Adsorption Characteristics of CH4 and CO2 in Shale at High Pressure and Temperature.

This work presents the adsorption behavior and appearance characteristics of CH4 and CO2 on the Longmaxi shale at high pressure and temperature. To investigate the variation of gas adsorption patterns under the constraint of pressure and temperature, the applicability of the theories of monolayer adsorption, multilayer adsorption, and micropore filling was discussed. The preferential selection coefficient of CO2 for CH4 under different conditions was characterized by the absolute adsorption capacity (V abs) ratio of CO2 to CH4 (αCO2/CH4). Moreover, the implication of the CO2 injection to enhance gas recovery and the CO2 capture and storage (EGR-CCS) process was analyzed. The results exhibit that the excess adsorption curves of CH4 are smooth, and the experimental temperature has no noticeable effect on the shape of curves. At the same time, a "sharp peak" is recorded in the excess adsorption curves of CO2 at low temperatures (30 and 55 °C) near the critical pressure, which is quite distinct from the smooth curves at high temperatures (80 and 100 °C). Correspondingly, there are two "jump pressure" values in the density curves (30 and 55 °C) of the adsorption system and the density curves are divided into three stages. The Dubinin-Astakhov and Brunauer-Emmett-Teller (BET) models show an optimum degree of fit for CH4 and CO2 adsorption curves under all experimental temperature and pressure conditions. The Langmuir model fits the adsorption curves of 80 and 100 °C better, while the BET model is appropriate for 30 and 55 °C. The adsorption affinity of CO2 is higher than CH4, with the value of αCO2/CH4 in the range of 2.47-12.16. The value of αCO2/CH4 increases with a rise in pressure but is inhibited by high temperatures, while the inhibition is negligible when the experimental temperature exceeds 80 °C. The adsorption preferential of CO2 is stronger in the shallow reservoir (αCO2/CH4 > 10.5), and the application prospect of the EGR process is promising. In contrast, the adsorption preferential is slightly weakened in the deep reservoir (αCO2/CH4 < 4.5), which can be considered for CO2 capture, utilization, and storage. Results from this investigation provide novel insights on the adsorption characteristics of CH4 and CO2 on the shale matrix at high pressure and temperature. They are also expected to give certain enlightenment for the EGR-CCS process.

Introduction

Shale gas is an unconventional clean energy source, with a wide distribution, having multiple strata and high reserves, and showing enormous development potential, and is receiving increasing attention worldwide.[1,2] Shale with “low porosity and low permeability” limits its natural productivity. The stimulation effect of a traditional hydraulic fracturing technology is remarkable.[3,4] However, a series of deficiencies need to be solved by alternative methods. For instance, the fracturing fluid may trigger a drop in the groundwater levels. The flowback gives rise to pollution, uncontrollable local earthquakes caused by the fracturing process, the mechanically complexity of the fracturing process, and the sensitivity of shale reservoirs to the fracturing fluid.[5−7] Besides, the hydraulic fracturing technology is getting worse with the rise in burial depth, and the development cost is enormous. It is a significant factor limiting the commercial development of deep shale gas reservoirs. The research of a new clean stimulation technology is crucial for the development of deep shale gas reservoirs. On the other hand, CO2 emission is the dominant proportion of global carbon, which caused a severe greenhouse effect. Emission reduction and even carbon neutralization have become the main target of atmospheric environmental protection in the world. CO2 capture and storage (CCS) is a promising way to achieve CO2 emission reduction, and the widely distributed shale reservoir is one of the potential geological bodies.[8,9]

