Experimental Study of the Influence of Moisture Content on the Pore Structure and Permeability of Anthracite Treated by Liquid Nitrogen Freeze-Thaw.

The liquid nitrogen freeze-thaw (LN2-FT) method has been widely used to improve the coal permeability in the coalbed methane (CBM) production. However, the influence of moisture content on the permeability of coal treated by LN2-FT remains unclear, limiting the broad application of this technique. A novel seepage system was proposed to analyze the permeability evolution of anthracite coal samples treated by LN2-FT. Moreover, variations of the pore structure were analyzed using scanning electron microscopy (SEM), mercury intrusion porosimetry (MIP), and low-field NMR. The results showed that pores and fractures appeared on the coal surface after the LN2-FT treatment. As the moisture content of the coal increased, more pores and fractures tended to be formed during the LN2-FT treatment. The total pore volume, porosity, and average pore diameter of the anthracite coal after the treatment were 1.77, 2.44, and 5.58 times higher, respectively, than that of the raw coal. The change in the specific surface area exhibited three trends as the moisture content of the coal samples increased: a slow descent, a steady increase, and a rapid descent. Moreover, it was found that the LN2-FT treatment increased the connections between pores and fractures, improving gas migration in the coal. Furthermore, the LN2-FT treatment significantly increased the permeability of the anthracite coal samples. The higher the coal moisture, the higher the permeability of the coal samples after the LN2-FT treatment. Hence, the LN2-FT technique can substantially improve the permeability of coal reservoirs, providing essential information for the efficient utilization of CBM.

Introduction

China is rich in coalbed methane (CBM) reserves, and shallow CBM resources within 2000 m depth comprise 36.7 × 1012 m3. The efficient extraction of CBM, an efficient and clean source, can substantially reduce China’s dependence on fossil energy.[1−3] However, China’s CBM reservoirs usually are characterized by low permeability, which reduces the CBM extraction efficiency.[4−7] Accordingly, the permeability improvement of coal seams to increase the CBM extraction efficiency is a hot research topic. Many scholars have proposed techniques for increasing the permeability of high-gas low-permeability coal seams.[8−11] These techniques include the following types: protective layer mining, hydraulic techniques (hydraulic fracturing, hydraulic slotting, etc.), physical field excitation techniques (ultrasonic, microwave, etc.), and gas injection techniques (CO2, N2, etc.).[12−18] Although these measures have achieved satisfactory results, they also have disadvantages, such as a small zone of influence, large quantities of products, water pollution, and uneven pressure relief.[19,20]

Therefore, it is urgent to develop new techniques to increase permeability and improve CBM mining efficiency. In 1997, McDaniel et al.[21] carried out a field test using liquid nitrogen to induce cracking in the San Juan CBM production area. The research showed that the thermal stress formed by a sudden drop in temperature after the injection of liquid nitrogen into the coal seam transformed the internal pore structure of the coal body and caused damage. Grundmann et al.[22] found that the ultralow temperature of liquid nitrogen caused the tensile stress of the rock to exceed its tensile strength at the crack tip, leading to the failure of the rock fracture surface and the expansion and extension of the rock fracture. Subsequently, liquid nitrogen freeze–thaw (LN2-FT) fracturing has attracted extensive attention as an effective means to improve coal seam permeability.[23−25] Liquid nitrogen has a very low temperature, stable chemical properties, low production costs, and is easy to obtain.[26−28] The LN2-FT technique causes three types of damage to coal: in situ stress, expansion due to liquid water gasification, and freeze–thaw damage.[29] Freeze–thaw damage is the primary failure mode of coal. When the coal contacts the injected liquid nitrogen, the temperature of the coal decreases sharply, resulting in rapid freezing shrinkage of the coal matrix.[30] After liquid nitrogen injection, the coal absorbs the surrounding heat and melts. Thus, the liquid nitrogen cracking process consists of repeated freeze–thaw cycles of the coal seam. This process produces two types of failure forces: the expansion force of the gas–liquid and liquid–ice phase transition and uneven thermal stress of the coal due to low temperatures. The combined action of these two forces causes damage to the coal body, transforming the coal reservoir and improving its permeability. Multiple freeze–thaw cycles cause freeze shrinkage and tension fractures in the coal body, resulting in fatigue damage and propagation of internal fractures, which form a fracture network, thus improving the permeability.[31−34]

