6. Examining scenarios involving hydrogen leakage

Hydrogen is a crucial element for the energy transition towards a low carbon economy. It can contribute significantly to the reduction of carbon emissions, which in turn could mitigate potentially catastrophic climate related disasters. A successful increase in hydrogen adoption would also play a key role in meeting the goals of the European Green Deal. Hydrogen can be used in several sectors, including transportation, industrial and domestic use. However, safety concerns and a lack of national-level safety regulations could potentially hinder its widespread use. The goal of this reform project and the report on literature review is to support the Dutch government to speed up its energy transition, and to develop country-specific recommendations for the safer and sustainable use of hydrogen.

This report consists of a study into the likelihood of a number of ignition/effect scenarios as resulting from hydrogen leakage. It consolidates existing knowledge, research, and data, on hydrogen leakage and ignition risks. It also aims at improving knowledge in relation to the risks associated with the use of hydrogen in small-scale applications. It provides main findings and guidance with regard to adequate risk-management of hydrogen applications in different scenarios, especially for the development of appropriate regulations and regulatory processes for the safe use of hydrogen technologies.

The report, which builds on both numerical and experimental research in the field of hydrogen safety and risk assessment, intends to help identify current gaps in hydrogen safety – to further help local authorities clarify risks related to hydrogen-based technologies in relation to the issuance of efficient, risk-based permits for their applications.

The literature review covers 99 scientific articles, divided into six distinct scenarios, covering potential sources of accidents in production, transportation, fuelling stations and residential use. The scenarios described in this report have been selected at the behest of the Dutch Ministry of Economic Affairs and Climate Policy. They are of particular interest as they cover use of hydrogen technology in densely populated areas requiring safety and risk management techniques.

The study of the six scenarios describes the evolution of hydrogen safety requirements over the years and the acceptable risk standards. For instance, it studies the risk associated with hydrogen production by electrolysis, focusing on pipes connected to electrolysers. Given that alkaline electrolysers are a very mature technology,1 the risk is considered acceptable should current risk measures be followed. When dealing with the domestic use of hydrogen for heating, preliminary results from on-going large-scale pilot projects indicate that such use could be as safe as using natural gas for heating, if the necessary mitigation measures are put in place.

This chapter presents lessons learnt and recommendations on key safety elements for hydrogen technologies that new/revised regulations could consider in order to achieve better outcomes. The findings are presented in separate sections for each scenario with a synthesis of the review of findings from research data and relevant safety recommendations for that scenario. These recommendations are based on the OECD research findings and should be considered as a list of options to reduce the risks related to hydrogen technologies. The recommendations are focused on six scenarios/applications that cover a wide spectrum of the hydrogen supply chain.

Improvements in technological standards and better risk-management studies show that hydrogen is not as risky as previously perceived to be. This finding runs parallel to the fact that no fuel is 100% risk free. With more understanding of the manner in which hydrogen functions and improvement in certain technologies such as those related to sensors, ventilation, and storage materials have improved, regulators should strive to improve the public perception of hydrogen against the greater risks of climate change. As the evaluation of safety codes and regulations on a rolling basis and pilot projects already shows, it can be stated that with adequate safety protocols in place hydrogen fuel can be used safely for commercial and small-scale private purposes.

Chapter 7 Hydrogen Safety Aspects presents a brief summary of scientific articles on existing knowledge by theoretical, experimental and numerical research on hydrogen safety in general.

Chapter 8 Mapping Exercise summarises and discusses existing knowledge from scientific research on previously defined scenarios. It focuses on data regarding safety of hydrogen as a fuel in six scenarios, covering the hydrogen lifecycle in its various phases. The literature review covers 99 scientific articles, divided into six distinct scenarios, covering potential sources of accidents in production, transportation, fuelling stations and residential use. A brief description of these scenarios is provided below:

This section presents a discussion on safety aspects of hydrogen production from water electrolysis, with a focus on leakage from pipes connected to electrolysers. It starts with a brief introduction on up-to-date electrolysis technology, which is followed by a holistic picture of risks associated with hydrogen production sites. With the major risk contributor (compressor) identified, a discussion on hydrogen pipeworks2 Within a production site focusing on those connected to electrolysers is presented: Incident database records together with experimental and computational work by independent authors conclude that the risk associated with pipework leakage is acceptable should current safety measures be followed.

