Taking the No. 26 coal seam in the 72605 return air roadway as the research object, 10 m long core samples collected from this area were analyzed to clarify the structural assemblage characteristics of the roof and floor strata of the roadway. On this basis, laboratory experiments were further conducted to investigate the physical and mechanical properties of the rock. The core samples obtained from the roof and floor are shown in Figure 1 and Figure 2.

Figure 1. Structural characteristics of roadway roof strata.

As shown in Figure 1, the roof core of the No. 26 coal seam can be divided into three lithological sections with distinctly different characteristics. The interval from 0 to 4.1 m consists mainly of gray sandy mudstone, in which silty particles are uniformly distributed, with a slight coarsening of grains in the middle part, and plant debris fossils are locally observed. Although more than ten core segments longer than 10 cm were recovered within this section, the overall integrity of the core is poor, exhibiting a fragmented structure.

The section from 4.0 to 9.07 m is dominated by dark gray argillaceous sandstone with evenly distributed argillaceous cement. Horizontal bedding is clearly developed, accompanied by thin interlayers of mudstone, and occasional sandstone interbeds occur in the lower part. The core from this interval also shows a relatively high degree of fragmentation.

The interval from 9.07 to 10.0 m corresponds to black massive semi-bright coal, characterized by irregular fracture surfaces. Due to mechanical disturbance during drilling and the scouring effect of drilling fluids, the integrity of the coal core in this section was significantly damaged, and no intact coal core longer than 10 cm was recovered.

Figure 2. Structural characteristics of roadway floor strata.

As shown in Figure 2, the floor core of the No. 26 coal seam is overall highly fragmented. Within several intervals, including 0.72 - 1.90 m, the rock exhibits a grayish-white color, moderate cementation, and poor sorting. It is composed mainly of quartz and feldspar grains, with a small amount of clay or other cementing materials binding the sand particles into a consolidated rock mass; therefore, these sections are identified as medium-grained sandstone. The overall integrity of the core is poor, and no intact rock segments longer than 10 cm were observed.

Within the ranges of 0 - 0.72 m,1.90 - 5.72 m,7 - 10 m, the rock is mainly gray in color, locally transitioning to dark gray. Silty particles are generally uniformly distributed, with slight grain coarsening in the middle portions, and plant debris fossils are commonly present. The argillaceous components are unevenly distributed, and these sections are identified as silty mudstone. In the interval of 5.72 - 7 m, the rock is predominantly dark gray, with uniformly distributed argillaceous material and thin mudstone interlayers, and is therefore identified as mudstone.

In summary, the thickness distribution of roof lithologies in the No. 26 coal seam shows relatively minor variation. The roadway roof is mainly composed of sandy mudstone and argillaceous sandstone. In contrast, the floor lithology is more complex and exhibits rapid lithological changes, consisting primarily of medium-grained sandstone, silty mudstone, and mudstone.

Two vertical exploratory boreholes with a depth of 8 m were drilled into the roof of the 72605 transportation roadway. A digital borehole panoramic imaging system was employed to inspect and characterize the internal structural features of the rock mass within the boreholes. The imaging results of Borehole No. 1 are presented in Figure 3.

Figure 3. Borehole observation results of roof No. 1.

Based on the roof imaging data obtained from Borehole No. 1, the revealed lithological assemblage is relatively simple, consisting mainly of alternating silty mudstone and argillaceous sandstone, with a single, extremely thin coal interlayer developed locally. Notably, the argillaceous sandstone contains a relatively high proportion of clay minerals.

In terms of specific geological features, interbedded coal seams were identified at borehole depths of 0.52 m and 9.56 m. Within the upper 2.0 m of the roadway roof, multiple transverse fractures are observed, while a longitudinal fracture is developed in the silty mudstone at a depth of 5.15 m, with an extension length of approximately 1.8 m. Overall, the fracture distribution revealed by this borehole exhibits a clear depth-dependent concentration, mainly confined to the zone within 2.0 m above the roadway roof. It is worth noting that the two coal interlayers constitute potential structural weak planes. In addition, seepage has already been observed in the roof area corresponding to this borehole. Considering the scale and distribution of the fractures, the roof fractures at present can be characterized as being at a moderate development level.

The borehole imaging of the No. 2 roof borehole is shown in Figure 4.

Based on the roof imaging data from Borehole No. 2, the revealed lithological structure is relatively simple, consisting mainly of alternating silty mudstone and argillaceous sandstone. The argillaceous sandstone is characterized by a relatively high clay content.

