Temperature Distribution Regularity and Dynamic Evolution of Spontaneous Combustion Coal Gangue Dump: Case Study of Yinying Coal Mine in Shanxi, China

Abstract

The serious environmental pollution caused by the spontaneous combustion of coal gangue has become a problem which cannot be ignored in the world mining industry. It is urgently necessary to clarify the law of the temperature distribution of spontaneous combustion in gangue dumps and to grasp its future dynamic evolution of spontaneous combustion. In this study, the internal temperature of the second platform of the Yinying coal mine gangue dump was monitored on the basis of the self-developed wireless temperature monitoring system. Its temperature distribution was analysed, and areas of low (<80 °C), medium (80~280 °C), and high temperature (>280 °C) were delimited. The finite element method was also used to simulate its internal temperature development of 1–5 years. The results show that: (1) The high temperature area is mainly distributed on the side close to the slope. In the area 3 m deep, the high temperature started to propagate quickly. At a depth of 4 m, medium and high temperature represented 90% of the platform’s total surface area. At 6 m deep, temperature peaked at 667 °C. (2) The conduction of the internal temperature of the spontaneous combustion of gangue discharge is a non-linear conduction, and the conduction of heat in the horizontal direction is lower than the vertical direction. (3) During the spontaneous combustion of gangue discharge over the next five years, the overall temperature increase is faster at first, then it decreases and eventually stabilises. The high-temperature zones extend 10 m from the slope to the interior in five years, and the high-temperature zones are oval-shaped. This study provides a theoretical reference for the prevention and control of spontaneous combustion of coal gangue dumps.

1. Introduction

Coal gangue is a solid waste product of coal which is generated during the mining process, namely, during the construction, extraction as well as washing of mines [1,2,3,4]. Coal gangue is dark grey in colour and harder in texture than coal. In terms of chemical composition it contains less carbon than finished coal. The emissions account for 10–20% of the carbon production, with an average of 12% [5,6]. It has become the largest industrial waste in the world in terms of annual emissions and cumulative inventories [7]. However, the overall utilisation rate of coal gangue is low, and most of the gangue is stacked as solid waste to form gangue dumps. Years of accumulation under the gangue dump will often result in spontaneous combustion, a re-ignition phenomenon which has become a global problem [8,9,10]. During its combustion, it emits carbon dioxide, sulfur dioxide, and nitrogen oxides. Not only does it affect the air quality of the area, but the acidic water formed by rainwater drenching and melting it also seriously affects the safety of groundwater. This has created incalculable risks for the residents of the mine. The impact of this phenomenon on the ecology of mining areas has become a bottleneck which hampers the achievement of sustainable development in mining areas around the world [11,12,13,14]. Therefore, in order to promote the sustainable development of mining areas and ecological environment recovery, it is necessary to grasp the internal temperature distribution pattern of spontaneous combustion of gangue dumps and its future spontaneous combustion development trend in time, so as to effectively reduce or avoid the phenomenon of spontaneous combustion of gangue dumps.

The researchers are more likely to predict the temperature of the heat source and analyse the distribution pattern of the gangue’s deep temperature using simulation experiments in the interior [15,16,17,18]. Hao Chuanbo [19] monitored the temperature at each depth of the gangue dump using experiments inside and identified five areas of self-ignition. Some researchers [20,21] have studied the composition of spontaneously combusted gangue and the characteristics of temperature changes in spontaneously combusted gangue dumps.

At the end of the 20th century, the use of the finite element method to simulate the dynamic evolution of spontaneous combustion began to be applied to the mining area, which is the basis for the study of the dynamic evolution of spontaneous combustion of coal gangue. Mingming Li [22] built an experimental bench of the active temperature field of a loose coal body and used the Fluent software to simulate it, and the model was proved to be effective by comparing the simulation results and experimental results. Minggao Yu [23] simulated the transient temperature field of spontaneously burning gangue discharge using ANSYS software. In 2011, Misz Kennan [24] investigated the spontaneous combustion of a coal gangue dump in Poland and found that the fire source had moved inside the gangue landfill. Qing Xia [25] used the finite difference method to resolve the deep temperature field of the gangue and study the depth temperature heat transfer law. Li Bei [26] studied the coupling mechanism between the temperature field and the heat transfer flow field in gangue dumps using Fluent numerical calculation software simulations.

Although much work has been done in the study of coal gangue spontaneous combustion, the temperature distribution inside gangue dumps is not clear. Based on field test data on spontaneous combustion of gangue piles, a temperature field for the dynamic simulation of the study is also lacking. This study analyses the internal temperature distribution of spontaneous combustion gangue dump in the Yinying coal mine in Yangquan, Shanxi, through wireless temperature data collection and transmission to a cloud platform system, divides different temperature zones, establishes a multi-field coupling model of spontaneous combustion gangue dump, and predicts its spontaneous combustion development change process in the next five years. The results of this study can provide a scientific reference for the location and prevention of fire sources for the spontaneous combustion of gangue dumps and provide a theoretical basis for the planning of gangue dumps based in sustainable development.

2. Materials and Methods

2.1. Overview of the Study Area

The Yinying coal mine is located in the northwestern part of the mining area of Yangquan City, Shanxi Province, and the gangue dump site is in the gunpowder ditch about 1.5 km north of Yinying town, with coordinates of 113°33′28″ E and 37°57′16″ N. The region has a temperate monsoon continental climate: little rain, sandy, controlled by ocean currents, a warm and humid climate in summer, cold and dry in winter. The multi-year annual average temperature in the study area is 11 °C, with a minimum temperature of −20.5 °C and a maximum temperature of 40 °C. The main winds throughout the year are from the north-west and south-west, with secondary winds from the east and a multi-year average wind speed of 2 m/s.

The gangue dump covers an area of about 16.3 hm², the bottom elevation is 875 m, the top elevation is 925 m, the pile of gangue has reached the capping elevation and has stored about 4 million tons of gangue. The gangue field has three platforms; the study area is the second platform. The local surface has a yellow-brown solid and the air has a pungent odour. Geographical location is shown in Figure 1.

2.2. Analysis of Coal Gangue Composition

In order to find out the composition and characteristics of the gangue, this study uses a high temperature furnace, a muffle furnace, electronic balance and other instruments to test the composition of the gangue. The gangue of the three platforms is dumped by the same coal mine; the coal type is the same. The three platforms are only stacked at different times. As the first platform has completely combusted spontaneously, there is a distinct pungent smell when approaching the platform. For safety reasons, we took samples from the second and third platform to do the test and took their average values. The results are shown in

Table 1. Coal gangue composition.

Ingredients	SiO2	Fe2O3	TiO2	P2O5	CaO	MgO	Al2O3	S	K2O	Scorch Reduction
	%
Value	17.12	3.34	0.18	0.08	39.4	1.24	3	0.96	0.21	35.16

.

As can be seen from Table 1, the gangue contains Si, Al, Fe, Ca, Mg, S, other oxides, and C. The main compositors are calcium oxide, silicon dioxide and iron oxide, which account for 59.56%, followed by aluminum oxide, magnesium oxide and sulfur. The main compositors that affect the spontaneous combustion of coal gangue are carbon and sulfur. The content of fixed carbon and total sulfur are 11.17% and 1.01%, respectively. The element sulfur is mainly attached to pyrite (FeS₂) and the sulfur in the composition is also the product of pyrite oxidation reaction.

2.3. Monitoring Point Layout

In order to monitor the internal temperature of gangue dumps, a wireless temperature monitoring system has been developed for this study. The buried temperature measurement tubes are mechanically formed holes and filled with fine sand to seal the tube holes. K-type high temperature thermocouples are used to measure the temperature at different depths inside the gangue pile. The collected data are received by the lora4g gateway and transmitted to mobile phones and PC display terminals after processing by the SaaS cloud service platform, as shown in Figure 2.

