Modeling and machine learning-based optimization of a hybrid biomass/solar-driven polygeneration energy system

This paper presents a novel hybrid energy system that combines biomass and solar energy in a polygeneration configuration to enhance flexibility and overall performance. The system utilizes anaerobic digestion, a solar power tower, a Brayton cycle, steam and organic Rankine cycles, a proton exchange membrane electrolyzer, and Reverse Osmosis (RO) to provide electricity, hydrogen, oxygen, fresh water, and domestic hot water. A comprehensive 4E (energy, exergy, exergoeconomic, and exergoenvironmental) analysis is conducted to assess key performance indicators. The proposed system is optimized using an artificial neural network and the multi-objective grey wolf optimizer, considering seven different scenarios. The optimization results of the best scenario indicate that the energy and exergy efficiencies are 34.5 % and 46.2 %, respectively, with a 9.46 % reduction in CO2 emissions compared to the reference system. At the component level, the greatest exergy destruction was identified in the solar heliostat field and combustion chamber, with values of 14,895 kW and 6,448 kW, respectively, which can be improved in the future. From an economic standpoint, the system is feasible, with a total cost of $582.4/h and levelized costs of electricity and hydrogen of $37.6/MWh and $1.8/kg, respectively. The highest total environmental impact is observed with the RO unit, followed by the solar heliostat field, with values of 7.5 Pt/h and 2.7 Pt/h, respectively.