Control of spontaneous combustion of coal in goaf at high geotemperatureby injecting liquid carbon dioxide: inertand cooling characteristics of coal

The spontaneous combustion of coal in goaf at high geo temperatures is threatening safety production in coalmine. The TG-DSC is employed to study the variation of mass and energy at 4 atmospheres (mixed gases of N2, O2 and CO2) and heating rates (10°C/min) during oxidation of coal samples. The apparent activation energy and pre-exponential factor of coal oxidation decrease rapidly with increasing theCO2 concentration. Furthermore, its reaction rate is slow, its heat released reduces. Based on the conditions of 1301 face in the Longgucoalmine, a three-dimensional geometry model is developed to simulate the distributions stream field and temperature field and the variation characteristics ofCO2 concentration field after injecting liquidCO2. The results indicate that oxygen reached to depths of˜120m in goaf, 100m in the side of inlet air, and 10m in the side of outlet air before injecting liquidCO2. After injecting liquidCO2for 28.8min, the width of oxidation and heat accumulation zone is shortened by 20m, and the distance is 80m in the side of working face and 40˜60m in goafin the direction of dip affected by temperature.


Introduction
Coal fires are caused by two types of ignition sources: forced and spontaneous. Spontaneous combustion may be the initial cause of a fire which is then spread by conduction or convection to other areas of a mine [1]. Therefore, a large number of researches have been conducted for the evolution of the spontaneous combustion of coal.With respect to the molecular structure of coal, Jones [2] et al. studied the functional groups of coal using FT-IR in the low-temperature oxidation of coal and analyzed the variation of hydroxyl group.Kök [3]et al.studied on the thermal analysis and kinetic characteristics for coal combustion.Numerousresearches have conducted numerical simulations for studying the evolution of the spontaneous combustion of coal. Zhu [4] et al. studied the self-ignition characteristics for coal stockpiles by the method of numerical simulation. Wen [5] simulated the change in temperature field and distribution trend in exploitation process under different advancing velocities of working face.Zhang [6]et al. established a three-dimensional (3D) mathematical model for the spontaneous combustion of coal, simulated the distribution of seepage velocity field,concentration field, and temperature field, concluded the relationship between the oxidation zone height with temperature rising time and air supply volume, and calculated the temperature variation regularity under different propulsion velocities. Compared to the measurements of the pre and post spontaneous combustion of coal, Qin [7]et al.using the injection of liquid nitrogen technology successfully extinguished large-area spontaneous combustible fire inthe goaf. Zhu [8]determined the cooling range by injecting nitrogen in working face and applied Fluent to analyze the range and distribution of spontaneous fire zone by pre and post nitrogen injection.Zhu [9],Hao [10] andZhou [11] et al. simulated O2 concentration field by injecting nitrogen in goaf and original situation and calculated the air leakage quantity of oxidation zone by the inversion method.
In recent years, with the increase in depth and geotemperature, the risk of spontaneous combustion has increased, and the difficulty of control has also increased. The conventional method of injecting nitrogen cannot reduce sufficient amount of heat and injecting colloid cannot stop air leakage. With a high gasification speed and heat absorption of liquidCO2, it can cool and inert coal rapidly and efficiently. In this study, the TG-DSC is employed to study on the variation of mass and energy at 4 atmospheres (mixed gases of N2 、 O2 and CO2) and different heating rates during oxidation of coal samples, and determines the effect to restrain coal oxidation byCO2. The flow field and cooling region distribution and regulation in goaf at high geotemperature were evaluated by the numerical simulation, thus providing a basis for the prevention and control of the spontaneous combustion of coal.

Experimental conditions
The instrument of Netzsch 449C TG-DSC is selected to measure the characteristics of mass and energy of coal samples at different atmospheres (mixed gases of N2 、 O2 andCO2)and the reaction kinetics parameters of coal oxidation and decomposition are calculated with it.
(1) The flow rate of the mixed gases: 50 ml/min.  Table 1.The detailed experimental conditions were described inTable 2.   Figure 2 displayed the DSC curves of the coal samples at differentCO2concentration.The curve shows a double step shape, and has two summits during the coal sample release heat, which mainly reflect two stages of the oxidation resolve and severe burning, and its resolve reaction is done in phases. It is related to the coal construction, there is a part of active group with forceful reaction activity but bad heat stability in the coal, it quickly comes into oxidizing reaction in a low temperature, and thus, the first summit occurs by the 350K, with increasing coal temperature, much aromatic construction take part in the process of coal oxidizing burning, which make two summits occur in the process of coal oxidation. With the lower ofCO2concentration, The DSC curve isshaper,and the reaction of coal oxidation isfiercer, furthermore, its intensity of heat emission is greater. The concentration ofCO2 is higher which results the DSC curve to change more flat, the quantity of heat emission is smaller. Therefore, the reaction of coal oxidation is effectively restrained, which the peaks of temperatures for heat emission are backward.

Reaction kinetics for the suppression of coal oxidation withCO2
Thekinetics parameters of coal oxidation can be calculated according to the data from TG curves.In Coats-Redfern model,we can get the slope and interceptthrough the function curve of 2 ( ) ln g a T and 1 T ,which the kinetics parameters of activation energy and pre-exponential factor (Table3) can be calculated.    (Figure 3).The activate energy E is decreased with increasing concentration ofCO2, and it at air atmosphere is greater than at fillingCO2.The activation energy and frequency in the process of the coal sampleburning reduce with the increase ofCO2 volume fraction; the average apparent activation energy is lower after injectingCO2 in the coal, and it reduces with the increase ofCO2 volume fraction, which is easier to generate the burning phenomenon in theCO2 environment. According to the Arrenhnius theory, the rate of reaction is closely related to the apparent activation energy E and pre-exponential factor A, average apparent activation energy is lower in theCO2 environment, but the corresponding pre-exponential factor decreases with the magnitude rate, reaction rate is lower if theCO2 volume is bigger. The decrease of the average apparent activation energy just indicates that the sensibility of the oxidation burning reactionto the temperature reduces, but pre-exponential factor (frequency factor) A has astrong influence on the reaction rate from. In brief, the activation energy of the same coal sample and the sensibility to the temperature reduce if theCO2 concentration increases, but the pre-exponential factor changes a lot, however the reaction rate actually reduces soon; with the increasingCO2 concentration, the characteristic temperature moves backward, and the reaction difficulty increases.

In-situ region of calculation model
The average geothermal gradient of 1301 working face is 3.23  C/100m, the ground temperature reached 42  C, the coal face strike was 2700m, and the tendency was 220m. The fully mechanized sublevel caving technology was used in this coal, the height of caving coal was 5m, and the height of exploit was 3.5m. The air volume of working face was 2400m 3 /min, and the downward ventilation technique was used. The remaining coal thickness from coal wall to goaf within 10 m was 7.14 m, and that in other place was 1.82 m. The vadose rupture zonewas in each side of inlet air and outlet air in goaf, which exceededthe coal seam floorby >15 m range. Because the air flow entered into the goaf from the working face, the width of working face and each lanewas 10m, the height of working face was 3.5 m, and the breadth of working face was 220 m.The height of each lanewas 3.5 m, and the breadth of each lanewas 4 m.The goaf width was 220 m, and its depth was 300 m.The amount of air in the designed face was 2400m 3 /min, and the oxygen concentration was 20.98%.The calculated region was a wall except two side lanes, 0  Q .The mesh was generated with a grid step of 0.2 m, perpendicular tothe working face direction in the infiltration areaof air leakage.The grid step in the zone of working face and two lanes was 0.4m. On this basis, a 3D model of calculation area was established as shown in Figure 4.

Hypothesis and boundary definition
The goafwas assumed to be an isotropicandhomogeneous medium, and the temperature was set at42C, which is close to a stable thermal field. Affected by many factors such as pressure drop and viscous resistance, fresh air andCO2reachedthe goaf by a stable percolation way, with a convection heat transfer and removing some heat from the goaf. LiquidCO2 was injected into the inlet upper corner, the flow was 1200m 3 /h, the outlet pressure was 0.6MPa, and the temperature is~5  C.The relevant physical parameters of coal and rock are shown in Table 4.
Inlet 2: LiquidCO2concentration: mass percentwas 100%.Pipe line (4 inches) velocity: decreasing the temperature of goaf. By interceptinga bottom plane of 0.5 m at the bottom of the plane, the oxygen concentration fieldand temperature field are shown in Figure 5and 6. The depth of oxygeninfiltration in goaf was 120m,the input air side was 100 m, the return air side was only 10 m, and the mass concentration decreased gradually with the increase in depth. The oxygen concentration within the range from 0 to 0.02 was 65%, and that within the range from 0.208 to 0.23 was 20%. Figure 6 shows that the range of exchanged temperature field in goaf is small. The energy exchange of the fresh air in goaf was limited because of its own limited cooling capacity by the energy conservation law. The energy exchange was obvious in the intake side goafof 20~30 m, the temperature was  311K (38  C), and the temperature of deep goaf area was 314K (41  C) and remained constant. The percentage of zone with temperature between 313.4 K and 315 K was 67.5%, and that between 299 K and 300.6 Kwas 2.5%.

