Integration Of Control & Monitoring Tools And Strategies For Improved Fuel Cell System Durability & Reliability

Expected Outcome:

The transition to sustainable mobility is accelerating the adoption of hydrogen-based powertrains, integrating fuel cells, batteries, electric motors and power electronics. In particular, fuel cells (low and high temperature) represent a very promising technology for decarbonising various sectors of mobility, including maritime applications, heavy-duty vehicles and aeronautics. However, remaining aspects that hamper the large-scale diffusion of Fuel Cell Systems (FCS) include their limited lifetime and operational reliability. Moreover, higher fuel cost puts even more stringent requirements and emphasis on system efficiency.

To increase the useful life and reliability of FCS, not only improvements in material and designs are required, but also on control, diagnostic, failure prediction and energy management strategies. It is thus key to integrate advanced prognostic and health management strategies into real operating FCS, suitable not only to enable the detection of anomalies, but also the real-time assessment of the current state of health, developing and integrating proper adaptive recovery strategies to mitigate FCS components’ deterioration. These considerations are generally valid for different technologies (PEMFC, SOFC…), which need to be optimally managed: all the components, both in the stack and Balance of Plant (BoP), should be continuously monitored to maximise the FCS lifetime, reliability as well as efficiency.

Project results are expected to contribute to all of the following expected outcomes:

• Improvements of control and energy management that enable increased lifetime and operational reliability at system level, while retaining high system efficiency;
• Providing new product ideas and solutions addressing monitoring and diagnostics of FCS, including hardware and software;
• Enhanced durability and reliability of fuel cell systems, including Balance-of-Plant (BoP) components, through predictive maintenance and operational optimisation;
• Development of specific hardware and software platforms, with modular and adaptable architecture enabling application across multiple FCS technologies, to implement diagnostic methodologies for identifying technology-related issues;
• Developing a European open-source platform for sharing data, simulation models, and strategies for control and diagnostics of different FCS technologies (PEMFC, SOFC…) in maritime transport, heavy-duty vehicles, and aviation;
• Building confidence in FC technology for all of the mobility applications, thus accelerating the market uptake;
• Supporting manufacturers, researchers, system integrators, and policymakers and promote technological sovereignty, industrial innovation, and decarbonisation in Europe.

Project results are expected to contribute to the following objectives and KPIs of the Clean Hydrogen JU SRIA:

• Reduction of FCS OPEX through predictive maintenance and improved fault detection coupled with fault mitigation;
• Improvement in dynamic operation and efficiency of the FCS, with high durability and reliability, especially when operating dynamically following an established load profile (representative of the application considered);
• Development of tools and methods for monitoring, diagnostics and control of fuel cell systems;
• Increase in FCS durability with regards to the established KPI targets (depending on the selected application) for the End-of-Life (30,000 hours for heavy duty vehicles, 80,000/100,000 hours for marine applications or 50,000 hours for aeronautics);
• Increase in Mean Time Between Failures (MTBF) of critical components and FCS;
• Increase in lifetime of the fuel cell stack or Balance-of-Plant (BoP) components by 15%, directly attributable to the implementation of predictive maintenance and anomaly mitigation strategies, supported by diagnostic tools capable of identifying and isolating at least 95% of degradation mechanisms and malfunctions under representative operating conditions;
• Contribute to increased deployment and adoption of fuel cell technologies by demonstrating long-term system robustness;
• Enable cross-sectoral applicability of solutions across different fuel cell technologies (e.g., PEMFC, SOFC) and fuel types.

Scope:

The scope of this topic is on the development, integration and validation of robust online monitoring and control algorithms for FCS, targeting prediction and mitigation of degradation, faults and failures throughout the system lifetime. Focus shall be on Balance of Plant (BoP) components rather than fuel cell stack. Activities should combine experimental testing, digital modelling and may use AI-based techniques to develop and validate predictive tools for health management. Proposals should clearly explain how they go beyond the achievements of previous projects (namely Ruby [1] and Giantleap [2]), identifying specific technical or methodological advances.

