(133b) Kinetic Modeling of Biomass Torrefaction

Thermochemical conversion of lignocellulosic biomass is a promising technology for production of renewable power, fuels, and chemicals. Both pyrolysis and gasification of biomass have significant technical barriers that must be eliminated for successful widespread commercialization. Feedstock handling is complex due to the diverse nature of important feedstocks, including commercial timber, slash removed due to fire hazard reduction, switchgrass, and agricultural residues as diverse as rice hulls, corn stover and wheat chaff. Seasonal availability and low forest density make feedstock logistics complex and expensive. A process to homogenize the feedstocks that will simultaneously produce a stable energy-dense fuel is needed. Dry torrefaction is a low-temperature pyrolysis process in which the biomass is heated in a chemically inert environment at temperatures ranging between 200 and 300 ˚C. About 70 ? 80% of the biomass feedstock is retained as an upgraded solid fuel with 90% of the fuel value. The remainder is converted to gases. The reaction produces solid and a gas. The solid has about 70% - 80% of the mass, and approximately 90% of the fuel value of the original biomass. The solid fuel product is friable and hydrophobic, and is well suited to be made into pellets that might be stored.

We have completed an experimental investigation into measuring and modeling the rate of the torrefaction reaction. Samples of pure xylan and cellulose were subjected to torrefaction in a Thermo gravimetric analyzer (TGA) to measure the weight-loss kinetics at 250 ˚C, 275 ˚C, and 300 ˚C.

Hemicellulose reacts to form gas and solid products, and the solid product decreases as temperature increases. The rate constant was measured to be first order. Cellulose reacts more slowly, and the rate is apparently zero order.

We have also measured the weight-loss kinetics of torrefaction of lignocellulosic biomass. In our model, the lignocellulosic biomass consists of three components: hemicellulose, cellulose, and inerts. The experimental measurements done in the TGA were compared with that predicted by torrefaction of hemicellulose and cellulose individually. Without any adjustment, the rate predicted by parallel torrefaction of cellulose and hemicellulose was somewhat greater than the measured rate. With some minor modification to the stoichiometric constants the predicted rate matched the measured rate quite well.