The rapid scale-up of electrolyser deployment has exposed a critical weakness in hydrogen production, with responsibility for process safety often remaining unclear.

Despite mature standards and well-understood hazards, process safety is too often treated as a documentation exercise rather than a design discipline. The industry does not lack the tools to address this problem. What it lacks is ownership. Unless project developers take explicit responsibility for defining process safety requirements, critical risks can remain unaddressed.

Piloting and industry collaboration have done much to improve the efficiency and financing of clean hydrogen. In my experience, the discipline of process safety has not kept pace.

This is not to say that the industry is operating in the dark. Hydrogen is one of the best-understood industrial materials, having been used for well over a century in town gas, chemical and process applications. The standards, methodologies and regulatory frameworks needed to manage it safely are mature and readily available.

What I often see is uncertainty across the supply chain about what process safety actions should be taken and, crucially, who is responsible for taking them. The result is an accountability gap that can leave important risks unaddressed.

Understanding the role of electrolyser process safety
Process safety, which aims to prevent and control major industrial accidents, remains one of the most widely misunderstood engineering disciplines, even among experienced engineers. That challenge will only grow as the hydrogen sector expands. DNV predicts that clean hydrogen production will increase 100-fold by 2060, helping to mitigate more than 2Gt of annual emissions from sectors that are difficult to electrify.

A 2023 study by the Institute of Sustainable Process Technology identified the need to maintain process safety across design, construction and operational phases as one of the industry’s key challenges.

Every process safety assessment starts with identifying hazards. Once hazards have been recognised and understood, the safeguards needed to manage them can be defined. Documentation, standards and assessment methods come afterwards, providing evidence that those safeguards have been specified and implemented.

Process safety is often reduced to a compliance exercise. An organisation may produce a Hazard and Operability (HAZOP) study and a full suite of process safety documentation but still be unable to explain which hazards those documents are intended to address. Producing documentation is not the same as understanding and managing risk.

Identifying the ownership gap
Responsibility for process safety often becomes blurred as projects move from design into procurement and delivery. Developers typically assume process safety is embedded within the systems delivered by manufacturers and engineering, procurement and construction (EPC) contractors. Manufacturers and EPC contractors, in turn, rely on developers to define the process safety requirements they are expected to meet.

This is not a question of intent. Every party is working to deliver a safe, successful project. But procurement specifications define what suppliers are expected to deliver. If process safety requirements are not clearly specified at that stage, they are unlikely to form part of the final delivery.

What I see is a systemic accountability gap, and one that poses real risks to project performance, asset integrity and ultimately human safety.

Many organisations still approach safety primarily as an engineering problem to be solved, producing the documentation needed to satisfy customer requirements or achieve certification, rather than as a discipline that should shape the design of a project from the outset.

Procurement is where responsibility can be exercised most effectively. Project developers ultimately carry the operational, commercial and legal consequences of a safety failure. They therefore have both the greatest exposure and the greatest influence. Procurement specifications should clearly define the process safety evidence expected from manufacturers and EPC contractors rather than relying on whatever information suppliers choose to provide.

There is a systemic accountability gap, and one that poses real risks to project performance, asset integrity and ultimately human safety.

Understanding electrolyser-specific risks
Documentation can address many common hazards, but electrolyser-specific risks are still often overlooked. Effective process safety follows a clear sequence:

• Identify hazards – pinpoint what can go wrong.
• Define safeguards – determine how escalation will be prevented.
• Demonstrate performance – provide evidence that those safeguards are effective.

Hydrogen understandably attracts most attention, but substantial volumes of high-concentration oxygen, for example, are often treated as a secondary hazard. Many organisations still assume oxygen presents little risk because it is unlikely to ignite under their operating conditions and any release will simply disperse into the atmosphere.

This assumption overlooks an important hazard. High-concentration oxygen requires strict material selection because contaminant particles within an oxygen stream can become self-combustible. Materials that perform safely under normal atmospheric conditions can also degrade rapidly when exposed to oxygen concentrations beyond those for which they are certified.

Safeguards illustrate the wider problem. One common example is cell stack monitoring, where sensors monitor the operating condition of the electrolyser stack, triggering alarms or shutdowns if abnormal conditions are detected. These systems are often cited in documentation as safeguards against a range of failure modes. A monitoring system, however, can only be regarded as effective if rigorous testing has demonstrated that it can detect the relevant failure modes early enough to prevent harm. Unless developers require that evidence during procurement, suppliers are unlikely to provide it, leaving the effectiveness of the safeguard unverified.

Factory acceptance testing raises similar concerns. In many industrial sectors, a factory acceptance test means a system has been fully tested before leaving the manufacturer’s facility. Within the electrolyser industry, factory acceptance testing often consists of pressure testing to confirm there are no leaks together with electrical checks. This system is not operated under live production conditions, however, meaning an electrolyser can arrive on site without producing a single cubic metre of hydrogen. The first production run may therefore become the first meaningful demonstration that the system performs as intended.

Bridging the gap
Effective process safety should not begin with a list of standards or regulatory requirements. It should begin with a structured assessment by the project developer of what the system produces, what can go wrong and what the consequences would be.

Hazards must be identified before safeguards can be defined. Only then can the appropriate standards and verification methods be selected to demonstrate that those safeguards are effective.

Independent hazard assessment and verification help ensure that hazards are identified consistently, risk reduction measures are proportionate to the risks involved, and responsibilities are clearly allocated across developers, manufacturers and contractors.

Industry standards set out requirements for electrolyser design, construction and operation, together with methods for verifying that systems perform as intended. They also establish common approaches to defining system boundaries and assessing performance, allowing technologies to be evaluated on a consistent basis. DNV’s electrolyser safety and performance standards are one example of this approach.

Standards, however, cannot compensate for requirements that were never defined. Responsibility rests with the project developer to issue technically grounded procurement specifications that identify the hazards to be addressed, the safeguards required and the evidence needed to demonstrate that those safeguards will perform as intended. In a commercial environment where suppliers deliver against defined specifications, those requirements establish accountability across the supply chain.

Standards are already available. The challenge is ensuring developers define the hazards, specify the safeguards and require evidence that those safeguards will perform as intended. Until ownership of process safety is clearly defined, accountability gaps will remain, regardless of how much documentation accompanies a project.

The views and opinions expressed in this article are strictly those of the author only and are not necessarily given or endorsed by or on behalf of the Energy Institute.

• Further reading: ‘Going beyond understanding ‘what went wrong’. Find out how the Energy Institute’s updated and expanded version of EI 3295: Reporting, investigating and learning from incidents, accidents and events, can make operations safer.
• ‘Is electrolyser technology good enough to build a large-scale hydrogen economy?’. Development of electrolysers which split water into hydrogen and oxygen will be critical to achieve net zero emissions by 2050. In particular for decarbonising hard-to-abate industries, with electrolysis powered by renewable energy. Although electrolysis capacity is growing from a relatively low base today, discover why there is need for significant acceleration to reach the net zero target.