Prediction of the Energy Properties of Charcoal Obtained from Eucalyptus and Corymbia Biomass Using Portable and Benchtop NIR Spectrometers

Abstract

Monitoring charcoal quality is essential for the industry. Near-infrared (NIR) spectroscopy enables fast and accurate predictions of key properties. This study evaluated the use of benchtop and portable NIR sensors to predict charcoal characteristics from woody biomass of 15 commercial clones (11 Eucalyptus and 4 Corymbia). Two trees per clone were sampled at six stem positions, generating 30 composite wood samples. After carbonization and grinding, spectral data were collected, totaling 600 spectra per sensor. Partial least squares regression was used to develop models for gravimetric yield (GY), apparent relative density (ARD), fines content (FC), volatile matter content (VMC), ash content (AC), and fixed carbon content (FCC). For Eucalyptus clones, the benchtop sensor outperformed the portable one for GY (R²p = 0.74; RPD = 2.02), ARD (R²p = 0.87; RPD = 2.82), VMC (R²p = 0.72; RPD = 1.92), AC (R²p = 0.72; RPD = 1.92), and FCC (R²p = 0.63; RPD = 1.64). The portable sensor was better only for FC (R²p = 0.64; RPD = 1.60). Similarly, for Corymbia clones, the benchtop sensor performed better for GY (R²p = 0.79; RPD = 2.15), ARD (R²p = 0.87; RPD = 2.77), FC (R²p = 0.69; RPD = 1.73), and AC (R²p = 0.61; RPD = 1.62). The portable sensor showed better results for FCC (R²p = 0.61; RPD = 1.48) and VMC (R²p = 0.64; RPD = 1.40). Overall, benchtop and portable NIR spectrometers showed similar performance in estimating charcoal parameters.