Advertisement

Physical simulation research on evolution laws of clay aquifuge stability during slice mining

  • Shizhong Zhang
  • Gangwei Fan
  • Dongsheng Zhang
  • Qizhen Li
Original Article

Abstract

The aquifuge stability is the key to study the impacts of coal mining on the aquifer. Based on the geological conditions of a mine in Yili of Xinjiang, China, this paper has studied the stability evolution laws of clay aquifuge during extremely thick coal seam mining by similar material simulation experiment in the laboratory. For the water-swelling and expansion property of clay aquifuge, the reasonable proportion of the similar material is firstly determined by taking the uniaxial compressive strength and the permeability coefficient as core indexes. Then, the overlying strata movement coupled solid–liquid physical model is established. In addition, the aquifuge deformation, the water level changes of the aquifer, and the height of fracture zone in overburden are analyzed during the slice mining. The research results indicate that the clay aquifuge will gradually occur instability failure during the mining of the working face, and the aquifuge stability has the threshold effects. When the ratio of the vertical displacement of the aquifuge to the thickness is Dv/T ≤ 58.0%, the ratio of the horizontal displacement to the thickness is Dh/T ≤ 17.0%, and the height of fractured zone in overburden is below the aquifuge, the mining-induced fractures may be closed and the aquifuge stability could be maintained. If Dv/T ≥ 75.0%, Dh/T ≥ 23.9%, and the height of fractured zone in overburden is within the aquifuge, the fractures will develop and connect the aquifuge and the stability failure of the aquifuge will occur, which has a direct correlation with the mining height.

Keywords

Similar material Solid-liquid coupling Slice mining Aquifuge Overlying strata movement 

Introduction

Water is a valuable resource; however, underground mining often has an adverse effect on water resources (Goswami 2013; Katoria et al. 2013). Moreover, large-scale and intensive underground mining is frequently accompanied by violent overburden rock caving and wide fracture development that can lead to the formation of highly permeable zones of groundwater flow. Such conditions can cause rapid groundwater inrush and flooding problems in mines (Xiao et al. 2010; Chen and Yang 2011) and damage to human life and property. Furthermore, the dewatering of coal mines can lead to the loss of surface water or groundwater and cause a series of ecological environmental problems such as the loss of river flow and reduced biodiversity (Hasan et al. 2013; Rashid et al. 2014; Li et al. 2015).

A previous study found that the aquifuge below an aquifer generally has a compact structure with poor water permeability or rich clay minerals with water-swelling and expansion properties that could block water infiltration or run off from the aquifer (Fan and Zhang 2015). In the process of coal mining, aquifuge stability can directly affect the loss of water in the upper aquifer or surface water and thus could influence mine safety and the regional ecological environment. Therefore, research on the stability of an aquifuge during mining has great importance on mining safety and regional environmental protection. Water resources are extremely valuable in Northwest China, a region with an arid or semiarid climate and a fragile ecological environment, particularly in the Xinjiang region. Coal seams in this region are characterized by shallow depth, large thickness (tens of meters to hundreds of meters), and wide distribution. These factors have made these deposits the focus of China’s coal exploitation.

However, sediments associated with coal deposits in the region have been subjected to only a short period of diagenesis; consequently, they are weakly consolidated and cemented and are highly sensitive to engineering disturbance. The inconsonance of the coal mining demand and the ecological environment in Northwest China makes the study of the evolution laws of aquifuge stability and their influence on water resources an important research topic.

Many experts and scholars have conducted related research studies on aquifuge (or rock mass) stability. Zhang et al. (2017b) proposed brittleness evaluation indices of the mechanical characteristics before and after rock failure based on the energy variation in the complete process of rock failure to evaluate the brittle failure mode and the brittle strength. Using a numerical simulation, Yao et al. (2011) found that coal mining can affect the impermeability of rock by changing the permeability coefficients of the overburden rock, and the permeability coefficients gradually decrease up from the immediate roof. Cheng et al. (2017) found that a thick aquifuge and low water pressure can decrease the water inrush risk on coal seam floors and that the seam thickness has only a small influence on the floor-water inrush rate. Based on field measurements, Zhang et al. (2010, 2011) and Booth et al. (2000) determined that when the aquifuge is not completely damaged by mining of the working face, the aquifuge can be stably maintained to a certain degree of stability, and that the groundwater level can be recovered over time. Some scholars have conducted similar material simulation tests to study the effects of coal mining on an overburden aquifer. Huang (2014) conducted a coupled solid–liquid physical simulation and determined that upward cracks and downward cracks are developed in the aquifuge during the mining process. The downward cracks are the main channels of groundwater loss. Using physical simulations, Feng et al. (2010) and Fu and Tu (2015) found that the structural instability of the floor aquifuge or seepage-induced instability causes the aquifuge to lose its water-resistant ability. They also found that the dynamic instability of the floor aquifuge and mine water inrush can be sudden. Therefore, as a common research method, the simulation of similar materials is one of the key means to study overburden (including the aquifer and aquifuge) deformation and failure characteristics during the mining process.

