Effect of Oil Composition on Minimum Miscibility Pressure-Part 1: Solubility of Hydrocarbons in Dense CO2

Abstract

Summary. This paper examines the effect of oil composition on the phase behavior of CO2/hydrocarbon mixtures and, hence, on the development of miscibility in a CO2 flood. Results of component-partitioning measurements are reported for mixtures of CO2 with five synthetic oil systems: normal alkanes, branched alkanes, naphthenes, aromatics, and a mixture of all four molecular types. The results of the experiments indicate that unsubstituted ring structures are less soluble in dense CO2 than branched or normal alkanes with the same number of carbon atoms, but that the addition of alkyl side chains to ring structures improves their solubility. Also reported are component-partitioning measurements for mixtures of CO2 with three crude oils: Rock Creek (paraffinic), Maljamar (more aromatic), and Rock Creek plus 15 wt% of a mixture of aromatic components. The experimental results suggest that of the many factors that influence extraction of hydrocarbons by dense CO2, the distribution of molecular weights present in the oil is the most important. Introduction The relationship between phase behavior and the development of miscibility in CO2/crude-oil displacements is by now well established. Hutchinson and Braun and Rathmell et al. argued and Helfferich proved that preferential extraction of hydrocarbons present in oil can generate composition paths that avoid two-phase present in oil can generate composition paths that avoid two-phase regions in displacements with ternary systems. Gardner et al., Orr et al., and Sigmund et al. showed that such concepts can be applied to produce quantitative predictions of oil recovery for CO2/crude-oil displacements in slim tubes. Thus, there is considerable experimental evidence that a CO2 displacement is efficient when some of the components present in the crude oil partition relatively efficiently into the displacing CO2-rich partition relatively efficiently into the displacing CO2-rich phase. Holm and Josendal argued that the pressure required to phase. Holm and Josendal argued that the pressure required to produce efficient extraction of hydrocarbons by CO2 is that produce efficient extraction of hydrocarbons by CO2 is that required to make the CO2 dense enough to be an effective solvent for some of the hydrocarbons present. Phase-composition data presented by Orr et al. confirmed that relatively dense liquid presented by Orr et al. confirmed that relatively dense liquid CO2 extracted more and heavier hydrocarbons than did low-density CO2 vapor. Slim-tube displacements conducted by Holm and Josendal also indicated that size and, to a lesser extent, structure of hydrocarbon molecules influence the density required for efficient hydrocarbon extraction. This paper examines directly the effects of molecular weight and structure on component partitioning in CO2/hydrocarbon mixtures. In a companion paper, we apply that analysis to the prediction of minimum miscibility pressure (MMP). Here we summarize recent reviews of the solubility of hydrocarbons in dense CO2 and then compare the measurements of component partitioning in four synthetic oil systems of similar overall molecular weight but composed of normal alkanes, branched alkanes, naphthenes, and aromatics. Also reported are component-partitioning measurements for mixtures of CO2 with a complex synthetic oil containing paraffins, naphthenes, and aromatics. Finally, phase-composition paraffins, naphthenes, and aromatics. Finally, phase-composition measurements for three CO2/crude-oil systems, including one to which an aromatic hydrocarbon mixture has been added, are presented. presented. Binary Systems Francis studied the phase behavior of a wide variety of binary and ternary systems containing CO2. The experiments were performed at about 77F [25C] and 950 psia [6.55 MPa] or at a performed at about 77F [25C] and 950 psia [6.55 MPa] or at a CO2 density of about 0.69 g/cm3. Table 1 presents some of his binary component-partitioning data to illustrate the effects of both molecular weight and type. The lightest alkanes through n-C12 are completely miscible. As molecular weight increases for alkanes larger than n-C12, the solubility of the hydrocarbon in CO2 decreases. For example, Francis found that CO2 extracts 16 wt% n-C 14, 8 wt%-n-C 6, and 3 wt% n-C18. Benzene and cyclohexane, like normal hexane, are completely miscible with CO2 at those conditions and therefore provided no additional evidence concerning the effects of molecular type. Hydrocarbons containing two rings were not miscible, however. CO2 solubilized 22 wt% decalin compared with 12 wt% tetralin and only 2 wt% of the entirely aromatic molecule naphthalene. Furthermore, only 2 wt% of the entirely aromatic molecule biphenyl partitioned into the CO2-rich phase, while 8 wt% of the partially. naphthenic molecule phenylcyclohexane did so. Apparently, naphthenes or aromatic phenylcyclohexane did so. Apparently, naphthenes or aromatic molecules containing naphthenic structures experience significantly higher extraction by CO2 than do purely aromatic molecules with the same number of carbon atoms. Normal alkanes remain the most efficiently extracted. The data presented by Francis also provide some evidence concerning the level of extraction for branched alkanes and direct evidence that alkyl groups improve hydrocarbon extraction, at least when they are attached to aromatic molecules. As Table 1 shows, naphthalene, a dicyclic aromatic, constitutes only 2 wt% of the CO2-rich phase. The addition of a methyl group, as in 1-methylnaphthalene, increases the solubility to 6 wt%. In addition, attaching the methyl group at a different location on the molecule (Position 2) further increases solubility to 9 wt%. Di-sec-butylbenzene and tri-sec-butylbenzene contain 14 and 18 carbon atoms, respectively. Yet, unlike normal paraffins with the same number of carbon atoms, these aromatics are completely miscible with CO2 at the experimental conditions. Apparently, the presence of several alkyl groups counters the negative effects of the benzene rings on solubility for these molecules. The miscibility of these two molecules suggests that highly branched alkanes would experience better extraction than straight-chain alkanes. Multicomponent Systems Holm and Josendal presented limited evidence suggesting that a more aromatic crude oil exhibited a lower MMP than did a similar paraffinic oil. They speculated that "although aromatics and paraffinic oil. They speculated that "although aromatics and naphthenes are less compatible with CO2 than paraffin hydrocarbons, aromatics and naphthenes are better solvents for petroleum heavy ends." SPERE p. 468