Investigation of Low-Density CO2 Injection for Enhanced Oil Recovery

Article pubs.acs.org/IECR

Investigation of Low-Density CO2 Injection for Enhanced Oil Recovery Seyyed Mehdi Seyyedsar,* Seyed Amir Farzaneh, and Mehran Sohrabi Heriot-Watt University, Edinburgh EH14 4AS, U.K.

ABSTRACT: The density of the CO2-rich phase in a reservoir would play a crucial role in the performance of an enahcned oil recovery (EOR) scheme. Many oil reservoirs are located in deep formations; hence, they have high temperatures. Moreover, the pressure of reservoirs decreases because of natural depletion. Under the conditions of those reservoirs, CO2 would be a lowdensity gas. A series of coreflood experiments were performed to evaluate the potential of low-density CO2 EOR. The experiments are intermittent CO2 injection, continuous tertiary and secondary CO2 injection, and water alternating CO2 injection followed by the coinjection of a surfactant and CO2. The same oil and gas were mixed to prepare live oil for all the experiments. The initial rate of oil recovery during secondary waterflood was high, but the efficiency of the process decreased after the breakthrough. Three pore volumes (PVs) of secondary CO2 injection resulted in the recovery of around 50% of the initial oil in place, which was 27% higher than the oil recovered during 1 PV of water injection. It was also observed that CO2 injection can improve the recovery factor after waterflood. However, the performance of tertiary CO2 injection is reduced because of the presence of water in pore spaces, which likely makes the oil less accessible to CO2. Waterflood after a period of CO2 injection recovered 20% of initial oil in place mainly because of the dissolution of CO2 in the oil and the resultant oil viscosity reduction. The impact of the rate of CO2 injection on the efficiency of oil recovery was investigated, and it appears that the dissolution of CO2 in the oil is the main mechanism of enhanced recovery. The reduction of oil viscosity as a result of the dissolution of CO2 in the oil as well as the low density of CO2 improved the effect of gravity drainage on oil production. In addition, it was observed that the mechanism of solution gas drive plays an important role in the process of oil recovery. The analysis of the physical properties of the core effluent reveals that CO2 can also improve the quality of produced oil compared to that of the original oil in the rock. The results of this study provide experimental evidence of the potential of low-density CO2 EOR.

■

immiscible.12 The mechanisms by which immiscible CO2 injection could enhance oil recovery have been investigated for many years. However, the impact and significance of them have not been well-documented in the literature. For instance, the mechanism of extraction of hydrocarbons by CO2 is generally neglected for heavy oil reservoirs because it is

INTRODUCTION The feasibility of CO2 injection for either light or heavy oil recovery has been shown by various laboratory investigations.1−7 Several field trials have also shown a considerable increase in oil production after CO2 injection.8−11 In light oil systems, miscibility between CO2 and the oil is usually a target, and it is often achievable by increasing reservoir pressure or enriching the injection fluid.11 However, in the case of medium or heavy oil reservoirs, the decrease in the interfacial tension (IFT) is not as significant as that in the light oil systems, and the contact between the oil and CO 2 would remain © XXXX American Chemical Society

Received: January 21, 2017 Revised: March 31, 2017 Accepted: April 20, 2017

A

DOI: 10.1021/acs.iecr.7b00303 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research perceived as a slow process.7,13 These mechanisms include oil viscosity reduction,6,14 oil swelling,15 IFT reduction,12 extraction of hydrocarbons,16,17 and asphaltene precipitation.18 In many of the CO2 EOR studies or field trials in the literature, CO2 has been a relatively dense fluid, mainly a supercritical fluid and in a handful of cases a liquid fluid.9,11 Thus, significant volumes of CO2 need to be compressed and injected into the reservoirs when economically feasible.7,19 That, of course, is in addition to the problems stemming from the high mobility of CO2 in porous media.20 However, many oil reservoirs all around the world have high temperatures.21,22 Moreover, oil reservoirs usually experience a natural decline in their pressure by beginning production. Under the conditions of those reservoirs, CO2 would be a vapor gas with a low density.23 CO2 injection into those reservoirs could also be a promising process mainly for of two reasons: First, a lower CO2 density means that lower volumes of CO2 need to be compressed, and that can make the process feasible in locations that a large volume of CO2 is not available. Second, the costs of compression are reduced because the injectant density is lower. In our previous studies, it was observed that not only the rate of CO2 dissolution and diffusion in the oil is affected by CO2 density, it also determines the impact of mechanism of extraction on the process of oil recovery.7,13 Accordingly, it is crucial to investigate the performance of oil recovery by lowdensity CO2 injection. To accomplish this task, the same core and fluids (e.g., oil, brine, CO2) used in our previous experiments7,13 were used in the experiments reported in this work to be able to compare the behavior of oil recovery at different conditions. An important purpose of a coreflood experiment is to understand or predict the behavior of fluids and rocks during a variety of field processes such as production or injection schemes, of course, on a small scale. Although a test with a core cannot completely mimic the complexity of processes in an actual reservoir, the information obtained from that can be used to predict the behavior of a reservoir. Among those advantages, however, one critical disadvantage of a coreflood experiment is that the trace of reservoir time cannot be seen in a coreflood run; thus, in processes in which time-dependent phenomena take place (e.g., diffusion), that could lead to incorrect translation of the coreflood results. The practice of CO2 injection for enhancing oil recovery is one of the processes in which time plays a crucial role in the performance of oil recovery.13 In this study, the results of four coreflood experiments which were performed at 50° C and 600 psi will be presented and discussed. It should be mentioned that all the live fluids were also prepared at the temperature and pressure of the experiemnts. The first experiment is an investigation of intermittent CO2 injection. The injection strategy is particularly suitable for systems in which mass transfer between the displacing and displaced fluids plays a crucial role in the efficiency of a process. In the second experiment, the tertiary injection of CO2 was investigated. In many reservoirs around the world, water has been injected as the first intervention method for improving the recovery factor.24 Therefore, it is necessary to evaluate the efficiency of low-density CO2 injection in waterflooded porous media. The third experiment is an investigation of secondary continuous CO2 injection. In the fourth experiment, the performance of water and CO2 injection in an alternating manner was first investigated. Then, the coinjection of a surfactant solution and CO2 was performed

to further enhance oil recovery. In addition, the compositional analysis of the core effluent and the measurements of PVT properties of fluids were performed to identify the underlying mechanisms of the process of oil recovery.

