Effective interfacial area in agitated liquid–liquid continuous reactors

Abstract

A chemical method has been used to quantify the e ective interfacial area in a ba ed continuous stirred liquid-liquid reactor. Two and four straight paddle impellers were used in the experimental runs, at 34• C, with hold-up fractions of dispersed organic phase between 0.061 and 0.166 and stirring speed ranging from 360 to 1500 rpm. In uence of the residence time on the formation of the interfacial area generated in this system was not registered; however, di erences were reported between continuous and batch mode operations.The interfacial area was correlated to hold-up fraction and Weber number by a new empirical model proposed in this work. This model allows to use only one equation to calculate the interfacial area in this continuous stirred reactor in the wide range of operating conditions tested (490 ¡ We ¡ 9600), which include di erent ow regimes. This is a relevant contribution as previous studies in this ÿeld only contemplate turbulent ow. In the transitional regime the mean drop size diameter decreases abruptly with Weber number, but this pattern changes in the higher range of Weber where the dispersed drops become smaller very smoothly. This pattern does not depend on the agitator used or hold-up fraction. The mean drop size diameter is smaller for the four paddle impeller and increases with hold-up fraction. The model developed may be applicable to dispersions in aromatic nitration reactors, improving its operation and design. ?

Representative drop sizes and drop size distributions in A/O dispersions in continuous flow stirred tank

“…Majority of them (Vermeulen et al, 1955;Rodger et al, 1956;Weinstein and Treybal, 1973;Mlynek and Resnick, 1972;Fernandes and Sharma, 1967;Brown and Pitt, 1974;McManamey, 1978;Calabrese et al, 1986a,b;Wang and Calabrese 1986;Nishikaw et al, 1987a,b;Laso et al, 1987;Chatzi et al, 1991;Zhou and Kresta, 1998;Pacek et al, 1999;Ruiz et al, 2002;Desnoyer et al, 2003;Giapos et al, 2005;Sechremeli et al, 2006) have been done on batch vessels. Only few of them have been done on continuous flow stirred tanks (Wienstein and Treybal, 1973;Fernandes and Sharma, 1967;Quadros and Baptista, 2003). In most of the cases, the experimental data have been correlated using the functional form developed by Hinze (1955) and modified by subsequent researchers (Shinnar and Church, 1960;Doulah, 1975).…”

“…In most of the cases, the experimental data have been correlated using the functional form developed by Hinze (1955) and modified by subsequent researchers (Shinnar and Church, 1960;Doulah, 1975). In some cases, altogether different functional forms are reported (Wienstein and Treybal, 1973;Quadros and Baptista, 2003). Since model of Hinze (1955) has been used in majority of cases, it is briefly discussed below for sake of completeness.…”

“…The first equation is for NiCl 2 as the dispersed phase and second for HCl as the dispersed phase. Quadros and Baptista (2003) Experiments to determine interfacial area in continuous flow stirred tank with di-isobutylene diluted with benzene as the dispersed and sulfuric acid as the aqueous phase. Impellers used were straight blade paddles with two or four number of blades.…”

Continuous adiabatic industrial benzene nitration with mixed acid at a pilot plant scale

“…Previous work carried out in this pilot plant, for a reacting system identical to the one used in this study, led to [33]:…”

Different Modeling Approaches for a Heterogeneous Liquid−Liquid Reaction Process

“…The analysis is based on the variation inflation factor VIF, eq 8, and the procedure consists of eliminating, at each step, one of the variables that presents a higher VIF or a known relationship with other variables. In this process, the stirring speed and the holdup fraction have a major influence on the effective interfacial area; 28 thus, only the effective interfacial area was selected to enter the model, to keep this modeling approach comparable with the mechanistic ones, which also assume the explicit knowledge of this parameter. If this model, based on the effective interfacial area, presents better predictive performance versus the mechanistic alternatives, then one also has justified the expectations that a model involving the manipulated variables (stirring speed and the holdup fraction) would also perform better.…”