Study on Utilization Boundaries and Contributions of Pore Throats of Different Scales in Low-Permeability Reservoirs

Abstract

Low-permeability sandstone oil reservoirs, as an important type of oil and gas resource, feature high reservoir density and low permeability. The utilization of pore throats of different scales during their development process is crucial for enhancing oil recovery. Based on nuclear magnetic resonance and CT scanning techniques, this paper systematically studies the utilization limits and energy contribution of pore larynx under different displacement methods. The results show that during the water injection development process, the main pore–throat radius used by water flooding is between 1 and 20 μm. Among them, the contribution of the small pore tends to stabilize after the pressure rises to a certain stage, the contribution of the medium pore increases with the rise in pressure, while the contribution of the large pore gradually decreases with the increase in pressure. After switching to CO₂ gas flooding, the application range of the pore throat was further expanded to a smaller scale. The contribution of the small pore and the middle pore significantly increased in a specific pressure range, while the large pore made a greater contribution at a lower pressure. This paper has certain reference significance for the study of the limit and contribution of pore–throat exploitation in low-permeability sandstone oil reservoirs.

1. Introduction

As conventional oil and gas field resources gradually decline, low-permeability sandstone reservoirs, as an important source of oil and gas, account for approximately 40% of the total oil and gas resources, demonstrating significant development potential [1,2,3,4]. However, due to the reservoir’s low porosity and permeability characteristics, many technical challenges arise during extraction. The pore–throat structure characteristics within the reservoir directly affect the flow behavior of oil and gas, which in turn determines the recovery efficiency of the reservoir. Therefore, for the effective development of low-permeability reservoirs, it is crucial to identify the utilization boundaries of pore–throats at different scales and their contribution during the development process, as this is key to improving recovery rates and optimizing development plans [5,6,7].

Nuclear magnetic resonance (NMR) is a powerful tool for quantitatively characterizing the internal pore structure of rocks and is widely used in laboratory research. Only fluids within the pore–throats of cores generate NMR signals, which provide insight into the distribution of pore fluids [8,9,10,11,12]. Researchers, both domestically and internationally, have established a strong mathematical relationship between NMR relaxation time and pore–throat radius. Using high-pressure mercury injection data, the relaxation time can be directly converted into the pore–throat radius via a conversion factor. Several studies have extensively explored the correlation between relaxation time and pore radius. By applying the pore–throat ratio, the pore radius can be converted into the throat radius. These NMR pore–throat radius distribution curves have been incorporated into oilfield development evaluations, where the accuracy and effectiveness of NMR-based analysis of core samples from low-porosity, low-permeability reservoirs have been discussed [13,14,15]. In a study by Wang et al., nanostructure morphology and pore size distributions (PSDs) of 10 samples from the Lower Cambrian Niutitang Formation in northwestern Hunan were investigated using field emission scanning electron microscopy (FE-SEM), high-pressure mercury intrusion (HPMI), low-pressure nitrogen gas adsorption (LP-N2GA), and carbon dioxide gas adsorption (LP-CO₂GA). In combination with the geochemical parameters and mineral composition, the factors influencing the nanoscale pore structure were analyzed [16]. In the work of Zhu et al., through a series of rock physics experiments, including gas measurement of porosity and permeability, casting thin sections, scanning electron microscopy, and high-pressure mercury injection, the quality of reservoir properties and microscopic pore–throat structure characteristics were systematically studied. Combined with fractal geometry theory, the effects of different pore–throat types, geometric shapes and scale sizes on the fractal characteristics and heterogeneity of sandstone pore–throat structure are clarified [17]. While previous studies have primarily focused on the distribution of pore–throat and pore radii across different types of reservoirs using NMR technology, there has been limited research on the activation thresholds and contributions of pore–throats under varying development conditions, such as during water flooding and gas injection.

In the gas injection phase following water flooding, the activation scale and degree of pore–throat utilization in core samples are critical factors influencing oil displacement efficiency [18,19,20,21,22]. Nuclear magnetic resonance (NMR) technology, an effective tool for reservoir characterization, provides quantitative data on the fluid distribution within pore–throats. This makes NMR essential for studying pore–throat activation during the gas injection process after water flooding. Research indicates that crude oil trapped in micron-scale pores is the primary contributor to water flooding recovery [5,23,24,25]. Moreover, variations in gas injection rates and pressures lead to noticeable changes in both the activation threshold and the degree of pore–throat utilization [26,27,28,29,30].

This paper investigates the tight oil reservoir of the Permian Wutonggou Formation in the Shanan Oilfield. By combining nuclear magnetic resonance (NMR) testing and CT scanning experiments, it analyzes the variations in NMR signal amplitudes from different pore–throat radii during the gas injection process following water flooding [31,32,33]. The study identifies the activation boundaries of pore–throats under various development methods. Using high-temperature, high-pressure physical flow simulation experiments, real-time monitoring of fluid distribution within core samples is conducted during the displacement process. The research examines the activation thresholds and contribution patterns of different-scale pore–throats under water injection and CO₂ injection, providing theoretical insights for studies on pore–throat activation boundaries and contributions in low-permeability sandstone reservoirs [34,35].

2. Experimental Process

2.1. Experimental Materials

Six core samples from the Wutonggou Formation sandstone of the Shaqiu 5 Well Area in the Junggar Basin were chosen for the study. The basic physical properties of the cores are summarized in

Table 1. Information core samples.

Core ID	Depth/m	Stratigraphy	Length/cm	Diameter/cm	Porosity/%	Permeability × 10−3 μm2
1	1742.2	P3wt13−2	5.05	2.51	13.55	0.94
2	1952.3	P3wt13−2	5.03	2.50	14.11	0.89
3	2322.1	P3wt13−2	5.02	2.52	13.71	2.85
4	2311.3	P3wt13−2	5.00	2.51	13.86	1.26
5	1895.6	P3wt13−2	5.03	2.50	13.89	1.18
6	2031.1	P3wt13−2	5.04	2.50	13.87	1.09

. The porosity ranged from 13.55% to 14.11%, and the permeability varied between 0.89 mD and 2.85 mD, typical of low-permeability sandstones with strong heterogeneity. The crude oil samples were taken from the same stratigraphic layer as the cores, and the experimental oil was prepared by mixing crude oil with refined kerosene at a 1:3 volume ratio. The viscosity of the degassed oil at 50 °C was 7.71 mPa·s, with an average density of 0.8446 g/cm³. The formation water used in the experiments was of the CaCl₂ type, with an average chloride (Cl⁻) concentration of 4222.53 mg/L and a total salinity of 7646.41 mg/L.

2.2. Experimental Equipment

The experimental setup for displacement is shown in Figure 1. The displacement pump used is an ISCO-260D (Teledyne ISCO Company, Lincoln, NE, USA) model, made in the USA, with a pressure range from 0 to 51.7 MPa and a continuous flow rate range from 0.001 to 80 mL/min. The vacuum pressurization saturation device is the NM-V model, with a working pressure range of 0 to 40 MPa. All of the equipment was manufactured by Jiangsu Huaxing Petroleum Instruments Co., Ltd. (Nantong, China). The nuclear magnetic resonance (NMR) instrument was manufactured by Suzhou Newmai Analytical Instruments Co., Ltd. (Suzhou, China), and its model is PQ001.

2.3. Experimental Principle

This experiment integrates low-field nuclear magnetic resonance (NMR) technology with CT scanning methods to develop a mathematical model that converts pore–throat radius to NMR T₂ spectra. This model allows for the precise calculation of the upper and lower boundaries of pore–throat radii in core samples after water flooding and gas injection. It also enables the quantitative evaluation of pore–throat scale limits under various displacement methods in low-permeability sandstone reservoirs.

In the NMR spectrum, the T₂ relaxation time on the horizontal axis is positively correlated with the pore–throat radius. The signal intensity at each point reflects the corresponding pore–throat volume and fluid distribution at that scale. The transverse relaxation time of fluids in porous media can be expressed as follows:

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

(1)

${\rho }_2$ represents the transverse relaxation intensity in μm/ms. S is the surface area of the pores (in cm²), and V denotes the pore volume.

The pore radius is the product of the pore–throat ratio and the rock’s pore–throat radius, Therefore, Equation (1) can be rewritten as follows:

$$T_2={\displaystyle \frac{{\left(R_{\mathrm{t}}{R}_1\right)}^{\mathrm{n}}}{{\rho }_2{F}_{\mathrm{s}}}}$$

(2)

In Equation (2), $R_{\mathrm{t}}$ refers to the average pore–throat ratio of the rock. $R_1$ is the pore–throat radius of the rock (in μm). $F_{\mathrm{s}}$ is the pore shape factor. Equation (2) can then be further transformed as follows:

$$B_1=K\cdot T_2^{{\displaystyle \frac{1}{\mathrm{m}}}}$$

(3)

Figure 2a quantitatively represents the heterogeneity of the reservoir’s pore–throat system using a bimodal curve of instantaneous frequency and cumulative frequency. The instantaneous frequency distribution highlights the peak characteristics of the pore–throat proportion (with the main peak radius reflecting the dominant pore–throat size). The cumulative frequency curve depicts the cumulative distribution of pore–throat sizes. Together, these two curves allow for the clear identification of the spatial proportions of high-permeability channels and low-permeability regions within the reservoir. Figure 2b, based on the conversion model between NMR T₂ relaxation time and pore radius, maps the relaxation time (T₂) to a mathematical model of physical pore radius distribution, which indirectly reflects the graded characteristics of multiscale pores. Figure 2c presents the cumulative distribution curve of T₂ signal intensity, revealing the graded characteristics of the reservoir’s pore system in terms of pore volume. T₂ relaxation times greater than 10 ms correspond to large pores (which contribute the majority of the signal intensity), while T₂ relaxation times less than 1 ms represent small pores or micropores. Figure 2d, through the development of a power-function model between T₂ relaxation time and pore radius, enables the conversion of relaxation time into pore–throat radius, offering theoretical support for the precise delineation of pore–throat activation boundaries.

Figure 3 illustrates the principle behind calculating the contribution of pore–throats with pore diameters ranging from 0.001 ms to 1 ms. The T₂ value of the initially saturated crude oil in pore–throats with relaxation times between 0.001 ms and 1 ms is represented by (S_o), while the T₂ value of the residual oil after displacement is denoted by (S_w). For pore–throats with relaxation times ranging from 0.001 ms to 10,000 ms, the initial amount of saturated crude oil is represented by (S₁), and the T₂ value of the residual oil after displacement is represented by (S₂). By comparing the frequency–area differences in the NMR T₂ spectra before and after the displacement, the pore–throat contribution (I) can be calculated:

$$I={\displaystyle \frac{S_o-S_w}{S_1-S_2}}\times 100\%$$

(4)

In Equation (4): I is the pore–throat contribution (%); S_o − S_w represents the difference in T₂ spectrum frequency area between the initial saturated oil volume and the residual oil volume after displacement for pore–throats with pore diameters ranging from 0.001 ms to 1 ms. S₁ − S₂ represents the difference in T₂ spectrum frequency area between the initial saturated crude oil volume and the residual oil volume after displacement for pore–throats with pore diameters ranging from 0.001 ms to 10,000 ms.

2.4. Experimental Procedure

The experimental procedure is as follows:

(1)

Clean, desalinate, and dry the natural reservoir core samples. Measure their gas permeability and porosity, then apply vacuum extraction and saturate with simulated formation water. After saturation, perform NMR T₂ spectrum sampling on the cores with the saturated formation water.

(2)

Prepare a 15,000 mg/L Mn²⁺ aqueous solution. Inject the manganese solution into the core at a constant flow rate of 0.05 mL/min to displace the simulated formation water, injecting 3 PV to 4 PV. Perform NMR T₂ spectrum sampling on the manganese-displaced core samples to assess the effectiveness of eliminating water signals.

(3)

Inject experimental crude oil into the core at a constant flow rate (0.05 mL/min) to displace the formation water until the crude oil content at the outlet reaches 100%. Establish the original oil–water distribution in the formation, and perform NMR T₂ spectrum sampling on the crude oil-saturated core samples.

(4)

Prepare a 2% potassium chloride + 0.5% organic anti-swelling agent solution. Inject this solution into the core at a constant flow rate of 0.05 mL/min to displace the crude oil, injecting 3 PV to 4 PV. Perform NMR T₂ spectrum sampling on the core samples that underwent water flooding.

