Model fitting kinetic analysis and characterisation of the devolatilization of coal blends with corn and sugarcane residues

Single and multi-component model fitting was used to determine the pyrolysis reaction kinetics of coal, corn cobs, and sugarcane bagasse, as well as blends of coal with each of the biomasses. The results showed that single component kinetics were a poor representation of the decomposition behaviour of all of the samples, regardless of whether 1st or nth order reaction models were assumed. Conversely, reasonably approximate simulations of reaction rates could be obtained by adopting the simpler 1st order model when 3 or more parallel reactions where assumed in the biomass fuels. However, not all the reactive pseudocomponents in biomass strictly followed the first order model. In comparison, the nth order model was found to be a more robust and flexible approach providing simulations and predictions with better fits to the experimental data, particularly for coal were a larger deviation from 1st order reaction behaviour was observed. Apparent activation energy values obtained for nth order model fitting with 3 pseudocomponents were 212, 188, and 94 kJ mol(-1) for sugarcane bagasse; 215, 189, and 99 kJ mol(-1) for corn cobs: and 252, 147 and 377 kJ mol(-1) for coal. Corresponding pre-exponential factor values obtained were 3.6 x 10(17),8.5 x 10(16),3.3 x 10(8) min(-1) for bagasse; 7.2 x 10(18),2.6 x 10(17),2.2 x 10(9) min(-1) for corn cobs; and 2.2 x 10(18), 1.5 x 10(9), 2.5 x 10(20) min(-1) for coal, respectively. These results, along with the corresponding reaction order values, produced reaction rate simulations with less than 2% deviation from experimental observations. The activation energies obtained also compared well with values derived previously in a model free analysis of the same data (174 kJ mol(-1), 184 kJ mol(-1), and 246 kJ mol(-1) for CC, in the 0.2-0.8 conversion range). Kinetic analysis of the coal blends with biomass revealed non-additive tendencies as indicated by the relatively poor quality of fit achieved when the parameters of the contributing single fuels were used to predict co-pyrolysis reaction rate curves. (C) 2011 Elsevier B.V. All rights reserved.