Research on Multi-Cycle Injection–Production Displacement Characteristics and Factors Influencing Storage Capacity in Oil Reservoir-Based Underground Gas Storage

Abstract

In order to clarify the feasibility of constructing a gas storage reservoir through synergistic injection and production in the target reservoir, micro-displacement experiments and multi-cycle injection–production experiments were conducted. These experiments investigated the displacement characteristics and the factors affecting storage capacity during the multi-cycle injection–production process for converting the target reservoir into a gas storage facility. Microscopic displacement experiments have shown that the remaining oil is primarily distributed in the dead pores and tiny pores of the core in the form of micro-bead chains and films. The oil displacement efficiency of water flooding followed by gas flooding is 18.61% higher than that of gas flooding alone, indicating that the transition from water flooding to gas flooding can further reduce the liquid saturation and increase the storage capacity space by 2.17%. Single-tube long-core displacement experiments indicate that, during the collaborative construction of a gas storage facility, the overall oil displacement efficiency without a depletion process is approximately 24% higher than that with a depletion process. This suggests that depletion production is detrimental to enhancing oil recovery and expanding the capacity of the gas storage facility. During the cyclic injection–production stage, the crude oil recovery rate increases by 1% to 4%. As the number of cycles increases, the incremental oil displacement efficiency in each stage gradually decreases, and so does the increase in cumulative oil displacement efficiency. Better capacity expansion effects are achieved when gas is produced simultaneously from both ends. Parallel double-tube long-core displacement experiments demonstrate that, when the permeability is the same, the oil displacement efficiencies during the gas flooding stage and the cyclic injection–production stage are essentially identical. When there is a permeability contrast, the oil displacement efficiency of the high-permeability core is 9.56% higher than that of the low-permeability core. The ratio of the oil displacement efficiency between the high-permeability end and the low-permeability end is positively correlated with the permeability contrast; the greater the permeability contrast, the larger the ratio. The research findings can provide a reference for enhancing oil recovery and expanding the capacity of the target reservoir when it is converted into a gas storage facility.

1. Introduction

Guided by the philosophy that “lucid waters and lush mountains are invaluable assets,” and in pursuit of energy structure reform aimed at achieving the “dual carbon” goals (carbon peak and carbon neutrality), China is set to witness a substantial increase in its demand for natural gas [1,2]. Underground gas storage facilities, characterized by their large storage capacity, high storage pressure, and low storage costs, have become an important means for natural gas storage and urban gas supply peak-shaving, primarily serving to ensure the long-term safety of urban gas supply. With increasing national gas consumption, gas reservoir-type storage facilities can no longer meet the gas demand. Therefore, it is necessary to repurpose other types of gas storage facilities to regulate the imbalance between natural gas supply and demand, thereby satisfying peak-shaving requirements [3,4,5]. The integrated construction of oil reservoir-type gas storage facilities in conjunction with natural gas enhanced oil recovery represents a novel development model that has emerged in recent years both domestically and internationally. In China, many oil-bearing structures possess the conditions for conversion into gas storage facilities, offering broad application prospects in further enhancing oil recovery rates from Chinese oil reservoirs, diversifying the types of gas storage facilities, and meeting regional gas supply demands [6].

The research on converting oil reservoirs into gas storage facilities can be traced back to the early 21st century, with the conversion of the Jing 58 gas-cap oil reservoir into a gas storage facility [7]. Since then, the conversion of oil reservoirs into gas storage facilities has seen significant advancements in both theoretical understanding and experimental practices. Currently, domestic and international scholars have primarily focused their research on collaborative gas storage reservoirs toward aspects such as the technical principles of reservoir construction, gas injection for enhanced oil recovery techniques, seepage patterns and phase behavior changes, and multi-cycle stress sensitivity [8,9,10,11]. However, relatively less research has been conducted on displacement characteristics and reservoir capacity laws [12,13]. Physical simulation experiments have mostly been limited to short core sizes under normal temperature conditions. Meanwhile, the optimization of injection–production methods and the study of expansion laws during multi-cycle operations are crucial components in the conversion of oil reservoirs into gas storage facilities. Many experts and scholars have primarily conducted experimental research on factors such as reservoir physical properties, gas injection rates, the number of mutual displacement cycles, pressure prediction, and the buried depth of gas storage reservoirs [14,15,16]. The main experimental methods employed include visualization experiments, core displacement experiments, and physical modeling experiments such as two-dimensional flat plate models. Jiao et al. (2023) [17] conducted gas reservoir injection–production simulation experiments using one-dimensional core and two-dimensional flat models. They established the correlation between the effective permeability of gas and water during different gas injection stages and accurately described the multi-cycle gas injection operation process. Ding et al. (2024) [18], addressing the scarcity of research on the impact of medium and low water saturation on free gas storage capacity, conducted laboratory experiments to simulate multiple cycles of gas injection, soaking, and gas production. Based on the principle of similarity, they utilized the experimental data to establish a quantitative characterization model for the relationship between free gas storage capacity, initial water saturation, and the number of injection–production cycles. Zhang et al. (2020) [19], based on the Wen 23 gas storage reservoir, investigated the flow patterns during multi-cycle injection and production processes as well as the utilization of pore volume in underground gas storage facilities. The research findings indicated that the changes in porosity and permeability caused by variations in injection and production pressures in the Wen 23 gas storage reservoir could be considered negligible. Liu et al. (2024) [20] conducted multi-cycle gas injection–production experiments for gas flooding, clarifying the feasibility of constructing oil reservoir-type gas storage reservoirs and the seepage flow laws during the construction process. Geng et al. (2024) [21] employed a combined experimental and numerical simulation approach to model the integrated construction plan for gas storage facilities involving simultaneous gas injection and production. Their study confirmed the feasibility of this integrated construction method. Bao et al. (2023) [22] employed gas reservoir engineering methods to conduct a dynamic analysis of the actual production and operation of the Wen 96 gas storage reservoir. Based on this analysis, they proposed corresponding technical countermeasures to achieve the designed storage capacity and production targets. Zhang et al. (2024) [23] conducted displacement experiments on water–gas dispersed systems using two etching models: the ball-and-stick standard pore channel model and the pore structure model. They observed the morphological changes and movement characteristics of microbubbles and combined numerical simulation methods to quantitatively analyze the experimental phenomena and the microscopic oil displacement mechanisms. Zhang et al. (2023) [24] utilized microscopic models to investigate the distribution of the remaining oil during water flooding and the microscopic oil displacement mechanisms of emulsion flooding. This research provides a scientific basis and theoretical support for further studies on the in situ emulsification flooding of conventional heavy oil using emulsions. He et al. (2024) [25] conducted multi-cycle mutual displacement relative permeability experiments and long-core injection–production experiments to clarify the effective storage capacity space for converting the Pugu 2 reservoir into a gas storage facility. Ren et al. (2024) [26] employed numerical simulation methods to analyze the impact of salt deposition on the flowing bottom-hole pressure, productivity, and storage capacity. The simulation results indicate that under initial water-bearing conditions, reservoir desiccation can enhance the gas storage capacity of the underground gas storage.

Building upon previous research, this study focuses on reservoir-type gas storage facilities, utilizing actual cores collected from the target reservoir to conduct injection–production experiments through indoor physical simulations. The actual reservoir injection–production processes are simulated through microscopic visualization core displacement experiments and multi-cycle injection–production core displacement experiments under high-temperature and high-pressure conditions. The research findings elucidate the oil displacement efficiency and displacement characteristics during the integrated construction of gas storage facilities. They also reflect the gas-displacing-liquid efficiency during the gas injection and production processes from both macroscopic and microscopic perspectives and analyze the factors influencing storage capacity during multi-cycle injection–production processes. This study provides a basis for optimizing reasonable injection–production schemes for gas storage facilities, thereby achieving the goals of enhancing oil recovery and expanding storage capacity.

2. Basic Overview of the Oil Reservoir

2.1. Characteristics of the Oil Reservoir

The crude oil in the target reservoir is located at the boundary between volatile oil and black oil, classifying it as a weakly volatile oil reservoir. The properties of the crude oil are those of conventional light oil. The average porosity of this reservoir is 23.2%, with an average permeability of 145 mD, and the rock wettability exhibits strong hydrophilic characteristics. The geological reserves amount to 1264.83 × 10⁴ t, with an oil-bearing area of 3.34 km². The target reservoir currently has 146 oil wells and 75 water injection wells. It produces 370 t/d and 8.7 × 10⁴ m³/d of gas per day, with a daily water injection volume of 2430 cubic meters. The recovery factor of crude oil is 10%, and the comprehensive water cut is 65.1%. The oil production rate from the geological reserves is 0.4%. Overall, the reservoir is in a stage characterized by “low-speed stable production with high water cut.”

2.2. Fluid Properties

The surface crude oil in the target reservoir has a density of 0.83 g/cm³, a viscosity of 4.6 mPa·s, a sulfur content of 0.11%, a pour point of 29.36 °C, a wax content of 11.81%, and a resin and asphaltene content of 9.39%. The formation crude oil has a density of 0.66 g/cm³, a viscosity of 1.36 mPa·s, a single-stage gas–oil ratio upon degassing of 185 m³/m³, and a formation volume factor of 1.53. The crude oil properties are characterized as light, low-viscosity, low-sulfur, and high-wax contents. The average relative density of the natural gas is 0.73, with an average methane content of 96.33%, an average ethane content of 3.40%, and an average CO₂ content of 1.70%. It does not contain hydrogen sulfide. The formation water has a salinity of 5462 mg/L, with chloride ion concentrations ranging from 173 to 669 mg/L, and it belongs to the NaHCO₃ water type.

2.3. Temperature and Pressure Systems

The initial formation pressure of the target reservoir is 27.3 MPa, with a pressure coefficient of 1.05 and a saturation pressure of 24.5 MPa. The current formation pressure is 23.2 MPa, the formation temperature is 100 °C, and the geothermal gradient ranges from 3.1 to 4.2 °C per 100 m. The reservoir belongs to a normal pressure and temperature system. The upper operating pressure limit for converting the oil reservoir into a gas storage facility is set at 27.3 MPa, and the lower pressure limit is determined to be 12 MPa.

