My 1 Kg Test Batch Is Stable but the Same Formula Sweats When I Manufacture 20 Kg — What Scale-up Variables Should I Investigate?

What actually changes between 1 kg and 20 kg

A 1 kg development batch is forgiving. You weigh 920 g of wax and 80 g of oil on the same scale, the wax temperature is easy to read, a stirring stick reaches every corner of the jug, and five candles are poured within a couple of minutes. At 20 kg every one of those conditions shifts. You are now handling 18,400 g of wax and 1,600 g of oil for one hundred 200 g candles, and the batch has more places for an error to hide.

So when the small test stays clean and the production run shows oil droplets, the honest starting assumption is that the formula held and the process did not. That assumption saves money. A new wax or a new fragrance oil costs a full re-validation; a process fault costs a thermometer reading and a revised batch sheet.

Recheck the arithmetic and the weighing basis

In this series fragrance load is a percentage of total candle weight. At 8% a 200 g candle is 184 g wax plus 16 g oil, so a 20 kg batch is 18,400 g wax plus 1,600 g oil. If your 1 kg test followed that rule but the 20 kg sheet did not, the two batches were never the same formula.

The most common scale-up error is not a wrong percentage but an estimated wax weight. Suppose the melter was filled to a line that someone believed meant 18.4 kg, but only 17,000 g of wax was actually in it. With 1,600 g of oil added, the real load is 1,600 ÷ 18,600 = 8.60% of total weight, not 8%. That may still sound modest, yet for a dense oil in a soy wax it can be the difference between a clean top and a greasy one.

Keep the wax page's basis in mind too. CSI 464's page says "up to 10%", which is a wax-weight figure: 10 ÷ 110 = 9.09% of total. A batch that drifted from 8% to 8.60% has eaten a large share of the headroom between your test and that ceiling, and a stated maximum is a ceiling for typical oils, not a promise for yours.

Addition temperature and mixing reach

At both scales the oil is the same 8% of the batch, so the average temperature drop when room-temperature oil goes in is roughly proportional. What changes is where that oil sits before your stirring reaches it. Pour 80 g into a jug and it is dispersed in seconds. Pour 1,600 g into one spot of a 20 kg pot and a cooler, oil-rich layer can sit on the surface or slide to one side while the stick works only the middle.

Read the temperature properly as well. The Electric Wax Melter page states a temperature display of 60°C–100°C with a stated ±10°C deviation. A display reading of 85°C could therefore be anything from 75°C to 95°C in the wax. In a 1 kg jug you read the wax directly; in the big pot people often trust the display. Take a separate thermometer reading at the centre, the edge and near the base before the oil goes in, and stay below the oil's published flashpoint.

Mixing time does not scale in a straight line, and longer is not automatically better. The question is reach: does your stirring lift wax from the bottom and pull it from the walls? A stirring stick that served the 1 kg jug perfectly may simply be the wrong tool for 20 kg. Change the method — a longer paddle, stirring in a figure-of-eight that reaches the base — and record how long it took, rather than adding minutes blindly.

Holding time, pour order and cooling mass

Five candles are poured before anything can settle. One hundred candles take long enough that the batch sits, cools at the edges and may start to separate if it is not restirred. Mark the first 20 candles, the middle 20 and the last 20 with a dot on the base. If only the last group sweats, the batch became non-uniform in the pot; if all groups sweat equally, look earlier in the process.

Cooling is the variable nobody writes down. Five jars spaced on a bench lose heat freely. A hundred jars packed shoulder to shoulder on trays warm each other, so the candles in the middle of the tray cool much more slowly than those on the edge. Slower or uneven setting can change the surface and how firmly oil is held as the wax sets. Check whether sweating follows tray position.

Isolate the cause with one controlled day

What a failed 20 kg run costs — and what to buy

Using CSI prices as an example, the wax in this batch is 18,400 × ₹0.5074 = ₹9,336.16 of CSI 464 (10 kg at ₹5,074.00). If the oil were Dark Vanilla Oud at ₹5,691.00 per kg, the oil is 1,600 × ₹5.691 = ₹9,105.60. With a matte 200 ml jar at ₹70.80 and an Eco C1 wick at ₹5.90, jars are 100 × ₹70.80 = ₹7,080.00 and wicks 100 × ₹5.90 = ₹590.00, so materials for 100 candles come to ₹9,336.16 + ₹9,105.60 + ₹7,080.00 + ₹590.00 = ₹26,111.76, before packaging, labels, labour, shipping and GST.

Against that, the investigation needs a ₹354.00 thermometer, a properly weighed batch and a day of careful observation. Buy wax for the corrected run as 2 × ₹5,074.00 = ₹10,148.00 for 20 kg of CSI 464, which leaves 1.6 kg for the control. Do not buy a different wax or oil until the control and the corrected run have told you which side of the line the fault is on. And if the investigation shows the oil only stays put below 8%, use less of it — that is cheaper for you even though it means you buy less from us.

Whatever the outcome, candles from the sweating 20 kg run are not for dispatch while you investigate. Hold them, label them by pour group, and decide on rework or scrap once you know the cause.

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