An Innovative Approach for Component-Based Power Consumption Prediction of Balance of Plant in Fuel Cell Systems

Abstract

In this study, a new semi-empirical modeling approach was developed to predict the power consumption of the balance-of-plant (BoP) and auxiliary subsystems of fuel cells, namely the compressor, fans, and pumps. Unlike regression-based methods commonly employed in the literature, this study proposes a single semi-empirical equation that is physically interpretable, with its parameters estimated for the first time using one of the heuristic optimization techniques, the Coati Optimization Algorithm. The proposed approach was validated by comparing experimental data with model outputs, and performance evaluation was conducted based on the Sum of Squared Errors criterion. According to the optimization results, the SSE values were calculated as 0.0557, 0.0540, 0.0736, and 0.0293 for the total BoP, compressor, fans, and pumps, respectively. Although the proposed model for BoP produced higher error values compared to existing studies in the literature, it demonstrated a physically meaningful, realistic, and system-level applicable structure. Furthermore, for the first time in the literature, separate semi-empirical models were formulated for the compressor, fans, and pumps, and the results were shown to be consistent with experimental data. Overall, the application of the Coati Algorithm revealed that parameter estimation is effective in enhancing both the accuracy of semi-empirical models and their physical interpretability. Ultimately, this study provides a new methodological framework and contributes to the performance analysis and optimization of fuel cells in the energy sector. Thus, the present work not only addresses a gap in the current literature but also lays the groundwork for future physics-based models and artificial intelligence-driven optimization studies. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.