How to Fix a Candle That Is Sweating Fragrance Oil

Triage first: beading, film or layer

This page does not diagnose. The sibling guides already do that job properly: why is my candle sweating oil covers the mechanism, why fragrance oil sits on top of the candle covers the surface case, separation in the jar covers what happens below the surface, and hot weather and sweating covers the Indian-summer case specifically. Read whichever fits, then come back here. What follows is repair.

Triage properly and you will usually find your batch splits three ways rather than being uniformly ruined. Do the inspection cold — a candle brought in from a warm van and inspected at 34°C will look far worse than the same candle at 24°C — in daylight, with the lids off, and tilt each jar about 20 degrees to see whether anything actually moves.

One more sorting step before you repair anything: weigh what you actually poured. Pull your batch sheet, or reconstruct it from the wax bag and the oil bottle. You are trying to answer a single yes/no question — was the fragrance load inside the wax's stated ceiling or not? Luxury Soy Wax Chunks are formulated to hold up to 13%. Luxury Soy Wax CSI 464 is formulated to hold up to 10%. Those are the wax manufacturer's formulation limits, not regulatory limits, and they are not interchangeable between waxes. The answer to that one question decides which of the three repairs below can possibly work. If you are not sure, work through the over-load check before you melt anything.

Fixing the candles you already have

Three repairs, in ascending order of effort and descending order of how often you will need them. Each one works on a specific cause and fails on the others, so match the repair to the triage result rather than working down the list.

The dilution arithmetic, worked at real wax prices

Most makers discard an over-loaded batch because they have never sat down and costed the alternative. It is worth doing once, properly, because the answer is not close. All prices below are CSI live pricing, September 2026, and the linked product pages are authoritative.

The worked case. A 20-jar festive run, poured as one melt: 5 kg of Luxury Soy Wax CSI 464 and 700 g of Sandalwood Fragrance Oil, filled at 200 g per jar. Total mass 5,700 g. The load is 700 ÷ 5,700 = 12.28% against a stated ceiling of 10%. Three days after pouring, every jar has a continuous film and four of them have a visible layer. Nobody measured anything by weight; the oil went in "by the bottle".

Apply the formula at a 10% target: 700 ÷ 0.10 = 7,000 g of finished candle, against 5,700 g in hand, so you add 1,300 g of wax. At CSI 464's kilogram price of ₹507.40 that is ₹659.62 of wax. In practice you buy in the sizes that exist — one 1 kg at ₹507.40 plus one 500 g at ₹260 — so ₹767.40 spent and 200 g left in the bag. Land at 10% exactly and you have no margin for the same measuring error twice; adding the full 1.5 kg you bought takes the batch to 700 ÷ 7,200 = 9.72%, which is where you actually want to be.

Read the last row twice. Diluting this batch does not merely salvage it — it produces a lower materials cost per candle than the batch would have had if nothing had gone wrong, because the fragrance is now spread across 35 candles instead of 28 and fragrance was the expensive half. You spend ₹1,139.10 to avoid writing off ₹6,055.76 and you finish with seven more units to sell. That is the arithmetic that should decide this, not the sinking feeling in the workshop.

The same arithmetic on the higher-ceiling wax: a batch of 3 kg of Luxury Soy Wax Chunks with 600 g of oil is 3,600 g at 16.67% — well past the stated 13%. At a 13% target the formula gives 600 ÷ 0.13 = 4,615 g, so 1,015 g of wax. Buy 1 kg at ₹599 and you land at 600 ÷ 4,600 = 13.04%, which is over. Buy 1 kg plus 500 g — ₹898.00 — and you land at 600 ÷ 5,100 = 11.76%, which is inside with room to spare. Always buy the next size up. The wax you do not use keeps; the batch you land at 13.04% does not.

Fixing the recipe so it does not happen again

Rescuing the batch on the bench is the smaller half of this job. If the recipe stays as it is, you will be back here in six weeks with the next run. Six steps, in order, once per scent.

That is not a throwaway line, so here is the number attached to it. Take a 200 g candle poured at 12% on a 10% wax — the exact recipe that produces the sweating this page exists to fix. Bringing it to 10% removes 4 g of oil per candle. On Sandalwood at ₹5.03 per gram that is ₹20.11 less oil in every candle, and across a 500-unit festive run it is ₹10,053.60 of fragrance oil we do not sell you. We would rather lose that than have you lose the run — a maker whose Diwali batch sweats in the carton does not order again in January. The oil sitting on top of the wax was never in the candle in any useful sense anyway; you paid for it, it separated, and it made the candle worse. Buying better oil and less of it beats buying more of it every time.

The equipment that prevents the repeat

Every failure mode on this page traces back to a number nobody measured: the grams of oil, the temperature of the melt, or the length of the stir. Three instruments fix all three, and two of them cost the same.

₹708 is less than 12% of the materials value of the single batch costed earlier in this guide, and it is a one-time purchase that applies to every batch you pour afterwards. Note that the weighing scale was down to two units at the last stock pull, which is worth knowing if you are ordering for a festive run. Beyond the three essentials, the optional upgrades that genuinely reduce sweating are the ones that stabilise temperature and mixing: a Mini Electric Wax Melter at ₹2,360 holds a melt in band instead of chasing it on a stove, a 600 ml Steel Pouring Pitcher at ₹708 pours without the dribbles that force a reheat, a CSI Pro Mixer at ₹349 makes a genuine two-minute incorporation painless, and bamboo stirring sticks at ₹118 for 5 are the cheapest thing on the bench. Everything here lives in Candle Accessories.

What not to do

Three of these circulate constantly in maker groups. All three make the situation worse, and one of them is how a fire-behaviour risk gets sold to a customer.

Two more worth stating quickly. Do not mix waxes in a rescue melt — a blend of two waxes has neither of their stated ceilings, so you have replaced a known problem with an unknown one. And do not skip the test burn after any remelt or dilution: cure the repaired batch properly, then burn one candle from it end to end before the rest go anywhere near a customer. If the repaired batch still sweats, the load was not the cause and the answer is in the separation guide rather than in more wax. If you are pricing the labour of a rescue melt into your unit cost — and you should — the 2026 pricing guide shows where it belongs.

Frequently asked questions