To meet the demands for efficient excavation in deep coal mine roadways, Tunnel Boring Machines (TBM) have become a pivotal solution due to their safety, high efficiency, and environmental benefits, enabling synchronized multi-stage operations. However, as the primary rock-breaking component, disc cutters face significant service-life challenges when operating in the complex conditions of coal-bearing strata. Consequently, investigating cutter load, rock-breaking efficiency, and tool wear is of paramount theoretical and practical significance for optimizing cutter designs. Drawing on a large-scale coal mine TBM construction project, this study analyzes the mechanical characteristics of cutters during excavation using the Mohr-Coulomb failure criterion. A theoretical model for the cutting forces of circular-edged disc cutters is established and validated using field data. Utilizing ABAQUS finite element software, the mechanical response of the cutter is investigated across varying cutting depths. The results indicate that circular-edged cutters are more conducive to the initiation and propagation of lateral cracks. Analysis of cutting forces and Specific Energy (SE) at various rock temperatures further demonstrates that thermally assisted rock breaking significantly enhances excavation efficiency and extends tool life. Finally, a comparative validation between the mathematical model, numerical simulations, and field measurements shows that the error between the simulation results and theoretical predictions is within 10
Accurate determination of cross-sectional mean wind speed is essential for mine ventilation network calculation and air velocity sensor placement. This study proposes a characteristic-point method based on the mean wind speed contour. Simulations of rectangular, trapezoidal, and straight-wall semicircular-arch roadways were conducted to examine the effects of inlet wind speed, cross-sectional geometry, and wall roughness on the mean-wind-speed point position, characterized by c and c/de. A random forest model was used to identify the main factor controlling airflow development length, and a double-exponential function was established to determine the point position. Results show that inlet wind speed has little effect under fully developed airflow conditions, whereas cross-sectional geometry and wall roughness significantly influence the point position. Across a large number of random cases, the coefficients of variation of c and c/de were both 0.36
In coal-gas overlay areas, coal mining causes surface subsidence due to strata movement, damaging buried pipelines. Using the Xinjie mining area as background, this study links the primary key stratum (PKS), surface, and pipeline (PSP) via theoretical and numerical methods. Monitoring lines are set for the PKS, surface, and pipeline, and FLAC3D simulations are conducted. Results show: the PKS has the largest subsidence (1.87 m), the surface the strongest horizontal sliding (0.57 m). Minimum differential subsidence between pipeline and surface is 3.846
Growing interest in expanding mining operations in the Arctic region has raised concerns about its potential environmental impact. Indeed, xanthates commonly used as flotation collectors of sulfide minerals are also among chemicals of which persistence and toxic effects on the environment can worsen in a sensitive ecosystem such as the Arctic. This article investigated the effects of some physical and chemical variables on the decomposition of xanthates. Temperature revealed to be the key variable given that considered effects have not yet been extensively studied. The decomposition of potassium ethyl xanthate (KEX) was studied first using statistical experimental design to screen the effects of time, temperature, initial KEX concentration, and sulfate concentration. Due to its most significant effect, temperature was studied in more detail in ultrapure water and real samples from a mine site using KEX and sodium isobutyl xanthate (SIBX). The results show that decreasing the temperature increased the half-lives of the studied xanthates. In most cases, the real sample matrix did not significantly affect decomposition rates compared to ultrapure water, except for a sample from a clarified water sump where notable differences were observed: The decomposition of SIBX remained temperature-independent and faster than KEX. The present study indicates that the effect of temperature should be considered when mining activities are done in the Arctic region. This study presents for the first time that temperature significantly decreases the decomposition rate of xanthates in real case study samples.
