Thermal Displacement with CO₂ for E-CBM Recovery: Mechanisms and Efficacy of Temperature–Pressure Synergy in Permeability Enhancement

Abstract

The efficient development of coalbed methane (CBM) faces persistent challenges due to low recovery rates. While CO₂ thermal displacement offers a promising approach, the pore–fracture structure (PFC) evolution and gas displacement mechanisms under temperature–pressure coupling remain insufficiently clear. To address this knowledge gap, the in situ, dynamic quantification of pore–fracture evolution during CO₂ displacement was achieved by an integrated system with NMR and CT scanning, revealing the expansion, connection, and reconfiguration of coal PFC under temperature–pressure synergy and establishing the intrinsic relationship between supercritical CO₂ (ScCO₂)-induced permeability enhancement and methane displacement efficiency. Experimental results identify an observed transition in permeability near 80 °C under the tested conditions as a critical permeability transition point: below this value, permeability declines from 0.61 mD to 0.49 mD, reflecting pore structure adjustment; above it, permeability rises markedly to 1.18 mD, indicating a structural shift toward fracture-dominated flow. A “pressure-dominated, temperature-assisted” mechanism is elucidated, wherein pressure acts as the primary driver in creating macro-fractures and forming percolation pathways, while temperature—mainly via thermal stress—promotes micro-fracture development and assists gas desorption, offering only limited direct contribution to permeability. Although elevated injection pressure enhances permeability and establishes fracture networks, displacement efficiency eventually reaches a physical limit. To transcend this constraint, a synergistic production mechanism is proposed in which pressure builds flow channels while temperature activates microporous desorption. This study provides an integrated, in situ quantification of the pore–fraction evolution under high-temperature ScCO₂ conditions. The elucidated synergy between pressure and temperature offers insights and an experimental basis for the design of deep CBM recovery and CO₂ storage strategies.

1. Introduction

China has vast CBM resources, with approximately 36.8 trillion cubic meters within 2000 m depth, ranking third globally [1,2,3]. Current extraction mainly uses hydraulic fracturing and dewatering for pressure reduction, creating single seepage channels to boost gas production. However, the generally low permeability of coal reservoirs severely hinders gas desorption, diffusion, and flow, leading to low recovery rates [4,5,6]. Furthermore, about two-thirds of China’s CBM resources lie deeper than 1000 m, where increased in situ stress and gas content further reduce permeability, raising extraction difficulties [7]. Therefore, effectively enhancing permeability in low-permeability coal seams is a critical challenge for China’s CBM development [8].

Extensive research shows that ScCO₂ exhibits properties of both gas and liquid phases, characterized by low viscosity, high diffusivity, and excellent fluidity, enabling it to rapidly penetrate micro-fractures and pores in porous media and enhance permeability [9,10]. When injected into low-permeability coal seams, ScCO₂ not only displaces CH₄ like non-supercritical CO₂ but also significantly improves permeability by extracting organic matter, disrupting the molecular structure and ordered aromatic lamellae within coal, and inducing matrix swelling and deformation [11,12]. These processes alter the PFC evolution, mechanical strength, and gas adsorption capacity of coal, thereby increasing permeability and enhancing CBM recovery [13,14]. Additionally, elevated coal seam temperature activates CH₄ molecules, promoting desorption and migration while also improving permeability, thereby strengthening gas displacement efficiency and CO₂ sequestration [15,16,17]. Thus, heated ScCO₂ injection combines the dual advantages of heating and CO₂ injection, emerging as a promising technique for enhanced CBM recovery.

In recent years, extensive experimental research has been conducted to understand the evolution of seepage pathways and underlying mechanisms in coal reservoirs following CO₂ injection. Liu et al. [18] observed a continuous decline in coal permeability with increasing temperature during supercritical CO₂ (ScCO₂) stimulation at 35–65 °C and 7–10 MPa, with reported reductions ranging from approximately 20% to 60% depending on coal type and conditions, attributing this to pore compression caused by thermal expansion. In contrast, under non-supercritical conditions (0.21–1.8 MPa, 20–50 °C), Zhang et al. [19] reported a V-shaped permeability trend: decreasing by up to ~40% below 40 °C due to thermal expansion and increasing by ~15–30% above 40 °C as thermal stress-induced fractures developed. Li et al. [20] identified a similar but higher transition temperature near 55 °C under comparable conditions, with permeability variations in the order of 35–50%, reflecting variability across coal ranks. Further diverging from these trends, Liu et al. [21] monitored permeability evolution during pyrolysis up to 400 °C under unconfined conditions, revealing an N-shaped trend: a slight increase of ~5–10% up to 100 °C from water evaporation and gas desorption, a decline of ~20–25% at 100–200 °C due to asphaltene clogging, and a sharp rise exceeding 200% at 300–400 °C resulting from extensive fracture formation. While these studies highlight diverse thermal effects on permeability, variations in experimental conditions, coal types, and control methods have hindered a unified understanding of PFC evolution during heating [22,23,24]. In particular, the behavior of coal under ScCO₂ injection at temperatures above 70 °C remains unclear, warranting further systematic investigation to elucidate the governing mechanisms.

Furthermore, the dynamic evolution of coal’s micro-fracture and pore structure remains unclear, with no consensus established [25,26]. For instance, Zhang et al. [27], using high-pressure gas injection coupled with micro-CT, found that under 7–10 MPa injection pressure, the proportion of micropores (<2 nm) initially increased and then decreased with injection time, while mesopore (2–50 nm) volume expanded significantly during early stages before stabilizing. They attributed this reorganization to ScCO₂’s extraction of organic matter. In contrast, Liu et al. [28] observed under near-critical conditions (31 °C, 7.2 MPa) that fracture aperture initially decreased due to matrix swelling, reducing permeability by approximately 20%, but after 6 h of continuous injection, average aperture increased by 15% from mineral dissolution and structural relaxation, revealing time-varying and nonlinear characteristics. Li et al. [29] further highlighted rank-dependent responses: medium-rank bituminous coal exhibited a 12.8% increase in total pore volume after ScCO₂ injection, whereas high-rank anthracite showed only a 3.5% increase, with newly formed fractures mainly isolated [30,31,32]. These studies collectively demonstrate limited systematic understanding of the evolution pathways, spatial reconfiguration, and subsequent impacts on seepage capacity of coal fractures and pores under ScCO₂ [33,34,35]. Particularly within the 70–150 °C range, the response under thermo–mechanical–chemical coupling remains unclear, and it is still debated whether the dominant factor in PFC evolution is temperature or pressure. Clarifying the dominant control mechanism and advancing in situ observation and multi-scale characterization are urgently needed to reveal the underlying mechanisms.

This study investigated the evolution of coal micro-fractures and pores under supercritical CO₂ at 35–150 °C through high-pressure thermal stimulation experiments. The fundamental novelty of this work lies in the in situ, same-sample, multi-modal characterization (NMR and CT) of pore–fracture evolution, enabling a quantitative comparison of temperature versus pressure effects within a controlled, high-temperature ScCO₂ experimental framework. By comparing NMR and CT scans of identical samples before and after treatment, this work tracks structural changes and clarifies the temperature–pressure coupling effects on pore–fracture evolution. The analysis clarifies temperature–pressure coupling effects on fracture propagation and pore reconfiguration, contributing to a clearer understanding of the relationship between ScCO₂-induced permeability enhancement and gas displacement efficiency. The findings address key mechanisms of ScCO₂–coal interaction at elevated temperatures and provide data and mechanistic support for enhancing deep CBM recovery and CO₂ geological storage.

2. Experiments Methods

2.1. Coal Sample Preparation

The coal samples used in this experiment were collected from the No. 5 coal seam in the Sanjiao Block, Ordos Basin, China. This medium-rank bituminous coal was selected for its representativeness of a major CBM-producing reservoir in China and its suitability for investigating temperature–pressure induced pore–fracture transformations under ScCO₂ conditions. The samples were precision wire-cut into standard cylindrical specimens (φ25 mm × 50 mm) with parallel end faces and no visible lateral fractures.

The key parameters presented in

Table 1. Maceral analysis, proximate analysis, and initial porosity–permeability of the coal.

