biot theory. This development marked a significant expansion of the theory of seepage in porous media24. In recent years, Zhou et al.25 developed an anisotropic permeability model for methane-bearing coal by considering three-dimensional stress conditions and the coupled effects of stress, methane adsorption, and temperature. Gu proposed a dual-pore fluid–solid coupled model based on the Biot linear elastic mechanics mechanism to reveal the migration characteristics of unconventional gases26. Yang et al.27 established a gas–solid coupled model for methane-bearing coal by incorporating factors such as adsorption, porosity, and permeability. Rong et al.28 investigated the effects of adsorption and effective stress and then constructed an anisotropic seepage model based on coal body damage. These findings provide a robust theoretical foundation for understanding the mechanisms of coalbed methane seepage. Research on pressure-relief mining of protective seams and the mechanisms of permeability enhancement has clarified how fracture development and stress redistribution increase coal-seam permeability. Wang et al.29 characterized the fracture and movement patterns of overlying strata during protective seam mining and demonstrated that stress release in the unloaded coal and rock masses generated new fractures that enhanced permeability. Liu et al.30 noted that mining of the protected seam generates bedding fractures—predominantly vertical or inclined—which form effective gas flow pathways. Xue et al.31 reported that pressure-relief mining of a lower protective seam at the Wulan Mine increased the permeability of the overlying protected seam by a factor of 2054 relative to its initial value. Zhang et al.32 found that, under constant stress conditions, the influence of the Klinkenberg effect and gas adsorption on permeability diminished as gas pressure increased. Zhang et al.33 discovered that the extraction of upper overlying seams subjects the underlying coal seam to cyclic loading–unloading stress paths, where elevated stresses promote the propagation of internal fractures and thereby substantially increase gas permeability. Zhang et al.34 investigated gas seepage under varying fracture conditions and stress paths and concluded that vertical fractures exert the greatest impact on gas permeability. In terms of research on factors influencing gas permeability, numerous studies have examined a wide range of factors, including surrounding rock stress35–39, pore structure40–42, water content43,44, gas desorption effects45–47, coal body damage48–50, and temperature51–53. Hatami et al.54 explored the impact of gas slippage effect and pore elasticity characteristics on the permeability of coal containing methane under constant confining pressure and effective stress conditions. Liu et al.55 conducted a series of experiments focusing on volume stress and pore pressure as key variables to elucidate the impacts of gas slippage effect and effective stress on methane permeability during the migration of methane in deep coal seams. Wang et al.56 utilized a nuclear magnetic resonance testing system equipped with a loading device to analyze permeability changes in coal throughout the stress–strain process. Furthermore, they investigated the influence of stress damage on coal permeability and concluded that poor connectivity of new pores resulted in a lag in coal permeability increase relative to the growth of coal pores. Wang et al.57 investigated pore distribution and characteristics utilizing a nuclear magnetic resonance permeability prediction model and revealed a negative linear relationship between coal permeability and the fractal dimension of non-adsorbed pores. They proposed a permeability prediction model that incorporates a fractal dimension factor for non-adsorbed pores. Teng et al.58 established a multi-physics coupling model considering coal deformation, gas flow, and water content. Also, they found that a decrease in water content increases the porosity and permeability of coal. Xiao et al.59 studied the deformation characteristics of water and gas adsorption expansion in coal fractures and matrix pores under the influence of water content. They also analyzed gas flow patterns of coal under varying water contents. Xiao et al.60 utilized low-frequency nuclear magnetic resonance and pulse pressure decay methodologies to investigate the pore structure characteristics and seepage patterns of coal bodies under varying water contents. Additionally, they integrated water content, effective stress, and gas desorption effects to construct an enhanced apparent permeability model and identified the critical pore pressure that enhances the effects of gas desorption. In summary, extensive studies have been conducted on the gas permeability in coal seams under mining-induced stress conditions. However, there is limited research on the low-permeability gas seepage mechanism during the pressure-relief mining of protective coal seams. To address these limitations, this study elucidated the gas seepage mechanisms in coal seams under the synergistic effects of mining-induced stress and gas seepage pressure during protective seam mining. The findings are expected to provide theoretical support for the pressure-relief mining of deep protective coal seams and hold significant implications for the safe and efficient extraction of depressurized gas during the mining of deep protective coal seams. Seepage modeling of gas-containing coal Porosity