Safe Mining Technology for Steeply Inclined Unstable Coal Seams

Steeply inclined thin coal seams [1], [2], [3] refer to seams with an inclination greater than 45° and a thickness of less than 1.3 m [4]. Although the total reserve of steeply inclined coal seams in China is not large, they are widely distributed, especially in the southern mining regions, where 80% of mines contain such seams. The complex geological structure of these seams, combined with the inclined layering of both the coal seam and its overlying strata, creates significant challenges for mining. The heterogeneity of the rock in the horizontal direction, the stress conditions of the strata, changes in the force transmission mechanism of the overlying rock after mining, and variations in mining depth along the dip direction all contribute to the increased difficulty of extracting steeply inclined coal seams [5].

Zhao [6] introduced the primary mining methods for steeply inclined coal seams and their applicability, which included the horizontal stratified mining method [6], [7], [8]. Cui et al. [9] and Zhang [10] explored the applicability of this method for steeply inclined coal seams, achieving safe, stable, efficient, and high-quality extraction of unstable coal seams. Lai et al. [11] described the mechanical model of rock pillars and roofs during steeply inclined mining, explaining the mechanical response behaviour of key disaster-prone zones in deep working faces under the influence of deep mining dynamics. By analysing the spatial correlation of a large amount of microseismic data, the time-space response relationship controlling coal-rock dynamic failure during the advancement of the steeply inclined working face was revealed. The movement of the rock strata is highly complex during the horizontal stratified mining of steeply inclined coal seams [12], [13], [14]. The working face can be affected by shock pressure caused by mining-induced seismic activity [15]. Cao et al. [16] conducted a case study using discrete element simulation and field investigation, studying the evolution of shock pressure from the perspectives of stress distribution, crack propagation, and coal body fragment ejection, thereby providing important references for understanding the mechanism of shock pressure-induced damage at the working face.

Taking the complex geological conditions of the Puxi Mine as an example, extensive and indepth research has been conducted on the tunnel layout of the mining area [17] and the mining technology [18]. The horizontal stratified mining method has been identified as the most suitable approach. Jia [19] explored the fundamentals and significance of rock control technology, emphasising its critical role. Therefore, a mechanical analysis of the stress conditions in the mining area was conducted in this study using FLAC3D software, with a numerical model established based on the mechanical parameters of the rock strata at the Puxi Mine. This model was used to assess the applicability and safety of applying the horizontal stratified mining method to steeply inclined and unstable coal seams [20], [21].

The current mining methods at the mine include the strike longwall mining method and the horizontal stratified mining method. The former is the primary mining method at the mine, accounting for more than 90% of the total production, while the latter contributes only 2.1–3.73% of the total production, being applied exclusively for the extraction of steeply inclined unstable coal seams. The horizontal stratified mining method involves dividing the steeply inclined coal seam into several layers parallel to the horizontal plane. The height of each layer is controlled within a specific range, with no coal left in the roof during mining, typically around 2 m. The geological structure of the Puxi Mine is complex, with a series of synclines and anticlines forming the strike and dip directions. The coal seam exhibits a wavy pattern. The area for steeply inclined mining is small, and the coal seam is unstable and complex, preventing the formation of a conventional wall face. As a result, mining is challenging, and it is economically unfeasible. However, considering regional outburst prevention measures, Jiahe Company has decided to extract this portion of the coal seam. Based on the conditions of the coal seam, the steeply inclined coal seam is mined first. After the completion of the steeply inclined mining, the resulting horizontal stratified tunnels will be used as the return airways for the longwall mining working face.

The stress environment in steeply inclined coal seams differs significantly from that in near-horizontal coal seams. Due to the influence of horizontal geological stress and geological structure, the thickness of the coal seam in steeply inclined layers varies considerably, and even the coal quality can change.

The gravity of the strata overlying the steeply inclined coal seam is a significant source of stress. As shown in Figure 1, the self-weight G of any particle in the overlying strata of the steeply inclined coal seam can be decomposed into two components: A vertical force $F_1$, perpendicular to the plane of the coal seam, and a force $F_2$, along the dip direction. As the dip angle of the coal seam increases, the force $F_2$ increases, while $F_1$ decreases. Consequently, the tunnel experiences greater lateral pressure. The displacement and collapse of the roof and floor strata generally exhibit a downward trend.

$F_1=G \cos \alpha \quad F_2=G \sin \alpha$

In the equation, $\alpha$ represents the dip angle of the coal seam.

The horizontal stratified mining method typically features a smaller length of the stratified level tunnel or mining face. The roof and floor strata of the coal seam act as the two sides of the roadway, with the overlying strata composed of collapsed loose gangue [22]. The load conditions on the support in the mining face are shown in Figure 2.

a) Calculation method for the stratified level tunnel or mining area and the mining area support load

The basic assumptions and conditions are as follows:

The gangue above the stratified level tunnel or mining area is treated as a loose medium.

Stability condition for the tunnel roof: No relative movement occurs between the gangue particles.

The dip angle of the coal seam is 90°.

b) Stress limit equilibrium theory for loose media

Under the conditions outlined above, the stress limit equilibrium theory for loose media was introduced to calculate the support load in the stratified level tunnel or mining area. The stress limit equilibrium state for loose materials refers to a state of equilibrium in which the initial shear stress and internal friction within the entire material or a specific region of it are just overcome. The appearance of this stress state causes the movement of the loose material.

At any point in the loose medium, suppose there is an arbitrary infinitesimal plane passing through this point with a normal. On this plane, normal and tangential stress components, denoted as $\sigma_n$ and $\tau_n$, act. In the case of equilibrium failure, the shear strength of the loosely bound medium along this plane follows a linear relationship.

where, $c$ is the cohesion of the loose medium.

If the following basic conditions are satisfied at any point in the loose medium:

Then the material will not undergo sliding.

For convenience, the stress limit equilibrium state of the loose medium is represented using the Mohr stress circle. The graphical method was then applied to determine the parameters that constitute the equilibrium conditions.

Figure 3 shows the Mohr stress circle. In order to simplify the problem, the envelope of all the curves was approximated by the straight line AD. From the figure, it can be observed that the shear strength line AD represents the limiting equilibrium equation.

The region above the line AD corresponds to the sliding zone, while the region below it represents the equilibrium zone. The angle of friction $\phi$ is the angle between the line and the horizontal axis. The initial shear strength $c$ is represented by the line segment OA.

Further analysis of Figure 3 reveals that the sine of the maximum inclination angle can be expressed as:

and

Thus,

From the figure, it is clear that the maximum inclination angle equals the internal friction angle $\phi$. Therefore, when $\sin \theta_{\max }=\sin \phi$, another stress limit equilibrium equation can be derived:

c) Lateral pressure coefficient under the stress limit state

The lateral pressure and pressure coefficient under the limit state were calculated using the relationship between the principal stresses in the loose medium’s limit equilibrium state.

By substituting $\sigma_1$ and $\sigma_3$ in Eq. (3) with $\sigma_z$ and $\sigma_x$ and letting $\sigma_c=c \cot \phi$, the following equation can be obtained:

Solving for $\sigma_x$ in this equation yields:

By rewriting $\sigma_z$ and the coefficient containing the internal friction angle $\phi$ following 2$c$ in Eq. (5) and simplifying, the expression becomes:

The coefficient following 2$c$ can be expressed as:

The substitution of Eqs. (6) and (7) into Eq. (5) results in:

For the loose medium with cohesion $c=0$, the following expressions apply:

Thus, the pressure coefficient under the loose medium’s limit equilibrium condition is given by:

d) Calculation of vertical pressure at the bottom of loose medium

In order to calculate the pressure at the bottom of a loose medium container, the Yang-Xin formula is still widely used to calculate the vertical pressure on a unit area.

where, $P_z$ represents the vertical pressure on the unit area of the bottom plate of a container with a filling depth of $Z$; $K_w$ is the friction coefficient between the bulk material and the container walls; $S$ is the horizontal area of the bulk material; and $l$ is the perimeter of the container's horizontal cross-section.

e) Calculation of stratified tunnel or mining area and support loads

Based on rock mechanics test parameters and combined with field experience data, the following parameter values were adopted:

