Sensitivity Analysis of Geological–Engineering Parameters and Injection Optimization for CO₂-ECBM in Coal Seams Based on Numerical Simulation

Abstract

CO₂-enhanced coalbed methane recovery and storage (CO₂-ECBM) is a promising approach for improving methane recovery and increasing CO₂ storage in low-permeability coal seams. However, limited injectivity and insufficient criteria for injection parameter optimization remain major constraints. Taking the No. 11-2 coal seam of the Zhangji Coal Mine in the Huainan mining area as the study object, this study established a thermo–hydro–mechanical coupled model that considers CO₂/CH₄ competitive adsorption, matrix diffusion, fracture seepage, gas–water two-phase flow, coal deformation, and porosity–permeability evolution. A 10-year numerical simulation was conducted to evaluate the effects of initial porosity, initial permeability, elastic modulus, CO₂ injection pressure, and injection scheme on CO₂-ECBM performance. The comprehensive sensitivity results show that initial porosity, CO₂ injection pressure, and initial permeability are the dominant controlling factors, whereas elastic modulus has a relatively weak influence. Initial porosity mainly determines reservoir storage space and CO₂ sequestration potential; permeability controls pressure propagation and gas migration; and injection pressure directly affects CH₄ displacement intensity, CO₂ storage capacity, and reservoir safety margin. Multi-objective evaluation indicates that the injection pressure should be controlled within 8.0–9.0 MPa, with 8.0–8.5 MPa recommended for long-term stable operation. When the engineering objective prioritizes CO₂ storage or CH₄ recovery and sufficient safety margin is confirmed, the injection pressure may be increased to approximately 9.0 MPa. Continuous constant-pressure injection favors cumulative CH₄ production and CO₂ storage, whereas stepwise pressurization reduces early pressure disturbance and improves later-stage injectivity. Therefore, an injection strategy combining early-stage stepwise pressurization with middle- and late-stage constant-pressure injection is recommended. These results provide a reference for injection parameter optimization in similar low-permeability coal reservoirs.

1. Introduction

Under the background of carbon-emission reduction and low-carbon coal utilization, CO₂-enhanced coalbed methane recovery and storage technology (CO₂-ECBM) has received increasing attention as a pathway that links geological CO₂ storage with coalbed methane recovery [1,2,3]. Compared with storage in deep saline aquifers and depleted oil and gas reservoirs [4,5], CO₂ storage in coal seams is unique in that it involves not only CO₂ injection and retention but also CH₄ desorption, diffusion, seepage, and production [6,7]. As a result, coal seams exhibit more complex reservoir responses and require more intensive engineering regulation. In particular, typical coal reservoirs in China, represented by those in the Huainan mining area, are commonly characterized by “three high, three soft, and one low” [8,9]. These characteristics significantly restrict CO₂ injectivity and the enhancement of coalbed methane production in low-permeability coal seams, making them a major bottleneck for the large-scale application of CO₂-ECBM [10]. Accordingly, optimizing CO₂ injection parameters to overcome the engineering constraints of low-permeability coal reservoirs and thereby improve CO₂-ECBM performance has become a key research focus.

Previous studies on injection parameter optimization for CO₂-ECBM have extensively investigated injection pressure, reservoir properties, gas composition, and injection schemes. For injection pressure, Yin et al. [11] examined the effects of different CO₂ injection pressures on CH₄ production and CO₂ breakthrough through coal sample displacement experiments and a dual-porosity numerical model. Their results indicated that higher injection pressure could enhance the early-stage CH₄ flow rate and production rate, but may also lead to earlier CO₂ breakthrough, thereby affecting cumulative gas production at later stages. Ma et al. [12] developed a thermal–hydrological–mechanical coupled model for CO₂-ECBM and validated its reliability by comparing TOUGH-FLAC and COMSOL 6.2 simulations, demonstrating that coal deformation, permeability evolution, and injection pressure jointly control the engineering response of CO₂-ECBM. Hou et al. [13] further constructed a THM coupled model using COMSOL Multiphysics to analyze the effects of CO₂ injection temperature and pressure on CO₂-ECBM, showing that variations in injection pressure significantly affect the CH₄ displacement process and CO₂ sequestration performance. Field pilot tests in the Allison Unit and Fruitland Coal Fairway of the San Juan Basin, USA, also showed that CO₂ injection can enhance coalbed methane production, although injectivity is often constrained by coal matrix swelling, pressure limitations, and permeability reduction [14]. Records from NETL further indicated that the actual amount of injected CO₂ was lower than expected and that production enhancement was limited [15], suggesting that simply increasing injection pressure at the field scale cannot guarantee satisfactory engineering performance.

Regarding injection schemes and multi-parameter optimization, recent studies have begun to focus on unconventional injection strategies and the synergistic effects of multiple factors. Bai et al. [16] proposed a stepwise CO₂ injection pressure method and compared it with constant-pressure CO₂ injection, showing that stepwise pressure-increasing injection could improve CH₄ recovery to some extent and reduce the amount of CO₂ injected per unit coal mass. Fan et al. [17] optimized CO₂-ECBM from the perspectives of injection timing and gas mixture proportion, indicating that both CO₂ geological sequestration and CH₄ recovery are controlled by injection timing and injected gas composition. In addition, reservoir permeability and its heterogeneity have also been recognized as important factors affecting CO₂-ECBM performance. Kumar et al. [18] demonstrated that initial permeability heterogeneity influences CO₂ migration, CH₄ production, and permeability evolution, while Li et al. [19] revealed that coal adsorption swelling, effective stress variation, and fracture structure adjustment during CO₂ injection jointly lead to dynamic permeability changes. Furthermore, Fan et al. [20] established a THM model for ECBM using CO₂/N₂ mixtures and optimized the injected gas composition and injection scheme, suggesting that an appropriate CO₂/N₂ ratio could help mitigate coal matrix swelling and injectivity decline caused by pure CO₂ injection. Sayyafzadeh et al. [21] employed a parallel genetic algorithm to optimize gas mixture injection schemes, providing a methodological reference for intelligent optimization of ECBM parameters.

Overall, existing studies have identified the key controlling factors of CO₂-ECBM from the perspectives of injection pressure, gas composition, injection timing, permeability evolution, and injection schemes. Nevertheless, several limitations remain. On the one hand, many studies have focused mainly on a single parameter or a single mechanism, while systematic comparisons of the relative importance of geological and engineering parameters remain insufficient. On the other hand, most existing optimization studies have primarily used cumulative CH₄ production or cumulative CO₂ sequestration as the main evaluation indicators, with limited consideration of engineering efficiency, stable injection capacity, and reservoir safety risk. Li et al. [22] pointed out that low injectivity and CO₂ injection-induced injectivity decline are major technical challenges facing the engineering application of CO₂-ECBM. This further indicates the necessity of establishing a comprehensive evaluation framework that integrates geological adaptability, engineering feasibility, and safety controllability.

Based on the above considerations, this study takes the No. 11-2 coal seam of the Zhangji Coal Mine in the Huainan mining area as the research object and establishes a coupled numerical model for CO₂-ECBM. The effects of initial porosity, initial permeability, elastic modulus, CO₂ injection pressure, and injection scheme on CH₄ production, CO₂ storage, injection capacity, and reservoir safety are systematically investigated. Different from studies that mainly evaluate CO₂-ECBM performance using single indicators, such as cumulative CH₄ production or cumulative CO₂ storage, this study further introduces stable injection capacity and reservoir safety risk into the engineering evaluation framework. On this basis, the relative importance of geological and engineering parameters is identified, and the injection pressure range and injection strategy suitable for the target low-permeability coal seam are proposed. The results can provide a reference for reservoir screening, injection parameter selection, and injection scheme optimization in similar low-permeability coal reservoirs.

2. Mathematical Model of Multiphysical Field Coupling for CO₂-ECBM

Based on the gas–water migration, gas competitive adsorption, coal structure evolution, and energy transfer in the reservoir, the following assumptions are made [23,24]: (1) the coal reservoir is represented by a single-permeability, dual-porosity elastic medium comprising matrix and fracture systems, and both media are assumed to be isotropic and homogeneous; (2) CH₄/CO₂ competitive adsorption–desorption in the matrix is described by the extended Langmuir model, and component diffusion from matrix to fractures follows Fick’s first law; (3) gas and water mainly flow in fractures and their flow satisfies Darcy’s law; and (4) coal deformation follows the small deformation assumption, and variations in gas pressure and adsorption–desorption cause volumetric strain changes in the reservoir. Figure 1 shows the gas–water migration process during CO₂-ECBM.

Figure 1. Schematic illustration of the gas–water migration process during CO₂-ECBM [23].

2.1. Seepage Field Equations

(1) Gas transport equation in the matrix

During CO₂-ECBM, CH₄ adsorbed in the coal matrix is released and diffuses toward fractures driven by the concentration difference, while injected CO₂ moves from fractures into matrix pores and becomes adsorbed on matrix surfaces. The matrix component-balance equation includes gas storage, adsorption/desorption variation, and matrix–fracture mass exchange [23,24,25]:

$${\displaystyle \frac{\partial }{\partial t}}\left\{{\displaystyle \frac{V_{L\mathrm{i}}{P}_{mi}/P_{Li}}{1+\frac{P_{mi}}{P_{Li}}+\frac{P_{mi}}{P_{Li}}}}\exp \left[-{\displaystyle \frac{d_1}{1+d_2{P}_m}}\left(T-T_r\right)\right]{\rho }_s{\rho }_{si}+{\varphi }_m{\displaystyle \frac{M_i}{RT}}{P}_{mi}\right\}=-{\displaystyle \frac{3{\pi }^2{D}_i}{n^2}}{\displaystyle \frac{M_i}{RT}}\left(P_{mi}-P_{fi}\right)$$

(1)

where the subscript i identifies the gas species, in which i = 1 corresponds to CH₄ and i = 2 corresponds to CO₂. M_gi denotes the molar mass of species i (g/mol); R is the universal gas constant (J/(mol·K)); and T is the gas temperature (K). The matrix porosity is represented by φ_m. ρ_i and ρ_s refer to the density of gas component i and the coal skeleton, respectively (kg/m³), while ρ_si is the density of component i under standard conditions (kg/m³). V_Li and P_Li are the Langmuir volume and Langmuir pressure constants of component i, respectively, with units of m³/t and MPa. d₁ and d₂ denote the pressure-related and temperature-related coefficients, respectively. P_mi is the partial pressure of component i in the coal matrix, and P_fi is the corresponding gas pressure in the fracture system. P_m represents the total gas pressure in the matrix, defined as the sum of all gas component partial pressures. T_ref is the reference temperature used for adsorption measurement (K). D_i is the diffusion coefficient of gas species i (m²/s), and n denotes the cleat spacing (m).

(2) Fluid transport equations in fractures

In fractures, gas and water coexist and migrate as a two-phase gas–water system. Desorbed CH₄ from the matrix acts as a source term for fracture CH₄ transport, whereas CO₂ transferred from fractures into the matrix provides a sink term associated with adsorption. Accordingly, the fracture mass-balance equations for gas and water are expressed as [25,26,27]:

$$\left\{\begin{array}{l}{\displaystyle \frac{\partial \left(S_{\mathrm{g}}{\varphi }_f{\rho }_{fi}\right)}{\partial t}}+\nabla \cdot \left({\rho }_{fi}{q}_i\right)+{\displaystyle \frac{\partial \left(S_{\mathrm{w}}{\varphi }_f{\rho }_{fdi}\right)}{\partial t}}+\nabla \cdot \left({\rho }_{fdi}{q}_w\right)=Q_{gi}\\ {\displaystyle \frac{\partial \left(S_w{\varphi }_f{\rho }_w\right)}{\partial t}}+\nabla \cdot \left({\rho }_w{q}_w\right)=Q_w\end{array}\right.$$

(2)

where S_g and S_w denote the gas saturation and water saturation, respectively, with S_g + S_w = 1. φ_f represents the fracture porosity. ρ_fi and ρ_w are the densities of the gas component i and water in the fracture system, respectively (kg/m³). q_gi and q_w denote the flow velocities of the gas component i and water, respectively (m/s). ρ_dgi is the density of the gas component i dissolved in water (kg/m³), and H_gi is Henry’s law coefficient of component i; in the present model, only CO₂ dissolution is considered. Q_gi and Q_w represent the source–sink terms of the gas component i and water, respectively. Gases and water are coupled through capillary pressure in fracture, and the change in relative permeability is determined based on saturation [25]:

$$\left\{\begin{array}{l}{P}_w=P_f-P_c\\ P_c=P_e{\left({\displaystyle \frac{S_w-S_{wr}}{1-S_{gr}-S_{wr}}}\right)}^{-1/2}\\ k_{rg}={\left[1-\left({\displaystyle \frac{S_w-S_{wr}}{1-S_{wr}}}\right)\right]}^2\left[1-{\left({\displaystyle \frac{S_w-S_{wr}}{1-S_{wr}}}\right)}^2\right]\\ k_{rw}={\left({\displaystyle \frac{S_w-S_{wr}}{1-S_{wr}}}\right)}^4\end{array}\right.$$

(3)

where P_c denotes the capillary pressure (MPa) and P_e represents the entry capillary pressure (MPa). S_gr and S_wr correspond to the residual gas saturation and irreducible water saturation, respectively. k_rg and k_rw are the relative permeabilities of the gas phase and water phase, respectively.

