Experimental Study on CO₂–Crude Oil Phase Behavior, Miscibility and Threshold Pressure Gradient for CCUS in Ultra-Low Permeability Reservoir

Abstract

The Chang 8 reservoir in the Baima Middle Block of the Ordos Basin is an ultra-low-permeability reservoir that has faced formation-pressure depletion, a rising water cut, and poor remaining-oil recovery during long-term water-injection development. CO₂ flooding in CCUS can simultaneously enhance oil recovery and achieve carbon sequestration; however, comprehensive and integrated research on the CO₂–crude oil phase behavior, miscibility characteristics, and nonlinear flow in porous media for this block is still lacking. To address this gap, this study used crude oil and natural cores from the block and natural cores to conduct a series of experiments at 70 °C, including PVT high-pressure physical-property tests, CO₂ constant-mass-expansion tests, slim-tube minimum-miscibility-pressure (MMP) tests, pendant-drop oil–gas interfacial-tension measurements, and CO₂-flooding threshold-pressure-gradient measurements in cores of different permeability. The results showed that as the CO₂ mole fraction increased from 0% to 61.78%, the bubble-point pressure of the crude oil rose from 17.424 to 22.575 MPa, the viscosity decreased from 9.026 to 4.906 mPa·s, and the volumetric expansion coefficient increased from 1.0000 to 1.1454. As the water cut increased from 0% to 80%, the apparent MMP of CO₂ increased from 15.83 to 17.69 MPa. The oil–gas interfacial tension gradually decreased with increasing system pressure, and droplet instability and atomization occurred under near-miscible high-pressure conditions. The CO₂-flooding threshold pressure gradient decreased with core permeability following a power-law relationship, and a knee-like trend appeared near 0.5 mD within the limited permeability dataset of this study; this trend cannot be generalized without further experimental evidence. This study comprehensively reveals the coupled mechanisms of phase behavior, miscibility, and nonlinear flow during CO₂ flooding in the target block, providing experimental and theoretical support for optimizing the injection pressure and reservoir-stimulation schemes of CCUS-EOR in ultra-low-permeability reservoirs.

1. Introduction

Against the backdrop of global climate-change governance and the transition toward low-carbon development, carbon capture, utilization, and storage (CCUS) has become a key technology pathway for China to achieve its “dual carbon” goals while ensuring the clean and efficient development of fossil fuels [1,2]. As an important application form of the CCUS system, CO₂-enhanced oil recovery (CO₂-EOR) exploits the interaction between CO₂ and formation crude oil to improve oil recovery and, at the same time, to achieve the deep geological storage of CO₂, which delivers both economic benefits for resource development and environmental benefits through carbon reduction. The Ordos Basin contains abundant low-permeability to ultra-low-permeability oil and gas resources and is surrounded by numerous high-emission enterprises, such as coal-chemical and thermal-power plants, which can provide stable CO₂ sources [3,4,5]. The close match between the CO₂ sources and the reservoir space makes the basin an advantageous area for large-scale CCUS engineering applications in China [6,7,8,9,10,11].

However, the CCUS development of low-permeability and ultra-low-permeability reservoirs differs substantially from that of conventional medium- and high-permeability reservoirs and faces additional technical challenges. Such reservoirs are characterized by complex pore structures, fine pore throats, high flow resistance, and strongly nonlinear fluid-migration behavior, which greatly increase the difficulty of mixing CO₂ with crude oil and impose stringent requirements on the reservoir conditions. The oil reservoirs of the Changqing Oilfield in the Ordos Basin generally exhibit the “three lows” of low permeability, low pressure, and low abundance [12,13,14]. Because of the insufficient natural energy of the formation, these reservoirs rely on long-term water injection to supplement energy and maintain stable production. The main oil-bearing interval of the Baima Middle Block investigated in this study is the Chang 8 reservoir, with an average effective thickness of 11.2 m, an average effective porosity of 10.5%, and an average permeability of only 1.21 mD, which is a typical ultra-low-permeability reservoir. During long-term water-injection development, the block has progressively encountered problems such as uneven formation-pressure distribution, a continuously increasing water cut, and difficulty in stabilizing production [15,16,17,18,19], making it difficult for the traditional water-flooding development model to effectively exploit the remaining-oil potential. Therefore, CO₂ displacement development based on CCUS technology, together with an accurate understanding of the phase-behavior laws, miscibility characteristics, and porous-media flow mechanisms of the CO₂-crude oil system, is a key prerequisite for optimizing injection parameters and improving development performance in the block [20,21].

