Experimental Study on the Optimization of CO₂ Displacement and Huff-n-Puff Parameters in the Conglomerate Reservoirs of the Xinjiang Oilfield

Abstract

Addressing the issue of poor water injection development effectiveness caused by strong water sensitivity damage in the conglomerate reservoirs of the Xinjiang Oilfield, this paper carries out experimental research on CO₂ displacement and CO₂ huff-n-puff to improve oil recovery in reservoirs under the conditions of reservoirs (86 °C, 44 MPa) by using a high-temperature and high-pressure large physical modeling repulsion device based on the artificial large-scale physical modeling of conglomerate oil reservoirs in the Xinjiang oilfield. The results showed that at any displacement rate, CO₂ displacement exhibits the trend where oil production initially increases and then decreases. The higher the gas injection rate, the higher the initial oil well production, and the shorter the time it takes for CO₂ to break through to the bottom of the well. After a breakthrough, production declines more rapidly. The oil recovery rate varies with different gas injection rates, initially increasing and then decreasing as the injection rate changes. The highest oil recovery rate was observed at an injection rate of 1.5 mL/min (equivalent to 38 t/d in the field). The efficiency of CO₂ displacement with multiple injection-production cycles is low; on the same scale of gas injection, single-cycle injection and production were more effective than multiple-cycle injection and production. CO₂ huff-n-puff can improve oil recovery, with a higher CO₂ injection pressure and a longer shut-in time leading to greater oil recovery. As the shut-in time increases, the efficiency of CO₂ oil exchange also improves. The strong supply capacity of the large physical model results in a tendency for the oil production curves of multiple huff-n-puff cycles to converge.

1. Introduction

Against the backdrop of intensified global climate change and the goals of “carbon peak and carbon neutrality,” a large-scale application of CO₂ resources is becoming one of the key technologies for zero and negative carbon emissions [1]. Injecting CO₂ can achieve effects such as reducing interfacial tension, decreasing crude oil viscosity, improving the mobility ratio, and conducting miscible extraction [2]. Consequently, CO₂ displacement and CO₂ huff-n-puff are increasingly used to enhance oil recovery, with more noticeable effects from CO₂ injections [3,4]. Both domestic and international scholars have conducted extensive research on this topic from different perspectives through physical experiments and numerical simulations.

Regarding CO₂ displacement technology, many researchers have conducted long-core displacement experiments and numerical simulations to study the mechanisms and efficiency of CO₂ displacement under various conditions, including different displacement pressures [5], CO₂ concentrations [6,7], displacement methods [8], and properties of crude oil and reservoir rocks [9]. Studies have found that higher displacement pressures and higher CO₂ concentrations lead to improved oil recovery efficiency, and segmented plug displacement can effectively enhance recovery rates. Fakher et al. used NMR technology to monitor CO₂ displacement in real time, revealing the migration patterns of CO₂ and studying the efficiency of CO₂ displacement from a microscopic perspective [10]. Talapatra et al. investigated the impact of layered rock characteristics on oil recovery, studying how factors such as permeability and porosity affect CO₂ displacement [11]. Wang et al. clarified the adaptability and mechanisms of CO₂ displacement in reservoirs, examining the effects of the CO₂ injection sequence, plug size, and injection rate on oil recovery [12]. Li et al. introduced a multi-agent deep deterministic policy gradient algorithm to study the recovery rates of CO₂ displacement under different injection and production rates [13]. Ahmadi and Chen et al. conducted alternating experiments with supercritical CO₂ and polymers, exploring the impact of different displacement media on oil recovery [14,15]. Ghasemi et al. compared the effects of nitrogen and CO₂ on oil recovery, identifying extraction as the primary reason for CO₂’s superior performance [16]. Andreeva et al. used Monte Carlo methods to simulate CO₂ displacement efficiency at the core scale for different oil compositions [17]. Wang et al. employed molecular dynamics to study the diffusion and mass transfer of CO₂ from crude oil during CO₂ displacement, examining the effects of the temperature, gas-to-oil ratio, and water content on recovery efficiency [18].

In addition to enhancing oil recovery through CO₂ displacement, CO₂ huff-n-puff can also effectively improve oil recovery. The mechanisms by which CO₂ huff-n-puff enhances oil recovery include energy replenishment, viscosity reduction through expansion, gas dissolution drive, extraction and leaching, and the reduction of oil–water interfacial tension [19,20,21]. Pu et al. conducted laboratory CO₂ huff-n-puff experiments on tight oil reservoirs and found that the oil swelling coefficient and CO₂ solubility increased with pressure, with higher injection pressures resulting in better huff-n-puff performance [22]. Song et al. performed oil recovery experiments with water flooding, non-miscible CO₂ huff-n-puff, near-miscible CO₂ huff-n-puff, and miscible CO₂ huff-n-puff, and observed that miscible CO₂ huff-n-puff yielded the highest recovery rate [23]. Sun et al. used numerical simulations to study the effects of the CO₂ diffusion rate, cycle number, CO₂ injection rate, and shut-in time on CO₂ huff-n-puff oil recovery [24]. Wei and Zhang suggested that diffusion is a key factor in enhancing oil recovery during CO₂ huff-n-puff [25,26]. Xue et al. conducted CO₂ huff-n-puff experiments on cores with different permeabilities, examining the recovery rates under various pressure conditions [27]. Tang et al. studied the phase behavior and microscopic flow mechanisms of CO₂ huff-n-puff in porous media [28]. Song et al. compared the oil recovery effects of N₂ and CO₂ huff-n-puff and found that CO₂ huff-n-puff performed better during multiple cycles [29].

Based on the above investigations and analyses, domestic and international scholars have studied the influencing factors and mechanisms for enhancing oil recovery through CO₂ displacement and CO₂ huff-n-puff techniques from both experimental and numerical simulation perspectives. Most of the experiments used one-dimensional, long-core displacement, which tends to overestimate the oil displacement efficiency. Therefore, there is a need to conduct three-dimensional displacement experiments to better simulate the actual reservoir conditions for CO₂ displacement and CO₂ huff-n-puff oil recovery. A conglomerate reservoir in the Xinjiang Oilfield, with a reservoir temperature of 86 °C, formation pressure of 44 MPa, and fracture pressure of 48 MPa, was studied. The well’s natural productivity is low, with rapid production decline (Figure 1), a short natural flow production period, and poor crude oil mobility in the reservoir. Due to the strong water sensitivity of the reservoir, water injection has a very poor effect on enhancing oil recovery, making it necessary to use more efficient displacement media for enhanced oil recovery and to clarify the mechanisms. Therefore, this study uses real rock fragments from the study area to establish a large-scale physical model that closely matches the physical properties of the area. Under reservoir conditions, CO₂ displacement and CO₂ huff-n-puff experiments were conducted. The experiments compared and analyzed the effects of these techniques on enhancing oil recovery, determined the optimal CO₂ injection rate for the study area, and identified strategies for improving oil recovery through CO₂ huff-n-puff. This provides some guidance for formulating enhanced oil recovery schemes for similar reservoirs.

2. Large-Scale Physical Displacement Model and Fluid Properties

2.1. Large-Scale Physical Displacement Model

A large-scale physical displacement model was constructed using real reservoir rock cuttings, combined with sand, cement, and adhesive solutions. The physical properties of the model were controlled by adjusting the ratio of cement to adhesive. The model’s dimensions were scaled down proportionally according to the study area, and the well locations were arranged to match the actual well distribution in the reservoir (Figure 2). The porosity and permeability values were designed based on the actual reservoir properties (

Table 1. Design parameters of the large-scale physical model.

Parameter	Geometric Dimensions	Permeability	Porosity	Permeability Variation
Model	Diameter: 0.44 m, Thickness: 0.3 m	0.01~1.75 mD	7.8%	1750

).

