1 Introduction

Since the 1850s, ground subsidence induced by mining activities have become more and more serious (Feng et al. 2018; Shen et al. 2018), which is considered responsible for the perturbation of the natural ground surface, damage to public facilities, farmland, and aquifers (Can et al. 2012; Li et al. 2016; Newman et al. 2017; Tzampoglou and Loupasakis 2018). The profile, depth and influence scope of subsidence area mainly depend on the mining methods used, coal types, rock type, geological settings, geo-mechanical behaviors, and so on (Holla and Buizen 1991; Falcón et al. 1996; Zhang and Nemcik 2013; Sasaoka et al. 2015; Xia et al. 2016; Do et al. 2017).

In order to investigate the law of ground subsidence induced by underground mining, foreign and domestic scholars and engineers have conducted a great number of research work (Can et al. 2012; Park et al. 2012; Bian et al. 2014). Meanwhile, for developing subsidence theory and improving the prediction system, in recent years, a number of advanced research methods, such as the similarity simulation (Ghabraie et al. 2015; Xuan et al. 2015), numerical simulation (Zuo et al. 2009; Deck and Anirudh 2010; Liu et al. 2011; Shabanimashcool and Li 2012; Salmi et al. 2017), and artificial intelligence (Li et al. 2007; Hyun and Saro 2011; Shahab 2011; Torres and Rodriguez 2017), have been widely employed to analyze the mechanism of ground deformation. Besides, more sophisticated mechanism models and their corresponding calculating formulations have been set up using aforementioned methods and a body of knowledge about rock mechanics, fracture mechanics, solid mechanics, etc. (Suchowerska et al. 2016; Sepehri et al. 2017; Hamdi et al. 2018). These mechanical models and research methods are well validated in specific conditions and used to guide fieldwork.

However, the ground deformation problems caused by multi-seam mining might be more complicated, since the mechanical properties of rock strata overlying the lower coal seam would be seriously changed because of previous mining (Ding et al. 2017). Thus, the above mechanisms are not entirely applicable for ground deformation when the lower seam mining (e.g., conducting mining operations under the previous mined area). The characteristics of subsidence from sequential extraction of multi-overlapping coal seams has been studied using physical and numerical models, and the results show that the substrata movement profile, after lower seam extraction, can be separated into three different zones (i.e., virgin ground, disturbed zone, and caved zone) (Ren et al. 2014; Ghabraie et al. 2015; Suchowerska et al. 2016). Subsidence after mining from under a previously mined area is enhanced, which results from the consolidation of the ground, closure of existing cracks, and the stress release in the previous mined area (Suchowerska et al. 2013; Zhao et al. 2014; Guo et al. 2017a, b). During the process of multi-seam mining, the upper seam mining softens the overlying substrata, which makes the whole overburden rock more vulnerable to break (Ghabraie et al. 2017; Wang and Zhang 2018). Furthermore, the activation of previous gob formed by the upper seam mining may intensify the subsidence to some extent. Since the rock structure in the previous gob will become loose and have a small load-bearing capacity during lower seam mining, though it has been compressed over a long period of time after upper seam mining (Xie et al. 2014; Li et al. 2018; Zhu and Tu 2017). Consequently, the mechanical properties of rock strata play an important role for the deformation of the substrata and ground surface (Zhu 2010; Can et al. 2013; Sasaoka et al. 2015; Singh 2015; Manekar et al. 2017; Mckee et al. 2018; Hu et al. 2018; Zhang et al. 2018; Zou and Lin 2017).

Ground cracks as the other product of ground deformation not only have a great impact on the stability of surface structures, but also threaten the productivity of land and even seriously affect the safety of residents in a mining area (Li et al. 2010). Therefore, it is necessary to explore the mechanism of crack formation and development. Ground cracks caused by mining fundamentally result from that stress transmitted through rock strata from the bottom to the surface and are also influenced by the rock properties and other characteristics of substrata. For instance, a thick layer of hard rock in the overburden could greatly reduce the mining-induced stress and slow down the dramatic surface cracks’ occurrence (Wu et al. 2009; Ghosh and Sivakumar 2018). Mining methods and mining direction (including downward mining and upward mining) also affect the characteristics (such as width, depth, and shape) and the distribution of ground cracks induced by mining activities (Wu et al. 2009). The development period of ground cracks is consistent with the breaking cycle of the immediate roof. In addition, mining under a previous mined area may induce the redevelopment of the previous cracks caused by previous mining in the fracture zone. In order to clarify the formation mechanism of ground cracks, some field work has been conducted (Zhu and Lian 2012; Wang et al. 2017; Zou and Lin 2018; Wu et al. 2018). Although numerous research work related to ground subsidence and surface cracks induced by mining has been done, the research focusing on the mechanisms of ground subsidence and formation of ground cracks induced by shallow-buried close-distance multi-seam mining has been rarely reported. Thus, corresponding field work and theoretical analysis should be done to clarify the evolution law of ground subsidence and the formation mechanism of ground cracks in under such geological conditions.

