Research on Enhanced Gas Recovery by CO₂/N₂ Injection in Tight Sandstone Gas Reservoirs

Abstract

CO₂-enhanced gas recovery (CO₂-EGR) is a crucial technology for achieving both natural gas production increase and CO₂ geological storage. While pure CO₂ flooding demonstrates favorable recovery performance, the technical challenges and high costs associated with purifying CO₂ remain significant. CO₂ purification from exhaust gas incurs prohibitive costs, while direct injection of an unpurified CO₂–N₂ mixture can greatly cut engineering expenditure. Nitrogen also provides synergistic pressure support, working with CO₂ to drive natural gas displacement. Therefore, from an economic and practical standpoint, employing impure CO₂ mixtures (e.g., CO₂–N₂) for flooding presents a more advantageous approach. To clarify the factors influencing the recovery enhancement in tight sandstone gas reservoirs using CO₂–N₂ mixtures, long-core flooding experiments were conducted at 100 °C. This study systematically investigates the impact patterns of three key factors—injection timing, injection rate, and injection gas composition—on the enhanced recovery of tight sandstone gas reservoirs. The experimental results indicate that: (1) Advancing the injection timing significantly improves the recovery performance for both CO₂ and N₂ flooding. However, the cumulative recovery factor (sum of the depletion recovery and the incremental recovery from gas injection) shows a declining trend. (2) The enhanced recovery effect exhibits a trend of first increasing and then decreasing with the increase in injection rate. When the injection rate exceeds 0.05 mL/min, it tends to cause premature breakthrough of the injected gas, thereby reducing the displacement efficiency. (3) As the proportion of CO₂ in the injected gas increases, the enhanced recovery effect shows a nonlinear rise. The highest incremental recovery (17.02%) was achieved with pure CO₂ flooding, while pure N₂ flooding yielded the lowest result (14.64%). The research findings, from a macroscopic perspective, elucidate the influence patterns of three distinct factors on enhancing gas recovery in tight sandstone reservoirs, thereby providing theoretical foundation and scientific guidance for the development of such reservoirs. In summary, the injection timing, injection rate and CO₂ proportion in injected gas are the key controlling factors for gas flooding enhanced recovery in tight sandstone reservoirs. This study clarifies the macroscopic influence law of each factor, and the optimized development parameters proposed can provide direct theoretical support and technical guidance for the on-site application of gas flooding in tight sandstone reservoirs.

1. Introduction

As the cleanest and lowest-carbon fossil energy source, natural gas serves as an indispensable component in the construction of China’s new energy system. In 2024, China’s natural gas output attained 2464 × 10⁸ m³, with apparent consumption surpassing 4000 × 10⁸ m³ and natural gas imports totaling 1817 × 10⁸ m³, which resulted in an external dependence rate of 42%. Accordingly, the task of safeguarding energy security remains arduous [1]. Given this context, accelerating the development of natural gas has become a core priority in China’s energy development strategy [2]. As a key constituent of unconventional natural gas, tight sandstone gas has gradually emerged as an increasingly prominent strategic element in the natural gas resource structure. Enhancing its recovery efficiency is widely acknowledged as a critical breakthrough for the efficient exploitation of unconventional natural gas resources in China [3,4]. Nevertheless, the development of tight sandstone gas reservoirs poses three major technical bottlenecks: inadequate recovery efficiency, rapid production decline, and limited economic viability. Therefore, systematic investigation into the key influencing factors and their intrinsic action mechanisms for improving tight sandstone gas recovery holds profound significance for the effective development and rational utilization of China’s tight gas resources [5]. CO₂-Enhanced Gas Recovery (CO₂-EGR) represents a critical technological pathway for realizing the synergistic development of natural gas production enhancement and CO₂ geological sequestration. However, the utilization of pure CO₂ flooding is accompanied by notable challenges, including high purification difficulty and substantial costs. Consequently, the adoption of a CO₂/N₂ mixed system for flooding demonstrates greater advantages in terms of economic feasibility [6].

Scholars at home and abroad have conducted extensive research on the mechanism of enhancing gas reservoir recovery efficiency via CO₂/N₂ injection. The physical properties of the injected gas exert a direct influence on the gas reservoir recovery performance. CO₂ and CH₄ exhibit significant differences in density and viscosity: the injected CO₂ tends to settle preferentially at the bottom of the gas reservoir and drives natural gas toward production wells through gravitational differentiation, thereby improving recovery efficiency [7]. Studies have indicated that the diffusion coefficient of N₂ is higher than that of CO₂, and nitrogen is more likely to mix with methane. Therefore, the injection of CO₂/N₂ mixed gas can also effectively enhance CH₄ recovery efficiency [8].