2.2. Properties of Experimental Fluid

In this experiment, the oil sample used was crude oil obtained from the conglomerate reservoirs of the Xinjiang Oilfield. Under reservoir temperature and pressure conditions (86 °C, 44 MPa), the crude oil had a viscosity of 1.58 mPa·s and a density of 0.813 g/cm³. The CO₂ used in the experiment was industrial-grade gas, and the minimum miscibility pressure of the CO₂ with the reservoir crude oil was determined to be 26.2 MPa through slim tube experiments (Figure 3). The pressure and temperature of the experiment meet the conditions for miscible displacement. The water used in the experiment was formulated formation water of the CaCl₂ type, with a salinity of 25,227 mg/L and a density of 1.03 g/cm³.

2.3. Conversion of Reservoir and Experimental Flow Parameters

To ensure that the flow behavior in a large-scale physical model is consistent with that of the reservoir, and to predict prototype flow conditions based on the experimental results of the large-scale physical model, it is necessary to determine the experimental parameters and conditions using similarity criteria. The flow of the model is primarily influenced by viscous forces, and in this study, the Reynolds similarity criterion was used to convert the flow parameters. The geometric scale ratio between the study area and the model was calculated to be 17,500. Based on this geometric scale, the similarity conversion was carried out, where the flow rate scale was equal to the length scale. A converted experimental flow rate of 1 mL/min represents an actual field production rate of 25.2 m³/d. Similarly, an experimental CO₂ injection rate of 1 mL/min corresponds to a field CO₂ injection rate of 20.48 t/d (with a density of 0.813 g/cm³). The viscosity and density of the experimental fluid, the experimental temperature and pressure conditions, the permeability, the porosity values, and the permeability variation range all match the actual production conditions in the reservoir, with a scale ratio of 1. The large-scale physical model in this paper is constructed by mixing and ramming real reservoir cuttings, binder solution, and quartz sand. Based on the physical properties of the reservoir, the proportion of the binder solution and the number of cuttings used are determined, followed by the consolidation and ramming of the model.

3. Experimental Conditions and Procedures

In this experiment, a self-developed three-dimensional large-scale core gas displacement experimental device (Figure 4) was used to study the feasibility of enhancing oil recovery in the conglomerate reservoirs of the Xinjiang Oilfield through CO₂ displacement and CO₂ huff-n-puff. This was done by measuring the changes in oil production during the displacement and huff-n-puff processes.

The specific experimental steps for CO₂ displacement are as follows: (1) Place the large-scale physical model into the experimental device, connect the apparatus, heat the system to 86 °C, and clean and dry the large-scale physical model. (2) Apply a confining pressure of 44 MPa and conduct a leakage test on the equipment. (3) Saturate the large-scale physical model with formation water under ambient temperature and pressure conditions, record the volume of water saturation in the rock, calculate the porosity of the core based on the saturated water volume, and use field crude oil to displace the formation water until the oil saturation reaches 50%. (4) Increase the temperature and pressure, setting the experimental temperature to 86 °C and the initial pressure to 44 MPa. (5) To eliminate the impact of depletions in production on CO₂ displacement, set the outlet backpressure to 44 MPa. (6) Conduct CO₂ displacement experiments with one injection and one production, one injection and two productions, and one injection and three productions. The specific experimental parameters are shown in

Table 2. Parameters of the CO₂ displacement experiment.

Experiment Type	Injection Well	Production Well	Displacement Rate	Cumulative Das Injection Volume At Experiment Stop
One Injection, One Production	well 5	well 8	0.5, 11.5, 2 mL/min	0.3 L
One Injection, Two Productions	well 5	well 8, 9	1.5, 2 mL/min	0.6 L
One Injection, Three Productions	well 5	well 6, 8, 9	1.5/min	1.2 L

.

