Study on the failure characteristics of overburden and the evolution law of seepage field in deep buried thick coal seam under aquifers

Water inrush at roof area seriously affects the safety of coal mines. The characteristics of aquifer and aquiclude at Wutongzhuang Mine are analyzed. Considering the effect of seepage field, a formula for calculating the height of water-conducting fractured zone (HWCFZ) in deep buried thick coal seam mining is derived. A damage-seepage coupling model with rock porosity and damage factor as independent variables is established. FLAC3D is re-developed by using FISH language, and the fluid–solid coupling calculation model of deep buried thick coal seam mining is established. The evolution law of the plastic zone, seepage field and water-conducting fractured zone (WCFZ)of the overburden in the gob with the advancement of the working face is analyzed, the main conclusions are as follows: With the continuous advancement of the working face, the distribution shape of the plastic zone and seepage field has changed from a trapezoidal to a saddle shape; when the working face reaches full mining, the maximum heights of the caving zone, fractured zone and HWCFZ are 24 m, 113 m, and 123 m, respectively; The 50 m-thick sandy shale aquifer is penetrated by the WCFZ, and the WCFZ on the side of the working face above the gob is the main water channel when the working face is advanced to 220 m. The on-site monitoring results showed that mine water inflow is not affected by surface rainfall and the 50 m-thick sandy shale is successively connected by the WCFZ. The results of comprehensive research showed that the HWCFZ cannot be calculated by traditional formulas when mining deep buried thick coal seams.


Introduction
Water inrush disaster is one of the five major disasters in mines, and the occurrence of water inrush accidents seriously threatens the safety of mining production (Keegan-Treloar et al. 2021;Bai et al. 2021;Wang et al. 2022a, b) The main reason for water inrush and sand inrush in the working face is that the aquiclude performance of the rock formation between the aquifer and the gob is damaged (Wang et al. 2021c;Xue et al. 2022;Li et al. 2023). Accurate judgment and effective monitoring of the performance change of the aquiclude is an important guarantee to ensure the safety of coal mining working face under aquifers (Bharti et al. 2021;Zuo et al. 2022).
Scholars have carried out many studies on the mining of coal seams under aquifers, using numerical simulation, theoretical analysis, laboratory tests, on-site measurements and other methods (Schulze et al. 2001;Majdi et al. 2012;Rezaei et al. 2015;Guo et al. 2018a, b;Yin et al. 2019Yin et al. , 2023Cao et al. 2022). Accurately analyzing and understanding the various influencing factors of the development of the WCFZ plays an important role in the establishment of computational and predictive models. Influencing factors of water inrush at working face include: vertical seepage, surface water outcrop recharge, coal seam thickness, water permeability of the aquiclude, mining depth, water head pressure, working face parameters, etc. Zheng et al. 2020;Xu et al. 2022). When the deep buried thick coal seam is mined under the aquifer, the reason for the water inrush at working face is the sharp drop of the water head in the aquifer caused by the sudden fracture of the upper rock formation .
For the calculation and analysis of the HWCFZ, scholars have obtained a large number of results with reference significance. During the mining process of the 207 working face of Huangling No. 2 Coal Mine, the HWCFZ is the largest at the open cut and stop line, and the ratio to the thickness of the coal seam is 21.7 (Yan 2021). When slicing and top coal caving mining are used to mine thick coal seams under rivers, the HWCFZ was 15.8 and 21 times of mining height, respectively (Du et al. 2017). The calculation formula for the HWCFZ in thick coal seam using large cutting height mining is proposed (Liu 2016). The HWCFZ in the deep longwall working face was calculated by analyzing the change of overburden permeability before and after coal seam mining (Mills et al. 2016). Numerical simulation, theoretical analysis and on-site measurement are used to analyze the HWCFZ in layered overlying rock of longwall coal mining (He et al. 2020(He et al. , 2021Wang et al. 2021d;Wang et al. 2022a, b). The mechanism of water inrush and the development of water-conducting fissures in multiple aquifers have been comprehensively studied (Yin et al. 2019). The HWCFZ was measured using the drilling flushing fluid consumption method and the transient electromagnetic method, and pointed out that the traditional empirical formula cannot be used to calculate the HWCFZ when the mining height is greater than 3 m (Chang et al. 2018;Li et al. 2020).
The law of water seepage in the fractured rock mass is the key basic theory to study the safe mining of coal seams under aquifers. The permeability of the aquifer has a negative exponential relationship with the burial depth, and different aquifers have different seepage laws of vertical heterogeneity (Yu et al. 2018;Guo et al. 2019). In order to solve the problem that the existing prediction models cannot accurately predict the risk of water inrush in deep buried coal seam mining, based on hydrogeological analysis, scholars have established a fault anti-seepage evaluation model with higher prediction accuracy (Cao et al. 2019), and the T-q method of water inrush risk combined with water inrush coefficient and water inflow (He et al. 2021).
The numerical simulation software FLAC3D was used for dynamic simulation of deep buried thick coal seam mining, the development characteristics of the WCFZ are studied (Liu et al. 2020. FLAC3D is used to analyze the coupling mechanism of filling body, the change of pore water pressure and the development of WCFZ during backfill mining (Sun et al. 2021). The seepage coupling model that predicting the threat of delayed water inrush from faulted structural belts was verified using FLAC3D numerical simulation software (Wu et al. 2011). Based on the dynamic damage mechanism, the HWCFZ and the water inrush mechanism of the roof were studied by FLAC3D . In addition, FLAC3D has been widely used to study the development characteristics of WCFZ, the fracture characteristics of overburden, the movement rules of overburden, surface subsidence, etc. (Wang et al. 2021d;Sikora and Wesołowski 2021;Niu et al. 2022;Ding et al. 2022).
The complex environment composed of high in-situ stress, high geothermal temperature, high karst water pressure and strong mining disturbance has brought great challenges to the prediction and measurement of the HWCFZ in the mining of deep buried thick coal seams Yuan 2021). At present, the calculation of the HWCFZ usually adopts the traditional empirical formula (Liu 1981), which only considers the influence of mining thickness. Due to the occurrence conditions of coal seams in different regions are very different, and the properties of rock formation vary widely, so there is no uniformity for the calculation of the HWCFZ in the mining of deep buried thick coal seams (Khanal et al. 2022;Wang et al. 2022a, b). Although many scholars have revised the traditional empirical formula (Guo et al. 2018a, b;Liu et al. 2019;He et al. 2020;Wang et al. 2021d), but it has limitations for deep buried thick coal seam mining Zhao et al. 2015;Yuan 2021). The HWCFZ can be more accurately determined by comprehensively considering the influence of the size of working face, stress transfer state, mining depth, seepage field, and rock mechanical properties on the development of the WCFZ in deepburied thick coal seam mining. Scholars usually take the development height of the plastic zone as the HWCFZ when FLAC3D numerical software simulation is used to study for the impact of underground engineering excavation on the aquifer (Wang et al. 2003;Sikora et al. 2021;Li et al. 2023). The height of the plastic zone cannot be regarded as the HWCFZ, mainly because the porosity and damage factor of the rock in the initial stage of plastic failure are very small, the fractures are not connected to form a water-conducting channel (Chen et al. 2014;Yi et al. 2020;Sun et al. 2021;Chen et al. 2022b, a). When the FLAC3D numerical simulation method is used to judge the HWCFZ, more accurate results can be obtained by comprehensively analyzing the variation law of pore water pressure and porosity, as well as the distribution characteristics of the plastic zone. As the WCFZ and the aquifer are connected, the water source of working face will increase, and the water inflow will increase suddenly (Qu et al. 2020;Xue et al. 2021). Therefore, it can be judged whether the WCFZ penetrates the aquifer by monitoring the change law of water inflow at the working face .
This paper adopts the research method combining theoretical analysis, numerical simulation and field measurement, and takes the 182706 working face of Wutongzhuang Mine as the engineering background. The hydraulic characteristics of the roof aquifer and aquiclude are analyzed. A calculation formula for the HWCFZ in deep buried thick coal seam mining is established considering the influence of size of working face, stress transfer state, mining depth, seepage field, and rock mechanical properties. Based on the damage-seepage coupling model, FLAC3D was re-developed by using FISH language. The bearing structure of the overburden, the evolution law of fissure and seepage field, and the HWCFZ were studied in the mining of the deep buried thick coal seam under aquifers. The results were 1 3 verified by monitoring and analyzing the water inflow on site. It provides a new method for calculating HWCFZ for deep buried thick coal seam mining. The research results are of great significance for releasing high-quality coal resources under water bodies, increasing non-renewable energy reserves, alleviating the contradiction between coal supply and demand, optimizing energy structure, and extending mine service cycles.