2.2. Stress Field Equation

Coal reservoir deformation during the CO₂-ECBM process is jointly controlled by effective stress, pore-fluid pressure, gas adsorption/desorption, and temperature. Based on the elastic theory of porous media, the governing equation for the stress field of the coal reservoir can be expressed as follows [26,27,28]:

$$G{u}_{i,jj}+{\displaystyle \frac{G}{1-2\nu }}{u}_{j,ji}+F_{,i}={\alpha }_m{P}_{m,i}+{\alpha }_f{P}_{f,i}+K{\alpha }_T{T}_{,i}+K{\epsilon }_{s,i}$$

(4)

where i and j denote the x-, y-, and z-directions in the three-dimensional coordinate system. u represents the displacement component (m), and F denotes the body-force term. G and D are the shear modulus and effective elastic modulus, respectively (MPa), while ν is Poisson’s ratio. K_n represents the normal stiffness of the fracture (GPa/m), and E is the elastic modulus (MPa). K and K_s denote the bulk modulus and skeletal bulk modulus, respectively (MPa), whereas E_s is the skeletal elastic modulus (MPa). α_m and α_f are the Biot effective stress coefficients for the coal matrix and fracture system, respectively. P_f represents the gas–water mixture pressure within the fracture (MPa), α_T is the thermal expansion coefficient (K⁻¹), and ε_s denotes the volumetric strain caused by matrix swelling or shrinkage induced by gas adsorption and desorption.

2.3. Temperature Field Equation

During CO₂-ECBM, changes in reservoir energy are mainly governed by adsorption/desorption heat, elastic deformation of the coal rock, heat conduction, convective heat transfer, and gas dissolution. According to the principle of energy conservation, the governing equation for the temperature field can be written as follows [26,27,28]:

$${\displaystyle \frac{\partial \left[{\left(\rho C_p\right)}_{eff}T\right]}{\partial t}}+{\eta }_{eff}\nabla T-\nabla \cdot \left({\lambda }_{eff}\nabla T\right)=K{\alpha }_{T}T{\displaystyle \frac{\partial {\epsilon }_V}{\partial t}}+{\displaystyle {\displaystyle \sum _{i=1}^2}\frac{e_{gi}}{M_i}}{\displaystyle \frac{\partial {\rho }_{dgi}}{\partial t}}+{\displaystyle {\displaystyle \sum _{i=1}^2}{q}_{sti}}{\displaystyle \frac{{\rho }_s{\rho }_{si}}{M_i}}{\displaystyle \frac{\partial V_{si}}{\partial t}}$$

(5)

where (ρC_P)_eff denotes the effective volumetric heat capacity of the coal reservoir containing gas and water (J/(m³·K)). η_eff represents the effective convective heat-transfer coefficient (J/(m²·s)), and ε_v is the volumetric strain of coal. λ_eff denotes the effective thermal conductivity of the reservoir (W/(m·K)). q_sti is the heat released or absorbed during adsorption of component i (J/mol), whereas e_gi represents the heat associated with dissolution of component i in water (J/mol). (ρC_P)_eff, η_eff, and λ_eff can be written as [29]:

$$\left\{\begin{array}{l}{\left(\rho C_p\right)}_{eff}=\left(1-{\varphi }_f-{\varphi }_m\right){\rho }_s{C}_s+S_w{\varphi }_f{\rho }_w{C}_w+\\ {\displaystyle \sum _{i=1}^2\left(S_g{\varphi }_f{\rho }_{fi}+{\varphi }_m{\rho }_{mi}+S_w{\varphi }_f{\rho }_{fdi}\right)}{C}_i\\ {\eta }_{eff}=-{\displaystyle \sum _{i=1}^2\left({\rho }_{fi}\frac{k{k}_{rg}{C}_i}{{\mu }_i}\left(1+\frac{{\mathrm{b}}_k}{P_{fi}}\right)\nabla P_{fi}+{\rho }_{fdi}\frac{k{k}_{rw}{C}_i}{{\mu }_i}\nabla P_{fw}\right)}\\ -{\rho }_w{\displaystyle \frac{k{k}_{rw}{C}_w}{{\mu }_w}}\nabla P_{fw}\\ {\lambda }_{eff}=\left(1-{\varphi }_f-{\varphi }_m\right){\lambda }_s+{\varphi }_m{\lambda }_{mgm}+{\varphi }_f\left(S_g{\lambda }_{fgm}+S_w{\lambda }_{fw}\right)\end{array}\right.$$

(6)

where C_s, C_gi, and C_w denote the specific heat capacities of coal, the gas component i, and water, respectively (J/(kg·K)). λ_s, λ_mgm, λ_fgm, and λ_fw represent the thermal conductivity coefficients of coal, mixed gas in the matrix, mixed gas in the fracture network, and water, respectively (W/(m·K)).

2.4. Porosity and Permeability Equations

Matrix and fracture porosities are calculated using the following expressions [23,26]:

$$\left\{\begin{array}{l}{\varphi }_m={\displaystyle \frac{1}{1+S}}\left[{\varphi }_{m0}\left(1+S_0\right)+{\alpha }_m\left(S-S_0\right)\right]\\ {\varphi }_f={\varphi }_{f0}+{\displaystyle \frac{{\varphi }_{f0}{K}_{\mathrm{m}}}{3K_f}}\left({\alpha }_T\Delta T+\Delta {\epsilon }_s+\Delta {\epsilon }_V+{\displaystyle \frac{{\alpha }_m\Delta p_m}{K_m}}\right)\\ S={\epsilon }_V+{\displaystyle \frac{P_m}{K_s}}-{\alpha }_{T}T-{\epsilon }_s\\ S_0={\epsilon }_{V0}+{\displaystyle \frac{P_{m0}}{K_s}}-{\alpha }_T{T}_0-{\epsilon }_{s0}\end{array}\right.$$

(7)

Based on the cubic relationship between porosity and permeability in coal fractures, the matrix and fracture permeabilities are updated as [23,26]:

$$\left\{\begin{array}{l}{k}_m=k_{m0}{\left({\displaystyle \frac{{\varphi }_m}{{\varphi }_{m0}}}\right)}^3\\ k_f=k_{f0}{\left({\displaystyle \frac{{\varphi }_f}{{\varphi }_{f0}}}\right)}^3\end{array}\right.$$

(8)

In summary, Equations (1)–(8) form the thermo–hydro–mechanical (THM) coupled mathematical model used for CO₂-ECBM simulation in coal seams. The interactions among the physical fields are summarized in Figure 2.