Detailed geological observations show that transverse fractures are developed at borehole depths of 0.88 m, 1.09 m, and 1.82 m, and multiple transverse fractures are distributed within the zone extending 2.0 m above the roadway roof. Overall, the fractures revealed by this borehole exhibit a pronounced depth-dependent concentration, being mainly confined to the interval within 2.0 m above the roadway roof. In addition, seepage has been observed in the roof area corresponding to this borehole. Considering the scale and distribution of the fractures, the current roof fractures can be characterized as exhibiting a moderate level of development.

Figure 4.Borehole observation results of roof No. 2.

In summary, the roadway roof is primarily composed of a composite assemblage of silty mudstone and argillaceous sandstone, with a thin coal interlayer of approximately 1.0 m locally developed. Within the 0 - 2.0 m range above the roof, the rock mass is relatively fragmented, forming a fracture zone dominated by transverse fractures. Moreover, longitudinal fractures are developed within the argillaceous sandstone at depths of 5.15 - 6.95 m. Overall, the roof fractures are assessed to be at a moderate stage of development.

Laboratory tests were conducted on coal and rock samples to determine key mechanical parameters, including uniaxial compressive strength, elastic modulus, Poisson’s ratio, and cohesion. All tests were performed in accordance with the standard Methods for Determining the Physical and Mechanical Properties of Coal and Rock (GB/T 23561-2009). The test results are presented in Table 1.

Table1. Physical and mechanical properties of rock specimens.

To determine the stress state of the roadway surrounding rock, the hydraulic fracturing method was employed to measure the in-situ stress. The measurements were conducted at the R264 survey point of the 72605 transportation roadway in Baliancheng Coal Mine. The measurement boreholes were drilled vertically upward, with the borehole collar located at a burial depth of approximately 590 m.

Figure 5. Pressure variation curve of measuring point 3.

A total of three measurement boreholes were arranged, with a spacing of 10 m between adjacent boreholes. Single-interval hydraulic fracturing tests were completed in each borehole. During the fracturing tests, the breakdown pressure, reopening pressure, and fracture closure pressure recorded in different loading-unloading cycles exhibited clear and well-defined response characteristics, demonstrating good consistency and repeatability among the boreholes. The detailed measurement results of the three boreholes at Measurement Point No. 3 are presented in Figure 5 and Table 2.

Table 2. Hydraulic fracturing parameters and calculated in-situ principal stresses for three boreholes.

1) Pb—In-situ Rock Breakdown Pressure; Pr—Fracture Reopening Pressure; Ps—Instantaneous Fracture Surface Pressure; Sh—Minimum Horizontal Principal Stress; SH—Maximum Horizontal Principal Stress; Sv—Vertical Principal Stress. 2) The vertical stress Sv was calculated assuming an average bulk density of the overlying rock mass of 2.5 g/cm³.

Based on the pressure data listed in Table 2, the maximum and minimum horizontal principal stresses (S_H and S_h) in this area were determined from the fracture reopening pressure and fracture closure pressure, respectively. The calculated values of S_H range from 14.07 to 18.69 MPa, while Sh varies between 9.01 and 10.72 MPa.

According to the thickness of the overlying strata and assuming a rock bulk density of 2.5 g/cm³, the vertical principal stress (S_V) was calculated to be 14.01 MPa. The resulting stress relationship satisfies S_H > S_V > S_h, indicating that the stress field is dominated by horizontal stress.

Borehole imaging was further used to identify the fracture orientations in the three measurement boreholes, as shown in Figure 6. The corresponding orientations of the maximum horizontal principal stress (S_H) are N54.56˚E, N53.55˚E, and N33.67˚E, respectively.

Figure 6. Fracturing fracture.

Based on the engineering geological conditions of Baliancheng Coal Mine, a three-dimensional numerical simulation was conducted using FLAC3D to analyze the mining process of the No. 26 coal seam. During the simulation, the influence of coal seam extraction on the stress field of the surrounding rock was systematically investigated, and the evolution characteristics of the mining-induced stress field during the retreat mining process were obtained.

3.2.1. Model Establishment

Based on the geological conditions of the No. 26 coal seam at Baliancheng Coal Mine, a three-dimensional numerical model with dimensions of 400 m × 500 m × 282 m was established to simulate the mining-induced effects of the coal seam. The model coordinate system was defined such that the x-axis is aligned with the dip direction of the working face and the y-axis with the strike direction. A key refinement zone was defined within 40 m above and below the No. 26 coal seam, where the element size of the coal seam and the immediate roof and floor was refined to 3 m × 3 m × 2 m. Additional mesh refinement was applied along the advance direction of the working face and within its influence zone along the strike to ensure accurate determination of the peak abutment pressure and its spatial extent. The refined model consists of approximately 2.0 × 10⁶ elements, and the mesh quality satisfies numerical requirements, ensuring computational stability.

