Lumped chemical kinetic modelling of raw and torrefied biomass under pressurized pyrolysis

Accurate prediction of the yield and composition of pyrolysis products is an important requirement for the design and operation of pyrolysis reactors and gasifiers. In this paper, a new semi-global kinetic reaction scheme is proposed to predict the composition of pyrolytic volatiles (classified as main chemical family groups), non-condensable gases and char derived from both raw and torrefied biomass for a wide range of operating conditions. The model is based on an adjustable mechanistic reaction scheme, which includes a combination of three different sub-mechanisms for the primary pyrolysis of three reference biopolymers (cellulose, hemicelluloses and lignin) and the secondary pyrolysis of their respective intermediates. The extent of primary/secondary reactions is varied according to the main process features (linear heating rate, temperature, pyrolysis time, volatile residence time and pressure). The secondary reactions in the scheme involve liquid-phase reactions of high molecular weight intermediates (producing non-condensable gases, water vapour and secondary char) as well as homogeneous and heterogeneous gas-phase conversion of primary volatiles. The model predictions were validated using experimental data obtained from fast pyrolysis in different micropyrolyzers (at 500-600 degrees C and heating rates of 27 and 110 degrees C/s, respectively) and slow pyrolysis in a laboratory-scale high-pressure fixed bed reactor (at 400-600 degrees C, 7 degrees C/min and 1, 15 and 30 bar). In general, the comparison of model outputs and experimental data were satisfactory, and the model predicted accurate trends in product distribution for changes of the lignocellulosic composition (the pre-removal of hemicelluloses in the case of torrefied biomass), heating rate and pressure. The model correctly predicted a significant increase in char yield (14.6 wt%) when torrefied instead of raw biomass was pyrolyzed due to the significance of char-forming reactions during pyrolysis of torrefied biomass. Moreover, the model's reliability was proven through its accurate prediction of various condensate groups in bio-oil produced in a micropyrolyzer (maximum deviation < 4 wt%). Corresponding to the experimental data, the model predictions showed that the effect of pressure was most significant in the range of 1-15 bar (bio-oil yield decreased by 5.4 wt%), whereas no significant effect in bio-oil yield was evident for a pressure increase in the range of 15-30 bar. The CO2 and CO yields were slightly under-predicted by around 3 wt %, attributed to the catalytic effects of inherent inorganics on secondary cracking reactions which were not considered in the reaction scheme. Future work should focus on the validation of the model at temperatures below 400 degrees C.

Automated summary

Accurate prediction of the yield and composition of pyrolysis products is crucial for the design and operation of pyrolysis reactors and gasifiers. A new semi-global kinetic reaction scheme was proposed in this study to predict the composition of pyrolytic volatiles from raw and torrefied biomass. The model showed satisfactory results in predicting product distribution trends for changes in lignocellulosic composition, heating rate, and pressure.