Study area
Wutongzhuang Mine is located in the southeast of Fengfeng mining area, west of Cixian county, Handan city, Hebei province, and north of Zhanghe river and Yuecheng reservoir. The western boundary of the mine field is the F 25 fault, the southwestern boundary is the F 53 and F 26 faults, the southern boundary is the exploration line 28, and the eastern boundary is the F 5 fault. The mining area is about 11.5 km long from north to south and 0-5 km wide from east to west. The mining area is distributed between Dongwushi Reservoir and Yuecheng Reservoir, and the Dongwushi Reservoir and Yuecheng Reservoir are located in the northeast and southwest respectively. There are Fuyang River, Yuefeng Canal, Zhongpan Canal and Xiangshui River in the mining area. The Zhang River is located on the southern boundary of the mining area. There are 21 villages distributed in the mining area, and the villagers communicate with the outside world through roads and railways. The location of the study area and working face is shown in Fig. 1. The study area is the 182,706 working face, the eastern boundary is the designed 182,605 working face, the west is adjacent to the F 702-10 fault, and the south is the gob of the 182,705 working face. The strike length of the working face is 573-672 m, the average length is 600 m, the strike direction dip angle is 0-9°, the dip length is 189 m, the coal seam thickness is 3.3-4.2 m, the average thickness is 3.7 m, the bulk density is 1.5 t/m 3 , and the average buried depth is 749 m, which belongs to the deep buried thick coal seam . The working face adopted the longwall retreating mining method, large cutting height mining comprehensive mechanized coal mining, and the caving method to deal with the roof of the gob.

