Search Documents

By S. W. Thomas

One of the most difficult tasks associated with placer mining is the sampling of the deposit. More placer projects have failed due to inaccurate assessment of the reserves than to any other reason. Within the realm of placers, those containing valuable minerals with a high unit value .are more difficult to sample than those with larger bulk, lower unit value minerals. Some items to consider when sampling a placer deposit are as follows: A) A relatively large size sample is needed for accurate valuation of the ground being tested. Placers are composed of many sizes of gravel which makes a representative sample difficult to obtain. B) When sampling placers for high unit value minerals such as gold, any error in mineral content of the sample will be magnified in the calculation of reserves.

Jan 1, 1985

By Edmund W. Murphy

At the first workshop on Surface Subsidence Due to Underground Mining (1 981 ) , Yarbrough delivered a paper (1 ) describing the mine subsidence insurance program in Illinois operated as the Illinois Mine Subsidence Insurance Fund (IMSIF) . In this paper we will present claims data and describe changes that have occurred in the Illinois Insurance Code during the first six years the IMSIF has been in existence. The change in the original Illinois Public Act 80-1 41 3 that most affected the public was an amendment increasing the maximum amount of mine subsidence insurance avail- able on a structure from $50,000 to $100,000. Seventy-two counties in Illinois have been undermined for minerals, primarily bituminous coal. In Franklin County (.in the southern part of the state) 36 percent of the land area has been undermined for coal extraction. In St. Clair and Madison Counties , two counties that include the St. Louis Metropolitan area along the western border of the state, 10 percent and 7 percent of the land surface have been undermined. Most of the undermined area in the two counties is beneath urbanized land. In 34 counties, approximately 1 percent or more of the land surface has been undermined (2). The first shaft mine in Illinois was sunk in the city of Belleville (east of St. Louis) in 1842, and the mines of early vintage have initiated the majority of the mine subsidence claims. Data compiled from the first six years reveals that most Illinois mine subsidence events that affect structures occur over abandoned mines where early methods of high- extraction mining practices were used. Maps of these mines operated in the 1800s reveal irregular development of room and pillar areas and poorly defined production areas (panels). Rooms in the mines varied from 15 to 40 feet in width and the pillars varied in width from 5 to 40 feet. The pillars were often long and narrow. Recovery rates in mines that operated prior to 1900 generally averaged 50 per- cent, and extraction rates as high as 80 percent have been documented (2). Some pillar robbing occurred.

Jan 1, 1986

By Fernando Pendás, Jorge Loredo, Alumudena Ordóñez

Water derived from mine-tailings typically contains high concentrations of H+, SO4, Fe and other metals, and it frequently suffers a decline in pH and the development of acidic mine drainage (AMD). A paper describing the results of a series of laboratory experiments with real AMD and a subsequent field trial with a in situ pilot system to assess a potential method for remediating tailings-derived water is presented here. After testing different organic substrates, residence times and flow conditions during two years to determine the most appropriate treatment for a polluted Spanish AMD, a pilot scale system combining SAPS, ALD, oxidation/sedimentation ponds and an anaerobic wetland was designed. Preliminary analyses of effluent samples indicate acidity and metal removal is occurring within the system, although sulphate removal is not complete. The results of this small-scale field trial suggest a potential cheap, long-term alternative for remediation of acid mine drainage.

Jan 1, 2000

By V. Galla

In the present paper, two large diameter columns 2 m tall and 1.3 m diameter filled with crushed (finer) and run off mine (coarser) heap leach materials respectively were used to perform unsaturated flow studies. Instruments measuring moisture content and matric suction were placed in the columns at different depths as water was applied at calibrated flow rates from the top of the columns. The outflows from the columns were recorded using tipping bucket rain gauge placed at the bottom of the columns. Unsaturated hydraulic parameters were also obtained for the heap leach materials, scalped (passing 4.75 mm sieve size) and whole (passing 37.5 mm sieve size) sample. The measurements obtained from the columns were later compared with unsaturated flow modeling results. The comparison indicated sensitivity towards the sample size selection for unsaturated hydraulic parameters, and initial conditions considered in the model.

