Development of machine learning model for the prediction of selectivity to light olefins from catalytic cracking of hydrocarbons

• Published in: Fuel (Guildford) Vol. 381
• Main Authors: Mafat, Iradat Hussain, Sharma, Sumeet K., Surya, Dadi Venkata, Rao, Chinta Sankar, Maity, Uttam, Barupal, Ashok, Jasra, Rakshvir
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.02.2025
• Subjects: Artificial neural network; Catalytic cracking; Ethylene; Hydrocarbons; Machine learning; Propylene; Propylene; Machine learning; Ethylene; Hydrocarbons; Artificial neural network; Catalytic cracking
• ISSN: 0016-2361
• DOI: 10.1016/j.fuel.2024.133682

•Importance of Machine Learning modeling is explained.•Light Olefins prediction is done with the help of 6 input factors.•Developed artificial neural network based machine learning model.•Model achieved highest accuracy value of 80.6% and 82.1% on the testing dataset.•SHAP analysis shows that reaction temperature is crucial for light olefins production. Light olefins are the primary building block for the production of petrochemicals and polymers. Light olefins are largely produced from steam/catalytic cracking of naphtha or ethane/propane. Selectivity to light olefins is significantly dependent on the reaction conditions. In this article, several machine learning models are developed and tested to predict the selectivity of ethylene and propylene using seven input features. For this study, atotal of eight ML models consisting of adaptive boost, extreme gradient boost, categorical boost, light gradient boost, decision tree with bagging, random forest, k-nearest neighbour, and artificial neural models are developed. The extreme gradient boost model gave thehighest prediction accuracy for the ethylene selectivity, while the light gradient boost gave thehighest R2 for the propylene selectivity. The SHAP analysis showed the input parameter’s importance ranking for ethylene predictions as temperature > number of carbon atoms > Si/Al ratio > acidity > weight hourly space velocity > effect of diluent > number of hydrogen atoms. The importance ranking of input parameters for propylene selectivity was observed as weight hourly space velocity > acidity > temperature > Si/Al ratio > effect of diluent > number of carbon atoms > number of hydrogen atoms.