CO2 injection to reservoir enhanced gas recovery, and realized CO2 CCS is a potential high-quality development project.[2,10] Its essence is to inject CO2 into the shale gas reservoir, and CO2 competes with CH4 for adsorption sites on the shale matrix, which boosts CH4 desorption and enhances the recovery. Meanwhile, the excellent sealing properties of shale allow for the storage of CO2, thereby being mutually beneficial in improving shale gas recovery and alleviating global warming. The adsorption capacity and priority of CO2 in shale are higher than that of CH4, which is the basis of the exhaust gas recirculation (EGR)–CCS process.[2,10] The proportion of adsorbed gas in shale is in the range of 20–85%,[11] so that its successful desorption is crucial to the stable production of shale gas reservoirs. The adsorption characteristics, controlling factors, and controlling mechanisms of CH4 and CO2 in shale have attracted extensive attention. There is an apparent impact of in situ temperature and pressure, mineral composition, organic matter content, and pore structure of the reservoir on the gas adsorption process.[12,13] Wang et al.[14] simulated the adsorption behavior of CH4 (0–10 MPa at room temperature) in the Longmaxi and Niutitang shales using the Langmuir and Brunauer–Emmett–Teller (BET) models. Zhou et al.[15] conducted the adsorption experiments of CH4 and CO2 (0–12 MPa, 35–55 °C), utilizing the Ono–Kondo and Dubinin–Astakhov (D–A) models. Abdul Kareem et al.[16] performed the adsorption of CO2 (0–15 MPa, room temperature). Bemani et al.[17] and Eshkalak et al.[8] carried out the adsorption of CO2 and CH4 on shale and discussed the superiority of CO2 displacement of CH4. Gu et al.[18] investigated the relationship of αCO2/CH4 versus mineral composition and pore structure in marine shale. Overall, the adsorption of CO2 and CH4 in shale is controlled mainly by temperature, pressure, mineral composition, and total organic carbon. The adsorption capacity of CO2 and CH4 increases gradually with a rise in pressure until the adsorption equilibrium is reached.[8] Increasing temperature inhibited the adsorption, and the adsorption capacity decreased with a rise in temperature.[15] The value of αCO2/CH4 also increased with an increase in pressure and reduced with a rise in temperature.[17,18]

However, the maximum experimental pressure of CO2 and CH4 adsorption mainly was less than 10 MPa in a previous work and rarely reached 15 MPa. Experimental pressures over 20 MPa were exceedingly rare. The exploitation of shale gas has been developed for deep reservoirs, where the reservoir pressure is greater than 20 MPa. Research on the adsorption behavior of CH4 and CO2 under high pressure is therefore lacking. Additionally, the experimental temperature of CO2 and CH4 was predominantly performed at room temperature. A few scholars discussed the influence of temperature on the adsorption behavior within a specific range, generally below 55 °C, but the maximum temperature of deep shale gas reservoirs is commonly over 100 °C. The research on CO2 and CH4 adsorption under high temperature is surprisingly limited, and the effect of temperature and pressure on αCO2/CH4 also requires further investigation. Furthermore, the investigation of a proper model fitting of CO2 and CH4 adsorption under high pressure and temperature is lacking. Besides, it needs to be verified whether the appropriate models under low pressure and temperature also have a high degree of fit in high pressure and temperature. It is, therefore, necessary to perform the model fitting of the isothermal adsorption models for CO2 and CH4 under high pressure and temperature.

This investigation focused on the Longmaxi shale in south Sichuan, China, and simulated the adsorption behaviors of CH4 and CO2 under high pressure and temperature (30–100 °C, 0–25 MPa) through isothermal adsorption experiments. To explore the gas adsorption characteristics of deep shale gas reservoirs under in situ conditions and the variation of CO2 adsorption superiority, the Vabs was corrected by the Vex. The Langmuir, BET, and D–A isothermal adsorption models were employed to fit the adsorption data of CH4 and CO2. The applicability of the models was evaluated using R2 values as the standard of the degree of fit. The adsorption mechanism and primary occurrence pattern of CH4 and CO2 in shale were also discussed. Additionally, αCO2/CH4 were calculated to analyze the preferential selection coefficient of CO2 for CH4 under different pressures and temperatures. The research results provide references for adsorption behaviors of CH4 and CO2 on shale and have theoretical significance for the promotion and development of the EGR–CCS process in shale gas reservoirs.