The characteristics of aquifers and aquicludes in the roof
According to the comprehensive exploration data of the mine, the 182,706 working face is located in the Carboniferous-Permian coal-bearing strata. There are 4 groups of aquifers from the roof of the coal seam to the surface, and the aquicludes layers are distributed between the aquifers, from bottom to top: The first group of aquifers is sandstone fractured aquifer on the roof of the coal seam of the Shanxi Formation Lower series Permian system, it is the direct or indirect roof of the mining coal seam. The total thickness of the aquifers is 0-30.88 m, and the average is 24 m. According to the results of the borehole pumping test, the unit water inflow of the aquifer is 0.561 L·s −1 ·m −1 , and the permeability is 2.04 m·d −1 , which belongs to the medium waterrich confined fractured aquifer. The aquifers are exposed when the coal seam is mined, and the water mainly entered the mining space in the form of water spraying and dripping from the coal seam roof, and the amount of water is not large .
The second group of aquifers is the sandstone fractured aquifer is located at the bottom of the Lower Shihezi Formation. The total thickness of the aquifers is 2-22 m, an average thickness is 14.0 m, unit water inflow is 0.017-0.22 L·s −1 ·m −1 , permeability is 0.2-1.3 m·d −1 , and water rich property is weak to medium. The distance between the lower part of the aquifer and the roof of 2# coal seam is 25-34 m. When the coal seam is mined, the WCFZ of the roof of the gob is connected with the aquifer, which is an indirect waterfilled aquifer at the working face, and each aquifer is indirectly or connected to cross-layer replenishment.
The third group of aquifers is the sandstone fissure aquifer is located in the upper Shihezi Formation. The aquifers are divided into three sections, with thicknesses is 5-12 m, 50 m and 10.0 m from bottom to top, respectively. The maximum water inflow per hole of the 50 m-thick aquifer is 20 m 3 /h, the permeability is 0.1-0.7 m·d −1 , and the distance between the bottom and the roof of the 2# coal is 78-264 m; The sandstone aquifer with a thickness of 10 m has a unit water inflow of 0.018-0.14 L·s −1 ·m −1 , permeability of 0.1-0.7 m·d −1 , and the water richness is medium. The distance between the bottom and the 2# coal roof is 334 m. Because the HWCFZ is much less than 334 m, the sandstone aquifer with a thickness of 10 m has no effect on the mining of the coal seam.
The fourth group of fractured aquifers is located in the third and fourth systems. The third system is composed of sandstone and conglomerate, and the fourth system is composed of cultivated soil, red clay, sandy clay, sand, and gravel, with a permeability coefficient of 35-202 m·d −1 and a total thickness of 0.5-233.06 m. Due to the large mining depth of the mine, the WCFZ is not connected to the aquifer.
Due to the depositional characteristics and regularities of the Carboniferous and Permian strata, there are aquicludes between the aquifers. The aquiclude is mainly composed of bauxite mudstone, siltstone, mudstone, and a small amount of fine-grained sandstone, it has favorable impermeability in natural state (Wang et al. 2021a, b, c, d). During the period from June 2015 to December 2022, the results of the mine water inflow and surface precipitation were counted. The statistical results are showed in Fig. 2. The average mine water inflow was 6.85 m 3 ·min −1 , and maximum mine water inflow was 11.56 m 3 ·min −1 . From 2015 to 2022, there were two extreme rainfall events on the surface. During this period, the mine water inflow did not change significantly compared with other periods. The water inflow of the mine is almost unaffected when the surface precipitation increases. Therefore, there is no correlation between the amount of water inflow in the mine and surface precipitation, and the upper aquifers of the mining coal seam is the main source of water inflow in the mine.

