My Candle Sweats Even Though the Fragrance Load Is Within the Wax Supplier’s Recommendation — Could My Fragrance Addition Temperature Be Wrong?

Within the recommendation is a claim worth checking

You followed the number. The candle sweats anyway. Before you suspect the temperature, confirm the number you followed is the number you think it is, because the most common mismatch in this series is not chemistry at all. It is arithmetic.

CSI wax pages phrase fragrance load as a percentage of wax weight. This series, and most careful production sheets, use a percentage of total candle weight. The two are not interchangeable. A wax-basis figure always looks larger than the same oil expressed against the whole candle.

Take the worked example used on this page: a 200 g candle at 9% of total weight. That is 18 g of oil and 182 g of wax. Against the wax alone, 18 ÷ 182 × 100 = 9.89%. CSI 464's page says 'up to 10%'. So yes, you are inside it, by about a tenth of a percentage point. Inside, but at the edge.

If your own load converts to above the page's figure, stop reading about temperature; you have a load problem first. If it converts to inside, keep going. A wax maker's maximum describes what the wax can typically hold when oil is properly dissolved into it. It does not describe what happens when oil is stirred into wax that is already starting to set, or when the temperature at the moment of addition moves around from batch to batch.

How addition temperature can undo an in-range load

Fragrance oil has to disperse through the molten wax and stay evenly held as the wax crystallises. That depends on the wax being fully liquid and warm enough, for long enough, while you stir. Add the oil when the wax is close to setting, or when the pot has cooled unevenly, and you risk oil-rich zones. Those zones are where the wax cannot hold what it has been given, and they are where droplets tend to appear once the candle has set and cured.

Cooling shape matters too. A small pot loses heat quickly at its walls and surface, so even when the centre reads warm the wax touching the steel may already be thickening. Oil poured onto that skin, then stirred briefly, has less chance of dissolving evenly. This is why one probe reading near the edge and one near the centre can tell you more than a single number.

Notice what this means. The average load in the pot can be 9%, perfectly legal, while small regions are far richer than that. The wax never sees your average. It sees the local mixture.

There is also a quieter version of the same problem: inconsistent addition temperature. If one batch got its oil at a warm, fully liquid point and the next got it after the pot had sat during a phone call, you can produce two different candles from one recipe. The page on separate batches covers that in detail; here the question is narrower. Was the temperature at the moment of addition wrong for this wax-oil pair?

First, find out what temperature the oil actually met

Most makers who suspect temperature have never measured it at the moment of addition. They read a dial. The CSI electric melter page states a display range of 60–100°C with a ±10°C deviation, which means a display reading tells you a band, not a number. A pot showing 80°C could plausibly be anywhere from 70°C to 90°C by the page's own figure.

So the first job is to measure the wax itself with a separate probe, stirred through, at the second you pour the oil in. Write it down. Do the same for the next three batches without changing anything else. If those numbers wander by a wide margin, you have probably found the reason your results wander.

The paired test that separates temperature from load

You want to know whether temperature is the cause without also changing the amount of oil. That means two groups of candles that differ only in addition temperature.

Read the result honestly. If Group A sweats and Group B does not, the addition step is your problem and the load can stay. If both groups sweat, temperature was not the cause, or not the only one, and the load comes down. If neither sweats, your earlier batches probably had something else going on, such as a cold shelf, a draught or a storage spell in a hot room.

What the test costs and what it can save

Using CSI 464 at the 10 kg rate of ₹5,074.00 (₹0.5074 per gram) and Bergamot Amber at the 1 kg rate of ₹5,437.00 (₹5.437 per gram), each 9% candle carries 182 × ₹0.5074 = ₹92.35 of wax and 18 × ₹5.437 = ₹97.87 of oil. Six test candles therefore use 6 × ₹97.87 = ₹587.22 of oil. That figure excludes jars, wicks, packaging, labour, shipping and GST.

Compare that with the alternative many makers jump to: cutting load to 7% without testing. At 7%, oil per candle is 14 × ₹5.437 = ₹76.12. It is cheaper per candle, and it might be the right answer, but if temperature was the real cause you have also lost scent strength you did not need to lose. Test first. Then cut, if the test tells you to.

And this is where we argue against our own oil sales: if the paired test shows both groups sweat, reduce the load, even though that means buying less Bergamot Amber. A 9% candle that sweats is not a stronger product. It is a returned one.

The decision rule

Use one of three outcomes. Temperature fixed it: write your measured addition band into the batch sheet and probe every batch. Temperature helped but did not clear it: keep the controlled addition and run a load ladder at 9%, 8% and 7%. Temperature changed nothing: keep the controlled addition anyway, step the load down, and if 7% still sweats, test a second wax such as CSI 468 or a different oil.

Whatever you find, the candles that sweated are not stock. Hold them, keep them as reference samples, and dispatch nothing from that method until a fresh batch passes the same observation schedule.

Where this page fits

CSI already has guides that cover parts of this question. This page answers one sweating decision; these cover the neighbouring questions:

Three things to buy for this fix

Frequently asked questions