Co-Optimization of Operational and Intelligent Completion Parameters of CO2 Water-Alternating-Gas Injection Processes in Carbonate Reservoirs

Abstract

Recently, maximum reservoir contacting (MRC) wells have attracted more and more attention and have been gradually applied to CO₂ WAG injections. During the use of MRC wells for CO₂ WAG injections, intelligent completions are commonly considered to control CO₂ breakthroughs. However, the design of the operational and intelligent completion parameters is a complicated process and there are no studies on the co-optimization of the operational and intelligent completion parameters for CO₂ WAG processes. This study outlines an approach to enhance the oil recovery from CO₂ WAG injection processes through the co-optimization of the operational and intelligent completion parameters of MRC wells in a carbonate reservoir. First, a simulation method is developed by using Petrel and Intersect. Then, a series of simulations are performed to prove the viability of intelligent completions and to investigate the effects of the timing and duration of the CO₂ WAG injection, as well as the type, number, and placement of intelligent completion devices on the performance of a CO₂ WAG injection by MRC wells. Finally, the imperialist competitive algorithm is used to co-optimize the operational and intelligent completion parameters for MRC wells. The results show that compared with the spiral inflow control device (SICD), autonomous inflow control device (AICD), labyrinth inflow control device (LICD), and annular interval control valve (AICV), the nozzle inflow control device (NICD) is the best type of intelligent completion device for MRC wells. There is an optimal installation timing, inflow area, and number of NICDs for a CO₂ WAG injection by MRC wells. The NICDs need to be placed based on the permeability distribution. The oil recovery for the optimal case with the NICDs reached 46.43%, which is an increase of 3.8% over that of the base case with a conventional completion. In addition, compared with the non-uniformity coefficient of the base case (11.7), the non-uniformity coefficient of the optimal case with the NICDs decreased to 4.21. This is the first time that the co-optimization of the operational and intelligent completion parameters of a CO₂ WAG injection has been reported, which adds more information about the practical applications of MRC wells in CO₂ WAG injections for enhancing oil recovery in carbonate reservoirs.

1. Introduction

CO₂ flooding, an efficient technology for enhancing oil recovery, is widely used in the U.S.A., Canada, and the Middle East [1,2], and can effectively enhance oil recovery and reduce green gas emissions [3]. However, due to the low density and viscosity of CO₂, CO₂ flooding has limitations from gravity overriding and viscous fingering [4,5,6]. Therefore, a CO₂ water-alternating-gas (CO₂ WAG) injection was proposed to overcome the limitations of CO₂ flooding [7,8]. Recently, maximum reservoir contacting (MRC) wells have attracted more and more attention and have been gradually applied to CO₂ WAG injections. Compared with a conventional horizontal well, an MRC well can maximize reservoir contact to obtain better well performance [9,10]. However, frictional pressure drops commonly result in an uneven inflow distribution along MRC wells’ wellbores, resulting in CO₂ and water breakthroughs. In addition, reservoir heterogeneity makes the uneven inflow distribution along MRC wells’ wellbores more serious. Intelligent completions can balance the inflow by adjusting the pressure profile along the wellbore, which mitigates the CO₂ and water breakthroughs and enhances oil recovery [11,12,13]. Therefore, intelligent completions need to be considered during the application of MRC wells to CO₂ WAG injections.

Some authors have reported that intelligent completions can effectively control CO₂ and water breakthroughs in CO₂ WAG processes. Specifically, Abdou et al. (2010) successfully designed and installed inflow control devices (ICDs) in a horizontal well in a Middle East oilfield and proved the advantages of the technology [14]. Ibrahim and Fuzi (2016) conducted a well completion design to improve the oil recovery of two mature oilfields by using interval control valves (ICVs) or ICDs [15]. Shahkarami et al. (2020) investigated the impact of the ICD design (ICD locations, ICD type, and number of ICDs) on well production [16]. Wu (2011) reported that the performance of horizontal wells exceeded expectations due to the installation of ICD completions [17]. Minulina et al. (2012) introduced the principle of three different types of ICD, and selected the appropriate ICD type, ICD location, and ICD size to delay gas and water production [18]. Lim (2017) evaluated the sensitivity of each ICD design method under reservoir uncertainty and heterogeneous conditions. The improvement of the existing intelligent completion device setting method was proposed to adapt to heterogeneous reservoirs [19]. Although the aforementioned results have verified that intelligent completion can be applied to horizontal wells to enhance oil recovery, its underlying mechanisms and potential to be applied to MRC wells are still not clear. In addition, the effects of intelligent completion parameters, such as the type, number, and placement of intelligent completion devices, on the performance of MRC wells needs further study.

In a CO₂ WAG injection, inappropriate operational conditions (injection rate, CO₂ slug size, water slug size, etc.) can result in low oil recovery [20,21,22]. The determination of the operational conditions by optimization algorithms is widely considered as an effective method to maximize oil recovery [23,24,25,26,27,28]. A few researchers have studied the optimization of CO₂ storage and oil recovery in CO₂ injection processes. Ampomah et al. (2017) used the polynomial response surface method to identify the optimal operational conditions for a CO₂ WAG injection to maximize oil recovery [24]. Sun et al. (2020) used a genetic algorithm to optimize a CO₂ WAG injection by maximizing the techno-economic objective functions [25]. Enab and Ertekin (2021) reported an optimization procedure for a CO₂ WAG injection into a low-permeability reservoir by using a novel artificial neural network-based toolbox [26]. Chen et al. (2010) developed an orthogonal array and then introduced the technology into a genetic algorithm to determine the optimal injection and production parameters for WAG production [27]. Mohaghe et al. (2018) used a genetic algorithm and particle swarm optimization to optimize CO₂ WAG operation parameters, such as the water injection rate, gas–water ratio, and production cycle. The oil recovery was increased by 13.8% [28].