Extensive experimental research on CO₂ flooding has been carried out worldwide, and substantial progress has been made in individual aspects, including the high-pressure physical properties of crude oil, the minimum miscibility pressure (MMP), and the permeability characteristics of reservoirs [22,23,24,25]. With regard to the high-pressure physical properties of crude oil, PVT experiments have clarified the influence of CO₂ dissolution on physical parameters such as viscosity, the volume coefficient, and the dissolution characteristics of crude oil. With regard to miscibility, slim-tube experiments and bubble-point tests are the mainstream methods for determining the MMP of CO₂–crude oil systems [26,27,28,29,30,31,32]; for example, previous studies in the Changqing Oilfield block measured the MMP of the regional crude oil as 22.12 MPa. With regard to flow mechanisms, it has been confirmed that CO₂ flooding in low-permeability reservoirs exhibits a significant threshold pressure gradient, and the seepage curve presents a two-stage “curve segment + straight-line segment” characteristic [33,34,35,36,37]; in addition, the oil-swelling and viscosity-reduction effects caused by the dissolution of CO₂ in crude oil can effectively improve the fluid seepage capacity. Nevertheless, several gaps remain in existing research. First, most experimental studies are fragmented, with results obtained from different blocks, rock cores, and crude-oil samples, and systematic and integrated testing data for the same target reservoir with the same set of core and fluid samples are lacking. Second, there is little research establishing the intrinsic relationship between the evolution of the crude-oil phase and the seepage response in porous media, and in particular, quantitative comparisons of the threshold pressure gradient between pure CO₂ flooding and CO₂-water flooding are scarce. As a result, the influence of the injection-medium type on the threshold pressure gradient remains unclear, and accurate theoretical support for the optimization of parameters such as the injection-production well spacing and the injection pressure in ultra-low-permeability reservoirs cannot be provided, which seriously restricts the refined design and efficient implementation of the CCUS-EOR scheme in the Baima block [38,39,40,41,42,43].

To fill the above gaps, this study took the Chang 8 reservoir in the Baima Middle Block of the Ordos Basin as the research object and, using on-site crude oil and natural core samples, systematically conducted high-pressure PVT physical-property tests, CO₂ expansion experiments, slim-tube MMP tests, pendant-drop interfacial-tension measurements, and threshold-pressure-gradient experiments on natural cores. The results reveal the phase evolution, miscibility mechanism, and nonlinear flow law of CO₂ flooding in ultra-low-permeability reservoirs, providing experimental and theoretical support for the optimization of CCUS-EOR injection schemes in the target block.

2. Experimental Apparatus and Steps

2.1. CO₂–Crude Oil High-Pressure Physical Property Experiments

CO₂–crude oil high-pressure physical property experiments were conducted to determine the basic PVT parameters of the crude oil in the Baima Middle Block formation. According to the GB/T26981-2020 standard [44], the processed crude oil was mixed with gas to a gas–oil ratio of 77.59 m³/m³ (49.5% CH₄, 15.5% C₂H₆, 17.9% C₃H₈, etc.), and single degassing, constant mass expansion, and viscosity tests were carried out at 70 °C and 21.3 MPa to obtain key physical property indicators such as the fluid composition, bubble point pressure, and volume coefficient.

2.1.1. Single Degassing Experiment

The single degassing experiment used a high-temperature and high-pressure PVT analyzer to flash the formation fluid from reservoir temperature and pressure to atmospheric conditions. The volumes of the separated gas and oil were measured with the separation bottle and the gas meter, and the chromatographic data of the single-degassed gas and liquid phases were obtained (as shown in Figure 1). Parameters such as the formation-fluid composition, fluid volume coefficient, density, and gas-oil ratio were then calculated.

2.1.2. CO₂ Injection Constant Mass Expansion (CCE) Experiment

The formation fluid was sealed in a PVT container under formation temperature and pressure, and the pressure–volume relationship of the constant-mass fluid was tested at constant temperature to obtain the saturation pressure, relative volume, single-phase oil density, Y function, and isothermal compressibility coefficient. As shown in Figure 2, gas was injected into the fluid in multiple stages, and the pressure was raised each time so that the gas was completely dissolved; the saturation pressure, solution gas–oil ratio, volume coefficient, viscosity, density, and fluid composition corresponding to different injection volumes were measured.

CO₂ injection ratios of 20 mol%, 40 mol%, and 60 mol% were used to conduct expansion experiments on the composite crude-oil samples, and the changes in the PVT phase parameters of the formation crude oil under the three CO₂ injection ratios were clarified.

2.2. Minimum Miscibility Pressure Experiment

2.2.1. Slim-Tube Experiment Instruments and Scheme

The slim-tube experiment mainly used high-pressure displacement pumps, intermediate containers, thin tubes, constant temperature chambers, back pressure pumps, etc. At the same time, some measuring devices such as pressure gauges, thermometers, and flow meters were also used (Figure 3). The thin tube used stainless steel coil with an inner diameter of 4 mm, a length of 12 m, and a sand particle range of 200 mesh. The detailed parameters are shown in

Table 1. Slim-tube parameters.

Slim-Tube Parameters	Maximum Working Pressure	Temperature	Length	Outer Diameter	Inner Diameter	Permeability
Value	70 MPa	200 °C	1200 cm	0.6 cm	0.4 cm	1 D

.

Slim-tube experiments were conducted with the composite crude oil under simulated formation conditions at 70 °C to determine the minimum miscibility pressure of the crude oil and CO₂. The experimental temperature was maintained at 70 °C, the experimental pressure ranged from 12 to 21 MPa, and the water cut ranged from 0% to 80%. The detailed parameters are shown in

Table 2. Slim-tube experiment conditions.

Water Cut	Experimental Pressure (MPa)
0%	12	13.5	15	16.5	18	19.5	/
20%	12	13.5	15	16.5	18	19.5	/
60%	12	13.5	15	16.5	18	19.5	21
80%	12	13.5	15	16.5	18	19.5	21

.

2.2.2. Experimental Steps of Slim-Tube Experiment

(1)

Preparation of the slim tube and airtightness test: Sand with a mesh size of 160–200 was sieved and dried, and the clean, dry slim tube was filled with the sand under continuous vibration to ensure even and dense packing. The experimental pipeline was connected, and high-pressure nitrogen was introduced for 1 h; if the pressure drop was ≤0.15 MPa, the airtightness was considered qualified.

(2)

Evacuation and back-pressure setting: A vacuum pump was used to evacuate the slim tube and the downstream pipeline until the vacuum degree at the front and rear ends of the slim tube was consistent, and the pressure was held stable for 6 h. The back-pressure pump was then adjusted manually to set the back-pressure valve to the target experimental pressure

(3)

Oil-sample saturation: Under constant-temperature conditions, the oil sample was transferred from the sampler to an intermediate container and stabilized to the experimental pressure with a high-pressure displacement pump. The valve was then slowly opened, and the water in the slim tube was displaced with the oil sample at a low flow rate of 0.1 mL/min to establish a water-cut system of 0–80%.

(4)

CO₂ oil displacement experiment: CO₂ oil displacement was injected under constant temperature, constant pressure, and constant flow rate of 0.1 m/min; the oil production, displacement pump readings, and inlet and outlet pressures of the thin tube were recorded for every 0.1 PV injection. When the cumulative gas injection was ≥1.4 PV or oil production was completely stopped, the displacement was terminated.

(5)

Reset and repetition: CO₂ was removed from the pipeline, and the slim tube and accessories were disassembled, cleaned, and dried (or replaced); the above steps were then repeated for the next pressure point.

(6)

Recovery rate calculation: After injecting 1.4 PV CO₂, the final crude oil recovery rate was calculated by dividing the volume of degassed crude oil extracted by the volume of saturated oil sample.

2.3. Experiment on Interfacial Tension of CO₂ and Crude Oil at Different Pressures

Pendant-Drop Method: Procedure and Principle

Measure the interfacial tension between crude oil and CO₂ using the hanging drop method. After cleaning and vacuuming the high-pressure hanging drop kettle, fill it with CO₂ at a constant temperature, gradually increase the pressure, and let it stand stable. Use a high-pressure micro pump to extrude 1–3 mm³ of crude oil at low speed, forming symmetrical suspended droplets at the probe tip. After the gas–liquid two-phase equilibrium, capture the droplet profile.

By using the empirical formula for calculating interfacial tension using the Andreas hanging droplet method (a general engineering formula for measuring high-pressure interfacial tension in petroleum), gravity and interfacial tension cancel each other out when the hanging droplet is in equilibrium. The dimensionless coefficient H is obtained from the characteristic size d_s and d_e of the droplet by looking up a table. Combined with parameters such as droplet volume and two-phase density difference, the interfacial tension between CO₂ and crude oil can be calculated by substituting them into Formulas (1) and (2). The principle is shown in Figure 4. The experimental temperature is maintained at 70 °C and the experimental pressure is 2–12 MPa. The droplet morphology image is collected by a camera system, and the oil–gas interfacial tension is calculated through image processing and analysis.

$$H=f(d_g/d_e)$$

(1)

$$\sigma =∆\rho d_c^{2}g/H$$

(2)

where σ denotes oi–gas interfacial tension, mN/m; $∆\rho$ is the density difference between gas and liquid, g/cm³; d_e represents the maximum diameter of the droplet, mm; d_g is the droplet diameter at the selected interface, mm; g stands for gravitational acceleration, cm/s²; H is the shape correction factor, a function related to d_g and d_e; f is the fitting coefficient.

2.4. CO₂ Drive Start Pressure Gradient Experiment

2.4.1. Experimental Steps and Plan

(1)

Experimental plan

The experiment used high-temperature and high-pressure rock core displacement equipment, selecting rock cores with different permeability ranging from 0.1 to 10 mD. The rock cores taken in the experiment were all natural rock cores from the Baima Middle Block, with a length of about 6 cm. The experimental temperature was maintained at 70 °C, and CO₂ displacement was used as the displacement method, with an injection rate of 0.01–1.5 mL/min and a water content of 60%, as shown in

Table 3. Experimental plan for CO₂ drive start pressure gradient.

Experimental Temperature (°C)	Water Cut (%)	Penetration Rate (mD)	Inject Medium
70	60	0.1	CO2 displacement
		0.3
		0.5
		1
		10

.

(2)

Experimental steps

−

Pre-treatment before the experiment

Select ten natural rock cores with different permeabilities, and measure porosity and permeability after high-temperature oil washing and drying. Load the core into the gripper and apply a constant confining pressure, then vacuum and saturate the formation water for at least 6 h. After saturation, oil-drive the core at a flow rate of 0.1 mL/min, stop the pump after the effluent state stabilizes, and age at 70 °C for 48 h to establish the bound water saturation of the core.

After the core preparation is completed, water flooding was carried out at a flow rate of 0.1 mL/min. After the water content stabilized at 60%, the flooding was stopped and aged for 6 h to complete the experimental preparation.

