Study on Main Controlling Factors of CO₂ Enhanced Gas Recovery and Geological Storage in Tight Gas Reservoirs

Abstract

Tight gas reservoirs, as important unconventional natural gas resources, face low recovery rates due to low porosity, low permeability, and strong heterogeneity. CO₂ Storage with Enhanced Gas Recovery (CSEGR) technology combines CO₂ geological storage with natural gas development, providing both economic and environmental benefits. However, the main controlling factors and influence mechanisms remain unclear. This study utilized the PR-EOS to investigate CH₄, CO₂, and natural gas physical properties, established a numerical simulation model considering CO₂ dissolution and geochemical reactions, and explored the influence of injection scheme, injection rate, production rate, and shut-in condition on CO₂ enhanced recovery and storage effectiveness through orthogonal design. Results show that CO₂ exhibits significant differences in compressibility factor, density, and viscosity compared to natural gas, enabling piston-like displacement. Intermittent injection slightly outperforms continuous injection in recovery enhancement, while continuous injection provides greater CO₂ storage capacity. The ranking of the significance of different influencing factors for enhanced oil recovery is as follows: injection rate > production rate > injection scheme > shut-in condition. For the effect of geological storage of CO₂, it is as follows: injection rate > injection scheme > production rate > shut-in condition. During gas injection, supercritical, ionic, and dissolved CO₂ continuously increase while mineral CO₂ decreases, with storage mechanisms dominated by structural and residual trapping. The study provides scientific basis for optimizing CO₂ flooding strategies in tight gas reservoirs.

1. Introduction

Tight gas reservoirs, as important unconventional natural gas resources, play a significant role in the global energy structure [1]. However, due to their characteristics of low porosity, low permeability, and strong heterogeneity, tight gas reservoirs commonly face technical challenges of low recovery rates during development [2,3,4]. Currently, the main technical means for enhanced gas recovery in tight reservoirs include hydraulic fracturing, horizontal well technology, and refracturing [5,6,7]. While these technologies have improved gas reservoir development effectiveness to some extent, technical bottlenecks still exist. China’s tight sandstone gas reservoirs contain abundant reserves, but they are characterized by strong reservoir heterogeneity, rapid production decline, and a lack of effective enhanced recovery methods [8].

Carbon Capture, Utilization and Storage (CCUS) technology, as an important technical pathway for addressing climate change, has received widespread attention and application globally [9,10]. Among these, CO₂ Storage with Enhanced Gas Recovery (CSEGR) technology [11,12,13] integrates CO₂ geological storage with natural gas development, not only achieving long-term geological sequestration of CO₂ but also enhancing natural gas recovery through CO₂ displacement effects, demonstrating significant dual economic and environmental benefits [14,15,16,17]. Many scholars have confirmed through experimental, theoretical calculation, and numerical simulation methods that CSEGR technology has unique advantages in tight gas reservoirs [18,19,20,21,22,23,24]. In tight reservoirs, the main CO₂ storage mechanisms are structural trapping, solubility trapping, residual (capillary) trapping, and mineral trapping [25,26,27,28]. However, the CO₂ flooding process in tight gas reservoirs involves complex physicochemical processes. Research results from many scholars indicate that after CO₂ dissolves in formation water, it reacts with reservoir minerals, causing dissolution or precipitation [29,30,31], thereby affecting enhanced natural gas recovery and storage effectiveness.

Currently, the main controlling factors and influence mechanisms of CO₂ injection for enhanced gas recovery are not yet fully understood. This study uses Computer Modeling Group (v. 2022.1) software, aiming to clarify the PVT property variation patterns of the CO₂-natural gas system under gas reservoir conditions through theoretical analysis and numerical simulation methods, systematically analyze the influence patterns of injection-production parameters during the CO₂ enhanced recovery and geological storage process in tight gas reservoirs, identify the main controlling factors affecting recovery enhancement and CO₂ storage effectiveness, and provide scientific basis for the optimal design of CO₂ flooding injection-production strategies and field applications in tight gas reservoirs.

2. Methods

2.1. PR-EOS

The Peng-Robinson equation of state (PR-EOS) was proposed by Peng and Robinson [32] in 1976 and is currently widely applied in the petroleum industry. The basic form of PR-EOS is:

$$P={\displaystyle \frac{RT}{V-b}}-{\displaystyle \frac{a(T)}{V(V+b)+b(V-b)}}$$

(1)

$$a=(T,\omega )=a_0\alpha$$

(2)

The compressibility factor Z is given by:

$$Z^3-(1-B)Z^2+(A-3B^2-2B)Z-(AB-B^2-B^3)=0$$

(3)

$$A={\displaystyle \frac{aP}{R^2{T}^2}}$$

(4)

$$B={\displaystyle \frac{bP}{RT}}$$

(5)

