Interactions of overburden failure zones due to multiple-seam mining using longwall caving

The Cuizhuang Coalmine is located in the Weishan county, 260 km south of the capital of the Shandong province, Jinan, in China (Fig. 1). Annual mine production was 0.6 million tons from 1998 to 2002 and its annual output has doubled since 2003. The upper and lower seams of No. 3 are the productive coal seams in the Permian Shanxi Formation. The mine was developed by using two vertical shafts and a single mining level at an elevation (EL.) of −250 m below the mean sea level (msl). The Cuizhuang Coalmine has an area of approximately 11.9 km², with a length of 4.2 km in the east–west direction and a width of 2.5–3.3 km in the north–south direction. About two-thirds of the total geological coal reserves were deposited under Weishan Lake, the largest freshwater lake in northern China with an area of 1,266 km², an average water depth between 1.5 and 3 m, a maximum reservoir of 4.73 billion m³ and a mean water level of EL +32.23 to +34.21 m.

Table 1 lists the stratigraphic and hydrogeological characteristics of the Cuizhuang Coalmine. The coal-bearing strata in the mine, the Carboniferous and Permian Systems, are typical coal-measures in the North China coal basin. The Permian Shanxi Formation bears 2–3 coal seams, including the upper and the lower seams of No. 3. The upper seam has a thickness of 0–6.57 m and the lower seam has a thickness of 0–5.17 m, where zero means sedimentary absence in some of the locations. The Shanxi Formation mainly consists of fine sandstone, mudstone, silty mudstone, siltstone.

The total water inflow in the Cuizhuang Coalmine has been stable with a maximum value of 40 m³/h, mainly from sandstone aquifers on the roof. The hydraulic relationships between the rock and soil aquifers are separated by several aquifuge layers. However, when coal mining takes place under the Neogene aquifers, the height of the fractured water flow zone is carefully controlled to avoid water and/or sand inrush from unconsolidated aquifers. Therefore, an accurate prediction for the height of overburden failure has become one of the critical issues for risk assessment and mining safety, especially in multiple seam mining.

The Cuizhuang Coalmine is geologically located on the southern rim of the Tengxian anticline hydrogeological setting, which tectonically belongs to the eastern margin of the Western Shandong Block of the North China Platform.

The geological structure of the mine is generally a monocline structure with a set of secondary gently undulating folds, varying in occurrence along the strike and dip directions (Fig. 2). The Cuizhuang anticline and the Chaili synclines are the main structures, accompanied by several normal faults. There are six large faults in the mine, which are boundary faults except for the Tiangang fault and SF1 fault across the mine. Among them, two are hydraulically permeable, two are impermeable, according to groundwater tests, and the remaining two have not been identified. The throw of the Tiangang fault is between 80 and 180 m, extending to the southwest beneath Weishan Lake. It is covered by the Neogene System with a bottom clay 5.31–23.18 m thick. The bottom clay layer protects the coal mine from water seepage from sand aquifers along the faults in the coalmine.

Figure 2 also shows the four main study panels of this paper. In-situ measurements of overburden failure were conducted in Panel 3301S and Panel 3302S and results were used to verify the scale models and numerical models. The scale model test prototypes are from geological cross-sections in Panel 3302 and are shown in Fig. 3. These four panels have similar geological and mining conditions. Figure 3 is an illustration of the cross-sections (section lines A–A′, B–B′ and C–C′ are shown in Fig. 2) examined for overburden failure and Table 2 lists the geotechnical properties of typical geological engineering types in the Permian Shanxi Formation. Mudstones in the geological column occupy about one-third of the overburden of the No. 3 upper seam. Mudstones, sandy mudstones or siltstone on the roof can be easily damaged, resulting in roof cave-ins that fill the mined-out area. The overburden strength varies from soft to medium; nevertheless, the strength of the rocks adjacent to the unconsolidated Neogene layers is significantly decreased because of strong weathering. This kind of overburden structure is helpful for resisting the development of fractured water flow zones due to mining. The thickness of the overburden of the No. 3 upper seam varies from 36.4 m to more than 200 m. The bedrock surface underlying the Neogene sediments is between −100 and −70 m msl with some topographical fluctuations in the coal mining area.

In-situ measurements have been one of the most important and direct methods for the investigation of the height of overburden failure zones. In this traditional measurement method, water loss is investigated by the drilling of boreholes. Currently, underground water pumping, ultrasonic imaging and other geophysical approaches have been adopted in overburden failure measurement. Figure 4 shows the principles of ground and underground measurement methods using boreholes.

