Multi-aspect assessment and artificial intelligence-based optimization of a sustainable solar-to-power/hydrogen process using three-state numerical simulation and thermal energy storage

Solar thermal systems face significant challenges due to daily solar variability, requiring advanced simulation and design methods. This study addresses these issues using a three-state numerical simulation, thermal energy storage, and artificial intelligence-based analysis. The proposed system comprises a solar power tower and an energy storage option designed to support a supercritical CO2 plant. Additionally, this plant is integrated with reverse osmosis desalination and a proton exchange membrane electrolyzer. A three-mode numerical simulation—solar, solar-storage, and storage—is conducted, and an artificial intelligence-driven multi-objective optimization using the NSGA-II algorithm is applied to enhance performance and cost efficiency. Under optimal conditions, the system exhibits a hydrogen production capacity of 54203 m3/day, a net power output of 4415 kW, an exergy round-trip efficiency of 23.01 %, and a unit cost of products (UCOP) of 33.41 $/GJ. The levelized cost of hydrogen is calculated to be 5.06 $/kg, with a net present value of 214.5 M$ and a payback period of 9.4 years. In addition, the entire installation indicates a cost rate of 1436 $/h. Compared to the base case, hydrogen output is improved by 29.7 %, and UCOP is reduced by 12.5 %. These results highlight the system's potential for cost-effective, zero-carbon hydrogen production using solar energy.