■

EXPERIMENTAL SETUP AND PROCEDURES A temperature-controlled air oven was used to house all the components of the rig other than the pumps [e.g., core-holder, storage vessels, back-pressure regulator (BPR), pipelines, and connections] at a constant temperature. During the experiments, all the required valves could be controlled from outside of the oven. Accordingly, that has ensured that the experiments were performed under constant and well-controlled temperature. To minimize laboratory artifacts associated with using small core plugs, a one-piece large sandstone core with a length of 32.1 cm was used for the experiments. The dimension and properties of the core are given in Table 1. Table 1. Physical Properties of the Core property

core

diameter, cm length, cm pore volume (PV), cm3 porosity (Φ), % permeability to brine (K), D

5.13 32.10 151.5 22.84 2.73

The complete properties of the coreflood rig and the core have been described elsewhere.7,13 Fluids. The same live viscous oil was used to saturate the core in all the experiments reported here in order to highlight the potential of CO2 injection for improving oil recovery under the conditions of significant adverse viscosity ratios. The live oil samples were prepared by mixing dead crude oil and hydrocarbon gas in a heated rocking fluid recombination cell. The dead oil was a reservoir stock-tank crude oil sample which has a density of 0.9908 g/cm3 at 25° C and ambient pressure. The composition of the dead oil is reported in Table 2. The mixing of the oil and the gas was performed at a pressure and a temperature higher than the bubble point, in order to ensure a good mixing of them. Then the temperature of the cell was slowly decreased to the experiment temperature, but the pressure was still kept high to avoid degassing. Eventually, the pressure of the cell was slowly reduced to the experiment pressure. To measure the impact of CO2 dissolution on the oil viscosity, another live oil sample was prepared by mixing dead crude oil and CO2 under the conditions of the coreflood experiments. The oil viscosities were calculated by measuring the pressure drop through an in-line calibrated capillary tube. The ratio of dissolved gas in the oil (GOR) samples was also measured. Table 3 gives the measured physical properties of the oil samples. A synthetic brine solution with a total salinity of 10 000 ppm was used in this study. The brine was made of 8000 ppm sodium chloride (NaCl) and 2000 ppm calcium chloride (CaCl2), and it was degassed before saturating the core. Furthermore, live (methane-saturated) brine was prepared to inject into the core prior to oil injection and CO2 injection. Nonetheless, CO2-saturated brine was prepared at the conditions of the experiments for injection into the core after each period of CO2 injection. Using another calibrated capillary tube within the setup, the viscosities of the brine samples were measured too. The properties of the brine samples are given in B


Article

Industrial & Engineering Chemistry Research Table 2. Composition of Original Dead Crude Oil

Table 4. Physical Properties of Brines

component

Mw (g/mol)

Z (mole %)

C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25 C26+ total

86 96 107 121 134 147 161 175 190 206 222 237 251 263 275 291 300 312 324 337 839 542

0.09 0.09 0.46 0.52 0.94 1.48 2.30 3.13 3.24 3.85 4.06 3.53 3.90 3.68 3.64 3.35 3.02 3.11 2.86 2.62 50.12 100

temperature (°C)

saturation pressure (psi)

dissolved gas in brine

GWR (scm3/rcm3)

50 50

600 600

CH4 CO2

1.75 15.85

Table 5. Properties of CO2 under the Conditions of the Experiments in This Study25 density (g/cm3)

viscosity (mPa·s)

state

0.082

0.017

vapor

Figure 1. Oil recovery and DP across the core during secondary CO2 injection.

Table 4. Table 5 shows the physical properties of CO2 under the conditions of our experiments. Procedure. The clean core was saturated with dead brine, and the permeability of the core was calculated (2.73 Darcy) by measuring the differential pressure (DP) across the core at several brine injection rates. Then, the core was flooded with methane-saturated brine to avoid methane diffusion from live oil into the brine in the core. Subsequently, live oil was injected through the core, and irreducible water saturation (Swi) of 8% was achieved in all the experiments. After Swi was established, injection of fluids into the core was started. After each experiment, the core was cleaned by injection of several cycles of toluene and methanol in succession. It should be mentioned that all the fluids were injected into the core from the top to its bottom. Unless otherwise stated, all the fluids have been injected at 7 cm3/h, which equals the frontal velocity of 1 ft/ day.

continuous injection. By the start of CO2 injection, it was observed that the heavy and viscous oil in the core could not be produced as fast as the injection rate. However, the oil production rate increased gradually until the breakthrough. At 0.14 PV of injection, CO2 had reached the production end, which resulted in oil recovery of 6.5% of the original oil in place. In an immiscible displacement, because the mobility of gas is significantly higher than that of heavy oil in porous media, a dramatic drop in the oil rate would be expected to happen after the breakthrough of the gas.26,27 However, that was not the case in this experiment, and it was noted that the oil rate reduction was relatively slow. The main reason for this observation probably was the impact of the mechanism of solution gas drive as a result of CO2 dissolution in oil until the breakthrough. The speed of nucleation of the dissolved gas of heavy oil has an inverse relationship with the viscosity of the oil, and also the gas bubbles evolved throughout the oil phase are generally longlived in high-viscosity oils.28 After 0.3 PV of injection, oil was recovered at a relatively constant rate until the end of continuous injection. Figure 2 reveals the concentration of methane in the produced gas during various stages of the period of continuous CO2 injection. As can be seen, the concentration of methane in the produced gas was relatively high after the breakthrough because of relatively high oil rates. However, the fraction of methane reduced as the oil rate decreased. Further reduction of the fraction of methane is attributed to the presence (dissolution) of CO2 in the produced oil and perhaps the liberation of methane of oil in the core.

