Analysis of Secondary Fracture Law of Roof Strata and Water Inrush Potential in Close-Distance Coal Seam Mining

Abstract

Close-distance multi-seam mining frequently induces secondary surface deformation and subsidence. Extracting a lower coal seam beneath an existing goaf repeatedly disturbs the overburden, often leading to roof collapse and the expansion of vertical water-conducting fractures that connect the working face to aquifers. Furthermore, the overlying goaf increases the risk of water inrush into active lower workings. This study investigates the mechanisms of strata reactivation and fracturing within an overlying goaf during lower seam extraction at a mine in Northwest China. Using theoretical analysis, numerical simulation, and microseismic monitoring, the research examines the secondary fracture mechanisms of the goaf roof and the resulting water-inrush potential. Research Findings: Strata Instability: Analysis of the key sandstone strata indicates that subsidence (W) of the key rock blocks satisfies 3.17 < W₁ = 4.61 m < 18 m for the lower seam and 3.17 m < W₂ = 5.31 m < 69.6 m for the 3-1# seam. These values confirm that key rock blocks in the basic roof undergo “reactivated” instability following fracture during lower seam mining. Pressure Relief and Fluid Dynamics: Mining-induced fracture initiation and propagation trigger strata reactivation. As the distance to the center of the goaf decreases, the subsidence of the overburden increases, ultimately resulting in a “trapezoidal” bending deformation pattern. Due to secondary activation, the roof subsidence 30 m above the 221 coal seam increased from 1.89 m to 5.475 m. The layers of high-strength, medium-grained sandstone and siltstone overlying the 317 coal seam and beneath the 221 goaf serve as high-strength material for the overlying rock formations. This suppresses the development of the caving zone and fracture zone, leading to subsidence failing to reach the sum of the heights of the two coal seams (6.8 m) and only reaching a value of 5.475 m. During extraction, the stress field undergoes a distinct evolution: it transitions from an initial “regular triangular” pressure-relief zone into a tripartite “weak–strong–strong” distribution. Furthermore, fluid discharge in the overlapping zone between the 317 working face and the 221 goaf increased sequentially, displaying an “alternating” pattern of peak vector variations as the face advanced. Microseismic Activity: Monitoring within the 300–500 m range identified frequent low-energy events and high-magnitude events (10⁴ J, 10⁵ J). These findings demonstrate that secondary excavation directly impacts the aquifer, creating a significant water-inrush hazard for the active working face.

1. Introduction

China’s coal reserves are characterized by a high proportion of closely spaced coal seams [1]. These fractured rock masses often contain significant groundwater. As extraction efficiency and socioeconomic demand grow, mining multi-seam configurations have become an unavoidable engineering challenge [2]. In this context, water-inrush hazards [3] increasingly threaten production safety [4], introducing complex technical problems. Fundamentally, extracting a lower seam disturbs the overlying goaf. Pressure relief on the goaf roof expands fracture zones [5] and triggers secondary deformation. Due to the minimal spacing between seams, stress concentrates around residual coal pillars from previous extraction [6]. This alters the stress environment of the lower mining area, potentially inducing abnormal roof collapse and creating vertically connected water-conducting fractures between the working face and aquifers [7]. Furthermore, water stored within these fractured rocks exerts additional loads through fluid flow forces and self-weight, jeopardizing structural stability [8]. These secondary water-inrush hazards are often delayed and subtle [9], complicating prediction and increasing the risk of disasters [10]. Consequently, investigating seepage behavior in fractured rock masses is now a critical research priority for preventing accidents in closely spaced seam groups. Research has primarily focused on the stability of surrounding rock during multi-seam extraction. Christopher Mark et al. [11] employed LaM2D to estimate multiple seam stresses, ALPS and ARMPS to determine pillar stability factors, and the CMRR to measure roof quality. Similarly, Miao [6] integrated analytical methods, numerical simulations, and field investigations at the Hongliulin Coal Mine. They developed an elastic mechanics model to estimate roof instability depth—adjusting for mining height—and used 3DEC modeling to simulate overburden deformation. Behrooz Ghabraie et al. [12] studied significant differences between single- and multi-seam subsidence parameters in a multi-seam case study in Australia. Using 3D discrete element modeling, Wang et al. [13] systematically investigated fracture propagation and surface subsidence during staggered overburden extraction. Additionally, Yao et al. [14] identified fracture propagation mechanisms in collapsed overburden and developed a damage constitutive model that accounts for plastic strain and porosity. Finally, Zou et al. [15] demonstrated that multi-seam mining transforms surface subsidence from a broad “basin” shape into a single-inflection “funnel” shape, emphasizing that early extraction significantly influences subsequent mining stages. Studies on rock permeability have further clarified these risks. Milsch et al. [16] distinguished how brittle and elastic fractures respond to pressure, while Wang et al. [17] used particle motion visualization to study mass loss in porous rocks caused by seepage. Ma et al. [18] addressed permeability anisotropy by combining acoustic emission data with microseismic monitoring to map fracture development. Yang et al. [19] proposed a coupled fault–stress–damage model to simulate groundwater flow during mining. Since most water-inrush incidents are triggered by faults, groundwater often breaches the mine floor via these weak zones under hydraulic pressure. Beyond faults, microscopic joints and fractures significantly influence seepage [20]. Faulted zones frequently generate structural coal bands that, due to their connectivity and water storage capacity, behave like joint fractures [21]. These fractures serve as the primary conduits for aquifers [22], making their study essential for understanding water-inrush mechanisms.

While researchers have explored activation mechanisms and permeability in porous rocks, most studies focus on the dynamic properties of intact rock. In real-world mining, activities inevitably create damaged zones, such as goafs and caving areas. Fluid migration must traverse both intact and damaged rock masses, where water-inrush hazards are most likely to manifest. The specific coupling mechanism governing fluid flow through rock failure zones remains insufficiently explored. Therefore, this study aims to address the following research questions: (1) What are the mechanical criteria for the “activation” and instability of roof strata in close-distance coal seam mining? (2) How do rock failure zones evolve incrementally to form water-conducting pathways during the extraction of the lower working face, and what are the dynamic response patterns induced by coal seam mining? What is the specific mechanism of fluid migration through rock failure zones? (3) Can microseismic monitoring accurately and effectively predict the occurrence of such rock failure zones? To address these questions, a comprehensive methodology integrating theoretical analysis, 3DEC numerical simulation, and field validation is employed.

