Numerical modelling of fractures induced by coal mining beneath reservoirs and aquifers in China

Transcription

1 research-articleResearch ArticleXXX /qjegh X. Wu <italic>et al</italic>.permeable Fracture Zone 2013 Numerical modelling of fractures induced by coal mining beneath reservoirs and aquifers in China Xiong Wu 1*, Xiao-Wei Jiang 1, Yu-Fu Chen 1, Xiao-Lei Wang 1 & Shu-Hong Tan 2 1 School of Water Resources and Environment, China University of Geosciences, Beijing , China 2 School of Engineering Technology, China University of Geosciences, Beijing , China *Corresponding author ( wuxiong@cugb.edu.cn) Abstract: The increasing demand in recent years for energy in China has resulted in coal being mined in complex geological conditions. This has included the mining of coal seams located beneath reservoirs, aquifers and flooded abandoned mine workings. Throughout China the extraction of coal in these circumstances has resulted in groundwater inrushes into working mines. This has caused loss of lives, disruption or abandonment of coal mining, and the sterilization of coal resources. This paper demonstrates that numerical modelling using ANSYS-2D, FLAC-2D and FLAC-3D combined with field observations may provide a method for the modelling of fractures that develop in the rock mass above a roof of coal mines. These fractures may increase significantly the permeability of the strata and groundwater flows into mine workings. Coal is important in many countries as a source of energy and for sustainable economic growth and development. For example, in South Africa, Poland, India and Australia, coal is the main source of energy, accounting for 77.1%, 68.1%, 56.8% and 44.5% of the primary energy consumption, respectively (Fan 2004). In China, the consumption of coal constituted about 70% of primary energy in Coal is likely to remain a major energy source in China for the foreseeable future. By 2050, it has been estimated that the demand for coal in China will account for 50% of the total energy consumption (Zhou 2009). With the increasing demand for Chinese coal, those deposits situated in more complex geological conditions are now being mined. This includes coal located beneath large water-bearing bodies, such as reservoirs, aquifers and flooded previously mined workings. For example, coal resources beneath Xiaolangdi Reservoir and Nansihu Hydropower Station are estimated to be c t (Wang 2002; Wu & Chen 2003) and t (Lu et al. 2003) respectively. Aquifers pose particular problems for mining, as the groundwater may flow into mine workings along fractures induced or exacerbated by mining and subsidence. These aquifers are Cenozoic in age and include karst and fissured strata as well as former worked-out areas. The mining of coal in these circumstances has caused inrushes and mining disasters in the past in China. This has resulted in flooding and the deposition of large quantities of fine sediments and debris in the mine workings. In some cases inrushes have resulted in fatalities, economic losses or delays in coal production. At Xinhe mine (National Coal Corporation of China 1992), which is located in Xuzhou Province, an inrush resulted in the deposition of huge volumes of silt, the blockage of a 1200 m section of mine tunnel and 58 days loss of coal production. Similar inrushes at Daxing (in Meizhou and Guangdong), Xinjing (at Zuoyun and Shanxi) and Nadu (in Youjiang and Guangxi) mining areas caused 121, 56 and 36 deaths respectively, in addition to significant economic losses (Zhao et al. 2006; Bai 2009). Further examples of inrushes in Chinese coal mines are presented in Table 1. Brief overview of coal mining beneath water bodies and aquifers Since the 1850s, some the world s leading coal-producing countries including Poland, Russia, the UK, Germany, Japan and Canada have investigated the deformation and failure mechanisms of the strata overlying coal mines, when extraction has taken place beneath water bodies. In the UK, Japan and Canada, undersea coal mining began in 1857, 1863 and 1874 respectively. In Nova Scotia, Canada, undersea mining regulations were established by the Canadian mining industry to stipulate the mining methods (Cui et al. 1997; Chen & Li 2010). When the mining depth to extraction thickness ratio exceeds 100 longwall or sublevel caving may be used. However, if this ratio is less than 100 partial extraction, using the room and pillar method, is recommended (Gu & Shen 1973). The depth at which coal located beneath water bodies should be mined by partial extraction varies from 46 m in Australia to 93 m in Japan (Table 2). In Japan, there have been five recorded incidents where seawater has flowed into coal mines. The most serious resulted in 10 deaths and 30 injuries (Cui 2008). China became the sixth country to allow coal mining below the sea bed (after the UK, Australia, Chile, Japan and Canada). This took place at the Longkou Mining Bureau s Beizao mine, located in Shandong province (Cui 2008). Since the 1960s, China has successfully mined coal situated beneath water bodies such as aquifers, flooded mine workings, and large rivers such as the