−

CO₂ flooding start-up pressure gradient test

Load CO₂ into an intermediate container and inject it into the core at a constant flow rate of 0.5 mL/min. Monitor the inlet and outlet pressures and obtain the minimum start-up pressure difference for CO₂ flooding after the pressure stabilizes. Set four levels of back pressure, namely, 2 MPa, 4 MPa, 6 MPa, and 8 MPa, and test the stable displacement pressure difference under each back pressure to determine the starting pressure under different operating conditions. Calculate the corresponding CO₂ drive starting pressure gradient.

2.4.2. Experimental Principle

The experimental measurement was conducted using a combination of flow pressure difference method and unstable equilibrium method. Firstly, the core was placed in a stable seepage state, and the seepage velocity at both ends of the core was changed. After the displacement pressure difference had stabilized, the displacement pressure difference at both ends of the core was measured at different seepage velocities to draw a flow–pressure gradient relationship curve. By reverse extension, the intercept of the curve on the pressure gradient coordinate axis was found, which was the proposed starting pressure gradient of the core, as shown at point B in Figure 5. After the pressure gradient measurement was completed, the inlet end of the rock core was sealed to allow the pressure difference between the two ends of the rock core to naturally decrease. When the final pressure stabilized, the pressure difference between the two ends of the rock core was the true starting pressure of the rock core, and the corresponding pressure gradient was the true starting pressure gradient of the rock core, as shown at point A in Figure 5.

3. Results and Discussion

3.1. CO₂–Crude Oil High-Pressure Physical Property Analysis

3.1.1. Relationship Between Different CO₂ Injection Amounts and Bubble Point Pressure

Through the constant mass expansion (CCE) experiment and CO₂ injection expansion experiments, as shown in the following figure, it was found that as the CO₂ injection ratio increased, the bubble point pressure of crude oil gradually increased, and the fluid degassing expansion effect was significantly enhanced when the pressure was lower than the bubble point.

As shown in Figure 6, the crude oil without gas injection is in a single state under formation pressure (a). When the crude oil reaches the bubble point pressure, small bubbles are released (b), and the recorded bubble point pressure is 17.424 MPa. After the first stage of CO₂ injection (21.16%), the bubble point pressure of the blended crude oil increases to 17.690 MPa. The bubble point pressure of the second stage CO₂ injection (40.88%) is 19.951 MPa. The third stage CO₂ injection (61.78%) has a bubble point pressure of 22.575 MPa.

The inherent mechanism is that the large amount of CO₂ dissolved in crude oil increases the dissolved gas content of the system, thereby increasing the bubble point pressure. When the formation pressure is lower than the bubble point, the CO₂ system is prone to produce a large amount of free gas, which changes the fluid flow characteristics of the reservoir. This has important reference significance for maintaining formation pressure and optimizing production systems in the CO₂ flooding development process.

3.1.2. Phase Diagram Analysis of Crude Oil with Different Gas–Liquid Ratios and CO₂ Injection Amounts

Crude oil prepared with different gas–liquid volume ratios, as shown in Figure 7, has a gas–liquid two-phase zone inside the curve, where each ratio curve intersects at the same critical point; As the gas–liquid volume ratio increases from 0.6 to 1.0, the overall phase envelope expands outward, and the temperature and pressure range of two-phase coexistence continues to increase. At the same temperature, the required pressure for phase separation is higher, and the phase boundaries of each curve in the low-temperature range differ significantly. As the critical temperature approaches, the curves gradually converge. The increase in the proportion of gas phase increases the probability of gas–liquid two-phase formation in crude oil, which can easily cause two-phase seepage in the reservoir. This law can provide a basis for reservoir pressure control and gas-injection drive-phase state control.

Figure 8 shows the temperature–pressure phase-envelope diagram of crude oil systems with different CO₂ molar injection ratios. The interior of the curve is the gas–liquid two-phase region, and the triangle marks the critical points of each system. As shown in the table, with the increase in CO₂ mole fraction from 0% to 61.78%, the bubble point pressure of the formation crude oil increases, the volume expands, and the viscosity decreases, indicating that CO₂ can effectively supplement energy, improve the flow capacity of crude oil, and provide basic parameters and relevant basis for CO₂ injection development of the target reservoir; The phase envelope expands outward, and the temperature and pressure range of the coexistence of the two phases continues to widen. The critical point shifts synchronously towards high temperature and high pressure, and the regulatory effect of CO₂ on the phase state is more significant in the low temperature range. The dissolution of CO₂ increases the gas-liquid phase separation range of crude oil, which can easily cause two-phase flow in the reservoir, providing a phase basis for maintaining formation pressure and controlling phase mixing in CO₂ flooding (see

Table 4. PVT analysis data table for CO₂ injection.

Stage	CO2 Injection Mole Fraction (mol%)	Bubble Point Pressure (MPa)	Solution Gas–Oil Ratio (m3/m3)	Oil Formation Volume Factor at Bubble Point Pressure	Density at Bubble Point Pressure (g/cm3)	Viscosity (mPa·s)	Volume Expansion Coefficient
No Gas Injection	0.00	17.424	77.59	1.1963	0.7761	9.026	1.0000
Stage 1	21.16	17.690	93.70	1.2698	0.7713	7.824	1.0534
Stage 2	40.88	19.951	124.08	1.3615	0.7735	6.227	1.1015
Stage 3	61.78	22.575	161.37	1.4596	0.7777	4.906	1.1454

).

3.2. Analysis of Minimum Mixed-Phase Pressure in Slim-Tube Experiment

Analysis of Minimum Mixed-Phase Pressure of CO₂ Under Different Moisture Content Conditions

Using a thin-tube model experiment to calculate the mixed-phase pressure, thin-tube displacement experiments were conducted at different pressure points of 12.0 MPa, 13.5 MPa, 15.0 MPa, 16.5 MPa, 18.0 MPa, 19.5 MPa, and 21.0 MPa with a formation temperature of 70 °C and water content of 0%, 20%, 60%, and 80%. By measuring the oil displacement efficiency under different water content conditions and fitting the data (as shown in Figure 9), the minimum miscibility pressure under different water content conditions was obtained.

As shown in

Table 5. Minimum miscibility pressure under different displacement pressures and water-content conditions.

Displacement Pressure (MPa)	Water Content
	0%	20%	60%	80%
	Oil Displacement Efficiency (%)
12	55.2	54.8	52.7	49.6
13.5	70.5	60.2	60.5	57.2
15	80.3	78.7	76.3	73.6
16.5	89.7	88.2	80.3	77.5
18	90.1	88.5	87.8	87.5
19.5	91.4	89.7	88	87.9
21	/	/	88.5	88.2
MMP (MPa)	15.83	16.44	17.31	17.69

, when the experimental pressure is lower than the apparent MMP, the oil-displacement efficiency increases significantly with increasing experimental pressure. When the experimental pressure exceeds the apparent MMP, the experimental pressure has no significant influence on the oil-displacement efficiency. As the water cut gradually increases from 0% to 80%, the apparent minimum miscibility pressure (apparent MMP) obtained from slim-tube displacement-efficiency fitting rises monotonically from 15.83 MPa to 17.69 MPa. This elevated apparent MMP originates from complicated multiphase-flow effects inside the slim-tube system under high water-cut conditions, rather than representing an increase in the intrinsic thermodynamic critical miscibility pressure of the pure CO₂–crude-oil fluid system.

Pressure < MMP (non-miscible drive): There is a clear gas–liquid interface between CO₂ and crude oil, which relies on extraction and evaporation to achieve mass transfer of light components. The interface tension is high, and a large amount of residual oil is retained by capillary action in the pores and throats. As the pressure increases, the density of CO₂ increases, the interfacial tension between oil and gas decreases, the ability of CO₂ to extract crude oil increases, and the oil washing efficiency continues to improve.

Pressure ≥ MMP (mixed phase drive): CO₂ and crude oil reach dynamic mixing, the interface between the two phases disappears, there is no capillary binding, and the crude oil in the matrix pores can be fully driven out. When continuing to increase pressure, the differences in oil and gas composition are no longer significant, and the recovery rate is no longer significantly improved.

3.3. Analysis of Interfacial Tension Between CO₂ and Crude Oil

By using a camera system to capture images of droplet morphology under different pressures, and after settling into two-phase equilibrium, a suspended droplet contour image was collected (as shown in Figure 10). As the pressure increased from 2 MPa to 12 MPa, the oil droplets became full and regular, and the interfacial tension under that pressure was measured by parameters such as the volume of the suspended droplet.

The interfacial tension between oil and gas was calculated through image processing and analysis (see

Table 6. CO₂ crude oil interface tension at different pressures.

Temperature °C	Pressure (MPa)	CO2 Density (kg/m3)	Interfacial Tension (mN/m)
70	2	63.012	19.69
	3	63.012	16.59
	4	63.012	15.51
	5	100.14	14.02
	6	100.14	12.50
	7	142.15	10.95
	8	142.15	10.45
	9	189.99	7.72
	10	189.99	6.25
	11	244.29	4.84
	12	244.29	4.48

). As the experimental pressure continued to rise, the density of CO₂ continued to increase, and the interfacial tension between CO₂ and crude oil gradually decreased. As the pressure increased from 2 MPa to 10 MPa, the contour of the crude oil droplet was smooth and fully pear shaped, and the bottom of the droplet was rounded without stretching, shrinkage, or dispersion. The overall volume of the droplet was stable, the boundary was clear and complete, and there was no significant elongation of the droplet below the needle tip; As the pressure increased to 18 MPa, the bottom of the droplet became completely unstable, and the crude oil dispersed and atomized downwards from the tip of the needle. The original intact hanging droplet shape was completely broken, and stable and intact droplets could not be formed, with flocculent dispersed fluid appearing. No direct experimental measurements of interfacial tension were available for the pressure range of 12–18 MPa. Relevant values in this interval were derived from polynomial extrapolation based on valid experimental data from 2–12 MPa and should not be interpreted as experimentally observed decreasing trends.

Under high pressure, a large amount of CO₂ dissolved in crude oil, and the properties of the oil phase and CO₂ gas phase components were significantly similar. The interfacial energy between the two phases sharply decreased, and the interfacial tension dropped to an extremely low level (near mixed-phase interface state). When the interfacial tension was too low, a stable liquid gas interface could not be maintained, and droplets could not maintain a complete suspended droplet structure, resulting in shrinkage, atomization, and dispersion. That pressure was close to the minimum miscible pressure of CO₂ and crude oil, and the two phases tended to be miscible.

3.4. Analysis of CO₂ Drive Pressure Gradient

We selected rock cores with different permeabilities ranging from 0.1 to 10 mD, obtained data through experiments, and used the flow pressure difference method to fit the curve. Then, we determined the magnitude of the proposed and actual starting pressure gradients for water/CO₂ flooding under different permeability conditions.

CO₂ Flooding Start-Up Pressure Gradient Under Different Permeability Conditions

➢

Analysis of the starting pressure gradient of CO₂ flooding in rock cores with different permeabilities:

As shown in Figure 11, different permeability cores of 0.1 and 10 mD were selected, and data were obtained and fitted through experiments. The permeability was 0.1 mD, and the simulated pressure gradient for CO₂ flooding was 0.2416 MPa/cm, with a true pressure gradient of 0.0212 MPa/cm. The permeability was 10 mD, the simulated pressure gradient for CO₂ flooding was 0.1260 MPa/cm, and the actual pressure gradient was 0.0021 MPa/cm (see

Table 7. Starting pressure gradient of CO₂ flooding for different permeability rock cores.