The selection of gas injection parameters exerts a significant influence on the effectiveness of CO₂/N₂ in enhancing gas reservoir recovery efficiency. Gas injection timing has been identified as one of the key factors affecting recovery efficiency, and an earlier injection timing generally contributes to better development performance. In particular, when CO₂ is in a supercritical state, its recovery enhancement effect on tight gas reservoirs is superior to that in a non-supercritical state [9,10,11,12,13,14]. Furthermore, when the injection pressure exceeds a certain threshold, the gain in recovery efficiency is no longer significant, and it is necessary to optimize pressure control to achieve a balance between economic feasibility and recovery performance [15]. An increase in injection rate can expand the sweep range of CO₂ in the core and more effectively displace natural gas in micro-pores, thereby increasing the ultimate recovery efficiency [16,17,18]. However, there exists an optimal threshold for the gas injection rate; exceeding this threshold will reduce the recovery efficiency due to premature gas breakthrough [19].

Reservoir conditions and injection schemes also have a significant impact on the effectiveness of CO₂/N₂ injection in enhancing gas reservoir recovery efficiency, with complex action mechanisms that jointly determine the ultimate efficiency of the displacement process. Luo et al. [20] pointed out through numerical simulation that strong reservoir heterogeneity can accelerate the formation of CO₂ preferential channels, leading to shortened breakthrough time and reduced CH₄ recovery efficiency. In contrast, Zhao et al. and Wu Jiaqi et al. [21,22] found that the development of a complex fracture network can expand the sweep volume, significantly improve CH₄ recovery efficiency, and delay CO₂ breakthrough. Fan et al. [23] proposed a strategy of progressive CO₂ injection increment through numerical simulation, which can maintain a high CH₄ production rate while inhibiting matrix swelling, resulting in an 8% to 12% increase in recovery efficiency.

In response to the above issues, this study takes the tight sandstone gas reservoir of the Sulige Gas Field in the Ordos Basin as the research object. Through long core flooding experiments of CO₂/N₂ injection under different conditions, the impact characteristics of different gas injection parameters on recovery efficiency enhancement are systematically analyzed, aiming to provide a theoretical basis for the efficient development of tight sandstone gas reservoirs by CO₂ injection.

Although CO₂/N₂ injection has been extensively studied in kerogen-rich shales, tight sandstone reservoirs exhibit distinct geomechanical responses (e.g., negligible matrix swelling). Furthermore, previous experimental studies heavily relied on crushed rock samples, which fail to capture the complex multiphase flow dynamics in intact, ultra-low permeability pore-throat networks. This study uniquely employs a 35.8 cm intact long-core physical simulation at actual reservoir conditions (100 °C) to quantitatively evaluate the macroscopic displacement efficiency, specifically focusing on the critical operational parameter of injection timing during depletion.

2. Overview of the Study Area

2.1. Geological Characteristics of the Sulige Gas Field

The Ordos Basin is a large-scale sedimentary Basin characterized by relatively weak tectonic deformation, multi-cyclic evolution, and the development of diverse sedimentary systems, covering an area of approximately 25 × 10⁴ km². The regional structure of the Sulige Gas Field is gentle, presenting an overall west-dipping monocline with a Formation dip angle generally less than 1°. The major pay zones of the gas field are the 8th member of the Shihezi Formation (He 8 Member) and the 1st member of the Shanxi Formation (Shan 1 Member). The sand bodies of the main pay zones have a width ranging from 0.5 to 3 km, predominantly exhibiting two superposition patterns: lateral juxtaposition and vertical incised stacking. Within the area, the gas-bearing intervals feature good continuity, while the boundary reservoirs are tight, rendering the gas pools relatively isolated. The regional caprocks consist of widely distributed Mudstones and sandy mudstones of the Upper Shihezi Formation, with a cumulative thickness of over 100 m. Meanwhile, the locally developed thick interlayers of the Lower Shihezi Formation serve as the direct caprocks. The basement is composed of lacustrine dark mudstones, coal seams, and tight limestones, which together constitute an effective sealing system. The study area is located in the eastern region, where the target formation is the He 8 Member of the Lower Shihezi Formation in the Upper Paleozoic, one of the main pay zones of the Sulige Gas Field in the Ordos Basin. The reservoirs of the He 8 Member exhibit strong heterogeneity, which is mainly reflected in the pore types dominated by primary residual intergranular pores and secondary lithic dissolution pores. The pore-throat structure is complex, belonging to the low porosity-fine to micro-throat type. According to the high-pressure mercury intrusion data, the throat radius is mainly distributed between 0.05 and 1.0 μm, with a median radius generally less than 0.5 μm. Observations of cast thin sections show that the pore connectivity is moderately poor, and the microscopic heterogeneity is significant [24]. In terms of reservoir physical properties, the overall physical properties of the He 8 Member reservoirs in this area are poor. Based on the statistics of a large number of core experimental data, the porosity ranges from 6% to 12% with an average of 8.5%, and the permeability ranges from 0.1 to 2.0 mD with an average of 0.6 mD. The porosity-permeability relationship shows a positive correlation, but the data points are scattered, indicating strong physical property heterogeneity within the reservoirs. The rock types are dominated by Quartz sandstones, followed by lithic quartz sandstones. From west to east, the quartz content gradually decreases, while the lithic content increases correspondingly. The clay minerals are mainly Illite and Kaolinite, with a small amount of Chlorite.