2. Overview of Mining

As shown in Figure 1, the study is conducted at a coal mine in Northwest China with a designed production capacity of 8.0 Mt/a. The primary extraction target is the 3-1 coal seam, which is characterized by a gentle inclination of 1–3°, an average burial depth of 698.2 m, and a mean thickness of 4.9 m. The 317 working face, situated in the No. 11 mining district, has a length of 320 m and a total advance length of 2919 m. It is extracted using a longwall fully mechanized method with full-seam height extraction in a single pass. This face is overlain by the previously exhausted 221 goaf (2-2# coal seam), which sits at an average vertical distance of 29 m above it. The 221 goaf begins 67.9 m from the 317 open-off cut, and its mining direction is perpendicular to that of the 317 working face. The 2-2# middle coal seam varies in thickness from 0.84 to 2.6 m (average 2.1 m). Crucially, an aquifer is located 49.4 m vertically above the 2-2# coal seam, posing a potential water-inrush hazard during the secondary disturbance caused by mining the 3-1 seam. The layout of the working faces is illustrated in Figure 2 and Figure 3. Figure 2 presents a 3D model diagram, which better illustrates the positional relationships. The stratigraphic parameters from a nearby borehole are illustrated in Figure 4. This flowchart clearly outlines the three main pillars of our study: Theoretical Analysis, Numerical Simulation (3DEC), and Field Verification, as shown in Figure 5.

3. Mechanical Criteria for Roof “Activation” and Instability in Close-Spaced Coal Seam Mining

3.1. Two-Zone Calculation

Research indicates that the maximum heights of the caving zone and the water-conducting fracture zone (collectively referred to as the “two zones”) are primarily governed by lithological properties and the coal seam dip angle. Given the overburden conditions of the 317 working face, the roof strata are predominantly sandstone, based on geomechanical strength characteristics.

The overburden of the 3-1# coal seam is classified as a “medium-hard” roof. To ensure the monitoring equipment range covers the full extent of the disturbed strata, preliminary calculations were performed based on medium-hard rock parameters. The immediate roof of the 3-1 coal seam consists mainly of sandstone with an average compressive strength of 37.2 MPa and an average tensile strength of 3.45 MPa. For medium-hard rock strata, the maximum heights of the caving zone and the water-conducting fracture zone are calculated using Equations (1) and (2) [23,24]:

Collapse zone height:

$$H_{\mathrm{m}}={\displaystyle \frac{100\sum M}{4.7\sum M + 19}} \pm 2.2$$

(1)

Fissure zone height:

$$H_{\mathrm{li}}={\displaystyle \frac{100\sum M}{1.6\sum M + 3.6 }} \pm 5.6$$

(2)

In the formula: ∑M—Cumulative thickness, (m); H_m—Collapse zone height; H_li—Fissure zone height.

Based on the empirical formulas for medium-hard rock strata, the predicted heights for the caving and water-conducting fracture zones are as follows: Using a maximum mining thickness of 3.4 m, the calculated heights are: Caving Zone Height 6.76~10.87 m; Water-conducting Fracture Zone Height 16.52~28.37 m. Using a maximum mining thickness of 5.8 m, the calculated heights increase significantly: Caving Zone Height: 15.97~21.37 m; Water-conducting Fracture Zone Height: 58.52~77.72 m.

3.2. Mechanical Identification of Roof Reactivation in the 221 Goaf

Previous research indicates that the fracture position of the main roof occurs at the location of peak abutment pressure inside the coal rib. Based on the limit equilibrium analysis of the yield zone [25], the distance x₀ from the coal wall to the peak stress point can be derived as:

$$x_0={\displaystyle \frac{\lambda m}{2\mathrm{t}\mathrm{a}\mathrm{n}{\phi }_0}}\mathrm{l}\mathrm{n}\left({\displaystyle \frac{k\gamma H+\frac{c_0}{\mathrm{t}\mathrm{a}\mathrm{n}{\phi }_0}}{\frac{c_0}{\mathrm{t}\mathrm{a}\mathrm{n}{\phi }_0}+\frac{p_0}{\lambda }}}\right)$$

(3)

In the formula: λ is the lateral pressure coefficient, taken as 1.8; m is the mining height of the 221 working face, taken as 2 m; φ₀ is the internal friction angle of the coal seam, taken as 30°; k is the stress concentration coefficient, taken as 3; H is the roadway burial depth, 492.9 m; γ is the average unit weight of the overlying strata, 0.026 MN/m³; c₀ is the cohesion of the coal seam, 5 MPa; and p₀ is the support resistance of the roadway sides, with anchor bolt support taken as 0.2 MPa.

By substituting the mechanical parameters of the 2-2# coal seam and its roof strata into the analytical model, the fracture distance is determined to be x₀ = 4.27 m. This indicates that following the extraction of the 317 working face, the lateral key rock block B fractures approximately 10 m from the small coal pillar. Owing to the significant span of key block B, key block C enters a suspended state, whereas block D maintains contact with the gob (caved zone). The resulting structural configuration and the mechanical model of the overlying roof along the goaf side of the 317 working face are characterized by these segmented interactions, as shown in Figure 6.

After the immediate roof collapses, the gap between it and the main roof $W_{n1}$ [26] is:

$$W_{n1} = m - {\displaystyle \frac{1}{2}}\sum \mathrm{h}(k_p-1)$$

(4)

In the formula: m is the mining height of the 317 working face, taken as 4.9 m; $\sum \mathrm{h}$ is the thickness of the immediate roof, taken as 4.62 m; kₚ is the bulking coefficient, taken as 1.3.

Substituting the relevant parameters into the analytical model yields a subsidence value of W_n1 = 4.72 m. Consequently, following the collapse of the immediate roof, a gap of at least 4 m persists beneath the hinge point of key rock blocks B and C along the goaf side. During the extraction of the 317 working face, as the exposed area of the main roof reaches its span limit, key rock block B undergoes significant rotation. This rotation intensifies the vertical stress acting on the coal pillars. Once these pillars reach their ultimate load-bearing capacity, the fractured overburden within the goaf begins to support the residual load. When this secondary loading reaches a critical threshold, structural instability occurs. The resulting impact disturbance creates pathways that facilitate the migration of fluid into the goaf.

3.3. Force Analysis of the Broken Block

Building on the fundamental studies by Qian et al. [27,28,29], when overburden strata fracture into discrete segments, they typically form a stable “masonry beam” structure through interlocking forces. To investigate the secondary activation mechanisms triggered by lower-seam extraction, a mechanical model of fractured goaf roof “activation” was established. The structural configuration, illustrated in Figure 7, represents the equilibrium state of the articulated rock blocks and the potential for instability under secondary mining disturbances.

From the geometric relationships of the rock blocks after rotation [30], as shown in the figure, it can be seen that:

$$W_{n1} = l_{n1}\sin {\theta }_1$$

(5)

$$W_{n2}=l_{n2}\left(\sin {\theta }_1+\sin {\theta }_2\right)$$

(6)

$$a={\displaystyle \frac{1}{2}}(m_4-l_n\mathit{\text{sin}}{\theta }_1)$$

(7)

The variables used in the geometric and mechanical equilibrium equations are defined as follows: θ₁, θ₂: Rotation angles of key blocks B and C, respectively (rad). W_n1, W_n2: Separation distances (gaps) beneath blocks B and C in the fourth rock layer (the 221 goaf) and the overlying stratum following the extraction of the 3-1# coal seam (m). l_n1, l_n2: Horizontal lengths of blocks B and C in the fourth rock layer (the 221 goaf), respectively (m). q_n: Average distributed load acting on the rock blocks (N/m). a: Height of the contact surface between the articulated blocks (m). m₄: Thickness of the rock blocks within the fourth rock layer (the 221 goaf) (m).