3. Materials and Methods

The synergistic application of microscopic visualization experiments and long-core displacement experiments is a core approach for validating the multi-cycle injection–production mechanism during the conversion of oil reservoirs into gas storage reservoirs. Microscopic visualization experiments reveal gas–water–rock interactions, residual oil film behavior, capillary force hysteresis effects, etc., providing microscopic mechanistic explanations and guiding the design of parameters for long-core displacement experiments. In contrast, long-core displacement experiments simulate macroscopic responses such as seepage flow laws during multi-cycle injection–production in the reservoir, the evolution of gas–water fronts, pressure transmission, and changes in cyclic efficiency, thereby validating the applicability of microscopic mechanisms at the reservoir scale and quantifying engineering parameters. These two types of experiments, respectively, investigate pore-scale dynamics and near-reservoir-scale seepage flow laws, collaboratively validating the multi-cycle injection–production mechanism. Prior to conducting these experiments, the relative permeability of the oil–gas two-phase system during multi-cycle injection was measured. After five cycles of gas–oil mutual displacement, the residual oil saturation decreased by approximately 6 percentage points, while the residual gas saturation increased by about 3.25 percentage points. Gas–oil mutual displacement can continuously mobilize the remaining oil, resulting in a 2.75% increase in storage capacity.

3.1. Microscopic Displacement Experiment

3.1.1. Experimental Preparation

(1) Experimental apparatus

The experimental apparatus (Figure 1) consists of a Nikon SMZ1270 microscope, a desktop computer (Nikon Corporation, Kumagaya, Japan), a core thin-section holder, a micro-displacement pump, a temperature-control system, and relevant data acquisition software. Figure 2 illustrates the flowchart of the experimental testing process. The microscopic distribution and seepage characteristics of fluids within a real core thin-section model are observed using a reflective microscopic magnification testing method, and dynamic images of the microscopic oil displacement process are automatically captured by a computer.

The analysis process also utilized a high-magnification metallographic microscope, the Nikon LV150N from Japan, as illustrated in Figure 3a. As shown in Figure 3b, the LV150N high-magnification metallographic microscope employs image reconstruction processing technology with varying depth-of-field values to enable magnified observation of rock particles, pore structures, and the distribution of oil, gas, and water on different observation planes with varying depths of field within the thin-section core model. The microscope offers a fine-tuning precision of up to 2 μm for depth-of-field adjustments, with magnification capabilities ranging from 500× to 1000×. It can also superimpose numerical images from different observation planes with varying depths of field to form a quasi-three-dimensional (3D) image of the thin-section core model, thereby obtaining a quasi-3D image of the remaining oil, gas, and water distribution, as illustrated in Figure 3c. Meanwhile, NIS-Elements 6.50.00 software can be utilized for image data processing and the quantification of fluid saturation (with a measurement error within 5%), as illustrated in Figure 3d. Its core principle involves conducting quantitative analysis on microscopic images captured under different conditions (typically before and after water flooding or gas flooding), accurately measuring the proportion of pore area (or volume) occupied by the oil phase in 3D imaging. The detailed steps are as follows:

(a)

Image Acquisition and Preprocessing: Microscopic imaging, adding different colored tracers to the oil and water phases, image calibration, and preprocessing.

(b)

Phase Separation: Accurately classify the pixels in the images into the oil phase, water phase, and rock matrix (boundary of pore space).

Quantitative Calculation of Saturation:

The total pore volume is determined as the volume of the oil phase mask on the initial state image before displacement.

The residual oil volume is the volume of the oil phase mask on the image after water flooding or gas flooding.

The displaced oil volume is calculated by subtracting the residual oil volume from the total pore volume.

Saturation calculations include the following:

The initial oil phase saturation is calculated based on the initial oil-saturated image.

The residual oil phase saturation, which is the ratio of residual oil volume to total pore volume. Finally, the displacement efficiency can be calculated.

Figure 3. Image processing equipment: (a) high-magnification microscope; (b) two-dimensional distribution map of remaining oil and water; (c) three-dimensional distribution map of remaining oil and water; (d) processing the saturation of remaining oil distribution within the field of view.

Figure 3. Image processing equipment: (a) high-magnification microscope; (b) two-dimensional distribution map of remaining oil and water; (c) three-dimensional distribution map of remaining oil and water; (d) processing the saturation of remaining oil distribution within the field of view.

(2) Selection and preparation of core samples

The thin sections used in microscopic displacement experiments are obtained and prepared from natural short core samples taken from the target reservoir. The cores are cut into thin sections with dimensions of 1 cm × 1 cm × 0.3 cm and set aside for future use. The initially prepared core thin sections are shown in Figure 4. The initially prepared core thin sections are then placed inside transparent rubber sleeves, as illustrated in Figure 5. Figure 6 displays the core thin sections that have been successfully mounted. The basic porosity and permeability parameters of the cores are presented in

Table 1. Basic porosity and permeability parameters of the core.

Core No	Permeability (mD)	Porosity (%)
1	156	20.3
2	93	17.2

.

(3) Fluid Selection

The fluids required for the experiment include the injection gas, formation water, and simulated oil. Among them, nitrogen is used as a substitute for the injection gas to facilitate the observation of displacement phenomena and image processing in the later stages of the experiment. The formation water is dyed with methylene blue, as shown in Figure 7a, with a salinity of 5462 mg/L. The simulated oil is prepared by blending aviation kerosene dyed with oil-soluble Sudan Red and hydraulic oil, with a viscosity of 1.36 mPa·s, as illustrated in Figure 7b.

