Search Results (2,188)

Ensuring the sustainability of deep coal mining requires a comprehensive understanding of hydrogeological risks, particularly fault-induced water inrush, which threatens human safety, depletes freshwater resources, and causes irreversible ecological damage. This study addresses the sustainability gap in managing heterogeneous fault fracture zones by conducting coupled loading–seepage experiments on representative limestone, sandstone, coal, and coal–rock mixtures from the Zhaogu No. 2 Mine. Results demonstrate that seepage behavior follows the Forchheimer non-linear regime (E = 0.2–0.95), deviating significantly from Darcy’s law. We quantified that effective stress induces particle crushing and rearrangement, leading to a drastic porosity reduction (up to 97.52% in coal). Crucially, lithology dictates permeability evolution: coal and mixtures exhibit exponential decay, whereas sandstone and limestone follow quadratic functions. The fractal dimension of particles correlates negatively with permeability (R² > 0.95). These findings provide a quantitative framework for predicting water inrush channels, enabling proactive strategies to prevent catastrophic groundwater loss and ensure the long-term viability of mining operations. This research supports SDG 6 (Clean Water) and SDG 12 (Responsible Consumption and Production) by offering scientific guidance for balancing resource extraction with hydrogeological integrity. Full article

To achieve safe mining and precise gas control in Liuzhuang Coal Mine, the dynamic evolution of mining-induced overburden fractures at the 150,502 working face was systematically investigated via theoretical analysis, FLAC3D numerical simulation, and field measurements. A 3D model (600 m × 300 m × 154 m) was established to simulate the plastic zone, displacement, and stress fields during face advancement from 50 to 400 m. Strata damage modes were evaluated, and the “two zones” heights were determined based on plastic criteria. The results show that fracture development exhibits distinct stages. The plastic zone displays a “spoon-shaped” distribution, with damage more concentrated on the open-off cut side than the working face side. The caving zone height is approximately 12 m, and the maximum fractured zone height reaches 60 m. Based on key strata theory, a 7 m thick fine sandstone layer 24–31 m above the roof acts as the key stratum controlling overburden deformation, offering stable lithological conditions for gas accumulation and borehole integrity. Field monitoring of cross-stripping boreholes demonstrates that the No. 6 drilling site at the 25 m horizon achieves the highest gas extraction concentration of up to 11%, significantly outperforming the 15 m and 20 m horizons. By integrating multiple methods, the optimal horizon for large-diameter directional boreholes is finalized at 24–31 m, providing a reliable scientific basis for efficient gas drainage under contiguous extra-thick coal seam mining conditions. Full article

Coal-mining-induced goaf areas are widely distributed in China and pose potential risks to high-speed railways. Repeated train-induced cyclic loading may further disturb the already weakened ground and aggravate deformation of the subgrade. A study on the goaf ground soil along the Taijiao High-Speed Railway utilized a GDS dynamic triaxial apparatus and controlled variable method to examine how waveforms, cyclic stress ratios (CSRs), effective confining pressures, vibration frequencies and cycles affect soil dynamics. The results show that cumulative plastic strain and residual pore pressure ratio generally tended to stabilize after rapid early development; however, both responses increased markedly within the 2–3 Hz frequency range. Among the investigated variables, the cyclic stress ratio (CSR) exerted the most significant influence and showed an exponential relationship with cumulative plastic strain and residual pore pressure ratio, whereas effective confining pressure produced a nearly linear decreasing trend. A dynamic stress–strain backbone curve was constructed, and by introducing the influence of vibration cycles into the H-D framework, the modified model achieved better agreement with the experimental backbone curves than the conventional H-D model. Furthermore, ABAQUS simulations further demonstrated that train speed and subgrade form significantly influence the distribution of dynamic stress and vertical displacement in goaf ground, with a more severe response in the cutting section. The optimized constitutive model and numerical results provide theoretical support for foundation design, long-term stability assessment and settlement prediction of high-speed railways constructed over goaf ground. Full article

Continuous spatial reconstruction in highly fragmented landscapes remains a persistent challenge in remote sensing and geographic information systems. Standardized products such as MODIS Net Primary Productivity (NPP) provide temporally consistent ecological baselines, but their moderate spatial resolution can obscure abrupt transitions in disturbed environments. This study develops a biophysical index-guided 1 × 1 Convolutional Neural Network (CNN) framework for NPP reconstruction over a mega-scale open-pit mining landscape. The framework uses five Landsat-derived biophysical tensors—Vegetation Moisture Stress Index (VMSI), Anti-Vegetation Environment Index (AVEI), Soil Adjusted Vegetation Environment Index (SAVEI), AAI-VHI, and Tasseled Cap Wetness (WET)—as point-wise predictors aligned to the MODIS NPP baseline grid. By restricting convolutional kernels to 1 × 1, the model performs nonlinear channel-wise mapping without incorporating neighboring grid cells, thereby reducing boundary mixing that can occur in conventional multi-pixel CNNs. Benchmark comparisons with Random Forest and a standard 3 × 3 CNN showed that the 3 × 3 CNN achieved slightly higher global accuracy, whereas the 1 × 1 CNN provided stronger gradient correspondence and lower full-domain spatial error in the 2020 spatial fidelity assessment. The results indicate that point-wise convolution guided by physically interpretable indices provides a conservative and interpretable option for standardized ecological reconstruction at the MODIS baseline grid scale. Full article

Student stress is a significant mental health issue in educational settings; therefore, developing reliable, calibrated, and interpretable predictive models can support the classification of observed stress levels. This study analyzed the public Student Stress Factors dataset, comprising 1100 records, 20 predictors, and one target variable, using a supervised machine learning pipeline designed to reduce information leakage. The pipeline included stratified data partitioning, encapsulated preprocessing, nested cross-validation restricted to the training data, and independent holdout evaluation. Nine ensemble and boosting algorithms for tabular data were compared: AdaBoost, Gradient Boosting, Random Forest, Extra Trees, Bagging, Voting, Stacking, XGBoost, and LightGBM. Model performance was assessed using key discrimination and calibration metrics, together with the nonparametric Friedman test for statistical comparison. Gradient Boosting achieved the best average performance in nested cross-validation, with an accuracy of 89.55 ± 3.16%, F1-weighted of 89.54 ± 3.17%, MCC of 0.845 ± 0.047, and ROC-AUC weighted of 98.59 ± 0.92%. XGBoost and LightGBM showed comparable performance. In the independent holdout set, the final calibrated model maintained robust predictive performance, achieving an accuracy of 0.8818, F1-weighted of 0.8818, MCC of 0.8237, and ROC-AUC weighted of 0.9861. Although the overall results indicate stable and high predictive performance, the Friedman test did not identify statistically significant differences among the algorithms, χ² = 10.953, p = 0.204. Therefore, model selection should consider not only predictive accuracy but also computational efficiency, calibration, interpretability, and implementation feasibility. Despite the internal stability of the pipeline and satisfactory holdout performance, the public and cross-sectional nature of the dataset limits causal inference and model transferability. Consequently, external and prospective validation is required before integration into institutional early warning systems. Full article

Rockburst risk in seismically active mines poses a significant threat to underground safety. This study aims to improve the prediction of rockburst-induced damage by addressing the challenge of class imbalance, which is commonly encountered in rockburst datasets. Various data-balancing techniques, including the Synthetic Minority Oversampling Technique (SMOTE) and its variants, namely SMOTE-Tomek, KM-SMOTE, SMOTE-ENN, SVM-SMOTE, Borderline-SMOTE, and Adaptive Synthetic (ADASYN) sampling, were applied to a dataset containing rockburst damage scales and selected influencing parameters. The data were collected from Canadian and Australian underground mining operations affected by mining-induced seismicity. Three machine learning classifiers, namely Random Forest (RF), CatBoost (CB), and Gradient Boosting (GB), were trained and evaluated using accuracy, recall, precision, and F1-score metrics. The five best-performing models were SMOTE-ENN-GB, SMOTE-ENN-CB, SMOTE-ENN-RF, KM-SMOTE-CB, and Borderline-SMOTE-CB, achieving testing accuracies ranging from 70% to 88%. SHAP analysis further revealed that stress conditions and peak particle velocity (PPV) are the dominant factors controlling rockburst severity, while geological factors and support conditions act as secondary contributing factors. Compared with previous studies using the same dataset, the proposed approach achieved substantial improvements in predictive performance, particularly for the minority and severe rockburst classes. It is concluded that SMOTE-based balancing techniques, when combined with ensemble learning algorithms, can significantly improve rockburst damage prediction and contribute to safer and more effective risk management in deep, seismically active mining environments. Full article