To address the critical production constraints posed by coal wall spalling, roof caving, and support crushing—triggered by the fracturing-induced impact of thick-hard rock groups (TRG) in ultra-large mining height working faces (UMHWF)—this study first analyzed the characteristics of strong mine pressure from TRG fracturing and the fracturing-induced instability structures in UMHWF. It was found that UMHWF with multi-layer TRG exhibit distinct strata behavior characteristics: “rapid resistance growth, high static load, multiple dynamic loads, and high-frequency cyclic loads”. Additionally, roof fracturing forms a structural configuration of “low cantilever beam + middle-upper stepped rock beam + high stable masonry beam”, where fracturing of the middle-upper TRG (at 40–60 m depth) acts as the key stratum governing strong mine pressure. Theoretically, a mechanical model was established to characterize the load-bearing behavior of TRG in UMHWF, encompassing short beams (cantilever segment) and semi-infinite elastic foundation beams (embedded segment). This model enabled the derivation of the mining-induced load-bearing state and energy expression of TRG prior to periodic strata weighting. Based on the law of conservation of energy, the dynamic and static load expression for hydraulic supports in the working face was further derived; substituting on-site parameters yielded a maximum working resistance of 32019.7 kN acting on the supports. This value was consistent with the on-site measured working resistance of supports during strata weighting periods, thereby verifying the theoretical validity to a reasonable extent. Building on this theoretical framework, the control variable method was employed to investigate the influencing factors governing periodic strong mine pressure in TRG. The results indicated that three factors exert the most significant effects on the total dynamic and static loads acting on working face supports during the fracturing of middle-upper TRG: equivalent immediate roof (EIR) thickness, overhanging length of the cantilever segment of TRG, and thickness of the middle-upper TRG. On-site, support-unloading coordination measures were implemented, and a “site-specific strategy” zone-fracturing concept was proposed for UMHWF with ultra-long advance distances in TRG. This approach effectively reduced the initial roof weighting interval and weighting intensity, improved the operational performance of working face supports, and ultimately ensured the safe and efficient extraction of the working face.
Nickel is critical for stainless steel production, batteries, and next-generation clean energy technologies; its sustainable recovery is becoming increasingly important as primary ore grades decline globally. Once considered waste, mine tailings are now recognized as a valuable secondary resource due to their residual nickel content. The main nickel minerals in the tailings sample used in this study are pentlandite and pyrrhotite, both of which are sulfide minerals. However, because mine tailings are left in open pits for long periods, surface oxidation and hydration occur, making sulfide mineral surfaces more hydrophilic and therefore more difficult to float. This study investigated advanced flotation strategies for recovering nickel from mine tailings and low-grade nickel ores. Bench-scale flotation experiments were conducted to evaluate activation flotation, activation–sulfidization flotation, mixed-collector flotation, and CO₂-assisted flotation. The addition of (NH₄)₂SO₄ as a secondary activator improved flotation selectivity, likely by dispersing slime coatings and exposing active sulfide sites. The use of AERO OX 100 as a co-collector with sodium isobutyl xanthate (SIBX) further enhanced nickel recovery, particularly during scavenger flotation. CO₂ flotation also improved flotation performance under optimized activation–sulfidization conditions, increasing the concentrate Ni grade from 0.92
Dolomite is an important raw material for refractory and steelmaking industries; however, large quantities of low-grade dolomite fines are discarded as mine waste because of their high silica content. The present study presents a systematic mineralogy-guided beneficiation strategy for valorizing low-grade dolomite fines by integrating detailed characterization with process flowsheet development. The scientific novelty lies in establishing the relationship between mineral liberation characteristics and the selection of an optimum beneficiation route through the comparative evaluation of three alternative flowsheets: (i) gravity separation-flotation, (ii) dry air classification-flotation, and (iii) two-stage reverse-direct flotation. Mineralogical investigations revealed quartz as the dominant siliceous gangue, with significant interlocking that necessitated controlled fine grinding for effective liberation. Although the hybrid dry classification-flotation route reduced water consumption and produced a concentrate containing approximately 4.2