3.1.2. Experimental Procedure

In this microscopic oil displacement experiment, a real core thin-section model was employed to investigate the displacement characteristics. Gas flooding experiments were conducted on Core Sample 1, while water flooding followed by gas flooding experiments were carried out on Core Sample 2. The specific experimental procedures are as follows:

(a)

Cut core samples to make thin sections, which are then placed into the holder and connected to the pipelines.

(b)

Apply confining pressure and saturate the core thin sections with light-colored formation water to establish the initial water saturation.

(c)

Adjust the equipment and select the imaging area of the core model.

(d)

Displace the core thin section saturated with formation water using simulated oil and observe the images until there is no further change in the images.

(e)

Gas flooding for oil displacement: Inject N₂ into the thin-section core model saturated with simulated oil, and continuously capture images of the designated imaging area of the thin-section core model. Observe the dynamic process of the images.

(f)

Water flooding followed by gas flooding for oil displacement: Inject dyed formation water into the thin-section core model saturated with simulated oil, and continuously capture images of the designated imaging area of the thin-section core model. Observe the dynamic process of the images and record them. Stop the injection when there is no further change in the images. Then, switch to N₂ flooding for displacement, observe the dynamic process of the images again, and record them. Stop the N₂ injection when there is no further change in the images.

3.2. Long-Core Displacement Experiment

Through single-tube and parallel double-tube long-core displacement experiments, the injection–production operation mode of gas storage reservoirs is simulated to analyze the effects of different injection–production methods and permeability contrasts on displacement efficiency and storage capacity.

3.2.1. Experimental Preparation

(1) Experimental apparatus

The multifunctional long-core displacement experimental setup is illustrated in Figure 8. This apparatus primarily consists of an intermediate container (1000 mL), pipelines, a core holder (70 MPa), pressure sensors (70 MPa), a back-pressure valve, an oil–water separator, a gas flowmeter, and a constant-pressure constant-rate pump (70 MPa). The experimental procedure for single-tube long-core displacement is depicted in Figure 9, while the procedure for dual-tube long-core displacement is shown in Figure 10.

(2) Selection and preparation of core samples

The core samples obtained from the field were extracted, cleaned, and dried in accordance with the national standard GB/T 28912-2012 [27]. Subsequently, parameters such as the length, diameter, permeability, and porosity of the core samples were measured.

The long core is assembled by splicing short cores. The total length of the single-tube long core is 57.87 cm, with an average permeability of 126 mD and an average irreducible water saturation of 40%. A photograph of the single-tube core is shown in Figure 11, and the basic parameters of the core are listed in

Table 2. Basic parameters of the single-tube long core.

Core No	Core Length (cm)	Permeability (mD)	Porosity (%)
1	4.50	71.6	21.63
2	4.07	68.83	21.59
3	3.43	63.37	21.42
4	3.98	60.42	21.67
5	4.20	93.78	22.58
6	3.80	45.65	21.70
7	4.34	104.37	23.33
8	3.68	113.19	23.55
9	4.47	33.82	20.79
10	4.33	121.95	23.85
11	4.29	28.72	21.79
12	4.34	231.29	22.58
13	3.87	205.29	22.25
14	4.57	521.05	24.67

.

A control experiment was set up when conducting a dual-tube parallel displacement experiment. The total length of the cores is 55.55 cm, with an average permeability of 220.1 mD, which is identical to the permeability of the relatively high-permeability cores selected for comparison. This setup is utilized to contrast and analyze the impacts of homogeneity and heterogeneity on displacement efficiency. The total length of the relatively high-permeability core is 55.56 cm, with an average permeability of 219.2 mD, while the total length of the relatively low-permeability core is 55.37 cm, with an average permeability of 57.64 mD. The permeability contrast ratio is 3.8, and the average irreducible water saturation for both tubes is 40%. A photograph of the dual-tube core is shown in Figure 12, and the basic parameters of the core are listed in

Table 3. Basic parameters of dual-tube parallel long cores (Control Group, 220.1 mD).

Core No	Core Length (cm)	Permeability (mD)	Porosity (%)
1	5.55	223.3	24.97
2	5.63	205.1	23.22
3	6.21	225.7	24.45
4	5.12	241.3	26.86
5	5.47	196.4	22.46
6	6.08	192.6	22.38
7	5.97	189.7	21.10
8	5.11	257.1	23.29
9	5.18	176.3	21.90
10	5.23	293.4	22.15

,

Table 4. Basic parameters of the dual-tube parallel long core (relatively high permeability, 219.2 mD).

Core No	Core Length (cm)	Permeability (mD)	Porosity (%)
1	5.51	221.4	24.54
2	5.72	205.6	24.44
3	6.13	223.4	22.13
4	5.38	239.1	20.56
5	5.61	198.4	25.02
6	5.92	192.7	24.57
7	6.02	191.2	25.07
8	5.04	256.3	25.66
9	4.83	175.9	26.95
10	5.40	290.1	22.08

and

Table 5. Basic parameters of the dual-tube parallel long core (relatively low permeability, 57.64 mD).

Core No	Core Length (cm)	Permeability (mD)	Porosity (%)
1	5.55	40.28	22.11
2	5.61	46.35	19.21
3	5.31	29.94	19.19
4	5.54	64.81	22.76
5	6.40	11.85	20.61
6	5.26	68.04	22.90
7	5.33	82.90	26.83
8	5.44	66.99	25.60
9	5.40	66.61	21.40
10	5.52	98.67	18.51

.

