Proximate analysis, backwards stepwise regression between gross calorific value, ultimate and chemical analysis of wood

The gross calorific value (GCV), proximate, ultimate and chemical analysis of debark wood in Portugal were studied, for future utilization in wood pellets industry and the results compared with CEN/TS 14961. The relationship between GCV, ultimate and chemical analysis were determined by multiple regression stepwise backward. The treatment between hardwoods–softwoods did not result in significant statistical differences for proximate, ultimate and chemical analysis. Significant statistical differences were found in carbon for National (hardwoods–softwoods) and (National-tropical) hardwoods in volatile matter, fixed carbon, carbon and oxygen and also for chemical analysis in National (hardwoods–softwoods) for F and (National-tropical) hardwoods for Br. GCV was highly positively related to C (0.79∗∗∗) and negatively to O (−0.71∗∗∗). The final independent variables of the model were (C, O, S, Zn, Ni, Br) with R2 = 0.86; F = 27.68∗∗∗. The hydrogen did not contribute statistically to the energy content.