A New Method for Predicting Coal and Gas Outbursts

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
From the perspective of energy demand, the dominance of coal in the energy structure will not change in the near future in China. e proportion of coal consumption will remain at about 50% for a long time and about 80% of this coal is mined in underground wells. In recent years, with the largescale exploitation of coal, mining depth of mines in the central and eastern China has rapidly increased and extends to the deep parts at the rate of nearly 20 m per year. Western resources have also been developed in a large area, with coal production increasing in these regions, such as Shaanxi Province, the Inner Mongolia Autonomous Region, and the Xinjiang Uygur Autonomous Region. is leads to an increasingly serious threat posed by dynamic disasters, such as coal and gas outbursts, to safe production in coal mines. Statistical data in recent years show that although a series of effective measures have been taken to prevent disasters and accidents in coal mines, the number of outburst-prone mines still increases by hundreds every year [1], and the number of outburst accidents and deaths has not decreased significantly. erefore, effectively preventing gas accidents in coal mines, especially coal and gas dynamic disasters represented by coal and gas outbursts, has been the primary task of engineers working in the field of coal mine safety. e mechanism of occurrence of coal and gas outbursts refers to the cause, condition, and process of initiation, excitation, development, and termination of coal and gas outbursts. e comprehensive theory of action is generally accepted in terms of occurrence mechanism and prediction of coal and gas outbursts. e representative theories include energy theory, theory of crushing wave, theory of uneven stress distribution, theory of initiation centre, the rheological hypothesis of Zhou and He [2], the spherical shell destabilisation concept of Jiang and Yu [3], instability theory of flow fixation coupling of Liang et al. [4], and the catastrophe theory of Wang and Yu [5]. In addition, the stick-slip mechanism of Guo and Han [6], dimensional analysis method of Zheng [7], mechanical theory of Hu et al. [8], deductive method of Xian et al. [9], hierarchical control theory of geological structure of Zhang [10], main controlling geological body theory of the authors [11], and key structure body theory of Shu et al. [12] are also included. Other scholars have done physical simulation experiments of coal and gas outbursts [13][14][15][16][17][18][19][20][21][22][23][24][25], Xu et al. [13] have obtained that the exposed area of coal seam has some effect on coal and gas outburst; Tang et al. [14] have obtained that high-pressure gas is not only outbursting power but also damage power for powdered coal; Zhang et al. [15] have investigated the distribution and characteristics of gas pressure and stress ahead of the crosscut uncovering the closed type tectonic fracture zone; Lei et al. [16] have acquired the frequency characteristics in the four stages; Zhao et al. [17] have designed the simulation test system of coal and gas outburst including single gas pressure test machine and gas pressure and ground stress comprehensive test machine; Nie et al. [18] have conducted an artificial outburst with absorption equilibrium pressure of 0.30 MPa and inducing pressure of 0.53 MPa; Li et al. [19] have studied the judgment criterion of the comprehensive effect for coal body stress and gas pressure; Wang et al. [20] have studied the influences of gas adsorption on coal and gas outburst; Wang et al. [21] have established a model of energy conditions and a model of intensity assessment for outburst and have verified the model with real cases; Li et al. [22] have established the prediction models of outburst energy and taken Pingdingshan mining area as an example; Xu et al. [23] have analyzed the evolution process of temperature-pressure-stress during outburst; Wen et al. [24] have analyzed the theoretical requirements of the over-kilometer deep mining environment on the indexing requirements of the outburst simulation system loading capacity, system stiffness, gas pressure, and temperature and have proposed its critical indicators; and the outburst simulation tests have verified theories of relevant mechanisms of occurrence. Hu et al. [25] used the drilling cuttings index to predict coal and gas outbursts; Dong [26] used the drilling cuttings index and acoustic emission index to predict coal and gas outbursts; Li et al. [27] predicted coal and gas outbursts based on spatial chaos theory using gas desorption index of drill cuttings; Tang et al. [28] used the expansion energy of the initially released gases to predict coal and gas outbursts; Chen et al. [29] studied an FDA-based multiple indicators discriminant model of coal and gas outburst and applied the discriminant model to predict coal and gas outbursts; Qiu et al. [30] used electromagnetic signals to predict coal and gas outbursts; Li et al. [31] used the acoustic emission index to predict coal and gas outbursts; Dong et al. [32] used the gas concentration extraction indices to provide an early warning of atypical outbursts. Wang et al. [33] used the microseismic index to predict coal and gas outbursts; Lu et al. [34][35][36] have studied the microseismic location method and multiparameter characteristics of microseismic events. Current predictions of coal and gas outburst mainly depend on indices related to drilling cuttings, initial velocity of gas emission from a borehole, thickness of the soft layer, coal temperature, gas concentration extraction, acoustic emissions, and microseismic event counts. Research results have been adopted in relevant laws and regulations, such as Coal Mine Gas Grade Appraisal Method, Detailed Rules and Regulations for Prevention, and Control of Coal and Gas Outburst and Coal Mine Safety Regulation.
Although much research has been conducted on the mechanism of occurrence and prediction of coal and gas outbursts, how to use indices of gas pressure, drilling cuttings index, initial velocity of gas emission from borehole, thickness of soft layer, precursor signals, and so on to predict coal and gas outbursts is rarely reported. Aiming at this scientific problem, a new method for predicting coal and gas outbursts based on the mechanism of occurrence thereof was proposed in terms of coal mass strength, gas pressure, and in situ stress. By means of theoretical analysis, laboratory experiments, mathematical treatment, and industrial testing, coal and gas outbursts were predicted. Moreover, the method has been verified on-site.

