Every fragrance-loss conversation ends at temperature. It should end at relative humidity, too. Temperature sets how fast fragrance oil leaves the wax. Humidity sets how much of what it leaves survives the air. Most warehouses run a 5 to 10 point RH swing that nobody logs, and that swing is usually what separates a candle that throws on month one from one that throws on month ten. Not the wax. Not the oil.
The band, and why it moves the number
The working band we run is 45% to 60% RH at 15 to 24C. Most retail storage guidance lands on the same 60 to 75F (15 to 24C) temperature range. The humidity half of the recommendation is a bare “dry” with no number attached. It should have one.
Fragrance oil is a mix of volatiles. The rate at which the lightest ends leave the wax into the head-space is a function of temperature. Once they are in the head-space they either re-adsorb back into the wax, or they stay in the air and leave the jar. Humidity decides which of those two paths dominates. Above 70% RH the head-space is close to saturated, so the re-adsorption rate drops and net loss accelerates. Below 45% RH the same oil dries out of the wax faster than the container can seal it back in, and the wax surface goes waxy and dull.
What happens when you step out of the band
In our factory runs we have tracked two failure modes that do not show up on a bench burn test, but show up on month four in a customer kitchen.
Above 60% RH, the first symptom is a soft wax surface. Soy in particular takes on a tacky film within two or three weeks in a 65% room. That film is where the lightest fragrance molecules sit. It is also the film that catches dust. The “dusty jar” complaint and the “smell is thin” complaint arrive together. They are one problem, not two.
Below 45% RH the failure looks cleaner, and is worse. The wax surface stays dry, so the month-one smell test passes. By month three the top centimetre of wax has lost enough light molecules that cold-throw at the shelf is fine and hot-throw on the first burn is under. That is the “smells strong on the shelf, weak in the living room” pattern. Humidity is the only variable that changes between production and storage on most batches.
The 12-month curve
The 12-month number is not a marketing figure. It is the point at which a candle stored inside the band has lost roughly 15% to 20% of its top-note load and is still inside the design spec of the blend. Top-note load is the sum of the volatiles in the top third of the oil blend. Losing 15% of it is a spec. Losing 35% of it is a redesign. Outside the band, the same candle sits at 30% to 40% loss by month six. No month-one burn test will catch that, because the light ends have not left the wax yet. They are being moved.
The control is not the wax. The wax is fixed after the pour. The control is the log. We keep a data logger in the middle of the racking, not at the door and not by the HVAC return, and we flag any hour above 60% or below 45%. The monthly review is one sheet: hours outside band, per month, per racking location. A month where the flag count crosses the threshold gets a re-sniff of ten random jars two hours after burn. If the hot-throw is under the batch baseline by more than the design tolerance, that lot is regraded before it ships. The regrade is not a scrap. It is a channel change. A month-six candle that still smells right ships as-is. One that has lost more than 25% of top-note moves to short-turnover retail, not to the 12-month shelf.
That is the whole spec: two numbers, one log, one monthly review.