Mining-induced strata stress changes, fractures and gas flow dynamics in multi-seam longwall mining

International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

Contents lists available at SciVerse ScienceDirect

International Journal of Rock Mechanics & Mining Sciences journal homepage: www.elsevier.com/locate/ijrmms

Mining-induced strata stress changes, fractures and gas flow dynamics in multi-seam longwall mining Hua Guo a,n, Liang Yuan b, Baotang Shen a, Qingdong Qu a, Junhua Xue b a b

The Commonwealth Scientific and Industrial Research Organisation (CSIRO), PO Box 883, Kenmore, Queensland 4069, Australia National Engineering Research Centre for Coal Mine Gas Control, Huainan, Anhui 232001, China

a r t i c l e i n f o

abstract

Article history: Received 15 April 2011 Received in revised form 15 March 2012 Accepted 18 May 2012 Available online 4 July 2012

This paper presents key findings from a recent comprehensive study of longwall mining-induced strata movement, stress changes, fractures, and gas flow dynamics in a deep underground coal mine in Anhui, China. The study includes field monitoring of overburden displacement, stress and water pressure changes at the longwall panel 1115 (1) of the Guqiao Mine. In addition, 3D modelling of strata behaviour at the longwall panel using a 3D finite element code and goaf gas flow simulations with a CFD code are carried out. This research has resulted in many new insights into the complex dynamic interaction between mining induced strata stress changes, fractures, and gas flow patterns. Based on the findings from the field monitoring and numerical modelling, a three-dimensional annular-shaped overlying zone along the perimeter of the longwall panel is identified for optimal methane drainage during mining. A practical method that helps define the geometry and boundary of this zone is proposed. This study provides a new methodology and a set of engineering principles for the design of optimal co-extraction of coal and methane. Crown Copyright & 2012 Published by Elsevier Ltd. All rights reserved.

Keywords: Mining Coal Strata stress Permeability Methane Gas drainage Ground water

1. Introduction The concept and practices of co-extraction of coal and methane have emerged in recent years to improve coal mining safety and productivity, utilise coal seam methane resources, and reduce fugitive emissions [1–3]. The system of the co-extraction effectively integrates the two previously separate operations of coal extraction and methane drainage. The coal mining operation reduces the coal seam gas pressure and creates strata fractures that enhance gas migration and capture from working and surrounding coal seams. Effective methane drainage operation produces a source of clean energy and reduces the methane concentration in underground workings and the methane content in the adjacent coal seams. This is a significant benefit in preventing gas explosions and outbursts and promotes a safer and more productive coal mining environment. In addition, utilisation of coal mine methane directly leads to a reduction in Green House Gas emissions. To implement the co-extraction of coal and methane, a clear understanding of mining-induced strata stress changes, fractures, and gas flow are essential. To date, extensive studies in these areas have been carried out for various purposes such as mine ground control, water inrush prevention, and gas control, respectively. Front and side abutment stresses about a longwall (LW) panel,

n

Corresponding author. Tel.: þ61 7 3327 4608; fax: þ61 7 3327 4455. E-mail address: [email protected] (H. Guo).

de-stressing behind the coal face, and goaf consolidation processes have been investigated and characterised by many researchers [4–7]. Overburden strata fractures are generally classified into three zones vertically: the caved zone, the fractured zone, and the continuous deformation zone, the heights of which are normally expressed as a function of mining thickness [7–10]. Palchik [11,12] recently measured the height of the interconnected fracture zone which is connected to mine workings, and the locations of horizontal fractures along rock layer interfaces in the overburden for goaf gas drainage, using a method of methane emission measurement from vertical boreholes. LW goaf flow mechanisms have been investigated by many researchers using gas tube bundle systems to collect and analyse goaf gas samples, underground tracer gas tests, and computational fluid dynamics (CFD) numerical modelling techniques [13–16] for gas drainage as well as fire control. The understanding of spatial distribution of goaf gas concentration and pressure has been used to improve the design and operation of goaf gas wells and goaf inertisation (for heating control). These studies have contributed to improved understanding of complex coal mine gas flow mechanisms in the recent years. However, to integrate and optimise coal extraction and methane drainage design, particularly in the multiple coal seam environment with low permeability conditions common in China, it is important to understand the dynamic interaction between strata, water, and gas desorption and migration during mining. Huainan Coal Mine Group (HCMG) in China and The Commonwealth Scientific and Industrial Research Organisation (CSIRO) in

1365-1609/$ - see front matter Crown Copyright & 2012 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijrmms.2012.05.023

130

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

Goafside retained roadway

Tailgate

Concrete wall 220m

Maingate

LW face

Goaf

Start-up line

Outflow

Inflow

Fig. 1. Plan view of Panel 1115(1).

Australia have recently completed a major collaborative research programme. This programme used Panel 1115(1) of the Guqiao Mine at Huainan, China as the experimental site and carried out a systematic and integrated study into mining-induced strata stress changes, fractures and gas flow dynamics by means of field measurements, numerical modelling, and theoretical development. This paper describes key results from that study.

2. Mining conditions of Panel 1115(1) The Guqiao Mine is located in the central to western part of the Huainan Pan-Xie Coal Field, 20 km west of Fengtai City of Anhui Province, China. It currently produces 12 Mt of raw coal annually. The key geological features at the Guqiao Mine include thick alluvium layers, deep coal seams, high methane content in minable seams and high geothermal temperatures. The coal-bearing geological section has a total thickness of 734 m, and it contains thirtythree coal seams, nine of which can be mined with a total thickness of 24 m. The five key minable seams include Seams 13-1, 11-2, 8, 6-1, and 1. Currently the mine is extracting Seams 11-2 and 13-1, both classified as low permeable seams with a methane content of 1.96–13.2 m3/t and 2.7–12.9 m3/t, respectively. The working seam at Panel 1115(1) is Seam 11-2. It has an average thickness of 3.0 m, an average dip angle of 51 ranging from 31 to 81. Overburden depth ranges between 640 and 760 m with a 400–450 m thick alluvium section at the top. The panel length is 2600 m and panel width is 230 m. It was mined by the retreat LW mining method with full seam extraction. The LW face was ventilated using the ‘‘Y’’ type of ventilation system with the maingate (belt road) and inby part of the tailgate (material transport road) being the intake roadways and the outby part of the tailgate being the return roadway. The tailgate behind the LW face was kept open by constructing a 2.2 m wide concrete wall to replace the mined-out roadway wall on the goaf side, to form ‘‘goafside retained roadway’’ (see Fig. 1). The coal seam within a distance of 200 m from either side of the panel was not mined. The working seam is overlain by Seam 13-1 (thickness¼3.5 m) at a distance of 75 m and Seam 17-2 (thickness¼1.4 m) at a distance of 179 m. The plan view of this panel is shown in Fig. 1. Simplified geological settings from a typical borehole log of this panel are listed in Table 1.

3. Field monitoring of strata movement, stress changes and water pressure at Panel 1115(1) 3.1. Monitoring design For the purpose of an in-depth investigation of mininginduced strata movement, stress changes, fractures, gas flow

Table 1 Simplified geological settings in Panel 1115(1). Layer number

Lithology

Thickness (m)

Floor depth (m)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Alluvium Mudstone & coarse sandstone Coal seam 17-2 Siltstone & mudstone Fine sandstone Mudstone & coarse sandstone Mudstone Siltstone & mudstone Coal seam 13-1 Mudstone & siltstone Fine sandstone Mudstone Fine sandstone Siltstone Medium sandstone Siltstone Fine sandstone Mudstone Coal seam 11-2 (Working seam) Mudstone Coarse sandstone & siltstone Mudstone & coarse sandstone

453.3 102.0 1.4 25.7 14.8 5.7 42.4 12.1 3.5 22.3 14.2 3.9 8.0 8.1 4.4 6.1 2.1 5.5 3.0

453.2 555.3 556.6 582.4 597.2 602.9 645.3 657.4 660.9 683.3 697.4 701.3 709.3 717.5 721.8 728.0 730.0 735.5 738.5

6.4 25.3 20.0

744.9 770.2 790.2

20 21 22

and their dynamic interaction, an integrated real-time monitoring system was designed (Fig. 2) based on the specific mining and geological conditions at Panel 1115(1). The monitoring programme focused on the part of the panel that covered a retreat distance of 1300 m from the LW start-up. The monitoring programme included: (a) Monitoring stress change and displacement of the roadway roof: 10 monitoring stations were set up in a 450 m section of the tailgate, located at a retreat distance of 1100–1550 m from LW start-up. At each station, one multi-point roof extensometer was installed which has five anchors located at depths of 8 m, 6 m, 4 m, 2 m and 1 m into the roof. At every second monitoring station, three uniaxial stressmeters were installed in vertical and inclined upward boreholes that monitored the stress changes in the roof of the roadway and sidewalls. (b) Monitoring overburden strata movement: Two multi-point surface extensometers were installed in the panel at distances of 90 m and 30 m, respectively from the tailgate. Twenty anchors of each extensometer were installed in the key overburden strata layers to measure their displacements. (c) Monitoring water pressure change in overburden strata: Seven piezometers were installed in two deep boreholes

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

131

BH#4 (1 piezometer) Data logger

50m

100m

30m

BH#3 (7 piezometers)

BH#2 (20 anchors)

retreat direction

90m Panel 1115(1)

BH#1 (20 anchors)

450m 1300m Surface piezometer boreholes

Surface extensometer boreholes

Monitoring stations for both stress and displacement of roadway roof Monitoring stations for only displacment of roadway roof Fig. 2. Integrated real-time monitoring plan at Panel 1115(1).

direction of 60 degrees to the vertical, as shown in Fig. 3a, indicated that the abutment stress extended 300 m ahead of the LW face. At a distance of about 28 m ahead of the face, the abutment stress change reached its peak and then decreased. The horizontal stress change in the roadway direction, as shown in Fig. 3b, kept decreasing as LW face retreated.

Stress change/MPa

2 1.5

~

1 Test location and direction 0.5 0 -350

-300

-250

-200

-150

-100

-50

0

50

-0.5 Distance ahead LW face/m

0.5 Distance ahead LW face/m -300

-250

-200

-150

-100

-50

0 -0.5 -1 -1.5 -2

Test location and direction

Stress change/MPa

0 -350

50

-2.5 -3

3.2.2. Relative displacements in the roadway roof Fig. 4 shows the results from the roadway roof displacement monitoring. Within the measurement depth of 8 m in the roof, bed separations occurred at a distance of 25 m behind the LW face. At a retreat distance of 170–200 m, roof displacement showed another obvious increase before stabilising.

3.2.3. Movement of overburden strata The two surface deep-hole extensometers had both been affected by borehole instability. As a result, only limited data were obtained for the depth range of 473–595 m (i.e., 152–274 m above the working seam); see Fig. 5. The limited data indicate that: (1) the trend of strata movement in the panel centre is generally consistent with that in the panel side; (2) strata within a vertical distance of 152–274 m above the working seam have a similar displacement trend; (3) most of the strata displacement occurred within 170 m behind the LW face; (4) there was a 650 mm relative displacement between the overlying rock strata and the ground surface, after mining.

Fig. 3. Measured stress change in the roadway roof.

drilled at a distance of 50 m outside the LW panel to avoid being damaged by mining-induced caving. The piezometers were located in aquifers in the rock and alluvium strata between the seam floor and the ground surface. The boreholes were fully grouted using special grout designed to isolate any hydraulic linkage between the piezometers. All monitoring systems included automatic data acquisition and recording with a logging interval of 1 h. 3.2. Key monitoring results 3.2.1. Stress changes in the roadway roof Fig. 3 shows typical monitoring results of the stress changes in the roadway roof at Panel 1115(1). The stress changes in the

3.2.4. Water pressure change in surrounding strata Heights of the caved and fractured zones have been investigated in numerous studies. Peng [7] suggested that combined thickness of the caved and fractured zones ranges from 30 to 50 times the mining thickness. Palchik [11] pointed out that the thickness of the fractured zone varies greatly from 20 to 100 times the mining thickness, and the measured results showed the heights of the zones of interconnected fractures and horizontal fractures may reach 19–41 and 53–92 times of the mining thickness, respectively. In China, LW induced fractures with high vertical conductivity are usually called water-conductive fractures, which form channels for ground water to flow down to mine workings. The maximum height of this water-conductive fractured zone is suggested to be from 9 to 28 times the mining thickness, depending on the stiffness of the overlying strata [5]. An investigation at 25 LW panels of Huainan Mining Group

132

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

100 8m Sandstone

80

8m 6m

60

4m

Displacement/mm

6m

4m Claystone 2m 1m Claystone

40 2m

20

1m Gateroad

0 -100

-50

0 -20

50

100 150 200 250 300 Distance behind LW face/m

350

400

450

Dispacement in BH #1/mm

200

0 BH #2 Roof 192m

0

BH #1 Roof 202m

-30

BH#2 Roof 192m

-200

-60 BH #2 Roof 222m

-400 BH #1 Roof 152m -600

-90

BH #2 Roof 274m

-120

BH #1 Roof 274m

-800 -50

0

50

100 150 200 250 Distance behind LW face/m

300

350

Displacement in BH #2/mm

Fig. 4. Measured relative displacement in the roadway roof.

-150 400

Fig. 5. Measured overburden displacements.

showed that the maximum height of the water-conductive fracture zone ranges from 7 to 20 times the mining thickness [17]. The change in water pressure in the overlying strata has a direct link with the type of strata fractures, making it possible to estimate the heights of fracture zones by measuring water pressure changes in overburden strata during mining. The measured water pressure change and associated overburden strata layers are shown in Fig. 6. The key observations from the measurement results are as follows. The strata with a significant water pressure drop can extend up to 145 m (48 times of mining thickness) above the working seam, indicating that the zone of strata fractures has extended to this height. This result is in reasonable agreement with findings by other researchers [7,11]. Strata at different heights showed a different trend in water pressure change, indicating the effect of mining-induced strata stress changes and rock fracturing. It is postulated that the roof strata within 7–43 m (2–14 times of mining thickness) have been fractured vertically or subvertically, and that the fractures are linked hydraulically between different strata layers. This zone can be called the cross-strata fracturing zone. Within the height range of 88–145 m (29–48 times of mining thickness) is the bed separation zone where fractures, mainly horizontal, developed. At a height of 145–237 m (48–79 times of mining thickness), rock mass deformed continuously without significant fracturing. This zone is commonly referred as the deformation zone. Further up in the alluvium layers, the key aquifers were not affected by mining. Water pressure started to increase at a distance of 300 m ahead of the LW face, indicating that mining induced stress can extend to 300 m ahead, which is consistent with the stress

monitoring result in the roadway roof. Water pressures dropped rapidly between 100 m ahead of the LW face and 170 m behind the LW face, implying that mining-induced strata fractures and stress changes are developed within this region. Further behind the LW face the water pressure was either stabilised or started to recover, suggesting that the goaf are being consolidated and the mining induced fractures are being compacted in this zone.

