Desorption characterization of methane and carbon dioxide in coal and its influence on outburst prediction

Abstract

Coal and gas outburst is a dynamic phenomenon with violent eruptions of coal and gas from the working coal seam. It has been proved that the rapid desorption within a short period is necessary for the occurrence of an outburst. Due to limitation of the present test condition, gas desorption characterization for the first 60 s has not been researched sufficiently. In the present study, an experimental apparatus with the ability of high-frequency data collection was developed. Initial desorption characterization of methane and carbon dioxide in coal was experimentally studied. Both the initial desorption characterization of methane and carbon dioxide were experimentally studied with different equilibrium pressures. The desorbed gas pressure was measured at desorption time phase of 0–10 and 45–60 s, besides the initial amount of desorbed gas and initial diffusion velocity of coal gas were calculated to assess their risk of outburst. The results show that the gas pressure for both methane and carbon dioxide increases sharply in the initial time and then levels off, and the total amount of desorbed gas increases with the increasing desorption time. Although the amount of desorption methane is slightly larger than that of carbon dioxide at the beginning, the total amount of desorbed carbon dioxide is significantly larger than that of methane at the desorption process. Therefore, it can be concluded that the coal and carbon dioxide outburst is more dangerous than the coal and methane outburst based on the obtained experimental results.

Keywords

Methane carbon dioxide coal and gas outburst gas pressure initial speed of diffusion gas

Introduction

Coal and gas outburst is a dynamic phenomenon with violent eruptions of coal and gas from the working coal seam, which would result in economic loss and casualties. It is becoming more and more serious as coal mining is extended into deeper seams (Fisne and Esen, 2014; Zou et al., 2015). The coal and gas outburst is closely related to the coalbed gas reserved in the coal. Coalbed gas is an associated product in the coal formation process (Tahmasebi et al., 2012), which is deposited in coal seam in several forms: adsorbed in the pores surface of the coal, as part of the coal molecular structure, existing in the fissures and larger pores as free gas, and dissolved in the water (Liu et al., 2018; Zarębska and Ceglarska-Stefańska, 2008; Zou and Lin, 2018). The adsorbed and free gas take up the most of coal bed gas, which dominates the occurrence of coal and gas outburst. In order to prevent the occurrence of coal and gas outburst in the process of uncovering coal in deep mines, numerous sensitivity indexes attached to the gas desorption properties were established, including the variation of gas emission (K_V) (De Vegeron and Belin, 1966a, 1966b; Somnier, 1960; Yan et al., 2017), prediction method attached to the gas adsorption index K₁ and Δh₂ of drilling cuttings (Zhu et al., 2017), and the index initial velocity diffusion of coal gas (IVDCG) (ΔP) (Lama and Bodziony, 1998; Skoczylas, 2012).

Although the indexes mentioned above played an important role in coal and gas outburst prediction, none of them can accurately reflect the desorption characterization for the first 60 s due to the restriction of testing apparatus, which is the most important for the prediction of outburst risk. However, most research results show that the longer exposure time under atmospheric pressure would lead to greater error. Besides, the gas desorption properties of coal particles in the first 60 s are usually ignored in most of these sensitivity indexes. In theory, even the date of K1 could be used to calculate the gas desorption properties by measuring the rate of gas desorption at any time. But the actual time is longer than 1 min for the rate of gas desorption in the first 60 s, thus it could not be detected. Although researchers tried to study the desorption characterization and many data acquisition systems (DASs) were developed for recording pressure data, the shortest time interval were half a second in the first 10 s set by Pillalamarry et al. (2011) and 2 s in the first 60 s set by Li et al. (2017). The amount of data is not sufficient for the analysis of pressure variation in the first 0–10 s. Thus, it is necessary to decrease the time interval value because of the extremely high desorption rate during the initial period. Furthermore, gases involving in this type of disaster mainly are methane and carbon dioxide. Several researches show that gas composition has great effect on outburst risk (Lama and Bodziony, 1998), and the index ΔP increases with the increasing carbon dioxide percentage (Wu et al., 2010). Researches emphasizing on the sorption capacity of these two types of gas predicted that coal and carbon dioxide outburst is more dangerous than coal and methane outburst (Sobczyk, 2011, 2014). Few studies focused on the desorption characterization of carbon dioxide; therefore, it is very important to analyze the reason from the point of their desorption characterization.

