What purity does the standard require, why is it so hard to maintain, and what does a specialised laboratory contribute? Here we answer the questions operators, planners and buyers ask us most often.
Fuel cells react sensitively to impurities. Even traces of water, sulphur compounds, carbon monoxide, ammonia, oils or particles can reduce performance and permanently damage the cell stack. Limit values are therefore in the range of single parts per billion – an order of magnitude that can only be demonstrated by measurement.
The analysis demonstrates whether the hydrogen meets the requirements of ISO 14687 and DIN EN 17124 for vehicle applications. For operators this evidence matters twice over: it safeguards reliable vehicle operation – and it documents the quality delivered should a damage claim arise.
Since 2021 we have specialised in hydrogen analytics to ISO 14687 and DIN EN 17124. In Berlin we operate one of the most advanced laboratories in Europe for this purpose; with our two mobile laboratories we test hydrogen purity on site across Europe.
We take care of sampling as well: we provide professional sampling kits and suitable sample containers and, on request, handle the entire transport logistics. For production sites we also plan and implement process monitoring systems that monitor gas quality 24/7.
Between 2022 and mid-2026 we produced 804 test reports for around 80 clients. 732 of these samples can be assessed against ISO 14687 – and 44.7 % of them failed at least one limit value.
That figure should be read with care. It is not an average of European hydrogen but the result of a testing laboratory’s work: we are often called in when a problem is already suspected – and then repeatedly, until it is resolved. What carries meaning is therefore less the absolute level than the comparison between installation types, because the same laboratory applied the same methods throughout.
0 %60 %
732 samples assessable against ISO 14687, out of 804 test reports, 2022 to mid-2026.
The pattern is unambiguous: the further from production a sample is taken, the more often it falls outside specification. At electrolysis plants it is 21.4 %, at trailers 53.6 %. The hydrogen leaving the electrolyser is in specification – what arrives after transport and interim storage is a different product.
Across all installation types the most frequent cause is water. At trailers, nitrogen (28.8 %) and oxygen (19.2 %) are added to it; the ratio between the two points to residual air. The catalyst poisons – carbon monoxide, sulphur, ammonia, halogenated compounds – stay below three percent almost everywhere. They are rare, but they damage the fuel cell irreversibly.
Share of non-conforming samples at hydrogen refuelling stations. The 2022 value is based on 19 samples, the 2025 value on more than 270. The year 2026 is not yet complete.
And the encouraging side: at hydrogen refuelling stations the share of non-conforming samples fell from 86 % in 2022 to 3 %, while the number of samples grew strongly. This is not proof – the client base also changed over the years. But the direction is unmistakable and matches what operators tell us: once the critical points are known and checked regularly, the rate goes down.
The problem is therefore manageable – it is not a fundamental limit of the technology. It does not resolve itself, though, and continuous measurement of every component at every transfer point is economically impractical. What remains is a risk-based monitoring concept: measure where the risk is.
Your question is not listed? Write to us – we answer personally and are glad to add the question here.