At the critical point, a pure substance reaches the maximum point on the P-V phase diagram, and the phase envelope exhibits an inflection point at the critical point. Therefore, the critical conditions are:

$${\left({\displaystyle \frac{\partial P}{\partial v}}\right)}_{P_C,T_C}={\left({\displaystyle \frac{{\partial }^{2}P}{\partial v^2}}\right)}_{P_C,T_C}=0$$

(6)

For gas mixtures like natural gas, parameters ${\left(\partial p/\partial v\right)}_{T_c}, {({\partial }^{2}p/\partial v^2)}_{T_c}$ are non-zero. By applying the critical conditions to PR-EOS, we obtain:

$$a(T_c)=0.45724{\displaystyle \frac{R^2{T}_c^2}{P_c}}$$

(7)

$$b(T_c)=0.07780{\displaystyle \frac{R{T}_c}{P_c}}$$

(8)

$${\mathrm{Z}}_{\mathrm{c}}=0.307$$

(9)

2.2. Henry’s Law Model

Gas dissolution in the liquid phase is commonly described by Henry’s law [33]. Under the assumption that the fugacity of components in the gas phase and aqueous phase are equal, and the gas phase and aqueous phase are in thermodynamic equilibrium, we have:

$$g_i=f_{ig}-f_{iw}=0,i=1,\cdots ,n_c$$

(10)

where $f_{ig}$ and $f_{iw}$ are the fugacity of component i in the gas phase and liquid phase, respectively, and nc is the number of gas components.

For gas components soluble in the aqueous phase, the fugacity fiw is calculated using Henry’s law, i.e.,

$$f_{iw}=y_{iw}\cdot H_i$$

(11)

where $H_i$ represents the Henry constant for component i, and $y_{iw}$ denotes the mole fraction of component i in the aqueous phase.

The Henry constant depends on pressure, temperature, and salinity. CMG-GEM employs the following equation to calculate the Henry constant at various pressures:

$$\ln H_i=\ln H_i^{*}+{\displaystyle \frac{{\overline{v}}_i(p-p^{*})}{RT}}$$

(12)

where $H_i$ is the Henry constant for component i at pressure $p$ and temperature $T$; $H_i^{*}$ is the Henry constant for component i at reference pressure $p^{*}$ and temperature $T$; ${\overline{v}}_i$ is the partial molar volume of component i at infinite dilution in water, L/mol; $p^{*}$ is the reference pressure for the Henry constant of component i, $\mathrm{Pa}$; T is the absolute temperature, K; R is the gas constant, taken as 8.314 $\mathrm{J}/(\mathrm{mol}\cdot \mathrm{K})$.

Due to the salting-out effect, as the concentration of inorganic salts (NaCl, CaCl₂, etc.) in formation water increases, water molecules bind more strongly with salt ions, resulting in a reduction in free water molecules available for gas dissolution and a decrease in solubility. The salting-out coefficient in GEM is defined by the following relationship [34]:

$${\log }_{10}\left({\displaystyle \frac{H_{salt,i}}{H_i}}\right)=k_{salt,i}{m}_{salt}$$

(13)

where $H_{salt,i}$ is the Henry constant of component i in the saline solution; $H_i$ is the Henry constant of component i at zero salinity; $k_{salt,i}$ is the salting-out coefficient of component i; $m_{salt}$ is the molality of dissolved salt, $\mathrm{mol}/\mathrm{kg}{\mathrm{H}}_2\mathrm{O}$.

2.3. Geochemical Reaction Kinetics Model

Following CO₂ injection into the formation, it can react with formation water to form carbonic acid. Under acidic conditions, the rock undergoes dissolution and new precipitates may form. The primary factors governing chemical reaction rates are reactant concentration and temperature. Reaction rates between gases and aqueous solutions are typically described by the Arrhenius equation [35], which mainly accounts for the effects of temperature and reactant concentrations on reaction rates. Due to its few parameters and simple form, the rate constants can be measured in the laboratory, and, through fitting, the following model can be established:

$$r=F\exp \left[-{\displaystyle \frac{E_{\mathrm{a}}}{R}}{\displaystyle \frac{1}{T}}\right]{\displaystyle \prod _{i=1}^n{c}_i^{a_i}}$$

(14)

where $r$ denotes the reaction rate, $\mathrm{mol}/({\mathrm{m}}^2\cdot \mathrm{s})$; $r$ is the pre-exponential factor, with units determined by the overall reaction order; $r$ is the activation energy, $\mathrm{J}/\mathrm{mol}$; $c_i$ is the molal concentration of the reactant, $\mathrm{mol}/\mathrm{kg}$; $n$ is the number of reactant species and $a_i$ is the overall reaction order, defined as the sum of the exponents of the concentration terms in the rate law. A larger reaction order indicates a stronger dependence of the reaction rate on concentration.