4. Modelling the effect of mining on surrounding rock mass and gas flow In order to extrapolate the localised monitoring results to the entire Panel 1115(1), a 3D numerical study was carried out. The numerical investigation employed software called COSFLOW [18,19], developed by CSIRO, NEDO and JCOAL of Japan, and the commercial CFD software Fluent. The numerical studies aimed to investigate the zones of mining-induced strata stress changes, rock fractures, permeability changes, and goaf gas flow mechanisms. 4.1. COSFLOW COSFLOW is a three dimensional finite element code which employs the Cosserat Theory and has a special advantage in efficiently modelling stratified rock mass. A brief description of COSFLOW is provided as follows. 4.1.1. Mechanical model A unique feature of COSFLOW is the incorporation of Cosserat continuum theory in its formulation. In the Cosserat model,

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

Piezometer BH

Pressure/mH20

200

133

Surface

100 0 -400 -250 -100 50 200 350 500 650 800 950 -100 Roof 466m -200

Alluvium

-300 -400 200 Pressure/mH20

Piezometer

Distance behind LW face/m

Roof 466m

454m

Claystone Sandstone

100 Roof 237m

0 -400 -250 -100 50 200 350 500 650 800 950 -100 -200

Roof 237m

Clay&Sand Coal 17-2

555m

-300

Pressure/mH20

-400

Roof 145m

Distance behind LW face/m

Clay set

200

Roof 88m

657m

100

0 -400 -250 -100 50 200 350 500 650 800 950 Roof 18m -100 Roof 7m Roof 15m Roof 43m Roof 88m Roof 145m

Sand set

-200

Sandstone Coal 13-1 Clay set

Roof 43m Roof 15m Roof 7m

Sand set

735m

Coal 11-2

-300 Floor 18m

-400 Distance behind LW face/m

Fig. 6. Measured water pressure change and its relationship with rock fracturing characteristics.

inter-layer interfaces are considered to be smeared across the mass, i.e., the effects of interfaces are incorporated implicitly in the choice of stress–strain model formulation. An important feature of the Cosserat model is that it incorporates bending rigidity of individual layers in its formulation and this makes it different from other conventional implicit models. In comparison to the conventional continuum model which has three independent degrees of freedom and six independent stress components in a three dimensional case, the Cosserat model for the stratified material will have six independent degrees of freedom and ten independent stress components. 4.1.2. Permeability equations In COSFLOW, rock mass permeability is formulated on the basis of the mine induced strain, proposed in [20]. The equations of permeability change (Dk11, Dk22, and Dk33) used in COSFLOW are: i k 1h ð1 þ b2 De22 Þ3 þ ð1 þ b3 De33 Þ3 ð1Þ Dk11 ¼ 11 ¼ ini 2 k11

Dk22 ¼

Dk33 ¼

k22 ini

k22 k33 ini k33

¼

i 1h ð1 þ b1 De11 Þ3 þ ð1 þ b3 De33 Þ3 2

ð2Þ

¼

i 1h ð1 þ b1 De11 Þ3 þ ð1 þ b2 De22 Þ3 2

ð3Þ

where kii (i¼1, 2, 3) are the three components of mining-induced absolute permeability, kini ii (i¼1, 2, 3) are the three components of absolute initial permeability, Deii are the normal strain components, and bi are expressed as:

bi ¼ 1 þ

1Rm ðFai =Fsi Þ

ð4Þ

where Rm is the modulus reduction ratio (ratio of rock mass modulus to rock matrix modulus), the term Fai/Fsi may be defined

9000m

1200m

10000m

Fig. 7. COSFLOW model and mesh.

as a function of equivalent fracture porosity. If Rm equals 1.0 then bi equals 1.0, resulting in minimal strain induced permeability changes. When Rm tends to 0.0 (i.e., the case of highly fractured rock), bi will attain the maximum value and hence will induce large changes in permeability.

4.2. COSFLOW model The finite element mesh was based on a 3D geological model of the Guqiao Mine, constructed using the data from 136 boreholes with geological software such as Minescape and Gocad. The final COSFLOW model is shown in Fig. 7. It simulates a region of size 10,000  9000  1200 m (length  width  height), with mesh of 1245,951 elements.

134

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

Table 2 Basic rock properties used in COSFLOW. Lithology

D (kg/m3)

E/GPa

g

UCS (MPa)

sc (MPa)

st (MPa)

j (o)

c (o)

Js (m)

2 Kini h (m )

ini Kini v / Kh

Rm

P

Alluvium Mudstone Siltstone Coarse sandstone Medium sandstone Fine sandstone Coal

2050 2500 2550 2574 2499 2650 1400

0.01 7 20 33 30 30 3.5

0.35 0.22 0.22 0.22 0.22 0.22 0.25

20 27.5 57.5 56 74.5 10

0.02 5.21 7.16 14.02 14.58 18.17 2.33

0.001 2 2.75 5.75 5.6 7.45 1

35 35 35 38 35 38 40

4.5 4 5 5 6 6 4

0.1 0.2 0.5 1 2 4 0.1

/ 5e-18 5e-15 1e14 2e14 1e14 1e17

/ 0.1 0.1 0.1 0.1 0.1 1

/ 0.6 0.5 0.52 0.54 0.54 0.4

/ 0.06 0.04 0.08 0.07 0.06 0.04

n D—density; E—Young’s modulus; g—Poission’s ratio; UCS—uniaxial compressed strength; sc—cohesion; st —tensile strength; j—friction angle; c—dilatancy angle; ini Js—joint spacing; p—porosity; Kini h —initial horizontal permeability; Kv —initial vertical permeability; Rm—modulus reduction.