Samples	R0,max/%	Maceral Content/%			Porosity/%	Permeability (mD)	Density (g/cm3)
		Vitrinite	Inertinite	Minerals
CL-1	1.222	75.6	21.8	2.6	6.1	0.023	1389.6
CL-2	1.235	74.3	22.4	3.3	5.8	0.031	1392.1
CH-1	1.288	78.3	20.3	1.4	5.5	0.024	1401.4
CH-2	1.246	73.8	23.5	2.7	5.7	0.025	1396.3

provide the essential baseline for interpreting the experimental outcomes. The vitrinite reflectance (R₀,_max) confirms the medium rank of the bituminous coal samples, which is representative of significant deep CBM resources in China. The high vitrinite content is relevant to gas adsorption and mechanical behavior during stimulation. The low initial porosity and permeability are characteristic of the tight coal reservoirs targeted for enhancement. The samples were specifically selected with closely matched properties to minimize the influence of sample heterogeneity, allowing the effects of temperature–pressure coupling to be isolated and analyzed. These property ranges are consistent with reported data for similar rank coals in the Ordos Basin, ensuring the relevance of our findings to the region.

Prior to thermal stimulation experiments, the initial porosity and permeability of all samples were measured using a gas porosimeter–permeameter [36]. To control systematic errors and ensure experimental reliability, six samples with closely matched porosity and permeability values were selected from the total set as valid experimental subjects for subsequent supercritical CO₂ thermal stimulation tests. The reported initial porosity and permeability values represent the mean of triplicate measurements, with standard deviations of less than 5%. Additionally, the reported permeability values are Klinkenberg-corrected absolute permeabilities, derived from measurements at multiple mean pore pressures to account for gas slippage effects, which are significant in this low-permeability range. Table 1 summarizes the maceral analysis, proximate analysis, and initial porosity–permeability data of the four coal cores.

2.2. Experimental Procedure

2.2.1. CO₂ Thermal Displacement with Online NMR Monitoring

The experiment employed a custom-integrated, in-house designed variable-temperature high-pressure nuclear magnetic resonance (NMR) imaging analysis system, as illustrated in Figure 1. The setup primarily consists of a constant-pressure/constant-flow pump, a high-low temperature and pressure circulation system, a heating and temperature control unit, pressure transducers, and a Suzhou Niumag variable-temperature high-pressure NMR imaging analysis system.

The temperatures of 35, 80, 120, and 150 °C were selected to achieve the following objectives: 35 °C establishes a supercritical baseline near common reservoir conditions; 80 °C allows for investigation of potential transition behavior; 120 °C and 150 °C enable the study of pronounced thermo-mechanical effects and enhanced desorption kinetics, exploring the potential of thermal stimulation for recovery enhancement.

(1)

CH₄ Saturation and Adsorption

The dried coal sample was placed in the NMR holder and saturated with CH₄ at set temperatures (35, 80, 120, and 150 °C). The injection pressure was gradually increased from atmospheric pressure to the reservoir pressure of 8 MPa at a rate of 0.1 MPa/min. While maintaining the target temperature, CH₄ was replenished slowly whenever the pressure dropped by 0.015 MPa until equilibrium was achieved (pressure fluctuation ≤ 0.015 MPa over 8 h), and the total injected CH₄ volume was recorded. Confining pressure was increased synchronously with injection pressure to prevent sample damage, reaching 8 MPa simultaneously.

(2)

Online NMR CO₂ Displacement under High Temperature and Pressure

① CO₂ was heated to the target temperature (35, 80, 120, or 150 °C) and injected at a constant pressure of 8 MPa. Injection began at a low flow rate of 0.01 mL/min, which was gradually adjusted based on outlet flow and pressure to avoid rapid pressure changes and sample failure. Injected CO₂ volume was recorded in real time.

② A back-pressure regulator maintained a constant outlet pressure of 6 MPa, stabilizing the core pressure at 8 MPa. Throughout the experiment, gas flow and pressure at both the injection and outlet ends were continuously monitored. The flow rate was carefully controlled to ensure stable, laminar flow through the low-permeability core. Prior to each NMR measurement, the system was allowed to reach steady-state flow under these constant pressure conditions. This ensured a stable fluid pressure distribution within the sample and minimized transient compression/decompression effects on the acquired T₂ signal.

2.2.2. Pore–Fracture Characterization via NMR Before and After Displacement

Nuclear Magnetic Resonance (NMR) provides an effective approach for quantitatively characterizing the internal pore and fracture structure of coal. This study employs the transverse relaxation time (T₂) distribution to reveal pore structure characteristics of coal samples. The principle relies on the fact that the relaxation rate of hydrogen nuclei in the pore fluid is governed by surface interactions, where the relaxation time correlates directly with pore size, as expressed by

$${\displaystyle \frac{1}{T_2}}={\rho }_2{\displaystyle \frac{S}{V}}$$

(1)

Here, T₂ (in ms) is the measured transverse relaxation time, ρ₂ (in µm/ms) is the surface relaxivity of the coal, and S and V (in cm² and cm³) represent the surface area and volume of a single pore, respectively. For spherical pores, the S/V ratio simplifies to 3/r, while for slit-shaped fractures, it approximates 2/r, where r denotes the characteristic pore radius.

The experiments were conducted using a variable-temperature high-pressure NMR imaging analysis system, with a main magnetic field strength of 0.3 T and an RF frequency range of 1–42 MHz. Prior to testing, standard coal rock cores were subjected to pretreatment: they were placed in a vacuum pressurized saturation device and saturated with distilled water for over 8 h to ensure full pore filling. The saturated samples were then positioned at the magnet center and scanned using the CPMG sequence to acquire raw echo train data. The T₂ spectrum, representing pore size distribution, was obtained by applying an inversion algorithm to the data. After all NMR measurements, the samples were transferred to a vacuum drying oven and dried for 12 h for subsequent experiments.

It is explicitly acknowledged that the transverse relaxation time (T₂) distribution is a non-unique indicator of pore structure. The measured T₂ values are influenced by a combination of factors including pore size, surface relaxivity, fluid properties, and diffusion effects. Therefore, in this study, absolute T₂ values are not used as direct, quantitative proof of specific pore-to-fracture conversion. Instead, the relative changes observed in the T₂ spectra of the same sample before and after treatment, under rigorously controlled experimental conditions, are interpreted as qualitative trends consistent with alterations in the pore–fracture architecture. These interpretations are further validated and substantiated by direct, three-dimensional visualization from X-ray micro-CT scans (Figure 2).

2.2.3. Pre- and Post-Displacement 3D CT Scanning

The experiments were performed using a NanoVoxel-350E 3D computed tomography system (Tianjin, China). This instrument achieves a maximum spatial resolution of 0.5 μm, with a maximum scanning voltage of 190 kV and a maximum power of 25 kW. A stepwise scanning mode was employed (Figure 3).

To characterize the three-dimensional spatial distribution of mesopores and fractures within the sample, the two-dimensional slices underwent preprocessing and three-dimensional reconstruction. To enhance the quality of the original CT slices, a non-local means filtering method was applied for 3D denoising. This approach combines the advantages of Gaussian filtering in preserving image details and edges with the pronounced smoothing effect of mean filtering.

After applying the median filter (Figure 4), the grayscale CT images with intensity values ranging from 0 to 255 were segmented using Equation (2):

$$v\left(i,i\right)=\left\{\begin{array}{l}0\begin{array}{c}\mathit{\hspace{1em} l}(i,j)⩽T_1\end{array}\\ 1\begin{array}{c}{\mathit{\hspace{1em} T}}_1<l(i,j)⩽T_2\end{array}\\ 2\begin{array}{c}{\mathit{\hspace{1em} T}}_2<l(i,j)\end{array}\end{array}\right.$$

(2)

where l (i, j) represents the grayscale value of the pixel, T₁ and T₂ are the segmentation thresholds. v (i, j) denotes the three segmented primary components, with 0 representing voids, 1 representing matrix, and 2 representing minerals.

During data analysis of the segmented images, the volume fraction of each component was calculated using Equation (3):

$$\phi =N_c/N_t\times 100\%$$

(3)

where N_c is the number of voxels belonging to a specific component, and N_t is the total number of voxels.

3. Experimental Results and Analysis

3.1. Temperature-Dependent Permeability Enhancement During Thermal Flooding

3.1.1. PFC Evolution at Different Temperatures

To systematically investigate the impact of thermal CO₂ flooding on the pore structure of coal, this study conducted tests on coal samples before and after thermal CO₂ treatment using an NIUMAG™ (Shanghai, China) low-field nuclear magnetic resonance analysis system. Based on Hodot’s pore classification scheme [37], the pore structures were categorized into four types: micropores (<10 nm), mesopores (10–100 nm), macropores (100–1000 nm), and fractures (>1000 nm).

