In defence of the smell test
Every so often a customer asks, politely, whether we are seriously identifying residues by smell in 2025. Yes. Here is why it is the right tool, and exactly where it stops being one.
- Sheet
- 77
- Revision
- 2026-10
- Scale
- 1:1
- Drawn at
- Peoria IL
The question behind the question is fair: is this a craft tradition or an actual control? It is an actual control, with a defined scope, and the scope is the important half.
What it is genuinely good at
- Detecting a mismatch between a declaration and reality. This is its primary job. If the paperwork says glycol and the container smells of solvent, everything stops.
- Separating glycol, vegetable oil, surfactant, fuel, fermentation residue and caustic in under five seconds each, with an accuracy no affordable instrument matches at that speed.
- Finding absorbed aromatics in a bottle that looks spotless. This is the one that saves us from selling a container that can never return to product service.
- Catching a container that was sealed damp and has grown something, which a visual inspection misses entirely.
On a pad handling 150 containers in a shift, a control that takes five seconds and catches declaration mismatches is worth more than a control that takes twenty minutes and is run on a sample. Both have a place; only one of them scales to every container.
The four things it must never decide
- Food-grade status. That requires four documents and a lab result, and no nose substitutes for any of them.
- Whether a residue is safe to wash. That comes from the declaration. The wash bay operator is the person standing next to the container, and they are not being asked to gamble on a guess.
- The grade, on its own. Odour is one of eighteen checks and it is the most subjective one. Since 2024, any container where odour is the deciding factor gets a second grader.
- Regulatory classification. Nothing about whether a residue is RCRA-regulated can be determined by smell, and we route anything uncertain to a licensed partner rather than forming an opinion.
“The nose is a detector, not a decision. It tells you to stop and look harder.”
— What we tell new graders in their first week
How we keep it honest
- Smell at the top fitting after the cap is cracked, never inside the opening, and never on a container with an undeclared residue.
- Two graders on any container where odour decides the outcome. Ninety seconds, and it removed our most-disputed check.
- Always compared against the residue ring in the valve body, which is a visual check of the same question.
- Recorded against the serial as an observation, not as a finding. If it disagrees with the declaration, that disagreement is the record.
- Never used as a reason to upgrade a container. It can stop a release; it cannot authorise one.
Used that way it is an ordinary industrial inspection technique, no stranger than a machinist listening to a bearing. Used the other way — as a substitute for documentation — it is exactly the sort of shortcut that puts a reconditioner in somebody else's audit finding.
The other seventeen checks
Published in full, in the order a grader runs them, with what each one fails to.