A Gas hold‐up model for slurry bubble columns

Abstract

in Wiley InterScience (www.interscience.wiley.com).A phenomenological model of the gas bubble hold-up in slurry bubble columns is presented. The slurry moves upward in the core of the column and downward in the annulus, resembling the riser and downer of an air-lift loop reactor, respectively. The model virtually divides the column in a riser and a downer. The cross-sectional areas of the riser and the downer are experimentally found to be approximately equal. A macroscopic momentum balance over the riser and downer gives the liquid circulation velocity. The gas bubble hold-up is modeled for two size classes: small and large. In the homogeneous regime only small gas bubbles are present. Above the regime transition velocity the churn-turbulent regime exists, characterized by the presence of small and large gas bubbles. High speed video imaging shows that the gas flow up to the regime transition velocity generates small gas bubbles and that the gas flow in excess of the regime transition velocity generates large gas bubbles. Small gas bubbles are present in the riser and in the downer and have a constant slip velocity. The Wallis driftflux model describes the small gas bubble hold-up in the riser and in the downer. Large gas bubbles are only present in the riser. The rise velocity of the large gas bubbles is the sum of the liquid velocity and the slip velocity according to the Davies-Taylor equation. The large gas bubble hold-up is obtained from the volume of gas flowing in large gas bubbles and the rise velocity of the large gas bubbles. For a given column diameter, column height, regime transition velocity, and physical properties of a gasliquid system, the model predicts the total gas hold-up, the hold-up of the small and the large gas bubbles, and the average superficial liquid velocity in the riser and in the downer. Two parameters in the model are determined from experiments in a 29-cm ID bubble column: the ratio of the small gas bubble hold-up in the downer to that in the riser, and the energy dissipation at the top and bottom of the column. The model describes the gas hold-up and liquid velocity data from literature with these parameter values within an error of 6%. The model shows that there is no effect of scale on the gas hold-up for columns larger than 15 cm in diameter.

Gas Holdup, Axial Dispersion, and Mass Transfer Studies in Bubble Columns

“…Gas holdup data obtained in the empty bubble column can be correlated by using one of the three types of correlations proposed in the literature. The first type correlates the gas holdup directly to the gas velocity as a homographic function. ,,,, The second type correlates the gas holdup as a function of the slip velocity between the gas and the liquid phase. , The third type is the gas holdup correlated with dimensionless numbers. , In this work, the gas holdup correlation is expressed as a function of Froude and Galileo dimensionless numbers. Alternatively, the Reynolds number can be used to describe the liquid phase.…”

“…Note that at gas velocities above 5 cm/s the bubble column is in the coalescing regime where large gas bubbles are formed by coalescence of small gas bubbles, and this is generally considered as representative for the churn-turbulent regime. Therefore, the correlations could be extrapolated to higher gas velocities . In total, 30 experiments consisting of various combinations of liquid flow, gas flow, and liquid viscosity are performed for each structured packing, each type of partition tray, and the empty bubble column.…”

Backmixing Characterization of a Bubble Column with Short Venturi Throats by Multipoint Internal Tracer Injections

“…Although they can monitor the variation of tracer concentrations for the bubble column with SVT #1, the mean liquid residence time and the variance of the response curves of these two probes are smaller than that for the empty column, especially for the dimensionless variance, as listed in Table . Since the strong backmixing of the liquid phase in a bubble column is originated from the liquid circulation due to the upward flow in the column central region and the downward flow in the annular wall region. − Therefore, the above phenomena mean that the liquid backmixing of the column is substantially reduced by the introduction of SVT #2 through preventing the downward liquid circulation flow. In addition, the probes C and D respectively located at the middle and top stages of the column can monitor the tracer response signals for the column with and without SVTs.…”

Determination of Liquid Multiscale Circulation Structure in a Bubble Column by Tracing the Liquid Flowing Trajectory

“…In order to eliminate small-scale circulation loops from the large ones, draft tubes with different lengths were installed coaxially into the column. Since the axial liquid velocity profile has an inversion point at a dimensionless radius of 0.7, − we chose this ratio, which is 35 cm, as the diameter of the draft tubes to minimize the impact of the internal on the liquid axial flow pattern in the column. The validity of our choice on draft tube diameter has been verified by the literature report.…”