In the process of coal mining, the general rules of aquifuge deformation and failure have been studied using various research methods. However, the dynamic evolution laws of aquifuge stability during mining and the effect of the aquifuge movement on the water level of the overlying aquifer remain unclear. Therefore, it is necessary to conduct an in-depth study of aquifuge stability. The use of physical models to simulate this behavior has the advantages of being straightforward to undertake and capable of providing visual results; thus, such models can more accurately reflect real situations, provided that the physical properties of materials in the model are comparable with the deformation behavior of sediments in the field, particularly if there are similarities between materials in terms of their uniaxial compressive strength, elastic modulus, and permeability coefficients, among other factors. Suitable materials for solid–liquid coupling have been successfully fabricated (Zhang and Hou 2004; Li et al. 2012). However, some limitations remain deficiencies in considering the ability of model systems to simulate the effects of weak cementation and water swelling, for example. Therefore, ratios of similar materials for clay aquifuge should be developed and determined first. Furthermore, existing similar material simulation systems could be used to study the solid–fluid coupling, overburden fractures, and seepage laws (Jin et al. 2014); however, this approach also has limitations in the areas of visual tracking and discrimination of fractures. The use of common pigment tracers, such as ink (Zhang and Hou 2004), often cannot reveal whether fractures are open or closed due to pigment residues, particularly when the early open fractures gradually close with the advancement of a mining operation, making the true fracture state challenging to distinguish. Therefore, it is necessary to improve the simulation system before performing experiments.

Based on previous research (Huang and Zhang 2011; Zhang et al. 2011; Fan and Zhang 2015) and according to the water-swelling and expansion properties of clay aquifuge in China’s Xinjiang region, fine river sand, bentonite, gypsum, and methyl silicone oil were selected as the main raw materials to simulate the behavior of the clay aquifuge in this study. The uniaxial compressive strength and permeability coefficient were taken as the core indices. The single-factor method was used to determine the reasonable proportion of the clay aquifuge. On this basis, a physical model for solid–fluid coupling in overburden rock movement was established in the first mining area of a coal mine in Yili. Visual tracking and discrimination modules of the fracture growth simulation were introduced. Moreover, aquifuge deformation, water level changes in the aquifer, and the height of the fracture zones in the overburden were analyzed during the process of coal seam mining. These research results provide a reference for further simulation using similar materials.

Methods and materials

Materials used to simulate aquifuge properties

Fine river sand, bentonite, gypsum, and methyl silicone oil were selected as the main ingredients of the material used to simulate the deformation behavior of the clay aquifuge. The fine river sand was used as aggregate with a size of less than 1.18 mm. The bentonite, used as the additive, was composed primarily of water-swelling montmorillonite. The bentonite was used to simulate the water-swelling and expansion properties of the clay aquifuge. The gypsum was used as an intensity regulator to adjust the uniaxial compressive strength of similar materials. The methyl silicone oil was used as a hydrophilic material to adjust the permeability coefficient of the material used to simulate the deformation behavior of the aquifuge. Measurements were made of the uniaxial compressive strength and the permeability coefficients of the material. Taking the uniaxial compressive strength and the permeability coefficient as core indices and using the single-factor method, the effects of changing the contents of various groups on the physical and mechanical properties of similar materials of the aquifuge were studied. The proportion schemes are shown in Table 1. In each scheme, water composed one-ninth of the total mass of the fine river sand, bentonite, and gypsum.
Table 1

Proportion scheme design

Scheme

Unchanged content

Altered content

Change ratio

Remarks

1

Ws:Wb+g = 8:1

Wo:Wsum = 1:9

Wb:Wg

1:9, 3:7, 5:5, 7:3, 9:1

Wb is the content of bentonite; Wg is the content of gypsum; Ws is the content of fine river sand; Wb+g is the total mass of bentonite and gypsum; Wo is the content of methyl silicone oil; and Wsum is the total mass of fine river sand, bentonite, and gypsum

2

Wb:Wg = 9:1

Wo:Wsum = 1:9

Ws:Wb+g

6:1, 8:1, 9:1

3

Ws:Wb+g = 8:1

Wb:Wg = 9:1

Wo:Wsum

1:5, 1:7, 1:9

Prior research has shown that scale dependence of strength and stress–strain relationship of rock is ubiquitous, and it is found that large-scale load tests is strongly retarded and negligible for stresses lower than about one-third of the failure load (Pusch and Weston 2012). To eliminate the effect of scale dependence between the material used to simulate the aquifuge and aquifuge at the field scale, and to satisfy the similarity theory, the strength and permeability coefficient of both were tested in the laboratory at the same standard size.

The results of the specimens’ uniaxial compressive strength tests indicated that as the content changed in each group, the strength ranged between 43.56 and 220.41 kPa. As the contents of the bentonite and fine river sand increased, the uniaxial compressive strength of the material used to simulate the aquifuge gradually decreased, the extent to which this material was subjected to brittle fracturing also gradually decreased and the extent to which it underwent plastic deformation under stress gradually increased. As the proportion of the silicone oil increased, the uniaxial compressive strength gradually decreased. The σε curves of the material used to simulate the deformation behavior of the aquifuge in each scheme are shown in Fig. 1.
Fig. 1

The change curves of the material σε with the content changes in each group. a The change curves of the material σε with Wb:Wg. b The change curves of the material σε with Ws:Wb+g. c The change curves of the material σε with Wo:Wsum

The results of the permeability coefficient tests indicated that permeability coefficient of the material used to simulate the aquifuge varied in the range of 4.49–32.41 × 10−7 cm s−1. As the bentonite content increased, the permeability coefficient of the material showed a logarithmic increased, and the R2 value of the fitting function was 0.991. As the content of the fine river sand and the methyl silicone oil increased, the permeability coefficient decreased exponentially, and the R2 value of the fitting function was more than 0.95. The best-fitting curve for the permeability coefficient in each scheme is shown in Fig. 2.
Fig. 2

The curves of specimens’ permeability coefficients in each scheme. a The change curves of the material σε with Wb:Wg. b The change curves of the material σε with Ws:Wb+g. c The change curves of the material σε with Wo:Wsum

According to the above-mentioned general laws of the ratios of the clay aquifuge similar materials and based on the real geological conditions in the first mining area of a coal mine in Yili, Xinjiang Province, a physical model was established for solid–liquid coupling for overburden rock movement in slice mining of a extremely thick coal seam. The model was used to study the clay aquifuge stability in the mining process. The aquifuge of the coal mine in Yili consisted of clay siltstone with uniaxial compressive strength of 12.6 MPa and a permeability coefficient of 9.66 × 10−6 cm s−1. The model’s geometric similarity ratio was 1:150. According to similarity theory, the time similarity ratio was 1:12.25. By comparison, when Wb:Wg = 9:1, Ws:Wb+g = 8:1, Wo:Wsum = 1:9 and Wwater:Wsum = 1:9, the errors of the aquifuge similar material’s uniaxial compressive strength and permeability coefficient and the equivalent value of the aquifuge in the model were all within 10%; thus, the proportions were reasonable. Table 2 presents the error statistics of the performance indices of the similar materials.
Table 2

Error statistics of the performance indices of the similar materials

Core indices

Rock mass

Similar conversion

Similar material

Error (%)

Uniaxial compressive strength (kPa)

1.26 × 104

84.00

79.25

5.65

Permeability coefficient (cm s−1)

9.66 × 10−6

7.89 × 10−7

8.56 × 10−7

8.49

Experimental method

According to the resulting reasonable proportions described above and based on the real geological conditions of the first mining area in the Yili no. 1 coal mine in Xinjiang, a physical model was established for simulating solid–fluid coupling in the overburden rock movement for slice mining of extremely thick coal seams. The model was used to analyze aquifuge deformation, water level changes of the aquifer, and the height of the fracture zone in the overburden. The model also used to study how the deformation behavior of the clay aquifuge stability would vary during the mining of such seams.

