Why Is Fragrance Oil Sitting on Top of My Candle?

Beads, film or layer — decide which one you have

Almost every message we get about this starts the same way: "there is oil on top of my candle". That sentence covers three genuinely different conditions, and the decision you make next — wipe it, remelt it, or destroy it — depends entirely on which one you are holding. Do the grading in daylight, with the candle at room temperature, and tilt the jar slowly. What matters is whether the oil is discrete, continuous, or free-flowing.

The reason to be strict about the grading is that the three faults teach you different things. Beads tell you your load is at the edge of what the wax will carry, which is useful information but not an emergency — it is the same conversation as frosting, a surface condition on an otherwise sound candle. A film tells you your process slipped on one batch. A layer tells you the batch was never a candle at all. Treating all three as "sweating" is why makers keep wiping the top of a jar that should have gone back into the melter a week ago.

One more distinction worth making early: this page is about oil that appears on the surface. Oil that has pooled in a cavity under the surface, or a crater that opened as the candle set, is a cooling-and-shrinkage story rather than a solution story, and it belongs with sinkholes and pouring temperature. A candle can have both faults at once. Diagnose them separately or you will change two variables and learn nothing.

Why a distinct layer usually means the oil never went into solution

Molten wax and fragrance oil are not automatically compatible liquids. They form a solution — a single phase where the oil molecules sit distributed through the wax and stay there when it sets — only if two conditions are met at the moment of mixing. The wax has to be hot enough that the oil disperses freely rather than being stirred through as suspended droplets, and the mixture has to be agitated long enough for that dispersion to become uniform all the way through the pitcher, not just around the stirrer.

Miss either condition and you do not get a solution. You get an emulsion that looks like one for about ninety seconds. It pours beautifully. It looks fine in the jar. Then the wax begins to crystallise, the crystals reject the oil they never absorbed, and the oil — lighter and slower to set — migrates upward through what is still liquid. By the time the candle is cool the separation has already happened; you are just waiting to notice it.

There is a third route to a layer, and it is the one that cannot be stirred away: a load well past the wax's stated ceiling. Those ceilings are wax-manufacturer formulation limits, not regulatory limits — Luxury Soy Wax Chunks are rated up to 13%, and Luxury Soy Wax CSI 464 up to 10%. They describe how much oil that specific wax has been formulated to carry. Pour at 18% into a wax rated for 10% and the surplus eight points have nowhere to go. Perfect technique will not change that, because the limit is not about your stirring.

The add-temperature and stir-time evidence

Here is the diagnostic question, and it is deliberately blunt. What temperature was the wax when the oil went in, and how long did you stir it? Not the target — the reading. Not "about two minutes" — the timed figure.

Eco Soy CSI 400 specifies adding fragrance at 80–90°C with two minutes of stirring. Luxury Soy Wax CSI 464 melts and pours in the same 80–90°C band. Those are the numbers on the products you bought. They are not aspirational.

Most makers cannot answer either question, and that inability is itself the finding. If the answer is "I waited until it looked ready" then the add temperature was almost certainly below 80°C, because wax that looks calm and clear in a pitcher is already dropping fast — a 600 ml steel pouring pitcher off the heat loses temperature quickly in an air-conditioned room. If the answer is "I stirred it well", it was between twenty and forty seconds; almost nobody stirs for two minutes unless they are watching a clock, because two minutes with a bamboo stirring stick in your hand feels absurdly long.

Both of those are free to fix and neither requires a different wax or a different oil. That is the good news buried in a bad batch. Before you conclude that the oil is at fault, run the causes in the order set out in the wax, load, mixing or temperature elimination guide — it exists so you change one variable at a time instead of four.

Is the candle safe to burn?

A candle with a visible liquid layer of fragrance oil on its surface should not be lit, and should not be sold. Fragrance oil is more flammable than wax. A candle works because the wick draws molten wax up by capillary action and burns it at a controlled rate at the flame; the wax is the fuel and the wick meters it. A free pool of fragrance oil sitting on top of the wax is not metered by anything. It is a volatile liquid in an open dish with an ignition source standing in it. The flame behaviour you get is not candle behaviour — it can flare, it can run along the surface, and it can involve the whole top of the jar rather than a controlled pool around the wick. That is not a risk worth taking to test a theory about your stirring.

The same logic applies with more force to selling. You do not know what the customer's shelf looks like, whether the jar will be lit near a curtain, or whether it will be left unattended. A maker who ships a jar with free oil in it has transferred an unquantified fire risk to somebody who has no idea it is there. Do not sell it, do not send it as a sample, and do not include it in a hamper as a freebie. The commercial cost of writing off a batch is always smaller than the cost of the alternative.

