The Effects of Physical Property Estimation on Packed Column Distillation Simulation with a Rate-based Mode

Abstract

Background/Objectives: we investigate the effects of estimation models for physical properties on the rate-based simulation and discuss what we should be aware of for obtaining better simulation results.

Methods/Statistical analysis: Three kinds of physical property models considered are an activity coefficient, a thermodynamic factor, and a liquid excess molar volume. We use experimental data obtained from a pilot-scale plant for comparison with numerical examples. We investigate the effects of estimation models for physical properties on the rate-based simulation and discuss what we should be aware of for obtaining better simulation results.

Findings: The thermodynamic factor models do not affect largely the simulation results in the distillation simulation where the mass transfer resistance of the vapor phase film is dominant.

The liquid excess molar volume of the ternary mixture of Methanol, Ethanol, and Water shows at most 3% deviations from the averaged molar volume. It affects the simulation results markedly where the composition of the liquid mixture passes through the region of the large excess molar volume.

Improvements/Applications: The simulation results indicate that even small difference in continuous operation becomes much pronounced in batch operation with multicomponent systems.