Energy optimization from a binary mixture of non-edible oilseeds pyrolysis: Kinetic triplets analysis using Thermogravimetric Analyser and prediction modeling by Artificial Neural Network

Abstract

Pyrolysis kinetics and thermodynamic parameters of two non-edible seeds, Pongamia pinnata (PP) and Sapindus emarginatus (SE), and their blend in the ratio of 1:1 (PS) were studied using the thermogravimetric analyzer. Kinetic triplets were determined using both model-free [Starink (STR), Friedman (FRM), Iterative Kissinger-Akahira-Sunose (IT-KAS), Iterative Ozawa-Flynn-Wall (IT-OFW), Vyazovkin (VYZ), and Master plot (MP)] and model fitting Coats-Redfern (CR) methods at three different heating rates 10, 30 and 50 °C/min. Activation energies were 192.66, 179.44, and 163.25 kJ/mol for PP, SE, and PS, respectively. It was found that the blend of the two-biomass (PS) showed promising results with lower activation energy compared to the individual biomass. Thermodynamic parameters (ΔG, ΔS, and ΔH) were obtained using the model-free isoconversional method. The three hidden layers of complex neuron topology are well fitted to the experimental DTG curves by artificial neural network (ANN). The study confirmed that the heating rate had a significant impact on the kinetics and thermodynamic parameters. The reaction mechanism was also in consonance with the experimental data. The study suggests that the PP and SE seeds can be an appropriate feed for pyrolysis, and their blend (PS) can be a viable alternative in optimizing the entire process.