A discrete element cohesive particle collision model for the prediction of ash-induced agglomeration

The utilisation of energy crops or residues in fluidised beds shows large potentials as a sustainable source of renewable energy. The compositional range of their ashes, however, can lead to the formation of large agglomerates, causing operational problems or defluidization. In this paper a novel cohesive contact model was developed that extends previously published wet collision models to account for the transition of highly viscous liquids to amorphous solids as the dominant agglomeration criteria. Previously published thermochemical equilibrium calculations using FactSage 6.2 were incorporated for the determination of required melt fractions and material properties. The contact model evaluates viscous dissipation within the molten ash coatings, capillary forces in liquid bridges and elastic forces of the solid particle cores. Finally, two-particle-collisions were simulated in a discretized VBA-script incorporating this contact model. A primary validation was performed with controlled agglomeration tests (350 runs, 83 fuels, 7 reactors). The presented model is able to reproduce the experimental parameter influences with good accuracy and shows predictive performance on par with previously published models (prediction error of +/- 30%).