The Properties of a Coal Body and Prediction of Compound Coal-Rock Dynamic Disasters

School of Energy Engineering, Xi’an University of Science and Technology, Xi’an 710054, China !e State Key Laboratory of Coal Resources and Safe Mining (CUMT), Xuzhou 221116, China Key Laboratory of Western Mine Exploitation and Hazard Prevention of the Ministry of Education, Xi’an 710054, China China Coal Technology and Engineering Group Corp Chongqing Research Institute, Chongqing 400039, China National Key Laboratory of Gas Disaster Monitoring, Preventing and Emergency Controlling, Chongqing 400039, China


Introduction
Due to the prevailing demand for energy resources, the coalbased energy pattern of China will not change in the foreseeable future. e proportion of coal in the energy system will remain at around 50% in the long-term, and 80% of coal is mined from a deeper underground, especially in mideastern China. Due to deep mining, the features of coalrock dynamic disasters become more complicated and diverse; thereby, a new compound coal-rock dynamic disaster occurs. Besides, in some middeep regions with abnormal gas geology, new compound coal-rock dynamic disasters also occur.
is new compound coal-rock dynamic disaster presents a dual dynamic feature combining bump proneness and outburst risk; that is, the disaster exhibits the features of both typical coal-gas outburst and rock burst ( Figure 1).
In recent years, some of China's coal mines such as those in the Pingdingshan mining areas, Huainan mining areas, Xinji mining areas, Fengcheng mining areas, Fuxin mining areas, and Fushun mining areas took the lead in deep mining, where some compound coal-rock dynamic disasters already occurred or are likely to occur.
At present, many scholars have studied the mechanism of compound coal-rock dynamic disasters, their prediction, early warning, and prevention: Petukhov [1], Zhang et al. [2], He et al. [3], Li et al. [4,5], Li [6], Zhang et al. [7], Pan [8], Yin et al. [9], Dou et al. [10], and Dong et al. [11] have studied the occurrence mechanism of compound coal-rock dynamic disaster. Zhang and Li [12], Zhen et al. [13], Li et al. [14], Yuan [15], Li et al. [16], Qi et al. [17], Pan [8], Yin et al. [9], and Dou et al. [10] studied the occurrence conditions or types of compound coal-rock dynamic disasters. Zhang [18], Yuan [15], Jianwei [19], Jiang et al. [20], Dong et al. [11], Sheng et al. [21], and Ouyang et al. [22] have studied the prediction, early warning, and evaluation of compound coal-rock dynamic disasters. Qi et al. [23] investigated the theory and technical framework of multiscale and multisource prevention and control of coal-rock dynamic disasters in deep coal mines. e prediction and early warning of a compound coal-rock dynamic disaster mainly relies on the direct or indirect parameters such as the amount of drilling cuttings [5-7, 10, 11, 15, 18, 20, 21, 23, 24], the size of the drilling cuttings [8,23], the temperature of the drilling cuttings [8,23], the desorption index of the drilling cuttings [5,7,11,15,18], the initial velocity of gas emission from boreholes [5,7,10,11,15,18], the stress [7,19,20,23], the displacement [23], the gas concentration extraction index [11,20], coal temperature [8], coal charge [8], electromagnetic radiation [10], microseismic event count [10,19,20,23,25], acoustic emission event count [10,11,20,21], shock wave CT [10], and seismic wave or elastic wave CT [10,23]. However, the coal physical-mechanical characterisation and prediction method of a compound coal-rock dynamic disaster still contain many deficiencies; therefore, in this study, a new coal characterisation and prediction method of compound coal-rock dynamic disasters were studied, based on the actual occurrence data of compound coal-rock dynamic disasters. Under certain confining pressure (CP) and gas pressure (GP) conditions, the desorption features of gas at different temperatures were analysed to reveal the controlling effect of temperature on gas desorption. Under a specific confining pressure and temperature conditions, the influence of gas pressure on the coal desorption regularity was discussed. e influence of confining pressure on the gas desorption behaviour was also studied under specific pore pressure and temperature conditions. During the experiment, the Darcy steady flow method and triaxial penetration test device were used. e outlet pressure is the atmospheric pressure, while the inlet pressure is adjustable. e desorption and penetration experiments of coal samples under different gas pressures, temperatures, 2 Shock and Vibration and confining pressure conditions were conducted. e main device and experimental system used are displayed in Figure 2.

