My Fragrance Addition Temperature Is Correct at the Top of the Melter but the Bottom Is Hotter — Can Temperature Stratification Occur in a Large Batch?

Yes, a large batch can layer by temperature

Temperature stratification simply means the wax sits in layers at different temperatures. In a small pitcher it barely has room to happen; you stir once and the whole mass is the same. In 15 or 20 kg of melted wax it can persist for a surprisingly long time, because wax is viscous, conducts heat poorly and is rarely stirred between melting and fragrance addition.

Two situations produce a hotter bottom. The first is active heating: if the element in or under the base is still working, the wax nearest it is the hottest in the vessel, and the natural circulation that would carry that heat upward is slow in a thick liquid. The second is surface cooling: with the lid off, the top layer loses heat to the room, so the top reads lower even when the bulk has not changed. Often both happen together, which is exactly the picture you describe: the top says the wax is ready for fragrance, while the bottom has not yet come down.

What stratification does not mean is that the melter is faulty or the wax is behaving strangely. It is ordinary physics in a large, still, heated liquid. The problem is only that a reading at the top is being used as though it described the whole batch.

Why it matters at the moment fragrance goes in

Your addition temperature is chosen so the oil meets wax in a known range. If the top is in range and the bottom is several degrees hotter, the oil you pour in and stir down will meet wax outside that range. How much that matters depends on how far apart the layers are and on the oil.

Take the worked example for this page: a 15 kg scented batch at 8% of total weight in the 20 kg melter, made from 13.8 kg of CSI 464 and 1.2 kg of Royal Oud, filling 75 candles of 200 g. The wax costs 13.8 × ₹507.40 = ₹7,002.12. Royal Oud's 100 g size was out of stock when prices were taken, so the 1.2 kg is bought as a 1 kg bottle at ₹6,400.00 and a 500 g bottle at ₹3,300.00, ₹9,700.00 for 1.5 kg, with 300 g kept for the next batch. That is a lot of money in oil. Royal Oud's page states its flashpoint only as above 93°C. If your addition temperature is in the 80s at the top and the bottom is running ten degrees hotter, the lower layer is close to that stated figure, and the melter's own display, with its stated ±10°C deviation, cannot tell you which is true.

Royal Oud's page also says oud takes longer to express and suggests a 72–96 h cure. That matters later, when you judge the candles: a slow-developing oil is harder to assess, so you want the process to be as controlled as possible before the oil goes in, not argued about afterwards.

Measure the profile in your own melter

Before changing anything, find out how much layering your melter produces at your fill level. Use a separate thermometer long enough to reach the base without your hand going near the rim, keep the tip off the metal, and hold it still until each reading settles.

The readings are illustrative, but the pattern is typical. Before stirring the batch looks ready from the top and is not. A slow stir that reaches the base narrows the spread a great deal, and a second stir finishes the job. Your own profile may be steeper or flatter; the point is to know it.

Fill level changes the picture

Repeat the profile at the fill levels you actually use. A 20 kg melter holding 15 kg has a deep column of wax above the base, and a steep gradient has room to form. The same melter holding 6 kg has a shallow column that evens out faster but may overshoot more easily, because the element is working on far less mass. Makers who profile only their usual batch are caught out the first time they run a small top-up batch or a large seasonal one. Two profiles, one near your largest fill and one near your smallest, are usually enough to show the range. Keep both on file beside the melter so the operator can see which one applies. If you ever change where the melter sits, say from a warm corner to a spot under a fan, profile it again: the surface cooling rate, and with it the whole gradient, depends on the air around the vessel as much as on the element beneath it.

Prove whether the layering affects your candles

The profile tells you the layers exist. A control tells you whether they matter. Make a 1 kg batch from the same CSI 464 and Royal Oud lots at the same 8%. Heat it evenly in a small vessel, stir, and divide it. Add the oil to one half at the top-of-melter reading from your profile and to the other half at the bottom reading, stirring both the same way. Pour both at the same pour temperature, cure them together for the page's stated 72–96 h and again at a later point you choose, such as day 7, as a suggested starting point. Compare blind, by smell only.

If the two halves are indistinguishable, the layering in your melter is a safety and consistency concern but not a scent problem at the spread you measured. If the hotter half is flatter or different, stratification at addition is a real cause, and your production candles may be uneven depending on where in the batch the oil met the hot layer.

What to do with a batch scented on a top-only reading

Controls that remove the layering risk

The fixes are procedural and cost almost nothing. Bring the batch to temperature, then lower the setting to hold before approaching the addition point. Stir slowly to the base, wait about a minute (a suggested starting point), and read at two depths: just under the surface and near the base. Add oil only when both are inside the addition range and the difference between them is within a spread you have validated, with about 2°C as a suggested starting point. Keep the lid on between steps to slow surface cooling. After adding oil, stir to the base again, so the oil does not stay in the top layer.

Write two depths on the batch sheet, not one. That single change turns a hidden layer into a visible number, and it is the cheapest control on this page. Never add oil with the melter at its maximum setting, and keep a large melter on its own socket, since its electrical load is well above that of a hotplate or mini melter.

Where this page fits

CSI already has guides that cover parts of this question. This page covers the failure investigation for the question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions