Different Effect Mechanisms of Supercritical CO2 on the Shale Microscopic Structure.

To better understand how supercritical carbon dioxide (CO2) enhances shale gas production, it is necessary to study the interaction of supercritical CO2 with shale and its impact on shale microstructure. The different mechanisms by which supercritical CO2 changes the shale pore structure were studied by X-ray diffraction analyses, scanning electron microscopy (SEM), nuclear magnetic resonance spectroscopy, and low-pressure nitrogen gas adsorption tests on shale samples before and after treatment with different pressures and gases (CO2 and Ar). The results showed that after treatment with CO2, the mineral content of shale changed significantly, and in particular, the proportions of calcite and dolomite decreased. The mineral content of shale changed the most after treatment with supercritical CO2, and the microscopic pores were most observable by SEM. In a gaseous CO2 environment, the effect of CO2 adsorption on shale pores is greater than the effects of gas pressure and dissolution reactions. However, in a supercritical CO2 environment, the changes in shale pore structures are mainly controlled by extraction and dissolution reactions. When shale is exposed to supercritical CO2, the fractal dimensions of adsorption pores and seepage pores decrease, indicating that the specific surface area and roughness of adsorption pores decrease. This implies that the adsorption capacity decreases, and that the complexity of the seepage pores declines, which is conducive for gas migration.

Introduction

Shale gas reservoirs contain spontaneously generated natural gas stored within shale rocks of low porosity and low permeability,[1−3] where methane (CH4) exists mainly as free gas and adsorbed gas.[4,5] Horizontal well fracturing is currently the main technique used for shale gas production.[6−9] However, because of the very high water consumption involved in this process,[10−12] which is typically in the range of 15,000–30,000 tons of water per shale gas well,[13] its application is severely limited in regions with water scarcity.[8,14] As an alternative, supercritical CO2 can be used as a new anhydrous fracturing fluid to extract shale gas.[15−21] It has the characteristics of a low-viscosity gas and a high-density liquid, with superior fluidity, permeability, and transmissivity.[21] Further, CO2 is more easily adsorbed than CH4[22−25] by shale rock, which is not only a source of CH4 but also a gas reservoir. Studies have shown that using supercritical CO2 to extract shale gas can not only improve the recovery rate of shale gas but also effectively store CO2 underground, which alleviates the greenhouse effect.[26−33]

When CO2 is injected into a shale gas reservoir, shale is exposed to supercritical CO2 at high temperature and pressure. On the one hand, this causes supercritical CO2 to react with carbonates and clay minerals within the reservoir,[34−37] resulting in changes in mineral contents; on the other hand, a large amount of injected CO2 increases the pore pressure and changes the pore structure of shale.[38,39] Microscopic pores adsorb gas and are channels for gas migration; thus, the study of the effect of supercritical CO2 on shale microstructure is of much significance for the efficient production of shale gas and CO2 storage. At present, several test methods are used to analyze the interaction between supercritical CO2 and the shale microstructure, such as X-ray diffraction (XRD), scanning electron microscopy (SEM), nuclear magnetic resonance (NMR), and low-pressure nitrogen gas adsorption (N2GA).[40−43] Some studies have shown that supercritical CO2 can extract water molecules from minerals to form carbonic acid and dissolve carbonate minerals.[44] Lyu et al.[45,46] found that new pores appeared on the surfaces of shale slices after being saturated with sub-/super-critical CO2, and an increase in carbon content confirmed a chemical reaction between CO2 and shale. Ao et al.[47] found that supercritical CO2 causes strain because of swelling and that high CO2 pressure causes shale deformation; however, deformation also occurs at low pressure, mainly because of CO2 adsorption. Yin et al.[48] observed that after treatment with supercritical CO2, the proportions of clay minerals, calcite, and dolomite in shale decreased, and the primary macropores and microcracks in shale were narrowed because of swelling induced by CO2 adsorption. Lu et al.[38] found that after CO2 saturation, shale porosity and volume of macropores increased, but the volume of micropores and mesopores decreased, and that the effect of gaseous CO2 on the pore structure of shale was weaker than that of supercritical CO2. Jiang et al.[49] noted that original pores and pre-existing fractures in shale could be eroded because of the effect of dissolution by supercritical CO2, and new pores could form. Zhou et al.[50] remarked that dissolution due to supercritical CO2 led to a reduction in the number of shale micropores, and that the effect of swelling transformed macropores into mesopores; also, supercritical CO2 could decrease the surface roughness of shale pores and the complexity of pore structures.

