Bed expansion and gas holdup characteristics of bubble–assisted fluidization of liquid–particle suspensions in a HydroFloat cell

Abstract

The HydroFloat cells made by Eriez are designed for floating coarse particles using fluidization involving bubble–liquid–particle suspensions. The operational optimization and control of these cells can be a challenging task and require various correlations that are difficult to establish by theory. Here, experiments were performed to study the effect of SDS (sodium dodecyl sulfate) on the liquid–particle and bubble–liquid–particle fluidization processes in a laboratory-scale HydroFloat cell. Air was supplied to a perforated circular tube at the bottom to generate the bubbles while irregular coarse quartz particles were fed to the top by a vibrating plate as the solid phase. In the liquid–particle fluidization, the bed expansion (BE) increased with increasing superficial liquid velocity and agreed with the modified Richardson-Zaki equation (within 4% relative error). BE decreased to 12.5% in the presence of SDS at low concentrations. In the bubble–liquid–particle fluidization, BE was reduced by 8.46 – 2.89% and 15.1% with increasing superficial gas velocity and decreasing superficial liquid velocity in the absence of SDS, respectively. In the presence of SDS, BE increased with increasing superficial gas and liquid velocities. High SDS concentrations generated expanded beds with higher heights. Low SDS concentration (0.2 mM) increased the gas holdups by an average of 20% with increasing superficial gas velocity from 1.1 to 3.4 mm/s. Both BE and gas holdup were compared with the available empirical correlations. The estimated BE showed smaller mean absolute relative errors (MAREs) in the absence of SDS than the presence of SDS, while gas holdups yielded large MAREs. The particle sphericity and bubble Reynolds number were incorporated into the Begovich-Watson and Ramesh-Murugesan equations to estimate the BE and gas holdup, respectively. The modified equations significantly improved the predictions. The empirical correlations developed and investigated here would be useful for the operational optimization of the fluidized bed HydroFloat cells.