■

RESULTS AND DISCUSSION The main objective of the experiments reported here is to evaluate the feasibility of low-density CO2 injection for oil recovery. It will be discussed that whether the results and observations of core-scale experiments could be translated to actual reservoir conditions. Furthermore, the repeatability of the experiments reported in this study is shown. Experiment 1: Intermittent CO2 Injection. First Continuous (Secondary) CO2 Injection. When the initial water and oil distributions in the core were established, continuous CO2 injection was started. Figure 1 shows the recovery profile and DP across the core during the period of Table 3. Physical Properties of Oil Samples temperature (°C)

saturation pressure (psi)

associated gas of oil

GOR (scm3/rcm3)

viscosity (mPa·s)

oil swelling (rcm3/scm3)

28 50 50

1500 600 600

− CH4 CO2

− 10.60 30.63

277 000 3 530 733

− 1.035 1.040

C


Article

Industrial & Engineering Chemistry Research

Figure 3. Core pressure change during shut-in periods.

Figure 2. Compositional analysis of produced gas during the period of secondary CO2 injection.

The viscosity of oil recovered during the periods of CO2 injection, after complete separation of gas from oil, was measured at 50° C, and the results were compared with the viscosity of the original crude oil (Table 6). It is shown that oil Table 6. Viscosity of Oil Recovered during Different Periods of Experiment 1 sample dead crude oil 1#1CO2#1 mixture of (1#2CO2#1, 1#4CO2#1, 1#6CO2#1, 1#7CO2#1)

PV of Injection

viscosity ratio (%)

− 0.35−0.97 (1.3−1.6, 1.9−2.2, 2.5−3.1)

100 86 79

Figure 4. Incremental oil recovery during each period of intermittent CO2 injection.

happened at early times of injection periods. Although CO2 dissolution in oil was dominant during halt periods, the viscosity difference between CO2 and oil in contact would have been significant. The main mechanisms of oil recovery were oil viscosity reduction as a result of the dissolution of CO2 in the oil; gravity drainage; and, of course, viscous forces. Despite the relatively low amount of oil recovery in each cycle of intermittent injection, Figure 5 clearly shows that the

recovered after the breakthrough of CO2 had a notable lower viscosity than the original oil in the core. In this experiment, the density of CO2 was relatively low. However, it appears that the mechanism of extraction of hydrocarbons by CO2 is responsible for the alteration of the viscosity of oil recovered after the breakthrough of CO2. Another possible reason for the abovementioned observation could be the precipitation and deposition of asphaltene of the oil which have been reported in various CO2 EOR studies.18,29,30 However, several results presented in previous investigations using the same crude oil used in this study and under conditions similar to those of the experiments reported here have shown that the contact of CO2 and the oil did not result in the precipation or deposition of the asphaltene fraction of the oil in porous media.26 Therefore, it appears that the precipitation or deposition of asphaltene would have not been the case in our experiments. Intermittent CO2 Injection. After the period of continuous injection, the core was shut-in for a period of 24 h. Then, 0.3 PV of CO2 was injected into the core to improve recovery factor. This process of halt−injection was repeated for 7 cycles. The pressure of the core was monitored during the shut-in periods, and it was observed that the core pressure decreased during each of the shut-in periods (Figure 3). This observation is in contrast with the observations made during the halt periods of the experiments where CO2 was a more dense fluid.13 The reduction of pressure during the shut-in periods is an indication of dissolution of CO2 in the oil, and as the oil in contact with CO2 became relatively saturated with that, the rate of pressure reduction decreased notably in successive cycles. The profiles of incremental oil recovery of each slug of CO2 injection of intermittent injection are shown in Figure 4. Before the beginning of injection, the core pressure was increased up until the flowing pressure of the system. No oil production

Figure 5. Cumulative oil recovery during the periods of continuous and intermittent CO2 injection.

changing of injection scheme has improved oil recovery. This significant additional oil recovery was achieved mainly because of increasing the residence time of CO2 in a porous medium, which facilitated the dissolution of CO2 in oil. Tertiary Waterflood. After the seventh cycle of intermittent CO2 injection, CO2-saturated brine was injected through the core. Water advancement in the core resulted in the displacement and accumulation of oil ahead of the water front, and an oil bank was formed. The oil bank then reached the outlet of the core and led to a considerable amount of oil recovery (Figure 6). At a certain point, water reached the D


Article


Figure 6. Incremental oil recovery and DP across the core during tertiary waterflood.

Figure 8. Oil recovery and DP across the core during secondary waterflood.

production outlet, but the oil recovery continued at high water cuts after the breakthrough. Moreover, gas production decreased sharply when the oil bank reached the production outlet. Summary. Figure 7 illustrates the cumulative oil recovery during different stages of experiment 1. The CO2 dissolution in

Figure 9. Incremental oil recovery and DP across the core during tertiary CO2 injection.

the oil, either by direct contact or by diffusing from the water around oil, has resulted in oil viscosity reduction as well as the swelling of oil. Consequently, oil was recovered at a relatively constant rate until the end of the period of CO2 injection. Despite the production of oil at a constant rate, the DP is decreasing, which means that the viscosity of fluids contributing to flow (mainly oil) is also decreasing. Other mechanisms such as gravity drainage would have also contributed to oil recovery. The composition of oil recovered during different stages of experiment 2 was analyzed and is reported in Table 7. Additionally, the viscosity of oil recovered by tertiary CO2 injection was measured at the temperature of the experiment (Table 8). It is shown that the physical properties of oil have been altered by the injection of CO2 into the core, albeit to a small extent. These observations are mainly attributed to the mechanisms which were also active during the injection of dense CO2 for heavy oil recovery.7 The lower density of CO2 in the vapor state is believed to be the main reason for the small extent of the in situ changes of the properties of the oil. Another reason could be due to the impact of gravity drainage, which was more significant on oil recovery in this experiment. Second Waterflood. After the period of tertiary CO2 injection, CO2-saturated brine was injected through the core for a period of 1 PV of injection (Figure 10). The injected water followed the least resistance paths of gas in the core and then reached the production outlet. A small fraction of oil was recovered before the breakthrough; however, the oil rate increased sharply after the breakthrough, and oil production continued until the end of waterflood. The compositional analysis of the oil displaced by water confirms that the injected water contacted the oil which was in contact with CO2. Summary. Figure 11 shows the cumulative oil recovery during different periods of experiment 2. The performance of waterflood for oil recovery is a function of the viscosity ratio of