Although several studies have been conducted on the effect of supercritical CO2 treatment on shale microstructures, the factors that affect these changes differ and have not been adequately studied. In addition to differences in shale specimens and experimental conditions, the mechanisms of microstructure changes are important and have not yet been well-understood. For example, it is unclear whether CO2 pressure or mineral dissolution has a greater effect on porosity. To explore the different mechanisms by which supercritical CO2 changes shale microscopic structure, several methods including XRD, SEM, NMR, N2GA, and fractal theory were used in this study to determine the effects of different gas and pressure conditions on the shale microstructure. The results and significance for improving shale gas recovery are discussed in this paper.

Methodology

Specimen Preparation

Shale for this study was obtained from the Fuling area of the Sichuan Basin in China, which is currently the largest shale gas producing area in China, and the shale has an average total organic carbon content of 3.47% and a vitrinite reflectance (R0) of 2.56%.[48,51] The shale was processed into standard cylindrical specimens with a diameter of 50 mm and a length of 25 mm (Figure ). The parallelism of the end faces of each cylindrical specimen was not greater than 0.01 mm, and the vertical deviation was not greater than 0.05°. To eliminate experimental errors caused by differences among the specimens, an NMR analysis was conducted to select specimens with the closest pore characteristics and thus maximize the reliability of the test results (Figure ). To meet different experimental requirements, some select shale samples were crushed, and the powder was mixed evenly. Powder finer than 200 mesh was used for XRD, powder in range of 20–60 mesh was used for N2GA tests, and cuboid samples of size less than 8 mm × 8 mm × 3 mm were used for SEM tests. All processed shale samples were wrapped and sealed with a plastic wrap to prevent moisture and contamination.

Figure 1

Shale samples: (a) φ 50 mm × 25 mm cylindrical specimens; (b) powder samples; and (c) cuboid samples.

Figure 2

Initial transverse relaxation times T2 of shale specimens.

Experimental Procedure

Experiments with the shale samples were conducted in a high-pressure reactor system designed at our facility (Figure a). It consists of a high-pressure vessel with a maximum working pressure of 32 MPa, a Teledyne ISCO 260D syringe pump (Teledyne, Thousand Oaks, CA, USA) with a maximum working pressure of 51.7 MPa and a flow rate range of 0.001–107 mL/min, and a thermostatic water bath with a fluctuation of less than 0.1 °C. The experimental temperature was set at 60 °C to simulate real reservoir environment, and the exposure time was set to be 10 days, as per the time required for the interaction between shale and CO2 to reach equilibrium.[48] Argon (Ar) gas has also been used in these experiments. Being an inert gas, Ar does not react with shale and can thus be used to isolate the effect of pressure from a gas on the pores in shale, as opposed to the effects of CO2, which not only include pressure from the gas, but also adsorption, dissolution, extraction, and other factors. To study the mechanisms by which supercritical CO2 changes the shale pore structure, the exposure conditions (Table ) of shale to gas treatments were set as follows: untreated sample (reference group), 4 MPa Ar, 16 MPa Ar, 4 MPa CO2 (gaseous state), and 16 MPa CO2 (supercritical state). Before each experiment, the shale sample chamber was subjected to vacuum for at least 2 h to avoid the experimental error caused by air.

Figure 3

Experiment apparatus: (a) high-pressure reactor system; (b) “JSM6610LV” scanning electron microscope; (c) “MacroMR12-150H-I” NMR core analysis system; and (d) Micromeritics ASAP 2020 system.