Transition state theory (TST) is a fundamental framework in chemical kinetics for explaining reaction rates. It posits the existence of a high-energy transition state between reactants and products, and assumes that the reaction rate is determined by the frequency with which reactants traverse this state. TST refines the Arrhenius pre-exponential factor and replaces the activation energy with the Gibbs free energy of activation, thereby providing concrete physical meaning to the empirical parameters in the Arrhenius equation, as follows:

$$r=sgn\left[1-\left({\displaystyle \frac{Q}{K_{eq}}}\right)\right]\widehat{A}{S}_w\left[k_0+{\displaystyle \sum _{i=1}^{nct}{k}_i{a}_i^{wi}}\right]{\left|1-{\left({\displaystyle \frac{Q}{K_{eq}}}\right)}^{\xi }\right|}^{\varsigma }$$

(15)

$$\widehat{A}={\widehat{A}}_0\times {\displaystyle \frac{N_m}{N_{m0}}}$$

(16)

$$k_0=k_0^{*}\exp \left[-{\displaystyle \frac{E_{\mathrm{a}}}{RT}}\left({\displaystyle \frac{1}{T}}-{\displaystyle \frac{1}{T_0}}\right)\right]$$

(17)

where $Q$ denotes the activity product, i.e., the product of the activities of all reactants in the transition state; $K_{eq}$ is the chemical equilibrium constant, $\mathrm{mol}/({\mathrm{m}}^2\cdot \mathrm{s})$; when $Q$ < $K_{eq}$, the reaction proceeds in the forward direction, whereas $Q$ > $K_{eq}$ favors the reverse direction; $\widehat{A}$ and ${\widehat{A}}_0$ denote the reactive surface area of the mineral per unit volume at the current time and at the initial time, respectively, ${\mathrm{m}}^2/{\mathrm{m}}^3$; and $N_m$ and $N_{m0}$ denote the amount of the mineral at the current time and at the initial time, respectively. $S_w$ denotes the water saturation; $k_0$ denotes the standard transition-state-theory rate constants; $T_0$ is a reference value whose initial value depends on the specific reaction; $k_i$ denotes the TST reaction constants for each species involved in the liquid phase and in the equation of state during the reaction; $\xi$ and $\varsigma$ are reaction-related exponents, determined experimentally and typically taken as 1; and $T_0$ is the initial temperature.

2.4. Numerical Model Parameters

The target block produced water is characterized as CaCl₂-type with a total dissolved solids content of 42,694.42 mg/L.

Table 1. CO₂ mole fraction in formation water at different experimental pressures.

Experimental Pressure (MPa)	CO2 Mole Fraction in Aqueous Phase (%)
5	0.46139
15	0.98926
25	1.24225
35	1.40589
45	1.53514

presents the experimental data of CO₂ solubility in formation water across different pressure regimes. The fluid model Henry’s law parameters were calibrated using these experimental data. For the natural gas composition analysis, only CO₂ dissolution in water was considered. The fitted parameters yield a reference Henry’s constant of 6.21 × 10⁵ kPa and an infinite dilution molar volume of 0.03127 L/mol for CO₂.

Figure 1 illustrates the relative permeability curves for the gas-water two-phase system. The reservoir rock types in the target block are primarily lithic quartz sandstone reservoirs, with quartz sandstone as the secondary lithology. The clay mineral assemblage is dominated by illite and kaolinite, while chlorite is scarcely present. Consequently, the model incorporates five mineral phases: calcite, anorthite, orthoclase, kaolinite, and illite.

Table 2. Principal mineralization reactions considered in the model.

Reaction Classification	Reaction Equations
Phase Equilibrium Reaction	${\mathrm{CO}}_{2\left(\mathrm{g}\right)}\rightleftharpoons {\mathrm{CO}}_{2\left(\mathrm{aq}\right)}$
Chemical Equilibrium Reactions	${\mathrm{CO}}_{2\left(\mathrm{aq}\right)}\rightleftharpoons {\mathrm{HCO}}_3^{2-}{+\mathrm{H}}^{+}$
	${\mathrm{HCO}}_3^{-}\rightleftharpoons {\mathrm{CO}}_3^{2-}{+\mathrm{H}}^{+}$
	${\mathrm{CaCO}}_3{+\mathrm{H}}^{+}\rightleftharpoons {\mathrm{HCO}}_3^{-}{+\mathrm{Ca}}^{2+}$
	${\mathrm{Al}}_2{\mathrm{Si}}_2{\mathrm{O}}_5{\left(\mathrm{OH}\right)}_4{+6\mathrm{H}}^{+}\rightleftharpoons {2\mathrm{Al}}^{3+}{+5\mathrm{H}}_2{\mathrm{O}+2\mathrm{SiO}}_2$
	${\mathrm{CaAl}}_2{\mathrm{Si}}_2{\mathrm{O}}_8{+8\mathrm{H}}^{+}\rightleftharpoons {2\mathrm{Al}}^{3+}{+\mathrm{Ca}}^{2+}{+4\mathrm{H}}_2{\mathrm{O}+2\mathrm{SiO}}_2$
	${\mathrm{KAlSi}}_3{\mathrm{O}}_8{+4\mathrm{H}}^{+}\rightleftharpoons {\mathrm{Al}}^{3+}{+\mathrm{K}}^{+}{+2\mathrm{H}}_2{\mathrm{O}+3\mathrm{SiO}}_2$
	${\mathrm{KAl}}_2\left[{\left(\mathrm{SiAl}\right)}_4{\mathrm{O}}_{10}\right]{\left(\mathrm{OH}\right)}_2\cdot {\mathrm{nH}}_2{\mathrm{O}+8\mathrm{H}}^{+}=2{.3\mathrm{Al}}^{3+}{+5\mathrm{H}}_2{\mathrm{O}+6\mathrm{K}}^{+}+0{.25\mathrm{Mg}}^{2+}+3{.5\mathrm{SiO}}_2$

