Evaluation of Miscible Gas Injection Strategies for Enhanced Oil Recovery in High-Salinity Reservoirs

Abstract

This study presents a comprehensive evaluation of miscible gas injection (MGI) strategies for enhanced oil recovery (EOR) in high-salinity reservoirs, with a focus on the Raleigh Oil Field. Using a calibrated Equation of State (EOS) model in CMG WinProp™, eight gas injection scenarios were simulated to assess phase behavior, miscibility, and swelling factors. The results indicate that carbon dioxide (CO₂) and enriched separator gas offer the most technically and economically viable options, with CO₂ demonstrating superior swelling performance and lower miscibility pressure requirements. The findings underscore the potential of CO₂-EOR as a sustainable and effective recovery method in pressure-depleted, high-salinity environments.

1. Introduction

The global energy landscape continues to rely heavily on hydrocarbon resources, necessitating the development of advanced recovery techniques to extract remaining oil from mature and challenging reservoirs. Among tertiary recovery methods, miscible gas injection (MGI) has emerged as a highly effective Enhanced Oil Recovery (EOR) strategy, particularly in light and medium oil reservoirs. MGI involves injecting gases such as carbon dioxide (CO₂), nitrogen (N₂), or hydrocarbon-rich gases into the reservoir at pressures exceeding the minimum miscibility pressure (MMP), enabling the gas and oil to mix into a single phase. This miscibility eliminates interfacial tension and significantly enhances oil displacement efficiency [1,2].

The determination of MMP is essential for optimizing EOR processes. According to Song et al. [3], MMP can be determined through two primary approaches: experimental and theoretical methods. Experimental techniques such as the slim tube test (STT), rising bubble apparatus (RBA), and vanishing interfacial tension (VIT) are widely used in laboratory settings [4]. Among them, STT is considered the most accurate and reliable, though it is time-consuming and costly. RBA and VIT offer faster and more economical alternatives but are more susceptible to human error and experimental limitations [5,6].

Theoretical calculation methods are increasingly favored for their cost-effectiveness and ability to estimate MMP dynamically during reservoir development. These include empirical correlations, artificial intelligence algorithms, and the Equation of State (EOS) approach. EOS methods are particularly valued in the oil industry for their ability to simulate phase behavior in complex multicomponent systems. Models such as the van der Waals, Redlich–Kwong (RK), Soave–Redlich–Kwong (SRK), and Peng–Robinson (PR) equations are based on thermodynamic principles and phase equilibrium theory [3]. They classify crude oil components into light (C₁), intermediate (C₂–C₆), and heavy (C₇⁺) fractions and use parameters like temperature, pressure, and molecular interactions to predict MMP [7].

Among these, the Peng–Robinson EOS (PR-EOS) is particularly noted for its improved accuracy in predicting fluid densities and phase behavior in systems with polar components. While EOS models offer a practical and relatively accurate alternative to experimental methods, they require iterative computation and may be less sensitive to variations in crude oil composition and reservoir conditions. Despite these limitations, EOS-based approaches remain a cornerstone in MMP prediction due to their balance of accuracy, flexibility, and computational efficiency [8].

MMP is a critical parameter in MGI for EOR processes. While MGI has been extensively applied in conventional reservoirs, its application in high-salinity reservoirs presents both challenges and opportunities. High-salinity environments, characterized by elevated concentrations of dissolved salts in formation water, can adversely affect rock wettability, fluid flow, and chemical interactions. These conditions often reduce the effectiveness of water-based EOR methods, such as polymer or surfactant flooding, due to issues like polymer degradation and surfactant precipitation. In contrast, gas-based methods like MGI are less sensitive to salinity, making them a promising alternative [9].

