How Long Should I Mix a Fragrance Into Wax Before Assuming It Has Been Uniformly Incorporated?

Why there is no universal answer

Ask ten makers how long they stir and you will get ten answers. Most of them are right for their own pot. A small jug of wax stirred by hand and a 20 kg melter stirred with a long paddle are different problems. The volume to move, the depth, the area of wall and base where cooler wax and oil can linger, and how vigorously each person stirs all change what counts as enough.

CSI publishes one stir instruction among its waxes: the Eco Soy CSI 400 page says add fragrance at 80–90°C and stir for 2 minutes. Treat that as a sensible order of magnitude from a wax maker. It does not cover CSI 464, Chunks, CSI 468 or Pearl, none of whose pages publish a stir time, and it cannot know the size of your pot.

So the honest answer to 'how long?' is: long enough that the top and the bottom of your pot make the same candle. The rest of this page shows you how to find that point.

It also helps to separate mixing from sweating. A well-mixed batch can still sweat if the load is too high for that wax-oil pair, and a poorly mixed one can look fine for days. The test below answers only the mixing question, which is exactly why it is useful: it removes one suspect cleanly before you look at the others.

What 'uniform' means in practice

You cannot measure fragrance concentration in a workshop. What you can measure is behaviour. If wax from the top of the pot and wax from the bottom, taken at the same moment, produce candles with the same surface after curing, the same cold scent and the same colour if dyed, then for practical purposes the batch was uniform at that moment.

That definition is useful because it tests the thing you care about. A batch can look even on top and still have oil-rich wax near the base. The base sample is where under-mixing shows.

The stop-and-sample test

Run it on a real production-sized batch. The worked example is a 5 kg batch: 5,000 g total at 8% of total weight, which is 400 g of Sandalwood and 4,600 g of CSI 464. Sandalwood's page states up to 10% in soy and 0% vanillin, so colour shifts will not confuse the comparison.

Sampling without spoiling the test

Draw the base sample with a clean, pre-warmed ladle or small pitcher lowered straight down to the bottom, so you are not collecting top wax on the way. Pour the samples promptly; a sample that sits in a cold ladle cools unevenly and adds its own error. Label each votive on the base before pouring, because labelling hot glass afterwards is awkward and mistakes are easy. Keep the votives in pairs on the rack so you compare like with like.

A visual tracer, if you use dye

If your product is dyed anyway, add the dye with the oil at the start of the test. Streaks or a different shade between top and base samples are an early, visible sign that mixing is incomplete, often before any surface oil appears. Dye is not a measure of fragrance, and CSI publishes no dye dosage, so use your normal amount and treat colour as a clue rather than proof. If your product is undyed, do not add dye just for this test; it introduces a material your real candles do not contain.

Your stir time is the first checkpoint at which the pair matches and every later pair also matches. If the 1-minute pair matches but the 2-minute pair does not, something else went wrong, such as a sample drawn from a cooler edge; repeat the test rather than trusting a single odd result.

Add a margin, then write it down

A time that only just works on a good day will fail on a busy one. Add a margin above the first matching checkpoint. If your pair first matched at 1 minute, writing 2 minutes into the method gives you room for a slower hand, a cooler workshop or a slightly larger batch. If nothing matched even at 3 minutes, extend the ladder at the next test and look at your stirring technique: are you reaching the base and walls, or circling the centre?

Then record it in the batch sheet with the stirring pattern and the batch size it was validated at. A stir time without a batch size is half a specification.

When you add a new oil to the range, you do not need the full ladder again. Run a quick spot-check: stir for your validated time, sample top and base once, and observe the pair. If they match, the time holds for that oil. If they do not, run the ladder for that oil and note its own stir time on its line of the sheet.

What the validation costs

The samples are small. A pack of ten 70 g clear votive jars costs ₹318.60, so ₹318.60 ÷ 10 = ₹31.86 per jar, and eight of them hold the test samples with two spare. The batch itself is production you were going to make anyway: wax 4,600 × ₹0.5074 = ₹2,334.04 (CSI 464 at the 10 kg rate of ₹5,074.00), oil 400 × ₹5.0268 = ₹2,010.72 (Sandalwood at the 1 kg rate of ₹5,026.80). The 5,000 g fills 5,000 ÷ 200 = 25 candles of 200 g, less the small amount drawn into the votives. Packaging, labour, shipping and GST are excluded.

Hold the main batch until the samples have been read. If the early checkpoints fail and the later ones pass, the main batch, which was stirred to the end of the ladder, is probably sound, but let its own observation confirm it before anything ships.

Uniform once does not mean uniform forever

A batch that was uniform when you stopped stirring can become less so if it sits. If pouring takes a long time, give the pot a gentle full-depth stir before each refill of the pitcher. If your first and last candles behave differently despite a validated stir time, the problem is during filling rather than mixing, and needs its own test.

Revalidate whenever the pot, the batch size, the wax or the stirring method changes. It is one batch and eight votives. It is much cheaper than finding out from a customer.

Where this page fits

CSI already has guides that cover parts of this question. This page answers one sweating decision; these cover the neighbouring questions:

Three things to buy for this fix

Frequently asked questions