A New Method for Predicting Coal and Gas Outbursts
For the new method for predicting coal and gas outbursts, many approaches such as theoretical analysis, laboratory experiments, mathematical treatment, and industrial testing were used to predict coal and gas outbursts. e method revealed that the rate of occurrence of coal and gas outbursts in mines was between 5% and 10% and gas pressures for coal and gas outbursts in shallow and deep mines separately were above 0.74 and 0.6 MPa in China [37]. During field measurement in a mining roadway, the prediction indices for coal and gas outbursts in the mine were determined according to geologic tectonic region, presence or absence of precursor areas of outbursts, changes before and after taking measures for preventing outbursts, and percentages of indices (i.e., normalised values thereof ). e generally used prediction indices included gas desorption of drilling cuttings (K 1 ), weight of drilling cuttings (S), initial velocity of gas emission (q), and thickness of a soft layer. e prediction index for coal and gas outbursts based on gas factor was gas desorption index K 1 of drilling cuttings, which denoted the gas content desorbed from the coal mass in the first minute of drilling. Firstly, the relationship between K 1 and gas pressure P was established through adsorption and desorption experiments in the laboratory and the initial prediction index critical value of K 1 was determined at gas pressures of 0.6 and 0.74 MPa during outbursts in the mine. Secondly, K 1 data were collected through in situ testing. When the proportion of a certain value among the collected data was 30% (considering a triple safety factor) [26], the value was taken as the initial prediction index critical value of K 1 . irdly, in accordance with that value of K 1 corresponding to outburst precursors collected in situ, the initial prediction index critical value was determined. e smallest value of the three was taken as the prediction index critical value of coal and gas outbursts in the coal seam. e prediction index for coal and gas outbursts based on coal mass strength was the thickness of the soft layer that could be twisted into powder by hand. Based on many cases of coal and gas outbursts, the prediction index critical value of the thickness of the soft layer was 0.2 m [37]. Furthermore, the prediction index for coal and gas outbursts based on in situ stress was the weight of drilling cuttings, which indicated the weight of drilling cuttings per linear metre of boreholes with diameters of 42 or 75 mm. Firstly, through the industrial test conducted on-site, the data pertaining to the weight of drilling cuttings were collected. If the proportion of a certain value accounted for 30% of the collected data [26] (considering a triple safety factor), the value was adopted as the initial prediction index critical value of the index. Secondly, according to the weight of drilling cuttings corresponding to outburst precursors collected in situ, the initial prediction index critical value was determined. e smaller one of the two was the critical value of prediction index of coal and gas outbursts in the coal seam. Finally, the new method for predicting coal and gas outbursts and critical values were verified based on industrial application.
Compared with other methods, the new method is simple and practical.

Overview of the Mine.
e No. 2 well in Enhong Coal Mine, Yunnan Province, China, has been in production since 1958, with a designed production capacity of 6 × 10 5 t/a and the actual production capacity of about 4 × 10 5 t/a. It is a coal and gas outburst mine. e mine is mainly controlled by the Cathaysian tectonic system. Primary structures in the mine including fault from Fuyuan County to Mile City, fault from Yingshang Township to A'gang Township, and Enhong compound syncline control the occurrence, metamorphism, and failure of coal-bearing strata. e coal seams in the mine belong to the Late Permian Longtan Formation. e coal seams in the mine field extend about 6 km from southwest to northeast, while they dip about 3 km from the northwest to the southeast. e dip angle of the coal seams ranges from 5°to 12°. Of them, the C9 coal seam, containing low ash, low sulphur, and high calorific value coking coal, is mainly mined.
Inclined shaft development was used in the No. 2 well of Enhong Coal Mine. Two mining levels were set, that is, the levels at elevations of +1940 m and +1800 m. ere were two mining areas, namely, the north wing and south wing. e stopping face in the south wing at the level of +1800 m was under active mining, while the mining in the north wing at the same level was about to be completed. About nine tunnelling faces were arranged in the mine. e stopping face was subject to longwall retreating mining along the strike and the coal blasting process, and the roof was controlled through the total subsidence method: blasting tunnelling technology and I-shaped steel supports were used in the tunnelling faces.
In the No. 2 well in Enhong Coal Mine, 23 coal and gas outbursts have occurred since 1980, with the maximum amount of outburst coal of 235 t and average amount of 16.55 t. e maximum amount of gas emitted was 8.3 × 10 4 m 3 . e outburst sites showed elevation of +1932 to 1874 m at vertical distances of 120 to 416 m from the surface (Table 1). e gas desorption index of drilling cuttings and critical values recommended by the Detailed Rules and Regulations for Prevention and Control of Coal and Gas Outbursts were adopted as the current prediction indices and critical values for the test working face of the mine. ey are as follows: the thickness of the soft layer was 0.2 m, K 1 � 0.5 mL/g.min 1/2 , and S � 6 kg/m (in a borehole with a diameter of 42 mm) or 18 kg/m (in a borehole with a diameter of 75 mm).