analysis of porous media Gas-containing coal is defined as a porous medium composed of solid skeleton particles, intergranular pores, and gas61. It has been established that gas-containing coal exhibits a unique deformation mechanism grounded in the concept of double effective stress. In this mechanism, the body effective stress governs body deformation, while the structural effective stress dictates structural deformation62. The skeletal stress σs of gas-bearing coal is influenced by several factors, including external load, seepage air pressure, and temperature. The interaction of these factors determines whether the coal body undergoes deformation. When the contact stress σc between the skeleton particles exceeds a certain threshold, it induces further structural deformation of the coal body. The combined effect of these two types of effective stresses promotes the deformation and damage of the coal body. Both the stress characteristics of the protected coal seam and mining-induced overlying rock migration characteristics are illustrated in Fig. 1. Diagram illustrating stress characteristics of the protected coal seam and mining-induced overlying rock migration characteristics. In this study, the following assumptions were made: There is only a single saturated gas fluid in the coal seam; There is only a single saturated gas fluid in the coal seam; The gas-containing coal is a homogeneous, continuous, and isotropic linear elastic body; The gas-containing coal is a homogeneous, continuous, and isotropic linear elastic body; The relationship between the strain resulting from the deformation of gas-containing coal and the effective stress follows the generalized Hooke’s law; The relationship between the strain resulting from the deformation of gas-containing coal and the effective stress follows the generalized Hooke’s law; The deformation of gas-containing coal is minimal; The deformation of gas-containing coal is minimal; The gas pressure, expansion stress, and strain in an isotropic linear elastic coal body are uniform in all directions. The gas pressure, expansion stress, and strain in an isotropic linear elastic coal body are uniform in all directions. The porosity φ of the coal body can be defined as follows: where VP is the pore volume; ΔVP is the change in the pore volume of the coal body; VB is the total appearance volume of the coal body; ΔVB is the change in the total appearance volume of the coal body; VS is the volume of the coal skeleton; ΔVS is the change in the volume of the coal skeleton; εV is the volumetric strain; φ0 is the initial porosity. The incremental compressive deformation of skeleton particles resulting from seepage pressure is expressed as follows: The incremental expansion and deformation of skeleton particles due to adsorption is expressed as follows: εp is defined as follows: The effect of temperature on the thermal expansion and deformation of skeleton particles can be expressed as follows: Substituting Eq. (1) into (5) yields the total strain of the coal skeleton particles as follows: Substituting Eq. (6) into Eq. (1) yields: The volume compression coefficient is defined as follows: Equation (9) can be derived based on the principles of elastic mechanics: Substituting Eqs. (8) and (9) into Eq. (7) yields: It can be demonstrated that the definition of the thermal expansion coefficient can be derived from the definition of the volume compression coefficient. Furthermore, the elastodynamics can be derived from the thermodynamic parameters that reflect the change in the average effective stress per unit volume in a gas-containing coal micrometamorphic body using the volumetricstrain relationship equation. Substituting Eq. (11) into Eq. (10) yields the porosity as follows: where KY is the volume compression coefficient in MPa-1; △p is the gas pressure change (p-p0) in MPa; β is the volume thermal expansion coefficient of coal m3/m3·K; △T is the absolute temperature change (T − T0) in K; a is the ultimate adsorption volume of coal per unit mass under reference pressure in m3/t; b is the adsorption equilibrium constant of coal in MPa-1; ρ is the apparent density of coal in t/m3; R is the universal gas constant, which is 8.3143 J/(mol·K); K is the bulk modulus of coal in Pa; Vm is the molar volume of gas, which is 22.4 × 10−3 m3/mol. Permeability model for porous media based on capillary features The interrelationship among permeability, porosity, and specific surface area can be derived from the Kozeny–Carman equation based on capillary modelling. where where k is the permeability in mD; KZ is a dimensionless constant, with a value of approximately 5; SP is the pore surface area per unit pore volume of the coal body in cm2; AS is the total surface area of the pores in the coal body in cm2; Ψ is the increment of surface area. The initial permeability can be expressed as follows: The following expression can be obtained from Eq. (6): The pore volume increment of gas-bearing coal is defined as follows: The new porosity can be expressed as follows: The new surface area can be expressed as follows: The ratio of the new permeability to the initial permeability is expressed as follows: Integrating the equations mentioned above, the final formula can be derived as: It is reasonable to assume that the change in the total surface area of coal particles per unit volume of the coal body can be neglected in the stress–strain relationship. Consequently, the following equation can be derived63: When the temperature is constant (i.e., ∆T = 0), there is: When the pore gas pressure is constant (i.e., ∆p = 0), there is: When both the temperature and pore gas pressure are constant, there is: Test apparatus and program Test setup and sample preparation Anhui University of Science and Technology has developed a microcomputer-controlled electro-hydraulic servo system for dynamic-static adaptive coupling tests of coal and rock (ZST-1500). This equipment is capable of facilitating thermal-flow-solid three-phase coupling experiments to investigate the permeability characteristics of the coal. The system features a self-balancing axial test force of 1500 kN, a radial test force of 140 kN, and a measurement accuracy of ± 1%. It allows for a maximum axial deformation of 10 mm, a maximum radial deformation of 5 mm, a gas pressure of 10 MPa, a stabilized pressure accuracy of ± 2%, and a temperature of 120 °C with a temperature resolution of 0.1 °C. It is noteworthy that this study specifically investigated the synergistic mechanisms by which mining-induced stress and gas seepage pressure jointly influenced coal permeability under the cyclic loading–unloading stress conditions encountered in protective seam mining. To streamline the experimental procedure, nitrogen was employed as the test gas, and all tests were conducted at a constant room temperature of 25 °C. This deliberate simplification excluded the effects of gas adsorption/desorption and thermal variations on coal permeability, thereby enabling a focused examination of the coupling between stress and gas pressure. The test device is depicted in Fig. 2. ZST-1500 microcomputer-controlled electro-hydraulic servo system for dynamic-static adaptive coupling tests of coal and rock. The coal samples utilized in the tests were sourced from the Luling Coal Mine, which is a part of Huaibei Mining Group. They were prepared in accordance with the sampling guidelines established by the International Society for Rock Mechanics and shaped into cylindrical samples with a diameter of 50 mm and a length of 100 mm. The selection of coal samples followed a rigorous, multi-step procedure to ensure consistency and reliability. First, ultrasonic wave testing was employed to assess fracture development and internal homogeneity. This operation ensured that all selected specimens exhibited comparable geomechanical properties. Subsequently, the end-face flatness was controlled to within 0.02 mm to prevent stress concentration-induced failure during loading. Previous seepage tests demonstrated that, under identical stress paths and gas pressures, coal specimens with mass deviations below 2 g exhibited negligible variability in both strain behavior and seepage behavior. Consequently, the final selection required all specimens to satisfy both the end-face flatness (< 0.02 mm) and mass deviation (< 2 g) criteria. The test procedure is outlined as follows: Installation. The coal sample and fixture were fixed using a dense latex sleeve. A heat gun was utilized to shrink the heat-shrink tube around the sample. Subsequently, the axial and radial deformation transducers were installed, and the sample was then repositioned to the base of the triaxial self-balancing pressure chamber. Finally, a sealing ring was employed to ensure the airtightness of the triaxial self-balancing pressure chamber. Installation. The coal sample and fixture were fixed using a dense latex sleeve. A heat gun was utilized to shrink the heat-shrink tube around the sample. Subsequently, the axial and radial deformation transducers were installed, and the sample was then repositioned to the base of the triaxial self-balancing pressure chamber. Finally, a sealing ring was employed to ensure the airtightness of the triaxial self-balancing pressure chamber. Oil filling. The pressure chamber of the triaxial pressure chamber was filled with pressure-resistant silicone oil. The valve of the pressure chamber was closed after the completion of oil filling. This marks the conclusion of the test preparation stage. Oil filling. The pressure chamber of the triaxial pressure chamber was filled with pressure-resistant silicone oil. The valve of the pressure chamber was closed after the completion of oil filling. This marks the conclusion of the test preparation stage. Pressurization. The sample was loaded to a horizontal stress using a fully digital servo controller (EDC). Pressurization. The sample was loaded to a horizontal stress using a fully digital servo controller (EDC). Seepage. The steady-state method was employed to assess the sample’s gas permeability. The gas pressure cylinder at the inlet of the gas pressure cabinet was pressurized with a specific gas. The hand valve at the inlet was opened to allow the gas to infiltrate the sample and subsequently flow into the gas cylinder at the outlet. The gas permeability of the coal sample was then measured when the gas pressure at the inlet of the gas cylinder remained constant. Seepage. The steady-state method was employed to assess the sample’s gas permeability. The gas pressure cylinder at the inlet of the gas pressure cabinet was pressurized with a specific gas. The hand valve at the inlet was opened to allow the gas to infiltrate the sample and subsequently flow into the gas cylinder at the outlet. The gas permeability of the coal sample was then measured when the gas pressure at the inlet of the gas cylinder remained constant. Variable load. The axial, circumferential, and seepage pore pressures