2 Research Objectives

In terms of ground subsidence and cracks caused by underground mining, many experts have conducted a number of field measurements (Hu et al. 2014; Guo et al. 2017a, b; Qian and Fang 2018) and investigated the related law and mechanisms (Ju and Xu 2015; Lu et al. 2017; Zhou et al. 2017). Considering current theoretical defects and technique requirements from field operations, the principal objectives of this research are as follows:

  1. 1.

    Obtaining the basic features of ground subsidence induced by shallow-buried close-distance multi-seam mining;

  2. 2.

    Analyzing the dynamic process of ground subsidence and calculating the related parameters;

  3. 3.

    Exploring the evolution law of ground subsidence by comparing the features of ground subsidence because of mining under the overlying goaf zone and overlying pillars zone based on the field data;

  4. 4.

    Acquiring the basic shape features, distribution characteristics and evolution law of ground cracks;

  5. 5.

    Investigating the formation mechanics and making reasonable prediction for the development of ground cracks caused by multi-seam mining.

3 Research Area and Field Monitoring Methods

3.1 Basic Information of Bulianta Coal Mine

Bulianta Colliery is located in the southeast of the Inner Mongolia Autonomous Region, belongs to the Bulianta Township (Fig. 1). It is the biggest modern underground coal mine in the world (Fan and Zhou 2010), which was constructed by the Shendong Coal Corporation. The main geological characteristics of the mining area include thick coal seams, shallow buried depth and thick loose soil layers. Currently, the mine is extracting 2-2 coal seam as the extraction of 1-2 coal seam at the upper panel has been finished by room and pillar mining method 3 years ago. However, because of geological conditions and mining methods, large numbers of coal pillars were left during the 1-2 coal seam mining period. During the observation period, many special phenomena of ground deformation occurred with the extraction of 2-2 coal seam, which is obviously different from the extraction of 1-2 coal seam. Some infrastructure was seriously damaged (Fig. 2), which directly influence the life of local residents.

Fig. 1
figure 1

Study site. a Location of Bulianta coal mine in China; b roadway layout of 22307 working face and positional relationship between 1# and 2# coal seams

Full size image
Fig. 2
figure 2

Damage to buildings and roads in mining areas

Full size image

The working coal seam at panel 22307 (Fig. 1b) is 2-2 coal seam, which is typically shallow with average depth of 100 m. It has an average thickness of 7.25 m ranging from 6.8 to 7.7 m, an average dip angle of 3° ranging from 1° to 3°. The tendency retreat longwall mining method was employed for a mining height of 6.8 m. The width (toward direction) and length (e.g., advancing direction) of working face are 300 m, 4954 m, respectively. Besides, thick soil layers with thickness of 8–23 m mostly cover the whole panel. Table 1 lists the simplified geological settings of the working panel.

Table 1 Simplified geological settings in panel 22307
Full size table

3.2 Field Monitoring Methods

3.2.1 Ground Subsidence Monitoring Method

The observation station was set on the ground surface covering the working face 22307. It included three observation lines, e.g., a tendency observation line (named line Q) along the working face direction, two toward observation lines (named line Z and line C) perpendicular to the working face direction. The tendency line was set on the surface corresponding to the central line of the working face, partially located above the previous gob and partially located above the left pillar with a critical point of Q12. Besides, it contained 27 observation points named Q1–Q27, with a fixed interval of 15 m and a total length of 395 m. Moreover, Q1 was set as the control point of line Q. Line Z, one of the toward observation lines, above the previous gob, contained 31 observation points named Z1–Z31, with a fixed interval of 30 m and a total length of 900 m. Meanwhile, Line C, above the left pillar zone, had the same arrangement as Line C. In addition, Q1, Z1, and C1 were separately set as fixed points (e.g., control points) in their lines (Fig. 3).

Fig. 3
figure 3

22307 working arrangement of measuring points

Full size image

Generally, the observation points should be set up on the ground surface within the influence scope caused by mining. Meanwhile, the following rules should be obeyed:

  1. 1.

    Observation points should be located on the main section of the ground subsidence basin if possible.

  2. 2.

    The length of the observation lines should sufficiently cover the whole influence scope of the ground subsidence basin.

  3. 3.

    Observation points should be set with proper distances from each other and not be affected by the adjacent working face according to the field conditions and the research purpose.

  4. 4.

    Control points should be set at the both ends of the observation lines. Generally, they are set outside the influence scope of the ground subsidence basin. If there is a frozen soil area, they should be laid below the frozen soil layer 0.5 m to guarantee their stability.

3.2.2 Ground Cracks Monitoring Method

Underground activities will not only cause surface subsidence, but also ground cracks, which directly or indirectly influence the human activities and even damage the ground structures (Zhou et al. 2017). Thus, it is significant to investigate the formation mechanics and the development law of mining-induced ground cracks and further put forward corresponding protective measures. In this paper, in order to obtain the distribution situation of surface cracks along the working face, related observations were conducted for several times, including measuring and recording the position, width and length of the cracks. Meanwhile, the temporal and spatial variation characteristics of typical cracks should be recorded to explore the relationship of cracks development versus the mining activities. In addition, considering that the cracks development may be closely related to the movement of overlying strata, related data which can reflect the overlying strata movement should be recorded.

3.3 Measurement Work

After establishing the observation stations, the measurement work of ground subsidence and cracks should be periodically and simultaneously conducted during the recovery of working face 22307. It mainly consisted of measuring and recording the height variation of observation points, the width variation and positions of ground cracks, the profile and the distribution of ground cracks. During the observation period, a Topcon DL-501 electronic level and an M-500 GNSS receiver were employed to measure the ground deformation. In addition, the measurement work continued until that the ground deformation was stable. Figure 4 shows the daily measurement work in the stopping duration.

Fig. 4
figure 4

Daily measurement work. a Height measurement for observation points; b width measurements for ground fissures

Full size image

4 Monitoring Results and Analysis

4.1 Characteristics of Ground Subsidence Basin After Multi-seam Mining

4.1.1 Subsidence Characteristics of Tendency Measurement Line

The final subsidence of line Q (see Fig. 3) is plotted in Fig. 5, which reflects the characteristics of tendency section of the ground subsidence basin. From Fig. 5, we can see that the maximum value of ground subsidence above the previous gob was 5847 mm and occurred at the point Q8 with a distance of 105 m from the fixed point (Q1). While the maximum value of ground subsidence above the pillar zone, occurred at the Q12 with a distance of 165 m from the fixed point (Q1) was 3638 mm. Overall, the ground subsidence above the previous gob was deeper and steeper than that above the pillar zone. This phenomenon may result from the significant reduction of bearing capacity of the overlying strata in the previous gob induced by lower seam mining together with the effect of its own weight.

Fig. 5
figure 5

Subsidence curve of line Q

Full size image

Figure 6a shows the tilt value of the observation points in line Q, which describes the features and severity of ground subsidence. The maximum tilt value of the ground above the previous gob was 53.6 mm/m at Q5, while the maximum tilt value of the ground above the pillar zone was 46.1 mm/m (absolute value) at Q11. Besides, we can see that tilt value of the ground above the previous gob was at a higher level. Meanwhile, the tilt value was the larger at the center of the tilt curve and decreased from the center to the brim. This reflects that the ground subsidence was the result of transmission of strata deformation induced by mining activities from bottom to the top.

Fig. 6
figure 6

a Tilt curve of line Q; b curvature of line Q

Full size image

The curvature of the observation points in line Q is presented in Fig. 6b, which reveals the unevenness of the ground surface. The maximum curvature value of the ground above the previous gob was 3.2 mm/m2 at Q2, and then there was a dramatic decrease at Q3. While, the maximum curvature value of the ground above the pillar zone was 2.2 mm/m2 at Q13. In addition, from Q9 to Q23, the curvature variation was a severe oscillations type, which means that the ground surface deformation in this region was more complicated.