The steps for the CO₂ huff-n-puff experiment are as follows: Steps (1) to (3) are the same as those for CO₂ displacement. (4) Connect the backpressure valve to the inlet, increase the temperature and pressure, set the experimental temperature to 86 °C, and the initial pressure to 44 MPa. (5) Conduct the CO₂ huff-n-puff experiment. For easier measurement, gradually reduce the outlet backpressure by 2 MPa each time to allow depletion in production until the outlet backpressure reaches 30 MPa. (6) Stop the experiment when there is no oil production. The specific experimental parameters are shown in

Table 3. Parameters of the CO₂ Huff-n-Puff experiment.

Shut-In Time	Huff-n-Puff Well	Injection Pressure	Remarks
2 h	10	46, 48, 50 MPa	Different injection pressures
2, 4, 6 h	10	46 MPa	Different shut-in times

.

4. Experimental Results and Analysis

To determine the feasibility of enhancing oil recovery in the conglomerate reservoirs of the Xinjiang Oilfield using CO₂ injection, a large-scale physical model CO₂ displacement simulation experiment was conducted under reservoir conditions. This experiment aimed to establish the relationship between the oil production rate and cumulative oil production, and to study the oil recovery efficiency of CO₂ flooding at different displacement rates. The goal was to determine the optimal CO₂ displacement rate for the conglomerate reservoirs in the Xinjiang Oilfield. Additionally, a large-scale physical model CO₂ huff-n-puff simulation experiment was conducted under reservoir conditions. This part of this study focused on examining the variations in oil production in different huff-n-puff cycles, and analyzing the impact of different injection pressures and shut-in times on the oil well recovery efficiency.

4.1. CO₂ Displacement Experiments

Measure the cumulative mass of the liquid produced at different time intervals at the outlet. After reading the produced water volume, calculate the produced water mass at the outlet, and then calculate the cumulative oil production at different times. Based on the cumulative oil production, calculate the average oil production during each measurement period. Subsequently, plot the oil production and cumulative oil production graphs for CO₂ flooding under different displacement rates for a single injection with a single production well, a single injection with two production wells, and a single injection with three production wells (Figure 5, Figure 6, Figure 7 and Figure 8). Finally, calculate the recovery factor under different injection and production scenarios based on the model’s reserves (

Table 4. Statistics of continuous CO₂ displacement recovery rates for one injection with one production well.

Displacement Rate	0.5 mL/min	1 mL/min	1.5 mL/min	2 mL/min
Cumulative Oil Production (g)	67.04	59.13	73.9	67.57
Recovery	4.53%	4.00%	4.99%	4.57%
Cumulative Gas Injection (L)	0.28	0.27	0.27	0.27
Oil Replacement Ratio	29.52%	26.82%	33.29%	30.64%

and

Table 5. Statistics of recovery rates for single injection with two production wells and single injection with three production wells.

Experiment Type	2 mL/min (5 Injections, 89 Productions)	1.5 mL/min (5 Injections, 89 Productions)	1.5 mL/min (5 Injections, 689 Productions)
Cumulative Oil Production (g)	70.68	81.03	121.87
Recovery	4.78%	5.48%	8.24%
Cumulative Gas Injection (L)	0.63	0.59	1.12
Oil Replacement Ratio	13.69%	16.97%	13.35%

).

Based on the results of the CO₂ displacement experiment with a single injection with a single production well, it is observed that without the impact of depletion, at any displacement rate, as gas injection proceeds, the oil well production first increases and then decreases. The higher the injection rate, the stronger the oil displacement capability, and the higher the peak oil production. When the injection rate is 2 mL/min, the peak oil production is 5.44 g/h, which is 4.73 times higher than the peak oil production at an injection rate of 0.5 mL/min. The higher the injection rate, the shorter the breakthrough time of CO₂ through the bottom of the well, and the faster the production decreases after the breakthrough. After the displacement process ends, the cumulative oil productions at injection rates of 0.5 mL/min, 1.0 mL/min, 1.5 mL/min, and 2.0 mL/min are 44.93 g, 43.91 g, 69.88 g, and 52.56 g, respectively. Given that the model’s reserves are 1479.7 g, the calculated recovery factors at these rates are 4.53%, 4.00%, 4.99%, and 4.57%, respectively. The highest cumulative oil production and recovery factor occur at an injection rate of 1.5 mL/min, indicating that the optimal gas injection rate for this Xinjiang conglomerate reservoir with one injection and one production well is 30.72 t/d. The recovery factor of the oil well initially decreases and then increases with the increase in the injection rate. A higher injection rate leads to a higher displacement efficiency but a shorter CO₂ breakthrough time and lower sweep efficiency. Conversely, a lower injection rate results in a lower displacement efficiency but a higher sweep efficiency, leading to a higher recovery factor at 0.5 mL/min compared to 1.0 mL/min.

