A Study on the Movement and Deformation Law of Overlying Strata and the Self-Healing Characteristics of Ground Fissures in Non-Pillar Mining in the Aeolian Sand Area

Abstract

The mining area in western China is ecologically sensitive. Coal mining can cause the formation of ground fissures, leading to geological disasters and further accelerating the process of land desertification. In this study, the working face of non-coal-pillar mining in the aeolian sand area was considered as the research object. The movement and deformation law of overlying strata were investigated through field measurements, theoretical analysis, and numerical simulation, and the mechanism governing the self-healing characteristics of ground fissures was revealed. The results demonstrated that the surface angular parameters were lower. This implies that the surface movement and the degree of deformation in non-coal-pillar mining in the aeolian sand area are significant, with a large mining influence range and rapid surface subsidence speed. After the mining of the working face, the resulting failure form of the overlying rock was asymmetric. Boundary ground fissures are typically located within the boundary of the working face, and no outward expansion is primarily observed. Dynamic ground fissures have “waviness” morphological characteristics and asymmetric “M” type development characteristics. A location model as well as a development cycle model of dynamic ground fissures were established for the first time, which can be used to predict the location and period of ground fissures. Based on the motion characteristics of hinged rock block structures, the mechanical mechanism of the self-healing phenomenon of dynamic ground fissures was revealed. A partition monitoring mode of working faces without coal pillar mining was proposed for the first time, which can reduce a lot of manpower and material resources. The coal mining subsidence basin is divided into a natural restoration area and an artificial restoration area. The combination of natural restoration and artificial guidance was used to control the ground fissures and reduce the associated costs. The research conclusions can provide a basis for mining damage evaluation and ecological environment protection in the aeolian sand area.

1. Introduction

Coal is the major fuel for energy consumption in the process of industrialization globally, which provides a strong foundation for the social and economic development of all countries. The top ten countries in terms of coal consumption are China, India, the United States, Japan, South Africa, Russia, Indonesia, South Korea, Vietnam, and Germany [1,2]. By the end of 2022, global coal production was still increasing and is expected to remain the same in the foreseeable future [3]. From 1960 to 2022, coal always ranked first in China’s primary energy consumption and continues to occupy the leading position. At present, coal consumption has reached 3.04 billion tons. The rapid growth of the domestic economy necessitated the shift of the center of coal development to the western mining area (Shaanxi, Shanxi, and Inner Mongolia provinces). It accounts for 70% of China’s coal production and will be the major region of coal supply for China in the future [4,5,6]. The northern Shaanxi mining area has attracted worldwide attention because of its advantages such as high coal quality, shallow burial depth, and simple geological and mining conditions.

Coal mining destroys the stress equilibrium of the rock strata, resulting in the movement and deformation of the overlying strata [7]. For several years, researchers have conducted extensive studies concerning the movement and deformation law of overlying strata. Singh et al. [8], Das [9], and Palchik [10] investigated mining areas in India and Ukraine. Mining subsidence is considered to be an unbalanced subsidence that occurs due to the inconsistency in rock strata structure as well as the rock strata expansion coefficient. Qian [11], Huang, and Xu [12,13] proposed the “masonry beam”, “step rock beam”, “key layer”, and other structural models using field measurements and theoretical analyses. These models can reasonably explain the phenomena of mine pressure and surface damage caused by coal mining (shallow depth). Considering the Shendong mining area as the research object, Wang and Bai [14,15] proposed a strip interval filling mining technology, and its ability to slow down the surface subsidence was verified through experimental simulation.

The mining area in northern Shaanxi is located in the arid area of Northwest China, which is mostly situated in the Maowusu Desert and is adjacent to the Loess Plateau. It is a typical ecologically sensitive area. Coal mining in the northern Shaanxi region is characterized by strong surface movement and deformation as well as short and intense overlying strata activity. Furthermore, the surface can be subjected to a large number of ground fissures, leading to soil nutrient loss, surface runoff depletion, vegetation withering, and death, as well as endangering the ecological environment [16,17,18]. This ground fissure disaster is a global issue. In recent years, there has been no decrease in heat, indicating its extremely detrimental effects. Kalogirou et al. [19] and Abdallah et al. [20] analyzed the influence of mining-induced ground fissures on surface buildings and structures. They demonstrated that ground fissures inflict great damage to buildings, which provided an engineering basis for protecting ground structures. Zhang and Singh et al. [2,3,8] combined remote sensing technology and particle flow numerical simulation to analyze the development process of mining-induced ground fissures. The breaking of the key strata will lead to the formation of step cracks on the surface. Based on an extensive measured dataset, Dai et al. [21] and Guo et al. [22] determined that the fissure distance in the Shendong mining area was consistent with the periodic weighting step of the working face, and the number and width of the ground fissures had a certain degree of regularity. Previous studies have demonstrated that ground fissures in aeolian sand landforms exhibit self-healing characteristics, and the non-pillar mining method can maximize the recovery rate of coal, unavoidably leading to serious ecological and environmental impacts on the vulnerable mining areas in the west [23]. Currently, studies on the movement of overlying strata and the self-healing mechanism of ground fissures resulting from non-pillar mining in aeolian sandy areas are limited.

Therefore, this study considered the non-pillar mining face in China’s aeolian sand area as the research object and analyzed the mechanism of failure of the overlying strata as well as the characteristics of surface movement and deformation. A mechanical model of overlying strata was constructed, and the self-healing characteristics of ground fissures were reasonably revealed. A location model and development cycle model of dynamic ground fissures were established for the first time, which can be used to predict the location and period of ground fissures. A partition monitoring mode of working faces without coal pillar mining was proposed for the first time, which can reduce a lot of manpower and material resources. The research results can provide a basic for the prevention of geological disasters and ecological restoration in mining areas.

2. Materials and Methods

2.1. Study Area

The study area is located in the Shennan mining area, Yulin City, Shaanxi Province, China. The terrain of the mining area is complex, with the development of a large number of valley gullies. The terrain is low in the east (942.2 m) and high in the west (1364.4 m). The mining area is located adjacent to the Maowusu Desert, and the surface is covered with aeolian sand. The area is characterized by four distinct seasons and less rainfall, mostly concentrated between July and September. It experiences a semi-arid continental desert climate. Figure 1 provides an overview of the study area.

The mining coal seam of the 12,013 working face in the Shennan mining area is the 2# coal seam, and the mining size of the working face is 1782 × 334 m². The average mining depth is 148 m, the coal seam dip angle is 2°, the coal seam thickness is 4 m, and the mining speed is 10 m/d. The roof and floor of the working face are medium-grained sandstone (19 m) and siltstone (4 m), respectively. The topsoil thickness varies from 15 to 77 m. The method of roof cutting and pressure relief without coal pillar retreat mining was employed for the working face. Non-pillar mining is conducive to the rational development of coal resources, the improvement of the coal recovery rate, and the improvement of economic benefits by not leaving coal pillars. It is beneficial to reduce the tunneling rate and improve the maintenance of roadways. Because non-pillar mining greatly reduces the amount of coal loss, the probability of natural fires is greatly reduced, which is conducive to the safety of mining production.

2.2. Mechanical Model of Overlying Strata Structure for Non-Pillar Mining in the Aeolian Sand Area

2.2.1. Mechanical Model of Overlying Strata Structure in the First Working Face Mining

In the process of non-coal-pillar mining, the roof is pre-split and cut with blasting technology, and the height of the roof cutting is estimated using the following equation:

$$h_c=\frac{\left(M-\Delta h_1-\Delta h_2\right)}{k_p-1}$$

(1)

where h_c indicates the height of the roof cutting, $\Delta h_1$ and $\Delta h_2$ are the amount of roof subsidence and floor heave, respectively, M is the thickness of the coal seam, and k_p represents the coefficient of broken expansion.

It is assumed that n rock strata exist within the range of the roof cutting height, the thickness is D_i, and the theoretical expansion coefficient of the rock is represented as k_pi. Considering the thickness of the roof as the weight, the coefficient of rock expansion is calculated using the following equation:

$$k_p={\displaystyle \sum _{i=1}^n{k}_{pi}}\frac{D_i}{M}$$

(2)

Figure 2 shows the mechanical model of the overlying strata structure in the first mining face. Rock block A is not affected by mining, and rock block B is fractured through tensile action, resulting in the formation of an articulated rock block structure with rock block A. Rock blocks B and C are subjected to the combined action of the overlying strata pressure and the bearing capacity of the rock blocks collapsed at the bottom, leading to compression among the upper sections and cracking at the bottom. As the bulking gangue in the goaf is gradually compacted, rock block C leads to the rotation of rock block B and its deformation to the side of the goaf through the hinge effect.

Salamon proposed the stress–strain relationship for the compression process of broken rock mass [24]:

$$\sigma =\frac{E_0\epsilon }{1-\raisebox{1ex}{$\epsilon $}\!\left/ \!\raisebox{-1ex}{${\epsilon }_m$}\right.}$$

(3)

where σ is the vertical stress of the caving rock; E₀ indicates the initial tangent modulus of rock; ε is the strain of the caving rock block; and ε_m indicates the maximum strain of rock subjected to rapid expansion. The expression of E₀ obtained via a regression analysis is as follows:

$$E_0=\frac{10.39{\sigma }_c^{1.042}}{k_p^{7.7}}$$

(4)

ε_m can be estimated with Equation (5):

$${\epsilon }_m=\frac{k_p-1}{k_p}$$

(5)

Equations (3)–(5) were combined to estimate the vertical stress of caving rock within the range of roof cutting, which is given in Equation (6):

$$\sigma =\frac{10.39{\sigma }_c^{1.042}\epsilon }{k_p^{7.7}(1-\frac{k_p}{k_p-1}\epsilon )}=q={\displaystyle \sum \gamma H}$$

(6)

where q indicates the load on the rock block, $\gamma$ is the bulk density, and H is the height of the overlying strata. Assuming no compression deformation in the rock strata within the roof cutting range, the subsidence of rock block C can be considered equal to the compression displacement caused by the caving gangue:

$$S=\epsilon (h_c+M)$$

(7)

Equations (6) and (7) were combined to estimate the subsidence (S) of rock block C:

$$S=\frac{(h_c+M)(k_p-1){\displaystyle \sum \gamma H}}{k_p{\displaystyle \sum \gamma H}+\frac{10.39{\sigma }_c^{1.042}}{k_p^{7.7}}(k_p-1)}$$

(8)

Assuming the length of rock block B as l, the expression can be given as

$$l=l^{\prime }[-\frac{l^{\prime }}{S_0}+\sqrt{\frac{{l^{\prime }}^2}{S_0^2}+\frac{3}{2}}]$$

(9)

where S₀ is the length of the working face and l′ represents the horizontal step distance of rock block B, which is determined through measurements or the following equation:

$$l^{\prime }=2h_m\sqrt{\frac{R_t}{3n}}$$

(10)

where R_t is the tensile strength.

Equations (8)–(10) were combined to obtain the rotation angle (θ) of rock block B to the goaf:

$$\theta =\arctan \frac{S}{l}$$

(11)

2.2.2. Mechanical Model of Overlying Strata Structure in the Second Working Face Mining

Following the mining of the adjacent working face, the overlying strata structure undergoes alteration. As the direct roof of the first mining working face eliminates the connection with the goaf via the pre-splitting method and no coal pillar or other support exists at the coal seam roadway, the rock beam at the first mining face remains unconstrained and is in a free state. As working face 2 undergoes mining, the rock strata in the pre-splitting range descend to the floor along the cutting surface without collapsing into a loose structure, and the burden of the overlying strata pressure is completely carried by the cutting rock layer (Figure 3). Rock block A gradually sinks to the top of the cutting rock strata under the pressure of the overlying strata, and the subsidence represents the mining height of working face 2. Due to the high stability and large rigidity of the cutting rock strata, there is only a slight compression displacement (which is neglected) and no change in the height of the cutting rock strata at working face 2. With the continuous subsidence of rock block A, the rotational deformation of rock block B along the new goaf side is driven by a hinged action.

A comparison of the support bodies in the two goaves indicated that one is the loose gangue formed by the collapse of the cutting rock strata, and the other is the overall cutting rock strata. The rock strata exhibit the same properties but different states. Based on the bulking characteristics of the caving gangue, it can be observed that, following the restoration of the original rock stress, the height (h_z) from the bottom of rock block C to the coal seam floor is larger than the height (h_c) from the bottom of rock block A to the coal seam floor, i.e.,

$$h_z>h_c$$

(12)

At the same time, the subsidence (S) of rock block C is less than the subsidence (M) of key block A, i.e.,

$$S<M$$

(13)

The subsidence (S₂) of rock block A, relative to rock block C, is given as

$$S_2=M-S$$

(14)

The angle (θ₂) of the reverse rotation of rock block B is estimated as follows:

$${\theta }_2=\arctan \frac{S_2}{l}$$

(15)

Briefly, following the mining of the first working face, the relative position of rock block A is higher than that of rock block B, and the fracture between the two rock blocks exhibits upper cracking–lower extrusion. Following the mining of the second working face, the relative position of rock block A is lower than that of rock block B, and the fracture between the two rock blocks is transformed into the lower cracking–upper extrusion. At the same time, the rotation of rock block B to goaf 2 leads to the shifting of the crack between rock block B and C from lower-end cracking–upper extrusion to upper cracking–lower extrusion. This offers a reasonable explanation for the phenomenon of ground crack closure. The conventional single-coal-seam mining method only results in a disturbance to the overlying strata, while non-pillar mining can cause a secondary disturbance to the overlying strata on the adjacent working face, resulting in an enhanced surface influence range of the mining activity.

2.3. Field Measurement and Experimental Simulation

2.3.1. Field Measurement

Considering the characteristics of the aeolian sand landform in the study area (with significant surface movement and deformation and a short duration), the observation station for surface movement was arranged based on the profile line. Two observation lines, which are arranged based on the trend and tendency of the working face, are observed (Figure 1c). Each observation line is composed of multiple observation points located 15 m apart, and the observation points were identified based on the positioning of wooden piles. During the process of measurements along the working face, the movement of each observation point was measured regularly and repeatedly using the method of total station and level measurement. The location and width of ground fissures were determined using GPS-RTK and tape. Figure 4 is a photo of the field measurement. It can be seen that the experimenters are using the instrument for leveling to obtain the movement and deformation of the surface observation points.