The block length governs the stress distribution characteristics. Owing to the inherent uniformity of periodic fracturing behavior across stratified rock masses, the equivalence l_n1= l_n2 is justified. Under conditions of simple contact between adjacent blocks—where a plastic hinge mechanism develops at the interface—the horizontal thrust T acts at the centroid of the compressive contact surface, i.e., at its geometric midpoint.

$$Q_A+{ Q}_B+R-F_1-F_2=0$$

(8)

Based on the global structural analysis of the masonry beam, equilibrium requires R = F₁ =F₂. Prior experimental and theoretical studies indicate that θ₁ and θ₂ are sufficiently small to justify the kinematic approximation θ₂ ≈ θ₁/4. Under this small-angle assumption, the trigonometric terms simplify as follows: cos θ₁ ≈ cos θ₂ ≈ 1 and sin θ₂ ≈ 0. Substituting these approximations into the governing equation yields:

$$Q_B = {\displaystyle \frac{ T\left(W_{n2}-W_{n1} + \frac{1}{2}a\right)}{l_n-\frac{1}{2}a\tan {\theta }_2}}$$

(9)

$$Q_{A }=F_1-Q_B$$

(10)

Solving the coupled system of equations yields analytical expressions for the horizontal compressive force T acting between the fractured blocks and the frictional force Q_B mobilized along the contact interface of rock block B:

$$T = {\displaystyle \frac{2F_1{l}_n}{2m_n-W_{n1}}}$$

(11)

$$Q_B={\displaystyle \frac{F_1\left(m_n-W_{n1}\right)}{2\left(2m_n-W_{n1}\right)}}$$

(12)

The frictional force Q_A mobilized along the contact interface of rock block A is given by:

$$Q_A={\displaystyle \frac{F_1\left(3m_n-W_{n1}\right)}{4m_n-2W_{n1}}}$$

(13)

where T denotes the horizontal thrust acting between the fractured rock blocks (in N); Q_A and Q_B represent the frictional forces mobilized along the contact interfaces of rock blocks A and B, respectively (in N); F₁ is the self-weight of block B (in N); and γ is the unit weight of the rock stratum (in N/m³).

F₁ = γm₄l_n

(14)

In the formula: F₁ is the self-weight of block B, γ is the bulk density of the rock layer, N/m³.

As the key rock blocks rotate, the horizontal thrust (T) at the interlocking contact surface (a) increases progressively. Rotational instability occurs when the average compressive stress acting on this contact surface (T/a) exceeds the critical compressive strength of the rock mass at the block corners. Consequently, the stability condition for the “masonry beam” structure is expressed as follows [31]:

$$T \le a\eta {\sigma }_{\mathsf{ϵ}}$$

(15)

In the formula: η is the extrusion coefficient, take 0.3; σ_c is the compressive strength of the key layer rock block (MPa). It can be obtained as follows:

$${\displaystyle \frac{2F_1{l}_n}{2m_n-W_n}} \le 0.15\left(m_n-W_n\right){\sigma }_c$$

(16)

Substituting (14) into Equation (16) simplifies to:

$${\displaystyle \frac{2\gamma m_n{l}^2\mathrm{n}}{2m_n-W_{n1}}}\le 0.15 \left(m_n-W_{n1}\right){\sigma }_c$$

(17)

Simplified to:

$$W_{n1} \le {\displaystyle \frac{3}{2}}{m}_n-\sqrt{{\displaystyle \frac{17}{4}}{m}_n^2-{\displaystyle \frac{2m_{n}y{l}_n^2}{0.15{\sigma }_c}}}$$

(18)

From this, it can be inferred that the “activation” instability of the broken roof in the lower goaf occurs when the amount of the separation layer of the broken roof in the goaf is greater than the minimum sinking amount of the stable broken roof $W_{n1}\ge {\displaystyle \frac{3}{2}}{m}_n-\sqrt{{\displaystyle \frac{17}{4}}{m}_n^2-{\displaystyle \frac{2m_{n}y{l}_n^2}{0.15{\sigma }_c}}}$.

According to the “masonry beam” mechanics model of the key rock block, when the shear force Q_A at the interlocking point a exceeds the frictional constraint force at that point, the structure will experience sliding instability, namely:

$$T \tan \varphi \ge Q_A$$

(19)

In the formula: φ is the friction angle between the key rock blocks, °; φ takes the value of 0.3.

Substituting Equation (12) gives:

$$W_1\ge { 3m}_n-{1.2l}_n$$

(20)

The sliding instability of fractured roof structures within a goaf is primarily dictated by the thickness and length of the rock blocks, as well as the separation distance between these blocks and the overlying strata. Specifically, when the separation distance between fractured blocks in a large goaf exceeds the threshold defined by 3 m4 − 1.2 ln, sliding instability occurs, triggering roof “reactivation. “For the 3-1# coal seam, the subsidence of the key rock block associated with the failure of the low-position fine-grained sandstone key stratum is 4.61 m, satisfying the condition: 3.17 < W₁ = 4.61 m < 18 m. Similarly, the key rock block of the fractured key stratum composed of the middle-positioned medium-grained sandstone, its average thickness m₁ = 40 m, length l₁ = 42 m, compressive strength 53.56 MPa, and unit weight γ = 24,228.35 N/m³. By substituting these parameters into Equations (1)–(13) and (1)–(14), respectively, the subsidence amount of the key rock block in the fractured key stratum of the middle-positioned medium-grained sandstone above the 3-1# coal seam is calculated to be 3.17 m < W₂ = 5.31 m < 69.6 m. This implies that when the 3-1# coal seam is mined, the key rock block in the fractured key stratum of the middle—positioned fine—grained sandstone, which serves as the main roof of the overlying strata, will experience “activation” instability.

4. Determination of Fundamental Mechanical Properties of the Coal Seam Roof

4.1. Preparation of Coal-Rock Block Specimens and Testing Apparatus

In situ coal and rock blocks exhibited irregular geometries and thus required machining to meet test specifications. Core samples were drilled and precision-machined into cylindrical specimens of three dimensions: Φ50 mm × 100 mm (for uniaxial compression tests), Φ50 mm × 50 mm (for direct shear tests), and Φ50 mm × 25 mm (for Brazilian splitting tests). The final specimens are shown in Figure 8.

Testing was conducted using a static resistance strain gauge, a KYAW-200S microcomputer-controlled electronic universal testing machine (MCEUTM), and a variable-angle shear apparatus. The MCEUTM was used for uniaxial compressive strength (UCS), uniaxial tensile strength (UTS), and direct shear tests. Radial strain was measured with the strain gauge; shear strength, cohesion, and internal friction angle were determined using the variable-angle shear apparatus. The MCEUTM is shown in Figure 9.