of the coal samples were sequentially varied based on the designed experimental paths. This approach allows for the measurements of the gas permeability, axial deformation, and circumferential deformation of the samples at different experimental stages. Variable load. The axial, circumferential, and seepage pore pressures of the coal samples were sequentially varied based on the designed experimental paths. This approach allows for the measurements of the gas permeability, axial deformation, and circumferential deformation of the samples at different experimental stages. Methods for calculating gas permeability The ZST-1500 microcomputer-controlled electro-hydraulic servo system for dynamic and static adaptive coupling tests of coal and rock employs sophisticated mathematical calculations to determine the gas permeability of various coal body configurations. The gas permeability assessment was achieved through meticulously monitoring of pressure changes in a gas cylinder at the inlet. Given the known height and cross-sectional area of the specimen, the gas permeability could be calculated by measuring the pressure change ΔP of the gas cylinder at the inlet over a designated time period Δt. The gas pressure underwent significant fluctuations during the initial phase of gas infiltration. Once the cylinder pressure stabilized, the infiltration data were collected for further analysis. The pressure of the gas cylinder at the inlet following the stabilization of gas pressure is denoted as P1. The change in gas pressure at the inlet after the time interval Δt is referred to as ΔPA. The average flow rate of gas through the coal specimen: The gas permeability k1 at the inlet is then given as follows64: where μ is the gas viscosity coefficient, μ = 1.14 × 10–5 Pa s, V0 is the volume of the gas cylinder, A is the cross-sectional area of the specimen, L is the length of the specimen, P0 is the atmospheric pressure in the standard state. Test path The integration of protective layer pressure-relief mining with gas extraction technology has emerged as a pivotal strategy for ensuring safe co-mining of coal and gas in deep and strongly protruding coal seams. Safely extracting the protective coal layer enables the overlying protected layer to experience mining-induced stress, which promotes the release of both the original stress and the stored gas in that layer. This process effectively mitigates the coal and gas outbursts in the protected layer. The following study aimed to characterize the transport dynamics of the coal body within the protected coal seam in relation to the mining process65. The overlying protected coal layer in the middle of the goaf is subject to axial tension and radial compression due to mining-induced stress in the protected layer. This phenomenon is referred to as the “axial pressure-relief zone (APRZ)”. In contrast, the overlying protected layer on the coal wall side of the working face and the cut-off side experiences axial compression and radial tension, which is designated as the “radial decompression zone” or “radial pressure-relief zone (RPRZ).” Various stress paths were designed to investigate the deformation characteristics of the APRZ of the protected layer, as illustrated in Fig. 3. Constant gas pressure test: Initially, axial and circumferential pressures were loaded to 12.5 MPa at a rate of 50 kN/min to simulate the original stress state. Subsequently, they were varied at a rate of 50 kN/min until the axial pressures reached 12.5, 11.25, 10, 8.75, and 7.5 MPa, while the circumferential pressures reached 12.5, 13.75, 15, 16.25, and 17.5 MPa, respectively. A constant gas pressure of 4 MPa was maintained at each stage. The sample’s gas permeability was assessed when the infiltration flow of permeable gases stabilized. Constant gas pressure test: Initially, axial and circumferential pressures were loaded to 12.5 MPa at a rate of 50 kN/min to simulate the original stress state. Subsequently, they were varied at a rate of 50 kN/min until the axial pressures reached 12.5, 11.25, 10, 8.75, and 7.5 MPa, while the circumferential pressures reached 12.5, 13.75, 15, 16.25, and 17.5 MPa, respectively. A constant gas pressure of 4 MPa was maintained at each stage. The sample’s gas permeability was assessed when the infiltration flow of permeable gases stabilized. Gas pressure reduction test: The stress path for this test was identical to that for the constant gas pressure test. The gas pressure at each stage was set at 4.5, 4.75, 5, 5.25, and 5.5 MPa, respectively. The sample’s gas permeability was rigorously tested following the stabilization of the permeate gas in the seepage. Gas pressure reduction test: The stress path for this test was identical to that for the constant gas pressure test. The gas pressure at each stage was set at 4.5, 4.75, 5, 5.25, and 5.5 MPa, respectively. The sample’s gas permeability was rigorously tested following the stabilization of the permeate gas in the seepage. Test program for APRZ. Various stress paths were designed to investigate the deformation characteristics of the RPRZ of the protected layer, as depicted in Fig. 4. Test program for RPRZ. Constant gas pressure test: Initially, axial and circumferential pressures were loaded to 12.5 MPa at a rate of 50 kN/min to simulate the original stress state. Subsequently, they were varied at a rate of 50 kN/min until the axial pressures reached 12.5, 13.75, 15, 16.25, and 17.5 MPa, while the circumferential pressures reached 12.5, 