三維物理模型在長臂采煤工作面采空區通風研究中的應用

摘要：煤礦長臂工作面采空區中的風流分布對危險區域識別起到重要作用，并且受到上覆巖層垮落效果和井下通風方式影響巨大。為了探究U型通風工作面采空區中風流分布，在相似比為1∶300的物理模型中開展了實驗研究。該物理模型主要由4部分組成：開采系統、由相似模擬材料堆砌的上覆巖層、通風系統和數據采集系統。隨著模型開采和上覆巖層垮落，在采空區形成后引入示蹤氣體。通過預埋的監測管路測試模型中示蹤氣體在各點的濃度，從而得到示蹤氣體濃度分布圖。該實驗平臺和研究方法通過相似材料和模擬開采系統可以獲取較為合理的物理上覆巖層和長臂采空區模型。通過示蹤氣體手段可以得到可視化的采空區風流分布。通過改變通風方式，該系統可應用于針對其他不同通風條件下采空區風流狀況研究。關鍵詞：長臂工作面采空區；物理模型；通風；示蹤氣體

中圖分類號：X 936文獻標志碼： A

0IntroductionDistributions of methane and air movement in longwall gobs are closely connected with mine fire and explosions.Nearly all U.S.longwall mines take advantage of bleeder ventilation systems to keep the zone of potentially explosive methane concentration in gob a safe distance away from the working face except the mines with serious spontaneous combustion issues[1].However，the bleederless ventilation systems are commonly used in longwall coal mines in other major coal producing countries

[2].Regardless of the ventilation systems used，it is necessary to gain a good knowledge about the methane concentration distribution and air movement in the gob.Such knowledge allows us to identify the zones of potential spontaneous combustion as well as zones of explosive air behind the longwall face.The air movement in the gob is not only governed by the ventilation pattern，but also largely governed by the gob permeability.

The characteristics and some key parameters of the inaccessible gob are difficult to measure in the field.Attempts have been made to determine the air leakage using methods such as tracer gas technique，but the results are limited[3-4].Currently，the controlled modelling techniques are viewed as the only rational way to assess ventilation in gob areas[5].Yuan[6] has carried out a computational fluid dynamics study on the ventilation flow paths within the longwall gob for three different ventilation systems.The gas flow in the caved gob is simulated as a laminar flow through porous media.The gob permeability was assigned based on geotechnical modelling of longwall mining and the associated stressstrain changes/rock failure using FLAC code[7].Air flow pattern and velocity data inside the gob are obtained from this study.More ventilation scenarios，gob drainage arrangements and gob hazard zones have been studied using such approach by other researchers[8-9].No physical gob models have been utilized in ventilation study，but in the field of ground control，physical models of simulant material have been well established and proven to be an accurate and reliable approach.Ventilation studies on a 3D physical model of mining entries，gob and overburden which have been conducted as the simulant material can realistically simulate the movement and failure mechanism of the roof strata associated with longwall mining.A unique mining system has been built in the model that allows the excavation process of a longwall face being simulated inside a simulated overburden model.It means that as the face advances，the gob behind the face was also formed in a natural manner.By introducing air flow in the entries，the longwall ventilation system can be simulated.The 3D physical model with a scale of

1∶300 was constructed for modelling a longwall gob using Uventilation system.To the authors’ knowledge，this research is the first

attempt of studying gob ventilation using such physical gob model.

1MethodologyAs the schematic shown in Figure 1，the physical model consists of the following four parts：mining system，overburden strata，ventilation and data acquisition system.The process of the experiment generally includes the following five stages as shown in this figure.The mining system（a series of steel bars）and entries（a tube）were first positioned on the coal seam level as shown in Figure 2.Note that the stress sensor，small white plate is used for monitoring the overburden deformation during mining process.The overburden strata in the model are constructed with simulant materials according to the geological column of the simulated site.After the model has been properly dried，experiment started.Simulated mining activity starts from the setup entry to the recovery line.As mining progresses，the overburden strata subsides behind the face.At the end of mining operation，caved and fractured debris fill the longwall gob.At that time，tracing gas is introduced into the entries and gas samples will be collected in the designated points for analysis.All the details of each of the system component and function will be discussed in the following sections.

