Real-time measuring energy characteristics of cane bagasse using NIR spectroscopy

Abstract

Online measurement of moisture content (MC) and higher heating value (HHV) is desired for supporting the thermal conversion process in power plants. The main objective of this study was to develop a reliable and accurate method for online measuring of the HHV and MC of dried and as received sugarcane bagasse, respectively, using near-infrared (NIR) spectroscopy. Simulation of the conveyor belt in power plants and online energy quality analysis of bagasse for combustion process were designed. One hundred samples of bagasse were collected from different sugar mills at various locations of large piles of bagasse stored in both indoor and outdoor buildings. A long-wave spectrometer with a wavelength range of 860–1755 nm was used for scanning using diffuse reflection mode. The models were developed by partial least squares (PLS) regression using spectral sets obtained from the raw spectra, the preprocessed full spectra from the traditional approach, and the preprocessed spectra from the multi-block technique. The multi-blocks of the spectral pre-treatment technique provided effective models for both MC and HHV prediction. The best model for MC had a coefficient of determination of prediction (R2p) and the root mean square error of prediction (RMSEP) of 0.90 and 3.9% wb respectively. In the meantime, the best model for HHV had R2p and the RMSEP of 0.71 and 188.7 J g−1. This research demonstrated that NIR spectroscopy provides a feasible and reliable method for online analysis of MC and could be used for screening of HHV of bagasse.