Figure 2. Schematic diagram of the thermo–hydro–mechanical (THM) coupling relationships in CO₂-ECBM multiphysical fields.

2.5. Model Validation

Model validation was conducted by reproducing the simulation settings reported in Ref. [30] using the governing equations of Equations (1)–(8). It should be noted that the purpose of this comparison was to perform a benchmark validation of the numerical implementation and reliability of the governing equations rather than to directly validate the geological and engineering conditions of the Zhangji Coal Mine. Therefore, the simulation parameters, operational scheme, boundary conditions, and simulation period in the validation case were kept consistent with those reported in Ref. [30]. CH₄ production rate and CO₂ injection rate under different injection pressures were selected as validation indicators, and the simulation period was 3650 days.

As shown in Figure 3, the simulation results obtained in this study are generally consistent with those reported by Fan et al. [30]. To further quantify the validation accuracy, representative data points were extracted from the published curves in Ref. [30], and the relative errors between the present simulation results and the reference results were calculated. The mean relative error of the CH₄ production rate is approximately 6–8%, while that of the CO₂ injection rate is approximately 3–5%. In addition, the simulated curves exhibit consistent temporal variation trends with the reference results. Both the CH₄ production rate and CO₂ injection rate decrease rapidly during the early stage and then gradually approach a stable state, and their magnitudes remain at the same order of 10³ m³/day. The small deviations are mainly related to the additional consideration of temperature effects and CO₂ dissolution in the present model. Therefore, the benchmark comparison indicates that the present model can reasonably capture the main dynamic characteristics of CO₂-ECBM.

Figure 3. Comparison between the simulation results of this study and those reported in Ref. [30]. (a) CH₄ production rate; (b) CO₂ injection rate.

3. Numerical Simulation Study of CO₂-ECBM in Coal Seams

3.1. Physical Model and Key Parameters

3.1.1. Physical Model

By taking the No. 11-2 coal seam, the main mining coal seam of the Zhangji Coal Mine in the Huainan Coalfield, as the research object (Figure 4) and drawing on domestic CO₂-ECBM pilot test projects, a single-well gas injection and multi-well gas production mode is adopted [23,24], with four CH₄ production wells arranged around the CO₂ injection well within a range of approximately 400 m for coalbed methane extraction. Considering the limitations of computational speed and memory, a quarter region (the lower-right corner of the CO₂-ECBM layout) is selected for numerical simulation based on the symmetry principle. The actual simulated dimensions are approximately 200 m × 200 m × 2.64 m (Figure 5). In the model, the injection well (IW) is located in the lower-left corner, and the production well (PW-2) is located in the upper-right corner. The average coal seam thickness is 2.64 m, and the wellbore radius is 0.01 m. The simulation duration is set to 3650 days (approximately 10 years).

Figure 4. Location of the study area.

Figure 5. Schematic of the injection–production well pattern in the CO₂-ECBM simulation. (a) Planar layout; (b) 3D model of numerical simulation.

3.1.2. Key Parameters

The core parameters for this numerical simulation (Table 1) are derived from three main sources: (1) actual engineering data from the No. 11-2 coal seam and coalbed methane well development at the Zhangji Coal Mine, such as reservoir pressure and temperature; (2) experimental test data from coal rock samples of the No. 11-2 coal seam, including porosity, permeability, and elastic modulus; and (3) relevant references [23,24,25,26,27,28,29], providing parameter values such as gas dynamic viscosity and the Klinkenberg factor.

Table 1. Core parameters for numerical simulation.

Variable	Value	Units	References	Variable	Value	Units	References
ρc	1.48 × 103	kg·m−3	Lab measurement	ρs1	0.717	kg·m−3	[23]
ϕm	0.040	-	Lab measurement	ρs2	1.977	kg·m−3	[23]
ϕf	0.025	-	Lab measurement	k0	3.5 × 10−16	m2	Lab measurement
μ1	1.03 × 10−5	Pa·s	[24]	VL1	17.6	m3·t−1	Lab measurement
μ2	1.84 × 10−5	Pa·s	[26]	PL1	3.13	MPa	Lab measurement
μw	1.01 × 10−3	Pa·s	[26]	VL2	21.5	m3·t−1	Lab measurement
eg2	14.12	kJ·mol−1	[23]	PL2	1.92	MPa	Lab measurement
qst1	33.4	kJ·mol−1	[28]	εL1	0.0128	-	[26]
qst2	19.2	kJ·mol−1	[23]	εL2	0.0237	-	[26]
T0	308	K	Site data	sw0	0.7	-	Lab measurement
E	3.14	GPa	Lab measurement	swr	0.05	-	[25]
Es	6.33	GPa	Lab measurement	sgr	0.05	-	[23]
Kn	2.86	GPa·m−1	[27]	bk	0.75	MPa	[25]
ν	0.30	-	Lab measurement	Pfg20	0.1	MPa	Site data
Cs	1550	J·kg−1·K−1	[27]	Pmg20	0.1	MPa	Site data
Cw	4350	J·kg−1·K−1	[27]	Pfg10	4.83	MPa	Site data
Cg1	2260	J·kg−1·K−1	[27]	Pmg10	4.83	MPa	Site data
Cg2	844	J·kg−1·K−1	[26]	τ1	0.221	d	[26]
αT	2.4 × 10−5	K−1	[26]	τ2	0.334	d	[26]
d1	2.30 × 10−2	K−1	[23]	λs	0.1913	W·m−1·K−1	[26]
d2	7.30 × 10−2	MPa−1	[23]	λg1	0.031	W·m−1·K−1	[28]
Tref	300	K	[23]	λg2	0.015	W·m−1·K−1	[26]
Hg2	0.0347	-	[29]	λw	0.652	W·m−1·K−1	[26]

3.2. Boundary Conditions

Seepage field: The initial reservoir pressure was set to 5.79 MPa and was assumed to be uniformly distributed in the simulated domain. A constant-pressure boundary was applied at the injection well, and the production well pressure was set to 0.1 MPa. The remaining external boundaries were treated as no-flow boundaries. This setting was adopted because the simulated domain represents a quarter symmetric region of the injection–production well pattern. Therefore, the no-flow boundaries were used to represent the symmetry condition and to focus on pressure propagation and gas migration between the injection well and the production well within the local drainage area.

Stress field: The initial stress state was simplified as a geostatic stress condition mainly controlled by the overburden load. A vertical downward load was applied on the upper boundary to represent the mechanical effect of the overlying strata. The lower boundary was fixed to represent the support from the underlying strata, and the lateral boundaries were set as roller-supported boundaries to provide lateral constraints while allowing vertical deformation. This mechanical boundary condition can reasonably describe the deformation response of a local coal seam section under reservoir-scale stress constraints.

Temperature field: The initial reservoir temperature was set to 308 K and was assumed to be uniformly distributed. The upper, lower, and lateral boundaries were treated as thermally insulated boundaries. The injection and production wells were assigned the initial reservoir temperature.

These boundary conditions are simplified representations of the actual reservoir conditions and are suitable for comparing the relative effects of different geological and engineering parameters under the same numerical framework. However, if strong regional hydraulic communication, spatial heterogeneity, or external heat exchange exists in the actual reservoir, the no-flow and thermal insulation assumptions may influence the predicted CO₂ migration range, pressure propagation, and production response.

4. Sensitivity Analysis of Geological and Engineering Parameters for CO₂-ECBM

4.1. Influence of Geological Parameters on CO₂-ECBM

4.1.1. Influence of Initial Reservoir Porosity on CO₂-ECBM

Figure 6 presents the variation trends of CH₄ production rate, cumulative CH₄ production, CO₂ injection rate, and cumulative CO₂ storage under natural production and different initial porosity conditions. Initial porosity has a significant effect on both CH₄ enhancement and CO₂ storage. After CO₂ injection, the CH₄ production rate under all porosity conditions is markedly higher than that under natural production (536 m³/d), indicating that CO₂ injection effectively promotes CH₄ desorption and production. When the initial porosity is 0.025, 0.035, and 0.045, the stable CH₄ production rates are 1232, 1478, and 1601 m³/d, respectively, which are 129%, 175%, and 198% higher than those of natural production. The cumulative CH₄ production reaches 773,899 m³, 1,006,069 m³, and 1,160,849 m³, showing a continuous increase with increasing porosity. This indicates that a higher porosity provides greater storage space and more favorable gas migration conditions, thereby enhancing the displacement effect of CO₂ on CH₄.

Figure 6. Under different initial porosities: (a) CH₄ production rate; (b) cumulative CH₄ production; (c) CO₂ injection rate; (d) cumulative CO₂ storage.

4.1.2. Influence of Initial Reservoir Permeability on CO₂-ECBM

Initial permeability has a significant effect on CO₂ injectability, CH₄ enhancement, and CO₂ storage (Figure 7). When the initial permeability is 3.5 × 10⁻¹⁶, 6.5 × 10⁻¹⁶, and 10.0 × 10⁻¹⁶ m², the stable CH₄ production rates are 1232, 1694, and 1572 m³/d, respectively, which are 129%, 216%, and 193% higher than those under natural production. The cumulative CH₄ production reaches 773,899 m³, 1,161,094 m³, and 2,168,075 m³, increasing significantly with permeability, which indicates that a higher permeability favors higher pressure transmission efficiency and gas migration capacity, thereby enhancing the displacement effect of CO₂ on CH₄. However, the stable CH₄ production rate is higher at k₀ = 6.5 × 10⁻¹⁶ m² than at k₀ = 10.0 × 10⁻¹⁶ m², indicating that the influence of permeability on the instantaneous gas production rate is not simply linear. The CO₂ injection rate under different initial permeabilities generally exhibits a trend of first decreasing and then stabilizing, which is attributed to the swelling deformation of the coal matrix induced by CO₂ adsorption, leading to a gradual deterioration of reservoir porosity and permeability conditions and inhibiting the later-stage injection capacity. When the initial permeability is 3.5 × 10⁻¹⁶, 6.5 × 10⁻¹⁶, and 10.0 × 10⁻¹⁶ m², the stable CO₂ injection rates are 738, 1034, and 1477 m³/d, respectively, and the cumulative CO₂ storage amounts are 8,216,442 m³, 10,311,634 m³, and 12,733,369 m³, respectively, indicating that a higher initial permeability results in a stronger CO₂ injection capacity and a larger cumulative storage amount.

Figure 7. Under different initial permeabilities: (a) CH₄ production rate; (b) cumulative CH₄ production; (c) CO₂ injection rate; (d) cumulative CO₂ storage.

4.1.3. Influence of Initial Reservoir Elastic Modulus on CO₂-ECBM

With increasing elastic modulus, the CH₄ production rate and cumulative production decrease slightly (Figure 8). When the elastic modulus E is 3 GPa, 4 GPa, and 5 GPa, the stable CH₄ production rates are 1273, 1252, and 1232 m³/d, respectively, which are 137%, 133%, and 129% higher than those under natural production. The cumulative CH₄ production amounts are 801,198 m³, 787,441 m³, and 773,899 m³, representing increases of 273%, 267%, and 260% compared with natural production (214,436 m³). This indicates that, within the simulated range, a lower mechanical strength is more favorable for CH₄ enhancement. The differences in CO₂ injection rate and cumulative CO₂ storage under different elastic modulus conditions are very small. When E is 3 GPa, 4 GPa, and 5 GPa, the stable CO₂ injection rates are 736, 737, and 738 m³/d, respectively, and the cumulative CO₂ storage amounts are 7,699,870 m³, 7,720,622 m³, and 7,732,819 m³, respectively, indicating that the mechanical strength of the reservoir has a minor effect on CO₂ injection capacity and storage amount.

Figure 8. Under different initial elastic modulus: (a) CH₄ production rate; (b) cumulative CH₄ production; (c) CO₂ injection rate; (d) cumulative CO₂ storage.

4.2. Influence of Engineering Parameters on CO₂-ECBM

4.2.1. Influence of Injection Pressure on CO₂-ECBM

Overall, CO₂ injection significantly enhances CH₄ production, and a higher injection pressure leads to a more pronounced enhancement effect (Figure 9). After CO₂ injection, the CH₄ production rate is significantly higher than that under natural production and increases with increasing injection pressure, because a higher injection pressure strengthens the pressure gradient around the wellbore and the CH₄ migration capacity, thereby enhancing the competitive adsorption and displacement of CH₄ by CO₂. At injection pressures of 7 MPa, 8 MPa, and 9 MPa, the stable CH₄ production rates are 987, 1108, and 1232 m³/d, respectively, which are 84%, 106%, and 129% higher than that under natural production (536 m³/d). The cumulative CH₄ production amounts are 655,370 m³, 710,619 m³, and 773,899 m³, representing increases of 204%, 231%, and 260% compared with natural production (214,436 m³). These results indicate that increasing the CO₂ injection pressure is beneficial for CH₄ enhancement.

