(29b) Modeling of Mass Transfer of Random Packings for Gas-Liquid Systems

This paper presents a new relationship for the prediction of the efficiency of random packings of various shapes and sizes from 15 to 90 mm for distillation and absorption systems, based on the droplet flow model in packed beds.

The new relationship in connection with the model of the instationary diffusion for short residence times [1] allows the prediction of the separation efficiency values HTUOV or nt/H up to the flooding point for classical packing types such as Raschig rings, Intalox saddles or Pall rings, as well as for modern lattice-type packings with partially perforated wall surface such as Hiflow rings, VSP rings, Mc-Pac, Tellerette, Envipac, Dtnpac, Ralu-Flow, RSR, CMR rings and others [2].

The presented model has been derived from an experimental data bank containing data from 30 different test systems under vacuum and normal pressure and for above mentioned packings [2] and with mass transfer systems with major gas-side resistance, such as NH3-air/water, NH3-air/1n H2SO4, SO2-air/water and SO2-air/1n NaOH and CO2-water/air in columns with diameters dS= 0.15-1.4 m.

The model was verified by means of more than 60 different packings and for a total number of more than 4000 experimental mass transfer absorption and distillation data points for above mentioned systems. The new model described the experimental data very well with a mean relative error of ±15-below the loading line Fv<0,65 FV,Fl and ± 23% below the flooding point. The satisfactory accordance of calculated data with experiments is confirmed by an additional comparison with the model of Billet and Schultes [3] for chosen packing types such as metallic and plastic Pallrings and Hiflowrings. This additionally confirms the general validity of the new relationship. In comparison to other known correlations from literature, the new model does not contain any characteristic, empirical packing specific constants.

The limitations of the new model are discussed and the calculation examples and comparisons with other models are presented.

With the presented model, it is possible to predict the volumetric mass transfer coefficient in the gas phase βV.a and finally, in combination with [1], the efficiency of any random packing type and size, nt/H- value or HTUOV-value, in the whole loading range up to the flooding point.

[1] MAÆKOWIAK, J., Model for the prediction of liquid phase mass transfer of random packed columns for gas–liquid systems, Chem. Eng. Res. Des. 2011, 89, 1308–1320.

[2] MAÆKOWIAK, J., Fluid Dynamics of Packed Columns - Principles of the fluid dynamic design of columns for gas/liquid and liquid/liquid systems, Springer, Berlin, 2009.

[3] BILLET, R., SCHULTES, M., Prediction of mass transfer columns with dumped and arranged packings, Trans IChemE 1999, 77, 498-504.