Statistical description of biomass blends devolatilization

Biomass is widely used as a renewable energy source either as a raw material or as a form of pellets, which are very often produced of mixture of different sorts of biomass. This gives the opportunity to produce pellets which are characterized by desired properties. However there is a question of interactions between sorts of biomass which may occur during pyrolysis. The main scope of this paper is to investigate devolatilization of biomass mixture and to approximate amount of released volatiles as a function of biomass composition. Five types of biomass has been selected for measurements: two types of wood - oak and pine, two types of agricultural wastes - rape straw and wheat straw and willow as popular energy crop. It has been assumed that the amount of volatiles released in a given temperature range can be expressed as a function of elemental composition of biomass mixture. Different forms of function have been tested and correlation coefficients as well as average and maximal approximation errors were examined as defining the quality of approximation. It has been found that linear form of a function gives the best quality of approximation characterized with high correlation coefficient and relatively small maximal approximation.