Experimental and modeling evaluation of packed column flotation with reflux

Flotation is a relatively effective and economic method for selectively separating fine particles from liquid. Experimental goals were to examine the packed flotation column (PFC) behavior as a function of reflux ratio, column length, coal type, etc. while modeling goals were to predict the solid concentrations and liquid flowrate of exit streams and estimate the rate constants of separation of each solid phase in a two component fine particle system in various sections of the column. Experiments using both an Illinois No. 5 and No. 6 coal were carried out to examine the effect of reflux ratio and other operating variables on PFC performance. A phenomenological, steady state (rate based) model was further developed, based on the algorithm suggested by Idlas et al. (IECR, 29, 6, pp. 943-949, 1990), where rate constants were assumed the only major model parameters needed to describe reflux column performance. Model extensions particularly analyzing important hydrodynamic effects in flotation column kinetics but requiring added parameterization were also discussed.; While the ash grade was somewhat improved as the wash water was increased in non-reflux flotation, the ash grade was improved with no decrease of coal recovery at {dollar}R approx{dollar} 0.6 in reflux flotation. When PFC process performance was compared with other processes, PFC and its predecessor NUCOL (Northwestern Univ. Flotation Column) produced grade/recovery that may approach release analysis (optimum grade for the given liberation).; Model testing was performed to compare model forms including one or two solid species (coal and/or ash) and one or two mechanisms (attachment and/or detachment) in 1, 2 or 3 sections of the column. Comparison with experiment for two parameter model forms suppressing detachment of coal and ash was satisfactory to predict each solid concentration and liquid flowrate of product (three dependent variables) although minor (ash) component was not as well predicted overall.; An estimation of macroscopic kinetic constants for PFC using microscopic variables was conducted following the approach of Li et al. (IECR, 29, 6, pp. 955-967, 1990) showing reasonable consistency to measured kinetic constants. Sensitivity of recovery and grade to a change of model parameters and selected input data was analyzed to determine what resources are needed to utilize and improve the model.