(332e) Modeling of Liquid Distribution in Random Packed Columns

The phenomenon of liquid maldistribution in random packed columns has long been known and has been the subject of numerous scientific investigations. In many cases, maldistribution affects the separation efficiency of packed columns and must be taken into account in the design. So far, more accurate predictions about the maldistribution can only be made based on experimental studies. A new approach in this field is the WelChem cell model [1]. The idea behind this approach is to accurately predict the maldistribution based on virtual irrigation experiments that require only the CAD data of a single packing element.

The validation and further development of this model is the aim of the project "Cell Model for Design of Packed Columns and Liquid Distributors" [2-4] funded by the Bavarian Research Foundation (AZ-1033-12). Within the project maldistribution is experimentally investigated by varying the experimental parameters column diameter, packing type, packing element nominal size, packing height, initial liquid distribution, liquid load and gas load. Most of the experiments are performed in a column operated with air/water with a diameter of 1.2 m and a packing height up to 3.0 m where the liquid distribution is measured in a liquid collector at the bottom of the column [5]. The experimental raw data gained in the project are also publicly available for research purposes (www.apt.mw.tum.de/maldistribution).

The original version of the WelChem cell model [1] predicts liquid distribution from cell to cell using directional dispersion coefficients independent of the liquid load. These are obtained from a virtual 3D irrigation experiment using the CAD model of a single packing element.

In order to map the influence of the liquid load in the WelChem cell model, which has not been considered previously, the original approach is extended by an additional distribution effect, which occurs in packing areas with a high local hold-up. The influence of the gas load is implemented locally via an amplification of the distribution effects coupled to column fluid dynamics. Furthermore, the dimensions of the unit cell for each packing type are defined uniformly as a function of the bulk density. The influence of the column wall is included by void cells randomly distributed in the outermost ring of the cell layer. These void cells have no distribution function and pass all liquid to the cell directly below.

Comparative calculations with experimental data show that the liquid distribution for the considered packing types can be predicted in the technically relevant operating range with very good agreement. The modified model is implemented in the column design software TrayHeart and can be applied to a wide range of random packing types.

[1] A. Wild and V. Engel: Ein neuartiges Zellenmodell zur Maldistributions- und Verteilgütenbestimmung. FVT-Meeting, Karlsruhe, Germany, 2007.

[2] F. Hanuš, R. Hoffmann, S. Rehfeldt und H. Klein: Experimentelle Untersuchung der Flüssig­keits­mal­distribution in Füll­körper­ko­lonnen mit unterschiedlichen Durchmessern. FVT-Meeting, Bremen, Germany 2015.

[3] F. Hanuš, S. Rehfeldt and H. Klein, Chem. Ing. Tech. 2015, 87, No. 8, 1063-1064. https://doi.org/10.1002/cite.201550041

[4] F. Hanuš, J. Solá, K. Hoffmann, S. Rehfeldt und H. Klein: Experimentelle Untersuchung der Flüs­sig­keitsverteilung in Füll­körper­packungen mit verschiedenen Nenngrößen und Ko­lon­nen­durch­mes­sern. FVT-Meeting, Garmisch-Partenkirchen, Germany, 2016.

[5] F. Hanusch, S. Rehfeldt and H. Klein, Liquid Maldistribution in Random Packed Columns: Experimental Investigation of Influencing Factors. Chem. Ing. Tech. 2017, 89, No. 11, 1550-1560. https://doi.org/10.1002/cite.201700015