Characterization of biodiesel based on plastic pyrolysis oil (PPO) and coconut oil: Performance and emission analysis using RSM-ANN approach

Asif Afzal,Roji George Roy,Chacko Preno Koshy,Mohamed Abbas,Erdem Cuce,Abdul Razak RK,Saboor Shaik,C. Ahamed Saleel,Alex Y
DOI: https://doi.org/10.1016/j.seta.2023.103046
IF: 7.632
2023-01-18
Sustainable Energy Technologies and Assessments
Abstract:The work presents comparative study of the production and characterisation of sustainable biodiesel fuel made from waste plastics and virgin coconut oil. For characterisation, various chemical tests were performed, including Fourier Transform Infrared (FTIR) spectroscopy, Thermogravimetric analysis (TGA), and Gas chromatography–mass spectrometry (GCMS). The emissions and performance of this fuel were investigated to determine whether it could be utilised in diesel engines without modification. RSM (response surface method) was used to design the experiments. ANN (artificial neural network) was used to model the relationship between the input and output parameters. It was seen that, 20 % hybrid blend with diesel showed a better output (33 %) than 10 % and 30 % blend. For 75 % of load, a value of 0.12 % (minimum) CO emission was obtained for the same blend with diesel fuel. Low proportion levels blends (10 %) have less amount of oxygen content, hence reduced in NOx (400 ppm). The ANN and RSM models were found to be fitting correctly with the experimental readings, with R 2 ratios varying from 90 % to 93.5 %, respectively. The outcomes demonstrated that RSM and ANN were excellent modelling techniques with good accuracy. In addition, ANN's prediction performance (R 2 = 0.9978 for BTE ) was somewhat better than RSM's (R 2 = 0.960 for BTE ).
energy & fuels,green & sustainable science & technology
What problem does this paper attempt to address?