21. Mobility and partially confined spaces: Accident at a hydrogen fuel station

To develop the hydrogen-powered vehicles market beyond the limitations of battery electric vehicles a well-structured network of refueling stations is needed at national and international level (across neighbouring countries). As of today H2 fuel stations are very constrained in many countries. In 2021, there were 492 operating hydrogen refuelling stations (HRS) worldwide (Statista, 2022[1]) with most of them (154) located in Japan.

Hydrogen refuelling stations can operate either with liquid hydrogen (LHRS) or with compressed hydrogen (GHRS) that can be produced onsite or be transported (mainly by road transport). Both types of refuelling stations raise specific safety concerns.

An international standard, ISO 19880-1:2020, covers the specifications for outdoor public and non-public fuelling stations that dispense gaseous hydrogen. This ISO standard defines the minimum safety design, installation, commissioning, operation, inspection and maintenance requirements, and, where appropriate, for the performance of GHRS.

In the United States, NFPA-2 code provides separation distances from certain group of exposures for GHRS and recent study by (Hecht and Ehrhart, 2021[2]) revised these distances for LHRS) (National Fire Protection Association, 2023[3]). For most systems the separation distances to most of the groups of exposures based on NFPA-2, were reduced in LHRS compared to GHRS. All separation distances were lower than 30 m.

Most types of the reported incidents in Hydrogen Refuelling Station (HRS)1 involve small leakages of hydrogen or no release at all. The 18.5 % of incidents led to serious consequences, such as fire and explosion. Most of the leakages occurred at the joint parts due to inadequate torque and inadequate sealing. Other causes include design error of the main bodies of apparatuses and human error.

The most severe leak scenario corresponds to a leak size equal to 100% of the pipe diameter connecting the components of the system. However, leaks equal to or less than 0.1% of the flow area of several components (compressors, cylinders, hoses, joints, pipes, valves) are estimated to represent 95% of the system leakage frequency (LaChance et al., 2009[4]). For a 0.1% diameter leak size, the system leakage frequency is 3∙10-2/year and 6∙10-2/year for 20.7 Mpa and 103.4 Mpa systems, respectively. What emerges from these values is that a 0.1% diameter leak would be anticipated during the lifetime of these facilities. Larger and less frequent leak sizes of at least 1% should be used as the basis for separation distances to reduce the likelihood of accidents.

A leak frequency analysis at HRS (Kodoth et al., 2020[5]) using Bayesian and frequentist methods estimated that the leak rate is 0.16/year, 0.20/year and 0.42/year based on the time-based, leak-hole-size, and non-parametric method, respectively. One of the possible solutions is to consider a conservative value for the design, in which case, the leak rate of 0.42/year can be used. The base value selected can be used in design to set performance standards for the availability and reliability in the operation and maintenance of HRSs.

A comparative risk assessment conducted by (Yoo et al., 2021[6]) indicated that the LHRS has a lower risk than the GHRS, but with small differences. Based on another quantified risk assessment, QRA performed by the National Institute for Public Health and the Environment in the Netherlands in 2016 the estimated distances2 to 10-6 per year risk of a single fatality were:

  • 30 m for the liquid hydrogen delivered via a tank and for the gaseous hydrogen dispensing system supplied by pipeline or local production, and

  • 35 m for gaseous hydrogen delivering via tube or cylinder trailer.

For gaseous hydrogen with delivery via pipeline or tube trailer, the risk of single fatality beyond 50 metres was 10-9 per year. For liquid hydrogen supplied by a tanker a risk of single fatality of 10-9 per year was reached at 270 m. However, it should be noted that these distances were estimated with an overly conservative ignition probability of gaseous hydrogen equal to 1. Moreover, the risk contours can be further reduced by the use of proper safety measures.

For hydrogen refuelling stations with an onsite production facility risk studies show that water electrolysis presented lower individual, societal and environmental risk than methane reforming (Dash, Chakraborty and Elangovan, 2023[7]). This is because methane reforming involves other flammable gases such as methane, hydrogen and carbon monoxide. On the contrary, the major safety risk associate with water electrolyser is only the leakage related to the hydrogen produced. Another study (Pan et al., 2016[8]) indicated that compressor is the major risk contributor among HRS elements. Khalil 2017 showed that a small leakage from the compressor is associated with intolerable single death risk frequency, which exceeds both the acceptance criterion at 1.0∙10-4 /year and NFPA’s guideline at 2.0∙10-5/year (Khalil, 2017[9]).

  • Design

    • For on-site hydrogen production, water electrolysis is the recommended production process as it presents lower risk than steam methane reforming.

    • Limit the inventory in the storage facility as low as practical based on the average daily number of fillings of the HRS.

    • Transportation of hydrogen through high-pressure pipelines allows station to dispense fuel without onsite compression and storage and reduce the risk. However, this system should additionally consider the risk of operating high-pressure pipelines in residential areas.

    • A QRA study indicated that liquid hydrogen refuelling stations entail less risk than compressed hydrogen refuelling stations, but the differences were small. Based on that the use of liquid hydrogen instead of compressed hydrogen could be recommended, but further research on that topic is highly advised.

  • Site layout

    • Hydrogen processing systems, high pressure storage containers and generators should be sited outdoors in well ventilated areas.

    • Implementation of safety and separation distances:

      • Separation distances from exposures in GHRS can follow the NFPA-2 code.

      • Hydrogen storage tank (up to 40 Mpa) should be configured 5 m from the location of the hydrogen onsite production facility.