presents the principal mineralization reactions considered in the model, the reaction-related parameter settings can be found in Ref. [36].

To conduct further research, a five-spot well pattern mechanistic model was constructed. The model consisted of a grid with dimensions of 50 × 50 × 10 cells in the X, Y, and Z directions, respectively, with a horizontal grid spacing of 10 m and a vertical grid spacing of 2 m. The production wells are positioned at the four corners, while the CO₂ injection well is centrally located. Initial conditions include a formation pressure of 20 MPa and a water saturation of 0.3. Figure 2 illustrates the heterogeneous porosity and permeability distributions, with mean values of 0.1 and 0.77 mD, respectively, the model employs heterogeneous porosity–permeability fields that are statistically consistent with a real gas reservoir, to capture the impact of in situ heterogeneity on flow.

3. Results and Discussions

3.1. Physical Properties of CH₄, CO₂, Natural Gas, and CO₂-Natural Gas Mixtures

The composition of the natural gas used is listed in

Table 3. Natural gas composition in the study area.

Composition	Mole Fraction (%)
CO2	0.91
N2	1.41
C1	90.61
C2	5.01
C3	0.97
iC4	0.25
nC4	0.24
iC5	0.13
nC5	0.07
C6	0.40

. The reservoir gas is methane-rich (>90%), with 7% C₂–C₆ intermediate hydrocarbons and approximately 1% each of CO₂ and N₂. Using the PR-EOS, we computed the compressibility factor (Z), density, and viscosity of pure CH₄ and of the reservoir natural gas over a range of temperatures and pressures; the results are shown in Figure 3. The panels are arranged with rows (top to bottom) representing CH₄, natural gas, and their comparison, and columns (left to right) showing the compressibility factor (Z), density, and viscosity. Each subplot depicts variations with pressure across temperatures of 20~120 °C. Z quantifies deviation from ideal-gas behavior (for an ideal gas, Z = 1). Because the natural gas contains about 10% non-methane components with stronger intermolecular interactions, at the same temperature and pressure it departs more from ideality than pure CH₄, and its Z exhibits a larger dependence on pressure, indicating higher compressibility. As temperature increases, thermal motion intensifies and the relative influence of intermolecular forces diminishes; consequently, the compressibility factors of both natural gas and CH₄ increase with temperature, and their difference at a given pressure decreases.

Due to the presence of heavier C₂–C₆ components and a small amount of CO₂, the density and viscosity of natural gas are consistently higher than those of CH₄. With increasing pressure, intermolecular forces and frictional resistance strengthen, and the viscosity difference between natural gas and CH₄ becomes more pronounced.

Overall, the natural gas in the research area contains more than 90% CH₄, so its Z, density, and viscosity are generally close to those of pure methane. The remaining approximately 10% non-methane components increase pressure sensitivity making the gas more compressible and raise both density and viscosity.

Using the PR-EOS, we computed the Z, density, and viscosity of CO₂ at various temperatures and pressures and compared them with those of natural gas. The results are shown in Figure 4. The figure separately depicts, for CH₄ and for natural gas, the pressure dependence of their physical properties over 20–120 °C. As pressure increases, the Z values of both fluids exhibit a trend of initial decrease followed by increase. Higher temperatures generally result in larger Z values. The Z of CO₂ is significantly lower than that of natural gas at the same temperature and pressure. Near the critical point, CO₂ shows more dramatic changes in Z, and compared to natural gas, CO₂’s Z is more sensitive to temperature and pressure variations, deviating more from ideal gas behavior and exhibiting higher compressibility.

Comparing density and viscosity variations reveals that CO₂ properties are more sensitive to temperature and pressure changes than natural gas. Near the critical point, CO₂ properties change more dramatically. Beyond the critical pressure, significant differences exist in density and viscosity between the two fluids. Under reservoir conditions, the high density of CO₂ facilitates downward migration, reducing convective diffusion with natural gas when CO₂ is injected in the lower part of the reservoir, thereby preventing large-scale mixing at the displacement front.

Similarly, when reservoir pressure exceeds the CO₂ critical pressure, CO₂ viscosity becomes significantly higher than natural gas at the same temperature and pressure. This creates favorable mobility ratios during displacement, resulting in piston-like displacement behavior that reduces viscous fingering and enhances displacement front stability.

Using the PR-EOS, we computed the Z, density, and viscosity of CO₂ at various temperatures and pressures and compared them with those of natural gas. The results are shown in Figure 5. As pressure increases, the Z values of both fluids exhibit a trend of initial de.

Since partial mixing occurs between the CO₂ displacement front and natural gas during the displacement process, creating a transition zone at the displacement front, it is essential to study the physical property variations of CO₂-natural gas mixtures at different proportions. Phase diagrams of CO₂-natural gas mixtures with varying CO₂ concentrations (10%, 30%, 50%, 70%, 90%) were calculated using PR-EOS, as illustrated in Figure 6. As the CO₂ content in the mixture system increases, the two-phase region narrows, and the critical point gradually approaches CO₂’s critical temperature and pressure. The compressibility factor (Z), density, and viscosity of CO₂-natural gas mixtures at different proportions were computed as shown in Figure 1. With increasing CO₂ proportion, the physical properties of the mixture system progressively converge toward those of pure CO₂.

In this study, we employ the PR-EOS for phase behavior calculations to ensure computational efficiency and reproducibility; however, it does not explicitly account for confined-space effects (e.g., capillary pressure and critical-point shifts), which may introduce systematic bias in phase equilibria and saturation distributions in shale/tight porous media [37]. Future work will adopt improved EOS formulations to incorporate these effects.

3.2. Effects of CO₂ Injection Scheme on CSEGR

To compare the enhanced recovery effects of gas injection, a complete depletion development was designed as the baseline comparison case. During the first 10 years of depletion development, each production well produced at a gas rate of 2000 m³/d. In the subsequent 10 years, production continued at 1000 m³/d until abandonment pressure was reached. After the first 10 years of depletion development, the average bottomhole flowing pressure of all production wells decreased to 10 MPa, with CH₄ recovery reaching 46.63%. The pressure decline rate slowed after reducing the gas production rate and continuing depletion development for another 10 years, with the final depletion development CH₄ recovery reaching 69.92%.

The continuous CO₂ injection development scheme, based on 10 years of depletion development, involved deploying one injection well at the model center for continuous CO₂ injection over 10 years, with a total injection volume of 1 HCPV. During the gas injection development period, each production well produced at a gas rate of 3000 m³/d, with a minimum bottomhole flowing pressure limit of 7.35 MPa (ensuring CO₂ remained in supercritical state in the reservoir). Wells were shut in when the CO₂ mole fraction in the produced gas reached 20%. The continuous injection scheme achieved a 36.15% increase in CH₄ recovery over the first 10 years of depletion development, representing a 12.86% improvement compared to the complete 20-year depletion development. This result is consistent with the findings reported in Ref. [38], indicating good agreement and reliability of our conclusions. To further demonstrate the superiority of CO₂ as an injection medium for enhanced recovery, we simulated injection of an equivalent volume of N₂. N₂ readily mixes with natural gas, leading to early breakthrough of the injected gas. The continuous CO₂ injection over 20 years achieved a 4% higher recovery improvement compared to N₂ injection.

To further compare the differences between continuous and intermittent gas injection methods, the intermittent ratio was set at 1:1, meaning injection time equaled shut-in time. Since the total gas injection development period was fixed at 10 years, with a total injection volume of 1 HCPV, the number of intermittent cycles was controlled by varying the single injection duration. For convenient comparison, depletion development was designated as case 1, continuous gas injection as case 2, and intermittent cycles of 2, 5, and 10 times as cases 3, 4, and 5, respectively. Production wells maintained a gas production rate of 3000 m³/d, with a minimum bottomhole flowing pressure limit of 7.35 MPa. The shut-in condition was when the CO₂ mole fraction in produced gas reached 20%.

Figure 7 presents the final recovery rates for various scenarios. Intermittent gas injection scenarios demonstrate superior recovery enhancement over continuous gas injection. Recovery enhancement increases slightly with the number of intermittent cycles. This phenomenon occurs because extended injection periods accelerate formation energy replenishment and CO₂ migration toward production wells, leading to earlier attainment of shut-in limit conditions. The shut-in periods in intermittent injection effectively delay gas breakthrough. Increased intermittent cycles result in prolonged shut-in periods, extended production well operation time, and consequently enhanced CH₄ production.

Since the injection rate during the injection period of intermittent gas injection equals that of continuous gas injection, the total CO₂ injection mass for intermittent injection methods is substantially reduced compared to the continuous injection scenario. Owing to shut-in conditions, when all production wells are closed, the gas reservoir enters a pure CO₂ storage phase. Figure 8 illustrates the cumulative CO₂ injection volume and gas enhancement per ton of CO₂ for different injection methods. From a recovery enhancement standpoint, intermittent injection methods achieve higher recovery rates with reduced cumulative CO₂ injection, resulting in superior gas enhancement per ton of CO₂. However, from a storage perspective, continuous injection methods offer greater CO₂ storage capacity.

