Project Summary
This project will develop AI-assisted safety tools to support the safe design and operation of liquid hydrogen (LH2) bunkering infrastructure in UK maritime ports. By combining high-fidelity simulations from CryoSMART—a specialist in cryogenic thermofluids—with physics-informed machine learning, the tools will rapidly predict dispersion, ignition risk and safe zoning under realistic release scenarios. The solution will accelerate the safe adoption of hydrogen fuel in ports and inform future regulations, codes, and safety standards.
Project Achievements
This project developed an AI-assisted safety assessment framework to support the safe deployment of liquid hydrogen (LH₂) bunkering infrastructure in maritime environments. The work combined high-fidelity computational fluid dynamics (CFD) simulations with machine learning techniques to accelerate hazard prediction for cryogenic hydrogen releases. The project successfully demonstrated a prototype digital safety tool capable of rapidly evaluating explosion risks associated with hydrogen leaks in representative port scenarios. The modelling framework captured key physical processes including cryogenic dispersion, ignition behaviour, flame propagation, and pressure wave generation. The project also engaged with academic experts, regulators, and industry stakeholders through an advisory panel and technical workshops, ensuring that the research remained aligned with real-world operational and regulatory needs. These outcomes demonstrate the potential of AI-enabled digital modelling tools to support early-stage safety assessment and infrastructure planning for hydrogen-based transport systems.
Conclusions
The project demonstrates that AI-assisted digital modelling can significantly enhance the speed and accessibility of safety assessments for liquid hydrogen infrastructure. By combining physics-based simulations with machine learning, the developed framework can reduce the time required to evaluate complex hydrogen release scenarios while maintaining a high level of physical fidelity. The results highlight the importance of integrating advanced modelling tools into early infrastructure planning and risk management processes. Such tools can help regulators, infrastructure developers, and operators better understand potential hazards and design safer hydrogen systems. Beyond the maritime sector, the methodologies developed through this work are transferable to other hydrogen applications including aviation, rail transport, and hydrogen-enabled energy systems. Overall, the project demonstrates how digital engineering approaches can support the safe and scalable adoption of hydrogen technologies within the UK’s net-zero transport transition.
Next Steps
The next phase of the project will focus on further developing the prototype into a more comprehensive digital safety assessment platform for hydrogen infrastructure. This will include expanding the scenario library, improving model automation, and integrating additional machine learning capabilities to enable faster hazard prediction. Future work will also involve validation against experimental datasets and engagement with infrastructure developers and regulators to ensure the tool meets operational and certification requirements. The research team is currently pursuing follow-on funding to extend the concept toward a hydrogen ecosystem digital twin framework capable of supporting safety analysis across multiple transport sectors. In parallel, the modelling methodologies are being applied to related projects in hydrogen aviation, rail transport, and built-environment safety, demonstrating the broader potential of the approach. These developments aim to move the concept toward higher technology readiness levels and real-world deployment.
