My 9 cm Jar Works with a Single Wick for Lavender but Not for Vanilla — Can One Fragrance Push the Same Container from Single-Wick to Double-Wick Territory?

Why vanilla and lavender can need different builds

A wick is fed by the melted pool, and the pool is mostly wax with 8% fragrance in it. Change the fragrance and you change what the wick draws up. Light herbal and citrus oils often burn easily; heavier, vanillin-rich, resinous and gourmand oils are commonly linked to faster carbon on the wick, smaller or sootier flames and, sometimes, a pool that reaches the glass less well. These are tendencies reported by makers, not measurements CSI publishes, and they vary oil by oil.

The pages give you the evidence you have. Vanilla Bean's page states "Vanillin Content: 2.5%" and warns of "Potential discoloration"; it also says "Wick testing is recommended based on your container size and wax type". Lavender's page states "Recommended fragrance load 6-10% for candles" and, in the text we captured, gives no vanillin figure. Neither page says one needs a bigger wick than the other, so treat the vanillin line as a reason to test, not a verdict. The first-level version of this question, whether heavy oils need a larger wick at all, is do heavy fragrance oils need a larger wick.

Now the jar. CSI lists no jar at 9 cm, so this is your own; measured 9 cm across the outside, on an assumed 0.35 cm wall, it is 8.3 cm inside with a 54.11 cm² pool. That is a lot of surface for one wick, which is why your single is probably Cotton Braided: its page describes it for "large-diameter candles exceeding 8 cm". Lavender may sit comfortably on it. Vanilla, asking more of the same wick, may not, and because there is no larger ready-made cotton wick in CSI's range, "more wick" at 9 cm means two wicks. Each wick of a pair serves 27.05 cm², half the pool.

In a 7 cm jar the same effect looks different: vanilla commonly asks for one rung up on a single, and the 7 cm offset bracket tests exactly that. At 9 cm the rung up is a pair, which is why this page is about crossing a line rather than climbing a ladder.

Two oils, two builds, eight candles

You could simply pour vanilla on a pair and see. But that answers only half the question. If the pair works for vanilla, can lavender move to it too, so the jar keeps one build? Or does lavender run too hot on two wicks, so the jar needs two? You need both oils on both builds to know, and you need your working lavender candle remade in the same pour, so it acts as the control.

Materials per candle are oil, wax and wick only, excluding your jar, labour, freight, packaging and GST. The worked example uses a 300 g fill at 8% of total weight: 24 g of oil and 276 g of CSI 464. Keep the load the same for both oils. It is tempting to "help" the vanilla with a lower load; do that and you have changed two things, and the plot can't tell you which one mattered.

"Not working", turned into four flags

Makers say a candle "doesn't work" for several different reasons, and they point to different fixes. A vanilla candle on a single wick can fail by leaving a ring at the glass, by growing a carbon cap, by sooting, or by a flame that is too tall or too small. Score each separately, the same way on every candle.

The 5 cm figure is the top of a range many makers use (about 2.5–5 cm), a rule of thumb rather than a CSI figure; the ruler method is set out in the 9 cm flame-height protocol. The mushroom scale is ours: crude, but the same crude scale on eight candles is better than eight impressions. A flag counts for the cell if either candle raises it. That is strict on purpose: a build you will sell has to pass on every candle, not on average.

Reading the plot: cross, swap, gap or flat

Plot the worse candle of each cell: build across the bottom, flags up the side, one line per oil. The four shapes in the diagram are the four answers you are likely to get.

A · Vanilla crosses, lavender doesn't mind. Vanilla raises flags on the single and none on the pair; lavender is clean on both. Vanilla has pushed this jar into double-wick territory, and lavender can follow it there. Move both oils to the pair and carry one candle of each through its full life before production. You have kept one build for the jar, which is the cheapest outcome to run. The pair is also ₹0.20 cheaper per candle in wicks than Cotton Braided, though you pay that back in time setting two tabs.

B · The lines swap. Vanilla needs the pair; lavender is fine on the single and raises a flag on the pair, usually a flame over the band or a pool that runs deep. Lavender now needs less wick than the pair gives, vanilla more than the single gives. The jar carries two builds. That is workable, and many makers run it, but it is a production decision, not a wick decision; the next section is what it involves.

C · A parallel gap. Vanilla raises the same flags on both builds; lavender is clean on both. The number of wicks is not what is failing vanilla. Common next moves: the same vanilla at a lower load inside its page's "Up to 10% for soy wax" (at 6% a 300 g candle takes 18 g, saving ₹47.40 of oil per candle), or another vanilla. Choosing between vanillas is the 6.5 cm vanilla swap, and the principle carries to 9 cm.

D · Flat at zero. Everything passes, including vanilla on the single. Your original vanilla problem did not come from the jar: common culprits are a candle burned before cure, a long trim, a draught, a different vanilla batch or a different load. Keep the single for both and write down what was different last time.

One jar, two builds: what it really costs

If your plot looks like B, the same glass leaves your pour table as two different products: a lavender candle on one Cotton Braided and a vanilla candle on two C1. Once poured and labelled they look the same from the outside, so the controls have to live in your process. The full split-production plan for a range, with holders and ₹ per build across six oils, is the six-fragrance 9 cm plan; for two oils, these five controls cover it.

Argued against our own interest: if the plot is A, resist keeping two builds "because lavender was fine before". One build means one wick bin, one holder setup and nothing to mix up, and a pair that passes lavender's full life costs ₹0.20 a candle less in wicks than the Cotton Braided you would stop buying. Pattern A is also the case for standardising a group of oils on one wick, which is grouping wicks by fragrance behaviour.

And if the plot is C, don't create a second build at all. A second build for an oil that fails on both builds adds all of the controls above and fixes nothing.

Grams, wicks and rupees

If you already own Lavender and your jars: Vanilla Bean 100 g (₹790.00), three 1 kg bags of CSI 464 (₹1,522.20, 792 g left over), a Cotton Braided 10 (₹120.00) and a C1 20 (₹118.00) come to ₹2,550.20. If you also own the wax and Cotton Braided, the test costs ₹908.00. Both 100 g bottles leave 4 g after four candles; that is no margin at all, so weigh carefully or add a 15 g bottle.

Per candle, oil, wax and wick only: lavender ₹281.64 on the single, ₹281.44 on the pair; vanilla ₹341.64 on the single, ₹341.44 on the pair. The build changes the cost by ₹0.20; the oil changes it by ₹60.00. Buying larger sizes narrows the oil cost a little (24 g costs ₹125.76 of Lavender or ₹184.32 of Vanilla Bean at 500 g prices), but only once a build has passed. All exclude jar, labour, freight, packaging and GST.

What to buy, and what to skip

Skip: a second vanilla before the 2×2 has run. Dark Vanilla is priced the same as Vanilla Bean at every size and its page notes "Vanillin present", so it is not a way round the question; it is a different oil to test if pattern C appears. Skip a bigger Cotton Braided pack too: if vanilla needs the pair, you will be buying fewer of them, not more. And skip Eco C2 for now; it becomes relevant only if the vanilla pair is short at the edge.

Frequently asked questions