Why Is My Fragrance Oil Sitting on Top of the Wax?

First: is this pour-day oil, set-day oil or sweating?

Before you change a recipe, fix the timing in your head, because the three faults share one appearance and share almost nothing else. Write down when you first saw the oil, in hours or days, and where on the surface it sat. That single note will save you a fortnight of changing the wrong variable.

The third row belongs to a different diagnosis and the live guides on candles sweating fragrance oil and the eight real causes of sweating cover it properly. The fourth row is the useful one to rule out cheaply: if only the last jars off a melt are affected, you have a distribution problem in the jug rather than a load problem in the recipe, and no amount of dilution will fix it.

Cause one: the load is above the governing ceiling

Every CSI wax grade that publishes a fragrance figure publishes a maximum as a percentage of wax weight, and every oil page that publishes a candle usage range publishes it the same way. Both apply, and the lower of the two governs. Luxury Soy Wax CSI 464 states up to 10%. Luxury Soy Wax Chunks and Premium Coconut Soy Wax CSI 468 state up to 13%. Eco Soy CSI 400 publishes no ceiling at all, so on that grade the oil page's number is the only published figure — worth confirming on WhatsApp.

Beyond the governing figure, the wax is being asked to hold oil that its manufacturer does not claim it holds. The oil does not vanish; it separates, and because it is lighter than wax it rises. On pour day that shows as a film. The arithmetic is unforgiving and it is worth doing once, slowly, in grams rather than percentages.

Read the fourth column and you can see the single most common way makers arrive here: a recipe written for one wax, poured into another. A 13% formula is entirely legitimate on Luxury Soy Wax Chunks, which states up to 13%. Move it unchanged onto Luxury Soy Wax CSI 464 and you are 6 g over the stated maximum on every 200 g candle before you have made a single error at the bench. Nothing about the process changed. The number on the page did.

The oil page can lower the ceiling further, and this catches even careful makers. Fresh Strawberries states a candle range of 6–9% of total wax weight. On Luxury Soy Wax Chunks, whose page allows up to 13%, that pairing still governs at 9% — 18 g in a 200 g candle, ₹63.72 of oil. Blue Ocean states up to 10% for soy, paraffin and natural wax blends; on CSI 468, whose page allows 13%, the governing figure is 10%.

Cause two: the oil went in outside the published window

This is a smaller contributor than the ceiling but it is real, and it produces the same film. The Luxury Soy Wax Chunks page asks for fragrance at 80–90°C and warns in its own pro tip that adding too cool causes fragrance to seize; the Eco Soy CSI 400 page names 85°C within an 80–90°C range. The Luxury Soy Wax CSI 464 page gives 80–90°C as the working range for melting and pouring. The Premium Coconut Soy Wax CSI 468 page publishes a melting point of 50–53°C and a pour temperature of 58–60°C and does not publish a fragrance-add figure at all.

The practical consequence for surface oil is simple: below the window, some of the oil is never taken into solution and stays as fine dispersed droplets that rise as the candle cools. You can be at a perfectly acceptable 8% and still get a film if the wax was at 65°C when the oil went in. Take the reading at mid-depth with a Pen Thermometer rather than judging by how the melt looks — wax at 95°C and wax at 78°C are indistinguishable by eye.

Two of CSI's oils publish notably low flash points and deserve a mention here because the handling advice runs the other way. Palo Santo publishes a flash point of 93°C and Gardenia Tuberose publishes 98°C. Flash point is a transport, storage and handling property, not a ranking of throw and not the temperature at which fragrance burns off — but with those two, adding at the lower end of the 80–90°C window rather than the top is sensible handling. Most CSI oils that publish a figure sit far higher: Sandalwood 203°C, British Rose 192°C, Roasted Coffee approximately 191–193°C.

Cause three: dye, additive and the other liquids in the jar

A wax page's stated maximum is a figure for fragrance load. It is not a budget for every liquid you add. Liquid dye, an additive and in some cases a second fragrance all go into the same wax, and a recipe sitting at exactly the stated maximum with a generous pour of dye on top is asking more of the wax than the page describes. Liquid dyes are concentrated — CSI dye reviewers note that a little gives a good pastel shade — so the correction is almost always a drop fewer rather than a different product.