In the CO₂ displacement experiment with a single injection with two production wells, the production trends are consistent with those observed in the single injection with single production well experiment, showing an initial increase followed by a decrease. The larger the injection volume, the higher the peak oil production. After the displacement process ends, the cumulative oil productions at injection rates of 1.5 mL/min and 2 mL/min are 81.03 g and 70.68 g, with recovery factors of 5.48% and 4.78%, respectively. The highest recovery factor occurs at an injection rate of 1.5 mL/min, consistent with the findings from the one injection and one production well experiment. Low-speed displacement can increase the sweep area and, to some extent, improve the oil well recovery factor. In the experiment with one injection and three production wells, using an injection rate of 1.5 mL/min resulted in a cumulative oil production of 121.87 g and a recovery factor of 8.24%.

The oil exchange rate and recovery factor in the single injection with single production well experiment show consistent trends, with an average oil exchange rate of 30.07%. The average oil exchange rates for the single injection with two production wells, and single injection with three production wells are 15.33% and 13.35%, respectively. The highest oil exchange rate is observed in the single injection with single production well experiment, indicating a better displacement efficiency and a better oil displacement effect with the same volume of CO₂ injections. Therefore, in the actual development of oil reservoirs, the single injection with single production well method should be prioritized when formulating CO₂ displacement schemes.

4.2. CO₂ Huff-n-Puff Experiment

The cyclic injection test was conducted on a single well with an abandonment pressure set at 30 MPa. Four rounds of CO₂ huff-n-puff experiments were carried out for each set of tests to study the effects of different injection pressures and soaking times on the CO₂ huff-n-puff efficiency. The oil recovery curves were plotted under varying injection pressures and shut-in times for the four rounds of CO₂ huff-n-puff (Figure 9, Figure 10, Figure 11, Figure 12 and Figure 13). The cumulative oil production results from the experiment are shown in

Table 6. Results of the CO₂ huff-n-puff experiment.

Shut-in Time (h)	Injection Pressure (MPa)	Cumulative Oil Production (g)				Total (g)
		First Cycle	Second Cycle	Third Cycle	Fourth Cycle
2	46	14.62	11.55	10.59	9.82	46.57
2	48	14.93	11.98	10.98	10.38	48.26
2	50	16.83	13.59	12.05	11.24	53.71
4	46	15.32	11.77	10.66	10.81	48.56
6	46	15.98	11.85	11.67	11.50	50.99

.

Based on the results of the CO₂ huff-n-puff experiment, it is evident that with a shut-in time of 2 h, the cumulative oil production at injection pressures of 46 MPa, 48 MPa, and 50 MPa were 46.57 g, 48.26 g, and 53.71 g, with recovery factors of 3.15%, 3.26%, and 3.63%, respectively. At an injection pressure of 48 MPa, the cumulative oil production for shut-in times of 2 h, 4 h, and 6 h were 46.57 g, 48.56 g, and 50.99 g, with recovery factors of 3.15%, 3.28%, and 3.45%, respectively.

