Assessment of energy content of industrial woodchip based on prediction models of moisture content using portable NIR spectrophotometer

According to the fundamental transition to renewable resources, the UE is investing many efforts in terms of economic policy and political approach to minimize environmental impact and ensure energy provision. In Italy the bioenergy sector represents an important contributor to the energy production from renewable resources with the promotion of low-carbon economy, so primary solid biofuels are considered the main competitor to the total share of electricity generation worldwide. Woodchip plays a fundamental role in the present political and economic background, considering as a good candidate for the transition from fossil to renewable fuels, because widely available and distributed as a natural resource. Woodchip presents a high inherent variability of properties, derived from different conditions and its variability influences its technical and energetic properties. For this reason, the requirement of quality characterization could be considered a necessary step to price setting and combustion management and efficiency. Near Infrared (NIR) spectroscopy is considered a reliable alternative method compared to the standard laboratory analysis because provide real time results in a rapid and cost-effective approach. Coupled with chemometrics to extract useful information from spectral data, it allows the development of prediction and/or classification models. The aim of this study is the evaluation of the energy content (GCV) of industrial woodchip samples based on moisture content (MC) values predicted from PLS regression models implemented in a portable NIR spectrophotometer (MicroNIR,Viavi Solution). After the use of 800 woodchip samples as test set to assess prediction models performances, the best one, according to MC prediction, is used to GCV indirect estimation on 193 new woodchip samples. The results highlight the possibility to use NIR spectroscopy for assessing the indirect estimation of energy content of woodchip, turning out to be an advantage for sector operators, preventing time-consuming lab analysis.