Assessment of plant biomass for pellet production using multivariate statistics (PCA and HCA)

Abstract

Multivariate statistics can be a powerful tool in the assessment of energy properties of lignocellulosic materials and it is fundamental to estimate the theoretical, technical and economic potentials of these biomasses for bioenergy production. In this research, it was used to select the most favorable vegetable biomass for the production of biofuel pellets, through two techniques: Hierarchical Clustering Agglomerative and Principal Components. Six types of biomasses (pinus wood, eucalyptus wood, sugarcane bagasse, bamboo, sorghum, and elephant grass) and three blends were used. The immediate, elemental and thermochemical analyzes provided 16 variables of each of the 9 types of pellets. The dendrogram highlighted the group of forest biomass as the most suitable for the production of pellets and the principal component factors produced two bioenergetic indicators; one of general performance and other of combustibility. The forest biomass pellets was highlighted as potential for the production of biofuel pellets because they have energy properties with low levels of ash (0.54%) and nitrogen (0.83%), associated with high fixed carbon content (20.89%) and higher heating value (20.71 MJ kg−1), as well a higher energy density (13.95 GJ m−3). The multivariate analysis was efficient and can be used to classify lignocellulosic materials.