Huaihe, Taizihe, Xiaowenhe, Puhe and Mianhe. As a result, there now exists in China improved knowledge and experience of coal mining beneath such water bodies and the associated failure of the overlying rock mass (Wang et al. 2010; Hu & Zhao 2011; Zhang 2011). Investigations have used empirical equations (State Bureau of Coal Industry 2000), in situ observations and monitoring (Hatherly et al. 1995; Cheng et al. 2001; Han et al. 2001; Song 2007; Cui 2008), physical models (Hu et al. 2003; Gao et al. 2004) and numerical modelling (Bower & Zyvoloski 1997; Rutqvist et al. 2002; Minkoff et al. 2003). The use of empirical equations is considered to be subjective and not suitable for calculating the extent of water flow through fractured strata. In situ monitoring has limitations, as it can provide specific observations for only selected parts of a mine. Physical modelling also has limitations, as this method cannot consider the complex rock mass discontinuities that may often prevail in Coal Measures strata. An alternative approach for the investigation of fracture development above coal mines may be the application of numerical modelling. The increasing development of computer technology has allowed numerical modelling to become a useful technique for the investigation Quarterly Journal of Engineering Geology and Hydrogeology, Vol. 46, 2013, pp doi: /qjegh Published Online First on March 25, The Geological Society of London

2 238 X. WU ET AL. Table 1. Flooding accidents that have occurred in China Name of mining area Location Pit water inflow (m 3 h 1 ) Suncun Shandong Province 1860 Fengfeng Hebei Province 2400 Yanmazhuang Henan Province Jinggezhuang Hebei Province 2400 Hanqiao Jiangsu Province 4440 Qingshangquan Jiangsu Province 5709 Daxing Guangdong Province Xinjing Zhangjiachang Town, Datong City, Shangxi Province Nadu Guangxi Province Anjialing Shanxi Province Table 2. Findings of previous work from various countries Country Minimum thickness of overlying rock (m) Remark UK 60 Partial extraction Chile 70 Partial extraction Japan 93 Partial extraction Canada 55 Partial extraction Australia 46 Partial extraction of rock mass deformation and fracture zone development. Numerical simulation is a method used to calculate and analyse the mechanical properties of strata. This method requires the establishment of boundary conditions and empirical equations, which can simulate the in situ rock mass behaviour. In recent years, several types of commercial software have been applied to the investigation of permeability and rock mass fractures (Wu et al. 2009). However, the results of these models have been influenced by variations in the judgements used to determine geotechnical parameters and the geometry of the fracture zone induced above mining voids. Such variations have caused discrepancies between the results of in situ observations and modelling. Numerical simulation by using FLAC-2D and ANSYS-2D Geological overview of study area Shenjiazhuang Colliery, a key state-owned coal mine, is located in the town of Huangsha, Ci County, in Hebei Province, to the SW of Beijing. Yuecheng Reservoir is located nearby and has a capacity of m 3. This is the main water source for domestic use in the cities of Handan and Anyang. In 1996, coal mining at Shenjiazhuang Colliery started to extend beneath Yucheng Reservoir (Fig. 1). The Carboniferous Permian coalbearing strata contain several coal seams, of which the No. 2 seam, in the lower part of the Shanxi Group, was the only coal seam that was extracted. This seam varies from 3.15 to 5.14 m thick, with an average thickness of 4.2 m. The seam was extracted from a depth of between 190 and 338 m. The coal seam has a simple structure and the roof strata comprise medium- to fine-grained sandstone averaging c. 9.7 m thick. The strata beneath the floor of the mine consist of siltstone about 4.2 m thick (Fig. 2). The regional geology of the coal basin where Shenjiazhuang Colliery is situated is dominated by a monoclinal structure, with the strata dipping c. 17 towards the general direction of the reservoir s dam. There are several principal high-angle normal faults that strike NE SW and NNE SSW. These faults obliquely intersect the No. 2 coal seam with displacement of between 20 and 50 m. Mechanical parameters The mechanical properties for the strata at Shenjiazhuang Colliery were interpolated from parameters obtained for the nearby Huangsha Colliery, located in the same coal basin (Table 3) (Wu et al. 2004, 2007, 2009). Model calculation The boundary of the model used to analyse the changes in stress and strain of the strata overlying the coal mine had a representative cross-section of 1750 m and a depth of 560 m; it was subdivided into elements and was cubic in shape. The model considered the worst-case scenario when the Yuecheng reservoir was at maximum capacity. Considering the geostatic stress of the rock mass and the water loading of the reservoir at its highest level (159.1 m), a vertical stress of 0.2 MPa was applied to the model. The base of the model was fixed to restrict displacement in two directions and rotation in the horizontal