Figure 7. Cumulative oil recovery during different periods of experiment 1.

the oil and the resultant effects of that on the physical properties of oil enhanced the efficiency of oil recovery. The increasing of the residence time of CO2 in porous media facilitated the contact between CO2 and oil while the utilization of gas was significantly reduced. In addition, lower-viscosity oil could be recovered as a result of the injection of CO2. It was observed that improving the viscosity of displacing fluid led to the recovery of a significant fraction of oil that has been in contact with CO2. Experiment 2: Tertiary CO2 Injection. First (Secondary) Waterflood. Following the procedures to establish initial oil and water saturations, methane-saturated brine was injected through the core. Figure 8 shows the profile of oil recovery and DP within the core during the period of secondary waterflood. The recovery profile is similar to waterflooding of heavy oil systems. An early breakthrough and production of a considerable fraction of oil at high water cuts are the main characteristics of those systems. Hence, a significant volume of the core remained unswept, and continuing waterflood could not target the remaining oil. Tertiary CO2 Injection. Figure 9 shows the profiles of oil recovery and DP during the period of tertiary CO2 injection. A significant fraction of the water in the core was displaced during the period of CO2 injection. However, because of the dissolution of CO2 in oil and water, water could not be displaced as fast as the injection rate. The dissolution of CO2 in E


Article

Industrial & Engineering Chemistry Research Table 7. Composition of Oil Recovered during Different Periods of Experiment 2 sample

PV of injection

≤C12 (mole %)

>C12−≤C16 (mole %)

>C16−≤C20 (mole %)

>C20−≤C23 (mole %)

>C23−≤C29 (mole %)

>C29−≤C45 (mole %)

>C45−≤C100 (mole %)

2#1WF#1 2#1CO2#2 2#2WF#1

0.0−0.07 3.0−4.2 0.24−0.28

12.8 15.9 9.4

22.9 21.8 24.1

28.9 28.6 26.0

12.2 11.7 12.0

12.7 11.9 13.5

8.5 8.0 11.5

2.0 2.0 3.4

recovery. Here, the injection rate of CO2 was set at 1 cm3/h to evaluate the effects of time and rate of injection on the process of oil recovery. Figure 12 shows the profiles of oil recovery and

Table 8. Viscosity of Oil Recovered during Tertiary CO2 Injection sample

PV of injection

viscosity ratio (%)

dead crude oil 2#1CO2#1 2#1CO2#2

− 1.8−3.0 3.0−4.2

100 81 83

Figure 12. Oil recovery and DP across the core during secondary CO2 injection.

DP across the core during the period of secondary injection of CO2. It was observed that the oil in the core could not be produced as fast as the injection rate. The relatively low rate of injection of CO2 would have assisted the low oil rate before the breakthrough by reducing the impact of viscous forces. Nonetheless, for a short period of time before the breakthrough, oil was produced at high rates. The early breakthrough of CO2 demonstrates that the rate of injection does not have a significant impact on the sweep efficiency and that the instability in the flood front dominates the flow in the systems of adverse viscosity ratio. After the breakthrough, oil production continued at a relatively constant and low rate until around 3 PVs of injection. The main mechanism of oil recovery was the dissolution of CO2 in the oil and the resultant oil viscosity reduction. After 3 PVs of injection, the oil rate was further decreased, which was noticed by the rising GOR until the end of the injection of CO2 (Figure 13). The reduction of oil rate during late times of the run was perhaps due to the lower saturation of oil in the core as well as the reduction of the rate of diffusion of CO2 in the oil. It should be noted that in comparison with the previous experiments, the impact of gravity drainage on oil recovery

Figure 10. Incremental oil recovery and DP across the core during second Waterflood.


the oil and water in porous media. Therefore, the sweep efficiency of waterflood is generally poor for viscous oil recovery. The injection of low-density CO2 after waterflood can still enhance oil recovery mainly by mechanisms associated with the dissolution of CO2 in the oil. It was known that the dissolution and diffusion of CO2 is a slow process, particularly in low-pressure systems. Therefore, time is an important factor determining the efficiency of enhanced oil recovery by CO2 injection. The physical properties of oil can be altered by vapor CO2. This alteration is more dominant in viscosity of produced oil. The injection of a fluid with mobility lower than that of CO2 can improve the sweep efficiency significantly. This is further facilitated by the lower viscosity of the oil in porous media because of the dissolution of CO2 in the oil. Experiment 3: Secondary Continuous CO2 Injection. Continuous (Secondary) CO2 Injection. A secondary injection of CO2 followed by a period of waterflood was performed to evaluate the potential of low-density CO2 injection for oil

Figure 13. Produced GOR during secondary CO2 injection. F


Article

Industrial & Engineering Chemistry Research Table 9. Composition of Oil Recovered during Different Periods of Experiment 3 sample

PV of injection

≤C12 (mole %)

>C12−≤C16 (mole %)

>C16−≤C20 (mole %)

>C20−≤C23 (mole %)

>C23−≤C29 (mole %)

>C29−≤C45 (mole %)

>C45−≤C100 (mole %)

