How Do I Test Whether Fragrance-Addition Temperature Rather Than Fragrance Percentage Is Causing My Candles to Sweat?

Why a grid beats two separate guesses

The usual response to sweating is to change several things at once: lower the load a little, add the oil a bit warmer, stir longer. If the next batch is dry, you are relieved and none the wiser. You do not know which change mattered, so you cannot protect it, and you may be carrying a cost, such as a weaker scent, that you never needed.

A 2 × 2 grid fixes that. You pick two levels of addition temperature and two levels of load, and you make every combination. Four cells. Each cell differs from its neighbour by exactly one variable. When you look at which cells sweat, the pattern points at the cause.

It also catches something a one-at-a-time test misses: interaction. Some wax-oil pairs hold a higher load perfectly well when the oil goes in warm and weep only when it goes in cool. In that case neither variable alone is the problem, and fixing either may be enough.

Setting up the example grid

The worked example uses Luxury Soy Wax Chunks, whose page states a maximum of 13% of wax weight. Converted to this series' basis, that is 13 ÷ 113 × 100 = 11.50% of the candle. Both test loads below sit well inside it, which is the point: you are testing a sweat that happens at a load the wax page says is fine.

The oil is Lavender, whose page states a 6–10% candle load. The two loads are 9% (your current, in this example) and 7% (a clear step down). On a 200 g candle, 9% is 18 g oil and 182 g wax; 9% of total is 18 ÷ 182 = 9.89% of wax weight. 7% is 14 g oil and 186 g wax, or 7.53% of wax weight.

Lavender's page does not publish a flashpoint, so pick your controlled addition point conservatively and write down the probe reading. Your 'usual' point should be what you really do, measured, not what you intend to do. If your usual habit is to add oil when the pot 'looks ready', measure what that means three times before the test and use the average.

Choosing the gap between the two temperatures

Make the gap large enough to matter and small enough to be realistic. If your two addition points are only a couple of degrees apart, the difference will be lost inside your thermometer's error and the melter's drift, and the grid will report that temperature does not matter when you simply never tested it. If the gap is enormous, you test a condition you would never use in production. A practical approach is to set the 'usual' level at what your staff actually do on a busy day, and the 'controlled' level at the measured point you would be willing to write into a method.

Choosing the gap between the two loads

The same logic applies to load. A step from 9% to 8.5% is too small to read reliably against normal candle-to-candle variation. A step from 9% to 7% is large enough to see, and 7% still sits inside Lavender's stated range, so the lower cell is a candle you could genuinely sell if it wins. Weigh every portion: at 200 g, the difference between the two loads is only 4 g of oil per candle.

Running it without contaminating the result

Reading the grid

Observe on suggested days 1, 2, 3, 7 and 14, at the same time of day. Look first without touching, then blot each surface with a fresh tissue and note whether the mark is greasy and smells of Lavender. Record each candle as dry, faint film, droplets or pooling.

Do not call it on one candle. If two of three in a cell sweat, the cell sweats. If one of three shows a faint film and the others are dry, note it and look at that candle's position on the shelf before drawing conclusions.

Scent is part of the result

Sweating is the main read, but do not ignore throw. After the oil page's stated soy cure of 48–72 h and again at a later suggested point, smell each coded jar cold, then run an attended, supervised test burn on one candle per cell. If the 7% controlled cell is dry and smells as good as the 9% cells, you have learned that the extra 4 g per candle was buying nothing but risk.

What the grid cannot tell you

It tests one wax with one oil in one jar. It does not tell you how the winning cell behaves in a delivery van in May, or with a different oil, or in a wider jar. Treat the result as an answer to the question you asked, and run separate checks for the others.

What the grid costs

Using Chunks at the 5 kg rate of ₹2,995.00 (₹0.599 per gram) and Lavender at the 500 g rate of ₹2,620.00 (₹5.24 per gram): wax 2,208 × ₹0.599 = ₹1,322.59; oil 192 × ₹5.24 = ₹1,006.08. Twelve matte jars at ₹70.80 each are 12 × ₹70.80 = ₹849.60, and twelve Eco C1 wicks at ₹5.90 are 12 × ₹5.90 = ₹70.80. One 5 kg Chunks pack and one 500 g Lavender pack cover the grid with plenty left over. Packaging, labour, shipping and GST are excluded.

If the grid says load, you will buy less Lavender from now on. That is a perfectly good outcome for you, and we would rather you knew it.

After the grid

Write the winning conditions into your batch sheet as a range you will hold. If your candles travel or sit in hot rooms, run a short supervised warm-room or cabinet check on the winning cell at a test condition you label yourself; never a kitchen oven or open flame. Then, and only then, scale it up and dispatch.

Where this page fits

CSI already has guides that cover parts of this question. This page answers one sweating decision; these cover the neighbouring questions:

Three things to buy for this fix

Frequently asked questions