4E assessment and neural network optimization of a solid oxide fuel cell-based plant with anode and cathode recycling for electricity, freshwater, and hydrogen production

Abstract

This study presents a novel hybrid system based on solid oxide fuel cell (SOFC) to produce electricity, heating load, and freshwater. A heat recovery method also provides electricity for hydrogen production through proton exchange membrane (PEM) electrolysis. A parametric study and sensitivity analysis investigated the impact of design parameters on the system outputs. Energy, exergy, economic, and environmental (4E) analyses were conducted. The results obtained from the base system show that the net power, freshwater and hydrogen production are 938.94 kW, 360.45 kg/h, and 11.94 kg/h, respectively. Also, the total exergy efficiency, total cost rate, and CO2 emission index are 51.46%, 53.57 $/h, and 0.075 tons/GJ, respectively. This system is optimized by integrating artificial neural networks with the genetic algorithm based on four objective functions. The optimization results reveal that in scenario a, considering the exergy efficiency and the total cost rate as the objective functions, optimal values are 48.69% and 43.98 $/h, respectively. In scenario b, when the objective functions are freshwater production and the total cost rate, the optimal values reach 731.62 kg/h and 67.42 $/h, respectively. Finally, the optimization outcomes in scenario c show that the hydrogen production and total cost rate reach 9.72 kg/h and 46.11 $/h, respectively.