Cyclic Distillation Technology

Distillation processes are one the most widespread separation technologies that are used in the chemical and biochemical industry. It is a simple, well known and effective method for separating various mixtures. It is also a high energy-demanding technology with a high environmental impact as the separation mixtures are boiled in the bottom of the column and condensed in the top. In an attempt to intensify the process, different alternatives have been proposed. One
of which is the cyclic distillation, where the phase movements inside the column are separated, giving a high separation efficiency and thus allowing for a reduction in energy demand, number of stages or an increase in throughput and conversion for reactive distillation processes.

In this thesis, the cyclic distillation technology has been studied and analysed in order to be able to understand the process. It is easier to propose a cyclic distillation as an alternative to conventional distillation with a higher process understanding.

A mass and energy balance stage model is proposed, which is a high fidelity model accounting for timedependent temperature and vapour flow rate and allowing for multiple feed locations, and side draws. The model performance was evaluated and compared to previous models. It was shown that if there is a high development in the stage temperature or vapour flow rate over a vapour flow rate, the proposed mass and energy balance model would be a suitable choice.

The presented model is further expanded to account for reactive cyclic distillation processes, including the reaction heat. Different reactive cyclic distillation cases are presented, and a detailed analysis of the stage behaviour over the vapour flow period is shown for the methyl tertbutyl ether case. This stage behaviour analysis showed that as the vapour flow period progresses, significant changes in the reaction and the separation affects the process.

The performance of three different reactive cyclic distillation cases was also evaluated for a disturbance in the inputs. Furthermore, the feasibility of a reactive cyclic distillation was discussed. Three performance indicators are proposed: the extent of reaction over a vapour flow period, the relative distance to equilibrium at the end of a vapour flow period and the mean Damköhler number over a vapour flow period. Of these, the extent of reaction and the Damköhler number is useful for investigating the performance and feasibility of a reactive cyclic distillation process. The distance to equilibrium could indicate whether an assumption of chemical equilibrium is valid or not. Based on existing feasibility conditions for reactive and cyclic distillation, some helpful observations are made that could facilitate part of a design method. However, this design method still requires iterations to find some of the important specifications as currently available methods.

All in all, the work presented in this thesis shows development of cyclic distillation technology and how new models can help to make more high fidelity studies useful for process analysis in terms of designing, control strategies and process feasibility.