The bottom boundary of the model was fixed in all directions, while normal displacement constraints were applied to the lateral boundaries in the x and y directions. A vertical stress of 9.725 MPa was applied to the top boundary to represent the overburden load. The Mohr-Coulomb constitutive model was adopted for all rock strata, with a bulk density of 2.5 g/cm³. According to the in-situ stress measurement results (Table2), the lateral pressure coefficients in the x and y directions were set to 1.4 and 1.2, respectively. The mechanical parameters of each rock layer were assigned based on laboratory test results (Table 1). The initial stress field was balanced and calibrated in terms of magnitude and orientation prior to excavation. Subsequently, the working face was excavated, and the distribution range of abutment pressure around the face was obtained and compared with field observations to further calibrate the model, thereby improving the reliability of the simulation results.

3.2.2. Distribution Characteristics of Mining-Induced Stress

Under natural conditions, the coal-rock mass exists in a state of three-dimensional stress equilibrium, and its physical and mechanical properties remain relatively stable. However, mining-induced disturbances disrupt this original equilibrium, leading to stress redistribution, fracture propagation, and degradation of mechanical properties, which in turn result in deformation and instability of the surrounding rock.

A systematic numerical simulation was conducted to analyze the excavation process of a working face with a maximum width of 140 m in the No. 26 coal seam of the Baliancheng Coal Mine. The stress contours around the goaf and the distribution of advanced mining-induced stress in front of the working face are illustrated in Figure 7 and Figure 8, respectively.

Figure 7. Stress contour map around the goaf of the No. 26 coal seam working face.

Figure 8. Advance abutment stress curve of the No. 26 coal seam working face.

After retreat mining of the coal seam working face, both the horizontal stress and the vertical stress in front of the working face exhibit an evolutionary trend characterized by an initial increase followed by a subsequent decrease, as shown in Figure 8. The specific characteristics are described as follows.

Variation characteristics of horizontal stress: Within the range of 0 - 30 m ahead of the working face, the horizontal stress shows a gradual increasing trend, reaching a peak value of 24.39 MPa, which is approximately 1.17 times the original in-situ stress. As the working face advances into the zone 30 - 130 m ahead, the horizontal stress gradually decreases, with values ranging from 22.11 to 24.39 MPa, corresponding to 1.06 - 1.17 times the original in-situ stress. Overall, the increase in horizontal stress is relatively limited, and its variation remains relatively gentle.

Variation characteristics of vertical stress: The vertical stress exhibits a more pronounced response to mining disturbance. In the zone 0 - 20 m ahead of the working face, the vertical stress increases progressively and reaches a peak value of 32.50 MPa, approximately 2.18 times the original in-situ stress. In the range of 20 - 130 m ahead of the working face, the vertical stress gradually decreases, varying between 16.52 and 32.50 MPa, equivalent to 1.11 - 2.18 times the original in-situ stress. Notably, within the interval of 20 - 30 m ahead of the working face, the vertical stress remains at a relatively high level, ranging from 24.91 to 32.50 MPa (1.67 - 2.18 times the original in-situ stress), indicating that this zone is still strongly influenced by mining activities.

Comprehensive analysis: The results indicate that, in the advance zone of the No. 26 coal seam working face, the magnitude of vertical stress increase is significantly greater than that of horizontal stress, and its variation trend is more pronounced. Overall, the region most strongly affected by mining-induced stress is concentrated within 0 - 30 m ahead of the working face. In particular, the high vertical stress zone in this range provides an important reference for surrounding rock stability assessment and support design.

Based on a comprehensive review of surrounding-rock classification methods both in China and abroad and targeted investigations, and considering practical applicability in production, a surrounding-rock stability classification scheme for retreat-mining roadways in gently inclined and inclined coal seams has been established in China; the corresponding classes and indices are summarized in Table 3.

Table 3. Classification of surrounding rock stability for retreat-mining roadways in gently inclined and inclined coal seams.

