CO2

The results of studies conducted to evaluate the development of the technology for utilizing entrainers for carbon dioxide (CO2) flooding are described. Experiments were conducted to determine the extent of the fluid-property enhancement (gas-phase density and viscosity) of CO2 + hydrocarbon and CO2 + crude oil systems in the presence of selected entrainers (cosolvents). The improvement in C02-rich phase hydrocarbon extraction capacity was also determined in the presence of added cosolvents. The hydrocarbon systems used in the study were n-hexadecane and a crude oil sample from Bloomington (TX) field with an API gravity of 22.7°. The results of the experiments showed that the addition of cosolvents contributed to a significant enhancement in extraction of high-molecular weight components of the oil into the C02-rich phase. The degree of enhancement in component extraction was also dependent on the system pressure and amount of cosolvent added. Preliminary entrainer screening studies were conducted utilizing gas and liquid chromatographic approaches to determine relative affinities of certain cosolvents with target high-molecular-weight hydrocarbon solutes. These solutes ranged in molecular weights from n-decane (n-C-io) up to n-tetracontane (n-C4o). These qualitative screening methods were useful in identifying potential cosolvents that may be used for future study. A method of utilizing entrainers to enhance the extraction of polymeric materials into the C02-rich phase was also investigated. This approach utilized a simple extraction apparatus to determine relative degrees of polymer extraction in CO2 as a function of pressure and the degree of enhancement of extraction in the presence of added cosolvent. The results indicate the potential of using cosolvents in enhancing the solubility of polymeric materials into the C02-rich phase. The solubility of a polymer sample was increased by as much as two times when a cosolvent was added to the system, compared to conditions when using CO2 alone. Additional work will be needed in order to properly select cosolvent-polymer combinations and to determine optimum conditions where enhancement in polymer solubility entrainers. The studies were performed by a method utilizing gas and liquid chromatography methods to determine the relative affinities of candidate entrainers with different high-molecular-weight hydrocarbons ranging from n-decane (n-C-io) up to n-tetracontane (n-C4o). An initial evaluation of the potential for the use of entrainers to improve the solubility of polymers in CO2 was also investigated. The use of polymers as direct thickeners for CO2 flooding have a significant potential in improving gas mobility. The primary difficulty in this mobility control approach is finding a suitable polymer that is soluble enough in CO2 to impart a significant enhancement in the gas phase viscosity and density. Extensive studies have been reported in the literature without any success in finding a commercially viable direct thickener for CO2 1 " 2 Our present study was conducted to evaluate and CO2 + isooctane formed a single phase fluid under the conditions tested. The results of the experiment showed that the presence of the entrainers (isooctane or 2-ethylhexanol) enhanced the viscosity of the C02-rich phase. The plots show the extent of the difference in C02-rich phase viscosity by comparing literature values for the viscosity 9 of pure CO2 to that of CO2 + n-hexadecane system with and without entrainers. systems in the presence of entrainers (cosolvents). The results showed that the presence of the entrainers (isooctane


INTRODUCTION
The successful application of gas flooding processes, such as CO2 flooding, for oil recovery depends on the ability of the injected gas to come in contact with the oil. Carbon dioxide flooding processes have a potentially high displacement efficiency. However oil recovery efficiency can be significantly hindered by the unfavorable mobility of the gas with respect to that of the oil. Under normal reservoir conditions, because the injected gas is considerably less viscous and has a lower density than 1 the target oil being displaced, which has a tendency to by-pass the oil which results in poor sweep efficiency and early gas breakthrough.
Carbon dioxide flooding has been used to recover both light and heavy oils. The CO2 gas flooding processes involve two variations: miscible and immiscible. Both processes suffer significantly from the unfavorable mobility ratio of the gas with respect to the oil. The mobility control problem is even compounded whenever a highly viscous (heavy) oil is the target for recovery. There is a greater need for mobility control when CO2 is used to recover the more viscous crude oils. Research efforts within the petroleum industry have been focused on finding ways to mitigate the problem of unfavorable gas mobility. Methods discussed in the literature include (1) utilizing the water-alternate-gas (WAG) process; (2) using foams; (3) viscosifying the C02-phase by means of adding polymers as direct thickeners, 1 " 2 or (4) in situ polymerization of monomers that may be soluble in supercritical carbon dioxide. 3 More recently, methods involving reverse micellar formations have also been investigated in supercritical fluid systems. 4 Our efforts to mitigate the problem of gas mobility control have been focused on the method of adding a small amount of a soluble component, an entrainer, to pure supercritical solvents such as CO2 in an effort to improve the solvent power of the gas. Preliminary results have shown the potential of utilizing entrainers in improving the recovery efficiency of CO2 flooding. 5 " 7 The results indicated that the addition of suitable entrainers preferentially increased the extraction of the heavier components of a synthetic oil as well as the density and viscosity of the C02-rich phase.
This status report summarizes the results of studies conducted to evaluate the development of the technology utilizing entrainers for CO2 flooding. Experiments were conducted to test the potential of enhancing the density and viscosity of CO2 by the addition of representative entrainers. The systems tested include a CO2 + hydrocarbon system and CO2 + crude oil system. The results of both studies indicate that the addition of entrainers significantly improved the extraction of the high-molecular-weight hydrocarbon components. Screening studies were conducted to evaluate additional cosolvents for their potential use as entrainers. The studies were performed by a method utilizing gas and liquid chromatography methods to determine the relative affinities of candidate entrainers with different highmolecular-weight hydrocarbons ranging from n-decane (n-C-io) up to n-tetracontane (n-C4o).
An initial evaluation of the potential for the use of entrainers to improve the solubility of polymers in CO2 was also investigated. The use of polymers as direct thickeners for CO2 flooding have a significant potential in improving gas mobility. The primary difficulty in this mobility control approach is finding a suitable polymer that is soluble enough in CO2 to impart a significant enhancement in the gas phase viscosity and density. Extensive studies have been reported in the literature without any success in finding a commercially viable direct thickener for CO2 1 " 2 Our present study was conducted to evaluate 2 the potential merit of utilizing entrainers to improve the solubility of polymeric materials in the C02-rich phase.

Experimental Apparatus and Procedure
An apparatus was constructed to measure the solubility of hydrocarbons and polymers in CO2 and the resulting increase in viscosity of the C02-rich phase after the addition of an entrainer (cosolvent) at high pressures and temperatures. The new equipment provided the capability to measure compositions of both gas and liquid phases. Measurements of the viscosities and densities of the two phases were also possible. This new design provided a significant improvement to the experimental set-up used for the FY88 work. 7 The new design allowed for both liquid and gas phase properties to be determined.
Provision for polymer solubility measurements have also been incorporated in the design.
The apparatus, as shown in figure 1, consisted of a high-pressure dual-drive pump, high-pressure floating piston vessels, high-pressure windowed cells, capillary tubings, density meter, magnetic circulating pump, high-pressure sampling valve assembly, pressure gauge, pressure transducers, and transmitters. Major components of the apparatus were housed in a constant-temperature oven.
Temperatures at various points within the oven were measured using copper/constantan thermocouples connected to a digital temperature monitor. A high-pressure windowed cell (100 cm 3  The experiments were conducted by injecting a measured amount of the target solute (hydrocarbon or crude oil) into the evacuated equilibrium cell. A known amount of the selected entrainer was then injected into the equilibrium cell. The system's pressure was adjusted by injecting compressed CO2 into the cell at the test pressure conditions. Upon reaching equilibrium, the gas phase viscosity and density were measured using the capillary viscometer and the remote density cell. Several measurements were conducted for each experiment to determine an average measurement for the gas phase properties.
