Permeability Evolution of Naturally Fractured Coal Injected with High-Temperature Nitrogen: Experimental Observations

: The permeability of more than 70% of coal seams in China is less than 1 mD, creating difﬁ-culties in recovering underground coal methane. Therefore, a new technology of high-temperature nitrogen (HTN 2 ) injection into the coal seam was proposed to improve the coal permeability and gas extraction rate. In this paper, the effects of the N 2 temperature, injection pressure and cycle number on the permeability of naturally fractured coking coal has been investigated. When HTN 2 was injected into coal samples, the results indicated that the permeability decreased over time in the beginning, suddenly increased to a large value, and was subsequently maintained in a relatively stable range. The maximum permeability ratio increased with the rise of the N 2 temperature and injection pressure. An analysis indicated that the increase of coal permeability was the result of the increase of the global coal strain caused by thermal expansion and the adsorption-induced expansion. The maximum permeability ratios in various cycles of multicycle N 2 injection into the coal sample were all greater than 1.0 while progressively declining. Obviously, the alternating stress was conducive to the further expansion of the coal fractures to increase the coal permeability. However, on the basis of the ﬁrst period of expansion, re-expansion was difﬁcult and required more energy. The effects of multicycle N 2 injection on coal permeability have been considerably improved when compared with N 2 injection with only one cycle. The research results are helpful for rapidly extracting methane and guaranteeing mine safety.


Introduction
Coalbed methane (CBM) is a precious nonrenewable energy [1]. The extraction of CBM can control gas disasters in mines and protect the environment [2]. However, more than 70% of Chinese coal seams' permeability is lower than 1 mD [3], and it is difficult to extract CBM. Plenty of studies on the drainage efficiency improvement of CBM have been conducted [4,5]. Gas injection is one of the main methods for increasing CBM production [6], and CO 2 and N 2 are the most common gases for injection into coal seams. CO 2 -Enhanced CBM (CO 2 -ECBM) production involves the injection of CO 2 into coal seams to promote the desorption of CBM while simultaneously reducing the output of greenhouse gases into the atmosphere [7,8]. However, CO 2 -ECBM can only be used in deep un-mineable coal seams [9]. Zhou et al. found that N 2 injection caused moderate increases in coal permeability [10]. Studies have shown that the coal desorption capacity for CBM increased with an increase of temperature [11][12][13]. Thus, a new technology of HTN 2 injection into the coal seam has been proposed.
The temperature effect on the desorption and adsorption of CBM has been studied. Li et al. found that the methane permeability increased with the temperature when the effective stress was lower than the thermal stress; otherwise, the methane permeability increased when the temperature decreased [14]. Hu et al. discovered that the N 2 permeability decreased at first, then rose, and finally decreased when the coal temperature was within 100 • C. 50 • C and 80 • C were the lowest and the highest temperature, respectively [15]. Zhao et al. found that the N 2 permeability changed during the heating process; it increased rapidly with the temperature when the percolation threshold value of the coal mass was smaller than the overall porosity [16]. Perera et al. found that the effect of the temperature on the N 2 permeability was not obvious [17]. Zhu et al. found that the thermal expansion and sorption-induced swelling caused by temperature change could affect coal permeability significantly [18]. Peng et al. found that when the diffusivity was low, the change of coal permeability with time displayed a "V" shape. Existing studies mainly focus on the effects of normal temperature N 2 injection or coal temperature on permeability, and the question of how coal permeability changes under N 2 injections with different temperatures and pressures into coal is still not studied [19].
In this study, the effects of the gas temperature, pressure and cycle number on the permeability of naturally fractured coking coal with HTN 2 injections were investigated. A testing system for HTN 2 injections was developed. We tested six specimens from the Pingdingshan Coalfield in China.

Testing System
A testing system for HTN 2 injection for permeability tests was established, as shown in Figure 1. The system can continuously monitor the temperature and pressure of the HTN 2 injection, the axial stress and the lateral stress, and the deformation displacement of the coal sample, etc. During the experiment, a mass flowmeter was used to measure the flow rate at the outlet.

Sample Preparation
The naturally fractured coking coal samples used in the test were taken from the #J15 coal seam of the Pingdingshan Coalfield in China. The physical properties of the coal are shown in Table 1. The large coal blocks were cored, cut and grinded, after which samples with a diameter of 50 mm and a length of 100 mm were obtained. Some of the natural fractures in a typical coal sample are shown in Figure 2a, and fractures distribution of the coal sample after HTN 2 injection are shown in Figure 2b.

Experimental Process
According to the actual situation at the Pingdingshan Coalfield, the geostress pressure is 10 MPa. The experimental conditions of the coal samples are shown in Table 2. The testing process was as follows: (1) Coal sample S-60-4 was placed on the indenter located at the bottom of the testing machine and connected to the HTN 2 pipelines.
(2) The pressure chamber was sealed. After the pressure chamber was filled with oil, the oil was heated to 30 • C by the heating ring. Then, hydrostatic pressure was applied to the coal sample S-60-4 at 3 MPa/min until it reached 10 MPa; the confining pressure and axial pressures were kept constant. The heating controller and pressure-regulating valve were adjusted to ensure that the N 2 temperature was 60 • C and that the N 2 injection pressure was 4 MPa.
(3) The flow rate and gas injection time were recorded during the test. The beginning time was defined as when the flow rate of the outlet was observed and remained constant. When the flow rate at the outlet suddenly increased and remained constant (the change of flow rate was within 5%) for at least 15 min, the N 2 injection was stopped with the temperature and pressure. The test was then completed. (4) Steps (1), (2) and (3) were repeated with the remaining samples. Currently, the measurements of the coal sample permeability are mainly via steadystate and transient methods. The flow of N 2 in the coal sample is laminar, and Ranjith et al. used a steady-state method to measure the N 2 permeability of low-permeability coal. Based on this, the steady-state method was chosen to perform this experiment [20]. The HTN 2 permeation through the coal sample is assumed to be an isothermal process, and the ideal gas law is applicable. The N 2 permeability is calculated by [21]: In this paper, the permeability k is defined in SI units of m 2 . The conversion factor from m 2 to mD is expressed by 1 mD = 10 −15 m 2 .

Permeability Evolution under N 2 Injection
Figure 2b presents the surface fracture growth of sample S-100-4 after the N 2 injection. New fractures were visually observed from the sample surfaces. The evolution of the coal permeability ratio with time is shown in Figures 3 and 4. As seen from the figure, during the N 2 injection process, the permeability ratio decreases over time in the beginning, suddenly increases to a large value, and is then maintained within a relatively stable range. To explain the experimental phenomena, the existing model has been cited and revised. The permeability of a natural fracture depended on the fracture permeability, so the fracture permeability was used to analyze the evolution of the coal permeability [19]. Appendix A presents the effective strain of the fractures and the permeability variation equations for N 2 .  When the N 2 injection begins, the matrix pressure is lower than the fracture pressure. Thus, the coal global strain is lower than the fracture local strain [22]. On the fracture surface, the adsorption expansion and thermal expansion occur. According to Equations (A1), (A2) and (A4), the rise of the sorption-induced strain of fracture ∆ε f s and the thermal strain of fracture ∆ε f T cause the decline of the fracture local strain ∆ε f l . This decline then leads to the decrease of the effective volumetric strain ∆ε f e and the decrease of the fracture permeability k f , forming a decline phase with a "V" shape.
According to Harpalani and Chen, the coal effective stress subjected to fluid pressure can be calculated using Equation (2) [23]: With a continuous HTN 2 injection, the HTN 2 pressure propagates into the coal matrix, and the matrix pressure rises while the effective stress decreases (Equation (2)). According to Equations (A1), (A3) and (A4), both the thermal expansion of the matrix α mT (1 − φ f )T and the adsorption-induced expansion of the matrix (1 − φ f )ε ms increase, resulting in the increase of the coal global strain ε v . Thus, the effective volumetric strain ∆ε f e and fracture permeability increase, forming an increase phase with a "V" shape. Therefore, the test results of the coal permeability evolution obey the "V" shape.

Effect of N 2 Temperature on Permeability Evolution
With a continuous HTN 2 injection and under the interaction of HTN 2 and the temperature difference of the coal sample, heat transfer occurs between the coal and HTN 2 . The temperature of the coal sample gradually increases. The results revealed that when the coal temperature increased from 30 • C to 70 • C, the triaxial compressive strength of coal was reduced by more than 42%, and the axial strains were reduced by more than 35% [24]. That is to say, the capacity for ductile deformation of coal was weakened with the increase of the temperature.
The higher the N 2 temperature is, the more intense the thermal expansions of the fracture surface and the matrix become. In accordance with Equations (A1), (A2) and (A4), due to the thermal expansion of the fracture surface, the fracture aperture is reduced. The greater permeability attenuation in the initial stage is therefore a result of the higher N 2 temperature. When the N 2 injection pressure is 4 MPa, the permeability attenuation values in the initial stage, which correspond to N 2 temperatures of 60 • C, 80 • C and 100 • C, are 34.3%, 41.3% and 46.9%, respectively ( Figure 5). With a continuous N 2 injection, the coal global strain increases. According to Equations (A1), (A3) and (A4), the higher the N 2 temperature, the more intense the integral thermal expansion becomes, which increases the fracture aperture. Thus, the higher the N 2 temperature, the greater the maximum permeability ratio. When the N 2 injection pressure is 4 MPa, the coal maximum permeability ratio values at N 2 temperatures of 60 • C, 80 • C and 100 • C are 4.48, 8.38 and 14.25, respectively ( Figure 5).

Effect of N 2 Injection Pressure on Permeability Evolution
The greater the N 2 injection pressure is, the greater and the more intense the compression degree of the matrix in the initial stage will become. p m is the matrix pressure of coal, p f is the fracture pressure of coal, and K m is the bulk modulus of the coal matrix. According to Equations (A1), (A2) and (A4), the smaller the magnitude of the fracture, the smaller the aperture and the permeability attenuation. When the N 2 temperature is 60 • C, the permeability attenuation in the initial stage is 34.3% at a 4 MPa injection pressure, 30.5% at a 6 MPa injection pressure and 25.4% at an 8 MPa injection pressure ( Figure 6). The greater the N 2 injection pressure, the greater the fracture pressure and matrix pressure; thus, the effective stress will be smaller, while the adsorption expansion of the fracture surface and the matrix (i.e., ε ms and ε f s ) will be larger. According to Equations (A1), (A3) and (A4), the coal global strain increases and the fracture aperture rises, resulting in an increase of the coal permeability. Therefore, the greater the injection pressure, the greater the maximum permeability ratio. When the N 2 temperature is 60 • C, this corresponds to a maximum permeability ratio of 4.48 for a 4 MPa injection pressure, 9.07 for a 6 MPa injection pressure and 10.37 for an 8 MPa injection pressure ( Figure 6).

Effect of Cycle Number on Permeability Evolution
Figure 7a-d shows the permeability ratio evolution with time under an N 2 injection with three cycles. As seen from the figure, the variation of the coal permeability ratio during various cycles of N 2 injection is the same during the test; they are all in the shape of a V, but the degree of change is constantly reduced. Under the multicycle N 2 injection, the coal sample will be affected by alternating stress. Practice shows that the failures and the static stress caused by the alternating stress are completely different. The alternating stress, the stress that is generated when a fatigue rupture of the coal occurs, is always lower than its static strength. For coal with initial pores and fractures, under the effects of alternating stress, fractures will be further produced and extended, which may well increase the coal permeability. The Paris law, which was proposed by Paris and Erdogan, is used to predict the fracture growth [25]: From this, the growth of the fatigue fractures is related to the maximum and minimum of the alternating stress. In the multicycle N 2 injection process, when alternating the stress amplitude, the coal suffered changes and the expanded degree of fractures could be studied by using the accumulated damage theory.
Palmgren-Miner's rule determines the fatigue strength of coal. When there are f different stress ranges, S i (1 ≤ i ≤ f ) acts on a coal structure, and each S i contributes n i cycles. Then, the sum of the fatigue damage is [26]: During the test, f = 1. In other words, the expansion of the coal fracture always occurs when N 2 is injected into the coal sample. According to Palmgren-Miner's rule, the multicycle N 2 injection is propitious to the re-expansion of coal fractures and the increase of coal permeability. The maximum permeability ratio of every period is greater than 1.0. However, based on the original fractures, the re-expansion of the coal volume requires more energy, and thus the mutation degree is declined.
The final permeability ratio of the coal samples S-60-4 and SU-1 was compared. It was observed that the final permeability ratio of sample SU-1 was increased by 101% when compared with sample S-60-4. In other words, the coal permeability increased with the increase of cycle numbers. This is meaningful for the fast gas extraction of low-permeability coal seams. After extracting CBM for a long time, the CBM concentration of boreholes will decrease from 6% to 20%. Such a low CBM concentration can lead to CBM not being usable, which will cause greenhouse effects if CBM is released directly. Therefore, the HTN 2 can be repeatedly injected into the coal seam to realize a CBM extraction with a high CBM concentration and a large flow until the CBM extraction rate meets the national standards.
The permeability of the #J15 coal seam of the Pingdingshan Coalfield in China is 0.0019 md (less than 1 md), the relative gas emission is 123.376 m 3 /t, the absolute gas emission is 9.573 m 3 /min, the gas pressure is 1.5-2.0 MPa, the gas content is 20-22 m 3 /t, and the gas drainage is difficult. In order to make the coal seam meet the requirements of safe mining as soon as possible, the conventional HTN 2 injection method can be used to increase the permeability of coal, followed by gas drainage. When the gas concentration is lower than 20%, HTN 2 can be injected into the coal seam many times, and the permeability of the coal seam can be further increased so as to realize the safe and rapid mining of coal.

Conclusions
The effects of the N 2 temperature, injection pressure and cycle number on the permeability of naturally fractured coking coal were investigated. Our main results are as follows: (1) When HTN 2 was injected into the coal sample, the matrix pressure was lower than the fracture pressure in the beginning, and the coal global strain was lower than the local fracture strain. On the fracture surface, adsorption expansion and thermal expansion occur. Then, the rise of the sorption-induced strain of the fracture and thermal strain of the fracture cause the decline of the fracture local strain, which leads to decreases in the effective volumetric strain and fracture permeability. With a continuous HTN 2 injection, the gas pressure propagates into the matrix, and the matrix pressure rises. Both the thermal expansion and the adsorption expansion of the matrix increase, resulting in the increase of the coal global strain. Thus, the effective volumetric strain and fracture permeability increase. The maximum permeability ratio increases with the rise of the N 2 injecting pressure and temperature.
(2) When multicycle N 2 was injected into the coal sample, the maximum permeability ratios in various cycles were all greater than 1.0 while progressively declining. It is thus clear that alternating stress is conducive to the further expansion of coal fractures and an increasing coal permeability. On the basis of the first period of expansion, re-expansion is difficult and requires more energy. The effects of multicycle N 2 injection on coal permeability has been considerably improved when compared with N 2 injection with only one cycle.