Info

For corrugated-sheet structured packings at less than 85 percent of flood only a portion of the surface is wetted, and this portion is a function of the surface Reynolds number. For design purposes, Bravo et al. recommended estimating the wetted fraction from p = 0.50 + 0.58 (US/iîs.F,)

Bravo et al. (124) demonstrate that their correlation gives good predictions of published data for the Koch-Sulzer BX (wire-mesh) packings. For corrugated-sheet structured packings, they state that their effective area equation [Eq. (9.32)] is an oversimplification (25). Surface characteristics influence interfacial area and its rate of change with column loadings (125,1256), and these are not accounted for by the simple Eq. (9.336). Interpolating data in Ref. 1256 can give closer estimates for p. The interfacial area prediction is currently undergoing further study (125,125a,1256).

Others. For random packings, correlations by Onda et al. (123) and by Bolles and Fair (55,96) were mentioned earlier. Predictions from the Bravo and Fair correlation were shown to be better than from those correlations. A new recent random packing correlation by Bornhutter and Mersmann (126) is the first to account for mass transfer in drops as well as in the liquid film. The authors show experimentally that with large (2 to 3 in) packings, drops often provide more mass transfer surface than the liquid film. A new efficiency correlation that applies both for random and structured packings has just been reported by Billet and Schultes (126a). The correlation is based on the two-film model. Although it does not directly account for drop mass transfer, it does so indirectly by accounting for surface tension gradients (these gradients influence drop formation). Initial testing of the correlation against a wide data bank appears very promising (126a). An efficiency correlation for Sulzer's BX and Mellapak structured packings was developed by Spiegel and Meier (21). The correlation is based on a model similar to that of Bravo and Fair (25,124), except that it neglects the liquid mass transfer coefficient. Unfortunately, Spiegel and Meier did not fully define some of their constants, making their correlation difficult to apply.

9.1.5 HETP prediction—rules of thumb

Because there are only few variables that significantly affect random-packing HETP (Sec. 9.1.3), and due to the unreliability of even the best mass transfer model, rules of thumb for HETP successfully compete with mass transfer models. It has been the author's experience that rules of thumb for HETP prediction are more accurate and more reliable than mass transfer models. A similar conclusion was reached by Porter and Jenkins (127).

Table 9.2 lists the rules of thumb available in the published literature. Most are based on second- and third-generation random packings and should not be applied to first-generation packings. The majority of rules are in very close agreement with each other. Porter

TABUE 9.2 Published Rutfts oi Thumb tor Random-Packing HETP
0 0

Post a comment