Process intensification is a design philosophy of chemical and reactor engineering aiming at significant energy, capital, environmental and safety benefits. One of the most promising ways to the process intensification is to deliberately integrate different phenomena or operations. This helps to lower equipment in number and size and to improve the process efficiency. Many successful examples of such integrated processes can be found among the so-called reactive separations combining reaction and separation steps in a single unit. For instance, simultaneous reaction (either homogeneously or heterogeneously catalyzed) and distillation give rise to the reactive distillation process, with several important advantages as compared to traditional reactor/distillation sequence. These advantages comprise increased yield due to overcoming chemical and thermodynamic equilibrium limitations, increased selectivity via suppression of undesired consecutive reactions, reduced energy consumption via direct heat integration in case of exothermic reactions, avoidance of hot spots by simultaneous liquid evaporation and ability to separate close boiling components (Agreda et al., 1990).

Along with reactive separations, there also exist other possibilities of process integration, for example by linking different separation units together. For the distillative separation of ternary mixtures, the well known Petlyuk column configuration consisting of two fully thermally coupled distillation columns was developed (Petlyuk et al., 1965). By this concept, condenser and reboiler of the first column (prefractionator) are replaced by additional thermal links with the second column (main column), which decreases both capital and operating costs. A further cost reduction is possible by the realization of the Petlyuk configuration in a single shell. In the so-called dividing wall column (DWC), a vertical partition (wall) is introduced to arrange a prefractionator and a main column into a single structure (Kaibel, 1987). In comparison with conventional distillation column sequences, the DWC offers the following advantages:

- high purity for all three product streams reached in only one column, - high thermodynamic efficiency due to reduced remixing effects with respect to the middle component, - reduced number of equipment units.

Both reactive distillation columns and dividing wall columns are further developments of a conventional distillation unit, and, at the same time, they represent two different ways of integration. The advantages of both integrated processes could be enhanced, if they are combined via a further integration step. The resulting new and highly integrated configuration comprises the prefractionator and main column together with the reacting section(s) in a single-shell column. Such a unit is called reactive dividing wall column (RDWC) (Mueller et al., 2004). Of course, adequate modeling of RDWC, with their multiple and complex interactions, represents a demanding task.

In the literature, non-reactive DWC have been described by combination of conventional column models (e.g., combined in Petlyuk configuration) based on the equilibrium stage concept (Becker et al., 2001). However, it was shown that the application of the equilibrium stage concept to complex systems even in conventional distillation is non-trivial (Taylor et al., 2003). The so-called rate-based approach taking actual mass and heat transfer rates and process hydrodynamics into account is preferable for the modeling of reactive separations (Kenig et al., 2004). A detailed model for heterogeneously catalyzed reactive distillation processes has been suggested by Górak & Hoffmann (2001) and further developed to cover different reactive and non-reactive separation processes and various column configurations (Kloeker et al., 2003).

In the present work, this model has been extended to cover both the DWC and RDWC configurations. Among others, important particular features of these columns as compared to conventional (reactive) distillation units are taken into account, namely, the vapor distribution below the dividing wall and the heat transfer through the dividing wall. These effects were usually neglected in previous publications. In reality, however, the vapor distribution is not an operating parameter to be easily controlled, rather it represents a self-adjusting parameter, since the vapor distribution in a RDWC has to fulfill the condition of equal pressure drop at both sides of the dividing wall. This issue deserves close attention. The heat transfer through the dividing wall is caused by the temperature difference between the prefractionator and main column sides. This phenomenon can also affect the column efficiency (Lestak et al., 1994).

Discretization of the dividing wall column in axial direction in combination with the detailed rate-based approach results in a large and highly non-linear algebraic system of equations. Therefore, a proper implementation and solution of the model equations is important. The rate-based model described above is implemented into the simulation environment ASPEN Custom Modeler. The model has been successfully validated for a non-reactive DWC (Mueller et al., 2004) using literature data for a ternary alcohol mixture.

