In-situ stress field inversion and its impact on mining-induced seismicity

Abstract

The magnitude and direction of the in-situ stress field significantly influence the Geodynamic disaster. Taking the Gaojiapu Coal Mine in the northwest Binchang mining area as the study area, the distribution law of the in-situ stress field and its relationship with microseismic events were analyzed and discussed through field geological investigation, in-situ measurement, and numerical simulation. The results show that: (1) The Maximum horizontal principal stress (MaxHPS) is mainly concentrated in the turning end of the fold, the axis of the syncline, the undulation end of the anticline, the footwall of the fault, and the two ends of the fault. (2) The MaxHPS in the footwall area of the fault is greater than that in the hanging wall area. (3) The MaxHPS at the axial part of the syncline is greater than that at the wing part, and the MaxHPS at the wing part is greater than that at the anticline. (4) In areas with high MaxHPS and Lateral pressure coefficient (LPC), the frequency of high-energy microseismic events is high. (5) The 10⁵J high-energy microseismic events more easily occur in areas with MaxHPS greater than 38 MPa and LPC greater than 1.50.

HIGHLIGHTS

• Based on the stereographic projection analysis of the outcrop trace of the field rock stratum, combined with the in-situ stress test and structural evolution analysis, the regional in-situ stress field’s evolutionary history and distribution characteristics are analyzed.

• The three-dimensional geometric simulation model is established by the numerical analysis method, the three-dimensional field stress field is inversed by multiple linear regression method, and the inversion results are tested and verified.

• According to the mine seismic monitoring results and inversion results of the in-situ stress field, the relationship between the in-situ stress field and mining-induced seismicity was analyzed and discussed.

Keywords:

• In-situ stress field
• Rockburst
• field geological investigate
• inversion analysis
• microseismic events

Introduction

In-situ stress affects engineering activities such as coal mining and rock tunnelling and is the fundamental force leading to the deformation and failure of underground engineering construction. The direction and magnitude of the in-situ stress field also have a significant influence on rock bursts (Hoek and Marinos 2009). In the high tectonic stress area, the rock mass would be deformed, destroyed, and broken during the excavation, and microseismic events and rock bursts usually occur in some areas (Gong et al. 2010; Xia and Ren 2013). Rock burst is a typical dynamic disaster that occurs in coal mines; it is characterized by abrupt, explosive, and urgent. It often causes heavy casualties and serious damage to the shaft and roadway (Dou et al. 2020). In the past decade, according to incomplete statistics, rock bursts have occurred in more than 100 mines in China, causing more than 300 deaths and nearly 1,000 injuries (Yuan et al. 2018). Therefore, it is essential to study the state of regional stress and analyze the stability of the surrounding rock to achieve safe excavation and underground engineering works.

Many studies have shown that in-situ stress measurement is a direct and effective method for studying the present in-situ stress field ( Zhao et al. 2013). Many scholars studied the distribution characteristics of regional stress through the field measurement of in situ stress (Kang et al. 2010; Zhang et al. 2020). However, the measurement of in-situ stress is usually expensive and time-consuming, and the causes of the in-situ stress field are complex; there are many influencing factors, and the measuring points are relatively scattered, which makes the measured results of the in-situ stress field prone to errors. To better analyze the distribution of the in-situ stress field, numerical modelling works have been widely carried out to understand the relation between regional stress regimes and scattered measuring points (Stephansson and Zang 2012; Ning et al. 2021). There are many methods for inverting the initial stress field, including multiple linear regression analysis, support vector machine, and genetic algorithm optimization (Li et al. 2012; Samui et al. 2015). Of course, The approximation degree of the value of inversion stress and the actual stress is an important criterion for validating inversed stress field results (Zhao et al. 2012).

