Why Do Some Fragrance Oils Work Better in Candles Than Others?

The five factors that decide whether an oil works in your candle

When a maker says an oil "doesn't work", the oil is rarely the only thing that changed. Wax, load, wick, vessel diameter and cure length usually moved with it. The five factors below carry the result — and four can be settled before you spend anything.

Read the right-hand column as a map of the gaps. Structure is nearly always published; load usually is; cure is on well over half the range. Wick and vessel never are — 12 pages suggest a wick type but none can size one for your jar, because that depends on your vessel diameter, your wax and your load. The honest division of labour: the page narrows the field, the burn test picks the winner.

Note structure: why a published pyramid predicts a burn

Here is the general principle, stated carefully because it is often overstated. Perfumery organises aroma materials by volatility — how readily a material evaporates at a given temperature. The convention is three tiers: top notes are the most volatile and register first, heart notes carry the body, base notes are the least volatile and last longest. That is a descriptive framework used across the fragrance trade, not a measurement of any particular oil, and no CSI page publishes evaporation data. What the pages do publish is which tier each material sits in for that specific oil, and that is enough to form a prediction.

The prediction: a candle presents fragrance two ways. Unlit, only the most volatile materials reach you from a cold surface, which is why a jar's cold throw is dominated by top notes. Lit, a melt pool holds the whole blend at a raised temperature for hours — the conditions heavier materials need. A front-loaded oil tends to sell itself on the shelf; one built on woods, amber, musk, tonka or vanilla tends to win the second hour. Neither is better; they are better at different jobs.

The same oil, two waxes — and why a ceiling exists at all

Twenty of the 107 oil pages state a candle load per wax type, and on every one the paraffin figure is lower. Fourteen use "Up to 10% for soy wax" with "Up to 6% for paraffin wax"; six use "6-10% for soy wax candles" with "6-8% for paraffin wax". So an oil is not simply "good in candles" — it is stated at one level in one wax and lower in another. The waxes differ in the two things a fragrance cares about: how much oil the formulation holds, and how hot the pool runs. CSI's wax pages state both.

The CSI 468 row answers a question makers ask constantly. An oil with a low stated flash point — Gardenia's page states "Flashpoint 98°C" and instructs you to "add this oil to your wax at a temperature below 98°C" — is easy in a wax whose stated pour is 58–60°C and needs care in one whose stated add band is 80–90°C. The oil did not change; the wax did. See does flash point determine when oil should be added to wax and best temperature for adding fragrance oil to soy wax.

A stated ceiling is not a recipe preference either. It describes how much oil a solid wax holds in solution once set. Below the limit the oil is distributed through the wax and evaporates from the melt pool as the candle burns. Above it the surplus has nowhere to go: it migrates to the surface and the glass, shows as an oily film or beads under a lid, and feeds the wick something it was not sized for. That is the mechanism behind every sweating, separating, drowned-wick jar. The difference between a load that works and one that sweats is small in grams and obvious in the jar — here is the same 200 g candle at five loads, costed at three real price points.

Vanillin: the one failure your customer photographs

Discolouration is the best-documented of these factors and the most misunderstood. Vanillin is an aroma material associated with vanilla and with many gourmand, tobacco, caramel and amber accords; in wax it can tone white or pale wax toward cream, yellow or brown, usually over weeks. It is cosmetic, not a performance fault — the pages say so plainly — but a white candle gone yellow in clear glass on a customer's shelf reads as a defect whatever the burn is doing. Sixty-two pages take a position, and the wording matters: the gap between "0%" and "Present" is the gap between a white-wax range and an amber-glass one.

The three-oil comparison that actually answers "why"

Everything above narrows the field; this settles it. The design compares three oils of deliberately different structure in one wax, at one load, on one wick, at one cure length — so the only thing that differs is what you are trying to learn. One evening of pouring, one of burning a fortnight later.

Two oils whose pages let you predict the burn

These are not "the best oils in the range" — that claim is unfalsifiable and we are not making it. They are the two opposite corners of the structure argument, and both pages publish enough to form a prediction before you spend anything: a stated structure, stated loads for soy and paraffin, a stated cure window, a stated vanillin position and a measured flash point. Each card shows every size, the price per 100 g, and how many 200 g candles at 10% each pack covers — 20 g of oil and 180 g of wax, the stated soy ceiling on both pages. The 15 gm trial reads ≈0 because it is a smelling size, not a full candle.

Frequently asked questions