4.2. Analysis of Compressive Strength Testing for Coal and Rock Specimens

Based on the experimental results, stress–time and stress–strain curves corresponding to the uniaxial failure of selected coal-rock specimens were plotted, as detailed in Figure 10 and Figure 11.

The typical stress–strain curves of specimens in the elastic modulus and Poisson’s ratio tests are shown in Figure 11.

4.3. Strength Test Results of Coal and Rock Specimens

In this chapter, uniaxial compressive strength (UCS), uniaxial tensile strength (UTS), and shear strength tests were conducted on the roof siltstone, medium-grained sandstone, and 3-1# coal using a KYAW-200S microcomputer-controlled electronic universal testing machine, combined with strain gauges and variable-angle shear apparatuses. Based on the measured results, the roof siltstone exhibited an average UCS of 51 MPa, average elastic modulus (E) of 7.7 GPa, average Poisson’s ratio (ν) of 0.26, average UTS of 2.4 MPa, average internal friction angle (φ) of 30.4°, and average cohesion (c) of 10.3 MPa. For the roof medium-grained sandstone, the average UCS was 52.65 MPa, average E was 8.6 GPa, average ν was 0.25, average UTS was 6 MPa, average φ was 37.26°, and average c was 16.90 MPa. The 3-1# coal specimen showed an average UCS of 20.4 MPa, average E of 3.4 GPa, average ν of 0.3, average UTS of 1.15 MPa, average φ of 32°, and average c of 6.35 MPa.

5. Research on the Development Pattern of the Mining Roof Based on 3DEC Numerical Simulation

5.1. Model Building

Based on the geological and mining conditions of the 317 working face, a computational model was developed to analyze the evolution of overburden fractures and the development of the fractured zone during face advancement. To eliminate boundary effects, buffer zones were incorporated on all sides. The model dimensions were established as X × Y × Z = 1100 × 320 × 167.5 m, where the X-direction represents the strike (advance direction) and the Y-direction represents the dip of the working face. The mesh density was varied to balance computational efficiency with accuracy. The top roof strata were discretized into blocks of 40 × 10 × 35 m, while the bottom strata blocks were 40 × 10 × 35 m. For the target strata of interest, a finer resolution of 5 × 10 × 5 m was applied. This planar model simulates roof caving, displacement characteristics, and stress evolution during the formation of the goaf. During numerical excavation, 50 m-wide boundary coal pillars were reserved to ensure model stability. The model geometry is illustrated in Figure 12. The 317 working face was excavated in four steps: an initial 50 m advance followed by three subsequent increments of 200 m, reaching a total cumulative distance of 650 m. The mining height was set to 4.4 m, utilizing a single-pass full-seam extraction method.

To simulate the overburden weight, an equivalent vertical load of 16.5 MPa was applied to the upper boundary. The lateral pressure coefficient was set to 1.0, and gravitational acceleration was defined as 9.8 m/s². The model was constrained by appropriate boundary conditions, including in situ stress, shear modulus, Poisson’s ratio, density, and displacement constraints. The specific physical and mechanical parameters for the various stratigraphic units are summarized in

Table 1. Rock Mechanics Parameters.

Rock Formation Names	Density/kg·m	Compressive Strength/MPa	Tensile Strength/MPa	Cohesion /MPa	Internal Friction Angle/°	Poisson’s Ratio	Elastic Modulus /GPa
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7
Fine-grained sandstone	2407.0	53.0	6.2	17.0	34.8	0.2	8.9
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7
Medium-grained sandstone	2407.0	53.0	6.2	17.0	34.8	0.2	8.9
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7
Medium-grained sandstone	2407.0	53.0	6.2	17.0	34.8	0.2	8.9
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7
2-2 coal	2597.0	21.1	1.3	6.3	31.7	0.3	3.7
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7
Fine-grained sandstone	2407.0	53.0	6.2	17.0	34.8	0.2	8.9
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7
3-1 Coal	1390.0	12.8	0.6	2.4	28.4	0.2	3.3
Siltstone	2574.0	52.8	2.5	10.3	31.4	0.2	7.7

.

In the numerical model, non-coal strata were first discretized into Blocks (discrete elements used to simulate macroscopic cracking, sliding, and compaction). Each block was further subdivided into multiple Zones (deformable elements) to capture internal rock mass deformation. To accurately reflect the heterogeneity of the coal seam, a Trigon block system was employed. Since coal mass failure in laboratory uniaxial compression tests typically occurs along pre-existing joint planes, these random natural joints and fractures were simulated using randomly generated contact elements. These elements support tensile and shear failure, effectively reproducing the stochastic failure characteristics of the coal mass. The Mohr–Coulomb constitutive model was applied to all rock strata to represent the matrix behavior, while the Coulomb slip model was utilized for all joint interfaces to govern shear and tensile behavior. Boundary conditions were applied to the model to reflect in situ constraints: Bottom Boundary: Fully constrained (u =0, v = 0); Lateral Boundaries (Left/Right): Partially constrained to allow vertical movement while preventing horizontal expansion (u = 0, v ≠ 0); Upper Boundary: Defined as a free boundary. To account for the missing overburden above the simulation domain, a vertical stress equivalent to the self-weight of the overlying strata was applied as a uniform external load. Block sizes were optimized based on the required sensitivity in different regions. Fine-mesh discretization was applied to the coal seam, immediate roof, and caving zone, where deformation is most volatile. Conversely, larger block sizes were used for the floor and main roof to improve computational efficiency. The strata were modeled as staggered, interbedded structures containing both horizontal and vertical joints. Physical and mechanical properties for the rock mass and joints are detailed in Table 1 and

Table 2. Physical and Mechanical Parameters of joints.

Rock Strata	Normal Stiffness/GPa	Shear Stiffness GPa	Joint Cohesion/MPa	Joint Internal Friction Angle/MPa	Joint Tensile Strength/MPa
Siltstone	13.3	8.3	1.6	35.0	1.0
Fine-grained sandstone	13.3	8.3	1.6	35.0	1.0
Siltstone	13.3	8.3	1.6	35.0	1.0
Medium-grained sandstone	13.3	8.3	1.6	35.0	1.0
Siltstone	13.3	8.3	1.6	35.0	1.0
Medium-grained sandstone	13.3	8.3	1.6	35.0	1.0
Siltstone	13.3	8.3	1.6	35.0	1.0
2-2 coal	4.3	1.1	0.1	18.0	0.1
Siltstone	13.3	8.3	1.6	35.0	1.0
Fine-grained sandstone	13.3	8.3	1.6	35.0	1.0
Siltstone	13.3	8.3	1.6	35.0	1.0
3-1 Coal	4.3	1.1	0.1	18.0	0.1
Siltstone	13.3	8.3	1.6	35.0	1.0

, respectively.