The Takhte-Gonbad porphyry copper deposit, situated within the prolific Kerman porphyry copper belt of southeastern Iran, was investigated through an integrated geophysical survey comprising electrical resistivity tomography (ERT), induced polarization (IP), and magnetics. This study employed a novel inversion scheme, processing two-dimensional geoelectric data alongside three-dimensional magnetic data using a mixed-norm approach. The inversion results delineated two principal east–west trending Cu-related anomalous zones characterized by elevated chargeability responses and variable resistivity signatures associated with hydrothermal alteration and sulfide mineralization. In several areas, relatively high resistivity values were interpreted to reflect silicification and quartz-rich alteration assemblages. The 3D magnetic inversion results revealed high susceptibility zones associated with magnetite-bearing domains and locally reduced copper enrichment. To validate the geophysical interpretation, borehole data were integrated to construct block models of copper grade, lithology, and alteration using ordinary kriging. The geological models confirmed economically significant mineralization, with copper grades locally exceeding 1
Dust pollution in surface mining poses significant environmental and occupational health challenges and conventional water-based suppression methods offer only limited and short-term effectiveness. This study evaluates the performance of polymer-based dust suppressants in reducing particulate matter (PM) concentrations under controlled laboratory conditions. Polymer blend ratios as 25:75, 30:70 and 35:65 were tested and compared to traditional water spraying. The polymer formulations achieved substantially higher reductions across all PM size fractions (PM₁₀.₀, PM₄.₀, PM₂.₅ and PM₁.₀) and with the greatest suppression observed for the coarser particles. Among the tested blends, the 30:70 blend ratio demonstrated the highest overall efficiency, outperforming both lower and higher concentrations. This indicates the importance of achieving an optimal dosage rather than relying on maximum polymer content. The enhanced performance is attributed to particle agglomeration, surface binding and crust formation, which minimize re-entrainment of fine particulates. Statistical validation using paired t-tests and one-way ANOVA confirmed that the observed reductions were significant and highly reproducible (p < 0.05). These findings highlight the potential of polymer-based suppressants as a sustainable and efficient alternative to water-intensive methods for dust control in mining operations. The study underscores the need for further field-scale research on long-term environmental safety, soil interaction and integration with real-time monitoring technologies. Overall, polymer-based dust suppression offers a promising pathway toward cleaner, safer and more sustainable surface mining practices.
The safety problems of underground confined space have become increasingly prominent. Nevertheless, due to the characteristics of low illumination, narrow and complex spatial structures, and weak Global Navigation Satellite System (GNSS) signals in underground confined space such as coal mine roadways and underground tunnels, it’s difficult to achieve accurate 3D reconstruction of underground confined space for traditional solutions and visual Simultaneous Localization and Mapping (SLAM). Furthermore, the inner surfaces of most underground confined space commonly consist of concrete structures characterized by a limited number of structural features that closely resemble each other. Laser SLAM solutions are easily degraded in such environments. In this paper, we propose a robust SLAM method based on non-repetitive scanning Light Detection and Ranging (LiDAR). This method only uses non-repetitive scanning LiDAR as the sole sensor. Initially, point cloud acquired through scanning undergoes filtering to eliminate distortions and noise, followed by feature extraction from the filtered point cloud. Secondly, the voxel mapping method is used to match point cloud features and obtain rough odometry. Finally, by performing the Bundle Adjustment (BA) method and loop closure optimization based on the scan context method, accurate odometry is obtained, achieving a 3D accurate reconstruction of the underground confined space. Extensive experiments conducted within various typical underground confined space scenarios demonstrate the superior robustness of the proposed method compared to LOAM-Livox, Livox-Mapping, and BALM methods. This method provides technical support for on-site applications such as 3D accurate reconstruction and intelligent inspection of underground confined space, as well as underground emergency rescue.
During steel manufacturing, the presence of Zn reduces product quality and increases production costs. Therefore, understanding the mineralogical distribution of Zn is essential to evaluate potential removal treatments. A mineralogical characterization of a Zn-contaminated Fe concentrate was carried out using microscopy and X-ray diffraction, identifying magnetite as the dominant Fe-bearing mineral. Zinc was identified as native zinc, willemite, zinc in silicates, sphalerite, and zinccopperite, a rarely reported Cu–Zn intermetallic phase. Native zinc and zinc hosted in silicates were associated with magnetite; sphalerite occurred as inclusions within magnetite, whereas willemite and the zinccopperite particles directly observed by microscopy were predominantly liberated. Selective dissolution studies revealed an additional fraction of insoluble zinc-bearing phases occluded within the magnetite matrix. These results demonstrate that sulfide species and zinccopperite are mostly occluded, making their removal difficult through magnetic concentration processes.