Figure 9. Under different injection pressures: (a) CH₄ production rate; (b) cumulative CH₄ production; (c) CO₂ injection rate; (d) cumulative CO₂ storage.

Under different injection pressures, the CO₂ injection rate generally exhibits a pattern of being high in the initial stage, followed by a rapid decline and then stabilization. This is because the pressure gradient is large in the early injection stage, resulting in a strong injection capacity. As time progresses, continuous CO₂ adsorption induces coal swelling and reduces porosity and permeability; meanwhile, the water saturation near the gas migration front increases and the gas relative permeability decreases, leading to a gradual stabilization of the injection rate. At injection pressures of 7 MPa, 8 MPa, and 9 MPa, the stable CO₂ injection rates are 714, 747, and 738 m³/d, respectively, indicating that a higher injection pressure can enhance the injection capacity, but the later-stage stable injection rate does not increase strictly linearly with pressure. This is because a higher pressure, on the one hand, enhances the driving effect, while on the other hand, it exacerbates the permeability inhibition caused by adsorption-induced swelling. Regarding cumulative CO₂ storage, under the conditions of 7 MPa, 8 MPa, and 9 MPa, the values are 6,862,021 m³, 7,588,879 m³, and 8,216,442 m³, respectively, indicating that increasing the injection pressure can significantly increase the scale of CO₂ storage.

4.2.2. Influence of Injection Mode on CO₂-ECBM

In terms of CH₄ production performance (Figure 10a,b), the CH₄ production rate under continuous constant-pressure injection is generally higher than that under stepwise pressurization injection. Within the simulated time scale, the stable CH₄ production rates under continuous constant-pressure injection and stepwise pressurization injection are 1232 m³/d and 1193 m³/d, respectively, which are 129% and 122% higher than that under natural production (536 m³/d). The cumulative CH₄ production amounts are 773,899 m³ and 718,272 m³, representing increases of 260% and 234% compared with natural production (214,436 m³). This indicates that continuous constant-pressure injection is more favorable for CH₄ enhancement. In terms of CO₂ injection and storage performance (Figure 10c,d), the stable CO₂ injection rate under continuous constant-pressure injection is 738 m³/d, which is lower than the 818 m³/d under stepwise pressurization injection, indicating that stepwise pressurization injection has a higher stable injection capacity in the later stage. However, the cumulative CO₂ storage under continuous constant-pressure injection is 8,216,442 m³, which is approximately 3% higher than that under stepwise pressurization injection (7,973,020 m³), indicating that continuous constant-pressure injection is more beneficial for increasing cumulative CO₂ storage.

Figure 10. Under different injection methods: (a) CH₄ production rate; (b) cumulative CH₄ production; (c) CO₂ injection rate; (d) cumulative CO₂ storage.

4.3. Sensitivity Analysis of Geological and Engineering Parameters for CO₂-ECBM

To quantitatively compare the influence of different geological and engineering parameters on CO₂-ECBM, cumulative CH₄ production, cumulative CO₂ storage, and stable CO₂ injection rate are selected as sensitivity evaluation indicators [31,32,33]. The injection mode, being a discrete variable, is treated separately as an engineering scheme for comparative evaluation. To avoid the influence of different parameter dimensions and ranges on the evaluation results, a normalized sensitivity coefficient is used to evaluate each parameter:

$$S_{ij}={\displaystyle \frac{(Y_{ij,\max }-Y_{ij,\min })/{\overline{Y}}_{ij}}{(X_{i,\max }-X_{i,\min })/{\overline{X}}_i}}$$

(9)

where S_ij is the sensitivity coefficient of the i-th parameter with respect to the j-th evaluation indicator; Y_ij,max, Y_ij,min, and ${\overline{Y}}_{ij}$ are the maximum, minimum, and average values of the j-th evaluation indicator as the i-th parameter varies, respectively; and X_ij,max, X_ij,min, and ${\overline{X}}_i$ are the maximum, minimum, and average values of the i-th parameter, respectively. A larger S_ij indicates a more significant influence of the parameter on the corresponding evaluation indicator.

Based on the single-indicator sensitivity evaluation, a comprehensive sensitivity index is further constructed as [31]:

$$S_i = w_1{S}_{i,q_{C{H}_4}}+w_2{S}_{i,Q_{C{H}_4}}+ w_3{S}_{i,Q_{C{O}_2}}+ w_4{S}_{i,q_{C{O}_2}}$$

(10)

where S_i is the comprehensive sensitivity index of the i-th parameter; $S_{i,q_{C{H}_4}}$, $S_{i,Q_{C{H}_4}}$, $S_{i,Q_{C{O}_2}}$, and $S_{i,q_{C{O}_2}}$ are the sensitivity coefficients of this parameter with respect to the stable CH₄ production rate, cumulative CH₄ production, cumulative CO₂ storage, and stable CO₂ injection rate, respectively; and w₁, w₂, w₃, and w₄ are the weight coefficients of the corresponding evaluation indicators. In this study, the weight coefficients were determined based on engineering experience and the relative importance of different indicators for CO₂-ECBM operation in low-permeability coal seams. Because the evaluation emphasizes CO₂ storage while also considering CH₄ recovery and injection stability, cumulative CO₂ storage was assigned the highest weight of 0.40, followed by cumulative CH₄ production with a weight of 0.25, and then stable CO₂ injection rate with a weight of 0.20. Stable CH₄ production rate was assigned a weight of 0.15. Therefore, the weight coefficients were set as w₁ = 0.15, w₂ = 0.25, w₃ = 0.40, and w₄ = 0.20. It should be noted that this weighting scheme reflects a storage-oriented engineering judgment and has a certain degree of subjectivity; for projects with different priorities, the weights can be adjusted accordingly.

The comprehensive sensitivity results for each indicator (Table 2 and Figure 11) show that initial porosity, CO₂ injection pressure, and initial permeability are the main factors affecting CO₂-ECBM performance in the study area (Figure 11a,b). According to the comprehensive sensitivity index, the ranking of the influencing factors is initial porosity > CO₂ injection pressure > initial permeability > elastic modulus. Among these, initial permeability and initial porosity primarily determine the CO₂ injectability and storage potential of the coal reservoir; thus, these are key geological parameters in reservoir screening. CO₂ injection pressure is the most important adjustable parameter in field engineering, directly influencing the CH₄ displacement intensity and cumulative CO₂ storage. In contrast, the elastic modulus of the reservoir has a relatively weak effect on the stable CO₂ injection rate and cumulative CO₂ storage, indirectly affecting CH₄ production mainly through coal deformation and permeability evolution. The injection mode exerts a certain regulatory effect on engineering performance (Figure 11c), but its overall influence is weaker than that of reservoir flow capacity and injection pressure.

Table 2. Sensitivity coefficients of different indicators for geological–engineering parameters.

Parameter	${\mathit{S}}_{{\mathit{q}}_{\mathit{C}{\mathit{H}}_4}}$	${\mathit{S}}_{{\mathit{Q}}_{\mathit{C}{\mathit{H}}_4}}$	${\mathit{S}}_{{\mathit{q}}_{\mathit{C}{\mathit{O}}_2}}$	${\mathit{S}}_{{\mathit{Q}}_{\mathit{C}{\mathit{O}}_2}}$	S
phi0	0.7704	1.1842	1.0000	1.7554	1.1627
k0	0.2923	0.9671	0.4112	0.6474	0.5796
E	0.1146	0.1213	0.0021	0.0095	0.0503
Pinj	0.8837	0.6647	0.7170	0.1801	0.6216

Figure 11. Sensitivity of different geological–engineering parameters to CO₂-ECBM performance. (a) Sensitivity coefficients of parameters to each indicator; (b) comprehensive sensitivity coefficients of parameters to indicators; (c) sensitivity coefficients of injection methods to each indicator.

5. CO₂ Injection Optimization

The engineering performance of CO₂-ECBM is difficult to evaluate using a single indicator. Although an excessively high injection pressure can enhance the ability of CO₂ to enter the coal reservoir and improve the CH₄ displacement effect, it may also increase the risks of local reservoir fracturing, the formation of preferential flow pathways, and early CO₂ breakthrough. Conversely, an excessively low injection pressure makes it difficult to achieve the desired storage and enhancement effects. Meanwhile, as injection time increases, the CO₂ injection rate gradually decreases and stabilizes, and the marginal increments of CH₄ enhancement and CO₂ storage decline, so the engineering benefits of continued injection will gradually diminish. Therefore, injection optimization requires a comprehensive trade-off among CH₄ enhancement, CO₂ storage, engineering efficiency, and reservoir safety.

5.1. Optimization Method and Evaluation Indicators for CO₂ Injection

Cumulative CH₄ production, cumulative CO₂ storage, engineering efficiency, and reservoir safety risk are selected as the evaluation indicators for CO₂ injection optimization. Among these, the engineering efficiency indicator is primarily characterized by the stable CO₂ injection rate, which reflects the ability of the coal reservoir to maintain continuous CO₂ entry during the later injection stage. A higher value of this indicator indicates better injectability and operational stability of the reservoir during the long-term injection phase. Reservoir safety risk mainly characterizes the risk of reservoir fracturing and the formation of preferential flow pathways that may occur when the injection pressure approaches the minimum principal stress [34].

To eliminate dimensional differences among the various indicators, a linear normalization method is used to perform dimensionless processing on each evaluation indicator [35]. For positive indicators, such as cumulative CH₄ production, cumulative CO₂ storage, and stable CO₂ injection rate, the normalization formula is:

$$X_N={\displaystyle \frac{X-X_{\min }}{X_{\max }-X_{\min }}}$$

(11)

For negative indicators, such as reservoir safety risk, the normalization formula is:

$$C_{\mathrm{risk},N}={\displaystyle \frac{P_{\mathrm{inj}}-P_0}{{\sigma }_{\min }-P}}$$

(12)

where C_risk,N is the reservoir safety risk index; and σ_min is the minimum principal stress of the reservoir in the study area, in MPa, with a value of 11.48 MPa. On this basis, a comprehensive engineering performance evaluation function for CO₂-ECBM is constructed as [36,37]:

$$V={\beta }_1{Q}_{C{H}_4,N}+{\beta }_2{Q}_{C{O}_2,N}+{\beta }_3{E}_N-{\beta }_4{C}_{risk,N}$$

(13)

where V is the comprehensive engineering performance value for CO₂-ECBM; Q_CH4,N is the normalized indicator of cumulative CH₄ production; Q_CO2,N is the normalized indicator of cumulative CO₂ storage; E_N is the normalized indicator of engineering efficiency; and β₁, β₂, β₃, and β₄ are the weight coefficients of the corresponding evaluation indicators. Under the same engineering objective, a larger V value indicates a better overall performance of the injection scheme.

Based on the different emphases of CO₂-ECBM engineering on CH₄ enhancement, CO₂ storage, engineering efficiency, and reservoir safety, four CO₂ injection optimization objectives are established: (1) comprehensive consideration of CH₄ enhancement, CO₂ storage, engineering efficiency, and reservoir safety; (2) CO₂ storage priority; (3) CH₄ enhancement priority; and (4) engineering efficiency priority, which focuses on diminishing returns in the later injection stage and operational benefits. The weight coefficients for each parameter corresponding to these objectives are listed in Table 3.

Table 3. Weight coefficients of objective functions.

Engineering Objectives	β1	β2	β3	β4
Comprehensive consideration	0.30	0.40	0.20	0.10
CO2 storage priority	0.20	0.55	0.15	0.10
CH4 enhancement priority	0.55	0.25	0.10	0.10
Engineering efficiency priority	0.20	0.30	0.40	0.10

5.2. Optimization Analysis Under Different Injection Pressures and Injection Schemes

5.2.1. Injection Pressure

From the engineering performance (Figure 12a), under the CH₄ enhancement-priority and CO₂ storage-priority objectives, the engineering effect value generally increases with injection pressure. Particularly under the condition of 9 MPa, the enhancement and storage effects are especially significant. In contrast, for the engineering efficiency-priority objective, 8 MPa is more advantageous, especially because the stable CO₂ injection rate is relatively high while the reservoir safety risk is lower than that at 9 MPa, making 8 MPa the optimal choice for long-term stable operation. Under the comprehensive multi-objective scenario, the performance within the range of 8–9 MPa is considered satisfactory. However, considering the minimum principal stress constraint of the study area and the potential risks associated with long-term high-pressure injection, it is recommended to control the injection pressure within the range of 8.0–9.0 MPa rather than selecting the maximum value.