Sample, Experiments, and Methods

Sample Information

In this investigation, drilling cores of the Longmaxi Formation shale gas reservoir in the Changning area of the southern Sichuan Basin (Figure ), one of the shale gas exploration and development demonstration areas in China, are selected as the research objects. Four black organic-rich shale samples were collected from the lower part of the Longmaxi Formation in the Lower Silurian in Well X, which are YS-13, YS-14, YS-15, and YS-16 from bottom to top, and the corresponding burial depths are 1243.77–1244.04, 1241.74–1242.01, 1239.72–1239.99, and 1237.96–1238.23 m, respectively.

Figure 1

Schematic diagram of the sampling point tectonic location.

Isothermal Adsorption Experiments of CH4 and CO2

In this investigation, the ISOSORP-SC high-pressure isotherm adsorption instrument was employed to test the adsorption behavior of CH4 and CO2 by using the gravimetric method. The adsorption system consists of a control cabinet system and magnetic suspension balance system (Figure ). The former can be used to transport adsorbate, vacuumize, and control experimental pressure. The latter completes the sample adsorption, temperature control, adsorption system parameters, and result recording. The experiments were conducted according to the standard of NB/T 10117-2018. Before the experiments, the shale samples were ground to 60–80 mesh. During the screening process, the particles less than 80 mesh should be fully screened out and each sample should weigh 10 g. The prepared samples were heated and dried in a vacuum at 110 °C for 12 h to remove the adsorbed water vapor. According to the adsorption requirements of CH4 and CO2, the gas path was connected, and the working state of the balance suspension system was checked after startup.

Figure 2

ISOSORP-SC high-pressure isotherm adsorption instrument.

After the airtightness test, blank test, pretreatment experiment, and buoyancy test, the isothermal adsorption tests were exerted. Specific experimental details are as follows:

First, the CH4 adsorption isotherm experiments were conducted. The experimental pressure was in the range of 0–25 MPa, 13 pressure points of 0, 2, 4, 22, and 24 MPa were designed totally and exerted at 30, 55, 80, and 100 °C (YS-16 at 30 °C, YS-15 at 55 °C, YS-14 at 80 °C, and YS-13 at 100 °C), respectively. The experimental temperature was controlled by a high-precision oil bath thermostat. After the CH4 isothermal adsorption experiments at all four temperatures were complete, the samples were vacuumed at 110 °C for 12 h to remove the residual CH4. After that, the abovementioned procedures were repeated to conduct the isothermal adsorption experiments for CO2.

Correction of Adsorption Capacity

The test result of the isothermal adsorption experiment is excess adsorption capacity (Vex), which is the excess capacity of the adsorption phase density that exceeds the bulk phase density. This is also known as the Gibbs adsorption capacity. During the calculation process, the volume of the adsorption phase is ignored, which leads to a calculated value that is lower than the actual adsorption capacity. It is, therefore, significant to correct the actual adsorption capacity, namely, the Vabs.[15,19,20] The high temperature and pressure adsorption model is appropriate for fixing the gas adsorption capacity in shale (eq ).where Vabs is absolute adsorption capacity, cm3/g; Vex is excess adsorption capacity, cm3/g; ρa is adsorption phase density, g/mL; and ρg is bulk phase density, g/mL.

Previous studies attempted to assume the density of adsorbed gas as the density of the boiling point in liquid at atmospheric pressure or the Van der Waals density of the gas.[1,14,21] For this study, CH4 and CO2 will be transformed into the supercritical state (CO2: 31.04 °C, 7.38 MPa;[22] CH4: −82.59 °C, 4.59 MPa[23]). It is challenging to meet the requirements of the isothermal adsorption experiments at different temperatures by simply setting ρa as a constant. Therefore, scholars proposed that the Vex will linearly decrease during the high-pressure stage. The bulk density can be taken as the abscissa, and the Vex in the descending section can be taken as the ordinate. The intercept of the fitting curve on the X-axis is ρa.[24,25]

