My New Wick Works with Lavender but Not Vanilla — Do I Need to Qualify Replacement Wicks Across Multiple Fragrance Types?

Why one wick can behave differently in two fragrances

A wick is only approved for a system: this wax, this jar, this load, and this fragrance. In a lavender candle the wick is burning a pool of wax that holds 8% lavender oil. In a vanilla candle it is burning a different mixture — the same wax, the same 8%, but a fragrance with a different composition. Fragrance oils are blends of many materials, and makers commonly find that different oils change how freely the molten pool feeds the wick, how cleanly it burns and how much carbon builds on the wick head. So a wick that burns cleanly in one fragrance and struggles in another is not a faulty wick. It is a wick that has been approved in one system and not yet in the other.

Sweet, vanilla-style fragrances are the type makers most often mention in this context — heavier burning, more mushrooming, a wick that seems to need to be a size up. Treat that as a hypothesis to test, not a fact about any particular oil. CSI does not publish composition, vanillin content or burn data for Vanilla Bean, for Lavender or for any other oil, and nothing on this page claims how either will behave. Your own observation — works in lavender, not in vanilla — is the data. The screen below is how you find out whether it is a vanilla problem, a one-oil problem, or a sign the replacement wick is marginal across the whole range.

So do you need to qualify across fragrance types?

Yes — every fragrance type you will sell on the replacement wick needs its own evidence. The lavender result tells you the wick can work in your wax and jar; it does not transfer to vanilla, and the vanilla result shows it clearly does not. What you do not need is a full-life test of every single oil from a standing start. A screen sorts the range cheaply by fragrance type; a full-life confirmation then proves each type that is going to ship. The eight-fragrance post takes the next step for a fixed range: how to use your old burn records to decide which fragrances carry the most risk.

The screening design

Choose one representative per fragrance type you sell, plus a second oil in the type that failed. The second vanilla-type oil is the key design choice: if Cinnamon Vanilla also struggles on the new wick, you are probably looking at a type-level pattern and should treat every sweet vanilla-style SKU as unqualified. If it burns cleanly, the problem may be specific to Vanilla Bean in your system, and only that SKU needs a separate answer. The types below are the way makers commonly group fragrances by character — they are a sorting device, not a statement about any oil's chemistry.

Each oil gets two candles on the replacement wick and one control on the approved wick. The control is not optional even for lavender, where you already believe the new wick works: the screen compares within fragrance, not across fragrances, because a citrus candle and a woody candle may simply burn differently on any wick. The comparison that matters is new wick versus approved wick in the same oil. Three candles per oil is a labelled working practice for a screen; the full-life stage uses three burned candles per variant, the series' working minimum.

The screening sheet

Every line is scored as new wick relative to the control in the same oil, so the sheet works for any fragrance without needing an absolute standard — which is fortunate, because CSI does not publish one. CSI also publishes no numeric flame-height or jar-temperature limit; "no worse than the control", within your jar supplier's guidance and the fire-safety standard you work to, is the line.

Reading the screen: vanilla problem, one-oil problem, or marginal wick?

The design lets three different stories show up, and each points to a different decision. Work through them in this order, because the first one — a marginal wick — makes the other two irrelevant.

Whatever the pattern, the vanilla SKU has a problem that a formula change might seem to solve — a lower load, say. Resist that inside this test. A load change is a second variable; it changes throw, cost and the wick's fuel all at once, and it makes the screen unreadable. If you want to explore it, do it as a separate, one-variable test after this one. The one-variable post explains why, and the 5,000-wick post prices exactly this kind of formula adjustment when a wick has already been bought.

Two tempting shortcuts sit behind the vanilla result, and both are covered elsewhere in the series. The first is "just go a size up for vanilla" — possible, but a bigger wick is a new wick with its own full life to prove, as the size-up post sets out, and a stronger or hotter burn is never accepted as the price of making it work (the hotter-jar post). The second is trusting a supplier's word that a wick is "equivalent" across your range — the equivalent-size post explains why that word ends no conversations about wicks, and the vanilla direct-replacement post makes the same point from the fragrance side.

Stage 2: full-life confirmation per type

A screen that stops at half-weight cannot release a wick, because late-life and end-of-life behaviour have not been seen. So every type that will ship on the replacement wick gets a full-life confirmation: three candles on the new wick and one control, burned to the end of usable wax, with throw scored blind against the control. That is four candles per type. If the four clear types in the example below go forward, stage two is 16 candles.

Bought as one combined order for the four clear types in the example — Lavender, Citrus Lemon, Woody Oud and Fresh Strawberries — stage two costs ₹5,421.76, because the wax and jars are shared. Screen plus stage two comes to ₹11,447.10. For comparison, skipping the screen and running a full life on all six oils from the start would cost ₹7,808.56 and would take the same calendar time as stage two — but it would not tell you anything the screen does not, and it would spend full-life candles on oils you are about to exclude.

The priced stage-1 screen

All sizes are listed and in stock on the 24 September 2026 pull, with a 5% pour-loss allowance (a working assumption). Three 200 g candles at 8% need 48 g of oil; with the allowance that is 50.4 g, which is just over a 50 g bottle — so each oil is priced at the cheapest listed cover, which usually adds a 15 g bottle. Weigh carefully and you may manage without it, but do not plan on it. Luxury Soy Wax Chunks are rated up to 13% fragrance load by the wax maker, so 8% is well inside that stated maximum.

The Cotton Braided Wicks Ready Made line stands in for your replacement wick and the Eco Thin C1 line for your approved wick, purely so the kit can be priced. CSI publishes no burn data, construction details or size charts for either, and does not claim one is equivalent to the other or to any other brand's wick.

Release by type, and record the split

The likely outcome of this exercise is not "the new wick is approved" but "the new wick is approved for these types, and the old wick stays on those". That split is where production mistakes happen, so the record matters more than usual.

If your range is small and fixed — say eight fragrances on one wick — and the question is how much of it to retest after a supplier change, the eight-fragrance post turns the same logic into a bracketing plan. If the vanilla problem only appears half-way down the jar, the half-way post explains why full-life testing is repeated per replacement fragrance. For a wax change across a large range the same screening logic appears in the 20-fragrance wax post.

Frequently asked questions