Studies have shown that the physical characteristics of coal, including the development degree and connections between pores and fractures, significantly affect the migration and utilization of CBM.[35,36] Several scholars studied the effect of the LN2-FT treatment on the mechanical properties and permeability in detail. It was found that liquid nitrogen affects the pore structure at the microscale.[37,38] When the low-temperature liquid nitrogen reaches the coal and rock, the water in the pores of the coal and rock rapidly condenses into ice due to the ultralow temperature. This phase change of the pore water can result in a 9% volumetric expansion inside the coal pores. Theoretically, this volumetric expansion can produce a stress of 207 MPa on the pore wall, exceeding the tensile strength of the coal and significantly expanding the pores.[39] This response improves the pore connectivity and pore size distribution (PSD) and increases the coal pore volume, resulting in desorption and diffusion of CBM. Under standard conditions, the volumetric expansion of liquid nitrogen can reach 694 times, causing damage to the coal and propagating coal fractures. Such fractures usually have a complicated fracture network, thereby increasing the permeability of the coal seam.[40]

Further investigations have revealed that several parameters, including the frequency and duration of freeze–thaw cycles, and the injection pressure, affect the pore structure and permeability of coal. As the frequency and duration of the freeze–thaw cycles increase, the crack width on the coal surface increases. Meanwhile, as the frequency of the freeze–thaw cycles increases, the damage to the pore structure increases, the compressive strength of the coal decreases, and the mechanical properties of coal degrade.[41−44] Moisture is an inherent part of coal and affects the ability to increase the permeability of LN2-FT cracked coal. In the field engineering of LN2-FT cracked coal, the water content in coal is different. However, the effects of moisture on the pore structure, freeze–thaw damage, and the migration characteristics of coal during LN2-FT cracking have been rarely investigated. The main target of the present research is to analyze the permeability evolution of anthracite coal with different moisture contents via a seepage experimental system. The pore structure of the samples is analyzed using scanning electron microscopy (SEM), mercury intrusion porosimetry (MIP), and low-field NMR. This study is expected to provide insights into the mechanism and application of the LN2-FT technology in CBM development engineering use.

Results and Discussion

SEM Results and Analysis

The pore structure of the raw coal samples and those subjected to the LN2-FT treatment were analyzed by a scanning electron microscope (Quanta 250, FEI Co.).[45,46] The surface of the raw coal samples was relatively smooth and had no pores and fractures, as shown in Figure . In contrast, pores and fractures appeared on the surface of the samples with various moisture contents subjected to the LN2-FT treatment, and the surface was rough and broken, as shown in Figure . The likely reason is that the coal matrix is affected by the low temperature. During the freezing process, shrinkage stress appeared in the coal sample, resulting in shrinkage deformation of the coal matrix. During the thawing process, expansion stress appeared in the coal sample, resulting in expansion deformation of the coal matrix. The cyclic treatment process of shrinkage and expansion resulted in different degrees of damage in the coal sample and destroyed the pore structure. The shapes and sizes of the newly generated pore differed from the original pores. They were mostly irregular circles or ellipses with flat and narrow pores. The distribution pattern was variable; some pores occurred in groups, some occurred in zones, and some single pores were observed. In addition, under the same LN2-FT treatment conditions, the pores and fractures were more abundant in the samples with more moisture after the LN2-FT treatment. It was found that the corresponding width and length of the newly formed fractures in the coal increased with the water content of the coal, and small branch fractures also formed around the main fractures. In addition, the higher the moisture content, the more connected fractures were observed. When the liquid nitrogen contacted the coal, the water in the pores rapidly froze, resulting in frost heave stress. The stress outstrips the tensile strength of the coal, thereby expanding the diameter and number of pores in the coal. Therefore, the more moisture in the coal, the greater the frost heave stress and more pores and fractures formed. These pores and fractures were conducive to gas desorption and migration, substantially enhancing the permeability of coal seams and increasing the rate of gas extraction rate.

