A green hydrogen system that looks cheap at pilot scale stays cheap at industrial scale only if it avoids three specific failure points: a scarce catalyst metal, degradation from renewable-driven on/off cycling, and a lack of proof that satisfies project lenders. This piece breaks down each one, then compares how three named electrolyzer companies are actually positioned against them.
Why Cheap at Pilot Scale Doesn’t Mean Cheap at Industrial Scale
Scaling an electrolyzer is not the same problem as building a bigger copy of the same machine. At a few hundred kilowatts, the stack itself is nearly the entire cost story, so a clever design choice reads as an unambiguous win. Stretch that design toward tens or hundreds of megawatts, and the balance of the equation shifts: materials supply chains, balance-of-plant equipment, and financing terms all start carrying more weight than they did at pilot scale. A detail that barely registered on a small budget, such as dependence on a rare catalyst metal, can become the exact reason a much larger project fails to get financed.
The scale of this gap is documented at the industry level. Fewer than 4 percent of the roughly 520 gigawatts of green hydrogen capacity announced worldwide has actually reached construction, according to project-tracking data. That is not a rounding error. It is evidence that the gap between a promising pilot and a financeable industrial-scale project is the rule in this industry right now, not the exception.
Do you know what it actually takes for an electrolyzer to skip batteries and connect straight to renewable output? It’s the same pilot-to-industrial gap discussed here, examined specifically through the cycling-tolerance question.
The Materials Constraint: Iridium and PEM
Alkaline and PEM electrolyzers were engineered for steady, grid-connected, baseload-style power. Running them on cheap but genuinely variable solar or wind exposes a materials constraint that a pilot-scale test rarely surfaces.
PEM electrolyzers require iridium catalysts. Global iridium output runs only around 7 to 8 tonnes a year. That figure functions as a structural ceiling on how much PEM manufacturing capacity the industry can realistically build out this decade, independent of how large order books get. A vendor’s cost roadmap that assumes PEM capacity scaling well past what current iridium supply supports has a gap in it that no amount of pilot-scale success will close.
The Cycling Constraint: Membranes and Intermittent Power
Both alkaline and PEM architectures rely on a membrane to keep hydrogen and oxygen separated. Steady baseload power barely stresses that membrane. The constant on/off switching that solar and wind output create is a different kind of stress entirely, and it accelerates membrane degradation in ways published specifications typically don’t fully account for.
A design that handles this cycling poorly is left with two options, and both are expensive. It can add batteries or grid backup to smooth its power input, which reintroduces cost the design was supposed to eliminate. Or it can run at reduced capacity to protect the membrane, which quietly raises the real cost per kilogram of hydrogen it produces regardless of how cheap the input electricity is.
For more on how this cost pressure compounds at utility scale, see The Line Item Utility-Scale Hydrogen Developers Keep Underestimating, which looks at the balance-of-plant side of this same scaling problem.
The Financing Constraint: Why Bankability Decides the Outcome
Materials supply and cycling tolerance are engineering constraints that a design team can plan for. Capital is the constraint that actually ends projects. With fewer than 4 percent of announced capacity reaching construction, lenders financing industrial-scale green hydrogen are not accepting a pilot’s cost projection at face value. They want evidence a design performs under real operating conditions over multiple years, not a favorable number from a small demonstration.
A system that cannot cycle cleanly with renewable output either needs added buffering hardware, which raises capital cost, or accepts lower effective output, which raises cost per kilogram anyway. Both outcomes undercut the low-cost case that justified the project in the first place, which is exactly why bankability, rather than any single engineering choice, ends up deciding which designs actually scale.
How Three Companies Compare on These Constraints
The table below compares three electrolyzer developers on the three constraints described above: platinum-group metal dependence, how each architecture handles renewable-driven cycling, and how far each has proven its design beyond pilot scale.