Monitoring the activation of concealed water-conducting structures and predicting the evolution of mining-induced floor failure above a confined aquifer are critical for ensuring the safety and sustainability of deep coal mining. The present study formulates an optimized hydrophobic similar material, improves the bidirectional four-face stress-adjustable loading test platform, integrates water pressure-flow and excitation current monitoring systems, and innovatively introduces the electric current change rate (K value) as a core analytical indicator to systematically conduct physical simulation experiments on floor failure during coal seam mining above a confined aquifer containing concealed water-conducting structures. The results demonstrate the successful development of similar materials with tunable properties (density: 1605–1994 kg·m⁻³; uniaxial compressive strength: 0.07–0.41 MPa; water absorption: 0.2–3%; permeability: 6.8 × 10⁻⁶–7.68 × 10⁻⁴ cm·s⁻¹), effectively replicating the mechanical and seepage characteristics of the prototypical rock strata. The spatiotemporal evolution of the mining-induced fracture field was identified to occur in two distinct stages: “horizontal–vertical evolution” followed by “horizontal periodic evolution”, with a failure depth stabilizing above the No. 9 lower coal seam and a horizontal lag of 4.3–10.1 cm behind the working face. The K value parameter proves highly sensitive in dynamically characterizing the multi-field coupling process of stress–damage–seepage, enabling the clear delineation of the floor’s “six horizontal zones” and “three vertical zones” structure. Crucially, the K value analysis revealed the underlying mechanism of confined water conduction, showing a significant upward migration in the concealed structure area that approached, but did not breach, the key aquifuge layer. The present study provides a novel geophysical perspective and an effective technical parameter (K value) for deciphering the failure mechanism of mining-disturbed coal seam floors, thereby offering a diagnostic framework and a theoretical basis for water hazard early warning and the promotion of green and safe mining practices. Full article

To reveal the stress transfer mechanism of overlying coal pillar loads in hard–soft composite floor strata during close-distance coal seam mining, this study comprehensively employed theoretical analysis, similar material simulation, and numerical simulation to systematically investigate the floor stress distribution characteristics under different pillar widths and rock combinations. This study focuses on the instantaneous elastic response of hard–soft composite floor strata under static coal pillar loading, providing a theoretical foundation for pillar design and roadway layout in multi-seam mining. The limitations and future research directions are also discussed. First, based on the elastic layered half-space theory, mechanical models for stress transfer in the floor under narrow coal pillars (unimodal load) and wide coal pillars (bimodal load) were established. Analytical expressions of stress at any point in the floor were derived, and the influence laws of key parameters, including Poisson’s ratio, interlayer spacing ratio, and shear modulus ratio, were clarified. Second, two typical physical models, namely “hard–soft–hard” and “soft–hard–soft”, were constructed. Experimental results revealed that the weak interlayer exhibits a significant “barrier effect” in the hard–soft–hard combination, causing the stress contours to contract in a “bulb-like” shape; whereas the hard rock layer plays a “bearing effect” in the soft–hard–soft combination, leading to stress contours diffusing in a “gourd-like” shape. Furthermore, numerical simulation revealed the controlling mechanisms of rock combination and thickness ratio: the hard rock layer dominates stress concentration, with the peak stress zone evolving from an “inverted water droplet” shape to a “platform” shape as the thickness increases; the soft rock layer governs stress diffusion and buffering. The depth of the plastic zone significantly decreases with increasing hard rock thickness ratio, achieving a reduction of 44.4%. Full article

Hydraulic fracturing is a key stimulation technique for enhancing the permeability of tight sandstone-hosted uranium deposits. However, existing hydraulic fracture network evaluation methods are primarily applicable to stress regimes in which the vertical principal stress exceeds the horizontal principal stress, making them unsuitable for evaluating low-angle or subhorizontal hydraulic fractures formed under high-horizontal-stress and low-vertical-stress conditions. To address this limitation, this study develops a semi-quantitative method for evaluating hydraulic fracturing effectiveness under stress regimes characterized by high horizontal stress and low vertical stress. The proposed method introduces a Stoneley-wave attenuation index and combines it with Stoneley-wave chevron-apex responses to identify hydraulically fractured intervals. By further integrating conventional well-log data to reduce interference from borehole enlargement, lithological boundaries, and natural fractures, the method supports the identification of hydraulically induced fractures and provides a semi-quantitative assessment of their development. The method was applied to a hydraulic fracturing pilot test for in situ leaching uranium mining in the Bayingobi Basin, Alxa, China, where it supported the identification of hydraulically induced fractures and fractured intervals in both stimulation and monitoring wells. Field application results support the engineering applicability of the proposed method and provide preliminary validation of its effectiveness. The results indicate that this method provides an effective logging-based approach for evaluating hydraulic fracturing performance and investigating fracture propagation in tight sandstone-hosted uranium deposits. Full article