Figure 12. Performance of CO₂-ECBM under varying injection pressures and injection modes. (a) Injection pressure; (b) injection scheme.

5.2.2. Injection Scheme

Beyond the effect of injection pressure, the injection method also influences the pressure disturbance caused by CO₂ entering the coal reservoir and the subsequent stable displacement performance (Figure 12b). Continuous constant-pressure injection helps maintain a high pressure gradient and is suitable for CH₄ enhancement-priority and CO₂ storage-priority objectives, favoring the increase in cumulative CH₄ production and CO₂ storage. Stepwise pressurization injection can effectively reduce the pressure impact at the initial injection stage and gradually establish the reservoir pressure field, making it suitable for engineering efficiency-priority objectives and conditions with strong reservoir safety constraints. Considering factors such as the stabilization of the CO₂ injection rate and the recovery of the CH₄ production rate after a trough, a reasonable injection method should be the combination of “stepwise pressure ramping in the early stage followed by constant pressure injection in the middle and late stages”. This method can not only reduce reservoir pressure disturbance and the risk of localized fracturing but also enhance CO₂ storage and CH₄ displacement performance. If the cumulative CH₄ production and CO₂ storage show a gradual slowdown in the later stage, this can serve as a reference for adjusting the injection scheme or terminating injection.

5.3. Optimization Scheme for CO₂ Injection in the Study Area

Integrating the sensitivity analysis and multi-objective comprehensive evaluation results, this study proposes that the optimization of CO₂-ECBM injection in the study area should follow the principles of: screening reservoirs based on geological parameters, optimizing injection schemes based on engineering parameters, controlling risks under safety constraints, and dynamically adjusting injection timing through performance evaluation. Specifically, initial porosity and initial permeability are key intrinsic factors determining the CO₂ injectability and storage potential of the coal reservoir, and should serve as important criteria for target coal seam selection. CO₂ injection pressure and injection scheme are the core variables for engineering optimization, directly affecting CH₄ displacement efficiency and CO₂ storage capacity.

Based on the simulation results and the reservoir safety conditions of the study area, the CO₂ injection pressure should be controlled within the range of 8.0–9.0 MPa. Within this range, 8.0–8.5 MPa is recommended as the preferred pressure interval for long-term stable injection. When the engineering objective is to maximize CO₂ storage or CH₄ production, and the minimum principal stress and pressure monitoring results indicate a sufficient safety margin, the injection pressure can be increased to approximately 9.0 MPa. When the engineering objective emphasizes long-term stable operation and engineering efficiency, an injection pressure in the range of 8.0–8.5 MPa is recommended. Regarding the injection scheme, a combined strategy of “stepwise pressure ramping in the early stage followed by constant pressure injection in the middle and late stages” is advised. In the early stage, a relatively low pressure with a gradual increase can be adopted to reduce reservoir pressure disturbance and the risk of localized fracturing. After the reservoir pressure field and CO₂ migration pathways have stabilized, the injection pressure is maintained within 8.0–9.0 MPa to enhance the cumulative CO₂ storage and CH₄ displacement performance. Meanwhile, the injection scheme should be dynamically adjusted based on indicators such as the stabilization time of the CO₂ injection rate, the trough time of the CH₄ production rate, and the marginal increments of cumulative production and storage metrics.

It should be noted that the initial reservoir pressure, temperature, and water saturation were assumed to be uniformly distributed in the present model. This simplification was adopted to isolate the effects of key geological and engineering parameters under the same numerical framework. However, actual coal reservoirs usually exhibit spatial heterogeneity in pressure, temperature, saturation, permeability, and fracture development. Such heterogeneity may influence CO₂ migration pathways, CH₄ desorption and production behavior, local pressure propagation, and CO₂ breakthrough risk. Therefore, the simulation results in this study should be regarded as the average response of the target coal seam under representative conditions. Future studies should further incorporate spatial heterogeneity of reservoir properties and initial conditions to improve the predictive reliability of CO₂-ECBM simulations. In addition, this study focused on deterministic sensitivity analysis and did not conduct probabilistic uncertainty analysis. Parameters such as permeability, adsorption coefficients, and mechanical properties may contain spatial variability and measurement uncertainty, which may affect the predicted CH₄ production, CO₂ storage, injection capacity, and reservoir safety risk. Future studies should combine Monte Carlo simulation, Latin hypercube sampling, or other probabilistic methods with the coupled model to further evaluate the uncertainty and robustness of the simulation results.

6. Conclusions

(1)

A coupled numerical model for CO₂-ECBM in low-permeability coal seams was established by considering CO₂/CH₄ competitive adsorption, matrix diffusion, fracture seepage, gas–water two-phase flow, coal deformation, and porosity–permeability evolution. The benchmark validation indicates that the model can reasonably reproduce the main dynamic characteristics of CH₄ production and CO₂ injection reported in previous studies, providing a basis for analyzing the coupled response of low-permeability coal reservoirs under CO₂ injection.

(2)

The comprehensive sensitivity analysis shows that initial porosity, CO₂ injection pressure, and initial permeability are the dominant factors controlling CO₂-ECBM performance in the study area. Initial porosity mainly determines reservoir storage space and CO₂ sequestration potential, while initial permeability controls pressure transmission and gas migration capacity. Therefore, porosity and permeability should be regarded as key geological indicators for target reservoir screening. CO₂ injection pressure is the most important adjustable engineering parameter, directly affecting CH₄ displacement intensity, CO₂ storage scale, and reservoir safety risk.

(3)

The multi-objective evaluation results indicate that the CO₂ injection pressure in the study area should be controlled within 8.0–9.0 MPa. For long-term stable operation, 8.0–8.5 MPa is recommended as the preferred pressure range because it can balance CO₂ storage, CH₄ enhancement, injection stability, and reservoir safety. When the engineering objective prioritizes CO₂ storage or CH₄ production and a sufficient safety margin is confirmed, the injection pressure may be increased to approximately 9.0 MPa.

(4)

In terms of injection strategy, continuous constant-pressure injection is favorable for increasing cumulative CH₄ production and CO₂ storage, whereas stepwise pressurization can reduce early-stage pressure disturbance and improve later-stage injection stability. Therefore, a combined injection scheme involving early-stage stepwise pressure ramping followed by middle- and late-stage constant-pressure injection is recommended. Overall, this study provides an engineering-oriented reference for reservoir screening, injection pressure design, and injection scheme optimization in similar low-permeability coal seams.