Abstract

Conventional hydraulic fracturing techniques typically consume large amounts of water when producing shale gas. Fracking fluids may cause environmental pollution. In contrast, supercritical carbon dioxide (scCO2) (above 31.8°C, 7.29 MPa) can displace CH4 in shale reservoirs. Achieve CO2 sequestration while increasing the shale gas production. We studied the mechanical properties and fracture characteristics of a shale under the action of scCO2, nitrogen, helium, and water by comparing the triaxial compression tests of shale samples with seven coring angles. The results show that: (1) scCO2 effectively reduced compressive strength of the shale and weakened the anisotropy of shale; (2) scCO2 caused the content of dolomite, calcite, and illite to decrease by 4.7%∼13.5%, respectively; (3) scCO2 produced micropores and microfractures 10 times larger than the original size in the microstructure. These microstructures can help improve the seepage and gathering of shale gas, leading to enhanced shale gas recovery and CO2 storage.

1. Introduction

Current hydraulic fracturing of shale needs thousands of tons of water and proppant. Besides, the flow-back fluid is typically harmful to the environment, making its handling difficult and costly. In addition, recent research showed large-scale multistage hydraulic fracturing will generate microseismic [1], which may further trigger stronger geological movements and be a serious threat to people’s lives and properties. China is a country with scarce water resources per capita, and the distribution of water resources is uneven. Most of the shale gas-rich areas in China are water-scarce, such as mountains and hills [2], and conventional hydraulic fracturing requires large amounts of water. Shale gas extraction will increase water pressure on local water resources. The large amounts of water injection into shale formations may cause ground sliding and trigger earthquakes. This dramatically increases the cost of hydraulic fracturing methods; therefore, new nonwater fracturing methods for shale gas production become very attractive in these regions. On the other hand, carbon capture and storage (CCS)-technologies are becoming more practical in reducing greenhouse gas emissions [3]. Carbon capture has already been applied commercially in certain industry branches, and CO2 is widely used in the natural gas processing industry [47]. Supercritical CO2 has the characteristics of low viscosity, low-surface tension, high diffusivity, no hydration with clay, nontoxic. Injecting the industrial liquefied CO2 into shale can realize the following: (1) scCO2 jet fracturing; (2) effective increase of shale permeability; (3) replacement or displacement of shale gas to enhance its recovery; (4) carbon geological sequestration [811]. Currently, scCO2 injection has become an environment-friendly alternative to the conventional hydraulic fracturing. Hence, to better understand its application in shale gas production and carbon sequestration, it is important to study the interaction between scCO2 and shale.

1.1. Breaking Rocks with scCO2 Jet

Kolle and Marvin [12, 13] introduced scCO2 in the coiled tubing drilling method. He found that scCO2 jet has a stronger permeability in the reservoir rock and can effectively reduce the fracture pressure of the rock. Du et al. [7, 14] carried out an experimental study on sandstones breaking with the scCO2 jet. Their study indicated that scCO2 caused the phenomena of large volumetric layered broken. Liu et al. [15] studied the fracture extension behaviour under the influence of supercritical CO2 jets and different fracture types.

1.2. Chemical Reaction between scCO2 and Minerals in Sandstone

Many researchers [1625] studied the chemical reactions of CO2-sandstone-brine under the supercritical condition of CO2. They found that scCO2 corroded the mineral surface and disrupted the original pore structure. Irregular etching marks were found on the surface of the mineral crystals under the scanning electron microscope.

1.3. Chemical Reactions between scCO2 and Minerals in Shale

Lahann et al. [26] put shale caprock in CO2-brine under high temperature and pressure to study their reactions. Results showed that the relative contents of some elements in the filtrate were higher than that in the control case. Xu et al. [27] found that kerogen decomposed stably and continuously in CO2 with high pressure. In addition, Angeli et al. [28] detected hydrogen (1% in content) at the CO2 outlet during his experiments on the scCO2-shale caprock, which confirmed shale organic matter decomposed under scCO2 condition. Moreover, Allawzi et al. [29] found that kerogen debris also decomposed under the scCO2 condition.

1.4. Changes of Rock Physical Properties

Lu et al. [30] found that the shale treated with supercritical CO2 slickwater had a significant increase in the number of micropores, and its pore area and volume increased. Yang et al. [31] investigated the effect of porosity change on the adsorption performance of CH4 after supercritical CO2 action on the shale at different temperature and pressure conditions. Fatah et al. [32] tested the effect of different mineral contents and temperatures on the hydrophilicity of shale under the action of supercritical CO2. The results show that shales with high-quartz content are highly hydrophilic. Wollenweber et al. [33] studied the carbon sequestration efficiency and found that CO2 could decrease the breakthrough capillary pressure of the caprock by two-thirds. In addition, repeated CO2 treatments led to decreased capillary sealing efficiency and increased shale caprock permeability. Vialle and Vanorio [34] observed permanent change in microporous structures when studying the reaction between carbonate rocks and injected CO2-saturated water.

While the study of the properties of scCO2 and its effect on the microstructure of different types of rocks is important, its influence on the properties of shale gas formations cannot be ignored. Shale from Longmaxi shale gas formation in the Fuling shale gas field, Chongqing City, China, is rich in organic matter with an average content of 2.5%, thermal maturity of 2.5%, and porosity of 2.5% [3537]. Due to its special characteristics, it is critical to clear the interactions and effects between scCO2 and shale. In this study, we collected specimens from the aforementioned location and carried out strength tests at 7 dissimilar coring angles under scCO2 condition. The mechanical characteristics and fracture properties were studied systematically at the microscopic scale.

2. Experimental Section

2.1. Preparation of Shale Samples

The Longmaxi Formation rocks and minerals are highly brittle, mainly because they contain a large number of silicified graptolites, radiolarians, and other fossils [38]. Collect original shale outcrops and remove the regolith.

Figures 1 and 2 show that the test samples are drilled from the same outcrop shale to reduce errors caused by differences in mineral compositions. Set the coring angle β as 0°, 15°, 30°, 45°, 60°, 75°, and 90°, respectively. These samples were further cut into standard cylinders to ensure smoothness, parallelism, and perpendicularity to the axis on both end faces.

Then the porosity of specimens is measured by helium, which is based on the principle of pressure pulse [39], and the test pore pressure and confining pressure are 2 MPa and 5 MPa, separately. The results show that most of the specimens belong to the bedrock (the density is 2.55 g/cm3 on average). Porosity is 3.5% on average, and the permeability is on the level of nanodarcy and varies from 141 nD to 323 nD with an average of 200 nD. So, the sample-to-sample variation is so small that the specimens can be considered as the same.

2.2. Experimental Apparatus and Method

The experimental equipment is shown in Figure 3. To create scCO2 condition (above 31.8°C, 7.29 MPa), the following improved experimental system main components and workflows are used [40]. (1) An air-driven and liquefied gas booster are used to increase the gas pressure, as shown in Figures 3(a) and 3(b). Then, they are connected to the inlet line at the top of the specimen through pipelines. Finally, pumped high-pressure CO2 into the sealed specimen through the confining pressure barrel. (2) A heating ring is installed around the confining pressure barrel to increase temperature. A temperature sensor is located beside the specimen to measure the temperature of the specimen, as shown in Figure 3(b). (3) A confining unit provides the pressure required to fracture the specimen by pressurizing hydraulic oil in the confining pressure barrel.

The specimen is placed into the triaxial apparatus as shown in Figure 3(c). The experiment procedures are listed as follows: (1) wrap the specimen with heat shrinkable tubing; (2) install the displacement sensor and temperature sensor; (3) seal the barrel; (4) adjust experimental conditions; (5) maintain the experimental conditions to ensure shale pores filled with scCO2, (6) impose the axial load to begin the test under the uniform loading rate of 0.04 mm/min until specimen damages.

Table 1 shows that the experiments are designed to include four test groups. Each group were carried out at seven different coring angles (0°, 15°, 30°, 45°, 60°, 75°, and 90°), with 28 triaxial compression experiments in total. Nitrogen gas, Helium gas, and clear water at the same temperature and pressure ([a] Confining pressure. [b] Pressure of the test gas in Table 1). were selected as the control groups.

3. Results and Discussion

3.1. Rock Mechanics and Characteristics of the Longmaxi Shale

The shale’s mechanical anisotropy is notable. The triaxial compression strength in the helium group (with an average of 310 MPa) is lower than that in the scCO2 experimental group (with an average of 314 MPa), in the water control group (with an average of 321 MPa), and in the nitrogen control group (with an average of 328 MPa). The main reason is that gas injection at high pressure leads to high pore pressure, which decreases the absolute value of the confining pressure. Some literature [41, 42] showed that natural damages are compressed more heavily under higher confining pressure, which increases the triaxial compression strength of the shale. In contrast, hydration leads to a reduction in the compression strength of the shale.

Table 2 shows the results of core strength tests with different coring angles under the action of different media. Compared to the nitrogen control group, the compression strength of the shale decreased by between 2% and 20% under the influence of scCO2, with a maximum absolute decrease of 49 MPa. The reduced compression strength of the Longmaxi shale is 2–5% greater than that found in the Zheng’s et al. experimental results [41] under the CO2–NaCl solution environment. The compression strength of the shale under the influence of scCO2 decreased by between 1% and 6% more than the water control, with a maximum absolute decrease of 15 MPa. Some studies [43, 44] have shown that the hydration of shales leads to a reduction in strength due to the dissolution of some minerals and the dislodging of particles from microperspectives. This study aims to investigate the effect of scCO2 on the mechanical properties of shale. The mechanical properties of shale under the action of clear water will change due to the influence of hydration and other factors. To exclude interference, this study will mainly conduct a comparative experimental study with inert gases.

Figure 4 shows that the relation of triaxial compression strength and coring angle can be fitted by a sine curve with high R2. Triaxial compression strength increases gradually with β and reaches its maximum at β = 15°. Then, it decreases and reaches the minimum value at β = 60°. However, it increases again when β is added from 60° to 90°.

Figures 58 show that there are no distinct phases representing fractures and pores being compacted on the curves. The stress-strain curves are approximately straight until the peak stress is reached. As the stress increases, the stress-strain curve begins to bend. This is because the shale is fractured after the peak stress is reached. A clear brittle fracture sound can be heard when the shale is fractured. After peaks, the stresses decrease rapidly to the lowest stress points. These results show that Longmaxi shale has high brittleness under all three experimental conditions.

By comparing the stress-strain curves in Figures 510, the scCO2 experimental group has the lowest yield strain into the damage phase. Meanwhile, the peak stresses of the scCO2 experimental group are lower than that in the control groups. After the maximum stress points, the stresses of the scCO2 experimental group drop sharply than that in the other two groups.

3.2. The Failure Modes of Specimens

Rock failure modes are affected by many factors. Among them, test conditions are the main factors [45]. Besides, different coring angles also lead to different failure modes. The failure modes can be mainly divided into splitting failure and shear failure modes.

Figure 11 shows the shale failure modes under diverse experiment conditions and coring angles. When the coring angle is less than 15°, the failure modes of the shale specimen contain both splitting and shear modes (Figure 12) and have formed Y-type fracture geometry. The failure condition is similar in the other two control groups. However, the specimens in the scCO2 experimental group break more thoroughly at 15° coring angle, showing more fractures. At the coring angle of 15°, the scCO2 group has a higher strain value than the nitrogen control group before reaching peak stress (Figure 6). This means shale failed after a longer period of compression under the action of scCO2. During this process, scCO2 repeatedly acted on the fractured fractures, producing more complex fractures.

At β = 30°, 45°, or 60°, the failure modes are the single shear or the double shear modes due to the slipping between the bedding planes (Figure 13). Some graptolites are observed on the fracture surfaces. The splitting failure modes of rock samples with low coring angles are not obvious due to the confining pressure and high coring angle. In addition, smooth shear fractures were formed on the bedding planes with severe slip. Therefore, the triaxial compression strength of the shale decreases as the angle increases.

At β = 75° or 90°, the failure modes mainly follow the splitting mode. The main reason is that the direction of axial loading is almost parallel to the shale bedding planes.

In summary, the combined effects of scCO2 and bedding planes led to various failure modes in the experiments. At low coring angles (0°–15°), the specimens in the scCO2 experimental group produced more fractures compared to the other control groups due to more reactions between scCO2 and shale. At moderate coring angles (30°–60°), the fractures could easily cross the specimens along the bedding planes. At high coring angles (75°–90°), some fractures were found on the specimens, due to the tensile damage.

3.3. Changes of Mechanical Anisotropy

Figures 14 and 15 show that the mechanical parameters of the Longmaxi shale changed with the internal bedding directions. At low coring angles (0°–15°), the elastic modulus of the scCO2 experimental group is lower than that of the N2 control group. However, at moderate and high coring angles (45°–90°), the elastic modulus of the scCO2 experimental group is higher than that in both control groups. However, at moderate and high coring angles (45°–90°), the elastic modulus of the scCO2 experimental group is higher than that in both control groups with an exception at the 60° coring angle, where it is slightly smaller than that of the N2 control group.

To describe the anisotropic characteristics of the shale, an anisotropic in-dex Rc is defined as follows [46]:where Xci(0) and Xci(90) are the shale mechanical parameters at the coring angle of 0° and 90°, respectively. Anisotropic indexes of triaxial strength, elastic module, and Poisson’s ratio are listed in Table 3. The anisotropic indexes of triaxial strength and elastic modulus in the scCO2 experimental group are much lower than those in the N2 control group, whereas the anisotropic index of Poisson’s ratio in the scCO2 experimental group is much higher than that of the N2 control group. The decreased shale mechanical anisotropic indexes of triaxial strength and elastic modulus in the scCO2 experimental group indicate that scCO2 tends to make the shale more isotropic.

3.4. Mineralogical Changes

The components of shale minerals in each experimental group have been obtained by the X-ray diffraction experiments. According to Kaszuba’s et al. test method [47], the samples were analyzed by XRD. In addition, analyzed the clay composition of the samples.

The several tests were performed within each group to reduce the error induced by rock heterogeneity. The mineral components and the clay components are shown in Tables 4 and 5.

