Optimization of a proton-conducting solid oxide fuel cell-based energy system using machine learning techniques

The growing need for clean energy and effective waste management highlights the importance of integrated systems like waste gasification combined with fuel cells, offering a sustainable solution to reduce emissions and convert waste into useful energy. This study presents a sustainable energy system based on a proton-conducting solid oxide fuel cell, in which the required fuel is supplied by gasifying municipal solid waste. The main goals are to maximize the net power output and minimize carbon dioxide emissions. The performance of the system was evaluated under varying operating conditions, including current density, inlet temperature, and fuel utilization ratio. A thermodynamic model of the system was developed using an engineering equation solver, and its accuracy was validated by comparing the results with data from previous studies. The comparison showed good agreement, confirming the reliability of the model. To further analyze the system, machine learning techniques were used to create regression models that predict the outputs based on the input parameters. These models helped examine the combined influence of the operational variables and supported a multi-objective optimization approach. The optimization results showed that higher current densities generally lead to increased power output. At high current densities, increasing the inlet temperature significantly raises CO2 emissions, which may rise from about 1085 kg/MWh to nearly 4468 kg/MWh. In contrast, when the system operates at current densities below 3500 A/m2, carbon dioxide emissions remain in a lower and more stable range (between 500 and 800 kg/MWh), regardless of the fuel utilization ratio. The optimal operating point for the system was found at a current density of 5798 A/m2, an inlet temperature of 800 °C, and a fuel utilization ratio of 0.80. Under these conditions, the system generates a net power output of 315.3 kW, while emitting 1001.4 kg of carbon dioxide per megawatt-hour of electricity produced. These findings highlight the potential of combining waste-to-energy technologies with advanced fuel cell systems to achieve both high efficiency and reduced environmental impact.