(109e) Improved Improved Mass Transfer Correlations for Random and Structured Packings

The reliability of rate-based calculations for packed columns operated countercurrently is known to be limited by the accuracy of the correlations used to estimate the mass transfer coefficients and the interfacial area (?vandová, et al., 2008; Adler, 2000). Typically the mass transfer coefficient matrices for multicomponent systems are based upon generalizations of mass transfer coefficient expressions developed from data on binary systems. Clearly, the successful analysis of experimental data for binary systems is crucial to the acceptance of rate-based modeling as an improvement over the traditional equilibrium stage model for multicomponent systems.

This paper explores the development and limitations of today's public domain mass transfer correlations for packed columns. First, we shall consider the types of data generally available to the modeler. Next, we demonstrate that traditional attempts to decompose binary HETP data into fits for a vapor-side and a liquid-side mass transfer coefficient equation along with an equation for the mass transfer area are, at best, accurate to within a multiplicative constant. This finding implies that other types of experiments are needed to unambiguously define the vapor-side and liquid-side contributions to the total resistance. We also demonstrate that it is sometimes not possible to find a unique set of fitting parameters from efficiency data alone; either theoretical arguments are needed to set some parameters or, once again, some other type of experiment is needed to resolve the impasse. Our studies further reveal that physical property models do play a substantial part in the success or failure of any particular mass transfer correlation and that care is therefore warranted in their selection.

Finally, we present new mass transfer correlations for Pall ring style metal random packings, IMTP style metal random packings, and Mellapak ?Y? style metal structured packings. The development of these correlations will be discussed, their limitations examined, and their performance relative to several public domain mass transfer correlations presented.

Z. ?vandová, et al., Chem. Eng. J., 140 (2008) 381?390. Adler, S., et al., Vision 2020: 2000 Separations Roadmap, AIChE Center for Waste Reductions and U.S. DOE, 2000.