(91a) Experimental and Theoretical Process Analysis of Membrane-Assisted Reactive Distillation

Experimental and Theoretical Process Analysis of
Membrane-Assisted Reactive Distillation

Johannes Holtbruegge, Sebastian Heile, Philip Lutze, Andrzej Górak

TU Dortmund University, Laboratory of Fluid
Separations

Emil-Figge-Strasse 70, D-44227 Dortmund, Germany

Phone: +49 (0) 231 755-2342, E-Mail: johannes.holtbruegge@bci.tu-dortmund.de

Keywords: Process Intensification, Hybrid Process, Reactive
Distillation, Membrane Separation, Modeling, Experimental Investigation

Due to decreasing fossil fuel
reserves, the development of innovative process concepts that lead to
economical and ecological benefits is absolutely necessary. Promising concepts
to increase sustainability are integrated and intensified processes. Those
processes integrate different phenomena or operations within one apparatus or
allow for strong interactions between unit-operations with different separation
characteristics in different apparatuses. Two examples are reactive
distillation which superimposes chemical reaction and distillation in one
apparatus as well as membrane-assisted hybrid separation processes that make
use of a membrane to overcome thermodynamic limitations. As reactive
distillation (RD) is nowadays the front-runner to overcome chemical
equilibrium-limited reaction systems the combination of RD with a membrane
separation is applicable for equilibrium-limited systems that exhibit strong
thermodynamic non-ideality thus having a high potential for process
improvement. To establish membrane-assisted reactive separation processes in
industry, detailed experimental and model-based investigations of hybrid
processes have to be performed. Thus it is possible to get on the one hand important
process know-how and based on this on the other hand process design tools for
membrane-assisted reactive separation processes. One of those tools is a
process analysis tool which supports the user to the decision whether the
process window of the considered phenomena for integration may happen at the
same time and place as well as which process parameters are the sensitive ones
to be intensified/improved to reach a defined process improvement.

In this work, both the process analysis
tool as well as the experimental results in a pilot plant for the development
of detailed models will be highlighted for a transesterification process.

Therefore, the transesterification
of propylene carbonate (PC) with two molecules of methanol (MeOH) to form the
target product dimethyl carbonate (DMC) and the byproduct 1,2-propanediol
(PDO) in a chemical equilibrium-limited reaction is chosen as case-study.

Reaction kinetics and chemical
equilibrium for this chemical system were determined using sodium methoxide as
homogeneous catalyst since no suitable heterogeneous catalyst has been found.
The chemical equilibrium constant was found to be very low (K_eq =
0.2 at T = 343 K) yielding in low reactant conversions. By process analysis on
the reaction, it was observed that by ensuring a molar excess of methanol or
selective removal of one product species an almost complete propylene carbonate
conversion can be obtained. To establish this while also having low recycle
streams, a reactive distillation process is suggested for this system. With a
molar excess of methanol in the feed stream it is possible to reach a complete
propylene carbonate conversion while obtaining an azeotropic mixture consisting
of dimethyl carbonate and large amounts of methanol in the upper part of the
column. To overcome this azeotrope the integration with a membrane at the top
of the column is identified. Hence, this mixture is fed to a vapor permeation
membrane that separates unreacted methanol from the mixture so that dimethyl
carbonate is enriched on the front side of the membrane. Unreacted methanol can
be recycled afterwards to the RD column to guarantee a high methanol excess. A
flowsheet of the membrane-assisted reactive distillation process is shown in
Figure 1.

Figure 1: Simplified flowsheet of
the membrane-assisted reactive distillation process

The integrated process is determined
to be promising enabling the need for detailed models. Hence, experimental
investigations of reactive distillation were performed in a pilot-scale
RD-column with an inner diameter of 0.05 m that was equipped with Sulzer BX
packing elements. The total packing height was 5.2 m and the total feed flow
rate was 4 kg/h. The achieved composition and temperature profiles along the
column were used successfully to validate the rate-based simulation model.

The vapor permeation was
investigated in a lab-scale plant equipped with a flat-sheet membrane module.
The separation characteristics of the dense, hydrophilic membrane Sulzer PERVAP^TM
1255 were determined by investigating the influence of feed composition, feed
pressure and feed temperature on the transmembrane flux and membrane selectivity.

For the process analysis, a detailed
simulation model comprising the rate-based model for reactive distillation and
an empirical solution-diffusion model was set up. The model parameters for the membrane
separation were determined from the lab-scale permeation experiments.

A process analysis on the detailed simulation
model is used to show the influence of important operational parameters like
reflux ratio and reactant ratio on the propylene carbonate conversion, the
product purities and the necessary membrane area.

Acknowledgements:

The financial
support of the German Federal Ministry of Education and Research for the
project ?Energy Efficiency Management and Benchmarking in the Process Industry?
is gratefully acknowledged.