My Formula Works When I Add Fragrance at One Temperature but Sweats When Production Staff Add It at Another — Should Temperature Become a Controlled SOP Parameter?

Your two results are the evidence

You have already done the hardest part of the investigation without meaning to. Your formula, made your way, gives dry candles. The same formula, made by production staff, sweats. Assuming the weights and materials really are the same, the difference lives in the method, and addition temperature is one of the most likely places for it to hide.

Confirm it once before you write rules around it. Stand beside the person whose batches sweat and measure what actually happens: the wax temperature at the second oil goes in, how long they stir, how hot they pour, where the candles cool. Then do the same for your own batch. Write both sets of numbers side by side. If addition temperature is the clearest difference, you have your answer.

A one-day shadow test

Run it in a way that respects the person being watched. The point is not to catch anyone out. In most workshops the staff member has been given an instruction that sounded clear, such as 'add when it is hot', and has followed it faithfully. The fault is in the instruction.

Choosing the target and the band

Where the wax manufacturer publishes an addition instruction, start from it. Among CSI waxes, only the Eco Soy CSI 400 page does: add fragrance at 80–90°C and stir for 2 minutes. If you use CSI 400, that is a ready-made SOP line. For CSI 464, Chunks, CSI 468 or Pearl, no addition temperature is published, so your band comes from your own testing.

Validate the edges, not just the middle. Make candles with oil added at the bottom of your proposed band and at the top, everything else fixed, and observe both on your normal schedule. If both ends give dry candles with acceptable scent, the band is real. If the bottom edge sweats, raise it. A band you have only tested in the middle is a guess with a range attached.

Then check the top edge against every oil you run. The band must sit below each oil's published flashpoint. Most CSI oils with a published figure sit well above typical addition temperatures, but Neroli's is 85°C and Gardenia's page says add below 98°C. Those oils get their own upper limit written on their line of the sheet.

Equipment that makes the band enforceable

A band nobody can measure is decoration. The CSI electric melter page states a 60–100°C display with a ±10°C deviation. That is fine for melting and holding, but a display reading of 85°C could mean wax anywhere from 75°C to 95°C, which straddles CSI 400's whole band and spills past both ends. The SOP must require a separate probe in the wax.

Give every station its own thermometer so readings are not shared, borrowed or remembered. Three stations at ₹354.00 each is 3 × ₹354.00 = ₹1,062.00. CSI does not publish the pen thermometer's range or accuracy, so compare all station probes against each other in the same pot of melted wax when they arrive, and periodically after that. If one reads differently, label it and correct for it, or replace it.

One sheet per oil, not one sheet for all

If your range covers many oils, keep the wax band on a master sheet and add an oil line for each fragrance: its published flashpoint or 'not published', its validated load, and any lower addition limit it needs. Citrus Lemon's page, for example, does not publish a flashpoint but warns that citrus top notes are volatile, which is a good reason to stay in the lower half of the band for it. A single generic sheet for every oil hides exactly the differences that cause trouble.

What an out-of-band batch costs, and why you hold it

Here is the example batch priced from CSI's live rates. Wax: 4,650 g of CSI 400 from the 5 kg pack at ₹1,996.00, which is ₹0.3992 per gram, so 4,650 × ₹0.3992 = ₹1,856.28. Oil: 350 g of Citrus Lemon at the 1 kg rate of ₹2,832.00, so 350 × ₹2.832 = ₹991.20. That batch fills 5,000 ÷ 200 = 25 candles. Jars: 25 × ₹70.80 = ₹1,770.00. Eco C1 wicks: 25 × ₹5.90 = ₹147.50. Total materials: ₹1,856.28 + ₹991.20 + ₹1,770.00 + ₹147.50 = ₹4,764.98, excluding packaging, labels, labour, shipping and GST.

Against that, three thermometers cost ₹1,062.00, roughly a fifth of one batch. The thermometers do not need to prevent many failures to pay for themselves.

The hold rule matters as much as the band. A batch made outside it is not automatically bad, but it is unproven. Label it, keep it apart, observe it on the same schedule as a test batch, and release it only if it passes. If it sweats, it goes to rework or scrap. It does not go out because a customer is waiting. A candle that sweats in the warehouse is not for dispatch, whatever the order date.

Training, review and when to change the band

Show staff why the band exists. Keep one sweated candle from the problem batches beside a dry one from the same recipe, and let people see and smell the difference. A reason is easier to follow than a rule.

Train on the probe, too. Show where to place it, how long to wait for the reading to settle, and why it should be stirred through the wax rather than resting against the hot pot wall. Two people using the same thermometer differently can produce readings several degrees apart, which quietly undoes the band you worked to validate.

Review the batch sheets monthly. Look for readings drifting towards one edge of the band, stir times getting shorter on busy days, or a station whose readings are always slightly different. Those patterns warn you before the candles do.

Change the band only with a new validation test. Seasonal changes in your workshop, a new melter, a new wax lot or a new oil are all reasons to recheck. When you change it, date the new sheet, retire the old one, and keep both on file so a future sweating batch can be traced to the method in force on that day.

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