Little difference was found in the mineral components of the N2 and Helium control groups. However, compared with the two control groups, the relative contents of calcite, dolomite, and illite in the scCO2 group reduce significantly. The content of calcite decreases about 5.6% on average (the maximum is 7.1%, and the minimum is 4.7%). The content of the dolomite and illite decreased by 6.9% (the maximum is 8.6%, and the minimum is 5.1%) and 10.6% (the maximum is 13.5%, and the minimum is 8.5%), respectively. The relative content of quartz in the scCO2 group increases due to very little reaction of quartz with CO2 in the short term and the decrease of other minerals.

Wu et al. [48] tested the water content of shale. CO2 gas becomes corrosive when contact with water. As a result, shale minerals and organic matter will be corroded by H+ [47, 49]. The following equation [49] can express the dissolutions of the calcite and dolomite minerals:

Meanwhile, the scCO2 condition facilitates the decomposing of minerals and organic matters through the generation of connected corrosion micropores and microcracks seen in Figures 16(c) and 16(d). This is the reason why notably changes of mineral components are observed in the scCO2 experimental group. In contrast, mineral components in the other two control groups show few changes.

3.5. Microstructure Changing of Fracture Surface

The microscopic structure of shale plays an important role in shale gas production. The specimens were carefully observed with a Quanta-450 SEM device to study the microscopic structure of Longmaxi shale under three experimental conditions. Their microscopic structures are summarized in Figure 16, with each row representing each group.

It has been shown that [35, 36] the composition of the Longmaxi Shale includes dolomite, quartz, muscovite minerals, organic matter with a small number of micropores and microfractures of 0.1 to 5 μm. In the control group of helium, it can be seen that the brittle minerals are closely bonded to the clay minerals from Figures 16(i) and 16(l); Meanwhile, the experimental results show that the natural microcracks are less affected by the axial load (Figures 16(j) and 16(l)). The shale microstructure of the helium control group changes little during the experiments.

Figures 16(e) and 16(l) show that no significant differences in the shale microstructure are observed between the N2 control group (Test 2) and the Helium control group (Test 3). Since the clay minerals are carried by N2, more of them are found around the surface of the quartz, as shown in Figures 16(h) and 16(l).

However, the shale microstructure in the scCO2 experiment (Test 1) corrodes severely (Figures 16(a)–16(d)). Due to the low viscosity and high diffusion rate of scCO2, it is easy to invade the micropores and microfractures during the experiment, which increases the fracture connectivity. The scCO2 forces some fractures to continually expand, generating long complex fractures (Figure 16(a)), and delamination sheets (Figure 16(b)). On the other hand, the high-speed scCO2 fluid in the micropores leads to the denudation and migration of clays, generating complex microscopic structures (Figure 16(c)). Therefore, scCO2 jets can effectively reduce the threshold pressure of breaking and fracturing rock and improve the degree of rock fracturing.

The microstructures of shale corrode obviously in the scCO2 condition. Some corrosive voids and grooves are observed. Figure 16(d) shows that the diameters of the corrosive voids and grooves are more than 10 μm, 10 times larger than the original size of microholes and microfractures.

Complex chemical and physical reactions between scCO2 and shale can generate numbers of corrosion voids and grooves, which are conducive to the seepage and accumulation of shale gas, enhancing recovery. Meanwhile, these reactions severely damage the original microstructures of shale, reducing its strength and mechanical anisotropy.

4. Conclusions

In our experiments, scCO2 reduces the triaxial strength of Longmaxi shale by 2% to 20% compared to the other two gases. It also decreases the shale mechanical anisotropic indexes. The combined effects of scCO2 and the bedding plane lead to various failure modes in the experiments. At low coring angles of less than 15°, more fractures are observed on the specimens in the scCO2 experimental group; at moderate coring angles (30°–60°), fractures cross the specimens more easily along the bedding planes; at high coring angles (75°–90°), some splitting fractures produced by tensile damage are found on the specimens.

The calcite, dolomite, and illite contents of Longmaxi shale decreased by 5.6%, 8.6%, and 10.6% in the scCO2 experimental group, respectively. The microstructures of shale are remarkably corroded under the scCO2 condition due to the complex chemical and physical reactions. Numbers of corrosive voids and grooves are produced with diameters of more than 10 μm, 10 times larger than those of the original microholes and microfractures. These microstructures can help improve the seepage and gathering of shale gas, leading to enhanced shale gas recovery.

Data Availability

Study data not yet publicly available.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

The authors would like to thank the National Natural Science Foundation of China (Grant no: 52004225), Application Foundation Project in Sichuan Province (2022NSFSC1273), and Open Fund (PLN2020-1) of State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation (Southwest Petroleum University) for financial support and permission to perceive this study.