In addition to optimizing the aforementioned operating parameters, designing intelligent completion parameters is also very important for CO₂ WAG injections with MRC wells. However, the uncertainty about reservoir characteristics and operational constraints adds complexity to the intelligent completion design and complicated optimization processes. There are no studies on the co-optimization of the operational and intelligent completion parameters and there is still much to study when CO₂ WAG injection processes involve MRC wells with intelligent completion.

This study outlines an approach to enhance the oil recovery of CO₂ WAG injection processes through the co-optimization of the operational and intelligent completion parameters of MRC wells in carbonate reservoirs. First, a simulation method is developed by using Petrel and Intersect. Then, a series of simulations are performed to prove the viability of intelligent completions and to investigate the effects of the timing and duration of the CO₂ WAG injection, as well as the type, number, and placement of intelligent completion devices on the performance of the CO₂ WAG injection by the MRC wells. Finally, the imperialist competitive algorithm is used to co-optimize the operational and intelligent completion parameters for the MRC wells to maximize oil recovery. This is the first time that the co-optimization of the operational and intelligent completion parameters of a CO₂ WAG injection has been reported, which adds more information about the practical applications of MRC wells in CO₂ WAG injections for enhancing oil recovery in carbonate reservoirs.

2. Assessment of Intelligent Completions for Enhanced Oil Recovery

2.1. Reservoir Model Description

The R oilfield is a giant carbonate reservoir in the Middle East, which belongs to a homogenous pore-type reservoir, and few natural fractures are observed. Zone B is the main pay zone, which is subdivided into eight main subzones: BI, BII, BIIIU, BIIIL, BIVU, BIVL, BV, and BVI, with a total thickness of about 150ft. The reservoir depth is 2700 m. The average porosity and permeability are 14–17% and 1 mD–3 mD, respectively. The reservoir pressure and temperature are 27.5 MPa and 129.4 °C, respectively. The oil viscosity has been 0.26 cP at reservoir conditions and the oil density has ranged from 570 kg/m³ to 690 kg/m³ at surface conditions since 2017. Zone B has been developed with water-alternating-miscible hydrocarbon gas injections by the line-drive horizontal wells. The average length and spacing of the horizontal wells are 1200 m and 750m, respectively. Up to 2020, the oil recovery was 26.35% with a water cut of 17.5% and a gas-to-oil ratio (GOR) of 1.37 MSCF/STB.

In this study, the numerical simulations were conducted by using Intersect. The three-dimensional geological model had dimensions of 3000 m × 2600 m × 12 m, discretized into 200 × 180 × 96 grid blocks, as depicted in Figure 1a. The oil was split into five pseudo-components (CO₂, C₁–N₂, C₂–H₂S, C₃–C₅, and C₆–C₁₀). The mole percentages for the five pseudo-components were 0.453, 42.039, 24.486, 24.701, and 8.321, respectively. A detailed pressure–volume–temperature model was built through a fluid module to match the fluid component properties, multiple-contact miscibility, differential liberation, constant volume depletion, and swelling tests. The basic parameters used in the simulation model are illustrated in Table 1, Table 2 and Table 3, which were updated from the history-matching process (Figure 1b).

The history-matching parameters included the water cut and cumulative oil production (COP). A total of 25 horizontal wells were used to implement the water-alternating-miscible hydrocarbon gas injection processes. To reduce CO₂ consumption, and improve the well performance and ultimate oil recovery, the 10 infill MRC wells and CO₂ WAG injection were planned and conducted instead of the initial water-alternating-hydrocarbon gas injection. Because the amount of CO₂ supplied by the R oilfield could meet the requirements for a CO₂ WAG injection in 2034, the timing of the CO₂ WAG injection was planned for 2034. A consistent set of constraints was utilized throughout the modeling process. The constraints used in the model were representative of the facility architecture, formation limitations, and the initial design flow rate to test the limits of the expected well performance, and are listed in Table 3. The simulation was implemented for 21 years.

2.2. Inflow Profile Analysis of MRC Wells

To select the MRC wells that require intelligent completions, a cross plot was made to analyze the relationships between the GOR and oil recovery (Figure 2a). According to the established criteria (GOR > 8, oil recovery < 0.25), two MRC wells (B1 and B2) were selected for the design process of the intelligent completions. As shown in Figure 2b, the selected MRC wells are mostly distributed in high positions of the reservoir.

To quantitatively describe the uniformity of CO₂ production in the horizontal section of the MRC wells, the non-uniformity coefficient αg was proposed and calculated as follows:

(1)${\alpha }_{\mathrm{g}}=\sqrt{\underset{i}{\Sigma }{(x_i-\stackrel{-}{x})}^2}/\stackrel{-}{x}$

where x_i is the value of the cumulative CO₂ production (CCP) in the i section of an MRC well, and $\stackrel{-}{x}$ is the average value of the CCP along an MRC well. The smaller the α_g is, the more uniform the CO₂ production along the MRC wells is.