Raw coal samples primarily displayed bimodal or trimodal T₂ distributions, with the main peak in the 0.1–1 ms range reflecting micropores, and a secondary peak in the 1–100 ms range indicating meso- and macropores. While minor variations existed among initial samples, all spectra consistently exhibited peaks within 10–100 ms, approximating a bimodal profile. These results indicate that despite subtle differences in peak features, the overall pore–fracture architecture remains broadly consistent across coal samples (Figure 5).

The T₂ spectrum of coal reveals distinct temperature-dependent evolution after CO₂–thermal treatment. At elevated temperatures (120 °C and 150 °C), the signal intensity in the 10–1000 ms range increases markedly (Figure 6a,c), demonstrating substantial growth in macropores and fractures. Concurrently, the peak in the 0.1–1 ms range weakens and shifts rightward, indicating micropore enlargement and a reduction in volume. Quantitative analysis of pore–fracture distribution further confirms a more pronounced increase in fractures proportion under higher temperatures. Specifically, samples CH-1 and CH-2 exhibited fractures proportion increases of 16.1% and 25.5%, respectively, accompanied by a decrease in micropores (32.0% and 41.8%) and an increase in macropores (10.7% and 8.9%). The observed conversion of micropores to macropores and the development of fractures at elevated temperatures are consistent with the effects of thermo-mechanical stress. We infer that non-uniform thermal expansion likely induces localized tensile stress, promoting micro-crack formation and propagation.

Compared to pre-treatment conditions, T₂ spectral changes in coal samples subjected to lower-temperature CO₂ thermal treatment were statistically insignificant (35 °C and 80 °C). The spectrum shifted slightly rightward, with reduced amplitude in the 0.1–1 ms range and lower peaks between 1 and 100 ms. This reflects general pore enlargement, a decrease in micropores, and a slight increase in meso- and macropores. Notably, no clear peak formed in the 100–1000 ms range—unlike under high temperatures—indicating no significant fractures development. These results confirm that weaker thermal stress at lower temperatures is inadequate to substantially modify the fractures system or induce extensive micro-fractures.

Figure 7 illustrates the evolution of pore-fissure distribution following CO₂ thermal treatment. At lower injection temperatures (35 °C and 80 °C), the increase in fractures proportion remained limited, rising by only 10.3% and 8.6%, respectively. At lower treatment temperatures, the limited thermal stress applied to the coal is insufficient to induce significant tensile failure, resulting in a low degree of fracture development.

Compared to the fractures system, low-temperature CO₂ thermal treatment exerts a more consistent influence on the pore structure. After treatment, the proportion of micropores decreased by 29.1% and 24.3%, while that of macropores increased by 15.8% and 13.0%, respectively (Figure 8). These alterations in the pore system are primarily attributed to structural readjustments induced by thermal expansion and CO₂–coal matrix interactions. Under identical thermal treatment conditions, the expansive response of the coal matrix remains relatively uniform, leading to a consistent evolutionary trend in the pore system.

3.1.2. Evolution of Poro-Permeability Across Thermal Flooding Temperatures

Figure 9 illustrates the variations in NMR porosity of coal samples after CO₂ thermal treatment at different temperatures. Experimental results show a significant increase in total porosity after treatment, with the growth ranging from 94.8% to 238%. Under high-temperature conditions (120 °C and 150 °C), the coal matrix experienced substantial thermal stress, inducing numerous newly generated fractures and resulting in pronounced porosity increases of 193% and 238%, respectively. In contrast, under medium-low temperature treatments (35 °C and 80 °C), porosity also increased—by 120% and 94.8%, respectively—though the mechanism was dominated by macropore expansion, with limited fractures development induced by thermal stress.

Figure 10 further reveals that the porosity increase induced by high-temperature thermal treatment primarily originates from the development of fracture systems. Concurrently, as the thermal treatment temperature rises, the proportion of micropores in the coal decreases, reflecting the systematic impact of thermal transformation on the pore structure. In contrast, under medium-low temperature conditions, the enhancement in porosity is mainly attributed to the expansion of macropores, with fractures contributing minimally to the overall porosity increase (Figure 10).

These results demonstrate a distinct threshold effect of temperature on coal modification, with different dominant mechanisms across temperature ranges. Below 80 °C, CO₂ thermal treatment primarily induces pore structure rearrangement, characterized by a reduction in micropores and an increase in macropores. In contrast, at elevated temperatures (120 °C and 150 °C), intensified thermal stress leads to extensive fissure development, becoming the dominant factor enhancing permeability.

This mechanistic transition corresponds to a nonlinear permeability evolution, with values of 0.61 mD (35 °C), 0.49 mD (80 °C), 0.97 mD (120 °C), and 1.18 mD (150 °C) displaying a V-shaped trend centered at 80 °C (Figure 11). These data reveal a clear non-monotonic, V-shaped trend in permeability as temperature rises, with a minimum value observed at 80 °C. Given the sample-specific nature of each temperature path, the observed V-shaped trend is presented as a clear and consistent observational trend across the systematically varied conditions. By extending the temperature range to 150 °C, this study confirms 80 °C as the critical transition temperature for pore-permeability evolution, resolving previous debates over temperature thresholds and providing key insights for optimizing thermal-enhanced coalbed methane recovery.

3.2. Pressure-Dependent Permeability Enhancement During Thermal Flooding

3.2.1. Pore–Fracture Evolution Under Different Injection Pressures by NMR

The evolution of T₂ spectra and pore–fracture structure in coal under CO₂-thermal treatment was systematically compared at 150 °C with injection pressures increased from 8 MPa to 10 MPa. Enhanced injection pressure was observed to significantly strengthen the T₂ signal beyond 1000 ms, indicating the development of larger-scale macro-fractures. Correspondingly, the fracture proportion rose from 31.7% at 8 MPa to 47.4% under the elevated pressure condition (Figure 12).

This change is primarily attributed to the enhanced mechanical expansion effect under elevated injection pressure. As pressure increased, pre-existing pores and micro-fractures in the coal were progressively expanded and interconnected, gradually forming a macroscopic fracture system with certain aperture and extension. The development of such fractures not only directly increased the fracture proportion within the pore structure but also significantly improved the overall flow conductivity of the reservoir, serving as the key structural foundation for permeability enhancement.

3.2.2. Pore–Fracture- Evolution Under Different Injection Pressures by C

The development characteristics of coal fractures under CO₂–thermal treatment exhibit significant variations with increasing injection pressure. As shown in the 3D reconstruction results (Figure 13), at an injection pressure of 8 MPa, a dense network of short, morphologically complex fractures is predominantly formed. When the injection pressure rises to 10 MPa, the fracture system evolves further, developing more dominant macro-fracture channels on the basis of the existing short-fracture network (Figure 14).

The observed evolution is closely related to the action mechanisms of ScCO₂ in coal. At 8 MPa, CO₂ primarily induces microporous restructuring and short-fracture formation through physicochemical interactions with the coal matrix. When pressure increases to 10 MPa, the enhanced fluid potential and penetration pressure subject the coal to stronger tensile stress, which not only extends existing fractures but also promotes the generation of macro-fractures on the 10⁴–10⁵ nm scale. Although fracture directionality remains poor at 10 MPa, the significant increase in macro-fracture proportion substantially improves network connectivity, establishing multi-scale flow pathways from micro- to macroscopic levels.

A comparative analysis of pore–fracture structure evolution under 8 MPa and 10 MPa injection pressures reveals significant reorganization of the coal pore system as pressure increases. As pressure rises from 8 MPa to 10 MPa, the proportion of macro-fractures (10⁴–10⁵ nm) increases markedly from 4.7% to 24.5%, a rise of 19.8 percentage points (Figure 13 and Figure 14). The extensive development of such fractures enhances and interconnects the existing fracture network, forming a more continuous and efficient flow pathway system, thereby significantly improving the overall permeability and fluid transport efficiency of the reservoir. The corresponding signal intensity in the NMR T₂ spectrum for the macro-fracture range also strengthens, further confirming these structural changes. This indicates that the macro-fracture system formed under high injection pressure is a key structural element controlling the enhancement of reservoir seepage capacity.

3.3. Permeability Enhancement Mechanism During ScCO₂ Thermal Flooding

Under a heating temperature of 150 °C, as the injection pressure increases gradually from 6 MPa to 8 MPa, and further to 10 MPa, the porosity and permeability of coal show a clear linear growth trend. Experimental data from pre- and post-flooding indicate that porosity increases linearly from 9.6% at 6 MPa to 16.7% at 8 MPa, reaching 34.8% at 10 MPa (Figure 15). Similarly, permeability exhibits a linear rise from 0.54 mD at 6 MPa to 1.18 mD at 8 MPa, and further to 2.37 mD at 10 MPa (Figure 16).

This linear growth pattern is primarily attributed to the direct correlation between injection pressure and macro-fracture development. As injection pressure increases, the permeation pressure and fluid wedging effect of CO₂ in coal are continuously enhanced, promoting the propagation and interconnection of micro-fractures to form a more developed macro-fracture network. These newly formed macro-fractures not only expand storage space but, more importantly, significantly improve pore connectivity, establishing efficient fluid transport pathways. The continuous expansion and enhanced connectivity of the macro-fracture system provide a reliable physical mechanism and structural basis for the linear improvement in coal porosity and permeability.

Based on systematic experimental data, the contribution rates of temperature and pressure to coal permeability were first quantified, revealing a “pressure-dominated, temperature-assisted” mechanism for CO₂-enhanced displacement. Between 80 and 150 °C, an 87.5% temperature increase raised permeability from 0.49 mD to 1.18 mD, corresponding to a 0.008 mD increase per 1% temperature rise (Figure 10). Over 6–10 MPa, a 66.7% pressure increase elevated permeability from 0.54 mD to 2.37 mD, equivalent to a 0.027 mD gain per 1% pressure increase (Figure 16).

The quantitative relationship confirms the dominant role of injection pressure in enhancing coal permeability. Increased pressure directly strengthens CO₂ permeation and fracturing effects, promoting macro-fracture propagation and interconnection. In contrast, temperature elevation mainly facilitates micro-fracture development through thermal stress, with limited contribution to permeability. The established “pressure-dominated, temperature-assisted” mechanism provides a theoretical basis for optimizing thermal-enhanced CBM extraction. Practical applications should prioritize injection pressure enhancement over temperature modulation to maximize energy efficiency and recovery performance.

4. Discussion

The identification of 80 °C as a critical transition point is supported by the confluence of permeability data and pore–fracture evolution. The V-shaped permeability trend (minimum at 80 °C) coincides with a shift in the dominant structural modification mechanism. At and below 80 °C, permeability reduction is primarily associated with pore restructuring (micropore decrease, macropore increase) without significant fracture development, likely due to thermal expansion effects. Above 80 °C, the sharp permeability increase correlates with the onset of thermal stress-induced fracturing, where the generation and connectivity of fractures become the principal factor enhancing flow capacity. We interpret that, under our experimental conditions, the thermo-mechanical stress surpasses a threshold near 80 °C, initiating the transition from a pore-adjusted to a fracture-dominated flow regime.

CO₂ displacement tests conducted at 8 MPa and various temperatures demonstrate a linear increase in methane desorption efficiency between 35 °C and 150 °C. The efficiency rose from 57% at 35 °C to 68% at 80 °C, and further reached 76% at 120 °C and 82% at 150 °C. This represents a 44% improvement relative to the baseline at 35 °C, establishing 150 °C as the optimal injection temperature within the 0–150 °C range (Figure 17).

Based on linear regression of experimental data, the temperature dependence of desorption efficiency was quantified, producing corresponding linear equations. These equations allow for accurate prediction of methane desorption efficiency and volume at any temperature between 0 and 150 °C, offering a practical model for optimizing field injection temperature and forecasting production.

Further analysis at 150 °C across varying injection pressures shows that raising pressure from 6 MPa to 8 MPa markedly increased methane desorption efficiency from 59% to 82%. However, further increasing pressure to 10 MPa only raised efficiency marginally to 84%, revealing a distinct efficiency plateau in the 8–10 MPa range. This result contradicts the conventional view that higher pressure consistently yields significant efficiency gains (Figure 18).

Although increasing injection pressure enhances macroscopic permeability and creates fracture networks for gas transport, its displacement efficiency exhibits a distinct physical limitation: beyond 8 MPa, efficiency plateaus with minimal further improvement. This bottleneck stems from the strong capillary forces and adsorption potentials in isolated micropores, which tightly retain CH₄ molecules. Macroscopic fluid pressure neither transmits effectively into such nanoscale pores nor provides sufficient energy to release methane from its adsorbed state.

The limited efficacy of pressure in displacing residual methane from isolated micropores during CO₂ flooding arises from two physical constraints. First, the nanoscale confinement in micropores induces strong capillary forces and adsorption potentials that firmly trap CH₄ molecules, creating an energy barrier that fluid pressure alone cannot overcome. Second, pressure transmission becomes inefficient at the nanoscale: while CO₂ under high pressure can propagate through macroscopic fractures, it fails to effectively penetrate or fill isolated micropores, leaving the trapped methane unaffected by the applied pressure (Figure 19).

Heating effectively overcomes the limitations of pressure-driven displacement. Elevated temperature not only improves CO₂ diffusion into isolated micropores but also supplies adsorbed CH₄ molecules with adequate thermal energy to facilitate desorption, thereby mobilizing residual gas inaccessible to pressure alone. The synergy is further enhanced by optimizing pressure application. Cyclic injection—alternating between injection and equilibration—ensures uniform reservoir pressurization and promotes complete gas penetration into micropores. This strategy increases the efficiency of pressure utilization at the microscopic scale, working in concert with thermal effects.

In summary, the TAPM (Thermally Augmented, Pressure-Managed) approach establishes a synergistic framework for enhanced coalbed methane recovery. This methodology employs cyclic pressurization to ensure macroscopic sweep and efficient transport, while thermal activation releases gas from micropores through enhanced desorption. The integrated process systematically improves overall extraction efficiency.

5. Conclusions

Integrated NMR and CT scanning revealed the dynamic evolution of coal’s fracture network under supercritical CO₂ conditions. Temperature–pressure decoupling demonstrated their distinct roles in fracture propagation and connectivity, elucidating a mechanistic link between permeability enhancement and methane displacement based on the current experimental findings. These insights address fundamental knowledge gaps and support optimized deep CBM extraction with CO₂ storage.

(1)

Based on the experimental data, a temperature-dependent shift in the dominant modification mechanism during thermal CO₂ displacement was identified. Below 80 °C, pore restructuring dominates, characterized by a reduction in micropores and an expansion of macropores. Beyond this temperature, thermal stress becomes predominant, converting micropores to macropores and generating thermal fractures, thereby shifting the system from a pore-adjusted to a fracture-dominated regime. Consequently, the permeability trend exhibits a distinct transition near 80 °C under the tested conditions, marking the change from pore-adjusted to fracture-dominated flow.

(2)

NMR and CT scans demonstrate that elevated injection pressure promotes the expansion and interconnection of micro-fractures, forming dominant macro-fracture networks. Fracture proportion increased from 31.7% to 47.4%, with corresponding micropore reduction and macropore growth. This structural reorganization enhanced permeability from 0.54 mD to 2.37 mD (339% growth). The multi-scale analysis reveals the transition from localized propagation to global connectivity under pressure, providing key theoretical support for deep CBM extraction.

(3)

Based on the experimental data, this investigation supports a “pressure-dominated, temperature-assisted” synergistic mechanism for permeability enhancement under the tested conditions. The experimental data demonstrate that the absolute increase in permeability driven by elevated injection pressure (e.g., from 0.54 to 2.37 mD between 6 and 10 MPa at 150 °C) was substantially greater than that induced by temperature increase alone (e.g., from 0.49 to 1.18 mD between 80 and 150 °C at 8 MPa). This supports the dominant role of pressure in creating and interconnecting macro-fractures, establishing efficient percolation pathways. Temperature primarily assists by promoting gas desorption and facilitating the initiation of micro-fractures through thermal stress.

(4)

Experimental results demonstrate a clear temperature–pressure synergy during CO₂ displacement. At 150 °C, permeability increased by 339% as pressure rose from 6 to 10 MPa, while displacement efficiency improved by only 2.4% beyond 8 MPa, indicating a pressure-driven limitation. Thermal activation effectively overcomes this constraint by enhancing CO₂ diffusion and providing activation energy for CH₄ desorption from isolated micropores, thereby mobilizing residual gas unreachable by pressure alone. This finding establishes a new theoretical framework for efficient coalbed methane extraction, challenging conventional pressure-centric approaches.

Limitations and Future Work: The conclusions of this study are drawn within the scope of the presented experimental system. It should be noted that this work did not incorporate numerical modeling, comparative analysis with other enhancement techniques (e.g., cyclic injection, N₂ flooding), or an energy balance assessment. These aspects represent important avenues for future research to further generalize and optimize the proposed mechanism.