The rock parameters derived from experiments are the elastic modulus (E) and Poisson’s ratio (μ). In contrast, the 3DEC deformable material model is defined by two key parameters: bulk modulus (K) and shear modulus (G), which are the inputs required for numerical computations. Thus, K and G were converted from E and μ using Equations (21) and (22). The physical and mechanical properties of each rock stratum were acquired from the geological department of the mine, with detailed values presented in Table 2

$$K={\displaystyle \frac{E}{3(1-2\mu )}}$$

(21)

$$G={\displaystyle \frac{E}{2(1+\mu )}}$$

(22)

The normal stiffness (K_n) and shear stiffness (K_s) were derived from the bulk modulus (K) and shear modulus (G) via Equations (21) and (22) [32].

$${\mathrm{K}}_{\mathrm{n}}=-7.15+1.75\mathrm{J}\mathrm{R}\mathrm{C}+2\left(\mathrm{J}\mathrm{C}\mathrm{S}/\mathrm{J}\mathrm{R}\mathrm{C}\right)$$

(23)

$${\mathrm{K}}_{\mathrm{s}}=100/\mathrm{L}\cdot \mathrm{J}\mathrm{C}\mathrm{S}\cdot \mathrm{t}\mathrm{a}\mathrm{n}{\mathsf{\phi }}_{\mathrm{r}}$$

(24)

where JRC—joint roughness coefficient; JCS—joint compressive strength; L—trace length; $\mathsf{\phi }$—friction angle.

Furthermore, to guarantee the computational efficiency and stability of the explicit finite difference scheme, the calculated stiffness values were validated to lie below the numerical threshold recommended by Itasca [33].

$$k_n, k_s\le 10\left[{\displaystyle \frac{K+\frac{4}{3}G}{\mathsf{\Delta }{z}_{\mathit{\text{min}}}}}\right]$$

(25)

$\mathsf{\Delta }{z}_{\mathit{\text{min}}}$—Minimum zone dimension, m

Following model construction, the simulation was executed according to the following computational sequence:

• Initial In Situ Stress Equilibrium: The model was first brought to an initial equilibrium state under gravitational loading (self-weight) to establish the virgin in situ stress field before any mining activities.
• Extraction of the Upper Seam (221 Coal Seam): The 221 coal seam was excavated to simulate the formation of the overlying goaf. The model was allowed to reach a new equilibrium state, capturing the primary disturbance and the resulting settled state of the overburden.
• Stepwise Advancement of the 317 Working Face: The 317 working face was excavated progressively to capture the dynamic evolution of the strata. Initial Step: An advance of 50 m from the open-off cut. Subsequent Steps: Three increments of 200 m each, totaling a 650 m advance. After each advancement, stress equilibrium calculations were performed. Throughout this process, key metrics were monitored, including vertical and horizontal stress, strata displacement, and the vertical height and lateral extent of the plastic failure zone.

5.2. Vertical Displacement Distribution of the Roof in the 221 Goaf During the Recovery Period of the 317 Working Face

5.2.1. The Distribution Characteristics of the Collapse

Figure 13 shows the distribution characteristics of the collapse during the mining process of the 317 working face.

5.2.2. Conclusions

Analysis of the vertical displacement and fracture development (Figure 13) reveals a four-stage evolution of the overburden structure as the 317 working face advances:

When the 317 working face advances by 50 m, the main roof undergoes pronounced subsidence and initial caving. The fractures propagate upward, establishing a gradual connection between the 317 goaf and the overlying 221 goaf. This interaction results in the formation of a hinged cantilever beam structure, marking the transition from isolated to coupled strata behavior.

At an advancement of 250 m, the disturbance reaches the strata above the immediate roof, leading to bed separation between the immediate roof and the overlying siltstone. Subsidence Characteristics: The overlying strata exhibit large-scale, low-amplitude bending subsidence. Structural Impact: The extraction induces extensive caving that reactivates the 221 goaf roof, intensifying its settlement. Secondary Activation: Secondary fracturing and caving occur within the 221 roof; the resulting rock blocks undergo mutual compression and interlocking to form a stable masonry rock-beam structure, which further increases the subsidence height.

As the face reaches 450 m, the entire stratigraphic column between the 317 face and the 221 goaf undergoes overall bending deformation. Compaction: The previously caved gangue in the lower goaf is compacted under the increasing load of the rotating overburden. Fracture Propagation: The main roof of the 317 face fractures and interlocks, while the main roof of the 221 goaf undergoes secondary fracturing. This synchronized failure induces a composite subsidence effect, where the upper coal seam’s main roof follows the settlement trajectory of the lower seam.

Upon reaching an advance of 650 m, the bending subsidence of the 221 goaf roof reaches a state of relative equilibrium. Activation Height: The height of roof activation stabilizes, and the 317 face transitions into a phase of periodic caving. Development of the “Two Zones”: The fracture zone propagates upward until it reaches the immediate roof of the 221 face. At this stage, the heights of the caving and fracture zones are fully developed, collectively forming a stable “trapezoidal” goaf structure across the multi-seam system.

5.2.3. Vertical Displacement Contour Plot

Figure 14 presents the vertical displacement contour plot during extraction of the 317 working face, while Figure 15 shows the monitoring curve of roof subsidence at a position 30 m above the 221 coal seam.

5.2.4. Vertical Displacement Contour Plot Analysis

Synthesizing the data from the displacement contours (Figure 7) and the monitoring curves (Figure 8), the evolutionary characteristics of the overburden can be summarized into four key advancement phases: At a 50 m advance, the immediate roof of the 317 face reaches a maximum subsidence of 5.0 m (located 23 m from the open-off cut). Simultaneously, the horizon 30 m above the 221 working face records an initial subsidence of 1.89 m. This displacement is concentrated 21.5 m above the coal seam, corresponding to the central axis of the existing 221 goaf.

As the advancement reaches 250 m, the maximum subsidence of the 317 immediate roof remains 5.0 m within the central goaf. However, subsidence 30 m above the 221 working face increases to 2.8 m, representing a 1.4-fold increase over the previous stage. This indicates that higher-level roof strata are undergoing pronounced, large-scale caving. The expanded subsidence range of the main roof at this location significantly promotes the upward propagation of the water-conducting fracture zone.

By the 450 m mark, subsidence of the 317 immediate roof and the strata 30 m above the coal seam both reach the mining height of 5.0 m. The higher overlying strata exhibit marked bending deformation and fracture development. At this stage, maximum subsidence 30 m above the 221 goaf reaches 4.75 m. The activation height of the 221 roof stabilizes at approximately 35 m, or a vertical distance of 72 m above the 317 coal seam. While displacement contours show that the siltstone subsides slowly as the caving zone propagates upward, the 10.27 m-thick siltstone layer is partially supported by the collapsed gangue of the immediate roof, restraining its total caving failure.

At the final 650 m advancement, the maximum subsidence 30 m above the 221 goaf increases to 5.475 m. The activation height stabilizes at 37.3 m (75 m above the 317 coal seam). High-strength medium-grained sandstone and siltstone layers provide structural support, inhibiting further development of the caving and fracture zones. Consequently, subsidence does not reach the cumulative height of both coal seams. At this stage, the caving zone height is recorded at 34.19 m.

