My Coffee Fragrance Makes the Flame Much Larger Than My Lavender Fragrance in the Same Candle System — Should Every Fragrance Have Its Own Wick Test?

Why the coffee flame grows on a lavender wick

You approved a wick with lavender. The flame was steady and the pool reached the edge of the jar at a sensible pace. Then you poured coffee into the same jar, with the same wax and wick at the same 8%, and the flame was visibly taller and livelier. Nothing mechanical changed, so the difference has to come from the oil.

A candle flame is fed by capillary action. Melted wax and fragrance climb the wick, vaporise near the flame and burn. Two things decide how big the flame becomes: how much fuel reaches it, and what happens to the wick tip over time. Fragrance oil affects both.

Fuel delivery

Oils differ in thickness and in how readily their components evaporate. An oil that thins the liquid fuel or travels up the wick more freely can raise the fuel supply, and more fuel means a bigger flame. Heavier oils can do the opposite. Either way, the same wick delivers a different amount of fuel depending on what is dissolved in the wax.

Carbon on the wick

CSI Roasted Coffee's page describes fresh coffee beans at the top, dark chocolate and espresso in the middle, and vanilla and caramel at the base, and states that vanillin is present. Rich gourmand materials tend to burn less completely than light herbal ones. As unburnt residue collects, the wick tip can swell into a carbon cap. A capped wick presents more surface to the flame, so the flame itself can grow and flicker, especially late in a session.

CSI Lavender, by contrast, is described as a single-note lavender with a soft musk base. It is a lighter fuel. That is why the same wick behaves calmly with lavender and more aggressively with coffee.

Does every fragrance need its own wick test?

Yes, but not the same depth of test. If you treat every oil as a brand-new wick project, you will spend months testing and still fall behind your launch dates. If you treat none of them individually, you will eventually ship a candle like your coffee one. The practical middle is a tiered policy.

The numbers in that table, such as three candles, are suggested starting points to validate in your own workshop, not rules. What matters is that each oil gets checked against a known reference before it goes into production.

Running the coffee confirmation burn

In your case, coffee has already shown that it behaves differently from lavender, so it has escalated. Put it in a gourmand family group and give it its own short ladder.

If the one-size-down coffee candle matches the lavender reference for steadiness while still giving a clear coffee throw, you have a gourmand wick. If only the 6% variant settles, load is the bigger lever. If neither settles, the oil may not suit this jar and wax without a broader rethink.

What a confirmation test costs against a batch

Makers skip per-oil tests because they feel expensive. The arithmetic says otherwise. Here is a three-candle Roasted Coffee test in 200 g candles at 8% using CSI 464 and Eco Wick Thin C1, before jars.

The 80-candle figure echoes the CSI Electric Wax Melter page's statement that it can 'create 80 candles at a time'. On those numbers, the confirmation test costs about 3.9% of the oil and wax in a single batch (₹477.20 ÷ ₹12,253.57). If the test stops you pouring 80 coffee candles that flare and soot, it has paid for itself many times over, and that is before counting jars, labour and customer returns.

Three conclusions not to draw from a bigger flame

Not: the coffee oil is faulty. A gourmand oil that asks for a different wick is doing what its structure suggests. Faulty would be an oil that separates, sweats or soots on every wick and load you try. One bigger flame on a borrowed wick is simply information.

Not: the lavender wick was wrong all along. The lavender candle passed its own test and still burns well. Changing the lavender wick to suit coffee would trade a working product for a compromise. Leave approved combinations alone and add a new one beside them.

Not: more coffee oil will calm it down. Some makers reason that a stronger oil load will make the fuel heavier and the flame smaller. In practice, adding more of a material that is already building carbon usually makes the cap worse, not better. If load is involved, the test direction is down, not up. The Roasted Coffee page gives a soy range of 6–10%, so there is room to test 6% without leaving the stated range.

Keeping these three misreads out of your notes stops a single odd flame from turning into a range-wide re-wick, which is the most expensive way to solve a one-oil problem.

Keep the policy light enough to follow

A wick policy only works if it is used. Keep one sheet per jar listing each family, its reference oil, its approved wick and its approved load, and add a line whenever an oil passes its confirmation burn. Coffee now becomes the gourmand reference for that jar, and the next chocolate or vanilla oil starts its test on the coffee wick rather than the lavender one.

That single change, testing within families and escalating only when a flame looks different, gives you most of the safety of per-oil testing for a fraction of the effort. The lavender candle that started all this stays exactly as it was.

Where this page fits

CSI already has guides that cover parts of this question. This page answers the specific oil-in-wax question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions