Hydrogen use is a key component of decarbonization, but ensuring the long-term operation of electrolysers and fuel cells is challenging. Fluctuating operating conditions lead to wear and performance losses. Fraunhofer IFAM has now achieved a decisive breakthrough with dynamic impedance spectroscopy: This diagnostic method enables precise analysis during operation, detects overloads in real time, and identifies weak points at an early stage. This allows for the targeted optimization of resilience and efficiency, extends the service life of the systems, and reduces operating costs – a technological leap with enormous potential for the hydrogen economy.

Material aging and performance degradation pose technical challenges for electrolysers and fuel cells. Varying operational loads, such as those caused by fluctuating renewable energy feed-in, can lead to significantly shortened service lives and high maintenance costs. To address this issue sustainably, researchers at Fraunhofer IFAM have made decisive advancements in the real-time diagnostic system for dynamic impedance spectroscopy.

Real-time analysis: From laboratory measurements to AI-supported industrial applications

State-of-the-art measurement methods require an interruption of operation with corresponding downtime to reach a steady state – such as classical impedance spectroscopy. These measurements are time-consuming and can hardly reflect the actual, fluctuating operating conditions. “With our new dynamic measurement method, we are obtaining meaningful data directly from the actual running process for the first time, using significantly higher current levels than in the previous laboratory setup,” says Dr. Hermann Pleteit, project manager at Fraunhofer IFAM.

Dynamic impedance spectroscopy is a new method that allows the internal state of an electrolyzer or fuel cell to be examined under non-steady-state conditions during operation. To do this, a small multi-frequency signal is superimposed to the normal operating current. Simultaneously, it is measured how the system reacts to this – i.e., how current and voltage change. The ratio of these quantities is used to calculate what is known as complex impedance, which can be understood as the system’s frequency-dependent “electrical resistance.” “It is precisely this frequency-dependent nature that makes impedance so valuable for the applications mentioned and so interesting to us. Unlike simple resistance value, it provides a comprehensive picture of the processes occurring inside an electrochemical system. This is because the various physical and chemical processes – such as the reactions at the electrodes, the transport of ions through the membrane, electrical contact resistances, or the replenishment of the substances involved – each have an effect at different frequencies. “This allows us to distinguish these processes from one another, evaluate them individually, and, thanks to the detailed real-time resolution, see the state of the individual components, such as the membrane or electrodes,” Pleteit explains.

Specific conclusions about the condition and aging of the system can be drawn from the temporal evolution of these spectra. If, for example, a certain component of resistance increases, this may indicate deteriorated electrode reactions, corrosion, or deteriorating electrical contacts. Shifts in phase behavior, on the other hand, may indicate membrane wear, problems with water balance, or restricted mass transfer. This makes it possible to identify the actual causes of performance losses – whether due to materials, interfaces, contacts, or operating conditions.

A real-time diagnostic system uses this information to continuously monitor the cell’s condition. This allows for the detection of incipient changes long before they lead to noticeable performance losses or even a failure. “Instead of reacting only after the system has already shut down, we can precisely predict maintenance needs and avoid costly downtime,” Pleteit continued. Dynamic impedance spectroscopy thus enables accurate condition assessment, targeted fault diagnosis, and predictive maintenance of electrolysers and fuel cells.

The basis for these advances is the scaling up of dynamic impedance spectroscopy from the laboratory scale to real-world application dimensions, so that operating currents of up to 30 amperes can now be superimposed on the measurement signal. This represents a decisive step toward industrial application. Through integrated online data processing, the temporal evolution of the impedance is calculated and displayed in real time. “This allows us, for the first time, to measure our mathematical parameters for larger systems during ongoing processes. In combination with AI-based models, the results can be applied to changing environmental conditions – such as pressure and temperature fluctuations – as well as to new material compositions,” Pleteit summarizes.

Integration into plant control systems and practical benefits

The practical advantage of the expanded system lies in its direct implementation into plant control systems, such as the energy management system. Operators of hydrogen systems – such as PEM fuel cells and alkaline electrolysers – thereby receive continuous, precise status data that reflects the plant’s current state of health. On this basis, critical processes such as catalyst degradation can be detected early, predictive maintenance intervals can be planned, and operating strategies can be dynamically adjusted to minimize degradation processes. “This allows us to significantly reduce lifetime costs and ensure the economic viability of green hydrogen projects – a key component for the competitiveness of the hydrogen economy,” says Pleteit.