Advances in gas content based on outburst control technology in Huainan, China

International Journal of Mining Science and Technology 24(2014)385-389Contents lists available at Science DirectInternational Journal of Mining Science and TechnologyELSEVIERurnalhomepagewww.elsevier.com/locate/ijmstAdvances in gas content based on outburst control technologyin huainan ChinaXue Sheng d, D. C, Yuan Liang, Xie Jun Wang Yucang aKey laboratory of Integrated Coal Exploitation and Gas Extraction, Anhui University of Science and Technology, Huainan 232001, ChinaSchool of Mining Engineering, Taiyuan University of Science and Technology, Taiyuan 030024d National Engineering Research Centre for Coal Mine Gas Control, Huainan 232001,ChinaARTICLE INFOA BSTRACTThe sudden and violent nature of coal and gas outbursts continues to pose a serious threat to coal mine10 October 2013fety in China. One of the key issues is to predict the occurrence of outbursts. Current methods that arerevised form 15 November 20138 December 2013used for predicting the outbursts in China are considered to be inadequate, inappropriate or impracticalAvailable online 29 April 2014in some seam conditions. In recent years, Huainan Mining Industry Group(Huainan)in China and theCommonwealth Scientific and Industrial Research Organisation(CSIRO) in Australia have been jointlyKeywords:developing technology based on gas content in coal seams to predict the occurrence of outbursts in Hua-Coal mininan. Significant progresses in the technology development have been made, including the developmentOutburstof a more rapid and accurate system in determining gas content in coal seams, the invention of a samOutburst predictionpling-while-drilling unit for fast and pointed coal sampling, and the coupling of DEM and LBM codeGas contentfor advanced numerical simulation of outburst initiation and propagation. These advances are describedin this paper.@2014 Published by Elsevier B V on behalf of China University of Mining Technology1 Introductionindices(K1, Ah2, and AP), gas flow indices(g and o), gas pressure(P), and Protodyakonov index ()[1, 2 Practical experience revealsCoal and gas outburst is that a large quantity of gas in conjunc- that there are three main shortcomings in the use of these indicestion with the ejection of coal and associated rock is rapidly re- The first issue is to set outburst threshold values for seams ofeased into the working face or mine workings. Outbursts are ious geological, and mining conditions. Experience has shown thathazardous through the mechanical effects of particle ejection and the outbursts do occur while the indices are below the presetby asphyxiation and possible explosion from the gas produced. threshold values. The second shortcoming is the impracticality ofWith an increase in depth of mining and production, outburst some indices in some seam conditions. Some indices, when appliedtensity and frequency tend to increase, although this trendin certain seams, lack reproducibility, which makes the use of suchsomewhat masked with advances in understanding the phenom indices inappropriate and impractical. the last challenge is severena and application of new prediction technologies and control impediments to development rate in the use of some indices. Themeasures. One of the key challenges to effectively manage and current prediction methods can predict less than 10 m ahead acontrol outbursts is to develop practically appropriate technologies development face, which makes it almost impossible to achieve ato predict the occurrence of outburstshigh development rate.Based upon the understanding that the outbursts occur as a re-In comparison, the method used to predict the occurrence ofsult of mutual interaction of a number of factors such as rock pres- outbursts in Australia is based on the rapid direct measurementsure, gas present in coal, and physical and mechanical properties of of gas content in coal seams. This method enables the predictioncoal, a number of indices have been developed and are in use to distance reaching 50-200 m in a coal seam and has been practicedpredict the occurrence of the outbursts in China. These indices in- successfully in Australia for 18 years. Though gas content, as a ba-lude mainly cutting volume (S), gas desorption and diffusion sic coal seam parameter, has been used and practiced for outburstprediction in Australia, surprisingly this has not been widely usedin practice elsewhere This, mav partially be due to different coalCorresponding author. Tel. +61 7 33274443.seam conditions or中国煤化工 f the developmentE-mail address: ShengXue@csiro. au(S. Xue)CNMHGd and directionalhttp://dx.doiorg/10.1016/j.ijmst.2014.03.0172095-2686 2014 Published by Elsevier B V on behalf of China University of Mining Technology386S Xue etborehole drilling technology. Because of its expediency and less Association of Australia(AS3980-1999)recommended a gas conimpact on mining operation (coal sampling can be taken well tent test procedure based on the USBM direct method and adahead of a working face for determination of its gas content), Aus- dressed equipment