Recent studies have demonstrated the potential of MGI in high-salinity settings. Davarpanah et al. [1] found that rich gas injection outperformed lean gas and CO₂ in a naturally fractured, high-salinity reservoir, achieving the highest oil recovery due to superior miscibility and delayed gas breakthrough. Pourhadi and Fath [10] showed that CO₂ injection achieved full miscibility across a compositionally graded reservoir, with water-alternating-gas (WAG) injection further enhancing sweep efficiency. Wei et al. [11] observed that while CO₂ injection improved sweep efficiency in carbonate reservoirs, it also triggered asphaltene instability, necessitating careful monitoring and chemical treatment. Based on recent advances in micro-nano bubble science have revealed that salinity plays a critical role in modulating gas–liquid interfacial behavior, Jia et al. [12] demonstrated that salt ions influence CO₂ dissolution, bubble stability, and interfacial tension through complex ion–interface interactions. These effects are particularly pronounced in high-salinity environments, where ionic strength alters the zeta potential and rheological properties of the gas–liquid interface.

The choice of gas and injection strategy plays a critical role in determining the success of MGI. CO₂ is widely favored due to its high solubility in oil, ability to swell the oil phase, and relatively low MMP requirements [13]. However, other gases such as methane, nitrogen, and hydrocarbon mixtures have also been explored. Masalmeh et al. [14] highlighted that CO₂ and rich gas injections were more effective than lean gas in heterogeneous carbonate reservoirs, particularly under high-salinity conditions. Al-Bayati et al. [15] confirmed that miscible CO₂ injection consistently outperformed immiscible scenarios in sandstone formations with varying salinity levels.

To optimize MGI performance, advanced modeling and predictive tools have been developed. Chen et al. [16] introduced a machine learning model that accurately predicts MMP by incorporating variables such as salinity, temperature, and oil composition. This enables more precise design and implementation of gas injection projects. Masalmeh et al. [14] and Ebadati et al. [17] emphasized the effectiveness of rich gas and WAG injection strategies in improving sweep efficiency and delaying gas breakthroughs in high-salinity carbonate and fractured reservoirs.

Recent advances in confined fluid phase behavior have highlighted the effect of different parameters on MGI. Chen et al. [18] developed a three-phase equilibrium algorithm that incorporates capillary pressure and critical point shifts, demonstrating that confinement suppresses phase regions and alters miscibility conditions in shale reservoirs. Similarly, Yang et al. [19] extended the Peng–Robinson equation of state to account for molecule-wall interactions and capillary effects, showing significant deviations in bubble-point pressures and phase envelopes for unconventional fluids. These studies provide a microscopic theoretical foundation for optimizing MGI strategies in high-salinity porous reservoirs.

MGI represents a viable and adaptable EOR strategy for high-salinity reservoirs. Its ability to overcome the limitations of water-based methods and adapt to complex reservoir conditions makes it a focal point of current research and field development. This paper aims to explore various MGI scenarios, evaluate their effectiveness in high-salinity reservoir located in Raleigh Field, and identify best practices for maximizing oil recovery while mitigating associated risks. The Raleigh Field, located in Smith County, Mississippi—approximately 50 miles southeast of Jackson—was discovered in January 1957 by the California Company and commenced production in March 1958. Oil is produced from a sandstone reservoir at an average depth of 12,600 feet. The field covers an area of approximately 40 acres, with an average porosity of 12% and permeability of 34 millidarcies.

At the onset of production, the reservoir exhibited an initial pressure of 5783 psig and a temperature of 258 °F, supported by a weak natural water drive. The initial water saturation was calculated at 42.3%, with a water density of 77.38 lb/ft³ and a salinity of 283,700 ppm. The reservoir’s saturation pressure was determined to be 3250 psi. Within just 1.5 years of production, reservoir pressure declined sharply from 5783 psig to below 3500 psig [20].

Conventional water flooding, typically employed for pressure maintenance, was deemed unsuitable for the Raleigh Field due to the high formation water salinity, which posed a significant risk of scaling. As a result, CO₂ injection emerged as a more promising alternative. This study presents a simulation-based evaluation of the CO₂-EOR method for the Raleigh Field, focusing on its potential to stabilize reservoir pressure and enhance oil recovery.

2. Methodology

This study presents a numerical investigation of eight gas injection scenarios for EOR using an Equation of State (EOS) model developed in CMG WinProp™. The simulations evaluate phase behavior under varying pressure conditions to assess the performance of different injection gases. The scenarios include the injection of methane (CH₄), nitrogen (N₂), liquefied petroleum gas (LPG), CO₂-propane mixtures, methane-propane blends, and separator-produced (enriched) gas, with carbon dioxide (CO₂) serving as the baseline for comparison.