As a test system for the investigation of the RDWC configuration, the transesterification of dimethyl carbonate (DMC) with ethanol to diethyl carbonate (DEC) and methanol was found to be interesting (Mueller et al., 2005). The reaction system consists of two consecutive reactions with the intermediate product ethyl methyl carbonate (EMC). Both reactions are equilibrium limited reactions, which reduces the DEC yield in conventional reactors. A reactive distillation process with this system has been recently studied by Richter et al. (2006). Here we consider the reactive distillation in the prefractionator equipped with a catalytic packing and combine this process with a simultaneous separation of reactants and products in a main column, thus building a RDWC configuration.

The simulation results demonstrate that a high DEC yield can be reached within the RDWC and, besides, the product streams are of sufficient purity. The high purity value product DEC is obtained at the bottom of the column, while EMC and non-reacted ethanol are removed in the side-draw stream which can then be recycled. The distillate stream at the column top contains the binary azeotrope methanol/DMC.

Further, the influence of vapor distribution and heat transfer through the dividing wall has been investigated. For this test case, no significant effect of the heat transfer has been found. On the contrary, a proper description of the vapor distribution in the column has been shown to be essential. Finally, important design and operating parameters (e.g., reflux ratio) have been identified to allow for the future alternative RDWC configuration development.

Acknowledgement

The support of the European Commission in the context of the 6th Framework Programme (INSERT– Integrating Separation and Reaction Technologies, Contract No. NMP2-CT-2003-505862) is greatly acknowledged

References

Agreda, V. H.; Partin, L. R; Heise, W. H. High-purity methyl acetate via reactive distillation, Chem. Eng. Prog., 1990, 86(2), 40-46.

Becker, H.; Godorr, S.; Kreis, H. Partitioned distilliation columns - why, when & how. Chem. Eng., 2001, January, 68–74.

Górak, A.; Hoffmann, A. Catalytic distillation in structured packings: Methyl acetate synthesis. AlChE J., 2001, 47, 1067–1076.

Kaibel, G. Distillation columns with vertical partitions. Chem. Eng. Technol., 1987, 10, 92–98.

Kenig, E. Y.; Pyhälachti, A.; Jakobsson, K.; Górak, A.; Aittamaa, J.; Sundmacher, K. Advanced rate-based simulation tool for reactive distillation. AIChE J. 2004. 50, 332–342.

Kloeker, M.; Kenig, E. Y.; Schmitt, M.; Althaus, K.; Schoenmakers, H.; Markusse, P.; Kwant, G. Influence of operating conditions and column configuration on the performance of reactive distillation columns with liquid-liquid separators. Can. J. Chem. Eng., 2003, 81, 725–732.

Lestak, F.; Smith, R.; Dhole, V. R. Heat transfer across the wall of dividing wall columns. Trans IChemE, 1994, 72, Part A, 639-644.

Mueller, I.; Kloeker, M.; Kenig, E. Y. Rate-based modelling of dividing wall columns - a new application to reactive systems. Proc. PRES 2004, Prague (Czech Republic), August 2004, vol. 4, 1325-1326.

Mueller, I.; Kenig, E. Y. Integration of reaction and separation in a dividing wall column. Proc. CAMURE-5 & ISMR-4, Portoroz (Slovenia), June 2005.

Petlyuk, F. B.; Platonov, V. M.; Slavinskii, D. M. Thermodynamically optimal method for seperating multicomponent mixtures. Int. Chem. Eng., 1965, 5, 555–561.

Richter, J.; Górak, A.; Kenig, E.Y. Catalytic Distillation, In: Integrated Reaction and Separation Operations: Modelling and Experimental Validation (Eds. H. Schmidt-Traub & A. Górak), Springer, 2006

Taylor, R.; Krishna, R.; Kooijman, H. Real-world modeling of distillation. Chem. Eng. Prog. 2003, 99(7), 28–39.