11.25, 10, 8.75, and 7.5 MPa, respectively. It should be noted that a constant gas pressure of 4 MPa was maintained across all stages. The sample’s gas permeability was tested when the infiltration flow of permeable gases stabilized. Gas pressure reduction test: This test followed the same stress path as the constant gas pressure test. The gas pressure at each stage was set at 3.5, 3.25, 3, 2.75, and 2.5 MPa, respectively. The sample’s gas permeability was tested when the infiltration flow of permeable gases stabilized. Mechanical characterization of test results Strain characteristics Figure 5 illustrates the changes in strain of the APRZ of the protected layer over time during the simulation experiments. Variation of strain of the APRZ over time. Figure 5 indicated that during both the constant gas pressure stage and the gas pressure increase stage, the axial strain of the APRZ gradually reduced, while the circumferential strain gradually increased, under a stress path where circumferential pressure increased while axial pressure applied to the sample decreased. Specifically, the axial strain of the sample during the constant gas pressure stage decreased from 6.034 × 10–3 to 5.505 × 10–3, representing a decrease of 8.77%. In contrast, the circumferential strain increased from 1.137 × 10–3 to 2.431 × 10–3, reflecting an increase of 113.8%. The axial strain of the sample decreased from 7.212 × 10–3 to 6.047 ε × 10–3 during the gas pressure increase stage, indicating a reduction of 16.15%. Concurrently, the circumferential strain varied from 1.099 × 10–3 to 2.338 × 10–3, representing an increase of 112.74%. Notably, for the same stress path, the axial strain in the gas pressure increase stage is larger than that in the constant gas pressure stage, whereas the circumferential strain is smaller. For samples experiencing axial pressure-relief and circumferential pressure-increase, as the permeable gas pressure increased, the change in axial strain during the gas pressure increase stage was less than the circumferential strain observed during the constant gas pressure stage. This is due to the fact that the axial force, influenced by the sample’s axial permeable gas pressure, increased during the gas pressure increase stage compared to the constant gas pressure stage, resulting in a greater axial strain. Furthermore, the permeable gas pressure within the sample was relatively high during the gas pressure increase stage. This led to significant expansion of the pores and microscopic fissures, alongside a relatively large circumferential strain in the sample. Figure 6 illustrates the changes in strain of the RPRZ of the protected layer over time during the simulation experiments. Variation of strain of the RPRZ over time. Figure 6 revealed that during both the constant gas pressure stage and the gas pressure decrease stage, the axial strain of the RPRZ gradually increased, while the circumferential strain gradually decreased. Additionally, radial expansion and deformation were observed in the sample. Specifically, the axial strain of the sample during the constant gas pressure stage increased from 6.00 × 10–3 to 7.539 × 10–3, indicating an increase of 25.65%. In contrast, the circumferential strain changed from 1.123 × 10–3 to − 0.469 × 10–3, indicating a decrease of 141.76%. The axial strain of the sample during the gas pressure decrease stage increased from 5.6 × 10–3 to 6.705 × 10–3, demonstrating an increase of 19.73%. In contrast, the circumferential strain varied from 1.305 × 10–3 to − 0.051 × 10–3, reflecting a decrease of 103.91%. Notably, for the same stress path, the axial strain in the gas pressure decrease stage is smaller than that in the constant gas pressure stage, whereas the circumferential strain is larger. The RPRZ of the protected layer exhibited axial compression and circumferential expansion under the axial pressure-relief condition. Under the same stress path, the axial strain of the sample decreased, whereas the circumferential strain increased during the gas pressure reduction stage compared to the constant gas pressure stage. This is due to the fact that, under the same stress path, a reduction in the gas pressure at the top of the sample diminished its effect on axial deformation. Additionally, the expansion of microscopic cracks within the coal body decreased during the gas pressure reduction stage. Consequently, the sample exhibited a trend of decreased axial strain and increased circumferential strain. Gas permeability characteristics Figure 7 illustrates the changes in gas permeability of the APRZ of the protected layer over time during the simulation experiments. Variation of gas permeability of the APRZ over time. As illustrated in Fig. 7, the gas permeability of the APRZ decreased over time during both the constant gas pressure stage and the gas pressure decrease stage. Specifically, the gas permeability during the constant gas pressure stage decreased by 23.08% from 1.93587 × 10–20 m2 to 1.48913 × 10–20 m2. In contrast, the gas permeability during the gas pressure increase stage declined by 10.26% from 0.89862 × 10–20 m2 to 0.80638 × 10–20 m2. Under the same stress path, the gas permeability during the gas pressure increase stage was 50.328% lower than that observed during the constant gas pressure stage. This indicated that higher gas pressure within the APRZ was associated with decreased sample’s gas permeability. This study introduced the concept of an “equivalent stable seepage channel” to