plane. The lateral sides were also fixed to restrict rotation in this plane and horizontal displacement, but vertical movement was allowed. Simulation results using FLAC-2D Fig. 1. Location of the area of study, showing Shijiazhuang, the Yuecheng Reservoir, Xuzhau and the Nansihu Hydropower Scheme. Immediately following the cutting of coal and the removal and advancement of the roof supports, movement and deformation of the roof rocks begins as the strata cave, and propagates upwards. These areas of deformation have been divided into three distinct zones based on the intensity of deformation, known as the caving zone, the fracture zone and the subsidence zone (Fig. 3) (Peng & Meng 2002). The caving zone includes roof and overlying rocks that fall into the void created by mining following the advancement of the roof supports. The rocks are characterized by their irregularity, variable sizes, high void ratios and low density. The strata in the caved zone may have enhanced permeability. Roof strata tend to be exposed to the greatest amount of deformation, including dilation and bending caused by large incremental strain accumulations. Therefore this may permit any groundwater to flow into the mine workings.

3 Permeable Fracture Zone 239 Fig. 2. Comprehensive stratigraphic column for Shenjiazhuang Colliery and the Liuhe Industry Limited Company. The fracture zone is located above the caved zone. Here, the strata maintain a higher degree of integrity and layering, although any original stratification may be disrupted by subvertical fissures and the dilation of joints and bedding planes. The characteristic geometry of this zone displays an increased density of fractures located closer to the caved zone with fracture attenuation occurring with increasing height. The subsidence zone includes the rock mass from the uppermost limits of the fracture zone to the ground surface. The strata in this zone experience the maximum subsidence above the centre of the mine workings, with the ground movements spreading outwards from the area of extraction. The caved and fracture zones are referred to collectively as the permeable fracture zone, and this is significant during mining owing to the

4 240 X. WU ET AL. Table 3. Parameters for rock mechanical properties Strata name Elastic modulus E (MPa) Poisson ratio μ Volume weight γ (KN m 3 ) Internal friction angle φ (deg.) Cohesive strength C (MPa) O C P 1 X P 1 S P 2 S P 2 S P 2 S P 2 S Q Coal seam Mudstone Fault The parameters were obtained from the results of back analysis, which used the results of underwater topographic survey before and after excavation of the colliery. Fig. 3. Schematic cross-section showing the failure characteristics following coal extraction: 1, irregular caved zone; 2, regular caved zone; 3, intensive fracture zone; 4, general fracture zone; 5, microfracture zone; 6, caved zone; 7, fracture zone; 8, subsidence trough; 9, destructive area (permeable fracture zone); 10, non-destructive area. increased porosity and permeability. If the permeable fracture zone propagates to intersect the base of a reservoir, aquifer or flooded mine workings, this may cause a rapid inflow of water, via the induced fracture network, into the active mine workings. If the amount of water flowing into the mine cannot be managed by drainage or pumping, this may cause flooding. The development of a permeable fracture zone at Shenjiazhuang Colliery, which has been modelled using FLAC-2D, is shown in Figure 4. The plastic zone is divided into three distinct areas comprising a shear stress zone (shown in red), an initial stress zone (shown in green) and a tensile stress zone (shown in purple).[helen: NB - colour mentioned here] According to the characteristics of the permeable fracture zone, the strata in this zone are likely to be in a state of tension and the tensile stresses in the overlying strata correspond to the geometry of the permeable fracture zone. The results from the FLAC-2D model show that the largest concentration of tensile stresses occurs near the uppermost, right side of the model, which corresponds to a maximum height of 100 m above the roof of the mine. Simulation analysis using ANSYS-2D For the ANSYS-2D simulation, the mechanical properties, boundary conditions and loading conditions were consistent with the above model. The simulated strain distributions of the overlying rock mass immediately following the excavation of coal seams are shown in Figure 5. Generally, strains greater than 3% may be considered as large deformation, whereas strains less than 1% may be regarded as small deformation. Strain values ranging between 1 and 3% are generally regarded as medium deformation. The model shows that the maximum height of large strain increments in the permeable fracture zone reaches m above the roof of the mine (Table 4). Numerical simulations by using FLAC-3D The study area The Liuhe Industry Limited Company (hereinafter referred to as the Liuhe Company ) located in Ci County, Hebei Province, operates a state-owned coal mine at the town of Guanta, situated on the south bank of the Yuecheng Reservoir (Fig. 1).