2.2. Brief Overview of Gas Reservoir Development

The Sulige Gas Field is situated in the northwest of the Yishan Slope in the Ordos Basin, serving as a typical representative of multi-gas layer superposition and development in the Upper Paleozoic. After the launch of systematic exploration in 1999, a major breakthrough was achieved in this area via Well Su-6 in 2000. This discovery marked the official entry of the Sulige Gas Field into the stage of large-scale exploration and development, laying a resource foundation for the development of onshore tight gas in China. The development of the Sulige Gas Field can be divided into three key stages: Appraisal Stage (2001–2006): During this stage, the geological characteristics of the gas field, namely low permeability, low pressure and low abundance, were clarified. A technical system centered on well location optimization and rapid drilling was established, providing a basis for the economic and efficient development of the gas field. Production Ramp-up Stage (2007–2013): The gas production achieved leapfrog growth, with an annual output increase of over 30 × 10⁸ m³ for six consecutive years. By popularizing horizontal well technology and standardized construction mode, the annual gas production reached 135 × 10⁸ m³ in 2012, making it the largest onshore integrated gas field in China. Stable Production Stage (2014–present): The gas field has entered a period of high and stable production. In 2024, the annual gas production stabilized at 300 × 10⁸ m³, accounting for approximately 58% of the total domestic tight gas output. It has maintained an annual gas production exceeding 30 billion cubic meters for three consecutive years, continuously guaranteeing the security of energy supply [25].

As the largest tight Sandstone gas field in China, after achieving the leap of an annual gas production exceeding 30 billion cubic meters, the Sulige Gas Field has now entered the stage of intensive development, with the core objectives of extending the stable production period and improving the ultimate recovery factor. At present, production maintenance and potential tapping mainly rely on conventional technical means such as well pattern infill and optimization, reservoir stimulation technologies including multi-stage fracturing of horizontal wells, water drainage gas recovery, and surface pressure boosting. Although these technologies have achieved remarkable results, they still face fundamental challenges, such as low sweep efficiency caused by strong reservoir heterogeneity, scattered distribution of remaining gas that is difficult to produce, and the balance between the cost and benefit of technical measures. In particular, existing technologies are difficult to effectively mobilize a large amount of isolated remaining gas between wells and layers, which restricts the further improvement of the recovery factor. Therefore, developing new enhanced recovery technologies such as gas injection (e.g., CO₂, N₂) for energy supplement has become an inevitable trend. By supplementing formation energy and effectively displacing and replacing remaining gas, these technologies are expected to increase the recovery factor by an additional 5% to 7%. Moreover, combined with carbon capture, utilization and storage (CCUS), they can provide a key technical replacement for the green, efficient and sustainable development of the gas field.

3. Experimental Materials and Methods

3.1. Experimental Materials

The cores (Figure 1) used in this study were collected from the Sulige Gas Field in the Ordos Basin, with the specific strata being Member He 8 and Member Shan 1, which are classified as Class I reservoirs in the gas field. The quartz content of the cores ranges from 88.7% to 92.4%, and the clay mineral content varies between 11.3% and 7.6%. In addition, the porosity of the cores is in the range of 6.4% to 9.33%, and the permeability is 0.199–0.48 mD. Detailed parameters are listed in

Table 1. Properties of the Sandstone Core.

Core Sorting	Length/mm	Diameter/mm	Dry Weight/g	Permeability/mD	Porosity/%	Core Pore Volume/(cm3)
1	44.03	24.92	51.03	0.377	9.2	1.976
2	43.77	24.92	51.66	0.213	8.05	1.719
3	42.51	24.99	50.88	0.41	7.06	1.472
4	45.45	24.92	53.77	0.208	8.69	1.926
5	43.29	25.00	51.97	0.459	6.4	1.359
6	45.31	24.95	53.30	0.199	8.36	1.851
7	49.94	24.89	57.85	0.48	8.83	2.146
8	43.89	24.92	50.94	0.444	9.33	1.997

.

The gas components employed in the experiments include high-purity CO₂ (with a purity of 99.95%), CH₄, N₂, and three types of CO₂–N₂ mixed gases with different proportions, and the uncertainty of the gas components is less than 2%.

3.2. Experimental Instruments and Methods

The equipment for the long-core displacement experiment consists of three constant-rate and constant-pressure pumps, a long core holder, a thermostatic oven, a back-pressure valve, a gas flowmeter, a gas chromatograph, three piston-type intermediate containers, gas cylinders, temperature and pressure sensors, and a data acquisition system. A schematic diagram of the experimental setup is shown in Figure 2, and the detailed experimental procedures are as follows.

(1) Core Sorting and Packing

The core sorting was determined according to the iterative calculation of the harmonic mean of permeability, as expressed in Equation (1). Specifically, the harmonic mean of the permeability of all core samples was first calculated, and the core sample with permeability closest to this value was placed at the initial position at the outlet end of the core holder. After removing the selected core sample, the new harmonic mean of permeability of the remaining core samples was calculated iteratively for sequential sorting until all core samples were arranged in a permeability gradient. Microporous filter papers were placed between adjacent core samples to mitigate the end effect. During packing, the sorted core samples were encapsulated in a fluororubber sleeve and then fixed in the core holder. The long core sample was a dry core spliced from eight core plugs, with a total length of 35.819 cm.

$${\displaystyle \frac{L}{\overline{K}}}={\displaystyle \frac{L_1}{K_1}}+{\displaystyle \frac{L_2}{K_2}}+\cdots +{\displaystyle \frac{L_i}{K_i}}+\cdots +{\displaystyle \frac{L_{\mathrm{n}}}{K_{\mathrm{n}}}}={\displaystyle \sum _{i=1}^n\frac{L_i}{K_i}}$$

(1)

where L is total length of the core, cm. $\overline{K}$ is harmonic mean permeability of the core, mD. $L_i$ is length of the ith core samples, cm. $K_i$ is permeability of the ith core samples, mD.

(2) Experimental System Connection and Initialization

The inlet end of the core holder was connected to piston-type intermediate containers loaded with different gases, and the bottom of each container was linked to a displacement pump. The outlet end was connected in sequence to a back-pressure valve (connected to a back-pressure pump), a gas chromatograph, and a gas flowmeter. The top of the core holder was connected to a confining pressure pump to simulate the overburden pressure. During system initialization, the confining pressure was set to be 3 MPa higher than the inlet pressure at all times, and CH₄ was introduced into the core holder to test the airtightness of the system. The specific connection mode is shown in Figure 2.

(3) Temperature and Pressure Equilibration

The temperature of the oven was set to the reservoir temperature of 100 °C, and the core was heated uniformly by holding this temperature for at least 4 h. Subsequently, the displacement pump pressurized CH₄ to 27 MPa (initial formation pressure) in constant pressure mode, and this pressure was maintained for more than 8 h to ensure that the core pores were fully saturated with CH₄.

(4) Depletion Recovery Simulation

The CH₄ injection was shut off to simulate the formation depletion process, and the system pressure was reduced to the target pressure (e.g., 11 MPa or 5 MPa), providing initial conditions for the subsequent CO₂ displacement.

(5) CO₂ Injection and Displacement

CO₂ was injected into the piston-type intermediate container and pressurized to the experimental pressure, after which the displacement pump was switched to constant flow mode (e.g., 0.05 mL/min). The inlet valve was opened to initiate CO₂ displacement, during which the back-pressure control mode was adopted to maintain a constant pressure at the production end (e.g., 5 MPa). The variations in pressure and flow rate were monitored in real time. (It should be noted that the pressure difference referred to in this experiment specifically denotes the pressure differential between the gas injection end and the production end of the long-core model. Among these, the “maximum pressure difference” refers to the peak pressure differential observed during the gas injection and displacement stage.)

(6) Experimental Data Acquisition and Breakthrough Determination

During the displacement process, time, inlet pressure, flow rate, and produced gas composition (measured by gas chromatograph) were continuously recorded. The onset of CO₂ breakthrough was marked when the volume fraction of CH₄ in the produced gas decreased to 95%. The experiment was terminated when the CH₄ volume fraction dropped to 5%, which was regarded as a complete breakthrough (i.e., the experimental abandonment condition was reached). At this moment, the pressure at the production end of the system was defined as the abandonment pressure of this experiment.

A total of 17 groups of long-core displacement experiments were designed and conducted at 100 °C, including experiments with different CO₂ injection timings, different CO₂ injection rates, and different gas injection compositions. Detailed experimental parameters are listed in

Table 2. Experimental Design.

No.	Influencing Factors	Gas Injection Timing/MPa	Gas Injection Rate/mL/min	Gas Injection Components
				CO2	N2
1	Gas Injection Timing	5	0.1	100%	0%
2		7	0.1	100%	0%
3		9	0.1	100%	0%
4		11	0.1	100%	0%
5		5	0.1	0%	100%
6		7	0.1	0%	100%
7		9	0.1	0%	100%
8		11	0.1	0%	100%
9	Gas Injection Rate	7	0.02	100%	0%
10		7	0.05	100%	0%
11		7	0.2	100%	0%
12		7	0.02	0%	100%
13		7	0.05	0%	100%
14		7	0.2	0%	100%
15	Gas Injection Components	7	0.1	66.56%	33.44%
16		7	0.1	49.98%	50.02%
17		7	0.1	33.33%	33.67%

.

4. Analysis of Effects of CO₂/N₂ Injection on Enhancing Recovery Efficiency of Tight Sandstone Gas Reservoirs

4.1. Effects of CO₂/N₂ Injection Timing

In the process of CO₂/N₂-enhanced gas recovery (CO₂/N₂-EGR), the selection of gas injection timing directly affects the cumulative gas recovery factor. As the formation pressure at the time of injection increases, the physical property differences between CO₂/N₂ and CH₄ increase significantly, which is conducive to forming a stable displacement front and thereby improving displacement efficiency. However, premature gas injection may lead to insufficient utilization of original reservoir energy, while increasing development costs and reducing economic benefits [26]. Under the condition of reservoir temperature of 100 °C, eight groups of displacement experiments with different gas injection timings were conducted to systematically analyze the influence law of gas injection timing on CH₄ recovery factor.

