I Changed from 6% to 10% Fragrance in the Same 7 cm Jar and the Flame Became Smaller — Why Can Higher Fragrance Loading Change Wick Performance?

Why more oil can mean less flame

A container candle's flame is fed by capillary action: liquid wax and oil climb the wick fibres from the melt pool, vaporise near the flame and burn. The wick controls how much fuel arrives; the fuel controls how it burns. When you raise the load from 6% to 10% of total weight, the pool changes from about 6 parts oil in 100 to about 10 in 100. That is a small change in the jar and a big one at the wick: 9 g of oil becomes 15 g per 150 g candle, 67% more.

What happens next is described in candle-making practice as tendencies, not rules, and they are worth stating plainly:

Two cautions keep this honest. First, not every oil does it. A light citrus like the worked example's Citrus Lemon may show a small change or none; heavier, vanillin-rich or resinous oils are the ones commonly linked to bigger shifts (vanilla, coffee and oud each have a post in this cluster). Second, the ladder measures your oil, not oils in general. CSI publishes no wick or flame data for any oil, and the Citrus Lemon page says nothing about wicks in the text we captured. The only way to know what your 10% does is to put it next to your 6% on the same wick.

What this post does not cover: if your candle burned perfectly unscented and misbehaved once scented, the unscented wick was only a baseline, and the scented re-bracket from an unscented baseline owns that. If the flame shrinks session after session at one load rather than across loads, that is a clogging pattern read with the flame-shrink curve.

Is 10% allowed? Two ceilings, one answer

Two limits apply to every load: the wax page's and the oil page's, and the lower one governs. They are formulation limits, not regulatory ones. Load here is always a percentage of total candle weight, CSI's convention: 150 g at 10% = 15 g oil + 135 g wax. The load-basis post explains the conversion if your supplier uses the other basis.

Two things follow. The ladder's rungs (6, 8, 10) are the bottom, middle and top of the oil page's own range, so none of them is improper. And there is no 11% or 12% rung to try "to push throw" with this oil, even though the wax would take it: the oil page stops at 10%. If you use Luxury Soy Wax CSI 464 instead, its page states "Fragrance Load Up to 10%" and the same answer holds. Max load is not best-performing load: the page ranges tell you where to test, not where to stop.

The load ladder: six candles on one wick

Hold everything constant except the Citrus Lemon load: the same wick (the one you used at 6%; C1 in our worked example), the same jar batch, the same 150 g fill, the same pour temperature, the same cure and the same room. Pour all six on the same day from the same wax. Two candles per rung is enough to spot an odd candle; it is not enough to approve a wick, which is why the winner still goes to full life.

Materials consumed: ₹216.00 of oil, ₹495.97 of wax (₹0.60 a gram), ₹566.40 of jars and ₹35.40 of C1 wicks, ₹1,313.77 in all. Candle materials each include the jar (₹94.40) and wick (₹5.90). These are worked examples that exclude labour, freight, packaging and GST. A 50 g bottle would not do: the ladder needs 72 g, so it would be 22 g short.

Measuring a flame against a ruler

"Smaller" needs a number. We measure trimmed flame height: from the surface of the melt pool to the visible tip, on a wick trimmed to the same length before lighting. Many makers use about 2.5–5 cm as a practical range for a container candle's flame: a commonly used rule of thumb, not a CSI figure. You need a steel ruler, a phone and a still room.

Keep hands and paper well clear of the flame; the card goes above, not into it. Never touch the glass to judge heat: this ladder is about flame size, and if you want glass readings, use a non-contact thermometer (CSI does not list one). The session length follows a rough rule of thumb of about one hour per 2.5 cm of width, 2.5 h on a 6.3 cm inside (2.8 h on the listed 7 cm), so a 3-hour session gives the pool a fair chance.

Reading the grid: lower load or re-bracket?

Lay the twelve photos out as the diagram does: loads across, times down. Read the 10% pair against the 6% and 8% pairs, and read both twins. One short flame in a pair is noise; two are a result. In the worked example the 6% and 8% flames sit in the range at both times and both pools meet the glass by 3 hours. The 10% flames are 12–14 mm shorter than the 6% flames at 2 h, drop under 2.5 cm by 2 h, carry a carbon bead, and the pool stops about 4 mm short of the glass (76% of the surface melted). That is a 10% fail on this wick.

The blind throw test, 8% v 10%

Use the passing 8% pair and the failing 10% pair in the same session. Someone else codes and places them; you walk in after 1 hour and after 2 hours with a fresh nose, and say which smells stronger, or "can't tell". Three separate sessions or three noses. If 10% is not picked at least 2 times out of 3, the extra oil is not buying you anything you can smell, and 8% wins on burn, cost and margin. How throw is judged in a room is covered in depth in the formulation series; the less-oil-burns-better log is the one to read if 8% turns out to burn better overall.

If you do re-bracket at 10%

Keep 10% fixed and change only the wick: two candles on C1 (the control) and two on C2, fresh pours, same everything else. That is four candles: 60 g of Citrus Lemon and 540 g of wax. After the ladder the 100 g bottle holds 28 g, 32 g short, so add a 50 g bottle (₹170.00; 18 g left afterwards), one more 500 g bag of Chunks (132 g left afterwards) and a C2 10-pack. The four spare jars from the 10-pack and 14 spare C1 cover the rest. Top-up cash: ₹563.40; materials consumed: ₹924.54.

Then the honest limit. At 7 cm CSI's tabbed rungs are C1 and C2, with no sized wick in between; the next rung, Cotton Braided Wicks Ready Made, is described by its own page as "engineered for large-diameter candles exceeding 8 cm". If C1 fails at 10% and C2 overshoots (a flame well above 5 cm, fast pool, a very deep pool, heat near the glass), your 10% wick sits between CSI's rungs. Choose 8% on C1, or ask us on WhatsApp and consider another supplier's size. Whether two rungs close a bracket is the adjacent-sizes post; a winner at 10% still needs the full life.

What the higher load costs

Of the wax and oil in a candle, oil is the dearer by far per gram (₹3.00 against ₹0.60 here), and the load sets how much of it you pay for. The table uses the 100 g price you will test with and the 1 kg price you might produce with. The wax each candle uses falls slightly as the load rises, which offsets a little of the extra oil.

Going from 8% to 10% adds ₹9.00 of oil per candle at the 100 g price and ₹8.50 at the 1 kg price. From 6% to 10% the net material change, after the 6 g of wax it displaces, is ₹14.41 more per candle at test prices. Over 100 candles at the 1 kg price, 8% uses 1,200 g of oil (₹3,398.40) and 10% uses 1,500 g (₹4,248.00): ₹849.60 for the choice of load alone, before any wick.

If you are changing the load across a whole range of approved candles, not one jar, the requalification plan is the collection-wide 8% to 10% post. If the 10% wick builds carbon every session, score it with the mushroom-scoring post; the map of this cluster is in the 6–7 cm hub.

What to buy, and what to skip

Five items, in the order the ladder uses them, chosen on page statements, price and stock on 29 September 2026. CSI publishes no wick or flame data, so nothing here is ranked on burn results. Candle counts use 12 g (8%) and 15 g (10%) of oil per 150 g candle.

Frequently asked questions