Penetration Rate (mD)	Proposed to Initiate Pressure Gradient (MPa/cm)	Threshold Pressure Gradient (MPa/cm)
0.1	0.2416	0.0212
0.3	0.2603	0.0138
0.5	0.2103	0.0073
1	0.2333	0.0050
10	0.1260	0.0021

).

As shown in Figure 12, the starting pressure gradient of CO₂ flooding decreases as a power law with increasing permeability. A knee-like trend can be observed near 0.5 mD based on the limited number of permeability data points in this study. This should not be interpreted as a well-defined universal critical threshold. For cores with permeability lower than 0.5 mD in our tests, the threshold pressure gradient rises sharply as permeability decreases. The pore throats of the core are small, and the pore properties are poor. The pore walls adsorb crude oil, CO₂ and the interfacial tension between crude oil produce strong capillary resistance, and the near wall viscous boundary layer almost blocks most of the flow channels. When the permeability is greater than 0.5 mD, the pressure gradient tends to flatten, the pore space is spacious enough, and the capillary and adsorption binding effects weaken. The flow resistance of CO₂ mainly comes from the viscosity of the gas itself and fluid friction. When continuing to increase the permeability, the drag reduction effect brought by the improvement of the flow channel decreases marginally, the starting pressure gradient slowly approaches zero, and the curve tends to flatten.

4. Conclusions

This study focused on the ultra-low-permeability Chang 8 reservoir in the Baima Middle Block of the Ordos Basin. Through fluid-property testing and core-displacement experiments using natural cores from the study area, the phase behavior, miscibility, and nonlinear-flow characteristics of CO₂ flooding were systematically analyzed. The results provide technical guidance for the CCUS-EOR development of this block and similar ultra-low-permeability reservoirs.

• The injection ratio of CO₂ directly affected the physical properties and displacement effect of crude oil. As the molar fraction of CO₂ injection increased from 0 to 61.78%, the system bubble point pressure rose to 22.575 MPa, the volume expansion coefficient increased to 1.1454, and the viscosity of single-phase crude oil decreased to 4.906 mPa·s. This confirms that CO₂ dissolution has significant solubilization, expansion, and viscosity reduction effects on formation crude oil, expands crude oil, and broadens the gas–liquid two-phase zone. On site, it is necessary to maintain the formation pressure above the bubble point pressure to avoid the influence of free gas on the displacement effect.
• The results of the slim-tube experiment showed that an increase in moisture content significantly increased the difficulty of mixing, and the minimum mixing pressure continued to increase with the increase in moisture content, rising from 15.83 MPa under 0% moisture content to 17.69 MPa under 80% moisture content.
• The measurement of interfacial tension under different pressures using the hanging drop method showed that the oil–gas interfacial tension gradually declined with the increase in pressure. At low pressure of 2–10 MPa, crude-oil droplets were full and regular. When the pressure approached the apparent minimum miscibility pressure at high pressure, the droplets broke and atomized. Under high-pressure conditions, a large amount of CO₂ dissolved into crude oil, the physical properties of the two-phase fluids converged, the interfacial energy decreased significantly, and it was difficult to maintain a stable oil–gas interface. Low pressure 2–10 MPa crude oil droplets were full and regular. When the high pressure approached the minimum mixed-phase pressure, the droplets broke and atomized. Under high pressure, a large amount of CO₂ was dissolved in crude oil, and the physical properties of the two phases converged. The interfacial energy was significantly reduced, and the interfacial tension almost disappeared, making it difficult to maintain a stable oil–gas interface.
• The core-flooding experiments demonstrated that the threshold pressure gradient decreased with permeability following a power-law relationship. Within the limited dataset of this study, a knee-like trend appeared near 0.5 mD: for the measured samples with K < 0.5 mD, the threshold pressure gradient increased sharply as the permeability decreased, whereas for the samples with K > 0.5 mD, the decreasing rate became much slower, and the gradient gradually approached zero. This practical threshold was derived from the present core-flooding measurements and requires further experimental data for broader validation