For a maximum stratified tunnel or mining area height of 15 m and a maximum roof distance of 3.2 m, the calculations are as follows:

$\phi=25^{\circ} ; \quad \gamma_k=25 \mathrm{kN} / \mathrm{m}^3 ; \quad l=10.4 \mathrm{~m} ; \quad S=6.4 \mathrm{~m}^2 ; \quad K_w=0.15 ; \mathrm{and} \,\, Z=15 \mathrm{~m}$

Therefore, the following values are obtained:

$K_c=0.4058, \mathrm{and}\,\, P_z=195 \mathrm{kN} / \mathrm{m}^2$

For the unit hydraulic support in the mining area and stratified tunnel, with a canopy spacing of 0.8 m, column spacing of 0.5 m, and support density of 2.5 supports per square metre, the supporting force per pillar was calculated to be 78.0 kN. This calculation indicates that the force acting on the existing single pillar is significantly below its rated value, ensuring that the strength requirements for support are met.

Research on rock layer control for steeply inclined coal seam mining is limited, both domestically and internationally. The study of rock layer control in the caving zone for the tunnel horizontal stratified mining of steeply inclined coal seams is, however, a novel area of research [23], [24]. The control of the caving zone rock layers is crucial to fully exploit the mining-induced pressure for breaking the roof coal, while also protecting the horizontally stratified tunnel.

When horizontal stratified mining is conducted, and the roof layer is not sufficiently filled with fallen gangue, if the coal seam roof and floor are relatively hard, the roof generally will not cave in until the fractured limit of the exposed roof layer is reached [24], [25], [26]. According to mechanical theory, the ultimate height for roof failure in the caving zone, when the roof is unsupported, is calculated as follows:

where, $A$ is the thickness of the immediate roof's first layer (m); $\beta$ is the angle of coal seam dip (degrees); $B$ is the thickness of the immediate roof (m); $\sigma_p$ is the ultimate tensile strength of the immediate roof (MPa); and $\gamma$ is the bulk density of the roof rock layer (kg/m$^3$).

The caved gangue in the caving zone, due to the effect of its own weight, generally does not remain in place but immediately slides downward. The necessary condition for sliding to occur is $f<$tga, where $f$ represents the dynamic friction coefficient of the fallen gangue along the coal seam floor and a is the coal seam dip angle. For mining areas, f=0.65 and $a$=60$\sim$900, which evidently satisfy this condition. However, the sliding of the fallen gangue does not solely depend on the dip angle, but also on the ratio of the thickness of the rock layer prior to falling, $h$, to the coal seam thickness, m. Studies have shown that when $h/m>$0.4$\sim$0.5, as shown in subgraph (a) of Figure 4, after the rock falls to the floor, it does not break into small pieces or roll but piles up or remains close to its original state. When $h/m<0.4\sim$0.5, as shown in subgraph (b) of Figure 4, due to the large fall height, the rock fragments are more likely to break into smaller pieces, which are arranged chaotically and irregularly, and slide violently downward. This results in the filling of the lower part of the caving zone, preventing further collapse of the roof rock in that region. Meanwhile, as the free height in the upper part of the caving zone increases, the fall height may also increase (Figure 5), thereby enhancing the filling effect on the lower part [20], [21], [22], [23], [24], [25], [26].

It is assumed that the shattered gangue and loose coal behave as elastic bodies, and thus, the roof can be regarded as a beam supported on an elastic foundation. Accordingly, a mechanical model of the mining area after the roof collapse was established, as shown in Figure 6.

Based on the theory of an elastic foundation beam, for the region with $x>0$, the differential equation of the roof deflection curve is given by:

where, $E$ and $I$ represent the elastic modulus and the moment of inertia of the roof, respectively; $q_0$ is the uniform load on the roof, with $q_0=H \gamma \cos \alpha$; and $k_1$ is the Winkler foundation coefficient for the real coal body.

The deflection curve equation for the roof is given by:

where, $C_1$ and $C_1$ are integration constants; and $\alpha$ is the resistance coefficient, with $\alpha=\sqrt[ 4]{k_1 / 4 E I}$.

For the region $x<0$, i.e., on the side of the loose roof coal, the differential equation of the roof deflection curve is:

where, $k_2$ is the Winkler foundation coefficient for the loose coal body; and $y_0$ is the roof deflection before the collapse of the top coal.

The deflection curve equation for the roof is then:

where, $D_1$ and $D_2$ are integration constants; and $\beta$ is the resistance coefficient, with $\beta=\sqrt[ 4]{k_2 / E I}$.

When $x$=0, Eqs. (22) and (23) lead to:

At $x$=0, the roof deflection curves on both sides smoothly connect, and the corresponding derivatives of the curves at this point are equal. This condition allows for the calculation of the in-tegration constants.

Solving Eq. (25) yields the following values:

$C_1=\frac{\beta^2(\beta-\alpha) q_0}{\alpha^2(\beta+\alpha) k_2}, \quad C_2=\frac{\beta^2 q_0}{\alpha^2 k_2}, \quad D_1=\frac{(\alpha-\beta) q_0}{(\alpha+\beta) k_2}, \quad D_2=-\frac{q_0}{k_2}, \quad y_0=\frac{q_0}{k_1}+\frac{\beta^2 q_0}{\alpha^2 k_2}$

Substituting these integration constants into Eq. (21) and simplifying results in Eq. (26). Multiplying both sides of Eq. (26) by $k$ gives the pressure acting on the real coal body roof:

From Eq. (27), it can be observed that when $x$=0, the maximum value of $p_m$ occurs:

When $x \rightarrow \infty$, then $p_m=q_0$.

Substituting $D_1$ and $D_2$ into Eq. (28), both sides are multiplied by $k_2$ and simplified, yielding the pressure acting on the loose coal seam as:

When $x$=0, the minimum value of $p^d$ obtained is $p_{\text {min }}^{d_0}$.

When $x \rightarrow-\infty$, then $p^d=q_0$.

From the above analysis, it can be deduced that at $x$=0, the pressure acting on the loose coal body is zero, as shown in Figure 7. As the various sections of the roof collapse and the collapse process is completed, the space above the tunnel becomes filled with debris, forming a filling zone that resupports the roof and plays a role in supporting the sides of the tunnel. At this point, the tunnel is in a reduced pressure or pressure-free state.

The following calculations were carried out in accordance with the problems that need to be addressed in this study:

a) Stability of the surrounding rock in the mining area after the upper stratigraphic excavation and support.

b) Stability of the surrounding rock in the mining area during the lower stratigraphic excavation and support process.

The parameters obtained from rock mechanics experiments are the mechanical properties of the rock blocks, while the mechanical parameters required for numerical simulation are those of the rock mass. Therefore, the mechanical parameters derived from the rock mechanics experiments need to be processed. The specific values are shown in Table 1.

In order to minimise the impact of boundary constraints on the simulation results and to meet the requirements of the study, the width of the computational model was set to 50 m, with the model height adjusted according to the specific simulation needs. The model employs stress boundary conditions. A uniform vertical compressive stress was applied to the upper boundary of the model, considering the self-weight of the overlying rock mass, denoted by , where is the average volumetric force of the overlying rock layer, and is the distance from the model's upper boundary to the surface. A horizontally varying compressive stress, depending on the depth, was applied to both sides of the model. The vertical displacement at the lower boundary was fixed, while horizontal displacements were constrained on the left and right boundaries, as shown in Figure 8.

The coal seam was modelled using the Mohr-Coulomb plasticity model. The roof and floor rock layers of the coal seam were also modelled using the Mohr-Coulomb plasticity model. Excavation was modelled using the null model, with beam elements used to simulate individual props and $\pi$ beams, while liner elements were used to simulate the stratified timber supports.

Figure 9 shows the displacement of the tunnel roof and floor, as well as the two sidewalls, after the upper layer mining under the current mining conditions. The results indicate that if the upper layer mining is not immediately supported, significant subsidence of the roof occurs, which severely affects the normal working space. Figure 9, Figure 10, Figure 11 to Figure 12 present various distributions under the condition of no support after upper layer mining, including the displacement distribution of the roof and floor, the plastic zone distribution of the roof and floor, the displacement distribution of the sidewalls, and the stress distribution of the sidewalls in the mining area.

