Comparison of Control Strategies for the Industrial Use of PEM Fuel Cells

"Hydrogen technologies are expected to play a key role in future energy systems by supporting the integration of renewable energy sources and increasing the flexibility of industrial power supply. In this context, efficient supervisory control concepts are required to coordinate the system components while considering technical and economic objectives. This thesis investigates the design of a goal-oriented supervisory control concept for an integrated proton exchange membrane fuel cell (PEMFC) system for industrial power supply. A hybrid energy system consisting of a PEMFC system, a battery, the electrical grid, and an industrial consumer is modeled in MATLAB/Simulink. Two energy management strategy (EMS) approaches are implemented and compared: rule-based (RB) and model predictive control (MPC). While the MPC approach utilizes load forecasts, operational constraints, and an optimization-based objective function, the RB strategy relies on predefined rules. A one-year simulation study is conducted using a standard load profile and time-varying day-ahead electricity prices. In addition to operating costs, PEMFC degradation is considered through a semi-empirical model accounting for operating time and start-stop cycles. Technical, economic, and computational performance indicators are evaluated. The results show that neither EMS is universally superior. The MPC approach achieves higher average PEMFC efficiency, smoother fuel cell operation, and fewer start-stop cycles. However, these benefits are accompanied by increased operating costs and stronger degradation under the applied degradation model. In contrast, the RB approach results in lower operating costs, lower degradation, and significantly lower computational effort, but exhibits higher grid dependency and less coordinated system operation. Overall, the results demonstrate that supervisory control design for integrated PEMFC systems requires balancing technical performance, degradation behavior, economic performance, and implementation complexity. The developed simulation framework enables the systematic evaluation of EMS concepts and provides a basis for future investigations of alternative operating conditions and control strategies to optimize system operation".