Figure 3 shows the CCP profiles, COP profiles, permeability profiles, and calculated αg of the selected MRC wells and several MRC wells that were not included in the intelligent completion design process.

Figure 3 shows that the CCP profile of the two MRC wells is not uniform. The uneven inflow issues are caused by heterogeneous permeability. Taking the B1 well, for example, the high CO₂ inflow zones are located at 17,500 to 19,000 ft, which have high permeability. Therefore, the permeability along the wellbore has a significant impact on the uniform inflow. In addition, as shown in Figure 3, the α_g values of the two MRC wells (B1 and B2) are 11.70 and 12.07, respectively, which are higher than those of the other MRC wells that were not included in the intelligent completion design process. The results prove the necessity of selecting these two MRC wells to be installed with intelligent completion devices to delay CO₂ production and equalize the inflow from the reservoir to the wellbore.

2.3. Sensitivity Study

To guide the design and installation of the intelligent completion devices, a sensitivity study was conducted to investigate the effects of the operational parameters, such as the type, number, inflow area, and placement of the intelligent completion devices on the performance of the CO₂ WAG injection by the MRC wells. In this study, five types of intelligent completion devices (autonomous inflow control device (AICD), labyrinth inflow control device (LICD), spiral inflow control device (SICD), nozzle inflow control device (NICD), and annular interval control valve (AICV)) were considered. The operational parameters used in the simulation cases are shown in Table 4.

2.3.1. Effects of the Intelligent Completion Type

Figure 4 compares five cases with different intelligent completion devices (NICD, AICD, LICD, SICD, and AFCV) to a conventional completion case (Case 1). The different devices result in different results (the oil production rate, COP, water cut, and GOR). As shown in Figure 4a, the COP is higher in Cases 2–5 with intelligent completion devices than in Case 1, the conventional completion case. Specifically, the COP in Case 2 (NICD) is higher than that in Case 1, with a 31% increase in the COP at the end of the simulation. Case 2 produces more oil and less CO₂ (the GOR is 21% lower) and water than Case 1 (Figure 4b–d). In addition, Case 2 illustrates that the CO₂ inflow is reduced in the identified breakthrough zones (17,500~19,000 ft) and the inflow is more uniform along the wellbore (Figure 4e). The αg value of the B1 well is 8.26, which is lower than that of Case 1. These results indicate that the intelligent completion devices (NICD, LICD, SICD, and AICV) can improve the performance of MRC wells compared with the conventional completion.

Although the performance improves with the use of intelligent completion devices, the performance varies with the different intelligent completion types. A comparison of the results for Cases 1–6 clearly show that the NICD (Case 2) results in a more favorable performance compared with that of the other intelligent completion devices. Compared to the other cases, Case 2 has a higher COP (Figure 4a), a higher oil production rate (Figure 4b), a lower GOR (Figure 4c), a lower water cut (Figure 4d), and a lower αg value. These findings indicate that the NICD is the best type of intelligent completion device for the MRC wells in this study. It is noted that Case 6 has the lowest COP, GOR, and water cut, indicating that the AICD simultaneously limits the oil, water, and CO₂ production. Considering the better performance of the NICD, the NICD was selected for the follow-up study.

2.3.2. Effects of Installation Timing of NICDs

Determining the installation timing of the NICDs is an important consideration when designing a field treatment for a CO₂ WAG injection by MRC wells. To investigate the effects of the installment timing of the NICDs, Cases 2, 7, 8, and 9 were conducted under similar conditions, except for the different installment timing of the NICDs (three MSCF/STB, four MSCF/STB, six MSCF/STB, and nine MSCF/STB for Cases 2, 7, 8, and 9, respectively).

The effects of the installation timing of the NICDs on the performance of CO₂ WAG injection processes are shown in Figure 5.

As shown in Figure 5, the NICDs were installed at a GOR of 4MSCF/STB in Case 7, resulting in a higher COP, a higher oil production rate, a lower GOR, a lower water cut, a lower αg value, and a more even CCP profile. When the NICDs were installed too early (Case 2), the oil production was decreased due to the restriction of the NICDs. However, when the NICDs were installed too late (Cases 8 and 9), CO₂ channeling was formed, resulting in lower oil production. Therefore, there is an optimal installation timing of NICDs for CO₂ WAG injection processes by MRC wells.

2.3.3. Effects of the Number of NICDs

The determination of the optimal number of NICDs is of vital importance for ensuring the economic viability of intelligent completions in CO₂ WAG injection processes. As shown in Figure 6, the simulation conditions for Cases 7, 10, and 11 are the same, but the number of NICDs are different (6 NICDs for Case 7, 8 NICDs for Case 10, and 12 NICDs for Case 11).

Figure 7 shows the effects of the number of NICDs on the performance of the CO₂ WAG injection processes. As shown in Figure 7, compared with Case 7 (6 NICDs) and 11 (12 NICDs), Case 10 (8 NICDs) results in the highest COP and oil production rate (Figure 7a,b). The COP for the MRC wells completed with eight NICDs is approximately 2.5% more than the MRC wells completed with a conventional completion. In addition, the GOR, water cut, and α_g value for Case 10 are relatively low (Figure 7c–e). Therefore, there is an optimal number of NICDs for CO₂ WAG injection processes (the optimal number of NICDs is eight in this study). It is noted that Case 11 (12 NICDs) has the lowest GOR, water cut, and α_g value. This indicates that more NICDs more effectively delay CO₂ breakthroughs and lead to a more even CO₂ production profile. However, more NICDs also restrict oil production, resulting in a lower COP and oil production rate.

2.3.4. Effects of the Inflow Area of the NICDs

In this study, Cases 10 and 12–15 were conducted to investigate the effects of the inflow area of the NICDs on the CO₂ WAG performance. As shown in Figure 8, the operational parameters of Cases 10 and 12–15 are the same. However, the inflow area of the NICDs in Cases 10 and 12–15 are 0.000541, 0.000141, 0.000341, 0.000403, and 0.001141, respectively. An increase in the inflow area of the NICDs increases the inflow into the wellbore of the MRC wells.

Figure 8 illustrates the effects of the inflow area of the NICDs on the performance of the CO₂ WAG injection processes. As shown in Figure 8, Case 13 has the optimal inflow area of the NICDs for the MRC well because of the highest COP and oil production rate, in combination with a relatively low GOR, water cut, and αg value. Case 12 yields both the lowest GOR and water cut. However, the smallest inflow area of the NICDs also restricts the oil production, resulting in a lower COP and oil production rate compared with Case 13.

2.3.5. Effects of the Placement of the NICDs

The placement of the NICDs is based on the reservoir properties. The permeability is the main driver of the intelligent completion design. The placement of the NICDs in Case 13 and Case 16 is shown in Figure 9. As shown in Figure 9, the NICDs are uniformly distributed along the wellbore of the MRC well in Case 13, while the NICDs are placed in the areas with high permeability in Case 16.

Figure 10 shows the effects of the placement of the NICDs on the performance of the CO₂ WAG injection processes. As shown in Figure 10a–d, compared to Case 13, the COP and oil production rate are higher, and the GOR and water cut are lower for Case 16. The COP of Case 16 is 6.84 × 106 m³, which is 5.8% higher than that for Case 13.

In addition, the αg value is lower for Case 16 compared to Case 13 (Figure 10e). The results indicate that the placement of the NICDs along the wellbore of the MRC well based on the permeability distribution can delay CO2 breakthrough, resulting in a uniform CO2 production profile.

2.3.6. Effects of the Timing of the CO₂ WAG Injection

The timing of the CO₂ WAG injection is critical for the operational optimization of a CO₂ WAG injection by MRC wells. To investigate the effects of the timing of the CO₂ WAG injection, Cases 10, 17, and 18 were conducted under similar conditions, except for the different timing of the CO₂ WAG injection. The timing of the CO₂ WAG injection for Cases 10, 17, and 18 was 2034, 2039, and 2044, respectively.

The effects of the timing on the performance of the CO₂ WAG injection processes are shown in Figure 11.

As shown in Figure 11, compared with Cases 17 and 18, Case 10 (the CO₂ WAG injection implemented in 2034) results in a higher COP, a higher oil production rate, a lower GOR, a lower water cut, and a lower αg value. Therefore, a CO₂ WAG injection process is recommended to be implemented as soon as possible. The reason for this is that CO₂ has a lower miscible pressure with oil compared to hydrocarbon gas. The early implementation of a CO₂ WAG injection process could achieve early miscibility with oil and improve the performance.

2.3.7. Effects of the Duration of the CO₂ WAG Injection

The duration of the CO₂ WAG injection significantly affects the performance of the CO₂ WAG injection by MRC wells. To investigate the effects of the duration of the CO₂ WAG injection, Cases 10, 19, and 20 were conducted under similar conditions, except for a different duration of the CO₂ WAG injection. For Cases 10, 19, and 20, the CO₂ WAG injection was implemented in 2034, but the duration of the CO₂ WAG injection was 21 years, 16 years, and 11 years, respectively. The effects of the duration of the CO₂ WAG injection are shown in Figure 12.

A comparison of the results of Cases 10, 19, and 20 clearly show that Case 10 has the highest COP and oil production rate, as well as the lowest GOR, water cut, and αg value. Therefore, it is recommended that a CO₂ WAG injection has a long duration. This is attributed to the fact that increasing the duration of a CO₂ WAG injection can effectively enhance the displacement efficiency, reduce the oil viscosity, and maintain reservoir pressure.

3. Co-Optimization of Operational and Intelligent Completion Parameters

In this part of this study, an optimization algorithm called the “imperialist competitive algorithm” is used to find the optimal operational and intelligent completion parameters that maximize the oil recovery for the R reservoir [29,30,31,32,33,34]. Table 5 shows the lower and upper bounds for the operational and intelligent completion parameters. Figure 13 shows the optimization processes for the operational and intelligent completion parameters. As shown in Figure 13, the average and maximum oil recovery increase with increasing the number of iterations, and tend to the maximum values. In addition, the operational and intelligent completion parameters change within the lower and upper bounds to reach the maximum oil recovery during the optimization processes. Table 6 and Table 7 show the operational and intelligent completion parameters for the optimal case. The results prove that the optimization method can be applied to determine the optimal operational and intelligent completion parameters that maximize the oil recovery in the R reservoir.