Based on several visits to the site, it is known that the first platform has completely spontaneously combusted without an inch of grass. There was a distinct pungent smell as we approached. The platform is entirely in a hot area. The third platform has good vegetation growth. The second platform has a distinct pungent odour locally, with good vegetation growth in some areas. The second platform is very representative of the dump site and was selected for this study. A total of 45 temperature measuring tubes were installed on the second platform, distributed around the scrub in the area. The locations of the monitoring tubes are shown in Figure 3. Each tube was fitted with one or more monitoring devices and was tested at depths of 1–8 m. For construction safety reasons, there were more testing points at 1–4 m depth and fewer points at 5–8 m depth. This study began on 21 November 2019 with automated monitoring of the second platform.

2.4. Modeling

COMSOL Multiphysics software has outstanding advantages in solving multi-field coupling problems. This study uses the software to establish a three-dimensional four-pronged gangue dump geometric model. The direction pointed by the x-axis is north; the model size is set as follows: the upper bottom is 50 m × 64 m, the lower bottom is 60 m × 64 m, the height is 10 m, and the inclination angle of the slope is 45°. As the southeast side of the gangue field has a fissure, set with 1 cm in diameter, the height of a 10 m cylindrical gap, is a column and air connection as a representative of the fissure. In addition, due to the gangue top cover being 0.5 m thick, the top 0.5 m range material is set as soil. For the rest of the gangue, the soil and gangue interface is set as continuous, the whole determined as a joint, to ensure the continuity of the physical interface domain inside.

2.4.1. Fundamental Assumptions

Spontaneous combustion of coal gangue dumps is a very complex three-dimensional non-stationary problem. In order to facilitate its simulation, we need to make some assumptions in the model. The basic assumptions are as follows.

(1)

Do not consider solar radiation and the gangue between the radiation heat transfer.

(2)

The gangue dump is a uniform porous medium and is isotropic.

(3)

The gangue dump percolating gas satisfies the ideal gas equation of state.

(4)

There is uniform flow of external air.

(5)

The heat transfer process between the gangue skeleton and the void is a quasi-static process and the temperature of the gas and the solid remains the same.

(6)

In the warming process, the physical parameters of the gangue, such as thermal conductivity, specific heat capacity, convective heat transfer coefficient, etc., remain unchanged by the temperature.

(7)

Do not consider the effect of moisture on the spontaneous combustion of the gangue dump.

(8)

The internal structure of gangue dump is stable.

2.4.2. Computational Model

In this study, the finite element method is used for mesh dissection. The geometric model is mainly divided into tetrahedral mesh and the mesh cell size is set to be refined; the boundaries of local areas can be further refined in manual form, such as at the corners, slopes, and fissures. It includes 105,800 tetrahedral cells, 24,768 prisms, 130,568 total cells, 471 edge unit mass, 16 vertex cells, a minimum cell mass of 0.1773, an average cell mass of 0.6791, grid volume ratio of 3.953 × 10⁻⁵, grid volume of 35,200 m³, and geometric shape function order for linear cells. The calculation model is shown in Figure 4.

The model research process is set to transient calculation mode and the solver is set to transient solver. The time unit is set as year; the calculation time frame is from November 2019 to November 2025, a total of 5 years, the time step is 1 year, and the relative tolerance is 0.01. The calculation time period is divided into two stages: the first time period is from November 2019 to November 2020, which lasts 1 year, for the identification and verification stage. The second time period is from November 2020 to November 2025, for the prediction stage.

2.4.3. Calculation Parameters

In this study, the gangue dump temperature monitored on 21 November 2019 was used as the initial value of the temperature field. The multi-year average wind speed value of 2 m/s was used as the initial wind speed and the initial oxygen concentration was 9.375 mol/m³.

(1)

Temperature field boundary conditions

The top, windward side and the fissure having direct contact with air, can have convective heat exchange with the external environment, as the third type of boundary conditions.

$$\mathrm{Convection} \mathrm{Heat} \mathrm{Transfer}:-n(-\lambda \nabla T)=h(T_0-T)$$

(1)

where (1) λ is the thermal conductivity in W/(m·K)), T₀ is the atmospheric temperature and T is the solid temperature in K.

The remaining boundaries are adiabatic.

$$\mathrm{No} \mathrm{convection} \mathrm{heat} \mathrm{transfer}:-\lambda \nabla T=0$$

(2)

(2)

Air seepage velocity field boundary conditions

The windward side is the first type of boundary conditions according to the average annual wind speed of Yangquan City, assuming that the external wind flow is always uniform towards the windward slope of the gangue dump in x direction:

The windward side: v = 2 m/s

Top and exit at the fissure, free flow of gas, pressure of one atmosphere, type II boundary conditions:

Free flow: P = 1 atm

The remaining boundaries are zero flux boundaries: v = 0 m/s

(3)

Oxygen concentration transport field boundary conditions

Oxygen with air flow into the gangue mountain’s internal, windward side for the first type of boundary conditions:

The windward side: c = 9.375 mol/m³

Top as well as being a free flow of oxygen at the fissure, as a second type of boundary condition.

$$\mathrm{Free} \mathrm{flow}:-n{D}_i\nabla c_i=0$$

(3)

where (3) D_i is the air diffusion coefficient at the fracture in m²/s. c_i is the oxygen concentration at the fracture in mol/m³.

The remaining boundaries are zero flux boundaries:

$$\frac{dC}{dn}\left|{}_s\right.=0$$

(4)

where (4) C is the remaining boundary oxygen concentration in mol/m³.

(4)

Parameter selection

The model involving porous media within the heat transfer, air percolation velocity field and oxygen component transport, and other processes, needs to set some parameters that reflect the environmental conditions, gangue thermal properties, and gangue characteristics in the simulation. The selection of such parameters are shown in

Table 2. Physical parameters.

Object	Density [kg/m3]	Specific Thermal Capacity [J/(kg·K)]	Thermal Conductivity [W/(m·K)]	Porosity
Air	1.43	1000	0.75
Coal gangue	2500	1450	0.37	0.3
Loess	1600	1696	0.48	0.1

.

3. Results and Discussion

3.1. Horizontal Temperature Distribution

The temperature monitoring data on 21 November 2019 were selected for this study to analyse the internal temperature distribution of the gangue dump. Since the temperature monitoring points mainly covered the depth of 1–4 m in range and there were fewer measurement points in the depth range of 5–8 m, the temperature data of 1–4 m were selected for analysis; the temperature distribution is shown in Figure 5. One sulfide quality in the gangue in 80 °C is rapid warming. The ignition point of carbon materials is 280 °C. The study of the spontaneous combustion of the gangue mountain will be divided into three temperature zones: low-temperature zone (<80 °C), medium-temperature zone (80~280 °C), high-temperature zone (>280 °C). The area of the low-, medium- and high-temperature zones at each depth is shown in Figure 6.

From Figure 5, it can be seen that: (1) There exists an overall consistency and similarity among the four plots. The change trend of temperature and isotherm morphological characteristics are similar. The high temperature areas of the gangue dump layers are mainly distributed in the north and east sides; and the eastern area of the north side is the most concentrated, where there are two high-value centres, which are more intensive isotherms and have a large temperature change gradient, distributed in the southeast and northeast. This is due to the fact that the north side of the study area is the airward side and also the windward side; it has sufficient oxygen supply, precipitation can infiltrate it through the side slopes, and the side slopes are covered with loose soil, which has certain heat storage conditions, so the temperature is higher. In the southeast high temperature area, due to the gangue inside the region having large fissures and, oxygen being more adequate, the temperature of spontaneous combustion is higher. (2) At a depth of 2 m, two high-temperature areas with a maximum temperature of 400 °C appeared, distributed in the northeast of the platform. At a depth of 3 m, the high-temperature area spread rapidly. At a depth of 4 m, the medium–high-temperature area accounted for 90% of the entire platform area and only a small low-temperature area existed in the southwest corner. This is due to the high oxygen concentration in the shallow layer; while the greater the depth, the better the heat storage conditions, and thus the temperature continues to rise.

From Figure 6, it can be seen that: At a depth of 1 m, only low-temperature zones and medium-temperature zones exist; all have not reached the ignition point. At depths of 1 m and 2 m, the spontaneous combustion tendency is low. At a depth of 3 m, the area of the two high-temperature zones expands 14 times. At a depth of 4 m, the area of the high-temperature zone expands 3 times again. On the one hand, this is due to the shallow layer where, although oxygen is sufficient, the convective heat exchange is more intense and the heat storage capacity is poor. On the other hand, due to the shallow measurement point being closer to the ground soil, the soil and gangue thermal physical parameters have a large difference; the soil absorbs part of the heat, resulting in a lower temperature.

3.2. Temperature Distribution in the Horizontal Layer Profile

In order to have a comprehensive and integrated analysis of the temperature distribution in the horizontal direction especially in the high-temperature region, two line segments were selected as profiles in the high-temperature region and the temperature profiles of the 1–4 m horizontal layer were plotted. The location of the profiles is shown in Figure 7. The temperature profiles at different depths are shown in Figure 8.

It can be seen from Figure 8 that: (1) The temperature of Profile 1 is higher than that of Profile 2 as a whole, indicating that the temperature at the side slope is higher than that of other areas. From Figure 8ⅱ, it can be seen that within 10 m from the side slope, the temperature drops rapidly and the temperature drop is fast, which is mainly governed by the oxygen concentration. (2) Comparing the temperature profiles of the 1 m horizontal layer and the rest of the horizontal layers, it can be found that the change at the 1 m depth with respect to distance is relatively smooth, and the temperature difference is less than 100 °C. The deeper the depth, the larger the temperature difference is; the maximum temperature can reach 400 °C. This is due to the fact that 1 m is close to the surface, which is influenced by the atmospheric temperature, and it is not easy to collect heat and exchange heat with the soil, making the temperature tend towards being balanced. While at the deepest part, it is less influenced by the external environment, and the influences range at the high-temperature concentration is about 10 m, that is, it has almost no influence on the temperature of the measurement points above 10 m distance from the high temperature point. (3) In the vertical direction, the higher the temperature, the faster the temperature rise, indicating that the temperature conduction is a nonlinear conduction. In the high-temperature point, for every 1 m depth, the temperature increases by about 100 °C; in the horizontal direction, in order to reach a 100 °C difference in temperature, there needs to be about 10 m in distance, which indicates that the horizontal direction of self-combustion heat conduction in a gangue mountain is weaker than the vertical direction.

3.3. Vertical Temperature Distribution

In order to study the vertical distribution pattern of temperature for the spontaneous combustion of a gangue dump, six monitoring points at more comprehensive depths were selected for monitoring. The depth and temperature profiles of each monitoring point are shown in Figure 9.

From Figure 9 it can be seen that: (1) In the depth range of 1–8 m, the trend of temperature variation with depth is consistent overall, with a non-linear variation along the vertical direction, increasing first and then slightly decreasing or slowly increasing with increasing depth. After 6 m, there are two main situations: firstly, the temperature reaches its peak at 6 m and starts to decrease after 6 m, such as T207, T216 and T221; secondly, the temperature increases slowly after 6 m and the growth rate is less than 8%, such as T241, T243 and T244. (2) Monitoring points T207, T216, T221 are located close to the side slope. The phenomenon that the temperature rises first and then falls, indicates that the gangue hill in the vertical direction also has the characteristics of the fire source centre; the centre is mainly concentrated at 6 m. This is due to the surface layer not being conducive to heat storage and the lack of oxygen in the deep layer. At 6 m the oxygen and heat storage conditions are met, hence the highest temperature. (3) Monitoring points T241, T243, T244 are located in the middle of the platform, 6 m after the temperature grows slowly. This is due to the gangue’s natural accumulation of particle size deviation; near the slope the particle size is fine; at the centre, the particle size is relatively coarse; in the void, it is larger. At 6 m the oxygen is sufficient, can be connected deep through the void, and the deep heat storage capacity is higher, hence the shown slow rise. In addition, because internal temperature of the gangue is high, the hot air will show an upward trend. After 6 m depths, due to the lack of oxygen, spontaneous combustion is not easy, which will also lead to the downward trend in temperature. (4) At 667 °C, the point of the highest temperature in the whole platform is located at the measurement point T207, which is 6 m in the ground, below the horizontal line, and at the slope below the intersection of the 6 m parallel line. This is due to the proximity of this measurement point to the slope, which is windward and has sufficient oxygen concentration and good thermal storage conditions with loose cover.