Proposals should address the following aspects:

• Develop innovative control, monitoring and diagnostic strategies to enhance FCS durability and reliability, leveraging predictive tools;
• Contribute to the development of standardised protocols for State of Health (SoH) diagnostics and components interoperability in FCS;
• Integrate developed strategies in an open framework, with adaptable hardware/software platforms as core systems for diagnostics and prognostics, adaptable to different types of FCs and other components of the powertrain (i.e., battery, motors, etc.)
• Identify and quantify how BoP components influence FCS durability and reliability, and integrate measures to mitigate degradation and FCS fault and failures;
• Improve the FCS reliability (e.g. increasing Mean Time Between Failures, MTBF) through approaches such as predictive maintenance and condition monitoring applied to key components, using additional (e.g. vibration) or standard (e.g., current/voltage, temperature measurements) or virtual sensors (e.g., estimation of local compositions/conditions);
• Use of artificial intelligence and data-driven methods is encouraged, especially when combined with physics-based modelling. Physical modelling can be used both to develop real-time diagnostic/prognostic strategies and to implement reference models that justify the adoption of specific diagnostic/prognostic strategies;
• As far as possible, all tools and methodologies developed should be made available through open data infrastructure, using common/open data formats and providing interfaces for data access and visualisation.
• Regardless of the type of diagnostic/prognostic algorithm developed (model-based or data-driven), these should be validated using experimental data obtained from real demonstrators. As in previously funded projects such as Helenus [3], experimental data could be obtained from small-scale components (down to low power levels, around 5–25 kW), which may be subjected to specific duty cycles and stress tests (corresponding to at least 1000 hours of accelerated aging tests, run in a laboratory environment or in real vehicles).
• Development of two power system demonstrators integrating FCS, software/hardware for diagnostic/prognostic functionalities and power electronics. Each demonstrator should be based on a relevant FC technology (e.g., one based on a PEMFC and the other on a SOFC system, with single or multiple stacks), with a power output representative of at least 100 kW and taking into account environmental and duty cycle constraints coming from potential OEMs involved. If aging and malfunction models are available (validated on real components of a specific FCS technology), one demonstrator could be replaced by a Hardware-in-the-Loop (HiL) system. In this case, diagnostic/prognostic methodologies should be run on a dedicated real-time target (the same used to monitor the real application);
• Demonstration of the versatility of the methodologies for different fuel cell technologies and with different fuels (when applicable); The demonstrators should be tested under representative duty cycles and operating conditions, providing sufficient evidence to support the achievement of TRL6.
• Creation of an open-access, anonymised datalake, integrating both experimental and high-fidelity simulated datasets (e.g., from digital twins), organised with open standards and anonymised metadata, to enable reproducible research, AI-based diagnostics training, and broad industrial adoption.

Projects should build on the results obtained in previous projects funded in this area of research, such as Ruby^[1], GiantLeap^[2], Helenus^[3] or Virtual-FCS^[4], developing and testing (on real systems) innovative methodologies that can be adapted to different FCS technologies.

Furthermore, depending on the application addressed, synergies with the relevant partnerships such be explored, e.g EU-Rail JU (rail), 2ZERO Partnership (road) or Zero Emission Waterborne Transport Partnership (waterborne).

For activities developing test protocols and procedures for the performance and durability assessment of fuel cell components proposals should foresee a collaboration mechanism with JRC^[5] (see section 2.2.4.3 "Collaboration with JRC"), in order to support EU-wide harmonisation. Test activities should adopt the already published EU harmonised testing protocols^[6] to benchmark performance and quantify progress at programme level.

For additional elements applicable to all topics please refer to section 2.2.3.2

Activities are expected to start at TRL 4 and achieve TRL 6 by the end of the project - see General Annex B.

The JU estimates that an EU contribution of maximum EUR 4.00 million would allow these outcomes to be addressed appropriately.

Technology Readiness Level - Technology readiness level expected from completed projects

[1] https://cordis.europa.eu/project/id/875047

[2] https://cordis.europa.eu/project/id/700101

[3] https://cordis.europa.eu/project/id/101056784

[4] https://cordis.europa.eu/project/id/875087

[5] https://www.clean-hydrogen.europa.eu/knowledge-management/collaboration-jrc-0_en

[6] https://www.clean-hydrogen.europa.eu/knowledge-management/collaboration-jrc-0/clean-hydrogen-ju-jrc-deliverables_en