Recently, CO2 geological sequestration combined with enhancing deep saline water/brine recovery is regarded as a potential strategic choice for reduction of CO2 emissions. This technology not only achieves the relatively secure storage of CO2 which was captured during industrial processes but also can enhance the recovery of water for drinking, industrial, and agricultural utilization. However, the impact of CO2-water-rock reactions on the shale reservoir in the system is unclear and the sealing performance of mudstone caprock has not been investigated. For analyzing the mechanism of mineral alteration in the shale reservoir, a three-dimensional injection-production model in the double-fractured horizontal well pattern is established according to actual parameters of shale and mudstone layers. In addition, mineral alteration was characterized and caprock sealing performance was also assessed. Numerical results showed that the presence of CO2 can lead to the dissolution of k-feldspar, oligoclase, chlorite, and dolomite and the precipitation of clay minerals such as kaolinite, illite, and smectite (Ca-smectite and Na-smectite). Due to positive ion released by dissolved primary minerals, the precipitation of secondary carbonate occurs including ankerite and dawsonite, which induces the mineral sequestration capacity of the shale reservoir. The amount of CO2 sequestration by mineral is 51430.96 t after 200 years, which equals 23.47% of the total injection (219145.34 t). Besides, the height of the sealing gas column is used for evaluating the sealing performance of the shale-mudstone interface. Results show that the height of the sealing gas column at the interface above the injection well is lower but the maximum value of CO2 gas saturation is only 0.00037 after 200 years. The height of the sealing gas column at the interface is greater than 800 m, which can be classified as level II and guarantee the security of the CO2 storage. The analysis results provide reliable guidance and reference for the site selection of CO2 geological sequestration.
Carbon dioxide (CO2) has contributed to the major increase in the atmospheric concentration of greenhouse gas, which is mainly caused by the combustion of fossil fuels in power plants and other industrial processes. In recent years, carbon capture and storage (CCS) has been widely regarded as one of the effective methods to achieve the reduction of CO2 emissions . Instead of injection into saline aquifers, carbon capture utilization and storage (CCUS) is considered as a better option for economic benefit, such as CO2-EGS and CO2-EOR [2, 3] Nowadays, a novel approach of CCUS is proposed, which is called the CO2 geological storage combined with deep saline water/brine recovery (CCS-EWR). This technology has shown the potential of simultaneously increasing the CO2 storage and producing the underground water from the aquifers .
Due to CO2 injection into the formation, CO2-water-rock chemical interactions will induce mineral alteration. Cui et al.  established a comprehensive reactive transport model. Meanwhile, they carried out a series of simulation to analyze the effect of CO2-water-rock geochemical reactions on reservoir physical properties. Actually, initial mineral composition plays a vital role in CO2-water-rock reactions during CO2 geological sequestration . On the basis of a dataset from the test well of Knox saline reservoirs in Kentucky, Zhu et al.  used 2-D radial-reactive transport models to evaluate mineral dissolution and precipitation. Results from the kinetic models suggested that if dawsonite is suppressed, the mineral trapping is essentially nonexistent. Wang et al.  developed radial models to determine the impacts of mineralogical compositions on trapping amounts for the CO2 geological storage. Their study indicated that dissolution of chlorite is beneficial to the formation of secondary carbonate minerals and CO2 mineral trapping. Xu et al.  employed a 2-D horizontal model in a five-spot well pattern and studied on the rock-fluid interaction for a CO2-based geothermal system.
Acting as a seal for CO2 rising from the reservoir, the caprock must be able to withstand the changes in physical and chemical properties due to the CO2-brine-rock mineral interactions . Gherardi et al.  presented 1-D and 2-D numerical simulations of reactive transport in the process of CO2 geological storage. The simulation results demonstrated that the occurrence of CO2 leakage from the reservoir may have an influence on the geochemical evolution of the caprock. Tian et al.  considered clay-rich shale and mudstone as two types of caprock for modeling analyses. Their results indicate that the mudstone is more suitable to be used as a caprock. Na et al.  used reaction experiments under in situ reservoir pressure conditions of 35 MPa and various temperatures (from 150 to 170°C) to analyze mineral alteration induced by mixtures of CO2 and water. After 12 days CO2-brine-rock reaction, XRD and SEM analysis results indicated that the existence of CO2 induced the dissolution of primary rock minerals such as calcite and feldspar. Thus, the precipitation of secondary carbonate (calcite and ankerite) was observed. Both geothermal energy recovery and CO2 geological storage could be achieved by the above geochemical processes. Khather et al.  carried out core-flood experiments to evaluate changes in the properties of dolostone samples. Overall, a slight increase in the porosity was observed in most samples, which might be caused by the dissolution of carbonate (dolomite and calcite). Wollenweber et al.  studied the effects of high-pressure CO2 exposure on fluid transport properties and mineralogical composition of two pelitic caprocks, a limestone, and a clay-rich marl lithotype. In repetitive gas breakthrough experiments, the capillary CO2 sealing efficiency of the initially water-saturated sample plugs was found to decrease for both caprocks. Espinoza and Santamarina  measured the breakthrough pressure and ensuing CO2 permeability through homogeneous specimens made of sand, silt, kaolinite, and smectite. Their experimental results and data gathered from previous studies highlight the inverse relationship between breakthrough pressure and pore size, as anticipated by Laplace’s equation.