Rapid assessment of underground conditions is critical after a mine disaster, yet direct entry into damaged roadways can delay rescue and expose personnel to secondary hazards. This study presents an offline reconstruction workflow that combines an unmanned aerial vehicle with a red-green-blue and depth camera and evaluates it in a 10 m simulated disaster tunnel. Color images were enhanced with a halo-suppressed dark channel prior method that uses quad-tree atmospheric-light estimation and homomorphic transmission refinement. Missing depth values were completed with a luminance-guided joint bilateral filter accelerated by a truncated fast Gauss transform. On four Middlebury scenes, the depth method reduced mean processing time by about 45
Studying the performance of dry-type exhaust system for explosion-proof diesel engines is critical for ensuring safe operation and emission compliance under complex conditions. Based an explosion-proof diesel engine test bench with three exhaust configurations, the engine performance, emission characteristics, and heat transfer characteristics are systematically studied: System A represents a conventional exhaust pipe, System B incorporates a heat exchanger, flame arrester, and spray device to form a dry cooling system, and System C further includes a diesel oxidation catalyst (DOC) and a selective catalytic reduction particulate filter (SDPF). In addition, machine learning prediction and correlation analysis are employed to evaluate parameter coupling relationships and model performance. Results indicate that the three systems have limited effects on engine power and torque. However, System C exhibits higher exhaust back pressure and exhaust temperature, while reducing brake specific fuel consumption (BSFC). Compared with System A, System C reduces the instantaneous total emissions of NOx, CO2, CO, and HC by approximately 96.1
Recovering nickel from iron-rich mine tailings by bioleaching is held back by the co-dissolution of iron, which gives pregnant leach solutions with high Fe/Ni mass ratios and heavy downstream purification costs. This study shows that running bioleaching at 0 °C under a CO2-enriched atmosphere, with no post-leach iron removal step, lowers the Fe/Ni mass ratio in the leachate from 13.87 at 25 °C to 2.25 at 0 °C, a 6.2-fold improvement against a feed Fe/Ni ratio of 119:1. Nickel recovery stayed near 45
In this paper, the challenges associated with the assessment of environmental and human health risk for two high sulfur tailings (HST), produced after flotation of sulfide ores in Finland, are highlighted and discussed. First, chemical and mineralogical analyses of the tailings are performed using an X-ray fluorescence energy dispersive spectrometer (XRF-EDS) and an X-ray diffractometer (XRD). For the risk assessment study the following tests are used (i) the EN 12457-2 and Toxicity Characteristics Leaching Procedure (TCLP) which assess the leaching rate of potentially harmful elements (PHEs) present in the tailings, (ii) the pH dependent test which assesses the leaching behavior of the main elements over a wide pH range and (iii) the sequential extraction procedure (SEP) which determines the exchangeable and other fractions of PHEs. For the assessment of human health risk the in vitro Physiologically Based Extraction Test (PBET) that simulates the gastric and intestinal tract and determines the bioaccessible fraction of PHEs that are readily available for uptake by humans through ingestion is used. Health risk indicators including carcinogenic risk (CR) and hazard quotient (HQ) are also calculated. The results indicate that (i) the environmental risk for most PHEs is relatively low due to the implementation of well-designed tailings disposal practices at both sites, and (ii) CR slightly higher than 10− 4 and HQ less than 1 are calculated under the worse-case scenario, indicating no actual adverse effects for adults and children.
Open-pit mining operations are often plagued by mill feed variability and economic loss as a result of the difference between the static block models and the actual ore grade. Traditional truck dispatch systems are based on heuristic rules or open-loop optimization and cannot adjust for real-time grade fluctuations that are recognized during the extraction process. This paper proposes a novel closed-loop Digital Twin (DT) framework for dynamic ore blending and dispatch optimization by integrating Internet of Things (IoT) sensor streams with Deep Reinforcement Learning (DRL). We propose a Residual Bayesian LSTM module to predict ore grades with uncertainty quantification from noisy sensor telemetry, enabling high-fidelity virtualization of the mining process. This digital twin is used to feed a Proximal Policy Optimization (PPO) agent, which is trained using a multi-objective curriculum to make real-time dispatch decisions that balance production throughput and strict blending quality constraints. Validated on the MineLib benchmark under realistic sensor-noise scenarios, our framework outperforms industry-standard heuristics and open-loop Deep Q-Networks (DDQN). Experimental results indicate that the proposed system achieves a 22.3
