How Do I Scale a Candle Formula From Laboratory-Size Testing to Commercial Production Without Assuming That Simply Multiplying the Recipe Is Enough?

A formula is four layers, and only one of them multiplies

A laboratory-size test, say two 200 g candles made from 400 g of scented wax, carries a lot of knowledge that never gets written down. You knew how the wax looked when it was ready, how long you stirred, how quickly you poured, where the jars cooled. At that size, all of that lives in your hands. At commercial size it has to live on paper, because the process is now longer, heavier, and often run by more than one person.

A useful way to structure the move is to split what you know into four layers. Only the first one multiplies.

Most scale-up failures come from treating the third layer as if it were the first. Multiplying stir time by the batch ratio, or assuming the pour window does not matter, turns a guess into an instruction. The method below keeps the layers separate.

Translating a lab card into a master batch record

Take a real case. A maker's lab card for Eucalyptus Spearmint in CSI 464 reads: 368 g wax, 32 g oil, 8% of total, two 200 g jars, stirred until clear, poured when it felt right. The commercial target is 20 kg: 18.4 kg of wax and 1.6 kg of oil, filling 100 candles before losses and about 95 after the 5% planning allowance.

Notice that half of the lab card's lines had no number at all. That is normal at 400 g, and it is precisely where commercial batches go wrong. The job of the batch record is to turn every unwritten habit into an explicit instruction, and to label each number you choose as a starting point until a batch has proved it.

A good test of the finished record is to hand it to someone who has never seen you make the candle and watch them follow it on a small batch. Every question they ask marks a line that still depends on your memory. Rewrite those lines until the record answers them, because at commercial scale the person pouring the last pitcher at the end of a long day needs the same clarity as the person who wrote it.

Define the pass before you make the batch

Writing acceptance criteria after seeing the batch invites generosity. Write them first, from your lab candles, while they are still the standard.

Eucalyptus Spearmint is rated moderate to strong on its product page. That is the page's own description, not a measurement, so your acceptance criterion should be comparative, against your control, rather than an absolute judgement of strength.

Treat the first commercial batch as the proof

The first 20 kg batch made to the new record is a validation batch. Make a 1 kg control on the same day from the same wax bag and oil bottle. Follow the record exactly, even where it feels over-careful, and write down every deviation.

The material on the line is real money. The 20 kg batch uses 18.4 kg of CSI 464 at ₹507.40 per kg, which is ₹9,336.16, and 1.6 kg of Eucalyptus Spearmint bought as the 1 kg, 500 g and 100 g packs: ₹3,068.00 + ₹1,652.00 + ₹354.00 = ₹5,074.00. That is ₹14,410.16 in total, or about ₹151.69 per candle over 95 sellable jars. A validation batch that passes is stock worth that much; one that fails is a lesson you want to have paid for only once.

If it passes against the criteria you wrote in advance, release it and mark the record as validated at 20 kg. If it fails, hold it and use the retains to find where. A pattern by pour order points to hold time; random spread to stirring; uniform weakness to addition temperature or heat exposure. Change one scale-dependent parameter and repeat.

Keep the record alive with change control

Once validated, the master batch record becomes the product. Any change to it, whether a new wax lot, a different oil, a bigger melter, a new operator, or a batch size above the validated one, reopens the relevant layer. A new oil reopens the formula. A new melter reopens the scale-dependent parameters. A new lot of the same material usually needs only a 1 kg control against the retained standard.

Keep the lab retains, the validation batch retains and the batch records together. When a customer later says a candle smells weaker, you will have what you need to check it: the standard, the record of what was done, and candles from the same position in the same batch. That is the difference between a formula that has been multiplied and a process that has been scaled.

Where this page fits

CSI already has guides that cover parts of this question. This page covers the failure investigation for the question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions