A numerical model for the fractional condensation of pyrolysis vapours

Experimentation on the fast pyrolysis process has been primarily focused on the pyrolysis reactor itself, with less emphasis given to the liquid collection system (LCS). More importantly, the physics behind the vapour condensation process in LCSs has not been thoroughly researched mainly due to the complexity of the phenomena involved. The present work focusses on providing detailed information of the condensation process within the LCS, which consists of a water cooled indirect contact condenser. In an effort to understand the mass transfer phenomena within the LCS, a numerical simulation was performed using the Eulerian approach. A multiphase multi-component model, with the condensable vapours and non-condensable gases as the gaseous phase and the condensed bio-oil as the liquid phase, has been created. Species transport modelling has been used to capture the detailed physical phenomena of 11 major compounds present in the pyrolysis vapours. The development of the condensation model relies on the saturation pressures of the individual compounds based on the corresponding states correlations and assuming that the pyrolysis vapours form an ideal mixture. After the numerical analysis, results showed that different species condense at different times and at different rates. In this simulation, acidic components like acetic acid and formic acids were not condensed as it was also evident in experimental works, were the pH value of the condensed oil is higher than subsequent stages. In the future, the current computational model can provide significant aid in the design and optimization of different types of LCSs. - Highlights: • CFD modeling of the fractional condensation of the pyrolysis vapour compounds. • Multiphase multicomponent vapour transport with phase change is simulated. • Corresponding states method is used for saturation vapour pressure calculations. • Heat transfer and flow dynamic effects on the phase change phenomena are studied. • The model is in good agreement with experimental observations