Physicomechanical Properties of Coal in a
By using a customised rock coring and detection device, the raw coal was cut into cylinders measuring Φ50 mm × 100 mm. After that, the samples were ground using an electric grinder, to meet test requirements. Experimental results and trend analyses are demonstrated in  (1) Variation regularity of raw coal-gas desorption velocity (GDV) with confining pressure e gas desorption velocity curves under different confining pressures are shown in Figure 3. During the test, the gas adsorption pressure is 2 MPa, and the temperature is 35°C. As shown in Figure 3, when the confining pressure is 6 MPa, the gas desorption velocity is greater than those under confining pressures of 10 and 12 MPa, in both the initial stage and steady stage. Meanwhile, the gas desorption velocities under confining pressures of 10 MPa and 12 MPa remain similar. As shown in Figure 4, when the confining pressure increases from 6 to 8 MPa, the gas desorption velocity (after entering the steady stage) sharply decreases from 2.72 to 1.51 cm 3 /s. Meanwhile, the gas desorption velocities at 10 MPa and 12 MPa are 1.31 cm 3 /s and 1.22 cm 3 /s, respectively. at is, under specific gas pressure and temperature conditions, the influence of confining pressure on the gas desorption velocity decreases gradually, after it increases to a certain value.
(2) Variation in raw coal-gas desorption velocity with gas adsorption pressure e gas desorption velocity curves under different gas pressures are shown in Figure 5. During the test, the confining pressure is 6 MPa, and the temperature is 35°C. As displayed in Figure 5, when the gas adsorption pressure reaches 4 MPa, the gas desorption velocity is beyond the range of the instrumentation used, so its values are not illustrated in the figure. Hence, no matter in the initial stage or the steady stage, the gas desorption velocity increases with the improvement of gas adsorption pressure.
(3) Variation in raw coal-gas desorption velocity with temperature e gas desorption velocity curves under different temperature conditions are shown in Figure 6. During the test, the gas adsorption pressure is 1 MPa, and the confining pressure is 6 MPa. As illustrated in Figure 6, in the initial stage, the raw coal-gas desorption velocity at low temperatures is higher than that at high temperatures. Meanwhile, the final desorption velocity at low temperatures is lower than that at high temperatures.

Experimental Study of the Coal-Gas Permeability in a Compound Coal-Rock Dynamic Disaster.
e variation in coal permeability (gas permeability coefficient) under different conditions was characterised. e main device and experimental system used are displayed in Figure 2.
(1) Variation in raw coal permeability with changing confining pressure e variation in permeability during continuous unloading is shown in Figures 7-10. As shown in the figure, the coal permeability gradually increases during unloading and the rate of increase also increases. In the initial unloading stage, i.e., high confining pressure stage (high burial depth), the     permeability rises slowly during the unloading of confining pressure. Meanwhile, in the mid-late stage, i.e., low confining pressure stage (low burial depth), the increase rate of coal permeability is accelerated during unloading. For example, when the adsorption pressure is 1 MPa, the rate of increase increased under confining pressures is lower than 6 MPa. Meanwhile, when the adsorption pressure is 4 MPa, the rate of increase rate is increased under confining pressures is lower than 8 MPa. In a word, the variation in permeability is closely related to the initial confining pressure (burial depth or tectonic stress) and gas pressure.    According to the above analysis, when the adsorption pressure and temperature are constant, upon the reduction of confining pressure, the permeability increases gradually; in the initial stage of unloading, i.e., the high confining pressure stage, the increasing amplitude of permeability is relatively small; meanwhile, in the later stage of unloading, i.e., the low confining pressure stage, the rate of increase of permeability is increased; the gas flow changes in a way similar to the permeability.
(2) Variation in raw coal permeability with gas pressure changing e permeability of the same sample under different gas adsorption pressures was characterised.
e results indicate that, with the increase in gas adsorption pressure, the permeability decreases. On the contrary, the gas desorption velocity; i.e., the gas flow increases accordingly (Figures 11 and 12).
(3) Variation in raw coal permeability with changing temperature  show the variation between the sample permeability and temperature under different confining pressures when the gas adsorption pressure is 1 MPa and 4 MPa, respectively. As shown, the coal permeability increases with increasing temperature because as the temperature rises, the activity and internal energy of gas molecules increase. Some of the adsorbed gas is desorbed. e seepage channels expand. e diffusion of gas molecules is accelerated; thus, the permeability of coal gas increases accordingly. Under high confining pressure conditions, the increase in permeability with temperature is relatively small yet becomes more significant at low confining pressure. In particular, when the gas adsorption pressure is 4 MPa, the seepage flow velocity exceeds the measuring range of equipment and cannot be precisely determined. According to the above analysis, when the adsorption pressure and confining pressure are constant, with the increase in temperature, the coal permeability rises slightly, but the variation is smooth (Figures 13 and 14); under high confining pressure, the increase in permeability as the temperature rises is relatively small (a trend that strengthens under low confining pressure conditions, as shown in Figure 14); when the gas adsorption pressure is 4 MPa, the seepage flow velocity and permeability (confining pressure of 6 MPa and temperatures of 35°C and 40°C) exceed the measuring range of equipment and cannot be accurately measured. Meanwhile, Figure 15 indicates that the variation in gas flow is similar to that of the permeability.

Coal Mechanical Properties in a Compound Coal-Rock
Dynamic Disaster. By using gas-tight equipment [26], the approximate-uniaxial test method for the characterisation of gas-containing coal mechanical properties was developed. Based on experiment, the influence of gas on the coal mechanical properties and mechanical response under zero effective confining pressure condition was obtained.

Stress-Strain Curves of Coal Samples under Uniaxial
Compression. In uniaxial compression tests, the gas pressures were set to 0 MPa, 1 MPa, and 2 MPa, respectively. Under each condition, at least three samples were characterised. e experimentally obtained stress-strain curves and related data are displayed in Figures 16 and 17. Figure 16 shows the uniaxial compression stress-strain curve of coal samples without gas being present. As shown in the figure, although the samples are cut from the same coal block, the uniformity of bedding direction is ensured, so some discreteness of measured data is unavoidable. Figure 17 demonstrates the stress-strain curves of coal samples uniaxially compressed under various gas pressure conditions. e gas in pores and fissures not only changes the mechanical properties and mechanical response of the coal body in the plastic deformation stage and after peak strength deformation stage but also affects the elastic deformation of the coal body; i.e., mechanical properties such as elastic modulus no longer remain constant. In other words, the gas pressure changes the mechanical response features of coal in all deformation stages, including its elastic deformation stage.

Influence of Adsorbed Gas on Coal Mechanical
Properties. Within the coal, gas exists in the free and adsorbed states. Under specific temperature and pressure conditions, a dynamic equilibrium is reached. e physicomechanical properties of coal containing gas are affected by these two gas states. e free gas affects the mechanical properties of coal via pore pressure which affects the volumetric strain. Meanwhile, adsorbed gas changes the mechanical properties and mechanical response of coal via the adsorption and desorption processes, which further affect the constitutive relationship of the coal.
(1) Elastic modulus (EM) decreases with the increase in gas pressure As shown in Figure 18, the elastic modulus decreases with the increase in gas pressure. Gas molecules adsorbing on the coal particle surface and interparticle space will reduce the cohesion of the coal samples, so the macroscopic deformation resistance of coal sample is decreased; i.e., the elastic modulus is reduced. (2) Plastic deformation at the peak point decreases with the increase in gas pressure    Shock and Vibration As shown in Figure 19, at the peak point stage, the plastic deformation of coal sample gradually decreases with the increase in gas pressure.

(3) Influence of gas on the bump proneness index
Experimental results of coal bump proneness index under different gas pressures are listed in Table 1.
As shown in Figure 20, the dynamic fracture time, i.e., the time required for the occurrence of brittle fracture, increases with the increase in gas pressure. e bump energy index and elastic energy index both decrease with the increase in gas pressure. Comparing to coal body skeleton frame bearing load alone, the coal body skeleton frame and gas bear the external load together after gas adsorption. us, the elastic deformation energy accumulated within the coal body is significantly reduced, making it difficult for the formation of coal body bump burst. Besides, according to the method used for calculation of bump energy index and elastic energy index, the reduction of uniaxial compressive strength and elastic modulus will result in the reduction of bump energy index and elastic energy index with the increase in gas pressure, as shown in Figures 21 and 22. Based on above analyses, as the mine gets deeper, the gas extraction is likely to increase the bump proneness of coal with the increase in coal strain energy. us, the risk of compound coal-rock dynamic disaster is increased. (4) Brittleness decreases with the increase in gas pressure As shown in Figures 16 and 17, the presence of gas makes the postpeak curve more shallow; in particular, the brittle fracture, that stage with a sharp instant reduction in strength, is less significant. is is mainly attributed to the existence of gas increasing the ductility of coal and reduces the possibility of brittle fracture. As shown in Figure 23, after the fracture of samples without gas adsorption effect, the sample fractures into a relatively complete split line; meanwhile, for samples affected by gas adsorption, they fragment, exhibiting poor completeness after fracture. Also, it can also be found that the trend in coal mechanical properties and mechanical response slows as the gas pressure increases. In other words, under the uniaxial compression condition with zero confining pressure, the influence of gas on the coal mechanical properties and mechanical response gradually diminishes when the gas adsorption reaches saturation. Gas extraction will increase the risk of compound a coal-rock dynamic disaster.  Figure 24. Contorted fold structures are folds transiting from normal incline to inverted incline or compress shear and noncylindrical folds caused by trend transition and increasing dip angle. According to the formation area and morphological features of contorted fold structures, the fold structures can be classified into reverse or kinked structures.