Calculation of the HWCFZ
According to the damage and movement characteristics of the rock formation, the overburden of the gob is divided into the caving, fractured, abscission and curved subsidence zone when the coal seam is mined. The WCFZ described in coal seam mining under water body or aquifer mainly includes caving and fractured zone (Huang and Xia 2016;He et al. 2020;He et al. 2021). The abscission zone formed in the lower part of the aquifer during the mining of deep buried thick coal seams is regarded as a waterconducting channel. Therefore, when mining deep buried thick coal seam, the height of the WCFZ includes the height of the caving, the fractured and the abscission zone (Shi et al. 2012;Liu et al. 2021).
Coal seam thickness, mining depth, mining method, method of dealing with the roof of gob, faults, collapse columns, lithology and thickness of overburden, and watercontaining (impermeable) property of rock stratum are the influencing factors of water inrush in the roof of the working face. When the mining height and burial depth are less than 3 m and 300 m, respectively, according to the empirical formula, the HWCFZ is calculated as follows (China National Bureau of coal industry 2000). According to the above traditional empirical formula, the heights of the water-conducting fracture zones in the 182,706 working face coal seam mining are calculated to be 66.4 m and 67.7 m, respectively. The minimum height between the 50 m-thick sandy shale aquifer and the roof of the coal seam is 78 m. It can be seen that coal seam mining has no effect on the 50 m-thick sandy shale aquifer.
However, Peng (1984), Palchik (2003) and other scholars pointed out that the traditional empirical formula cannot accurately calculate the HWCFZ when the mining height is greater than 3 m (Ragon et al.2013).
The abutment stress of the overburden in the gob is equal to the weight of the upper stratum when the influence of seepage field on the HWCFZ is not considered. When the abutment stress generated by the overburden exceeds the shear strength of the rock, the rock is cracked and destroyed.
where m is abutment stress, MPa; max maximum shear strength of rock, taking 25.6 MPa for max .
The HWCFZ is calculated according to the following formula (Fang 2013; Li 2019; Wang et al. 2021d): where H w is the HWCFZ, m; L is the inclined length of working face, taking 189 m for L ; C x is the stress transmission coefficient between rock formations, taking 1.4 for C x ; is the average bulk density of the overburden of the coal seam, taking 25000 kN/m 3 for ; H is the average buried depth of coal seam, taking 749 m for H ; is the attenuation exponent of roof stress peak, taking 2.3 for , its value is related to the combination of the thickness of the coal seam and the rock properties of the roof.
The abutment pressure of the overburden is equal to the sum of the weight of the overburden and water pressure of the aquifer when considering the influence of seepage field on the calculation of the HWCFZ. At this time, the conditions for rock failure is (Xiao et al. 2020;Fu et al. 2022;Shao et al. 2022): where P w is the confined water pressure of an aquifer, MPa.
Due to the action of water, the maximum shear strength of the rock formation is equal to the strength after water softening (Liu et al. 2008;Yu et al. 2020;Xie et al. 2021). At this time, the HWCFZ is calculated as follows: where w is the softening factor, taking 0.6 for w .
In order to verify the correctness of the proposed calculation formula, the observed values of the HWCFZ in mines with similar conditions were collected, and the proposed calculation formula is used to calculate the HWCFZ. The observed and simulated values of the HWCFZs of 39 mines are obtained, and the results are shown in the Fig. 3 (Shi et al. 2012;Fan 2013;Guo et al. 2018a, b;Chang et al. 2018;Ma et al. 2019;He et al. 2020;Zhang et al. 2021;Liu et al. 2021). According to the statistical results, the performance metrics of the calculation model include Nash-Sutcliffe Efficiency (NSE), Root Mean Square Error (RMSE), Ratio of Standard Deviation (RSD), Mean Absolute Error (MAE), Volume Error (VE), Relative Error (RE) and Percent bias (PBIAS) are calculated. The optimal value of PBIAS is 0, and lower values represent more accurate calculation models. A positive value indicates that the model underestimates the bias, and a negative value indicates that the model overestimates the bias (Gupta et al. 1999;Moriasi et al. 2007;Burgan et al. 2022). The evaluation criteria of other performance metrics are shown in Table 1 ( Burgan et al. 2022). and the results are shown in Table 2.
According to the statistics in Fig. 3 and calculation results in Table 2, compared with the observed values, the simulated values is generally overestimated, this is beneficial to   ). Mining of extra-thick seams which can be seen as the limitation of the model. However, the model is a development in mining science, and it is particularly useful for the practice of the HWCFZ in deep buried thick coal seam mining. Therefore, despite the limitations for mining of extra-thick coal seams, it should be emphasized that the model is an important progress in the mining of deep coal seams under aquifers, and can be utilized with known buried depth, mining method, and seepage field. Through calculation, the WCFZ of coal seam mining in 182,706 working face without considering the seepage field and considering the effect of seepage field are 108.9 m and 111.6 m, respectively.
When mining deep buried thick coal seams, if the calculated critical value of the HWCFZ is equal to the distance between the bottom of the aquifer and roof of the coal seam, the influence of the height of the curved subsidence zone and thickness of the aquifer should be considered (Duan et al. 2006). At this time, the HWCFZ is calculated as follows: where H wt is the HWCFZ considering the thickness of the aquifer and the height of the curved subsidence zone; h w is the thickness of the aquifer, taking 50 m for h w ; h c is the value calculated using Eq. (6) coincides with the height of the aquifer, taking 33.6 m for h c .
The distance between the bottom of the sandy shale aquifer with thickness of 50 m and the top of the coal seam is 78 m, so the height of the water-conducting fracture zone is 128 m.

Numerical simulation analysis
Analyzing the failure characteristics of the overburden and the evolution law of the seepage field is of great significance for studying the HWCFZ. High cost, low efficiency and low accuracy are the shortcomings of field test when mining deep buried thick coal seams (Chen et al. 2022a, b). With the rapid development of computer technology and the advantages of numerical simulation technology, numerical simulation technology is widely used in underground engineering (Sikora and Wesołowski 2021;Ding et al. 2022). For example, Tang (2002) established a seepage-stress-damage coupling model for heterogeneous rock mass based on the analysis results of triaxial seepage experiments, and used F-RFPA2D numerical simulation to carry out fluid-solid coupling calculation analysis for different heterogeneity coefficients. According to the basic information of the coal seam and rock formation of the roof and floor of the 182,706 working face, the FLAC3D numerical simulation software was used to establish a calculation model, and the failure characteristics of the overlying rock, the change of pore water pressure and the seepage field of the 182,706 working face were calculated and analyzed. Combined with the theoretical calculation results, it is beneficial to comprehensively and accurately judge the HWCFZ (Sun et al. 2021;Wang et al. 2021a, b, c, d).

Calculation of damage variables for rocks
According to the Mohr-Coulomb criterion, assuming that the tensile stress is negative, the failure criterion of the rock is showed in Eqs. (8) and (9) (Rutqvist et al. 2002;Zhu et al. 2004): where is the rock internal friction angle, °; f t0 is the uniaxial tensile strength, MPa; f c0 is the uniaxial compressive strength, MPa.
The damage variables of elastic and plastic stages in the process of rock failure are expressed as (Bunger et al. 2013; where t0 is the maximum tensile strain of the elastic stage; c0 is the maximum shear strain of the elastic stage; 1 is the first principal strain; 3 is the third principal strain; n is the constants related to rock strength.

