Counterfeit Refrigerant: The One Purchase Nobody Inspects

Every other line on an aircon invoice describes something a person can look at. A capacitor carries its rating on the case, a fan motor has a part number, and a filter can be held up to the light. Refrigerant is the exception, built into the product rather than into anyone's conduct.

It arrives as a weight out of a container the household never handled. The cylinder comes off a van, a hose goes on, and the contents move into a circuit that closes behind them. Nothing stays on the ledge to be examined, and the gas has no colour that would help.

Evidence of what went in is therefore whatever the technician chose to make visible. On most visits there is nothing, because nobody thought of the moment as something to show.

An invisible product that is also getting dearer is the kind substitution finds. These gases sit inside an international phase-down, with each country working to a quota on how much may be brought in. Supply that tightens on a schedule is worth faking, and the faking happens far upstream of any ledge.

Distance up the chain accounts for the rest. The cylinder on the van came from a wholesaler, which took it from an importer, which bought it further out again. Every handover is a document rather than an inspection, so a container that entered the chain wrong carries on wrong.

What the machine was built to hold is settled separately by the plate on the condenser casing; the guide on how to identify refrigerant type covers reading it. This page picks up the remainder: what actually enters the circuit at this visit, and how much of that an owner can establish.

The word suggests a watered-down version of the genuine article, and that is not what the published findings describe. What turns up is a different chemical altogether, inside a cylinder wearing a legitimate label.

Canada's federal environment department issued a bulletin on cylinders sold as R410A that turned up inside the country. Their contents were methyl chloride, designated R40, blended with other fluorinated gases. Methyl chloride has no place in modern domestic cooling equipment: it costs less to make than the gases it stands in for, and inside a running system it behaves enough like them to pass.

Britain reached a similar place from a different direction. The national health and safety regulator, working with local trading standards officers, seized cylinders marked R134a, a refrigerant used in vehicles and commercial plant; under test the contents caught fire. Bottles seized elsewhere in Europe held almost none of the named gas, one holding mostly an older refrigerant and another mostly flammable hydrocarbon.

Shipping produced the earliest heavily recorded wave. Refrigerated containers were re-gassed with counterfeit product at overseas depots, and the contamination travelled on with the boxes. Compressors exploded during service work and technicians were killed. The insurers who examined the cases called for cylinders already in use to be tested.

None of this is finished business. A refrigerant producer and a distributors' trade association in the United States have both warned that cylinders packaged to resemble established brands are turning up in ordinary distribution, with contents that may not match the packaging.

None of that evidence is Singaporean, and implying otherwise would be dishonest. No published finding places counterfeit refrigerant inside the local trade. What the record does establish is a practice that is international, repeated over many years, and aimed squarely at a product nobody downstream can check.

Weak cooling is the mild end, and by a distance the likeliest. A system carrying something other than its design fluid works at pressures and temperatures its published data was never written against, so output drifts and the flat no longer settles where it used to.

Component wear sits in the middle of the range. Compressor oil is chosen to travel with one particular refrigerant, and a fluid it was never matched to alters how that oil moves and collects. Machinery running short of lubricant rarely fails on the day; it wears, and the bill turns up later.

The safety end is narrow and documented. Methyl chloride reacts with aluminium, which sits inside cooling circuits, and what that reaction produces catches fire on contact with air. That is the mechanism behind the compressors that burst in the shipping cases, and why a government published an advisory.

Flammability is the second strand of the safety question. A cylinder holding hydrocarbon under a label promising a non-flammable gas places a fuel inside equipment designed around a charge that cannot burn. Refrigerant flammability carries a formal grading, covered separately, and the point here is narrower: a grading protects anybody only while the label is true.

The difference from a deliberate swap shows itself when somebody tries to unpick the result. Where a substitution was decided, one person knows what went in and can say so. Where the label lied, that knowledge exists nowhere along the chain: the paperwork is confidently wrong instead of merely silent, and confident wrongness is harder to argue with than a blank.

Four things, and all four happen before a hose is connected: a cylinder that has not been opened, a label that can be read, a product designation that matches the outdoor unit, and a quantity that gets written down.

An unopened cylinder is the plainest of the four. A part-used container carries a history nobody present can account for, including whatever it held before. Watching the seal break proves nothing about contents, but it does close the shorter question: whether today's charge came out of some earlier undocumented job.

A legible label beats an impressive one. It should name the product, name whoever filled it, and carry a batch or lot marking. Those markings exist so a particular filling can be traced back if questioned, and an anonymous cylinder offers nothing to trace.

Matching the designation to the machine is the check with the sharpest edge. The outdoor unit names one refrigerant and the cylinder names one refrigerant. Two answers that agree end the matter; two that differ need an explanation before work goes ahead.

The written quantity is the check that outlives the visit. Refrigerant is sold and charged by weight, so a figure exists whether or not it reaches the paperwork. Asking for the product name and amount on the invoice produces the only durable record the day generates.

Ask during the visit: an amount written while the scale is in the room is a measurement, one reconstructed from memory afterwards is a recollection.

None of the four asks for anything unusual. Somebody charging genuine product loses nothing by opening the cylinder in view or writing its name onto the job sheet. These are reasonable to put to anyone, and they need no framing as a challenge.

None of those four checks establishes what is inside the cylinder, and saying so plainly is the honest position. Every documented case describes product that looked correct to whoever bought it, and looking correct is the object of the exercise.

Establishing composition takes an instrument or a laboratory. Refrigerant identifiers exist, and the trade holds purity specifications a sample can be measured against, but both belong to whoever handles the gas professionally. An owner on a ledge has no version of that test.

Traceability rather than certainty is what those four checks buy. A named product, a named amount and a date turn an unanswerable question into a followable one. Where a system misbehaves afterwards, that record is what the next technician works from instead of nothing.

The matter then folds back into the question of who is doing the work. Refrigerant is an unusually clean test of that, because the product gives an owner no help whatsoever. Separating a contractor who can prove things from one who is only confident is settled elsewhere, under signs you can trust an aircon company.

Nothing here asks a household to meet a routine gas visit with suspicion. It asks for two questions and one line on an invoice. Where those arrive easily, the subject closes; where they do not, the difficulty in answering is itself the information.