The Refinery Additive Follows the Unit, Not the Barrel

A refinery process additive is not consumed per barrel. It is consumed per unit, per operating envelope, at a dose set by the feed that unit happens to be running. Which is why the most important fact about the market is not a demand forecast but a list of refineries that no longer exist in the form they were built.

Europe has removed or converted nearly three in ten of its refineries

FuelsEurope’s Statistical Report 2025, drawing on Concawe data, states it without adornment: “Since 2009, out of close to 100 refineries operating in Europe, 28 refineries (threshold > 30 kbbl/d or 1.5 Mt/a) were closed or transformed. Currently, at least eight refineries in Europe underwent a transformation process, moving away from oil and converting into biorefineries. In 2024, ENI started the process of transforming the refinery of Livorno into a biorefinery.”

Twenty-eight out of something close to a hundred. And eight of those twenty-eight are conversions rather than demolitions, which is the detail that matters most, because a conversion does not remove an additive customer. It replaces one.

Why the closure number is not a demand forecast

Trace a single additive through a refinery and the reason becomes obvious. A desalting aid is dosed against the salt and solids in a specific crude slate at a specific desalter. A hydroprocessing catalyst charge is designed to a feed’s sulphur and nitrogen level and a target cycle length. An FCC additive is bought against a metals level and a gasoline sulphur specification. Each is a chemistry qualified against a piece of equipment and the envelope it runs in.

So a closure takes the additive programme with it. A conversion swaps the programme: the crude unit’s desalting chemistry leaves, and feedstock pretreatment, metals removal, dewaxing and isomerisation chemistry arrive. Same fence line, different customer, frequently a different supplier — because the incumbent’s qualification was written against a unit that is no longer there. The installed base does not shrink smoothly. It is deleted unit by unit and rebuilt unit by unit.

The molecule’s biggest customer is not a refinery

Our Hydrogen Sulfide (H2S) Scavengers forecast runs from USD 3.2 billion in 2025 to USD 5.1 billion in 2033 at a headline 6.0%. The endpoints compound at 6.00%, so the rate is internally consistent (calculation). The largest segment is triazine-based scavengers; the fastest-growing is eco-friendly non-triazine formulations; the dominant region is North America and the fastest-growing is Asia-Pacific.

The line that reclassifies the whole market, though, is the end-use one: the leading end-use segment is upstream oil and gas production. Refining appears on the list. It is not the customer. The scavenger protecting a refinery’s overhead system is bought off a supply chain built for the wellhead, from the same counterparties — Baker Hughes, Halliburton, SLB, BASF, Kemira, Ecolab — at a price the wellhead sets. Downstream additive demand is therefore a residual of upstream demand for the same chemistry, and it inherits the upstream price rather than setting its own.

Two switches, one of them a reformulation

Geography. The European installed base is shrinking and converting while Asia-Pacific grows fastest. Additive demand follows sour gas and new refining capacity east, not the legacy Atlantic basin. A supplier whose field-service model was built around a North American or European site list is defending a declining estate.

Chemistry. A market whose fastest-growing segment is explicitly labelled “eco-friendly non-triazine” is telling you that the incumbent chemistry — triazine, still the largest segment — is under substitution pressure from inside its own category. That is a reformulation race, not a volume race. Reformulation races are won on registration, toxicology and field data rather than on reactor capacity, which favours the suppliers who already run the field-service relationship and disadvantages anyone trying to enter on price.

What a supplier should take from it

The instinct when a customer base shrinks is to sell more per site, and here that instinct happens to be right — but not for the reason it is usually offered. Value per site rises because each surviving unit is dosed harder, monitored more closely and re-qualified more often. Service intensity rises while site count falls. That is a field-service business with a chemistry attach, which is precisely how the large suppliers describe themselves and precisely why the small ones struggle to hold a position.

It is also why a per-barrel model of refinery chemicals gives the wrong answer in both directions. It understates what one complex refinery is worth, because the unit count and the qualification load never appear in it. And it overstates what an entire region is worth once three in ten of its refineries have closed or changed product. The additive follows the unit. Any model that starts from crude runs will be wrong about both the numerator and the denominator.