My Previously Smooth Candle Tops Became Rough After Opening a New Wax Batch — What Controlled Comparison Should I Run Before Changing My Formula?

Leave the formula alone until the comparison is done

A smooth top that turns bumpy, pitted, cratered or cauliflower-textured is one of the most visible changes a candle can make, and when it coincides with opening a new bag of wax, the instinct is to act. Raise the pour temperature, lower it, change the oil load, try an additive, reach for the heat gun. Each of those might help. Each of them also destroys your ability to learn what actually happened, and if the new lot was not the cause, you will have changed a working formula to fix a problem that was somewhere else.

The goods-in approach is the opposite. Quarantine the new lot, keep your formula frozen, and run one small, designed comparison that tells you whether the roughness comes from the wax itself, from the wax interacting with your fragrance, or from something in the process that changed at the same time as the lot.

Rough tops generally trace back to how the surface cooled and crystallised. Pour temperature, cooling speed, air trapped by stirring, jar and room temperature, fragrance behaviour and the wax lot all feed into that. A well-designed comparison isolates the lot from the others.

The comparison design: two lots, two conditions

The most informative small test separates two questions at once. First, is the roughness present in the wax on its own, without fragrance? Second, if not, does it appear only when your fragrance is added? That gives a simple grid of four cells, with jars in each so a single odd jar cannot mislead you.

Everything other than the two variables in the table is held fixed: the same jars from the same box, the same melt temperature checked by thermometer, the same pour temperature, the same stirring method, the same bench, and the same curing shelf. Pour in an alternating order — A, B, C, D, then repeat — so no cell gets all the warm first pours or all the cool last ones. Label bases only and have someone else arrange the jars before you assess them.

Running it without introducing new variables

If you normally finish tops with a heat gun or a second pour, that is fine for production. It is not fine for this test, because a repair step can hide a real difference between lots. Once you know the cause, you can decide whether the repair step is acceptable as a permanent part of the process with the new lot.

Wick all twelve jars exactly as you would in production, with the same wick, tab and centring method. A wick changes how heat leaves the surface as the wax sets, and a jar poured without one can set differently from the candle you actually sell. The comparison is only useful if every jar in it is a real product in all respects except the two variables under test.

Worked example and how to read the grid

Say your product is a 200 ml salsa jar with a golden lug cap. At the 100-piece rate those jars cost ₹2,950.00 ÷ 100 = ₹29.50 each, so the twelve-jar comparison uses 12 × ₹29.50 = ₹354.00 of glass, plus wax, oil and wicks for twelve candles. Most of that can be sold as seconds or used as your own reference candles, so the true cost is lower still.

Now suppose the scores come back like this. Cells A and C, the old lot, are smooth. Cell B, new lot unscented, is smooth. Cell D, new lot with fragrance, is rough in all three jars. The wax is not rough on its own, and your process still produces smooth tops with the old wax. The roughness only appears when the new lot and your oil meet. That is an interaction, and the next step is a narrow one: a small pour-temperature ladder on the new lot with fragrance only, not a new formula.

Other patterns point elsewhere. If B and D are both rough while A and C are smooth, the new wax sets rough by itself, and the lot is the cause. If A and C are rough too, the old lot does the same thing today, and the change is in your process or conditions — room temperature, a drifted thermometer, a new stirring habit, jars stored somewhere cold — not in the wax. If everything is smooth, the rough batch you saw earlier was a one-off, and the comparison has cost you twelve candles to prove the lot is fine.

From grid to goods-in decision

Match the decision to the pattern. Accept when all four cells are smooth, or when only A and C are rough, which clears the lot and points you at your process. Conditional release when only D is rough and a narrow pour adjustment restores a smooth top in a follow-up pour; release the lot with that adjustment written into the batch sheet for this lot only, and keep checking. Quarantine when B and D are rough, because the new wax sets rough on its own. Send your grid scores, photos of all twelve tops together, the invoice number and the date received to the supplier — for CSI on WhatsApp at +91 7397976926 — and record the reply. CSI does not publish crystallisation, surface or per-lot tolerance data, so your grid is the evidence. A lot that sets rough on its own goes back as a question whoever supplied it, including us.

Only after the grid has told you where the problem lives should you consider changing the formula, and even then, change one thing at a time with its own small comparison. Keep the twelve jars, labelled, for at least as long as you use this lot. When the next delivery arrives, a single jar from cell A beside a fresh pour of the new lot will tell you in a day whether you are back to smooth or starting again.

Where this page fits

CSI already has guides that cover parts of this question. This page covers the goods-in check for the question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions