My 8 cm Soy Candle Burns Correctly When the Wick Is Trimmed to 5 mm but Becomes Smoky When Customers Leave It Longer — How Should I Account for Real-World Wick Trimming?

Why 5 mm and 12 mm are different candles

A wick is a fuel pump. Molten wax climbs it and vaporises in the flame; the length of wick standing above the pool sets how much fuel reaches the flame and how big the flame gets. Trim to 5 mm and you have set the pump to the rate you tested. Leave 12 mm and, commonly, the flame grows taller and draws more fuel than it can burn cleanly: the tip glows, carbon collects in a cap (mushrooming) and the flame starts to flicker and smoke. These are tendencies, not laws, and some wicks tolerate extra length far better than others. That difference is exactly what this test measures.

Your 5 mm result is real, but it answers a narrow question: does this wick burn cleanly when it is looked after? You cannot count on a customer looking after it. A wick that is clean at 5 mm and smoky at 10 mm has no margin. A wick that is clean at 5 mm and still acceptable at 12 mm has margin, and margin is what survives a customer's living room. Many makers trim to about 5–6 mm; one CSI trimmer page puts it as "Trim your wick to 1/4 inch before every burn", which is 6.35 mm. Your 5 mm is the careful end of that habit.

Why does this bite harder at 8 cm than at 6 or 7? Because a wider pool needs a livelier wick to reach the glass, and a livelier wick has more to give when it is left long. An 8 cm jar is also where CSI's range is at its edge: in this cluster's worked examples the house wick at 7.5–8 cm is a C2, and the next rung, Cotton Braided Wicks Ready Made, is described by its own page as "engineered for large-diameter candles exceeding 8 cm". A wick picked to just reach the glass at 5 mm is often working near its limit. Leave it long and it goes over. This is judgement from the geometry, not CSI data.

Oil matters too. Heavier, vanillin-rich or resinous oils are commonly linked to more carbon on the wick; lighter herbals and citrus often burn more easily. Our worked example uses Bergamot Amber at 8%, whose page lists a "smooth vanilla base". A vanilla-heavy or oud oil at 10% may lose margin sooner than a lavender at 6%, so run the test in the formula you sell, not in a stand-in.

The length customers really relight at

A customer who never trims doesn't relight at some random length. They relight at whatever the last session left: the wick that stood above the pool when the flame went out, often with a carbon cap on top. You can measure that on your own candles, and it turns "customers leave it longer" into a number.

On the two 5 mm twins, after each session has cooled, straighten the wick gently with a cocktail stick (don't pull it), lay a steel ruler against it and read the height from the wax surface to the tip, cap included. Write it down before you trim back to 5 mm for the next session. Three sessions on two candles give six readings. In our worked example they were 9, 11 and 12 mm on one twin and 10, 12 and 13 mm on the other: a mean of 11.2 mm, about 2.2 times the 5 mm trim. Your numbers will differ with wick, wax, oil and session length, which is why you measure rather than borrow ours.

Two cautions. First, the carry-over grows with session length: a 4-hour session commonly leaves more wick than a 2-hour one, so measure after the session length your candle's label will print (the printed length itself is the label-policy decision). Second, lower in the jar the wick can sit in a deeper pool and behave differently; this test deliberately reads the top half, and the full-life burn catches the rest.

The untrimmed-tolerance test, step by step

Measure the jar first. The title says 8 cm, but it doesn't say inside. On a 0.35 cm wall (our assumption) an 8 cm outside is 7.3 cm inside: a 41.85 cm² pool, not the 50.27 cm² an 8 cm inside would give (20% more). Measure yours at rim, middle and near the base (why the outside figure misleads). CSI lists no 8 cm jar; our nearest listings are the 7 cm Clear Glass Jar with Golden Lid (150 gram) (outer diameter approx. 7 cm by its page) and the 10 cm Apothecary Glass Candle Jar with Dome Lid. Neither is your jar, so this test runs in yours.

Stage 2, only if the house wick fails at the carry-over length, is the same six candles on C1: another 96 g of oil (a second 100 g bottle), 1,104 g of wax (396 g left from stage 1 plus a new 1 kg bag, leaving 292 g) and six C1 wicks from the 20-pack. That is ₹1,138.16 of materials and ₹1,185.40 in cash. Both stages from nothing come to ₹2,902.20. Worked-example costs exclude labour, freight, packaging, GST and your jars. Already have six unburned candles from the production batch? Use them for stage 1 and the only purchase is a trimmer, if you need one.

Reading the twins: acceptable, borderline, fail

These thresholds are our judgement, chosen because they can be read from a photo and a tissue. The flame figure uses the practical range many makers work to, about 2.5–5 cm, as a rule of thumb, not a CSI figure. If your brand's standard is stricter, use it, but fix it before session 1 and use it for every candle.

A setting passes when both twins score acceptable on every signal in all three sessions, or borderline on one signal once. A setting fails when either twin fails any signal in any session with the air still. Twins that disagree (one pass, one fail) are not a pass: pour a third candle at that setting before deciding.

Read the pattern, not a single cell. In the worked example the house wick is fine at 5 mm and fails at both longer lengths: its tolerance window closes somewhere between 5 and 10 mm, below the 11.2 mm carry-over. One rung down stays acceptable at 10 mm and is borderline at 15: its window covers the carry-over. That is the diagram above in numbers. The flames are illustrative; your wick, wax and oil will give their own.

When soot first appears within a session is a different measurement, the soot-onset clock. Here a fail anywhere in the session is enough, because the question is only whether the wick tolerates the length.

The robustness rule and stage 2

The rule is our judgement, stated so you can apply it without us: choose the wick whose tolerance window covers your measured carry-over length, on both twins, and still reaches the glass in a normal session. A wick that is perfect at 5 mm and fails at the carry-over is a wick that works for you and not for your customer.

Stage 2: the rung below has to pass two tests

Stepping down buys margin at the long end and costs heat at the short end. The smaller wick has to pass the tolerance test and still melt to the glass in the session customers actually burn. In an 8 cm jar that second test is not automatic: a C1 that is beautifully calm at 10 mm may leave a ring at the 3-hour mark. So stage 2 logs one more thing on the 5 mm twins: the ring of solid wax at the glass at 2 and 3 hours, at four points. Whether a slow-but-clean wick is acceptable at the session length customers really burn is the customer-session coverage post; whether a ring that closes by burn 3 passes is the catch-up log in the hub.

Whichever wick wins, the production candle still gets a full-life burn to about 1 cm left, trimmed as your label will instruct, because the lower third brings its own heat (why the flame turns aggressive near the bottom). The tolerance test clears the top of the candle for the customer who doesn't trim; it doesn't approve the bottom.

What not to buy, and when the answer is a label

And sometimes the right purchase is a label, not a wick. If the house wick passes at the carry-over length and fails only at 15 mm, the candle forgives ordinary neglect; what it can't forgive is a customer who lets the wick grow for days. That candle needs a clear trim instruction and a sensible printed session length, which is the label-policy decision. A Candle Care Stickers Roll is ₹413.00 for 500 (₹0.83 a sticker) if you want a separate care sticker; the words matter more than the sticker. We sell the stickers, and we would still rather you fixed a fragile wick than printed a warning over it.

What to buy for the test

Four items, in the order the test uses them, 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 16 g (200 g at 8%) and 12 g (150 g at 8%).

Frequently asked questions