Two things this section is not saying. It is not saying that a candle that once beaded is dangerous — beads are a surface condition and a wiped, cured candle with no free liquid burns normally. And it is not saying that a film means a fire. A film means the batch is not finished; it means remelt, not incinerate. The line is free liquid. If oil moves independently of the wax when you tilt the jar, that jar does not get a flame put into it.

Can the batch be saved?

Often, yes — and the deciding factor is not how bad the layer looks. It is whether the fragrance load is inside the wax's stated ceiling. Work that out before you switch the melter on, because it determines whether you are doing a remelt or a dilution.

The honest caveat, stated once and clearly: if the load is over the ceiling, remelting will not fix it. No amount of heat or stirring makes a wax hold more oil than it has been formulated to hold. You have exactly two options at that point — add wax until the load falls back inside the ceiling, or discard the batch. Anyone who tells you a hotter melt or a longer stir will absorb an 18% load into a 10% wax is describing something that does not happen.

The dilution arithmetic — how much extra wax brings it back inside the ceiling

This is simple arithmetic and it is worth doing before you decide to bin anything, because the cost of the rescue is usually a fraction of the cost of the write-off. The formula has one step.

The same batch on a wax with more headroom needs much less added. Luxury Soy Wax Chunks are rated up to 13% and cost ₹599 per kg, so 150 g of oil in a 1,000 g batch needs 150 ÷ 0.13 = 1,153.85 g of finished mix, which is 153.85 g of extra chunks at ₹92.16. At 180 g of oil it is 384.62 g of wax at ₹230.39; at 200 g of oil it is 538.46 g at ₹322.54. That three-point difference between a 10% and a 13% ceiling is the entire reason it is worth knowing which wax you are actually pouring.

Dilution versus write-off, in rupees

Take the 15% case and price the materials in it. The batch is 850 g of CSI 464 at ₹0.5074 per gram, which is ₹431.29, plus 150 g of Woody Oud Fragrance Oil at 1 kg pricing of ₹4,531.20, which is ₹4.5312 per gram and therefore ₹679.68. Total materials in the pitcher: ₹1,110.97, before jars, wicks, gas and your time.

Rescuing it costs 500 g of CSI 464 — ₹253.70 of wax, or ₹260.00 if you buy the 500 g pack. That is roughly 23% of what you would be throwing away, and it does not merely recover the batch: it turns five 200 g candles into seven, because you now have 1,500 g of mix instead of 1,000 g. Total materials across the rescued batch come to ₹1,364.67 for 1,500 g, which is ₹181.96 of wax and oil in each 200 g candle. Anyone doing the sums in how to price candles in India will recognise that as the difference between a batch that earns and a batch that does not.

Preventing it: a thermometer, a timer and a recorded load

Everything in the two sections above is remedial. The prevention is three items long and two of the three are free.

Record four numbers per batch and this fault stops being mysterious: grams of wax, grams of oil, the thermometer reading when the oil went in, and the timed stir in seconds. Write them at the bench, during the pour, not from memory afterwards. Four numbers on a scrap of paper is enough. When a batch separates, those four numbers tell you which of the three causes you are looking at in under a minute — and when a batch is perfect, they tell you how to repeat it. Add the pour date and you also know exactly which jars to pull if you have to.

That is deliberately against our own interest and it is still the correct advice. CSI sells fragrance oil — 92 of them — and the two changes that most reliably stop oil pooling on your candles are to buy less of it per candle and to spend ninety more seconds with a stirrer. A maker pouring at 8% with a timed two-minute stir will use less of our oil per candle than one pouring at 14% badly, and will also stop writing off batches, stop replacing jars and keep customers who reorder. We would rather sell you steady volume for years than a spike of oil that ends up in a sealed container in your bin. If you are still working out what your annual figure should look like, how much fragrance oil a candle business needs works it through, and premium versus cheap fragrance oils covers what you actually get for a higher per-gram price.

One last point about wax choice, since it comes up whenever a batch separates. Moving to Premium Coconut Soy CSI 468 at ₹325 per 500 g, or to Luxury Soy Chunks at ₹299 per 500 g for the higher stated ceiling, is a reasonable decision on its own merits. It is not a fix for a mixing failure. If the oil went in at 65°C and got stirred for twenty seconds, it will separate in any wax on the shelf. Fix the process first, then choose the wax for the candle you want to make.

Frequently asked questions