Abstract
Enhanced oil recovery (EOR) methods often involve water-based flooding where surfactants, co-solvents and/or other chemicals are present in the aqueous phase being injected into the oil-bearing rock. The chemical flood may facilitate oil recovery by altering physical and chemical properties of the water-oil-chemical(s) system present in the porous medium such as interfacial tension, phase partitioning and liquid-liquid mixing. Prior to testing the efficiency of a chemical flood, initial laboratory experiments are routinely conducted to investigate the phase behaviour and thermodynamic properties (e.g. brine compatibility) of the present fluid system. However, such laboratory screening experiments consume significant time that scales linearly with the number of chemicals and combinations hereof to be characterized.
The Conductor-Like Screening MOdel (COSMO) has been developed over the last decades with the aim of predicting thermodynamics of complex mixtures based on single-molecule charge distributions. The charge distributions are calculated based on the molecular structure using density functional theory (DFT) to estimate molecular energies of quantum mechanical origin. Despite the fact that applications of COSMO have predominantly been used for biomedical, environmental, and chemical engineering research, applications of the theory show great promise for EOR research.
In this paper we demonstrate the versatility of applying the COSMO-theory to predict thermodynamic properties of water-oil systems, such as liquid-liquid mixing, phase partitioning, solubility, interfacial tension and ternary phase diagrams. Moreover, we describe an in-house developed database of 3D charge density landscapes of 100+ surfactants and present an in-depth analysis hereof with respect to single molecule properties as well as water-oil-surfactant system properties. Strikingly, we show that COSMO-theory in combination with a specific application of probability theory and the method of moments (MoM) can be used to estimate system properties such as interfacial tension, critical micelle concentration and phase partition coefficients.
We demonstrate that COSMO-theory provides a powerful framework for large-scale, fast and inexpensive initial computational prediction of single-molecule and continuum system properties. In particular we foresee that future applications of COSMO-theory to accurately estimate model input-parameters for chemical EOR simulators, such as UTCHEM, would be highly beneficial to the EOR research community.
