Experimental Study on Pore Structure and Gas Desorption Characteristics of a Low Rank Coal: Impact of Moisture

s: Coal and gas outburst is one of the most serious disasters for underground coal mining. The 1 water adsorbed on coal can leads to that the pore structure of moist coal is different from that of dry 2 coal, thereby affecting methane desorption characteristics of coal for the outburst risk prediction. In this 3 paper, the impact of moisture on pore structure and methane desorption performance were investigated. 4 The analysis on low - temperature nitrogen gas adsorption tests show that the micropores (pore diameter 5 < 10 nm) are most affected by the adsorbed water. In particular, for water - equilibrated coal sample at 6 98% relatively humidity, the micropores less than 4 nm analyzed by DFT pore size distributions almost 7 disappear probably due to the blocking effect of the formed water clusters and capillary water. In this 8 case, the micropores can still contributes most sites for gas adsorption. Furthermore, the fractal 9 dimension at relative pressure of 0–0.5 ( D 1 ) and 0.5–1 ( D 2 ) calculated by the Frenkel - Halsey - Hill 10 model indicates that, when moisture content is less than 4.74%, D 1 decreases rapidly while D 2 shows a 11 slight change; whereas, further increases in moisture content results in that D 2 decreases significantly 12 and D 1 remains at about 2.32. Further investigation shows that, below the equilibrium moisture content, 13 the ultimate desorption volume ( A ) and initial desorption rate ( V 0 ) are closely related to D 1 , while the 14 desorption constant ( K t ) mainly depends on D 2 . Therefore, the adsorbed moisture has significant 15 negative impact on methane desorption performances by affecting characteristics of coal’s pores.

desorption capacity of coal, and many methane desorption indices are used to predict the outburst risk 23 (Wang et al. 2020b; Xue et al. 2020). Therefore, understanding the methane desorption characteristics 24 of coal is critical for the prediction of coal and gas outburst. Coal is a porous medium with a highly 25 developed pore system (Cheng and Pan 2020; Lu et al. 2021). Water in coal can be divided into two 26 types: inherent moisture in coal matrix and free water in cleat system (Pan et al. 2010). The inherent 27 moisture mainly exists in the absorbed state, which can occupy the surface sites of coal pores due to the 28 physical adsorption and oxygen-containing functional group effects ). Moreover, the 29 existing of oxygen-containing functional groups (e.g. carboxylic and hydroxyl groups) adsorption sites   on methane adsorption capacity is also related to the metamorphic degree of coal. Dry coal shows a 41 trend of first falling and then rising with increasing coal rank, whereas the methane adsorption capacity 4 bituminous coal and the non-adsorbed water has no effect; in contrast, for anthracite coal, both the two 45 form of water have remarkably weaken on methane adsorption, which is considered to be related to the 46 difference in pore structure between the two coal samples (Wang et al. 2020a). Besides, water also has great weakening impact on methane desorption and diffusion capacities of coal (

57
Generally, coal with Ro,max less than 0.65% is classified as low-rank coal in China (Wang et al.

58
2017), which includes lignite and some long-flame coal. Low-rank coal generally features a higher 59 adsorbed water content than middle-high rank coals, because low-rank coal has a larger porosity,

64
In this paper, a typical long-flame coal was used to perform the following studies: (1) water adsorption characteristics of coal and pore structure of water-equilibrated coal at different relatively humidity, (2) the relationships between adsorbed moisture content and fractal dimension of coal pores,

67
(3) the impact of adsorbed moisture on methane desorption performances and its relations with fractal 68 dimension. The low rank coal sample was selected from the No. 2 coal seam in the Yuanzigou coal mine, coal seam on air dry basis is 3.60-11.28%. Several kilograms of fresh lump coal were collected from 75 the working face, and various sizes of coal particles were prepared by crushing and screening.

76
Following the ISO 17246:2010 standard, the coal sample with sizes of 0.074-0.20 mm was selected to 77 perform the proximate analysis by an automatic proximate analyzer. The vitrinite reflectance of coal 78 reflects the coalification degree, which was determined by following the ISO 7404-5:2009 standard.

79
The ash content on air dry basis (Aad), volatile matter on dry ash free basis (Vdaf), fixed carbon on air 80 dry basis (FCad), maximum vitrinite reflectance (Ro, max) are shown in Table 1.

81
Where M is the moisture content, %; mmoist is the weight of moist sample at a certain RH, g; and 96 mdry is the weight of dry sample, g. The water adsorption test was repeated twice and the mean values 97 of moisture content at different relative humidity were determined for further study.

98
Furthermore, the water adsorption characteristic of the studied coal was analyzed by a modified

106
Where M0 is the monolayer adsorption capacity, x is relative humidity, C and K are the adsorption 107 constants related to primary sites and secondary sites, respectively, and α represents the heterogeneity 108 of the adsorption system. The water adsorption amounts of primary and secondary sites have the

119
Under these conditions, the moist samples were frozen so that the loss of pre-adsorbed water could be 120 ignored. At the same time, the vacuum pumping step was omitted to avoid the loss of water in the 121 tested samples. The bulk desorption method (Zhang 2008) was applied to perform the methane desorption test of 125 coal samples with particle sizes of 0.2-0.25 mm. First, approximately 50 g of coal sample was put into 126 a coal sample tank. To reduce the loss of water from the coal sample as much as possible, the vacuum 127 pumping time was operated for no longer than 30 min. Then, the methane gas with a purity of 99.99% 128 was pumped into the coal sample tank to a gas pressure, and the coal sample tank was put into a stable 129 temperature water bath at 303.15 K. Subsequently, the gas pressure of the coal sample tank was 130 adjusted to a predetermined pressure (1 MPa). When the pressure gauge remained constant for 8 h, it is 8 free gas of the coal sample tank was removed, and then a methane desorption test of coal sample can be 133 performed. The test time was conducted for 120 min, and the methane desorption volume of coal at 134 regular intervals and the ultimate desorption volume (Q∞) were recorded. The moisture content of the 135 test sample was measured by the weighing method after the methane desorption test.
where Q t is the gas desorption volume at time t, cm 3 /g; A is the ultimate desorption volume, 144 cm 3 /g; t0 is the desorption time constant; and n is a coefficient.

145
In Winter's theory, when the gas pressure was removed, the change in the gas desorption rate with 146 time can be described by the following power function (Winter and Janas 1975): After mathematical integration, the relationship between methane desorption volume and time can 149 be obtained, as shown in the following equation: where Qt is the cumulative desorption volume at time t, cm 3 /g; V1 and Va are the methane 152 desorption rates at times t1 and ta, respectively, cm 3 /(g·min); and kt is a constant that reflects the degree

197
The fit results of the Airey-type and Winter-type equations are shown in

207
The errors of parameter A are 0.22-0.85% for methane pressure of 1 MPa, so parameter A is very close 208 to the measured value Q∞. Therefore, the Airey-type equation is more suitable to describe the methane 209 desorption behavior of coal.
210 Table 3 Results of methane desorption amounts in different desorption periods.

284
With a further increase in moisture content, the multilayer adsorption will occurs and the thickness of

302
Where V is the gas adsorption amount at adsorption equilibrium pressure P; P0 is the saturated gas is to 2, the smoother the pore surface is; and the closer D is to 3, the more complex the pore surface is.

307
Low-temperature nitrogen desorption isotherms is generally used to calculate the fractal 308 dimension, because the corresponding adsorption state is more stable (Li et al. 2015). In the region of  Table 5.

314
The calculation results obtained by the equation 'A=D-3' are between 2 and 3, which is more 315 reasonable than the calculation from the equation 'A=(D-3)/3'. It can be seen that D1 is 2.296-2.534 316 and D2 is 2.651-2.912. The relationships between the fractal dimensions D1, D2 and moisture content are shown in Figure   320 8. When the moisture content is less than 4.74% (corresponding to RH of 43%), D1 decreases rapidly,      software, as shown in Figure 9. The values of γ for the correlations of A and D1 or V1 and D1 are larger than 0.8, which are slightly higher than that for the correlations of A and D2 or V1 and D2.

398
The methane desorption parameters A and V1 reflect the ultimate and initial methane desorption

421
(1) Both N2-SSA and N2-PV decrease significantly with equilibrium moisture content increases, 422 and the adsorbed moisture has greater influence on coal's micropores with pore diameter less than 10 423 nm. In particular, when the adsorbed moisture content increases to 10.88% that is attained at 98% 424 relatively humidity, the micropores less than 4 nm almost disappear in the DFT-PSDs probably due to 425 the blocking effect of the formed water clusters and capillary water. However, the N2-SSA of 426 micropores (pore diameter < 10 nm) shows that it can still contributes most sites for gas adsorption on 427 water-equilibrated coal at 98% relatively humidity.

428
(2) The fractal characteristics analyzed by the FHH model shows that, when the equilibrium 429 moisture content is less than 4.74% (corresponding to a RH of 43%), D1 decreases rapidly while D2 430 shows a slight change, which is mainly due to the water adsorption on coal mainly occurs on the