For a long time, roadway stability in coal mines has been evaluated largely through experience-based and qualitative (fuzzy) judgments, and unified quantitative criteria have been lacking. To scientifically characterize roadway stability, key classification indices were determined with reference to Qian’s classic work on mine pressure and its control ([11]) and the study by Li and Hou on the characteristics and application of rock-bolt support in roadways ([7]). The selected indices include the strengths of the roof, floor, and coal seam (σD, σd, σM), roadway burial depth (H), first caving interval of the immediate roof(L), and the mining influence index (N) ([12]; [1]).

1) Three surrounding rock strength indicators (σ_roof, σ_coal, σ_floor): The uniaxial compressive strengths of the roof and floor rocks were determined as weighted average values within ranges equivalent to two times and one time the roadway width above the roof and below the floor, respectively. The uniaxial compressive strength of the coal seam was taken as the strength of the coal on both sides of the roadway. All strength values are expressed in MPa.

2) Roadway burial depth: The burial depth of the roadway refers to the vertical distance between the roadway location and the ground surface. For inclined roadways, the burial depth can be determined by segmented averaging. The unit is meters (m).

3) Rock mass integrity index (L): The rock mass integrity index is determined by converting the initial caving step distance of the immediate roof. The unit is meters (m).

4) Mining influence indicator (N): The mining influence refers to the effect of advanced abutment pressure induced by working face extraction. The mining influence coefficient is defined as the mining-induced stress concentration factor.

Based on comprehensive field investigations, in situ monitoring, and laboratory tests, the roof of the No. 26 seam roadway is predominantly a medium-hard composite strata composed of silty mudstone and argillaceous sandstone, with an average uniaxial compressive strength (UCS) of 51.37 MPa. The floor consists of a single medium-hard silty mudstone layer, with an average UCS of 50.63 MPa. The coal seam has a Protodyakonov coefficient of 0.90. The surrounding rock mass shows moderately developed fracturing. The ratio of the maximum horizontal principal stress to the vertical stress is 1.33, and the mining-induced stress concentration factor is 2.18. The average first weighting interval of the immediate roof is 14 m. The maximum roof-floor convergence is 290 mm, with a relative convergence of 11.15%. According to the classification table for surrounding-rock stability of recovery roadways in gently inclined and inclined coal seams (Table 3), the surrounding-rock stability at Baliancheng Coal Mine is classified as Class III (moderately stable), as listed in Table 4.

Table 4.Classification results of surrounding rock stability of mining roadways in baliancheng coal mine.

Given that the surrounding-rock stability of the No. 26 seam retreat-mining gateroad at Baliancheng Coal Mine was assessed as Class III (moderately stable), the original support system was systematically optimized to improve the integrity and load-bearing capacity of the support structure. In the original scheme, roof bolts were 2000 mm long with 5 bolts per row; cable bolts (anchor cables) were 6500 mm long with 3 cables per row; and rib bolts were 1800 mm long with 4 bolts per row.

According to borehole imaging observations, the roof strata are mainly a composite of silty mudstone and argillaceous sandstone. A thin coal interlayer of approximately 1.0 m is present. Within 0 - 2.0 m above the roofline, the rock mass is relatively fractured, forming a jointed zone dominated by transverse discontinuities. In the argillaceous sandstone, longitudinal fractures are developed at depths of 5.15 - 6.95 m. The coal on both ribs is relatively weak, with a Protodyakonov coefficient f = 0.9.

In the optimized scheme, the roof-bolt length was increased to 2200 mm with 6 bolts per row; the cable-bolt length was increased to 7300 mm with 3 cables per row; and the rib-bolt length was increased to 2000 mm with 4 bolts per row. The row spacing remained the same as that in the original scheme. The optimized support layout is shown in Figure 9 and Table 5.

Table 5. Parameters of roadway support design.

Figure 9. Roadway support cross-section.

Figure 10. Displacement monitoring diagram of measuring point 1.

To evaluate the performance of the roadway support design, convergence monitoring was conducted during retreat mining by tracking the temporal evolution of roadway deformation, including roof-floor convergence and rib-to-rib convergence. A cross-measurement layout was adopted for convergence monitoring. Two monitoring stations were installed in the roadway. The deformation histories recorded over approximately 40 days are shown in Figure 10 and Figure 11. The results indicate that, as the working face advanced, both roof-floor convergence and rib convergence at the two stations increased noticeably. In general, the ribs exhibited larger deformation amplitudes than the roof and floor. After the implementation of the optimized support, the convergences of both the roof-floor and the ribs decreased significantly, suggesting an evident reinforcement effect. Therefore, the surrounding-rock stability classification and the corresponding support design are proven to be effective for maintaining roadway stability.

Figure 11. Displacement monitoring diagram of measuring point 2.