(5)

Set the experimental temperature to simulate reservoir conditions. Increase the pressure of the intermediate container filled with CO₂ to the preset experimental pressures (6 MPa, 10 MPa, 14 MPa, 18 MPa, 22 MPa, and 26 MPa). Open the core holder inlet valve and inject CO₂ into the core at a flow rate of 0.1 mL/min. Adjust the backpressure valve to maintain a constant backpressure.

(6)

Perform NMR T₂ spectrum sampling on the CO₂-displaced core samples to examine the oil–water distribution characteristics. Record the oil production and calculate oil recovery efficiency.

(7)

Wash the CO₂ gas-flooded core samples with petroleum ether and benzene for 120 h. After washing, dry the core samples at 80 °C for 24 h. Repeat the above steps with cores of different permeabilities.

3. Experimental Results and Analysis

3.1. Characteristics of Sandstone Samples

The lower section of the Wutonggou Formation (P₃wt₁) in the Permian of the Junggar Basin is divided into four sandstone layers from bottom to top. The P₃wt₁³⁻² sandstone layer is the most developed and serves as the primary oil-bearing layer. The core samples for this experiment were obtained from the Shaqiu 5 area of the Shanan Oilfield. X-ray diffraction (XRD) whole-rock analysis was conducted on the Permian Wutonggou Formation reservoir of the Shanan Oilfield, with the test results presented in

Table 2. X–diffraction whole rock analysis of core samples.

Number	Formation	Mineral Content (%)
		Quartz	K-Feldspar	Plagioclase	Laumontite	Clay Minerals
1	P3wt13−2	62.6	5.2	17.1	/	15.1
2	P3wt13−2	42.1	23.0	21.4	/	13.5
3	P3wt13−2	60.5	2.4	15.5	/	11.5
4	P3wt13−2	58.2	1.8	24.0	2.1	13.9

and

Table 3. Relative clay mineral content of core samples.

Number	Formation	Relative Content of Clay Minerals (%)
		S	It	K	C
1	P3wt13−2	51	5	35	9
2	P3wt13−2	79	15	3	3
3	P3wt13−2	22	11	59	8
4	P3wt13−2	45	5	39	11

S: Soapstone type; It: Illite; K: Kaolinite; C: Chlorite.

.

To provide a clearer comparison of the proportions of different rock minerals in the reservoir of the study area, a histogram of the rock-mineral types in the reservoir is shown in Figure 4.

The experimental results indicate that quartz has the highest content among the minerals, with an average value of 55.85%. This is followed by feldspar minerals, with potassium feldspar making up 8.1% and plagioclase making up 19.5%. The clay mineral content is also relatively high, averaging 13.5%. The clay minerals are primarily composed of montmorillonite (49.52%) and interlayered kaolinite (34%), with sheet-like chlorite (7.75%) as the second most abundant. These parameters reflect the typical mineral composition of low-permeability sandstones in the samples selected for this experiment. They are essential for analyzing the pore–throat size and connectivity, providing a foundation for determining the activation boundaries of pore–throats at different scales in subsequent studies.

3.2. Waterflood to CO₂ Gas Injection: Pore–Throat Activation Boundaries

Waterflood followed by CO₂ gas injection experiments were performed on core samples (1 to 6). Using the NMR T₂ spectra of these six core samples, the crude oil recovery under different injection methods was quantitatively calculated. The calculation results are presented in

Table 4. Crude oil recovery from post-water-flooding cores at different CO₂ injection pressures.

Core Sample ID	Displacement Medium	Permeability /×10−3 μm2	Porosity /%	Liquid Displacement Pressure/MPa	Gas Displacement Pressure/MPa	Waterflood Efficiency /%	Gas Injection Efficiency /%	Total Oil Recovery Efficiency /%
1	CO2	0.94	13.55	6	6	10.71	33.89	44.60
2	CO2	0.89	14.11	10	10	12.27	41.96	54.23
3	CO2	2.85	13.71	14	14	15.54	49.86	65.40
4	CO2	1.26	13.86	18	18	19.21	50.78	69.99
5	CO2	1.18	13.89	22	22	18.83	49.87	68.70
6	CO2	1.09	13.87	26	26	15.17	46.82	61.99

. The oil recovery efficiency parameters provide a comprehensive reflection of the connectivity and seepage capacity of pore–throats at various scales.

Figure 5 presents the NMR T₂ spectra curves for waterflooding and CO₂ gas injection at various pressures. For core sample 1, after waterflooding, the injection of 6 MPa CO₂ increased the overall crude oil recovery by 23.18%, reaching 33.89%. For core sample 2, following waterflooding, the injection of 10 MPa CO₂ increased the crude oil recovery by 29.69%, reaching 41.96%. For core sample 3, after waterflooding, CO₂ injection at 14 MPa enhanced crude oil recovery in small pores by 19.82%, reaching 31.05%, while in large pores, the recovery increased by 38.62%, reaching 56.88%. The overall recovery compared to waterflooding increased by 34.32%, reaching 49.86%. For core sample 4, after waterflooding, the injection of 18 MPa CO₂ raised the crude oil recovery in small pores by 31.80%, reaching 46.15%, and in large pores by 34.76%, reaching 54.85%. The overall recovery increased by 31.57%, reaching 50.78%, compared to the recovery after waterflooding.

For core sample 5, after waterflooding, the injection of 22 MPa CO₂ increased crude oil recovery in small pores by 33.14%, reaching 43.43%, and in large pores by 29.64%, reaching 55.2%. The injection of CO₂ significantly increased oil recovery in the core’s small pores. Compared to waterflooding, the overall crude oil recovery increased by 31.04%, reaching 49.87%. For core sample 6, following waterflooding, the injection of 26 MPa CO₂ resulted in a 35.25% increase in recovery in small pores, reaching 57.74%, while recovery in large pores increased by 6.55%, reaching 16.41%. CO₂ injection significantly enhanced oil recovery in the small pores of the core. Compared to waterflooding, the overall crude oil recovery increased by 31.65%, reaching 46.82%. By comparing the NMR T₂ spectra curves for waterflooding, followed by CO₂ injection at different pressures, it is evident that CO₂ injection at 18 MPa yields the highest final recovery, with the highest oil recovery in both small and large pores and the best oil recovery efficiency.

The calculation results for the pore–throat activation radii and crude oil recovery in the Wutonggou Formation of the Shaqiu 5 area are presented in

Table 5. Radius of water-to-CO₂-gas flooding borehole throat mobilization and degree of crude oil mobilization.

Core Sample ID	Displacement Medium	Displacement Pressure/MPa	Pore–Throat Activation Radius/μm		Crude Oil Recovery Degree/%
			Lower Limit	Upper Limit	Small Pores	Large Pores	Overall
1	Water	6	1.571	17.004	4.26	12.71	10.71
A-1	CO2	6	0.348	19.231	17.95	43.97	33.89
2	Water	10	1.620	18.307	6.28	15.06	12.27
A-2	CO2	10	0.339	19.711	10.26	63.68	41.96
3	Water	14	1.437	18.764	11.23	18.26	15.54
A-3	CO2	14	0.331	15.793	31.05	56.88	49.86
4	Water	18	1.282	23.418	14.35	20.09	19.21
A-4	CO2	18	0.307	20.706	46.15	54.85	50.78
5	Water	22	1.161	20.706	10.29	25.56	18.83
A-5	CO2	22	0.315	16.590	43.43	55.2	49.87
6	Water	26	1.161	20.202	22.49	9.86	15.17
A-6	CO2	26	0.323	17.004	57.74	16.41	46.82

. The minimum lower limit for activation by waterflooding is 1.161 μm, while the maximum upper limit is 23.418 μm. The average lower limit is 1.372 μm, and the average upper limit is 19.733 μm. For waterflooding followed by CO₂ gas injection, the minimum lower limit is 0.307 μm, and the maximum upper limit is 20.706 μm. The average lower limit is 0.327 μm, and the average upper limit is 18.172 μm. Waterflooding primarily activates pore–throat radii between 1.372 μm and 19.733 μm, whereas CO₂ gas injection activates pore–throats from 0.327 μm to 18.172 μm. As shown in Figure 6 and Figure 7, the activation boundaries for CO₂ gas injection are higher than those for waterflooding, indicating a broader activation range and greater potential to enter smaller pore spaces and displace crude oil.

