Resende F. L. P., Sánchez C. G., Ferreira M. M. C., "Pyrolysis of Biomass: Prediction of Kinetic Behavior in Fluidized Bed From Thermogravimetric Data". Alicante, Spain, 23-27/05/2004: 16th International Symposium on Analytical and Applied Pyrolysis. Book of Abstracts, (2004). Oral L27.

BIOMASS PYROLYSIS: PREDICTION OF KINETIC BEHAVIOR IN
FLUIDIZED BED FROM THERMOGRAVIMETRIC DATA

Resende, F. L. P.*, Sánchez, C. G., Ferreira M. M. C.**

*Faculdade de Engenharia Mecânica da Unicamp, Caixa Postal 6088 - Campinas - SP -
CEP: 13083-970. e-mail: caio@fem.unicamp.br
** Instituto de Química da Unicamp, Caixa Postal 6154 - Campinas - SP - CEP: 13084-971
 
 

            The goal of this work was  to improve the achievement of kinetic parameters  for pyrolysis reactions  of
biomass. Such parameters are important in the design and simulation of reactors for biomass thermal degradation
with energetic purposes.   One technique usually  performed  to obtain the  kinetic parameters is  the widely used
Thermogravimetric Analysis (TG). The kinetic parameters can also be measured with the use of a Fluidized Bed,
that  performs in laboratorial scale a process similar  to those that carry out in large scale reactors.   Nevertheless,
these  two measuring ways have been showing different results,  since  the reaction conditions  in  a fluidized bed
reactor are very different from those in a TG equipment.   In  this work,  Sugarcane  Bagasse, Grass  and  Waste
Wood pyrolysis data from both experiments were used to  propose a mathematical model based on chemometrics
methods to relate the data.   The model intends to allow that  TGA data are enough to predict kinetic behavior of
biomass when pyrolyzed in  Fluidized Bed Reactors,   in order to turn wider the applicability  of the large amount
of TGA data available in literature.

             Fluidized bed  experiments  were  performed  from  350º  to  900ºC,   and  CO  emission vs.  time  was
measured.   TG dynamic runs were performed varying the heating rate from  5ºC/min  to  20ºC/min.

             The analysis used  include  the weight loss data  from  TG in a matrix  X,  and  the rate constant of  CO
emission obtained in the fluidized bed in a matrix  Y.   The vector b  relating the two matrix is called a regression
vector.  For the range of temperature from 500ºC to 900ºC,  it was possible to use  Multivariate  Analysis  to find
regression  vectors   relating   the  kinetic   results  in   TG   experiments  with   those  obtained  in  fluidized  bed
experiments.   The regression vectors  were found  to be given  by  functions of  the type   f(x) = A sen(w x + x₀),
where  x  is the line in the vector.   The knowledge of  the  regression  vector  allows  the  prediction  of  the  rate
constant of  CO  emission in  this temperature  range.
 

References
[1] Ferreira, M. M. C., Antunes, A. M., Melgo, M. S. and Volpe, L. O., Quím. Nova, 22, 5, 724-729 (1999).
[2] Simmons, G. and Sanchez, M., J. Anal. Appl. Pyrol., 3, 161-171 (1981).
[3] Orfão, J. J. M., Antunes, F. J. A. and Figueiredo, J. L., Fuel, 78, 3, 349-358 (1999).