In this paper, gas desorption characterization for the first 60 s is studied from the following three aspects: gas pressure, initial amount of desorbed gas, and initial diffusion velocity of coal gas. The effects of different gas types and different gas pressures on coal and gas outburst risk were analyzed. And the experimental process is followed by “Determination method for index (ΔP) of initial velocity of diffusion of coal gas”, which has been defined that the gas desorption process for the first 60 s can be divided into three steps: the releasing of free gas steps (0–10 s), the gas filled in the space between coal sample and pressure tank is the source of released gas; the gas desorption step (10–45 s); and the releasing of desorption gas step (45–60 s).

Experimental apparatus and procedures

Experimental apparatus

The schematic of the experimental setup is shown in Figure 1. It mainly consists of a pressure container, a constant temperature water bath, a vacuum system, an inflation system, and a self-developed real-time DAS. The pressure container is connected to the diffusion tank through the electromagnetic valve. The pressure regulating valve is installed in the inflation system to achieve the designated gas pressure. The application of the real-time DAS is the most remarkable characteristic of the experimental setup.

Experimental procedures

In order to acquire enough gas pressure data in the initial desorption process, time interval is brought down to about 10 ms. The ultimate pressure almost decreases to the level of an absolute vacuum (less than 10 mmHg). For the isovolumetric variable pressure measuring instrument, the pressure of the diffusion space is degassed to be less than 1 mmHg before the sample is loaded. Then coal samples were placed into the pressure container to vacuumize for 1.5 h. After the vacuum process, the gas was injected into the pressure container with predefined pressure and sealed until the pressure remains stable. After the adsorption process reaches balance, the electromagnetic valve between pressure container and diffusion tank is opened, the DAS is turned on at the same time to record the real-time pressure data of the diffusion tank at desorption time process of 0–10 and 45–60 s. At desorption time process of 10–45 s, the electromagnetic valve is closed.

Experimental samples

Coal samples recovered from Fuxiang coal mine of Shandong Energy Guizhou Co. Ltd of China were employed in this study. Samples were ground and sieved to achieve the desired size of 0.20–0.25 mm. Each sample was weighted 3.5 g. The industry analysis of the coal samples is as follows: the moisture content is 1.53%, the volatile matter content is 9.98%, the ash content is 26.44%, the fixed carbon is 78.2%, the maximum vitrinite reflectance is 3.03%, and the total sulfur content is 0.56%. Both of the desorption characterization of methane and carbon dioxide were studied with different equilibrium pressures. Experiments were carried out with equilibrium methane pressures of 0.073, 0.080, 0.087, 0.093, 0.100 MPa and carbon dioxide pressures of 0.074, 0.079, 0.088, 0.092, and 0.100 MPa. For comparison, the injection gas pressure is adjusted by the pressure regulating valve under different gas conditions. Since the adsorption or desorption characterization is extremely sensitive to temperature, the entire container was placed into water bath with the water temperature of 25°C (298.15 K) during the experiment process.

Results and discussion

Gas pressure

Figure 2 shows the typical pressure variation of the diffusion tank. It can be found that gas pressure increases sharply in the initial time and then it tends to be gentle. Besides, the pressure rise rate at 45–60 s is lower than that of 0–10 s, which attributes to the rapid desorption of the adsorbed gas on micropore surfaces under the influence of gas pressure gradient and concentration gradient (Czerw, 2011; Siemons et al., 2007; Wang et al., 2015; Xu et al., 2014). The desorption characterizations of methane and carbon dioxide with different adsorption pressures are shown in Figure 2. Results show that the pressure rise rate of methane is always lower than that of carbon dioxide under the same adsorption pressure. Thus, the carbon dioxide pressure keeps a high level for a longer time. Furthermore, the desorption of a large amount of carbon dioxide after the exposure of coal contributes to producing enough energy for an outburst. The research result is consistent with both the practical situation and laboratory experiments (Lama and Bodziony, 1998; Sobczyk, 2014; Wu et al., 2010).

Initial amount of desorbed gas

The amount of gas diffusion consisted of gas flux from the void volume and the desorbed gas from the coal particles. The gas diffusion mass flow (m) can be calculated as follows (Xu et al., 2014)

m= \frac{P σ^{*}}{\sqrt{T}} \sqrt{\frac{2 γ}{γ - 1} [{(\frac{P_{2}}{P})}^{\frac{2}{γ}} - {(\frac{P_{2}}{P})}^{\frac{γ + 1}{γ}}]}

(1)

where P is the absolute pressure (Pa); T is the temperature (K); P₂ is the atmospheric pressure (Pa);

γ

is the adiabatic exponent, which is dimensionless; and

σ^{*}

is the cross-sectional area of the nozzle orifice (m²). Then the total amount of gas can be calculated as follows

n= \frac{1}{M} \int_{0}^{t_{0}} mdt = \frac{P V}{298.15 Z R}

(2)

where M is molar mass of the gas (g/mol); R is molar gas constant (J/(mol·K)); V is volume of the diffusion space (m³); and Z is the compressibility factor, which is dimensionless. The calculation results are shown in Figure 3. It shows that the total amount of desorbed gas increases with time at a great speed in the initial stage. Then, the increasing rate continuously decreases. Methane has a lower diffusivity, which is the reason why its desorption pressure is slower than that of carbon dioxide (Busch et al., 2004; Cao et al., 2018; Cui et al., 2004). Molar mass of gas fluxed from the pressure container is equal to the gas increase in the diffusion space volume, which can be calculated as formula (3)

Δ n = \frac{V}{298.15 R} (\frac{P_{2}}{Z_{2}} - \frac{P_{1}}{Z_{1}})

(3)

where P₁ is the measured pressure at the end of first 10 s, and P₂ is the measured pressure at the end of first 60 s, both of them are the absolute pressure (Pa); V is volume of the diffusion space (m³); and Z is the compressibility factor, which is dimensionless. The gas desorbed velocity from coal particles can be calculated as follows

υ = Δ n / Δ t

(4)

As shown in Figures 4 and 5, the gas desorbed velocity decreases dramatically at the beginning and then it continuously reduces with the increasing desorption time both in the gas desorption phase of 0–10 and 45–60 s. All lines have shown that the gas desorbed velocity of carbon dioxide is lower than that of methane at the beginning.Afterward, it appears a cutoff point, and the gas desorbed velocity for carbon dioxide is significantly faster than that of methane, that causing the larger total amount of desorbed carbon dioxide.

Initial diffusion velocity of coal gas

The index (ΔP) of IVDCG is widely used to identify the outburst-prone coal seams. It has been proved that the occurrence probability of outburst increases with the increase of IVDCG, and the threshold for outburst coal seams is 10 mmHg. According to the research of Yang and Wang (1986), when the coal particle size is less than the limit size, the coal particles are basically composed of pore structure. Nevertheless, there are little effects of coal particle size on IVDCG when it reaches the limit particle size. The critical particle size varies with the coal, about 0.5–10 mm (Han and Jiang, 2005). Therefore, the mass flow of gas at the initial velocity measurement will be proportional to the gas concentration gradient, which is consistent with the diffusion law. The law of Fick is as follows

J = - D \frac{\partial C}{\partial X}

(5)

where J is the diffusion velocity (m³/(m·s)); D is the coal gas diffusion coefficient (m²/s); C is the coal gas concentration (m³/t); and X is the distance from the center of the particle (m). Large number of experimental results show that when the coal adsorbs gas, the adsorption isotherms conformed to the Langmuir equation (Xu, 2015)

W = \frac{a b P}{1 + b P}

(6)

where W is the adsorption equilibrium gas pressure P at a certain temperature (m³/t); a is the adsorption constant, which is the maximum amount of adsorbed gas when the gas pressure approaches infinity at a certain temperature (m³/t); and b is adsorption constant and the reciprocal of Langmuir pressure (MPa⁻¹).

Then the ΔP is calculated based on the assumptions of: (1) The coal sample is composed of spherical particles, (2) the coal is a homogeneous material, (3) the gas flow conforms to the law of conservation of mass and the continuity principle, and (4) the average radius (R) of the coal particles is 0.125 mm.Thus, the desorption of coalbed methane from coal particles in the first class of boundary conditions of the approximate solution can be written as follows (Kang, 2010)

\frac{C - C_{P}}{C_{0} - C_{P}} = - \frac{2 R {(- 1)}^{n}}{π rn} \sin \frac{n π r}{R} \exp (- n^{2} π^{2} \frac{D t}{R^{2}})

(7)

According to the determination method for index ΔP, it can be obtained as follows

Δ P = \frac{3.5}{27 ρ (\frac{4}{3} π R^{3})} \int_{0}^{60} (∭ V (C_{0} - C_{P}) d V) dt

(8)

where ρ is the coal particle density (g/cm³) and C_P is the gas concentration. The approximate solution of the index ΔP can be obtained by formulae (7) and (8)

\begin{matrix} Δ P = \frac{3.5}{27 ρ} (\frac{ab P}{1 + b P} - C_{P}) {\sum_{n= 1}^{\infty} \frac{6}{π^{2} n^{2}} [\exp (- π^{2} n^{2} \frac{60 D}{R^{2}}) - \exp (π^{2} n^{2} \frac{10 D}{R^{2}})]} \end{matrix}

(9)

The adsorption pressure on the determination of a greater impact can be obtained from formula (9). Because the desorption time is very short, and the temperature is controlled by the water bath, we assume that the gas diffusion coefficient D of the coal is constant in the desorption process. Hence index ΔP for different coal samples is only affected by the adsorption pressure, which can be confirmed in Table 1. As shown in Figure 6, when the equilibrium gas adsorption pressure reduces 26%, the reduction rate of the index ΔP (methane) is 54%. The greater the amount of gas adsorption is, the greater the effects of the adsorption pressure on the IVDCG are. There is a certain relationship between the IVDCG and the adsorption constants of coal. According to the IVDCG, the ultimate adsorption capacity of coal can be estimated.

Table 1.

Index ΔP measurement results.

CH₄	P₀ (MPa)	0.073	0.080	0.087	0.093	0.100
CH₄	ΔP (mmHg)	9.68	12.56	15.98	17.95	20.12
CO₂	P₀ (MPa)	0.076	0.081	0.09	0.095	0.100
CO₂	ΔP (mmHg)	12.35	15.29	18.24	21.22	24.56

The relationship between IVDCG and equilibrium gas pressure is shown in Figure 6. Do (1998) has proved that IVDCG varies directly with the methane pressure. But there are no experimental results of coal containing carbon dioxide. The relationships can be adequately described by the linear equations as

Δ P = 407 . {54P}_{0} - 2 0. 129 ({for CH}_{4})

(10)

and Δ P = 444 .0 {4P}_{0} - 2 0. 461 ({for CO}_{2})

(11)

Thus, the positive correlation is not only suitable for the coal containing methane, but also for the coal containing carbon dioxide. Moreover, it can be concluded that the initial speed of carbon dioxide diffusion is larger than the initial speed of methane diffusion (k_CO₂>k_CH₄). According to the outburst threshold of 10 mmHg (Zhu et al., 2017), and the linear equations as shown in equations (10) and (11), coal seams should be identified as outburst-prone seams when carbon dioxide pressure is higher than 0.069 MPa and methane pressure is higher than 0.074 MPa, which confirms that the coal and carbon dioxide outburst is more dangerous than the coal and methane outburst. In addition, the IVDCG (ΔP) and the adsorption constants (a and b value) of coal have a certain relationship (Tang, 2014). Therefore, the ultimate adsorption capacity of coal can be approximately estimated according to the initial speed of gas diffusion.