Fitting Models of the Adsorption Results

Monolayer adsorption, multilayer adsorption, and micropore filling theories are widely accepted in the research of gas adsorption behavior on solid adsorbent surfaces and the internal and external surfaces of porous media. To explore the adsorption pattern of CH4 and CO2 under different temperatures and pressures, the most representative isothermal adsorption model of the three theories was selected, which are the Langmuir model, BET model, and D–A model. The details of calculation principles and methods are as follows:

Langmuir Model

The Langmuir model[26] is a widely accepted isotherm adsorption equation (eq ). The Langmuir model assumes that the adsorption energy on the adsorbent surface is uniform and constant. Only a monolayer of adsorption is formed on the adsorbent surface, and the adsorption capacity reaches the maximum when the monolayer adsorption is saturated. Therefore, adsorption is a monolayer. Also, if the adsorption mechanism is equivalent, no intermolecular force exists among the adsorbates, which means that only a gas molecule will be adsorbed at each adsorption site.where P is the experimental pressure at adsorption equilibrium, MPa; VL is the Langmuir volume, cm3/g, which represents the maximum adsorption capacity of a monolayer and is determined by the property of shale and the type of adsorbate; and PL is the Langmuir pressure, MPa, the value is the pressure when the adsorption capacity equals to VL/2.

BET Model

The BET model was established by Brunauer et al.,[27] who expanded and extended the Langmuir model (eq ) and concluded that physical adsorption is caused by the Van der Waals forces. This model states that the intermolecular force existed among the adsorbates due to the Van der Waals force of the gas molecules, thereby formed multilayer adsorption. The BET model assumes that the adsorption heat of the first layer is constant. The adsorption heat of each layer is equal to the condensation heat after the second layer, and the adsorption layers are infinite.where Vm is the saturated adsorption capacity of a monolayer, cm3/g; p0 is the saturated vapor pressure at the experimental temperature, MPa; and c is a constant, dimensionless, related to the adsorption heat.

D–R and D–A Models

The D–R model[28] is based on the micropore filling theory and the Polanyi[29] adsorption potential theory (eq ), which is suitable for the adsorption behavior of gas molecules on the surface of porous adsorbents. It is suggested that the adsorption mechanism of gas on micropore adsorbents is quite distinct from that on open surfaces due to the small distances between the pore walls. The gas adsorption in the micropore fills the pores, different from the molecular layer adsorption described by the Langmuir and BET models. The D–A model (eq ) amends the D–R model and has one more fitting parameter, which significantly enhances the flexibility.where a is the structural heterogeneity constant of shale, with a value in 1–4 ranges.

Results and Discussion

Vabs and Vex

Adsorption Results of CH4

The morphology of the Vex curves of CH4 in four samples is analogous and exhibits a convex shape (Figure ). Vex increases fast from 0 to 4 MPa, and the increments gradually reduce with a rise in pressure (4–10 MPa). From 10 to 16 MPa, the Vex drops, and the slope gradually increases. For Vabs, the adsorption amount increases continuously with a rise in pressure until the adsorption equilibrium is reached. The discrepancy of Vabs and Vex grows linearly with an increase in pressure. This is because the volume of the adsorbed phase is not accounted for and the larger the bulk density is, the discrepancy will be more significant with a rise in pressure.

Figure 3

Vex, Vabs, and Vabs–Vex of CH4 adsorption results on shale.