The Binchang mining area is located in the Ordos Basin, which is one of the main mining areas of 13 coalfield bases in China-HuangLong Coal Field Base (Xu et al. 2012). Rock burst disasters have occurred frequently in recent years, along with the deep and large-scale development of coal resources. Since 2013, the Hujiahe coal mine, Tingnan coal mine, Gaojiapu coal mine, and Mengcun coal mines in the Binchang mine area have experienced different degrees of rock burst (Liu et al. 2017). Since the beginning of mining in July 2014, a coal burst has occurred many times in the Gaojiapu coal mine, and the situation of the coal burst disaster is very serious. This study takes the Gaojiapu coal mine in the Binchang mining area as the study area. The casing hole stress relief method was used to measure the in-situ stress field in the study area. Combined with the field measurement of stereo projection and the regional tectonic evolution analysis, the current in-situ stress field characteristics in the study area were determined. Considering the engineering geological conditions, the three-dimensional stress field in the study area was inversed, and the distribution law of the initial geo-stress field was analyzed. The relationship between the in-situ stress field and microseismic events was further investigated based on the results of microseismic monitoring in the mine area. The research results could serve as a reference for the design of good deep excavations and the evaluation of mine stability and provide support, which could also help predict and evaluate mine burst risks.

Study area

Engineering geological conditions and mining situation

Gaojiapu coal mine is located in the Binchang mining area, in the beam and gully region of the loess plateau of northern Shaanxi and eastern Gansu. According to the appraisal report on coal seam, roof, and floor impact tendency submitted by China Coal Research Institute, the uniaxial compressive strength of No.4 Coal Seam in Gaojiapu Coal Mine is 20.47 MPa, the dynamic failure time is 278.40 ms, the elastic energy index is 13.36, and the impact energy index is 3.20, respectively, and is identified as a coal seam with strong impact propensity; The bending energy index of the top and bottom strata of Coal Seam 4 is 54.52 kJ and 20.08 kJ respectively, which are identified as weak impact prone strata. The mine is seriously threatened by coal bursts during production. The geological structure of the study area is simple, with most of the strata being covered by the Quaternary loess layer, and the occurrence of the coal seam is relatively stable.

According to the planning and design data of the mine, the mining area is divided into four panels. Among them, the first and second panels have been mined. The 302 coalface of the third panel is presently being mined, the fourth panel is undergoing roadway excavation work, and the next replacement working face is the 301 coalface. There are 12 main faults and eight-fold structures in the mining area. Among them, nine faults with a throw of more than 10 m. The distributions of the main geological structures and the mining area are shown in Figure 1a. The strata in the study area from old to new are Jurassic Yanan Formation (J₂y), Zhiluo Formation (J₂z), Anding Formation (J₂a), Lower Cretaceous Yijun Formation (K₁y), Luohe Formation (K₁l), Huachi Formation (K₁h), and Quaternary strata. The research area’s engineering geological profile is shown in Figure 1b.

Figure 1. Engineering layout and geological profile in the study area.

Microseismic monitoring system

An ‘SOS’ microseismic monitoring system has been installed in the first and second blocks of the mining area, which can be used to monitor the seismic activity in real-time and determine the source position and energy of the seismic (Zhou et al. 2020). From May 2019 to January 2021, 542 high-energy microseismic events with the energy larger than 10⁵J occurred in the mining area (Figure 1c). Among them, there were 20 high-energy microseismic events above 10⁶J. As shown in Figure 1c, most microseismic events occurred in the main roadway area of the first plate area, the middle part of the 204 and 205 coalfaces, and the 302 coalface.

Methodology

Measurement of in-situ stress

There are many methods to measure the in-situ stress in a rock mass (Sakurai and Akutagawa 1994; Ljunggren et al. 2003; Zang and Stephansson 2006), such as the hydraulic fracturing method, the stress relief method, and the acoustic emission method. Among them, the stress relief method has been more widely used, and the technology is more mature. When a rock is taken out of a rock mass under force, the elastic deformation accumulated in the rock will cause it to expand and recover. The YHY16 monitoring substation (Figure 2a) and the KX2011 hollow core inclusion strain gauge (Figure 2b) can be used to measure the expansion deformation of the core before and after the stress is relieved, and the elastic modulus and Poisson’s ratio of the rock can be measured with the calibrator for transducer calibrating confining pressure. The main body of the KX2011 hollow core is a hollow cylinder made of epoxy resin. Three groups of strain flowers are embedded at an equal distance along the same circumference in the middle layer of the cylinder. Each strain group consists of four strain plates, with an interval of 45°, as shown in Figure 2c.

