Abstract Torrefaction and slow pyrolysis are processes that can be used for improving the fuel properties of low-quality biomass, which could be effectively used for improving fuel logistics and making biomass a fully tradeable commodity. Automation of these processes could prove vital for their economic viability and carbon footprint, giving the possibility to reduce manpower requirement and raining time for operators. Automation based on online control of the quality of biocoal is not possible. However, finding a suitable marker in torgas could make automation possible. Results presented in the study show that formaldehyde can be considered as a viable marker. Obtained results are limited to torrefaction and slow pyrolysis of lignocellulosic biomass at a maximum reactor temperature of 430 °C. Temperature measurements show that endothermal evaporation of the bound moisture delays the time necessary to reach the point where exothermal reactions start taking a leading role in the thermal decomposition process, which is extremely important in the context of thermal run-away of such installation. Artificial neural network proved to be an effective tool for determining of correlation between the heating value of pyrolysis gas and its concentration of formaldehyde. This could be used in advanced control solutions for pyrolysis gas combustion systems.
Read full abstract