Fig.1Schematic of the experiment model and flowprocess diagram

Fig.2Coal seam level structure layout

（ventilation tubes and mining units）

1.1Simulation of the Mining ProcessThe mining operation is simulated by the mining system，which consists of a set of steel bars.Each of the steel bars can be lifted and lowered individually.The lifted steel bars represent the coal seam in the original overburden as seen in Figure 3（a）.By sequentially lowering them，the longwall mining operation can be simulated as Figure 3.The deformation process of the overburden strata on the model is also schematically depicted in this figure.A waiting time of 2hour is used after lowering each bar to allow the failure of the overburden strata to simulate the face retreat rate.This waiting time is based on the monitored stress changing rate.After the “mining operation” has been completed at the recovery line，a 24hour waiting time was allowed for the overburden strata reaching its new equilibrium.

The model is to simulate the process of overburden strata deformation and breakage during longwall mining.In order to build an overburden which is able to represent the caving and fracturing process and the permeability changes in the formed gob for ventilation study，various scalar ratios must be maintained.The force term in the model is stress while the basic failure modes are shear and tension.The deformations are related to the elastic modulus and Poisson’s ratio of the simulant materials.Hence，the main intensity index to be similar is compressive strength，and the main deformation indices are elastic modulus and Poisson’s ratio.Consequently，the specific weight and the compressive strength，the elastic modulus and the Poisson’s ratio of the material should follow the following rules.

For a physical model with dimension ratio of

λ=lp/lm=300，the specific weight ratio is

β=δpδm

=

ρpgρmg

=1.5.

（1）

The compressive strength ratio is

σpσm

=δpδm

·

lplm

=

CpCm

=λB=450.

（2）The elastic modulus ratio is the same as the compressive strength ratio as

EpEm

=λB=450.

（3）In these relationships，l is length，δ is specific weight，ρ is density，g is the gravitational acceleration，σ is compressive strength，C is cohesive strength，and E is elastic modulus.The subscripts m and p represent the model and prototype，respectively.The majority of the overburden strata in the simulated case are sandstone and shale as shown in Table 1.The physical and mechanical properties of the strata are also listed in the table.Using the scale relationships above，the specific weight，elastic modulus and compressive strength，Poisson ratio of the simulant materials are determined from those of real strata and listed.Each of the strata is simulated by a simulant material with the mixture of sand，gypsum and lime in a predetermined ratio and the mixture information for the physical model is listed in Table 2[10].

Tab.1Rock properties between the in situ and model

Strata

Specific Weight/（kN·m-3）

In situModel

Elastic Modulus/MPa

In situModel

Compressive Strength/MPa

In situModel

Poisson’s Ratio

Mudstone20.8013.8720 01944.4920.50.050.20

Sandy mudstone26.4017.6026 76759.4848.80.110.28

Medium sandstone26.6017.7350 430112.0065.10.140.28

Fine sandstone26.2017.4743 02095.6069.00.150.26

Silt stone26.0017.3354 739121.6058.50.130.25

Coal14.609.7314 14231.4313.50.030.28

Tab.2Strata parameters and material proportion

Part of Real Seams Information

Serial No.StrataThickness/m

Model Seams Information

S-G-L RatioSand/kgGypsum/kgLime/kgCoal ash/kgThickness/cm

15K32.837∶5∶511.780.840.84-1.00

14Coal0.369∶2∶82.020.050.181.010.20

13Mudstone1.438∶3∶75.980.220.52-0.50

12Fine sandstone4.877∶4∶619.631.121.68-1.70

11Sandy mudstone1.008∶4∶63.990.200.30-0.30

10Mudstone4.578∶3∶719.940.751.74-1.70

9K2 4.617∶5∶519.631.401.40-1.70

8Coal0.46 9∶2∶82.020.050.181.010.20

7Mudstone4.638∶3∶719.940.751.74-1.70

6Silt stone7.047∶2∶827.470.793.14-2.30

5Fine sandstone2.227∶4∶67.850.450.67-0.70

4Mudstone5.458∶3∶721.930.821.92-1.80

3Fine sandstone2.147∶4∶67.850.450.67-0.70

2Medsandstone8.507∶3∶733.361.433.34-2.80

1Coal （mined）3.009∶2∶812.110.271.086.061.00

The strata above K3 is simulated by precalculated flat iron plate putting on the top of the overburden to serve as a simplified pressure device.In constructing the model，sand is used as aggregate while gypsum and lime are used as cement.The frangibility of material mixed with gypsum is close to rock.Such material mixtures provide sufficient adjustable ranges in elastic modulus and compressive strength.The simulated overburden materials are placed in the model frame layer by layer from coal bottom to the ground surface（Figure 4）.A drying time of more than 30 days is required for the materials to be hardened before experiments are conducted.