A so-called two-terminal plugging tool was used in the Cuizhuang Coalmine to measure water loss in different boreholes to determine the height of the overburden failure zone. The boreholes were designed and drilled before and after mining. The borehole was separated into 0.1 m or 0.2 m sections. The ends of each section were plugged with an expansion rubber inside the borehole. High pressure water was injected into the isolated section and water loss was measured prior to and after coal mining. The height of the fractured water flow zone was determined by comparing water loss along the boreholes pre- and post-excavation (Xu and Sui 2013).

Scale model testing is one of the common methods used in the research of overburden failure. In this study, the dimensions of the scale model test rig are 6 m × 3 m × 0.4 m (length × height × thickness). The model parameter scales are as follows: (1) 1:200 for geometry; (2) 1:1.7 for gravity; and (3) 1:200 for uniaxial compressive strength (UCS). Materials used for modeling comprised a combination of sand, barite, gypsum, mica, sawdust, etc. Mica powder was well distributed between the rock layers to simulate formation beddings. A distributed load was applied to the scale model to compensate for gravitational pressure of the overburden.

Table 3 lists the four types of scale modeling conducted in this paper to simulate different mining and geological conditions. The major differences among them are found in the cutting height and the interburden thickness. Scale Model (SM) No. 1 simulates Panels No. 3301S_upper and 3301S_lower to verify the scale modeling through a comparison of the height of the overburden failure zone with that from the in situ measurements. SM No. 2 is basically the same as SM No. 1 except that the interburden is changed to 14 m to study the influence of the interburden thickness on the height of the overburden failure zone. The other two models, SMs No. 3 and 4, are designed to simulate fracture propagation due to mining in the close-distance seams in Panel 3302 along the dip and strike cross sections (illustrated in Fig. 5) that correspond to those in Fig. 3. In SMs No. 3 and 4, the cutting height of the upper and the lower seams is 6.0 and 4.3 m, respectively; the thickness of the interburden is 10.6 m. Excavation was carried out in the mining entrance from left to right. Every day, a 4-cm seam (in model size) of coal was mined out; this equals 8 m in the prototype. The lower seam was mined by using the same process after excavation of the upper seam was completed. The displacement of the strata during the mining simulation was measured by using linear variable differential transformers (LVDTs). For every step of the excavation, the deformation and failure of the overburden were recorded and images were taken with a high speed camera. Then, the movement and deformation due to mining were analyzed by digital image processing.

Numerical modeling needs to be verified and validated against (1) the overburden failure measurement in the study area and (2) the overburden failure induced by fracture propagation in the scale experiments before it can be used to study multiple seam interactions in mining. It was found that the location of a sudden change of vertical displacement, which shows a concentration of vertical displacement isolines, is in good agreement with the height of the caving zone; the height of the zone with continuous tensile cracks is in an agreement with the height of the fractured water flow zone. Therefore, the propagation of tensile and shear failures, and the distribution of displacement can be used to evaluate the heights of the caving and fractured water flow zones.

A Fast Lagrangian Analysis of Continua (FLAC 2D) code (Itasca 2009) was used in this study. Boundary conditions for the plane strain models are free displacement and loading on the top, fixed at the bottom, fixed in the horizontal direction and free in the vertical direction on both side boundaries. Parameters of the different geological engineering types were carefully determined by using laboratory tests and geological observations in accordance with the Hoek–Brown criteria (Hoek and Brown 1997). Extraction of coal was conducted by the function of “model null” in the code. The maximum convergence between the roof and floor was limited to less than the cutting height of the seam.

Two panels of the Cuizhuang Coalmine were selected for overburden measurements: Panels No. 3301S_upper and 3301S_lower (locations shown in Fig. 2 and Panel No. 3301S denoted). The upper and the lower seams of No. 3 were separated by a 7 m thick interburden composed of grey, thin-layered silty mudstone and siltstone. Significant amounts of fossilized plant fragments were found along the bedding plane that grew on the horizontal bedding on the interburden, meaning that caving easily took place and filled the mine gob after excavation of the lower seam. The dip angle of the strata above the panels was 8° on average and varied between 0° and 11°.

The above-mentioned water loss method was used to measure overburden failure induced by upper seam mining with LTCC and carried out on Panel No. 3301S_upper with 3 underground boreholes. The measurements were completed on July 2007. Results showed the maximum height of the caving zone is 19.6 m; the maximum height of the fractured water flow zone is 59.59 m above the roof of the upper seam. The height ratio of the caving and fractured water flow zones against the cutting height of 5.0 m is 3.9 and 11.9, respectively. Measurements for the lower seam were conducted on Panel No. 3301S_lower just beneath Panel No. 3301S_upper. Measurements were completed in November 2008 and shortly thereafter mining was stopped on the panel. The maximum height of the fractured water flow zone after mining the 4.0 m lower seam was 66.0 m above the roof of the upper seam. An equivalent cutting height thickness of the upper and lower seams can be calculated by using Formula (1) in accordance with a code from the China State Bureau of Coal Industry (2000), which is 7.2 m and the ratio of the height of the fractured water flow zone against the equivalent cutting height of 7.2 m is 9.2.where M is the equivalent thickness of the upper and lower seams; M _upper and M _lower are the cutting height of the upper and lower seams, respectively; h _interburden is the thickness of the interburden; y _lower is the ratio of the height of the caving zone to the thickness of the lower seam; this study assumes that it has the same value of 3.9 as the upper seam does.

Therefore, predicting the height of the fractured water flow zone H _f for mining the upper and the lower seams of No. 3 in the Cuizhuang Coalmine can be calculated by using the formula H _f = 9.2 M that takes interactions into account with a rule of thumb. Here, please note that the heights of overburden failure are calculated from the roof of the upper seam.

For verification purposes, Table 4 shows a comparison of the physical and numerical simulations based on in situ measurements, numerical simulations and scale modeling. The results show the interactions lead to the superposition of overburden failure due to the mining of two close distance seams. The height of the fractured water flow zone is increased from 52.1 to 68.0 m because of superposition failure due to excavation of the upper and lower seams with an interburden thickness of 7 m, see SM No. 1. The results in Table 4 verify the feasibility of using numerical simulation and scale modeling to study overburden failure, caving and fracturing interactions in multiple-seam mining. The results from SM No. 2 show that the height of the fractured water flow zones is increased from 52.1 to 64.8 m above the roof of the upper seam, thus indicating the influence of the interburden thickness when it is changed to 14 m. These results indicate there is still interaction of rock caving and fracturing in close-seam mining when the interburden thickness reaches 14 m, but the interaction increases when the interburden thickness is 7 m.

Figure 6a–d presents pictures showing the performance of selected stages during and after mining of the upper seam of No. 3 from SM No. 3. The development of caving and fractured water flow zones can be interpreted through a detailed analysis of the mining process in a stepwise consecutive manner. The first caving occurred when panel length was advanced to 24 m (the prototype size, same below in this paper). After that, the roof periodically cracked and the height of the caving and fractured water flow zones gradually increased until rocks filled the mined out area behind the working face. Rocks almost fully filled the mined out area when the recovery length reached 84 m. After mining, the height of the caving and fractured water flow zones reached 17.6 and 48.0 m, respectively. The height of the fractured water flow zone was expected to increase if recovery of coal continued. The outline of the fractured water flow zone has the shape of an asymmetric ladder.

Figure 6e–h show development of the overburden failure zone during mining of the lower seam of No. 3 from SM No. 3. The interburden between the upper and lower seams did not cave until the distance of coal recovery reached 44 m; the fractured water flow zone also did not develop further until this length. The caving zone of the upper and lower seams merged into one another when the mining length was advanced to 56 m, and the height of the fractured water flow zone continuously increased until its ladder-shaped outline reached the bottom clay layer.

Figure 7 shows caving height variations and fractured water flow zones with coal recovery in the coal seam. Caving zone heights reaches the maximum value first, and then the fractured water flow zone reaches its highest position with a longer recovery distance, depending on the cutting height and overburden features. The maximum height of both zones is 27.0 and 69.0 m, respectively, above the upper seam after the excavation of the upper and lower seams of No. 3 along the strike direction.

Figure 8a–d show development of caving and fractured water flow zones due to the recovery of coal from the upper seam along the dip direction from SM No. 4. There is little caving and fracturing until the excavation length reached 28 m. The first caving occurred when the panel length was increased to 32 m and then the roof periodically cracked. Furthermore, the height of both zones gradually increased. When rocks completely filled the mined out area, the propagation of the caving zone stopped, while the height of the fractured water flow zone continued to slowly increase until a length of about 120 m. After mining, the heights of the caving zone and fractured water flow zone were 16.8 and 53.5 m, respectively.

Figures 8e–h and 9 show the overburden failure zone during mining of the lower seam of No. 3 from SM No. 4 after excavation of the upper seam. The interburden between the upper and lower seams did not cave until coal recovery reached a distance of 44 m and the fractured water flow zone also did not further increase. The caving zones of the upper and lower seams merged into one another when the mining length advanced to 56 m, and the height of the fractured water flow zone continuously increased until its ladder-shaped outline reached the bottom clay layer. The maximum height of the zones reached 25.1 and 68.5 m, respectively, above the upper seam after excavation was completed on the upper and lower seams along the dip direction.