The composition of the gas phase was determined using the in-line sampling valve connected directly to the GC apparatus. Different pressures were tested for each sample loading to determine the pressure dependence of the fluid properties and composition. Several concentrations of entrainers were also used to determine the dependence of the enhancement in solvent extraction power on entrainer concentration. The specifics of the procedure for determining viscosities and densities were discussed in a previous report. 7

Experimental Results and Discussion
The density and viscosity of the C02-rich phase for the system of CO2 + n-hexadecane was determined as a function of pressure at 60° C. The study was conducted to investigate the enhancement in bulk properties of the C02-rich phase with the addition of suitable entrainers. Experiments using n-hexadecane also served a two-fold purpose: (1) comparison with available literature data 8 on n-hexadecane and CO2 at 60° C; and (2) determination of multiple-phase properties at entrainer-  behavior relatively close to a transition from a two-phase system to a single-phase system. Single-phase data points are labeled accordingly. Table 1        with and without added entrainer. Figure 5 shows a plot of the C02-rich phase density as a function of system pressure. The presence of the cosolvent contributed to a slight increase in C02-rich phase density.
The composition of the C02-rich phase was also determined with the presence of entrainers. experiments. The extraction of the heavy molecular weight hydrocarbons was considerably improved with the addition of the entrainer. The amount of heavy hydrocarbon components extracted increased as the concentration of isooctane was increased. The results of the experiments showed that using CO2 alone, the highest hydrocarbon chain length detectable was the Cis group. The addition of isooctane at 2.5 and 8.0 mol % increased the highest detectable hydrocarbon chain length to C24 and C36, respectively.
There is a definite contribution from the addition of the entrainer to improving extraction of high-molecularweight hydrocarbon components of the crude oil.
A comparison of the effect of pressure and the presence of entrainer is shown in figures 7 and 8.
The cosolvent used in the experiments was isooctane. Figure 7 shows a plot of the GC detector signal level as a function of the retention time. The plot shows the hydrocarbon molecular weight distribution of the gas phase components. The plot shows that using CO2 alone did not significantly alter the molecularweight distribution of the components greater than n-octadecane, Cis-There was an indication of some difference in detectable signal levels for the higher-molecular-weight hydrocarbons from 1,798 to about 3,133 psi. Overall, despite of the difference in range of pressures, using CO2 alone did not markedly improve the extraction of the high-molecular-weight hydrocarbons. Figure 8 shows a plot of the GC detector response as a function of the retention time when 8.0 mol % cosolvent was added to the system of CO2 + crude oil. Compared to figure 7, the addition of the entrainer significantly improved the extraction of high-molecular-weight components. The addition of the entrainer enhanced the extraction capacity of the C02-rich phase to the extent that the highest extracted molecular weight hydrocarbon was shifted from about Cis to about C36 at a pressure of about 1,800 psi. The amount of heavy molecular weight components extracted was also significantly improved when the system pressure was increased from 1,778 to 2,850 psi, at the same entrainer concentration. The results indicate that increasing the pressure alone did not improve the level of heavy molecular weight component extraction when using only CO2. On the other hand, the addition of the entrainer to the system of CO2 + crude oil significantly increased the extraction of the high-molecular-weight components as a function of system pressure and entrainer concentration.

SCREENING METHOD FOR ENTRAINER EVALUATION
The selection of entrainers strongly depends on the reservoir operating conditions to be encountered. Cosolvents that have appreciable solubility in the gas phase and the ability to enhance the solubility of crude oil heavy components in the C0 2 -rich phase would be positive factors in the selection.