(3) Fluid Selection

The displacement water was formulated formation water (mineralization degree 5462 mg/L), the displacement gas was the on-site injection gas, and the displacement crude oil was the sample oil taken from the target block (density 0.83 g/cm³, volume factor 1.53) and the mixed associated gas, which was prepared according to the original formation conditions (27.3 MPa, 100 °C) and gas–oil ratio (185 m³/m³). The analysis data for formation water are shown in

Table 6. Stratigraphic water analysis data.

Cation			Anion				TDS (mg/L)	PH	Water-Based
Na+, K+	Mg2+	Ca2+	Cl−	SO42−	HCO3−	CO32−
1601	6	7	353	26	3286	60	5462	7~8	NaHCO3

.

(4) Experimental temperature and pressure

Under the experimental condition of 100 °C, the cyclic injection–production pressure window was maintained between 12 and 27.3 MPa, with an upper limit pressure of 27.3 MPa during the injection phase and a lower limit pressure of 12 MPa during the production phase.

3.2.2. Experimental Design

Scheme 1: First, conduct depletion production until the lower pressure limit is reached. Then, close the outlet end and inject gas through the inlet end until the upper pressure limit is achieved. Next, produce gas through the inlet end until the lower pressure limit is reached again. Repeat this process until no more oil is produced.

Scheme 2: First, conduct depletion production until the pressure drops to the lower pressure limit. Then, shut off the outlet end and inject gas through the inlet end until the pressure reaches the upper pressure limit. Subsequently, open both the inlet end and the outlet end simultaneously to produce gas until the pressure decreases back to the lower pressure limit. Repeat this process of injection and production until no further oil is recovered.

Scheme 3: Directly perform a gas drive until no more oil is produced. Then, shut off the outlet end and inject gas through the inlet end until the pressure stabilizes at the upper pressure limit. After stabilization, produce gas through the inlet end until the pressure decreases to the lower pressure limit. Repeat this process of injection and production until no further oil is recovered.

Scheme 4: Directly perform a gas drive until no more oil is produced. Then, shut off the outlet end and inject gas through the inlet end until the upper pressure limit is reached. Subsequently, open both the inlet end and the outlet end simultaneously to produce gas until the pressure decreases to the lower pressure limit. Repeat this process of injection and production until no further oil is recovered.

Scheme 5 (dual-tube parallel control group): Simultaneously subject both tubes (with the same permeability) to direct gas flooding until no oil is produced. Then, close the outlet end and inject gas through the inlet end until the upper pressure limit is reached. Subsequently, open both the inlet and outlet ends to produce gas until the lower pressure limit is achieved. Repeat this process until no more oil is produced.

Scheme 6 (dual-tube parallel experimental group): Simultaneously subject both tubes (with a permeability contrast of 3.8) to direct gas flooding until no oil is produced. Then, close the outlet end and inject gas through the inlet end until the upper pressure limit is reached. Subsequently, open both the inlet and outlet ends to produce gas until the lower pressure limit is achieved. Repeat this process until no more oil is produced.

During the cyclic injection–production stage, the cyclic injection–production process stops when no fluid is produced during injection or production. Injection stops when the upper pressure limit is reached, and production stops when the lower pressure limit is reached. Both the injection and production durations are set at 3 h. The specific simulation stages are detailed in

Table 7. Simulation stages.

Scheme No	Displacement Process
	Gas Flooding	Depletion Development	Cyclic Injection–Production
1	/	Pressure depletion to the lower limit	Cyclic gas injection–production at the inlet
2	/	Pressure depletion to the lower limit	Cyclic gas injection was conducted through the inlet terminal while simultaneous gas production was maintained at both the inlet and outlet terminals.
3	Gas flooding continued until oil production ceased	/	Cyclic gas injection–production at the inlet
4	Gas flooding continued until oil production ceased	/	Cyclic gas injection was conducted through the inlet terminal while simultaneous gas production was maintained at both the inlet and outlet terminals.
5 (Control Group)	Gas flooding continued until oil production ceased	/	Cyclic gas injection was conducted through the inlet terminal while simultaneous gas production was maintained at both the inlet and outlet terminals.
6 (Experimental Group)

.

3.2.3. Experimental Procedure

The experimental procedure for long-core flooding tests consisted of the following steps:

(a)

The irreducible water saturation was established through the following experimental procedure: First, the long-core samples were sorted according to harmonic mean permeability and loaded into a long-core holder. After vacuum pumping to remove residual gases, the cores were saturated with formation water under controlled pressure. Subsequently, surface-degassed crude oil was injected to displace the formation water under reservoir conditions, thereby establishing the target irreducible water saturation.

(b)

The fluid displacement experimental conditions were established by injecting reconstituted reservoir crude oil to displace surface-degassed oil under simulated reservoir temperature and pressure conditions. The displacement process was maintained until a constant producing gas–oil ratio was achieved at the outlet.

(c)

After establishing the experimental fluid conditions, the core flooding tests were systematically conducted. Following each experimental run, the core was thoroughly cleaned using toluene/methanol azeotrope (7:3 v/v) to restore initial wettability. The standardized procedures (a) vacuum saturation and (b) initial fluid saturation were then repeated prior to subsequent tests. This protocol was rigorously executed for all five experimental sets to ensure data reproducibility.

During the experiment, the main parameters listed in

Table 8. The main parameters.

