Applications of machine learning and multi-objective optimization in agricultural waste management: A techno-economic study of hydrogen production from olive waste via combined air-steam gasification

This study investigates the potential of olive cake, a widely available agricultural residue in Jordan, as a sustainable biomass feedstock for hydrogen production via advanced gasification techniques. The goal is to optimize hydrogen yield by employing a combined steam?air gasification method, addressing challenges such as high ash and lignin content, variable moisture, and complex chemical composition Advanced optimization techniques, including machine learning models (Extreme Gradient Boosting, XGBoost) and the modified fire hawk optimizer, are used to optimize key gasification parameters: steam-to-biomass ratio, temperature, pressure, and air-to-biomass ratio. The gasification process is modelled by focusing on key thermochemical stages like decomposition, pyrolysis, combustion, and reduction. The results show that optimizing these parameters significantly enhances hydrogen production, achieving a rate of 329.25 kg/hr and a gasification efficiency of 82.1 %. The optimal conditions include a biomass flow rate of 5000 kg/hr, a gasification temperature of 778 ?C, and a pressure of 2.2 bar. The system?s hydrogen yield is most sensitive to changes in steam-to-biomass ratio and gasification temperature, with higher steam flows resulting in increased hydrogen content. The XGBoost model demonstrates high accuracy, with a coefficient of determination (R2) of 0.995 for hydrogen production and 0.998 for methane production. Further analysis reveals that increasing biomass flow rate and gasification temperature improves hydrogen yield, while higher pressures at elevated temperatures slightly reduce production. The economic analysis reveals a levelized cost of hydrogen of $1.23/kg, a net present value exceeding $213 million, and a payback period of 1.26 years.