2.3.2. Experimental Simulation

• Similar material simulation

Similar simulation is an important scientific research method to make a model similar to the prototype in the laboratory, according to the similarity principle. The rock strata were reduced to a certain extent to develop a model that simulates the movement and deformation of the overlying strata. The experimental model must satisfy the criteria of bulk density, time, geometry, and stress similarity. The expression is provided as follows:

$${\alpha }_l=\frac{L_m}{L_p}$$

(16)

$${\alpha }_{\gamma }=\frac{{\gamma }_m}{{\gamma }_p}$$

(17)

$${\alpha }_{\sigma }=\frac{{\sigma }_m}{{\sigma }_p}={\alpha }_l\cdot {\alpha }_{\gamma }$$

(18)

$${\alpha }_t=\sqrt{{\alpha }_l}$$

(19)

where $\alpha$ is the similarity ratio; l, $\gamma$, $\sigma$, and t represent the size, bulk density, stress, and time, respectively; m is the experimental model; and p is the prototype.

The size of the experimental model was 500 × 200 cm², and the ratios of the geometric similarity and bulk density similarity corresponded to 1:100 and 1:1, respectively. A distance of 100 cm was left on both sides of the coal pillars, with each step involving mining 10 cm (actual mining corresponds to 10 m) to simulate the actual mining progress, resulting in a total mining distance of 300 cm. The experimental model is presented in Figure 5.

The primary material of the experiment was sand, and the cementing materials, calcium carbonate and gypsum, were reasonably mixed. Considering the factor of the aeolian sand layer, an appropriate amount of sawdust was added to the material to increase the looseness, and an appropriate amount of borax was added to delay the setting time of the material. The overlying strata structure was merged and removed to achieve optimal experimental results, and the observation lines were arranged in different rock strata to monitor its movement and deformation (Figure 5).

2.

Numerical simulation (UDEC 6.0 software)

The UDEC (Universal Distinct Element Code) software was developed by Itasca, which is a discrete element program. The UDEC software is capable of describing the mechanical behavior of discrete media in a two-dimensional space. It is widely used in the fields of geotechnical engineering, mining engineering, and others. It can be used to reconstruct the failure of overlying strata, as well as surface movement and deformation following coal seam mining. It is commonly used to solve practical engineering problems [3].

Combined with the mining parameters and rock strata occurrence conditions of the study area, a numerical model was developed. The rock strata from top to bottom were composed of an aeolian sand layer, a sandy soil layer, a red soil layer, sandstone layer 1, sandstone layer 2, a coal seam, and sandstone layer 3. The size of the numerical model was 820 × 200 m², and 100 m coal pillars were preserved on both sides to eliminate the influence of surface subsidence on the boundary. The numerical model employed the Mohr–Coulomb constitutive relation, and the fixed constraint conditions were applied on the left and right sides of the boundary as well as the lower boundary. In the absence of coal pillar mining, gob-side entry retaining via roof cutting and pressure relief were simulated by dividing the fracture unit. The width of the two working faces corresponded to 300 m (Figure 6), and the lithologic parameters were determined using the trial and error method in combination with

Table 1. Experimental ratio of similar materials.

Layer Number	Lithology	Observation Line Number	Thickness (m)	Total Thickness (m)	Compressive Strength (kPa)	Total Weight (kg)	Sand Weight (kg)	Gypsum Weight (kg)	Calcium Carbonate Weight (kg)
41–42	Aeolian sand	1	7.5	7.5	4	198	173.25	17.33	7.43
39–40	Sandy soil		8	15.5	4	211.2	184.80	18.48	7.92
31–38	Red soil	2	31	46.5	4	818.4	716.10	71.61	30.69
29–30	Fine-grained sandstone		6	52.5	244	158.4	135.77	6.79	15.84
28	Sandy mudstone	3	3	55.5	310	79.2	67.89	5.66	5.66
26–27	Fine-grained sandstone		7	62.5	320	184.8	147.84	11.09	25.87
25	Coarse sandstone		2	64.5	67	52.8	46.20	4.62	1.98
21–24	Siltstone	4–5	16	80.5	308	422.4	316.80	52.80	52.80
17–20	Medium-grained sandstone	8	13	93.5	153	343.2	300.30	21.45	21.45
15–16	Fine-grained sandstone		6	99.5	273	158.4	132.00	7.92	18.48
11–14	Siltstone	7–8	16	115.5	334	422.4	352.00	35.20	35.20
7–10	Fine-grained sandstone	9	9	124.5	310	237.6	203.66	16.97	16.97
3–7	Medium-grained sandstone	10	19	143.5	302	501.6	429.94	35.83	35.83
2	Coal		4	147.5	294	105.6	92.40	3.96	9.24
1	Siltstone		4	151.5	634	105.6	79.20	13.20	13.20

Note: In order to be consistent with the actual situation, according to previous experience, the thick rock layer is divided into multiple layers, and the thickness of each layer is roughly 4–5 cm. Numbers (1–42) in Table 1 refer not to the number of rock strata, but the number of layers of the model.

. In the numerical model, E and u could not reasonably reflect the physical properties of the rock joints, so they were expressed as normal stiffness and tangential stiffness. These joint parameters were mainly determined using empirical formula, which were roughly 1/10 of the mechanical parameters of the rock strata on both sides [2].

Figure 7 represents the subsidence curve and subsidence basin of each rock stratum obtained from the numerical simulation. Following the mining of the first working face, the subsidence of sandstone layer 2 was the highest at 4 m, while the subsidence of the aeolian sand layer was the lowest at 2.7 m. The subsidence trend of the gob-side entry-retaining side was not as significant as that of the left coal pillar side. Similarly, in the case of mining the second working face, a new subsidence basin was formed on the surface, and different degrees of protrusions appeared in the middle of the subsidence basin. The surface subsidence above the gob-side entry retaining varied between 2.3 and 2.5 m, which was significantly smaller than the maximum surface subsidence. The measured results were consistent, indicating that the roof cutting and pressure relief gob-side entry retaining mining method was capable of decreasing the surface subsidence, but still produced ground fissures.

3. Results and Analysis

3.1. Characteristic Analysis of Surface Movement and Deformation Parameters

Mining subsidence can result in surface movement and deformation. Few movement and deformation values were observed at the edge of the surface subsidence basin, which can be used to evaluate the degree of danger for buildings and determine if they are influenced by mining activity (Figure 8). There were variations in the surface movement and deformation parameters (surface movement angle parameters and probability integral parameters) estimated for various mining areas. These parameters depend on the geological and mining conditions as well as topography.

Based on the measured data and the probability integral method, the surface movement and deformation parameters were estimated for the study area. These parameters were then compared with the Shendong mining area to analyze the differences in their characteristics, as shown in

Table 2. Comparison of surface movement and deformation parameters.

Parameters	Shennan Mine 12013	Shennan Mine 1212	Shennan Mine 15201	Shendong Mine 12406	Shendong Mine 52304	Shendong Mine 22111
Mining depth (m)	148	190	128	200	225	250
Mining thickness (m)	4	4.8	6.2	4.5	6.9	2.8
Loose layer thickness (m)	48	95	70	17	30	8
Bedrock thickness (m)	100	95	58
Trend boundary angle (°) Tendency boundary angle (uphill/downhill) (°)	71.3/ 58.7/35.2	65.9/ 55.6/51.7	63/ 54/54	45/ 46.5/46.5	42.5/ 42.2/42.2	48/ 51.9/51.9
Trend displacement angle (°) Tendency displacement angle (uphill/downhill) (°)	58.7/ 63.1/51	70.3/ 62.3/57.6	71.2/ 74/74	81.7/ 85.1/80.6	--/ 82/82	74.3/ 75.9/75.9
Trend fissure angle (°) Tendency fissure angle (uphill/downhill) (°)	77/ 79/58	77.2/ --	80/ 90/90	87.6/ 87.6/93.3	--	--
Maximum subsidence angle (°)	89.4	83.9	--	--	--	--
Subsidence coefficient	0.66	0.78	0.71	0.55	0.61	0.51
Horizontal movement coefficient	0.41	0.44	0.47	0.26	0.32	0.36
Tangent of major influence angle	2.4	2.4	2.8	2.5	2.9	2.7
Inflection point offset (m)	45	45	26	40	38	40

.

(1)

The trend boundary angle of the Shennan mining area corresponded to 63–71.3°, and the tendency boundary angle (uphill and downhill) corresponded to 35.2–58.7°, which was higher than that of the Shendong mining area. When compared with conventional coal pillar mining, the downhill boundary angle obtained through non-coal-pillar mining was smaller, indicating a significant influence range of the surface subsidence basin.

(2)

The trend displacement angle of the Shennan mining area corresponded to 58–79°, which was smaller than that of the Shendong mining area. Due to no coal pillar mining and a thick loose layer, the tendency displacement angle (uphill and downhill) was smaller, indicating that the loose layer in the aeolian sand layer exhibits poor mechanical properties and is susceptible to movement and deformation under the influence of mining.

(3)

The fissure angle of the Shennan mining area was generally smaller than that of the Shendong mining area because of the greater thickness of the loose layer.

(4)

The horizontal movement coefficient and subsidence coefficient of the Shennan mining area were higher, and the tangent of the major influence angle was smaller, indicating a significant intensity of surface movement and deformation.

3.2. Analysis of Overlying Strata and Surface Movement and Deformation Process

3.2.1. Failure Process of Overlying Strata

The geometric similarity ratio of the experimental model corresponded to 1:100, with every 10 cm of the working face representing 10 m of actual mining. When the working face advances by 60–110 cm, the working face has experienced the process of “the first collapse of the main roof—the first breaking of the key strata”. The overburden breaking angle of the left is basically unchanged, and the caving height and breaking angle in front of the working face gradually increase, indicating that the failure pattern of the overlying strata is asymmetrically distributed. When the working face advances by 140–150 cm (close to the mining depth), the height of the caving zone increases by five times the mining thickness. Due to the different physical and mechanical properties of the overlying strata and the topsoil, the boundary between the two is more obvious. Meanwhile, the overburden breaking angle on both sides of the working face remains unchanged, implying no change in the overburden breaking angle following the breaking of key strata. At this time, the side fissures of the open-off cut were developed and belong to the severely damaged area, which was consistent with the measured results. The damages to the self-healing area of the rock and soil sand layer were observed above the goaf, and the self-healing effect of fissures in the area was significant.

The rock block in front of the working face undergoes a continuous cyclic process of hinge rotation–compaction, which results in the closure effect of the fissure. The overlying strata are continuously subjected to tensile compression deformation, which is transferred to the surface forming dynamic ground fissures, accompanied by the “cracking-closing” phenomenon. The advanced fissures were observed in the experimental model, indicating that the mining fissures were composed of both upward and downward fissures (Figure 9).

Briefly, following the mining of the working face, the overlying strata structure and fissure field of the stope tend to become stable, and the boundary of the breaking angle between the topsoil and the overlying strata is more evident. The breaking angle and influence range of the topsoil exhibited a significant increase in contrast to the interior part of the overlying strata. This is attributed to the poor mechanical properties of the topsoil. When the overlying strata is broken, the upper soil can rapidly compact the fissures, and to some extent weaken the development of the overlying strata fissures on the surface. Under the influence of mining driving forces, the overlying strata can form a stable mechanical structure, leading to the rapid healing of the ground fissures. This provides a reasonable explanation for the self-healing phenomenon of ground fissures that occurs in the aeolian sand area. Figure 10 shows the damage zoning of the overlying strata. It can be seen that after the mining of the working face, the failure form of overlying strata is “trapezoid” as a whole. The rock strata on both sides of the trapezoid (open-off cut and stopping line) are seriously damaged by the influence of mining, resulting in many fissures. The internal rock strata of the trapezoid can be divided into three areas according to the damage degree, namely, the caving zone (5M), the self-restoration area of the overlying strata damage, and the self-restoration area of the topsoil damage. The range of the self-restoration area increases with mining activities. At the same time, the boundary between the breaking angle of the topsoil and the rock strata is more obvious, indicating that their self-restoration degree is different.

3.2.2. Development and Evolution of Mining-Induced Fissures

The occurrence and development of fissures in overlying strata were closely associated with the underground mining activities, but also dependent on the physical properties, structural bedding, thickness, and other factors of the rock strata. Overlying strata fissures exhibited dynamic changes with the expansion of goaf. The thick and hard key strata in the mining area of northern Shaanxi control the movement and deformation of overlying strata. Before the breaking of key strata, the separation structure was generated in the overlying strata. However, following the breaking of key strata, the masonry beam structure was formed to support the overlying strata. In addition, the rock strata were sheared to form transverse and longitudinal fissures.

The development and extension of fissures in the overlying strata were associated with mining activities, and the formation of fissures was the direct result of rock mass failure and fracture. In the horizontal direction, the width and density of the fissures outside the goaf decreased gradually. On the contrary, in the vertical direction, the rock block in the caving zone had a high degree of fragmentation and a large caving range, resulting in fissures with a large width and density. In the fissure zone, the degree of fissure development was lower than that of the caving zone due to the structural integrity of the roadblocks supported by the hinged rock beam. In the curved subsidence zone, due to the physical properties of the surface soil layer in the aeolian sand area of northern Shaanxi, the fissures in the surface subsidence basin exhibit a self-healing function, and only large longitudinal fissures are observed along the open-off cut (Figure 11). Thus, the characteristics of overlying strata fissure zoning are evident. Under the combined influence of special landforms and rock strata structures in the aeolian sand area of northern Shaanxi, the phenomenon of synergistic “self-healing” is observed in rock strata, soil layers, and sand layers, which exhibited a strong correlation with the coordinated process of rock strata breaking, caving, and migration.