Large volumes of electrohydraulic control data generated at intelligent longwall faces remain underutilised for roof-dynamics management, periodic weighting prediction, and the predictive maintenance of hydraulic supports. To address this limitation, this study proposes a dual-stage prediction method that combines mathematical statistics with deep learning and develops an integrated platform for monitoring and early warning of roof dynamics and hydraulic support performance. At the core of the method is a dual-stage MS–LSTM architecture. In the first stage, a mathematical-statistical (MS) module extracts periodic features from historical pressure data and estimates a range for the location of the next periodic weighting event, thereby providing a coarse prediction. In the second stage, a long short-term memory (LSTM) network continuously tracks pressure trends. Once the longwall face enters the time window defined by the coarse prediction, the monitoring frequency and warning sensitivity are increased to provide a fine-grained prediction. By constraining the temporal search window, this mechanism narrows the prediction space and limits cumulative error. To assess hydraulic support performance, the platform employs a two-dimensional pressure–time matrix to derive dynamic threshold intervals from recursively updated statistical features within rolling windows. Combined with temporal analysis of support operating sequences, this approach enables real time health assessment and anomaly localisation. The platform was deployed across 12 intelligent longwall faces at Hongliulin Coal Mine and validated using three months of periodic-weighting data from the 35201 longwall face. The MS–LSTM model achieved a hit rate of 85.71
Area wide pillar failures, or “squeezes,” were once fairly common in US underground coal mines. While rare today, they do occur. This paper describes nine case histories that were investigated by MSHA Technical Support. They occurred across the range of US coalfields, in both longwall and room-and-pillar mines, and involved development mining, pillar recovery, longwall panel extraction, and multiple seam interactions. The analysis showed that just three of the nine squeezes were “false negatives” that occurred where the pillar stability factor exceeded the accepted design criteria. The observed mechanism of failure in these cases was sliding along bedding planes, highlighting the importance of their frictional strength in developing pillar confinement. The six other squeezes resulted from a variety of causes, including unanticipated multiple seam interactions and as-built pillar dimensions that differed from the design values.
Combined with the laws of roof fracture and structural evolution associated with rock burst hazards section fully-mechanized caving mining face of steep-thick coal seams, it is concluded that the impact load induced by structural instability of the “cantilever beam” in overlying strata is the key structural dynamic factor triggering rock bursts. Accordingly, the roof strata at the working face are simplified as a cracked “cantilever beam”, and a fracture mechanics model is established to describe “cantilever beam” stability controlled by structural planes. Equations are derived for the equivalent stress intensity factor at structural planes and the ultimate overhang length of the “cantilever beam”. The influencing factors of roof stability were analyzed. The results indicate that both concentrated stress and stress intensity factor at the crack tip increase with working face advancement. When the stress intensity factor reaches a critical value, the “cantilever beam” fractures, triggering rotational instability of the rock mass at the working face. Sufficient condition for rock beam instability: the length of branch cracks equals the thickness of the fracture protection layer; necessary condition: crack activation. As rock dip angle and broken overburden load increase, the unstable zone of the “cantilever beam” expands, while the ultimate length decreases (by approximately 12 m and 18 m, respectively). As structural plane length and rock thickness increase, the ultimate length rises (by approximately 10 m and 14 m, respectively), enhancing rock beam stability. The influence degree of each parameter on the ultimate length of the rock beam is as follows: overburden load q > structural surface length a > rock dip angle α > rock beam thickness h. Theoretical results are validated with an engineering example. Theoretical calculations showed that the limit length of the “cantilever beam” rock layer in the No. 412 working face is 14.9 m. A preventive measure of roof-cutting blasting is proposed, which reduced internal stress in the coal mass at the working face side by 66.9
Intelligent and autonomous systems are increasingly essential for decision-making in complex mining operations. This study proposes a simulation-based intelligent framework for short-term open-pit mine planning by integrating Agent-Based Discrete Event Simulation (ABDES) with Deep Reinforcement Learning (DRL). The framework incorporates a Deep Q-Network (DQN)-based Intelligent Supervisory Agent (ISA) that dynamically manages truck dispatching, shovel allocation, and mining block sequencing within a stochastic environment. The model is implemented in AnyLogic and validated using a real-world bauxite mine case study across four scenarios with increasing operational complexity, including uncertainty and equipment failures. The ISA continuously learns from system feedback to adapt its decisions in real time, improving coordination between mining and processing operations. Results show that the proposed framework enhances production stability, maintains consistent ore grade, optimizes stockpile utilization, and improves equipment synchronization. The ISA-enabled system achieves superior economic performance, with a maximum cumulative cash flow of 840.9 million over 104 weeks under the most complex scenario. The proposed approach provides a scalable and practical solution for bridging short-term scheduling and real-time operational control, enabling more resilient and value-driven mining operations.
