Extension of HLD-NAC Flash Calculation Algorithm to Multicomponent Mixtures of Microemulsion and Excess Phases

Khorsandi, Saeid (The Pennsylvania State University) | Johns, Russell T. (The Pennsylvania State University)



Current HLD-NAC theory and most simulators represent multicomponent mixtures with three lumped components, where the excess phases are also assumed pure. This can cause significant errors, and discontinuities in chemical flooding simulation for surfactant mixtures. We coupled the HLD-NAC and pseudo-phase models to develop an EOS for microemulsions where surfactant, polymer, alcohol, alkali and monovalent/divalent ions can partition differently into the excess phases and microemulsion phase as temperature and pressure are changed.

We develop a pseudo-phase model to calculate partitioning of components between lumped components or namely pseudo-phases. The pseudo-phase model is based on a transformed composition space. The partitioning model is based on different mechanisms such as cation exchange like reactions for ions and surfactant hydration properties. Next, the three-pseudo-component HLD-NAC EOS is used to calculate curvature of the interface and microemulsion phase composition based on pseudo-phases. That is, the microemulsion phase consists of a curved ruled surface between water and oil pseudo-phases. Polymer partitioning is updated based on micelle radius. Finally, the phase compositions are converted back from pseudo-phase space to the original composition space.

This model is the first comprehensive and mechanistic flash calculation algorithm based on HLD-NAC and pseudo-phase theory to calculate microemulsion properties for mixtures without the assumption of pure excess phases. This algorithm allows for modeling of the chromatographic separation of surfactant, soap, alcohol, alkali and polymer components in chemical flooding processes. Current microemulsion models usually ignore the differing partitioning of components between excess and microemulsion phases, generating discontinuities that slow computational time and adversely impact accuracy.