Calculation of the porosity of rocks
The effective stress principle is used to analyze the hydraulic coupling effect of porous media. According to Biot's (1955) research results.
where * ij is the effective stress matrix; ij is the stress matrix; b is the Biot coefficient; P is the pore water pressure, MPa; ij is the second-order identity matrix.
According to Lemaitre's equivalence principle of strain, the relationship between the damage variable and the effective stress is (Lemaitre 1992): where * ij is the effective stress tensor; ij is the stress tensor; E ij is the elastic modulus tensor.
According to formula (11) and (12) can be obtained: The relationship between stress and strain is: According to the stress-strain relationship in the three directions, the strains in the elastic and plastic stages of the rock are expressed as (Jiang et al. 2021). The volumetric strains of the elastic and plastic stages of rock are: The porosity in the elastic and plastic stages are calculated as follows: where n is the porosity; n 0 is the initial porosity; v is the volumetric strain.

Damage-seepage coupling model
According to the cubic relationship between permeability and porosity, the effect of damage variable on permeability is as follows  where k is the influence coefficient of damage on permeability; k is the permeability; k 0 is the initial permeability.

Numerical simulation calculation model
Based on the damage-seepage coupling model, the FLAC3D was re-developed by using the FISH language, and a fluid-solid coupling calculation model for deep buried thick coal seam mining was established (Niu et al. 2022;Chen et al. 2022a, b). According to the (20) hydrogeological conditions of the mine, the established coal seam and overburden are all horizontal models. The model includes a top stratum with a height of 231 m from the top of the coal seam, mined coal seam of 3.7 m, and bottom stratum with a depth of 58.3 m from the bottom of the coal seam. The self-weight stress generated by the top strata is expressed by applying a uniform load. The average bulk density of the stratum is 25 kN/ m 3 , the average depth is 518 m, and the uniform load is 12.95 MPa. The flow chart of the numerical modeling process is shown in Fig. 4. The upper part of the model is set as a free boundary. The remaining boundaries of the model are fixed boundaries, and the horizontal and vertical displacements of the boundaries are zero. The surrounding boundaries of the model are set as impervious boundaries, and the stratum are set as free permeable boundaries. The parameters of the rock formations in the model, the initial pore water pressure and permeability of the aquifer are based on field test. Its parameters are showed in the Table 3.
The simulated coal seam mining thickness is 3.7 m, mining step is 5 m, advancing length is 600 m, and mining method is full-height mining at one time. In order to eliminate the influence of the boundary effect, a 100 m coal pillar is left around the gob.