characterize the gas seepage behavior in coal. The observed increase in coal specimen permeability and the circumferential expansion during testing collectively indicated the progressive development and improved connectivity of these channels within the coal matrix. Specimens from the APRZ subjected to axial unloading and concurrent circumferential loading experience a stress state that causes contraction of their internal channels. Consequently, permeability decreased progressively as circumferential stress increased. The “gas slippage effect” occurs in gas flow through micro-/nano-scale porous media. It denotes the condition in which gas molecules exhibit a non-zero tangential velocity (“slip”) at the pore walls, which is contrary to the zero-velocity assumption in classic fluid mechanics. At low pore pressures, when the mean free path of gas molecules was comparable to the pore size, this slippage effect significantly enhanced the apparent flow capacity of the gas. Consequently, measured gas permeability exceeded the intrinsic (absolute) permeability of the rock matrix. During the gas pressure increase within the APRZ, the rising pressure shortened the mean free path of the seeping gas molecules and intensified intermolecular collisions. This reduced the contribution of the gas slippage effect to specimen permeability. Under identical stress paths, the average permeability during the pressure-increase stage was approximately 50.328% lower than that during the stable gas pressure stage. Figure 8 illustrates the changes in gas permeability of the RPRZ of the protected layer over time during the simulation experiments. Variation of gas permeability of the RPRZ over time. As depicted in Fig. 8, the gas permeability of the sample in the RPRZ increased gradually over time during both the constant gas pressure stage and the gas pressure decrease stage. Specifically, the gas permeability during the constant gas pressure stage increased by 35.64% from 1.81668 × 10–20 m2 to 2.46413 × 10–20 m2. In contrast, the gas permeability during the gas pressure reduction stage increased by 40.32% from 0.90946 × 10–20 m2 to 1.27614 × 10–20 m2. The average gas permeability during the gas pressure reduction stage was 47.434% lower than that during the constant gas pressure stage under the same stress condition. There is a significant development of “equivalent stable seepage channels” in the RPRZ during the axial pressure-relief and circumferential pressure-increase stage. This development resulted a gradual increase in the gas permeability of the sample as radial stress decreased. Moreover, under same constant gas pressure conditions, the gas permeability of the RPRZ was significantly higher than that of the APRZ. During the process of reducing the gas pressure, the average molecular free path of the gas increased, which enhanced the gas slippage effect on gas permeability. However, the abrupt decrease in gas pressure resulted in the radial strain of the sample under radial stress conditions being generally greater than that observed under constant gas pressure conditions. Consequently, some of the “equivalent stable seepage channels” gradually closed as the gas pressure decreased. This eventually led to the formation of a RPRZ under the same stress path. The sample’s gas permeability during the gas pressure reduction stage was, on average, 47.434% lower than that observed during the constant gas pressure stage. The fitting results, derived from the capillary-based gas permeability model for the porous media, are presented in Table 1. This model incorporates both the APRZ and RPRZ at each stage of the gas permeability tests. Table 1Validation of the fitting results for gas permeability tests conducted in the APRZ.Constant gas pressureIncreased gas pressureTime/hFitted gas permeability/10−2012 m2Experimental gas permeability/10−2012 m2Absolute error/10−2312 m2Relative error/%Time/hFitted gas permeability/10−2012 m2Experimental gas permeability/10−2012 m2Absolute error/10−2315 m2Relative error/%21.935871.935870020.898620.898620041.8151.806558.450.46840.8760.874351.650.18961.7231.726563.560.20660.8530.852070.930.10981.6181.6291711.170.68580.8290.8151813.821.708101.4891.489130.130.009100.8060.806380.380.047 Validation of the fitting results for gas permeability tests conducted in the APRZ. Table 1 demonstrated that the maximum fitting error for gas permeability in the APRZ was 0.685% during the constant gas pressure stage and 1.708% during the gas pressure increase stage. Table 2 revealed that the maximum fitting error for gas permeability in the RPRZ was 0.561% during the constant gas pressure stage and 0.361% during the gas pressure reduction stage. The porous media seepage model, developed based on capillary characteristics, quantifies the relationship between permeability and coal porosity under the coupled effects of stress, gas adsorption/desorption, and temperature. It clarifies how mining-induced stress and gas pressure alter the coal microstructure (porosity φ) and thereby affect macroscopic permeability. The model achieved a fitting error of less than 2% against experimental data from two different pressure-relief zones and along two distinct gas pressure paths. This finding demonstrated the model’s excellent accuracy and strong generalization capability. Its comprehensive consideration of mining-induced stress, gas adsorption/desorption, and temperature regulation renders it particularly suitable for practical engineering