Choosing a gas odorant: four properties that pull in different directions, and one change nobody should make on price

In short

Natural gas leaves the ground without a smell. Everything a distribution network relies on for leak detection is added deliberately, and the choice of what to add is made once and then lived with for decades. That makes odorant selection unusual among purchasing decisions: the material is cheap relative to the system around it, and the cost of getting it wrong is not commercial.

The components, and why blends exist

The odorants in common use are a short list, and most real odorants are combinations of them:

A blend of roughly 75–80% TBM with 20–25% DMS is a common choice for injection systems. THT is regularly used on its own, without blending.

Ethyl mercaptan is on this list for a narrower reason than the others. It is the standard odorant for LPG, propane and butane worldwide — that is where almost all of it goes — and in natural gas service it appears as a component of a blend alongside heavier thiols, not as a standalone choice. It is listed because a blend specification can call for it, not because it is an alternative to THT.

Blends exist because the properties that make an odorant work are not all maximised by the same molecule. There are four that matter, and a molecule that is excellent at one is often mediocre at another.

Vapour pressure

Vapour pressure decides how the odorant behaves inside the equipment rather than inside the pipe.

Too high and the material flashes in the injection line, in the pump head, or in the headspace of the storage tank — which means the dose that reaches the gas is not the dose that was metered, and the loss is largest in exactly the warm weather when tanks are hottest. Too low and the odorant is sluggish to disperse into the gas stream at the injection point, especially in a wick or bypass system that depends on evaporation rather than on a pump.

It is the property most tied to the system rather than to the gas: an injection skid and a bypass odoriser want different things from it, which is one reason a blend chosen for one plant is not automatically right for another.

Odour character — it has to smell like gas

This one resists measurement and matters more than any of the others.

The requirement is not that the gas smells strong. It is that it smells like gas — that a person who has never been trained recognises the smell as a gas leak and acts on it. The industry's term for that quality is a gassy odour, and it is a property of the molecule, not of the concentration.

Different sulfur compounds smell different in kind: some read as skunk, some as garlic, some as rotting vegetation, some as the smell people already associate with a gas leak. An odorant that is intensely detectable but smells like something else is technically doing its job and practically failing at it, because the person who smells it does not conclude gas.

Soil penetrability — the one that gets forgotten

Most leaks in a distribution network are not in a room. They are in a buried pipe, and the odorant has to travel up through soil before anyone smells anything.

Soil is not a neutral medium. It adsorbs organic sulfur compounds onto mineral and organic surfaces, and it hosts microorganisms and oxidising conditions that will convert some of them. A molecule that transmits well through wet clay is not necessarily the one that transmits well through dry sand, and a molecule that is easily oxidised loses ground twice — once in the pipe over time, and again in the metre of soil between the leak and the surface.

This is a large part of why blends are specified rather than single compounds. Different components carry the odour through different ground conditions, and a network that runs across varied soil is buying insurance rather than optimising an average.

Resistance to oxidation

An odorant that has been converted by oxygen is no longer an odorant.

Mercaptans carry an S–H bond and oxidise to disulfides, which smell far weaker than the mercaptans they came from. That reaction is the mechanism behind odour fade: gas leaves the plant correctly odorised and arrives at the end of the network smelling of much less. New pipe, oxygen ingress and long residence times all accelerate it.

THT has no S–H bond and no comparable pathway, which is why it is described as the most stable of the odorants in current use. That stability is the reason to choose it, and it is not the same as needing less of it — its odour threshold is higher, so the mass dose is higher. The trade runs in a definite direction and it is worth stating rather than glossing.

Changing odorant is change control

The most important sentence in any odorant selection is not about chemistry.

People are trained to a smell. Utility crews, emergency responders, and — in most countries — the general public, through decades of public safety messaging. A network that changes odorant changes the warning signal that all of those people have learned, at the same intensity and with no announcement reaching most of them.

That makes a change of odorant a decision with a public-safety dimension, and it should be handled the way other safety-critical changes are handled: assessed, documented, communicated, and phased. It is specifically not a decision that should be driven by a price difference on a tonne of chemical, and a supplier who presents it as one is misreading what they are selling.

The corollary matters for procurement too. If a network has settled on a component or a blend, "equivalent" is a claim to be tested rather than accepted. Two odorants can deliver comparable intensity and still differ in character, in soil behaviour, and in how they age.

Delivery is part of the specification

Odour incidents around odorant do not usually happen in storage. They happen during transfer, when material moves between a truck, a tank and a system, and vapour has somewhere to go.

That makes the delivery method part of what is being specified, not a logistics detail settled afterwards. Bulk delivery by tanker with vapour return, ISO tank, returnable drums, or a container the odorant is injected directly from are different answers with different exposure profiles, and the right one depends on volume, site layout and how often a transfer happens.

For smaller injection points the direct-injection container is often the better answer precisely because it removes a transfer step. Fewer connections broken is fewer opportunities for a plume.

What to ask for

The certificate of analysis for the lot, with methods named — see the THT specification note for what the individual lines control.

Blend composition as specified and as made, if you are buying a blend. A blend is a formulation, and the tolerance on each component is a real number that should be on paper.

The delivery method and vapour handling, agreed before the first shipment rather than at the gate.

Continuity of source. For a material where changing is a safety-managed event, the supply question that matters is not this quarter's price. It is whether the same material will be available on the same specification in five years.

Sources. The component list, the 75–80% TBM / 20–25% DMS blend ratio, and the four selection properties are retained from the article this page previously carried; they are the part of it worth keeping. THT's packaging in drums and ISO tanks and its description as the most stable odorant in current use are quoted from /tetrahydrothiophene-tht/. Oxidation of mercaptans to disulfides, soil adsorption behaviour and the effect of vapour pressure on injection systems describe general chemistry and engineering practice, not measurements on any particular lot or network. No odorisation rate, national requirement or approval status is stated here; dosing is set by the network operator under the rules that apply to them. Ethyl mercaptan was added to the component list on 2026-08-26, when its product page was written; it is the one entry not retained from the original article, and the paragraph above says why it is qualified rather than simply listed. Its boiling point of 35 °C, and the 0.36 ppb detection threshold quoted on that page, are reference values for the pure compound.