Evolution process of the plastic field
According to past research, the fracture of rock can be characterized by the plastic zone. After the coal seam is mined, the failure mode of the caving zone of the overburden in the gob is mainly tensile-shear mixed failure, and the fissure zone is mainly shear failure (Wang et al. 2003 Figure 5 shows the distribution characteristics of the plastic zone when the working face is advanced to 40 m, 200 m, 220 m and 320 m. As showed in Fig. 5a, when the working face advances to 40 m. The failure mode of the immediate roof and main roof rock is the Fig. 5 Distribution of plastic zone of rock formation with different advancing distances in working face 3.3.2 The evolution law of permeability tension shear mixed failure. and the upper part of the main roof rock is shear failure. The height of tension shear mixed failure zone is 9 m, and shear failure zone is 24 m. Therefore, the height of the caving zone is 9 m and the height of the fractured zone is 24 m. The shape of the overall distribution is trapezoid. As showed in Fig. 5b, With the advancement of the working face, the caving zone and the fractured zone continue to expand to the upper strata. When the working face is advanced to 200 m, the height of the largest caving zone is 27 m, and the height of the largest fractured zone is 78 m. The distribution pattern of the fractured zone is high on both sides and low in the middle. The fractured zone developed to the siltstone aquiclude with a thickness of 15 m. Therefore, when the working face is advanced to 200 m, coal seam mining has an impact on the sandy shale aquifer with a thickness of 50 m. As showed in Fig. 5c, the height of the largest fractured zone is 98 m, and the fracture zone develops to the sandy shale aquifer with a thickness of 50 m. The distribution characteristics of the fractured zone are high on both sides and low in the middle. As showed in Fig. 5d, When the working face is advanced to 320 m, the development height of the fractured zone reaches the maximum value is 113 m. With the continuous advancement of the working face, the maximum height of the fractured zone no longer increases with the advancement of the working face. Figure 6 shows the distribution of the variation characteristics of permeability when the working face is advanced to 40 m, 200 m, 220 m and 320 m, respectively. The permeability of the rock formation gradually increases with the continuous advancement of the working face. When the working face advances to 320 m, the permeability of the rock formation is hardly affected by mining when the vertical distance from the roof of the coal seam is greater than 123 m. The variation characteristics of rock permeability are: As the vertical distance between the rock formations and the roof of the coal seam increases, the amount of increase in permeability decreases gradually; Due to the different mechanical states and properties of rock formations, the change of permeability has obvious difference and zoning features. The zoning feature is that the increase of the permeability of the caving zone is obviously larger than that of the fractured zone, and the permeability of the rock formation on both sides of the gob is greater than that in the middle of the gob.
Combined plastic zone and permeability coefficient distribution characteristic pictures. The permeability of the rock stratum at different heights from the roof of the coal seam were monitored, and the distances between the monitoring point and the open cut were 20 m, 200 m, and 300 m, respectively. The results are showed in Fig. 7a.
As showed in Fig. 7a, the permeability of the floor strata in the gob decreases gradually with the increase of the floor depth, and the influence range of coal seam mining on the permeability of the floor strata is 22.7 m. Compared with the initial permeability, it can be seen that the maximum height of the influence of coal seam mining on the overburden in the gob is 122.9 m. The overall variation characteristics of permeability are: permeability on the side of the open cut > on the side of the working face > on the middle of the gob. Because the side of the open cut is affected by mining for a long time, so the fractures are sufficiently developed, the change value of permeability is greater than that of the side of the working face. Because the action of mine pressure re-compacts the fissures of the overlying strata in the middle of the gob, the change value of its permeability is smaller than that on both sides of the gob. When the height from the coal seam roof is greater than 102.5 m, the permeability of the overburden on both sides of the gob are equal, and the permeability decreases rapidly with the increase of height; When the height from the top of the coal seam is greater than 117.7 m, the permeability of the rock formation is consistent affected by mining. The main reason is that the fissured zone of the overburden in the middle of the gob is less affected by the compaction of the mine pressure. From the change 1 3 Fig. 6 Distribution of permeability coefficient at different advancing distances of working face trend of the curve, it can be known that due to the different lithology and stress state of the stratum, the variation law of the permeability of different stratum is different. However, as the height of the measuring point from the coal seam roof increases, the permeability shows a decreasing trend.
In order to study the variation law of the permeability of the key aquicludes affected by mining, when the working face is advanced to 320 m, the permeability of different horizontal distances of the aquicludes is monitored, the aquicludes include 4 m-thick siltstone, 4 m-thick aluminous mudstone and 15 m-thick siltstone, the distance from the seam roof are 63 m, 55 m and 37 m, respectively. The results are showed in Fig. 7b. The greater the distance from the coal seam roof, the smaller the permeability. For the aquiclude with a distance of 37 m from the roof of the coal seam, the permeability of the side of the open cut is obviously greater than that of the side of the working face. The main reason is that  Figure 8 shows the distribution of the variation characteristics of pore water pressure when the working face is advanced to 40 m, 200 m, 220 m and 320 m, respectively. As showed in Fig. 8b, as the working face continues to advance, the aquifer is affected by mining in sequence from bottom to top. The pore water pressure of the aquifers on the direct roof and basic roof decreases rapidly with the mining of the coal seam, and the water conduction channels are mainly the fractured zones of the overlying rock layers on both sides of the gob; When the working face advanced to 200 m, the pore water pressure of the 14 m-thick medium sandstone aquifer began to decrease, indicating that the lower aquiclude lost its water-resistance performance, and the WCFZ is connected to the aquifer. When the working face advances to 220 m, the pore water pressure of the 50 m-thick sandy shale aquifer began to decrease, and the relative change of pore water pressure is smaller than that of the lower aquifer. The main reason is that the vertical distance between the aquifer and the coal seam roof is large, the mining influence is small, the fractures are not fully developed, and the water permeability of the aquifer is weak. At this time, the water-conducting area is close to the side of the open cut. As the working face continues to advance, the WCFZ above the working face gradually connects with the sandy shale aquifer. When the working face advances to 320 m, the pore water pressure of the 50 m-thick sandy shale aquifer decreases significantly. The pore water pressure recovered to 81% of the initial pore water pressure near the side of open cut. The main factors were the re-compaction of the fractured zone, the restoration of the water-resistance performance of the aquicludes, and the lateral recharge of the aquifer water source. The pore water pressure of the 14 m-thick medium sandstone aquifer also recovered to 68% of the initial pore water pressure due to lateral and vertical water supply.

Evolution law of pore water pressure
Monitoring lines are arranged at positions 20 m, 100 m, 200 m and 350 m from the open cut to monitor the pore water pressure of rock formations at different heights from the roof of the coal seam. The results are showed in Fig. 9a. It can be seen from the figure that the pore water pressure of the aquifer in the range of 0 to 25 m from the roof of the coal seam is much lower than the initial value, the reason is most of the water in the aquifer flows into the gob. In the fractured zone with a distance of 28 m to 40 m from the roof of the coal seam, the pore water pressure in this area increases due to the inflow of water from the upper aquifer into the fractured zone. When the height of the measuring point from the roof of the coal seam is greater than 63 m, the distribution of pore water pressure showed obvious zoning feature. Except for the side of the open cut, the smaller the distance from the open cut, the higher the pore water pressure. The reason is that the fractured zone on the side of the open cut is highly compacted, and the permeability of the aquiclude in the fracture zone is small. It can be seen that time is one of the influencing factors of the waterresistance performance recovery of the aquiclude. The overall distribution characteristics of pore water pressure are: the side near the open cut > the middle of the gob > the side of the open cut > the side of the working face.
In order to study the evolution law of the pore water pressure of the aquifers, when the working face advanced to 400 m, the pore water pressure of different horizontal distances of the aquifers was monitored, and the results are showed in Fig. 9b. It can be 1 3 seen from the figure that the pore water pressures of the aquifers with heights of 43 m and 83 m from the roof of the coal seam are all lower than the initial value. Its distribution characteristics are that the two sides of the gob are smaller than the middle, the pore water pressure on the side near the open cut is greater than that of the working face in the central area of the gob, and the side of the working face is smaller than the side of the open cut. In the aquifer with a height of 61 m from the top of the coal seam, the pore water pressure increases significantly on the side close to the open cut. There are Fig. 8 Distribution of pore water pressure at different mining distances three main reasons: First, the WCFZ runs through the 50 m-thick sandy shale aquifer. Second, the aluminous mudstone aquifer in the lower part of the aquifer has good waterresistance performance. Third, the fissure zone was re-compacted under the action of mine pressure, and the aluminous mudstone aquifer recovered its water-resistance performance. The interaction of three factors results in an increase in the head of the aquifer. Compared with the initial pore water pressure, the pore water pressure of the aquifer in the caving zone is obviously reduced, and there is no recovery phenomenon. The main reason is that in order to simulate the drainage situation of the working face, the drainage method is used to calculate the fluid-solid coupling of water. Fig. 9 Results of pore water pressure monitoring 1 3