Figure 2. In-situ stress testing equipment and Structural diagram of sensors. (a) YHY16 monitoring substation, (b) KX2011 hollow core inclusion strain gauge, (c) Hollow inclusion strain cell (Modified from Liu et al. 2021).

When the core material was released from a drill, 12 groups of deformation sensor readings changed, as shown in Figure 3. Based on the elastic stress distribution of surrounding rocks in the circular cave, the least square method can be used to inversely calculate the magnitude and direction of the stress in the rock mass before the stress in the surrounding rock. The stress relief method was used to complete the testing of the eight measurement points (Figure 2a). These results are shown in Table 1.

Figure 3. Stress relief curve of the T1 test point.

Table 1. The measuring results in the study area.

Test points	Vertical depth (m)	σ1 (MPa)	σ2 (MPa)	σ3 (MPa)	Azimuth of σ1 (°)
T1	799.8	38.23	21.46	14.44	236.13
T2	888.3	41.98	22.23	20.60	235.34
T3	1012.5	44.04	24.29	20.79	240.82
T4	1077.6	45.16	24.29	19.47	217.44
T5	1065.7	45.15	23.92	14.47	325.83
T6	1080.4	44.59	24.66	14.66	220.40
T7	947	42.29	23.66	20.10	308.70
T8	914	42.03	25.06	18.25	282.46

Multiple linear regression

Currently, the multiple regression numerical calculation method of stress reverse analysis is widely used in engineering; it is a multiple linear regression method combining numerical calculation and multiple regression (Qiu et al. 2003). This method considers the gravity and geological structure effects of the initial stress field, and its reliability has been demonstrated in many engineering studies. Assuming that all influencing factors are linearly superimposed, the initial in-situ stress can be expressed by the following formula (Korkmaz 2021; Meng et al. 2021): $${\sigma }_0=\sum _{i=1}^n{w}_i{\sigma }_i+\epsilon +C$$ where σ₀ is the stress regression value, σ_i is the regression of stress under condition i, w_i is the regression coefficient, ε is the model error, C is the constant value, and n is the total number of basic boundaries.

To measure the significance of regression, the analysis is carried out through the test of the complex correlation coefficients R and F, and the correlation calculation formula is as follows (Wang and Li 2015; Deng 2018): $$S_r={\sum _{k=1}^m\sum _{j=1}^6\left({\sigma }_{jk}^{*}-{\overline{\sigma }}_{jk}\right)}^2$$ $$\overline{\sigma }=\frac{1}{6m}\sum _{k=1}^m\sum _{j=1}^6{\sigma }_{jk}^{*}$$ $$S_c={\sum _{k=1}^m\sum _{j=1}^6\left({\sigma }_{jk}^{*}-\sum _{i=1}^n{L}_i{\sigma }_{jk}^i\right)}^2$$ $$R=\sqrt{1-S_C/S_r}$$ $$F=\frac{\left(S_r-S_c\right)/n}{S_c/\left(mn-m-1\right)}\sim F\left(\mathit{\text{n,mn}}-m-1\right)$$ $$F>F_{1-a}\left(\mathit{\text{n,mn}}-n-1\right)$$

The complex correlation coefficient R is between 0 and 1, and the greater the magnitude, the better the regression effect. When the statistic value of F satisfies a certain confidence level a, the regression effect of the equation is remarkable.

In the inversion of the geo-stress field using the multiple linear regression method, the significance of the regression model does not mean that each independent variable has a significant impact on the dependent variable. Therefore, the t-test is used to test the significance of each independent variable. The correlation calculation formula is as follows (Meng et al. 2021): $$\left|T_i\right|=\left|\frac{w_i\sqrt{mn-1}}{\sqrt{w_{ii}{S}_r}}\right|\sim t_{1-a/2}\left(mn-n\right)$$ $$\left|T_i\right|>t_{1-a/2}\left(mn-n\right)$$ where w_ii is the inverse matrix element of the coefficient matrix of the normal equation.

