Determination of total carbon in biosolids using MID-infrared spectroscopy

This study was designed to determine if MID-Infrared (MID-IR) spectroscopy combined with Partial Least Squares (PLS) modelling could be used to predict total carbon (C) in biosolids generated from wastewater treatment plants. Biosolids samples were selected and analysed for total C, and MID-IR spectra were recorded. Using the total C and MID-IR data, a PLS model was developed using both the full spectral range and selected wavelengths regions. The PLS modelling showed that an R² of 0.97 was achieved for both approaches. To validate the modelling, PLS prediction was applied to a randomly selected validation set of 20 samples and prediction uncertainties determined using The Unscrambler software. A prediction error (or uncertainty) of approximately ±2% carbon was achieved using both full spectra and selected wavelengths regions. Based on the results of the study, MID-IR combined with PLS could be used as an alternate and inexpensive method to determine the total C of biosolids compared to conventional methods such as the Dumas combustion method using a LECO C analyser.