Our previous studies have concentrated on a limited number of cosolvents that have been found to be suitable for entrainer application. These cosolvents include hydrocarbons (i.e. isooctane) and alcohols (i.e. 2-ethylhexanol). In order to screen additional potential candidates as entrainers, quick screening methods were designed to provide a tool to select cosolvents that may be used in future experimental evaluations. The screening experiments were conducted investigate methods to determine cosolvent- contribute to some problem or error in determining or ranking the affinity of these cosolvents for the hydrocarbon packed columns tested, such that these cosolvents will not be used in the cosolvent ranking process. Further tests will be needed to determine the factors that may have contributed to this difference in retention time behavior. from the column packed with 100 wt % UCW-982. The results at first seems to indicate that, in general, most of the cosolvents tested followed the behavior of decreasing affinity in this order: n-C3o > n-C4o ^ n_ C25. This discrepancy in the affinity ranking may be attributed to the way the hydrocarbon packed columns may have been prepared. A difference in the manner of packing these columns, whether tightly or loosely packed, can contribute to a significant difference in retention times for each column. This difference will invalidate any possible ranking in hydrocarbon affinity order. Comparison of the cosolvents' affinity should be only be made with each individual high-molecular-weight hydrocarbon filled column. Cirj"Ci6 m i xture used was an equal volume mixture of C10, C12. C14. and Ci6-In all experiments, only one chromatogram peak was detected for this C10-C16 mixture. The results presented in figure 12 showed no significant difference in retention time when using the solvent mixtures tested on the Cs based column. The affinity of the injected hydrocarbon samples for the hydrocarbon based column and the solvent pairs were very similar, so that no difference in hydrocarbon retention times was observed.
The affinity for the stationary phase material and the mobile phase material was too close to result in a difference in hydrocarbon sample retention time. The results showed that Cs based column was not suitable for the experiments conducted. Figure 13 shows that by the use of a Cis based column, significantly higher retention times can be observed in comparison to that of a Cs based column. Figure   14 shows a comparison of the different combination of solvent mixtures tested in the Cis based column.
These experiments were conducted to qualitatively determine the effect of solvent pair concentrations on the hydrocarbon sample retention times. The results indicated that a slight difference in retention times was achieved using a solvent mixture of 100% ethanol. On the other hand, no difference was observed when using 100% n-hexane only. When using a mixture of 70% ethanol and 30% n-hexane as solvent, a significant difference in retention times was observed. The retention times using the solvent pairs (70% ethanol and 30% n-hexane) were significantly higher than when using any of the solvents individually. On the other hand, the combination of 25% ethanol and 75% n-hexane as solvent did not show any significant retention time difference. The retention times were only slightly longer than when using 100% n-hexane as solvent. The system almost behaved similar to 100% n-hexane. The results for this solvent pair combination (25% ethanol and 75% n-hexane) were not presented in figure 14.
Within the limits of the solvent pair combinations tested, the results indicated that the combination

ENTRAINER-ENHANCED POLYMER SOLUBILITY IN C0 2
Extensive studies have been reported in the literature without any success in finding a commercially applicable polymer as direct thickener for CO2 application. Our study provides an initial evaluation of the potential for the use of entrainers to improve the solubility of polymers in the gas phase. The objective would be to improve the solubility of the polymer in the C02-rich phase to an extent that a substantial increase in bulk C02-rich fluid properties can be achieved. This scoping study involved the determination of the extent of polymer solubility in the C02-rich phase with the addition of cosolvents. Figure 15 shows a schematic diagram of the experimental apparatus that was used to measure the amount of polymeric materials extracted under different conditions. High-pressure tubes packed with finely ground activated charcoal were used as the adsorption media for the extracted polymer materials. The experiments were conducted by injecting a measured amount of polymer into an evacuated high pressure floating piston vessel. A known amount of selected entrainer was then injected into the floating piston vessel when required. The pressure inside the vessel was adjusted by injecting compressed CO2 into the vessel at the test pressure conditions. After sufficient time had been allowed for the system to reach equilibrium, the C02-rich gas phase was bled-off under constant pressure through the high-pressure tubes packed with activated charcoal. A constant pressure C02-rich gas phase bleed was maintained throughout the sampling procedure by means of a high pressure direct drive pump injecting pressure fluid into the piston side of the floating piston vessel. The volume of CO2 gas bled during the experiment was measured using a wet test meter. The gain in weight of the charcoal column accounts for the amount of polymeric material extracted into C02-rich phase. In experiments where cosolvents were added, the filtering media was purged with nitrogen under the same experimental temperature of 50° C in order to gradually remove the more volatile cosolvent that was trapped in the filtering media during the extraction experiment.  