Similar material simulation experiments demonstrated that dynamic ground fissures were observed for the first time when the working face reached 0.88 H₀. With the continuous progression of the working face, multiple dynamic fissures were formed periodically on the surface and a self-healing phenomenon was observed. The number and extent of the ground fissures was limited by the size of the model. When the mining of the working face was completed, the overlying strata formed a “vertical three-zone” structure, exhibiting relatively wide vertical fissures with a long development distance, and relatively narrow horizontal fissures with a short development distance. The process of the development of vertical fissures was associated with the period of mining pressure, which represents the collapse period of the overlying strata. With the continuous progression of the working face, the height of the caving zone increased gradually and reached a state of stability. This, in turn, led to an expansion in the range of the caving zone. Once the ground fissure was subjected to tensile compression, the width exhibited the characteristics of initial cracking followed by closing (Figure 12b,c).

3.2.3. Overlying Strata and the Law of Surface Movement and Deformation

The surface subsidence curves of the experimental model at varying distances of the working face are shown in Figure 13. Once the working face reaches 110 m, the effect of the stope is transferred to the surface, initiating a gradual sinking process. With the continuous progression of the working face, the overlying strata experienced multiple breakages and caving, resulting in several subsidence curves, each corresponding to a distinct sinking event. The extension of the working face to 140 m allowed for the complete mining of the surface. Moreover, the surface subsidence achieved the maximum value under the geological and mining conditions. With the continuous progression of the working face, the maximum surface subsidence remained unchanged. Further, the scope of the subsidence basin continued to increase, with the surface reaching the stage of super-full mining.

Briefly, during the initial stage of working face mining, the subsidence curve was gentle and exhibited a narrow range. With the continuous progression of the working face, the surface continued to subside, resulting in the continuous expansion of the scope of the subsidence basin along the mining direction until the surface reached the stage of full mining. Once the surface entered the upper full mining stage, no further change in the maximum subsidence was observed. The range of the surface subsidence basin and the magnitude of the subsidence velocity were directly proportional to the violent movement of the overlying strata. Under the influence of gravity, the surface soil layer moves and sinks to the inside of the working face, which has significant convergence and is consistent with the measured data.

The surface deformation curves of the experimental model at varying distances of the working face are shown in Figure 14. Based on a large amount of measured data in the mining area, Dai et al. [21] hypothesized that the horizontal deformation value is the standard for measuring the occurrence of ground fissures, with its threshold generally varying between 2 and 4 mm/m. In combination with the field monitoring data, a horizontal deformation of 2 mm/m was used as the critical threshold for ground fissures. The horizontal deformation near the maximum subsidence point was primarily concentrated within the negative range, while the horizontal deformation near the boundary of the subsidence basin was nearly negligible, which conforms to the general law. With the continuous progression of the working face, the horizontal deformation values for several locations in the surface subsidence basin exceeded the critical threshold of fissure generation, indicating the occurrence of dynamic ground fissures in the subsidence basin. Dynamic ground fissures are affected by surface tensile and compressive stress and exhibit a self-healing effect. The horizontal deformation value was maximum on the side of the open-off cut, which was associated with the increase in subsidence and a rapid convergence at the edge of the subsidence basin, indicating the continuous exposure of the area to horizontal stretching. As shown in Figure 14, the horizontal deformation of the surface represents a cohesive continuum, rather than a single point change.

The subsidence curves of different rock strata in the experimental model are shown in Figure 15. A continuous decrease in the subsidence value and subsidence range from the goaf to the surface was observed. The subsidence values of observation line nos. 1–3 were similar to the shape of the subsidence curve, which was associated with the varying trend of the topsoil. Observation line no. 4, located close to the mudstone layer, represents the subsidence trend at the boundary of the rock and soil layer, with greater subsidence range. Observation line nos. 5–9 represent the synergistic subsidence trend of each rock strata, and the subsidence range was consistent with the subsidence curve shape. Observation line no. 10 is located in the caving zone. The rock mass in this area exhibits a high degree of breaking and concentrated fissure development. Therefore, the shape of the subsidence curve is more complex.

Based on the observation data obtained from the experimental model, a linear negative correlation exists between the subsidence value and the height of the rock strata, i.e., as the distance from the goaf increases, the subsidence value decreases. This confirms the zoning characteristics of the “vertical three zones” of the overlying strata (Figure 16).

3.3. Characteristics and Mechanism Analysis of Ground Fissure Self-Healing in Aeolian Sand Area

3.3.1. Distribution Characteristics of Ground Fissures

The surface soil layer is affected by the mining process, leading to the deformation of the stress condition of the internal points. This deformation is characterized by non-uniform subsidence and horizontal movement. Ground fissures can occur when the topsoil exceeds its plastic deformation limit. The coupling of overlying strata fracturing and surface soil deformation leads to the formation of ground fissures. It is a bottom-up dynamic development process, i.e., coal seam mining–rock strata failure–surface movement and deformation–ground fissure generation. Combined with the distribution of ground fissures, it can be divided into boundary ground fissures and dynamic ground fissures. The former are developed near the stopping line, open-off cut, and roadway of the working face, while the latter are developed in the working face and are generally aligned parallel to the open-off cut, as shown in Figure 17.

In contrast to the ground fissures distributed outside the boundary of the working face, the boundary ground fissures that form due to high-intensity mining activity are mostly situated within the boundary of the working face, with no outward expansion. Previous researchers have reported that boundary ground fissures are generally distributed in an arc shape [25].

The dynamic ground fissures initially formed near the open-off cut of the working face. However, the continuous progression of the working face led to the continuous development of new ground fissures at a certain distance. Moreover, the original ground fissures undergo continuous widening and expansion. When the working face reaches a certain distance, the original ground fissures undergo gradual closure. Monitoring the development process of dynamic ground fissures revealed that the location of the ground fissures was 10.36 m ahead of the working face on average. In addition, the width demonstrated the variational characteristics of “open first, close, and then open” (asymmetric “M” type). The development period was 18 days, and no significant relationship between the depth and width of the ground fissures was observed. The monitoring data differed from the development characteristics of ground fissures reported in previous studies, which depend on factors such as monitoring time intervals and geological mining conditions. As the Shennan coalfield is located in the mining area of northern Shaanxi and is adjacent to the edge of the Maowusu Desert, the dynamic ground fissures in the aeolian sand landform environment exhibit discontinuous morphological characteristics and are characterized by “waviness” (Figure 18). This is due to the poor physical and mechanical properties of the sand layer. Plastic deformation easily occurs under the influence of mining subsidence. However, it does not possess the specific characteristics of soil cohesion and loose flow but exhibits the characteristics of natural restoration.

A continuous observation of the location of dynamic ground fissures revealed that the ground fissures were consistently located ahead of the working face. To further study the location of the dynamic ground fissures, the relationship between the mining speed and advanced fissure distance was analyzed (Figure 19). The linear, exponential, logarithmic, and other regression models revealed a significant linear positive correlation between these two factors (correlation coefficient = 0.92), and the functional relationship is given in the following equation:

$$Y=0.637V+2.715+d$$

(20)

where Y indicates the position of the ground fissure from the working face, m; V is the mining speed of the working face, m/d; and d is the mining progress of the working face, m.

