This paper describes a study of the possibility of using point efficiencies, measured in a small laboratory column, in the scale-up of distillation efficiencies. The conditions on a large rectangular tray were closely duplicated in a small glass Oldershaw column with one sieve tray having similar hole size and percent free area. Point efficiencies on the large rectangular sieve tray were deduced from the composition profiles and compared with the point efficiencies measured in the small column. A simple relationship between the two columns was found in terms of the froth heights and liquid holdups. The results demonstrate that the slightly lower point efficiencies measured in the Oldershaw column could be used in conjunction with a mathematical mass transfer model based on the eddy diffusion concept, to relate the point efficiencies to the tray efficiences, resulting in a conservative design provided either that allowance is made for other factors, such as nonuniform flow patterns and stagnant zones, or that these are eliminated. A wide range of composition was covered using the two systems MeOH-nPrOH and EtOH-nPrOH operating in the froth region. M. M. Dribika, M. W. BiddulphUniversity of Nottingham Department of Chemical Engineering University Park, Nottingham, England SCOPEThe prediction of the performance of a large-scale distillation tray column is difficult, particularly if a new system is being processed. It is a problem that has attracted a considerable amount of attention over many years. However, the prediction of tray efficiency is still a very uncertain part of the design unless operating data from a similar size column distilling a similar system are available. The problem is that both mass transfer characteristics and tray hydraulic considerations are involved, and so the tray efficiency is a function of system and tray size. This paper is concerned with the mass transfer part of the problem. It describes a study of the relationship betwen point efficiencies, measured in a small glass Oldershaw Column, and the efficiencies inferred at points along a rectangular tray with a liquid flow path length of about 1 m. The tray used is a sieve type with small hole sizes typical of those used in the cryogenic distillation of air. The hole sizes and percent free area in the two trays being compared are similar. CONCLUSIONS AND SIGNIFICANCEThe relationship between point efficiencies measured in a small laboratory column and the point efficiencies that actually exist on a large rectangular tray were studied using two systems, MeOH-nPrOH and EtOH-nPrOH, and covering a wide range of composition. The composition profiles across the tray were measured and used to infer the values of point efficiencies using a mathematical mass transfer model based on the eddy diffusion concept.Similar conditions to those on the large rectangular 1864November 1986 sieve tray were used in a small Oldershaw column with one sieve tray and external downcomers. The results indicate that the point efficiencies obtained on the large rectangu...

No abstract

Vapor-Squid equilibrium data for the ternary system MeOH-EtOH-/-PrOH were determined at atmospheric pressure In an Ellis and Froome equilibrium still. Activity coefficients were calculated by assuming nonidealities In the vapor and liquid phases. The results showed the existence of slight nonkfeaWles among alcohols. The thermodynamic consistency of the data was considered by the McDermott and Ellis test. The ternary results were predicted satisfactorily by the use of the Wilson model.

No abstract

A study of the efficiencies of two plate distillation columns is reported. One column had three rectangular trays with a flow path length of about 1 m, the other was a modified glass Oldershaw column with one tray of 38 mm dia. Tray efficiencies were measured in both columns and point efficiencies were deduced for the rectangular column.The systems studied were ternary and quaternary mixtures of alcohols and water. The results showed that for thermodynamically nonideal systems significant differences existed among the individual component point efficiencies. In the case of an ideal system the component point efficiencies were effectively equal, but tray efficiencies were different due to differing volatilities and composition profiles across the large trays.This study illustrates the need to allow for the above effects in largecolumn designs for multicomponent systems. The possibility of using laboratory-scale point efficiencies in these designs is explored.