3.3. Effects of CO₂ Injection Rate on CSEGR

The injection rate affects both the gas injection energy replenishment effectiveness and the CO₂ front advancement velocity. To explore the influence of different injection rates during the injection phase on recovery enhancement and geological storage, various injection rate schemes were established under continuous gas injection conditions. Specifically, annual injection rates were configured at 0.02 HCPV/year, 0.04 HCPV/year, 0.06 HCPV/year, 0.08 HCPV/year, and 0.1 HCPV/year. Figure 9 illustrates the recovery enhancement for different injection rate schemes. Recovery enhancement initially increases and then decreases with increasing injection rate, with 0.06 HCPV/year achieving optimal recovery enhancement. At injection rates below 0.06 HCPV/year, insufficient total injected gas volume prevents adequate formation energy deficit replenishment and effective pressure drive system establishment. Conversely, at injection rates above 0.06 HCPV/year, excessive CO₂ front advancement velocity prematurely triggers shut-in conditions, reducing production well operation duration and consequently decreasing recovery rates.

The cumulative CO₂ injection mass and gas enhancement per ton of CO₂ for different injection rates are shown in Figure 10. Regarding gas enhancement per ton of CO₂, lower injection rates result in reduced cumulative CO₂ injection, yielding higher gas enhancement per ton of CO₂. While injection rates below 0.06 HCPV/year demonstrate higher CO₂ utilization efficiency, overall recovery enhancement remains limited. At injection rates exceeding 0.06 HCPV/year, recovery enhancement plateaus, and gas enhancement per ton of CO₂ decreases substantially. This suggests that elevated injection rates rapidly replenish formation energy and effectively enhance recovery but diminish effective gas enhancement per ton of CO₂. As injection rates exceed reasonable values under current production rates and shut-in limit conditions, gas breakthrough in production wells accelerates. From a storage perspective, increased injection rates hasten the pure storage phase arrival, with elevated formation pressure enhancing CO₂ storage capacity.

This indicates that optimal injection rates are established based on determined production rates and shut-in conditions. Changes in production rates or shut-in conditions will correspondingly alter optimal injection rates. Future research should address the reasonable formulation of injection-production schemes considering multi-parameter interactions.

3.4. Effects of Production Rate on CSEGR

Production well production rates influence gas breakthrough timing to a certain degree and directly determine recovery rate variation patterns. Various production rate schemes were established under continuous gas injection conditions. Specifically, production wells were configured to operate at gas production rates of 1000 m³/d, 2000 m³/d, 3000 m³/d, 4000 m³/d, and 5000 m³/d. Figure 11 illustrates the recovery enhancement for different production rate schemes. Recovery enhancement continuously increases with rising production rates, while production wells simultaneously reach shut-in conditions more rapidly. When production rates exceed 4000 m³/d, the recovery enhancement improvement rate diminishes.

Figure 12 presents the cumulative CO₂ injection volume and gas enhancement per ton of CO₂ for various injection rates. Gas enhancement per ton of CO₂ continuously increases with rising production rates. From a storage perspective, given unchanged injection rates, formation pressure decline rates accelerate continuously, and the mass corresponding to equivalent underground volumes of injected CO₂ continuously decreases. Further increasing production rates causes production wells to rapidly transition to constant pressure production, which offers no assistance for recovery enhancement while further reducing storage capacity. Consequently, under current injection rates, a production rate of 4000 m³/d is optimal.

3.5. Effects of Shut-In Condition on CSEGR

Shut-in condition directly determines the production duration of producers: the higher the CO₂ fraction in produced gas allowed before shut-in, the greater the corresponding CH₄ produced. However, engineering practice must also account for the treatment of high-CO₂ produced gas and associated issues such as pipeline corrosion. Therefore, a trade-off is required between maximizing recovery and corrosion control. Under continuous CO₂ injection, we configured different shut-in conditions whereby producers are shut in when the CO₂ fraction in produced gas reaches 10%, 15%, 20%, 25%, and 30%, respectively. The resulting recovery enhancement under different shut-in conditions is shown in Figure 13. As the shut-in CO₂ limit increases, production time progressively extends and recovery enhancement increases, because later shut-in leads to higher cumulative production.

The cumulative CO₂ injection mass and gas enhancement per ton of CO₂ under different shut-in conditions are shown in Figure 14. As shut-in conditions become more permissive, production time increases, the magnitude of formation pressure decline grows, and the mass corresponding to the same subsurface volume of injected CO₂ decreases. Additionally, the gas enhancement per ton of CO₂ increases gradually with increasing CO₂ mole fraction at shut-in. The influence of shut-in condition on recovery is straightforward and has significant implications for optimizing both injection and production rates.