4.1.1. Analysis of Enhanced Recovery Efficiency and Cumulative Recovery Efficiency Under Different Injection Timing

Figure 3 illustrates the effects of different gas injection timings on recovery enhancement and cumulative recovery factor. As the gas injection timing is advanced, the enhanced recovery effects of both CO₂ and N₂ are significantly strengthened. When the reservoir is depleted to 11 MPa (with an abandonment pressure of approximately 11.25 MPa), the depletion recovery factor reaches 59.82%, and the recovery increments induced by switching to CO₂ and N₂ injection are 24.03% and 23.00%, respectively. In contrast, when the depletion proceeds to 5 MPa (with an abandonment pressure of around 5.55 MPa), the depletion recovery factor rises to 82.41%, whereas the corresponding recovery increments of CO₂ and N₂ injection are only 13.22% and 9.64%, respectively. Elastic depletion development is the most efficient development method for tight sandstone gas reservoirs, which relies on the expansion energy of natural gas itself to produce reserves. This method has no problems of gas channeling and invalid circulation of injected gas, with the highest recovery efficiency and the lowest cost. The early gas injection scheme artificially terminates the depletion development when there is still a large amount of reserves that can be efficiently produced through elastic expansion, and switches to gas flooding with lower displacement efficiency. Although gas flooding brings a higher incremental recovery, it cannot make up for the loss of recovery in the depletion stage. In contrast, the late gas injection scheme first maximizes the utilization of formation elastic energy, produces the majority of reserves through high-efficiency depletion development, and then uses gas flooding to produce the remaining hard-to-recover reserves. The increase in recovery in the depletion stage far exceeds the loss of incremental recovery from gas flooding, so the total cumulative recovery is higher.

Figure 4 presents the differences in density and viscosity among CO₂, N₂, and CH₄ at various pressures under the temperature condition of 100 °C. With the increase in pressure, the differences in physical properties between the injected gases and CH₄ become more pronounced, which is conducive to inhibiting gas-phase miscibility [27,28], thereby forming a stable piston-like displacement front and improving the displacement performance [29]. At low formation pressures, not only is the remaining recoverable CH₄ reserve limited, but miscibility between gases is also prone to occur, resulting in poor displacement efficiency and thus restricting the amplitude of recovery improvement. When gas injection is performed earlier, the pressure at the injection end rises slowly and stabilizes at a low level, reflecting sufficient reservoir energy, low seepage resistance, and favorable gas mobility. Conversely, delayed gas injection leads to a rapid increase in the injection-end pressure to a higher steady value, indicating a significant rise in seepage resistance under the low-energy state. Nevertheless, this also forces the injected gases to overcome higher capillary forces, mobilizing the residual gas trapped in microscopic pores and thereby enhancing the CH₄ recovery factor. In addition, the overall enhanced recovery performance of CO₂ is superior to that of N₂.

The cumulative recovery factor exhibits a decreasing trend with the advancement of gas injection timing. Gas injection at the depletion pressure of 5 MPa yields the highest cumulative recovery factors, reaching 95.63% for CO₂ injection and 92.04% for N₂ injection, respectively. In comparison, the cumulative recovery factors are only 83.85% and 82.82% when CO₂ and N₂ are injected at the depletion pressure of 11 MPa. The experimental results demonstrate that although early gas injection is beneficial for enhancing displacement efficiency [22], it fails to fully utilize the original formation energy of the gas reservoir, resulting in a decline in cumulative recovery factor. Therefore, it is recommended that supercritical CO₂ be injected after sufficient depletion development of the gas reservoir, so as to synergistically optimize the utilization of formation energy and the effect of recovery improvement.

4.1.2. Analysis of Variation in CH₄ Gas Production Rate Under Different Injection Timing

Figure 5 illustrates the effects of different gas injection timings on the average CH₄ production rate and pressure differential. As the gas injection timing is advanced, the average CH₄ production rate increases significantly, whereas the pressure differential shows no obvious variation under different gas injection timings. When gas injection is conducted after depletion to 11 MPa, the average CH₄ production rates corresponding to CO₂ and N₂ displacement are 15.53 mL/min and 3.68 mL/min, respectively. In contrast, when gas injection is implemented after depletion to 5 MPa, the values decrease to 10.17 mL/min and 1.60 mL/min, respectively. During the injection of the two gases, the maximum pressure differentials between the injection end and production end are 1.0 MPa (for CO₂) and 0.4 MPa (for N₂), respectively. A comparison indicates that both the pressure differential and CH₄ production rate during CO₂ injection are significantly higher than those during N₂ injection. This is mainly attributed to the fact that the density of CO₂ is higher than that of N₂. In particular, when the gas injection timing is advanced, CO₂ exists in a supercritical state (as shown in Figure 6), with a density closer to that of liquids. As a result, a larger volume of CO₂ can be injected within the same time frame, leading to higher efficiency in replenishing reservoir energy and a more pronounced displacement effect. In addition, the increased viscosity of supercritical CO₂ helps to slow down its migration rate and mitigate gas miscibility, thereby enhancing the CH₄ production rate [30].

