Interpretation of the Change in Optimal Salinity With Overall Surfactant Concentration

Abstract

Abstract Background. For chemical flooding formulations, optimal salinity changes with overall surfactant concentration when the phase behavior is observed in test tubes. Applying these observations to the mathematical simulator is questionable because chromatographic mechanisms during displacement through porous media result in different compositions. Purpose. This work sought the mechanism for the observed change so that calculated optimal salinity can be expressed through the appropriate intensive variable rather than overall surfactant concentration. Method. Association of the alcohol has been described by partition coefficients for distribution of the alcohol among brine, oil, and surfactant. The alcohol was isopropanol (IPA), 1-butanol (NBA), or tertiary amyl alcohol (TAA) in the systems in which they were included and was used to represent a disulfonate in the system with Petrostep petroleum sulfonate. Association of sodium and divalent ions with surfactant has been described by the Donnan equilibrium model, which experimental observations show can be applied to microemulsions as well as to micelles. Conclusions. For the seven systems investigated, the change in optimal salinity is a function of (1) the alcohol associated with the surfactant and (2) the divalent ion fraction of the associated counterions. Introduction Reed and Healy reviewed the concept of optimal salinity for minimum inter-facial tension (IFT) and its relationship to phase behavior. They showed that, as a first approximation, phase behavior can be represented by electrolyte concentration and three pseudocomponents: brine, oil, and surfactant plus cosolvent. If the system actually contains three components plus sodium chloride, optimal salinity should be independent of overall surfactant concentration and WOR. However. in the system Reed and Healy investigated, optimal salinity changed with overall surfactant concentration and WOR, which indicates that the system did not contain just sodium chloride plus three additional components. To handle this problem, Vinatieri and Fleming suggested using regression analysis to determine the best set of pseudocomponents. Then alcohol can be included with the oil and brine as pseudocomponents. Blevins et al. examined the phase behavior of a quaternary system (with brine as a pseudocomponent) by examining pseudoternary planes on a quaternary diagram. Glover et al. showed that the change in optimal salinity of a system containing divalent ions can be modeled by (1) considering the equilibrium composition of the brine, and (2) describing optimal salinity as a linear function of the concentration of divalent ions associated with the sulfonate. They assumed that NEODOL 25-3S did not associate divalent ions. (NEODOL 25-3S is a sodium salt of C12-C15 alkyl ether sulfate, with an average ethylene oxide number of three. Hereafter in this paper it is abbreviated as N253S.) Pope and Nelson showed that phase behavior and IFT's can be modeled in a compositional simulator when optimal salinity and the upper and lower limits of the Type III environment are known. The purpose of this work is to model alcohol or multiple surfactant components and divalent ions so that they can be included in a compositional simulator. Thermodynamic Analysis The Gibbs phase rule is used to show that a four-component system of pure oil, surfactant, water, and NaCl has an optimal salinity that does not depend on overall surfactant concentration. SPEJ P. 971^