Figure 14 shows a performance comparison between Case 1 and the optimal case, indicating that the optimal case has a higher oil production rate, a higher oil recovery, a lower GOR, a lower water cut, a lower αg value, and a lower residual oil saturation. For instance, the oil recovery for the optimal case with the NICDs reached 46.43%, which is an increase of 3.8% over that of the base case with a conventional completion. In the future, more research, including developing modified methods to enhance the search capability of the imperialist competitive algorithm, comparisons with other available optimization algorithms, and verifying the performance of the optimal case in the practical giant carbonate reservoir in the Middle East, will be required.

4. Conclusions

In this study, we developed a simulation method by using Petrel and Intersect and performed a series of simulations to prove the viability of intelligent completions and to investigate the effects of the intelligent completion parameters on the CO₂ WAG performance by MRC wells. We also used the imperialist competitive algorithm to co-optimize the operational and intelligent completion parameters for MRC wells. This is the first time that the co-optimization of the operational and intelligent completion parameters for a CO₂ WAG injection has been reported. The following conclusions have been drawn from this research:

The CO₂ WAG performance improved with the use of four types of intelligent completion devices (NICD, LICD, SICD, and AICV). Compared with the AICD, LICD, SICD, and AFCV, the NICD is the best type of intelligent completion device for the MRC wells under the simulation conditions in this study.

When the NICDs were installed too early, the oil production was decreased due to the restriction of the NICDs. However, when the NICDs were installed too late, CO₂ channeling was formed. Therefore, there is an optimal installation timing of NICDs for CO₂ WAG injection processes.

There is an optimal number of NICDs and inflow area for CO₂ WAG injection processes. More NICDs or small inflow areas can more effectively delay CO₂ breakthrough, but also restrict oil production. The NICDs for an MRC well need to be placed based on the permeability distribution.

The imperialist competitive algorithm can be used to determine the optimal operational and intelligent completion parameters for a CO₂ WAG injection by MRC wells, which maximizes the oil recovery in the R reservoir.

The optimal case had a higher oil production rate, a higher oil recovery, a lower GOR, a lower water cut, and a lower αg value. The oil recovery for the optimal case with the NICDs reached 46.43%, which was an increase of 3.8% over that of the base case with a conventional completion.

The ICA has some defects, such as poor global exploration ability and premature convergence. More research, including selecting more reasonable parameters for the ICA to increase its optimization accuracy, comparisons with other available optimization algorithms, and verifying the performance of the optimal case in the practical carbonate reservoir, will be required in the future.

Author Contributions

Conceptualization, X.D.; Methodology, X.D. and J.W.; Writing—original draft preparation, Y.Z. and J.W.; Visualization, J.W. and X.S.; Investigation, X.Z. and L.R.; Resources, X.Z. and L.R.; Writing—review and editing, L.R.; Supervision, X.S. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The original contributions presented in this study are included in the article; further inquiries can be directed to the corresponding authors.

Conflicts of Interest

Authors Xiangguo Zhao and Liangyu Rao were employed by the company of International Hong Kong Limited—Abu Dhabi, China National Petroleum Corporation. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. (a) Three-dimensional geological model and (b) history-matching results.

Figure 2. (a) The cross plot of the GOR and oil recovery and (b) the location of the selected MRC wells in the reservoir.

Figure 3. CCP profiles, COP profiles, permeability profiles, and calculated αg of the selected MRC wells and several other MRC wells.

Figure 4. Comparisons of performances of Cases 1–6 in terms of CO2 WAG injection processes: (a) COP, (b) oil production rate, (c) GOR, (d) water cut, and (e) CCP profiles, COP profiles, and αg values.

Figure 4. Comparisons of performances of Cases 1–6 in terms of CO2 WAG injection processes: (a) COP, (b) oil production rate, (c) GOR, (d) water cut, and (e) CCP profiles, COP profiles, and αg values.

Figure 5. Effects of installation timing of NICDs on CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 6. The placement of NICDs in Cases 7, 10, and 11.

Figure 7. Effects of the number of NICDs on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 8. Effects of the inflow area of the NICDs on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 8. Effects of the inflow area of the NICDs on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 9. Placement of NICDs in Case 13 and Case 16.

Figure 10. Effects of the placement of the NICDs on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 11. Effects of the timing of the CO2 WAG injection on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 11. Effects of the timing of the CO2 WAG injection on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 12. Effects of the duration of the CO2 WAG injection on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 12. Effects of the duration of the CO2 WAG injection on the CO2 WAG performance: (a) COP; (b) oil production rate; (c) GOR; (d) water cut; and (e) CCP profiles, COP profiles, and αg values.

Figure 13. Optimization processes of the operational and intelligent completion parameters: (a) the optimization process of the oil recovery, (b) the oil production rate of the B1 well, (c) the oil production rate of the B2 well, (d) the CO2 injection slug size of the C1 well, (e) the base strength of NICD 1 of the B2 well, and (f) the inflow area of NICD 1 of the B1 well.

