Feasibility of Two MEMS-NIR Spectrophotometers for Characterizing Different Biofuel Origin

The EU common effort to reduce fossil fuel consumption paves the way to enhancing interest towards renewable resources, as solid biofuel. Its properties and origin are fundamental issues affecting energy conversion efficiency and supply chain sustainability. Laboratory conventional analyses are time consuming and require huge economic efforts, so it is essential to obtain rapid and real-time quality information along all the supply chain. Near Infrared Spectroscopy (NIRS) is able to satisfy these requirements, minimizing the time delay and promoting on site analysis along the whole process. The aim of the present work is to evaluate the performance of two different portable MEMS (Micro Electro Mechanical System) NIR spectrophotometers for identifying the raw biomass of which pellet is made. Thus, samples of different categories have been collected, particularly Italian softwood and hardwood species, wood processing industry residues and bioenergy crops. Ten replicates for each sample have been performed and different spectral pretreatments have been tested to reduce noise and scattering effects. Principal Component Analysis (PCA) has been used to investigate the spectral variability and look for groupings among the categories. According to the spectral range, the C12 and C13 spectrophotometers revealed different performance in groups separation due to the detection of different chemical bonds, for example glue presence. More interesting results could be obtained by widening the biomass samples to manage stronger dataset for chemometrics analysis. Nevertheless, results of the research aim at introducing NIR sensor, coupled with chemometrics, onboard pellet combustion devices to improve process efficiency and environmental performance.