Prediction Used in a Case Study of a Compound Coal-Rock Dynamic Disaster
e Heiqiyan contorted fold band in Donglin Mine is distributed approximately north to south, with a total length of 4 km, and Heiqiyan Matong marks the transition point. Its south trend is 1.5 km. e deep stratum trend transits into S30°W, and the incline is W (i.e., the east wing of the Ganjiaping syncline); the band trend between the north of Matong and boundary is 2.5 km. e stratum overturns into an east incline of 30°. Its shallow seam trend is N10°W, while the deep seam trend is N10°E. e ground elevation is +310 m, and the production level is -100 m. e initial crustal stress is 56.3 MPa. e initial gas pressure is 1.8 MPa. e initial bump proneness indices of coal seam are such that the dynamic fracture time DT is 323 ms, elastic energy index W ET is 2.16, bump energy index K E is 1.64, and the uniaxial compressive strength R c is 4.3 MPa, thus indicating bump proneness.
Donglin Coal Mine is located at the intersection of the north-east secondary fold and the north-south structure. Due to stress imbalance, contorted fold structures are formed in Heiqiyan.
During the formation of contorted fold structures in Donglin Coal Mine, severe damage to the coal body structure and intensive changes in coal attitude are caused by compression and interformational sliding. Furthermore, the contorted fold structures also form a closed structural unit with favourable gas occurrence: these geological factors lead to the serious compound coal-rock dynamic disaster accidents that have occurred in Donglin Coal Mine. To date, about 200 compound coal-rock dynamic disasters have occurred in Donglin Coal Mine, with an average intensity of 72.8 t each. e maximum intensity reached 3500 t, and the maximum gas emission was 200,000 m 3 . Altogether, 92 people have died in such gas accidents. As for the No. 4 coal seam, the maximum intensity was 3500 t, while that for the No. 6 coal seam was 160 t. e parameters of compound coal-rock dynamic disasters in Donglin Coal Mine are summarised in Table 2.

Study of the Prediction Index for Compound Coal-Rock
Dynamic Disasters in Donglin Coal Mine. In this study, the prediction indices of compound coal-rock dynamic disaster are studied mainly focusing on the No. 6 coal seam in the north wing of Donglin Coal Mine, via laboratory tests and field investigation.

Laboratory Study.
Soft coal samples were collected in the 2607-50 heading face and 2609-90 heading face. Risk parameters of compound coal-rock dynamic disaster were measured in the laboratory, with the results listed in Table 3. In the table, 0.5 MPa, 0.74 MPa, and 1.1 MPa are the assumed gas pressures used. e risk parameters of a compound coal-rock dynamic disaster include initial gas emission velocity ΔP and the consistency coefficient of coal f. e K 1 -P relationship was fitted according to related experimental data. As shown in Table 3, coal samples were collected at different sites in the No. 6 coal seam. e critical value of K 1 was determined to be 0.10 to 0.23 based on the minimum gas pressure method. e gas parameters and mechanical parameters of compound coal-rock dynamic disaster were mainly decided by laboratory investigation. e critical gas desorption index K 1 was determined based on the critical gas pressure of 0.74 MPa and the theoretical gas pressure of 1.1 MPa calculated according to the consistency coefficient.

