Mixing biomasses has a good potential to solve operational problems in thermochemical valorization processes related to ash behavior. Chemical reactions within the ash of the blend, and not only a mixing effect without reaction, need to take place to form new solid phases in the mixture at the expense of getting rid of the problematic liquid one to decrease the slagging tendency. The present work focuses on assessing the presence of a chemical reaction in comparison with a simple mixing effect within the ash of mixtures of wheat straw - oak bark biomass and ash in two laboratory setups and in a fixed and in a moving bed pilot combustion reactors. The operating conditions were varied to study their effects on the reactivity of the ash within the mixtures. This aimed to optimize the ash reactivity and to assess the capabilities of the laboratory test to predict the ash behavior in pilot-scale reactors.Mixing without reaction effect was more evident when mixtures of biomass were used on both laboratory and pilot scales. In this case, the simple mixing rule was able to simulate the general ash behavior of biomass mixture to a certain extent that exact prediction was always limited by the presence of certain chemical reactions. However, chemical reactions effect was dominant when mixtures of ash of biomass were applied as feedstock. Solid crystalline phases K2Ca2Si2O7 and K2Ca6Si4O15 were the direct end products of these chemical reactions. Their relative proportion was inversely proportional to the problematic amorphous phase in the final ash. Hence, optimizing the blend proportion to increase their concentrations has the potential to solve slagging problematic. These two compounds were mostly stable at 1000 ℃, at which equilibrium was reached after 6 h. However, they existed at a lower proportion at 850℃ and disappeared at 1200℃. The developed laboratory pellet test was also able to predict very efficiently the pilot ash behavior of both individual biomasses and their mixture in terms of crystalline and amorphous composition and proportion along with agglomeration distribution.
Read full abstract