Main Parameters	Unit
Inlet pressure	MPa
Outlet pressure	MPa
Confining pressure	MPa
Back pressure	MPa
Liquid production rate	mL
Gas production rate	mL

were measured and recorded every 0.1 hcpv.

4. Results

4.1. Multi-Cycle Seepage Characteristics

(1) Gas injection displacement

Gas flooding experiments were conducted under irreducible water saturation conditions. The critical pore structures and changes in fluid saturation during the displacement process can be observed from the red ellipses marked in the three images. Figure 13a illustrates the pre-flooding state of the formation water-saturated core, where micro-CT imaging reveals an interconnected water phase distribution throughout the pore network. As illustrated in Figure 13b, following the simulated oil flooding process, the formation water was substantially displaced with a rapid decline in water saturation. Microstructural analysis reveals that despite visible pore spaces, certain regions remain saturated with irreducible water (blue zones), where the trapped water phase exhibits a discontinuous distribution and is encapsulated by the invading oil. Notably, these water clusters maintain only a single pore-throat connection, exhibiting characteristic dead-end pore behavior that severely restricts producibility. Concurrently, the simulated oil (red/yellow phases) predominantly occupies the macroporous network with a continuous distribution, while forming isolated droplets in smaller pores. Quantitative analysis indicates an oil saturation of 59.76% under these conditions. Following core saturation with simulated oil, nitrogen flooding was conducted to displace the oil. The post-flooding fluid distribution (Figure 13c) demonstrates significantly enhanced displacement efficiency in zones with higher porosity and permeability, as evidenced by the preferential nitrogen migration through these favorable flow pathways. With prolonged gas injection duration, both the gas sweep efficiency and gas–liquid interfacial contact area exhibited progressive enlargement. This enhanced displacement dynamics resulted in the continuous production of simulated oil from the core, manifesting as a significant reduction in liquid saturation. During the nitrogen displacement process, the injected gas preferentially invaded macroporous networks, effectively displacing the majority of mobile oil. However, residual oil films persisted on pore walls due to wettability effects, while the remaining oil predominantly existed as disconnected ganglia and pendular rings within isolated pore bodies and constricted pore throats. Following the displacement experiment, specialized software was employed to quantify the gas displacement efficiency, which was determined to be 41.85%.

(2) Sequential water-to-gas displacement process

Under irreducible water saturation conditions, water flooding was carried out until no oil was produced, followed by a switch to gas flooding. The changes in critical pore structures and fluid saturation during the displacement process can be observed from the red ellipses marked in the four images. Figure 14a demonstrates that after core saturation with formation water, the aqueous phase maintained an interconnected distribution within the pore network. Figure 14b demonstrates that simulated oil displacement achieved irreducible water saturation, with relatively low oil saturation and high residual water saturation. Following water flooding (Figure 14c), the simulated oil was largely displaced by formation water, as indicated by the faded red/yellow zones. Residual oil predominantly existed as surface-adhered films and isolated clusters within constricted pore throats. Following gas flooding (Figure 14d), residual oil in partial small pores was further displaced, while the remainder primarily existed as minute oil films and discontinuous droplets adhering to pore surfaces and within constricted pore throats. Sequential water–gas displacement achieved substantial mobilization of the synthetic oil phase, leading to markedly diminished oil saturation levels throughout the core sample. In this experimental series, the water flooding displacement efficiency reached 42.77%, with gas flooding contributing an additional 17.69%, yielding an ultimate recovery factor of 60.46%.

(3) Comparative analysis of two displacement schemes

As shown in

Table 9. Comparison of recovery performance under different displacement methods.

Core NO	Experimental Protocol	Water Flooding Recovery Efficiency (%)	Gas Flooding Recovery Efficiency (%)	Ultimate Recovery Efficiency (%)
1	Gas flooding	/	41.85	41.85
2	Water flooding-to-gas flooding transition	42.77	17.69	60.46

, under irreducible water saturation conditions, the water flooding recovery factor is higher than that of gas flooding. This indicates that during the high oil saturation period, the displacement medium has a certain impact on crude oil recovery, primarily due to viscosity differences and gravitational segregation. The viscosity of gas is typically much lower than that of liquids (such as oil or water), which enhances the mobility of gas in the reservoir. Gas experiences buoyancy in the vertical direction, causing it to move upwards and more easily pass through the porous medium quickly, leading to gas channeling. Consequently, the injected gas breaks through earlier than the formation water, ultimately resulting in lower oil displacement efficiency compared with water flooding followed by gas flooding. After water flooding, when gas flooding is initiated, the oil recovery factor during the gas flooding stage increases by 17.69%, and the storage capacity expands by 2.17%. This suggests that gas injection can further displace the simulated oil that water flooding cannot reach, mobilizing the oil in the tiny pores of the core. In summary, during the oil production stage, the combined effect of water flooding followed by gas flooding yields better oil displacement results than pure gas flooding.