Adsorption Results of CO2

The Vex curves of CO2 at different temperatures exhibit significant discrepancies (Figure ). The experimental temperatures of YS-13 (100 °C) and YS-14 (80 °C) are higher, and the Vex curves are convex and smooth. Vex increases fast from 0 to 2 MPa, and adsorption incrementally reduces and increases linearly from 2 to 14 MPa. After that, Vex decreases linearly when P > 16 MPa. In contrast, the experimental temperatures of YS-15 (55 °C) and YS-16 (30 °C) are relatively lower. Although the Vex curves are also convex, a “sharp peak” (Figure ) existed, distinct from the high-temperature curves. In previous studies, the characteristics of CO2 excess adsorption curves can be divided into two categories: (i) type I adsorption isotherm in BDDT classification[30] and (ii) typical CH4 excess adsorption curves similar to Figure .[15] The adsorption experiments were repeated three times and similar experimental phenomena were obtained. This indicates that the “sharp peak” of CO2 excess adsorption curves at 30 and 55 °C is anomalous and should be deeply explored.

Figure 4

Vex, Vabs, and Vabs–Vex of CO2 adsorption results on shale.

The “sharp peak” of excess adsorption curves on YS-15 (55 °C) and YS-16 (30 °C) approximately starts at 6 MPa and ends at 14 MPa and reaches the maximum at 10 MPa (Figure ). Tsuzuki et al.,[31] Yang et al.,[32] and Nikolai et al.[33] reported that when CO2 is close to the critical state, the sensitivity of compressibility to pressure change is significantly improved, and the aggregation behavior is remarkable; and when far away from the critical pressure, the state of CO2 tends to stable again. The tiny change of pressure will have a significant impact on the density and viscosity of CO2 nearing the critical pressure. The characteristics of density variation in the adsorption system during the adsorption of CH4 and CO2 were analyzed (Figure ). The curves of CH4 density versus experimental pressure are characterized by a “single stage”, which is determined by the thermodynamic and transport properties of CH4. It is a gradual system in the adsorption process, corresponding to the smooth and continuous characteristics of CH4 excess adsorption curves in Figure . In contrast, the density curves of CO2 of YS-13 (100 °C) and YS-14 (80 °C) are characterized by “double stages”, and the “jump pressure” is about 7 MPa, nearing the critical pressure. For YS-15 and YS-16, the experimental temperature is relatively close to the critical temperature of CO2, and the density curves are characterized by “ternary stages” separated by two “jump pressure” values (Figure ). Namely, CO2 is in a subcritical state (0–7 MPa), the compressibility is relatively lower, resulting in a slow rise in the correlation curves of density versus experimental pressure in stage ①. Then, the property of CO2 becomes active nearing the critical pressure. The sensitivity of compressibility and density to pressure increases significantly, resulting in a sharp rise in the slope of stage ②. When the pressure is far away from the critical pressure, CO2 transforms into a stable supercritical state. The sensitivity of compressibility and density to the pressure becomes steady, resulting in a slower slope of the density curves of stage ③ (Figure ). In addition, although the experimental temperature of YS-16 is 30 °C, which is slightly lower than the critical temperature of CO2. Yang et al.[32] and Nikolai et al.[33] suggested that the CO2 properties also change significantly when it approaches the critical temperature. Overall, the adsorption capacity and density of the CO2 are obviously decreased at high temperature, especially when CO2 is in a supercritical state, the regularity is more prominent, which is consistent with the current general understanding.

Figure 5

Density curves of CO2 and CH4 in the adsorption system.

Model Fitting of the Adsorption Behaviors

In Figures and 4, the Vabs curves of CH4 and CO2 (especially the latter one) at different experimental temperatures are diverse, which indicates that the temperature has a particular influence on the adsorption characteristics and mechanism of gas on the shale matrix. Therefore, it is necessary to re-explore the applicability of monolayer adsorption (Langmuir model), multilayer adsorption (BET model), and micropore filling (D–A model) theories to the adsorption behavior of CH4 and CO2 at different temperatures, respectively. The specifications are as follows:

