Prediction of carbon, hydrogen, and oxygen compositions of raw and torrefied biomass using proximate analysis

Elemental compositions of biomass are essential for designing energy conversion systems. Only a few correlations to estimate the elemental compositions using the proximate analysis of raw biomass have been published so far. Recently researches on biomass torrefaction have been increasing significantly, which require performing an elemental analysis of the torrefied biomass. Torrefaction affects both the proximate and elemental analyses of biomass. Therefore, this study examines if the existing correlations can be deployed or not for estimating carbon, hydrogen, and oxygen compositions of the torrefied biomass. For this, estimation errors were calculated for the existing correlations using data from the torrefied biomass. Results suggest that existing correlations were not suitable for predicting the elemental compositions of the torrefied biomass. New correlations were proposed using a wide range biomass, including both raw and torrefied biomass (447 samples). New correlations C=-35.9972+0.7698VM+1.3269FC+0.3250ASH,H=55.3678-0.4830VM-0.5319FC-0.5600ASH, and O=223.6805-1.7226VM-2.2296FC-2.2463ASH were selected for future use. These correlations have the MAE of 2.58%, 0.41%, 2.60%, the AAE of 5.23%, 9.94%, 8.79%, the ABE of 0.45%, 2.82%, 2.01%, and the R2 of 0.83, 0.70, 0.84 corresponding to the measured values of C, H, and O, respectively. The selected correlations were also validated and compared with the existing correlations using another set of data that includes raw, washed, torrefied, and carbonized biomass. Selected new correlations could be used for predicting carbon, hydrogen, and oxygen compositions in the raw and torrefied biomass, especially those biomasses which have negligible nitrogen and sulfur contents.