Table 1

Experimental Scheme of Shale Exposed to Different Gas Conditions

specimen groups	gas	pressure (MPa)	temperature (°C)	time (d)	state
A					untreated
B	Ar	4	60	10	gaseous
C	Ar	16			gaseous
D	CO2	4			gaseous
E	CO2	16			supercritical

An “X-Max20/W500X” X-ray diffractometer (Co Kα radiation, 40 kV, 40 mA) was used for the XRD analysis of the 200 mesh shale powder to determine the change in the mineral content of shale before and after exposure. The micromorphology of shale was observed using a “JSM6610LV” scanning electron microscope (JEOL, Japan). The aperture distribution of shale was tested by a “MacroMR12-150H-I” low-field NMR core analysis system (Niumag Analytical Instrument Corp., China). The NMR analysis is based on the principle of using the hydrogen-containing fluid in pores to determine the pore structure of rocks.[52] Therefore, it was necessary to fully saturate the shale specimens with water before performing an NMR analysis. The N2 adsorption–desorption isotherms were obtained at a temperature of −196 °C at a relative pressure (p/p0) range of 0.01–0.99 using a Micromeritics ASAP 2020 system.

Results and Discussion

XRD and SEM

Table and Figure show the results for the mineral content in shale under different treatment conditions. It can be seen that the shale was mainly composed of quartz, dolomite, and calcite. The mineral content of shale did not change appreciably because of treatment with Ar, but did change significantly after treatment with CO2. The change in the mineral content of shale after treatment with CO2 is likely due to the chemical reactions of minerals in the CO2-water system (eqs and 2). The dissolution of calcite (CaCO3) in acid solution occurs before that of dolomite (CaMg(CO3)2) (eqs and 4), while quartz (SiO2) is stable and remains almost undissolved. Therefore, it is believed that the change in the proportion of calcite indicates the degree of dissolution, and the change in the proportion of quartz indicates the degree of the overall chemical reaction. The calcite content is the lowest, and the quartz content is the highest after treatment with 16 MPa CO2, indicating that when shale is exposed to supercritical CO2, the dissolution of calcite and the overall chemical reaction between shale minerals and the CO2–water system are the strongest compared to other conditions. This is because supercritical CO2 can extract bound water from clay minerals (TCCM), which increases the amount of CO2 in aqueous solution. The pH of the acid solution decreases, that is, the acidity increases, with an increase in the pressure of CO2.[53] Compared with a 4 MPa CO2 environment, the 16 MPa CO2 environment has a higher acidity, which results in a sudden decrease in calcite content and a sudden increase in quartz content.

Table 2

XRD Resultsa

	mineral content (%)
treatment conditions	quartz	dolomite	calcite	TCCM	feldspar	pyrite	hematite
untreated	42.7	30	16.2	9.3	0.8	0.5	0.5
4 MPa Ar	42.6	30.1	16.2	9.2	0.9	0.6	0.4
16 MPa Ar	42.7	29.8	16.3	9.3	0.9	0.5	0.5
4 MPa CO2	43.3	30.2	15.1	9.8	0.8	0.4	0.4
16 MPa CO2	45.5	30.1	12.4	10.4	0.6	0.5	0.5

TCCM represents clay minerals such as kaolinite, illite, and montmorillonite.

Figure 4

Mineral content of shale under different treatment conditions.

The microscopic morphologies (Figure ) of shale samples under different treatment conditions show that compared with untreated and other treatment conditions, the pores on the shale surface seem to develop the most because of the dissolution of minerals when shale is exposed to a 16 MPa CO2 environment.

Figure 5

1000× magnification of the microscopic morphology of different shale surfaces.

