(366e) Predicting the Properties of Torrefied Biomass with a Three-Stage Chemical Kinetics Model | AIChE

(366e) Predicting the Properties of Torrefied Biomass with a Three-Stage Chemical Kinetics Model

Authors 

Klinger, J. - Presenter, Michigan Technological University
Bar-Ziv, E., Michigan Technological University
Shonnard, D. R., Michigan Technological University

Torrefied biomass has been heavy investigation in recent years and has great potential as a renewable solid fuel and as an upgraded feedstock for subsequent processing through pyrolysis or gasification.  When raw biomass is torrefied, many of the properties change – but those of highest interest from a renewable fuel/feedstock standpoint include increased heating value (primarily through drying, and deoxygenation reactions), changes in elemental composition (C,H,O), and decrease energy consumption in subsequent size reduction steps by partially degrading the hemicellulose and structure of the biomass.  There is a wealth of empirical information on how these physical properties change with torrefaction severity, but because torrefaction is complex and is highly dependent on the starting material, little work has correlated these improved properties with a kinetic model.  This poster presents an approach to model these changes in physical properties with a published three-stage chemical kinetics model.  An extensive literature survey has been performed to collect empirical information on the properties of woody feedstocks at varying torrefaction severities (time, temperature).  The kinetic model is then validated against the empirical data.  Preliminary analysis shows close agreement between model predictions for aspen wood, and the trends observed in survey data for all woody materials.  The modeling predicts an increase of HHV by approximately 30-50%, and C/O increase >200% and H/O by >150% at 300C for 30min torrefaction.  After further validation, it can be used as a powerful engineering tool to predict operation under varying or transient torrefaction severities without having to perform time consuming and costly experimental work.