International Journal of Minerals, Metallurgy and Materials Volume 22, Number 12, December 2015, Page 1233 DOI: 10.1007/s12613-015-1190-z

Structure instability forecasting and analysis of giant rock pillars in steeply dipping thick coal seams Xing-ping Lai1,2), Huan Sun1,2), Peng-fei Shan1,2), Ming Cai3), Jian-tao Cao1,2), and Feng Cui1,2) 1) Energy School, Xi’an University of Science and Technology, Xi’an 710054, China 2) Key Laboratory of Western Mines and Hazard Prevention (Ministry of Education of China), Xi’an 710054, China 3) Bharti School of Engineering, Laurentian University, Sudbury, Ontario, Canada (Received: 6 June 2015; revised: 17 July 2015; accepted: 21 July 2015)

Abstract: Structure stability analysis of rock masses is essential for forecasting catastrophic structure failure in coal seam mining. Steeply dipping thick coal seams (SDTCS) are common in the Urumqi coalfield, and some dynamical hazards such as roof collapse and mining-induced seismicity occur frequently in the coal mines. The cause of these events is mainly structure instability in giant rock pillars sandwiched between SDTCS. Developing methods to predict these events is important for safe mining in such a complex environment. This study focuses on understanding the structural mechanics model of a giant rock pillar and presents a viewpoint of the stability of a trend sphenoid fractured beam (TSFB). Some stability index parameters such as failure surface dips were measured, and most dips were observed to be between 46° and 51°. We used a digital panoramic borehole monitoring system to measure the TSFB’s height (ΔH), which varied from 56.37 to 60.50 m. Next, FLAC3D was used to model the distribution and evolution of vertical displacement in the giant rock pillars; the results confirmed the existence of a TSFB structure. Finally, we investigated the acoustic emission (AE) energy accumulation rate and observed that the rate commonly ranged from 20 to 40 kJ/min. The AE energy accumulation rate could be used to anticipate impeding seismic events related to structure failure. The results presented provide a useful approach for forecasting catastrophic events related to structure instability and for developing hazard prevention technology for mining in SDTCS. Keywords: coal mining; structural instability; rock pillars; forecasting; acoustic emission (AE); steeply dipping coal beds

1. Introduction Prediction of structure instability related to the exploitation of coal mineral resources is a challenging task. Some catastrophic events such as roof collapse and mining-induced seismicity have occurred in coal mines in the Urumqi coalfield in western China [1−2]. These disasters occurred mainly because of structure instability of giant rock pillars sandwiched between steeply dipping thick coal seams (SDTCS). Techniques for predicting catastrophic rock failure of traditional long-wall mining in gently dipping coal seams have been developed [3−6]. However, these techniques are not applicable to mining in SDTCS. Hence, an urgent need exists to develop methods to forecast catastrophic structure instability of steeply dipping giant rock Corresponding author: Xing-ping Lai

pillars. The Urumqi coalfield has 30 coal seams with thicknesses varying from 20 to 50 m and dips ranging from 45° to 87° (Fig. 1 and Table 1). These coal seams are classified as SDTCS [7]. The present mechanized excavation depth is approximately 400 m. The horizontal section top-coal caving (HSTCC) method is normally used to recover SDTCS in underground coal mining. The length of a HSTCC working face is designed according to the inclination width of the coal seams. The maximum length of the HSTCC working face is 50 m, which is much shorter than that used in long-wall top-coal caving (LTCC) in gently dipping coal seams. The caving height is normally between 24 and 30 m, depending on the dip and width of the coal seams. The height of the undercut is approximately 3 m, and the upper coal layers are weakened by pre-blasting [8−9]. Multiple

E-mail: [email protected]

© University of Science and Technology Beijing and Springer-Verlag Berlin Heidelberg 2015

1234

Int. J. Miner. Metall. Mater., Vol. 22, No. 12, Dec. 2015

narrow voids are formed above the sub-horizontal section workings, which can be abruptly destabilized during coal

extraction. As a result, giant rock pillars on both sides of the stope can fail.

Fig. 1. Profiles of the Urumqi mining area, including simplified geological surface and sectional maps with main lithostratigraphic units. Table 1. Coal seam thicknesses and dips in six mines Mine Liudaowan

Weihuliang

Jiangou

Xiaohonggou

Dahonggou

Tiechanggou

Coal seam

Thickness / m

B1+2 B3+6 B7−B20 B21−B33 B1+2 B3+6 B7−B20 B21 B1+2 B3+6 B7−B20 B21−B34 B1+2 B3+6 B7−B20 B21−B32 B1+2 B3+6 B7−B20 B21−B33 B45 B42+43 B25−B40 B1−B22

24.7 39.0 36.3 15.6 28.3 41.1 39.1 22.3 33.8 49.2 52.7 39.5 32.4 44.9 49.8 40.2 30.7 57.5 48.1 51.2 28.4 36.9 47.3 50.2

Dip / (°) 60−70

54−72

80−88

86−88

82−88

39−53

Rocks between the roofs and floors of SDTCS at mines in the Urumqi coalfield form inclined giant rock pillars, which are sandwiched between the coal seams (Fig. 1). In situ stress in the pillars is high because of historical seism during the Jurassic period and because of the existence of the adjacent Xishan fault group [10]. Hence, the stresses in the pillars can easily increase abruptly as the coals are extracted. The reasons are mainly fracture propagation and even structure instability in the pillars. In fact, large-scale fractures have been observed on the ground surface in the mining area (Fig. 1). The giant rock pillars considered in this study have a width between 45 and 50 m and a dip between 45° and 87°. The height of instability block due to the giant rock pillars differentiation increases as mining progresses to deep levels. Numerous studies have been conducted to develop methods for rock mass stability analysis [11−16]. Rock mass stability analyses have been performed using physical model tests, numerical simulations, and other integrated intelligent testing technologies [11]. Andersson et al. [12−13] presented the results of the Äspö pillar stability experiment. Esterhuizen et al. [14] proposed a method of estimating pillar strength based on field observations of stable and failed pillars, supplemented by numerical modeling. Korzeniowski

X.P. Lai et al., Structure instability forecasting and analysis of giant rock pillars in steeply dipping thick coal seams

[15] used experimental results to develop a rheological Burgers model for hard rock pillars. Rinne et al. [16] presented modeling results showing fracture propagation and failure in a rock pillar under mechanical and thermal loadings. Some literatures have focused on coal pillar stability analysis [17−24]. Ghasemi et al. [17] proposed a new method to assess the overall risk of pillar recovery operation, which involves calculating the pillar recovery-risk (PR-Risk) indicator. Zhu et al. [18] reported the results of their numerical simulations research and experiments related to the effects of layer pressure relief under different coal pillar widths. Please et al. [19] performed fracture analysis of an Euler–Bernoulli beam in coal mine pillar extraction. Wattimena et al. [20] used the logistic regression method to develop a coal pillar stability chart. Singh et al. [21−22] assessed mining-induced stress development and dynamic loading in coal pillars during pillar recovery. Chen et al. [23] analyzed the rock burst danger when a fully mechanized caving coal face passed a fault with deep mining. Najafi et al. [24] predicted confidence intervals for stability analyses of chain pillars in coal mines, which they evaluated using the first-order second moment and the advanced second moment (ASM) methods. However, the literature contains little preliminary research on the stability analysis of rock pillars that are sandwiched between SDTCS. This study focuses on understanding a structural mechanics model of a giant rock pillar and presents a viewpoint of the formation of a trend sphenoid fractured beam (TSFB) and associated equilibrium equations. Some parameters in

1235

the equilibrium equations were determined from laboratory experiments, whereas others were determined from field monitoring. These equations provide a simple and efficient method for forecasting catastrophic events related to mining of SDTCS. FLAC3D was used to model giant rock pillars of different dips, which were disturbed by HSTCC. Vertical displacement contours in the giant rock pillars due to gravitation loading and extraction disturbance were obtained. In addition, statistical analysis of acoustic emission (AE) data obtained from field monitoring was conducted to reveal the AE energy accumulation characteristics of a giant rock pillar between the tail entry and the head entry at a mine site. The aim of these approaches was to develop a method for forecasting catastrophic events related to structure instability of the giant rock pillars and to establish hazard prevention technologies for mining in SDTCS.

2. Mechanics model 2.1. In situ geological and mining settings As previously mentioned, the giant rock pillars are sandwiched between SDTCS. The rock types of the giant rock pillars in this study are mainly siltstone, fine-grained sandstone, and locally shale (Fig. 2). The coals and rocks encountered are rigid compared with the same types of rocks elsewhere. The stability of the giant rock pillars depends on the geological conditions and on the mechanical properties of the coals and the rock masses, in situ stress, and the disturbance of extraction, ground support, etc. Hence, a mechanics model should take these factors into consideration.

Fig. 2. Lithology in the Xishanyao Fm (J2x) of the Jurassic series.

