Water management in highly dynamic PEM fuel cell system operation

Abstract

Polymer ElectrolyteMembrane Fuel Cells (PEMFCs) are a promising technology to achieve carbon-free mobility. However, multiple challenges must be overcome for a market breakthrough. In a fuel cell dominant vehicle, where only a small hybridization battery is installed, highly dynamic operation of the Fuel Cell System (FCS) is required: A driver expects power dynamics on a time scale of 1 s, which must be predominantly provided by the FCS. In addition, the system efficiency as well as the lifetime must be kept high.

This work subsequently aims at improving the dynamic capabilities of FCSs, while maintaining high system efficiencies and avoiding lifetime-limiting critical states. Therefore, the understanding of the key processes’ dynamics, proceeding on time scales spanning multiple orders of magnitude, is fostered. Special focus lies on the water management as cornerstone of highly dynamic and efficient operation. In contrast to the broad research published, this work features two insufficiently addressed fields. On one hand, highly dynamic operation on the time scale of 1 s is in focus. On the other hand, the subsystems with their own intrinsic dynamic limitations are also included in the scope of this work. As core contribution of this work, a suite of dynamic experiments is carried out with a sophisticated full-scale automotive fuel cell system.

Four research questions serve as a guidance throughout this work. First, the dominant dynamic processes are identified using targeted dynamic simulations and experiments. Second, the setup and potential pitfalls of dynamic experiments are described, as well as their contribution to advanced modeling. Third, stationary-based control strategies are evaluated, limitations are discussed and improved strategies for highly dynamic operation are developed. Lastly, the impact of two central degrees of freedom on system level, control strategy and system architecture, on efficiency and dynamics is discussed. Overall, the setup of this work follows the logic from one fuel cell to a fuel cell system to other fuel cell systems. Cell-internal processes are explained first, then the stack/system interplay is discussed and lastly the results are transferred onto other systems.

Membrane humidification and dry-out dynamics strongly impact the dynamic stack behavior. In addition, liquid water accumulation and drainage plays an important role, as liquid water limits the cross-sectional area available for reactant diffusion. In case of a highly dynamic power increase, too much liquid water quickly results in reactant starvation, while the same amount of liquid water is allowed during low load steady-state operation. As common short stack test rigs fail to resemble the dynamic capabilities of an automotive fuel cell system, the transferability of the results must be checked. The highly dynamic system operation is limited by the Air System (AirS). Changing the AirS architecture or control strategy to achieve higher dynamic capabilities worsens the system efficiency, resulting in a trade-off between the two conflicting goals.