Integrating Biogas of Municipal Wastes, Solar and Wind Sources for Sustainable Hydrogen Generation: A Techno-Economic Study by Optimization and Machine Learning Methods

• Published in: IOP conference series. Earth and environmental science Vol. 1520; no. 1; pp. 12003 - 12008
• Main Authors: Sekhar Santhappan, Joseph, Gopinath, Arun S, Boddu, Muralikrishna, Ahmed Yousif Al-Nofli, Maryam, Bora, Satyanarayana
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.07.2025
• Subjects: Alternative energy sources; Biofuels; Biogas; Datasets; Digesters; Gasification; Hydrogen; Hydrogen production; Learning algorithms; Machine learning; Meteorological data; Municipal wastes; Municipal wastewater; Net zero; Optimization; Parameters; Photovoltaic cells; Photovoltaics; Python; Renewable energy sources; Renewable resources; Sustainability; Wastewater; Wind power; Wind turbines; Oman
• ISSN: 1755-1307; 1755-1315; 1755-1315
• DOI: 10.1088/1755-1315/1520/1/012003

The combined use of biofuel from biowaste gasification and renewable energy sources is an attractive method for sustainable hydrogen generation. This strategy can contribute to the implementation of the net-zero emissions target of several countries, including Oman. Therefore, this study combines machine learning (ML) and HOMER Pro’s optimization methods for predictions. The software ‘OriginPro’ analysed the Pearson coefficient of data sets and removed the redundant input parameters. The Python code used in this study employed two machine learning (ML) methods to analyse the potential of ML techniques to interrelate the performance parameters. The daily hydrogen load considered is 200 kg, which can fuel 40 hydrogen cars in a fuel station. The biodigesters of municipal wastewater (MWW) supplied the fuel to biogas generators (BG). The electrolyzer used power from BG, a photovoltaic (PV) array, a battery bank (BB), and a wind turbine (WT). The simulation used 28 combinations of renewable energy systems. Based on weather data from Khasab, Oman, the levelized cost of hydrogen (LCOH), the net present cost of the system, biogas use, and the amount of wastewater that the system will need are estimated to be $3.06 to $5.34 per kg, $2,854,670 to $3,869,221, 232,268 to 2,494,004 m per year, and 2,476 to 23,537 m per day. The comparison of LCOH with literature proves the feasibility of MWW utilization in sustainable hydrogen generation. The selected ML models’ determinants (R²) are above 0.99, and the mean square, root mean square, and mean absolute errors are below 0.005. Thus, this analysis confirms the validity of the multiple input, single output regression ML approach for the selected numerical data sets.