The studies in this section analyse the transport of compressed gaseous hydrogen through pipelines and the safety measures that should be taken into account. The research focuses on ignition, leakage and explosion likelihood, potential damage to buildings, people and the necessary safety (viz., separation) distances to mitigate these hazards. Quantitative and experimental research alongside models verify the impact of the pipe material, nature of ground soil, the internal flow and the position of the pipeline (viz., above or below ground level) on the aforementioned hazards. It concludes that more detailed verification of relevant –– that can influence the hazard situations and the relative safety mitigation measures are needed.

This section presents a discussion of important safety aspects of hydrogen behaviour. This includes hydrogen-related risks arising from parking garages and accidents in urban areas, as well as comparisons between hydrogen fuel cells vehicles (FCVs) and compressed natural gas (CNG) cars from a safety perspective. One of the key considerations for policy makers regulating hydrogen use in confined spaces is the use and design of ventilation (natural and mechanical) systems. Experimentation and computational methods show that natural and/ or mechanical ventilation contribute to the reduction of risks associated with hydrogen leakage. Studies have shown that sensors and their placement in the HFCVs is an important consideration for regulators to reduce risks associated with hydrogen leakage. It can therefore be considered that if adequate precautions exist such as ventilation, sensors, and well-tested safety valves, the public perception for HFCVs could gradually be improved.

An important issue concerning the safe use of hydrogen-powered FCVs is the possibility of accidents inside tunnels resulting in the release of hydrogen.3 To understand the potential consequences, several experimental and theoretical studies as well as risk assessments have been conducted. The studies presented herein determine the severity of the predicted consequences. These studies analysed the behaviour of the flammable hydrogen cloud inside the tunnel, predicted the overpressures arising from accidental hydrogen releases in areas with no or limited ventilation and determined the probability of ignition and the possible delay before the cloud ignites. The height, shape and architectural design of the tunnel as well as different ventilation regimes were studied as potentially important parameters in determining explosion risks and appropriate mitigation measures. The role of the TPRD’s size and orientation were investigated: smaller sizes were recommended and vertically downwards releases were discouraged. The time delay prior to ignition, in case of a hydrogen leak, was found to be an important parameter, since: ignition delays of about 4 to 8 seconds can result in dangerously high overpressures, approaching or surpassing the fatality threshold level.

Through several studies, safety measures were gauged via risk-based approaches to prevent leak and explosion of hydrogen. These included studies aimed at determining the safety distances on hydrogen stations planned to be installed and the ignition likelihood in the station’s components. Additionally, aspects such as the nature of accidents and incidents at hydrogen fuelling stations over time were analysed to identify key safety issues. The catastrophic rupture of a tube trailer and a liquefied hydrogen tank were found to be the worst accidents of hydrogen refuelling stations.

In this scenario, the studies that have been conducted on the safe use of hydrogen in residential buildings, mainly for heating, are being investigated. The findings of large-scale projects, such as the Hy4heat project, the H100 project and the HyDelta project, which aim to investigate the possibility of substituting natural gas with hydrogen for heating, are presented and summarised in this section. Based on findings from these projects it has been concluded that the use of hydrogen for heating could be as safe as using natural gas, as long as a fit for purpose s distribution network is used and additional mitigation measures are implemented downstream of the gas metre.

Results and main takeaways from the scientific literature review for the six above-mentioned scenarios are presented in separate subsections in this report. Each study is briefly summarised, together with its main conclusion, and supported by relevant supplementary material. The mapping exercise provides important parameters of hydrogen in case of an accident, (release rate, dispersion, overpressure, heat flux, etc.), as well as possible prevention and mitigation measures (such as ventilation, safety distances etc.). Recommended actions are proposed based on the gaps identified in the scientific papers.

