Decision study and ANN-assisted multi-criteria optimization of a novel three-state solar-driven integrated process using energy storage for hydrogen liquefication: A case study for Malaysian solar status

This paper provides a comprehensive decision study of a novel heat integration process for a three-state heliostatbased power plant, utilizing thermal energy storage, to generate and liquefy hydrogen. The proposed system comprises a supercritical CO2 cycle, a solid oxide electrolyzer cell, photovoltaic-thermal modules integrated with reverse osmosis desalination to provide the required water for electrolysis, and a Claude cycle for hydrogen liquefaction. Based on global data, the variation in solar conditions can be categorized into periods of lowradiation, high-radiation, and without-radiation modes. The study examines thermodynamic, sustainable/exergoenvironmental, and economic perspectives using a sensitivity analysis followed by the consideration of three distinct solar radiation-independent optimization scenarios (exergy-cost, sustainably-cost, and environmentalcost). The optimization process involved the development of artificial neural network models to represent the objective functions, their integration into the MOGWO algorithm, and the application of the TOPSIS method to determine the ultimate optimal solution. Finally, a case study is conducted to assess the viability of the scheme in three different climatic regions in Malaysia, focusing on coastal areas. The sensitivity analysis identifies the mass flow split ratio of the molten salt as the most crucial parameter, demonstrating a sensitivity index of 0.27. Moreover, the first optimization scenario demonstrates the best feasibility, showing an exergetic round-trip efficiency and a liquefied hydrogen cost of 19.30% and 3.87 $/kg, respectively. The case study results indicate that Alor Setar is the most favorable city for establishing the system, achieving a liquefied hydrogen production rate of 1494 kg/day at a cost of 7.77 $/kg.