Equation (20) indicates that the location of dynamic ground fissures depends on the geological and mining conditions. Previous studies have demonstrated that dynamic ground fissures exhibit the characteristic of single peak development. Hu et al. [23] were the first to reveal that the dynamic ground fissures in the Shendong mining area exhibit the characteristics of “M”-type double cycle development and rapid closure. Additionally, a dynamic ground fissure development cycle model was constructed that includes geological and mining conditions. The monitoring data indicated that the dynamic ground fissure exhibits the characteristic of asymmetric “M”-type development. The second cracking closure period accounted for 28% of the initial cracking closure period. The data in

Table 3. Dynamic fissure parameters.

Working Face Name	Maximum Width (mm)	Ratio of Bedrock Thickness to Mining Thickness	Horizontal Deformation (mm/m)	Mining Speed (m/d)
Shennan Mine S12013	57	22	2.7~3.5	10
Shendong Mine 12406	31	37	4.05	12

show that the horizontal deformation of dynamic ground fissures is lower compared to in the Shendong area. During the development of ground fissures, the mechanical properties of overlying strata in different areas result in spatial and temporal differences in the formation of ground fissures, but generally meet the threshold criteria of horizontal deformation. Due to the evident impact of aeolian sand landforms on small fissures, dynamic ground fissures exhibit a significant self-repairing ability.

3.3.2. Development Cycle of Ground Fissures

The formation of dynamic ground fissures is the result of the progressive influence of the subsidence basin of coal mining. Once the horizontal deformation reaches a certain threshold within its range, ground fissures can be formed until the surface movement stabilizes and eventually closes (Figure 20).

A two-dimensional coordinate system corresponding to the coordinate orientation of the surface and the underground working face was established. It is possible to obtain the distribution characteristics and evolution of ground fissures in the aeolian sand area of northern Shaanxi through long-term continuous monitoring. Based on the measured data and the geological and mining conditions of the study area, a mathematical model of the dynamic ground fissure development cycle was developed (Figure 21).

The model includes the progressive fissure distance, the lag angle of the maximum subsidence velocity, the mining depth, the advanced influence angle, and the mining speed, which can quantitatively explain the process of the development of ground fissures and accurately predict the location and development cycle of ground fissures in the mining area of northern Shaanxi. In addition, it provides scientific technical support for land reclamation and ecological reconstruction in this area, as given in Equation (21).

$$T=H_0\left(\frac{1}{\tan \delta }+\frac{1}{\tan \phi }\right)/V$$

(21)

where T represents the development cycle of ground fissures, d; H₀ is the mining depth, m; $\delta$ is the influence angle of the progressive fissure; $\phi$ is the lag angle of the maximum subsidence velocity; and V is the mining speed, m/d. Based on the measured data, the lag angle of the maximum subsidence velocity can be estimated using the following equations:

$$\phi =69.01-0.01395\left(H_0-h\right)M\hspace{0.17em}({\mathrm{R}}^2=0.93)$$

(22)

where h is the thickness of the loose layer, m, and M is the coal thickness, m.

Substituting Equation (22) in Equation (21), the expression for estimating the development period of ground fissures was obtained:

$$T=\frac{H_0}{V}\left\{\frac{1}{\tan \delta }+\frac{1}{\tan \left[69.01-0.01395\left(H_0-h\right)\frac{V}{M}\right]}\right\}$$

(23)

The above equation validated the mathematical model of the dynamic ground fissure development cycle proposed by Hu et al. [23]. This model can accurately predict the development cycle of dynamic ground fissures and has the potential for a wide range of applications.

3.3.3. Characteristics and Mechanism Analysis of Ground Fissure Self-Healing

Previous research studies have proven that the physical and mechanical properties of topsoil and the degree of topographic relief are closely associated with the distribution of ground fissures. These factors exhibit zoning characteristics that contribute to this association [26]. Ground fissures are significantly affected by natural forces.

Soil biocrusts are formed due to rainwater splashing and soil clay dispersion due to the blocking of soil pores. In the rainy season, the accumulation of rainwater containing plant debris in the ground fissure area increases the water content of the surface sand layer. A large number of black biological crusts are formed on the surface due to complex biochemical processes, further promoting the closure of ground fissures (Figure 22).

Following the mass extinction event that occurred during the Permian and Triassic eras, the changes in the ecological environment of Earth resulted in the emergence of more adaptable Equisetum plants, which played a vital role in terrestrial vegetation during the early and middle Triassic eras. After hundreds of millions of years of evolution, Equisetum has now become Neocalamites [27]. It is a rare living fossil in the plant kingdom, but is much smaller in shape and size than it was millions of years ago. Equisetum plants have a strong adaptability as well as reproductive ability, and their rate of growth and reproduction is comparable to that of bamboo. These plants can grow in areas with abundant humus and water resources. Figure 22c shows Neocalamites grown in the study area, located 200 m away from the open-off cut of the 12,013 working face. This growth manifests as grass accumulation, indicating its rapid growth. This implies that the self-healing effect of the aeolian sand area is well established, and that the ecological environment is gradually progressing in a positive direction. Briefly, the aeolian sand area is composed mostly of sand layers, which is characterized by strong looseness, no plasticity, poor water stability, and strong wet sinking. Coal mining can lead to surface subsidence and local cracking. The ground fissures exhibit self-healing characteristics that represent rapid cracking–closing due to the physical properties of the aeolian sand layer, as well as the mining driving force, natural wind, rain invasion, and desertification. Soil nutrients can be restored to their original levels in a short time, and the surface ecology supports self-healing.

Based on the numerical simulation results, the first weighting distance of the main roof corresponded to 40 m, and the average periodic weighting distance varied between 10 and 20 m, which was generally consistent with the measured data (Figure 23). Therefore, the self-healing mechanism of ground fissures can be elucidated through numerical simulations.

(1)

Upon the initial weighting of the working face, the immediate roof experienced a comprehensive collapse, and a hinged rock structure was formed within the caving zone. The separation of the overlying strata was observed at the key strata, and several places were affected by the shear force. Stress concentration and tension occurred within the overlying strata, subsequently propagating through the entire surface. The stress concentration area emerged at the center of the subsidence basin.

(2)

The overall collapse of the overlying strata transpired in a backward manner, and the periodic weighting of the working face was evident. Under the influence of tension, ground fissures were formed directly above the open-off cut and the working face. Although the cracking phenomenon at the edge of the subsidence basin exhibited deceleration, the dynamic ground fissures that formed above the working face were affected by the movement of the loose layer.

(3)

With the continuous mining of the working face, the extent of the surface subsidence basin continued to increase. The original surface tensile stress zone was transformed into a compressive stress zone, and the dynamic ground fissures transformed from a tensile to a compressive state. The fissure width increased, followed by a decrease, representing a “closure effect”. Due to the overall collapse and compaction in the middle of the overburden, a new stress concentration zone was formed.

4. Discussion

4.1. Influence of Non-Pillar Mining on Surface Damage in the Aeolian Sand Area

The degradation of the ecological environment caused by non-pillar mining in the aeolian sand area is significantly less than that of open-pit mining. Its impact on the ecological environment of the mining area mainly focuses on the destruction of underground rock strata and groundwater circulation systems, resulting in geological environmental issues such as subsidence basins, ground fissures, and vegetation wilting.