      • Safety distances can be reduced when installed safety systems are effective and can be quickly activated, by employing for instance a dispenser which operates in parallel with an emergency shutdown valve.

      • To determine safety distances for facility layout and under specific operating conditions it is recommended to perform quantified risk assessment targeted to the facility’s specific parameters.

        • This requires a checklist of the HRS sub-systems and components and an extensive description of sub-systems, components, preventive and mitigation measures, configurations (including piping and instrumentation diagrams) and input parameters.

        • The estimated failure rate should be a function of number of fillings rather than based on survival time, as it is more reliable and realistic approach.

        • Establishing a national, independent review function for Quantitative Risk Assessments (QRAs) of HRSs is advisable (see Khalil, 2017). Such an entity would have the potential to become a centre of expertise that could collect existing and future QRAs of HRSs to monitor the latest developments and progress towards the consistent application of the approach as well as provide guidance to permitting authorities on how to apply the approach for HRSs.

    • Protective walls around the HRS can lead to reduced safety distance requirements if they are designed so that flammable concentrations will not reach outside these barriers. However, in case of ignition protective walls can act as obstacles and generate higher overpressures inside the facility. Therefore, their installation should be carefully examined and evaluated under the specific conditions of the facility.

    • Installation of a fire protection wall along station boundaries. This will also reduce the required safety distances.

    • A protective wall surrounding the production site and the storage tank can protect them from the impact from an explosion. Careful design of the protective wall is essential as its resistance to over pressure is another factor. A concrete wall without steel reinforcing bars can withstand a pressure of up to 0.2 bar. This limit may be exceeded under certain conditions if an explosion occurred, for example, in the dispenser. Thus, an additional distance of 2 m away from the dispenser is also recommended for the protective wall and the control room.3

  • Standards / materials

    • Use of equipment in compliance with ATEX to eliminate ignition sources.

  • Safety devices

    • Fit pressure relief valves to components that operate at high-pressure.

    • Provide hydrogen leak sensors and automatic shutdown systems as well as manual ESD buttons.

    • Use infra-red temperature sensors for compressor linked to a high temperature alarm.

  • Safety measures

    • Use proper ventilation if hydrogen equipment is located indoors

  • Practices

    • Install warning signs to prohibiting ignition sources at the HRS.

    • For physical security, install of CCTV surveillance system to act proactively in case of malicious actions.

    • Prohibit self-refilling. Refuelling should be undertaken by trained staff. Alternatively, similar to Japanese regulations, self-filling could be allowed if the driver receives safety education and training on how to mount and demount of the nozzle.

  • Controls

    • If the leak rate based on historical data is estimated to be high, inspections activities shall be more frequent to limit the unrevealed leak time (evaluated from the estimated leak frequency) and increase the process of safety.

    • In densely populated areas, where large safety distances may be impossible to achieve, stricter requirements for quality, inspection and protection of refuelling stations against impact should be implemented.


[7] Dash, S., S. Chakraborty and D. Elangovan (2023), “A Brief Review of Hydrogen Production Methods and Their Challenges”, Energies, p. 16.

[2] Hecht, E. and B. Ehrhart (2021), Analysis to support revised distances between bulk liquid hydrogen systems and exposures, ICHS 202, September 21-24.

[9] Khalil, Y. (2017), “A probabilistic visual-flowcharting-based model for consequence assessment of fire and explosion events involving leaks of flammable gases”, Journal of Loss Prevention in the Process Industries, Vol. 50(A), pp. 190-204.

[10] Kim, E. et al. (2013), “Simulation of hydrogen leak and explosion for the safety design of hydrogen fueling stations in Korea”, International Journal of Hydrogen Energy, Vol. 38/3, pp. 1737–1743.

[5] Kodoth, M. et al. (2020), “Leak frequency analysis for hydrogen-based technology using bayesian and frequentist methods”, Process Safety and Environmental Protection, Vol. 136, pp. 148-156.

[4] LaChance, J. et al. (2009), Analyses to Support Development of Risk-Informed Separation Distances for Hydrogen Codes and Standards, SANDIA REPORT SAND2009-0874.

[3] National Fire Protection Association (2023), Codes and Standards, https://www.nfpa.org/codes-and-standards/all-codes-and-standards/list-of-codes-and-standards.

[8] Pan, X. et al. (2016), “Safety study of a wind–solar hybrid renewable hydrogen refuelling station in China”, International Journal of Hydrogen Energy, Vol. 41/30, pp. 13315-13321.

[1] Statista (2022), Number of hydrogen fueling stations for road vehicles worldwide as of 2022, by country, https://www.statista.com/statistics/1026719/number-of-hydrogen-fuel-stations-by-country/.

[6] Yoo, B. et al. (2021), “Comparative risk assessment of liquefied and gaseous hydrogen refuelling stations”, International Journal of Hydrogen Energy, Vol. 46/71, pp. 35511-35524.


← 1. Based on hydrogen incident databases (see OECD Report – Review on incident database, 2022).

← 2. These distances should not be confused and compared directly with the values presented by (Hecht and Ehrhart, 2021[2]) (see Existing technical norms), because different assumptions on leak conditions were made in the two studies. Moreover, (Hecht and Ehrhart, 2021[2]) haven't calculated the distances based on risk contours, but based on the furthest distance to selective hazardous criteria of the exposure groups.

← 3. Based on scientific work by (Kim et al., 2013[10]).

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2023

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.