3#1CO2#1 3#1CO2#2 3#1CO2#5 3#1WF#1

0.15−0.34 1.1−1.3 2.9−3.3 0.26−0.37

13.3 13.1 13.9 12.0

24.2 24.5 24.1 23.2

29.1 29.2 29.5 28.9

11.8 11.6 11.9 12.7

12.1 12.0 12.1 12.8

7.9 7.9 7.5 8.4

1.6 1.7 1.0 2.1

would have been higher in this test mainly because of the significant time of the contact of oil and CO2 in the core. The compositional analysis of oil recovered during different stages of experiment 3 is reported in Table 9. After the breakthrough, the composition of oil recovered by CO2 injection was remained relatively unchanged at different steps of this run. This implies that the impact of CO2 dissolution and gravity drainage were probably significant during the period of CO2 injection. Moreover, the amount of free CO2 in pore spaces could be low because of the dissolution of CO2 in the oil. This perhaps has reduced the strength of the mechanism of extaction of hydrocarbons by CO2. However, this does not necessarily reflect that CO2 could not alter the physical properties of the initial oil in the core. The composition of oil recovered by waterflood changed slightly, reflecting that the remaining oil in the core was relatively heavier than the oil recovered during the previous period of CO2 injection. In addition, the viscosity of oil recovered during the period of CO2 injection was measured, and the results, given in Table 10,

did not have a notable impact on the amount of oil produced before the breakthrough. However, for a short period after the breakthrough, oil was recovered with a lower rate in the experiment in which the rate of injection was also lower. This observation highlights the impact of solution gas drive on oil recovery as a result of pressure rising before the breakthrough. The higher pressure of injection fluid until the breakthrough would have resulted in a higher amount of CO2 dissolution in the oil. The higher pressure gradient would also enhance the impact of viscous forces on the process of oil recovery. Tertiary Waterflood. CO2-saturated brine was injected through the core after the period of secondary CO2 injection. The injected water followed the least resistance paths of CO2 in the core, and it displaced gas toward the production outlet. A low fraction of the remaining oil in the core was recovered until the breakthrough (Figure 15). Oil production rate increased

Table 10. Viscosity of Oil Recovered during Secondary CO2 Injection sample

PV of injection

viscosity ratio (%)

dead crude oil 3#1CO2#3 3#1CO2#4

− 1.6−1.9 2.4−2.9

100 90 88

confirmed that higher-quality oil has been recovered by CO2 injection. All the viscosity measurements were performed at the temperature that the coreflood experiment was conducted. Effect of Rate of Injection (Time) on Performance of Secondary CO2 Injection. Figure 14 compares the performance of oil recovery by secondary CO2 injection in experiments 1 and 3. The rate of injection of CO2 was 7 times lower in experiment 3. The instability in the flood front, regardless of the rate of injection, dominated the flow until the breakthrough time in both experiments. Hence, the rate of injection of CO2

Figure 15. Incremental oil recovery and DP across the core during tertiary waterflood.

after the breakthrough because of the production of the oil bank ahead of the water front. Oil was produced continuously at high water cuts until the end of the waterflood. The relatively low amount of oil recovery is mainly attributed to the low saturation of the remaining oil in the core. The results of the coreflood experiments in this study show that the behavior of waterflood after an extended period of CO2 injection probably depends on the scheme of recovery prior to that waterflood. For example, a distinct difference is the production of a considerable fraction of oil before the breakthrough by waterflood after the period of intermittent injection while an insignificant amount of oil is recovered before the breakthrough during waterflood after continuous injection of CO2. The main reason for this observation would be related to the injection scenario of CO2 before the waterflood. CO2 during the shut-in periods could invade a higher area of the core; therefore, the potential pore volume for water to invade would have been higher after intermittent injection of CO2. Summary. Figure 16 illustrates the cumulative oil recovery during different periods of experiment 3. The breakthrough of CO2 was not affected by reducing the rate of injection of CO2. The main mechanism of oil recovery was the dissolution of

Figure 14. Comparison of profiles of oil recovery during secondary CO2 injection (experiment 1) and secondary low-rate CO2 injection (experiment 3). G


Article


Figure 18. Comparison of profiles of DP across the core during secondary waterflood (experiment 2) and secondary waterflood (experiment 4).


CO2 in the oil and the resultant oil viscosity reduction. An advantage of low-density CO2 injection compared to dense CO2 injection is that the mechanisms of extraction and liberation of methane of oil are less effective. The liberation of methane of oil by dense CO2 leads to an initial increase in the viscosity of the oil which later is compensated by the dissolution of CO2 in oil. However, in the case of low-density CO2 injection, the viscosity of the oil is being reduced from the beginning of the contact of CO2 and live oil. The performance of tertiary waterflood after an extended period of CO2 injection depends on the saturation of remaining oil in place as well as the scheme of injection of CO2. Experiment 4: Water Alternating CO2 Injection. First (Secondary) Waterflood. The first slug of water alternating CO2 injection was an injection of methane-saturated brine into the core saturated with live crude oil. Following the procedures to establish initial water and oil distributions in the core, the injection of water into the core was started. This experiment (experiment 4) was performed at the same conditions that the experiment of tertiary CO2 injection (experiment 2) had been performed. In both experiments, (methane-saturated) brine was the first injection fluid after establishing irreducible water saturation in the core. Figures 17 and 18 compare the profiles

experiments 2 and 4. The main difference in both experiments was the rate of CO2 injection. The injection of CO2 was continued for an almost equal period of time in both runs, as evident in Figure 20. It should be mentioned that the difference in the saturation of the remaining oil in the core was negligible before the beginning of CO2 injection.

Figure 17. Comparison of profiles of oil recovery during secondary waterflood (experiment 2) and secondary waterflood (experiment 4).

Figure 20. Comparison of performance of incremental oil recovery during tertiary CO2 injection (experiment 2) and first CO2 injection (experiment 4).

of oil recovery and DP across the core, respectively, during the period of first (secondary) waterflood in both experiments. It is ensured that the results and also the behavior of the fluids and the rock were repeatable in the coreflood experiments reported in this study. First (Tertiary) CO2 Injection. After the period of first (secondary) waterflood, 1 PV of CO2 was injected at 1 cm3/h through the core. Figure 19 compares the profiles of recovery of the remaining oil in the core by tertiary injection of CO2 in

The results of oil recovery reflect that higher amount of oil was recovered by a higher rate of injection of CO2 into the core. However, in the scale of the core used in this study, the volume of CO2 utilized in the run of high rate of injection is 6 times higher than the volume of CO2 injected at a lower rate. In an actual reservoir, the amount of time for CO2 to contact oil would be significantly higher; hence, a greater amount of oil can be recovered by increasing the rate of injection of CO2. It should be mentioned that none of the flow rates at which CO2

Figure 19. Comparison of profiles of incremental oil recovery during tertiary CO2 injection (experiment 2) and first CO2 injection (experiment 4).

