My 7.5 cm Coffee Candle Gives Excellent Hot Throw but Produces Soot After Three Hours — How Do I Keep the Fragrance Performance While Correcting Combustion?

Why the minute soot starts matters more than whether it starts

"Soot after three hours" is really two facts: a wick and oil that produce more fuel than the flame cleanly burns, and a clock. The first is a matter of degree. Soot is unburned carbon, commonly linked to a flame that is too large or unstable for its air supply: an oversized or untrimmed wick, a heavy fragrance load, draughts. Heavy gourmand oils such as coffee, with cocoa, tonka, vanilla and tobacco in their notes, are often linked to more carbon at the wick (a tendency; CSI has no measured soot data for any oil). The second fact, the clock, is what turns a tendency into a product decision.

Three reasons the clock matters in a 7.5 cm jar. First, onset usually moves earlier as the candle burns down: the glass above the pool holds heat, the pool deepens and exposed wick can grow, so a candle that soots at three hours near the top may soot at two and a half near the bottom. Second, customers burn longer than labels say. Third, a fix that delays onset by twenty minutes looks like success and is not. A measured minute, compared with a fixed target, avoids all three traps.

What this page does not do: decide what your label should say (that is the label-policy post), or read where the soot lands on the glass (that is the soot-band post). It measures when soot starts, on two wicks, against the label you already have.

Running the soot-onset clock

The target: onset later than the burn time on your label

The target is our judgement, labelled: onset later than the label's burn time plus a 30-minute margin, on both twins, in both readings. With a label that says 3 hours, that is 210 minutes. The margin is there because real candles vary, rooms vary and people are late to blow out candles. The reason for both readings is that onset commonly moves earlier low in the jar, and the lower third is where a customer complains.

Your wick against one rung down, in twins

The test holds everything but the wick: Coffee Vanilla Tobacco at 8% of total weight, one bag of CSI 464, one jar batch, 150 g fill, the same pour temperature (CSI 464 states 80–90°C; your jar's own heat wording applies, so pour at the low end and pre-warm clean jars; judgement), the same cure and room. Twins, not triplets, because this is a screen with a clear pass line and a follow-up: the winning wick still goes through a full life with three candles before production.

Why one rung down, not a smaller flame by trimming shorter? Trimming is the customer's job, not the product's, and trimming shorter than 5–6 mm tends to shift the problem rather than solve it. Why C1? Because it is the rung below C2 in CSI's range, and the only one: there is no sized, tabbed wick between them. If your current wick is Cotton Braided, the rung down is C2; if it is C1 already, CSI lists no smaller sized, tabbed wick, and the untabbed Eco Candle Wicks Loose is "designed specifically for 2 inch diameters", far narrower than a 7.5 cm pool.

From nothing but your jars and C2 wicks, the cash outlay is ₹1,074.40. The 50 g bottle leaves only 2 g; if a twin disagrees and you need a fifth candle, you need another bottle, so decide before pouring whether to buy 100 g instead (it is ₹0.01 a gram dearer than the 50 g). Costs exclude labour, freight, packaging, jars and GST.

Keeping the coffee throw while you fix the combustion

The fear behind the question is reasonable: a smaller wick makes a smaller pool, and hot throw comes from fragrance evaporating from a heated pool. A pool that is too small or too shallow releases less. So the one-down twins must pass two more checks before you switch.

Pool check. On burns 1–3, note whether the C1 pool meets the glass by the end of a 3-hour session; a rough rule of thumb puts a full first pool at about 3.0 hours for a 7.5 cm width (44.18 cm² if 7.5 cm is the inside). A ring that closes by burn 3 passes; a ring that persists fails, and the scoring method is the catch-up score in the 7.5 cm hub.

Blind throw check. On session 2, move one C2 and one C1 candle, in turn, into the same closed room (a worked example: 10 m² with a 3 m ceiling, 30 m³) for 45 minutes each, with the room aired between. Two or three people who did not pour the candles walk in from outside and say which of the two is stronger, or "same". That is a preference, not a score, and CSI has no throw data to compare it with. Pass (judgement): C1 is not called clearly weaker by most sniffers.

The oil page itself hints at why a smaller flame need not cost throw: Coffee Vanilla Tobacco says "strong throw at lower loads". That is the page's claim, not CSI measurement, but it matches the common experience that a cleaner, calmer flame often lets a gourmand's character come through instead of a burnt edge. What a hotter flame does to a heavy oil's character is the 8 cm oud off-note check.

If C1 is too small: the gap, and the other two levers

The awkward result is C1 clean on the clock but failing the pool or throw gate, while C2 soots before the label. The answer lies between C1 and C2, and CSI lists no sized, tabbed wick there. We would rather tell you that than sell you a pack of something that does not exist; ask us on WhatsApp, and consider another supplier's intermediate size. On CSI's range, the two levers left are these:

Do not go up in load to "get the throw back": 9% is Coffee Vanilla Tobacco's stated top (14 g per candle), and the common tendency is that more oil means more carbon, not less. A two-wick layout is another route for wide jars, but at 7.5 cm it changes the spacing and the whole bracket; that belongs to the double-wick cluster. The same soot problem in a 10 cm jar, with two wicks, is the 10 cm coffee soot matrix.

What the clock needs

Five items, chosen on page statements, price and stock on 29 September 2026. CSI publishes no soot or wick-performance data, so nothing here is ranked on burn results. Candle counts use 150 g at 8% (12 g of oil).

Frequently asked questions