applications. Table 2Validation of the fitting results for gas permeability tests conducted in the RPRZ.Constant gas pressureDecreased gas pressureTime/hFitted gas permeability/10−20 m2Experimental gas permeability/10−20 m2Absolute error/10−23 m2Relative error/%Time/hFitted gas permeability/10−20 m2Experimental gas permeability/10−20 m2Absolute error/10−23 m2Relative error/%21.816681.816680020.909460.909460042.0082.004873.130.15641.0431.046783.780.36162.0722.0604511.550.56161.0921.090641.360.12582.2892.283095.910.25981.2751.274810.190.015102.4642.464130.130.005101.2761.276140.140.011 Validation of the fitting results for gas permeability tests conducted in the RPRZ. Gas permeability mechanism of the protected layer based on mining-induced stresses Gas permeability stress/gas pressure sensitivity coefficient The variation of the gas permeability of the coal body in the protected layer is primarily influenced by the stress environment and gas pressure. Considering the evolution characteristics of stress paths for the APRZ and RPRZ of the protected layer, the gas permeability stress sensitivity coefficient66 was introduced to quantify the impact of mining-induced stress on gas permeability under constant gas pressure conditions. where αe is the gas permeability stress sensitivity coefficient; K0 is the initial gas permeability of the sample in m2; ΔK is the gas permeability difference in m2; Δσe is the axial-circumferential pressure difference in MPa. In the protected layer, both the APRZ and RPRZ follow the same stress paths. The sensitivity of gas permeability to gas pressure was further analyzed through variations in gas pressure. The formula for calculating the gas permeability gas pressure sensitivity coefficient is defined as follows: where βe is the gas permeability gas pressure sensitivity coefficient; K0 is the initial gas permeability of the sample in m2; ΔK is the gas permeability difference in m2; ΔPe is the gas pressure difference in MPa; The stress sensitivity coefficient characterizes the degree of influence of mining-induced stress on the gas permeability of the sample, while the gas pressure sensitivity coefficient quantifies the degree of influence of gas pressure on the gas permeability of the sample. A greater change in these sensitivity coefficients indicates a steeper gradient of gas permeability change in response to the respective influencing factors. The gas permeability stress sensitivity coefficients are presented in Fig. 9a, while the gas permeability gas pressure sensitivity coefficients are shown in Fig. 9b. Variations in gas permeability sensitivity coefficients. As depicted in Fig. 9, the gas permeability stress sensitivity coefficient and gas pressure sensitivity coefficient in the APRZ of the protected layer exhibited smaller and more stable variations. In contrast, both coefficients in the RPRZ displayed larger variations. This observation indicated that the gas permeability of the sample in the RPRZ is significantly influenced by the stress state of the surrounding rock and the permeable gas pressure. This phenomenon can be attributed to the fact that the sample in the APRZ experienced decreasing axial pressure and increasing circumferential pressure, resulting in a gradual contraction of the effective seepage channels within the coal body. Consequently, variations in gas pressure had a minimal impact on the gas permeability of the sample. Conversely, the coal samples in the RPRZ were in a pressure-relief state that facilitates the development of numerous seepage channels. The increase in gas pressure promoted the expansion of the coal samples. The gas slippage effect was synchronously intensified due to the expansion of the seepage channels. Consequently, the stress sensitivity in the RPRZ increased correspondingly. The stress sensitivity coefficient and the gas pressure sensitivity coefficient of the RPRZ were both notably high. Penetration rate increase To further characterize the influence of the mining-induced stress field and gas pressure intensity on the gas permeability in the APRZ and RPRZ of the protected layer, the gas permeability enhancement rate was introduced based on the previous study67: where DKi is the relative gas permeability enhancement; Dri is the absolute gas permeability enhancement; K1 is the initial gas permeability of the sample in m2; Ki is the ith gas permeability of the sample in m2; Ki+1 is the ith + 1 gas permeability of the sample in m2. As illustrated in Figs. 10 and 11, the relative gas permeability in the RPRZ was larger than that in the APRZ under both constant and variable gas pressure conditions. Furthermore, the absolute gas permeability in the RPRZ gradually increased with increasing axial pressure-relief amplitude. This observation reinforces the explanation that an increase in the radial stress within the coal body leads to a gradual contraction of the “equivalent stable seepage channels” in the sample. Consequently, the “expansion phenomenon” of the sample was suppressed. The gas permeability of the unloaded gas in the protected layer demonstrated a strong correlation with the radial stress in the coal body. Variations in gas permeability enhancement rates under constant gas pressure conditions. Variations in gas permeability enhancement rates under variable gas pressure conditions. Engineering implications and future perspectives The present study elucidated the gas seepage mechanisms in deep coal seams under the coupled effects of mining-induced