The diffusivity of gas in coal is determined by the physical and mechanical properties of coal and gas. Under the same gas content, the greater the IVDCG is, the greater the risk of coal and gas outburst is (Wu et al., 2010). The index ΔP is one of the indicators used in the coal and gas outburst forecast. When ΔP ≥ 10 mmHg, the coal seam has the outburst risk and vice versa. However, the atmospheric pressure varies from place to place, and if the critical value (10 mmHg) is used, it is necessary to make a deviation. Therefore, it should be adjusted according to the local atmospheric pressure. Due to the small amount of data in this experiment, more studies will be conducted in the future.

Conclusions

Both the desorption characterization of methane and carbon dioxide were experimentally studied with different equilibrium pressures. The desorbed gas pressure was measured and recorded at desorption time phase of 0–10 and 45–60 s. The initial amount of desorbed gas and initial diffusion velocity of coal gas were calculated. Experimental results show that the gas pressure for both methane and carbon dioxide increases sharply in the initial time and then levels off, which attributes to the rapid desorption of the adsorbed gas on micropore surfaces within the coal matrix when the container is filled with coal particles. The total amount of desorbed gas increases with the increase of time at a great speed at the initial stage, and then the increasing rate continuously decreases. The trend lines have a cutoff point and the total amount of desorbed methane is slightly larger than that of desorbed carbon dioxide before the cutoff point. Different from the desorbed methane, the total amount of desorbed carbon dioxide increases fast. With the increase of time, the gap becomes larger and larger. Meanwhile, pressure rise rate of carbon dioxide is significantly larger than that of methane, and carbon dioxide pressure can keep a high level for a longer time. It can be concluded that a large amount of carbon dioxide would be desorbed to free gas after the exposure of coal, which contributes to the enough energy for an outburst. The study results also show that the positive correlation between IVDCG and equilibrium gas pressure is not only suitable for coal containing methane, but also for coal containing carbon dioxide; besides IVDCG is greatly affected by the adsorption pressure. The index ΔP increases with the increasing adsorption pressure, and the initial velocity diffusion of carbon dioxide gas is larger than that of methane gas. In order to ensure the accuracy of measuring results, it is better to adjust the critical value (10 mmHg) according to the local atmospheric pressure.

Footnotes

Authors’ contributions

YL designed the experimental procedures. FW and YL conducted experiments in the whole process. And this paper was written by YL.

Acknowledgements

The authors also thank the editor and anonymous reviewers very much for their valuable advices.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is financially supported by the State Key Research Development Program of China (Grant No. 2016YFC0801404 and 2016YFC0801402), the National Science and Technology Major Project of China (Grant No. 2016ZX05043005), the National Natural Science Foundation of China (51674050), and the Fundamental Research Funds for the Central Universities (106112017CDJXY240001), which are gratefully acknowledged.

References

Busch

Gensterblum

Krooss

et al . (2004) Methane and carbon dioxide adsorption/diffusion experiments on coal: An upscaling and modeling approach. International Journal of Coal Geology 60: 151–168.

Cao

Sun

Dai

et al . (2018) Simulation research on dynamic effect of coal and gas outburst. Journal of China University of Mining & Technology 47: 113–120.

Cui

Bustin

Dipple

(2004) Selective transport of CO₂, CH₄ and N₂ on coals: Insight from modeling of experimental gas adsorption data. Fuel 83: 293–303.

Czerw

(2011) Methane and carbon dioxide sorption/desorption on bituminous coal experiments on cubicoid sample cut from the primal coal lump. International Journal of Coal Geology 85: 72–77.

de Vegeron

Belin

(1966a) Recherche d’un signe premonitoire et chronologie des degagements instantan es. Revue de l’industrie minerale 48(1): 60–80.

de Vegeron

Belin

(1966b) Etude des degagements instantan es de methane. Ann Mines Carb Paris. pp. 1–16.

(1998) Adsorption Analysis: Equilibria and Kinetics. London: Imperial College Press.

Fisne

Esen

(2014) Coal and gas outburst hazard in Zonguldak Coal Basin of Turkey, and association with geological parameters. Natural Hazards 74: 1363–1390.