The modeling approach enables a comprehensive analysis of phase behavior, miscibility, and swelling factors associated with each gas type when injected into the Raleigh reservoir. The methodology involves constructing a fluid model, determining saturation pressures, and performing miscibility and swelling factor calculations at multiple pressure levels (1000, 1500, 2000, and 2500 psi) to identify optimal injection conditions relative to CO₂.

CMG WinProp™ is employed to develop and calibrate the PVT model for the Raleigh field, using available experimental data such as Constant Composition Expansion, Differential Liberation (DL), and separator tests, along with the fluid composition and C₇⁺ properties (as detailed in

Table 1. Composition of reservoir fluid with C₇⁺ properties.

Component	Mole Composition%
CO2	1.19
N2	0.51
C1	45.21
C2	7.09
C3	4.61
iC4	1.69
nC4	2.81
iC5	1.55
nC5	2.01
C6	4.42
C7+	28.91
C7+ Properties
MW	190
SG	0.8142

). Following regression analysis and EOS model tuning, updated component properties are used to conduct two-phase flash calculations at reservoir pressures ranging from the saturation pressure of 3250 psia down to 1000 psi.

The eight gas injection scenarios evaluated in this study are:

• Pure Carbon Dioxide (CO₂);
• 50% CO₂/50% Propane (C₃H₈);
• Pure Methane (CH₄);
• Pure Nitrogen (N₂);
• Liquefied Petroleum Gas (LPG: 40% C₃H₈, 60% C₄H₁₀);
• Methane–Propane Blend (80/20);
• Methane–Propane Blend (60/40);
• Separator Produced Gas (Enriched Gas).

For each scenario, the updated EOS model in CMG WinProp™ is used to calculate the MMP and swelling factors at the specified pressures. A comparative analysis is then conducted to evaluate the effectiveness of each gas relative to CO₂, highlighting the potential of CO₂ injection for EOR in the Raleigh oil field.

3. Results and Discussion

The dataset utilized in this study comprises experimental results from Constant Composition Expansion tests (

Table 2. Constant composition Expansion values of reservoir fluid sample from Raleigh Oil Field.

Gauge Pressure (psi)	Relative Volume of Oil V/Vob at 258 °F	Viscosity of Oil at 258 °F (cp)
600	0.9387	0.119
5500	0.9471
5300		0.113
5000	0.9562
4590		0.107
4500	0.9666
4100		1.102
4000	0.9781
3800	0.9833
3720		0.099
3600	0.9988
3500	0.9918
3400	0.9948
3390		0.096
3300	0.9979
3235	1	0.093
3200	1.0047
3141	1.0128
3110		0.095
3094	1.0192
3039	10,273
2969	1.0387
2938

), Differential Liberation tests (

Table 3. Differential liberation data for reservoir fluid sample from Raleigh Oil Field.

Pressure (psig)	Gas Gravity	Oil Density (g/cc)	Daviation Factor (Z)
3236		0.5773
2938	0.870	0.5905	0.886
2607	0.846	0.6055	0.879
2301	0.833	0.6179	0.878
1903	0.830	0.6326	0.884
1505	0.835	0.6455	0.897
0	1.532	0.734

), and separator tests (

Table 4. Separator tests of reservoir fluid sample from Raleigh Oil Field.

Separator		GOR
Pressure (psig)	Temperature (°F)	Separator	Stock Tank	Stock-Tank Gravity (API at 60 °F)	Shrinkage Factor	Formation Factor	Gas Specific Gravity
0	75	1206	0	45.6	0.5456	1.833	0.942
50	74	1011	35	48.1	0.5872	1.703
100	75	950	68	48.5	0.5949	1.681
200	73	875	134	48.5	0.5974	1.674

). An EOS model was developed using CMG WinProp™, incorporating the aforementioned experimental data. The saturation pressure for the Raleigh field was determined to be 3250 psia. A regression process was conducted to align the EOS model with the experimental observations, following the methodology outlined in the CMG WinProp™ User’s Guide [21].

The regression parameters adjusted in this study include the critical pressure, critical temperature, acentric factor, and molecular weight of the C₇⁺ fraction. Additionally, the omega A and omega B parameters for methane, as well as volume shift coefficients for CO₂, N₂, and CH₄, were tuned. All available options for viscosity regression were also activated to enhance model accuracy. Figure 1, Figure 2, Figure 3 and Figure 4 illustrate a strong correlation between the predictions of the Peng–Robinson EOS and the experimental field data, demonstrating the robustness and reliability of the developed model.

Based on the successful match, the properties of the hydrocarbon components were updated using the optimized regression parameters. The two-phase envelope corresponding to the given fluid composition is shown in Figure 5. This phase diagram confirms that the fluid exhibits black oil behavior, with a critical point identified at 550 °F and 1800 psia.

As production increases in the Raleigh Oil Field, reservoir pressure is expected to decline. Therefore, understanding the variation in oil composition at pressures below the initial reservoir pressure of 5783 psig is essential. To evaluate this, two-phase flash calculations were performed in CMG WinProp™ at pressures of 2500, 2000, 1500, and 1000 psia. The results, summarized in

Table 5. Oil and gas composition at 2500, 2000, 1500, and 1000 psi.

	at 2500 psi		at 2000 psi		at 1500 psi		at 1000 psi
Component Percentage	Oil Comp.	Gas Comp.	Oil Comp.	Gas Comp.	Oil Comp.	Gas Comp.	Oil Comp.	Gas Comp.
CO2	1.14	1.375	1.0677	1.46	0.942	1.557	0.741	1.66
N2	0.365	1.065	0.277	1.024	0.196	0.975	0.121	0.918
C1	36.36	79.25	29.96	78.91	23.05	78.04	15.6	76.32
C2	6.81	8.139	6.45	8.5	5.82	8.97	4.74	9.56
C3	4.828	3.769	4.91	3.95	4.85	4.25	4.48	4.75
iC4	1.85	1.59775	1.956	1.12	2.03	1.18	2.02	1.34
nC4	3.125	1.06516	3.336	1.647	3.53	1.761	3.575	2.01
iC5	1.779	0.666	1.948	0.647	2.12	0.7	2.27	0.79
nC5	2.324	0.801	2.5577	0.779	2.81	0.83	3.03	0.93
C6	5.237	1.27179	5.86495	1.226	6.571	1.232	7.34	1.4
C7+	36.163	0.98839	41.67	0.69	48.09	0.492	56.07	0.37

, illustrate the behavior of oil and gas compositions under varying pressure conditions.

The analysis reveals that the concentration of heavier hydrocarbon components (e.g., C₆, C₇⁺) in the liquid phase increases as pressure decreases. Conversely, the gas phase remains dominated by lighter components (e.g., CO₂, N₂, CH₄), although their relative concentrations slightly decline with decreasing pressure.

The miscibility module in CMG WinProp™ was employed to calculate both First Contact Miscibility (FCM) and Multiple Contact Miscibility (FCM) for each gas injection scenario. A cell-to-cell simulation approach was used to model the miscibility behavior between the injected gas and reservoir oil. Additionally, the swelling pressure function in CMG WinProp™ was utilized to determine the swelling factor for each scenario. All simulations were conducted at a constant reservoir temperature of 258 °F.

A sensitivity analysis was performed based on the calculated contact miscibility and swelling factors. The swelling factor was computed using the following equation. The results, corresponding to various reservoir pressures, are summarized in

Table 6. FCM, MCM, and swelling factor calculation at reservoir pressure 2500 psi.

#	Injection Gas	Multi-Contact Miscibility Pressure	First Contact Miscibility Pressure	Swelling Factor
1	CO2	2750	2875	27.5%
2	(50% CO2, 50% C3H8)	2000	2500	31.8%
3	Methane, CH4	5125	6750	18.1%
4	Nitrogen, N2	5250	≥12,000	10.2%
5	LPG (40% C3H8, 60% C4H10)	2000	2375	39.1%
6	Methane-propane (80/20)	-	4750	21.9%
7	Methane-propane (60/40)	3000	3500	25.6%
8	Separator Gas	2875	3750	25.1%

The symbol ‘#’ denotes ‘number.