Jan 1, 2009

By L. Xu, N. I. Jayaraman, R. Bhaskar

The objective of this study is to develop a better understanding of the collection characteristics of dust measurement instruments used in mine dust control research with respect to properties of quartz and coal dust. Coal and quartz dust of narrow sues, with aerodynamic median diameters being approximately 2 µm, 4 µm and 10 µm, with quartz contents of l0%, 20%, 30%, 40% and 50%, were released in an aerosol characterization chamber using a fluidized bed aerosol generator. Dust was collected in MSA personal gravimetric samplers with l0-mm nylon cyclones, total dust samplers and Sierra 298 inertial cascade impactors. The samples were analyzed for mass and quartz content. A total of 21 experiments were performed for this study. The mass and quartz collection efficiencies of the total and respirable dust samplers and the impactor were determined. The data show that the amount of dust collected by the respirable dust sampler (respirable + oversize dust trapped in the cyclone pre-classifier) was, in general, less than the amount collected by the total dust sampler. The amount of dust collected by the impactor was higher than that collected by the total dust sampler. Quartz analyses of the airborne dust samples showed that the total dust and respirable dust samplers (including oversize dust) collect less quartz dust than at inertial impactor. The larger the median particle sire of the quartz and coal mixture, the lower the quart content of the dust collected in the samplers. For all three sires, the impactor was a better collector of quartz than the other instruments.

Jan 1, 1994

By Warnie Flint

According to statistics compiled by the National Safety Council, US Bureau of Mines, American Medical Association, and other agencies, accidental injuries cause more deaths than all the leading diseases. No one is immune to an accident. Probably the best way to describe an accident is to say that it is an unexpected happening. The automobile driver is not expecting a terrifying crash, the electrician does not expect to be electrocuted, nor is the coal miner expecting the roof to fall on him. All personal injuries will not result in death. There is a multitude of victims who are living proof of what accidents can mean to an individual-victims who are blind, paralytic, missing arms, legs, etc. In most instances the earning power of these individuals is gone, and they are utterly dependent upon family or welfare. For many years there seemed to be an attitude that prevailed in industrial America that accidents were inevitable, and that so many fatalities and injuries should be expected. This, of course, was utter nonsense. Since the birth of such organizations as The National Safety Council, American Society of Safety Engineers, and. the US Bureau of Mines, much progress has been made in reducing accidents. The enactment of compensation laws greatly contributed to the progress that has been made. Does this mean that all accidents can be prevented? Unfortunately, it does not. Human psychology indicates that men will work unsafely or perform unsafe acts for the following reasons : 1) Lack of knowledge or skill. 2) Conflicting motivations, such as: (a) to be more comfortable, (b) to save time or effort, (c) to be accepted by a group. 3) Physical or mental condition. Nearly everyone desires his own safety, but not all individuals are willing to do what they know to do to insure their own safety.

Jan 1, 1981

By S. Orr

Flow in heaps and dumps is a two-phase flow phenomenon, though for many applications it could be simplified as unsaturated flow. Unlike saturated flow (typical of ground water flow), where permeability is independent of other hydraulic parameters, unsaturated flow permeability depends on the degree of saturation and/or on capillary pressure. Several recent field studies suggest that flows in heaps and dumps tend to concentrate in preferred pathways, bypassing much of the ore. Different preferential flow phenomena are triggered, promoted and influenced by different heap structures, by pretreatment, by the composition of the leach solution and by application rates and schedules. The structure of heaps and dumps is determined and affected by every stage of their construction -from blasting to crushing to conveying and stacking of the material. History of heap construction and operation combined with monitoring, mapping and certain field tests can reveal patterns of flow and transport in a given heap and can shed light on the physical mechanisms behind it. Such insight can provide a first cut on heap leaching enhancement by either eliminating or bypassing physical barriers to flow and leaching. Based on the understanding of the heap structure, advanced flow and transport models can be constructed and further enhance the insight. This insight is essential for understanding cause and effect and can provide a basis for decision making. Further, based on the physically-based models, an intelligent control agent, such as MRDS1, can learn and build a self-learning representation (or a model) of the complex heap leaching system, followed by dynamic optimization and planning of heap leaching and heap rinsing.

