Canonical correlation analysis is a statistical multivariate procedure that allows analyzing linear correlation that may exist between two groups or sets of variables (X and Y). This paper aimed to provide canonical correlation analysis between a group comprised of lignin and total extractives contents and higher heating value (HHV) with a group of elemental components (carbon, hydrogen, nitrogen and sulfur) for lignocellulosic wastes. The following wastes were used: eucalyptus shavings; pine shavings; red cedar shavings; sugar cane bagasse; residual bamboo cellulose pulp; coffee husk and parchment; maize harvesting wastes; and rice husk. Only the first canonical function was significant, but it presented a low canonical R² . High carbon, hydrogen and sulfur contents and low nitrogen contents seem to be related to high total extractives contents of the lignocellulosic wastes. The preliminary results found in this paper indicate that the canonical correlations were not efficient to explain the correlations between the chemical elemental components and lignin contents and higher heating values.
Multivariate statistics; canonical function; bioenergy