(1)

Impact on the landscape

Mining activities lead to surface deformation, resulting in collapse pits, ground fissures, and steps, which in turn alter the original topography. The western coal mine area is primarily affected by ground fissures, which are progressively formed with the expansion of the mining area of the working face. These fissures are responsible for major influencing characteristics and hazards in the coal mine area. Based on the distribution and variational characteristics of ground fissures, it can be divided into dynamic and boundary ground fissures [23,28].

(2)

Impact on water resources

Mining activities lead to the deformation, fracture, and movement of overlying strata, which result in the formation of “horizontal three zones” and “vertical three zones”. A previous study demonstrated that during the process of coal mining subsidence, the groundwater level continues to decline, with overlying strata fissures and ground fissures being the major factors contributing to the changes in the water environment. Water flowing from the fractured zone and mine drainage can contribute to the depletion of groundwater resources and a decline in groundwater levels [29]. From 1986 to 1996, the surface water area of the Daliuta coal mine area decreased from 7.96 km² to 4.32 km², and the water resources were reduced by 46.7% of the original levels. The groundwater level decreased from 1–2 m to 4–6 m. It can be concluded that mining subsidence has a significant and long-term influence on the surface water in mining areas. The water content of the working face in the aeolian sand area was significantly affected by surface subsidence, and the soil moisture content in the mining area was much lower compared to the non-mining area.

(3)

Impact on the soil

Mining activities generate several ground fissures, which accelerate the seepage of surface water and atmospheric precipitation, resulting in the loss of soil nutrients and affecting the soil water content. The results of the time domain reflectometry (TDR) monitoring of the water content changes on both sides of the ground fissures demonstrated that the self-healing property of surface dynamic fissures can significantly reduce the impact on the water content of the topsoil [30]. The soil water content within 2 m of different coal mining subsidence areas was measured using a neutron meter for long-term monitoring. The results indicated that the soil water content above 60 cm increased linearly with depth, and the water content within 60–200 cm decreased slightly. Mining subsidence has a certain influence on soil water content. The water content in the subsidence basin is greater compared to in the surrounding area. The changes in water content of the basin area occur rapidly compared to the edge area, and the water content in the edge area requires at least 1–1.5 years of recovery [30].

(4)

Impact on vegetation

The geological disasters that are associated with ground fissures and collapse pits due to mining activities are significant, resulting in the tensile fracture of surface vegetation roots (Figure 24). As the water content of shallow groundwater and topsoil in the mining area decreases significantly, the ecological threshold required for vegetation growth is disrupted, which may lead to the withering or even death of vegetation [31]. There is a continuous decrease in the extent of vegetation, which directly leads to a decrease in vegetation diversity, thereby destroying the ecological system of plants. Additionally, it indirectly reduces the vegetation coverage, which intensifies the degree of soil erosion in the mining area, and creates a vicious cycle in the future. The contamination of water and soil could lead to an accumulation of heavy metals in vegetation.

Most of the ground fissures or collapse pits in the aeolian sand area are susceptible to rain as well as wind erosion. When the degree of mining subsidence is severe (surface step subsidence), harmful air leakage channels can be formed between overburden fissures and ground fissures, thereby leading to air leakage, sand inrush, water seepage, and other mine geological disasters [32].

Briefly, coal mining in the typical aeolian sand area of northern Shaanxi harms the geological environment of the mine. Due to the harsh natural environment, low vegetation coverage, and less surface runoff in the western aeolian sand area, coal development is almost certain to cause the deterioration of the exposed surface. Ground fissures are regarded as one of the major damage characteristics and indicators of coal mining in the western aeolian sand area.

4.2. Surface Movement and Deformation Characteristics of Non-Pillar Mining in the Aeolian Sand Area

The surface movement during the mining stage of the working face is considered a dynamic process. The change in the mining progression along the trend direction can affect the movement and deformation of the surface point (within the influence range of mining), leading to dynamic changes until it reaches a stationary point and there is no longer any movement. It is imperative to study the dynamic change law of various surface phenomena by elucidating the final shape of the coal mining subsidence basin for analyzing the shortcomings associated with them. The law of subsidence basins formed through non-pillar mining in the trend direction is consistent with the results of the research studies [33]. Due to the absence of the protective coal pillar in the dip direction, the surface points on the edge of the subsidence basin can be affected by secondary mining, resulting in continuous surface movement and deformation, with longer deformation times. Combining the overlying strata stress zoning with the surface point movement trajectory, a schematic diagram of the change characteristics of surface point movement for non-pillar mining is shown in Figure 25.

Figure 25 shows that the edge of the subsidence basin is affected by the mining of the first working face in the trend direction, with points A and E moving outside the working face and points B and D in the subsidence basin moving inside the basin. The possible reason is attributed to the compression effect of the subsidence basin, which produces a gathering effect.

Non-pillar mining is a mining technology that maintains roadways in other ways without retaining coal pillars in the process of coal mining. This technology is mostly suitable for thin and medium-thick coal seams. Through gob-side entry retaining and other methods, such as using the reserved protective coal pillars that can be recovered from traditional mining methods, certain technologies are used to re-support the crossheading of the previous working face to the next working face, which can achieve the purpose of reducing costs, recycling resources and reducing dynamic disasters.

According to Section 2.2, it can be seen that, due to the cancellation of the section coal pillar in the adjacent working face, the stress on the side of the next working face near the solid coal is greater, and the overlying strata collapse more fully. Previous studies have shown that, for non-pillar mining technology, the overlying strata on the adjacent two working faces will form an overall structure. The fissures on the side of the coal pillar on the first working face will gradually be compacted and closed with the mining of the next working face; that is, the surface will form a ring fissure area, indicating that the fissures are more fully developed. There is a ring fissure area in each of the two working faces under traditional mining technology (long-wall mining with coal pillars), which is significantly different from the two working faces under non-pillar mining technology [34].

It can be seen from Figure 7 that the surface subsidence is more uniform under the non-pillar mining technology, and there are fewer fissures between the two working faces. Under traditional mining technology, a non-uniform subsidence area will be formed between the two working faces, and it is easier to form permanent fissures.

4.3. Optimization of Surface Damage Monitoring of Non-Pillar Mining in the Aeolian Sand Area

By measuring and analyzing the surface movement and deformation parameters in the aeolian sand area, certain fluctuations within the range are expected. In the initial stage of surface movement and deformation, the degree of change in the respective parameter is small, with insignificant amplitude. When the working face reaches H₀, the surface enters a stage of drastic change, and simultaneously, the extension of the subsidence basin at the non-coal-pillar retaining roadway is evident. It indicates that, in the early stage of working face mining, the failure of the immediate roof does not occur. When a certain distance is mined, the main roof is subjected to pressure, and the surface enters a slow sinking state. Further, the surface movement and deformation parameters increase gradually. When mining enters the super-full mining stage, the overlying strata experience regular periodic pressure. This causes the surface subsidence to enter a violent period, and the parameter changes become stable within a certain range.