Engineering geological conditions

The no. 5# coal seam is the working coal seam in the first mining area of the Yili no. 1 coal mine. The average thickness of the coal seam is 19.2 m. The sequence of sediments that contains this coal seam exhibits low inclination angles. The 4.4-m-thick siltstone that overlies the coal seam is the main aquifuge of the mine. The siltstone is rich in clay minerals and exhibits water swelling and expansion, and the distance of the siltstone from the seam roof is 53.9 m. The potentiometric head of the aquifer in the region is between 5 and 8 m with an average of 6 m. Figure 3 shows the comprehensive stratigraphic column, and Table 3 shows the physical and mechanical properties of the coal seam and surrounding rocks.
Fig. 3

The comprehensive stratigraphic column

Table 3

Physical and mechanical parameters of the coal and rock (Zhang et al. 2017a)

Lithology

Thickness (m)

Density (kg m−3)

Cohesion (MPa)

Friction (°)

Uniaxial compressive strength (MPa)

Remarks

Loose sand layer

6.0

1500

1.1

38.0

5.0

Aquifer

Clay siltstone

4.4

2300

1.6

36.0

12.6

Aquifuge

Fine sandstone

2.3

2100

11.9

22.6

13.0

 

Siltstone

4.0

2300

1.6

36.0

12.6

 

Fine sandstone

2.8

2100

11.9

22.6

13.0

 

Siltstone

1.7

2300

1.6

36.0

12.6

 

No. 3 coal

3.6

1200

1.7

39.0

4.7

 

Mudstone

14.1

2150

1.0

38.0

2.9

 

Fine sandstone

2.4

2150

17.9

18.2

15.0

 

4-1 coal

1.2

1200

1.7

39.0

4.7

 

Sand–mud interbed

5.9

2220

1.0

38.0

11.0

 

4-2 coal

0.6

1200

1.7

39.0

4.7

 

Mudstone

2.1

2220

1.0

38.0

11.0

 

Fine sandstone

13.2

2180

24.7

17.8

14.7

 

No. 5 coal

19.2

1170

1.7

39.0

8.8

Mining coal seam

Establishment of the testing system

The established physical simulation system included: a sealed module to simulate solid–fluid coupling, a system for visual tracking of the development of fractures in the aquifuge, and apparatus to monitor the water level.

An 8-mm-thick transparent organic glass plate and 295-2 silicone grease were used in the sealed module to simulate solid–fluid coupling. The seal thickness was approximately 5 mm. This configuration maintained the plate glass strength while retaining the sealed nature and overall transparency of the model.

LUYOR-6200 fluorescent leak detection agent emits bright yellow–green fluorescence and leaves no pigment residue after liquid egress. The use of this agent can clearly and accurately show whether fractures are open or closed and reduce test errors caused by residual pigment. Therefore, LUYOR-6200 fluorescent leak detection agent was used to simulate the water in the aquifer. In addition, UV light with a wavelength of 365 nm and a UV intensity of 3400 MJ/cm2 was used to enable the visual tracing and discrimination of fracture growth.

The water level was monitored by mounting two water level monitors with a scale in the aquifer level in the middle and on one side of the model. Using this configuration, the water could be simulated, and water level changes could be recorded.

Figure 4 shows the key features of each module in the physical model.
Fig. 4

The sketch map of the physical simulation system

Experimental procedures

The size of the physical model was 1.3 m long, 0.12 m wide and 0.91 m high. When slice mining was implemented, the upper slice was 3 cm thick and the lower slice was 9.8 cm thick; the corresponding total coal thickness was 19.2 m. A 17-cm protective coal pillar remained at both sides of the model. Sequential excavation was conducted 12 times at a distance of 17 cm from the left side of the model, with each excavation extending 8 cm. The corresponding actual working face mining distance was 144 m. In the middle of the aquifuge in the model, a displacement monitoring point was laid every 15 cm at a distance of 20 cm from the left boundary of the model. There was a total of 7 displacement points (namely, the no. 1 to no. 7 monitoring points) used to monitor the aquifuge displacement variations during the excavation process.