Fitting Results of the Langmuir Model

The Langmuir model has an excellent degree of fit of CH4 adsorption behavior at different experimental temperatures (R2 = 0.9948, 0.9953, 0.9918, and 0.98003, respectively). The degree of fit of CO2 adsorption behaviors at different temperatures is significantly distinct (R2 = 0.9848, 0.9738, 0.93222, and 0.93437, respectively) (Figure ). For YS-13 and YS-14, the experimental temperatures are higher (100 and 80 °C), and the degree of fit is relatively favorable. Conversely, the degree of fit of YS-15 and YS-16 is relatively lower (R2 < 0.94), indicating that the monolayer adsorption theory is not the most suitable model to describe CO2 adsorption at 30 and 55 °C. The higher the temperature, the better the degree of fit, and the more closely it approaches the monolayer adsorption theory. The reason behind this phenomenon is that the gas diffusion ability in shale becomes more substantial, the adsorption stability becomes worse, and the total adsorption amount becomes smaller with the increase of experimental temperature.[9,34] In contrast, the adsorption energy between shale matrix and the first layer (monolayer) is higher than that of other adsorption layers (multilayer), results in relatively better adsorption stability at high temperature.

Figure 6

Fitting results of the Langmuir model of CH4 and CO2 adsorption behaviors on shale.

Fitting Results of the BET Model

The degree of fit of the BET model for CH4 is all favorable (R2 = 0.9898, 0.9903, 0.9976, and 0.9951, respectively) (Figure ). The degree of fit of YS-15 and YS-16 by using the BET model is higher than that of YS-13 and YS-14, which is contrary to the relevant results of the Langmuir model. Under high temperatures, the velocity and diffusion coefficient of gases are large, and the adsorption behavior of CH4 on shale is closer to that of monolayer adsorption. Therefore, the Langmuir model is more appropriate. Still, at 55 and 30 °C, the adsorption behavior of CH4 on shale is closer to multilayer adsorption of the BET model, with relatively lower molecular velocities and diffusion coefficients. The degree of fit of CO2 adsorption by using the BET model is also excellent (R2 = 0.9943, 0.9870, 0.9840, and 0.9967, respectively), and the variation of the degree of fit is similar to that of CH4.

Figure 7

Fitting results of the BET model of CH4 and CO2 adsorption behaviors on shale.

Fitting Results of the D–A Model

The fitting variable of the D–A model is P/P0 in eq , while the saturated vapor pressure (P0) changes with a change in the experimental temperature. In this work, the saturated vapor pressure was calculated by using eq . It was found that the experimental pressure is higher than the saturated vapor pressure for high-temperature CO2 adsorption experiments, which does not conform to the original theory of the D–A model. Therefore, referring to the treatment of Wang et al.,[35] the physical meaning of ρg/ρa is, in essence, synonymous to that of P/P0, and by replacing P/P0, a positive fit can be obtained.where Pc is the critical pressure of adsorbed gas, MPa; Tc is the critical temperature of absorbed gas, K; ΔH is the standard boiling point vaporization enthalpy of adsorbed gas, cal/mol; and R is the general constant of gas, J/mol·K.

The D–A model has the optimum degree of fit of the CH4 adsorption curves, where the R2 values are all greater than 0.99 (R2 = 0.9980, 0.9954, 0.9957, and 0.9914, respectively) (Figure ). The degree of fit of the CO2 adsorption curves also promises (R2 = 0.9939, 0.9864, 0.9935, 0.9964, respectively) (Figure ). There is no obvious regularity between the degree of fit vs experimental temperature, which indicates that the D–A model is not only proper for the fitting of CH4 and CO2 adsorption behaviors in shale but also has low sensitivity to temperature and wide application range. The adsorption pattern of CH4 and CO2 is better explained by the micropore filling theory correspondingly.

Figure 8

Fitting results of the D–A model of CH4 and CO2 adsorption behaviors on shale.

Applicability of the Adsorption Theories and Models

The degree of fit of the three models to the curves of CH4 absolute adsorption curves is favorable, with the R2 values of all samples are higher than 0.98 (Figure ). In general, the applicability of D–A > BET > Langmuir. Additionally, the performance of the three models is a discrepancy in the different experimental temperatures. For the Langmuir model, the degree of fit for high-temperature samples is better, which indicates that high-temperature promotes the formation of monolayer adsorption. However, the performance of the BET model is opposite to the Langmuir model, and it is more appropriate for low temperature samples, which indicates that low-temperature promotes the formation of multilayer adsorption. The degree of fit of the D–A model is less affected by temperature and has a high applicability. For CO2 adsorption, the fitting effect of the BET and D–A models (all R2 values are higher than 0.98) is better than that of the Langmuir model in the fitting of CH4 and CO2 adsorption behaviors (Figure ). The performance of the Langmuir model at high temperature (R2 > 0.97) is apparently more potent than that at low temperature (R2 < 0.94). In contrast, the BET and D–A models do not exhibit apparent regularity for temperature.

Figure 9

Degree of fit of the Langmuir, BET, and D–A models.

With the increase of experimental temperature, the gas diffusion ability increases and the stability of adsorbed gas on the pore surface decreases.[9,34,36] Therefore, monolayer adsorption is easier to form on the surface of the shale matrix, and high temperature has a particular hindrance to the multilayer adsorption. Additionally, Gu et al.[18] indicated that the gas adsorption tends to monolayer with the increase of temperature. Hence, the Langmuir model has better adsorption performance at high temperature, and the BET model is outstanding for low temperature. An et al.[37] demonstrated that gas in the micropore of coal reservoir occurs in overlapping filling adsorption. The excellent degree of fit of the D–A model for CH4 and CO2 adsorption in this investigation proves that this conclusion is also applicable in shale reservoirs, and it has promising applicabilities at different temperatures.

Preferential Selection Coefficient of CO2 for CH4

The αCO2/CH4 of the four samples are in the range of 2.64–4.29, 2.47–4.15, 3.72–11.35, and 4.46–12.16, respectively (Figure a). αCO2/CH4 increases with the rise in experimental pressure until it reaches equilibrium. However, the shape of high temperature curves (YS-13 and YS-14) is different from that of low-temperature curves (YS-15 and YS-16). The slope of high-temperature curves is slight, and the equilibrium pressure is about 18 MPa, while the slope of low-temperature curves is large, and the equilibrium pressure is about 14 MPa. The equilibrium values of the former (αCO2/CH4 > 10.5) is much larger than that of the latter (αCO2/CH4 < 4.5), which suggested that with the rise in experimental temperature, αCO2/CH4 decreases (R2 = 0.8609) and equilibrium pressure increases (Figure a,b). Notably, the αCO2/CH4 curves of YS-13 (100 °C) and YS-14 (80 °C) are similar. Most of the equilibrium values coincide. It can be inferred that when the temperature is higher than a certain value (80 °C in this work), the inhibition of temperature on the adsorption superiority of CO2 is negligible.

Figure 10

Variation of αCO2/CH4 at different experimental pressures and temperatures. (a) Curves of αCO2/CH4 versus experimental pressure and (b) correlation of αCO2/CH4 vs experimental temperature.

Implication to the EGR–CCS Process

Implication to the EGR Process

The adsorption capacity of CO2 is evidently stronger than that of CH4. The discrepancy of molecular structure, molecular size, boiling point, self-diffusion coefficient, and adsorption stability leads to the superiority of CO2.[38−40] Therefore, injecting CO2 into shale reservoirs to enhance gas recovery (CO2–EGR) is regarded as one of the most potential clean and efficient development schemes. In the 30 and 55 °C curves of Figure a, when the experimental pressure is in the range of 10–24 MPa, αCO2/CH4 is higher than 10.5. According to the characteristics of the hydrostatic pressure gradient (1 MPa/100 m) and conventional geothermal gradient (2.5–3.0 °C/100 m), the reservoir temperature and pressure conditions of 10 MPa and 30–55 °C can be reached with a burial depth of 1000 m. It is speculated that CO2 injection can significantly enhance the recovery of CH4. Furthermore, exploring a deep shale gas reservoir (with a burial depth > 3500 m) has captured attention worldwide. Immature exploited technology and huge development costs limit the commercial development of deep gas reservoirs. The CO2–EGR process can be regarded as a potential alternative development technology. In Figure a, the αCO2/CH4 values are lower than 4.5 in 80 and 100 °C. Although it is lower than that of low temperature, the CO2–EGR process still has considerable application potential.

