This paper aims to address the issue of hydraulic support crushing accidents or support failures in the retracement roadway (RR) that frequently occurs when a fully mechanized mining face is retraced during the end-mining stage. The deformation and instability mechanism of surrounding rock in the RR during the end mining of a fully mechanized mining face at the Hanjiawan Coal Mine located in the northern Shaanxi mining area is explored through field measurement, theoretical analysis, similar simulation, and numerical simulation. The results reveal that the stability of the remaining coal pillar (RCP) and the fracture position of the main roof are the main factors contributing to large-scale dynamic load pressure in the RR during the end-mining stage. The plastic zone width limit of the RCP is identified to be 5.5 m. Furthermore, the stress distribution within the RCP during the end-mining stage is determined, and the linear relationship between the load borne by the RCP and the strength of the coal pillar is quantified. A similar simulation experiment is conducted to examine the collapse and instability characteristics of the overlying rock structure during the end-mining stage. UDEC (v.5.0) software is utilized to optimize the roof support parameters of the RR. A surrounding rock control technology that integrates the anchor net cable and hydraulic chock is proposed to ensure RR stability. Meanwhile, a method involving ceasing mining operations and waiting pressure is adopted to ensure a safe and smooth connection between the working face and the RR. This study provides a reference for the surrounding rock control of the RR during end mining in shallow, closely-spaced coal seams under similar conditions.

In coal seam groups where the spacing between the upper and lower seams is small, the lower seam working face is significantly influenced by residual coal pillars from the upper seam and the void spaces created during mining. This presents considerable challenges for underground mining safety. Through field investigations, the layout of the coal seam quarry above the working face of the 3⁻¹ coal seam in Yanghuopan Mine was examined, along with the distribution of the residual coal pillars. This allowed for the identification of the interlayer rock strata characteristics. Subsequently, we analyzed the mechanism of directional hydraulic fracturing and decompression to determine the key parameters of the 3⁻¹ coal seam. Using the Rock Fracture Process Analysis 3D (RFPA 3D) numerical simulation, we evaluated the effects of various factors on the initiation and propagation of hydraulic fracturing-induced cracks, formulated the evolution law of these fractures, and incorporated the damage variables into the analysis. Additionally, we assessed the influence of different parameters on crack initiation and extension during hydraulic fracturing, using RFPA 3D simulations to derive the evolution law governing directional hydraulic fractures. This allowed us to define the hydraulic fracturing parameters for the 3⁻¹ interbedded rock layers by integrating the process parameter calculations with the damage variables. Based on these findings, an on-site implementation plan was developed and executed, followed by a comprehensive evaluation of the construction results. The study concludes that directional hydraulic fracturing and decompression effectively contribute to the prevention and control of roof-related disasters in the mining of lower coal seams where seam spacing is minimal. This research offers valuable theoretical insights and practical reference for disaster prevention and control in similar geological conditions.

The isolated working face is significantly impacted by the adjacent goaf and the mining activities of the working face itself, causing the overlying rock layers above the working face to exhibit far more intense activity compared to an ordinary working face. The stress levels are high, and the surrounding rock suffers severe damage, posing serious challenges to the safe and efficient extraction of the working face. Improving the service life of the retreating roadway in an isolated working face is a pressing technical issue that coal mining companies must address. Focusing on the characteristics of the strata and mining conditions of the 8213 isolated working face in the Yanjiahe Coal Mine, which features a three-soft coal seam, a combination of field investigation, theoretical analysis, on-site monitoring, and numerical simulation methods was employed. This approach aimed to analyze the fundamental laws of mine pressure behavior in the three-soft coal seam isolated working face as well as the deformation and failure mechanisms of the surrounding rock in the retreating roadway. Using elastic thin plate theory, it was determined that the basic roof periodic fracture step of the 8213 isolated face in the Yanjiahe Coal Mine is approximately 23 m. Field mine pressure monitoring on the 8213 isolated working face revealed that during non-periodic pressure events, the support resistance of the working face generally fluctuated stably below the rated working resistance. When the basic roof collapsed, the average working resistance of the support showed a significant increase with periodic pressure steps ranging from 16 to 27 m and an average of 22 m. Numerical simulations were further used to analyze the changes in stress and the plastic zone of the overlying rock on the 8213 isolated working face, clarifying the mechanism by which instability in the overlying rock structure leads to incidents. This analysis provides theoretical support for the safe mining of isolated working faces.