Working face water inflow measurement
According to the expected results from the data of the hydrogeological department, the maximum water inflow during the mining of the working face is 0.5 m 3 /min. Drainage channels are laid in the headgate and tailgate, and the drainage point is arranged in the low-lying part of the roadway, and the drainage capacity is not less than 2 m 3 /min. The MD155-30*7 multistage centrifugal pump is installed at the drainage point, and the technical parameters are: The maximum lift is 210 m, flow is 2.58 m 3 /min, and power is 160 kW. The pump lift and drainage capacity are both greater than the maximum water inflow of the working face, so it can meet the drainage demand of the working face. The water inflow of the working face was comprehensively measured by the pump discharge method and flow rate meter method during the mining period of the working face (Zhu 2007;Wang et al. 2016;Chen et al. 2019).
The basic principle of water inflow measurement method: (1) The basic principle of the pump discharge method. The principle of the pump discharge method is to using the actual displacement of the pump and the running time of the pump to calculate the displacement as: where Q is the water flow, m 3 /h; Q n is the nameplate displacement of the water pump, m 3 /h; is the actual efficiency of the water pump; t 1 is the water pump drainage time, s; t 2 is the time when the water gushes into the water bin, s; N is the number of pumps.
(2) The basic principle of flow rate meter method.
Observation principle: The instantaneous flow velocity is equal to the number of photoelectric pulses recorded per second multiplied by the coefficient of the propeller, the instantaneous flow is equal to the instantaneous flow velocity multiplied by the cross-sectional area of the open channel, and the average flow is obtained through the accumulation of time. The calculation formula of water inflow is: where v is the instantaneous flow velocity of water, m/s; v is the average flow of water, m/s; t 0 is the starting moment, s; t 1 is the termination time, s; S is the cross-sectional area, m 2 .
The water inflow of the working face was measured and counted when the 182,706 working face started to mining, and the statistical results are showed in Fig. 10a. From the figure, it can be seen that the minimum water inflow of the working face is 0.114 m 3 /min, maximum is 0.522 m 3 /min, and average is 0.349 m 3 /min. The statistical data is analyzed, and its distribution characteristic is an exponential distribution, as shown in Fig. 10b. The characteristic properties include skewness, coefficient of variation, confidence intervals, standard deviation and average absolute deviation as shown in Table 4.
Starting from the open cut, the water inflow gradually increased with the advancing distance of the working face. The water inflow increased from 0.114m 3 /min to 0.168 m 3 /min when the working face was advanced to 28 m. The water inflow jumpingly increased to 0.239 m 3 /min when the working face was advanced to 32 m. The reason is that the roof of the working face collapsed for the first time, and a large number of water-conducting cracks appeared on the roof of the coal seam ). After that, as the working face continued to advance, the water inflow remained between 0.213 m 3 /min and 0.256 m 3 /min. The water inflow of the working face increased sharply for the second time, and increased from 0.253 m 3 /min to 0.329 m 3 /min when the working face was advanced to 156 m. It can be judged that the WCFZ is connected to the medium-sandstone aquifer with a thickness of 14 m. The third water inflow increased sharply, and the water inflow increased from 0.344 m 3 /min to 0.446 m 3 /min when the working face advanced to 192 m. From this, it can be judged that the WCFZ is connected to the 50 m-thick sandy shale aquifer. Later, as the working face continued to advance, the water inflow increased slowly.
The main reason was that the area of the gob increased, which led to the increase of waterconducting fractured channels in the roof, which provided more water-conducting channels for the water source of the upper aquifer.