construction, sampling and testingtralian coal industry essentially bases its outburst management on procedures, and methods of calculating the final gas content regas content. This reliance on one index was criticised by some sults. Saghafi et al. proposed a method that places stainless steelresearchers; however its success at preventing outbursts cannot balls in a coal sample container for subsequent crushing of thebe disputed Over the last 18 years only a handful of small out- sample in the laboratory without transferring the sample to anbursts have occurred in Australia in extracting coal in which gas other container [6]. despite these significant developments, issuescontent was below preset threshold valuesremain about the rapidity and accuracy in measuring the gas con-To overcome some shortcomings of the prediction methods tent of coal in underground coal mining. One of the main issues inused in China and take advantages of the successful experience the estimation relates to soft coal seams In a soft or friable coalin Australia, Huainan and Csiro entered into an agreement a few seam, it is impossible to obtain a relatively complete coal core,years ago to undertake joint research to develop a gas content leaving coal cuttings as the only practical alternative. There arebased technology to predict the occurrence of outbursts in Hua- two obvious challenges in coal cutting: how to quickly obtain theinan. The main scopes of the joint research include: (1)establish- coal cutting at a given position and how to estimate gas lost duringment of the standard laboratory and calculation methods for sampling. a new sampling unit was invented for sampling and adetermining gas content with the rapid direct desorption method; computer program was developed to estimate the gas lost during(2)development of an innovative coal sampling technique suitable sampling. Due to its significance and relative independence fromfor soft coal seams; (3)laboratory and field investigations of gas gas content measurements, the new sampling unit is described indiffusion and desorption characteristics of coal seams; (4)develop- Section 3ment of a coupled numerical program to simulate the process ofThe gas content determination system was developed to notoutburst initiation and propagation; (5)determination of an out- only comply with gas content determination standards in Chinaburst threshold value of gas content for coal seams; and (6)devel-(GB/T 23250-2009)and Australia(as 3980-1999), but also to rapopment of a complete set of gas content based technologies to idly measure the gas content of coal (normally in a few hours) withpredict the occurrence of outbursts in Huainanimproved accuracy to meet the requirements of underground coalThe joint research effort has achieved significant results and mining. The system consists of several integrated components,continues to power ahead. These results include the development namely a sampling-while-drilling(SWD)unit for coal samplingof a more rapid and accurate method for determining the gas con- (refer to Section 3), a canister to contain the coal sample, a portabletent in coal seams, the invention of a sampling-while-drillingunit used underground for the initial gas desorption measurement,tem for fast and pointed coal sampling, and the coupling of discrete a surface-based unit for gas desorption measurement, a sampleelement method (DEM) and Lattice Boltzmann method(LBM) crushing and gas desorption measurement unit, a weighing device,codes for advanced numerical simulation of outburst initiation a gas composition measurement unit, and gas content calculationand propagation. These progresses are discussed below.software[7). Each of these elements except the SWd unit is brieflydescribed belowa canister is used to contain a coal sample during its transport2. Determination of gas contentand gas desorption. The canister body is made from stainless steelpipe. The diameter and length of the canister body depend on theMethods used for determining the gas content can be divided size and amount of samples, with minimum air space to reduce theinto two categories: indirect methods and direct methods. Indirect effect of changes in atmospheric conditions on the gas volumes tomethods are based upon either the gas adsorption characteristic of be measured. the canister diameter is in the range of 10-20 cmcoal under a given pressure and temperature condition, or other (internal diameter) and the canister body length is in the rangeempirical data that relate the gas content of coal to other parame of 50-80 cm. the canister is sealed at one end and the other endters such as gas pressure, gas emission rate, coal rank, or depth of has a removable gastight cap fitted with a bi-directional and gas-cover. Direct methods are based on the actual measurement of the tight valve. For convenience of transporting, a set of handles are fitvolume of gas released from a coal sample. Direct methods can be ted on the canister bodyfurther divided into slow and fast desorption methods. In the slowThe portable unit is used to measure the initial gas desorptiondesorption