Co-pyrolysis of Azadirachta indica non-edible seed and waste LDPE: Analysis of kinetic models using thermogravimetric analyser and prediction modeling with Artificial Neural Network (ANN)

Abstract

Energy production from waste biomass seeds and plastic wastes must be done immediately, keeping the future in mind, as waste plastic and seeds have become a major source of attractiveness for producing liquid fuel in a pollution-free environment and disposing of waste plastic, which causes serious environmental issues. In this study, theoretical modeling was used to assess the pyrolysis feedstock potential of Azadirachta indica, or Neem non-edible seeds, and waste LDPE (low-density polyethylene) in a set ratio. Thereafter, the TG-DTG data was applied into artificial neural network (ANN). The Azadirachta indica seed-waste LDPE mixture was TG-DTG examined at three heating rates of 10, 20 and 30 °C/min. The thermodynamic parameters were also obtained at the same heating rates. A blend sample requires less energy to pyrolyze than an individual sample. When ANN prediction modeling was used to predict TG-DTG data, R2 was always found unity in training, validation, and testing, and MSE was minimised for all heating rates. Levenberg-Marquardt trained the ANN model. According to the findings of this study, combining Azadirachta indica seed with waste LDPE could be a workable alternative that would result in an improvement to the process as a whole.