Compared with traditional CCUS methods, CCS-EWR has two advantages: (1) on the basis of reasonable pumping well engineering design, reservoir pressure release and water production can be controlled to achieve the safety and stability of large-scale CO2 geological sequestration and (2) deep saline water can be collected and treated for life drinking, industrial, and/or agricultural utilizations to alleviate water shortage as well as ecological impact . Shale has the potential to store a portion of the injected CO2 into the reservoir . Once shale gas is depleted from shale, CO2 could be trapped in the reservoirs by taking advantage of the newly available pore space and existing well infrastructure [19, 20]. In this study, a simulation model was established to perform the CCS-EWR process in the shale reservoir. However, Each CCUS system may behave differently due to differences in compositions of brine and rock. For this reason, it’s advisable to perform a specific case study for each type of reservoir before the CO2 sequestration process . Thus, the numerical simulation method was employed in this study to investigate the CO2-brine-rock interaction occurring in the CCS-EWR system. Meanwhile, mineral trapping capacity and sealing performance of caprock were assessed.
2. Simulation Approach
2.1. Governing Equations
Chemical transport equations can be written as equation (4), which are in terms of the total dissolved concentrations of chemical components that are concentrations of their basis species plus their associated aqueous secondary species. In equation (4), is component concentration, mol/L, is diffusion coefficient, m2/s, is medium tortuosity, and is the aqueous chemical component index. is the source/sink term of the th component, mol/m2/s, in which is the part which transforms into the solid phase. For other terms in equation (4), molar units are used instead of kg.
2.2. Model Setup
In this study, a three-dimensional model was developed with the size of . As shown in Figure 1, the formation consists of two parts, which are shale and mudstone. At their interface, a thin layer is divided from mudstone to monitor the evolution characteristic of the mudstone layer. Shale #2 is the injection reservoir which has already been hydrofractured. Considering the actual exploring mode of shale gas, the double-horizontal well pattern was employed for the current simulation study. The distance between the injection well and the production well is 300 m, and the spacing of the fracture is 100 m. With the injection rate of 0.3472 kg/s (30 t/day), the entire injection time is 20 years. After half-year injection, the production well is activated with the rate of 0.2 kg/s. When the injection period is ceased, we monitor the model for additional 180 years.
2.3. Initial Conditions and Parameters
Simulation was carried out for isothermal condition at 97.3°C, which corresponds to the values measured at the top of the shale gas reservoir in the site. According to the fluid flow test results of the core, the basic data (i.e., density, porosity, and permeability) was acquired. It should be noted that the core sample of shale came from three depths (2554.36 m, 2581.59 m, and 2593.87 m) but their results of porosity and permeability are very close. Therefore, we use the average value of porosity and permeability as shown in Table 1.
The pore pressure of the shale gas reservoir (shale #2) is 26.8 MPa which has been measured, and the pore pressure equilibrium is required before simulation solution. The pressure equilibrium simulation was conducted firstly, and the results are used for the initial pore pressure condition which is shown as Figure 2.
As shown in Table 1, rock grain density, porosity, and permeability are obtained by rock sample test. Parameters of relative permeability and capillary pressure for rock matrix are from Tian et al. .
Due to the existence of the hydraulic fracture, the MINC (multiple interacting continua) method is adopted which is used for modeling the fluid flow in fractured-porous media . On the basis of hydraulic fracturing design and microseismic data, fracture basic data was obtained as shown in Table 2.