NMR Analysis of Aperture Distribution

The aperture distribution of shale specimens was obtained from the distribution of the transverse relaxation time T2.[54,55] The pore size has a positive correlation with the transverse relaxation time. The relationship between pore size and transverse relaxation time can be expressed by eq as follows[56,57]where T2 is the transverse relaxation time, ms; Fs represents the shape factor of the pore; ρ is the relaxation strength of the transverse surface, nm/ms; r is the pore size, nm. For this study, Fs = 2 and ρ = 10 nm/ms.[38,43,58]

The aperture distribution of shale before and after exposure is shown in Figure . Before the tests were conducted, specimens with almost identical T2 curves were identified and selected for testing. Hence, the T2 curves for untreated shale in Figure are identical. It can be seen that the aperture distributions of shale samples show a bimodal distribution, with two peaks located in the pore size ranges of 1–250 nm (main peak) and 250–2000 nm (secondary peak). After treatment with 4 MPa Ar, the aperture distribution curve shows a slight shift to the left. Because Ar does not react with shale, the overall decrease in pore size for this condition is likely caused by the compression of shale pores because of the pressure of the gas. After treatment with 16 MPa Ar, the decrease in pore size corresponding to the main peak is more apparent, indicating that the increase in gas pressure causes a greater compression of shale pores. Also, some microcracks occur in shale because of the higher pressure of the Ar gas at 16 MPa; hence, the pore component corresponding to the secondary peak increases. It has been shown that when CO2 is adsorbed on a shale surface, it causes an expansion and deformation of shale and a reduction in pore size.[59−61] It can also be seen from Figure c that after treatment with 4 MPa CO2, the main peak decreases, while the secondary peak increases because of the dissolution of minerals such as calcite. In addition to adsorption and the dissolution reactions with shale minerals, supercritical CO2 also has the unique ability to extract organic matter, resulting in an increase in pore size and forming an acidic environment with higher solubility. Compared with other treatment conditions, when shale is exposed to a 16 MPa CO2 environment, the magnitudes of both the main peak and the secondary peak increase significantly, which indicates that the extraction reaction of supercritical CO2 has a greater effect on shale pore size than other factors.

Figure 6

Aperture distributions of shale under different treatment conditions: (a) 4 MPa Ar; (b) 16 MPa Ar; (c) 4 MPa CO2; and (d) 16 MPa CO2.

Porosity and Average Pore Size

To further reveal the effect of different mechanisms on shale pores, a cumulative porosity diagram (Figure ) was plotted based on NMR data. As seen in this figure, the porosity of untreated shale was 3.46%, and after the treatments with 4 MPa Ar, 16 MPa Ar, and 4 MPa CO2, the porosity had decreased in all cases, and the increments were −0.06, −0.15, and −0.25%, respectively. However, after being exposed to 16 MPa CO2, the porosity of shale increased from 3.46 to 3.63%, which represents an increase of 0.17%. Combining the data of Figures b and 7, it can be inferred that although a high gas pressure induces microcracks in shale, the porosity of shale decreases, indicating that the impact of gas pressure on shale pores is mainly compressive, regardless of the magnitude of the pressure. According to Table and Figure , it is seen that after treatment with 4 MPa CO2, the gas pressure causes a reduction in shale porosity of 0.06%, and the combined effect of CO2 adsorption and mineral dissolution causes the porosity to decrease by 0.19%. In general, the effect of CO2 adsorption on shale porosity is opposite to that of mineral dissolution, that is, adsorption-induced expansion reduces porosity, and dissolution increases porosity. Therefore, it can be concluded that in a gaseous CO2 environment, the adsorption of CO2 has a greater effect on shale pores than gas pressure and dissolution reactions. By comparing the change in shale porosity after treatments with 16 MPa Ar and 16 MPa CO2, it is found that after treatment with 16 MPa Ar, the shale porosity decreases by 0.15%, while after treatment with 16 MPa CO2, the shale porosity increases by 0.17%. The increase in porosity in the latter case is because the effects of the extraction and dissolution reactions of supercritical CO2 can not only overcome the effects of compression caused by gas pressure but also the compressive effects of CO2 adsorption; this indicates that after treatment with supercritical CO2, the change in shale porosity is mainly controlled by extraction and dissolution reactions. Consistent with the above effects on shale porosity, the average pore size of shale decreases after treatment with 4 MPa Ar, 16 MPa Ar, and 4 MPa CO2, but increases after treatment with 16 MPa CO2 (Figure ).

Figure 7

Cumulative porosity vs T2.

Table 3

Porosity and the Factors Affecting Porosity

treatment conditions	porosity/%	increment/%	factors affecting porosity
untreated	3.46
4 MPa Ar	3.40	–0.06	low gas pressure
16 MPa Ar	3.31	–0.15	high gas pressure
4 MPa CO2	3.21	–0.25	low gas pressure + adsorption + dissolution
16 MPa CO2	3.63	0.17	high gas pressure + adsorption + dissolution + extraction

Figure 8

Schematic diagram of shale porosity and the factors affecting porosity.

Figure 9

Average pore size of shale specimens.

Fractal Dimensions of T2 Spectra

According to the International Union of Pure and Applied Chemistry (IUPAC) classification of pore sizes,[62] pores of sizes 0–2 nm are defined as micropores, pores of sizes 2–50 nm are defined as mesopores, and pores larger than 50 nm are defined as macropores. Adsorption in rocks mainly occurs in micropores and mesopores, while macropores are the main channels for gas migration. Therefore, according to the T2 curves obtained from NMR tests, T2 = 2.5 ms (corresponding to a pore size of 50 nm) is chosen as the demarcation point, and the fractal dimensions of shale adsorption pores and seepage pores are calculated using eq .[63−66]Figure and Table present the results of the calculations of fractal dimensions. D obtained from adsorption pores represents the fractal dimension of the pore area, which reflects the roughness and heterogeneity of the surface of shale adsorption pores, and its value is typically between 1 and 2. D obtained from seepage pores is the fractal dimension of pore volume, which reflects the complexity of the flow channel of shale seepage pores, and its value is typically between 2 and 3. It can be seen from Table that both adsorption pores and seepage pores show a good fitting effect and that the fitting coefficient of adsorption pores is higher than that of seepage pores, indicating that adsorption pores have more fractal characteristics than seepage pores. Figure shows the variation of fractal dimensions of shale pores for different treatment conditions. For adsorption pores, D increases after treatment with 4 MPa Ar, 16 MPa Ar, and 4 MPa CO2, and the maximum value appears for the 16 MPa Ar environment, while D decreases only in the case of treatment with 16 MPa CO2, indicating that supercritical CO2 decreases the specific surface area and roughness of adsorption pores, which may weaken the adsorption capacity of shale. For seepage pores, there are no obvious changes in D after treatment with 4 MPa Ar, 16 MPa Ar, and 4 MPa CO2. However, after treatment with 16 MPa CO2, D decreases significantly, indicating that in a supercritical CO2 environment, the complexity of shale seepage pores declines, which is conducive for gas migration.where D represents the fractal dimension of pores; Sv represents the pore volume component occupied in the wetting phase; T2 is the relaxation time, ms; T2max represents the maximum relaxation time constant, corresponding to the capillary pressure. Besides the NMR method, it is worth pointing out that fractal theory has been effectively applied to natural porous media from different perspectives. Cai and Yu[67] introduced fractal to modify the classical Lucas–Washburn equation, and the proposed fractal time exponent model presented a theoretical insight to the effect of tortuosity on capillary flow. Xia et al.[68] employed three fractal structural parameters, fractal dimension, lacunarity and succolarity, to characterize scale-invariant complexity, heterogeneity, and anisotropy of rock microstructures, respectively.

Figure 10

Plots of ln(Sv) vs ln(T2) from NMR testing data under different treatment conditions: (a) untreated; (b) 4 MPa Ar; (c) 16 MPa Ar; (d) 4 MPa CO2; and (e) 16 MPa CO2.

