Rabat – Hydrogen is increasingly seen as a cornerstone of the global clean energy transition, as it offers abundant, low-carbon power for industries and transport sectors where batteries remain impractical. Yet, a major challenge that persists is safe and efficient storage. Because hydrogen must be stored at very high pressure or very low temperatures, storage tanks experience extreme mechanical stress.
Addressing this issue, Professor Mourad Nachtane from University Mohammed VI Polytechnic has developed an integrated approach that combines physics-based modeling, advanced simulations, and artificial intelligence to design hydrogen tanks that are lighter, safer, and more cost-effective.
In his interview with Morocco World News (MWN), Professor Nachtane shared his ambition to develop cutting-edge digital tools that revolutionize the design of hydrogen storage systems. His goal is to make this complex process faster, smarter, and more accessible for engineers, bridging the gap between research innovation and industrial application.
“A novel integrated framework was developed in this study to revolutionize the design and analysis of lightweight composite high-pressure hydrogen storage tanks. By synergistically combining advanced analytical techniques, sophisticated numerical simulations, and powerful deep learning, we achieved unprecedented accuracy and efficiency in predicting the tank performance,” reads the study.
Understanding the challenge
Modern hydrogen tanks are typically made of layered composite materials, which are lightweight, yet strong. However, these materials are anisotropic, meaning their strength varies depending on fiber orientation. As a result, even small changes in layer angles, wall thickness, or tank diameter can significantly affect stress distribution and failure patterns in the tank.
“As stated by Prof. Nachtane, among the critical design and operational parameters investigated namely internal pressure, composite layer thickness, and tank diameter. Internal pressure emerged as the most dominant factor governing the tank’s structural integrity and overall performance. He further elucidated that the direct mechanical effects of internal pressure on induced stresses and potential failure mechanisms significantly outweigh the influence of minor geometric or thickness variations. Therefore, accurately and comprehensively modeling these complex interactions between internal pressure and the anisotropic behavior of composite laminates is paramount to ensuring the safety and reliability of hydrogen storage systems.
While traditional methods relying on full-scale physical testing or detailed numerical simulations offer high accuracy, they are inherently time-consuming and cost-prohibitive. Nachtane’s research specifically addresses this challenge by aiming to achieve comparable accuracy with a substantial reduction in computational time, thereby enhancing the efficiency of tank performance prediction
Three tools, one workflow
Nachtane’s team combines three complementary methods into a single workflow. First, they use analytical physics, specifically, 3D anisotropic elasticity theory, to understand how each composite layer carries stress in different directions. Second, they apply finite-element to model stress distribution, strain, and failure mechanisms under realistic conditions.
Finally, they incorporate deep learning by training neural networks on extensive datasets generated from these detailed numerical simulations, enabling the AI to serve as a high-fidelity surrogate model capable of generating predictions significantly faster than running new, full-fidelity simulations for every design iteration.
“The idea is to combine these three powerful methods,” Nachtane explained. “Deep learning is crucial in this framework, specifically helping us overcome the high computational cost associated with running such detailed, full-fidelity simulations.”
Teaching the AI to approximate complex physics accurately allows the team to quickly test hundreds, or even thousands, of design variations, which is impossible with traditional methods alone. Such extensive design space exploration is practically unfeasible when relying solely on traditional detailed numerical simulations or physical testing due to prohibitive time and computational costs. Promising designs identified through this AI-accelerated screening process are then rigorously validated using full-fidelity finite-element, complemented by analytical models, to ensure their structural integrity and performance.
To identify the best tank designs, the researchers pair the AI model with an optimization algorithm known as differential evolution. This algorithm evaluates a wide range of potential configurations, retains the best-performing ones, and iteratively refines them until an optimal solution emerges.
This approach allows engineers to explore lightweight yet robust tank structures while maintaining compliance with strict safety standards.