3.3.2.
Flow field after injectingCO2The oxygen concentration field and temperature field changed after injectingCO2. Because of the low temperature of injectedCO2with a large velocity and momentum,the goaf temperature decreased, the oxygen concentration reduced, the zone narrowed, and the cooling and oxygen reduction effect occurred. TheCO2relief vent of 10m in the upper corner of goafwas taken as an example. By the numerical simulation ofCO2 components and energy exchange at different times, theCO2concentration field, temperature field, and the results of the oxygen concentration field after injectingCO2for 12 min and 28.8minare shown in Figure 7 to 9.   Figure 7 shows that in the area of goaf theCO2 seepagewas large, and the length was 20 m along the direction of working face and  30 m along the goaf direction in 12 min. After 28.8 min, the increase in depth was better, along the direction of working face was  70 m and along the goaf direction was 40 m. Therefore, after the injection ofCO2, the inerting effect was ideal. Figure 8 shows that the temperature of goaf was 42 C and that ofCO2was~5 C, which is a large temperature difference.Therefore, energy exchange occurred easily. After 12 min of injecting CO2, the cooling range was 25 m along the direction of working face and 10 m along the goaf direction. After 28.8 min, the cooling range further increased, the length along the working face was 35 m along the goaf direction, resulting in a significant cooling effect. . Concentration field of oxygen in goaf after injectingCO2. Figure 9 shows that thediffusion direction of oxygen concentration in goaf after injecting CO2. After 12 min, the reduction in oxygen concentration and area were limited. After 28.8 min, the inerting effect was 80 m along the direction of working face and 70 m along the direction of the goaf. Thus, the effect of inerting and decrease in oxygen was obvious.

In-site verification
Using SF6 as the tracer gas, 9L SF6was released into a pre-buriedpipeline with injectedCO2 by the transient releasing method. The releasing position and monitoring pointare shown in Figure 10 The test results are shown in Figure 10. middle part of stents to deep goaf, and the velocity was 3~4m/min, which is similar to the simulation results.
To determine the temperature variation, the advance speed was maintained slow during the turnaround direction of working face. The CO concentration of the lower-corner angle increased slowly at the beginning, the highest value obtained was 190ppm, and the results are shown in Figure  11. The highest temperature in goaf went up to 58C. The quantity of continuousCO2 injection was 40,000kg.Then,the CO concentration and temperature decreased steadily, and the temperature decreased after a 40 m distance, similar to the simulation results of CO2injection diffusion distance and influencing range (40m). The temperature and CO concentration of the suffocating zoneafter 60m decreased rapidly.
(a) ( b) Figure 11.Variation trend of CO concentration in face corner and temperature in goaf.

Conclusions
(1)The activation energy at air atmosphere is greater than fillingCO2 for coal sample. With increasingCO2 concentration, the activate energy and the pre-exponential factor are fast small, which restrain the coal combustion.
(2)The depth of oxygen permeation ingoafis 120 m before injectingCO2, air inlet side is 100 m, and air outlet side is 10 m.The mass concentration gradually increases with the increase in goaf depth; the temperature zone of the air inlet side in goafclearly changesto 20~30 m.
(3)The variation zone of injectedCO2 concentration in 12 min is  20 m along the direction of working face and 30 m along the goaf direction; the cooling zone is  25 m along the direction of working face and10 m along the goaf direction. After 28.8 min, the expanding zone is better,the length along the direction of working faceis  70 m, and the depth along the goaf direction is  40 m.The cooling range further increases, the length along the working face is almost all, and the depthis 35 m along the goaf direction. The reducing zone of oxygen concentration is 80 m along the direction of working face and 70 m along the direction of the goaf.
(4) Using SF6 as the tracer gas, the width of oxidation and heat accumulation zoneis shortened to 20m.The distanceis 80m in the side of working face and 40-60m in the goafdirection affected by temperature, and the affected distanceis  20 m, thus confirming the results of numerical simulation and indicating that the developed model is reliable.