stress and gas pressure. The findings provide a theoretical basis for delineating gas accumulation zones, designing gas extraction boreholes, and enabling safe and efficient co-extraction of coal and gas. However, the stress paths imposed on the coal samples represent a simplified approximation of the complex conditions encountered in actual mining environments. This simplification limits the generalizability of the findings. Future research will prioritize the investigation of gas seepage behavior under more complex multi-field coupling conditions, including anisotropic stress paths, in situ gas pressures, gas adsorption/desorption effects, and temperature variations. These factors will remain primary directions for our continued research. Conclusions The axial deformation of the coal body and the degree of expansion of internal microscopic cracks were positively correlated with the intensity of the seepage gas pressure at the inlet. An increase in the seepage gas pressure at the inlet led to an increase in the axial strain and a decrease in the circumferential strain of the sample. The axial deformation of the coal body and the degree of expansion of internal microscopic cracks were positively correlated with the intensity of the seepage gas pressure at the inlet. An increase in the seepage gas pressure at the inlet led to an increase in the axial strain and a decrease in the circumferential strain of the sample. The radial stress in the protected coal seam and the in-situ gas pressure are the primary parameters affecting the gas permeability of the sample. An increase in the radial stress led to the gradual contraction of the “equivalent stable seepage channels” within the coal body until they completely closed. Conversely, an increase in the seepage gas pressure at the inlet diminished the contraction degree of some “equivalent stable seepage channels” and reduced the promoting effect of the “gas slippage effect” on the gas permeability of the coal body. The gas permeability of the coal sample in the RPRZ reached its maximum during the constant gas pressure stage. The radial stress in the protected coal seam and the in-situ gas pressure are the primary parameters affecting the gas permeability of the sample. An increase in the radial stress led to the gradual contraction of the “equivalent stable seepage channels” within the coal body until they completely closed. Conversely, an increase in the seepage gas pressure at the inlet diminished the contraction degree of some “equivalent stable seepage channels” and reduced the promoting effect of the “gas slippage effect” on the gas permeability of the coal body. The gas permeability of the coal sample in the RPRZ reached its maximum during the constant gas pressure stage. The observed increase in coal permeability and the circumferential expansion indicated the formation of a dominant seepage network within the coal mass. Consequently, the development and connectivity of these equivalent stable seepage channels critically controlled the gas permeability of the coal. Furthermore, variations in gas seepage pressure modulated the degree to which the gas slippage effect increased the measured permeability. This led to significant differences in its contribution under different pressure conditions. The observed increase in coal permeability and the circumferential expansion indicated the formation of a dominant seepage network within the coal mass. Consequently, the development and connectivity of these equivalent stable seepage channels critically controlled the gas permeability of the coal. Furthermore, variations in gas seepage pressure modulated the degree to which the gas slippage effect increased the measured permeability. This led to significant differences in its contribution under different pressure conditions. The experimental gas permeability measured in the APRZ and RPRZ of the protected layer, as well as the theoretical gas permeability derived from the proposed gas permeability model, were all less than 2%. This indicates that the fitting accuracy is satisfactory. The experimental gas permeability measured in the APRZ and RPRZ of the protected layer, as well as the theoretical gas permeability derived from the proposed gas permeability model, were all less than 2%. This indicates that the fitting accuracy is satisfactory. Acknowledgements The authors would like to acknowledge the discipline innovation team of Anhui University of Science and Technology. Author contributions K. W. Conceptualization, Funding acquisition, Investigation, Writing—original draft. X. M. Resources, Funding acquisition, Supervision. G. Z. Resources, Supervision. X. C. Formal analysis, Funding acquisition, Validation, Writing—review & editing. G. H. Data curation, Investigation. A. W. Data curation, Investigation. Funding This work was supported by the Open Fund Project of Key Laboratory of Provincial-Ministry Coal Mine Safety and Efficient Mining under the Ministry of Education (No. JYBSYS202407); National Natural Science Foundation of China (No. 52474090, No. 52374074); Outstanding Young Teachers’ Cultivation Program in 2024 (No. YQYB2024020). Data availability All data generated or analysed during this study are included in this published article. Declarations Competing interests The authors declare that they have no conflicts of interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement All data generated or analysed during this study are included in this published article.