Table 4

Fractal Dimension of Shale Samples from NMRa

	region 1 (T2 < 2.5 ms)			region 2 (2.5 ms ≤ T2 < 100 ms)
treatment conditions	fitting equation	DN1	R2	fitting equation	DN2	R2
untreated	y = 1.797x – 1.902	1.203	0.89555	y = 0.056x – 0.215	2.944	0.67133
4 MPa Ar	y = 1.793x – 0.981	1.207	0.89045	y = 0.046x – 0.178	2.954	0.70342
16 MPa Ar	y = 1.718x – 0.905	1.282	0.87899	y = 0.048x – 2.952	2.952	0.85493
4 MPa CO2	y = 1.755x – 1.099	1.245	0.89046	y = 0.056x – 0.220	2.944	0.74793
16 MPa CO2	y = 1.845x – 1.537	1.155	0.90366	y = 0.091x – 0.353	2.909	0.681

T2 is the transverse relaxation time obtained from the NMR test.

Figure 11

Variation of fractal dimensions of shale pores.

N2 Adsorption–Desorption Isotherms

The N2 adsorption–desorption isotherms of shale samples under different treatment conditions are shown in Figure , where the shapes of the isotherms are found to be similar. When the relative pressure p/p0 is in the range of p/p0 = 0–0.45, the adsorption isotherms rise slowly with relative pressure, and the main type of adsorption is the monolayer adsorption of N2, or microporous filling. When p/p0 reaches 0.45, the monolayer adsorption of N2 is considered complete. When p/p0 exceeds 0.45, the adsorption isotherms continue to rise with relative pressure, the slope increases gradually, and the adsorption transforms from monolayer adsorption to multilayer adsorption. Also, when p/p0 is more than 0.45, the isotherms of desorption show hysteresis loops because of capillary condensation. Also, compared with other treatment conditions, the volume of adsorption in shale samples, that is, the adsorption capacity, significantly decreases after treatment with 16 MPa CO2. Figure shows the specific surface area of shale measured by N2 adsorption. It can be seen that the specific surface area of shale decreases significantly after supercritical CO2 treatment.

Figure 12

Adsorption and desorption curves of shale specimens under different treatment conditions: (a) 4 MPa Ar; (b) 16 MPa Ar; (c) 4 MPa CO2; and (d) 16 MPa CO2.

Figure 13

Specific surface area of shale specimens.

Conclusions

To explore the different mechanisms by which supercritical CO2 changes the shale pore structure, this study uses shale specimens from the Sichuan Basin, China, and utilizes several methods such as XRD, SEM, NMR, N2GA, and fractal theory to analyze and compare the changes in shale pore structures under different gases and pressures. An analysis of XRD results showed that almost no changes in shale mineral content occurred after treatment with Ar, while the mineral content changed significantly after treatment with CO2. The mineral content changed the most, particularly in the proportions of calcite and dolomite in shale, which decreased, after exposure to supercritical CO2; these conditions also resulted in the highest number of microscopic pores, as observed by SEM. The NMR test results revealed that gas pressure caused compression of shale pores. In a gaseous CO2 environment, the effect of CO2 adsorption on shale pores was found to be greater than the effects of gas pressure and dissolution reactions. In a supercritical CO2 environment, changes in the structures of shale pores were mainly controlled by the extraction and dissolution reactions of supercritical CO2. The results of N2GA tests and fractal dimensions of T2 spectra showed that adsorption pores have more fractal characteristics than seepage pores. After treatment with 16 MPa CO2, the fractal dimensions of the adsorption pores and seepage pores both decreased, indicating that when shale is exposed to supercritical CO2, the specific surface area and roughness of the adsorption pores decrease, the adsorption capacity is weakened, and the complexity of the seepage pores declines, which is conducive for gas migration. Therefore, when supercritical CO2 is used to extract shale gas, it reduces the adsorption capacity of shale for CH4 and promotes the conversion of adsorbed gas to free gas, thereby improving shale gas production.