Nachtane’s group is also developing a user-friendly graphical interface so that engineers outside academic settings can benefit from the research. “We want the tool to be accessible to non-specialists,” he said. “It should provide useful insights and recommendations without requiring users to be simulation experts.”
This work represents a foundational step towards the local manufacturing of advanced composite hydrogen tanks made in Morocco. Conducted within the LIMSET department at UM6P, I extend my sincere gratitude to the LIMSET direction for their unwavering support in realizing this ambitious objective. Our deliberate strategy of commencing with a comprehensive numerical approach, rather than immediate experimental fabrication, was driven by the imperative to optimize initial designs, minimize extensive experimental costs, and significantly reduce the time associated with iterative physical prototyping and testing. By leveraging high-fidelity simulations and AI-driven design exploration, we can rapidly identify the most promising and robust tank configurations, thereby streamlining the path toward efficient and safe localized manufacturing.
AI as a complement, not a replacement
While AI plays a central role in accelerating the process, Nachtane insists it cannot replace physical testing. “AI is a tool to simplify and minimize computational cost,” he explained. “But I don’t think AI can replace physical testing … it’s a complement, not a substitute.”
To guarantee accuracy and trust in their results, Nachtane’s team rigorously cross-checks every AI prediction with high-fidelity simulations and analytical models before moving to real-world applications. This careful balance between innovation and safety is crucial when designing high-pressure hydrogen vessels, where even the smallest detail can make a big difference.
Morocco has ambitious plans to become a key producer and exporter of green hydrogen and ammonia, positioning itself as a regional hub for clean energy. “I am convinced that Morocco is among the highly competitive countries capable of establishing itself in the global renewable energy market,” Leila Benali, Morocco’s Minister of Energy Transition and Sustainable Development, told Bloomberg earlier this year.
But, these large-scale projects will require robust storage and transport systems, along with local expertise to design and certify them.
“Our work is market-oriented,” explained Professor Mourad Nachtane. “The goal is to develop products that can be commercialized or used directly by startups.”
By transforming advanced research into practical design tools, Nachtane and his team aim to bridge the gap between academia and industry, empowering Moroccan engineers and entrepreneurs to innovate locally.
Their work has far-reaching benefits. Engineers can leverage faster and more reliable design tools to cut down on prototyping time and costs. Startups can commercialize homegrown hydrogen technologies, strengthening Morocco’s clean-tech ecosystem. Meanwhile, industry stakeholders and policymakers gain data-driven insights for safer and smarter investments. Ultimately, these innovations enhance public safety, ensuring that hydrogen storage systems are both efficient and secure—a cornerstone for Morocco’s clean energy future.
Aligning with Morocco’s vision
Professor Mourad Nachtane’s research focuses on a key challenge for hydrogen energy: storing high-pressure and liquid hydrogen safely and efficiently. His team, working closely with Morocco’s fertilizer and phosphate giant OCP is creating advanced composite materials and new structural designs that could change how hydrogen is stored and moved.
“After that, we plan to expand our efforts to support Morocco’s wider hydrogen and advanced materials industries,” Nachtane explained.
Beyond this, his work is also leading the way in replacing heavy, traditional metal tanks and pipelines with advanced composite structures. Unlike metal, these composites are much lighter, resist corrosion, and can handle extreme pressures. This makes them perfect for storing high-pressure and liquid hydrogen. This strategy offers sustainable solutions for Morocco and Africa’s growing hydrogen sectors, where corrosion and high running costs are significant problems.
Nachtane’s work combines deep scientific study with practical industrial needs. His team uses a blend of physics-based models, proven computer simulations, and AI (Artificial Intelligence) to quickly design composite hydrogen tanks and their parts. This ensures both speed and safety in development.
‘Our main goal,’ he stressed, ‘is to turn our research into useful products that businesses and startups can actually use.’ Through these efforts, Nachtane is helping build the foundation for new, safe, efficient, and locally made hydrogen storage technologies. This will help establish Morocco as a leader in composite innovation and clean energy infrastructure across Africa.”