Stress/MPa

30 25 20

Roof 7m Roof 30m Roof 88m Roof 226m

15 10 5 0 -400

-300

-200

-100

0 -5

Ahead LW face/m

100

200

300

400

500

Behind LW face/m

30

Stress/MPa

25 20 20m behind LW face 50m behind LW face 100m behind LW face 200m behind LW face 400m behind LW face

15 10 5 0

-300

-200

0

-100 -5

100 200 300 Lateral location of LW/m

400

500

r 0.8 0.6 0.2 -0.05

Fig. 8. Modelled vertical stress variation in overburden strata (a) Stress variation along the panel length direction, (b) Stress variation along the face direction, (c) Stress distribution in a plane section 30m above the mining seam, (d) Stress distribution in a vertical cross section along the length of the panel.

The initial vertical stress was assumed to be proportional to the overburden weight. The major and minor principal horizontal stresses were set to be 1.12 and 0.94 times of the vertical stress, respectively, according to the stress measurement data provided

by the mine. The outer boundaries (except the top boundary) of the model were assigned with roller boundary conditions, i.e., deformations in the direction perpendicular to the boundary faces were constrained to zero and deformations parallel to the

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

boundary faces were unconstrained. The deformations on the top surface were unconstrained. The mechanical properties such as UCS, Young’s modulus, and porosity were based on the results of recent laboratory tests on rock specimens collected from the panel area. The other strata properties, including initial permeability values, were estimated based on previous studies and experience at Huainan coal mines [1,21,22]. The strata properties used for COSFLOW modelling for typical rock layers at the experimental site are shown in Table 2.

r = 0.2

r = 0.2 r = 0.5

r = 0.5 r = 0.8

r = 0.8

B-B

A-A A

r = 0.2 r = 0.8

B

B Startup

LW face

A Fig. 9. Conceptual model of stress reduction factor in overburden strata.

Orders changed

8 7 6 5 4 3 2 1 0 -30 -1 0

135

4.3. Dynamic development of mining-induced stress During the process of coal gas adsorption–desorption, significant methane adsorption can be achieved if the stress in a coal seam is reduced below a critical value. In order to display the degrees of strata de-stressing, a parameter, r, was suggested and used by the project team to represent the ratio between the stress reduction during mining and the original stress before mining as follows: r ¼ 12ðsz =szo Þ

ð5Þ

where sz is the vertical stress during mining, sz0 is the original vertical stress before mining. Fig. 8 shows the modelled vertical stress distribution in the seam roof of Panel 1115(1) along the LW face and along the panel length. Key observations on the COSFLOW modelling results are as follows. Along the panel length direction, the vertical stress shows a four-stage process marked by increase–decrease–recovery– stabilisation (see Fig. 8a). The abutment stress can extend up to 300 m ahead of the face. The stress reduction zone is 0–80 m behind the face, followed by a stress recovery zone around 150– 200 m behind the face. Beyond 200 m, the stress begin to stabilise. Across the panel, the abutment stress zone can also extend up to a distance of 300 m outside the panel. The vertical stress in the central part of the panel recovers gradually as the mining retreats. However, a de-stressed zone remains at either side of the panel even after the completion of the mining (Fig. 8b). There exists an annular shaped zone where rock mass is fully de-stressed (see Fig. 8c), and its width reduces with the height into the roof (see Fig. 8d). Based on the COSFLOW modelling results, a conceptual model is constructed to demonstrate the stress reduction factor in the overburden strata of Panel 1115(1); see Fig. 9.

20m behind LW face 50m behind LW face 100m behind LW face 200m behind LW face 400m behind LW face

30

60

90 120 150 180 Lateral location of LW/m

210

240

270

Fig. 10. Modelled distribution of permeability changes in overburden strata. (a) Horizontal permeability changes across the panel width 30m above the working seam, (b) Distribution of horizontal permeability changes in a plane section 30m above the working seam, (c) Distribution of horizontal permeability changes in a cross-panel section 60m behind the face and (d) Distribution of vertical permeability changes in a cross-panel section 60m behind the face.

136

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

Horizontal permeable zone Cross-strata permeable zone

A-A

B-B

A

Horizontal permeable zone Start -up

B

Goaf consolidated zone

B LW face

A Horizontal permeability

Vertical permeability

Fig. 11. Conceptual model of permeability distribution in overburden fractured zone.

4.4. Permeability change of surrounding rock mass The success of coal seam methane drainage will not only depend upon the degree of methane desorption but also the permeability of the coal seam and surrounding rock mass. Fig. 10 shows the modelled permeability changes in horizontal and vertical directions in the overburden strata of Panel 1115(1). The horizontal permeability is predicted to increase significantly in the de-stressed zone. A horsesaddle shaped horizontal permeability distribution is predicted across the panel width, where the permeability in strata above the panel’s central region is predicted to increase by about 6 orders of magnitude, then decrease by 1 order of magnitude, and finally stabilise as the face advances. Near the sides of the panel, however, the permeability is again predicted to increase by about 6 orders of magnitude and be relatively stable at that level (Fig. 10a). In a plan view, an ‘‘annular’’ region with high permeability can be seen (Fig. 10b). Fig. 10c and d indicate how the horizontal and vertical permeability changes 60 m behind the face. COSFLOW modelling results indicate that mining-induced fractures open and close following the stress decrease and recovery, causing permeability increase and decrease in the same manner. Two key aspects of the permeability in overburden strata are its magnitude and direction, both playing an important role in methane drainage. There exist zones with high cross-strata permeability and zones with high horizontal permeability. In the cross-strata permeable zones, vertical or subvertical and horizontal fractures are both well developed. Therefore both the horizontal and vertical permeability are high, encouraging gas flow in both horizontal and vertical directions. In the horizontal permeable zone, bed separation fractures are dominant and the horizontal permeability is much higher than the vertical permeability. Gas will mainly flow horizontally in these zones. Based on the discussion above and the relationship between permeability and de-stressing and fracturing, a conceptual model is suggested to describe the characteristics of permeable zones in the overburden strata of Panel 1115(1); see Fig.11. 4.5. CFD modelling of gas migration Mining induced gas flow is a combined consequence of methane desorption, ventilation and gas drainage. To understand the gas flow process, a 3D numerical study was carried out using

