(550c) A Hybrid Solution for a Heat Transfer Model to Distributed Parameters in the Solid Phase of a Biomass Pyrolysis Reactor | AIChE

(550c) A Hybrid Solution for a Heat Transfer Model to Distributed Parameters in the Solid Phase of a Biomass Pyrolysis Reactor

Authors 

Wiggers, V. R. - Presenter, Regional University of Blumenau
Almeida Junior, J., Regional University of Blumenau
Bastos, J. C. S. C., University of Blumenau
Krebs de Souza, C., University of Blumenau
Schmitz, F. R. W., Regional University of Blumenau
Bertoli, S., Regional University of Blumenau
A new world energy matrix has been the subject of much research in recent years. In this scope there are several types of raw materials, equipment and processes that can be studied and used. In optimization studies, the main interest is the economic viability with the lowest possible environmental impact.
Biofuels are promising regarding this aspect. Among the processes for obtaining biofuels, pyrolysis stands out as being used as a solution for the disposal of waste. Pyrolysis is the irreversible transformation of substances rich in carbon and hydrogen by the action of heat in a reduced or no oxygen atmosphere. The chemical reaction occurs at high temperatures, from 300 to 1000 ° C, and is predominantly endothermic.
In the present work, is proposed a hybrid solution based on the Finite Analytic (FA) and Finite Difference (FD) methods, for a heat transfer model to distributed parameters in the solid phase, of a biomass pyrolysis reactor. In the heat transfer model, the temperature profile inside the particle is considered; the reactor operates at steady state and has adiabatic walls; the particles are considered as a heat carrier medium; the gas phase reactions are first order. The discretization of the model occurs through the combination of the Finite Differences Method with the Finite Analytical Method.

The results were verified by comparison with a known lumped parameter model solution, as well as with the results from the literature, allowing the validation of the model. Finally, the proposed hybrid method of solution is easy handled, allowing for instance its implementation in an Excel worksheet. In comparison with a Runge-Kutta-Fehlberg&FD method, the present FA&FD method shows a superior performance relative to the processing time and it is equivalent in accuracy.