Transient Interfacial Tension Behavior of Crude-Oil/Caustic Interfaces

Abstract

Summary A systematic study is conducted to describe the transient interfacial tension (EFT) behavior of crude-oil/caustic interfaces. The variation in this behavior with variation in the composition of the aqueous phase and other parameters is the focus of this study. It is found that at low pH values, the experimentally observed behavior is similar to that reported earlier. At high pH values, however, a maximum in IFT is observed, followed by a minimum at a late time. Sodium chloride is shown to have no significant effect on the shape of the transient IFT curve, but the absolute values of IFT are at lower NaCl concentrations. Divalent ions dramatically increase the IFT. A chemical model that takes into account the kinetics and the detailed chemistry of the process is used to explain the observed phenomena. The phenomenological surface phase approach is used to phenomena. The phenomenological surface phase approach is used to model the interface. The resultant set of ordinary differential equations is linearized by making certain justifiable, simplifying assumptions. The analytical solution to the set of equations yields the variation of the various concentrations with time. The IFT is then related to the interfacial concentration of the surface-active species, A, by the Gibbs equation. The variation in the transient IFT behavior on changing the kinetic constants, phase volumes, and ionic concentrations is discussed. Finally it phase volumes, and ionic concentrations is discussed. Finally it is shown that the extension of this model to a two-component model would successfully explain all the experimentally observed phenomena. Introduction The lowering of DT at crude-oil/caustic interfaces was observed by Atkinson and Nutting many years ago. This led to the use of caustic solutions as EOR agents. Much work has since been done to elucidate the mechanisms of such a process. It is now well established that the alkali-sensitive components of the crude oil diffuse out to the interface, where they react with the caustic to generate surface-active species. These surface-active species either adsorb at the interface to lower the IFT or diffuse out into the bulk aqueous phase. This sequence of events-convective-diffusion reaction-gives rise to interesting IFT maxima and minima as dictated by the relative kinetics of each step. Such a phenomenon is not restricted to the system studied here. Similar observations may be made in any system involving interphase mass transfer of a solute between two miscible fluid phases accompanied by a chemical reaction. The emphasis in this paper, however, is placed on understanding the mechanisms and kinetics of crude-oil/caustic interfaces, and the experimental results and kinetic models are presented with that objective in mind. From the onset, the chemistry of the process has been the focus of much attention and controversy. Attempts were made by Seifer and Howells, Seifert, Yen et al., Farmanian et al., and Wasan et al. to isolate and identify the surface-active components present at the interface. The interfacial region was shown to present at the interface. The interfacial region was shown to consist mainly of long-chain carboxylic acids with a wide range of molecular weights (- 300 to 400) and chemical structures. Although most of the long-chain acids found were saturated aliphatics, some unsaturated, substituted, and aromatic acids and diacids were also identified. Nitrogen and sulfur heteroatoms were also found to be concentrated at the interface. Dunning et al. showed that porphyrins and porphyrin/metal chelated complexes exhibit strong porphyrins and porphyrin/metal chelated complexes exhibit strong interfacial activity and film-forming characteristics. The presence of other metals like copper, zinc, and nickel in presence of other metals like copper, zinc, and nickel in oil-soluble forms as porphyrin/ metal chelate complexes led Dodd et al. to conjecture that the interfacial films were stabilized by resins, porphyrins, porphyrin ring oxidation products and protein/metal porphyrins, porphyrin ring oxidation products and protein/metal salts. Some recent results show how this complex chemistry might affect the dynamic IFT behavior. They separated the crude oil into three fractions on the basis of differences in boiling points, Fraction 1 being the lowest boiling. Fraction 2 exhibited points, Fraction 1 being the lowest boiling. Fraction 2 exhibited a minimum in IFT with time. Fraction 3, which presumably contained the higher-molecular-weight asphaltic components, however, showed no minima at all. This leads us to believe that more than one species is involved in the diffusion and reaction process occurring at the interface. The overall behavior of the crude oil will therefore be governed by the relative amounts of these species present and by their individual diffusion and kinetic constants. present and by their individual diffusion and kinetic constants. From this brief review, it is evident that the chemistry of the reaction between caustic and crude oil is extremely complex. This fact must be home in mind when certain simplifying assumptions are made later in this paper in developing the kinetic model. Ward and Tordai were the first to describe quantitatively the role of diffusion in the time-dependent IFT behavior of solutions. Later England and Berg 12 studied the same problem of transfer of surface- active solute across liquid/liquid interfaces-no interfacial reactions were involved. Most of the earlier attempts model the transient behavior of solute extraction from the oleic phase as a simple diffusional process where the flux into and out of the interfacial region is represented by Fick's law. This, however, is not the case for the system considered here, as has been demonstrated by many studies on liquid/liquid extraction of a solute from one phase into another. Wei accumulated qualitative data on a wide range of liquid/liquid extraction systems and showed that in almost every instance where either of the two phases contained reacting solutes, there was evidence of localized Marangoni disturbances and convection currents were spontaneously set up. In such systems, as Sternling and Scriven later showed, such interfacial turbulence was more intense whensolute was being transferred out of the phase of higher viscosity,solute was transferred out of the phase of low diffusivity,there were large differences in kinematic viscosity and solute diffusivity between the two phases,steep concentration gradients were present near the inter-face,IFT was highly sensitive to the concentration of solute,both phases had low diffusivities and viscosities,there were no surface-active species present, andthe interfaces were of large extent. For our system, all but Condition 7 are met. Indeed, the presence of surface-active species has been shown to dampen out presence of surface-active species has been shown to dampen out interfacial turbulence. For crude-oil/caustic systems, however, such interfacial turbulence is quite pronounced and has even been shown to lead to spontaneous emulsification. In addition to the interfacial turbulence generated by the gradient in chemical potential, the instrument used in this study to measure IFT-the spinning-drop tensiometer-generates its own fluid motions in the external (aqueous) and internal (oleic) phases. The detailed hydrodynamics of this instrument was recently described by Currie and Nieuwkoop, who showed theoretically and experimentally, by use of an n-butanol/water system, that at speeds of revolution less than 5,000 rev/min, the axis of the spinning drop in a spinning-drop tensiometer is displaced from the axis of rotation of the capillary tube because of buoyancy and Coriolis forces. This displacement gives rise to fluid convective movement in both phases. The Eckman cones and the flow-velocity regimes inside and in the vicinity of the drop were observed visually. However, to avoid interfacial deformation from these effects and to ensure no dependence of IFT on speeds of rotation, the IFT was measured at 7,000 rev/min. Even at this speed of rotation, it is conceivable that small convective currents may be present. It is evident from the preceding discussion that modeling the transport of solute from the crude oil to the interface and from the interface into the aqueous phase as a purely diffusional process is clearly a gross oversimplification. process is clearly a gross oversimplification. SPERE P. 228