My First Candles From a 10 Kg Batch Smell Weaker Than the Last Candles Poured — Could Fragrance Distribution Be Changing While the Wax Sits?

Read the direction before you guess the cause

When candles from one batch differ, most makers jump straight to the fragrance oil. But a batch has a sequence. The first candle was poured from the batch at one moment, from one part of the vessel, at one temperature, and the last candle was poured later, from a different part, at a different temperature. A consistent gradient from weak to strong is evidence about that sequence.

Start by asking what could make the late candles stronger. Only a few things can. There could be more oil in the wax they were poured from. They could contain more wax overall, and so more oil. Or they could be burning differently because of their wick or how the wax set. Anything that simply removes scent from the batch over time, like evaporation from the hot surface, would push the pattern the other way.

That single observation rules out a common worry. If you were losing fragrance from the melter while it stood, your last candles would be the weak ones. Yours are the strong ones, so the oil is not escaping; it is arriving unevenly.

The four suspects, and what each predicts

It is common for two suspects to work together. An incomplete mix followed by long standing time produces a stronger gradient than either would alone. That is why the run described below collects evidence on all four at once rather than testing one idea at a time.

The diagnostic run

You need one ordinary production batch and a little discipline. Nothing about the recipe or process changes for this run. You are observing, not fixing.

Make sure the test itself is fair

Before you trust any gradient, check that the way you compared the candles did not create it. Several ordinary habits can make identical candles seem different, and they tend to favour whichever candle you smell last.

Uneven cure

If the first candles were tested a day or two after pouring while the last ones were tested later, you are comparing different cure stages, not different candles. Test all kept candles on the same day after the same cure. Any number of cure days you choose for this is your own standard to validate, but it must be the same for every candle in the comparison.

Nose fatigue and order

Noses adapt quickly. The first candle you smell in a session gets a fresh nose and the fifth gets a tired one, which can flatten or exaggerate differences. Shuffle the order, take short breaks outdoors between rounds, and repeat the ranking on a second day.

Different rooms, different burns

Hot throw depends on room size, airflow and how long the candle has been burning. Burn all five in the same room, for the same time, with doors and fans in the same state. If you cannot burn them together, rotate them through the same room in a written sequence.

Only when the test is fair does the gradient mean anything. If a fair test shows no difference, your original impression came from the comparison, not the batch, and nothing in production needs to change.

Worked example: a 10 kg batch, 48 candles

Your 10 kg of scented wax at 8% contains 9.2 kg wax and 0.8 kg oil. At 200 g per candle that is 50 candles exactly, 48 after allowing 5% wastage. You number all 48, weigh each one and keep numbers 1, 12, 24, 36 and 48.

Suppose the weights all sit close to 200 g. Fill weight is ruled out. The wick positions look the same, so wick is unlikely. The blind ranking puts candle 1 weakest, candle 12 close to candles 24 and 36, and candle 48 slightly stronger again. That shape, a lean start, a stable middle and a rich end, suggests two things together: pouring began before the oil was fully mixed, and the last draws came from a remnant that had stood long enough to drift.

Now the fix is specific rather than general. Extend the full-depth mix before the first draw, purge the tap, and re-stir before each draw, especially near the end of the batch when the level is low. Run the same diagnostic on the next batch. If candles 1 to 48 now rank as indistinguishable, you have found your cause and you can write the change into the procedure.

When it turns out not to be distribution

One more thing worth doing, even though it costs a little: keep one or two numbered candles from every batch for a few weeks after it ships. If a customer later says their candle was weak, you can light its sibling from the same position in the run and know whether the complaint reflects the batch or something that happened after it left you. Label those retained candles with the batch code and pour number, and store them away from heat and sunlight so they stay a fair reference.

Where this page fits

CSI already has guides that cover parts of this question. This page takes the production-floor angle; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions