The Time-Varying Variation Characteristics of Methane during Nitrogen Injection Process: An Experimental Study on Bituminous Coals

,


Introduction
Coalbed methane is a fossil energy associated with the natural evolution of coal [1].The development of coalbed methane cannot only ensure the safe production of coal mines but also alleviate the increasingly severe energy crisis [2,3].Borehole gas extraction has become the main way to prevent gas disasters and develop coalbed methane [4][5][6].Coal seams are a dual porous medium composed of coal matrix blocks and interblock fractures [7,8].The adsorbed gas and the free gas in the cracks reach equilibrium under a certain pressure [9].The extraction drilling hole can form a gas pressure gradient, promoting the free gas in the fractures to flow towards the drilling hole [10][11][12].With the extension of the extraction time, the gas pressure in the coal seam fractures gradually decreases [13].The dynamic equilibrium between the adsorbed gas and the free gas is broken, and the adsorbed gas is desorbed and diffused from the matrix to the fractures [14,15].However, with the progress of gas extraction, the gas pressure gradient between the coal seam and the borehole also gradually decreases [16][17][18].
Conventional enhanced extraction technology can be divided into three categories: mechanical methods, physical methods, and chemical methods [19][20][21][22].The physical properties of coal and external technology in this process are transformed and utilized, which constitutes the main content of the enhanced extraction technology [23][24][25].However, low permeability, low reservoir pressure, and low gas content are the characteristics of coal reservoirs in China.Conventional enhanced extraction technologies are not effective in this type of reservoir.Due to its particularity, gas injection technology can effectively improve the gas extraction efficiency by increasing the gas driving force and reducing the effective stress and the gas partial pressure [26].Thus, the gas injection technology can rapidly reduce the gas content in the coal seam and ensure mining safety.
There have been several theoretical and experimental studies on gas injection technology.Longinos et al. examined the efficacy of liquid nitrogen in the coal fracturing process in coalbed methane reservoirs [27][28][29].Wu [31].Ozdemir established a mixedgas-coupled seepage model of porous media, taking into account the effect of moisture in coal [32].Huang et al. established a water-gas two-phase coupled seepage model which considered the influence of water on gas seepage [33].Seto et al. established a gas-water two-phase flow model which considered the interaction of gases (CO 2 , N 2 , CH 4 , and H 2 O) in the coal seam [34].Xia et al. used a coupled composition model to study the effect of borehole sealing on gas emissions [35].There are also several field trials of enhanced methane recovery involving gas injection.The United States conducted CO 2 -ECBM test and N 2 -ECBM test at the Allison Unit and Tiffany Unit, respectively [36].Canada had conducted a gas injection test for CO 2 storage in Alberta Province [37].The European Union's RECOPOL project was first implemented in Poland, with a gas injection depth of 1050 m [38].Japan had conducted a field test of gas injection to replace coalbed methane in Hokkaido [39,40].China and Canada jointly carried out a study that focused on the recovery enhancement of coalbed methane in the Qinshui Basin, Shanxi Province [41,42].
In this paper, we investigated the gas migration law and time-varying characteristics of CH 4 displacement by N 2 in coal under different injection pressures by a physical simulation experiment.Based on the above results, this research can provide a reference for the theoretical research of N 2 -ECBM-related technology in low permeability coal seams and the selection of injection pressure in the field technology implementation.

Sample Collection and Preparation.
The raw coal samples were collected from the Tashan mine in Shanxi, China.The raw coal samples were made into cylindrical coal samples with specifications of 50 mm in length and 25 mm in diameter.The raw coal and cylindrical coal samples are shown in Figure 1.
The instrument mainly includes the following systems: (1) core clamping system: the core is wrapped by polytetrafluoroethylene casing; (2) gas injection system: gas can be injected into the core holder under constant pressure; (3) pressure system: the back pressure is controlled by the back pressure valve, and the annular pressure is the pressure loaded around the core by the injection pump; (4) data metering system: the inlet and outlet ends of the core holder are, respectively, equipped with a pressure sensor for realtime monitoring of the pressure values at both ends, and the gas component detector and drainage method are used to measure the flow rate of the mixed gas; (5) temperature control system: temperature control adopts thermostat air bath heating, temperature range: 20 °C~120 °C.The schematic diagram of the device is shown in Figure 3.