Jan 1, 2003

By Hans-Christian Kurzweil, Klaus Mussger

CERN, the European Laboratory for Particle Physics, are upgrading their existing particle accelerator for the next generation of experiments and have three major packages of underground construction work in progress. Each package involves the construction of shafts, cavern, tunnels and surface buildings. This paper gives an overview of CERN, the LHC Project and a more detailed discussion of design process, design review and construction works of Experimental Point 5. The LHC works at Point 5 include two large shafts, two major parallel caverns with a total span of 50 m, which are separated by a 7 m wide and 28 m high concrete pillar, and a system of secondary tunnels and caverns. The overburden ground consists of around 50 m of water bearing moraine (silty, clayey gravels and sands) overlying 20 mof molasse, which consists of horizontally bedded marls and sandstones. The shaft shave been constructed with ground freezing, primary concrete lining, membrane and secondary slip formed concrete lining. The concrete pillar was then excavated and concreted in advance of excavation of the two caverns. Cavern construction includes primary support of shotcrete lining and rock bolts, followed by waterproof membrane and primary support of shotcrete lining and rock bolts, followed by water proof membrane and insitu concrete secondary lining.

Jan 1, 2003

By W. M. Ma, Daniel W. H. Su, K. Centofanti, Yi Luo, W. L. Zhong, Syd S. Peng

1.1 INTRODUCTION When underground mining involves total extraction, it induces overburden strata movements. If not properly planned it causes surface subsidence and affects surface environmental conditions. Total extraction usually refers to longwall mining and room and pillar mining with pillar extraction. Surface subsidence has long been a subject of intensive research for scientists all over the world and considerable achievements have been obtained. However due to its difficulties and complicated nature, research into overburden movement has been thus far incomplete as compared to that into surface subsidence. Since surface subsidence is a manifestation of the results of overburden movement, the processes and mechanisms of overburden movement must be fully understood in order to establish the mathematical prediction models of surface subsidence. In this chapter the processes of overburden movement and its zones of movement in the overburden will be discussed. 1.2 PROCESSES OF OVERBURDEN MOVEMENT When total extraction is used, it produces a large void in the coal seam and disturbs the equilibrium conditions of the surrounding rock strata. The roof strata bend downward while the floor heaves. When the excavated area (or gob) expands to a sufficient size, the roof strata will cave. As a result, the overlying strata continue to bend and break until the piles of the fallen rock fragments are sufficiently high to support the overhanging strata. At this tie the overhanging strata no longer cave, but bend and rest on the underlying strata. Strata bending and subsidence develop upward until reaching the surface and forming a subsidence basin. The whole overburden strata and the surface subsidence basin will further go through a period of compaction and gradually become stabilized. Current knowledge regarding the process of overburden movement has been derived from several sources. One is the direct observation of the mined sections and their surfaces, another from field monitoring of strata movements in the overburden, and others from computer modeling and scale modeling in the laboratory. For example, Figs. 1.1 and 1.2 show the results of a borehole monitoring experiment (Borehole B-2) using the full profile borehole inclinometer and full profile borehole extensometer for lateral displacements and vertical subsidence, respectively (Conroy and Gyarmaty, 1983). The panel, 400 ft wide by 5000 ft long and 630 ft deep, was extracted from the Pittsburgh No. 8 seam which had an average thickness of 54 in. The face advanced from east to west. Curve A in Fig. 1.1 shows the inclination of the borehole when it was 45 ft ahead of the face at Position A (i.e., on the solid coal side). The total deviation of the borehole was 1/7000. The first shear movement occurred at 130 ft above the coal seam and appeared to have occurred along the bedding plane with large contrast of rock strata on both sides, i.e., sandstone vs. shale. When the face had passed the borehole 79 ft at Position B, the borehole deformed to assume a bow shape with a maximum deviation of 2.8 in. from the center line. Numerous shearing planes occurred and extended further upward. When the face had passed the borehole by 75 ft at position C, strata subsidence was measured. Fig. 1.2 shows that a strata separation as large as 4 ft occurred at 40 to 100 ft above the coal seam. Above this level, the strata subsided more uniformly with bedding separations within 1 to 2 in. Another example is the longwall mining with complete caving in the Soviet Union's Karaganda Coalfield (Kolebaeva, 1968). The coal seam was 6 ft thick and 154 ft deep. The immediate roof was the fine-grained sandstone interbedded with coarse-grained sandstone. Above this, there was sandy shale and shale. In order to monitor the process of overburden movement, 15 stations were established in a borehole (Fig. 1.3). The elevations of those stations were measured when the face was at various distances from the borehole. Fig. 1.3 shows the movement history of each station. For convenience of comparison, the movement history of each station was plotted on a common reference point as shown in Fig. 1.4. Clearly, the movements of strata from Station # 1 to #6 were nearly simultaneous, i.e., they behaved as a single unit. The strata from Station #7 to # 15 had differential separations, the maximum of which was 40 in. As the face passed by and moved away, strata separations in general reduced gradually. But strata separation between Stations #12 and #13 remained the same until the end of the monitoring period. The total bed separation between Station #7 and # 15 reached 11 in., approximately 18% of seam thickness, when the face was 132 ft beyond. Based on the above observations, the author developed a conceptual model of overburden movement due to underground longwall mining (Fig. 1.5). The above-mentioned two case studies illustrated that subsidence in the overburden strata propagates upward and subsidence velocity decreases from the bottom to the top. When the subsidence velocity at the surface reaches the maximum value, the subsidence velocity at the bottom portion of the overburden strata has decreased considerably and the strata have begun to compact. Bed separation occurs within a certain distance above the coal seam and reduces from the bottom to the top. When the face has moved away, bed separation reduces gradually, so that eventually some beds completely close and others partially close. Bed separation reaches the highest level when the subsidence velocity at the surface is at its maximum value. For instance in Fig. 1.4 surface subsidence velocity was maximum when the face was 38 to 44 ft past the point. At this time the total bed separation was maximum. Overall the strata directly above the opening are subjected to tension in the vertical direction. But above a certain level, all strata move nearly simultaneously. There is also significant shear movement along some bedding planes. 1.3 ZONES OF MOVEMENT INTHEOVERBURDEN After the extraction of a longwall panel or room and pillar section of sufficient width the strata in the overburden are subjected to various degrees of movement from the bottom to the top. According to the movement characteristics, the damaged overburden can be divided into four zones (Fig. 1.6). Caved Zone. After the extraction of coal, the immediate roof caves irregularly and fills up the void. The strata in this zone not only lose their continuity completely, they also lose their stratified

Jan 1, 1992

By Qing-Wang Hao

In this paper, the authors attempt to develop engineering based approaches for estimating safety factors against pillar and floor failures and predicting likelihood of subsidence events over abandoned coal mines in Illinois. There are two critical problems involved in the analysis: determination of geotechnical properties, and consideration of time effect. The authors attempt to solve the problems by generating hypotheses, which are used to modify existing engineering models for estimating pillar and floor safety factors. The modified models are validated using data from two well established databases of subsidence events over active and abandoned mines in Illinois. The data analysis suggests that most of the trough type subsidence events in Illinois were caused by floor failures rather than pillar failures. The subsidence incubation period seems to be directly related to initial floor safety factors at the time of mining. This correlation can be used to predict the likelihood of subsidence and to assess subsidence risks within both spatial and temporal contexts where detailed mine maps are available.

Jan 1, 1986

By F. Duan, W. G. Pariseau, C. S. Schmuck

Three dimensional comparisons between calculated and measured displacements in the hanging wall of a rock mechanics study region at the Homestake Mine show correlation coefficients of 0.7 or higher. Two and three dimensional back analysis of the data lead to different scale factors for determining rock mass strengths and moduli from laboratory values and seemingly to different rock mass properties. A simple analytic model suggests that indeed not only are there two independent scale factors, one for strength and one for moduli, but that there is a logical relationship between the scale factors obtained from two and three dimensional back analyses of the same mine measurements. The three dimensional modeling procedure implemented in the finite element code UTAH3 thus seems to have a realistic potential for investigating underground hardrock mine questions not amenable to two dimensional analysis.