Determination of the boundary conditions of the in-situ stress inversion

The initial in-situ stress field is the stress state in the original rock mass itself and is generated by gravity and various tectonic stresses. It is also affected by external effects such as weathering and denudation. Energy is constantly released, and the stress is continuously redistributed, finally forming the present residual stress state. There are many causes of the initial in-situ stress field of a rock mass, such as its own characteristics, geological structure, temperature, and groundwater (Zhang et al. 2016). Actual engineering experience shows that the gravity of the rock mass and geological tectonics are the main factors affecting the initial in-situ stress in the rock mass (Hoek 2007). Therefore, in the inversion process, the following factors and calculation methods are mainly selected to invert the in-situ stress field (Hu et al. 2005; Su et al. 2011):

1. Gravity-weight stress: many actual engineering in-situ stress regression studies show that gravity stress is one of the = jm\][polkhgbv main causes of the formation of the in-situ stress field in a rock mass. Therefore, in the simulation of deadweight action, rigid displacement constraints are applied to all boundary surfaces other than the upper ground surface, and deadweight stresses are applied in the entire range of the calculation model (Figure 4a).

2. The horizontal uniform compression tectonic movement and the plane shear tectonic movement are important sources of tectonic stress and a key embodiment of plate movement. Therefore, five kinds of calculation methods are selected in the horizontal plane for simulation.

   • Horizontal compression tectonic movement in the X-axis direction of the horizontal direction: we select the boundary condition of applying displacement parallel to the X-axis direction. The displacement boundary conditions are shown in Figure 4b.

   • Horizontal compression tectonic movement in the Y-axis direction: the boundary condition of the displacement applied parallel to the Y-axis direction was selected. The displacement boundary conditions are shown in Figure 4c.

   • YZ plane shear tectonic movement (Figure 4d): we apply pairwise relative tangential displacement in the Z direction; a rigid displacement constraint was applied on the two boundary surfaces perpendicular to the X axis.

   • XZ plane shear tectonic movement (Figure 4e): we apply pairwise relative tangential displacement in the Z direction; Rigid displacement constraints were applied on two boundary surfaces perpendicular to the Y axis.

   • XY plane shear tectonic movement (Figure 4f): we apply pairwise relative tangential displacement in the Y direction; a rigid displacement constraint was applied on the two boundary surfaces perpendicular to the Z axis.

Figure 4. Schematic diagram of the displacement boundary conditions of the tectonic movement.

Analysis and significance test of the in-situ stress field inversion

Finite element software has been favoured by increasingly more researchers due to its powerful simulation and analysis capability, and is used to perform stress field inversion during the geotechnical engineering activity (Li et al. 2020). Based on the morphological characteristics of the main geological structures and the geological data of the study area, a three-dimensional geological model was established using the ANSYS finite element software. It included a total of 56995 nodes and 213272 units. Assign different rock mechanics parameters (Table 2) to different elements in the geometric model according to the indoor physical and mechanical index experiments.

Table 2. Physical and mechanical parameters of the rock mass.

Formation	Density (g/cm^3)	Internal friction angle (°)	Cohesion (MPa)	Elastic modulus (×10^4MPa)	Poisson’s ratio
Huachi	1.30–2.71(2.47)	33.34–42.30(39.6)	1.38–5.47(2.93)	0.10–4.65(1.66)	0.16–0.60(0.24)
Luohe	2.32–2.95(2.45)	32.42–42.56(38.9)	0.49–4.19(2.01)	0.02–4.10(1.18)	0.17–0.39(0.23)
Anding	2.34–2.63(2.55)	32.75–42.13(39.5)	1.02–4.60(2.91)	0.20–4.19(1.83)	0.18–0.21(0.19)
Zhiluo	2.44–2.54(2.49)	33.12–40.96(35.4)	1.41–4.34(2.77)	0.13–2.68(1.09)	0.19–0.30(0.23)
Yanan	2.34–2.67(2.53)	30.42–42.96(36.3)	1.12–3.95(2.80)	0.15–3.65(1.73)	0.14–0.31(0.21)
Coal seam	1.27–1.62(1.35)	35.12–38.24(36.72)	0.35–1.95(1.19)	0.32–0.95(0.51)	0.18–0.38(0.26)

The stress field under each boundary condition was calculated according to the selected boundary conditions and the ANSYS 3D numerical model. The measured results of in-situ stress at the measured point are the results of the joint action of these boundary conditions. Hence, the above six boundary conditions are superimposed. The complete initial in-situ stress field can be obtained by multiple linear regression between the simulated data and the measured data. The simulated data and the measured data were imported into the software for Multiple linear regression calculation, and the regression coefficients of each boundary condition were obtained.

