Hydrodynam cs and Mass Transfer in NonNewtonian Solutions in a Bubble ColumnUntil now the oxygen transfer in viscous nowNewtonian solutions has been studied only in bubble columns of about 0.14-m diameter. Recently Godbole et al. (1982) reported much smaller gas holdups in Carboxy Methyl Cellulose solutions (CMC) for a large-diameter column. Therefore, the gas holdups, volumetric mass transfer coefficients, and specific gas-liquid interfacial areas are measured in CMC solutions using a bubble column of diameter 0.305 m and height 3.4 m. The transition from churn-turbulent to slug flalw regime occured at higher viscosities and the gas holdups and volumetric mass transfer coefficients were lower in both flow regimes than reported for smaller column diameters. Empirical correlations are presented for the gas holdup, volumetric mass transfer coefficient, and specific gas-liquid interfacial area which would be suitable for the design of fermentors. SCOPEBubble column reactors are becoming increasingly popular in the biotechnological and pharmaceutical industry. The rheological behavior of microbiological cultures in a fermentation tower can be fairly well simulated by ithe solutions of carboxymethyl cellulose (CMC). All the literature data for the hydrodynamics and mass transfer in CMC solutions were taken in columns of up to 0.14-m diameter. Recently for viscous CMC solutions, Godbole et al. (1982) reported a strong decrease in the gas holdup with an increase in the column diameter. The strong dependency of the gas holdup on column diameter suggests a similar dependency for the volumetric mass transfer coefficient, i.e., the previous investigations in columns of 0.14-m diameter are insufficient for scale-up purposes. The correlation of Nakanoh and Yoshida (1980) even suggests an increase in volumetric mass transfer coefficient with increasing column diameter. Therefore, in this work oxygen mass transfer in CMC solution was studied in a 0.305-m diameter column.Volumetric mass transfer coefficients are measured by the dynamic method. Specific interfacial areas are determined by the sulfite oxidation technique used by . The kinetics of the cobalt catalyzed reaction is not affected by addition of CMC (Wesselingh and van? Hoog, 1970);Onken and Schalk, 1978;Poggemann, 1982) and the deviation of the chemically effective interfacial area from the geometrical one is small because of the small gas-phase conversion (Schumpe and Deckwer, 1980a,b). The gas holdups are measured using a hydrostatic head technique and fractional gas holdups are measured using the dynamic gas disengagement technique (Sriram and Mann, 1976). To study the influence of the added salt, volumetric mass transfer coefficients are determined in CMC/sodium sulfate solutions and compared with the kLa values obtained in pure CMC solutions. Fermentation media might be more complex than the model media used and hence the effect of surfactant (Triton X-114) on the hydrodynamics and mass transfer in CMC solutions is studied. The gas holdup, volumetric mass transfe...

No abstract

Gas holdup and axial dispersion coefficient data for dilute aqueous alcohol solutions and two different diameter columns at larger gas and liquid velocities compared to those of Schiigerl et al. (1977) are presented. Data for cocurrent and batch systems are qualitatively explained using Zuber and Findlay's theory (1965) and bubble structure, and quantified further using a dynamic gas disengagement technique. Unified empirical correlations for the gas holdup and axial dispersion coefficients are presented.Applications of bubble columns as bioreactors and for the process of coal liquefaction are relatively recent. Characteristics of the liquid-phase media in these two reactors can be fairly well represented by dilute alcohol solutions. The only work reported on this subject is by Schugerl et al. (1977) for relatively lower gas and liquid throughputs. They propose an empirical correlation for gas holdup involving the bubble diameter which is more difficult to estimate than the gas holdup. They do not propose any correlation for determining the axial dispersion coefficient.The experimental studies were carried out in two different diameter columns using cocurrent and batch systems. Five aliphatic alcohols (methanol, ethanol, n-propanol, i-propanol, and butanol) were investigated with concentration varying from 0.5 wt % to 2.4 wt %. The gas holdup was measured using a hydrostatic head technique. The axial dispersion coefficient was measured using heat as a tracer, and was based on the dispersion model. A dynamic gas disengagement method is applied to quantify the bubble-size distribution and bubble-rise velocities. This method involves a measurement of a decline in liquid height as a function of time, after gas flow is suddenly stopped. The experimental data are explained both qualitatively and quantitatively. Empirical correlations for gas holdup and the axial dispersion coefficient, applicable over a wide range of gas and liquid throughputs are presented.This infers that the dispersion coefficient and the gas holdup are interrelated.

No abstract

The viscosity of hydrocarbon mixtures, whether in the gas or liquid phase, is a function of pressure, temperature, and phase composition. This paper presents methods for the prediction of the viscosity of the gas or less dense fluid phase over the practical range of pressure, temperature, and phase compositions encountered in surface and subsurface petroleum production operations. The correlation necessary to predict the effect of pressure on viscosities is presented in Part I. Serious discrepancies in high pressure gas viscosity data in the literature are discussed.The application of the correlation to predict absolute v,iscosities is discussed in Part II. Auxiliary correlations are presented to enable calculations of viscosities from a knowledge of the pressure, temperature, and gravity of the gas phase.