Fig.4Constructed overburden model

1.3Ventilation and Data AcquisitionThe main purpose of the physical modelling is to investigate the ventilation in the longwall gob using tracer gas.The ventilation simulation system consists of gas cylinder，reduction valves，balance tank，flow control valves，precision flowmeter，ventilation pipe and negative pressure pump.Helium gas has been chosen as the tracer gas both for safe use and detection perspective[11].The gas concentration in the model is measured by vacuum gas analyzers through test tubes buried in the longwall gob during the model construction.The locations of the sampling points are showed in Fig.5 while the experiment parameters for the simulation study are shown in Table 3.

Fig.5Tracer gas measurement points（unit：mm）

Ventilation was started and the tracer gas was introduced into the intake air after the new equilibrium is reached in the overburden strata after “mining activities”.After starting the ventilation for a predetermined time，seven gas analyzers were connected to the first column of measurement points close to the longwall face and measured the helium gas concentrations along that line.The gas analyzer makes the measurement based on the difference in thermal conductivities between normal air（0024 W/（m·k））and helium gas（0142 W/（m·k））.Each of the analyzer has a vacuum pumping rate of 100 mL/min.The test at each point lasted for 15 seconds so that a steady concentration reading could be obtained.The tubes should be sealed properly after the testing.A complete set of tests lasted about 4 minutes.Then the points in each of the subsequent column were measured in the direction from longwall face to setup entry.Three complete sets of ventilation measurements have been conducted starting at 3，20 and 30 minutes after starting the ventilation，respectively.

Tab.3Experiment parameters

Dimensions of structures

L/mmW/mmH/mm

Frame1 2007001 000

Overburden1 200700600

Working panel4001616

Gob96040016

Steel bar4003040

Ventilation quantity/（L·min-1）

Intake air11 Intake tracing gas7

2Data Analysis and Results

2.1Overburden DeformationAs the first attempt of such experiment，it is valuable to observe the deformation condition inside the overburden.A quarter and then one half of the overburden model was cut and removed to show the strata deformations after the completion of the experiment as shown in Figs.6 and 7.The inserted drawings show the location of the displayed faces of the deformed overburden.Figure 6 shows the fracture zone along the panel transverse direction.Along this direction，the top of the archshaped fracture zone is located in the center while the arch foot is located at the edge of the gob.The pillar next to the gob was not been affected by the mining activity.Figure 7 shows the overburden deformation along the panel longitudinal direction.The most serious overburden fracturing occurred in the lower right corner（above or near the panel setup entry）.These deformation characteristics observed from the physical modeling conform with real mining environment and findings from previous research[12-13].Therefore，the caved and fractured zones formed over the longwall gob should be able to play their intended roles to affect gob ventilation.

Fig.6Observed overburden fractures along the panel direction

Fig.7Observed overburden deformation

along the mining direction

2.2Air Flow Distribution and AnalysisThe distributions of helium gas concentration and flow path in the gob area for the three sets of test are plotted and the first two sets are shown in Figs.8 and 9.No significant differences were shown between the two sets of measurements starting at 20 and 30 minutes after ventilation establishment，indicating that a steady state of ventilation has been reached after 20 minutes.In Figure 8，the gob area with detectable helium gas concentration after 3 minutes ventilation is relatively small while the affected area in Figure 9 has grown much larger and became steady.From the asymmetric distribution patterns in both plots，it is obvious that the ventilation air penetrated into the longwall gob much deeper on the headgate side than the tailgate side.The path lines also show that air leaked from the headgate side into the gob，but some was forced back into the face again through the shields.This reenter of leaked air starts from the middle of the face to the tailgate rib and these air goes into the gob as deep as 9.1 m.Moreover，it is also very important to note that the air travels through the gob from headgate to tailgate and reaches its deepest point at the middle line of the longwall face.

Fig.8Tracer gas concentration（in % volume）

contour in the gob after 3 minutes of ventilation

Fig.9Tracer gas concentration（in % volume）contour

in the gob after 20 minutes of ventilation

3ConclusionsThis paper presents a new 3D physical modeling approach to investigate the air flow pattern in the inaccessible longwall gob areas.Simulating of gob air flow can be satisfactorily achieved using well designed physical model.With the buildin mining system，this platform is capable of realistically simulating compaction of gob materials，as well as the overburden.The deformation characteristics observed from the model confirms that a rational physical overburden and longwall gob can be obtained by such approach.This is the first attempt of using such physical model for ventilation research，an oneentry bleederless ventilation system was employed in this experiment.Based on the tracer gas concentration distribution，the air flow patterns are visualized，and an asymmetric flow path was obtained and discussed.Air flow are mainly concentrated behind the shields and reenter through the shields on the latter part of the face.By rearranging the ventilation scheme，this platform can be used to study various ventilation scenarios，as well as bleeder system.This physical model approach can be expected to provide more detailed information of the flow characteristics with a larger model scale and have potential to validate numerical models.

References

[1]

Smith A C，Diamond W P，Mucho T P，et al.Bleederless ventilation systems as a spontaneous combustion control measure in US coal mines[M].US Department of the Interior，Bureau of Mines，Information Circular，1994.

[2]Schatzel S，Krog R，Dougherty H.Field study of longwall coal mine ventilation and bleeder performance[J].Trans Soc Min Metal Explor，TP-10-040，2012，330：388-396

[3]Thimons E D，Kissell F N.Tracer gas as an aid in mine ventilation analysis[R].Bureau of Mines，Washington，DC（USA），1974.

[4]Timko R J，Thimons E D.Sulfur hexafluoride as a mine ventilation research tool：recent field applications[M].US Department of the Interior，Bureau of Mines，1982.

[5]Gilmore R C.Computational fluid dynamics modeling of underground coal longwall gob ventilation systems using a developed meshing approach[D].Colorado School of Mines.Arthur Lakes Library，2015.

[6]Yuan L，Smith A C，Brune J F.Computational fluid dynamics study on the ventilation flow paths in longwall gobs[C]//Proceedings of the 11th US/North American Mine Ventilation Symposium，2006.

[7]Esterhuizen G S，Karacan C .A methodology for determining gob permeability distributions and its application to reservoir modeling of coal mine longwalls[C]//SME 2007 Annual Meeting，2007：07-078.

[8]Brune J F.The methaneair explosion hazard within coal mine gobs[J].Transactions of the Society for Mining Metallurgy and Exploration Inc，2014，334（1）：376-390.

[9]Ren T，Balusu R，Claassen C.Computational fluid dynamics modelling of gas flow dynamics in large longwall goaf areas[J].International Journal of Obstetric Anesthesia，2011，17（17）：374-375.

[10]Lin H.Study on the law of mininginduced fracture evolution of overlying strata and relieved methane delivery and its engineering application in fullymechanized top coal caving[D].Xi’an：Xi’an University of Science and Technology，2009.

[11]Persily A K，Grot R A，Ventilation system performance evaluation using tracer gas techniques，air infiltration centre，ventilation strategies and measurement techniques[C]//6th AIC Conference Proceedings，England，Air Infiltration Centre，1985：19-26.

[12]Yang W，Sui W，Xia X.Model test of the overburden deformation and failure law in close distance multiseam mining[J].Journal of Coal Science and Engineering（China），2008，14（2）：181-185.

[13]Singh M M，Kendorski F S.Strata disturbance prediction for mining beneath surface water and waste impoundments：Proc 1st conference on ground control in mining，Morgantown，27-29 July 1981，P76-89.Publ Morgantown：West Virginia University，1981[C]//International Journal of Rock Mechanics and Mining Sciences Geomechanics Abstracts.Pergamon，1983，20（1）：A13.