Countercurrent-flow columns are widely used in production processes in the chemical industry and their application in ecological engineering is of increasing importance. A theoretical model is presented here that allows mass transfer to be described in terms of packing geometry and physical properties which influence the gas-liquid or vapour-liquid systems in absorption, desorption and rectification columns. The relationships derived from the model can be applied to all countercurrent-flow columns, regardless of whether the packing has been dumped at random or arranged in a geometric pattern.

Predicting mass transfer in packed columns

Publisher Summary This chapter studies fluid–flow characteristics of liquids in layers, with and without a superimposed gas velocity. The types of turbulence in layers need to be investigated and also the nature of a laminar layer containing ripples. As regards the theoretical studies of film flow, it is shown in the chapter that it is possible to predict quite accurately the flow behavior in the smooth laminar flow regime of the film; unfortunately, this flow regime is not of great practical importance. The Kapitsa theory for wavy film flow appears to apply over only a very limited part of the total wavy flow regime. Several new experimental techniques for the study of film flow have been developed. Film flow is a special case of two-phase flow. It takes place along a solid surface of some sort, with only one free surface. The second phase in contact with the free surface of the film may be either a gas or a second liquid, which may be at rest or in motion relative to the solid surface on which the film flows. Film flow is distinguished from other forms of two-phase flow by the presence of large interfaces of basically simple geometrical configuration. Two-phase flows are also often further classified as single-component. The occurrence and applications of film flow in modern technology are numerous and important.

The Flow of Liquids in Thin Films

To evaluate quantitatively the results of gas-liquid absorption experiments, accurate liquid-phase diffusion coefficients and viscosities are needed. In this paper experimental values of these quantities will be reported for the binary systems carbon dioxide + water, carbon dioxide + methanol, ammonia + water, and ammonia + methanol. The diffusion coefficients have been measured using the Taylor-Aris dispersion method, and the viscosities have been measured with a falling ball viscometer at temperatures from 293 to 333 K. The ammonia mole fraction ranged from 0 to 0.312. The results have been correlated using Arrhenius-type equations and have been compared with literature data, where available. Furthermore, the measured diffusion coefficients are compared with values predicted by the modified Stokes-Einstein equation and the Wilke-Chang equation.

/pdf/diffusion-coefficients-and-viscosities-of-co2-h2o-co2-ch3oh-497u66iyed.pdf

Diffusion Coefficients and Viscosities of CO2 + H2O, CO2 + CH3OH, NH3 + H2O, and NH3 + CH3OH Liquid Mixtures

The evolution of solitary waves into three-dimensional (3-D) waves was experimentally observed on a vertically falling water film at mainly Re = 10–100. At Re greater than 40, 2-D solitary waves are very unstable in the face of spanwise perturbations of approximately 2-cm wavelengths and disintegrate into isolated horseshoe-shaped solitary waves and clusters of dimples between the horseshoes, whereas wavefront modulations are limited to low levels at Re below 40. Horseshoes of larger velocities have larger curvature heads, and extend longer oblique legs upward. Curving capillary ripples preceding each horseshoe widen their wavelengths with an increase in the wavefront inclination, showing that the ripples possess the characteristics of the shallow-water capillary waves. The horseshoes may hold vortices inside, and they have similarities in shape and size to hairpin vortices observed in laminar–turbulent transition regions of boundary layers on walls. The disintegration of waves into dimples is caused by a capillary instability similar to the one for breakup of a cylindrical soap film. This transition of wave dynamics at Re≈40 is associated with a drastic change in the mass transfer from the surface into the film.

Three‐dimensional wave dynamics on a falling film and associated mass transfer

Experiments are conducted for gas absorption in a long wetted-wall column. Liquid-side mass transfer coefficients are measured for absorption of CO2 and O2 into falling water films on the outside of a stainless steel pipe 2.72 cm OD and 183 cm absorption length. The liquid film Reynolds number ranges from 129 to 10500 which encompasses the wavy-laminar, wavy-transition and turbulent flow regimes. The experimental data are correlated by a dimensionless equation of the form kt = (ktD) (v2/g) 1/3 = a-Rep-Sc1/2. The correlation is well supported by a viscosity-damped turbulence model at the gas-liquid interface which tends to confirm that viscosity is probably the major mechanism causing eddy damping and not surface tension as proposed by Levich and Davies. The form of the above correlation also represents previous experimental work at different temperatures and for different gases quite well.

Gas absorption into wavy and turbulent falling liquid films in a wetted-wall column

This work was undertaken with a view to obtaining a general correlation for liquid phase mass transfer in wetted-wall columns. Studies were made on the absorption of pure solute gases by various solvents in columns of varied dimensions. The column dimensions, the systems employed and the ranges of the variables covered are given in Table 1.It was observed that the gas rate had no effect on the values of HL when the ReG was below 7, 000 or so, as shown in Fig. 3, and the difference obscrved by Emmert and Pigford4) between the HL valucs for absorption and those for desorption could not be noticed.The plots of HL vs. Re on logarithmic coordinates showed two distinctly different regions, A and B, as seen in Figs. 4-10. In region A, the values of HL varied with Re1.0 Sc0.5 and (μ/ρ)2/3 but were independent of z and σ. In region B, the HL values were proportional to Re0.5 Sc0.38, (μ/ρ)0.59, z0.12and σ0.15The data we e correlated with Eqs. 12 and 14, for region A and region B. The agreements of the data with the equations were good, as seen in Figs. 13, 14 and 15.

Liquid Phase Mass Transfer in Wetted-Wall Columns

The present work was undertaken for the purpose of obtaining a general correlation for the effective interfacial area in packed columns. Studies were made on the absorption of pure carbon dioxide by various solvents in a 7.0cm or 12.5cm column packed with Raschig rings and Berl saddles. The experimental apparatus employed is shown in Fig. 1, and the characteristics of packings used and the run series studied are listed in Tables 1 and 2, respectively.The values of HL are shown in Figs. 2 to 5, plotted against liquid rates on logarithmic coordinates. The effective interfacial areas for liquid phase mass transfer were determined from HL data by using Eq. (1). based on the assumption that the values of kL in a packed column are approximately equal to those for single piece of packings5), when their Reynolds numbers are the same. The fractional effective areas, aL/at, were plotted against liquid rates on logarithmic coordinates, as shown in Figs. 6, 7, 8, 10 and 11. respectively.The values of aL/at, varied with the liquid rate to the 0.455 power for all packings and systems used.The effects of the physical properties of liquid were nearly the same on the effective area and on the wetted area2); the viscosity of liquid had no significant effect on aL, as shown in Figs. 6 and 7, but the surface tension of liquid showed a considerable effect, varying with each packing as shown in Fig. 12.The present data on the effective area could be correlated by Eqs. (4) and (5) for Raschig rings and Berl saddles, respectively. Figs. 14 and 15 show the comparisons between the values obtained from these equations and the effective areas calculated, by means of Eqs. (1) or (2), from the previous data on liquid phase mass transfer rates in packed columns. Considerably good agreement is observed between the values obtained from the authors' equations and the previous data.

H. Hikita

Papers

Liquid Phase Mass Transfer in Wetted-Wall Columns

Effective Interfacial Area for Liquid Phase Mass Transfer in Packed Columns