My Large Batch and Small Control Both Suddenly Fail — Should I Investigate Raw-Material Batch Changes Before Changing the Production Process?

What it means when the control fails too

Most scale-up investigations begin with a simple comparison: make the approved formula again at 1 kg, the old way, and see whether it still works. If it does, the large batch's process is the prime suspect. You are in the other situation. The fresh control failed in the same way as the 20 kg batch, which means the thing that went wrong is present in both — and batch size is not one of the things they share.

That narrows the field considerably. The shared inputs are the wax, the fragrance oil, the wick, the jar, any dye or additive, the measuring instruments, and the way you judge the result. The process differences between 1 kg and 20 kg — heat retention, mixing energy, hold time, pour duration — are now unlikely to be the main driver, because the 1 kg control did not experience them and still failed.

Unlikely is not impossible. A process fault can hit both batches if it lives in something they share, such as a thermometer reading low so both received oil too hot, or a scale drifting so both were under-dosed. Those count as shared inputs too, and they go on the suspect list alongside the materials.

List every shared input and what changed

Before you make a single test candle, write down what is different between the approved candle and today. Go through your purchase records and your shelf, not your memory. A newly opened wax pack, the second kilo of an oil, a fresh box of wicks from a different order, a new jar delivery, a replacement thermometer — each is a candidate. Anything that has not changed since the last good batch drops down the list.

Then check the test itself. The approved reference candle may now be months older than the new ones. The burn room may be warmer or more ventilated. You may be testing at a different cure age. If the reference candle has aged while the new one is young, the comparison was never fair.

In the example that follows, two inputs changed: a new pack of CSI 468 and a new bottle of Bergamot Amber. Instruments come first because a wrong reading would contaminate every control you make.

Run a 2 × 2 swap with retained material

The cleanest test is a swap. If you have retained old-stock wax and oil — a sealed portion kept back from the pack and bottle used in the approved candle — you can make four controls that differ in only one respect each.

If A is good and D is weak, something changed. If B is weak and C is fine, the oil moved. If C is weak and B is fine, the wax moved. If all four are weak, including A, the materials are probably not the cause — look at the reference, the room, the wick box or your judging.

Worked example: 20 kg of coconut soy with Bergamot Amber

Suppose the failed batch was 20 kg of scented wax at 8% of total weight, poured into 200 g jars. That is 18.4 kg of wax and 1.6 kg of oil, filling 100 candles before losses.

Cost of the batch at risk. CSI 468 in the 10 kg pack is ₹6,370.00, or ₹637.00 per kg. 18.4 kg × ₹637.00 = ₹11,720.80. Bergamot Amber in the 1 kg bottle is ₹5,437.00, so 1.6 kg × ₹5,437.00 = ₹8,699.20. Wax and oil together: ₹11,720.80 + ₹8,699.20 = ₹20,420.00, before wicks, jars and labour.

Cost of the swap. Four controls use 4 × 460 g = 1.84 kg of wax and 4 × 40 g = 160 g of oil. At the same per-kg rates: 1.84 × ₹637.00 = ₹1,172.08 and 0.16 × ₹5,437.00 = ₹869.92, a total of ₹2,042.00. The swap costs about a tenth of one batch, and it stops you repeating a ₹20,420.00 mistake.

Now imagine the result: A and B are good, C and D are weak. The new wax pack is the common factor in both weak candles. Bergamot Amber is cleared. The product page's intensity guide says the bergamot top leads at 6% and amber comes forward at 10% — useful when you later judge character, but irrelevant to this finding because both oils were dosed identically.

What to do with the held batch

While the swap cures, the 20 kg batch stays on hold: set, covered, labelled with the batch code and the word HOLD, and kept away from finished stock. Do not remelt it to try a fix; every remelt spends more of the volatile top notes.

Once the swap reads out, the decision follows the evidence. If the new wax is implicated, a re-wick or a load adjustment might be tested on a small portion — but only as a fresh validation, not a patch. If the old reference also fails, the batch may be fine and your judging method was the problem; burn-test held candles against a freshly made reference before writing anything off.

The control that stops a repeat

The fix for next time is cheap. Keep a sealed retained portion from every wax pack and oil bottle you open for production, labelled with your own receiving code and date. Write that code on the batch record. Check thermometer and scale against a reference before every production day. And when you open a new pack or bottle, pour one 500 g control beside a retained-stock control before the first large batch uses it.

If this investigation keeps pointing at incoming material, move the approval step earlier: approve each new pack before it enters production rather than discovering the change in finished candles. WhatsApp CSI on +91 7397976926 for IFRA and MSDS documents on any oil you rely on, and keep them with the batch file.

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