The Urumqi coalfield is located in the south of the Dzungaria Basin of China and at the south edge of Urumqi city and is adjacent to the Xishan fault group (Fig. 1). The basin, with a southward nappe, is formed by the Xishan fault group where the depth of décollement, which is greater in the north than that in the south, is between 10 and 15 km. All faults in the Xishan fault group converge to the décollement whose overburden depth is approximately 11 km. The Urumqi coalfield is composed of the Badaowan syncline and the Qidaogou anticline. Xiaohonggou coal mine, which contains the giant rock pillar investigated in this study, is located in the Badaowan syncline, along with a few other

coal mines. The Urumqi coalfield has 30 steeply dipping coal seams with different thicknesses, and the region has no groundwater presence. The region is also ecologically fragile and is under ecological stress. The stratum thickness varies from 513.77 to 902.90 m. Among these strata, there are more than 40 layers of coal seams that can be classified into 4 groups and only 33 layers are minable seams with thicknesses that vary between 117.05 and 175.45 m. The coal seam inclination angles vary unevenly from 46° to 67° along the trend of the coalfield and from 63° to 88° from north to south. The main minable coal seams, i.e., the B1+2 and B3+6 seams, lie

1236

on the Xishanyao Fm (J2x) of the Jurassic series, with coal seam thicknesses between 24.70 and 33.81 m and between 39.04 and 57.50 m, respectively. The main and immediate roofs or floors are mainly composed of siltstone and shale (Table 2), and the false roofs or floors that are sandwiched between immediate roofs or floors and coal seams are mainly carbon mudstone with a uniaxial compressive strength less than 30.0 MPa. Exploration results indicate that the existing coal reserves among all mines combined is 1.893  109 t. Two large-scale V-shaped collapsed grooves (Fig. 1) are located on both sides of the giant rock pillar between the B1+2 and B3+6 coal seams. The collapsed groove on top of the B1+2 coal seam is deeper than that on top of the B3+6 coal seam because the mining depth is relatively large and the mechanical properties of the rocks are relatively poor in the B1+2 coal seam. Furthermore, the size of the V-shaped collapsed grooves can grow along the structure instability of the giant rock pillar caused by repeated HSTCC. Rocks can move into the large-scale V-shaped collapsed grooves on the surface as a consequence of the structure instability. Meanwhile, noxious gases can enter into the multiple narrow mine voids and the mining conditions are very challenging. 2.2. Model description The giant rock pillar sandwiched between the B1+2 and B3+6 coal seams is shown in Fig. 3(a). The B1+2 coal seam working face carries the burden of the giant rock pillar. As the working face of the B1+2 coal seam was advanced, residual coal and rock masses above the horizontal section caved in, resulting in the formation of multiple narrow mine voids. Meanwhile, the inclined rock pillar was hanging beside the large-scale V-shaped collapsed groove and shear failure occurred slowly under gravitational loading. Fractures within the giant rock pillars initiated and propagated along the trend of the coal seam. The in-plane outline of the giant rock pillar resembled a sphenoid shape, and it eventually evolved into a rock beam and fractured along the trend. This structure of the giant rock pillar is referred to as a TSFB (Fig. 3(b)). 2.3. Analysis of the trend sphenoid fractured beam The failure of rock pillars has been studied extensively in laboratories, with a focus on verifying numerical simulations. However, very few studies focused on steeply dipping rock pillars between thick coal seams have been reported. The TSFB is one part of the giant rock pillar and it can fail in shear as mining progresses. The Mohr–Coulomb failure criterion and the equilibrium

Int. J. Miner. Metall. Mater., Vol. 22, No. 12, Dec. 2015

of a simply supported beam were used to analyze the TSFB. A section of a unit thickness (1.0 m) of the beam was taken to research the failure mechanism along the y direction, as shown in Fig. 4. Rock failure was modeled using the Mohr–Coulomb failure criterion. If σps and τps are assumed to be the stresses acting perpendicular and parallel to the failure surface, respectively, then the Mohr–Coulomb failure criterion is expressed as

 ps   ps  tan   c

(1)

where φ is the friction angle and c is the cohesive strength of the rock mass, which can be obtained from laboratory tests. Stresses σps and τps are considered to result from the gravitation loading of the rock mass. If the failure surface dip is assumed to be α, then Eq. (1) can be expressed as (Gps  sin 2  ) H  (Gps  sin 2 2H )  tan   c

(2)

where Gps is the weight of the layer section and ΔH is the height of the unit section (1 m thickness) of the beam. Under the assumption that the rock mass is homogeneous along the trend of the beam, the weight of the unit section of the beam GTSFB according to mathematical integration theory can be calculated from L

GTSFB   Gps ·d y 0

(3)

where L is the beam length. When fractures propagate in the pillar, the TSFB will be formed. With respect to the structural mechanics, the beam can be considered as a typical simple supported beam. As shown in Fig. 5, the beam is under gravitational loading. Hence, the instability of the beam can be considered as a rock mass beam that cannot resist uneven distribution of gravity loading along the trend. One bracket of the beam is fixed and the other can slide (Fig. 5). Both of them carry the beam weight. Hence, the problem of the TSFB can be treated using the beam theory in structure mechanics. As shown in Fig. 5, the length of TSFB is L. Point A is hinge supported, whereas point B is roller supported; the loads at points A and B are VAx and VBx, respectively. The weight of the rock mass beam can be regarded as a uniform-distribution loading along the y direction; along the x direction, the load distribution is non-uniform. The load intensity is defined as qz. According to the equilibrium equations of moment and force, the rock mass beam maintains equilibrium after the following conditions are met:

 F  0    M  0 z

(4)

A

where ∑Fz is all of stress along the z direction, and ∑MA is

X.P. Lai et al., Structure instability forecasting and analysis of giant rock pillars in steeply dipping thick coal seams

1237

Table 2. B1+2 and B3+6 coal seams with all important roof and floor information in the study area Coal seam

B3+6

B1+2

Name

Lithology

Thickness / m

Description

Main roof

Siltstone

24.49

Immediate roof

Shale

2.15

Lamellar and joints are obvious, soaking loose, unstable

Hard and fine siltstone; rocks solidified entirely and stable

Main floor

Siltstone

4.00

Hard and fine siltstone; rocks solidified entirely and stable

Immediate floor

Shale

1.35

Lamellar and joints are obvious, soaking loose, unstable

Main roof

Siltstone

24.49

Immediate roof

Shale

2.15

Joints are obvious, fragile, soaking loose, unstable

Main floor

Siltstone

8.50

Fine siltstone with micro-granular structures, dark grey, hard and stable

Immediate floor

Shale

1.00

Joints are obvious, fragile, soaking loose, unstable

Fine siltstone with micro-granular structures, dark grey, hard and stable

Fig. 3. Structure description of the giant rock pillar: (a) working face and coal-rock mass distributions; (b) failure behavior and configuration of the trend sphenoid fractured beam.

1238

Int. J. Miner. Metall. Mater., Vol. 22, No. 12, Dec. 2015

Fig. 4. Analytical model of a unit section (1 m) of the trend sphenoid fractured beam.

Fig. 5. Rock mass beam mechanics model of the trend sphenoid fractured beam: (a) rock mass beam model in three dimensions; (b) rock mass beam model in a single plane.

all of torques for point A. Using qz, VAx, VBx, L, etc., we can rewrite Eq. (4) as VAx d x   VBx d x   qz d y  0  0 0   Δx L2 L L   VBx d x    qz d y  0 (5)  0 2 0  qz  Gps  sin    where VAx and VBx are the forces at the beam ends, qz is the load intensity, L is the length of the rock mass beam, Gps is the weight of the unit section of the beam, and α is the dip of the failure surface. Some parameters in Eq. (2) were obtained from laboratory tests conducted at the Key Laboratory of Western Mine Exploitation and Hazard Prevention at Xi’an University of Science and Technology, China. Facilities for coal and rock mechanics experiments have been constructed to obtain the



x

0

Δx

L

mechanical properties of rock samples taken from the field. Some parameters such as dip (α, depth (D) of fractures, and height (ΔH) of the TSFB were determined from field monitoring. Monitoring of the behavior of the giant rock pillar was conducted using a digital panoramic borehole monitoring system (DPBMS), which integrates imaging, digital, and intelligent identification technologies. The most important component of the monitoring system is a panoramic camera probe equipped with a magnetic compass to measure the orientation of the boreholes. The widths of the fractures can be identified from the images of the borehole walls. Meanwhile, the terminal computer can record the depth of the panoramic camera probe. All imaging signals and measurement data are sent to the terminal computer for processing and analysis. The digital panoramic borehole monitoring system was mainly used to measure the width and the angle of the frac-

X.P. Lai et al., Structure instability forecasting and analysis of giant rock pillars in steeply dipping thick coal seams

1239

ground surface as illustrated in Fig. 6. The parameters of the boreholes are listed in Table 3, along with the position of the large-scale fracture identified from the logging data. For example, the  of the fracture and the depth of the fracture from the collar identified in borehole 1 are 51.0° and 16.0 m, respectively.

tures formed in the giant rock pillar. Failure of the giant rock pillar is a combination of shear fracture with tensile failure (Fig. 3). We classify fractures that are over 300 mm in width on the ground surface as large-scale fractures. In situ logging using the digital panoramic borehole monitoring system was conducted in four boreholes drilled from the

Fig. 6. Logging of four blast boreholes in the giant rock pillar using the DPBMS system. Table 3. Parameters of boreholes used for logging by the DPBMS system and the dips and locations of the large-scale fracture Borehole No.

