Page 37 - Gas Purification 5E

Introduction 27

location of the reaction zone (and the value of the absorption coefficient) will depend pri-
marily upon the diffusion rate of reactants and reaction products to and from the reaction
zone, the concentration of solute at the interface, and the concentration of the reactant in the
body of the liquid. However, since the distance that the solute must diffuse into the liquid is
extremely small compared to the distance that it would have to travel for simple physical
absorption, a high liquid-film coefficient is observed, and, in many cases, the gas-film resis-
tance becomes the controlling factor.
Since the effect of chemical reaction is to increase the liquid film coefficient, k, over the
value it would have in the absence of chemical reaction, kL a common approach is to utilize
the ratio, kLkL in correlations. This ratio is called the enhancement factor. Both kL and k:
are affected by the fluid mechanics, but fortunately their ratio, E, has been found to be rela-
tively independent of these factors. It is primarily a function of concentrations, reaction rates,
and diffusivities in the liquid phase.
The theoretical evaluation of absorption followed by liquid-phase chemical reaction has
received a great deal of attention although the results are not yet routinely useful for design
purposes. Early studies of serveral reaction types were made by Hatta (1929, 1932) and Van
Krevelen and Hoftijzer (1948). This work has been expanded by more recent investigators to
cover reversible and irreversible reactions, various reaction orders, and reaction rates from
very slow to instantanmus. Important contributions have been made by Perry and pigford
(1953), Brian et al. (1961), Gilliland et al. (1958), Brian (1964), Danckwerts and Gillham
(1966), Decoursey (1974), Matheron and Sandal1 (1978), and Olander (1960). The applica-
tion of the theory to specific gas purification cases has been described by Joshi et al. (1981)
(absorption of C02 in hot potassium carbonate solution), and by Ouwerkerk (1978) (selec-
tive absorption of HIS in the presence of C02 into amine solutions).
Stripping in the presence of chemical reaction has been considered by Astarita and Savage
(1980), Savage et al. (1980), and Weiland et al. (1982). In general, it is concluded that the
same mathematical procedures may be used for stripping as for absorption; however, the
results may be quite different because of the different ranges of parameters involved. It is
always necessary to consider reaction reversibility in the calculation of stripping with chemi-
cal reaction.
It is beyond the scope of this intductory discussion to present even a listing of the numer-
ous mathematical equations developed to correlate the effects of chemical reactions on mass
transfer. Detailed equations and examples of their application are presented in comprehensive
books on the subject by Asta~ita (1967), Danckwerts (1970), and Astarita et al. (1983).

Column Diameter

Packed Columns

The diameter of packed columns filled with randomly dumped packings is usually estab-
lished on the basis of flooding correlations such as those developed by Sherwood et al. (1938),
Elgin and Weiss (1939), Lobo et al. (1945), Ekkert (197OA, 1975), Kister and Gill (1991),
Robbins (1991), and Leva (1992). According to Fair (1990), the currently used correlation for
packed tower pressure drop prediction--commonly called the Generalized Pressure Drop Cor-
relation (GPDCjshould be attributed to Leva (1954). Other investigators have developed
minor improvements. A generalized carelation for estimating pressure drop in structured pack-
ings is presented by Bravo et al. (1986). The Eckert (1975) version, which is the basis for the
approach given by Strigle (19%), is widely used and is therefore included here.