Biomass Fast Pyrolysis Fluidized Bed Reactor: Modelling and Experimental Validation

• Main Author: Matta, Johnny
• Other Authors: Mehrani, Poupak
• Language: en
• Published: Université d'Ottawa / University of Ottawa 2016
• Subjects: Biomass; Fast Pyrolysis; Thermochemical Conversion; Fluidized Bed
• Online Access: http://hdl.handle.net/10393/35516; http://dx.doi.org/10.20381/ruor-474

Of the many thermochemical conversion pathways for utilizing biomass as a renewable energy source, fast pyrolysis is a promising method for converting and upgrading carbonaceous feedstocks into a range of liquid fuels for use in heat, electricity and transportation applications. Experimental trials have been carried out to assess the impact of operational parameters on process yields. However, dealing with larger-scale experimental systems comes at the expense of lengthy and resource-intensive experiments. Luckily, the advances in computing technology and numerical algorithm solvers have allowed reactor modelling to be an attractive opportunity for reactor design, optimization and experimental data interpretation in a cost-effective fashion. In this work, a fluidized bed reactor model for biomass fast pyrolysis was developed and applied to the Bell’s Corners Complex (BCC) fluidized bed fast pyrolysis unit located at NRCan CanmetENERGY (Ottawa, Canada) for testing and validation. The model was programmed using the Microsoft Visual Basic for Applications software with the motivation of facilitating use and accessibility as well as minimizing runtime and input requirements. The application of different biomass devolatilization schemes within the model was conducted, not only for biomass fast pyrolysis product quantity but also liquid product composition (quality), to examine the effect of variable reaction kinetic sub-models on product yields. The model predictions were in good agreement with the results generated from the experimental work and mechanism modifications were proposed which further increased the accuracy of model predictions. Successively, the formulation of the modelled fluid dynamic scheme was adapted to study the effect of variable hydrodynamic sub-models on product yields for which no significant effect was observed. The work also looked into effect of the dominant process variables such as feedstock composition, bed temperature, fluidizing velocity and feedstock size on measurable product outputs (bio-oil, gas and biochar) and compared the results to those generated from the experimental fast pyrolysis unit. The ideal parameters for maximizing bio-oil yield have been determined to be those which: minimize the content of lignin and inorganic minerals in the feedstock, maintain the dense-bed temperature in a temperature range of 450-520 ºC, maximize the fluidization velocity without leading to bed entrainment, and limit the feedstock particle size to a maximum of 2000 μm.