My Candle Starts with a Clean Wick but Develops a Carbon Ball After Two Hours — Could Fragrance Load Be Contributing?

Clean at the start, a ball by hour two: what changes

Your symptom, stated exactly: you trim the wick, light the candle, and for the first hour or so the tip stays clean and the flame looks right. Somewhere around hour two a ball of carbon appears on the tip and grows from there. It does this session after session. You suspect the fragrance load, because it's one of the few things you can change without buying anything new.

Three things build up over those two hours. The pool widens and deepens, so the wick sits in more liquid fuel. The tip has been drawing that fuel continuously, so any carbon the flame doesn't finish has had two hours to collect. And the fragrance in that fuel is part of what has to be burned. A higher load, or an oil with heavier materials, commonly means more of the fuel arriving at the tip is the kind that leaves carbon behind, so the ball arrives sooner. That is the mechanism behind your suspicion, and it is plausible. It is not the only mechanism: a wick that draws too much for the jar will cap at hour two at any load, and an untrimmed start brings the ball forward too. The cluster hub, large mushroom after every three-hour burn, sets out the general picture; this post is about separating the load from the rest.

The causes, ranked: load, oil, wick or the burn

Our ranking for a clean start and a ball by hour two (judgement). Signs are common observations, not CSI burn data.

Why put load first for your pattern? Because "clean, then a ball at a fairly consistent time" with no other symptom fits a fuel that is slightly too rich for the wick. An oversized wick usually brings company by hour two: a tall flame, a deep pool, faster wax loss. If your candle has that company, swap rows 1 and 3, and start with stage 2. Higher fragrance loads changing how a wick behaves is a well-reported effect; the 7 cm sandalwood case at 10% asks your question from the jar side, and the higher-load, worse-burn post covers how excess fragrance can interfere with combustion and throw.

Time to mushroom: the measurement

"After two hours" is the most useful thing you've told us, because time is what the load changes. A lower load rarely makes a ball vanish; it delays it. So the measurement isn't "is there a mushroom?" but "when does it reach a size that matters?". Here's the in-house method (labelled; the sizes are yours to set).

Thirty minutes is the coarsest interval that still separates arms whose balls arrive an hour apart. A 2 mm onset is about the point where most makers would call it a ball rather than a darkened tip, and a 3 mm limit matches the example bands in the cluster hub; both are labels you can change, as long as you change them before the burn, not after. Where to set the limit properly is how much mushrooming is normal; how to record and average the readings across candles is the quantitative recording post. For the load comparison, use the sessions in the top third, before late-jar heat joins in, and take the earliest limit time among the three candles in each arm.

Stage 1: load arms on the same wick

The arms use the series' default 200 g test candle, labelled, with the load as a percentage of total weight: 10% = 20 g fragrance + 180 g wax; 8% = 16 g + 184 g; 6% = 12 g + 188 g. Use your current load as the top arm if it is 10% or less; 10/8/6 is our worked example. Check both ceilings first: the lower of your wax page's maximum and your oil page's maximum governs. Roasted Coffee states "candles (6-10%)" and Vanilla Bean "Up to 10% for soy wax"; Luxury Soy Wax CSI 464 states "Up to 10%" without saying whether that is of wax or total weight, so if you read it conservatively, run the top arm at 9% instead (the CSI 464 card below has the arithmetic).

In the diagram's worked example, the 8% arm reaches the limit 1.0 hour later than 10%, and 6% doesn't reach it within four hours: the load is clearly contributing. If 6% still throws well enough, that is the fix, and a cheaper candle; if it doesn't, 8% caps at 3:30, inside a four-hour label, so stage 2 tests a smaller wick at 8%. Choosing the best load on throw, cost and carbon together, rather than just the one that stops the ball, is optimising the load before changing the wick, and why less fragrance can burn better explains why a lower load can improve more than the wick.

Stage 2: wick arms at the load you keep

Stage 2 runs when the load you'd keep still reaches your limit before your label's burn time, or when stage 1 says the load isn't the driver. Two arms of three at that load: your current wick and one CSI rung down, 6 candles poured together, burned to the end in four-hour sessions with the same time-to-mushroom checks, judged in the top, middle and bottom thirds. At 8% in the worked example that is 96 g of oil and 1,104 g of wax; with Roasted Coffee, the stage 1 leftover plus one more 100 g gives 106 g, and the leftover wax plus 1 kg gives 1,344 g.

CSI's wick range is short, and that shapes stage 2. Our sized, tabbed wicks are C1 → C2 → Cotton Braided, with no sized, tabbed wick between them, nothing smaller than C1, and one wooden width. From C2 the arm is C1; from Cotton Braided it is C2; from a C2 pair, a C1 pair. If the smaller rung tunnels and the current one caps, the fine step you need doesn't exist as a wick here: it is the load, which is why stage 1 comes first. If you're already on C1, there is no smaller CSI wick to test; ask us on WhatsApp and we'll say so plainly rather than sell you something near.

The buy list and the step to production

Argued against our own interest: stage 1 needs no wicks, and if it says the load is contributing and a lower load keeps your throw, the purchase is less oil from us, not more wicks. Worked example, labelled: 200 Roasted Coffee candles a month at 6% use 2,400 g of oil (2 × 1 kg + 500 g, ₹9,366.00) instead of 4,000 g at 10% (4 × 1 kg, ₹14,952.00): ₹5,586.00 a month less spent on oil, or ₹29.90 a candle at the 1 kg price. And because C1 and C2 cost the same per wick at every pack size, a bigger wick pack never saves anything before approval.

The step to production quantity

If stage 2 passes on C1: production wicks = candles × wicks per candle × (1 + a defect allowance). At 200 candles a month with a 2% allowance that is 204 wicks; buy 2 × 100 + 20 = 220 C1 for ₹1,298.00. The step from C2 to C1 saves ₹3.54 a candle, a side effect, never the reason. Order the oil for the approved load, not the old one: weigh it by the batch, because "about 6%" drifts back towards 8% on a busy day. Reorder point = daily use × your lead time + safety stock; CSI publishes no lead times, so ask on WhatsApp. A new lot of the same oil is worth one timed session before it goes into production.

What to buy, and what to skip

Four items, chosen on page statements, price and stock on 7 October 2026. CSI publishes no carbon, soot or throw data, so nothing is ranked on performance, and each card says when not to buy.

Skip for now: wicks before stage 1 has answered; a different oil before you've timed your own at a lower load; anything above your wax's or oil's stated maximum; and production packs before the chosen configuration burns to the end.

Frequently asked questions