Implication to the CCS Process

CCS or CCUS (CO2 capture, utilization, and storage) process is an effective approach to achieve CO2 emission reduction. CO2 used for utilization and storage can be obtained from existing industrial processes to avoid further aggravation of the greenhouse effect. The widely distributed and huge shale gas reservoir is one of the potential geological reservoirs. In Figure a, the αCO2/CH4 values tend to be stable when reached equilibrium pressure, which indicated that once the underground storage of CO2 is completed, the possibility of leakage is low. Additionally, in the shale gas reservoirs with a burial depth of over 1000 m, CO2 behaves like a supercritical fluid and its injection has significant engineering advantages. In a pore system of shale dominated by slit pores, plate-shaped pores, and ink bottle pores,[41,42] the throat of an ink bottle pore is narrow, and the internal CH4 molecule is difficult to move out during the process of conventional depressurization and drainage. Wang et al.[43] revealed that supercritical CO2 would react with organic matter and minerals in the pore wall, which widens the pore throat and increases CH4 drainage. Additionally, hydraulic fracturing in the shale can be seriously affected by “water sensitivity” and “collapsibility” due to the high clay content of shale, which may cause the internal blocking of pores, damage pore connectivity and influence the effectiveness of hydraulic fracturing. In contrast, fracturing fluid based on supercritical CO2 has the characteristics of low viscosity, high diffusion, high density, low rock-breaking threshold, and fast speed.[44,45] Zhang et al.[46] also implied that the injection pressure of the CO2 fracturing process is lower and the fracture dispersion is wide. Therefore, it is easy for the CO2 to communicate with the original pore system and form a gas migration network channel.

Although the EGR–CCS process has great application potential, it is not mature yet and several critical bottlenecks exist in the application process. (i) the cost of CO2 acquisition, transportation, and injection is vast, which is regarded as an economic burden at present; (ii) the phase transformation of CO2 is fast and complex, and the change of its properties and state is not completely clear under high pressure and temperature. It is also essential to clarify the reservoir strain caused by the CO2 injection process, as well as the gas migration—displacement—desorption process, to ensure the displacement efficiency and stability of CO2 underground storage; and (iii) the sand carrying capacity of CO2 based fracturing fluid is lower than that of water-based fracturing fluids, and its viscosity is low, which may result in sand plugging the fractures. The above problems are the crucial directions in further work.

Conclusions

The adsorption capacity of CH4 is relatively low at high experimental temperature, and there was no apparent correlation between them; while a linear negative correlation was recorded between CO2 adsorption capacity versus experimental temperature. The D–A model has a perfect degree of fit for CH4 and CO2 adsorption behavior and has a high adaptability to temperature. Micropore filling theory is suitable for simulating gas adsorption behavior in deep and shallow gas reservoirs. The BET model also has a promising degree of fit for CH4 and CO2, while it is more applicable at low temperatures. Contrarily, the degree of fit of the Langmuir model is more suitable for high temperatures. The adsorption behavior of gas in shale tends to transform from multilayer to monolayer with increased temperature.

αCO2/CH4 of the four samples is in the range of 2.64–4.29, 2.47–4.15, 3.72–11.35, and 4.46–12.16, respectively. The values are positively correlated versus experimental pressure and negatively correlated versus experimental temperature. For YS-15 and YS-16, the equilibrium αCO2/CH4 is over 10.5, the injection of CO2 can promote the desorption of CH4. By contrast, the equilibrium αCO2/CH4 of YS-13 and YS-14 is below 4.5, which suggests that the injection of CO2 will enhance CH4 recovery and realize the geological CCS of CO2 with a low leakage risk.