Previous studies have reported that the mining area represents a severely surface-damaged area, with significant deformations. The phreatic water level in the aeolian sand area is low, and the surface does not allow for the accumulation of water. The primary failure mode is ground fissures, which indirectly disrupt the stable environment of the surface vegetation root system and the physicochemical properties of the soil. Currently, the surface damage monitoring method does not consider the differences between regions and the monitoring frequency is very high, which leads to errors in the obtained data. Therefore, the edge of the subsidence basin is divided into tensile stress areas, which belong to permanently damaged areas. The occurrence of a few phenomena such as a large ground fissure width, the extensive damage of vegetation root systems, severe soil erosion, and a decline in soil physicochemical properties is possible. Therefore, this area needs to be monitored emphatically, allocating reasonable monitoring time. For the subsidence basin undergoing repeated cycles of tension and compression, the monitoring area is identified based on the direction of the trend center line, and in situ, dynamic, and non-destructive accurate monitoring is conducted. The monitoring area is divided on the side of gob-side entry retaining in non-pillar mining. The full-cycle law and characteristics of surface damage resulting from secondary disturbance can be derived from the first working face to the end of the second working face (Figure 26).

Based on the precise monitoring of the working face without coal pillars, the characteristics of surface movement and deformation in different regions can be obtained. Based on the limited conditions of vegetation root strength in each region, the degree of land damage due to coal mining activities can be quantitatively analyzed. The manpower and material resources required for mining activities can be reduced to provide a scientifically guided approach for the optimization and modification of underground mining technology.

4.4. Ecological Restoration Model of Non-Pillar Mining in Aeolian Sandy Area

In the 2010s, the scientific mining of coal was formally proposed, and the technical approaches and methods to conduct the coordinated (green) mining of coal resources and environment were put forward. In recent times, with the increasing contradiction between the coal industry and the ecological environment, starting from the concept of harmonious coexistence between man and nature, the connotations and framework of scientific mining have been redefined and enriched. This has laid a theoretical foundation for the development of effective green mining technology for ecologically fragile mining areas in the western region [35]. Zhang et al. [36] proposed the connotations of low-damage mining on the ecological environment, and reported that the damage to the ecological environment should be restricted to the “ecological threshold”. Gu [37] discussed the concept of a coal mine underground reservoir with “guiding, storing, and using” as the core concepts. They provide technical support for the protection and proper utilization of groundwater resources in the western mining areas. Fan et al. [38] divided the geological zoning of water-preserved coal mining in northern Shaanxi. Based on the principle of ecological water level protection, water-preserved coal mining technologies such as underground gangue filling mining, aquifuge reconstruction, and bed separation grouting were developed.

The management of coal mining subsidence land involves the formulation of a scientific, reasonable, and standardized remediation measure that is suitable for the local conditions. It is significant to consider the comprehensive perspective of soil and water conservation as well as ecological restoration in the region, based on the topographical and geomorphological characteristics. Based on the measured data, the ground fissures in the aeolian sand area of northern Shaanxi generally represent the “waviness” development mode. Moreover, it also indicates that the center of the basin exhibits dynamic fissures with a self-healing function, with the boundary fissure of the working face representing the permanent fissure. Therefore, based on the degree of development and formation cycle of ground fissures, the coal mining subsidence basin can be divided into a natural restoration area (NRR) and an artificial restoration area (ARR), as shown in Figure 27.

The natural restoration area is characterized by a relatively uniform subsidence area (RUSR) as well as a completely uniform subsidence area (CUSR) [39], located at the edge of the subsidence basin and the super-full mining area in the subsidence basin. The degree of ground fissure development at the outermost edge area is small, and natural forces can easily bridge it. The dynamic fractures in the subsidence basin perform the function of “self-healing” without additional manual repair.

The artificial restoration area is characterized generally by a non-uniform subsidence area (NSR), located at the edge of the coal mining subsidence basin, the position of the stopping line, and the non-full mining area. These areas exhibit subsidence phenomena such as being prone to step fissures and severe disturbances to the surface soil and vegetation. Artificial guidance should be implemented to promote the repair.

To implement coordinated mining and reduce the proportion of NSRs from the source, the design of working faces should be optimized to increase the length of the trend as well as the tendency of the working face to the maximum possible extent. The promotion of the “mechanized replacement, automatic reduction” project can be more conducive to the construction of large mines in the west. Based on the calculations performed for 20m coal pillars and a working face size of 400 × 3000 m², the surface uniform subsidence area under non-coal-pillar mining was increased by nearly 50% compared to conventional coal pillar mining, and the loss was reduced from the underground source [28]. Based on the technical concept of “mining while recovering” in the underground mines, the overall protection, system repair, and comprehensive management can be implemented from the source using the method of non-pillar mining.

5. Conclusions

(1)

Compared with the Shendong mining area, the surface angular parameters of the study area are less pronounced on the side without coal pillars. This implies that the surface movement and the degree of deformation of the non-coal-pillar mining in the aeolian sand area is significant, with a large mining influence range and a rapid surface subsidence speed. Following the mining of the working face, the resulting failure form of the overlying rock was asymmetric.

(2)

The ground fissures can be divided into two types: boundary and dynamic ground fissures. The boundary ground fissures are mostly located within the boundary of the working face, with no outward expansion. The dynamic ground fissures exhibit the “waviness” type of morphological characteristics and asymmetric “M”-type development characteristics (development cycle of 18 d). The location model and development cycle model of dynamic ground fissures were constructed, indicating that their degree of development is associated with mining parameters. That is, the advanced fissure distance is positively correlated with the mining speed and the mining progression. The development cycle is positively correlated with the mining depth, and is negatively correlated with the mining speed, the advanced influence angle, and the lag angle of the maximum subsidence velocity.

(3)

The mechanical model of the overlying strata structure of non-pillar mining in the aeolian sand area was constructed to evaluate the self-healing characteristics of dynamic ground fissures under the influence of the secondary rotation of hinged rock block structures. Due to the physical characteristics of the aeolian sand layer, as well as the mining driving force, natural wind, rain invasion, and desertification, ground fissures can undergo quick recovery by “cracking–closing”. Moreover, the soil nutrients can be rapidly restored to their original levels, and the surface ecology exhibits a “self-healing” phenomenon.

(4)

The partition monitoring mode of working faces without coal pillars is proposed, which can acquire the characteristics of surface movement and deformation in different regions. Further, based on the degree of development and development cycle of ground fissures, the coal mining subsidence basin can be divided into self-restoration areas and artificial restoration areas. The control of ground fissures as well as a reduction in the overall cost can be achieved through the combination of natural guidance and artificial guidance.

This research conclusion is more suitable for the mining of shallow coal seams without coal pillars in the aeolian sand area. The mining method is roof cutting and pressure relief without coal pillar retreat mining. The mining depth is less than 200 m. The landform is aeolian sand. Unfortunately, in this paper, the traditional measurement method (using GPS-RTK to determine its location, and using tape to measure its width and height difference) is used to monitor ground fissures, which wastes a lot of manpower. At present, remote sensing technology is developing rapidly. How to propose new image recognition methods to efficiently monitor ground fissures is work for future research.