4.1.3. Analysis of Breakthrough Timing Under Different Injection Timing

Figure 7 depicts the variations in CO₂ and N₂ compositions under different gas injection timings, while Figure 8 presents the variations in CO₂ and N₂ compositions as well as the injection volumes at key nodes. The inset graphs in the upper left corner of Figure 8a,b illustrate the initial gas production time corresponding to different gas injection timings. It can be observed that as the gas injection timing is advanced, the breakthrough times of CO₂ and N₂ displacement are correspondingly delayed, whereas the initial gas production time is advanced with the earlier gas injection timing. The criterion is set as follows: gas breakthrough initiates at an injected gas content of 5%, and full gas breakthrough is achieved at 95%. When gas injection is conducted at 5 MPa, the breakthrough times of CO₂ and N₂ are 0.19 HCPV (Hydrocarbon Pore Volume) and 0.17 HCPV, respectively; in contrast, when injection is performed at 11 MPa, the breakthrough times are extended to 0.30 HCPV and 0.28 HCPV, respectively.

In low-pressure environments, CO₂/N₂ tends to be miscible with CH₄, resulting in slow pressure buildup at the displacement front and delayed initial gas production time. Meanwhile, earlier gas breakthrough occurs, which shortens the CH₄ production duration. On the contrary, at high pressures, the differences in physical properties between CO₂/N₂ and CH₄ are significantly enhanced. Such distinct physical property differences can effectively inhibit the mixing of CO₂/N₂ and CH₄. Additionally, the high-viscosity CO₂/N₂ exhibits reduced migration rate and diffusivity in the pores, thereby delaying its breakthrough and facilitating the improvement of CO₂/N₂ displacement uniformity. The results in Figure 6 indicate that the maximum pressure differentials during N₂ injection are generally lower than those during CO₂ injection, suggesting that N₂ displacement takes effect more rapidly. Correspondingly, both the initial gas production time and the breakthrough time of N₂ displacement are earlier than those of CO₂ displacement.

4.2. Effects of CO₂/N₂ Injection Rate

In the process of CO₂/N₂-enhanced gas recovery (CO₂/N₂-EGR), the selection of gas injection rate exerts a significant influence on enhanced recovery efficiency [32]. When the gas injection rate is excessively low, insufficient displacement energy will be supplied, resulting in an insignificant enhanced recovery effect [33]. Whereas an excessively high gas injection rate tends to induce premature gas breakthrough, which is unfavorable for uniform gas displacement and even inhibits the improvement of recovery factor. Therefore, optimizing the CO₂ injection rate is of great importance. Under the conditions of a reservoir temperature of 100 °C and an injection pressure of 7 MPa, 6 groups of displacement experiments with different gas injection rates were conducted to systematically analyze the influence law of gas injection rate on CH₄ recovery factor.

4.2.1. Analysis of Enhanced Recovery Efficiency Under Different Injection Rates

Figure 9 illustrates the influence of different gas injection rates on the enhanced recovery effect. With the increase in gas injection rate, the enhanced recovery effects of both CO₂ and N₂ exhibit a trend of first increasing and then decreasing. When the injection rate is set at 0.05 mL/min, CO₂ and N₂ achieve the optimal enhanced recovery performance, with the recovery increments reaching 17.29% and 14.89%, respectively. In contrast, either excessively low or high injection rates result in inferior enhanced recovery efficiency.

At relatively low injection rates, the contact time between CO₂/N₂ and CH₄ is prolonged, the miscible displacement range is expanded, and the interfacial tension is significantly reduced, thereby decreasing the flow resistance. Additionally, the low displacement velocity facilitates more sufficient displacement of CH₄ trapped in micropores by CO₂/N₂, which increases the proportion of injected gas occupying the pore throats and thus improves the ultimate recovery factor to a certain extent. At the optimal injection rate, the pressure at the injection end rises steadily, corresponding to efficient and stable displacement.

When the injection rate is excessively high, the injection pressure rises sharply and then quickly reaches a plateau, which directly verifies the occurrence of viscous fingering—the injected gas channels through a limited number of preferential pathways, leading to a large amount of energy being consumed in ineffective circulation. Macroscopically, this phenomenon is characterized by high pressure yet low sweep efficiency. Overall, the enhanced recovery performance of CO₂ is superior to that of N₂, which is mainly attributed to CO₂’s stronger adsorption and displacement capacity as well as its excellent miscibility, thus enabling it to exhibit more significant performance in enhancing gas recovery.

4.2.2. Analysis of Variation in CH₄ Gas Production Rate Under Different Injection Rates

Figure 10 depicts the variations in the average CH₄ production rate and pressure differential under different gas injection rates. The average CH₄ production rate specifically refers to the mean value of CH₄ production during the gas injection displacement stage, spanning from the initiation of gas injection to the point where the CH₄ content in the produced gas drops below 5%. The average CH₄ production rate increases significantly with the rise in gas injection rate, whereas the pressure differential shows no obvious variation under different gas injection timings. At a gas injection rate of 0.2 mL/min, the average CH₄ production rates corresponding to CO₂ and N₂ injection are 15.41 mL/min and 2.86 mL/min, respectively; in contrast, at an injection rate of 0.02 mL/min, the values decrease to 2.84 mL/min and 0.76 mL/min, respectively. The maximum pressure differentials during the injection of the two gases are 1.7 MPa (for CO₂) and 0.3 MPa (for N₂), respectively.