5 Permeable Fracture Zone 241 Fig. 4. The high plastic zone of the permeable fracture zone. The subhorizontal line indicates ground plane and tilted lines indicate coal seam; the horizontal range of coal extraction is from 830 to 2150 m. Fig. 5. The principal strain of the permeable fracture zone. The colours (tints in print version) indicate different values of the principal strain. Table 4. ANSYS-2D simulation results for the maximum height of the permeable fracture zone Strain increment (%) Maximum height of permeable fracture zone (m) > > > > > Mechanical parameters and calculations The geological conditions and mechanical properties at the stateowned coal mine of the Liuhe Company are similar to those described above for Shenjiazhuang Colliery (Table 3). The model developed was 5500 m wide, 1300 m deep and m long, and it was subdivided into elements. The loading conditions and boundary conditions were consistent with those used for Shenjiazhuang Colliery. Simulation results using FLAC-3D Following the excavation of the coal seams, the results of the model for the displacement vectors and calculated strain increment in the overlying rock mass are provided in Figure 6 and Figure 7, respectively. The zone with large strain increments in the overlying rock mass is curved. In areas where the curvature is large, the deformation of the rock mass is also large, and tensile fissures at right angles to the bedding plane can easily develop. In this model, the maximum height of the permeable fracture zone is 85 m.

6 242 X. WU ET AL. Fig. 6. Horizontal displacement vector diagram of surface collapse after coal mining by the Liuhe Company. Fig. 7. The high strain increment diagram of the permeable fracture zone in the area of the Liuhe Company. Comparison of field observations and empirical calculations at Shenjiazhuang From 1987 to 1989, the maximum height of the permeable fracture zone was measured at two locations at Shenjiazhuang Colliery. This was achieved by observations of drilling fluid losses from boreholes and was shown to be 42.3 and 50.5 m above the roof of the coal seam (Fig. 8). In October 2002, further observations were made at three drilling sites and the height of the permeable fracture zone was measured at 63.6, 42.1 and 52.1 m. These observations were made using double-ended bulkhead equipment and inclined drilling during fully mechanized sublevel caving. The results of water quality analysis during sublevel caving showed minewater compositions to vary from Na-HCO 3 type in the VII and VIII 1 bedrock aquifers to Ca-Na-HCO 3 type in the overlying VIII 2 aquifer. The groundwater discharges observed flowing

7 Permeable Fracture Zone 243 Fig. 8. Schematic diagram to show the location of boreholes drilled to perform leakage tests and monitoring work. Fig. 9. Schematic cross-section showing groundwater quality variations in the aquifer overlying the No. 2 coal seam. into the mine were Na-HCO 3 type. Therefore, we can infer that the height of the permeable fracture zone reached the VII and VIII 1 aquifers, but not the VIII 2 aquifer. According to the drilling results the distances from the VIII 1 and VIII 2 aquifers to the roof of the No. 2 coal seam were 43 m and 81 m, respectively, and the thickness of the VIII 1 aquifer was 5 m. The height of the permeable fracture zone ranges between 48 and 81 m, as shown in Figure 9. Two formulae for calculating the development height of the permeable fracture zone have been provided by the State Bureau of Coal Industry (2000), as follows: 100 M H = ± M H = 20 M + 10 where ΣM is the total thickness of the coal seam. Using these two equations, the height of the water flow fracture zone was calculated to be 46.3 m and 51.0 m, respectively. The results listed above show that during mining, the maximum height of the permeable fracture zone was calculated to be 58.5 m, whereas during fully mechanized sublevel caving the maximum height of the permeable fracture zone was found to be 81 m using water quality analysis and 100 m using FLAC-2D simulation models. Comparison of field observations and empirical calculations at Liuhe In September 1998, the maximum height of the