Before excavation, the water level monitoring module was used to inject the LUYOR-6200 fluorescent leak detection agent with a ratio of 1/500 into the aquifer position in the model. Because the dosage of fluid additive used was only 12.5 ml and the amount of water needed to simulate the water level of aquifer was 6240 ml, the change in density and viscosity of the fluid in the model was negligible. The height of water injection was 4 cm and was stable before the experiment.

The first excavation was implemented when the water level stabilized. Over 2 h, the aquifuge displacement changes, the water level changes in the aquifer, and the overburden caving characteristics were observed and recorded. Next, the whole model was observed under UV light. By observing the flowing and occurrence of the fluorescent leak detection agent, the opening and closing conditions of the fractures in the aquifuge were characterized.

After observation and recording of the effects of the excavation, the process described above was repeated until all the tests were completed.

Results

Upper slice mining

During the process of upper slice mining, V-type and inversed V-type fractures were observed in the simulated aquifuge. The V-type fractures were distributed predominantly along the top of the aquifuge or at both sides of the subsidence basin, whereas the inverted V-type fractures were distributed predominantly along the bottom of the aquifuge or in the middle of the subsidence basin, as shown in Fig. 5. As the working face advanced, the distribution of mining-induced fractures in the aquifuge also changed. The model indicated that when the working face advanced to 60 m, fractures first appeared and the aquifuge displacement was prominent. At this time, the maximum vertical displacement was − 1.05 m, accounting for 23.9% of the aquifuge thickness. Relative to the displacement amount before the upper slice mining, the maximum horizontal displacement was 0.6 m, accounting for 13.6% of the aquifuge thickness, as shown in Fig. 6a, b.
Fig. 5

The sketch map of the dynamic evolution of the fracture in the aquifuge

Fig. 6

The results of the up slice mining. a The vertical displacement of the aquifuge. b The horizontal displacement of the aquifuge. c The water level changes of the aquifer. d The overburden caving shapes

When the working face advanced to 72 m, the V-type fractures on the left side of the aquifuge continued to develop, and the fracture aperture increased. As the original V-type fractures on the right side of the aquifuge closed, new V-type fractures occurred on the right side of the subsidence basin. After upper slice mining was completed, the V-type fractures on the left side of the aquifuge continued to develop and the fracture aperture continued to increase. The fractures reappeared at the position of the original V-type fracture closure. The V-type fractures on the right side of the subsidence basin moved farther to the right. Figure 5 shows the dynamic evolution process of the mining-induced fractures in the aquifuge. Induced by coal mining, the V-type fractures on the left side of the subsidence basin continued to grow and expand, and the V-type fractures on the right side of the aquifuge moved farther to the right within the subsidence basin. Before the final formation of the subsidence basin, the V-type and inverted V-type fractures underwent a dynamical process of initiation, closure, reappearance, and reclosing. This process is the result of aquifuge movement with the overburden rock and the action of water swelling and expansion of clay minerals.

As the working face advanced, the overburden subsidence basin gradually increased and moved to the right. The water in the aquifer gradually flowed into the subsidence basin. When the water level at both sides of the aquifer gradually decreased, the central water level first decreased and then increased, as shown in Fig. 6c. When the working face advanced to 96 m, the maximum vertical displacement of the aquifuge reached to 2.55 m, accounting for 58.0% of the aquifuge thickness, as shown in Fig. 6a. At this point, the subsidence basin also reached the maximum subsidence. The central water level in the aquifer reached a maximum of 8.1 m, which was 1.35 times greater than the original water level. The water level on the left of the aquifer decreased to 0.75 m, accounting for 12.5% of the original water level, as shown in Fig. 6c.

After the completion of upper slice mining, the height of the fracture zone in overburden was approximately 48 m, approximately 5.9 m from the bottom of the aquifuge. The aquifuge was located in a bending subsidence zone with substantial plastic deformation, as shown in Fig. 6d. The maximum vertical aquifuge displacement was − 2.7 m in the central subsidence basin, accounting for 61.4% of the aquifuge thickness. The horizontal displacement at both sides of the aquifuge was larger than that in the middle. The maximum horizontal displacement was 0.75 m on the left side of the aquifuge, accounting for 17.0% of the aquifuge thickness, as shown in Fig. 6a, b. The maximum water level of the central aquifer was 8.1 m, which was 1.35 times the original water level. The lowest water level on the left side of the aquifer was 0.6 m, which was 10.0% of the original amount, as shown in Fig. 6c.