Two methods, the R test and the F test, were used to test the correlation between the independent and dependent variables. The significance of several influencing factors in the equation was obtained through regression. The correlation coefficients R = 0.934 and F = 46.97 were obtained by statistical calculation. The closer R is to 1, the stronger the correlation between the independent and dependent variables under the six calculation methods. In the F test, when the significance level is set as a = 0.05, the critical value F_a=(n,mn-n-1)=F(6,41)=2.33, F_a<F = 46.97 can be obtained from the distribution table of the F test, indicating that the independent variables selected in this article are significant as a whole. In addition, we used the t-test to test the significance of each independent variable further. The independent variable regression coefficient and t-test results are shown in Table 3.

Table 3. The independent variable regression coefficient and t-test results.

Independent variable	Ug	Ux	Uy	Uxz	Uyz	Uxy
Regression coefficient	0.9	−2.58	5.98	−1.33	2.57	3.36
T-value	7.14	−5.15	7.04	−2.62	4.76	4.98

At the level of significance α = 0.05, t_1-a/2 (mn-n)=t_0.975(42)=2.019. Therefore, according to the comparison of test results in Table 3, at the level of significance α = 0.05, $\left|T\right|$>t_0.975(42)=2.019, meeting the significance test of the regression coefficient. Hence, the regression equation of the in-situ stress field of the mining area was established as follows, $$\stackrel{⌢}{\sigma }=0.90U_g-2.58U_x+5.98U_y-1.33U_{xz}+2.57U_{yz}+3.36U_{xy}+5.75$$ where $U_g,$ $U_x,$ $U_y,$ $U_{xz},$ $U_{yz}$ and $U_{xy}$ represent the calculated results of the above six boundary conditions.

To further verify and analyze the accuracy and reliability of the inversion model, we compare the test data and regression data of Test points 1–3 of the above eight measured points. The comparison results are shown in Figure 5.

Figure 5. Comparison of in-situ stress test data and regression data.

It can be seen from Figure 5 that there are specific differences between the value of inversion stress and actual stress. Among them, the regression result of Test point 1 is greater than the test result, while the regression result of measuring point 3 is less than the test result. Overall, the regression results are very close to the test results, the error is basically within 10%, and the error between the regression data and the test data is within an acceptable range. Combined with the results of the above R test and F test, the in-situ stress at each node in the mining area can be calculated by regression formula. Therefore, the in-situ stress values of the 3 D model were calculated using the regression equation. The three-dimensional in-situ stress field distribution in the study area was finally obtained.

Results

Investigation and analysis of field structural features

The constructor is a rupture structure surface generated under the action of a constructive field effect or only a slight displacement. It can be used to restore the structural stress field. From the perspective of geomechanics, Sun et al. (2006) used the measurement of conjugate shear joints and fault plane occurrence, scratch angle, and displacement sign together with stereographic projection to determine the direction of the principal stress axis. Sun et al. (2008) conducted on-site in-situ stress testing using the casing hole stress relief method and analyzed the in situ stress field in combination with geomechanics and stereographic projection. The measured original rock stress in such studies agrees with the geomechanical analysis results. Lin and Xie (2018) analyzed the fracture surface scratch and other tectonic traces using the stereographic projection method, judged the stages of the tectonic stress environment in the study area, and analyzed the essential characteristics of the tectonic stress environment in the area.

Through field geological surveys, we found that there are mainly two groups of joints on the surface of the study area and the surrounding areas (Figure 6a). These two sets of joints intersect at an inclination angle of approximately 90°, and there is no filling; they also often cut into each other as conjugate joints. Among them, the strikes of the joint planes are Eastern Northeast (NEE) (65°±10°N) and Northern Northwest (NNW) (335°±10°N). In addition, the joint surface in the NEE direction is more densely developed (Figure 6b). Based on the collected data on the occurrence of this group of conjugate joints in the study area, the stereographic projection method is used to judge the direction of the stress axis in the structural joints in the study area. The results are shown in Figure 6c.