Figure 13. Optimization processes of the operational and intelligent completion parameters: (a) the optimization process of the oil recovery, (b) the oil production rate of the B1 well, (c) the oil production rate of the B2 well, (d) the CO2 injection slug size of the C1 well, (e) the base strength of NICD 1 of the B2 well, and (f) the inflow area of NICD 1 of the B1 well.

Figure 14. Comparison of the performance of the optimal case and Case 1: (a) oil recovery; (b) oil production rate; (c) GOR; (d) water cut; (e) CCP profiles, COP profiles, and αg values; and (f) the residual oil saturations.

Figure 14. Comparison of the performance of the optimal case and Case 1: (a) oil recovery; (b) oil production rate; (c) GOR; (d) water cut; (e) CCP profiles, COP profiles, and αg values; and (f) the residual oil saturations.

Table 1

Interaction coefficients between each pseudo-component.

Components	CO2	C1 to N2	C2 to H2S	C3 to C5	C6 to C10
CO2	×	×	×	×	×
C1 to N2	0.11911	×	×	×	×
C2 to H2S	0.11999	0.00037	×	×	×
C3 to C5	0.12	0.00058	0.00003	×	×
C6 to C10	0.0796	0.00089	0.00044	0.00016	×

× means no data.

Table 2

Water and oil relative permeabilities.

Sw	Krw	Kro
0.06	0	0.8
0.1155	0.001718	0.64311
0.208	0.015615	0.42829
0.3005	0.046556	0.26616
0.393	0.096819	0.15
0.4855	0.16807	0.072875
0.578	0.26164	0.027481
0.6705	0.37866	0.006076
0.763	0.52013	0.000182
0.96	0.90679	0

Table 3

The operational parameters.

Parameters	Values
Bottom hole pressure of producers (psi)	2500
Oil production rate of horizontal producers (STB/d)	3000
Oil production rate of MRC wells (STB/d)	6000
Liquid production rate of total producers in the reservoir (STB/d)	4000
Water injection rate (STB/d)	3000
Gas injection rate (STB/d)	7000
Hydrocarbon gas slug size (d)	182
CO2 slug size (d)	182
Water slug size (d)	182
Timing of CO2 WAG injection (year)	2034

Table 4

The operational parameters used in the simulation cases.

Simulation Cases	Intelligent Completion Type	Installation Timing (MSCF/STB)	Numbers	Inflow Area (ft2)	Placement	Timing of CO2 WAG Injection (Year)	Duration of CO2 WAG Injection (Year)
1	/	/	/	/	/	2034	21
2	NICD	3	6	0.000541	Uniform distribution	2034	21
3	AICD	3	6	0.000541	Uniform distribution	2034	21
4	LICD	3	6	0.000541	Uniform distribution	2034	21
5	SICD	3	6	0.000541	Uniform distribution	2034	21
6	AICV	3	6	0.000541	Uniform distribution	2034	21
7	NICD	4	6	0.000541	Uniform distribution	2034	21
8	NICD	6	6	0.000541	Uniform distribution	2034	21
9	NICD	9	6	0.000541	Uniform distribution	2034	21
10	NICD	4	8	0.000541	Uniform distribution	2034	21
11	NICD	4	12	0.000541	Uniform distribution	2034	21
12	NICD	4	8	0.000141	Uniform distribution	2034	21
13	NICD	4	8	0.000341	Uniform distribution	2034	21
14	NICD	4	8	0.000403	Uniform distribution	2034	21
15	NICD	4	8	0.001141	Uniform distribution	2034	21
16	NICD	4	8	0.000341	Non-uniform distribution	2034	21
17	NICD	4	8	0.000541	Uniform distribution	2039	21
18	NICD	4	8	0.000541	Uniform distribution	2044	21
19	NICD	4	8	0.000541	Uniform distribution	2034	16
20	NICD	4	8	0.000541	Uniform distribution	2034	11

Table 5

The lower and upper bounds of the operational and intelligent completion parameters.

Parameters	Lower Bound	Upper Bound
Oil production rate of MRC wells (STB/d)	2000	6000
CO2 slug size (d)	60	300
Water slug size (d)	60	300
The base strength of the 8 NICDs	10−9	10−6
Inflow area of the 8 NICDs (ft2)	0.0001	0.001
Flow coefficient of the 8 NICDs	1	1.5

Table 6

The operational parameters of the optimal case.

Parameters	MRC wells		Injectors
	B1	B2	C1	C2	C3	C4	C5	C6
Oil production rate (STB/d)	2730	2080	-	-	-	-	-	-
CO2 slug size (d)	-	-	246	220	256	210	222	249
Water slug size (d)	-	-	119	145	109	155	143	116

Table 7

The intelligent completion parameters of the optimal case.