These upper slice-mining observations indicate that when the vertical displacement of the aquifuge is 23.9% of the aquifuge thickness, that is, \(D_{\text{v}} /T \ge 23.9\%\), and the horizontal displacement is 13.6% of the aquifuge thickness, that is, \(D_{\text{h}} /T \ge 13.6\%\), then V-type and inverted V-type mining-induced fractures first occur in the clay aquifuge. During coal mining, the fractures undergo a dynamical process of initiation and closure and then reappear and close again due to the effects of aquifuge deformation and the action of water swelling and expansion of clay minerals. After mining ceases, if \(D_{\text{v}} /T \le 58.0\%\), \(D_{h} /T \le 1 7. 0 {\text{\% }}\), and the height of the fracture zone in the overburden is below the aquifuge, plastic deformation occurs in the aquifuge. Under these conditions, the generated V-type and inverted V-type fractures do not penetrate the aquifuge; thus, the aquifuge maintains relatively better stability. At this point, the water in the aquifer flows from both sides of the subsidence basin into the central basin without loss or damage; consequently, the water level can recover over time.

Lower slice mining

During the process of lower slice mining, when the working face advanced to 24 m, the V-type fractures on the left side of the aquifuge first developed and connected to the aquifuge. Next, the water in the aquifer flowed into the goaf through the water-transmitting fractures. At this point, the water level dropped sharply, decreasing by nearly 81.5% of the original amount, as shown in Fig. 7c. The maximum vertical displacement was − 3.3 m, accounting for 75.0% of the aquifuge thickness. The maximum horizontal displacement was − 1.05 m, accounting for 23.9% of the aquifuge thickness, as shown in Fig. 7a, b. At this point, water flowed through open fractures in the aquifuge. When the working face advanced to 48 m, the water level reached 0 m, and the aquifuge displacement was maximized when working face advanced to 96 m. The maximum vertical displacement was 15.75 m in the central aquifuge. The maximum horizontal displacement was 1.65 and − 1.35 m on both sides of the aquifuge, and the maximum horizontal displacement on the left side was slightly greater than that on the right side.
Fig. 7

The results of the lower slice mining. a The vertical displacement of the aquifuge. b The horizontal displacement of the aquifuge. c The water level changes of the aquifer. d The overburden caving shapes

After lower slice mining, the height of the fracture zone in overburden was approximately 58.5 m, and the aquifuge was completely in the fractured zone. The V-type fractures in the aquifuge gradually developed and formed fractures allowing flow. These fractures were distributed over both sides of the aquifuge. Next, the water in the aquifer flowed into the goaf through the fractures, and the aquifuge underwent stability failure. Observations of the subsidence basin in the model indicated that the subsidence was asymmetrical in nature. Both the vertical and horizontal displacements on the left side of the overburden were slightly larger than those on the right side, as shown in Fig. 7d.

The results of the lower slice mining show that if \(D_{\text{v}} /T \ge 75.0\%\), \(D_{\text{h}} /T \ge 23.9\%\) and the height of the fracture zone in the overburden is within the aquifuge, then V-type mining-induced fractures in the clay aquifuge develop and connect the aquifuge. Next, the water in the aquifer flows into the goaf through the water-flowing fractures, and aquifuge instability arises. This sequence is the result of both the strong clay aquifuge movement with the overburden as the mining height increases, and the action of water swelling and expansion of clay minerals. The large fractures continue to develop and ultimately connect the aquifuge, thus forming channels of water flow, and the aquifuge becomes unstable.

Discussion and conclusions

In the experiment, fine river sand, bentonite, gypsum, and methyl silicone oil were used as raw materials to simulate the deformation behavior of a clay aquifuge at a coal mine. Mixtures of these materials were made, and measurements of uniaxial compressive strength and the permeability coefficient were used as the core indices to fabricate materials similar to those of clay aquifuge at the mine site. Mixtures of these materials with appropriate physical properties were then used in a coupled solid–fluid physical model to simulate overburden rock movement during coal seam mining. The deformation and failure of the materials and their water-resisting properties were found to be highly consistent with the behavior of overburden rocks under field conditions, thus verifying the rationality and effectiveness of this approach to simulate the behavior of the clay aquifuge. However, the experiment did not consider the similarity between the similar materials and rock mass in terms of microstructure and component contents; thus, further studies should be undertaken to investigate these issues.