Hydrogen can contribute to the reduction of carbon emissions; however, its simple structure makes hydrogen more flammable, thereby raising more safety concerns. In order to evaluate the safety of hydrogen as a fuel, regulators must compare its safety risks, benefits, and disadvantages against other fossil fuels. It should be noted that no source of fuel is entirely safe. In fact, on some counts, hydrogen is found to fare better than other conventional fuels. For instance, while fossil fuels are carcinogenic and polluting, hydrogen is absolutely non-toxic and there is little evidence to suggest that hydrogen leakage will cause catastrophic environmental disasters like those arising from oil spills. On the other hand, hydrogen requires 18-59% oxygen for explosions as compared to just 1 to 3% in case of fossil fuels. Due to its low weight, hydrogen rises easily. This property reduces the probability of secondary fires.

General prevention measures for hydrogen applications include installation of pressure relief valves, flow restrictors and shut down emergency systems, regular maintenance of individual components, training of personnel, controlling ignition sources, limiting congestion in closed spaces and use safety distances. Possible mitigation measures are: proper installation of detection sensors at locations where hydrogen accumulation is expected, e.g. on ceiling, and limiting/stopping the hydrogen supply before concentration reaches 15 % v/v, natural ventilation with multiple vent configuration, large vent area and small aspect ratio (length/height) with no obstructions in front of them or mechanical ventilation with ATEX-compliant systems to prevent concentrations above LFL, when possible. The use of fire suppressing system, such as water mist, can also mitigate the consequences in case of fire and explosion and prevent the fire spreading.

Over the years, technological standards related to sensors, ventilations, and materials have improved. The same is the case with safety codes and regulations. Pilot projects have shown that with adequate safety protocols in place, hydrogen fuel can be used safely for commercial purposes. For instance, new research shows that hydrogen FCVs should be treated at par with CNG vehicles even in serious cases involving crashes. Given that CNG cars are now publicly accepted as safe, such kind of assurance could improve the regulatory case for hydrogen as well. Similarly, for parking garages especially those in confined spaces and basements, the use of ventilation has been found to be a good mitigation measure for accidents arising from hydrogen leakage. In addition to this, prominent signage informing the public about bans on smoking, mobile use or fire lighting are simple yet well-known steps for controlling ignition sources at refuelling stations. Incorporating a behaviour change using these measures is relatively less work due to similar global restrictions at fuel stations involving conventional fuels.

As far as hydrogen production via water electrolysis is concerned, alkaline electrolysers represent a mature technology with most large production plants built between 1920s-1980s. Risk analysis together with historical data from 3 databases (ENSAD,HIAD 2.0 and H2tools) show that most root events can be either reduced or eliminated following current risk measures. ENSAD reported no hydrogen release at production sites; HIAD2.0 data suggest the risk associated with electrolysers are small compared to compressors and pressurised storage; while H2tools reported no such accidents after 1990 - when modern valve design became available. In addition, calculations performed by Sandia based on historical hydrogen data suggest that the risk associated with leakage from pipeworks connected to the electrolyser is within the boundary set by the purple book. Therefore, we can conclude that existing knowledge of Scenario 1 is sufficient and we expect the risk to be acceptable.

Research concerning hydrogen pipeline transport has highlighted the consequences and risks related to pipeline failures. The two main incidents of leakage and rupture of a high pressure pipeline lead to potential damages to people and buildings with both individual and societal risks. Different QRA have been to define the probability of accidents to occur and the risks connected to ignition and explosion. With the release of hydrogen due to rupture the potential ignition and explosion leads to a maximum value of 1.65 10-3 death/year/1000 km. To mitigate this risk it is advisable to establish zoning in land use management to create a distance between the source of risk and nearby buildings and people. As well as the potential of external interferences with the pipeline site. A great influence to the scale of damage is linked with wind speed, ground roughness, tube pressure and leakage gap area on the diffusion distance and overpressure distance. Experiments on the surrounding feature can factor in other mitigation and safety measures.

The second major hazard is leakage of hydrogen which happens at a higher pace and is greater in volume flow in confined spaces compared to natural gas. It is highly influenced by contact with air. Conversely it has a lower ignition likelihood determining a lower probability of explosion damage. Being hydrogen gas odourless, colourless, and tasteless, leaks are not detected by human senses. Therefore, as a safety measure to counter major consequences from hazards, the use of hydrogen sensors is recommended to successfully detect hydrogen leaks. As well as a ventilation system that mitigates the potential damage by enabling hydrogen to escape to adjacent spaces.

