Before the First Ammonia Bunker: The Evidence Pack a UK Port Must Hold

Ammonia has spent several years as the most plausible of the alternative marine fuels on paper, and 2026 is the year in which the paper case starts meeting the physical one. Bunker suppliers have signed letters of intent, class societies have published rule sets, and a handful of UK ports have begun the quiet, unglamorous work of establishing what a commercial delivery would actually require. The fuel supply question is close to answered. The port safety question is not.

That distinction matters, because ammonia is not a drop-in liquid that happens to be cleaner. It is a toxic gas stored as a liquid under modest pressure or at around minus 33 degrees Celsius, and every part of the chain — the receiving ship, the bunker vessel, the loading arm, the quay, the control room, the emergency plan — has to be built around that fact. A pilot that succeeds because everyone involved is paying unusually close attention has proved very little. The test is whether the arrangements still hold when the novelty has worn off.

The hazard does not shrink to fit the pilot

There is a temptation to treat a first delivery as a small event: one vessel, one thousand cubic metres, one afternoon. The physics does not cooperate. A release of liquid ammonia produces a cold aerosol cloud that initially behaves as a dense gas, hugging the ground and following the contours of bunds, quay edges and vehicle ramps before it warms and rises. Ammonia vapour is lighter than air at ambient temperature, but that is a fact about the end of a release, not the beginning.

Nor is the hazard signalled in the way the public expects. Ammonia has a sharp, recognisable odour, but the smell threshold sits far below the level at which it becomes dangerous and olfactory fatigue sets in quickly. A worker who stops noticing the smell has not stopped being exposed. The Immediately Dangerous to Life or Health concentration is roughly 300 parts per million, and a cloud at that concentration is invisible.

Toxicity modelling that survives cross-examination

Port authorities should expect a dispersion study, not a diffusion diagram. The difference lies in the assumptions. A credible study for a UK harbour will use site-specific meteorological data rather than generic stability classes, will model the berth configuration and the surrounding built environment, and will set out both a defensible worst case and a realistic operating envelope. It should distinguish between a hose failure, a loading-arm disconnect with emergency release, and a catastrophic vessel-side rupture, because the response to each is different.

Two questions expose a weak study. First, what happens under a low-wind, stable inversion at two in the morning, when the cloud neither disperses nor lifts? Second, where does it travel if the wind carries it across a residential area, a retail park or a ferry terminal? If the answer depends on sheltering thousands of people in buildings with recirculating HVAC and no ammonia filtration, the study has described a problem rather than solved one.

Emergency response: who holds the capability at three in the morning?

A harbour authority will need evidence that the local fire and rescue service can operate in the hot zone, not simply attend it. That means gas-tight suits rated for ammonia, appropriate respiratory protection, water spray for vapour knockdown, and a plan for the alkaline runoff that knockdown inevitably produces. Ammonia is highly soluble in water, and suppressing the cloud simply moves the hazard into the drainage system unless containment has been designed in.

Mutual aid matters as much as equipment. Ports should be able to name the neighbouring terminal, contractor or industrial site that will lend capability, the time it will take to arrive, and the individual authorised to call them out. Under COMAH, arrangements of this kind are familiar territory for terminals handling bulk chemicals; the novel element is the interface between a regulated terminal and a bunker vessel that is, at the moment of transfer, connected to both a ship and a shore installation.

Competence: certificates are not competence

Ammonia bunkering sits awkwardly across existing training frameworks. Deck and engineering officers are trained under the IGF Code, and the IMO's interim guidelines have given the industry a structure to work with, but a certificate of competency does not by itself evidence familiarity with a particular loading arm, a particular emergency shutdown sequence or a particular control room. Ports should insist on vessel-specific and terminal-specific familiarisation that is recorded and assessed, rather than accepting a generic certificate as the whole answer.

The same discipline applies ashore. A bunkering supervisor who has spent a career on distillate fuels needs more than an induction video. Simulator time on the emergency release, drills run with the fire service present, and a documented assessment of decision-making under a simulated leak are all reasonable conditions before a bunkering licence is granted.

Equipment, and the vapour return question

Terminal equipment should be reviewed against the release scenarios the modelling has produced. Dry-break couplings, remotely operated emergency shutdown, gas detection with voting logic to limit nuisance trips, and transfer equipment rated for the temperature and pressure regime all belong on the list. So does the vapour return line: ammonia transfer displaces vapour, and a terminal intending to vent it is not operating a closed system.

Detection deserves particular scrutiny. Ammonia sensing is straightforward in principle and awkward in practice, because sensors drift, cross-react and lose sensitivity in cold, humid marine conditions. A calibration regime with independent challenge testing is not a technical nicety; it is the difference between a detection system and a set of boxes on a wall.

Fuel quality: a specification still being written

Unlike residual marine fuels, ammonia bunkers do not yet sit under a single settled international purchase specification. Standards work continues, and class societies have issued guidance, but a port and a supplier agreeing terms in 2026 are to some degree writing their own specification. That makes independent verification a commercial necessity rather than a technical adornment. Purity, water content, oil and particulate contamination, and residual lubricating oil from the production process all affect engine performance and emissions, and the sampling protocol needs to be agreed before the first delivery rather than after the first dispute.

A readiness checklist

• Site-specific dispersion modelling, independently reviewed, covering night-time inversion and off-site impact
• A documented emergency plan with named mutual aid, committed response times and joint drills
• Containment and treatment arrangements for alkaline runoff from vapour knockdown
• Vessel-specific and terminal-specific competence assessment, with records held by the port
• Verified gas detection subject to independent calibration and challenge testing
• Closed vapour return arrangements, or a permitted and documented alternative
• Agreed fuel specification, sampling protocol and independent verification
• A published route from pilot to routine operations, including the review points at which a licence is renewed

The test after the pilot

The real question for any UK port weighing ammonia is not whether it can host a single delivery in 2026. It is whether it can describe, in writing, the conditions under which it would stop. A safety case with no defined failure conditions is not a safety case at all. Ports that set those conditions early, and publish them, will find the move to routine operations considerably easier than those that treat a successful pilot as a permanent licence to scale.