permeable fracture zone was measured at two sites at Lihue by the monitoring of drilling fluid losses and the maximum height was found to be 54.0 m. Furthermore, the results of water quality analysis showed that the height of the permeable fracture zone ranged between 52.6 and 84.2 m during fully mechanized sublevel caving. The height of the permeable fracture zone was calculated to be 46.3 m and 51.0 m, respectively, according to the two formulae set out by the State Bureau of Coal Industry (2000). The results showed that during longwall mining the maximum height of the permeable fracture zone was calculated to be 65.8 m,

8 244 X. WU ET AL. whereas during fully mechanized sublevel caving and by the analysis of water quality the maximum height of the permeable fracture zone was found to be 84.2 m and 85 m respectively using FLAC-3D simulation. Analysis of the numerical simulation results The results from the drilling observations, the analysis of the chemical composition of mine water and numerical modelling have shown that the maximum height of the permeable fracture zone is largest when using FLAC-2D and FLAC-3D. The reasons for these findings are as follows. (1) Sublevel caving has a greater zone of influence on the overlying strata compared with longwall mining, which has also been verified by in situ drilling observations. (2) The modelling and calculations assume that total extraction has taken place. However, any support pillars used in practice may also influence the height of the permeable fracture zone. (3) The numerical simulation results from FLAC-2D and FLAC- 3D are applicable when all of the coal seams are mined. However, the in situ observations were conducted when only some of the coal seams had been extracted. (4) The boundary conditions for the 3D model are more complicated than those of the 2D model and therefore the simulated height of the permeable fracture zone may be larger when using FLAC-2D rather than FLAC-3D. The results of the three methods demonstrate that numerical simulation is a useful way to simulate the height of the permeable fracture zones. In ANSYS-2D, the permeable fracture zone is calculated from the principal strain, whereas FLAC-2D uses the range of plastic zones and FLAC-3D utilizes the strain increment (curvature). Conclusions The permeable fracture zone in the strata overlying a coal mine is considered in this paper to be represented by the caved zone and the fracture zone. Both of these zones underlie the subsidence zone and may have enhanced porosity and permeability. When coal is mined beneath reservoirs, aquifers or flooded abandoned mine workings this increased permeability may initiate or exacerbate the flow of groundwater into underlying mine workings. It is therefore desirable to determine the geometry of the permeable fracture zone and in particular the height above the roof of a mine to which the fractures may migrate. To calculate the height of the permeable fracture zone numerical modelling was used. ANSYS-2D uses principal strains, FLAC-2D uses plasticity zone and FLAC-3D uses strain increments. The model parameters chosen were based on extrapolation from similar geological and mining settings. In general, the results bear a similarity to the observations of the height of the permeable fracture zone that was determined based on borehole monitoring and the loss of drilling fluids into fractured strata. Therefore, a combination of numerical analysis with drilling observations provides a method to determine the height of the permeable fracture zone above coal mines in China. It is recommended that when mining takes place beneath reservoirs, aquifers, flooded workings, rivers or the sea a combination of numerical modelling and field testing may assist in determining the height on the permeable fracture zone. Acknowledgements. This work was supported by National Nature Science Foundation of China (Grant No ), Beijing Nature Science Foundation (Grant No ) and Program for New Century Excellent Talents in University (NCET ). We thank H. Floyd-Walker, L. Donnelly and Z. Lu for their work on our paper. References Bai, Y.J Causes analysis and control technology of