Modelling of reactive separation systems
No Thumbnail Available
Doctoral thesis (article-based)
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Chemical engineering report series, Kemian laitetekniikan raporttisarja, 50
AbstractThe first part of this study is concerned with setting up a reactive distillation process for production of tert-amyl-methyl ether (TAME). This work was linked to the development of etherification technology of Neste Oy. TAME production makes possible to upgrade some low value olefinic components to high value gasoline. Moreover, it has a significant impact in the reduction of the air pollution caused by the cars by introducing oxygen to the gasoline. However at the time of the study, there was no technology available for production of that component. Reactive distillation (RD) had been applied successfully to the production of the tert-butyl-methyl-ether (MTBE). Thus it seemed worth of trying to apply RD to TAME production as well. The actual work of setting up the process was accomplished using a simulation model of a reactive distillation column. Arrangement of the column and conditions of the experimental runs were determined with the model developed earlier by Aittamaa and Kettunen (1993). The pilot run was successful, so that ethers could be produced as planned and experiments verified with the existence of the operating regimes predicted by the model. The results of this study had a significant impact on the development of the highly successful NExTAME and NExETHERS technologies, even if the final solution was based on the Side Reactor Concept (SRC), i.e. a combination of a distillation column and a reactor connected to the column via side streams, rather than on RD. The second part is the development of a rate-based model of a reactive distillation column including the effects of incomplete lateral mixing on the trays. Most published tests with RD have been performed with small pilot or bench scale columns. In such columns vapour and liquid mixing is nearly complete. However, that is not the case in large industrial columns. On the other hand, making tests with reactive system in columns having diameter of two meter or more is very expensive and practically impossible for most research institutions. Not only the sheer size and utility consumption of such devices are large, but the feed and product volumes are huge even for a short run. With non-reacting systems it is often possible to recycle the products back to the feed, but when reaction takes place, that is much more difficult. If lateral mixing is suspected to have importance in some particular case, a mathematical model is probably the only viable way to estimate its significance before the full-scale plant is built. Two different models for the effects of the lateral concentration profiles on reactive distillation trays were developed. The first model is an eddy diffusion model, the other one is a mixed pool model of reactive distillation trays. The basic principles of both models are known already earlier but both include novel features. Similar models have not been applied earlier to the reactive distillation. The eddy diffusion equations are solved simultaneously and rigorously with the other equations of the equation group describing the column, instead of using solutions of strongly idealized problem as an approximation as has been usual with non-reactive distillation columns separating nearly ideal mixtures. The mixed pool model differs from the ones presented earlier with its ability to take different mixing cases of the vapour phase into account. The normal assumption has been that the vapour is fully mixed between the column trays. Here is presented a simple and efficient method for treating columns with unmixed vapour flow from tray to tray. Two types of the liquid flow are considered – liquid flow to same direction on adjacent trays or liquid flow to contrary directions on adjacent trays. In the third part a development of a combined model for SRC system is presented. When the SRC was found to be very useful in etherification processes, it was considered useful to develop a dedicated modelling tool for it. The part of the model describing the distillation column is in principle an equilibrium stage reactive distillation column model. The part of the model describing the side reactor(s) consists of a series of reactor segments. Each reactor segment is considered internally fully mixed, but using sufficient number of such segments in series, a plug flow reactor can be modelled with good accuracy. There is a significant flexibility in structure of the reactor sequence. The reactor system may operate in vapour or liquid phase or in co-current two-phase mode. Intercoolers or heat flows between surroundings and the reactor segments can be used. As examples of the application of this model a case study of comparing SRC and RD in the production of tert-amyl-methyl-ether (TAME) and diisobutene is presented. The study showed that SRC optimisation of the reaction conditions is of crucial importance to the performance of the process. The developed model was proofed to be an efficient tool for this purpose. The fourth part of this work was a part of development work of a novel tower packing by Sarvis Oy, The packing is continuously manufactured with the name HUFO. The results of the experimental work performed during the development project were used in order to develop an interfacial area correlation for a packed bed. This correlation attempts to take the structural detail of the packing into account by introducing the width of the surface elements of the packing type into the correlation. In most correlations the specific surface and nominal size of the packing have been used for this purpose. However, the development of random packings has gone towards more and more heavily perforated shapes. The actual width of the packing walls interacting with the fluids in a bed consisting of some modern packing of certain nominal size is very different from the nominal diameter. Thus various correlations based on different combinations of the specific surface and of the width of the packing surface elements were tried. The best fit between the correlation and the test results was achieved when only the width of the surface elements was used in the correlation. Another important result for practical chemical engineering is the observation that wettability of the plastic packing, known to be an important factor affecting the efficient interfacial area was greatly improved by the thin film of impurities deposited on the packing surface during the test program. The same thing may happen in industrial applications as well where impurities are common and run times long. Thus in many cases the efficiency correlations developed using brand new packings may be overly conservative.
distillation, absorption, simulation, etherification
- Bravo, J. L., Pyhälahti, A., and Järvelin, H., Investigations in a catalytic distillation pilot plant: vapor/liquid equilibrium, kinetics, and mass-transfer issues, Industrial & Engineering Chemistry Research 32 (1993) 2220-2225.
- Kenig, E. Y., Pyhälahti, A., Jakobsson, K., Górak, A., Aittamaa, J., and Sundmacher, K., Advanced rate-based simulation tool for reactive distillation, AIChE Journal 50 (2004) 322-342.
- Pyhälahti, A., and Jakobsson, K., Rate-based mixed-pool model of a reactive distillation column, Industrial & Engineering Chemistry Research 42 (2003) 6188-6195.
- Jakobsson, K., Pyhälahti, A., Pakkanen, S., Keskinen, K., and Aittamaa, J., Modelling of a side reactor configuration combining reaction and distillation, Chemical Engineering Science 57 (2002) 1521-1524.
- Ouni, T., Jakobsson, K., Pyhälahti, A., and Aittamaa, J., Enhancing productivity of side reactor configuration through optimizing the reaction conditions, Chemical Engineering Research and Design 82 (A2) (2004) 167-174.
- Pyhälahti, A., and Järveläinen, M., Correlation for interfacial area in a packed column, 1990, 10th International Congress of Chemical Engineering, Chemical Equipment Design and Automation (CHISA '90), 8 p.