My Wick Mushroom Is Large Enough to Fall into the Melt Pool — What Should I Change First?

Today: the immediate response

Your situation, stated exactly: during a burn, the carbon cap on the wick has grown large enough that it, or pieces of it, drop into the melt pool. Perhaps it happened at hour three or four of a long session, perhaps with a vanilla or other heavy scent, perhaps in a dyed candle, perhaps only near the bottom of the jar. Write down which, because each points somewhere different. Whatever the cause, the first change is the same.

Why a falling mushroom fails the candle

Carbon on the tip is something you can set a limit for (how much mushrooming is normal sets one from your own baseline candles). Carbon in the pool is different in kind. A glowing piece can keep burning on the surface or act as a second, unplanned wick; debris can flare as it meets liquid fuel; and every piece that sets into the wax is a black speck a customer will see the next time they light the candle. Those are the reasons we treat any fall as a fail in testing, whatever the cap measured a minute before. The full quality and commercial argument is 6955's to make; for this post, it's enough that a falling mushroom is a stop sign, not a grade.

Why does a cap get that big? It grows when the tip receives more fuel than the flame burns completely, and the longer the session, the more it collects (the cluster hub, large mushroom after every three-hour burn, covers the mechanism). A cap heavy enough to break off has usually been growing for hours, which is why falls cluster late in long sessions and lower in the jar, where heat builds and the pool deepens. A leaning or off-centre wick can carry its cap out over the pool and drop it sooner.

The causes, ranked, and which change comes first

The ranking is our judgement for a cap that grows until it falls. Signs are common observations, not CSI burn data; CSI publishes no carbon, soot or wick data for any product.

Causes 3 and 4 decide whether the formula needs changing at all: if the control never drops carbon from a 5–6 mm start within your label's burn time, the fix is the label, not the candle. But be honest about it: if it only behaves when trimmed shorter than real customers manage, the label is covering a formula problem (trimming versus a bad formula).

Wick first or load first: the pool decides

The load route is often missed because it doesn't feel like a fix, but it is the one change that reduces the fuel without reducing the wick. It also costs less oil per candle: 8% to 6% saves 4 g in a 200 g candle, 25% of the oil. With Vanilla Bean as a worked example at its 1 kg price, that is ₹30.58 a candle. Throw may drop too; whether it stays acceptable is your sniff test, and finding the lowest load that keeps your throw is optimising the load before the wick. When the carbon starts early, around hour two, the load-arm method is the carbon-ball post. The jar-diameter version of "reduce the load to cure the mushroom" is the 8 cm coffee case.

The three-arm test and the zero-fall rule

The test uses the series' default test candle, labelled: 200 g at 8% = 16 g fragrance + 184 g wax (load as a percentage of total weight), in your jar. Three arms of three, poured together: A your current wick at 8% (the control: it should reproduce the falls); W one CSI rung down at 8%; L your current wick at 6%, 12 g fragrance + 188 g wax. That is 9 candles, 132 g of your oil and 1,668 g of wax. If the candle is dyed, add a fourth arm, current wick at 8% undyed: 2,220 g of wax, 2 × 1 kg + 500 g (₹1,274.80).

Burn every candle to about 1 cm from the bottom: at a worked-example 7.0 g an hour, about 7 four-hour sessions each, so roughly 252 pool checks across the test. Cure first (3 days; Vanilla Bean's page, for example, states "48–72 hours for soy wax"), burn in still air, and mark each jar into thirds, because falls cluster late in the jar. A speck count of zero after a session is the result you want; any speck means carbon fell, even if you didn't see it happen.

Reading the result: if L passes and W tunnels, keep the wick and move the load. If W passes and L still drops carbon, the wick was too strong. If both pass, choose the one with the better full-life pool and the throw you want. If neither passes, combine them only now (one rung down at 6%), or look at construction and oil (smaller size or a different construction). A late-jar version of the same failure, falls only in the bottom third, is mushrooms only in the bottom third.

The CSI buy list and the step to production

CSI's sized, tabbed wicks are three rungs: C1 → C2 → Cotton Braided. One rung down means Cotton Braided → C2, C2 → C1, or a C2 pair → a C1 pair; there is no sized, tabbed wick between them, nothing smaller than C1, and one wooden wick with no width options. If you are already on C1, route W doesn't exist at CSI: the load, the oil or a different construction are your options, or ask us on WhatsApp, and we'll say plainly if we have nothing smaller.

Argued against our own interest: if the pool is normal, the most likely fix is route L, which needs no wicks and uses less of the oil we sell. And because C1 (₹5.90) and C2 (₹9.44) cost the same per wick at every pack, buying bigger now gains you nothing but shelf stock. The step from C2 to C1 saves ₹3.54 a candle; that is a side effect, never the reason.

The step to production quantity

Production wicks = candles × wicks per candle × (1 + a defect allowance). Worked example, labelled: 350 single-wick candles a month with a 2% allowance is 357 wicks. If route W passes on C1, 3 × 100 + 50 + 20 = 370 for ₹2,183.00. If route L passes on your C2, the wick order is unchanged (₹3,492.80 for 370) and the oil order falls: 1,400 g a month less oil at that volume, worth ₹10,704.40 with Vanilla Bean as the worked example at its 1 kg price. Reorder point = daily use × your lead time + safety stock; CSI publishes no lead times, so ask on WhatsApp. Release the quarantined batch only if it was made to the formula that passed; if not, see 6955.

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 flame data, so nothing is ranked on performance, and each card says when not to buy.

Skip for now: wicks if your pool is normal (start with route L); a new oil "that doesn't mushroom" while the test is running; Cotton Braided as a fix for falling carbon in a jar where C2 already caps; and any production pack before nine candles burn to the end.

Frequently asked questions