Discussion
The traditional empirical formula for the HWCFZ is mainly applicable to the mining of shallow coal seams. Although many scholars have revised the formula (Guo et al. 2018a, b;Liu et al. 2019;He et al. 2020;Wang et al. 2021d), but it has limitations for deep buried thick coal seam mining. A calculation model for the HWCFZ is proposed considering the working face size, stress transfer state, mining depth, seepage field and   (Moriasi et al. 2007;Burgan et al. 2022). When FLAC3D numerical simulation software is used to study the development of WCFZ, the height of plastic zones is used as the HWCFZ, which can intuitively describe the impact of coal seam mining on aquifers (Wang et al. 2003;Zhang et al. 2019;Sikora et al. 2021;Chen et al. 2022b, a). It is inaccurate to regard the height of the plastic zone as the HWCFZ, mainly because the porosity and damage factor of the rock in the initial stage of plastic failure are very small, and the fractures are not connected to form a water-conducting channel (Chen et al. 2014;Yi et al. 2020;Sun et al 2021). A fluid-solid coupling seepage model based on porosity and damage factors is established, the re-development of FLAC3D is carried out by using the FISH language. A numerical calculation model for the HWCFZ in deep buried thick coal seam mining was established, and the HWCFZ was determined according to the distribution characteristics of the plastic zone, the evolution law of the permeability coefficient and the change law of the pore water pressure. When the WCFZ is connected to the aquifer, the water inflow at the working face will suddenly increase (Qu et al. 2020;Xue et al. 2021). Therefore, it can be judged whether the water-conducting fracture zone penetrates the aquifer by monitoring the change law of water inflow at working face ). The results of theoretical calculation, numerical simulation and field measurement are mutually verified. The proposed method for predicting the HWCFZ in the mining of deep buried thick coal seams has high reliability, it provides a new method for the calculation of HWCFZ in coal mining and theoretical support for the safe and efficient mining technology of coal resources under water. The high-quality coal resources under the water body are released, the recovery rate of coal resources is significantly improved and the service cycle of the mine is extended. It has created good economic benefits for enterprises and made a significant contribution to the growth of the regional economy. The HWCFZ is accurately determined, which can guarantee the safe recovery of coal resources under the water, elevate non-renewable energy reserves, alleviate the contradiction between coal supply and demand, optimize the energy structure, and provide strong energy power for economic construction in the new era.
There are great differences in the distribution characteristics of WCFZ when geological conditions and mining methods are different (Wang et al. 2021d). According to the comparison results of the observed and simulated values, it is found that the error of the simulated value is large when the ultra-thick coal seam is mined. Therefore, the proposed calculation method has certain limitations for ultra-thick coal seam mining (Li 2019). Since the influence of faults on the HWCFZ on the roof is not considered, the prediction model also has certain limitations when there are faults on the roof of the working face (Wu et al. 2011;Cao et al. 2019;Wang et al. 2021d). Fault is an important factor affecting the safe recovery of underground working faces, and some water inrush accidents in mine working faces are associated with fault. In order to comprehensively judge and accurately prevent the occurrence of water inrush accidents in the mining process of the working face, when deep mining is affected by faults, it is necessary to further study the activation mechanism of faults, the failure form of overlying rock, the development and expansion of fractures, and the evolution of seepage fields, explore the distribution characteristics of the WCFZ in deep mining under the influence of faults.

Conclusions
(1) Mine water inflow is not affected by surface rainfall. The water inflow in the gob of the 182,706 working face mainly comes from the sandstone fissure aquifer of the Lower Permian Shanxi Formation, the sandstone fissure aquifer at the bottom of the Lower Shihezi Formation, and the 50 m-thick sandstone fissure aquifer of the Upper Shihezi Formation. The aquicludes has good water-resistance performance in the natural state.
(2) Taking shear strength as the criterion for rock failure, and considering the effect of working face size, stress transfer state, mining depth, seepage field, rock mechanical properties,, the formula for calculating the HWCFZ in deep-and-thick coal seam mining is derived. The model provides fairly acceptable results compared to observed values. It is calculated that the HWCFZ of the 182,706 working face is 128 m. (3) Based on the fluid-solid coupling seepage model considering porosity and damage factors, the fish language was used to carry out re-development of FLAC3D, and a numerical calculation model for the HWCFZ was established. The fissure field of overburden, seepage field, and HWCFZ were studied in deep-buried thick coal seam mining under the aquifer. (4) The numerical calculation results showed that: The main failure mode of the rock in the caving zone is the mixed tension-shear failure, and fractured zone is the shear failure; The distribution characteristics of plastic zone and permeability are trapezoidal in the early stage of mining. The distribution characteristics of plastic zone and permeability are saddle-shaped with high on both sides of the goaf and low in the middle when the working face is advanced to 200 m. The fracture zone develops to a maximum height of 113 m, and the maximum height of the caving zone is 27 m when the working face is advanced to 320 m. In the process of transition from trapezoid to saddle shape, the fractures on the side of the open cut are relatively developed, and the permeability is relatively large. The permeability of the caving zone is obviously larger than the fracture zone. The maximum height of the area with the increased permeability from the roof of the coal seam is 123 m. The pore water pressure of the aquifer decreases with the increase of the vertical distance from the coal seam roof. The overall distribution characteristics of pore water pressure are: the side near the open cut > the middle of the gob > the side of the open cut > the side of the working face. Compared with the initial pore water pressure, the pore water pressure in some areas of the aquifer will increase. (5) The monitoring results of the water inflow at the working face shows that the WCFZ is connected to the sandy shale aquifer with a thickness of 50 m, and the results are consistent with the theoretical calculation and numerical simulation. The research results provide theoretical support for the safe and efficient mining technology of coal resources under water, high-quality coal resources under water bodies are released, non-renewable energy reserves are improved, the contradiction between coal supply and demand is alleviated, the energy structure is optimized, and the service cycle of mines is extended. It has created good economic benefits for enterprises, made important contributions to the growth of regional economy, and provided strong energy power for economic construction in the new era.