Determination of the Uncertainty Bounds of a Continuous Distillation Code: Effect of Input Variability and Model Uncertainty

Abstract

In this work, the effect of input variability and model uncertainty on the distillate composition of a continuous distillation tower is studied. To do that, we developed a stationary distillation code by combining mass and energy balance equations with a liquid-vapor equilibrium model and tray efficiency correlations. Feed and model uncertainties were modeled by using normal and uniform distributions respectively. A Monte Carlo propagation method was used to determine the upper and lower uncertainty margins of the distillate composition. The results of the application to a methanol-water distillation showed that the model uncertainty is as high as that of the feed variability. The information can be useful for the robust design of distillation towers.