Figure 6. Analysis of field geological survey in the research area. a ‘X’ conjugated heritage of the research area, b Fissures in Cretaceous strata outcrops, c the Stereographic projection method, d the maximum horizontal principal stress (MaxHPS) direction at the measurement point.

As shown in Figure 6c, the research area is mainly subjected to a tectonic stress field in the direction of Northern Northeast-Southern Southwest (NNE-SSW) and Western Northwest-Eastern Southeast (NWW-SEE). Among them, the tectonic stress with the direction of Northern Northeast-Southern Southwest (NNE-SSW) is stronger. Judging from the in-situ stress measured results (Figure 6d), the MaxHPS directions of the measured in-situ stress points in the study area are Northern-Northeast and Northeast, which are consistent with the results of field investigation and analysis.

Gaojiapu coal mine is located on the southwest edge of the Northern Shaanxi Slope structural unit in Ordos Basin, with the Tianhuan depression in the West and Weibei uplift in the south. In recent years, there have been many studies on the regional tectonic evolution of the Ordos Basin (Ding et al. 2015). Based on the regional geological tectonic evolution of the Ordos Basin (Zhao et al. 2017; Li et al. 2018), the current tectonic stress field in the study area is mainly affected by the Yanshan movement and the Himalayan movement. Since the collision of the Indian plate with the Eurasian plate and intracontinental subduction after the collision period, and the westward movement of the Pacific plate, orogenic movement continued to make the study area compressed, forming a tectonic stress field with NNE-SSW as the main compression direction. The field geological survey results and in-situ stress test results in the study area are consistent with the regional stress field analysis results of the above tectonic movement evolution analysis. Among them, the trend of cretaceous outcrop fractures is also consistent with the direction of structural stress field in the Himalayan period (Figure 6b).

Distribution characteristics of the MaxHPS and horizontal difference stress (HDS)

The initial in-situ stress field in the fold and the fault structure area is unevenly distributed. The MaxHPS near the fold is approximately symmetrically distributed along the fold crank axis trace. The MaxHPS near the syncline axis and the undulating end of the anticline are relatively high, showing stress concentration. The magnitude of the MaxHPS gradually decreases from the X3 syncline axis to the B4 anticline (Figure 7b). That is, the MaxHPS in the syncline axis is greater than that in the syncline wing, and the magnitude of the MaxHPS in the syncline wing is greater than that in the back inclined shaft. This result is similar to the results of Kang et al. (2012), Han et al. (2013), and Zhao et al. (2019). It is worth noting that the stress concentration state of the MaxHPS has an obvious relationship with the angle of the fold structure fluctuation. The greater the wrinkle fluctuation, the more obvious the stress concentration and the greater the MaxHPS, while the side wall of the roadway will also have a certain degree of deformation (Figure 7c).

Figure 7. The distribution of MaxHPS in the coal seam.

The distribution of the stress field near the fault structure is extremely complicated (Huang et al. 1998; Su et al. 2003). It can be seen from the inversion results of the in-situ stress field that there is a significant difference in the MaxHPS value between the fault structure area and the footwall on both sides of the fault. Except for the F2, F3, and F7 faults, the MaxHPS values at the footwall of the other faults are higher than that at the fault hanging wall, and the MaxHPS concentration exists in the footwall of the fault. This result is similar to the results of Xie et al. (2019). This is because the stress on the hanging wall is a local stress that has been disturbed by the fault, while the footwall is less disturbed and should be consistent with the regional stress distribution. Among them, the F3 fault is caused by the influence of the X3 syncline, which makes the MaxHPS in the footwall lower than that in the hanging wall. However, the F2 and F7 may be affected by their own existence, resulting in a MaxHPS distribution that is not obvious. In addition, there is a stress concentration area at the end of the fault, which is consistent with the research results of Shen et al. (2007) and Weng et al. (2020). Combined with the effect of the F7 fault, the trace of the fault changes, and the MaxHPS also changes. In addition, the MaxHPS may vary when superposed by the action of faults (Shen et al. 2008).