After the upper layer mining, without support, the stress environment around the working face was altered due to the removal of coal, causing a transition from a three-dimensional stress state to a two-dimensional stress state. This leads to the subsidence of the roof above the working face. Figure 10 illustrates the significant deformation of the roof, while a large plastic zone appears in the coal layer of the roof. Additionally, a certain depth of plastic zone was formed in the floor. Figure 11 and Figure 12 demonstrate that stress concentrations occur both vertically and horizontally in the surrounding rock, but due to the high strength of the sidewalls, no significant deformation is observed in these areas.

If timely support is implemented after upper layer mining, as shown in the support structure diagram (Figure 13), a significant reduction in the amount of roof subsidence is observed. After the application of single pillar and $\pi$ beam support structures, From Figure 14, Figure 15, Figure 16 and Figure 17, the subsidence decreased substantially from 72 cm, which occurred without support, to just 5 cm. Furthermore, the plastic zone distribution on the coal roof virtually disappeared, with only a very shallow plastic zone remaining in the tunnel floor. The roof remained largely intact.

As shown in Figure 18 and Figure 19, Upon completion of the upper layer excavation, the lower layer was then mined, maintaining three rows of connected pillars between the upper and lower layers, thus forming a ventilation system with a forward and return airflow pattern. In terms of plastic zone distribution, a significant plastic zone was observed in the roof after removing the upper layer support, leading to roof collapse in this area. The collapsed roof material accumulated on the support structures in the lower layer, forming an artificial false roof, which ensures the stability of the lower layer support system. Additionally, a relatively deep plastic zone was also observed in the unmined coal body of the lower layer, which facilitates the subsequent mining of the lower layer.

Upon commencement of lower layer excavation, Figure 20 and Figure 21 demonstrate the presence of stress reduction zones around the mining area in both the upper and lower layers. This phenomenon facilitates the support process, as only a relatively small supporting resistance is required to maintain the stability of the mining area.

The displacement distribution around the mining area after lower layer mining indicates significant displacement in the roof after the removal of support in the upper layer. Additionally, at the step between the upper and lower layers, substantial axial deformation of the coal body occurs. This necessitates careful attention to the stability of the lower layer excavation face during mining to prevent sidewall collapse incidents.

A "top-down" problem exists at the junction of the upper and lower layers' support structures. As shown in Figure 22 and Figure 23, the support beams are located between the roof and floor of the coal seam. During mining, the stress in the roof and floor decreases due to coal excavation, thus maintaining relative stability. Consequently, the support beams remain stable due to the two fixed sections. In the direction of the face advance, the crossbeam is partly located on the floor of the upper layer and partly on the roof of the lower layer, forming a cantilever beam structure. Therefore, collapse is prevented in the direction of face advance. Despite the high height of the support structure at the junction of the upper and lower layers, it remains in a stable state.

In the stress environment of the mining area using the horizontal stratified mining method, in addition to bearing the self-weight stress, significant horizontal ground stress, and gas pressure from the overlying strata, concentrated stress is also experienced along the inclined and strike directions of the mining area. The most significant manifestation of mining-induced pressure occurs during the prop-drawing caving process. The dynamic pressure acts on the lower-layer horizontal tunnels. However, field production practice shows that the impact of dynamic pressure on the tunnels is relatively weak. Therefore, the rational determination of tunnel support structures is crucial to maintaining the stability of the horizontal stratified tunnels, reducing maintenance costs, and improving mining efficiency.

When mining is conducted using this method, the stability of the tunnels is generally good as the two sides of the mining area are composed of rock. During the extraction of the lower layer, the collapse of the roof in the upper layer becomes a significant factor affecting the safety of the lower-layer support structure, necessitating enhanced on-site support management. The mining face of the lower layer experiences some deformation, and measures must be taken to prevent sidewall collapse. Thus, the use of the horizontal stratified mining method at the Puxi mine for the extraction of its steeply inclined unstable coal seams is deemed feasible.