1

2

3

4

Borehole parameters Position (x, y, z)

DPBMS information

Dip / (°)

Radius / mm

Length / m

α / (°)

Depth / m

ΔH / m

90

150.0

230.0

51.0

16.00

60.50

90

150.0

230.0

49.0

28.34

56.37

90

150.0

230.0

46.0

39.90

50.74

90

150.0

230.0

49.0

50.30

56.37

x = 4864795.27, y = 29562067.95, z = 878.07 x = 4864788.06, y = 29562072.19, z = 878.17 x = 4864780.84, y = 29562076.36, z = 878.18 x = 4864772.19, y = 29562081.37, z = 878.11

Some measures of preventing the giant rock pillar failure, which include pre-blasting and hydro-fracturing, have been implemented successfully at Xiaohonggou coal mine. For example, a measure of blasting the giant rock pillar had been implemented to release ground pressure and the hy-

dro-fracturing technique was used in the inner giant rock pillar to prevent the occurrence of some structure failures. Meanwhile, we have obtained some data related to the giant rock pillar (Table 4) from laboratory testing and field monitoring. For example, the strength parameters (c and φ) were

Table 4. Theoretical computation results and parameters of the digital panoramic borehole monitoring system (DPBMS) testing of giant rock pillars Name

Lithology

φ / (°)

c / MPa

α / (°)

ΔH / m

L/m

Gps / (105 kN)

GTSFB / (105 kN)

Giant rock pillar

Siltstone

26.70

3.70

46.0 to 51.0

56.37−60.50

93.0

6.25−7.05

581.5−656.1

1240

Int. J. Miner. Metall. Mater., Vol. 22, No. 12, Dec. 2015

obtained from laboratory uniaxial compressive strength tests of rock specimens obtained from the mining area in the giant rock pillar at Xiaohonggou mine and the dip, depth, and length of the giant rock pillar were obtained from field monitoring. GTSFB was calculated using Eq. (3).

3. Numerical simulation 3.1. Constitutive model Numerical modeling was carried out using FLAC3D to study the deformation behavior of the giant rock pillar

sandwiched between the B1+2 and B3+6 coal seams. Three dips of the giant rock pillar, i.e., 87°, 63°, and 45° were considered in the modeling. The thickness of the B1+2 coal seam is 30.0 m, and the excavation depth is 80 m (Fig. 7). The thickness of the B3+6 coal seam is 49.0 m, and the excavation depth is 100.0 m. The model size is 600 m along the trend, 250 m in the height direction, and ranges from 210 to 446 m in the other horizontal direction. The depth of the collapsed area on the surface is 50 m. In the case of the collapsed area, the angles of the slope that lies on the side of the roof and the floor are 87° and 65°, respectively. All

Fig. 7. Models for FLAC3D simulation under different dip conditions: (a) 87°; (b) 63°; (c) 45°.

X.P. Lai et al., Structure instability forecasting and analysis of giant rock pillars in steeply dipping thick coal seams

1241

regular and sphenoid sections along the horizontal direction. As fractures propagate within the giant rock pillar, the TSFB is gradually formed. The dip of the giant rock pillar is large; however, the areas of the irregular and the sphenoid section are small, resulting in a small vertical displacement area and large vertical displacements. As a result, structure instability easily occurs in the giant rock pillar when the coal seems dip at 87°. In fact, field seismic monitoring results revealed that the structure instability problem exists in mines in the most SDTCS.

model dimensions are shown in Fig. 7. The physical and mechanical properties (Table 5) of the pillar and the roof and floor rocks were obtained from laboratory tests. 3.2. Caving disturbance to the giant rock pillar The FLAC3D model for the SDTCS mining simulated 100 and 80 m depths of mining along the 600 m trend of the B1+2 and B3+6 coal seams, respectively. Fig. 8 presents the contours of the vertical displacement in the giant rock pillar, showing the influence of the excavation disturbance of the coal seams. The giant rock pillar, which is under gravitation loading, is affected by the excavation disturbance. The pillar has ir-