100 m

Coal 13-1 Goafside retained roadway Coal 11-2

Tailgate

LW face

Maingate

Fig. 12. CFD model and mesh.

CFD code Fluent. The study was built on the monitoring results and the COSFLOW results, and focused on the dynamic process of gas flow with ventilation and drainage.

4.5.1. CFD model The CFD model of Panel 1115(1) is shown in Fig. 12. The dimension of the model is 700 m in length, 240 m in width, and 100 m in height. The ventilation system has two inlet entries (the maingate and the tailgate) and one return entry (the goaf side retained roadway). The size of ventilation tunnels was 4 m wide and 3 m high. The boundary conditions of the two inlet entries were assigned as velocity while the return was set as outflow. The simulation was of steady-state flow under the specific ventilation and drainage conditions at the moment of actual panel retreat distance, which is equal to the model length. The permeability distribution in the CFD model was based on COSFLOW results and the CSIRO’s past modelling experience. The other input parameters including ventilation parameters, and methane emission rates were based on the actual gas measurement at Panel 1115(1) and operational experience at Huainan Coal mines. The velocities at the maingate and tailgate in the model were from actual measurement data at Panel 1115(1) and were 3.05 m/s and 0.72 m/s, respectively. The total methane emission rate of the LW panel was measured at about 25 m3/min. The emissions came from

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

137

Tailgate outby

Tailgate

Maingate

25m Tailgate Fig. 13. Modelled methane concentration. (a) Plan view and (b) Cross section.

Tailgate outby

Drainage borehole Velocity m/s Start -up

Coal 13-1

Mining level Fig. 14. Modelled gas migration pattern under drainage condition. (a) Horizontal plane 50m above the working seam and (b) Cross-panel width section 350m behind face (50m to borehole).

three sources, 25% was from the LW face and coal production, 5% from the remaining coal in the goaf, and 70% from the overlying Seam 13-1, based on measurements and analyses at the experiment site and other previously mined panels of similar conditions at Huainan coal mines [1]. These methane emission rates from different sources in the model were then calculated and modelled by using source terms in the species transport equation. The time-dependent variations of methane emission rates were not considered in the model. A vertical borehole was constructed in the model to simulate the impact of gas drainage on the goaf gas flow pattern. The completion of the borehole was simplified as an open hole. The cells in the borehole zone were defined as non-porous while the cells in the surrounding goaf were defined as porous. Borehole hydraulics was not considered since the borehole was connected to the desaturated goaf as a result of mining. A constant flow rate of 14 m3/min was assigned at the borehole outlet as the boundary condition.

methane concentration ( o30%) exists in the vicinity of the LW face and the roadways. The modelled methane concentration distribution for ‘‘Y’’ type of ventilation is shown in Fig. 13. 4.5.3. Gas flow under drainage Fig. 14 shows the modelled gas flow pattern when drained from a vertical borehole in the goaf. The key findings are: gas flows mainly along the perimeter of the goaf zone, an annular flow channel is formed, and gas released from adjacent coal seam flows horizontally at first to the perimeter of the goaf zone and then flows through the annular flow channels.

5. The three-dimensional zone for optimal methane drainage of Panel 1115(1) 5.1. Optimal zone for methane drainage

4.5.2. Gas concentration The modelled methane concentration in Panel 1115(1) increases with the height, the distance behind the face, and the lateral distance from the panel sides (i.e., roadways). Low

Optimal methane drainage relies on two key factors: constantly high gas flow and high methane concentration. To achieve such a goal, the drainage holes should be located in the zone with high methane desorption, high permeability, and high methane

138

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

concentration. Based on the understanding of overburden de-stressing, permeability distribution, and gas flow processes under drainage, it is suggested that the following factors be considered to define the optimal methane drainage zone: (1) High methane desorption: Determine the stress reduction for significant methane desorption in the coal seams under their specific geological conditions; then create and use the de-stressing model to estimate the zone of high methane desorption. (2) High permeability: Define the overburden fractured zone, and then create and use the permeability distribution model to estimate the zones with high cross-strata permeability and horizontal permeability, respectively. The upper limit of the optimal zone is often confined by the top boundary of the mining-induced horizontal permeable zone. (3) High methane concentration: Define the minimum methane concentration required in the drainage stream, and then, based on the gas flow and concentration pattern together with the methane desorption zone and permeability distribution, determine the zone for optimal methane drainage. The above steps are considered to be a practical method for determining the optimal methane drainage zone for multi-seam mining in low gas permeability environment.

5.2. Determination of parameters of optimal methane drainage zone of Panel 1115(1) 5.2.1. Estimate of critical stress reduction factors Seam 13-1 which overlies the working seam at Panel 1115(1) has a methane content of 5.7 m3/t and a gas pressure of 1.4–5.9 MPa. The isothermal adsorption curve shows that methane desorption occurs only when the gas pressure is reduced to below 0.7 MPa. Hence, the required stress reduction in this seam is high for efficient methane drainage. Based on the past experience at Huainan [1] and COSFLOW modelling results, it is estimated that the critical stress reduction factor for Seam 13-1 is 0.8, which implies that the in situ vertical stress at the seam level needs to be reduced by 80% for efficient methane drainage. Mining-induced fractures at a distance of more than 170 m behind the LW face in Panel 1115(1) are mostly compacted due to consolidation. The stress reduction factor at this location is about 0.2. This value can be considered as the threshold value for consolidation.

5.2.2. Estimate of the upper boundary of the optimal methane drainage zone The field monitoring measurements showed that the fractured zone can extend to a height of 145 m from the mining roof. The COSFLOW modelling results indicate that the zone with a stress reduction factor of 0.8 is about 139–160 m above the working seam. The zone with high horizontal permeability is about 150 m. From both the field measurements and modelling results, it is estimated that the upper boundary of the optimal methane drainage zone is 145 m above Seam 11-2. 5.2.3. Estimate of the lower boundary of the optimal methane drainage zone CFD modelling results suggest that the zone with low methane concentration ( o30%) under drainage can extend 30 m above the working seam. The COSFLOW model indicates that the zone with high cross-strata permeability can extend up to 54 m. It is estimated that the lower boundary of the optimal methane drainage zone is 30 m above the working seam, to maintain methane concentration above 30% in the drainage gas. 5.2.4. Estimate of upper and side boundaries of the optimal methane drainage zone Using the critical stress reduction factors of 0.8 for drainage and 0.2 for consolidation, and based on COSFLOW modelling results and the conceptual model of overburden de-stressing, it is estimated the angle of de-stressed zone is 751 in the mining direction and 781 along the face direction and the angle of consolidation zone is 791 in the mining direction and 911 along the face direction. The optimal methane drainage zone has a length of 170 m along the panel length from the LW face and 75 m along the panel width at the goaf level. 5.3. The three-dimensional zone for optimal methane drainage of Panel 1115(1) Based on the discussion in Section 5.2, a three-dimensional zone for optimal methane drainage of the Panel 1115(1) can be defined as shown in Fig. 15. This zone has an annular shape and is confined in height and width and bounded by sides of a certain inclination angle. Along the panel length, the optimal methane drainage zone can be divided into three sections: front, central and rear sections. The front section, adjacent to the LW face, is the key region where major methane desorption occurs. The central section is linked with the front and rear section from its sides which are the major 17-2 145m

13-1

78°

78°

Rear section

30m

Start-up

11-2

230 m -

Central section

Front section

17-2

13-1

91° 78°

78° 91°

11-2 Maingate

75 m

75m 17-2

LW face Tailgate 79°

75°

13-1

91°

78°

11-2 170m Front section

-

75 m Rear section

Fig. 15. The three-dimensional zone of optimal methane drainage in the overburden strata of Panel 1115(1).

H. Guo et al. / International Journal of Rock Mechanics & Mining Sciences 54 (2012) 129–139

flow channels for large scale gas migration. The rear section is in the vicinity of LW installation road, and is smaller in size than the front section and similar to the central section in shape. In the cross section along the panel length, the optimal methane drainage zone is confined by inclined side boundaries. The internal and external side boundaries have different inclination angles. In the cross section along panel width, the zone is located at a certain height from the goaf, and is confined by inclined side boundaries. The front section is larger than the rear section. 5.4. Applications of this study to guide methane drainage in multiseam mining The three-dimensional optimal methane drainage zone developed at Panel 1115(1) and the methodology used in this study provide a valuable approach to design of co-extraction of coal and methane at the Guqiao mine or other mines with similar conditions. The parameters of the optimal methane drainage zone at a specific site should be determined by means of field monitoring, laboratory tests, numerical modelling, and the specific geological and mining conditions. The following parameters and processes are suggested, a. The height of the lower boundary can be determined from the methane concentration requirement from the drainage operations. The optimum height is often located at the mid-height of the zone of high cross-strata permeability. b. The height of upper boundary can be determined from the height of de-stressed zone and that of high mining induced horizontal permeability. c. Inclination angle of external side boundaries can be estimated from the shape of the de-stressed zone. d. Length and width of the zone and internal side boundaries can be estimated from the shape of the goaf consolidation zone. 6. Conclusions An integrated study of LW mining-induced strata movement, stress changes, fractures, and gas flow dynamics was undertaken at Panel 1115(1) of the Guqiao Mine, a deep underground coal mine in China, by means of field monitoring, numerical modelling, and theoretical analysis. The key conclusions from this study are as follows. Field monitoring results indicate the abutment stresses can extend 300 m ahead of mining, active overburden strata movement and mining-induced fracturing occurs within 170 m behind the LW face, and the height of this mining-induced fractured zone can extend up to 145 m above the working seam roof. Water pressure reduces significantly within this fractured zone. The mining-induced stresses, rock fractures, gas flow, and the dynamic interaction between them were systematically investigated further with extensive COSFLOW and CFD modelling work. The spatial shape, location, stress and permeability conditions of the fractured zone were examined in detail. In addition, gas flow mechanisms and gas concentration distribution were studied in the fractured zone under the influence of mine ventilation, and gas drainage. It was found that drainage efficiency is controlled by three key factors: the degree of de-stressing, permeability distribution, and the pattern of gas flow. It was also found that at the middle and upper part within the fractured zone, there exists an annular gas flow channel along the perimeter of the LW panel, from which high concentration methane can be captured with a relatively high and stable flow rate.

139

It must be noted that the findings of this study were based on the specific geological and mining conditions at the Guqiao Mine. Similar studies are being carried out by CSIRO for different conditions in Huainan and Australia, to develop more generalised design guidelines for co-extraction of coal and gas.

Acknowledgements This study was financially supported by Huainan Coal Mine Group Ltd (HCMG) and CSIRO. The authors wish to acknowledge the significant contributions to this study by Dr Sheng Xue, Dr Stuart Craig, Dr Johnny Qin, Dr Deepak Adhikary, Dr Andy Wilkins, Dr Tom Liu of CSIRO, Mr Li Ping and Mr Liao Binchen of HCMG, and Mr Qin Yongyang, Mr Gao Song and Mr Sun Xingping of the Guqiao Mine.

References [1] Yuan L. Theory and Technology of Gas Drainage and Capture in Soft Multiple Coal Seams of Low Permeability. Beijing: China: Coal Industry Publishing House; 2004. [2] Guo H, Ishihara N, Fujioka M, et al. Integrated simulation of deep coal seam mining – optimisation of mining and gas management In: Australia–Japan Technology Exchange Workshop in Coal Mining 2001, Hunter Valley, 2001. [3] Yuan L. Theory and practice of integrated pillarless coal production and methane extraction in multiseams of low permeability. Eng Sci 2009;11:72–80. [4] Brady BHG, Brown ET. Rock Mechanics for Underground Mining. 2nd ed. London: Chapman & Hall; 1992. [5] Qian MG, Shi PW. Ground Pressure and Strata Control. Xuzhou: The Press of China University of Mining and Technology; 2003. [6] Morsy K, Peng SS Numerical modeling of the gob loading mechanism in longwall coal mines In: 21st international conference on ground control in mining; 2002, pp. 58–67. [7] Peng SS, Chiang HS. Longwall Mining. New York: Wiley; 1984. [8] Qian MG, Li HC. The movement of overlying strata in longwall mining and its effect on ground pressure. J China Coal Soc 1982;2:1–12. [9] Kendorski FS Effect of high-extraction coal mining on surface subsidence and ground water, In: Proceedings 12th conference on ground control in mining, Morgantown, WV; 1993. [10] Sweby G Review the caving mechanisms around high extraction systems and determine the effect of the mechanism on the safety of the system SIMRAC Project No: COL 327, 1997. [11] Palchik V. Formation of fractured zones in overburden due to longwall mining. Environ Geol 2003;44:28–38. [12] Palchik V. Localization of mining-induced horizontal fractures along rock layer interfaces in overburden: field measurements and prediction. Environ Geol 2005;48:68–80. [13] Thimons ED, Bielicki RJ, Kissell FN Using sulphur hexafluoride as a gaseous tracer to study ventilation systems in Mines Rep Invest 7916, U.S Bureau of Mines, 1974. [14] Balusu R, Ren TX, Humphries P Proactive inertisation strategies and technology development, ACARP Project C12020: Exploration and mining Report no: 907 F, 2005. [15] Ting XR, Rao B, Hua G Advanced coal mine gas capture and utilisation technologies for gassy mines In: Proceedings of the 5th international symposium on CBM/CDM in China, Beijing, 2005. [16] Yuan LM, Smith AC. Numerical study on effects of coal properties on spontaneous heating in longwall gob areas. Fuel 2007;87(15–16):3409–19. [17] Li Y, Li WP, Liu DX. The regression analysis on the relation between waterconductive fractured zone and mining thickness in Huainan mining area. Earth Environ 2005;33(Suppl):66–9. [18] Guo H, Adhikary DP, Gabeva D Hydrological response to longwall Mining ACARP Project 14033, CSIRO exploration and mining. Report P2007/692; 2007. [19] Guo H, Adhikary DP, Craig MS. Simulation of mine water inflow and gas emission during longwall mining. Rock Mech Rock Eng 2009;42:25–51. [20] Liu J, Elsworth D. Three-dimensional effects of hydraulic conductivity enhancement and desaturation around mined panels. Int J Rock Mech Min Sci 1997;34(8):1139–52. [21] Zhang H, Wang SZ, Peng GL, et al. A study on the mechanism and geological model of CBM reservoir-forming dynamic system in Huainan Coalfield. Coal Geol Explor 2005;33(4):29–34. [22] Zhang YG. Study of CBM exploitation potential in Huainan Coalfield. Coal Geol China 2005;17(4):21–3.