Determination of Parameters for Coal and Gas
Outbursts in the C9 Coal Seam. By using the channel sampling method on-site, four samples were taken from the C9 coal seam: one from all levels and three from the soft layer. e gas parameters including proximate analysis of coal, initial velocity of gas emission, and Protodyakonov coefficient were determined in the laboratory (Tables 2 and 3).
e method for measuring pressure in boreholes was used on-site to determine gas pressures in the coal seam and other gas parameters (Table 4).

Determination of Sensitive Indices for Coal and Gas
Outbursts in the C9 Coal Seam. In the process of field measurement of the mining roadway, based on the index indicating the presence or absence of precursor areas of outbursts and change therein before and after taking measures for preventing outbursts, the sensitivities of index K 1 of drilling cuttings desorption, S, and initial velocity q of gas emission were investigated. Finally, the gas desorption index K 1 of drilling cuttings was selected as the main prediction index. Considering the increase of burial depth and geologic tectonic zone, the weightS of drilling cuttings was regarded as a reference index (Tables 5 and 6 and Figures 1  and 2).
According to Table 5 and Figure 1, the gas desorption index K 1 of drilling cuttings decreased to a significant extent (33% reduction); when the thickness of the coal seam changed likewise, the value of precursor areas of outbursts was large and measures for preventing outbursts were taken; namely, the index was sensitive to coal and gas outbursts.
Based on Table 6 and Figure 2, the index S of the weight of drilling cuttings reduced slightly when the thickness of the coal seam changed and values of precursor areas of outbursts were small after taking measures for preventing outbursts (e.g., drilling holes). is implied that the index was not sensitive to coal and gas outbursts; however, by taking the increase of burial depth and geological structural zone into account, the index S of the weight of drilling cuttings was considered as a reference index.
During the test, several tests were conducted for determining initial velocity q of gas emission: however, due to softness of coal seam, collapses in boreholes, and lack of a tight seal to such boreholes, test indices were generally small and the initial velocity of gas emission from boreholes could not be determined at most points; therefore, it was not used Shock and Vibration 3 as a predictive index for coal and gas outbursts and the test values are listed in Table 7. During field investigation, the thickness of the soft layer was not obvious, so it was not used as a predictive index for coal and gas outbursts.

Determination of Critical Values of Sensitive Indices for
Coal and Gas Outbursts in the C9 Coal Seam

Determination of Critical Value of Gas Desorption Index K 1 of Drilling Cuttings
(1) Preliminarily determining critical value of gas desorption index K 1 of drilling cuttings based on the modelled K 1 -P relationship e experimental process of assessing the K 1 -P relationship model is as follows: coal samples with particle sizes between 1 and 3 mm were taken and degassed for 2 h using a vacuum pump. After that, the coal samples were pressurised to the predetermined gas pressure for adsorption, to allow the coal samples to absorb gas for 24 h. e gas pressure under equilibrium was recorded. After exposing the coal samples to the atmosphere for 1 to 2 min, the value of K 1 was determined. At least five experiments were conducted on each coal sample. By fitting these data, the K 1 -P relationship model could be obtained. Figures 3-5 show the experimental results pertaining to the K 1 -P relationship model for coal samples from different sites in the C9 coal seam in Enhong Coal Mine.    Based on the K 1 -P relationship model, the critical value of gas desorption index K 1 of drilling cuttings was preliminarily determined as 0.62 mL/g.min 1/ 2 (gas pressure P � 0.74 MPa), as presented in Table 8. (2) Preliminarily determining the critical value of gas desorption index K 1 of drilling cuttings based on percentages of data collected on-site By cycling samples 109 times to collect prediction index K 1 on-site, the index and its percentages are shown in Figure 6 and Table 9; accordingly, when K 1max ≥ 0.7 mL/g.min 1/2 , the percentage was 29%. (3) Determining initial critical value according to the K 1 value corresponding to outburst precursors collected during field testing In accordance with Table 5, the minimum K 1 value corresponding to outburst precursors including coal burst and emission after pulling the drill pipe out collected during field testing was 0.72 mL/g min 1/2 , so the value was used as the initial critical value of K 1 .
By combining the initial critical values (0.62, 0.70, and 0.72 mL/g min 1/2 (occurrence of coal and gas outburst)) as separately determined from the K 1 -P relationship model, data percentages, and outburst precursors, 0.60 mL/g min 1/2 was selected as the critical value of gas desorption index K 1 of drilling cuttings after numerical rounding of the data.