2.4. Mineralogy and Aqueous Composition
According to X-ray diffraction (XRD) mineralogical analysis of core samples from mudstone and shale in the site, mudstone consists mainly of quartz and clay minerals (i.e., kaolinite, chlorite, illite, and smectite). For shale samples from three depths, their mineral compositions are obviously different from each other. In Table 3, primary minerals mean existing at the beginning of the simulation and secondary means minerals could be formed after simulations. For revealing the process of mineral evolution, three common secondary minerals are specified which could be formed after CO2 injection.
Initial aqueous composition used in the model (Table 4) is based on water samples taken from a nearby well completed in the site.
2.5. Thermodynamic and Kinetic Parameters
The present simulation was performed with the TOUGHREACT code, which is a comprehensive multicomponent reactive fluid flow and geochemical transport simulator . In this code, the primary source used for equilibrium constants of aqueous species and minerals is originally obtained from EQ3/6 V7.2b database . Moreover, a portion of thermodynamic properties of several minerals and aqueous species has been revised on the basis of Xu et al. .
The parameters for calculating kinetic rate constants of minerals are presented in Table 5. Note that (1) is the specific surface area, is kinetic constant at 25°C, is activation energy, and is the power term (equation (6)); (2) the power term for both acid and base mechanisms is with respect to H+; and (3) for pyrite, the neutral mechanism has with respect to O2(aq) and the base mechanism has two species involved: one with respect to H+ and another with respect to Fe3+. The reaction kinetic data of minerals are listed in Table 5, which are given by Tian et al. , Liu et al. , Yang et al. , and Xu et al. .
3. Results and Discussion
3.1. Mineral Alteration in the Shale Layer
As shown in Figure 3, with the injection of CO2 into the shale layer, CO2 saturation close to the injection well reaches the maximum value 0.54 after 20 years. Due to dissolution of CO2 into groundwater, pH decreases to 4.98 which will induce dissolution of primary minerals and precipitation of secondary minerals in the shale layer.
Figure 4 shows the dissolved primary minerals including K-feldspar, oligoclase, chlorite, and dolomite. Because of the larger kinetic constant, dolomite dissolves more severely and easier to reach the equilibrium state, which results in a smaller dissolution volume fraction in hydraulic fractures. Meanwhile, the change of Ca2+ concentration which is caused by dolomite dissolution affects oligoclase dissolution. Due to larger permeability, sufficient reaction cannot occur in fractures which leads K-feldspar and chlorite to dissolve obviously around fractures.
Figure 5 shows changes in the volume fraction of quartz and clay minerals. Precipitation of clay minerals requires SiO2(aq) and AlO2− provided by the dissolution of feldspar minerals and chlorite. Illite is precipitated in the period of injection. K-Feldspar provides K+ which is needed for illite precipitation, so the spatial distribution of changes in volume fractions of illite precipitation and K-feldspar dissolution negatively correlates. The dissolution of Na- and Ca-bearing oligoclase supplies Na+ and Ca2+ for the precipitation of smectite-Na and smectite-Ca; thus, both are formed with a volume fraction amount of about 0.0019 near hydraulic fractures. Meanwhile, a small amount of quartz is precipitated.
Figure 6 shows changes in the volume fraction of carbonate. Calcite dissolution occurs close to the injection side, which is with a volume fraction about 0.0069 near hydraulic fractures. Mg2+ is released by dissolution of chlorite and dolomite, and it causes magnesite precipitation. At the same time, precipitation of secondary carbonate ankerite and dawsonite is obtained. All of the precipitation of the above carbonate will eventually increase the effective mineral trapping capacity of CO2 in the shale layer.
As shown in Figure 7, because of mineral alteration in the period of injection, the porosity of the shale layer decreases slightly which causes small changes in permeability. But those tiny decreases will not significantly lower the connectivity of the shale layer.
3.2. CO2 Mineral Trapping of the Shale Layer
Carbonate dissolution enhances the effect on the CO2 trapping capacity of the shale layer, and it is of great significance to CO2 storage. According to the result of CO2 sequestration by mineral (SMCO2, kg/m3 medium) in Figure 8, the performance of CO2 mineral trapping is consistent with precipitation of carbonate which includes magnesite, ankerite, and dawsonite.
The amount of CO2 trapped in mineral increases significantly in the process of injection, and the rate of increase obviously reduces after injection stops. The amount of CO2 sequestration by mineral is 51430.96 t after 200 years which equals 23.47% of total injection (219145.34 t). The results show that the current CCS-EWR system has an excellent effect on CO2 mineral trapping.