Regarding hydrogen mobility inside tunnels, risk analysis studies showed that a hydrogen accident inside a tunnel will most likely not lead to ignition since there will be no release of hydrogen (probability of 94.1%). However, if hydrogen does ignite, fire can spread quickly inside the tunnel. An appropriate ventilation is key to help prevent ignition and reduce the chance of an explosion. Appropriate ceiling design and additional measures are also needed to reduce or mitigate potential damages. For example, attributes of the TPRD, such as diameter and orientation of gas release can make a difference when it comes to hazard mitigation. Storage systems involving more than one TPRDs should be designed to avoid simultaneous opening of all PRDs. The deliberate activation of TPRD can also mitigate the consequences from a tunnel accident. New technologies, like the leak-no-burst tank, that prevent tank rupture, can significantly reduce the risk and address the concerns of firefighters, especially in tunnels and other confined spaces. If hydrogen is ignited right after being injected in the tunnel it forms a jet fire with a heat release rate that gradually decays with the injection rate. In the case of delayed ignition however, a pressure wave propagates through the detonable hydrogen cloud resulting in a blast wave and overpressures that may approach the fatality threshold level. A potential failure of the TPRD failing is a hazard that should be taken very seriously as there can be severe consequences from the ensuing explosion.

Research within the area of accidents at hydrogen fuel stations highlighted the need to guarantee a high level of safety for hydrogen fuelling stations, in view of their increasing widespread construction across the world. Not only do correct and adequate sealing and torque need to be carefully considered, but also a set of safety recommendations. These include the installation of a protective wall surrounding the dispenser, limiting the inventory in storage facilities on-site, refuelling stations and setting appropriate safety distances not only between the station and residential area but also among different elements of the station, e.g. between dispenser and storage room. Moreover, the study, by looking into life parameters in QRA and fire spreading at both GHRS and LHRS could provide concrete information about risks and what can be done to reduce their likelihood.

Overall, recent research into the residential use of hydrogen for heating has shown that hydrogen can be as safe as natural gas, as long as the right mitigation measures are put in place such as appropriate pipeline components and leakproof design. The use of a 100% polyethylene network is proposed to minimise gas leaks and the installation of two emergency flow valves is recommended to cut off the flow of hydrogen before a hazardous scenario can occur (Mouli-Castillo et al., 2021[1]). By ensuring that all hydrogen appliances adhere to the proper specifications and are properly ventilated, the risks can be reduced even further.

In conclusion, there are gaps of existing knowledge relating to hydrogen safety. However, the success and learnings from existing projects is a positive sign that hydrogen could be a key pillar in the fight against climate change, and environmental disasters. The involvement and initiative of governments around the world, academic and scientific institutions as well as private firms shows that a scientific foundation for hydrogen-use already exists. However, more work is required to improve public perception on hydrogen safety.

Through numerous experimental and theoretical studies, the behaviour of hydrogen and its intrinsic properties are fairly well established. However, there are still some knowledge gaps on how accidents could lead to harmful consequences. This is particularly challenging because of the prevalence of uncertainties in the application of QRA. Moreover, current data collection efforts in the hydrogen fuelling industry do not impose the obligation of the quality necessary to perform QRA. Therefore, data coherence continues to pose challenges.

To offset the scarcity of hydrogen-related field failure data, scientists use statistical methods such as bounding analysis and uncertainty propagation techniques such as Monte Carlo sampling and the Latin Hypercube sampling methods so that the predicted outcomes can be represented by probability distribution as compared to single-point values.

Scenario 1, shows that alkaline type electrolysers, as a mature technology for hydrogen production, were thoroughly assessed in the last century. Based on entries in accidents databases and estimation using hydrogen specific historical data, we concluded that there is sufficient knowledge regarding Scenario 1. Nonetheless, continuous monitoring of hydrogen accident databases is recommended to gain further insights on the risks associated with Scenario 1.

There has been a large number of experimental research on Transport pipelines: leakage from high-pressure pipeline (scenario 2) both in lab and in external environments. It is advisable to deepen the knowledge, through experiments, of pipeline ruptures and possible ignition likelihood due to adverse and different weather conditions that can vary the impact and damages of accidents. In terms of safety measures the HSE and EIGA experiments can streamline a series of guidelines useful to mitigate the possible adverse effects.

For HFCVs (scenario 3), QRAs can help identify important hazards. Some hazards identified under HySafe studies include internal (random component failure) and external (crashes, high winds, floods, earthquakes) incidents, accident sequences where hydrogen leakage and ignition lead to fires and explosions, as well as consequences of such incidents due to thermal and pressure effects on neighbouring property and people. Current research is simplistic. For instance, current ignition probability models tend to ignore several complicated ignition-induced scenarios such as vehicle crash scenarios. A second example relates to sensors. While sensors and gas and flame detection equipment are established as necessary, there still isn’t enough information on the accuracy of these equipment, or for instance their proper placement to reduce the consequences of a crash involving vehicle roll-over. In addition to equipment failure, studies should also factor human, software and organisations errors. In Germany and Norway, several non-fatal hydrogen accidents were a result of human-related errors. While a full-scale QRA is always desirable, as it could help evaluate all the types of potential accidents, their consequences, frequencies, and it can also help prioritise the risks, a full-scale QRA could be costly and labour intensive to collect the needed equipment failure data.

There has been a large number of experimental and numerical studies of the risks of hydrogen vehicles in tunnels (Scenario 4). Additional research can be performed to understand the effects of deflagration or detonation inside tunnels of different configurations and for accidents involving different types and classes of hydrogen vehicles. Determining the extent in which obstructions inside the tunnel can raise additional hazards in the case of a detonation or deflagration could also be an important step forward to improve the current state of knowledge in this area. The now concluded EU-funded project, HyTunnel-CS, has answered several knowledge gaps around the safe use of hydrogen inside tunnels and other confined spaces, notably showcasing a TPRD-less tank that releases hydrogen in a controlled manner by turning porous under extreme heat. However, given that the Hy-tunnel project findings were released when this report was near completion, the present report does not delve deep into the findings from Hy-Tunnel.

There is quite extensive literature on postulated accidents in refuelling stations (scenario 5). However, some further research as well as actions to facilitate further collaboration and research between hydrogen industry and research organisations with the aim of transferring knowledge should be promoted. In addition, developing a more thorough knowledge on accidents and incidents that took place at hydrogen refuelling stations involving small leakages of hydrogen is needed and suggested safety measures shall be taken. Further analysis on QRA guidelines for HRSs is needed to facilitate the implementation of such recommendations.

For scenario 6, knowledge gaps are actively being filled-in by on-going demonstration / pilot projects. It would be useful to get in contact with the consortium of these projects to obtain up-to-date information on their research and on their future plans.

The findings arrived at in Part 1 for the various scenarios have been through an analysis of scientific articles and reports published over the last couple of decades. In addition to this, several ongoing projects in the field of hydrogen safety have also been referenced. These international projects are either ongoing or have concluded and help shed light on hydrogen safety including its safety features in the selected scenarios. The international projects referred to in this report are in Table 6.1 below.


[1] Mouli-Castillo, J. et al. (2021), “A quantitative risk assessment of a domestic property connected to a hydrogen distribution network”, International Journal of Hydrogen Energy, Vol. 46/29, pp. 16217–16231, https://doi.org/10.1016/j.ijhydene.2021.02.114.


← 1. Nonetheless, there is still room for innovation and newer technologies such as PEM (proton exchange membrane) and SOECs (solid oxide electrolysis cells) can provide either better control or efficiency.

← 2. Pipes within a site used for transfers internally and to the site boundary.

← 3. This scenario examines the possible hazards from a hydrogen vehicle crash inside a partially confined space. The example most prominently used here will be that of a collision involving a hydrogen vehicle driving inside a tunnel. Other partially confined spaces, such as parking garages are discussed elsewhere in the report as they are more relevant to Scenario 3 (“H2 leakage in a confined space/ built environment”).

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.