4.2. Impact of Injection–Production Strategies on Displacement Efficiency and Storage Capacity

4.2.1. Depletion Regime

Scheme 1 simulated cyclic gas injection and production at the inlet terminal under depletion conditions, while Scheme 2 modeled cyclic gas injection at the inlet with simultaneous production at both the inlet and outlet terminals under depletion. Figure 15 demonstrates that the cumulative oil recovery reaches 51.62% in Scheme 1 and 51.78% in Scheme 2. During the depletion phase, the oil displacement efficiencies are 50.53% and 49.52% for Scheme 1 and Scheme 2, respectively, showing a marginal difference between the two scenarios. As shown in Figure 16 and Figure 17, during cyclic injection–production, the first two cycles demonstrate significant displacement effects. However, with increasing cycle numbers, although oil continues to be displaced, the incremental displacement efficiency gradually decreases for both schemes, with improvements of only 1.1% and 2.26%, respectively. This indicates that five-cycle operation can further enhance oil recovery, with simultaneous production at both inlet and outlet showing better performance, achieving the gas storage expansion effect.

4.2.2. Non-Depletion Regime

Scheme 3 simulates cyclic gas injection–production at the inlet without depletion, while Scheme 4 models cyclic gas injection at the inlet without depletion coupled with simultaneous production at both the inlet and outlet. As shown in Figure 18, the cumulative oil recovery reaches 74.56% in Scheme 3 and 77.03% in Scheme 4, representing a 2.47% higher displacement efficiency for Scheme 4. During the gas flooding phase, both schemes exhibit comparable displacement efficiencies of approximately 73%. Figure 19 demonstrates that the first five cycles show significant displacement effects, with gradually decreasing efficiencies in subsequent cycles for both scenarios. Figure 20 reveals diminishing increments in cumulative displacement efficiency, confirming that dual-end production (Scheme 4) achieves a better displacement performance and more effectively enhances gas storage capacity expansion.

4.2.3. Comparative Analysis

Firstly, a comparative analysis of displacement efficiency with and without depletion was conducted (Figure 21). The overall displacement efficiency of scenarios without depletion (Schemes 3 and 4) is approximately 24% higher than that of scenarios with depletion (Schemes 1 and 2). The lower oil recovery degree in Scenarios 1 and 2 indicates that adopting a production strategy of primary depletion followed by cyclic injection–production significantly reduces the ultimate oil recovery [28,29,30,31,32,33,34,35,36,37]. This is because, under the depletion production regime, which relies solely on the natural energy of the reservoir to drive the flow of crude oil, it is impossible to effectively maintain the formation pressure over the long term to drive the crude oil towards the production wells. The flowability of crude oil is highly dependent on pressure; the higher the pressure, the easier it is for the crude oil to flow from the rock pores towards the production end. When the pressure drops sharply, the flow rate of the crude oil slows down rapidly, making it ineffective in displacing the crude oil. This ultimately leads to a sharp decline in production and is also detrimental to the expansion of the gas storage capacity. Under a non-depletion production system, the injection of gas is employed to replenish formation energy, and the physical mechanisms through which it enhances displacement efficiency primarily include the following aspects: Firstly, it maintains formation pressure, inhibits the release of dissolved gas, reduces crude oil viscosity, and sustains high oil mobility. Gas injection provides a continuous energy source to maintain a relatively high pressure gradient, overcoming capillary force constraints and driving crude oil from small pores. Secondly, the injected gas mixes with the crude oil, reducing interfacial tension. Thirdly, it prevents reservoir stress-sensitive damage caused by drastic pressure changes; high pressure maintains the rock pore structure, avoiding pore compression under low pressure. Gas flooding effectively overcomes the issues of energy attenuation, fluid degradation, and capillary confinement encountered in depletion production, thereby significantly improving oil displacement efficiency and facilitating the expansion of gas storage capacity. Therefore, during the oil production stage, the formation pressure should be maintained above a certain pressure point to ensure sufficient formation energy.

Secondly, a comparative analysis is conducted to explore the reasons why the cyclic injection–production effect with simultaneous gas production at both ends is better than that with gas production only at the outlet end. In the single-well gas production mode, gas is extracted solely from the production well end. As a result, gas tends to rapidly break through to the production well along high-permeability channels, while gas in low-permeability zones is difficult to extract effectively. This leads to a significant amount of crude oil being trapped in low-permeability areas or dead pores. Moreover, the pressure in regions far from the production well decreases slowly, making it difficult for gas to effectively penetrate these areas and mobilize the crude oil. In contrast, in the dual-well gas production mode, gas is extracted simultaneously from both the injection well and the production well. This creates a stable pressure gradient that traverses the entire core, compelling the gas to flow through more pore channels. It drives the residual oil in the deep parts of the core and near the injection well towards the production end, expanding the swept volume and significantly improving the crude oil mobilization efficiency.

4.3. Impact of Permeability Contrast on Displacement Efficiency and Storage Capacity

Experimental Scheme 5 for the one-dimensional long-core displacement simulation investigated the impact of uniform permeability on displacement efficiency and storage capacity. As illustrated in Figure 22, under conditions of equal permeability, the oil displacement efficiencies in both tubes were essentially the same during both the gas flooding stage and the cyclic injection–production stage. This indicates that a homogeneous formation has little impact on oil displacement efficiency and storage capacity. In contrast, Experimental Scheme 6 simulated the influence of permeability contrast on displacement efficiency and storage capacity. As shown in Figure 23, the final oil displacement efficiency of the high-permeability core was 9.56% higher than that of the low-permeability core, suggesting that oil displacement efficiency is positively correlated with permeability contrast. Consequently, it is evident that a heterogeneous formation significantly affects displacement efficiency and storage capacity. During the gas flooding stage, the high-permeability reservoir is mobilized first. The cumulative recovery degree of crude oil in the high-permeability zone reaches 79.26%, while that in the low-permeability zone is 69.70%. Several methods for improving oil displacement efficiency in low-permeability zones have been identified through a review of the domestic and international literature, including multistage fracturing and fracture network modification [38], temporary plugging and diverting fracturing [39], enhanced imbibition using nanofluids [40], surfactant imbibition [41], CO₂ miscible/near-miscible flooding [42], low-salinity water flooding [43], and deep profile control and oil displacement using nanomicrospheres [44]. Given the high pressure in gas storage reservoirs, fracturing techniques may easily lead to natural gas leakage and can thus be temporarily excluded from consideration. After taking into account cost control and the actual conditions of the oilfield, chemical regulation technologies can be employed to improve microscopic displacement efficiency.