3.3. Sealing Performance of the Mudstone Layer
Since the section of the caprock affected by geochemical reactions is very thin, mineral trapping of CO2 in the caprock can be negligible . The injected CO2 plume initially migrates towards the top of the shale reservoir due to buoyancy until it reaches the mudstone layer. Hence, the mudstone layer above the shale layer is the key to prevent CO2 from large CO2 leakage, whose sealing performance would ensure the security of the CCS-EWR system. In Figure 9, the simulation results show that the porosity of the shale-mudstone interface above the injection well increases from the initial value of 0.0127 to the maximum value of 0.0180 after 200 years. Those increases will induce small leakage of CO2. The results of CO2 gas saturation after 200 years show that maximum value is only 0.00037; hence, the large amount of CO2 leakage does not occur.
The HSGC of shale-mudstone interface after 200 years is presented in Figure 10, and the result shows that HSGCs are all greater than 800 m. The pore size of the shale-mudstone interface above the injection well increases; thus, HSGCs are smaller than those of the above production well. According to the evaluation index in Table 6 , generally, HSGC of the mudstone layer can be classified as level II. For maintaining the high sealing performance of caprock in the period of CO2 geological sequestration, it is crucial to take breakthrough pressure into consideration before site selection. Due to the existence of natural cracks, microfractures, and fault systems in caprocks, the actual sealing performance of caprock will be worse than theoretical value. Consequently, it is reasonable to take the sealing performance index as high as possible.
In this study, the conceptual model was established based on a particular depleted shale gas reservoir. The model was built on the assumption that the initial shale gas saturation was 0. In fact, there should have been residual shale gas in the reservoir. The existence of shale gas (CH4) would cause CO2-CH4 competitive adsorption and have an effect on CO2 transport [36–38]. However, this factor was not considered in this study due to the lack of shale gas saturation data.
Deformation of the formation will occur because of CO2 injection. Considering the existence of shale the bedding plane, natural fractures, and faults [39–53], the deformation behavior and the effect on fluid transport will be extremely complicated. The reactivation potential of faults does exist, which may provide a leak path for CO2 to escape the storage formation [54–56]. In addition, the acid mixture of CO2 and formation water will cause the influence on the strength and fracture behavior and induce the mechanical failure of the reservoir, caprock, or their interface [57, 58]. The fractures caused by mechanical damage will have an impact on the sealing performance of the CCS-EWR system. The next step is to further study the influence of rock mechanical property degradation, natural fracture properties, injection-production rates, and other factors on the sealing performance of the CCS-EWR system and further investigate effective measures for the security of CO2 storage.
In this paper, according to actual parameters of shale and mudstone layers in the CCS-EWR system, a three-dimensional injection-production model in the double-fractured horizontal well pattern is established to investigate the mechanism of mineral alteration. With the developed model, the mineral sequestration capacity of the shale reservoir is analyzed and the sealing performance of mudstone caprock is assessed. The main conclusions obtained from this research are as follows:
Mineral alteration is driven by pH decrease due to the presence of CO2. The dissolved primary minerals include K-feldspar, oligoclase, chlorite, and dolomite. Meanwhile, clay minerals such as kaolinite, illite, and smectite (Ca-smectite and Na-smectite) are precipitated. Because of large permeability, sufficient reaction cannot occur in fractures. Hence, significant mineral alterations are observed around fractures on the side of the injection well
Due to positive ion released by dissolved primary minerals, the precipitation of secondary carbonate occurs including ankerite and dawsonite, which induces the mineral sequestration capacity of the shale reservoir. The amount of CO2 sequestration by mineral is 51430.96 t after 200 years, which equals 23.47% of total injection (219145.34 t).
The height of the sealing gas column (HSGC) is used for evaluating the sealing performance of the shale-mudstone interface. Results show that HSGCs of the interface above the injection well are lower because of the decrease of porosity but the maximum value of CO2 gas saturation is only 0.00037 after 200 years. The entire HSGCs of the interface are greater than 800 m, which can be classified as level II and guarantee the security of CO2 storage
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare no conflict of interest.
The conceptualization was done by Yuwei Li; the methodology was done by Maosen Yan and Yuwei Li; the investigation was done by Jun Zhang, Xu Han, and Ziyuan Cong; writing and original draft preparation were done by Maosen Yan and Chi Ai; writing, review, and editing were done by Maosen Yan; visualization was done by Xiaofei Fu, Fahao Yu, and Wei Li.
This research was funded by the Opening Fund of Key Laboratory of Continental Shale Accumulation and Development (Northeast Petroleum University), Ministry of Education, Youth Backbone Training Program of Northeast Petroleum University (no. 15041260501), and the Research Start Project of Northeast Petroleum University (no. 1305021857).