3.3. Structure instability The disturbance to the giant rock pillar caused by

Table 5. Physical and mechanical parameters of the pillar and the roof and floor Name

Bulk modulus / GPa

Shear modulus / GPa

Tensile strength / MPa

Cohesion / MPa

Friction angle / (°)

Volume weight / (kNm3)

Main roof

8.30

4.10

2.39

3.70

30.33

24.83

Immediate roof

7.96

3.83

1.41

3.63

26.70

20.08

False roof

7.96

3.83

1.41

3.63

26.70

20.08

Coal

3.70

1.79

1.12

2.95

25.90

12.60

Giant rock pillar

8.30

4.10

2.39

3.70

26.70

24.83

Immediate floor

7.96

3.83

1.41

3.63

26.70

20.08

Main floor

8.30

4.10

2.39

3.70

30.33

24.83

Fig. 8. Vertical displacement contours of giant rock pillars with different dips, and (ac) represent that the dip angles of giant rock pillars are 87°, 63°, and 45°, respectively.

1242

Int. J. Miner. Metall. Mater., Vol. 22, No. 12, Dec. 2015

HSTCC demonstrated that the rock mass surrounding the mined-out area was destabilized dynamically because of the formation and instability of the TSFB. This destabilization can induce a roof fall of the tail entry and rib spalling of the head entry and can trigger localized pressure aberration of the working face leading to rock support failure. An integrated AE monitoring system (KJ-623) was installed at the Xiaohonggou coal mine. AE monitoring is an efficient technique to monitor the dynamic behavior of giant rock pillar structure deformation and instability. Before the structure instability of the giant rock pillar occurs, many micro-fracturing events are observed in the rock masses. The intensity of AE activities is measured using parameters such as the AE energy accumulation rate (kJ/min) and frequency (min1). These AE parameters reflect fracture conditions in the rock masses. All AE waves at the mine site were received from AE sensors installed in the rock masses. The frequency ranges of the AE sensors range from 300 to 2000 Hz, with a sensitivity of (65 ± 3) dB. This AE monitoring system covered the working faces of the B1+2 and B3+6 coal seams at multiple mining levels. The AE data obtained from the field monitoring were combined with some seismic ac-

tivity data to assess the rock mass stability conditions. Moreover, the layout of the AE sensors in the monitoring section was properly considered to ensure the quality of the data. The in situ AE monitoring program constitutes one of the efforts dedicated to understanding the process of giant rock pillar instability during coal mining. Fig. 9 presents the layout of the AE sensors in the monitoring section. Before the excavation of the B1+2 and B3+6 coal seams, 14 AE sensors were installed in the monitoring section (Fig. 9). The spacing between the sensors was 40 m, and the sensors were installed in the rock mass on both sides of the walls at chainages from 1300 to 1820 m along the trend of the giant rock pillar. The sensors were divided into two groups, with sensors A1 to A7 in the B3 head entry and sensors A8 to A14 in the B2 tail entry. The height difference between the two groups of sensors was 20 m. Data were acquired for three continuous hours using the AE monitoring system, and the recorded data were processed automatically to eliminate noise. A seismic event occurred at the +500 m level of the B2 tail entry at 11:30 AM on September 18, 2013. This event was considered to be associated with the giant rock pillar structure instability.

Fig. 9. Layout of AE sensors in the monitoring section.

Fig. 10 presents the AE energy accumulation rates along with the seismic event that occurred at the +500 m level of the B2 tail entry and the B3 head entry at the Xiaohonggou coal mine. Analyzing the AE energy accumulation rates could reveal some information about the giant rock pillar structure mobilization and instability. These pre- and post-event information parameters can be considered for forecasting catastrophic events. From the collected data, the AE energy accumulation rates were divided into different stages as shown in Fig. 10. On the basis of a detailed analysis of the data, the following conclusions were reached. (1) According to the AE energy accumulation rates of the B2 tail entry and the B3 head entry at the +500 m level, the rock deformation process can be divided into three stages:

normal, increasing, and fluctuation (Fig. 10). The energy accumulation rates at the normal stage in the B2 tail entry and the B3 head entry range from 10 to 20 kJ/min. However, the AE energy accumulation rates at the increasing stage range from 20 to 30 kJ/min in the B3 head entry and from 20 to 40 kJ/min in the B2 tail entry, respectively. The AE energy accumulation rates at the fluctuation stage range from 30 to 40 kJ/min and from 40 to 70 kJ/min in the B3 head entry and the B2 tail entry, respectively. The AE energy accumulation rates at the increasing stage increase sharply. In contrast, the AE energy accumulation rates at the fluctuation stage exhibit a decreasing trend with a large fluctuation. (2) The giant rock pillar can fail in shear and form a