water disaster in coal mine. Coal Technology, 11, Bower, K.M. & Zyvoloski, G A numerical model for thermo-hydromechanical coupling in fractured rock. International Journal of Rock Mechanics and Mining Science, 34, Chen, J. & Li, Q.S Research on status of coal mining technology under constructions and water bodies. Coal Technology, 12, Cheng, X.F., Liu, S.D. & Liu, D.X Sound-wave CT detection for failure patterns of surrounding rock after mining. Journal of China Coal Society, 26, Cui, C.X Feasibility analysis and practice of full-mechanized caving mining under sea. Coal Mining Technology, 13, Cui, D.R. & Yao, Q.H Development of special mining technology. Ground Pressure and Strata Control, 2, Fan, W.T The situation and developing trend of energy. Shanxi Energy and Conservation, 34, 1 7. Gao, J.H., Kang, T.H., Jin, Z.M. & Zheng, T.B Similar simulation study on crack development law of surrounding rocks of roadway with great-thick thin-bedded top roof. Chinese Journal of Rock Mechanics and Engineering, 23, Gu, Z.D. & Shen, H.M Coal mining under the sea. Mine Survey, 2, Han, X.S., Zhang, J.K., Liu, Z.X. & Lin, Z.T The identifying method on sound wave imaging of crack abscission layer and caving zones. Coal Geology of China, 13, Hatherly, P., Luo, X., McKavanagh, B., Dixon, R. & Devey, D Seismic monitoring for mapping longwall caving at Gordonstone Mine. In: Follington, I. L., Beeston, J. W. & Hamilton, L. H. (eds) Proceedings of the Bowen Basin Symposium. Geological Society of Australia, Coal Geology Group, Brisbare, Hu, C. & Zhao, J.M Analysis of the safety and technical measures for coal mining under water body. Science and Technology Information, 10, 93. Hu, Y.Q., Zhao, Y.S., Yang, D. & Duan, K.L Study of 3D simulation on breakage for coal floor in mining above aquifer. Chinese Journal of Rock Mechanics and Engineering, 22, Lu, M.S., Kong, F.S. & Zhuang, X.H Comprehensive environmental geological survey of the Nansi Lake drainage area, Southwestern Shandong. Geology in China, 30, Minkoff, S.E., Stone, C.M., Bryant, S. & Peszynsk, M Coupled fluid flow and geomechanical deformation modeling. Journal of Petroleum Science and Engineering, 38, National Coal Corporation of China A compilation of mine water disaster accident. Coal Science and Technology Publishing House, Beijing. Peng, S.P. & Meng, Z.P Theory and Practice of Mining Engineering Geology. Geological Publishing House, Beijing. Rutqvist, J., Wu, Y.S., Tsang, C.F. & Bodvarsson, G A modeling approach for analysis of coupled multiphase fluid flow, heat transfer, and deformation in fractured porous rock. International Journal of Rock Mechanics and Mining Sciences, 39, Song, J.S Analysis and distinguishing of water conducted zone height above roof in Qidong coalmine. Coal Geology of China, 17, State Bureau of Coal Industry Regulations for setting coal pillar and mining pressed coal under buildings, water bodies, railways, and main roadway. China Coal Industry Publishing House, Beijing. Wang, D.W., Gao, Z.P., Zhang, D.L. & Yang, N Feasibility of mining under water body in planning area. Coal Mining Technology, 15, Wang, E.Y Influences of Xiaolangdi Reservoir water storage on mining activity of adjacent coal mine. China Mining Magazine, 11, Wu, Q. & Chen, D.L The environmental engineering geological evaluation of Xiaolangdi Reservoir on the Yellow River. Soil Engineering and Foundation, 17, Wu, X., Yang, J., Duan, Q.W. & Wang, J.J Impact of coal mining on safety of reservoir. Journal of Hydraulic Engineering, 9, Wu, X., Wang, X.G., et al A study on coal mining under large reservoir areas. Environmental Geology, 57, Wu, X.J.J., Yu, Q.C., et al Exploitation of coal resources under surface water body. Chinese Journal of Rock Mechanics and Engineering, 26, Zhang, H.X On the technology of coal mining under buildings, railways and water body in future. Coal Mining Technology, 16, 1 3. Zhao, S.Q, Wu, Q., Yin, S.X Mechanism analysis of water inrush in Daxing coal mine. Journal of China Coal Society, 5, Zhou, D The research on energy consumption and energy-saving strategy of China. Journal of Lanzhou Commercial College, 25, Received 26 January 2010; accepted 17 January 2013.