Prediction of biomass pellet quality indices using near infrared spectroscopy

The potential of near infrared spectroscopy in conjunction with partial least squares regression to predict quality indices of biomass pellet blends was assessed. A diverse range of biomass was used including wood, Miscanthus and herbaceous energy grasses. The moisture, carbon and ash contents and gross calorific value were predicted with a root mean square error of cross validation of 0.73% (R2 = 0.85, range = 9.11%), 2.74% (R2 = 0.78, range = 19.83%), 0.62% (R2 = 0.82, range = 6.22%) and 0.24 MJ kg−1 (R2 = 0.94, range = 3.26 MJ kg−1), respectively. The moisture and gross calorific value models had good and excellent accuracy, respectively while the ash and carbon models were deemed good and fair, respectively. The results indicate that near infrared spectroscopy has the potential to predict quality indices of biomass pellets in a multi-biomass stream.