method, a coal sample is allowed to desorb until a low by the water replacement method in the field The method is sim-desorption rate cut-off point is reached or desorption stops, which ilar to the one that was described by Kissell et al. [5]. The unit con-can often take weeks or months. In the fast direct desorption meth- sists of two inverted and scaled cylinders, a water pan, a Pvc tubeod, instead of waiting weeks or months, the sample is crushed for gas flowing from the canister to the cylinders, and other fittingswhen received by the laboratory, and the gas content of the sample such as stands, bracket, switch, pump and battery. The size of thecan often be obtained in days. The direct desorption method is cylinders is in the range of 100-800 mL to cater for various gaswidely used in the coal mining industrydesorption rates. The water in the unit is fully saturated salt waterSince Bertard et al. introduced the direct gas content measure- to reduce the solubility of desorbed gas in water and coloured forment method in 1970, a number of systems have been developed easier reading against the measurement scales. The cylinders areto measure the gas content of coal by direct methods 3, 4. Bertard secured to the stands and filled with water by drawing water intoet al. used three containers (initial measurement, transport, and them by applying suction to the tube passing through the pan. thecrushing a U-tube manometer, and a crusher to measure the tube is raised to a level above the cylinder and supported horizongas desorption volume of a 10 g coal sample collected from an tally to reduce the likelihood of the water being drawn into thenderground borehole. Kissell et al. modified Bertard's devices, canister if the ambient temperature falls sharply. With this unit,eliminated the need for multiple containers, and used replacement the desorbed gas flovrement is direccontainers constructed from 0.3 m long sections of 0. 1 m diameter ted towards determi中国煤化工, subsequent toaluminium pipe 5]. These modified devices were used to measure being removed fromHCNMHGto its containthe gas content of larger sized virgin coal core samples. The meth- ment in the gas canisod associated with the modified devices is referred to as the unitedSimilar to the portable unit used in underground, this unit wasState Bureau of Mines (USBM)direct method. The Standards developed to measure gas desorption by the water replacementS Xue et aL. International Journal of Mining Science and Technology 24(2014)385-389387method in a surface-based laboratory. the notable additions in thelaboratory unit include temperature control, automatic data acquiswivel SaverHead rosition, and large measuring cylindersThe coal sample crushing unit is used to pulverise the coal samle into a fine powder to quickly release the remaining gas. theunit allows the gas to be bled off to the automated surface-basedgas desorption measurement unit during crushing( Fig. 1)2 NPIer componentsis new system include a weighinghose endsdevice, a gas composition measurement unit, and gas content calculation software. The mass of coal samples is determined byespectively weighing the canister with and without a coal sampleThe weighing device is capable of measuring the sample to accu-acy better than 1% A gas chromatograph is used for gas composi-Fig. 2. SWD unit.tion analyses. A special computer program based on the diffusionof gas from coal is developed to estimate gas lost during samplingCoal seamAir outhow3. Coal samplingInner tubesamples for direct gas content measurements are often ob-tained from conventional cores and drill cuttings from boreholesdrilled in coal seams. However in soft coal seams, where core sam-ples are difficult or sometimes impossible to obtain due to poorborehole stability leading to the deformation of large boreholesAir inflowOuter tubeand borehole collapse, drill cuttings have to be taken insteadDue to the faster initial gas desorption rates, the cuttings need tobe taken quickly during borehole drilling and at given positionsFig 3. Fluid circulations in the Swd unit.along the borehole to ensure the accuracy of the gas content measurements. In conventional borehole drilling underground, pressurised air is injected into the borehole through the inner space of tube, and the cuttings are sampled at the collar of the boreholedrill rods and the cuttings generated at the drill bit are flushed,A more detailed description of this unit can be found in a paperand carried out along the outer perimeter of the drill rods, and by Xue et al. [10]can be sampled at the collar of the borehole. With this conven-tional method the sampling time to obtain the cuttings increasesignificantly with the length of the borehole. It is also likely that 4. Numerical simulatorthe cuttings obtained with this method may be contaminated withcuttings from other positions along the borehole. Clearly, with theA fundamental challenge to effectively predict and control anconventional method, it is impossible to obtain drill cuttings from outburst is to understand its physical mechanism and factors conborehole drilled underground to