Based on the use of scaling ratios to match conditions between the physical conditions in the model and in the first mining area of Yili no. 1 coal mine in Xinjiang Province, a coupled solid–liquid physical simulation model of overburden rock movement was established to simulate the deformation behavior of rocks during the mining of an extremely thick coal seam at the mine site. The model was used to study the mining stability of the clay aquifuge. The general relationships among the aquifuge stability, height of the fracture zone in the overburden, aquifuge displacement, and water level changes were obtained. As an indoor research method, the coupled solid–liquid similar material simulation was demonstrated to be effective in the study of aquifuge stability. The physical simulation system was developed to introduce visual tracing and discrimination capability for improve accuracy when characterizing fracture initiation, development, opening, and closure. However, because UV light can be hazardous, the UV light was used only for observation and records. The use of intermittent UV light may allow certain details in the experiment to be lost; this aspect should be addressed in further investigations.

The results of the simulation indicate that V-type and inverted V-type fractures occur in the clay aquifuge during mining. The fractures first occur when \(D_{\text{v}} /T \ge 23.9\%\) and \(D_{\text{h}} /T \ge 13.6\%\). Before the final formation of the subsidence basin, the mining-induced fractures undergo a dynamic process of initiation, closure, reappearance and even reclosing. This behavior is a direct result of the clay aquifuge movement with the overburden rock and the action of water swelling and expansion of clay minerals.

During the process of slice mining, the analyses of the aquifuge deformation, changes in the water level of the aquifer, and the height of the fracture zone in the overburden indicated that, as the working face advances, the clay aquifuge gradually becomes unstable, and the aquifuge stability exhibits threshold effects. When \(D_{\text{v}} /T \le 58.0\%\), \(D_{\text{h}} /T \le 17.0\%\) and the height of the fracture zone in overburden is below the aquifuge, the mining-induced fractures may close, and aquifuge stability can be maintained. When \(D_{\text{v}} /T \ge 75.0\%\), \(D_{\text{h}} /T \ge 23.9\%\) and the height of the fracture zone in overburden is within the aquifuge, fractures develop and connect within the aquifuge, and aquifuge stability is lost. During upper slice mining of a 4.5-m-thick coal seam, the clay aquifuge does not reach the water-flowing threshold value, thus maintaining stability. During continuous lower slice mining with 14.7 m coal seam thickness, V-type fractures develop within the clay aquifuge and create conduits for groundwater flow, causing the aquifuge to undergo stability failure. Therefore, the threshold effect of clay aquifuge stability has a direct correlation with the mining height.

The use of a physical model is an effective method of studying the mining stability of a clay aquifuge with intuitive and easy-to-interpret and ease of implementation in the laboratory. However, a scale model of an aquifuge analysis may not consider the complete physics of the actual site. In this study, we analyzed the parameters of aquifuge deformation, water level changes in the aquifer, and the height of the fracture zone in the overburden before the clay aquifuge failed. However, the scale, size, and mass of a failing clay aquifuge were ignored. Further analysis in this area might capture the failing characteristics of an aquifuge during mining, such as the development depth and fracture aperture; these characteristics should be further considered in future investigations.

Notes

Acknowledgements

Funding was provided by the Fundamental Research Funds for the Central Universities of China (Grant No. 2017XKQY073).