Parameters	NICD 1	NICD 2	NICD 3	NICD 4	NICD 5	NICD 6	NICD 7	NICD 8
Base strength of B1 well (10−9)	12.28	107.27	650.21	979.83	374.25	998.69	115.85	950.59
Base strength of B2 well (10−9)	521.34	1	1000	182.93	740.76	546.32	840.76	864.69
Inflow area of B1 well (10−5ft2)	39.667	125.383	48.568	43.713	30.727	32.666	54.562	46.883
Inflow area of B2 well (10−5ft2)	61.393	26.198	35.384	144.1	29.458	27.888	32.122	48.593
Flow coefficient of B1 well	1.0	1.2	1.2	1.1	1.0	1.3	1.2	1.2
Flow coefficient of B2 well	1.2	1.0	1.3	1.5	1.4	1.0	1.3	1.0

References

1. Han, L.; Gu, Y. Optimization of miscible CO₂ water-alternating-gas injection in the bakken formation. Energy Fuels; 2014; 28, pp. 6811-6819. [DOI: https://dx.doi.org/10.1021/ef501547x]

2. Alvarado, V.; Manrique, E. Enhanced oil recovery: An update review. Energies; 2010; 3, pp. 1529-1575. [DOI: https://dx.doi.org/10.3390/en3091529]

3. Abedini, A.; Torabi, F. On the CO₂ storage potential of cyclic CO₂ injection process for enhanced oil recovery. Fuel; 2014; 124, pp. 14-27. [DOI: https://dx.doi.org/10.1016/j.fuel.2014.01.084]

4. Dellinger, S.E.; Patton, J.T.; Holbrook, S.T. CO₂ mobility control. SPE J.; 1984; 24, pp. 191-196. [DOI: https://dx.doi.org/10.2118/9808-PA]

5. Gong, Y.; Gu, Y. Miscible CO₂ simultaneous water-and-gas (CO₂-SWAG) injection in the bakken formation. Energy Fuels; 2015; 29, pp. 5655-5665. [DOI: https://dx.doi.org/10.1021/acs.energyfuels.5b01182]

6. Gong, Y.; Gu, Y. Experimental study of water and CO₂ flooding in the tight main pay zone and vuggy residual oil zone of a carbonate reservoir. Energy Fuels; 2015; 29, pp. 6213-6223. [DOI: https://dx.doi.org/10.1021/acs.energyfuels.5b01185]

7. Holt, T.; Lindeberg, E.; Berg, D.W. EOR and CO₂ disposal-economic and capacity potential in the north sea. Energy Procedia; 2009; 1, pp. 4159-4166. [DOI: https://dx.doi.org/10.1016/j.egypro.2009.02.225]

8. Ahmadi, Y.; Eshraghi, S.E.; Bahrami, P.; Hasanbeygi, M.; Kazemzadeh, Y.; Vahedian, A. Comprehensive water-alternating-gas (WAG) injection study to evaluate the most effective method based on heavy oil recovery and asphaltene precipitation tests. J. Pet. Sci. Eng.; 2015; 133, pp. 123-129. [DOI: https://dx.doi.org/10.1016/j.petrol.2015.05.003]

9. Saleri, N.G.; Salamy, S.P.; Al-Otaibi, S.S. The expanding role of the drill bit in shaping the subsurface. JPT; 2003; 55, pp. 53-56. [DOI: https://dx.doi.org/10.2118/84923-JPT]

10. Salamy, S.P.; Al-Mubarak, H.K.; Al-Ghamdi, M.S.; Hembling, D. Maximum-reservoir-contact-wells performance update: Shaybah field, saudi arabia. SPE Prod. Oper.; 2008; 23, pp. 439-443. [DOI: https://dx.doi.org/10.2118/105141-PA]

11. Li, Z.; Fernandes, P.; Zhu, D. Understanding the roles of inflow-control devices in optimizing horizontal-well performance. SPE Drill. Complet.; 2011; 26, pp. 376-385. [DOI: https://dx.doi.org/10.2118/124677-PA]

12. Da Cruz Schaefer, B.; Sampaio, M.A. Efficient workflow for optimizing intelligent well completion using production parameters in real-time. Oil Gas Sci. Technol.; 2020; 75, 69. [DOI: https://dx.doi.org/10.2516/ogst/2020061]

13. Afuekwe, A.; Bello, K. Use of smart controls in intelligent well completion to optimize oil & gas recovery. J. Eng. Res. Rep.; 2019; 5, pp. 1-14.

14. Abdou, M.; Kshada, A.; Shafiq, M.; Ogunyemi, O.; Chong, T.S.; Hadjar, K.; Leung, E. Applied production completion using optimum number of inflow control devices. Proceedings of the Abu Dhabi International Petroleum Exhibition and Conference; Abu Dhabi, United Arab Emirates, 1–4 November 2010.

15. Ibrahim, A.; Fuzi, N.A.M. Completion design for enhanced oil recovery programs in brown fields. Proceedings of the International Petroleum Technology Conference; Bangkok, Thailand, 14–16 November 2016.

16. Shahkarami, A.; Friedrichs, M.; Iyer, N.; Izadi, G.; Klenner, R.; Meyer, E.; Murrell, G. Utilizing bayesian optimization and machine learning to find the best inflow control design for horizontal wells. Proceedings of the Offshore Technology Conference; Houston, TX, USA, 4–7 May 2020.

17. Wu, R.; Turpin, A.; MacDonald, D.; Kavanagh, D. A procedure for the configuration of an inflow control device completion using reservoir modelling and simulation in the north amethyst pool. Proceedings of the SPE Reservoir Characterisation and Simulation Conference and Exhibition; Abu Dhabi, United Arab Emirates, 9–11 October 2011.

