Bed agglomeration characteristics and mechanisms during gasification and combustion of biomass fuels

Controlled agglomeration tests, using six representative biomass fuels (bark, Lucerne, reed canary grass, bagasse, olive flesh, and cane trash) were performed in a bench-scale fluidized bed (5 kW) during both gasification and combustion conditions. The resulting bed materials were analyzed using scanning electron microscopy/energy-dispersive spectroscopy (SEM/EDS), and chemical equilibrium calculations were performed to facilitate the interpretation of the experimental findings. Layers of fuel-ash-derived compounds were built up on the bed particles during processing of all studied fuels. The accumulated material was determined to consist of two layers: (i) an inner thicker and more homogeneous layer that consisted of mainly K-Ca-silicates and (ii) a thinner, particle-rich outer layer. For all fuels except Lucerne, no major differences in bed agglomeration tendencies or bed particle layer characteristics could be detected between gasification and combustion, which suggested no major difference in layer formation processes or bed agglomeration mechanisms between the two different operational modes. Thus, initial silicate layer formation followed by subsequent viscous flow sintering and agglomeration was identified as the bed agglomeration process in all cases except during the combustion of Lucerne. For combustion of the relatively sulfur-rich Lucerne fuel, the agglomeration was induced by a salt melt where the bed material particles were directly glued together by a separate ash-particle-derived melt.