Light the candle and the room fills. Ten minutes later it sounds like the flame gave up. Customers call this "faded" and blame the batch. It is not a defect. It is the note pyramid doing exactly what the chemistry says it will do, and the formulation never accounted for it.
Top notes are the volatile fraction. Citrus terpenes, light aldehydes, some green accords. They vaporize at room temperature, which is why the scent hits first and hardest. In a burning candle the effect is multiplied: the flame zone runs 600 to 1,000 degrees Celsius while the wax pool sits around 50 to 60. Top-note molecules spend minutes, not hours, in the air before the flame reclaims them as heat.
So what the customer calls throw is really two different events: a 10-minute top-note flash, then a long, quiet base-note burn. A formulation built around the flash gets reviewed as "smells weak after a while." A formulation built around the second event keeps the room honest for the whole session.
In our own test room we time this. We burn a 250 g jar on a controlled wick, log a blind sniffer score every 15 minutes, and watch the classic curve: peak at minute 3, drop of roughly half again by minute 15, then a slow floor for the rest of the burn. The curve shape tells you the blend ratio before any customer complaint ever arrives.
The ratio that actually holds
Working rule from the bench: for a candle that must smell like itself at hour two, build the blend so no more than 30 to 40 percent of the aromatic mass is top- and light-middle material. The rest carries the session. Vanilla, cedar, oakmoss, labdanum, a little sugar-type base. These are slow molecules, and they are what the room keeps smelling.
If the client wants a loud, bright opening, add the top notes, but dose them for a flash, not a promise. Three to five percent of the total fragrance blend, not 20. The opening sells the candle. The base holds it.
What the burn itself does to the notes
Three levers move the top-note burn-off, and all three are on the customer or the label, not in the formula:
Wick length. Trimmed to 3 to 5 millimeters before the burn, the flame is small and steady, and the pool heats gradually. Left long, the flame is bigger, the pool runs hot, and the volatile fraction goes up in the first minutes. We have measured a visible throw difference between a 5 mm and a 12 mm wick on the same jar, same blend, same room.
Burn session length. The first burn should run until the pool reaches the glass wall, typically 2 to 3 hours for a standard 70 to 80 mm jar. Shorter sessions tunnel the wax and trap base-note oil in the dead zone, so the next burn starts with even less going on.
Room temperature. Cold rooms slow the base notes more than the top notes, which makes the fade worse. A 16 degree room in winter is a different candle than a 22 degree room in August, same product.
One thing our lab watches closely: the first burn is not a burn test. It is a burn-off test. Everything that can go up quickly, goes up. What remains on day two and three is the real product. Judge the candle there, and the "it died after ten minutes" complaints stop arriving.
The customer does not need a louder candle. They need a formulation that keeps its shape for the hour they actually sit in the room. Build the base, dose the flash, and the review says "still smells good at hour two." That is the review that reorders.