3.6. Main Controlling Factors for CCUS-EGR

Based on the preceding single-factor analyses, we further considered interactions among parameters. Optimizing each parameter in isolation may yield only a local optimum; the optimal injection-production scheme and parameter combination may not coincide with the combination of single-factor optima. Therefore, we designed an orthogonal experiment to investigate how interactions among multiple development factors affect the performance metrics, and employed analysis of variance to assess the relative importance of each factor.

We adopted recovery enhancement, CO₂ utilization efficiency (gas enhancement per ton of CO₂), and CO₂ storage efficiency as evaluation metrics. The factors in the orthogonal design were Injection scheme, injection rate, production rate, and shut-in condition. The parameter settings for each case are summarized in

Table 4. Parameter settings for the orthogonal design.

Case ID	Injection Scheme	Injection Rate (HCPV/Year)	Production Rate (m3/day)	Shut-in Condition (%)	Gas Enhancement Per Ton of CO2 (m3/ton)	Enhanced CH4 Recovery Rate (%)	Cumulative CO2 Injection Mass (tons)
1	Intermittent	1.0	1000	15	436.15	23.22%	5.12 × 104
2	Intermittent	0.6	2000	20	533.04	35.01%	4.11 × 104
3	Continuous	0.6	4000	10	349.62	38.56%	6.16 × 104
4	Continuous	0.6	5000	30	135.51	41.53%	1.57 × 105
5	Continuous	1.0	3000	30	398.31	38.42%	5.34 × 104
6	Intermittent	0.4	5000	10	395.52	27.28%	5.25 × 104
7	Continuous	0.8	1000	20	123.26	23.18%	1.68 × 105
8	Continuous	0.4	2000	25	477.14	34.68%	4.25 × 104
9	Continuous	0.6	3000	15	665.64	39.51%	2.83 × 104
10	Continuous	0.8	5000	25	603.78	23.18%	3.09 × 104
11	Intermittent	0.6	1000	25	162.83	23.30%	1.13 × 105
12	Continuous	0.2	4000	20	621.97	24.95%	2.93 × 104
13	Continuous	0.2	5000	15	83.87	24.92%	1.89 × 105
14	Continuous	0.4	3000	20	978.14	33.87%	1.51 × 104
15	Intermittent	0.8	4000	30	983.75	40.69%	1.49 × 104
16	Continuous	0.2	1000	10	1085.95	23.26%	1.24 × 104
17	Intermittent	1.0	5000	20	1089.74	40.01%	1.23 × 104
18	Intermittent	0.8	3000	10	322.27	37.74%	3.89 × 104
19	Continuous	1.0	2000	10	373.31	29.45%	3.36 × 104
20	Continuous	0.4	1000	30	828.65	23.30%	1.51 × 104
21	Intermittent	0.2	2000	30	148.37	22.46%	8.42 × 104
22	Continuous	1.0	4000	25	92.46	39.54%	1.35 × 105
23	Intermittent	0.4	4000	15	92.46	27.37%	1.35 × 105
24	Intermittent	0.2	3000	25	1744.69	22.31%	6.93 × 103
25	Continuous	0.8	2000	15	1798.75	34.25%	6.67 × 103

.

Based on the analysis of variance results for recovery enhancement, the factors in order of decreasing significance are injection rate > production rate > Injection scheme > shut-in condition. Injection rate directly governs energy replenishment and determines whether the displacement pressure system can satisfy the daily gas offtake requirement. Production rate is secondary, as it controls the produced gas volume, while Injection scheme and shut-in condition are relatively less influential than injection and production rates.

For gas enhancement per ton of CO₂, the order is: injection rate > Injection scheme > production rate > shut-in condition. Injection rate is paramount because it directly affects CO₂ utilization efficiency. Injection scheme ranks second: with intermittent injection, shut-in periods allow production to be sustained by the pressure gradient established during injection, yielding higher CO₂ utilization than continuous injection. Production rate and the shut-in condition are comparatively least significant.

For cumulative CO₂ injection mass, the order is: injection rate > Injection scheme > production rate > shut-in condition. The injection rate directly determines the amount of CO₂ injected over the development period, hence the highest significance. Injection scheme is next because intermittent injection includes shut-in periods that materially affect the total injected volume. Production rate and the shut-in condition have the lowest significance for total injected CO₂.

Different schemes emphasize different objectives; there is no single “perfect” scheme that both minimizes CO₂ injection and maximizes recovery enhancement. If recovery enhancement is the sole objective, Case 4 is optimal. However, when the economic cost of CO₂ injection is considered, Case 15 is clearly more economical: recovery decreases by only 0.85%, while injected CO₂ is reduced by 19.7%. For Case 15, with an incremental gas production of 983.75 m³/ton CO₂ injected, a CO₂ injection cost of 180 ¥/ton (based on Ref. [39]), and a natural gas price of 1.575 ¥/m³, the net economic benefit per ton of CO₂ for enhanced recovery reaches 1369.41 ¥, excluding storage subsidies. We plot an X–Y scatter of cumulative injected CO₂ versus recovery enhancement for all orthogonally designed cases (Figure 15) and identify the Pareto front (the set of non-dominated solutions), thereby provide field-oriented parameter optimization guidance and implementable operating windows, which can directly inform option selection and injection–production parameter settings for the target block.

Considering the cost of gas injection, Case 15, which achieves higher gas enhancement per ton of CO₂, was selected for further analysis of CO₂ storage mechanisms in the gas reservoir. CO₂ geological storage mechanisms primarily include structural trapping, residual trapping, solubility trapping, and mineral trapping. Case 15 maintains production wells above the CO₂ critical pressure, ensuring all CO₂ in the reservoir remains in the supercritical state.

CO₂ dissolution in the aqueous phase refers to CO₂ gas entering and dispersing within the aqueous phase. When CO₂ contacts water, some CO₂ molecules enter the aqueous phase, forming a CO₂ aqueous solution. At this point, CO₂ exists in molecular form within the water. Beyond molecular existence, CO₂ also reacts chemically with water to form carbonic acid (H₂CO₃), which further ionizes to produce bicarbonate ions (HCO₃⁻) and carbonate ions (CO₃²⁻). Therefore, CO₂ dissolution in the aqueous phase includes both molecular CO₂ and the ionic forms (HCO₃⁻and CO₃²⁻) formed after CO₂ dissolves in formation water. Mineral trapping refers to CO₂ existing in the form of mineral precipitation.

The evolution curves of different CO₂ storage mechanisms for Case 15 are shown in Figure 16, it can be observed that as gas injection progresses, supercritical, ionic, and dissolved CO₂ continuously increase, while mineral CO₂ continuously decreases. This indicates that during the injection period, storage mechanisms are dominated by structural and residual trapping. Since Case 15 is an intermittent injection scheme, storage mechanism transformations occur during shut-in periods: some supercritical CO₂ continuously dissolves into formation water, with slight increases in ionic and dissolved CO₂ during the shut-in phase. Throughout the entire injection period, the amount of CO₂ stored in mineral form continuously decreases, indicating that mineral dissolution predominates.

Mineralization reactions primarily occur within the CO₂ swept zone, as illustrated by calcite in Figure 17. The changes in mineral quantities in the formation are shown in Figure 18. As injection progresses, calcite dissolution dominates the mineral changes in the formation, anorthite is continuously dissolved, orthoclase initially shows slight precipitation followed by continuous dissolution, while kaolinite and illite predominantly precipitate. The evolution of different CO₂ storage mechanisms indicates that during the CO₂ injection development phase, due to the short time scale, storage mechanisms are dominated by structural and residual trapping, while minerals primarily undergo dissolution. Achieving more stable CO₂ storage in mineral form requires a longer time scale.

4. Conclusions

This study utilized the PR-EOS to investigate the physical properties of CH₄, CO₂, and natural gas. A numerical simulation model for tight gas reservoirs was established, considering CO₂ dissolution in formation water and CO₂-water-rock geochemical reactions. Based on numerical simulation methods, we explored the influence patterns of injection scheme, injection rate, production rate, and shut-in condition on CO₂ enhanced recovery and storage effectiveness. An orthogonal design was conducted to investigate the main controlling factors affecting CO₂ enhanced gas recovery and storage effectiveness under multi-factor interactions. The main conclusions are as follows:

• Under reservoir temperature and pressure conditions, CO₂ exhibits significant differences in compressibility factor, density, and viscosity compared to natural gas. The higher density enables bottom injection to avoid large-scale mixing at the displacement front, while viscosity differences form favorable mobility ratios for piston-like displacement and improved front stability.
• Intermittent injection slightly outperforms continuous injection for recovery enhancement, with shut-in periods delaying CO₂ breakthrough and improving utilization efficiency. However, continuous injection stores more CO₂ mass during development. Injection rate affects energy replenishment and front advancement velocity, while production rates must balance recovery enhancement with formation energy constraints and corrosion prevention.
• For recovery enhancement effectiveness, factor significance ranking is injection rate > production rate > injection scheme > shut-in condition. For CO₂ storage effectiveness, the ranking is: injection rate > injection scheme > production rate > shut-in condition. Injection rate directly determines energy replenishment and total CO₂ mass injected, while injection scheme significantly affects total injection volume due to shut-in periods.
• During gas injection, supercritical, ionic, and dissolved CO₂ continuously increase while mineral CO₂ decreases. Storage mechanisms are dominated by structural and residual trapping, with supercritical CO₂ dissolving into formation water. Mineral changes are characterized by calcite dissolution, anorthite dissolution, orthoclase precipitation followed by dissolution, and kaolinite/illite precipitation.