Han

Jiang

(2005) Relationship between expansion energy of initial released gas and gas pressure. International Journal of Mining Science & Technology 34: 650–654.

10.

Kang

(2010) Influence study on initial speed of methane diffusion by adsorption pressure. Safety in Coal Mines 9: 4–5.

11.

Lama

Bodziony

(1998) Management of outburst in underground coal mines. International Journal of Coal Geology 35: 83–115.

12.

Cheng

Liu

et al . (2017) Research on gas diffusion model and experimental diffusion characteristic of cylindrical coal. Journal of China University of Mining & Technology 46: 1033–1040.

13.

Liu

Wang

Hao

et al . (2018) Constitutive model for methane desorption and diffusion based on pore structure differences between soft and hard coal. International Journal of Mining Science and Technology 27: 937–944.

14.

Pillalamarry

Harpalani

Liu

(2011) Gas diffusion behavior of coal and its impact on production from coalbed methane reservoirs. International Journal of Coal Geology 86: 342–348.

15.

Siemons

Wolf

KAA

Bruining

(2007) Interpretation of carbon dioxide diffusion behavior in coals. International Journal of Coal Geology 72: 315–324.

16.

Skoczylas

(2012) Laboratory study of the phenomenon of methane and coal outburst. International Journal of Rock Mechanics & Mining Sciences 55: 102–107.

17.

Sobczyk

(2011) The influence of sorption processes on gas stresses leading to the coal and gas outburst in the laboratory conditions. Fuel 90: 1018–1023.

18.

Sobczyk

(2014) A comparison of the influence of adsorbed gases on gas stresses leading to coal and gas outburst. Fuel 1(15): 288–294.

19.

Somnier

(1960) La desorption naturelle des charbons son application a la prevention des degagements instantanes et a l ’explication de leur mecanisme. Revue de l’industrie minerale 9(1): 776–784.

20.

Tahmasebi

Han

et al . (2012) A study of chemical structure changes of Chinese lignite during fluidized-bed drying in nitrogen and air. Fuel Processing Technology 101: 85–93.

21.

Tang

(2014) Research on influence of coal adsorption properties upon its gas diffusion characteristics. Mining Safety & Environmental Protection 6(41): 16–18.

22.

Wang

Elsworth

Liu

(2015) Rapid decompression and desorption induced energetic failure in coal. Journal of Rock Mechanics & Geotechnical Engineering 7: 345–350.

23.

Zhao

Cheng

et al . (2010) ΔP index with different gas compositions for instantaneous outburst prediction in coal mines. International Journal of Mining Science & Technology 20: 723–726.

24.

(2015) Experiment Study on Method of Initial Gas Emission from Borehole (IGEB) for Outburst Prediction in Cross-Cut Coal Uncovering. Xuzhou: China University of Mining and Technology.

25.

Jiang

Tian

(2014) Experimental study of the gas concentration boundary condition for diffusion through the coal particle. Journal of Nature Gas Science & Engineering 21: 451–455.

26.

Yan

Gang

Lin

et al . (2017) Direct fitting measurement of gas content in coalbed and selection of reasonable sampling time. International Journal of Mining Science and Technology 27: 299–305.

27.

Yang

Wang

(1986) Coal dust gas diffusion theory and its application. Journal of China Coal Society 3: 87–93.

28.

Zarębska

Ceglarska-Stefańska

(2008) The change in effective stress associated with swelling during carbon dioxide sequestration on natural gas recovery. International Journal of Coal Geology 74: 167–174.

29.

Zhu

Gao

et al . (2017) Effect of high-voltage thermal breakdown on pore characteristics of coal. International Journal of Mining Science and Technology 27: 1051–1055.

30.

Zou

Lin

(2018) Fluid-solid coupling characteristics of gas-bearing coal subject to hydraulic slotting: An experimental investigation. Energy & Fuels.

31.

Zou

Lin

Zheng

et al . (2015) Novel integrated techniques of drilling–slotting–separation-sealing for enhanced coal bed methane recovery in underground coal mines. Journal of Natural Gas Science and Engineering 26: 960–973.