X.P. Lai et al., Structure instability forecasting and analysis of giant rock pillars in steeply dipping thick coal seams

sphenoid rock ply section. Meanwhile, fractures in the giant rock pillar propagate as the coal seam extraction continues, eventually forming the TSFB. Using FLAC3D, we demonstrated that the TSFB can be formed as a consequence of the giant rock pillar vertical deformation caused by coal seam mining. These processes occurred mainly in the increasing AE energy accumulation rate of the fluctuation stage. Hence, the AE energy accumulation rates could potentially be used to forecast structure instability of the giant rock pillar. (3) The B2 tail entry and the B3 head entry in the +500 m

1243

level lie on the both sides of the giant rock pillar. Structure instability of the TSFB dynamically affected the rock mass stability of the B2 tail entry at the +500 m level when some seismic activities occurred. After the structure instability of the TSFB occurred, the giant rock pillar unloaded a portion of the stresses, which was beneficial for the stability control of coal and rock masses in the B3 head entry. This effect was obvious as shown in Fig. 10, where the AE energy accumulation rates in the B3 head entry were, in general, smaller than those in the B2 tail entry.

Fig. 10. (a) AE energy accumulation rates in the B3 head entry in the +520 m level; (b) AE energy accumulation rates in the B2 tail entry in the +500 m level.

4. Conclusions (1) Structure instability of giant rock pillars sandwiched between SDTCS can induce potentially catastrophic events. Seismic events occurred frequently because of structure instability of TSFBs. A TSFB can be considered as a simply supported beam, and the Mohr–Coulomb failure criterion and equilibrium equations of a simply supported beam can be used to analyze its structural stability. At Xiaohonggou coal mine, the dip (α) of the failure plane of the TSFB ranges from 46° to 51°, the length (L) of the beam is 93 m, and the height (ΔH) ranges from 56.4 to 60.5 m. (2) FLAC3D was used to simulate the distribution and evolution of the giant rock pillar’s vertical displacement with dips of 87°, 63°, and 45°. A steeper giant rock pillar results in smaller areas of the irregular and sphenoid sections, which means that the area of large vertical displacement is also small. However, the vertical displacements are larger if the inclination of the giant rock pillar is steeper. These results indicate that a seismic event is more likely to occur in the giant rock pillar with a dip of 87°. The AE energy accumulation rates show that structure instability occurs mainly at the stages of the increasing AE energy accumulation rate of the fluctuation stage. The change of the AE

energy accumulation rates, which commonly range from 20 to 70 kJ/min, can potentially be used for forecasting structure instability. (3) The structure instability of the TSFB had a destructive impact on the ground surface. Large-scale mine voids have evolved into V-shaped collapsed grooves, which can be observed on the surface. As coal seam mining progresses to deeper levels, the V-shaped collapsed grooves can grow in size and can potentially further affect the surface conditions. This issue needs to be addressed in future mining at the mine site.

Acknowledgements This work was financially supported by the Key National Basic Research Program of China (Nos. 2014CB260404 and 2015CB251602), the Key National Natural Science Foundation of China (No. U13612030), Shaanxi Innovation Team Program (No. 2013KCT-16), and the High Technology Development Program of XinJiang Municipality (No. 201432102).

References [1]

