My Candle Surface Becomes Oily When Stored Above 30°C but Returns to Normal When Cooled — Is the Product Still Stable Enough to Sell?

What recovery tells you — and what it does not

Credit where it is due: the usual version of this question is "my candles sweat in summer, what wax should I buy", which is unanswerable because it contains no information. You have watched the same candle through a temperature change in both directions and noticed that the change undoes itself — and that splits two mechanisms which look identical in a single photograph.

A container candle is a crystalline solid holding a liquid in solution — fragrance oil distributed through a wax lattice, in the sense that sugar is held in water. Warm it and two things change: the lattice becomes mobile, so molecules that were locked can move, and the amount of oil it holds in place at that temperature changes. Oil that can no longer be held migrates, and the shortest route out is straight up to the free surface. That is your film.

What happens next is the whole question. In the first mechanism the surplus was surplus only at that temperature: cool the candle, the lattice tightens back around the oil, the film is drawn into the body of the wax, and the surface reads clean. Nothing left the candle — mass unchanged, fragrance unchanged, finish restored. In the second, the oil that reached the surface does not return. It pools, wets the label from underneath, some evaporates, and where it sat the structure re-sets differently — a matte patch against gloss, a shallow crater, a faint ring. The candle is permanently marked and weighs slightly less.

So reversibility is good news. Now the honest part: recovery is not the same as no consequence. Four things can be true at once — the surface recovers, and this is still not a candle you want to post in May.

The twenty-cycle test, in full

A single cycle proves almost nothing, and that is the flaw in nearly every stability test a small brand runs: one cycle tells you the candle survived one cycle. You need to know whether the recovery is repeatable and whether anything is accumulating underneath it — which means the same units going round the loop many times while you write down both ends.

Beyond the film there is a checklist makers miss because they are watching the shiny part. Score each on your own scale at every cooled observation, the same way every time: surface gloss against the original, noting whether loss is cumulative; wet spots and whether the wax re-adheres to the glass; wax movement — slumping, a convex top; label and sticker adhesive at every corner; lid residue, because oil on a rim transfers to the customer's hands; colour, since liquid dye carried in migrating oil need not come back where it started; and cold throw character, sniffed at the same hour each cycle.

The output is your written pass or fail, and this page deliberately does not set it. Write the standard down before cycle 1, because a standard written after the results is not a standard. Something like: "passes if at every cooled observation the surface is indistinguishable from the control at arm's length in daylight, if mass loss across twenty cycles is under X grams, if no label lifts, and if the film has gone within Y hours." You choose X and Y — they belong to your brand and your price point. A ₹350 market-stall candle and a ₹1,500 corporate gifting candle do not share a pass mark, and pretending they do is how brands either scrap good stock or ship bad stock. Same discipline, buying context: how to verify wax quality before you buy in India.

Five questions that make the decision yours

With twenty cycles of data the remaining work is commercial, not physical, and it is genuinely your call. CSI will not tell you a product is or is not fit to sell — that belongs to you as the brand on the label, and no published standard exists that either of us could be measured against. Here is the frame, with costs attached where a cost exists.

The one specification you can actually fail

There is exactly one number here where CSI publishes a limit you can be measured against: fragrance load. Three of the sixteen waxes state a maximum; thirteen state nothing. Check yours before you cycle anything, because a formula above its published ceiling has failed a stated specification however gracefully it recovers, and no amount of cycle data makes that go away.

Work it in grams, not percentages, because percentages are where the slips happen. A 200 g candle at 12% is 24 g of oil and 176 g of wax. On CSI 464, whose listing states up to 10%, that is 4 g per candle above the published ceiling — with British Rose at ₹3,068 per kg, ₹12.27 per candle or ₹3,681.60 across 300, spent on oil the listing does not say that wax will hold. On Luxury Soy Chunks or CSI 468 the same 24 g is inside the stated 13% and fails nothing. The specification belongs to the wax you bought, not the candle you designed.

Now the harder half. If your load is inside the ceiling, or your wax publishes none at all — thirteen of the sixteen — there is no specification left to fail, and equally none your 30°C observation can be checked against. CSI publishes no softening point, no heat-resistance figure, no storage-temperature limit and no summer-stability claim for any wax. This page will not paper over that with a plausible number, because a plausible number is worse than none: you would plan a production run around it. Where there is no published standard, the only standard is the one you measure.

Two of the sixteen waxes publish a shelf life and both say the same thing — 60 months in cool, dry storage — on CSI 464 and CSI 468. Read it for what it covers: flakes in a sealed bag, not a poured candle in a customer's cupboard. The other fourteen publish none. And where a jar's fill capacity is not published — the containers collection runs to 80 products with uneven detail — pour and weigh: fill one jar to your intended line, weigh it, use your own number.

What the test costs against what it protects

Three weeks of cycling feels expensive with stock sitting in boxes, so here is the cost at live pricing. The recipe is a 200 g candle at a 10% load — 20 g oil, 180 g wax — in CSI 464 with British Rose, sitting exactly at that wax's published 10% ceiling. Substitute your own numbers; the shape does not change.

Now the run it protects. The same recipe at 300 candles needs 54 kg of wax and 6 kg of oil: wax, 54 kg at ₹507.40 = ₹27,399.60; oil, 6 kg at ₹3,068 = ₹18,408.00; jars, 300 at ₹47.20 = ₹14,160.00; wicks, 300 at ₹5.90 = ₹1,770.00; wick stickers, 8 sheets of 40 at ₹100 = ₹800.00. That is ₹62,537.60 of materials, or ₹208.46 per candle, before labour, packaging, courier, dye or overhead.

The last thing on cost is the part that sells you nothing. A recovering surface is frequently acceptable, and the honest advice to a good many makers reading this is that you may not need to change anything at all. If twenty cycles come back with stable mass, a baseline that does not drift, labels down, no wet spots and a recovered surface you cannot pick out beside the control, the candle is behaving the way a wax-and-oil solid behaves when warmed and cooled. Write the standard down, keep the recipe, stop worrying. That buys no wax, no oil, no additive and no jar — and a supplier who tells you otherwise before you have cycled anything is selling rather than advising. That includes us. CSI has published no heat-stability data, so nothing we said by way of reassurance would have a measurement behind it — the same reason to refuse any other supplier's. Further reading: scaling to 1,000 orders a month and where professional candle brands buy their raw materials.

Frequently asked questions