Components vs systems
Hydrogen is often positioned as a new energy vector, but much of the engineering behind it is built on familiar ground. Across production, compression, storage, and dispensing, systems are assembled from established components – valves, fittings, instrumentation – and designed according to conventional process-industry principles
At first glance, this seems entirely reasonable. Individual components are increasingly well-qualified for hydrogen service, with genuine advances in material selection, sealing technology and high-pressure performance. Certification standards have evolved, and manufacturers have responded with more rigorous testing and validation.
But optimising individual components does not automatically deliver reliable systems. That distinction matters more with hydrogen than with almost any other medium.
Hydrogen’s small molecular size makes leakage more likely than with larger molecules. High-pressure cycling – routinely in the 350 to 1378 bar range – places sustained, repeated demands on sealing and structural integrity. Embrittlement is not a single-variable problem; it is a function of stress, environment and material condition interacting over time. Under these circumstances, system performance cannot simply be read off from the sum of its parts.
Focus on the interface
This is most visible at the interface level. The connections between components – threads, seals, fittings – represent the majority of potential leak paths in any hydrogen system. Yet interfaces are routinely treated as secondary engineering concerns, even as system complexity grows and the number of those interfaces multiplies. More components mean more boundaries, and more boundaries mean more points where tolerance, alignment and sealing must all be managed correctly, every time.
The deeper problem is that hydrogen systems are often assembled rather than designed. Modular architectures offer real flexibility, but they can also lead to overcomplexity. Every additional component introduces not just a function but a boundary – another point of vulnerability to vibration, thermal cycling, and the ordinary variability of real-world installation.
A different approach is gaining ground, one that starts with architecture rather than component selection. The goal is to reduce the number of interfaces, integrate functions where possible, and simplify flow paths throughout. In practice, this frequently leads toward manifold-based designs, where multiple functions are consolidated into a single block. Fewer discrete components and connections mean fewer potential leak paths, more consistent assembly and measurably greater system robustness.
Sealing philosophy is shifting in the same direction. Traditional compression-based sealing relies on sustained mechanical loading applied at assembly – a condition that is difficult to verify and degrades over time. Pressure-energised sealing arrangements, by contrast, improve performance as operating pressure increases, making the seal more effective precisely when it matters most.
Defining what ‘performance’ means is itself changing. Hydrogen systems operate under variability – pressure cycles, transient states, long-term material interaction – that static component specifications were never designed to capture. The relevant metric is increasingly resilience: whether a system can tolerate and adapt to changing conditions, not just whether it passes a qualification test.
The question is not whether conventional approaches work. Many do. The question is whether they scale – and whether the assumptions embedded in them are appropriate for infrastructure that must operate reliably across decades and millions of cycles.
For more information visit: jbv.co.uk
Contributed by Jonathan Balmforth from JB Valves.