,

Table 7. FCM, MCM, and swelling factor calculation at reservoir pressure 2000 psi.

#	Injection Gas	Multi-Contact Miscibility Pressure	First Contact Miscibility Pressure	Swelling Factor
1	CO2	-	2750	24.9%
2	(50% CO2, 50% C3H8)	-	2000	28.8%
3	Methane, CH4	5375	6875	16.7%
4	Nitrogen, N2	5125	≥12,000	9.5%
5	LPG (40% C3H8, 60% C4H10)	-	2000	34.9%
6	Methane-propane (80/20)	-	4750	20%
7	Methane-propane (60/40)	3000	3375	22.3%
8	Separator Gas	2625	3500	23.5%

The symbol ‘#’ denotes ‘number.

,

Table 8. FCM, MCM, and swelling factor calculation at reservoir pressure 1500 psi.

#	Injection Gas	Multi-Contact Miscibility Pressure	First Contact Miscibility Pressure	Swelling Factor
1	CO2	-	2750	22.6%
2	(50% CO2, 50% C3H8)	-	2000	26.1%
3	Methane, CH4	5125	6875	15.4%
4	Nitrogen, N2	5875	Greater than 12,000	8.8%
5	LPG (40% C3H8, 60% C4H10)	-	2000	31.9%
6	Methane-propane (80/20)	-	4750	18.3%
7	Methane-propane (60/40)	3000	3375	21.2%
8	Separator Gas	2375	3250	21.9%

The symbol ‘#’ denotes ‘number.

and

Table 9. FCM, MCM, and swelling factor calculation at reservoir pressure 1000 psi.

#	Injection Gas	Multi-Contact Miscibility Pressure	First Contact Miscibility Pressure	Swelling Factor
1	CO2	2625	2750	20.3%
2	(50% CO2, 50% C3H8)	-	2000	23.4%
3	Methane, CH4	5500	7125	14%
4	Nitrogen, N2	8375	≥12,000	8.1%
5	LPG (40% C3H8, 60% C4H10)	-	2000	28.6%
6	Methane–propane (80/20)	-	4750	16.6%
7	Methane–propane (60/40)	3000	3375	19.1%
8	Separator Gas	2125	3000	20.3%

.

$$Swelling Factor={\displaystyle \frac{Volume of saturated gas injection at current temperature }{Volume of oil at current temperature }}$$

The results presented in Table 6, Table 7, Table 8 and Table 9 indicate that the calculated values for FCM and MCM remain unchanged at pressures of 2000 psi and below. A MMP of 2750 psi is required for both LPG injection and the 50% CO₂/50% C₃H₈ mixture. This value increases slightly when pure CO₂ is used, also yielding an MMP of 2750 psi. For separator gas injection, the MMP rises to 3000 psi—250 psi higher than that of pure CO₂. The highest MMP values are observed in scenarios involving CH₄ and N₂ injection. Methane–propane mixtures, regardless of their composition, also result in elevated MMP values.

The highest swelling factors are observed at a reservoir pressure of 2500 psi. Across all pressure conditions, the greatest swelling is achieved with LPG, the 50% CO₂/50% C₃H₈ mixture, pure CO₂, and separator (enriched) gas injection. In contrast, lower swelling factors are associated with methane–propane blends, CH₄, and N₂ injection, as illustrated in Figure 6 and Figure 7.

Based on the analysis of lower MMP values and higher swelling factors, the most technically favorable gas injection options for EOR in the Raleigh Oil Field are:

• LPG (40% C₃H₈, 60% C₄H₁₀);
• 50% CO₂/50% C₃H₈;
• Pure Carbon Dioxide (CO₂);
• Separator-Produced Gas (Enriched Gas).

