I’m Using a 7.5 cm Jar with Soy Wax and 8% Lavender Fragrance — The Candle Tunnels for Two Burns but Eventually Catches Up; Should I Change the Wick?

Two tunnelling burns, then catch-up: what the pattern means

Three things happen in the first sessions of a soy candle that make the pool look smaller than it will later. The wax and glass start cold all the way through. The top of a freshly poured candle can be harder or uneven after cure. And the first session is often shorter than it should be. By burn 3 the surface has melted and set twice, the glass has been warmed, and a wick that is sized correctly commonly reaches the edge. That is why a ring that shrinks each burn and then disappears is a tendency many makers accept in testing, and why this cluster never judges a wick on burn 1.

A 7.5 cm jar sharpens both sides of the question. The pool is larger than in the 6–7 cm jars of the previous cluster, so the same wick has more wax to reach; and a ring that looks thin is a big share of a wide surface. In our worked example the ring after burn 1 averages 5 mm. In a 75 mm pool that is 11.00 cm² of wax on the glass, 25% of the surface. Two burns like that and a quarter of the top of the candle has not melted twice.

That is also why "it catches up eventually" needs a number. A ring that closes at burn 3 and one that closes at burn 5 both "catch up", but the second one spends four sessions under-melting and commonly leaves a thicker wall behind; whether that wall ever fully clears is the wasted-wax audit. Our rule below separates them.

Your 7.5 cm jar, and what CSI actually lists near it

The question says 7.5 cm without saying inside or outside. It matters: on a 0.35 cm wall (our labelled assumption) a 7.5 cm outside is 6.8 cm inside, a 36.32 cm² pool, while a 7.5 cm inside is 44.18 cm², 22% more. Measure inside edge to inside edge at the rim, twice at right angles, then at mid-height and about 1 cm above the base with a card strip curled inside the jar (overlap length ÷ 3.14). The method, and why listings mislead, is the external-versus-internal diameter post; the pool-area effect of 5 mm is the 5 mm diameter maths. This page's worked example assumes you measured 7.5 cm inside and pour a 150 g candle at 8% (12 g Lavender + 138 g wax); weigh your own fill line.

The honest part. None of CSI's 72 container listings states an internal diameter. Three listings give 7.5 cm: the Glass Jar With Bamboo Lid and the Amber Glass Jar with Bamboo Lid ("H: 8.5 cm x W:7.5 cm", "Fill Weight200 gm") and the Cute Ceramic Candle Container ("diameter of 7.5 cm", 130 g). All three are out of stock on 29 September 2026, and none states a heat rating in the text we captured. No listing states 8 cm, and nothing sits between 7.5 and 10 cm. The in-stock glass nearest your jar is at 7 cm (19 of the 23 listings that give a 7 cm figure are in stock, such as the Clear Glass Jar with Golden Lid (150 gram) and the 7 cm matte jars) and at 10 cm (the Apothecary jar with dome lid). We will not pretend either is a 7.5 cm jar: moving to one means re-running this log, because the pool area changes. If you need a 7.5 cm restock date, ask us on WhatsApp.

The ring-width log: four points, every burn, in millimetres

This log is lighter than an hourly pool curve on purpose: it is read once, at the moment you put the candle out, so it fits real life. It answers one question: is the ring left on the glass getting narrower, and how fast? Use the candle you already poured, and if you can, a twin from the same pour (24 g of Lavender for the pair; the twin itself costs about ₹144.26 in oil, wax and a C2, excluding your jar).

Case A matches your question, and it is the good version: each burn narrows every point and the ring is gone by burn 3. Case B also "catches up eventually", but its burn-1 ring leaves 29% of the surface solid and the ring still averages about 3 mm at burn 3. Case C never catches up at all. Case D is a wick that is off-centre or burning in a draught: the S side lags every burn while N is clean. Different verdicts, from twenty readings.

The catch-up score and where each result sends you

Catch-up score = (R₁ − R₃) ÷ R₁ × 100, where R is the ring sum in millimetres after that burn. 100% means the ring was gone by burn 3. A negative score means the ring was wider at burn 3 than burn 1. The thresholds below are our judgement, chosen because they can be read from a photo and because they send each result to a different test; set your own if you prefer, but set them before you burn and keep them fixed.

Why score at burn 3 and not burn 2? Because burn 2 still carries the first burn's uneven top; in many soy candles the biggest single change comes between burns 2 and 3. And why let case B continue at all? Because a candle that closes at burn 4 or 5 is not necessarily under-wicked: it may be a wick that is right for the middle and the bottom and slow only at the top, where a bigger wick would help and later hurt. Only a full-life comparison tells you which, and that comparison is the one-rung-up full-life pair, not a guess.

If the score fails: one rung up from C2, honestly

When the score is below 50% on real sessions, hold everything but the wick: same wax bag, Lavender at 8% on total weight, same jar batch, 150 g fill, pour temperature, cure and room. Two cells: your current C2 (the control) and one rung up, three candles each, so one odd candle cannot decide.

What "one rung up" means at CSI. The tabbed Eco wicks come in two sizes, C1 and C2, with no jar range stated. The next sized CSI rung is Cotton Braided Wicks Ready Made, whose page says "engineered for large-diameter candles exceeding 8 cm". A 7.5 cm jar is under that wording, so treat Cotton Braided as a labelled test arm, not a match. Beyond that: Eco Braided Wicks ("the perfect solution for large diameter container jars", cut to length, no size stated), the Braided Wick Roll with 100 wick tabs (cut to length) and one listing of Performance Wooden Wicks with no widths. There is no sized, tabbed wick between C2 and Cotton Braided, and if your candle needs one we would rather say so. A two-wick arm (2 × C1) is another route for wide jars; it changes spacing and geometry, and it belongs to the double-wick cluster.

Per candle that is ₹144.26 on C2 and ₹146.82 on Cotton Braided, before your jar: ₹64.80 of oil, ₹70.02 of wax and the wick. Cotton Braided costs ₹2.56 more a candle. Costs are worked examples and exclude labour, freight, packaging, the jar and GST. From nothing, the cash outlay is ₹1,261.80; with wax and Lavender already on the shelf, it is ₹214.40 for the two wick packs.

Whether two rungs close a bracket is the bracket-closure rule; at 7.5 cm, if C2 passes, "nothing smaller" is proven by C1 only if you run it, and there is no smaller sized tabbed wick than C1 at CSI.

What not to change first

If the fragrance changes rather than the jar, the rest of this cluster takes over: oud mushrooming where lavender didn't is the oil-versus-load separation, and a vanilla flame that shrinks burn after burn is the flame-shrink curve.

What to buy, and what to skip

Five items, chosen on page statements, price and stock on 29 September 2026. CSI publishes no wick-performance data, so nothing here is ranked on burn results. Oil candles are counted at 12 g per 150 g candle at 8%.

Map of this cluster: 4221–4240

This page owns the short answer for 7.5–8 cm single-wick jars, the end-of-burn ring log and the catch-up score. Each other post owns one test for one situation in which, mostly, the fragrance is what changed.

For 6–7 cm jars, go back to the 6–7 cm hub. For 8.5–9 cm jars and the first double-wick questions, go on to the next hub.

Frequently asked questions