My Candle Flame Is Taller Than the Same Formula from My Previous Batch — Should I Compare Wick Lots Before Changing the Formula?

What can change between two batches of "the same" candle

Your formula says "200 g, 8% oil, C2, this jar". Your batch also contains a wick lot, an oil lot, a wax lot, a pour temperature, a fill weight, a trimmed wick height, a centring job, a cure age at the time you tested, and the room you tested in. Any of them can move a flame. The worked example in this post: last batch's flame sat around 2.9 cm at hour three; this batch's reaches about 4.0 cm, 1.1 cm taller, same recipe.

Wicks top the list for a physical reason: the flame's size depends on how much fuel the wick draws up, which depends on its construction (braid tightness, fibre, thickness, any coating). Two production runs of the same wick can differ slightly in any of these, and CSI's wick pages state no construction specification and no lot data, in the text we captured. We do not claim our lots are identical, and you should not assume any supplier's are.

The other materials matter too, more weakly. A new oil lot can differ in how it burns (a tendency; CSI publishes no oil batch data, and lot questions go to WhatsApp). A new wax lot can melt slightly differently. These are why stage 2 below pours both wick lots from one pot of today's mix: it removes the oil and wax lots from the comparison at a stroke. For fragrance-strength swings between batches, the stock audit in the batch-to-batch fragrance post covers oil age and storage.

The batch variables ranked: your test or your materials?

Variable 2 is the sneaky one. If the old batch was trimmed to 5 mm at finishing and the new one went out at 12 mm, the first burn's flame will be taller on every new candle and look exactly like a stronger wick. It settles after one trim; a stronger wick lot doesn't. Variable 3 is why stage 1 exists: a retained candle is the only fixed reference you have, so if it is tall today too, the batch didn't change; your test did. If you have no retained candle, the same logic applies to a known-good candle of any earlier batch.

Measuring two wick lots side by side

Take ten wicks from each lot. Lay them in two rows on white paper under the same light, and photograph them together with a ruler in the frame. CSI lists no ruler or calliper; a steel ruler and your phone's macro mode are enough to see most differences, and a vernier calliper of your own helps with width.

Visible differences are a reason to burn, not a verdict: two lots can look identical and burn differently, and two that look slightly different can burn the same. Hidden variables (fibre, ply count, core, coating chemistry) are covered in same length and thickness, different burn. If the tab looks different, the effect shows only at the end of the jar; the tab-change post has that end-of-life test.

The three-stage lot test and the equivalence rule

Argued against our own interest: stages 1 and 2 need no wicks from us at all, and if you already own centring bars, they cost nothing but wax, oil and time. That is the point: a lot comparison should be cheap, so you run it every time instead of guessing.

What to do with the answer, and the CSI buy list

Two honest limits. First, CSI publishes no lot data and cannot promise that two packs, or a large order, come from one lot. Send us your order numbers on WhatsApp and we will tell you what we know, but plan on testing every new lot rather than relying on continuity. Second, our range has no half-sizes: if a stronger lot pushes you between C1 and C2, we can't sell you the step in between, and we would rather say so. A different supplier's wick is a different wick, not a different lot; that is 6913's territory.

The production step, with retained wicks built in

Once a lot is approved: production wicks = candles × wicks per candle × (1 + a defect allowance), plus the wicks you retain from that lot. Worked example, labelled: 200 single-wick candles a month with a 2% allowance is 204 C2, plus 10 retained = 214. Buy 2 × 100 + 20 = 220 for ₹2,076.80, leaving 6 spare. The ten retained wicks cost ₹94.40 per lot, which is the cheapest insurance in candle making. Because C2 is ₹9.44 a wick at every pack, buying 500 at once (₹4,720.00) saves nothing per wick; it only ties more of one unknown lot to your shelf until you have tested it.

What to buy, and what to skip

Four items chosen on page statements, price and stock on 7 October 2026. CSI has no burn data, wick specifications or lot records to rank them on, so each card says what it does for this test and when not to buy it.

Skip for now: a new oil or wax "to bring the flame down" (one variable at a time, and it is probably the wick); Cotton Braided, which is a rung up, not a fix for a stronger lot; and any 500- or 1,000-pack until the new lot has passed stage 2.

Related in this series: if the flame is too large even on a correctly trimmed candle from a single batch, start with 6904; for the hour-by-hour diagnosis of a growing flame, the cluster hub; and for what a lot change does to burn time, the 200 g burn-time post. Outside it: a wick that leans and burns one-sided, investigate centring first.

Frequently asked questions