Development of a non-linear model for prediction of higher heating value from the proximate composition of lignocellulosic biomass

Abstract

ABSTRACT The determination of calorific value (in terms of higher heating value or HHV) is of much significance for the assessment of the energy recovery potential of biomass. The literature reported that HHV of the biomass can be computed using the proximate composition of the biomass. The prediction of HHV from the proximate composition of biomass is of great interest as it can save expensive and time-consuming laboratory trials. The present work evaluates the available prediction models for the estimation of HHV of the biomass from the literature. A thorough review showed that most literature reported models are limited in terms of applicability across different biomass categories. The present study proposed a non-linear prediction model that can be universally used to predict calorific value across biomass categories. Accordingly, the proposed model was validated using the proximate composition of various biomass specimens extracted from the literature to confirm the universal applicability of the model. The accuracy and precision of model prediction were assessed in terms of statistical indicators. The values of the statistical indicators showed that the proposed model had better prediction efficiency (in terms of accuracy and precision) and less bias compared to the other literature reported models for all biomass categories. The model predicted maximum energy recovery potential for biomass having high fixed carbon, high volatile matter, and low ash. The maximum energy recovery of 19.148 MJ/kg predicted by the model on validation was found to be associated with less than 2% (1.19%) percent error.