4.1.2 Subsidence Characteristics of Toward Measurement Line

Figure 7 presents the final profiles of line Z and line C (see Fig. 3). The maximum subsidence values of lines Z and C are 4381 mm, 2688 mm, respectively, and they were all located in the middle of the working face. Besides, the two curves were approximately symmetric about the center of the working face. However, the average subsidence of line Z was much bigger than line C. This phenomenon may have been caused by the activation of the previous gob, e.g., the overlying strata in the previous gob broken again because of lower seam mining, which resulted in the intensification of ground subsidence. Moreover, the affected scope of ground subsidence was bigger than the width of the working face (between the haul road and the ventilating road), which indicates the mining activities not only caused the vertical subsidence, but also induced the horizontal deformation of the rock strata and ground surface.

Fig. 7
figure 7

Toward sinking curves of lines Z and C

Full size image

Figure 8a demonstrates the tilt variation of lines Z and C, which reflects the change of subsidence value in unit distance. The maximum tilt values of lines Z and C were 21.6 mm/m and 16 mm/m, respectively, at a distance of 60 m from the haul road. Furthermore, the curves tilt value in the scope between the haul road and ventilating road was higher than outside in term of absolute value, which reveals that the ground in the affected scope of mining activities subsided to a greater extent. In addition, the tilt value of line Z was obviously higher than line C in terms of absolute value, which reflects that the profile of line Z is steeper than line C.

Fig. 8
figure 8

Toward a tilt curves; b curvatures of lines Z and C

Full size image

The curvature variation of lines Z and C is illustrated in Fig. 8b. The maximum curvature values of lines Z and C were 0.7 mm/m2 (absolute value) and 0.6 mm/m2, respectively, which separately corresponded to the points of maximum subsidence of the ground surface. The curvature values in the middle part of the working face (between the haul road and the ventilating road) were bigger.

4.2 Dynamic Variation of Ground subsidence Induced by Multi-seam Mining

Figure 9 demonstrates the process of ground subsidence with the working face proceeding. The subsidence gradually increased with the advancement of the working face, together with the enlargement of the deformation scope of ground. At the same time, the location of maximum subsidence of ground moved forward with the advancement of the working face (highlighted with broken black lines). When the working face advanced 220 m from the fixed point, the maximum subsidence reached 3948 at point Q8. Thereafter, the location of maximum subsidence of ground never changed with the advancement of the working face; however, the subsidence still continually increased until it reached 5847 mm with the working face approaching 450 m from the fixed point. Apparently, it occurred at the ground above the previous gob. For the subsidence above the left pillar, the maximum value was 3638 mm at point Q13. The subsidence velocity of the ground above previous gob was bigger than the part above left pillar.

Fig. 9
figure 9

Tendency sinking curves with advancement of the working face

Full size image

Figure 10a reflects the tilt variation of the ground surface with the advancement of the working face. The distribution of the tilt value along the observation line hardly changed with the advancement of the working face, but the tilt value of each point increased at various degrees with the advancing of the working face.

Fig. 10
figure 10

a Tilt curves of tendency line with the advancement of the working face; b curvature curves of tendency line with advancement of the working face

Full size image

As can be seen from Fig. 10b, in the early stage of mining, the curvature of the surface was relatively small. However, with the advancement of the working face, the curvature of each point gradually increased. When the working face approached 320 m from the fixed point, it never changed.

4.3 Comparison of Related Parameters About Subsidence

4.3.1 Advance Angle of Subsidence Influence

Generally, ground subsidence occurs beforehand with the advancement of the working face because of mining activities. In the condition of full mining or nearly full mining, the advance angle of subsidence influence is the angle between the line that connects the point of 10 mm subsidence on the main section of the ground subsidence basin and the position of the working face, and the horizontal line on the side of the pillar. As can be seen from Fig. 11, ω1 is the ground subsidence curve corresponding to the working face reaching A, on which the subsidence value of point 1 is 10 mm, and ϕ is the advance angle of influence when the working face reaches A. The advance angle of influence is an important parameter, which is mainly used to judge the scope affected by mining activities. It can be calculated as follows:

$$\varphi =\arctan \frac{{{H_0}}}{L},$$
(1)

where H0 is the average mining depth; L is the advance influence distance (see Fig. 11).

Fig. 11
figure 11

Schematic diagram of advance angle of influence

Full size image

Based on the recorded advance influence distances observed in different periods with the advancement of the working face, the advance angle of influence can be obtained by combining with Eq. 1. Thus, through calculation, the advance angle of influence above the previous gob is 44.52°, together with the advance influence distance of 101.70 m; and the advance angle of influence above the pillar zone is 56.84°, together with the advance influence distance of 65.33 m.