Experimental Scheme.
The influence of different injection pressures on the process of CH 4 displacement by N 2 was analyzed.The standard coal sample size is 25 × 50 mm.The experiment adopts a true triaxial loading: the confining pressure is 4.0 MPa, and the axial pressure is 2 MPa.The adsorption pressure of CH 4 is 1 MPa, and the adsorption time is 24 h, which ensures that the pressure remains stable and the flow remains unchanged within 30 min when the adsorption is saturated.The N 2 injection pressure is within the range of 0.5~2.5 MPa, and the ambient temperature is set at 25 °C.Specific experimental parameters are shown in Table 1.

Experimental Procedures.
The following are the experimental operation steps: (1) replace the test sample with a solid cushion block, connect the instrument, and check the air tightness of the test device; (2) put the coal sample (25 × 50 mm) into the holder; (3) add an axial pressure of 1 MPa first, and then the ring pressure of 2 MPa, and pressurize alternately step by step until the ring pressure is 4 MPa, and the axial pressure is 2 MPa; (4) vacuum the whole system for 2 hours; (5) inject 1 MPa methane into the holder through the pressurization system, close the downstream valve, and maintain it for 12 hours to ensure complete adsorption; (6) inject 0.5, 1, 1.5, 2, and 2.5 MPa nitrogen, respectively, open the downstream valve, and connect the gas detector to analyze the outlet gas concentration; (7) when measuring the flow, close the branch valve of the gas detector, open the flow valve, and measure the flow.Switch back to the gas detector branch in time after measurement; (8) collect the monitoring data, record the flow, and the gas detector shows that the concentration of each gas is not changing; (9) after the test, relieve the pressure step by step and disassemble the test device.3 Geofluids N 2 injection, the N 2 will remain in the coal samples and cannot be detected at the outlet.It can be seen from Figure 7 that the breakthrough time was 50 minutes, 45 minutes, 35 minutes, 25 minutes, and 20 minutes, respectively, at the level of 0.5 MPa, 1 MPa, 1.5 MPa, 2 MPa, and 2.5 MPa.In this process, the breakthrough time decreased by 10%, 30%, 50%, and 60%, respectively, with the increase in injection pressure.It can be concluded that the breakthrough  Geofluids time was gradually shortened and the change rate decreased with the increase of injection pressure.At this stage, the N 2 would diffuse into the micropores and occupy the adsorption vacancy in the coal surface.At this time, the seepage process was not dominant.Then, the seepage velocity of N 2 in the fractures increased with the increase of N 2 injec-tion pressure.Thus, the seepage velocity was faster than the diffusion velocity result in that the seepage process was dominant.Therefore, the N 2 would pass through the coal samples rapidly.Displacement time refers to the time that the concentration of CH 4 decreased from 100% to a particular value in the   5 Geofluids injection process.It can be seen from Figure 8 that the displacement time was 5 minutes at the level of 2.5 MPa, while the displacement time was 35 minutes at the level of 0.5 MPa.This indicated that the higher the injection pressure, the faster the variation of the CH 4 concentration.The results showed that the pressure gradient between the inlet and the outlet of the coal sample increased with the increase of injection pressure.Thus, the gas was easier to migrate.On the other hand, the high pore pressure offsets part of the effective stress, which reduced the pressure on the coal skeleton.Then, the porosity increased.Thus, the permeability of the coal sample was increased so that the gas migration velocity was accelerated.

Displacement Efficiency and Replacement
Ratio.According to the total mixed flow q of CH 4 and N 2 measured under different injection pressures and different gas volume fractions (φ CO2 , φ CH4 ), the output of CH 4 and N 2 can be calculated.Finally, the displacement efficiency η and replacement ratio μ under different pressure conditions are obtained.The calculation formula is as follows: where η is the efficiency of N 2 displacing CH 4 (%); Q CH4 is CH 4 output (mL); Q T CH4 is the total injection amount of CH 4 (mL); φ CH4 is the CH 4 concentration (%); q is the total flow of mixed gas (mL); μ is the replacement ratio (dimensionless); Q T N2 is the total injected amount of N 2 (mL); t is time (min).
According to the experimental results, the η was 52.3%, 55.9%, 62.9%, 79.9%, and 85.2%, respectively, under different injection pressures.With increase in pressure, the η 6 Geofluids increased by 6.8%, 20.2%, 52.7%, and 62.9%, respectively.The μ in different pressures was 10.9, 10.2, 9.1, 7.2, and 6.7, respectively.With increase in pressure, the μ decreased by 6.4%, 16.5%, 33.9%, and 38.5%, respectively.It can be concluded that η increased with the increase of injection pressure.And the greater the pressure, the more significant the η increased.The maximum increase was more than 60%.The μ decreased with increase in pressure.The maximum decrease of μ was more than 38%.

Flow Rate of CH 4 and Accumulated Production of CH 4 .
It can be seen from Figure 9 that the maximum flow rate of CH 4 was 0.085 mL/min, 0.110 mL/min, 0.130 mL/min, 0.222 mL/min, and 0.273 mL/min, respectively, at the level of 0.5 MPa, 1 MPa, 1.5 MPa, 2 MPa, and 2.5 MPa.The regression curve appeared wave top under different injection pressures.The greater the injection pressure, the sharper the waveform.While the less the injection pres-sure, the wider the waveform.According to the experimental results, the flow rate of CH 4 decreased after an initial increase under different injection pressures.The peak value of the flow rate of CH 4 increased with the injection pressure.But the flow rate of CH 4 decreased rapidly at the pressure above 1.5 MPa.This indicated that the N 2 flowed easily in pores and fractures with high injection pressure so that the breakthrough time decreased.But the N 2 do not remain in the coal matrix at high injection pressure.Therefore, the coal matrix cannot effectively adsorb N 2 at a pressure above 1.5 MPa, which results in a large amount of CH 4 remaining in the coal.It can be seen from Figure 10 that the accumulated production of CH 4 was 3.59 mL, 3.91 mL, 4.39 mL, 5.58 mL, and 5.94 mL, respectively, at the level of 0.5 MPa, 1 MPa, 1.5 MPa, 2 MPa, and 2.5 MPa.The results showed that the accumulated production of CH 4 increased with the increase in injection pressure.However, the increase of accumulated  7 Geofluids production of CH 4 decreased at the injection pressure above 2 MPa.This indicated that the amount of CH 4 displaced by N 2 increased with the increase of injection pressure.The increase of injection pressure will increase the permeability of the coal so that it is easier for the gas to migrate in the seepage channel.It can be concluded that at the beginning, N 2 diffused from fractures to micropores, and competitive adsorption occurred.A large amount of CH 4 had been replaced due to a decrease of the partial pressure of CH 4 .After reaching the critical time, CH 4 desorption and N 2 adsorption reached dynamic equilibrium.Therefore, the production of CH 4 will no longer increase.According to the displacement efficiency, the injection pressure can be divided into three stages.At the range of 0.5~1 MPa, it was the low-efficiency stage, and the displacement efficiency was about 50%; at the range of 1~2 MPa, it was the medium efficiency stage, and the displacement efficiency was about 60%; at the range of 2~2.5 MPa, it was the high-efficiency stage, and the displacement efficiency was 70~85%.
3.6.N 2 Injection Pressure.In the experiment, the displacement efficiency and replacement ratio were obtained according to different evaluation indicators.The two parameters cannot be directly compared and analyzed.In order to eliminate the impact of the difference in the dimension and value range among the indicators, it was necessary to carry out a standardization process and scale the parameters according to the proportion.The minimum-maximum normalization of the parameters, also known as discrete normalization, maps the parameter's value to (0, 1) according to the linear transformation of the original data according to formula (2).The data processing results are shown in Table 2.
It can be seen from Figure 11 that there is a crossover point between the displacement efficiency and replacement ratio with different injection pressures after the discrete standardization treatment of the two parameters.The injection pressure corresponding to the crossover point was 1.6 MPa, and the displacement efficiency was 45%.Theoretically, the injection pressure corresponding to the crossover point was the most economical.The displacement efficiency increased with the increase of the injection pressures, while the replacement ratio decreased with the increase of the injection pressures.Figure 11 shows that the displacement efficiency was low while the replacement ratio is relatively high at the range of 0.5~1.6MPa.This indicated that the amount of N 2 required for the displacement of CH 4 per unit volume was large.This pressure range was not ideal from a technical and economic perspective.At the range of 1.6~2 MPa, the displacement efficiency was high, but the replacement ratio is gradually reduced.This indicated that the amount of N 2 required for the displacement of unit volume of CH 4 was low.At injection pressure above 2 MPa, the increase rate of displacement efficiency was low, and the decrease rate of displacement ratio was reduced.This   8 Geofluids indicated that the flow velocity of N 2 and free CH 4 increased with the increase of injection pressure.At the same time, the displacement efficiency increased with the increase of injection pressure.Therefore, more N 2 enters into the tiny pores of the coal matrix so that more adsorbed CH 4 was displaced in the pores.The larger the amount of CH 4 displaced, the smaller the replacement ratio.When the injection pressure is greater than 2 MPa, the change rate of displacement efficiency and replacement ratio decreases with the increase of injection pressure, which shows that the influence of injection pressure on displacement efficiency and replacement ratio is weakened.

Discussion
The displacement of CH 4 by N 2 is a dynamic process.Most of the researchers [17,23,32] divide the process into three stages.This is also verified by our experimental results.According to our experimental results, the variation rate of the concentration of CH 4 showed a trend of increase before decrease.It was believed that the free CH 4 in the coal samples would be driven away to the outlet first after N 2 was injected into the coal samples.When the concentration of CH 4 in the fractures and macropores decreased, the desorbed CH 4 in the coal surface increased.Then, a large number of empty adsorption sites would be left [8,17].The N 2 would diffuse into the coal surface with the increase of injection [13].Thus, the N 2 molecules would collide with the empty adsorption sites which had not yet adsorbed CH 4 molecules [27].Then, the N 2 molecules would occupy the empty adsorption sites.In addition, the partial pressure of CH 4 decreased with the injection of N 2 .The partial pressure of CH 4 would promote the desorption of CH 4 , which provides more adsorption sites for N 2 [43].However, the adsorption of N 2 molecules on the coal surface was weak.The coal surface cannot absorb N 2 molecules anymore with continuous injection.Then, the injected N 2 would migrate directly to the outlet of the coal samples.At the inception of stage II, the concentration of CH 4 produced increased rapidly because there was a large amount of free CH 4 in the fractures.Then, both of the free CH 4 and adsorbable CH 4 in the coal samples were decreased.Thus, the rate of decrease of CH 4 concentration was increased first and then decreased.
Gas extraction undergoes a desorption-diffusion-seepage process [44].The transport of N 2 and CH 4 after N 2 injection into coal seams is an interactive process.After the injection of N 2 , the injected gas occupies a certain space and bears part of the pore pressure.The number of adsorption sites in coal is fixed.The number of adsorption sites occupied by the injected gas increases which results to the decrease of CH 4 -occupied adsorption sites [43].This means that the increased partial pressure after N 2 injection can promote the desorption of CH 4 .In addition, after gas injection, the reservoir pressure increases while the effective stress decreases.Thus, the permeability of the coal increases, which is conducive to the transport of CH 4 [42].These two points indicate that the higher the injection pressure, the more methane is produced.However, the results of Figure 10 show that when the injection pressure is increased to a certain level, the methane output increases insignificantly.We believe that the bidirectional diffusion processes of gases influence the output of methane.The flow rate of the injected gas increases as the injection pressure increases.When the flow rate of injection gas is too fast, the injected gas cannot diffuse sufficiently into the coal matrix, resulting in reduced methane production.Thus, there is a specific pressure range which is beneficial to the production of methane.

4
Displaced by N 2 .The displacement of CH 4 by N 2 is a dynamic process.The released gas volume changed constantly under different injection 2 Geofluids pressures and injection times.The results show that the displacement of CH 4 by N 2 results from the joint action of multiple mechanisms.It can be seen from Figure 4 that the whole process can be divided into three stages (taking the injection pressure of 1.0 MPa as an example).Stage I (original equilibrium stage): the volume fraction of N 2 was 0, and the volume fraction of CH 4 was 100%.At first, N 2 was injected into the coal samples saturated with CH 4 .The migration speed of N 2 through "seepage-diffusionadsorption" under the pressure gradient is relatively slow.It takes a specific time to break the original equilibrium state.Stage II (dynamic balance stage): N 2 had seeped from fractures with continuous gas injection.Then, N 2 diffused into the pores of the coal matrix, and the concentration of N 2 began to increase while the concentration of CH 4 decreased.There was a dynamic change relationship of "this and the other."This is because the partial pressure of CH 4 decreased, and CH 4 is desorbed from the coal matrix after N 2 injection.In addition, the concentration difference between macropores and micropores will be increased with the migration of free CH 4 in fractures.Thus, the CH 4 desorption was accelerated.In this stage, CH 4 desorbed from the coal matrix and diffused to the fractures under the concentration gradient.Stage III (new equilibrium stage): the concentration of N 2 increased to 100%, while the concentration of CH 4 decreased to 0 with continuous N 2 injection.The new equilibrium state was established, and the CH 4 was no longer desorbed.

Figure 4 :
Figure 4: Variation of concentration with gas injection time.

Figure 5 :
Figure 5: Variation of CH 4 concentration under different pressures.

Figure 9 :
Figure 9: The flow rate of CH 4 under different injection pressures.
et al. developed a dual pores model for enhanced methane recovery by CO 2 injection [30].Kumar et al. developed a coupled finite element (FE) model to study the heterogeneously permeable coal reservoirs

Table 1 :
Experimental parameters of CH 4 displacement by N 2 .

Table 2 :
Data standardization results under different pressures.