Why Is Fragrance Oil Separating From My Candle Wax?

Separation, oil sitting on top, or sweating — three faults, one family

Makers describe all three with the same sentence — "there's oil coming out of my candle" — and then get given advice for the wrong one. The distinction is not cosmetic. Each fault points at a different variable, and the fix for one does nothing for the others. Put plainly: separation is a mixing failure, oil on top is a single expulsion, and sweating is a repeating expulsion triggered by heat. If you can say which clock your fault is on, you have already narrowed six causes to two.

One honest qualification: none of these three faults proves any single cause. A ring at the glass can be the glass; a wet patch can be a pour into a cold jar. The clock only tells you which shortlist to work, and this page is the shortlist for the fault still visible in the jug.

What one kilo of wax is actually stated to carry

Solubility is the whole story, so it is worth having in grams rather than percentages. A percentage is an abstraction; grams are what your scale shows. The table below converts every stated load in the CSI wax range into grams of fragrance oil per kilogram of wax, which is the number you can weigh against. Where a page states nothing, the row says so — and a blank is not permission, it is an instruction to test.

There is a second ceiling most makers never look up, and when the two disagree the lower one is the one that binds. Your oil page publishes a candle rate of its own, and CSI's are specific: Jasmine Fragrance Oil states "Up to 10% for soy wax" and "Up to 6% for paraffin wax"; British Rose states the same pair; Lemongrass and Lime states "6-10% of total wax weight for candles. 10-15% for wax melts". A 13% wax does not license a 10% oil, and a 13% oil does not raise a 10% wax. Two published figures, lower one wins, every batch.

Six causes of separation, in the order worth checking them

Order matters because the cheap checks are also the common ones. Work down this table and stop at the first row that describes your bench. Every row has a test that produces a yes-or-no answer rather than an impression, because "it looked fine" is how a separated melt becomes a hundred jars.

Notice what is not on that list: the age of the bottle, or "a bad batch". Those are the explanations makers reach for first, because they are the only ones that are not the maker's fault — and they are the least likely. CSI's oil pages publish no shelf life and no storage specification, so a bottle that separates in wax is almost never a bottle that has gone off.

The jug test: find separation before it becomes stock

This is the whole commercial argument of the page. Separation discovered in the jug costs you the oil and five minutes. Separation discovered in a labelled, boxed, cured case costs you the oil, the wax, the glass, the wick, the label, the box, the hours and — if one got out — a customer. The procedure below adds about three minutes to a pour and is the reason experienced production benches almost never ship a separated candle.

Salvaging a separated batch — and knowing which salvage you need

If the cause was temperature or stirring, the oil is all still in the vessel and simply has not been distributed. Re-melting the candles inside the page's stated band and stirring properly can dissolve what was never dissolved, and nothing has to be added. If the cause was dose, re-melting achieves nothing at all, because the surplus will come straight back out — the only route is to add wax until the ratio is inside the stated figure, which is a different operation with a different cost. And if the cause was contamination, there is no rescue: water, a diffuser base or an unknown additive cannot be taken out of wax.

Oils whose pages give you the numbers a dissolution problem needs

When separation is the question, three things on an oil page earn their place: a stated candle load so you know the dose, a stated flash point where the page has one so you can talk to your wax page's add band intelligently, and a stated vanillin figure so you can tell a discolouration question apart from an oil-migration one. Per-candle figures below assume a 200 g candle at an 8% load — 16 g of oil — at 1 kg pricing, and scale proportionally if you pour to a stated ceiling instead.

The ₹826 of bench kit that turns separation into a variable

Every cause in the table above is a measurement problem wearing a chemistry costume. You cannot hold a temperature you are not reading, you cannot repeat a dose you did not weigh, and you cannot stir for two minutes without a clock. Three items fix all three, they are bought once, and they are reused for every batch for years.

Set that against the batch. Twenty-five 200 g candles on CSI 464 with Jasmine at a stated 10% load carry ₹4,498 of wax and oil before a single jar, wick or label is counted. The kit is 18% of one batch, spent once, against a fault that can take the whole batch. That is not a close call, and it is the same argument a production bench makes when it buys a second thermometer so that nobody has to share.

One last habit, and it is free: change one thing per batch. Raise the add temperature, extend the stir and switch the wax in the same pour and you have learned that something worked — not what, which means you cannot repeat it next season.

Frequently asked questions