A comparison demonstrates that both the pressure differential and CH₄ production rate during CO₂ injection are significantly higher than those during N₂ injection. This is mainly because the plug-type displacement mode of CO₂ enables it to sweep the reservoir more effectively, but it also means that a higher pressure differential is required to drive the relatively uniform displacement front. In contrast, N₂ is more prone to gas channeling, forming high-speed flow pathways, which conversely reduces the subsequent displacement pressure.

4.2.3. Analysis of Breakthrough Timing Under Different Injection Rates

Figure 11 and Figure 12 illustrate the compositional variations and breakthrough times of CO₂ and N₂ under different gas injection rates, respectively. Both the initial gas production time and the breakthrough initiation time are advanced with the increase in gas injection rate. It is analyzed that under high injection rate conditions, the injected gas tends to form preferential flow paths more readily, thus advancing the initial gas production time. At relatively low injection rates, insufficient displacement energy slows down the migration and diffusion of CO₂/N₂ in the pores, thereby delaying their breakthrough times. When the displacement rate increases but remains below the threshold value, the injection pressure gradient increases gradually, which is conducive to improving the displacement uniformity of CO₂/N₂, promoting the faster migration of CO₂/N₂ toward the production end, and shortening the breakthrough time.

4.3. Effects of Gas Injection Compositions

In the process of gas injection displacement, the selection of injected gas composition also exerts an influence on enhanced recovery efficiency. Carbon dioxide (CO₂) is typically derived from tail gas and flue gas generated by fuel combustion [34], which are mixed gases with high impurity content and complex compositions. The use of pure CO₂ for displacement would lead to greater difficulty and higher costs in CO₂ purification; thus, the adoption of impure CO₂ for displacement can effectively address this issue. To this end, under the conditions of a reservoir temperature of 100 °C, an injection pressure of 7 MPa, and an injection rate of 0.1 mL/min, five groups of displacement experiments with different gas compositions were conducted to systematically analyze the influence law of injected gas composition on CH₄ recovery factor.

4.3.1. Analysis of Enhanced Recovery Efficiency Under Different Gas Injection Compositions

Figure 13 illustrates the influence of injected gas mixtures with different proportions on the enhanced recovery effect. When pure CO₂ is used as the injected gas, the optimal enhanced recovery performance is achieved, with the recovery increment reaching 17.02%. In contrast, pure N₂ yields the poorest displacement effect, corresponding to a recovery increment of only 14.64%. With the increase in CO₂ proportion in the injected gas mixture, the enhanced recovery effect exhibits a gradual upward trend. Concurrently, the pressure at the injection end increases significantly as the CO₂ proportion rises.

Beyond the viscosity difference, this phenomenon is mainly attributed to the competitive adsorption effect of CO₂. CO₂ molecules are adsorbed on the surface of pore throats, which is equivalent to temporarily reducing the effective radius of pore throats and thereby increasing the flow resistance. This additional pressure “consumption” is precisely the macro-mechanical manifestation of the microscopic process where CO₂ displaces the adsorbed CH₄. Analysis indicates that an increase in CO₂ proportion elevates the CO₂ content in the pore fluid, which contributes to a more uniform advancement of the displacement front. Meanwhile, CO₂ displacement is characterized by a relatively long breakthrough time, which helps to expand the sweep range and mitigate the fingering phenomenon, thus displacing the CH₄ in pores more effectively. The higher the CO₂ proportion, the more pronounced the aforementioned effects, and consequently, the better the enhanced recovery performance. Overall, pure CO₂ delivers the optimal displacement effect. However, the engineering application of pure CO₂ is confronted with challenges of high purification difficulty and elevated costs. Therefore, the adoption of CO₂/N₂ mixed gases with an appropriate proportion for displacement can ensure the enhanced recovery effect while achieving cost reduction and efficiency improvement.

Our results show that pure CO₂ yields a higher incremental recovery (17.02%) than pure N₂ (14.64%) in tight sandstone. This directly contrasts with numerical simulations of shale reservoirs by Li & Elsworth (2019) [35], who reported that pure N₂ could increase shale gas recovery by 80%, compared to only 20% for pure CO₂ due to severe CO₂-induced matrix swelling restricting fracture permeability. Because tight sandstone lacks substantial kerogen, the matrix swelling effect is minimized; thus, the higher density and viscosity of supercritical CO₂ dominate the displacement efficiency. Additionally, Du et al. (2019) [36] reported absolute product recoveries of 52.90% (CO₂) and 41.60% (N₂) using crushed shale samples in a fixed bed. While our study confirms the positive correlation between CO₂ fraction and recovery, our absolute incremental values (14–17%) are significantly more realistic for field applications, as the 35.8 cm intact long-core preserves the natural seepage resistance and structural heterogeneity that crushed samples eliminate.

