On an NGL train the governing fire is a jet fire, not a pool fire

Construction began this week on a fifth natural gas liquids fractionation train at a gas processing complex on the Abu Dhabi coast, with a scope that includes a propane refrigeration system and storage facilities. On plant of that description the credible major fire is a pressurised release of a liquefied gas burning as a jet, and the passive fire protection specification most projects inherit is qualified against a pool fire. Those are not the same fire, and the second qualification does not cover the first.

What started this week

The groundbreaking on 24 September covers engineering, procurement and construction of a fifth NGL fractionation unit, a regeneration gas treatment unit, a propane refrigeration system, ancillary systems and storage, for completion in 2030 at an output capacity of 8 million tonnes a year. Elsewhere in the same week a production facility expansion was awarded in Oman and a high-pressure polyethylene technology licence in Saudi Arabia. The common feature across all three is inventory held as liquefied gas under pressure rather than as a liquid at atmospheric conditions.

That distinction decides the fire. A liquid hydrocarbon that escapes forms a pool and burns as a pool — buoyant, spreading to the limits of containment, radiating in all directions. A liquefied gas that escapes through a small hole at pressure does not form a pool. It issues as a momentum-driven flame that is narrow, directional, and capable of impinging on a vessel wall, a pipe support or a structural member several metres away for as long as the inventory lasts.

Two fires, two test curves

The methods that qualify passive fire protection reflect the difference, and reflect it in the numbers.

The hydrocarbon pool fire test used for onshore fireproofing, UL 1709, exposes the specimen in a furnace to a total heat flux of about 204 kW/m² at an average temperature of about 1093 °C. The jet fire test, ISO 22899-1, works to a nominal 250 kW/m² and a flame temperature near 1200 °C, applied as an impinging jet rather than an enveloping furnace. ISO 22899-3, published in January 2025 as an extended method complementary to Part 1, raises that to approximately 350 kW/m² and about 1300 °C, with a validity condition that at least two flame thermocouples reach 1200 °C within the first 300 seconds after ignition.

Heat flux is only half of it. A pool fire heats; a jet fire heats and scours. An intumescent coating protects by charring into an insulating foam, and that char has to survive mechanically. Under a high-momentum jet it can be eroded away from the steel it is meant to insulate, which is why jet fire resistance is a separate qualification and not an extrapolation from a pool fire result. Piper Alpha remains the reference case: the passive fire protection on the platform was destroyed within minutes, against the one to two hours for which such systems are normally rated.

Where the specification quietly fails

Onshore practice inherited the pool fire. The most widely cited fireproofing guidance for petroleum and petrochemical processing plants, API 2218, is a pool fire document in its main body; jet fires entered it as an annex recommending that owners consider them. That change followed a February 2007 refinery fire in Texas, where water trapped in a piping dead leg froze, cracked a 10-inch liquid propane line, and produced a fire that reached well past the envelope the fireproofing had been placed to cover. The fireproofing was installed. It was in the wrong place and to the wrong performance standard.

Three failure modes follow from that inheritance, and all three are settled at the point a project breaks ground rather than later.

The envelope is drawn from the wrong scenario. A fireproofing envelope derived from pool radii and containment kerbs will not include the structure a jet can reach from a flange at a different elevation several metres away. Jet fire cases have to be run as a separate exercise, with release direction treated as unknown.

The product is qualified for the wrong fire. A coating certified only to a hydrocarbon pool fire curve is cheaper, and on a bid comparison it appears to satisfy the same line item. On a gas processing unit the pool fire risk is low and the jet fire risk is high, so the cheaper product protects against the scenario least likely to govern.

The duration is not tied to depressurisation. A fire protection rating is only meaningful against the time the plant needs to isolate and blow down the affected inventory. Where the emergency depressurisation philosophy and the passive fire protection rating are set by different parties and never reconciled — the normal case — the rating is arbitrary.

Water does not fix a jet fire

The instinct to close the gap with deluge should be resisted on gas service. A jet carries enough momentum to punch through a water curtain and continue heating the steel behind it, to the point of vessel rupture. Worse, deluge can extinguish the flame without stopping the release, which converts a jet fire into an accumulating flammable cloud and a possible explosion — a heavier event than the one being suppressed. Water earns its place cooling adjacent equipment and providing exposure protection. It is not a substitute for passive protection on the item under impingement.

What to check now

On any gas processing, NGL or refrigerated storage package in front-end or detailed engineering:

Does a jet fire scenario set exist as a deliverable separate from the pool fire assessment, with release directions and impingement targets identified?
Is the fireproofing envelope on the drawings derived from that set, or from containment kerbs and pool radii?
Are the specified products certified to a jet fire test — ISO 22899-1, or the extended Part 3 — and is each certificate for the configuration actually being used, whether structural steel, pipe, valve or penetration?
What fire duration was assumed, and does it exceed the time the depressurisation study says isolation and blowdown will take?
Where deluge appears in the fire protection philosophy, does any line of it rely on water to protect an item against jet impingement?

If the first question has no answer, the remaining four have no basis. That is the state of a large share of packages at the point construction starts. It is cheap to correct now and expensive to correct in 2030.

Mainstream Petrochemical Instruments FZE provides fire safety engineering for high-hazard industrial facilities. This article is general technical commentary and is not a design recommendation for any specific facility.