H


Article

Industrial & Engineering Chemistry Research was injected into the core in these two experiments is considered as the optimum rate of injection. Many factors, such as availability of injection gas, injectivity of formation, reservoir heterogeneity, and capacity of production facilities, can affect the rate of injection of a gaseous fluid into an oil reservoir.11,20 Water Alternating CO2 Injection. After the first period of CO2 injection, two more cycles of water and CO2 injection were performed to improve the recovery factor (Figure 21). A

to drive oil toward the production outlet at this stage of the experiment suggests that improving sweep efficiency of CO2 injection by lowering its mobility in porous media may further enhance recovery factor from of the oil. Coinjection of Surfactant and CO2. After the last (third) slug of CO2 injection, the cumulative oil recovery of 57% of the initial oil in the core was achieved, meaning that the significant saturation of the remaining oil in place can still be a target of EOR techniques. Although CO2 injection could still recover the remaining oil in the core, the efficiency of this process was relatively low. A solution to reduce the mobility of CO2 in porous media is the combining of CO2 injection with a surfactant solution to form in situ foam. Table 11 provides the properties of the surfactant used in this study. Table 11. Physical Properties of Surfactant

Figure 21. Cumulative oil recovery during different periods of water alternating CO2 injection.

commercial name

type

active (wt %)

Petrostep C1

anionic

39

formula (C14) sodium alpha olefin sulfonate

Under the conditions of the coreflood experiments reported in this paper, Emadi26 performed two coreflood experiments using the same surfactant solution in this study. First, the clean core was saturated with the surfactant solution, and then CO2 and surfactant were simultaneously injected through the core (CO2 at 4.5 cm3/h and surfactant at 2.5 cm3/h). The behavior of DP was monitored and recorded during the experiment. After 2 total pore volumes (TPVs) of injection, it was observed that the DP was increased at a constant rate until 6 TPVs of injection and then remained relatively constant until the end of the run (8.3 TPVs). The results indicated that relatively strong and stable foam was formed during the late times of the period of coinjection. The apparent viscosity of the foam was also calculated by using the data of DP across the core. The core was then prepared for the second experiment by cleaning it with toluene and methanol injected in succession. In the second experiment, after a period of secondary waterflood, CO2 and the surfactant were injected simultaneously into the core. The waterflood after 1 PV of injection resulted in the recovery of 19% of the initial oil in the core. The 6 TPVs of the coinjection of CO2 and the surfactant (CO2 at 4.5 cm3/h and surfactant at 2.5 cm3/h) improved oil recovery by the production of 56% of the initial oil in place during this period. The DP across the core increased significantly before the breakthrough because of the formation of an oil bank. The production of the oil bank was continued until the breakthrough of CO2, which was accompanied by the reduction of the DP within the core. Later, the DP increased gradually almost until the late times of the period of coinjection. Accordingly, it was concluded that in situ formation of “CO2-foam” was the main reason for the rising DP at late times of the run. That is, the formation of foam in the core took place while a significantly high saturation of oil was still in the core. In this work, 0.1 PV of the surfactant solution was injected at 7 cm3/h into the core after the third slug of CO2 injection. This surfactant preflush was conducted to reduce the adsorption of the surfactant on the rock surface during the period of coinjection of surfactant and CO2. The injected surfactant displaced CO2 toward the production outlet, and a small fraction of oil (less than 1% of the remaining oil in place) was also produced during this period. Then, the coinjection of CO2

significant amount of the remaining oil in the core was recovered by (CO2-saturated) waterflood after the first period of CO2 injection. The behavior of this period of waterflood was notably similar to the performance of the period of second waterflood in experiment 2. The continuation of the experiment by CO2 injection after the second waterflood improved recovery factor as well. However, the impact was not as significant as it was during the period of first CO2 injection, perhaps because of the lower saturation of oil in the core and also a lower amount of oil being accessible to CO2 by watershielding. The performance of the third waterflood was significantly poorer; hence, the run was terminated after only 0.5 PV of injection. The reasons for poor recovery of waterflood could be due to the presences of established paths of water and CO2 within the core and the low saturation of oil in those paths of interest of water. The distributions of saturation of the fluids in the core at the end of each slug of injection are compared in Figure 22 in which the extent of reshuffling of water and CO2 is noted. It is significant that the oil was still recovered during the third period of CO2 injection. The dissolution of CO2 in oil causes oil swelling as well as oil viscosity reduction which would improve the mobility of oil in porous media. The ability of CO2

Figure 22. Distribution of fluid saturation at the end of each slug of water alternating CO2 injection. I


Article

Industrial & Engineering Chemistry Research and surfactant was started at a total rate of 7 cm3/h (CO2 at 4.5 cm3/h and surfactant at 2.5 cm3/h). The coinjection of CO2 and surfactant was continued at a constant total rate until around 5.8 TPVs of injection where the rates of injection (of both fluids) were doubled and continued until 7.4 TPVs of injection. Here, the injection was stopped and the core was shut to charge the storage vessels with the injection fluids. After 24 h, the coinjection of surfactant and CO2 into the core was continued but at a different total rate of injection of CO2 and surfactant: CO2 at 11.5 cm3/h and surfactant at 2.5 cm3/h. This run was continued at a constant rate until 2.7 TPVs of injection, where the rate of injection of CO2 was increased to 18.5 cm3/h. The results of the above-mentioned periods, the profile of incremental oil recovery and the behavior of DP across the core, are illustrated in Figures 23 and 24. Unlike the

dissolved gas in the oil could affect the phase behavior of the oil and the surfactant solution (i.e., optimal salinity).31,32 Therefore, it could be a reason for no propagation of foam in the coreflood experiment performed in our study. The presence of gas, in particular, CO2 in oil would also reduce the IFT between CO2 and the oil. Therefore, oil would tend to spread between CO2 and the aqueous solution in the pores, and this can delay the propagation speed of foam. The spreading behavior of oil can also cause fast destabilizing of foam in porous media.33,34 In spite of the coinjection of CO2 and surfactant into the core, the pressure gradient within the core was lower during this run compared to the third slug of CO2 injection. This observation may also show that the flow paths of the fluids were different in these two runs. In addition, the physical properties of the core effluent, such as its color, indicated that emulsification was significant between oil and the surfactant solution. The emulsion scavenged the surfactant; hence, a lower amount of surfactant was probably available for formation of foam in the core. The results of this study highlight the complexity of the processes in porous media. This becomes even more challenging in the processes in which mass transfer plays a key role in determining the performance. It is, therefore, important to fully understand these processes and carefully study their underlying mechanisms prior to making decisions on implementation of those processes in actual reservoirs. Summary. Figure 25 illustrates the cumulative oil recovery during different stages of experiment 4. A combination of CO2

Figure 23. Incremental oil recovery and DP across the core during the first period of coinjection of surfactant and CO2.