The failure of rock mass is usually caused by shear failure. Therefore, the maximum shear stress reflects the stability of deep surrounding rock to a certain extent (Seredin et al. 2013; Cheng and Wang 2018). Based on the Rock mechanics theory, the HDS is the difference between the MaxHPS and the minimum horizontal principal stress (MinHPS), and the maximum shear stress is one-half of the HDS. The HDS not only reflects the difference between the MaxHPS and the MinHPS, but also reflects the characteristics of the maximum shear stress.

As shown in Figure 8, in the fold structural area, the magnitude of the HDS is relatively low, and the local HDS in the axis of the X3 syncline and B2 anticline is larger than that in the two wings, indicating that the shear stress in the crankshaft of the fold is higher. The HDS is obviously higher in the fault structural area, especially in the middle of the two walls of the fault and at the end of the fault. From the middle of the fault trace to the end of the fault, the difference in the HDS values increases from gentle to sharp, indicating that there is a HDS concentration at the two ends of the fault, and the shear stress at the end of the fault is high.

Figure 8. The distribution of HDS in the coal seam.

Discussions

Analysis of the relationship between the mining-induced seismicity and the in-situ stress field

The seismicity is affected by many factors, such as in-situ and excavation induced stresses, etc (Naji et al. 2019). To further analyze the relationship between the mining-induced seismicity and the in-situ stress field, the plane distribution of microseismic events is analyzed and compared from three aspects: MaxHPS, HDS, and the lateral pressure coefficient (LPC). It is one of the control factors affecting the deformation and failure of underground surrounding rock (Gao et al. 2020). Most of the microseismic events are distributed in areas with high MaxHPS. This is especially near the 204 coalface and the 205 coalface, where there are more occurrences of microseismic events; the MaxHPS in this area is also relatively high.

It can be seen from the statistical data of seismic events and the MaxHPS in the study area (Figure 9) that in areas with lower MaxHPS, the number of microseismic events is relatively small. In areas where the MaxHPS is greater than 35.5 MPa, the number of microseismic events is 464, accounting for 85.61% of the total events. When the MaxHPS is greater than 39.5 MPa, there are fewer microseismic events. This is because most of the areas where the MaxHPS is greater than 39.5 MPa belong to the area to be mined, so there are obviously fewer microseismic events. As the MaxHPS increases, the number of microseismic events increases significantly. This indicates that it is easy for microseismic events to occur in areas with a large MaxHPS. With this, more microseismic events will occur in the future in the axis of the X3 syncline and the other regions where the MaxHPS is concentrated.

Figure 9. Statistics of microseismic events and the MaxHPS in the study area.

Looking at the distribution of the LPC and microseismic events in the study area, it can be seen that most high-energy microseismic events are distributed in areas with relatively high LPC (Figure 10). According to the statistics, there were 432 microseismic events in mining areas with a LPC greater than 1.48, accounting for 79.7% of the total number of microseismic events. Among them, 12 high-energy microseismic events of 10⁶J have a LPC greater than 1.53, accounting for 60% of the total number of high-energy microseismic events of 10⁶J.

Figure 10. Statistics of microseismic events and the LPC in the study area.

It can be seen from Figure 10 that when the LPC ranges from 1.48 to 1.68, a total of 385 microseismic events occurred. This is 3.5 times the number of microseismic events when the LPC ranges from 1.28 to 1.48. When the LPC is low, the number of microseismic events is significantly lower. With the increase in the LPC, the microseismic events increased significantly, showing an obvious proportional relationship. Generally speaking, microseismic events occur more easily in areas with a higher LPC.

Analysis of profile characteristics at 204 coalface and 205 coalface

Due to a large amount of microseismic data in 204 coalface and 205 coalface, the relationship between the initial in-situ stress field and microseismic events has been analyzed and discussed in the above chapters. There is an X2 syncline with NE-SW strike near the open-off cut of coalfaces 204 and 205, and there is an X3 syncline with NE strike in the middle of the coalface. Among them, the X2 syncline is almost consistent with the strike of the coalface, and the X3 syncline is almost perpendicular to the strike of the coalface. A profile line is arranged along the coal pillar in the middle of 204 coalface and 205 coalface. The comprehensive characteristic of the profile at 204 coalface and 205 coalface is shown in Figure 11.

Figure 11. Comprehensive characteristic of profile at 204 coalface and 205 coalface.

