Impact of solids loading and mixture composition on the classification efficiency of a novel cross-flow fluidized bed classifier

Calcium looping with indirect heat transfer between the main combustor and the calciner is a CO2 capture technology with low energy penalty. One way of transferring the heat, is heating of inert particles by hot combustion gases in the combustor followed by heating and calcination of sorbent (CaCO3) particles by direct contact with the inert heat transfer particles in the calciner. In this process, it is crucial to separate the sorbent and the heated inert solids at the end of the calcination process. Hence, a highly efficient classification device operating at temperatures above 900 °C is required. In this work, a downscaled cold-flow version of a cross-flow fluidized bed classifier was studied as novel classification method. The classification efficiency and the pressure drop in the classifier were measured when varying solids loading (mass flow of solid feed/mass flow of fluidization air) and mixture composition using ambient air as the fluidization medium. Different mixtures of zirconia (Dmean = 69 μm; ρ = 3800 kg/m3) and steel (Dmean = 290 μm; ρ = 7800 kg/m3) were used. The solids loading was varied between 2 and 99 kg/kg, either by varying the air supply rate and/or the solids feed rate. For a mixture of 28 wt% zirconia and 72 wt% steel, a zirconia bottom exit loss of 6.6% was achieved with a solids-loading of 4.6 kg/kg. Although higher air flow rates reduced the loss of lighter particles (zirconia) through the bottom exit, loss of heavier particles (steel) through the top exit was then increased. Reducing the content of zirconia from 28 to 14 wt% reduced the bottom exit loss to 5%. [Display omitted] •A particle classification method for high temperature CO2 capture is presented.•Such a classification is possible with a cross-flow fluidized bed classifier.•Carry light particles out via top exit is reduced with increased solids loading.•Reducing the lighter solids in the feed enhances the classification efficiency.