Towards digital twins of distillation columns

Distillation is the most frequently used thermal separation process and accounts for a remarkable share of the global energy consumption. Therefore, design of distillation columns needs to target highly efficient operation. The present contribution describes current achievements of modeling and predicting the complex two-phase flow in tray columns, which strongly affects the heat and mass transfer processes and thus the separation efficiency. The strategy for creating a digital twin is based on combining a morphology adaptive multifield two-fluid model with reasonable abstraction of the vapor injection through fixed and push valves. Validation is carried out against experimental data of threedimensional phase distribution around single valves. In a second stage a pre-processing tool was developed to automatically set up simulation cases for industrial-scale applications. The simulations will be compared against phase fraction and velocity measurements of a conductivity sensor array that was applied in a large-scale column mockup.