Figure 1

SEM image of the raw anthracite coal samples.

Figure 2

SEM images of the anthracite coal samples with various moisture contents after the LN2 -FT treatment: (a) 0.23%, (b) 0.95%, (c) 1.83%, (d) 2.97%, (e) 3.65%, (f) 5.12%.

MIP Results and Analysis

MIP is a common method for measuring coal porosity in a pore diameter within a range of 0.003–1000 μm. The pressurized mercury overcomes the surface tension and enters the coal pores. It should be noted that the amount of injected mercury depends on the pore size and PSD. The smaller the pore size, the greater the required pressure is. The pore characteristics of the coal, including the pore size, PSD, pore volume, and specific surface area, can be obtained through data analysis, calculation, and the mercury curve.[47−49] Many pore classification methods for coal exist. This paper used the Hodort classification method with four levels, including macropores (i.e., d > 1000 nm), mesopores (i.e., 100 < d < 1000 nm), transition pores (i.e., 10 < d < 100 nm), and micropores (i.e., d < 10 nm).[50−52] In this regard, a mercury intrusion porosimeter (PoreMaster-60, Quantachrome) with a working pressure of 0.1–60000 psi (0.0007-413.8 MPa) and a volume accuracy of fewer than 0.0001 mL was used.

Figure illustrates the PSD of the coal samples with different moisture contents. The results indicate that the LN2-FT treatment had a substantial impact on the PSD of the samples. Most of the pores can be classified as micropores and transition pores, which is consistent with the structure of middle-high rank coal. However, as the moisture content increased, the proportion of micropores and transition pores decreased, while the number of pores in the other classes of pores increased significantly. The micropores and transition pores are the primary locations for gas adsorption and storage, while the mesopores and macropores are the main channels for gas seepage.[53] The PSD changed from complex before the LN2-FT treatment to simple after the treatment. These changes in the pore structure are not conducive to gas adsorption but facilitate the desorption and migration of CBM.

Figure 3

PSD of the anthracite coal samples with different moisture contents: (a) 0.23%, (b) 0.95%, (c) 1.83%, (d) 2.97%, (e) 3.65%, (f) 5.12%.

Table lists the values of the pore structure parameters after the LN2-FT treatment. Figure illustrates the increased amplitude of the pore parameters in the anthracite coal samples. The results indicate that the pore structure of the coal with different moisture contents has changed after the LN2-FT treatment to varying degrees compared to the original sample. It is found that as the moisture content increased, the total pore volume after the LN2-FT treatment increased. The higher the moisture content, the more pronounced the change, and the larger the total pore volume after the LN2-FT treatment (Figure a). The total pore volume of the coal samples with a moisture content of 5.12% before and after the LN2-FT treatment was 239 × 10–4 and 424 × 10–4 cm3/g, respectively, indicating an increase of 1.77 times. Similarly, as the moisture content increased, the porosity increased. The porosity of the coal samples after the treatment was 2.44 times that of the initial sample, indicating an improvement in the connectivity between the pores in the coal and is conducive to the gas migration.[54,55] The trend of the average pore diameter was consistent with that of the total pore volume and porosity. The average pore diameter was 5.42 nm before the treatment and 17.11 nm after the treatment, and the rate of increase in the average pore diameter increased with the moisture content. The average pore diameter of the sample with 5.12% moisture after the LN2-FT treatment was 29.27 nm, which was 5.58 times that of the initial samples. Accordingly, it was inferred that the LN2-FT treatment resulted in the expansion of the pores at all levels.

Table 1

MIP Results of the Pore Parameter Values of the Anthracite Coal Samples with Different Moisture Contents

		total pore volume (10–4 cm3/g)		total porosity (%)		average pore diameter (nm)		total specific surface area (m2/g)
sample	moisture content (%)	raw coal	LN2-FT	raw coal	LN2-FT	raw coal	LN2-FT	raw coal	LN2-FT
QJ	0.23	229	261	5.21	6.43	5.25	7.61	15.23	13.32
	0.95	233	313	5.23	7.60	5.43	9.13	14.18	17.24
	1.83	242	357	5.26	8.99	5.51	13.34	16.08	17.56
	2.97	253	372	5.31	9.39	5.57	18.52	15.12	11.23
	3.65	245	391	5.28	10.07	5.47	24.83	13.32	8.45
	5.12	239	424	5.25	12.83	5.34	29.27	14.24	5.81

Figure 4

Increased amplitude of pore parameters in anthracite coal samples after LN2-FT treatment: (a) total pore volume, (b) total porosity, (c) average pore diameter, and (d) total specific surface area.

The change in the specific surface area of the samples with different moisture contents subjected to the LN2-FT treatment was relatively complex. As the coal moisture content increased, the following three trends were observed:

Slow descent:

At a low moisture content, the specific surface area exhibited a slow decline. The specific surface area of samples subjected to the LN2-FT treatment was slightly lower than that of the initial samples, but the difference was relatively small. The reason is that the water in the coal preferentially occupies the micropores and transition pores. At a low moisture content, most of the water is in the micropores and transition pores. The expansion stress resulting from the freezing of the pore water due to the low-temperature liquid nitrogen mainly destroyed the micropores and transition pores, resulting in the expansion of some of the micropores and transition pores and a decrease in their volume. It was also observed in Figure a that the micropore volume of the coal sample treated with liquid nitrogen was lower than that of the raw coal when the moisture content was low. Because micropores and transition pores comprised the largest proportion of the specific surface area of coal, the total specific surface area decreased.

Steady rising:

As the moisture content increased to 0.95 and 1.83%, the specific surface area exhibited a slow upward trend. Before the LN2-FT treatment, the specific surface area of the sample increased from 14.18 to 17.56 m2/g. The micropores and transition pores in the coal could not store all of the water; thus, some of the water occupied the mesopores. Since the frost heave force was stronger at a higher moisture content, some micropores expanded into transition pores and mesopores, and some transition pores changed into mesopores or macropores. Considering cyclic thermal stress of the coal, the ultralow temperature of the liquid nitrogen impacts the coal skeleton. However, the thermal conductivity of coal is low. When the internal temperature of the coal was uneven, the coal particles in the adjacent temperature zone deformed to varying degrees, leading to internal damages. Because of thermal stress and the freezing-induced expansion, the coal matrix opened due to repeated shrinkage and expansion. These changes increased the number of measurable micropores and the overall volume of the micropores in the coal.

Rapid descent: As the moisture content of samples continued to increase (>1.83%), water occupied all types of pores (including some macropores). When the coal contacted the liquid nitrogen, the water in the pores froze and solidified rapidly due to a large water volume. The expansion force was much greater than the thermal stress caused by the tensile strength and temperature change of the coal body. Therefore, most of the pores increased in volume. During this transformation, the increase in volume and number of newly formed micropores was smaller than the decrease caused by the transformation from micropores to larger pores, resulting in a reduction in the micropore volume and number. The increase in the number and volume of transition pores was attributed to the transformation of some micropores into transition pores. These changes resulted in the rapid reduction in the total specific surface area of the pores in coal. The decrease in the total specific surface area of the pores reduced the number of gas adsorption sites, which was not conducive to gas storage in coal.

Variation of the pore structure characteristics of the coal samples with increasing moisture content after LN2-FT treatment indicated that LN2-FT treatment increased the pore volume. It was found that as the moisture content increases, the corresponding effect is more pronounced so that the gas migration improves in the coal.

NMR Results and Analysis

A low-field NMR test is different from MIP. It measures the T2 relaxation time of the fluid in the coal pores to obtain the pore distribution and connectivity and various physical parameters of the coal pores.[56,57] This method is fast and nondestructive, provides rich information, enables continuous detection, and has a wide measurement range. The correlation between the pore size and T2 can be mathematically expressed in the form belowwhere T2 represents the transverse relaxation time (ms), ρ represents the transverse surface relaxation strength (μm/ms), S represents the surface area of pore (cm2), V represents the volume of pore (cm3), FS represents the pore shape coefficient (fracture, FS = 1; cylindrical pore, FS = 2, spherical pore, FS = 3), and r represents the pore diameter.

Equation indicates that the pore radius r of the coal is positively proportional to the transverse relaxation time T2. Therefore, the NMR T2 curve reflects the distribution of the pore diameter. The larger the T2 value, the larger the pore diameter is and vice versa. Three peaks typically occur in the T2 spectrum. The value range of the micropores and transition pores is the interval of the first peak (the T2 value is 0–10 ms), the value range of the mesopores and macropores is the interval of the second peak (the T2 value is 10–100 ms), and the value range of the fractures is the interval of the third peak (the T2 value is greater than 100 ms).[58,59] In addition, in the NMR T2 curve, the magnitude of the amplitude is related to the number of pores. The greater the amplitude of the curve, the larger the number of pores corresponding to a particular pore diameter. In this paper, a Meso MR23-060H-I low-field NMR spectrometer was used to measure the pore parameters of the coal samples having various moisture contents before and after the LN2-FT treatment. The results are shown in Figure .

Figure 5

T2 values of the anthracite coal samples with different moisture contents: (a) 0.23%, (b) 0.95%, (c) 1.83%, (d) 2.97%, (e) 3.65%, (f) 5.12%.

The T2 spectrum of the coal sample showed peaks corresponding to micropores and transition pores and a peak corresponding to mesopores and macropores before the LN2-FT treatment. The third peak corresponding to fractures was missing in the NMR T2 spectrum of the coal samples with moisture contents of 0.23, 0.95, 2.97, and 5.12%. In contrast, there were three peaks in the T2 spectrum of the coal samples with moisture contents of 1.83 and 3.65%. The signal amplitude of the peak indicated common characteristics. The signal amplitude of the first peak was the highest, followed by the second peak and the fracture peak. This result indicated that the micropores and transition pores of the raw anthracite were well developed, followed by the mesopores and macropores, whereas the fractures were not developed. The spectral peak amplitude differed for the coal samples because the coal is a multiporous and heterogeneous medium, and the PSD is different in different samples. The peaks indicating fractures in Figure c,e may have been caused by external factors during sample preparation.

Figure reveals that the T2 curves of different samples have different amplitudes of the peak after liquid nitrogen injection. It is observed that the number of micropores and transition pores in the coal samples decreased after the liquid nitrogen treatment. In contrast, the signal amplitude of the second peak of samples with different moisture contents was higher after the treatment, indicating that the micropores and transition pores expanded into mesopores and macropores, increasing the number of mesopores and macropores. A third peak occurred in all coal samples, indicating that the internal and surface pores of the anthracite coal samples formed a network, large fractures appeared, and the coal samples were broken. It was found that the higher the moisture content, the higher the signal amplitude of the third peak. This phenomenon may be attributed to damage of samples originating from the combined action of the frost heave force and thermal stress, increasing the compression degree. In addition, the continuity between the second and third peaks of the coal samples after the LN2-FT treatment was good, suggesting high connectivity between mesopores, macropores, and fractures, thereby desorbing the gas in coal.[60,61] The mesopores and macropores correspond to seepage pores, and the micropores and transition pores correspond to adsorption pores. The obtained results demonstrate that the corresponding gas diffusion in the coal improved as the number of seepage pores increased. An increase in the number of fractures provided channels for the migration and flow of gas in the coal. These changes were beneficial to the development of CBM.

Permeability Results and Analysis

Studies have shown that coal permeability is an important indicator of the difficulty of gas migration and largely depends on the pore size and connectivity.[60,62] Moreover, the freeze–thaw damage of the liquid nitrogen affects the permeability and the pore structure.

Figure illustrates the influence of the LN2-FT treatment on the permeability of the samples with different moisture contents, indicating that the LN2-FT treatment significantly affected the permeability. It was found that the permeability curve changed to a V-shape as the gas pressure increased, i.e., a decrease followed by an increase. The gas pressure at the inflection point was about 1.0 MPa. This phenomenon is caused by the Klinkenberg effect.[63−65] As the gas pressure increased from 0.5 MPa to the inflection point, the amount of gas adsorbed by the coal samples increased. The adsorbed gas layer on the surface of the pores thickened, thereby enhancing the slip flow of the gas molecules on the pore fissure wall of the coal body. The effective seepage channel of gas was reduced, the migration resistance of gas molecules was increased, and the gas flow speed was significantly slower, resulting in the decline of permeability. When the gas pressure reached the inflection point, the gas adsorption speed was equal to the gas desorption speed, reaching the equilibrium limit. As the gas pressure exceeded the inflection point, the Klinkenberg effect gradually lost its dominant position in controlling the permeability of coal samples relative to the larger gas pressure, so the permeability began to rise. In addition, at the same gas pressure, the greater the moisture content of the samples before the LN2-FT treatment, the lower the permeability. The reason is that water is preferentially adsorbed on the pore surface, blocking the gas flow channels. As the water content in samples increased, the corresponding gas flow resistance in the coal body increased, preventing the gas from escaping the samples. It was inferred that the LN2-FT treatment increased the permeability of the coal samples. Due to the nitrogen-induced freeze–thaw damages during the LN2-FT treatment, the pores in the coal samples were well developed, the cross-sectional area of the gas seepage channels was large, and gas migration occurred readily. The obtained results in this regard are presented in Figure . Compared to the initial samples, the average increase in permeability of the samples after LN2-FT treatment with moisture contents of 0.23, 0.95, 1.83, 2.97, 3.65, and 5.12% was 69, 103, 141, 192, 268, and 362%, respectively. The higher the water content of the coal samples, the larger the freeze–thaw damage by the liquid nitrogen and the better the development of the pore channels in the coal body. Thus, the larger the cross-sectional area of the gas seepage channel, the higher the gas flow.

Figure 6

Permeability of the anthracite coal samples with different moisture contents: (a) before and (b) after the LN2-FT treatment.

Figure 7

Increased amplitude of the permeability of the anthracite coal samples with different moisture contents after the LN2-FT treatment: (a) 0.23%, (b) 0.95%, (c) 1.83%, (d) 2.97%, (e) 3.65%, (f) 5.12%.

Conclusions

In this study, the effect of the moisture content on the evolution of the pore structure in anthracite coal samples subjected to the LN2-FT treatment was investigated using SEM, MIP, and NMR. A gas seepage experimental device was utilized to analyze the influence of the moisture content on the permeability of the anthracite coal samples in the LN2-FT treatment. The main conclusions of this paper were summarized as follows:

The SEM results revealed that some pores and fractures appeared on the surface of the anthracite coal samples after the LN2-FT treatment. As the moisture content increased, the size and width of the newly formed pores and fractures increased.

The MIP and NMR results demonstrated that the proportion of micropores and transition pores was lower after the LN2-FT treatment. The total pore volume, porosity, and average pore diameter of the anthracite coal samples after the LN2-FT treatment were 1.77, 2.44, and 5.58 times larger, respectively, than those before the treatment. The change in the specific surface area with an increase in the moisture content of the coal was relatively complex. In addition, LN2-FT treatment improved the connectivity between the pores and fractures. These changes were beneficial to the gas migration in the coal and critical for enhancing CBM recovery.

The performed experiment revealed that the permeability of the anthracite samples was significantly higher after than before the LN2-FT treatment under the same experimental conditions. The higher the moisture content in the coal, the higher the permeability of the coal samples was after the LN2-FT treatment. Hence, the moisture is conducive to improving the permeability of the anthracite samples during LN2-FT treatment.

The performed analyses demonstrate that the LN2-FT treatment improved the pore structure and permeability of coal reservoirs, which is conducive to CBM exploitation.

Experiment Section

Sample Preparation

To perform the experiments, anthracite samples were prepared from Qianjin coal mine, Bijie coalfield, Guizhou, China. To this end, large coal samples were packed in a cling film and quickly transferred to the laboratory for processing and analysis. A large coal block was selected as the sampling material, and its structure was not damaged during sampling. The Φ50 × 100 mm standard coal samples were drilled and cut using an automatic drilling and coring machine, and both ends of the prepared samples were ground to obtain a smooth surface. The height of the coal samples was compatible with the stratification direction to ensure that the liquid nitrogen was injected in the horizontal stratification direction of the coal seam. An HTGF-9000 automatic industrial analyzer was used for the elemental analysis of the coal samples on the basis of the “Chinese national standards GB/T 212-2008” and “GB/T 476-2001”. A Leica DM4500P polarizing microscope was used to analyze the vitrinite reflectance of the coal samples on the basis of the “Chinese national standard GB/T 6948-2008”. The properties of the coal samples are summarized in Table , and the mineral types and contents of the coal samples are presented in Table .

Table 2

Properties of the Coal Samples

			proximate analysis (wt %)				ultimate analysis (%)
sample	coal rank	R0max	Mad	Aad	Vdaf	FCad	Oad	Cad	Had	Nad
QJ	anthracite	3.03	2.32	16.17	6.44	75.07	2.96	92.63	3.09	1.32

Table 3

Mineral Type and Content of the Coal Samples

		mineral type and content (wt %)
sample	coal rank	kaolinite	calcite	pyrite	quartz	ankerite	marcasite	plaster
QJ	anthracite	11.32	3.61	2.44	1.65	0.96	0.63	1.32

In the present study, samples were prepared as follows:

Six groups of cylindrical standard coal samples (a–f) were prepared.

Using a vacuum drying oven for drying treatment, the dry weight md of the coal samples was obtained using a scale.

The dried coal samples were placed in a vacuum saturation machine for 48 h to ensure water saturation of the samples. The weight of the saturated coal samples ms was obtained.

The saturated coal samples were dried at 105 °C. The heating times of different groups of samples were set to 1, 2, 4, 6, 8, and 10 h, respectively. Subsequently, the weight mw of coal samples was acquired. The moisture content w of the samples was calculated as follows

Test Setup

All experiments were carried out in a custom-made coal and rock triaxial seepage experimental device, consisting of a gas cylinder, a vacuum pump, a coal sample gripper, a constant-temperature control device, a servo press, a digital gas flowmeter, and a precision pressure gauge. The gripper containing the coal sample consists of fluororubber inner and stainless-steel outer layers and holds the coal sample. The outer layer was a sleeve that had an air inlet, air outlet, and inlet holes for the liquid. During the experiment, a hydraulic oil pump was used to inject hydraulic oil around the inner rubber sleeve through the inlet hole for confining pressure loading. The confining pressure range was 0–30 MPa and the axial pressure was provided using a servo press. The axial pressure loading setting was controlled by a computer connected to the servo press. The axial pressure range was 0–70 MPa. All experiments were conducted in a vacuum environment. To this end, the gas was extracted to reach the desired vacuum level using a vacuum pump. A high-pressure gas cylinder was installed at the inlet port to inject gas into the gripper, and the air outlet on the coal sample gripper was connected to a digital gas flowmeter to measure the airflow at the air outlet. The working principle of the seepage experiment is illustrated in Figure .

Figure 8

Configuration of the experimental seepage system.

Experimental Process

The flowchart in Figure shows the experimental process. The coal samples were subjected to the LN2-FT treatment consisting of a cycle of cold immersion for 30 min and recovery at room temperature for 1 h. Some of the coal samples from each group were placed into the vacuum LN2 cup, quickly pouring the LN2 over the samples to immerse them and covering them with thermal insulation. The temperature of the coal body rapidly decreased from the surface to the inside, reaching −195.8 °C. After the designated cold-soaking period, the vacuum LN2 cup was opened to let the samples reach room temperature. The permeability of the samples was tested as follows:

Figure 9

Experimental procedure.

The coal samples were placed in the holder and the gas cylinder was connected. After ensuring that all valves in the system were closed, the valve on the He gas storage bottle was opened to fill the system with 3 MPa high-purity He, and let it sit for 24 h. If the pressure gauge was stable, the airtightness of the system was adequate.

The temperature of the constant-temperature water bath system was set to 30 °C. After the system reached this temperature, the predetermined axial pressure and confining pressure were applied using the liquid injection pressure servo machine and hydraulic oil pump. To prevent gas leakage during the experiment, the confining pressure should exceed the gas pressure.

The vacuum pump was started and ran for 12 h to ensure vacuum in the pipeline. Gas was injected from the gas cylinder into the experimental system. The injection pressure reached the predetermined value and remained stable for 12 h. The pressure changes of the gas adsorbed by the coal samples were observed through the pressure gauge. If the pressure decreased, the gas cylinder was opened to increase the gas adsorption pressure. This process continued until the pressure did not decrease within 30 min, indicating that the coal sample had reached the adsorption limit.

Subsequently, the valve was opened at the outlet of the gripper. After the gas outflow at the outlet was stable, the digital flowmeter software automatically acquired the flow data.

The gas pressure (pressures of 0.5, 0.8, 1.0, 1.2, and 1.5 MPa were used in the experiment) was changed and steps (3) and (4) were repeated.

Steps (1)–(5) were repeated using the remaining samples.

Analysis Method

Coal is an extremely complex porous medium with many pores, and the PSD ranges from the millimeter to the nanometer scale. Pores with different sizes affect the gas adsorption and desorption, gas diffusion, and gas seepage behavior, thereby affecting the permeability of the coal seam.[66−68] Numerous researchers investigated the pore structure for coal using image observations and fluid injection. Image observation methods include using image slices to obtain quantitative statistics on holes and fractures in a small field of view using electron microscope imaging techniques such as SEM, atomic force microscopy (AFM), and transmission electron microscopy (TEM).[69−71] It is worth noting that measuring accuracy is substantially affected by the mode of handling, field of view, and device resolution. The image observation method is typically used for the qualitative observation and analysis of pores. Fluid injection is mainly used for the quantitative measurements of the pore structure of open pores in coal. Common methods include MIP, gas adsorption methods (N2 and CO2), and NMR spectroscopy.[72−74] These methods have superior advantages such as a wide measurement range and high precision, but they do not provide information on closed pores. The ranges of pore sizes that can be obtained from different methods are variable, as shown in Figure . Accordingly, it is challenging to determine the properties of the coal pore structure using a single characterization method. Therefore, SEM, MIP, and NMR were applied to analyze the pore structure of anthracite samples before and after the LN2-FT treatment.

Figure 10

Classification of coal pore structures using different methods.

Coal permeability is a crucial index of gas flow in coal. In this regard, Darcy’s seepage law was applied to calculate the coal permeability and determine the treatment effect on the seepage performance.[18,75] The permeability K was calculated according to the difference between gas pressure P1 at the inlet and gas pressure P2 at the outlet of the coal sample, atmospheric pressure P0, gas viscosity μ, stable gas flow qv, length L, and cross-sectional area A. It was calculated as follows