The propagation behavior of hydraulic fractures in deep hard roofs is governed by the multi-factor coupling of injection rate, in situ stress, and fluid viscosity. Taking the Cuimu coal mine as the engineering background, this study systematically investigates the effects of injection rate, lateral pressure coefficient, and fluid viscosity on fracture propagation through numerical simulation. Theoretical derivations based on the KGD (Kristianovich-Geertsma-de Klerk) model are further integrated and validated by field tests. The results indicate a critical injection rate of 6 × 10⁻⁸ m³/s, above which the marginal increase in acoustic emission events declines significantly. Increasing the lateral pressure coefficient from 1.0 to 3.5 shifts the fracture pattern from relatively simple to increasingly complex and interwoven, accompanied by a logarithmic increase in fractal dimension from 1.38 to 1.80. The total acoustic emission count rises to a peak of 125,910 as viscosity increases from 0.001 Pa·s to 0.5 Pa·s, then drops to 44,061 at 1.0 Pa·s, showing a unimodal trend. Theoretical analysis shows that during the propagation stage, the fracture length follows $L\propto Q^{1/2}$, and the maximum fracture opening follows $w_{\max }\propto Q^{1/3}$. The lateral pressure coefficient controls the complexity of the fracture network through the directional distribution of stress intensity factors. Field tests at the Cuimu coal mine adopted a combination of stepwise injection rate and low-viscosity fluid, together with borehole densification and interval-skipping fracturing sequences. The effective fracturing radius reached 25~30 m, roof convergence was reduced by 31%, and the proportion of high-energy microseismic events decreased from 12% to 4%. This study establishes a complete theoretical framework from initiation theory to propagation dynamics and then to multi-crack competition, providing both a theoretical basis and engineering example for optimizing fracturing parameters in hard roofs under high stress anisotropy. Full article

Mining-induced land subsidence poses significant geohazard risks to critical on-site mining operational support infrastructure, such as tailings storage facilities (TSFs). This study investigates the Doornpoort TSF subsidence at the South Deep gold mine in South Africa, using multi-temporal Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) and Non-negative Matrix Factorisation (NMF) algorithm approach to split the superimposed subsidence contributing drivers, alongside the incorporation of Global Navigation Satellite System (GNSS) data and underground mining layout plans. 78 Sentinel-1A Satellite Aperture Radar (SAR) ascending acquisitions between May 2022 and December 2024 were obtained and processed to determine the average annual deformation rates and cumulative time-series displacement for the study area. The InSAR-derived subsidence rates at the designated three benchmarks on the embankments of the Doornpoort TSF and TSF 1&2 are −26.09 mm/year, −13.40 mm/year and −16.25 mm/year, while the maximum cumulative subsidence was −57.45 mm, −42.76 mm and −36.44 mm. A comparison of the InSAR results with the GNSS-derived subsidence results showed correlation standard deviations of 1.87 mm, 0.93 mm, and 1.05 mm, respectively. The InSAR results revealed spatially coherent subsidence patterns and a good correlation between deformation boundaries and underground mining layouts, suggesting that mining-induced stress redistribution is the primary driver of regional surface subsidence. The NMF decomposition of the InSAR-derived deformation result at a selected benchmark on the Doornpoort TSF embankment, whose average annual deformation with a cumulative time series deformation of −26.09 mm/year and −57.45 mm, respectively, revealed that 92% of the observed cumulative deformation is associated directly with the underground mining, whilst the remaining 8% is associated with TSF embankment consolidation. Furthermore, the selected decomposition benchmark within the TSF basin showed that underground mining alone accounted for 100% of the observed subsidence there. These findings support a coupled deformation framework in which deep mining activities influence regional subsidence, while localised geological conditions modulate its surface manifestation. Full article

Addressing the severe risks of floor instability and leakage in underground coal mine reservoirs in Western China under coupled mining and hydraulic pressures, this study developed a fully coupled stress-damage-seepage numerical model. Incorporating rock heterogeneity based on the geological conditions of the Shendong mining area, the model systematically simulates the evolution of floor damage under varying storage pressures, burial depths, mining heights, and lithologies. The simulation results demonstrate that storage pressure is the primary driver for deep damage propagation via a “hydraulic wedging” mechanism, governed by a critical activation threshold of 1.0 MPa. Specifically, before the water storage pressure reaches 1.0 MPa, the floor damage remains dormant and basically unchanged, stagnating at a shallow level. However, once the pressure exceeds this 1.0 MPa threshold, it overcomes the effective confining stress, abruptly shifting the failure mode from shallow discrete fracturing to deep penetrating failure, which is accompanied by an order-of-magnitude surge in permeability. Furthermore, a dimensionless sensitivity analysis reveals that burial depth and lithology strongly govern the failure path and depth. Notably, mudstone possesses a significantly lower intrinsic permeability, and even when subjected to damage, its water barrier performance remains superior to that of sandstone because its localized plastic shear characteristics highly restrict permeability mutations. In contrast, brittle sandstone is highly susceptible to tensile cracking and the formation of deep penetrating seepage channels. Additionally, mining height demonstrates weak sensitivity to the depth of floor damage due to an “equivalent unloading” mechanism, which validates the technical feasibility of constructing underground water reservoirs in ultra-thick coal seams. These findings provide a vital theoretical foundation for the scientific site selection of underground reservoirs and the precise determination of operational water level thresholds to ensure long-term stability. Full article