Figure 10 shows a simplified numerical model. The model is constructed in accordance with the cross-section of a prototype along the dip of Panel 3302S to simulate and compare the displacements, stresses and failures after coal recovery from the upper and lower seams. The length and width of the 2D model are 1068 and 167 m, respectively. The overburden is divided into 15 groups according to their geological engineering and geotechnical properties. Figure 11 shows comparisons of the vertical displacements, maximum and minimum principal stresses and tension and plastic zones along the strike directions. Figures on the left column of Fig. 11 are results after recovery of the upper seam of No. 3 and the right column is results for the lower seam of No. 3. Results indicate obvious interaction and superposition of the close-seam mining. For example, a cutting height of 6.0 and 4.3 m is modeled for the upper and lower seams, respectively, with an interburden thickness of 10.0 m. After excavation of the upper seam, the height of the caving and fractured water flow zones reached 22 and 54 m, respectively, and increased to 28.5 and 65.4 m in response to excavation of the lower seam. The results also show obvious interactions of close-seam mining on overburden failure and stress distributions.

Table 5 lists the results of the other 18 numerical models to investigate the influence of the ratio of the overburden thickness (h) to the cutting height (M) of the lower seam, i.e., h/M, on the interactions of overburden failure. These models have the same cutting height of the upper seam (5.3 m), varying cutting heights (2.5, 4.3 and 5.3 m) of the lower seam and varying thicknesses of the interburden (5.0, 7.0, 14.0, 17.0, 19.5 and 22.5 m).

Figure 12 shows the interactions of overburden failure influenced by the interburden thickness and the cutting height of the seams. Figure 12a shows the results of NMs No. 1–6 with varying interburden thicknesses between 5.0 and 22.5 m, a cutting height of 5.3 m for the upper seam, and a cutting height of 2.5 m for the lower seam. The height of the caving zone slightly increased with increases in the ratio h/M until it reached 6.8 with a corresponding interburden thickness of 17.0 m. This means that when h/M is greater than 6.8, the interactions of the caving zone induced by the recovery of coal in the lower seam and the recovery of the upper seam can be neglected without obvious calculation errors. The interactions of the fractured water flow zones induced by mining of the two seams showed the same performance; that is to say, when h/M is greater than 6.8, the interactions could be neglected and the mining of the lower seam barely increases the height of the fractured water flow zone induced by mining of the upper seam. Therefore, the height of the caving zone can be calculated in terms of single-seam excavation when the h/M ratio is greater than 6.8 in this type of geological and mining condition, while the height of the fractured water flow zone remains almost the same as that induced by upper seam mining. Figure 12b shows the results of NMs No. 7–12 for a cutting height of 4.3 m for the lower seam and the same conditions mentioned above. In this case, the turning point of h/M is 4.5 for the interactions of the caving zone and 3.9 for the fractured water flow zone. Moreover, Fig. 12c shows results for NMs No. 13–18; the h/M of the interactions in mining the upper and lower seams at a cutting height of 5.3 m is 3.7 and 3.2 for the caving and water flow fracture zones, respectively.

Figure 13a summarizes some of the results of this paper and those from the literature (Zhang 2010; Guo 2011; Xu et al. 2013; Luan et al. 2010) to illustrate the interactions of overburden failure due to recovery of coal from multiple seams, with varying interburden thicknesses and cutting heights of the lower seam. If a point that represents (M, h/M) falls under Line D, interaction or superposition should be taken into account to calculate the height of the overburden failure (Fig. 13b). Otherwise, the interaction can be neglected when the point drops above Line D; therefore, the height of the caving zone can be calculated by treating it single-seam mining, while the height of the fractured water flow zone will not be surpassed due to mining of the upper seam (Fig. 13c). Line D may be defined as the dividing line that determines whether interaction causes overburden failure superposition in close-seam mining.

The results of this paper could provide a helpful reference in the recovery of coal in close-distance seams by using the LTCC method. However, the interactions presented in this paper still need further investigation. An orthogonal array can be adopted to investigate the influence of factors (overburden and interburden geotechnical properties, cutting heights of multiple seams) on the interactions of multiple-seam mining in further research. Multiple-seam mining with more than two seams of different geological engineering structure, discontinuities, geo-stress environments and dip angles, especially for mining of steeply dipping seams, will also be further considered. Moreover, coupling the seepage, failure and stress fields should be taken into account for an intrinsic understanding of the mechanism of multiple mining interactions. Results need to be validated by more in situ measurements.