As the number of huff-n-puff cycles increases, the cumulative oil production per cycle decreases, with the rate of decline slowing. Under the same shut-in time, the cumulative oil production over four cycles increases with the higher injection pressure; the higher the injection pressure, the greater the gas injection volume, the wider the sweep area, and the better the oil displacement effect. Under the same injection pressure, the longer the shut-in time, and the greater the cumulative oil production over the four cycles. During the CO₂ huff-n-puff experiment with a 6 h shut-in time, the longer shut-in time allowed for more thorough diffusion and dissolution of CO₂ into the reservoir, and the strong supply capacity of the large-scale physical model resulted in the oil production curves for the last three cycles nearly overlapping.

Since the oil production curves of the subsequent cycles nearly overlap after a 6 h shut-in period, further extending the shut-in time has no significant impact on oil recovery. Therefore, the optimal shut-in time for the model is determined to be 6 h. Increasing the injection pressure and shut-in time effectively enhances the CO₂ huff-n-puff recovery factor. Given that the reservoir fracturing pressure is 48 MPa, the optimal parameter combination for the CO₂ huff-n-puff experiment is an injection pressure of 48 MPa with a shut-in time of 6 h.

4.3. Feasibility Analysis of Enhanced Oil Recovery

In the absence of depletion production effects, CO₂ flooding exhibits a favorable oil displacement performance, achieving an average oil exchange rate of 30.07% in the one injection, one production scenario, thereby effectively enhancing the reservoir’s recovery factor. A depletion production experiment was conducted on well 10, reducing the reservoir pressure to an abandonment pressure of 30 MPa, which resulted in a cumulative oil production of 6.33 g. The cumulative oil production from the four cycles of CO₂ huff-n-puff exceeded that of the depletion production, leading to a significant improvement in the recovery factor. Therefore, both CO₂ displacement and CO₂ huff-n-puff can effectively enhance the recovery of conglomerate reservoirs in the Xinjiang Oilfield.

5. Conclusions

In conclusion, this study addresses the challenges of poor water injection development effectiveness in the conglomerate reservoirs of the Xinjiang Oilfield due to strong water sensitivity damage. Through experimental research conducted using a high-temperature and high-pressure large physical modeling displacement device, we have gained valuable insights into the effectiveness of CO₂ displacement and CO₂ huff-n-puff techniques in improving oil recovery under reservoir conditions (86 °C, 44 MPa). Our findings highlight that the optimal gas injection rate for maximum oil recovery is 1.5 mL/min (equivalent to 38 t/d in the field). Higher CO₂ injection pressure and longer shut-in times resulted in greater oil recovery. These findings provide important guidelines for optimizing CO₂ displacement and huff-n-puff operations in conglomerate reservoirs. Future research will focus on the following: evaluating the long-term stability of the CO₂ displacement and huff-n-puff processes; the scalability of the experimental results to field-scale implementation; and the environmental impact. The following conclusions are drawn:

(1)

The recovery rates at injection rates of 0.5 mL/min, 1.0 mL/min, 1.5 mL/min, and 2.0 mL/min are 4.53%, 4.00%, 4.99%, and 4.57%, respectively. The oil recovery rate is highest at an injection rate of 1.5 mL/min. After adjusting proportionally, the optimal injection rate for the Xinjiang conglomerate oil reservoir for single injection and production is 30.72 t/d.

(2)

The average oil replacement ratio for single injection and production is 30.07%, for single injection and two productions is 15.33%, and for single injection and three productions is 13.35%. The highest oil replacement ratio is achieved with a single injection and production, indicating better displacement efficiency.

(3)

In the CO₂ displacement experiments, with the same soaking time, higher injection pressure results in a higher oil recovery rate. For a shut-in time of 2 h, the recovery rates at injection pressures of 46 MPa, 48 MPa, and 50 MPa are 3.15%, 3.26%, and 3.63%, respectively. At the same injection pressure, a longer shut-in time results in a higher oil recovery rate. At an injection pressure of 48 MPa, the recovery rates for shut-in times of 2 h, 4 h, and 6 h are 3.15%, 3.28%, and 3.45%, respectively.

(4)

Under experimental conditions, CO₂ in a supercritical state has a good displacement and viscosity-reducing effect on crude oil, effectively improving the oil recovery rate.