Field
Investigation. Some 70 prediction cycles were carried out in the 2607-50 heading face and 2609-90 heading    Shock and Vibration face during the study, with 58 sets of prediction data collected.
Among the 58 sets of data, on more than 26 occasions, the prediction rate of compound coal-rock dynamic disaster reached 45%. e averaged maximum values of K 1 and S in the 58 sets of data are 0.16 ml/g·min 1/2 and 14.3 kg/m, respectively. A check on the result cycle was conducted after 25 events (i.e., a check rate of 32%). e per-metre cuttings desorption index of prediction cycle, and the distribution of the maximum prediction cycle and check cycle values of this index, is listed in Tables 4-6, corresponding to  According to Tables 4 and 5, the maximum average value of per-metre drilling index K 1 is 0.08 ml/g·min 1/2 , while the maximum average value of S is 8.2 kg/m. Based on the above field measurements, it is implied that the following relationship exists between the desorption index of cuttings, kinetic phenomena, hole depth, and coal seam occurrence: for K 1max < 0.2, no spray hole phenomenon has been reported. When the coal body is relatively hard (failure type II and consistency coefficient of 0.7) and no drill jamming or coal-gunning occurs, the maximum drilling chip amount index S is usually around 6.0 kg/m. When the coal body is relatively hard (failure type II and firmness of 0.7) and there is drill jamming or coalgunning, the maximum drilling chip amount index S is usually above 15 kg/m. For soft coal (failure types III and IV and consistency coefficient of 0.2) with drill jamming or coal-gunning, the maximum drilling chip amount index S is even greater. e prediction or check cycle drill hole depth should be controlled to within 10 m; the maximum drilling chip amount index S occurs within a hole depth of 6 m, so the maximum index K 1 value occurs within a hole depth of 9 m.
In areas with significant coal seam occurrence (average seam thickness) variation, hole collapse, drill jamming, and coal-gunning occur frequently.
In ordinary mines, the occurrence rate of compound coal-rock dynamic disaster is between 5% and 10%. By considering a safety factor of three, the critical value of the compound coal-rock dynamic disaster prediction was theoretically determined; i.e., the number of an index exceeding a certain value accounts for 30% of the total number of statistical items of data. As shown in Table 6 and Figures 27  and 28, the ranges of different K 1 and S were obtained via statistical analysis of data obtained through 58 field tests.
According to the range of K 1max obtained via 58 field tests, the maximum value of K 1 exceeding 30% is 0.2 mL/(g·min 1/2 ), the proportion of which is 31% overall. As implied by field investigation results, when K 1 < 0.2 mL/(g·min 1/2 ), no spray hole phenomenon has been reported. With the increase in K 1 , the possibility of occurrence of kinetic phenomena including spray hole development increases.
According to the range of the prediction cycle drilling chip amount index S max obtained through 58 field tests, the

Prediction Index and Critical Value of the No. 6 Coal Seam of Compound Coal-Rock Dynamic Disasters in Donglin
Coal Mine. In combination with laboratory investigation, field data statistics, and field kinetic phenomena, the cuttings desorption index K 1 is taken as the gas prediction index, with a critical value of 0.2 mL/(g·min 1/2 ).
Based on the field test data statistics and actual kinetic phenomenon, the drilling chip amount S can be determined as the crustal stress index of compound coal-rock dynamic disaster for the No. 6 seam in the contorted fold structure band of Donglin Coal Mine, with a critical value of 15 kg/m. According to these prediction indices and critical values, 85 sets of prediction data and check cycle analysis were adopted during extensive field application. During the    excavation of a 257.5 m long roadway, no compound coalrock dynamic disaster occurred.

Conclusion
(1) e areas of occurrence of compound coal-rock dynamic disasters are those with high crustal stress confining pressure, high gas pressure, and low permeability; the coal mechanical properties of compound coal-rock dynamic disaster exhibit bump proneness and outburst risk. (2) e gas desorption velocity and permeability reduce with the increase in crustal stress, while increase and decrease with increasing gas pressure, respectively. A temperature variation within 30°C has limited influence on the gas desorption velocity. (3) As the mining depth increases, the crustal stress and gas pressure both increase. At the same time, compared to the gas pressure, the increase amplitude and velocity of crustal stress increase, resulting in the increase in coal strain energy, elastic modulus, bump energy index, and elastic energy index. Meanwhile, the dynamic fracture time is reduced. Gas extraction leads to the increase in bump energy index and elastic energy index, while reducing the dynamic fracture time; thus, the risk of compound coal-rock dynamic disaster is significantly increased. (4) Based on the actual geological conditions of Donglin Coal Mine, the prediction sensitivity index and critical value of compound coal-rock dynamic disaster were investigated. e cuttings desorption index K 1 and drilling chip amount S were set as the sensitivity index, with critical values of 0.2 mL/ (g·min 1/2 ) and 15 kg/m, respectively.

Data Availability
e data used to support the findings of this study are included within this article.

Conflicts of Interest
e authors declare that there are no conflicts of interest.

Authors' Contributions
Dong G and Wang Z conceived and designed the experiments. Wang Z performed the experiments. Wang Z and Liang X analysed the data. Dong G wrote the paper.