Modeling of Air Emissions From Co‐Pyrolysis of Polymeric Wastes: Simplex Lattice Design and Artificial Neural Network Approach

ABSTRACT

Co‐pyrolysis of plastics is a simultaneous thermochemical conversion of a mixture of plastics to value‐added products. This study quantified air emissions from co‐pyrolysis of low‐density polyethylene (LDPE), polystyrene (PS), and polyethylene terephthalate (PET) at varying mixture compositions. The emissions were modelled using simplex lattice design (SLD) and artificial neural network (ANN). A mixture containing 100% LDPE, 0% PS, and 0% PET emits the highest quantities of hydrocarbons (HCs), carbon monoxide (CO), and nitrogen oxide (NO) with values of 7980.46, 1793.67, and 41.41 mg/m ³ , respectively, with no indication of NO ₂ , SO ₂ , H ₂ S, or CO ₂ emissions. Meanwhile, mixture containing 0% LDPE, 0% PS, and 100% PET emits the lowest quantities of HCs and CO, with values of 200.47 and 44.69 mg/m ³ , respectively. Mixtures containing 17% LDPE, 17% PS, and 66% PET; 17% LDPE, 66% PS, and 17% PET; and 0% LDPE, 50% PS, and 50% PET showed no emissions of NO. The study generated predictive models for the emissions of HC, CO, and NO and statistically established that using SLD, a special quartic model best describes HC and CO emissions, while a quadratic model best describes NO emissions. Although ANN shows good modelling of the emissions, the coefficients of estimate were lower than that of simplex lattice design.