5.2.5. Conclusions

Summary of Evolutionary Patterns The continuous advancement of the 317 working face links the fractured lower main roof with the 221 goaf, forming a hinged cantilever beam structure. The overlying basic roof exhibits pronounced delamination, followed by progressive bending subsidence and localized fracturing. However, full strata collapse is prevented due to the load-bearing capacity provided by multiple competent layers—predominantly medium-grained sandstone and siltstone—situated immediately above the 3107 coal seam, which effectively support the overlying rock mass. This evolution triggers secondary activation of the 221 overlying strata, characterized by an “initial activation–stabilization–reactivation” pattern. Mutually compressed rock blocks form a rock-beam structure, while the upper main roof transitions from separation to bending subsidence and partial fracturing. Ultimately, the siltstone and fine-grained sandstone layers undergo complete caving, with the caving zone reaching a height of 34.19 m. The fracture zone extends into the bending zone at a height of 75.2 m. The increase in maximum subsidence from 1.89 m to 5.475 m aligns with theoretical predictions, confirming that mining the 317 face induces significant secondary movement in the overlying 221 goaf.

5.3. Vertical Stress Distribution of the Roof During the Recovery of the 317 Working Face

5.3.1. Vertical Stress Distribution of the Roof Analysis

As illustrated in Figure 16 and Figure 17, the initial extraction of the coal seam—prior to the 317 working face advancing beneath the 221 goaf—induces significant stress concentration in the surrounding coal–rock mass, establishing distinct abutment pressure zones. Following the extraction of the 317 working face, vertical stress peaks at approximately 29.39 MPa and 32.02 MPa within the coal pillars on either side, indicating the formation of pronounced abutment pressure zones. Conversely, the rock mass directly above the goaf exhibits an inverted “positive triangular” pressure-relief zone, where the vertical stress is reduced to approximately 0.89 MPa. As the 317 working face advances from 0 m to 450 m, the stress concentration zone is increasingly influenced by the overlying 221 goaf. During this stage, the stress field evolves into a tripartite pressure-relief pattern characterized as “weak–strong–strong.” Vertical stress concentrations remain evident on both sides of the 317 goaf; specifically, at the intersection zone between the 317 and 221 working faces, maximum vertical stresses are recorded at 19.94 MPa, 14.44 MPa, and 20.56 MPa. When the advancement reaches 650 m, vertical stress levels show a slight rebound, recovering to an average of approximately 23.5 MPa. This rebound is attributed to the gradual compaction of the rear goaf and the progressive closure of overlying separation layers. Despite this recovery, stress concentrations persist within the roof and floor strata flanking the working face.

5.3.2. Conclusions

The simulation results demonstrate that mining-induced fracturing and subsequent fracture propagation trigger roof activation. Deformation is minimal above the boundary coal pillars, which provide a protective effect for the surrounding strata within a range of approximately 45° inclined toward the goaf. Subsidence increases toward the center of the goaf, resulting in a “trapezoidal” bending deformation pattern. Within the overburden of the 317 working face, competent strata—such as thick medium-grained sandstone and siltstone—exert a controlling influence that inhibits the upward development of caving and fracture zones. Consequently, the total subsidence of the 221 working face reaches a maximum of 5.475 m, which is less than the combined mining height of the two seams. In summary, the stress distribution evolves from an initial “positive triangular” zone to a tripartite “initial activation–stabilization–reactivation” pattern, aligning with the roof stress-reduction zones observed during onsite mining. The simulated caving behavior and the resulting displacement and stress contours provide a consistent mechanical explanation for the secondary fracturing process of the roof.

5.4. Study on the Water Seepage Law of the Working Face Roof

Upper-seam extraction redistributes stress around residual coal pillars, significantly altering the roof structure and stress environment for the underlying coal seam. When the lower seam is mined under these conditions, the overlying strata undergo repeated failure, leading to abnormal roof caving. This process can trigger the development of vertically connected water-conducting fractures that link the working face to overlying aquifers, substantially increasing the risk of water inrush. Mining-induced rockbursts in the lower coal seam typically coincide with secondary variations in pore pressure and fluid discharge. These coupled hydro-mechanical processes are critical drivers of dynamic disasters at the working face. Lower-seam extraction disrupts the stress equilibrium of the coal–rock mass for a second time, resulting in the rapid accumulation and sudden release of strain energy. Consequently, spatial gradients in pore pressure and fluid discharge serve as effective indicators for characterizing fluid concentration and attenuation zones during migration. To investigate the mechanical response of strata under fluid influence, 3DEC software was utilized to generate pore pressure distribution contours and fluid discharge vector evolution diagrams. These models capture the various excavation stages and the resulting hydrological changes throughout the mining process.

5.4.1. The Evolution of Fluid Discharge Vectors Analysis

Figure 18 illustrates the evolution of fluid discharge vectors within the rock strata under dynamic loading. Analysis of the vertical velocity distribution reveals three distinct stages of downward fluid propagation.

• Stage I: Fluid Initiation

As the working face advances from the open-off cut to 50 m, the lower main roof connects with the 221 goaf. Under quasi-static equilibrium, initial fluid discharge occurs via limited water inflow along vertical fractures within the separated strata. In the overlapping zone between the 317 working face and the 221 goaf, the magnitude of the fluid discharge vectors increases slightly. Initial vector peaks emerge, characterized by irregular flow patterns. During this phase, fluid migration remains relatively stable and is minimally influenced by the surrounding medium. The maximum vertical extent of the fluid discharge reaches 24.97 m, though it does not yet reach the 317 working face.

• Stage II: Migration within the Rock-Beam Structure

Fluid propagation then extends into the relatively intact rock mass. Following the failure of the lower main roof, fractured rock blocks hinge together to form a stable rock-beam structure. In addition to inflow along down-dip fractures and separations, significant water inflow develops within lower-level separation fractures. This phenomenon is driven by increasing hydrostatic pressure within the separations beneath the aquifer. Once this pressure exceeds the shear strength of the underlying strata, the rock mass integrity is compromised, creating a connection with the water-conducting fracture zone and establishing a water-inrush pathway. At an advancement of approximately 170 m, a pronounced peak in fluid discharge occurs. This peak is spatially broader and extends further downward than the initial peak, creating a “dual-peak” vector structure with a spacing of roughly 100 m. As mining progresses, the rear goaf undergoes compaction and overlying separations close, reducing fluid discharge in that area. Increased porosity and fracture development allow the flow pattern to maintain symmetry. However, because the siltstone layer lacks sufficient caving space, its collapse is suppressed, resulting in slow subsidence and gradual downward expansion of the fluid-affected zone. The maximum vertical fluid discharge reach increases to 58.36 m, eventually infiltrating the 317 working face.

• Stage III: Migration within a Stable Plastic Rock Mass

As the working face nears the trailing boundary of the 221 goaf, extensive compaction of the rear goaf and the closure of separations significantly diminish fluid discharge. Continued advancement compresses the main and immediate roofs into a unified structure. Upon entering the 221 goaf, the fluid flow becomes increasingly dispersed due to the diffusion of dynamic loading across the contact area. In this final stage, the maximum vertical extent of the fluid discharge is recorded at 51.75 m.

