Modelling and Optimization of Batch Distillation

Abstract: The detailed dynamic modelling, simulation and optimisation of a batch distillation process at Bayer AG is presented. The model considers a mixture of 10 components separated in a 40 tray column. Simulations performed in the process modelling tool gPROMS proved that the model gives a very accurate description of the process behaviour. To increase the profit of the process an optimisation of the operating procedure was performed, with the reflux ratio as the only control variable. The optimisation resulted in an increased profit, i.e. the product yield was increased while at the same time production time was decreased. Another optimisation was performed to account for the periodicity of the process, by optimising the initial conditions of the batch. These depend on the amount of fresh feed and the amount and composition of material which is recycled from the intermediate cut and mixed with the fresh feed. This optimisation also resulted in an increased profit. Overall, gPROMS proved to be a powerful tool for solving optimisation problems of large scale models of great complexity such as the batch distillation process.