The results indicate that the addition of the entrainers did show a slight increase in the amount of material extracted. The extent of enhancement though may not be sufficient to significantly increase the viscosity and density of the gas phase. The results from the scoping study are preliminary, but they do indicate the potential of the use of entrainers in enhancing polymer solubility in the C02-rich phase. In order to determine optimum conditions and potential polymer-cosolvent combinations to make this method applicable, more thorough experimental studies will be necessary. These studies should involve (1) screening cosolvents for polymer solubility and compatibility; (2)

CONCLUSIONS
The objective of the study was to evaluate potential areas for the development of the technology for utilizing entrainers for carbon dioxide (CO2) flooding. Experiments were conducted to determine the enhancement of the fluid-phase properties of CO2 + hydrocarbon and CO2 + crude oil systems in the presence of entrainers (cosolvents). The results showed that the presence of the entrainers (isooctane or 2-ethylhexanol) can significantly enhance the viscosity and density of the C02-rich phase of the CO2 + n-hexadecane and CO2 + crude oil systems. The presence of the cosolvent contributed to an enhancement effect on the solubility of n-hexadecane and extraction of the high-molecular-weight components of the oil into the C02-rich phase. The enhancement in solubility was more pronounced in the case when using crude oil samples than when using single component high molecular weight hydrocarbons alone. The amount or concentration level of added entrainer also contributed to a significant difference in hydrocarbon extraction. The results showed that the addition of the entrainer at different concentrations increased the highest extractable hydrocarbon chain length from the crude oil.
The system pressure did not significantly affect the extraction potential of the CO2 phase without the addition of entrainers. With the addition of entrainers, the amount of heavy molecular weight components extracted was significantly increased when the system pressure was increased, even at the same entrainer concentration level. The results clearly showed that the addition of the entrainer to the system of CO2 + crude oil significantly improved the extraction of the high-molecular-weight components as a function of system pressure and entrainer concentration.
The GC and LC screening methods have been useful in possibly qualitatively identifying other potential cosolvents that have particular affinity for a certain range of high molecular weight hydrocarbons.
These initial experiments represent helpful steps in the effort to develop better and quicker means of evaluating potential entrainers. The results of the present study indicate that cosolvents such as petroleum ether and chloroform may have potential application as entrainers. The more polar cosolvents such as ammonia, methanol, ethanol, and acetone need to further studied. The difference in behavior for these cosolvents needs to be further investigated and additional experiments will be needed for that purpose. One limitation though of these initial screening methods was that they were designed primarily to evaluate only the affinity of a cosolvent to a particular solute. The method does not take into account the effect of the presence of the CO2 in the cosolvent-solute system. Additional experiments have to be conducted to determine the solubility of the cosolvents in CO2 and the degree of enhancement in gas phase density and viscosity that the cosolvent can impart to the C02-rich phase. Other screening methods have to be evaluated in order to develop better and faster means of testing other chemicals of cosolvent application potential.
Experiments were also conducted to determine the possibility of using entrainers to enhance the extraction of polymeric materials into the C02-rich phase. The results indicated the potential of using cosolvents in polymer solubility enhancement. The solubility of a polymer sample was increased by as much as two times when a cosolvent was added to the system, compared to conditions when using CO2 alone. In these experiments, the gas phase viscosity was not experimentally measured such that the amount of enhancement in polymer solubility may or may not be sufficient to increase the viscosity and 22 density of the C02-rich phase significantly. Additional experimental work will be needed in order to properly select cosolvent-polymer combinations and to determine optimum conditions where the enhancement in polymer solubility can impart a substantial improvement in the C02-rich phase bulk fluid properties.