Monday, May 2, 2011
Grand Ballroom C-D, 2nd fl (Sheraton Seattle)
Obtaining accurate chemical compositional analysis of biomass feedstocks in a time efficient manner is necessary for the commercialization of biofuels. NREL has developed several rapid analytical methods for biomass feedstocks utilizing Near Infrared (NIR) spectroscopy and Projection to Latent Structures (PLS) multivariate analysis. Some of these feedstocks include corn stover, sorghum, switchgrass, and miscanthus. While each of these methods functions well for a given feedstock type, we present the results of combined methods. Individual methods are compared to iterations of combined methods. Statistics are presented to evaluate the initial performance of the methods, and validation samples are discussed to measure the predictive quality of the methods.