While LPG offers excellent miscibility and swelling characteristics, its economic feasibility is limited due to its significantly higher market value compared to CO₂. According to industry data, the average cost of CO₂ for EOR operations—particularly when sourced from industrial emissions—ranges between USD 10 and USD 20 per metric ton [22]. In contrast, LPG, being a refined hydrocarbon product, can exceed USD $500 per metric ton depending on market conditions and regional availability [23]. This substantial cost differential, when scaled to the volumes required for field-scale injection, results in a markedly higher unit injection cost for LPG. Although a full net present value analysis was beyond the scope of this study, the relative cost disparity supports the conclusion that CO₂ and enriched separator gas are more economically viable options for the Raleigh Field, especially under current reservoir pressure and economic constraints.

As previously discussed, the Raleigh Oil Field is experiencing a rapid decline in reservoir pressure—from 5783 psig to below 3500 psig—along with reduced productivity. This pressure drop may lead to the cessation of oil production and result in high residual oil saturation. To mitigate this, it is critical to implement strategies that enhance oil recovery and maintain reservoir pressure.

Two potential approaches include water injection and immiscible gas injection. However, the high salinity of the formation water poses a significant risk of incompatibility with injected water, potentially leading to scaling and formation damage. Consequently, immiscible gas injection is recommended for this field.

The two most feasible gas injection options for EOR are CO₂ and enriched separator gas. Immiscible gas injection is planned to commence once the reservoir pressure reaches the bubble point of 3250 psi. At that stage, contact miscibility and swelling factor calculations will be repeated using CMG WinProp™ at 3250 psi. The updated results are presented in

Table 10. FCM, MCM, and swelling factor calculation at 3250 psi (bubble point pressure).

#	Injection Gas	Multi-Contact Miscibility Pressure	First-Contact Miscibility Pressure	Swelling Factor
1	CO2	2750	3250	31%
2	Separator Gas	3250	4125	28%

and Figure 8 and Figure 9.

Based on the analysis of Table 10 and Figure 8 and Figure 9, CO₂ injection is recommended over separator gas for EOR in the Raleigh Oil Field, for the following key reasons:

• Higher Swelling Factor: CO₂ demonstrates a greater swelling factor than separator gas, which enhances oil expansion and improves displacement efficiency, ultimately leading to increased oil recovery.
• Lower Miscibility Pressure Requirement: CO₂ achieves miscibility with reservoir oil at lower pressures compared to separator gas, making it more suitable under the current reservoir pressure conditions.
• Enhanced Performance with Pressure Decline: As reservoir pressure declines, the swelling factor associated with CO₂ injection increases, maintaining its effectiveness even at lower pressures.
• Environmental Advantages: Utilizing CO₂ captured from industrial emissions for EOR not only supports hydrocarbon recovery but also contributes to carbon sequestration efforts, reducing atmospheric CO₂ levels and mitigating climate change.

Given these technical and environmental benefits, CO₂ injection presents a more effective and sustainable EOR strategy for the Raleigh Oil Field.

4. Conclusions

In this study, the EOS model was developed using CMG WinProp™ to evaluate and identify the most suitable gas injection strategy for EOR in the Raleigh Oil Field, located in Smith County, Mississippi. The model was calibrated using experimental data and validated through regression analysis, with a particular focus on CO₂ injection performance. Based on simulation results and available production data, the following conclusions were drawn:

• Among the eight gas injection scenarios evaluated, CO₂ and enriched separator gas were identified as the most viable options for enhancing oil recovery.
• Under CO₂ injection conditions, the MMP was determined to be 3250 psi, with a corresponding swelling factor of approximately 31%.
• For enriched separator gas injection, the MMP was found to be 4125 psi, with a swelling factor of approximately 28%.
• Due to the rapid decline in reservoir pressure—from 5783 psig to below 3500 psig within 1.5 years of production—CO₂ injection is the most appropriate strategy for maintaining reservoir pressure and improving the recovery factor.
• It is recommended that CO₂-EOR operations commence once the reservoir pressure reaches the saturation pressure of 3250 psia.

This study highlights the technical and environmental advantages of CO₂ injection and provides a robust framework for implementing EOR strategies in pressure-depleted reservoirs.