In order to study the instability characteristics of interlayer rock strata (IRS) in shallow buried close-distance coal seams under insufficient mining areas, based on the background of interval mining under goaf in Nanliang Coal Mine, this paper studies the instability characteristics of interlayer strata in interval mining under goaf by means of similar simulation, numerical simulation, and field measurement. The results indicated that the first weighting interval of the main roof during mining in the lower coal seam was 49 m, while small and large periodic weightings with intervals of 10–14 m and 15–19 m were identified. During periodic weighting, the support resistance ranged from 6813 to 10,935 kN, with a dynamic load factor of 1.07–1.74, and the peak abutment pressure in front of the working face was 5.85–9.85 MPa. The mining under the interval coal pillar (ICP) was the ‘stress increase zone’, and the mining under the temporary coal pillars (TCPs) and the interval goaf was the ‘stress release zone’. During the working face mining out of the ICP, the support resistance reached 10,934 kN, the dynamic load factor reached 1.74, and the abutment pressure (AP) reached 9.85 MPa, which was 60% higher than the AP mining under the “stress release zone”. Analysis suggests that the cutting instability of the IRS was the root cause of the increased AP in the working face of the lower coal seam. A numerical simulation was performed to verify the instability characteristics of the IRS in the interval goaf. The relationship between support strength and roof subsidence during the period of the working face leaving the coal pillar was established. A dynamic pressure prevention method involving pre-splitting and pressure relief of the ICP was proposed and yields superior field application performance. The findings of the study provide a reference for rock strata control during mining under the subcritical mining area in shallow and closely spaced coal seams.

With the increasing depth of coal mining operations, the repeated extraction of multiple coal seams has led to serious safety threats to mines, including secondary roof fracturing, interlayer separation-induced water hazards, and intense mine pressure. Due to the limited research available on the roof failure laws of gently inclined coal seam groups, this study focuses on the Yindonggou Coal Mine and employs a comprehensive approach combining theoretical analysis, numerical simulation, and field measurement. Theoretical calculations indicate that after the mining of Seam 1, the caving zone height ranges from 6.69 to 11.09 m, and the height of the water-conducting fracture zone ranges from 29.59 to 40.79 m. After Coal Seam 2 is mined, the caving zone extends 24.05–33.47 m above the roof of Coal Seam 1, and the fracture zone develops for up to 74.10–94.94 m. Following the mining of Seam 4, the caving zone expands to 30.73–40.15 m above the roof of Coal Seam 1, and the fracture zone reaches 92.26–113.10 m. The numerical simulation results show that after mining Seam 1, the caving zone height is 8.4 m, and the fracture zone reaches 36 m. After Seam 2 is mined, the caving zone extends to 27 m above the roof of Coal Seam 1 and the fracture zone extends to 89 m. After Seam 4 is mined, the caving zone expands to 40 m above the roof of Coal Seam 1 and the fracture zone develops to 112.6 m. The field measurements validate the following findings: a loss of flushing fluid during drilling indicates that after Coal Seam 4 is mined, the fracture zone develops up to 110.5 m above the roof of Coal Seam 1, and the caving zone reaches 47.5 m. Optical imaging logging shows the fracture zone developing to 114.5 m and the caving zone extending to 48.1 m above the roof of Coal Seam 1. The results demonstrate good consistency among these theoretical calculations, numerical simulations, and field measurements. This study reveals a progressive development pattern of roof failure during the repeated mining of gently inclined coal seam groups, providing a theoretical foundation for water hazard prevention and mine pressure control in deep multi-seam mining operations.

Reliable roof-water-inrush risk zoning in coal mines remains difficult under small-sample conditions because geological, hydrogeological, and mining-induced factors interact nonlinearly, and field labels are often derived from engineering evidence rather than complete accident records. This study aims to develop and preliminarily validate a deployment-oriented workflow for case-specific roof-water-inrush risk-zone classification in the No. 4⁻³ coal seam of Shiyangou Coal Mine, Shaanxi Province, China. A total of 100 spatial evaluation samples were compiled using eight indicators grouped into water-resisting capacity, water-supply intensity, water-conducting pathways, mining-induced disturbance, and rock-mass integrity. The Safe-Zone and Risk-Zone labels were generated from mine-water-prevention documents, borehole and hydrogeological evidence, abnormal goaf-water information, and expert engineering judgment. A lightweight CNN-LSTM model was trained with leakage-controlled data splitting, early stopping, L2 regularization, gradient clipping, and model-capacity control. Because the inputs are static or quasi-static spatial indicators, LSTM was used only as a gated feature-dependency module for CNN-derived feature sequences, and its necessity was examined using ablation and feature-order perturbation tests. In repeated stratified five-fold cross-validation, the model achieved mean accuracy, F1-score, and AUC values of 0.792 ± 0.108, 0.633 ± 0.111, and 0.829 ± 0.097, respectively. On an independent prediction dataset, the accuracy and AUC were 0.750 and 0.743, respectively. After ONNX export, STM32CubeMX evaluation indicated 68,962 B Flash and 3252 B RAM consumption. These results indicate the preliminary feasibility of embedded roof-water-inrush risk-zone classification for this specific mine case. However, the findings do not demonstrate cross-mine transferability; further validation using external datasets, independent labels, and field deployment records is required.