Rock-mass instability in deep mining is a progressive process controlled by stress redistribution, fracture evolution, and excavation disturbance. Conventional risk assessment mainly identifies high-risk zones according to risk magnitude, but provides limited information on risk-transition boundaries and propagation directions. This study proposes a risk–gradient field method for identifying the propagation direction and migration path of rock-mass instability. An interpretable instability probability function was first established using the sure independence screening and sparsifying operator (SISSO), and the predicted probabilities were mapped into a three-dimensional probabilistic risk field. On this basis, the spatial risk–gradient magnitude and direction were calculated to identify high-gradient boundaries, potential risk fronts, and dominant propagation directions. The results show that the probabilistic risk field and risk–gradient field provide complementary information: the former indicates where instability risk accumulates, whereas the latter reveals where risk changes sharply. High-gradient zones were mainly concentrated near the margins of high-risk regions, suggesting that they represent transition zones rather than high-risk cores. Validation using subsequent microseismic activity showed that 72.53% of later events and 63.26% of released energy occurred within the high-gradient buffer zone, with an angular deviation of 37.9° between the gradient-centroid migration direction and the microseismic energy-centroid migration direction. These results indicate that the proposed method can extend rock-mass instability assessment from probability-based zoning to direction-oriented risk-evolution diagnosis, providing a useful basis for identifying potential instability migration zones in deep mining. Full article

Hybrid energy storage systems combining LiFePO₄ batteries and supercapacitors can reduce high-rate battery loading in battery electric mining dump trucks operating under intensive regenerative braking conditions. This study proposes a constrained multi-objective sizing methodology for a semi-active battery–supercapacitor hybrid energy storage system applied to a 65 t payload-class mining electric dump truck. The model combines segment-level mining duty cycles, longitudinal vehicle dynamics, a first-order Thevenin battery representation, a usable supercapacitor energy window, bidirectional DC/DC converter limits, and constrained supervisory power splitting. Three mining duty cycles are considered: production haulage, reclamation/backfill operation, and mixed operation. The final sizing result is reported using a dimensionless usable energy buffer ratio rather than a direct comparison between supercapacitor capacitance and battery energy capacity. The results show that the required supercapacitor buffer is strongly duty cycle-dependent. For the regenerative-dominant backfill cycle, the hybrid configuration reduced peak battery charging current from approximately −950 A to −180 … −280 A and reduced battery root mean square (RMS) current by 52–64% relative to the pure battery configuration. The constrained stored fraction of regenerative energy also increased when the supercapacitor branch was included, while non-accepted braking power was assigned to the residual braking channel. The proposed approach provides a physically consistent basis for preliminary hybrid energy storage system (HESS) sizing and clarifies that battery current reduction should be interpreted as a degradation-relevant stress indicator rather than as a direct quantified lifetime prediction. Full article

The intrusion of groundwater inevitably alters the environment within the caving zones of mined-out areas and affects the mechanical properties of the surrounding rock mass, thereby influencing surface deformation. To investigate the deformation mechanisms of overburden and surface under the influence of groundwater in caving zones, as well as the deformation patterns of abandoned mined-out areas under different geological and mining conditions, this study takes a coal mine in Jiangsu Province, China, as a case study. Based on FLAC3D (version 6.00) numerical simulations, the study analyzes the mechanisms and main influencing factors of overburden and surface movement in abandoned goafs under groundwater conditions. It further examines the deformation patterns under groundwater intrusion and explores the effects of different mining thicknesses and weakening of caving zone parameters on surface deformation. The results show that the weakening of mechanical properties leads to subsidence of the overburden and surface, while increased pore water pressure reduces effective stress, causing initial subsidence followed by uplift. Moreover, under varying conditions of coal seam thickness and weakening intensity of the caving zone, surface deformation in response to groundwater in abandoned goafs exhibits distinct patterns. These findings provide a theoretical basis for scientifically interpreting the propagation mechanisms and movement characteristics of surface deformation in abandoned mines under hydro-rock interaction. Full article