The pressure-dependent adsorption behavior of CH₄/CO₂ and CH₄/N₂ mixtures matched that of their respective single-component gases [37]. As shown in Figure 14a, increasing CO₂ proportion in CH₄/CO₂ mixtures enhanced CO₂ adsorption while suppressing CH₄ adsorption. For CH₄/N₂ mixtures (Figure 14b), CH₄ adsorption was higher than N₂ at CH₄/N₂ ratios above 1:3, equal at the 1:3 critical ratio, and lower than N₂ at ratios below 1:3. This crossover confirms that N₂, despite its weaker intrinsic adsorption affinity, can outcompete CH₄ and dominate the adsorption process once its relative concentration exceeds the 1:3 threshold [38].

4.3.2. Analysis of Variation in CH₄ Gas Production Rate Under Different Gas Injection Compositions

Figure 15 depicts the variations in the average CH₄ production rate and pressure differential under different injected gas compositions. With the increase in CO₂ content in the injected gas, the average CH₄ production rate increases gradually. The pressure differential reaches its maximum during pure CO₂ displacement, whereas it shows no obvious variation as the CO₂ content decreases in the mixed gas. The average CH₄ production rate is 12.07 mL/min for pure CO₂ displacement and only 2.44 mL/min for pure N₂ displacement.

The comparison results indicate that pure CO₂ displacement is significantly superior to impure CO₂ displacement in terms of both pressure differential and CH₄ production rate. This is mainly because CO₂ advances in a plug-type displacement mode, which enables it to sweep the reservoir more effectively, but it also requires a higher displacement pressure to drive the relatively uniform displacement front. In contrast, N₂ features a high diffusion coefficient and is prone to gas channeling, forming preferential flow paths and thus leading to a reduction in displacement pressure. Overall, the higher the CO₂ content in the injected gas, the better the displacement efficiency.

4.3.3. Analysis of Breakthrough Timing Under Different Gas Injection Compositions

Figure 16 and Figure 17 illustrate the variations in CH₄ content and breakthrough time under different injected gas compositions, respectively. With the increase in the CO₂ proportion in the injected gas, both the initial gas production time and breakthrough time are correspondingly delayed. Specifically, breakthrough occurs at 0.24 HCPV for pure CO₂ displacement, whereas it takes place at 0.19 HCPV for pure N₂ displacement.

Analysis indicates that as the N₂ proportion increases, the miscibility between the injected gas and CH₄ decreases, leading to premature gas channeling and shortened breakthrough time. The results presented in Figure 13 show that the maximum pressure differentials of mixed gas displacement with N₂ addition are generally lower than those of pure CO₂ displacement, which further demonstrates that mixed gas displacement with N₂ takes effect more rapidly, with both its initial gas production time and breakthrough time being earlier than those of pure CO₂ displacement.

This study macroscopically observed that N₂ breaks through earlier (0.19 HCPV) than CO₂ (0.24 HCPV). This quantitatively aligns with the microscopic mechanisms identified by Du et al. (2019) [36], who used the Coats-Smith dispersion-capacitance model to demonstrate that the longitudinal dispersion coefficient of N₂ is more than 5 times higher than that of CO₂, leading to premature gas breakthrough and excessive mixing. Our intact long-core tests successfully validate this rapid breakthrough phenomenon at a macroscopic engineering scale.

5. Conclusions

This study systematically uses CO₂–N₂ mixtures, and long-core flooding experiments were conducted at 100 °C to investigate the impact patterns of three key factors—injection timing, injection rate, and injection gas composition—on the enhanced recovery of tight sandstone gas reservoirs. The main conclusions reached are as follows:

(1) During the process of CO₂/N₂ displacement for natural gas recovery, the earlier the gas injection timing, the higher the CH₄ production rate and the greater the amplitude of recovery improvement, whereas the cumulative recovery factor decreases conversely. Therefore, it is necessary to maximize the synergistic effect of gas displacement while achieving high energy utilization efficiency, so as to realize the optimization of cumulative recovery factor. This finding provides a quantitative reference range for field determination of gas injection timing and corrects the one-sided understanding that “the earlier the gas injection, the better”, and premature gas injection fails to fully utilize the original formation energy of the natural gas reservoir for depletion development.

(2) During the process of CO₂/N₂ displacement for natural gas recovery, the average CH₄ production rate increases with the rise in gas injection rate, and the extent of CH₄ recovery improvement exhibits a trend of first increasing and then decreasing. This indicates that there exists an optimal injection rate range, which can provide a key direction for the optimization of field operation schemes: specifically, field design should determine the optimal injection intensity between “insufficient energy supplement” and “premature gas breakthrough” through numerical simulation or pilot tests.

(3) During the process of CO₂/N₂ displacement for natural gas recovery, both the average CH₄ production rate and the extent of recovery improvement show a nonlinear increasing trend with the rise in CO₂ proportion in the injected gas. Overall, pure CO₂ displacement delivers the optimal effect. However, the engineering application of pure CO₂ is confronted with the problems of high purification difficulty and elevated costs. Thus, the adoption of CO₂/N₂ mixed gas with an appropriate proportion for displacement can ensure the enhanced recovery effect while achieving cost reduction and efficiency improvement.