Jan 1, 1987

By Leonard G. Austin

INTRODUCTION Run-of-mine (ROM) coal generally falls into two major groups: that from underground mining (continuous mining machines) and that from strip mining. Continuous miners produce a finer product and strip mining a coarser product that is crushed to produce the desired size. Further size reduction depends on the proposed use or uses for the coal. For example, for pulverized coal power plants as a rough general rule, the coal is ground or pulverized to 80% less than 200 mesh (0.075 mm) before it is blown into the boiler furnace as a coal-air suspension. For solvent-refined coal processes, the desired size is about <30 mesh (< 0.60 mm); in fluidized bed uses, the coal is typically < 8, in. (< 1.6 mm); for Lurgi pressure gasifiers, the desired size is 1 to 2% in. (25 to 63 mm). Coal preparation plants typically work with a coarse size fraction of 6X 1/14 in. (150X32 mm). The past decades has shown a steady increase in the tonnage of coal cleaned, 1 but it seems probable that the next decades will produce much wider applica- tion of deep coal cleaning before use and that the stages of crushing and pulverization from ROM coal to end use will be integrated with coal cleaning processes. Since different cleaning processes work better with different ranges of coal size and the degree of cleaning possible is intimately connected with the size of coal, a systems design engineer will need as much information as possible concerning the size distributions produced by different types of machines and how these can be modified by different ways of operating the machines. In general, the following questions will be of concern: 1.What type of crusher or grinder is best, or most satisfactory, for reducing a given feed size and type of coal to a desired product size? Frequently, the desired product is smaller than some nominal size, but with as few fines as possible, or the fraction in some size range has to be as high as possible. 2. How big does the machine have to be for a given throughput rate? 3. How much electrical energy (or its equivalent) is required per ton of product? 4. How does the size distribution vary with change in throughput rate, and is a control scheme necessary or desirable? 5. Can the size distribution be readily varied if desired? 6. What is the optimum way to operate a machine in a given system? This section will emphasize certain general principles which guide us in answering these questions. PARTICLE SIZING Since particle size and size distribution are the raw information with which we work, it is worthwhile to review the terminology used. In this regard, see

Jan 1, 1991

By F. F. Aplan, C. E. Raleigh

Three run-of-mine bituminous coals, ranging in rank from low volatile to high volatile A, were floated at nominal -28 mesh (590 µm), -100 mesh (150 µm), and -400 mesh (37µm) feed particle sizes with selected pyrite depressants and a dispersant. For the three coals tested, pyrite and ash -forming minerals were usually rejected best at -100 mesh feed particle size. Depressant additions tended to improve flotation selectivity between pyrite and coal. When joint dispersant-depressant systems were used at -100 mesh (150 µm) feed particle sizes, the total sulfur content of the clean coal decreased further. For a comparable coal yield, less fuel oil collector was required for the flotation of -100 mesh coal than was required for the flotation of either -28 or -400 mesh feed.

Jan 1, 1992

By M. Popescu

In the southern part ?f the Occidental Carpathians an iron ore body embedded in crystalline schist was exploited. The presence of minerals such as magnetite, pyrite, pyro-chorine, allanite, monazite and xenotime, which also include within their structure substituted thorium isomorphic, was pointed out. The iron ore was processed by crushing, grinding and magnetic separation in order to obtain magnetite concentrates. The environment was affected by pollution generated by the mining activities. In 1995 due to the economic reasons, it was decided to cease mining exploitation and remediate the area. The paper presents the ecological impact caused by mining activities on a number of environmental targets and the population. The pollution sources, namely sterile dumps and mine waters are characterized (from the physical, chemical and radiometric point of view); the pollution pathways and theirs effects on potential targets, namely soil, vegetation, hydrographic network and population are identified as well. The degree of pollution due to all sources was assessed and maps of pollution sources distribution were generated. The supplementary effective doses for the population were assessed using rigorous scenarios and considering all exposure pathways. The rehabilitation plan includes solutions for dumps geometrization, covering and revegetation as well as mine waters management.

