Yousef, SamyEimontas, JustasStriugas, NerijusAbdelnaby, Mohammed Ali2024-01-152024-01-152024-01https://doi.org/10.1016/j.jaap.2023.106330http://repository.msa.edu.eg/xmlui/handle/123456789/5814This research aims to convert the resin fraction of waste wind turbine blades (WTB) into value-added aromatic chemicals using catalytic pyrolysis. The catalytic study on WTB made of glass fibre/unsaturated polyester resin (UPR) was performed on two different types of zeolite catalysts (ZSM-5 and Y-type) using a thermogravimetric (TG) analyser. The effect of catalyst and heating rate on the abundance and composition of the synthesised aromatic chemicals was observed using TG-FTIR and GC/MS. The kinetics and thermodynamic behaviour of catalytic pyrolysis of WTB was also studied using traditional modelling techniques (KAS, FWO, Friedman, Vyazovkin, and Cai) and an artificial neural network (ANN). TG-FTIR results showed that the gases released from the catalytic process were very rich in aromatic groups, while GC/MS analysis revealed that benzene, toluene, and ethylbenzene (BTE) were the main constituents of the synthesised aromatic chemicals with abundance estimated at 36% (ZSM-5 at 10◦C/min) and 64% (Y-type at 15◦C/min) accompanied by a significant reduction in styrene formation up to 16.2% (main toxic element in the UPR). Besides, it contributed to reduction of the activation energy of the reaction up to 126 KJ/mol (ZSM-5) and 100 KJ/mol (Y-type). The trained ANN model also showed high performance in predicting the thermal decomposition zones of WTB at unknown heating rates with R2 close to 1. Accordingly, the use of catalytic pyrolysis of WTB over a Y-type zeolite catalyst is highly recommended for decomposition of UPR to aromatic chemicals BTE and reduction of styrene in the produced fuel.enWaste wind turbine bladesCatalytic pyrolysisValue-added aromatic chemicalsArtificial neural networkCatalytic pyrolysis kineticsSynthesis of value-added aromatic chemicals from catalytic pyrolysis of waste wind turbine blades and their kinetic analysis using artificial neural networkArticlehttps://doi.org/10.1016/j.jaap.2023.106330