3.4. Temperature Evolution inside Gangue Dump

This study uses the finite element simulation method to simulate the temperature change within 5 years of the self-combustion gangue dump. Figure 10 shows the gangue dump’s self-combustion in the initial stage and 1–5 years after in the xz direction and yz direction on a temperature cross-section map. According to the initial temperature field simulation results and the initial measured temperature for comparison, the error is small, indicating that the model can be a more accurate simulation of the dynamic evolution of the gangue dump’s self-ignition.

From Figure 10 it can be seen that: (1) The distribution of the temperature field in the high-temperature area is continuously pushed from the slope to the inside in the x direction. Within 5 years, the high-temperature centre has moved a distance of 10 m. This is due to: air inflow from the side slope, the larger air filtration rate in the area near the side slope, gangue dumps’ internal pressure difference. Hence, the high temperature area continues its push inwards. On the other hand, in the high temperature area, the reaction rate is faster, the gangue layer of combustible material will gradually burn out, the fire centre will be pushed more towards the combustible material, and the heat storage capacity of the internal area is better. (2) By comparing the xz and yz plane slices, can be found in the xz-plane, the high-temperature area gradually radiates in a semi-circular fashion, constantly expands around, and the southeast high-temperature areas merge into one. In the yz plane, the range of the high- temperature area is also expanding. The y direction of the high-temperature range is greater than the x direction of the high-temperature range and is ellipsoidal. This indicates that the y direction of the temperature difference is less than the x direction of the temperature difference, which is the wind direction. Hence, the greater the wind speed, the faster the heat transfer.

4. Conclusions

This study takes the spontaneous combustion of the gangue dump in the Yinying coal mine in Yangquan, Shanxi Province, as the research object. In this study, site investigations, monitoring point arrangement and temperature monitoring of the spontaneous combustion of gangue hills were carried out. The distribution of its internal temperature field was studied. Moreover, the dynamic evolution of its spontaneous combustion was numerically simulated. The main conclusions were obtained as follows:

(1)

The internal temperature of the spontaneous combustion of gangue mountains increases with depth. At depths of 1–4 m, the temperature rises more rapidly. At 6 m the temperature is at its highest. This spontaneous combustion of the gangue mountain has two high temperature areas in the southeast and northeast, respectively. At 1 m depth, spontaneous combustion tendency is low. The high-temperature area depth began to form at a depth of 2 m. At 3 m depth the area quickly spread and expanded 14 times. At 4 m depth, high-temperature area again expanded 3 times. There is a small range of low temperature area only in the southwest corner. At 6 m depth temperature is the highest.

(2)

The range of heat transfer within a spontaneously combusted gangue dump is greater in the vertical direction than in the horizontal direction. In the vertical direction, the temperature increases by approximately 100 °C for every 1 m increase in depth from the hot spot. To achieve a temperature difference of 100 °C in the horizontal direction, a distance of approximately 10 m is required.

(3)

According to the predicted results, in the next 1–5 years, the temperature of the spontaneous combustion of gangue dump decreases from fast to slow and finally tends to stabilize. The high-temperature zone advances continuously from the side slopes to the interior. In the x direction the high-temperature zone advances 10 m to the interior within 5 years and the high temperature zone is spatially elliptical.

This study can provide a theoretical basis for the prevention of spontaneous combustion in coal gangue hills. However, there are certain problems in the study, which need to be further studied in the future. Therefore, in future work, the deep temperature monitoring points within the gangue dump should be arranged more densely in order to be able to understand its temperature distribution more accurately.