Jan 1, 2005

By G. M. Wenberg

Bacterial leaching of copper and uranium is practiced in environments containing available sulfur and iron and having a strong acidic reaction. However, many mineral deposits do not have these conditions; the chalcocite and native copper ores in Michigan&apos;s Upper Peninsula are examples. Bacterial and fungal populations were observed in samples taken from the White Pine Copper Company tailing pond and mine, and isolates were prepared of the dominant species. The isolates were then screened for solubilization of copper from tailing, ore, and concentrate during incuba¬tion in a carbon and nitrogen supplying medium. Certain Penicillium fungi solubilized significant amounts of copper under neutral to slightly acidic conditions. Leaching apparently was accomplished by release into the medium of metabolites which chelated with the copper.- The process was found to be highly dependent upon the medium and the concentration of copper in the source material. The paper describes sampling and screening techniques, presents the results of experimentation to standardize procedures and maximize the yield of copper, and discusses the nature of the leaching process. This preliminary research reveals a biological process which may find use in mineral extractions but whether it can be applied economically remains to be determined.

Jan 1, 1969

By J. C. Yingling, J. Sottile

This paper addresses detailed models of the temporal pattern of material flows produced by longwall mining units and procedures for inferring the parameters of these models from field data. First, a model is established that predicts material flows referred to shearer position and cutting element. This model is comprised of trend and fluctuation components. The trend component, which is invariant between repetitions of the cut cycle, accounts for deterministic effects on the flow pattern, such as web depth variation as a function of the cutting operation, systematic tendencies for seam height and cutting speed variations along the face, changes in the relative velocity between the shearer and face conveyor, etc. The fluctuation component models random effects on material flow and employs time series approaches. Subsequently, a method is given for converting the positionally-referenced flows to temporally-referenced flows. The model serves as a basis for generating detailed, accurate computer simulations of material flows from longwall units. As such, it provides a more rigorous foundation for analysis of cutting procedures and outby materials handling systems in longwall mines.

Jan 1, 1994

By Richard P. Killmeyer

In 1985 the Pittsburgh Energy Technology Center (PETC) of the U.S. Department of Energy (DOE) conducted a study to evaluate the emerging selective agglomeration processes (Killmeyer, 1985). This paper covers a recent similar study evaluating an associated technology, the emerging advanced froth flotation processes. This study has goals similar to those of the 1985 study: to evaluate the emerging froth flotation technologies collectively and determine how far they have progressed in their ability to clean ultrafine coal, and to assess the differences among the participants. The major concern again in this study was not the level of development, nor the economics of each individual process but only the current technical performance.

Jan 1, 1989

By David A. Gulley

One of the advantages of the market¬place, economists believe, is that it automatically generates information which would otherwise be hard to come by. For example, supply and demand signals imply something about private and social sacrifices and benefits. This information is seldom perfect, and can neglect some costs and benefits, but it nevertheless is crucial data. Moreover, marketplaces can also eliminate the need for some information. For example; since market participants are voluntary actors, only those transactions are culminated in which a mutual gain occurs. This eliminates the need for some cost-benefit calculations. A corollary to this standard economic doctrine is that whenever a government or other party supplants a marketplace, it must generate some of the information either lost or newly-needed. (Schultze, 1977) In 1981, following a ten-year moratorium the United States government resumed leasing of publicly-owned coal. When the government designed this program, the goal was to guide the marketplace, rather than merely to respond, as a landlord, to the market&apos;s wishes. This goal has been modified somewhat, due both to the changing political climate and to the practical difficulties of implementing so ambitious an undertaking. Nevertheless, during this period of program design and implementation, part of the challenge of creating and using the necessary information fell to the professional modelling fraternity.