It can be seen from Figure 11 that the mining method of 204 coalface and 205 coalface is prone to mining first and then overhand mining. The coal seam thickness first increases and then decreases, which is thicker in the middle of the coalface and thinner on both sides. Within 250 m from the cutting hole, due to the low MaxHPS, small LPC, and small mining range, microseismic events are mainly within a certain range of coal seam roof and floor, with few times and low height.

Due to the influence of the X2 syncline, the MaxHPS and LPC first decrease and then increase. Within 250 m-750m from the cut hole, the MaxHPS and LPC increase gradually because the distance from the X3 syncline axis is gradually closer. Moreover, due to the increase in mining range, the number of microseismic events increases, and the height of microseismic events increases. Due to the influence of the X2 syncline, three times 10⁶J high-energy microseismic events occurred near the end of the X2 syncline trace. Among them, within the range of about 750 m from the cutting hole, the MaxHPS is more than 38 MPa, and the LPC is more than 1.50. Many microseismic events have occurred within this range, the height of microseismic events is the largest, and 10⁵J high-energy microseismic events occur frequently. After passing through the X3 syncline axis, the MaxHPS and LPC gradually decrease with the distance from the X3 syncline axis. With the increase of MaxHPS and LPC, the frequency of microseismic events gradually increases. The frequency of high-energy microseismic events will be higher in the areas with high MaxHPS and high LPC.

Conclusions

Through the field investigation, in-situ stress test, and numerical simulation, an accurate in-situ stress inversion model was established in this study, and the characteristics of the current geo-stress field and its impact on mining-induced seismicity were analyzed. The main conclusions are as follows:

1. Based on the field geological survey and analysis of the stereographic projection method, the study area is mainly affected by the tectonic stress field in NNE-SSW direction; According to the analysis of tectonic evolution, due to the collision of the Indian plate with the Eurasian plate and intracontinental subduction after the collision period, and the westward movement of the Pacific plate, orogenic movement continued to make the study area compressed, forming a tectonic stress field with NNE-SSW as the main compression direction. The MaxHPS directions of the measured in-situ stress points in the study area are Northern-Northeast and Northeast, which is consistent with the results of field investigation and tectonic evolution analysis.

2. The numerical analysis method establishes the three-dimensional geometric simulation model. Combined with engineering geological conditions, with the aid of 3D finite element numerical simulation and multi-linear regression methods, the initial stress field in the study area was inverted, and the R, F-tests, and t-tests verified the accuracy of the inversion results.

3. The initial in-situ stress field in the fold and the fault structure area is unevenly distributed. The MaxHPS is mainly concentrated in the turning end of the fold, the axis of the syncline, the undulation end of the anticline, the footwall of the fault, and the two ends of the fault. Among them, the MaxHPS at the axial part of the syncline is greater than that at the wing part, and the MaxHPS at the wing part is greater than that at the anticline. The MaxHPS in the footwall area of the fault is greater than that in the hanging wall area. The variation of the HDS is not obvious in the fold structure area, and there is a certain increment of differential stress locally.

4. Based on a comprehensive analysis, the frequency of microseismic events gradually increases with the increase of MaxHPS and LPC. The frequency of high-energy microseismic events will be higher in the areas with high MaxHPS and high LPC. When MaxHPS is greater than 38MPa and LPC is greater than 1.50, 10⁵J high-energy microseismic events occur frequently.

Authors’ contributions

Xianggang Cheng performed implementation, organized the field test, and wrote the manuscript. Wei Qiao collected a part of the data and participated in drafting the manuscript. Linming Dou, Hu He, and Wei Ju participated in drafting the manuscript. Jinkui Zhang, Shikang Song, Heng Cui, and Hengzhi Fang collected and analyzed the data.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

Data availability statement

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Additional information

Funding

The research was supported by the Fundamental Research Funds of the National Natural Science of China (Gran 41772302) and the Priority Academic Program Development of Jiangsu Higher Education Institutions. This work was also supported by the Future Scientists Program of China University of Mining and Technology (Grant 2021WLKXJ007), the National Scholarship Fund (Grant 202206420039), and the Postgraduate Research & Practice Innovation Program of Jiangsu Province (Grant KYCX21_2323).