A comparison of the torrefaction behavior of wood, miscanthus and palm kernel shells: Measurements on single particles with geometries of technical relevance

•Single particle investigations (time resolved mass loss) under well-defined torrefaction conditions.•The results show the general tendencies known for the influence of temperature, residence time on the mass yield.•Process temperature and residence time can be varied.•Data processing allows to derive mass yield, energy yield and kinetic parameters. A torrefaction test rig was designed to investigate large single biomass particles up to characteristic sizes of 25 mm, typical for industrial reactors. Time-resolved mass loss for such particles is measured with a magnetic suspension balance at well-defined torrefaction conditions (temperature, residence time, gas atmosphere). This paper comprises the results of woody and non-woody biomass: pine, a coniferous, and beech, a deciduous, wood, palm kernel shells and miscanthus. Influence of process temperature (240 to 320 °C), residence time (up to 1 h) and type of solid biomass on time-resolved mass loss is presented. Additional tests with oxygen in the process gas (0–15 vol%), typical for industrial torrefaction systems, are carried out for selected samples of beech wood. The differences in torrefaction behaviour of bark, sap- and heartwood of pine are evaluated. Finally, it is shown that the torrefaction reactor developed allows to derive kinetic parameters for mass loss. At temperatures up to 300 °C the mass loss for palm kern shells is highest followed by miscanthus, and pine. By examining pine, as an example, it is shown that heartwood is significantly more reactive than sapwood and bark. Finally, it is demonstrated, that for the particle sizes considered here heat and mass transfer limitations can be neglected for the determination of torrefaction kinetics. Kinetic data agree well with data from literature.