Figure 25. Cumulative oil recovery during different periods of experiment 4. Figure 24. Incremental oil recovery and DP across the core during the second period of coinjection of surfactant and CO2.

injection with a higher viscous fluid injection (e.g., water) further improves the sweep efficiency in porous media. It is believed that the remaining oil in the core after three cycles of water and CO2 injection was mainly trapped by capillary forces. Thus, it was aimed to reduce the saturation of oil in the core by coinjection of CO2 and a surfactant solution. The successful coreflood application of the coinjection of CO2 and the surfactant solution used in this experiment has been reported in other investigations. 26,27 However, no sign of strong propagation of foam was observed in this study, mainly because of the presence of dissolved gas (methane and CO2) in the remaining oil in the core.

observation made by Emadi,26 no clear sign of formation of strong foam in the core was noted during the periods of coinjection of CO2 and surfactant in the experiment reported here. The cores used in both studies were taken from the same block of rock, and their physical properties were very similar; however, the core had been exposed to surfactant in the study reported by Emadi26 before performing the coreflood experiment with oil. Thus, the significance of adsorption of the surfactant on the surface of rock would have been lower in that study. However, a significant volume of surfactant was utilized in our study, which probably eliminates the role of adsorption in opposing the formation of foam in the core. The core in our study was initially saturated with live (methane-saturated) crude oil instead of dead oil in the study reported by Emadi.26 Furthermore, because of the injection of several cycles of CO2 and CO2-saturated brine, the remaining oil in the core would have been relatively saturated with CO2. The presence of

■

CONCLUSIONS Four coreflood experiments were performed at various injection strategies to evaluate the performance of low-density (vapor) CO2 injection for oil recovery. The mechanisms involved in these experiments were investigated and discussed. It was shown that the coreflood experiments performed in this study J


Article


(7) Seyyedsar, S.; Farzaneh, S.; Sohrabi, M. Experimental Investigation of Tertiary CO2 Injection for Enhanced Heavy Oil Recovery. J. Nat. Gas Sci. Eng. 2016, 34 (1), 1205−1214. (8) Attanucci, V.; Asbsen, K.; Hejl, K.; Wright, C. WAG Process Optimization in the Rangely CO2 Miscible Flood. SPE Annual Technical Conference and Exhibition, Houston, TX, October 3−6, 1993. (9) Stephenson, D.; Graham, A.; Luhning, R. Mobility Control Experience in the Joffre Viking Miscible CO2 Flood. SPE Reservoir Eng. 1993, 8 (3), 183−188. (10) Robie, D., Jr.; Roedell, J.; Wackowski, R. Field Trial of Simultaneous Injection of CO2 and Water, Rangely Weber Sand Unit, Colorado. SPE Production Operations Symposium, Oklahoma City, OK, April 2−4, 1995. (11) Moffitt, P.; Pecore, D.; Trees, M.; Salts, G. East Vacuum Grayburg San Andres Unit, 30 Years of CO2 Flooding: Accomplishments, Challenges and Opportunities. SPE Annual Technical Conference and Exhibition, Houston, TX, September 28−30, 2015. (12) Yang, D.; Tontiwachwuthikul, P.; Gu, Y. Interfacial Tensions of the Crude Oil + Reservoir Brine + CO2 Systems at Pressures up to 31 MPa and Temperature of 27 and 58 C. J. Chem. Eng. Data 2005, 50 (4), 1242−1249. (13) Seyyedsar, S.; Farzaneh, S.; Sohrabi, M. Enhanced Heavy Oil Recovery by Intermittent CO2 Injection. SPE Annual Technical Conference and Exhibition, Houston, TX, September 28−30, 2015. (14) Klins, M. Carbon Dioxide Flooding: Basic Mechanisms and Project Design, 1st ed.; D. Reidel Publishing Company: Boston, 1984. (15) Li, H.; Zheng, S.; Yang, D. Enhanced Swelling Effect and Viscosity Reduction of Solvent(s)/CO2/Heavy-Oil Systems. SPE Journal 2013, 18 (4), 695−707. (16) Hwang, J.; Park, S.; Deo, M.; Hanson, F. Phase Behavior of CO2/Crude Oil Mixtures in Supercritical Fluid Extraction System: Experimental Data and Modeling. Ind. Eng. Chem. Res. 1995, 34 (4), 1280−1286. (17) Wang, S.; Chen, S.; Li, Z. Characterization of Produced and Residual Oils in the CO2 Flooding Process. Energy Fuels 2016, 30 (1), 54−62. (18) Vazquez, D.; Mansoori, G. Identification and Measurement of Petroleum Precipitates. J. Pet. Sci. Eng. 2000, 26 (1−4), 49−55. (19) Gray, L.; S.G, G. Overcoming the CO2 Supply Challenge for CO2 EOR. Abu Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, UAE, November 10−13, 2014. (20) Lake, L. W. Enhanced Oil Recovery, 1st ed.; Prentice Hall: Englewood Cliffs, 1989. (21) Elhag, A.; Chen, Y.; Chen, H.; Reddy, P.; Cui, L.; Worthen, A.; Ma, K.; Hirasaki, G. J.; Nguyen, Q. P.; Biswal, S. L.; Johnston, K. P. Switchable Amine Surfactants for Stable CO2/Brine Foams in High Temperature, High Salinity Reservoirs. SPE Improved Oil Recovery Symposium, Tulsa, OK, April 12−16, 2014. (22) Orphan, V.; Goffredi, S.; Delong, E.; Boles, J. Geochemical Influence on Diversity and Microbial Processes in High Temperature Oil Reservoirs. Geomicrobiol. J. 2003, 20 (4), 295−311. (23) Vermeulen, T. N. Knowledge Sharing Report - CO2 Liquid Logistics Concept (LLSC): Overall Supply Chain Optimization. Tebodin Netherlands B.V: The Hague, The Netherlands, 2011. (24) Willhite, G. P. Waterflooding; Society of Petroleum Engineers: Richardson, TX, 1986. (25) NIST. Thermophysical Properties of Fluid Systems. http:// webbook.nist.gov/chemistry/fluid/ (accessed February 5, 2014). (26) Emadi, A. Enhanced Heavy Oil Recovery by Water and Carbon Dioxide Flood. Ph.D. Thesis, Heriot-Watt University, Edinburgh, U.K., 2012. (27) Farzaneh, S. A. Investigation of Enhanced Heavy Oil Recovery by CO2 Flood Under Various Injection Strategies. Ph.D. Thesis, Heriot-Watt University, Edinburgh, U.K., 2015. (28) Lillico, D.; Babchin, A.; Jossy, W.; Sawatzky, R.; Yuan, J. Gas Bubble Nucleation Kinetics in a Live Heavy Oil. Colloids Surf., A 2001, 192 (1), 25−38.