References

  1. Booth CJ, Curtiss AM, Demaris PJ, Bauer RA (2000) Site-specific variation in the potentiometric response to subsidence above active longwall mining. Environ Eng Geosci 6(4):383–394CrossRefGoogle Scholar
  2. Chen JM, Yang RH (2011) Analysis of mine water inrush accident based on FTA. Proc Environ Sci 11:1550–1554CrossRefGoogle Scholar
  3. Cheng T, Guo BH, Wang L, Yang XY (2017) Numerical simulation on water inrush from floor during mining above confined aquifer in Jiulishan coal mine. Electron J Geotech Eng 22(8):2907–2918Google Scholar
  4. Fan GW, Zhang DS (2015) Mechanisms of aquifer protection in underground coal Mining. Mine Water Environ 34(1):95–104CrossRefGoogle Scholar
  5. Feng MM, Mao XB, Zhu QH (2010) Effect of lithologic association of the water-resisting strata in coal seam floor on water insulating. J Min Saf Eng 27(3):404–409 (in Chinese) Google Scholar
  6. Fu BJ, Tu M (2015) Experimental research on fracture evolution law of water-resisting in coal seam floor during re-mining. J Min Saf Eng 32(1):54–58 (in Chinese) Google Scholar
  7. Goswami S (2013) Susceptible development: impact of coal mining on environment in India. Glob J Hum Soc Sci Geogr Geo-Sci Environ Disaster Manag 13(8):22–33Google Scholar
  8. Hasan MR, Rashid H, Paul AK (2013) Environmental impact of coal mining: a case study on Barapukuria Coal Mining Industry, Dinajpur, Bangladesh. J Environ Sci Nat Resour 6(2):207–212Google Scholar
  9. Huang QX (2014) Research on cracks zone of clay aquiclude in overburden. Appl Mech Mater 548–549(2014):1744–1747CrossRefGoogle Scholar
  10. Huang QX, Zhang WZ (2011) Research on downward crack closing of clay aquiclude in shallow coal seam safety mining. J Coal Sci Eng 17(3):349–354CrossRefGoogle Scholar
  11. Jin ZY, Ma LQ, Wang F (2014) Research progress and prospect of aquifuge in shallow coal seam: a case study in the Shendong coal field of China. Electron J Geotech Eng 19:2809–2819Google Scholar
  12. Katoria D, Sehgal D, Kumar S (2013) Environment impact assessment of coal mining. Int J Environ Eng Manag 4(3):245–250Google Scholar
  13. Li SC, Zhou Y, Li LL, Zhang Q, Song SG, Li JL, Wang K, Wang QH (2012) Development and application of a new similar material for underground engineering fluid–solid coupling model text. Chin J Rock Mech Eng 31(6):1128–1137 (in Chinese) Google Scholar
  14. Li Y, Peng SS, Zhang JW (2015) Impact of longwall mining on groundwater above the longwall panel in shallow coal seams. J Rock Mech Geotech Eng 7(3):298–305CrossRefGoogle Scholar
  15. Pusch R, Weston R (2012) Impact of scale on rock strength. Int Sch Sci Res Innov 6(8):1487–1492Google Scholar
  16. Rashid HO, Hossain MS, Urbi Z, Islam MS (2014) Environmental impact of coal mining: a case study on the Barapukuria coal mining industry, Dinajpur, Bangladesh. MiddleEast J Sci Res 21(1):268–274Google Scholar
  17. Xiao JY, Tong MM, Fan Q, Zhu CJ (2010) Fuzzy neural network model applied in the mine water inrush prediction. Int Conf Image Process Pattern Recognit Ind Eng 7820(1):78203O–78203O–7CrossRefGoogle Scholar
  18. Yao DX, Xu JY, Lu HF (2011) Nonlinear coupling analysis of coal seam floor during mining based on FLAC3D. J Coal Sci Eng (China) 17(1):22–27CrossRefGoogle Scholar
  19. Zhang J, Hou ZJ (2004) Experimental study on simulation materials for solid–liquid coupling. Chin J Rock Mech Eng 23(18):3157–3161 (in Chinese) Google Scholar
  20. Zhang DS, Fan GW, Liu YD, Ma LQ (2010) Field trials of aquifer protection in longwall mining of shallow coal seams in China. Int J Rock Mech Min Sci 47:908–914CrossRefGoogle Scholar
  21. Zhang DS, Fan GW, Ma LQ, Wang XF (2011) Aquifer protection during longwall mining of shallow coal seams: a case study in the Shendong Coalfield of China. Int J Coal Geol 86(2–3):190–196CrossRefGoogle Scholar
  22. Zhang DS, Fan GW, Zhang SZ (2017a) Annual report of the projects 973 of national key basic research program for basic theory of water protection during coal mining in northwest of China. Xuzhou, China (in Chinese)Google Scholar
  23. Zhang J, Ai C, Li YW, Zeng J, Qiu DZ (2017b) Brittleness evaluation index based on energy variation in the whole process of rock failure. Chin J Rock Mech Eng 36(6):1326–1340 (in Chinese) Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Shizhong Zhang
    • 1
  • Gangwei Fan
    • 1
  • Dongsheng Zhang
    • 1
  • Qizhen Li
    • 1
  1. 1.China University of Mining and TechnologyXuzhouChina

Personalised recommendations