The current paper aims to develop a multi-heat integration structure for a solid oxide fuel cell, focusing on methods that reduce thermodynamic irreversibility and address environmental concerns. Hence, the suggested method comprises a bi-evaporator refrigeration-organic flash cycle, a water electrolyzer cycle, a reverse osmosis cycle, and a Claude cycle producing electricity, cooling load, and liquefied hydrogen simultaneously. Furthermore, intelligent data-driven study/optimization focusing on thermodynamic, environmental, and economic aspects are performed to highlight potential areas for enhancement. Hence, two different multi-objective scenarios using a detailed sensitivity analysis are defined. Accordingly, artificial neural networks are developed for learning and verifying objectives related to energetic and exergetic performances, the cost of liquefied hydrogen, and the reduction of CO2 emissions. Subsequently, a multi-objective grey wolf optimization is used in energy-cost-environmental and exergy-cost-environmental scenarios. The results reveal a significant sensitivity index of 0.619 for fuel cell operating temperature. Notably, the first scenario provides the most appropriate optimization way, showing an energy efficiency of 62.91%, a liquefied hydrogen cost of 3.177 $/kg, and a CO2 emission reduction of 101.9 kg/MWh. Also, an exergy efficiency of 45.42% and a payback time of 2.45 years are the other notable findings.
Read full abstract