During the cyclic gas injection and production stage, the oil displacement efficiency in low-permeability zones is approximately 3% higher than that in high-permeability zones. The primary reasons for this are as follows: Firstly, it is the countercurrent displacement dominated by capillary forces. During the gas production phase, capillary forces drive the liquid phase in low-permeability zones to displace the gas in a reverse manner. As the system pressure decreases, gas in high-permeability zones, due to their good connectivity and low resistance, rapidly escapes through large pores, leading to a rapid depletion of the gas phase. However, the residual oil is trapped within the pore throats by capillary forces and remains difficult to mobilize. In contrast, within the small pore throats of low-permeability zones, the gas–liquid interface undergoes reverse migration driven by capillary forces, allowing the liquid phase to move from the small pores in low-permeability zones into the large pores in high-permeability zones. Secondly, there is the pressure transmission delay effect. When pressure is reduced during gas production, after the gas in high-permeability zones is rapidly produced, the pressure drops sharply. However, the low-permeability zones still maintain a relatively high pressure, forming a local high-pressure gas source that continuously supplies gas to the high-permeability zones, thereby pushing the liquid phase towards the high-permeability zones and carrying out more gas. Thirdly, low-permeability zones inhibit gas channeling and improve sweep efficiency. Consequently, during the gas production phase, the displacement effect in the low-permeability end is superior to that in the high-permeability end.

Furthermore, by observing the oil displacement efficiencies of Scheme 4 and Scheme 6 in the long-core displacement experiments along with the average core permeability (

Table 10. Average core permeability and oil displacement efficiency for different schemes.

Scheme No	Average Core Permeability (mD)	Oil Displacement Efficiency (%)
6 (Low Permeability)	57.6	69.7
4	126.0	77.0
6 (High Permeability)	219.2	79.3

), it can be observed that the oil displacement efficiency increases with the rise in average core permeability. However, as the average core permeability increases, the magnitude of the increase in oil displacement efficiency diminishes. A regression analysis conducted on this data reveals this relationship, as depicted in Figure 24.

5. Conclusions

This paper focuses on the conversion of an oil reservoir into a gas storage facility. By establishing microscopic displacement experiments and multi-cycle injection–production experiments, we investigated the displacement characteristics and factors affecting storage capacity during the multi-cycle injection–production process for converting the target reservoir into a gas storage facility. The specific conclusions are as follows:

(1)

Microscopic displacement experiments indicate that as the gas injection time increases, the gas sweep area expands, and the contact area between the gas and liquid phases also continuously enlarges. As a result, the simulated oil is progressively displaced, and the decline in liquid saturation is significant. The remaining oil is primarily distributed in the dead pores and tiny pores of the core in the form of micro-bead chains and films. After the displacement process, professional software was used to calculate the oil displacement efficiency, which was found to be 41.85% for gas flooding and 60.46% for the transition from water flooding to gas flooding. This suggests that gas injection can further displace the simulated oil that was not reached by water flooding, thereby reducing the liquid saturation and increasing the storage capacity space by 2.17%.

(2)

Single-tube long-core displacement experiments reveal that during the collaborative construction of a gas storage facility, the overall oil displacement efficiency without a depletion process is approximately 24% higher than that with a depletion process. This indicates that depletion production is detrimental to enhancing oil recovery and expanding the capacity of the gas storage facility. During the cyclic injection–production stage, the crude oil recovery rate increases by 1% to 4%. As the number of cycles increases, the incremental oil displacement efficiency in each stage gradually decreases, and so does the increase in cumulative oil displacement efficiency. Better capacity expansion effects are achieved when gas is produced simultaneously from both ends.

(3)

Parallel double-tube long-core displacement experiments demonstrate that when the permeability is the same, the oil displacement efficiencies during the gas flooding stage and the cyclic injection–production stage are essentially identical. When there exists a permeability contrast, the oil displacement efficiency of the high-permeability core is 9.56% higher than that of the low-permeability core. The ratio of the oil displacement efficiency between the high-permeability end and the low-permeability end is positively correlated with the permeability contrast; the greater the permeability contrast, the larger the ratio.

Based on the above experimental results, during the oil production phase of collaborative gas storage construction in the target reservoir, water flooding can be initially employed followed by gas flooding. In the collaborative phase, both gas injection and oil production should be considered. During the operation phase, only gas injection wells are used for gas injection, while both injection and production wells are utilized for gas production simultaneously. This approach can achieve the objectives of enhancing oil recovery and expanding storage capacity. The research findings can provide a reference for designing a reasonable injection–production scheme to converting the target oil reservoir into a gas storage facility.