When miners carve a roadway out of a coal seam that is not lying flat, the rock around the opening does not respond symmetrically. A new numerical study published in GeoEnergy Communications shows that the way damage spreads around such an excavation depends on a three-way interplay between the in-situ stress ratio, the dip angle of the coal–mudstone layers, and the strength of hydromechanical coupling between stress and fluid pressure in the rock. The work, led by Chenxi Zhao and Fuqiang Gao together with Qinghua Lei, Shuangyong Dong, Xiangyuan Peng and Rui Wu, was published on 11 September 2026 as an open-access Research article.
The team built a two-dimensional, fully coupled hydromechanical finite element model in COMSOL representing a vertical section of water-saturated coal–mudstone strata containing a network of natural fractures. The square simulation domain measures 25 metres on each side, with a 4-metre-long unexcavated roadway positioned at its centre. The coal seam is 6 metres thick and its dip angle was varied from 0 to 45 degrees, while the rest of the domain is mudstone. The intact rock is described by a Mohr–Coulomb plastic constitutive law with a tension cutoff, and the displacement of pre-existing fractures is governed by nonlinear stress–displacement relationships in both normal and shear directions. Fluid migration through the fractured porous medium follows Darcy’s law, with hydromechanical coupling capturing stress redistribution, pore pressure variation, plasticity development and fracture deformation during excavation.
The fracture network represents the face and butt cleats that characterise coal. These fracture sets are generally orthogonal to each other, with face cleats parallel to the bedding planes and butt cleats perpendicular to them, and they form as the coal matrix shrinks during maturation. In the model, cleat lengths follow a power law distribution with an exponent of 2.5 and a fracture intensity of 3 per metre, with the number ratio of face cleats to butt cleats set to 3:1. Material properties for the coal, mudstone and fractures were drawn from laboratory testing at Zhangcun coal mine, and the coal and mudstone matrices were assigned identical, extremely low hydraulic permeabilities, meaning that fluid transport is controlled predominantly by the fracture and cleat networks rather than by flow through the rock matrix itself.
The simulations were run in three stages. First, in-situ stress and pore pressure were ramped up to establish equilibrium, with a constant pore pressure of 1 megapascal imposed on all outer boundaries and a vertical stress of 5 megapascals at the top boundary. Second, excavation was simulated by gradually reducing the modulus of coal within the tunnel and decreasing the pore pressure at the tunnel boundary over 10 hours, using a time step of 180 seconds. Third, a 10-hour post-excavation drainage phase captured subsequent hydromechanical responses. The maximum horizontal stress was set to 5, 10 or 15 megapascals, corresponding to stress ratios of 1, 2 or 3, with the first two cases reflecting typical far-field stress ratios observed in coal mining and the third accounting for near-field stress concentration effects induced by longwall mining.
The results show that the in-situ stress ratio exerts a primary control over the magnitude and extent of excavation-induced responses. Higher stress ratios intensified plastic yielding, shear stress concentration and fracture slip. During excavation, the pore pressure field developed zones of both pressurisation and depressurisation, arising from drainage at the roadway boundary and the poroelastic response of the surrounding low-permeability matrix. The most pronounced pore pressure disturbance occurred at the end of the excavation, after which diffusion dominated and the pressurised zone gradually shrank. Stress concentration proved more pronounced in the surrounding mudstone than in the coal seam, owing to the higher stiffness of mudstone, and compressive zones developed above and below the roadway while tensile zones emerged along its lateral boundaries.
Stratigraphic dip strongly shaped the failure pattern. Under isotropic loading, plastic zones consistently emerged at the four corners of the roadway regardless of dip angle, but as the dip angle increased, the failure pattern shifted toward the upper-right and lower-left corners, aligned with the face cleat orientation. In inclined strata, triangular-shaped plastic regions developed along the coal–mudstone interfaces, indicating that cleat slip promotes shear-dominated failure processes. The authors describe this as a progressive transition of the disturbance behaviour from mechanically dominated to structurally controlled regimes. Notably, the steepest dip angle of 45 degrees did not always produce the greatest fracture sliding: moderate dip angles of 15 and 30 degrees showed more pronounced slippage zones, because steeply inclined cleats experience higher normal compression, which suppresses shear displacement even as shear stress rises.
