How Many Test Candles Should I Pour?

For one wick test, suppliers say 2 or 3 candles: same jar, same wax, same fragrance, and only the wick size changes. Then the count grows with every variable you add. NorthWood says that ideally you test each wax, wick, container, fragrance type and fragrance load you plan to use, and that it is wise to redo the test if you change any of them. So the real answer is a number you work out: count the variables, multiply, and stage the test.

What do suppliers say?

• CandleScience: make 2 to 3 candles using the same size container, wax and fragrance and dye. Put a different size wick of the same series in each. Wait at least 24 hours, then burn them at the same time.
• Makesy: burn-test at least three wick sizes in your vessel before a production run.
• NorthWood: test 2 to 3 sizes. You will probably need separate test candles for each fragrance, and new ones if you adjust the load. Ideally, do individual tests for each wax, wick, container, fragrance type and load.
• Midwest Fragrance: for a weak throw, it suggests a wick 1 to 2 sizes up or down, or testing 6, 8 and 10 percent loads, with 10 only if your wax allows it.

None of the sources I read says to pour two identical candles to check consistency. I could not find a source on that, so I do not state a rule.

How many burns does each candle need?

One candle gets several sessions. CandleScience says to wait until at least the fourth burn cycle before you draw conclusions about a wick. So a test candle is a candle you burn through, not a sample you light once. See our burn test guide.

Worked example: counting the candles

Say one jar, one wax, three fragrances and three wick sizes, all at an 8 percent load. Each candle holds 8 oz (227 g) of wax. All numbers are examples.

Full grid, as NorthWood describes it:

• Wick sizes x fragrances: 3 x 3 = 9 candles
• Wax: 9 x 8 = 72 oz (2,041 g)
• Oil at 8 percent of the wax: 72 x 0.08 = 5.76 oz (163 g). Check per candle: 8 x 0.08 = 0.64 oz, and 9 x 0.64 = 5.76.
• Add one "official" candle per fragrance, made exactly the way you will make them for sale, as NorthWood suggests: 9 + 3 = 12 candles

Add loads and it grows fast. Three wicks x three fragrances x three loads is 3 x 3 x 3 = 27 candles. That is why I stage it.

A staged test (my own habit, not a source's rule):

1. Pick the wick on one fragrance with 3 sizes: 3 candles.
2. For each of the other two fragrances, pour the winning size and the next size up: 2 x 2 = 4 candles. I go up because CandleScience says a dense fragrance or heavier load can need a bigger wick.
3. One official candle per fragrance: 3 candles.

Total: 3 + 4 + 3 = 10 candles, against 12 for the full grid.

The staged test is weaker, because a fragrance can throw you off in either direction. NorthWood asks for tests on each. With 5 fragrances the gap widens: full grid 3 x 5 + 5 = 20, staged 3 + 2 x 4 + 5 = 16. Pick the staged plan only if you are fine with that trade, and say so on the recipe card.

How do I keep the numbers down?

• Change one variable at a time, the way CandleScience's test changes only the wick. Do not test wick and load in the same round.
• Burn the candles side by side, as CandleScience says, so you only need one test window. Stay with them the whole time.
• Reuse a result only for the same wax, jar, wick and fragrance. NorthWood says it is wise to redo it when any of those changes.

Common mistakes

• Testing a single wick and calling it proven.
• Changing the wick and the load in the same candle.
• Pouring test candles in a different jar than the one you will sell.
• Testing with unscented wax and then adding fragrance later.
• Not labeling each jar with its wick, as CandleScience says to.
• Burning once and moving on, when the fourth cycle is the one that counts.

Questions

Do I need to test every fragrance?

NorthWood leans yes. It says you will probably have to make separate test candles for each fragrance you want to use. A staged plan like the one above is a shortcut that carries some risk.

How many test candles are too many?

When you cannot burn them side by side or keep track of them. If you cannot log each one, you have poured too many.

Do I need a new test when my wax supplier changes a batch?

CandleScience recommends burn testing periodically, at the start of each season or at least twice a year, because supplies vary, especially soy and natural waxes. Retest when the wax batch, oil or dye changes.