4.3.2 Lagging Angle of Maximum Subsidence Velocity

The lagging angle of maximum subsidence velocity is the angle between the line, connecting the point with maximum subsidence velocity on the main section of the ground subsidence basin and the position of working face, and the horizontal line on the side of the gob. For instance, VAB is the subsidence velocity curve corresponding to the working face reaching B (see Fig. 12), on which the subsidence velocity of point a is maximum, and then φ1 is the lagging angle of maximum subsidence velocity when the working face reaches a. Therefore, it can be defined as follows:

Fig. 12
figure 12

Angle of maximum subsidence velocity schematic diagram. a Temporary cracks, b permanent cracks

Full size image
$$\theta =\arctan \frac{{{H_0}}}{l},$$
(2)

where H0 is average mining depth, m; l is lagging distance of maximum subsidence velocity (Fig. 12), m. Through calculation, the lagging angle of maximum subsidence velocity above the previous gob is 54.46°, while that above the pillar zone is 41.99°.

4.3.3 Subsidence Coefficient

The subsidence coefficient reflects the relationship between the maximum of ground subsidence and mining height under the circumstance of full mining. Generally, it is stable in the same mining condition (full mining or non-full mining). The subsidence coefficient is calculated using Eq. 3 in the condition of full mining or super-full mining (Xu 2015).

$${W_{\text{cm}}}=qM\cos \alpha ,$$
(3)

where Wcm stands for the maximum of ground subsidence; M stands for the mining height; α stands for the dip angle of coal seam; and q stands for the subsidence coefficient.

Generally, the subsidence coefficient, q, is closely related to the property of immediately overlying strata, mining methods, and methods of controlling the roof. For instance, if the overlying strata are hard, the surface subsidence coefficient is small and vice versa. Based on the field data obtained, the working face 22307 was fully exploited, and the maximum ground subsidence above previous gob and that above left pillar are 5847 mm and 3638 mm, respectively; mining height is 6.8 m; tendency of coal seam is 3°. Therefore, the subsidence coefficient of ground above previous gob is 0.85, while that above the pillar zone is 0.53.

Parameters related to surface deformation caused by multi-mining under different strata are listed in Table 2. Ground subsidence above the previous gob is much greater than that above the left pillar. Furthermore, the advance angle of influence and the lagging angle of maximum subsidence velocity, caused by ground subsidence, above the previous gob are also bigger than that above the pillar zone, which indicates that mining under the conditions of the previous gob induces a wider range of ground deformation. The subsidence coefficient of the ground above the previous gob was greater. This may be due to that the immediately overlying strata above the previous gob have undergone the process of caving–fracturing–separation–deformation from upper seam mining, which destroyed the original state of the overlying strata and decreased its strength.

Table 2 Comparison of relative parameters of surface deformation
Full size table

4.4 Classification of Mining Ground Cracks

4.4.1 Temporary and Permanent Cracks

Various ground cracks were formed during lower seam mining. According to the development period, they can be defined as temporary cracks or permanent cracks.

Based on daily observation, temporary cracks were usually produced during the process of the advancement of the working face, as shown in Fig. 13a. Such cracks were mainly formed in the middle part of the working face and caused by the uncoordinated deformation of the ground surface resulting from the breakage of overlying rock strata with the advancement of the working face. However, most of them were closed because of the compression deformation of the ground surface after advancement of the working face. Therefore, the development of such cracks was synchronized with the advancement of the working face. Generally, it is difficult to recognize them after the advancement of the working face because of the rapid change.

Fig. 13
figure 13

Mining ground cracks classified by developing period. a Tensile cracks, b compressive cracks

Full size image

Permanent cracks usually occurred at the boundary of the working face, that is, the area where the horizontal deformation is the largest after ground deformation stabilized, as shown in Fig. 13b. From formation to stabilization, which is a slow process, the width and depth gradually increased and became permanent. If a large number of permanent cracks developed, they may well cause soil erosion, degradation of the surface vegetation, and destruction of the ecological environment of the western mining area. In an effort to protect the ecological environment in mining area, such cracks were usually backfilled after recovery.

4.4.2 Tensile and Compressive Cracks

Tensile cracks (Fig. 14a) were caused by the horizontal tensile deformation of the ground surface exceeding the limited tensile strain of the ground surface layer during advancement of the working face. They were mainly distributed in the tensile deformation zone of the ground surface, and generally before or after the working face. Besides, the depth and width of tensile cracks were usually small, and the development process was easily observed despite its rapid variation.