Jan 1, 1982

By M. J. Friedel

The scale-dependent hydraulic properties in a fractured porphyry copper oxide deposit are investigated for relevance to In Situ Leach Mining (ISLM) design. In most well field designs, mine operators assume that hydraulic properties represent a continuous spatial function, or continuum. The problem, however, is that the representative elemental volume, the volume over which hydraulic properties are described as a continuum, is a domain (if it exists) that occurs between end members representing varying degrees of heterogeneity. At the Cyprus Casa Grande mine, transmissivity fluctuations are spatially correlated. This suggests that, at the scale of ISLM, the hydraulic process is heterogeneous, and a typical continuum representation is inappropriate for design. Alternatively, a conditional stochastic approach using the Simulated Annealing (SA) technique is deemed useful for the imaging of heterogeneous hydraulic properties. Specifically, SA preserves transmissivity extremes and spatial complexity, both necessary for accurate conditioning of a flow simulator. Also, the stochastic continuum approach provides an avenue whereby uncertainty (or risk) may be addressed. Currently, the SA weaknesses are that it is computationally time consuming and that uncertainty may increase when simulating under conditions of statistical isotropy. Recognizing that ISLM exerts a dynamic influence on hydraulic properties with time, the mine design should be re-evaluated at intervals throughout the evolution of the leaching process.

Jan 1, 1994

By Wenbing Guo, Gaobo Zhao, Erhu Bai, Brijes Mishra

"Violent damage to overlying strata can occur above the large mined-out space left by high-intensity longwall mining. For the safe mining under the aquifer and surface water body, it is imperative to investigate the criteria of overburden failure degree due to high-intensity mining. In this paper we present the characteristics of high-intensity longwall mining panels and overburden critical failure. Based on the geological and mining conditions of a high-intensity mining panel, 3D numerical models that consider the maximum horizontal principal stress and minimum horizontal stress and their orientations with the panel advance direction were established by using the 3DEC software. Sizes of the panel reaching the critical failure of the overburden under different mining thicknesses (3.5 m, 6.5 m, 9.5 m, 12.5 m, and 15.5 m) were analyzed. Further, criteria of overburden critical failure were proposed for the high-intensity mining. The coefficient of overburden critical failure (Cf) is the ratio of the strike length (Ls) to the mining depth (H). Through curve fitting analysis, a fitting equation between Cf and mining thickness was obtained. The results showed that overburden failure criterion with high-intensity mining panel is 0.38 = Cf = 0.97 with different thicknesses. Field measurements validated the results from the numerical models.INTRODUCTIONHigh-intensity longwall mining refers to a coal mining method with thick coal seam, large panel width, fast face advance rate, and severe damage to the overburden and surface in China (Guo et al., 2018). Violent movement of the roof rock and severe damage to overlying rock strata occur in the large mined-out space left by a high-intensity longwall mining panel, which is a 3D (three-dimensional) structure layout with highly non-uniform stress distribution. Recognition of such fact a will facilitate this overlying rock strata failure problem investigation (Peng et al., 2019).The studies of overlying strata failure due to longwall coal mining have been reported by some scholars with different methods. Kang et al. (2018) studied the sudden massive roof collapse during longwall coal retreat mining by a large-scale two-dimensional physical model and a UDEC numerical model. Ghabraie et al. (2015) researched the mining-related strata movement by applying the three-dimensional laser scanner, optical transducer, and digital image processing techniques to two-dimensional physical models. Wang et al. (2015) analyzed the structural characteristics and fracture failure process of overlying hard and thick main roof by a 3DEC numerical model; this study also investigated the geostress reduction and permeability enhancement redistribution of the adjacent coal measure strata by FLAC3D. Guo et al. (2012) performed a three-dimensional finite element analysis and shared some insights into the dynamic interaction between mining induced strata stress changes, failures, and gas flow patterns. Meng, Shi, and Li (2016) proposed a 3D coupling model between stress and permeability, investigating the dynamics of a stress, permeability coefficient using FLAC3D. Khanal, Guo, and Adhikary (2019) used the 3D numerical modelling approach (COSFOW) to investigate the strata deformation and the change in permeability in the overburden strata due to the underground extraction. However, few studies considered the influence of the horizontal stress on overburden failure in a high-intensity longwall mining panel, and the critical failure criteria of overlying rock strata is unclear."

Jan 1, 2019