O

mass transfer kinetics chemical equilibrium phase and chemical equilibrium mass transfer kinetics reaction kinetics phase equilibrium reaction kinetics model complexity

Fig. 3.5. Model complexity in simulation of reactive distillation

3 Application of Reactive Distillation and Strategies in Process Design | 55 3.3.2.1 Catalyst Selection and Reaction Kinetics

By analogy with MTBE synthesis, the decomposition of MTBE is also an acid-catalyzed and equilibrium-limited reaction. Ion-exchange resins were found to be suitable for this reaction. Besides catalyst activity and selectivity, a major requirement is to provide a catalyst lifetime of the same scale of time as the period of plant turnaround. Otherwise, catalyst replacement would require a shutdown of the entire unit. However, the results of run-time testing experiments have shown that the ion-exchange catalyst used provides high stability for more than 8000 h. Only slight decreases in catalyst activity have been observed.

To determine reaction kinetics, batch experiments were performed in the temperature range 90-120 °C. Beginning with pure MTBE, or different mixtures of MTBE and methanol or isobutene, all experiments were started in the kinetic regime and were continued until chemical equilibrium was reached. Fitting the experimental data to the well-known kinetic approach suggested by Rehfinger and Hoffmann [11], good agreement was found between model prediction and experimental data (Fig. 3.6). This holds for both the description of the kinetic regime and the approach to chemical equilibrium. A consideration of the reaction medium influence on the chemical reaction as suggested by Fite et al. [12] gave no further improvement in the fit. All relevant side reactions have been included in the model, for example, diisobutene formation [12, 13], dimethyl ether formation, and hydration of isobutene.

Fig. 3.6. Results of a kinetic experiment performed in batch mode and comparison with model predictions based on the kinetic approach suggested by Rehfinger and Hoffmann [11] ([18], reprinted from Chem. Eng. Sci., Vol 57, Beckmann et al., Pages 1525-1530, Copyright 2002, with permission from Elsevier Science)

Fig. 3.6. Results of a kinetic experiment performed in batch mode and comparison with model predictions based on the kinetic approach suggested by Rehfinger and Hoffmann [11] ([18], reprinted from Chem. Eng. Sci., Vol 57, Beckmann et al., Pages 1525-1530, Copyright 2002, with permission from Elsevier Science)

3.3.2.2 Phase Equilibrium Model

The UNIQUAC model [14] was used to account for liquid phase non-idealities. The vapor phase was modeled using the Redlich-Kwong EOS [15]. The combination of both is implemented in Aspen Plus as the UNIQ-RK property model. UNIQUAC

parameters were fitted to experimental vapor-liquid equilibrium data from the literature and our own measurements. Special attention was paid to the exact description of the azeotropes (including the temperature dependence), as they were found to have a strong influence on the calculated conversion.

3.3.2.3 Steady-State Simulation

After the feasibility analysis had shown that the process for the decomposition of MTBE can be carried out using reactive distillation equipment, this conclusion had to be verified using a rigorous model. Since no information on mass-transfer limitations is available, the equilibrium stage approach is the model of choice. Therefore, such a simulation was set up in Aspen Plus (Version 10.2) with a column configuration containing a reactive section and non-reactive stripping and rectifying sections. The chemical reaction was implemented as a user subroutine. The amount of catalyst per theoretical stage was adjusted to match the properties of Katapak-S by Sulzer. The feed position was set to the top of the reactive section, since the reactant MTBE is high-boiling. The feed was pure MTBE. Its amount was varied in order to assess the influence on column performance. The distillate-to-feed ratio was adjusted to collect the isobutene-methanol azeotrope as top product. Two calculated concentration profiles are depicted in Fig. 3.7. It can be seen that both profiles are fairly similar, although MTBE conversion is very differ-

Fig. 3.7. Composition profiles of a reactive distillation column equipped with 10 reactive trays, and 10 non-reactive trays in the stripping and rectifying section each for high and low feed-rates. The feed position was located above the reactive section ([18], reprinted from Chem. Eng. Sei., Vol 57, Beckmann et al., Pages 1525-1530, Copyright 2002, with permission from Elsevier Science)

Isobutene

Isobutene

Fig. 3.7. Composition profiles of a reactive distillation column equipped with 10 reactive trays, and 10 non-reactive trays in the stripping and rectifying section each for high and low feed-rates. The feed position was located above the reactive section ([18], reprinted from Chem. Eng. Sei., Vol 57, Beckmann et al., Pages 1525-1530, Copyright 2002, with permission from Elsevier Science)

High Feed Rate O Low Feed Rate

0 0

Post a comment