Fig. 14
figure 14

Mining ground fissures classified by developing period. a Collapsed cracks, b slipped cracks

Full size image

Compressive cracks (Fig. 14b) mainly resulted from the compression deformation exceeding the limited compression strain of the ground surface. They were mainly distributed in the compression deformation zone of the ground surface, and their occurrence usually lagged behind the working face.

4.4.3 Collapsed and Slipped Cracks

The collapsed cracks were caused by the collapse of the ground surface resulting from the overlying strata fracturing due to mining activities. They were mainly located on the ground surface above the working face and developed with the advancement of the working face. Generally, the width and depth were large, as can be seen in Fig. 15a.

Fig. 15
figure 15

Mining ground cracks classified by developing period

Full size image

The slope slippage and the local breakage of the ground surface because of mining activities easily triggered the slipped cracks, as shown in Fig. 15b. In terms of formation mechanics, they were induced by stretching of the surface due to mining activities combined with the slope slippage, which is different from other types of ground cracks. Cracks were also greatly influenced by the geological and geomorphic surroundings. For instance, the topsoil layer was looser when the surface slope was greater, which resulted in cracks readily forming.

4.5 Distribution Characteristics of Ground Cracks

During the observation period, the position, width and depth of ground cracks and their variation with the advancement of the working face 22307 were recorded. Based on the data, combining with the exploitation ichnography, the spatial distribution of ground cracks along the working surface can be obtained, as shown in Fig. 16.

Fig. 16
figure 16

Ground cracks: partial distribution of 22307 working face

Full size image

As seen from Fig. 16, the ground cracks were almost located in the middle part of the working face and nearly perpendicular to the direction of the working face. Obviously, the ground cracks above a previous gob were denser and longer than that above the left pillar. The average interval of the ground cracks above the previous gob was 14.75 m, while that above the pillar zone was 27.8 m. This phenomenon indicates that mining activities caused greater deformation of the ground surface above the previous gob than of the ground surface above the pillar zone. Furthermore, the ground cracks above the previous gob still existed and hardly changed after advancement of the working face; most of the surface cracks above the pillar zone were closed after advancement of the working face, and only a few were observed. This may be related to the mechanical properties of the overlying strata.

4.6 Development Law of Ground Cracks

Generally, with the influence of the movement of overlying strata, the ground cracks induced by mining activities would experience formation, expansion, decrescence, and closure. Figure 17 shows the dynamic variation process of the development of typical ground crack CLF1 with the advancement of the working face. At the beginning of the advancement of the working face, the ground crack formed and its width was very small, which resulted in a slight deformation of the road (Fig. 17a). Then the width of the ground crack gradually increased with the advancement of the working face. Ultimately, the crack width and deformation reached the maximum when the working face exceeded the crack position 6 m (Fig. 17d). Since the expansion of the crack has seriously affected road transportation, workers made a timely backfilling, which prevented continued data collection from the crack. The ground crack variation was strongly associated with the advance rate of the working face. Table 3 lists the variation of relative parameters of the crack CLF1 with the advancement of the working face. A 50-mm-wide ground crack formed 35 m ahead of the working face. With the advancement of the working face, the crack width increased from 50 mm to 560 mm; the length of the crack increased from 10 mm to 55 mm; and the step height increased from 0 mm to 420 mm.

Fig. 17
figure 17

Crack CLF1 variation process. a August 10th, b August 11th, c August 12th, d August 13th

Full size image
Table 3 Crack CLF1 change process table
Full size table

CISMP software (Peng 2009) was employed to predict the horizontal deformation of CLF1 with the advancement of the working face. As can be seen from Fig. 18, at the beginning, the horizontal deformation increased with the increase of crack width. When the working face exceeded the crack position of 6 m, both of them reached the maximum with the width of 560 mm and the horizontal deformation of 4.88 mm/m. However, the horizontal deformation gradually decreased with the advancement of the working face because of surface compression associated with mining activities. When the working face exceeded the crack position of 52 m, the horizontal deformation reached the minimum of 0.07 mm/m. While the specific value of the crack width was unobtainable because of backfilling, based on the later observation data, the crack width did decrease after it reached the maximum.