were repeatable. Furthermore, the compositional analysis and the measurement of physical properties of the effluent (e.g., oil, gas) of the coreflood experiments were performed at several stages of the experiments. Based on the results of the experiments reported in this work, the following conclusions are drawn: • In an immiscible contact, the contact of low-density CO2 and oil leads to the dissolution of CO2 in the oil and hence the viscosity of the oil reduces. • Low-density CO2 is capable of in situ improvement of the physical properties (e.g., viscosity) of oil. • In a tertiary injection of low-density CO2, the amount of dissolution of CO2 in the water in porous media should be considered for determining the proper amount of CO2 for injection. • Although time is a dominant factor in determining the performance of oil recovery by CO2 injection, viscous forces also have crucial impact on the speed of oil production. Thus, increasing the rate of injection of CO2 would increase the rate of oil production. • The presence of dissolved gas in oil can affect the performance of coinjection of CO2 and surfactant. Therefore, it is essential to evaluate the phase behavior of surfactant and oil under reservoir conditions.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Seyyed Mehdi Seyyedsar: 0000-0002-4306-7374 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was carried out as a part of the Nonthermal Enhanced Heavy Oil Recovery joint industry project (JIP) in the Centre for Enhanced Oil Recovery and CO2 Solutions of Institute of Petroleum Engineering at Heriot-Watt University. The project was equally funded by Total E&P, ConocoPhillips, CONACyT-SENER-Hidrocarburos − Mexico, Pemex, Wintershall, and Eni, which is gratefully acknowledged.

■

REFERENCES

(1) Orr, F.; Yu, A.; Lien, C. Phase Behavior of CO2 and Crude Oil in Low-Temperature Reservoirs. SPEJ, Soc. Pet. Eng. J. 1981, 21 (4), 480−492. (2) Monger, T.; Coma, J. A Laboratory and Field Evaluation of the CO2 Huff ’n’ Puff Process for Light-Oil Recovery. SPE Reservoir Eng. 1988, 3 (4), 1168−1176. (3) Ghedan, S. Global Laboratory Experience of CO2-EOR Flooding. SPE/EAGE Reservoir Characterization and Simulation Conference, Abu Dhabi, UAE, October 19−21, 2009. (4) Abedini, A.; Torabi, F. Parametric Study of the Cyclic CO2 Injection Process in Light Oil Systems. Ind. Eng. Chem. Res. 2013, 52 (43), 15211−15223. (5) Lei, H.; Yang, S.; Zu, L.; Wang, Z.; Li, Y. Oil Recovery Performance and CO2 Storage Potential of CO2 Water-AlternatingGas Injection after Continuous CO2 Injection in a Multilayer Formation. Energy Fuels 2016, 30 (11), 8922−8931. (6) Miller, J.; Jones, R. A Laboratory Study to Determine Physical Characteristics of Heavy Oil after CO2 Saturation. SPE/DOE Enhanced Oil Recovery Symposium, Tulsa, OK, April 5−8, 1981. K


Article

Industrial & Engineering Chemistry Research (29) Zanganeh, P.; Ayatollahi, S.; Alamdari, A.; Zolghadr, A.; Dashti, H.; Kord, S. Asphaltene Deposition during CO2 Injection and Pressure Depletion: A Visual Study. Energy Fuels 2012, 26 (1), 1412−1419. (30) Mahdavi, E.; Zebarjad, F.; Ayatollahi, S.; Taghikhani, V. Experimental Investigation on the Effect of Asphaltene Types on the Interfacial Tension of CO2−Hydrocarbon Systems. Energy Fuels 2015, 29 (12), 7941−7947. (31) Southwick, J.; Svec, Y.; Chilek, G.; Shahin, G. Effect of Live Crude on Alkaline/Surfactant Polymer Formulations: Implications for Final Formulation Design. SPE Journal 2012, 17 (2), 352−361. (32) Roshanfekr, M.; Johns, R.; Pope, G.; Britton, L.; Linnemeyer, H.; Britton, C.; Vyssotski, A. Simulation of the Effect of Pressure and Solution Gas on Oil Recovery From Surfactant/Polymer Floods. SPE Journal 2012, 17 (3), 705−716. (33) Lau, H.; O’Brien, S. Effects of Spreading and Nonspreading Oils on Foam Propagation through Porous Media. SPE Reservoir Eng. 1988, 3 (3), 893−896. (34) Osei-Bonsu, K.; Shokri, N.; Grassia, P. Foam Stability in the Presence and Absence of Hydrocarbons: From Bubble- to Bulk-Scale. Colloids Surf., A 2015, 481, 514−526.

L


Investigation of Low-Density CO2 Injection for Enhanced Oil Recovery

Recommend Documents