As shown in the Figure 19, it is the vertical displacement cloud map of the working face 317.

5.4.2. The Contour Maps of Joint Pore Water Pressure Analysis

Figure 20 displays the contour maps of joint pore water pressure density within the rock strata under dynamic loading. As the 317 working face advances, the pore pressure density evolves, exhibiting distinct spatial distribution characteristics at different intervals.

• Initial Stage: Hinged Cantilever Beam Formation

During the early formation of the hinged cantilever beam, pore pressure above the 317 working face increases progressively in the upward direction. As the pressure front reaches the overlying 221 working face, localized accumulation occurs, eventually stabilizing at the ambient joint pore pressure of approximately 1.89 MPa. At a distance of 73.6 m from the observation point, the peak pore pressure drops sharply, signaling the onset of a stable, accelerated seepage regime. In this phase, seepage is dominated by interlayer flow. Within the pore pressure concentration zone, the minimum principal stress direction aligns with the excavation direction. The inter-seam seepage pressure peaks recorded at 10 m, 20 m, 30 m, and 40 m from the observation point are 0.39 MPa, 0.58 MPa, 0.96 MPa, and 1.89 MPa, respectively.

• Intermediate Stage: Fluid Propagation in the Rock-Beam Structure

During fluid migration within the rock-beam structure, diffusion-controlled seepage causes a slight increase in pore water pressure at the observation point, reaching a maximum of 2.12 MPa. As the lower main roof fails and the fractured rock blocks hinge together, the resulting structural looseness causes joint pore pressures to drop locally toward 0 MPa due to drainage effects. In this stage, the seepage pressure peaks at distances of 10 m to 40 m increase to 0.52 MPa, 0.79 MPa, 1.32 MPa, and 2.12 MPa, respectively.

• Final Stage: Migration within a Stable Plastic Rock Mass

Once the working face passes completely beneath the upper boundary of the 221 goaf, the rock-beam structure is maintained by the hinged blocks of the lower main roof. Further advancement leads to the collapse of the main and immediate roofs, which accumulate into an arch-like structure. The caved strata of the 221 goaf and the overlying strata of the 317 face become hydraulically connected, forming an integrated arched seepage zone. As the rear goaf is compacted and separation layers close, the downward pore water pressure exceeds the interlayer pressure. This inversion is likely caused by a “water-blocking” effect induced by localized compaction within transitional strata, which hinders seepage and triggers pressure accumulation.

The evolution of fluid discharge vectors progresses from single peaks to a stage-specific dual-peak structure, characterized by irregular flow behavior. The maximum vertical extent of fluid discharge vectors initially reaches 24.97 m, peaks at 58.36 m (as noted in earlier stages), and eventually stabilizes.

5.4.3. Conclusions

While pore water pressure peaks drop sharply during early excavation, the pressure in the rear area gradually recovers as the goaf compacts and separations close. From an engineering safety perspective, particular attention must be paid to water-inrush hazards within the 50 m zone behind the working face. This area is especially vulnerable to secondary strata reactivation during mining-induced stress redistribution.

6. Engineering Verification

6.1. Microseismic Monitoring Analysis

To detect secondary strata reactivation under field conditions, a microseismic (MS) monitoring system was deployed. This system enables the real-time assessment of rock mass deformation, strata subsidence, and potential water-inrush precursors. The resulting data provide a scientific framework for identifying instability in the surrounding rock during excavation, allowing for timely adjustments to support measures or construction protocols. Consequently, dynamic control over the excavation process is achieved. This mitigates the risk of structural failure while ensuring operational safety and cost-efficiency, ultimately enhancing the controllability of the engineering project. Furthermore, MS monitoring data serve to evaluate the effectiveness of existing support systems and guide the dynamic optimization of support parameters [34,35].

Monitoring results indicate that as the 317 working face advanced beneath the 221 goaf, high-energy microseismic events clustered within a zone extending approximately 50 m above the 221 goaf. These events were primarily concentrated within a 37 m-thick medium-grained sandstone stratum and significantly impacted the structural integrity of the roadways associated with the 317 working face. The spatial distribution of these microseismic events during the mining process is illustrated in Figure 21.

As illustrated in Figure 21 and Figure 22, the distribution of microseismic (MS) events along the strike of the 317 working face reveals that after an advancement of approximately 600 m, activity within the goaf was primarily concentrated between the 100 m and 600 m marks. These events were predominantly clustered within the middle- to high-level roof strata.

Low-energy events occurred mainly in the fine-grained sandstone and siltstone immediately above the coal seam. In contrast, medium- and high-energy events were distributed within the higher fine-grained, coarse-grained, and medium-grained sandstone strata. While the total event count was dominated by low-energy releases on the order of 10 J and 10² J, higher-magnitude fracturing was evident in specific zones.

• Spatial Distribution and Energy Characteristics

The spatial data show that MS events, including high-energy clusters, were concentrated within the vertical projection zone between the lower and upper working faces. This concentration indicates a significant increase in the roof caving height within this region. Notably, within the 300–500 m range, a high frequency of fourth-order (10⁴ J) and fifth-order (10⁵ J) energy events was recorded alongside numerous low-energy events. The prevalence of these 10⁴ J and 10⁵ J events suggests large-scale fracturing above the upper working face, representing secondary fracturing induced by mining activity.

• Strata Impact and Hydrological Risk

The primary concentration of MS activity was located within the lower roof strata of the 317 coal seam—specifically within the 3 m siltstone layer and the 3–29 m fine-grained sandstone layer. As extraction progressed into its middle stage, the zone of dense MS activity migrated upward into the 29–50 m siltstone layer.

These observations demonstrate that secondary mining activities have significantly affected the overlying rock mass. The upward expansion of the fracture network reached the aquifer, creating a water-inrush hazard that poses a direct threat to the safety of the 317 working face.

6.2. Borehole Peeping Analysis

To monitor the overburden, observation boreholes were drilled near the 317 main haulage roadway, as shown specifically in the Figure 23 below. The specific arrangements are as follows:

Observation Borehole 4#: Located in the 319 development roadway, Borehole No. 4 was positioned at a horizontal distance of 30 m from the 221 working face return airway. It was oriented perpendicular to the 317 main haulage roadway, directed toward the 317 goaf with an inclination angle of 45°. Observation Borehole 1#: Also situated in the 319 development roadway, Borehole No. 1 extended into the 221 working face. This borehole was drilled obliquely toward the 317 goaf at a 45° inclination. Its horizontal projection formed a 15° angle relative to the 317 main haulage roadway.

The drilling fluid-loss data for Borehole No. 4, shown in Figure 24a, reveals a distinct four-stage evolution across the 0–80 m depth interval. Stage I: Initial Support Zone. The borehole is located within the coal pillar and roadway support zone. Leakage is minimal and confined to a limited interval, reflecting low permeability in the supported rock mass. Stage II: Caving Zone Entry. As drilling enters the caving zone, the leakage rate increases significantly. This shift indicates the onset of severe rock mass damage and increased fracture connectivity. Stage III: Hard Rock Stratum Penetration. The leakage rate rises sharply, reaching approximately 1.3–1.8 times the levels observed in Stage II. This indicates substantial rock mass failure within the hard stratum, resulting in enhanced permeability and accelerated fluid loss. Stage IV: Post-Fracture Zone. The leakage rate decreases abruptly. While the borehole penetrated an aquifer, the lack of sustained high leakage suggests it did not intersect the primary water-conducting fracture zone (WCFZ).

• Comparative Analysis: Borehole No. 1

The data from Borehole No. 1 in Figure 24b follow a similar stage-wise trend but with critical differences in magnitude and behavior: Higher Damage Intensity: During Stage III, the leakage rate in Borehole No. 1 (369–392 L/min) is approximately 1.1 times higher than that of Borehole No. 4. This suggests a higher degree of rock mass fracturing in the vicinity of Borehole No. 1. Sustained Fracture Connectivity: Unlike Borehole No. 4, the leakage rate in Stage IV of Borehole No. 1 continues to increase rather than dropping. This indicates that the borehole remained within the influence of the fracture network and did not penetrate beyond it into the aquifer.

These findings demonstrate that secondary fracturing significantly compromises roadway stability. Lower-seam extraction triggers strata reactivation, which reduces the structural integrity of the surrounding rock and facilitates water-inrush pathways. Consequently, mining and construction operations must prioritize roof stability monitoring and water-inrush prevention to mitigate the risk of catastrophic instability.

6.3. Numerical Simulation and Field Monitoring Validation

We performed a quantitative comparison between the simulated results and theoretical/field data (Figure 25):

Subsidence Calibration: The simulated maximum roof subsidence of 5.475 m was compared with the theoretical calculation based on key strata theory (W₂ = 5.31 m). The relative error is approximately 3%, which is within the acceptable range for complex geological engineering.

Fracture Zone Validation: The simulated height of the water-conducting fracture zone (stabilizing at 75.2 m) was cross-referenced with microseismic event clusters and the simulated results (concentrated 27–50 m above the lower seam, corresponding to the same horizon).

These verifications are sufficient to prove that this model method is effective and can appropriately analyze the secondary fracture law of the roof strata and the potential for water inrush in close-distance coal seam mining.

7. Conclusions and Limitations

This study introduces “Analysis of Secondary Fracture Law of Roof Strata and Water Inrush Potential in Close-Distance Coal Seam Mining”. Under the influence of an upper goaf, the fracture mode of overlying strata above the lower coal seam departs fundamentally from the classic “O-X” pattern seen in single-seam mining. The resulting fractured structure differs from existing models—such as rock beam, block, and micro-scale seepage models—in morphology, load-bearing behavior, and water-inrush susceptibility [11,17]. Most prior studies ignore this distinction: they either assume beam-like behavior along the working face strike or use plane-strain physical simulations to analyze fracture propagation. These approaches lose validity when the study area’s spatial relationship with overlying residual coal pillars violates 2D assumptions—limiting their engineering applicability for close-distance mining. To bridge this gap, we integrate 3DEC numerical simulation, theoretical analysis, and field measurements. Our approach explicitly distinguishes the evolution of overlying strata movement and surrounding rock failure during lower-seam mining (under goaf influence) from that of single-seam mining. By quantifying displacement, stress redistribution, and water-inrush mechanisms across the roof strata above and below the seam, we characterize spatial variations in structural stability across the working face. Validation using site data shows 97.12% agreement between numerical and theoretical results. Microseismic, drilling, and simulation outputs also align closely—confirming the model’s effectiveness in assessing overlying strata movement and supporting safer, more reliable mining operations.

7.1. Conclusions

• For the lower coal seam, the subsidence of the key rock block associated with the failure of the low-position fine-grained sandstone key stratum satisfies 3.17 m < W₁ = 4.61 m < 18 m. For the 3-1# coal seam, the subsidence of the key rock block resulting from the failure of the mid-position medium-grained sandstone key stratum satisfies 3.17 m < W₂ = 5.31 m < 69.6 m. These results indicate that during extraction of the 3-1# coal seam, fracture of the mid-position fine-grained sandstone key stratum in the basic roof of the overburden leads to “reactivated” instability of the key rock blocks.
• Mining-induced fracture initiation and propagation trigger strata reactivation. As the distance to the center of the goaf decreases, the subsidence of the overburden increases, ultimately resulting in a “trapezoidal” bending deformation pattern. Competent overlying strata above the 317 working face exert a significant controlling effect. This structural stiffness suppresses the expansion of both the caving zone and the water-conducting fracture zone, restricting the maximum subsidence to 5.475 m. During extraction, the stress field undergoes a distinct evolution: it transitions from an initial “regular triangular” pressure-relief zone into a tripartite “weak–strong–strong” distribution. As the working face continued to advance, fluid discharge in the spatially connected and overlapping zone between the 317 working face and the 221 goaf increased sequentially, exhibiting an “alternating” pattern of vector peak evolution. This behavior indicates that rock mass damage zones significantly enhance the potential for water inrush.
• Within the 300–500 m range, numerous low-energy microseismic events were recorded, together with multiple fourth-order (10⁴ J) and fifth-order (10⁵ J) energy events. Events with energies of 10⁴ J and 10⁵ J occurred most frequently, indicating large-scale fracturing above the upper working face and the occurrence of secondary fracturing. The primary concentration zone of microseismic activity migrated upward from the low-level strata—comprising approximately 3 m of siltstone and 3–29 m of fine-grained sandstone above the lower coal seam—to the 29–50 m siltstone layer. These observations demonstrate that secondary excavation activities affected the aquifer, and that the associated water-inrush hazard poses a direct threat to the working face.

7.2. Control Strategies

• Source control: Pre-drain aquifers (e.g., via floor pre-drainage or directional drilling) to reduce stress concentration at key strata fracture points—lowering water-inrush intensity at its origin.
• Structural stabilization: Apply paste filling in rigid rocks (e.g., medium-grained sandstone, siltstone) to create buffer zones. This suppresses subsidence, limits the height of the water-conducting fracture zone, prevents water inflow into the goaf, and forms a functional water-resisting barrier.
• Roadway reinforcement: Increase support strength to enhance intrinsic resistance to water inrush and mitigate its mechanical impact on roadways.

7.3. Limitations

• The findings derive from one gently inclined coal seam in Northwest China; applicability to steeply inclined seams remains unverified.
• 3DEC modeling assumes homogeneous blocks and statistical joint distributions—potentially missing local geological anomalies.
• Microseismic data identify fracture zones, but the link between seismic energy and permeability evolution is still qualitative; no quantitative damage–permeability–flow constitutive relationship has been established.

Future work will extend the model to diverse geological settings and develop quantitative damage-permeability coupling models using long-term microseismic statistics.