Fig. 18
figure 18

The relationship between width and horizontal deformation of Fissure CLF1

Full size image

5 Discussion

5.1 The Evolution Law of Ground Subsidence Induced by Shallow-Buried Close-Distance Multi-seam Mining

The gob is formed after the upper seam extraction, and the initial stress of overlying strata is disturbed and redistributed, thus movement, deformation and destruction of the overlying strata would occur. With the expansion of the gob, there was an expansion of the range of rock strata movement. When the gob expanded, the movement of the rock strata spread to the surface, and caused ground surface movement and deformation. Therefore, ground subsidence is the result of the gradual development of overlying strata movement from the bottom to the ground surface after the mining activities.

According to the field measurement in Bulianta Colliery, the ground subsidence (5847 mm) above the previous gob was greater than that above the left pillar (3638 mm). Moreover, the final profile of the ground subsidence basin above the previous gob was steeper than that above the left pillar. The scope of ground surface deformation caused by mining activities above the previous gob is greater than that above the left pillar.

Based on the mining conditions in Bulianta Colliery, the above phenomena can be explained through the following mechanisms:

  1. 1.

    Lower seam mining activated the previous gob. The rock strata immediately overlying on the 2-1 coal seam was broken because of the previous mining activities, then, after a period, the broken rock strata would fall, sink, and accumulate in a loose state in the gob. Although the rock mass has been compacted naturally for a long time, the underground cavity, separation, fracture and pore would still exist for a long time. During repeated mining, the stress equilibrium would be broken again, which again resulted in instability and compaction of the rock mass in the previous gob. Thus, it may induce ground subsidence above the previous gob.

  2. 2.

    As we know, the lithology of overlying rock has significant influence on the dynamic process of ground subsidence and the characteristics of the subsidence basin. The practice shows that the key strata (Pu and Zhang 2011; Coggan et al. 2012) play a major role in controlling movement of overlying strata on the stope. According to the geological data, there were only about 30 m of medium-hard rock strata (Table 2), between the 2-1 coal seam and the 2-2 coal seam, controlling the overlying strata movement; therefore, the bearing capacity of key strata was low. Meanwhile, the previous gob would be activated because of the 2-2 coal seam mining operations, which made the compacted rock loose in the previous gob. Together with the effect of the deadweight of the loose rock, the ground subsidence above the previous gob would be intensified. Compared with the rock in the overlying gob zone, the overlying strata in the left pillar zone were relatively complete and had a large bearing capacity. Therefore, the ground subsidence on the left pillar was smaller than that above the previous gob.

  3. 3.

    In addition, the working panel was covered with 8–23 m of Aeolian sand. Compared with the bedrock, the Aeolian sand had low density, low strength, and no tensile strength, which made it easily deformed when subjected to an external force. Therefore, it did not have the bearing capacity similar to the “beam” structure. Under its own heavy weight, it would produce greater pressure on the lower bedrock. Thus, it sank along with the subsidence of bedrock, but also deformed under the additional stress from mining activities.

5.2 The Formation Mechanism of Ground Cracks Induced by Multi-seam Mining

Based on elastic finite element method analysis, the ground stress distribution in the overburden layer after the extraction is illustrated in Fig. 19. In the figure, (1) is the compressive stress area; (2a) is the tensile stress area; (2b) is the tensile stress area of positive curvature; (3a) is the compressive stress area of negative curvature; and (3b) is the supporting pressure area. The ground surface above the gob (e.g., the underground stope) was located in the horizontally compressive stress scope. Therefore, the compressive cracks would be mostly formed in this area because of the effect of compressive stress. However, previous tensile cracks would be closed because of the effect of compressive stress. Meanwhile, the ground surface above the both sides of the gob was located in the tensile stress area, which resulted in the formation of tensile cracks.

Fig. 19
figure 19

Stress distribution in the overburden layer after the extraction (Sui 1999)

Full size image

As can be seen from Fig. 19, within the scope of the influence of the mining subsidence, there are both compressive deformation and tensile deformation zone simultaneously on the ground surface. Generally, during the underground mining process, tensile cracks usually occur ahead of the working face, develop in the tensile deformation zone; while compressive cracks occur behind the working face, and develop in the compressive deformation zone on the ground surface.

During the observation period, when the 22307 working face reached 37.25 m from the left side of the border (between overlying goaf zone and overlying pillars zone), a tensile crack C1 was formed at the border; while the working surface advanced to 25 m from the left side of the border, a compressive crack C2 was formed lagging behind 25 m from the crack C1 (Fig. 20). With the continuous advancement of the working face, the surface near the cracks gradually subsided and evolved to a graben. When the working face advanced to 15.65 m from the left side of the border, the subsidence height of the graben was about 100 mm. However, when the working face advanced from 8.4 to 3.2 m from the left side of the border, a strong roof weighting phenomenon occurred and lasted 24 h, which resulted in that the subsidence height of the graben increased from 100 to 900 mm. During this period, the support resistance was in a high level about 18 000 N, which may result from the breakage of the immediate roof.