Two macroscopic indicators tracked the disturbance over time: the plastic area, summing the areas of elements in a plastic state, and the fracture slippage area, summing shear displacement weighted by fracture length. Both increased systematically with higher stress ratio. The most substantial growth in plastic area occurred towards the end of excavation, but the drainage stage continued to contribute to plasticity development at a slower, stable rate, indicating that depressurisation near the roadway progressively destabilises the surrounding rock mass even after the main excavation disturbance has ceased. Fracture slippage behaved differently: it initiated earlier than plasticity, becoming active during excavation, but increased only minimally during drainage, and in one case, at a dip angle of 30 degrees under a stress ratio of 3, the slippage area even decreased slightly, attributed to drainage-induced normal compaction suppressing further shear movement.
To isolate the role of hydromechanical coupling, the team performed additional simulations with different matrix Biot coefficients, ranging from zero coupling to full coupling, for the case with a stress ratio of 3 and a dip angle of 30 degrees. Larger Biot coefficients produced greater magnitudes and broader regions of pressurisation, higher maximum principal stresses, and more extensive cleat shear slippage and plastic failure. In models with higher coupling, high-pressure zones coincided with regions of substantial fracture displacement, consistent with the concept of pressurisation-induced fracture slip, an interpretation the authors note aligns with field measurements from the FEBEX tunnel at Grimsel in Switzerland. Mohr circle analysis showed that as the Biot coefficient increased, stress states moved closer to, or even intersected, the yield envelopes of coal and mudstone, explaining the greater fracture slippage and plastic deformation in those simulations.
The study also found that plastic zones can act as hydraulic bridges, coalescing with natural fractures to form connected flow pathways to the roadway, thereby enlarging the potential leakage zone for groundwater or methane migration. Because these plastic zones tend to develop along the dip direction of inclined strata, the connected network is strongly influenced by the structural anisotropy of the coalbed. The authors suggest that this supports customising roadway support, concentrating reinforcement at roof and floor corners and along coal–mudstone interfaces, and prioritising grouting and sealing of plastic damage zones that propagate along inclined lithological interfaces. They note that the predicted failure patterns qualitatively mirror observations such as asymmetric deformation recorded at the Wudong Coal Mine and in deep mines in Huainan City, while stressing that quantitative comparison with borehole imaging, convergence records, piezometer readings and microseismic data remains future work.
The authors are careful to state the limitations of the modelling. The simulations rely on a two-dimensional plane-strain assumption and therefore do not capture three-dimensional phenomena such as longitudinal stress redistribution driven by the advancing excavation face. Damage, slippage and hydraulic connectivity metrics were derived from a single representative fracture network, and the authors recommend that future research evaluate the stochastic variability of fracture geometry. They also caution that the normalized trends they report provide a preliminary basis for assessing mining under complex hydrogeological conditions, and that a broader dataset would be required before establishing a more general empirical correlation. Even so, the overarching qualitative trends, driven by the relative alignments of the stress field, bedding, cleat networks and excavation geometry, offer a mechanistic picture of why roadways in dipping, fractured coal seams fail asymmetrically, and why pore pressure transients in low-permeability rock deserve a central place in excavation design.
Subject of Research: Numerical simulation of coupled hydromechanical responses around roadway excavation in inclined coal–mudstone strata
Article Title: Coupled hydromechanical responses around a roadway excavation in inclined coal–mudstone strata
Article References: Zhao, C., Gao, F., Lei, Q., Dong, S., Peng, X., & Wu, R. (2026). Coupled hydromechanical responses around a roadway excavation in inclined coal–mudstone strata. GeoEnergy Communications, 2(1), Article 13. https://doi.org/10.1007/s44421-026-00026-5
Image Credits: AI Generated
DOI: 10.1007/s44421-026-00026-5
Keywords: roadway excavation, coal mining, hydromechanical coupling, pore pressure, fracture slip, plastic zone, coal cleats, mudstone, inclined strata, Biot coefficient, numerical modelling, excavation damaged zone
News Source: Denise Maddox. (October 28, 2026). Simulations reveal how dipping coal seams and pore pressure swings shape roadway damage. Scienmag.
