Skip to Content
HydrogenHydrogen
  • This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
  • Article
  • Open Access

26 September 2026

19 Pages

Influence of HRS Parameters During Parallel Sampling Events for Determining Hydrogen Fuel Quality

,
,
,
,
,
,
,
,
and
1
National Physical Laboratory (NPL), Hampton Road, Teddington TW11 0LW, UK
2
ZBT, Zentrum für BrennstoffzellenTechnik, Carl-Benz-Str. 201, 47057 Duisburg, Germany
3
Czech Metrology Institute (CMI), Okruzni 31, 638 00 Brno, Czech Republic
4
SINTEF Industry, Richard Birkelandsvei 2b, 7034 Trondheim, Norway

Abstract

This study investigates how hydrogen refuelling station (HRS) operating parameters influence the representativeness of hydrogen fuel samples collected during parallel sampling events. Parallel sampling methodology is regularly used in Europe because it can be performed during normal vehicle refuelling with minimal disruption, but uncertainty remains regarding the effects of refuelling protocol, flow profile, or storage-bank usage. Experiments were conducted at the ZBT test HRS using the HySaM sampling system and at operating European HRS using the NPL ParSAM. Controlled comparisons were performed between several flow-rate profiles, while comprehensive impurity analysis was performed against ISO 14687:2025 requirements. Under controlled conditions, the refuelling protocol (excluding temperature and storage bank) and flow-rate profile had limited influence on most contaminants, whereas water showed the greatest variability and requires further investigation. However, repetitive sampling at normally operating HRS revealed larger differences for several species, including nitrogen, oxygen, argon, methane, carbon dioxide and sulphur, indicating that factors beyond protocol parameters tested in this study and nominal flow rate, such as storage-bank switching, flow path, and station hardware, may affect measured composition. Modelling indicates that the filling rates of the sampling cylinder and FCEV may diverge depending on the sampling-system strategy, highlighting the need for further improvements. The results show that parallel sampling may represent the fuel delivered during a specific refuelling event (if the filling rate of the cylinder and FCEV are sufficiently similar). However, parallel sampling may not fully represent the overall HRS and all operating conditions. Further research involving a broader range of HRS technologies and operating conditions may lead to validation and standardisation of the finding and possibly dedicated sampling protocols.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.