Thermoeconomic modeling and artificial neural network-based optimization of a decarbonized combined heat and power plant with hydrogen re-electrification

Abstract

This study explores the performance of a decarbonized heat and power plant integrated with a hydrogen production and storage system through thermoeconomic modeling. By employing artificial neural network algorithms along with Genetic Algorithm (GA) and Simulated Annealing (SA) methods, the proposed system is optimized for multiple objectives, including exergy efficiency and costs associated with electricity and heating. The findings reveal that at peak power output from the solar panels, the electrolyzer achieves maximum waste heat recovery, approximately 4.33 kW. Simultaneously, the gas engine decreases its power output to 32.62 kW to minimize energy losses. The heat pump steps in to address any thermal power deficiencies when the gas engine cannot fulfill the thermal load demands. During peak demand periods, the heat pump supplies 50.7% of the total heating needs, while 5.8% comes from waste heat recovery from the electrolyzer, and the remaining 43.5% is fulfilled by the gas engine. The GA algorithm identified the optimal system configuration for exergy efficiency, achieving a value of 44.13% and most cost-effective system configuration with an annual product cost of $136,933 from the viewpoint of multi-objective optimization.