Absorption column design mass transfer coefficients

In 1966, in a paper that now is considered a classic, Danckwerts and Gillham [Tmns. Inst. Chem. Eng., 44, T42 (1966)] showed that data taken in a small stirred-ceU laboratoiy apparatus could be used in the design of a packed-tower absorber when chemical reactions are involved. They showed that if the packed-tower mass-transfer coefficient in the absence of reaction (/cf) can be reproduced in the laboratory unit, then the rate of absorption in the l oratoiy apparatus will respond to chemical reactions in the same way as in the packed column even though the means of agitating the hquid in the two systems might be quite different. 

Packed columns are often used for distillation, liquid-liquid extraction, and humidification as well as for gas absorption. The design can be based on overall transfer coefficients or on the number of transfer units and the height of a transfer unit. For distillation or humidification, where the gas phase is continuous and the liquid flows in rivulets over the packing, the mass-transfer coefficients and flooding characteristics are similar to those for gas absorption, and the same generalized correlations would apply. 

Chapters 7 and 8 present models and data for mass transfer and reaction in gas-liquid and gas-liquid-solid systems. Many diagrams are used to illustrate the concentration profiles for gas absorption plus reaction and to explain the controlling steps for different cases. Published correlations for mass transfer in bubble columns and stirred tanks are reviewed, with recommendations for design or interpretation of laboratory results. The data for slurry reactors and trickle-bed reactors are also reviewed and shown to fit relatively simple models. However, scaleup can be a problem because of changes in gas velocity and uncertainty in the mass transfer coefficients. The advantages of a scaledown approach are discussed. 

The values of interfacial area and of overall mass transfer coefficient increase with decreasing distance S between the spray nozzle and gas inlet, whatever the nozzle type, column dimensions and flowrates. Indeed the spray provides a large interfacial area in the vicinity of the nozzle, where there is intensive circulation. Then a decreases quickly away from the nozzle, as a result of both coalescence of droplets and collection of liquid on the column walls. kQa and a are approximately proportional to for absorption and desorption processes, which shows that kQ is practically independent of the column height. Moreover Mehta and Sharma indicate that a is unaffected by ionic strength and viscosity but may decrease about 20 % when solids are generated by the reaction of gas with the liquid. Thus the following correlations pay be used for design (112)... 

Although transport properties are, strictly speaking, not part of chemical thermodynamics, the most common correlation and estimation methods are briefly introduced here, as they are needed for process simulation steps that are determined by mass transfer (e.g., absorption columns) and for the design of the equipment (e.g., heat transfer). It was decided to present the methods for calculating diffusion coefficients here as well, although they refer to mixtures and not to pure compounds. 

The design of adsorption columns uses a breakthrough curve measured at small scale to estimate performance at a large scale. The experiment is essential because the effects of the isotherm and the dispersion cannot normally be predicted. In this regard, adsorption is less well understood than staged distillation or dilute absorption. For staged distillation, the key information is thermodynamic, summarized as a y-x diagram. For simple absorption, the information is both thermodynamic (a Henry s law coefficient) and kinetic (a mass transfer correlation), but the dilute analysis leads to simple, reliable answers. Adsorption is more complex and requires that first experiment. 

---