18. Minulina, P.; Al-Sharif, S.; Zeito, G.; Bouchard, M. The design, implementation and use of inflow control devices for improving the production performance of horizontal wells. Proceedings of the SPE International Production and Operations Conference and Exhibition; Doha, Qatar, 14–16 May 2012.

19. Lim, M. ICDS for uncertainty and heterogeneity mitigation: Evaluation of best practice design strategies for inflow control devices. Proceedings of the SPE Asia Pacific Oil and Gas Conference and Exhibition; Jakarta, Indonesia, 17–19 October 2017.

20. Wang, L.; Tian, Y.; Yu, X.; Wang, C.; Yao, B.; Wang, S.; Winterfeld, P.H.; Wang, X.; Yang, Z.; Wang, Y. et al. Advances in improved/enhanced oil recovery technologies for tight and shale reservoirs. Fuel; 2017; 210, pp. 425-445. [DOI: https://dx.doi.org/10.1016/j.fuel.2017.08.095]

21. Koyanbayev, M.; Wang, L.; Wang, Y.; Hashmet, M.R.; Hazlett, R.D. Impact of gas composition and reservoir heterogeneity on miscible sour gas flooding—A simulation study. Fuel; 2023; 346, 128267. [DOI: https://dx.doi.org/10.1016/j.fuel.2023.128267]

22. Wang, R.; Wang, L.; Chen, W.; Shafiq, M.U.; Qiu, X.; Zou, J.; Wang, H. Surrogate-assisted evolutionary optimization of co₂-esgr and storage. Energy Fuels; 2023; 37, pp. 14800-14810. [DOI: https://dx.doi.org/10.1021/acs.energyfuels.3c01682]

23. Janiga, D.; Czarnota, R.; Stopa, J.; Wojnarowski, P.; Kosowski, P. Performance of nature inspired optimization algorithms for polymer enhanced oil recovery process. J. Pet. Sci. Eng.; 2017; 154, pp. 354-366. [DOI: https://dx.doi.org/10.1016/j.petrol.2017.04.010]

24. Ampomah, W.; Balch, R.S.; Grigg, R.B. Co-optimization of CO₂-EOR and storage processes in mature oil reservoirs. Greenh. Gases; 2017; 7, pp. 128-142. [DOI: https://dx.doi.org/10.1002/ghg.1618]

25. Sun, Q.; Ampomah, W.; Kutsienyo, E.J.; Appold, M.; Adu-Gyamfi, B.; Dai, Z.; Soltanian, M.R. Assessment of CO₂ trapping mechanisms in partially depleted oil-bearing sands. Fuel; 2020; 278, 118356. [DOI: https://dx.doi.org/10.1016/j.fuel.2020.118356]

26. Enab, K.; Ertekin, T. Screening and optimization of CO₂-WAG injection and fish-bone well structures in low permeability reservoirs using artificial neural network. J. Pet. Sci. Eng.; 2021; 200, 108268. [DOI: https://dx.doi.org/10.1016/j.petrol.2020.108268]

27. Chen, S.; Li, H.; Yang, D. Optimal parametric design for water-alternating-gas (WAG) process in a CO₂-miscible flooding reservoir. J. Can. Pet. Technol.; 2010; 49, pp. 75-82. [DOI: https://dx.doi.org/10.2118/141650-PA]

28. Mohagheghian, E.; James, L.A.; Haynes, R.D. Optimization of hydrocarbon water alternating gas in the Norne field: Application of evolutionary algorithms. Fuel; 2018; 223, pp. 86-98. [DOI: https://dx.doi.org/10.1016/j.fuel.2018.01.138]

29. Dossary, A.; Mohammad, A.; Nasrabadi, H. Well placement optimization using imperialist competitive algorithm. J. Pet. Sci. Eng.; 2016; 147, pp. 237-248. [DOI: https://dx.doi.org/10.1016/j.petrol.2016.06.017]

30. Zadeh, M.R.D.; Fathian, M.; Gholamian, M.R. A new method for clustering based on development of imperialist competitive algorithm. China Commun.; 2014; 11, pp. 54-61. [DOI: https://dx.doi.org/10.1109/CC.2014.7019840]

31. Talatahari, S.; Kaveh, A.; Sheikholeslami, R. Chaotic imperialist competitive algorithm for optimum design of truss structures. Struct. Multidiscip. Optim.; 2012; 46, pp. 355-367. [DOI: https://dx.doi.org/10.1007/s00158-011-0754-4]

32. Bagheri, A.; Razeghi, H.R.; Amiri, G.G. Detection and estimation of damage in structures using imperialist competitive algorithm. Shock. Vib.; 2021; 19, pp. 405-419. [DOI: https://dx.doi.org/10.1155/2012/154987]

33. Zhou, W.; Yan, J.; Li, Y.; Xia, C.; Zheng, J. Imperialist competitive algorithm for assembly sequence planning. Int. J. Adv. Manuf. Technol.; 2013; 67, pp. 2207-2216. [DOI: https://dx.doi.org/10.1007/s00170-012-4641-y]

34. Ahmadi, S.; Forouzideh, N.; Alizadeh, S.; Papageorgiou, E. Learning fuzzy cognitive maps using imperialist competitive algorithm. Neural Comput. Appl.; 2015; 26, pp. 1333-1354. [DOI: https://dx.doi.org/10.1007/s00521-014-1797-4]