Determining the Critical Value of Index S of the Weight of Drilling Cuttings
(1) Preliminarily determining the critical value of index S of the weight of drilling cuttings according to percentages of data collected on-site A total of 68 cycles were conducted in situ to collect values for S as a prediction index. e index and its percentages are demonstrated in Figures 7 and 8 and Tables 10 and 11: when S max ≥ 3 kg/m (in a borehole with a diameter of 42 mm), the percentage was 38%; for S max ≥ 14 kg/m (in a borehole with a diameter of 75 mm), the percentage was 39%. e initial critical values of the index S of the weight of drilling cuttings were determined as S max = 3 kg/m (in a borehole with a diameter of 42 mm) or S max = 14 kg/m (in a borehole with a diameter of 75 mm); however, the indices of the weight of drilling cuttings were generally small.
(2) Determining the initial critical value of S in accordance with data corresponding to outburst precursors collected during field testing As listed in Table 6, the minimum S value corresponding to outburst precursors including coal burst and emission after pulling the drill pipe out collected     Shock and Vibration during field testing was 1.5 kg/m (in a borehole with a diameter of 42 mm). As no outburst precursors were observed for boreholes with a diameter of 75 mm, 1.5 kg/m (Φ42 mm) was regarded as the initial critical value of S. e initial critical values determined by combining with data percentages and outburst precursors and the S index were not sensitive to coal and gas outbursts in the C9 coal seam in this mine. Meanwhile, considering the increase of burial depth and the prevailing geological

Situation during Application of Indices K 1 and S for Coal and Gas
Outbursts. Based on the aforementioned research, gas desorption index K 1 of drilling cuttings was used as the main predictive index for coal and gas outbursts in the C9 coal seam in Enhong Coal Mine and its critical value was 0.6 mL/g.min 1/2 . e auxiliary index was the index S of the weight of drilling cuttings, and its critical values were 6 kg/m (in a borehole with a diameter of 42 mm) or 18 kg/m (in a borehole with a diameter of 75 mm). During field testing, 157 cycles were applied to collect values of K 1 and the percentages of the index are shown in Table 12 and Figure 9. It can be seen from the figure and the table that the percentage was 73% for K 1max ≥ 0.6 mL/ g.min 1/2 . Later, it was found that there was a large fault with the downthrow of 20 m near the 121901 working face, increasing the risk of an outburst. is further proved the sensitivity of the prediction index K 1 and the accuracy of its critical value.
During field testing, a total of 135 cycles were applied to collect the weight of drilling cuttings S. Percentages of the index are shown in Table 13 and Figure 10. e number of cycles exceeding the critical value was zero when S max ≥ 6 kg/ m.
During application, the roadway was safely tunnelled for about 1000 m according to the determined prediction indices for coal and gas outbursts, without a coal and gas outburst occurring.

Conclusion
(1) Based on the mechanism of occurrence of a coal and gas outburst relating to coal mass strength, gas pressure, and in situ stress, a method for predicting coal and gas outbursts was proposed. (2) During in situ measurement of the mining roadway, the specific prediction indices for coal and gas outbursts were determined according to indices including geological tectonic region, presence or absence of outburst precursors, changes before and after taking measures for preventing outbursts, and percentages of indices. e indices generally used were the gas desorption index K 1 of drilling cuttings, the weightS of drilling cuttings, initial velocity q of gas emission, and thickness of the soft layer.
(3) By using the new method for predicting coal and gas outbursts, the main prediction index for coal and gas outbursts in the C9 coal seam in Enhong Coal Mine was the gas desorption index K 1 of drilling cuttings and its critical value was 0.6 mL/g min 1/2 . e index S of the weight of drilling cuttings was used as the auxiliary index and its critical values were 6 kg/m (in a borehole with a diameter of 42 mm) or 18 kg/m (in a borehole with a diameter of 75 mm). (4) e new method of coal and gas outbursts can greatly improve the technical level of coal and gas outburst personnel and the prediction accuracy of coal and gas outbursts.
Data Availability e data supporting the conclusion of the article are shown in the relevant figures and tables in the article.