| Company | Platinum-Group Metal Dependence | Cycling Tolerance Approach | Proof Beyond Pilot Scale |
|---|---|---|---|
| H2Pro (Decoupled Water Electrolysis) | None — bi-functional nickel-based electrode | No membrane; hydrogen and oxygen produced at separate times, designed to switch on and off without degradation (self-reported) | Pre-commercial: 0.5 MW pilot in Israel; 5-to-50 MW off-grid demonstration underway in Spain |
| Hysata | None — capillary-fed alkaline design | Conventional alkaline separation; cycling-tolerance data at industrial scale not yet public | Early commercial stage, backed by a $111 million funding round |
| ITM Power | Yes — PEM requires iridium catalysts | Membrane-based PEM; exposed to the cycling-driven degradation described above | Commercially established manufacturer with a multi-year deployment history |
Reading this table, the pattern is that no company has cleared all three constraints simultaneously. H2Pro’s architecture avoids both the materials and cycling constraints on paper, but the company is still pre-commercial, so its claims haven’t been independently verified at industrial scale. Hysata avoids the materials constraint through its platinum-group-metal-free design but hasn’t yet published cycling-tolerance data at the scale this piece is about. ITM Power carries real commercial history, which matters to lenders, but that history comes with direct exposure to the iridium supply ceiling.
What Buyers Should Actually Ask a Vendor
Given how easily a pilot-scale result can mask an industrial-scale problem, a structured set of questions matters more than a single cost projection. Four are worth asking of any electrolyzer vendor before treating a low-cost claim as decided:
- What is the design’s platinum-group metal exposure, and at what price and supply risk? A roadmap that depends on iridium scaling faster than global mine output supports is a roadmap with an unpriced risk baked in.
- What cycling-tolerance data exists at a scale close to the proposed project, not just at pilot scale? A membrane’s degradation curve under constant on/off switching rarely shows up cleanly in a short pilot run.
- How does the design’s balance-of-plant cost change as module size increases? This is a separate question from the stack’s headline efficiency number, and it is where a lot of the real industrial-scale cost lives.
- What operating history exists beyond the vendor’s own pilot? Independent, third-party performance data carries more weight with lenders than a vendor’s internal projection, and it should carry more weight with buyers too.
None of these questions require specialized expertise to ask. What they require is treating a vendor’s pilot-scale cost claim as a starting hypothesis rather than a settled fact, which is a distinction that becomes expensive to ignore only after a project is already under construction.
Bottom Line
A low-cost green hydrogen system stays cheap at industrial scale only if its design clears three specific hurdles: it avoids a scarce catalyst metal, it tolerates renewable-driven cycling without expensive buffering hardware, and it has enough operating history to satisfy the lenders who ultimately decide whether a project gets built. None of the three companies compared here has cleared every hurdle at once, which reflects the state of the industry more than it reflects any single company’s shortcoming. Buyers and developers evaluating electrolyzer vendors should ask which of these three constraints a given design actually avoids, and treat any claim about the remaining ones as unproven until deployment data says otherwise.
FAQ
Q: Why doesn’t a cheap electrolyzer design at pilot scale stay cheap at industrial scale?
A: At pilot scale, the stack accounts for nearly the entire cost, so a clever design choice looks like an unambiguous win. At tens or hundreds of megawatts, materials supply, balance-of-plant equipment, and financing terms carry more weight, and a detail that barely mattered at kilowatt scale, like dependence on a scarce catalyst metal, can become the reason a much larger project doesn’t get financed.
Q: What is the materials constraint facing PEM electrolyzers at scale?
A: PEM systems depend on iridium catalysts, and global iridium output is only around 7 to 8 tonnes a year. That is a structural ceiling on how much PEM manufacturing capacity the industry can build this decade, regardless of order volume.
Q: How does H2Pro’s Decoupled Water Electrolysis address the materials and cycling constraints?
A: DWE replaces the membrane and platinum-group metal catalysts used in conventional alkaline and PEM designs with a bi-functional nickel-based electrode, producing hydrogen and oxygen at separate times instead of splitting them simultaneously across a membrane. That removes exposure to both the iridium supply constraint and membrane-related cycling degradation, though the technology is still pre-commercial and these claims haven’t been independently verified at industrial scale.
Q: How does H2Pro’s approach compare to Hysata’s and ITM Power’s?
A: Hysata’s capillary-fed alkaline design also avoids platinum-group metals but hasn’t published industrial-scale cycling data. ITM Power has real commercial deployment history but is directly exposed to the iridium supply ceiling because it builds PEM systems. H2Pro’s architecture avoids both constraints on paper but remains the least commercially proven of the three.
Q: What ultimately decides whether a low-cost green hydrogen project gets built at industrial scale?
A: Financing, more than any single engineering decision. With fewer than 4 percent of announced global capacity reaching construction, lenders are underwriting designs with demonstrated performance under real cycling conditions over multiple years, not a favorable cost projection from a small-scale pilot.