Enhanced Recovery and Concentration of Positively Buoyant Cenospheres from Negatively Buoyant Fly Ash Particles using the Inverted Reflux Classifier

• Enhanced recovery and concentration of cenospheres from fly ash was achieved.
• The high performance was attributed to the Boycott Effect and bulk streaming.
• A cenosphere recovery of 89.9% and a throughput advantage of 54 were achieved.

The enhanced separation of valuable positively buoyant cenosphere particles from negatively buoyant fly ash particles using an Inverted Reflux Classifier (IRC) was examined. The effect of the suspension density on the recovery and concentration was examined in the IRC by operating at different feed pulp densities ranging from 10 wt% to 46 wt%. Using a sufficiently high fly ash concentration, it was hypothesised that a powerful bulk streaming phenomenon develops (Batchelor and Van Rensburg, 1986) within the inclined channels, driving the segregation between the positively and negatively buoyant species. With the feed flow rate, fluidization rate, and flow split to overflow and underflow fixed, the recovery of the cenospheres increased from 61.7% (at 10.1% solids) through to an optimum recovery of 89.9% (at 38.1% solids), before declining rapidly to a recovery of 60.2% (at 46.4% solids). The performance at the optimum of 38.1% pulp density was remarkable, with 3.1 t/(m2 h) solids throughput, a single-stage cenosphere recovery of 89.9% and upgrade of 58.6, and throughput advantage over a conventional fluidized bed of 54. Detailed analysis indicated that the inclined channels produced an underlying throughput advantage of 18, with a further factor of 3 attributed to the bulk streaming phenomenon. The separations were also assessed in terms of the partitioning of the cenospheres between the overflow and underflow exit streams, with the sharpest size classification evident at the optimum feed pulp density, with the d25 = 31.5 μm, d50 = 36.5 μm, and d75 = 50.0 μm. The separation was then investigated using different feed flow rates, providing the basis needed for ensuring optimum performance in future pilot scale investigation of this novel technology.