A predictive model for multi-criteria selection of optimal thermochemical processing pathways in biorefineries

The rapid growth of the global economy, combined with the growing demand for energy, environmental degradation from greenhouse gas emissions, and fluctuating fossil fuel prices, have emphasised the importance of renewable sources of energy. Biofuels produced from biomass conversion processes accounts for at least 13% of the gross global energy consumption and 70% of the world renewable energy mix. The biomass resources, including municipal, industrial and forestry waste, were proven to have a great potential for deriving various forms of energy in an affordable and reliable manner. In this regard, several thermochemical technologies have been developed to convert biomass waste into energy. However, due to the extremely heterogeneous characteristics of biomass resources, the feasibility and efficiency of these processes may greatly vary, depending on the biomass category and composition. Intensive experiments, simulations and optimisation models were developed to select the optimal processing pathway for each feedstock, which generally consume significant time and effort. To address this issue, it is desirable to seek novel and accurate mathematical representations that enable rapid performance estimation and multi-criteria selection for the optimal biomass processing pathway based on the physical properties and chemical compositions of different biomass categories, without the need for expensive experimental setup or time-consuming simulations. The objective of this study is to develop a mathematical model which links the biomass’ proximate and elemental analyses to three crucial technology performance criteria: including the return on investment, energy efficiency, and carbon intensity. For this purpose, intensive simulations and sensitivity analyses were carried out using Aspen Plus to examine three main processes including gasification, pyrolysis and hydrothermal liquefaction (HTL). The comprehensive simulation data were subsequently used to develop and compare multiple meta-models for their accuracy of representation using regression algorithms. This model is believed to expedite the ongoing research on biomass thermo-processing and play a significant role towards enhancing biomass sustainability.