The Vybar Additive page is worth reading if you are working near a ceiling, because it claims to help wax hold above-standard fragrance rates and to prevent mottling and sweating. Two honest caveats. Those are page claims, not measurements taken here. And an additive is not permission to exceed a wax page's stated maximum — if you want to run close to the top of a ceiling, test it in a controlled batch and burn it before you sell it, exactly as the Chunks page asks you to test burn thoroughly at 13%.

The last liquid worth naming is water, and it is the one case where dilution does not help: a hazy, spitting melt has water in it, and that batch should be kept back as testers rather than poured for sale. It is rarer than the other causes and it looks different — cloudiness through the body of the wax rather than a clear film on top.

The batch rescue: the arithmetic, the jars and what it costs

The rescue is a remelt with dilution, and the only thing you need to get right is the arithmetic. Total oil ÷ governing ceiling = the wax you need. Everything else is handling. Work in grams, not percentages, and write the numbers down before you turn anything on.

The bottom row is the honest headline. ₹168.74 a jar is exactly what those candles would have cost you if you had weighed the oil correctly the first time — ₹101.48 of wax, ₹61.36 of oil and ₹5.90 of wick. The rescue does not lose you money per candle; it costs you ₹101.48 of extra flakes, one more jar and one more wick, and about forty minutes. Set against writing off ₹306.80 of oil, it is not a close call.

When to dilute, and when to write the batch off

A remelt is cheap in materials and expensive in two things a small candle business rarely has: bench time and cure time. Before you start, be honest about whether the rescue is the right call, because pouring a rescued batch into a festive order that ships in four days is a worse decision than losing the wax.

There is a fifth question that sits behind all of them, and it is about what the batch is for. Candles poured for a wedding hamper, a corporate order or a Diwali drop carry your name into somebody's home. If you cannot burn one of the rescued jars for three hours before the rest ship, do not ship them. Pour the rescue into testers, keep them, and make the customer's candles from a fresh, correctly weighed melt. The wax you lose is ₹405.92. The order you keep is worth considerably more.

Preventing it: four checks that live at the bench

None of these takes more than a minute, and together they remove almost every route to a tray of oily jars.

1. Weigh the oil on a scale, every time

Not by eye, not by counting seconds of pouring, not by emptying what is left in a bottle. Weigh into a cup, then add the cup's contents to the melt. A scale mistake costs you a re-weigh; a bottle mistake costs you a batch. CSI's Candle Making Weighing Scale is ₹354.00, though only two units are in stock at the time of writing — any accurate kitchen scale reading to the gram does the job.

2. Re-check the ceiling whenever you change wax grade

This is the single highest-value habit on the page. Recipes travel between waxes and ceilings do not. Write the governing ceiling on the recipe card next to the grade name, not just the percentage: “Luxury Soy Wax CSI 464 — 10% max — 20 g per 200 g jar” is a line that cannot be misread six months later. When you switch to Luxury Soy Wax Chunks at 13%, the line changes with it.

3. Pour one, wait ten minutes, then commit

Covered above, and worth repeating because it converts a tray-wide fault into a jug-wide one. The melt is still hot and still correctable for as long as it is in the jug.

4. Log the load, the temperature and the stir

Three numbers per batch: grams of oil per gram of wax, the temperature at which the oil went in, and the stirring time. Without them, a rescue in step one becomes guesswork and the whole arithmetic collapses. With them, any batch on this page can be diagnosed and costed in two minutes.

One last note on Indian benches. In a 40°C-plus workshop in Delhi, Ahmedabad or Nagpur the jars stay molten far longer after the pour than they do in a cool room, which gives any un-dissolved fraction more time to rise and gather. The same recipe that looked clean in a December pour can show a film in May without a single thing having changed in the recipe. If a batch that used to be fine starts showing surface oil as the season turns, check the load against the ceiling before you blame the oil — and if the oil appears days later rather than on pour day, you are looking at a storage and sweating question instead.

Frequently asked questions