meet sampling requirements for tributing to its initiation and development. Field observations andrapid and accurate gas content measurements, hence the need for a laboratory studies reveal that the occurrence of an outburst is theSWD unitresult of combined effects of stress redistribution, gas desorptionThe SWD unit uses a special design of double-tubing drill rods coal properties and time effects. The use of numerical simulationsnd a reversed circulation of pressurised air to flush and sample can provide useful insights on outbursts. Along this line, there havecoal cuttings at any given position during borehole drilling( Figs. 2 already been some attempts 111-13]. However, these models doind 3). The unit consists of drill bit, drill rods, side entry swivel, not simulate solid fracture and fragmentation explicitly. Free flowhead rod, top swivel, hose, air/water inlet and drum. Each drill of fluid and two-way interactions between the solid and fluid arerod is 1.5 m long, and made of double steel tubes, i.e., an outer tube also missingand an inner tube. during borehole drilling, pressurised air is inIn this study, a new outburst model which includes the mostjected through the space between the inner tube and outer tube recognised factors of outbursts is presented. The model couplesto the borehole, flushes some of the cuttings generated at the two well developed numerical approaches: the discrete elementdrill bit out of the borehole through the inner space of the inner method (DEM)and the Lattice Boltzmann method (lBm)he dEm is a powerful numerical tool and has been extensivelyused in many scientific and engineering problems. The ESyS-Particle is an open source DEM code and is used in this study Detailedinformation about the model and the code can be found in open lit-In recent years, the LBM has made brilliant progress as a newnethod in numerical modelling of fluid dynamics. The LBM solvesthe particle distribution function f. The completely discretizedequation, with the time step At and space step Ax, is given byH中国煤化工f(x+ex,t+△CNMHGwhere t is the lattice relaxation time: ex the discrete lattice velocityFig. 1. Coal sample crusher with an automated gas desorption system.in direction a; xi the point in the discretized physical space; and f:388S Xue et al International Journal of Mining Science and Technology 24(2014)385-389fr(x, t+At)=f(x, t)-(x, t)-f(x, t)Then the porous medium step has the form∫x(x,t+△t)=fx(x,t+△t)nf(x+ex,t+△t)-fx(x,t+△t)where ns is the damping parameter introduced by Dardis andMcCloskey [17 Diffusion is managed by the dEm part in the coupled scheme. It is assumed that there is an average and uniformpore pressure p, and concentration c for each particle i The fluid exhange between two contacted particles i and j is determined byFicks first law of diffusion AV/= D(C-C)At, where D is the diffu-Fig 4. Schematic diagram of excavation and outburst simulation.sion coefficient of the linkThe hydro-mechanical coupling can be implemented based orthe equilibrium distribution function. Eq (1) is usually solved in the Biot's linear pore-elastic theory [181following two steps: collision and streaming step(P-∞p)/Knf2(x1,t+△t)-1(x1,t)=-V1(X,t)-f(x,(P-p/B)/K,f2(x+ex,t+△t)=fx(X,t+△t)(3) where p is the pore pressure; P=-okk/3 the mean or total mechancal pressure (isotropic compressive stress ):a=Ekk=Av/v the voluwhere fa is the post-collision statemetric strain(positive for extension): s=V//V the variation of fluidIn this study, the open source Lattice Boltzmann code, Open LBcontent(positive corresponds to a"gain"fluid): a the biot coeffi-used to couple with the ESys-Particle code There are several is- cient: b the Skempton pore pressure coefficient; Km the drainedsues to be addressed for such coupling moving boundary condi- bulk modulus of the material; and v and V the volume of the matetions for a curved shape, momentum transfer between solid rial and fluid, respectivelparticle and fluid, and force from fluid to solid particlesDesorption is implemented in both the DEM and the lbm codesThe particle surface can intersect the link between two nodes at The reduction of concentration in a solid particle is described bybitrary distance with a ratio of 8=xr-xwl/x -Xb]. The reflecteddistribution function at fluid-solid boundary node x can be calcu- at-I lc-c(p)an interpolation scheme [16where c is the average matrix gas concentration; p the gas pressure(x,t+△D)=1(1-6:5(x,【+AD)+6:5xt+AD(Eq (8)); td the sorption time; and c(p)=vc/v∫(xf2t+△t)-6W2Pwhere wa is the weight factor; Pw the fluid density at node x, andUw the velocity of the solid particle. The fluid force acted on the par- where V is the volume of the particle; Vc the volume of adsorbed gasticle surface can be obtained usingadsorption isotherms; and Vi and PL Langmuir volume and LangF=∑∑e21(x2t)+1(x,t+△)△x/△t(5) muir pressureThe results of a small outburst simulation are shown in this pa-per as an example to illustrate the capability of this numerical simThe first summation is taken over all fluid nodes at Xb adjacent to ulator The model consists of 1003 equal-sized particles andthe particle and the second is taken over all possible lattice direc- 305 x 300 fluid grids. Constant confining pressures are applied attions pointing towards a particle cell. This force is added to the par- left, right, top and bottom boundaries. The soft and fragile middleticle force in the dem codepart(shown in dark and light grey in Fig. 4), sandwiched by a hardDarcy flow is modelled by the lbm code by considering ther and stronger rock, and models a coal seam Excavation startstranslational collision step as a second intermediate step after from the left side with constant speed by removing the dark greystreaming, denoted by fpart. Two supporting walls (horizontally dashed lines)are placedYH中国煤化工CNMHGFig. 5. Particle motion and fluid velocity during the outburst development.S Xue et aL. International Journal of Mining Science and Technology 24(2014)385-3893895.42c-0ll.63e-02Fig. 6. Gas concentration at different time stepsafter the excavation, growing and following the moving excavationable to simulate solid fracture and fragmentation, free flowwall(vertical dashed line)at the right end( Fig 4)of fluid, and two-way interactions between the solid andFig 5 shows the results of particle motion and fluid velocity atfluidseveral time steps. In the early stage of excavation, there is no fracture events, and desorption of gas from the exposed surface islearly observed. At the time step of 7000 fractures occur at the Referencesexcavation surface and excavation stops. After this, the whole sys-tem of particle and fluid evolves by itself, and eventually the frac- I1l Cheng wY. Wang YA, Wang K). Determination of the critical value oftured coal is ejected from the face under exertion force of fluid andFig 6 shows the concentration of gas adsorbed in the solid coalline. Int J Min Sci Technol 2012: 22(1): 89-933 Bertard C, Bruyet B. Gunther J Determination of desorbable gas concentrationat the same time steps with Fig. 5. Lower concentration at the sur-of coal(direct method ) Int J Rock Mech Min Sci 1970: 7: 43-65face is observed due to desorption. The concentration changes in[4] Diamond WP, Schatzel S). Measuring the gas content of coal: a review. Int Jthe inner particles are caused by the diffusion process. The parti-Coal Geol 1998: 35: 311cles of the ejected coal have the lowest gas concentration because5]Kissell F, McCulloch CM, Elder CH. The direct method of determining methanentent of coalbeds for ventilation design(Report of Investigationthey are totally exposed and have larger surface areas, thereforecontinue gas desorption while moving away from the face.6 Saghafi A, Williams D], Roberts DB. Determination of coal gas content byushing method (Investigation Report CET/IT391R) Sydne[7 Yuan L, Xue S, Xie j.5. Conclusions[8]Li YB, Xue S, Wang JF, Wang YC, Xie J Gas diffusion in a cylindrical sample-aneral solution, approximation and error analyses. Int J Min Sci TechneSignificant advances in gas content based outburst control tech-4;24(1):69-73nologies have been made in Huainan, China. These advancesI9 Li X, Nie B, Zhang R, Chi L. Experiment of gas diffusion and its diffusionmechanism in coal. Int J Min Sci Technol 2012: 22(6): 885-9[10 Xue S, Lix, Xie ]. A new coal sampling system for measurement of gas contentin soft coal seams. Appl Mech Mater 2012: 121-126: 2459-64.(1)A more rapid and accurate method for determining the gas lll Xue S, Wang YC, Xie J. Wang G A coupled approach to simulate initiation ofcontent in soft or friable coal seams have been developedoutbursts of coal and gas-model development Int J Coal Geol 2011: 86: 222-30[121 Xu T, Tang CA, Yang TH, Zhu WC, Liu J. Numerical investigation of coal and gIn comparison with traditional gas content methods, thisutbursts in underground collieries. Int Rock Mech Min Sci 2006: 43: 905-19improved method uses coal cuttings instead of coal core [13] Yang W, Lin BQ, Zhai C, Li X, Sun X, Zhang C A new technology for coal and gasfor gas content measurement and a special computer pro-control based on the in situ stress distribution and the roadway layout. Int Jgram to more accurately estimate gas lost during coalMin Sci Technol 2012; 22(2): 145-9[14 Wang YC. A new algorithm to model the dynamics of 3-d bonded rigid bodiewith rotations. Acta Geotech 2009: 4: 117-27(2)A sampling-while-drilling system has been developed for[15 Wang YC, Alonso-Marroquin F. A finite deformation method for discretefast and pointed coal sampling. This system is capable of taking coal samples quickly at given positions by using a special [16] Wang YC, Xue S, Xie J. A fully coupled solid and fluidor simulating coaldesign of double-tubing drill rods and a reversed circulationand gas outburst with DEM and LBM. AGH J Min Geoeng 2012: 36: 377-84flow. Geophys Res Lett 1998 25: 14713)A coupled solid-fluid model has been developed to simulate (18 Detournay E,AHD. Fundamentals of poroelasticity. Comprehensive rockutburst initiation and propagation. The model couples dEmengineering: principles, practice and projects, Vol. ll, Analysis and designand LBM, two well developed numerical approaches, and ismethod. New York: Pergamon Press, 1993中国煤化工CNMHG