Fig. 20
figure 20

(modified after Du et al. 2017)

Development law of surface cracks

Full size image

Therefore, it can be concluded that the ground cracks were formed ahead of the working face and developed rapidly during the period of the breakage of the immediate roof. Generally, the ground cracks above the overlying goaf zone were formed ahead 16.78 m of the working face, while that above the overlying pillars zone were formed ahead 33.5 m of the working face.

5.3 The Effect of Advance Rate of the Longwall Face on the Formation of Ground Cracks

Related studies have shown that advance rate of the longwall face is closely related to the periodic roof weighting and the breaking law of overlying strata, which may directly or indirectly influence the ground deformation and the formation and development of ground cracks (Krzysztof et al. 2018; Wang et al. 2018). Generally, with the increase of the advance rate of the working face, the peak distance and the influence range of the abutment pressure show a decreasing trend, as well as the concentration coefficient of in situ stress (Liu 2010), which is conducive to the stability of the overlying strata. However, the loading rate of mining pressure on overlying strata is also affected by the advance rate of the working face, which may intensify the damage of the overlying strata (Yang 2015). Thus, it is necessary to explore the relationship between the formation of ground cracks and the advance rate of the longwall face to guide the field work.

The ground cracks were counted during the advancement of the working face. The advance distance of the tensile cracks was closely related to the advance rate of the working face. The relationship between the advance distance of ground cracks and the advance rate of the working face was fitted in Fig. 21, using a quadratic function. The advance distance of the mining surface crack reached the minimum 13.6 m, when the advance rate of the working face was 12.8 m/day, which indicates that the ground surface was minimally damaged by mining activities.

Fig. 21
figure 21

Advance distance of ground cracks and advance rate simulation relation

Full size image

6 Conclusions

Ground crack and subsidence measurements were conducted after establishing an observation station in the Bulianta Colliery, which has a shallow-buried coal seam and thick, loose layer covering the ground surface. Measurements included the height variation of the observation points, the width variation and positions of ground cracks, the profile and distribution of ground cracks, and the development of ground cracks with the advancement of the working face.

Through analyzing the data observed from the field, the ground subsidence above the previous (5847 mm) was greater than above the left pillar (3638 mm). The profile of the ground subsidence basin above the previous gob was steeper than that above the left pillar. Mining activities under the previous gob induced more intense deformation and ground subsidence than under the pillar. The advance angle of influence and the lagging angle of maximum subsidence velocity of the ground surface above the previous gob were greater than that above the left pillar, which indicates that the influence scope was greater above the previous gob. Thus, the mechanism of ground subsidence induced by multi-mining has been discussed, mainly including the activation of a previous gob caused by mining activities, the low-bearing capacity of key strata, and the strong self-weight of thick loose layer covering the ground surface.

The ground cracks induced by mining activities were classified and summarized depending on their features as temporary cracks or permanent cracks, tensile cracks or compressive cracks, and collapse cracks or slipped cracks. Furthermore, the distribution of ground cracks above the previous gob varied from that above the left pillar. The average distance of the ground cracks above the previous gob was 14.75 m, while that above the left pillar was 27.8 m. Furthermore, the formation mechanisms of tensile cracks and compressive cracks were determined by analyzing the ground stress distribution in the overburden layer after the extraction. The relationship between the advance distance of ground cracks and advance rate of the working face was fitted with a quadratic function. Through calculation, when the advance rate of the working face was 12.8 m/day, the advance distance of the mining surface crack reached the minimum of 13.6 m. This finding is helpful for protection of the surficial environment during and after mining operations.

In addition to the results presented in this paper, there are a few issues that should be clarified in future work. First, the movement of overlying rock strata directly relates to the deformation of the ground surface. Thus, the different movement mechanisms of overlying strata should be explored and combined with the deformation of the ground surface. Second, although the mechanisms for ground subsidence and crack formation were identified for this site, they may not be applicable for other mining conditions. Therefore, different mining conditions should be considered, such as properties of the rock strata, mining depth, and mining methods, to improve and perfect the above mechanisms. Third, corresponding protection operations for the ground surface should be developed for the mechanisms, so that they are systematic and standardized for different mining conditions.