A Fragrance Oil Works Beautifully in Wax Melts but Gives Weak Throw in a Wicked Candle — Why Can the Same Oil Perform Differently?

Two delivery systems, one oil

Makers often treat a strong melt as proof that an oil is strong, and a weak candle as proof that something has gone wrong with the oil. Both conclusions skip a step. The oil sits in the same wax in both formats, but almost everything around the oil changes: where the heat comes from, how much of the surface is liquid, whether a flame is present, and how long the wax lasts. Each of those changes how much fragrance reaches the air and how far it travels. The table sets them side by side.

Read down the last column and a pattern appears. Almost every difference gives the melt an easier job. The warmer liquefies the whole surface without trying, it never burns any of the scent, and it is usually placed in a small room close to the person judging it. The candle has to earn its pool. If the wick is a size too small, the pool may stop well short of the glass for the first few burns, and much of the scented wax is locked in the solid wall. If the wick is a size too big, the pool gets hot quickly, the flame grows, and in general practice a larger share of fragrance near the flame is combusted. Neither of those is a fault in the oil.

There is also a habit hidden in most comparisons. Makers who pour melts tend to run them at a higher load than their candles, because melts do not have a flame to smother and do not need to burn cleanly. So a "same oil" comparison is often a 10% melt against a 7% or 8% candle. That is a comparison of two formulas. Before anything else, put both formats on the same load in the same wax. Whether every scent should share one load is its own question, but for this diagnosis they must match.

The pool arithmetic

It helps to put numbers on the two pools. The figures below are illustrative: a jar with a 7.5 cm internal diameter, a pool 1 cm deep, and a working density of 0.9 g per cm³ for liquid soy wax. Your jar and wax will differ, so measure your own. The page uses CSI's standard load convention throughout: load is the share of the finished weight, so oil = total weight × load.

On these assumptions a full candle pool holds 3.18 g of oil in its liquid, about twice what a 20 g warmer charge holds at the same load. Even a narrow pool holds 1.41 g. So a weak candle is seldom starved of fragrance. What changes is the open surface and the temperature pattern across it. A 5 cm pool in a 7.5 cm jar exposes only 44.4% of the surface a full pool would. The warmer, meanwhile, exposes its whole dish at an even temperature, with no flame consuming vapour at its centre.

This is why raising the candle's load is usually the wrong first move. Adding oil increases what is dissolved in a pool that may not be releasing it well. A candle at 10% with a narrow pool can smell weaker than a candle at 8% with a full one. Why an 8% candle can beat a 10% candle works through that effect, and whether to raise load or fix the wick first gives the order of operations.

The split test

The test below removes every difference except the one you are investigating. You make one pot of scented wax per oil and pour it into both formats, so the wax, the load, the mixing and the pour temperature are identical. Then you cure and judge both under the same conditions. For temperatures, follow the wax maker's instructions. The only CSI temperature on record is CSI 464's 80–90°C melt and pour guidance, so confirm your wax's figures on its product page.

A 1 kg bag of Luxury Soy Wax Chunks covers the 846.4 g needed with 153.6 g to spare. Plan 3–5% extra for pour loss as a working assumption; the spare wax covers it. The 60 g of melts is an illustrative quantity. CSI's Wax Melts Containers are listed without a published fill weight, so weigh one filled with water before you pour, and scale the melt row to match.

Fixing the candle side

Once the split test points to the candle, the fixes follow the order of the decision sheet. They are general candle-making practice, so treat each as a test to run rather than a guarantee.

Give the first burn enough time. A pool that stops short of the glass on the first burn tends to stay short. Burn long enough for the melt to reach the edge before extinguishing, and judge throw only once that happens. If the pool never reaches the glass in a reasonable burn, the wick is likely undersized for that jar. CSI does not publish wick sizes per jar diameter, so this has to be a burn test: two candles, same pot, two wicks.

Watch the flame, not just the nose. A tall, flickering or sooting flame suggests a wick too large for the wax and load, or a fragrance-heavy pool feeding the flame. That candle burns fast and hot, and the room may smell of candle more than of fragrance. When the wick struggles at 10% and why less fragrance can burn better cover the two sides of this.

Cure before you judge. Melts are often tested later than candles because they are poured as a side batch. If your melts were a week old and your candle was two days old, that alone can explain the gap. When to stop waiting on cure sets the point at which a candle has had a fair chance.

Only then, the load. If a full pool, a well-behaved flame and a fair cure still leave the candle behind the melt in the same room, run a small load ladder in the same wax. Luxury Soy Wax Chunks are rated up to 13% and CSI 464 up to 10%. Those are the wax makers' stated formulation maximums, not targets. Finding the lowest load that still throws shows how to ladder without wasting oil.

What to buy for the test

Two oils make the test more informative than one. If both behave the same way, the cause is your candle system. If one closes the gap and the other does not, the oil matters in this wax. We use one fruity-tropical and one gourmand, both in stock at 50 g and 1 kg, with prices from CSI live pricing, pulled 24 September 2026.

For context, one finished 200 g candle from this test costs ₹266.92 in materials with Mango Coconut at the 50 g rate (oil ₹80.00, wax ₹110.22, wick ₹5.90, jar ₹70.80). That excludes labour, packaging, freight and GST. At the 1 kg rate the oil in that candle drops to ₹63.42, and 60 g of melts at 8% carries ₹19.03 of oil (₹23.78 at 10%). If you would rather start smaller, the 15 g Mango Coconut at ₹93.22 scents 187.5 g of melts at 8%, enough for a melt-only check before committing to the full split test.

When the split test is done you will know which half of the problem you own. If it is the candle, the concentration-versus-diffusion test and an objective hot-throw method are the next steps. If it turns out to be the oil, why two equally strong oils throw differently explains what to look for in a replacement, and how to judge an oil bought online covers what to ask before you buy it.

Frequently asked questions