3.3. Contribution of Pore–Throats at Different Scales

The contribution results for the Wutonggou Formation in the Shaqiu 5 area are shown in

Table 6. Scale of pore–throat contribution of water flooding and water-to-gas flooding at different injection pressures.

Core Sample ID	Displacement Medium	Displacement Pressure/MPa	Pore–Throat Contribution /%
			Small Pores	Medium Pores	Large Pores
1	Water	6	38.71	18.25	43.04
A-1	CO2	6	25.26	37.88	36.86
2	Water	10	40.65	17.37	41.97
A-2	CO2	10	20.18	29.98	49.84
3	Water	14	44.36	15.86	39.78
A-3	CO2	14	38.84	23.26	37.90
4	Water	18	46.70	16.69	36.60
A-4	CO2	18	43.69	22.69	33.62
5	Water	22	45.28	17.55	37.17
A-5	CO2	22	41.05	24.54	33.61
6	Water	26	40.93	19.75	39.32
A-6	CO2	26	25.03	35.06	39.91

. It can be observed that during waterflooding, the contribution of large pores gradually decreases within the pressure range of 6 MPa to 18 MPa, from 43.04% to 36.60%, a reduction of 6.44%. At 22 MPa and 26 MPa, the contribution slightly increases. The contribution of medium pores is higher at lower pressures but decreases at higher pressures. The contribution of medium pores shows a pattern of first decreasing and then increasing within the pressure range of 6 MPa to 26 MPa. It reaches the lowest point of 15.86% at 14 MPa and then gradually rises to 19.75%. This is because medium pores are more easily displaced by water at higher pressures. For small pores, the contribution gradually increases from 38.71% to 46.70% within the pressure range of 6 MPa to 18 MPa and then decreases at 22 MPa and 26 MPa. This indicates that small pores contribute more to waterflooding at medium pressures, as they are more easily displaced by water at lower pressures. As shown in Figure 8, during the waterflooding process, the contribution of large pore–throats is initially higher. As the pressure increases, the contribution stabilizes, while the contribution of medium and small pore–throats gradually increases. Initially, the overall pore–throat contribution is uneven, but it becomes more uniform later in the process.

In the transition from waterflooding to gas injection in the Wutonggou Formation of the Shaqiu 5 area, the contribution of large pores first increases and then decreases within the pressure range of 6 MPa to 22 MPa. It reaches the maximum contribution of 49.84% at 10 MPa, followed by a slight increase at 22 MPa and 26 MPa. Large pores contribute more to gas injection at lower pressures, but their contribution decreases at higher pressures. The contribution of medium pores shows a trend of first decreasing and then increasing within the pressure range of 6 MPa to 26 MPa. It reaches the lowest point of 22.69% at 22 MPa, and then gradually increases to 35.06%. This may be due to the fact that medium pores are more easily displaced by gas in a finger-like form at higher pressures. The contribution of small pores significantly increases between 6 MPa and 14 MPa, from 25.26% to 38.84%, a 53.7% increase. However, the contribution decreases between 18 MPa and 26 MPa, suggesting that small pores contribute more to gas injection at medium pressures. As shown in Figure 9, during the transition from waterflooding to gas injection, the contribution of large pore–throats remains high as pressure increases, while the contributions of medium and small pore–throats gradually rise, leading to a more uniform distribution of overall pore–throat contributions. This trend helps improve the recovery factor of the reservoir.

4. Discussion

4.1. The Degree of Utilization of Water-Driven to CO₂ and Gas-Flooding Crude Oil

As shown in Figure 6 and Figure 7, through the experimental results of water flooding and CO₂ flooding after water flooding, it is concluded that when the injection pressure increases from 6 MPa to 26MPa, the overall utilization rate of water flooding reaches a maximum of 19.21%, the utilization rate of small pores reaches a maximum of 22.49%, and the utilization rate of medium pore and large pore reaches a maximum of 25.56%. The overall utilization rate of CO₂ displacement reached a maximum of 33.89%, the utilization rate of small pores reached a maximum of 57.74%, and the utilization rates of medium pores and large pores reached a maximum of 56.88%. The utilization degree of water flooding on pores of different diameters is not high, and the utilization degree of medium and large pores is slightly higher than that of small pores, which is in line with the characteristic that water flooding usually can more easily enter larger pores. The utilization degree of CO₂ flooding in the overall structure and in small pores, medium pores and large pores is much higher than that of water flooding, especially in small pores where the utilization degree is significantly increased.

The interfacial tension between CO₂ and crude oil is relatively low. After CO₂ injection, it can be miscible or nearly miscible with crude oil under high pressure, significantly reducing the viscosity of crude oil. This viscosity reduction effect enhances the fluidity of crude oil, enabling the crude oil in small pores to be utilized more effectively, thereby significantly increasing the overall utilization rate.

By comparing the data in Figure 8 with that in Figure 9, it can be seen that the maximum range for water drive is 1.106 μm to 20.706 μm, and the maximum range for CO₂ drive is 0.303 μm to 20.706 μm. The upper and lower limits of water CO₂ drive are higher and have a wider range than those of water drive. It indicates that CO₂ is more likely to extract crude oil from smaller pores than water flooding, thereby enhancing the crude oil recovery rate.

4.2. CO₂ Microscopic Oil Displacement Mechanism

During CO₂ displacement, the non-wetting phase characteristic causes CO₂ to preferentially occupy large pore–throats [36]. At low injection volumes, capillary resistance is high, limiting oil recovery. As the injection volume increases, the swept volume expands and the oil recovery efficiency improves (Figure 10). In high-permeability cores, pore–throat connectivity is good, leading to broad CO₂ sweep and low residual oil saturation, with high oil recovery efficiency when the optimal injection volume is large [37,38,39,40]. In low-permeability cores, dissolution pores dominate, resulting in poor sweep efficiency and wide distribution of residual oil, leading to higher saturation. Heterogeneity causes CO₂ to form dominant flow channels in large pore–throats, while oil recovery in small pore–throats remains low, creating a phenomenon where large pore clusters of residual oil coexist with small pore oil droplets.

Injection pressure significantly influences the oil displacement mechanism: At low pressures, the displacement is non-miscible, with CO₂ primarily diffusing along large pores, leading to insufficient oil recovery in small pores [41,42]. The recovery is mainly dependent on oil dissolution and oil swelling. Once the pressure exceeds the minimum miscibility pressure, CO₂ enters smaller pores, extracting light hydrocarbons and reducing interfacial tension, significantly improving sweep efficiency under miscible displacement. When the pressure exceeds the miscibility pressure, fluid mass transfer is enhanced, and the recovery rate reaches its peak.

5. Conclusions

This study analyzes core samples from the Wutonggou Formation in the Shaqiu 5 Well Area of the Shanan Oilfield. By combining low-field nuclear magnetic resonance (NMR) technology with CT experiments, the pore–throat activation boundaries and contributions in low-permeability sandstone reservoirs under varying injection conditions are quantitatively characterized. The key findings are as follows:

(1)

At an injection pressure of 6.0 MPa, the lower limit of pore–throat mobilization is 0.348 μm; at 10.0 MPa, it is 0.339 μm; at 14.0 MPa, it is 0.331 μm; and at 18.0 MPa, it is 0.307 μm. Higher injection pressures result in a greater overall degree of pore–throat mobilization in the core.

(2)

During water flooding, large pore throats initially exhibit the highest contribution rate, reaching 43.04%. As pressure increases, the contribution rate gradually stabilizes at 22 MPa, while the contribution rates of medium and small pore throats progressively increase. Initially, the overall pore–throat contribution shows significant heterogeneity, which tends to homogenize in the later stages. During the transition from water flooding to CO₂ flooding, the contribution rate of large pore throats remains high, peaking at 49.84% under 10 MPa pressure. Meanwhile, the contribution rates of medium and small pore throats gradually increase, leading to a more uniform distribution of overall pore–throat contributions, which aids in enhancing oil recovery.

(3)

In CO₂ flooding, CO₂ dissolves into the crude oil, causing volume expansion and viscosity reduction, accompanied by mass transfer and extraction processes. Injection pressure significantly influences these mechanisms. During the immiscible phase, the mobilization extent and range are limited, whereas miscibility substantially enhances both mobilization extent and range. In field applications, controlling the injection pressure to achieve miscible conditions with CO₂ can effectively improve crude oil recovery.