F.H. Ren, X.P. Lai, and M.F. Cai, Dynamic destabilization

1244

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

Int. J. Miner. Metall. Mater., Vol. 22, No. 12, Dec. 2015 analysis based on AE experiment of deep-seated, steep-inclined and extra-thick coal seam, J. Univ. Sci. Technol. Beijing, 15(2008), No. 3, p. 215. S.J. Miao, X.P. Lai, and F. Cui, Top coal flows in an excavation disturbed zone of high section top coal caving of an extremely steep and thick seam, Min. Sci. Technol. China, 21(2011), No. 1, p. 99. Z.Y. Tan and M.F. Cai, Measurement and study of the distributing law of in-situ stresses in rock mass at great depth, J. Univ. Sci. Technol. Beijing, 13(2006), No. 3, p. 207. T.K. Lu, Y.Z. Liu, and F.S. Xu, Deformation and failure of stratified weak roof strata of longwall roadway, J. Univ. Sci. Technol. Beijing, 14(2007), No. 5, p. 387. A. Vakili and B.K. Hebblewhite, A new cavability assessment criterion for longwall top coal caving, Int. J. Rock Mech. Min. Sci., 47(2010), No. 8, p. 1317. V.V. Khanzode, J. Maiti, and P.K. Ray, A methodology for evaluation and monitoring of recurring hazards in underground coal mining, Saf. Sci., 49(2011), No. 8-9, p. 1172. J.G. Zhang, Z.Q. Zhao, and Y. Gao, Research on top coal caving technique in steep and extra-thick coal seam, Procedia Earth Planet. Sci., 2(2011), p. 145. X.P. Lai, P.F. Shan, J.T. Cao, H. Sun, Z.Y. Suo, and F. Cui, Hybrid assessment of pre-blasting weakening to horizontal section top coal caving (HSTCC) in steep and thick seams, Int. J. Min. Sci. Technol., 24(2014), No. 1, p. 31. Y. Luo and Z.W. Shen, Study on orientation fracture blasting with shaped charge in rock, J. Univ. Sci. Technol. Beijing, 13(2006), No. 3, p. 193. Y.K. Ran, L.C. Chen, J. Shen, J. Li, and H.L. Gong, Xishan fault group near Urumqi city and paleo-earthquake identification on reverse fault, Seismol. Geol., 29(2007), No. 2, p. 218. C.X. Yang, Z.Q. Luo, G.B. Hu, and X.M. Liu, Application of a microseismic monitoring system in deep mining, J. Univ. Sci. Technol. Beijing, 14(2007), No. 1, p. 6. J. Christer Andersson and C. Derek Martin, The Äspö pillar stability experiment: Part I. Experiment design, Int. J. Rock Mech. Min. Sci., 46(2009), No. 5, p. 865. J. Christer Andersson, C. Derek Martin, and H. Stille, The Äspö pillar stability experiment: Part II. Rock mass response

[14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

to coupled excavation-induced and thermal-induced stresses, Int. J. Rock Mech. Min. Sci., 46(2009), No .5, p. 879. G.S. Esterhuizen, D.R. Dolinar, and J.L. Ellenberger, Pillar strength in underground stone mines in the United States, Int. J. Rock Mech. Min. Sci., 48(2011), No. 1, p. 42. W. Korzeniowski, Rheological model of hard rock pillar, Rock Mech. Rock Eng., 24(1991), No. 3, p. 155. M. Rinne, B.T. Shen, and T. Backers, Modelling fracture propagation and failure in a rock pillar under mechanical and thermal loadings, J. Rock Mech. Geo. Eng., 5(2013), No. 1, p. 73. E. Ghasemi, K. Shahriar, and M. Sharifzadeh, A new method for risk assessment of pillar recovery operation, Saf. Sci., 48(2010), No. 10, p. 1304. H.Z. Zhu, P. Liu, and Z.Y. Tong, Numerical simulation research and application on protected layer pressure relief affection under different coal pillar width, Procedia Eng., 84(2014), p. 818. C.P. Please, D.P. Mason, C.M. Khalique, J.M.T. Ngnotchouye, A.J. Hutchinson, J.N. van der Merwe, and H. Yilmaz, Fracturing of an Euler-Bernoulli beam in coal mine pillar extraction, Int. J. Rock Mech. Min. Sci., 64(2013), p. 132. R.K. Wattimena, S. Kramadibrata, I.D. Sidi, and M.A. Azizi, Developing coal pillar stability chart using logistic regression, Int. J. Rock Mech. Min. Sci., 58(2013), p. 55. R. Singh, A.K. Singh, J. Maiti, P.K. Mandal, R. Singh, and R. Kumar, An observational approach for assessment of dynamic loading during underground coal pillar extraction, Int. J. Rock Mech. Min. Sci., 48(2011), No. 5, p. 794. A.K. Singh, R. Singh, J. Maiti, R. Kumar, and P.K. Mandal, Assessment of mining induced stress development over coal pillars during depillaring, Int. J. Rock Mech. Min. Sci., 48(2011), No. 5, p. 805. X.H. Chen, W.Q. Li, and X.Y. Yan, Analysis on rock burst danger when fully-mechanized caving coal face passed fault with deep mining, Saf. Sci., 50(2012), No. 4, p. 645. M. Najafi, S.E. Jalali, A.R. Yarahmadi Bafghi, and F. Sereshki, Prediction of the confidence interval for stability analysis of chain pillars in coal mines, Saf. Sci., 49(2011), No. 5, p. 651.