If Actual Fragrance Consumption Is Higher Than Formula Consumption, What Production Practices Should I Audit?

The order and the overrun, as numbers

The title gives no figures, so this is a labelled worked example. A corporate client wants 1,200 candles for Diwali gifting (32 days from 7 October 2026 to 8 November 2026), delivered by the end of October. The approved sample is a 150 g candle in the Clear Glass Jar with Golden Lid (150 gram) at 8% Cinnamon Vanilla on the total-weight basis: 12 g of oil and 138 g of Luxury Soy Wax CSI 464. Production runs in 9 kg melts (720 g oil + 8,280 g wax, 60 pours), about twelve melts a week, so the order is two weeks of pouring.

After week 1 the stock count tells the story. 12 melts poured 720 candles, 690 of them good and 30 rejected. The standard for 690 good candles is 690 × 12.6 g = 8,694 g of oil. The oil stock went down by 9,454 g, outside the 10 retains. That is 760 g over standard, 8.7%, or 13.70 g per good candle against 12.6. At 4.55 a gram it is ₹3,461.65 of oil in one week, small next to the order, but large next to the stock: at that rate the oil runs out before the last candle.

The seven-practice audit list

Each line below says what goes wrong, how to check it on the day, what the worked audit found, and where the oil ended up. The last column is the one that decides how urgent the fix is.

The order matters. Scale first, because every other measurement depends on it. Then watch before you ask, because people change how they weigh as soon as they know they are being watched, and the first five weigh-outs are the honest ones. A daily per-batch log, which turns this one-day audit into a standing check, is the post on tracking consumption after every batch.

The three that change the candle

Topping up "for strength"

The commonest and the most human. A pot of hot wax with oil in it does not smell like a cured candle, and someone adds a splash to be safe. In the worked audit it happened on two melts, 60 g each: 120 g. A topped-up melt is at 8.61% instead of 8%. That is inside CSI 464's 10% ceiling, so the candles are not unsafe on that count, but they are not the approved formula, and their hot throw, wick behaviour and colour may differ from the sample. Hot-wax smell is not a test of a candle's throw; cure and a burn test are. The fix is a rule that nobody adds oil except to the weighed charge on the sheet, and an answer for the worried pourer: "the sample was approved after cure".

Load-basis slips

8% of the total candle weight is 8.70% of the wax weight. Both figures are right for their own base, and either can end up on the wrong one. In the worked audit a helper had written "8.70%" from an earlier conversion and applied it to the whole 9 kg mix: 783 g instead of 720 g, 63 g too much per melt, 126 g over two melts, and a load of 8.64%. The fix is a charge table on the wall in grams per melt size, so nobody converts percentages at the melter. The basis arithmetic itself is in the post on whether 8% is calculated from the wax or the total.

Re-dosing remelted rejects

A reject dug out of its jar is wax and oil. Fourteen rejects were remelted in their own pot with 12 g of fresh oil each "to refresh them": 168 g of new oil on top of the 168 g already in them. That pot was at 14.8%, above CSI 464's stated 10% maximum (and the lower of the two ceilings governs, even though Cinnamon Vanilla's page says "Up to 13% load for candles"). The 15 candles poured from it are not shipped. Rejects should be remelted into a normal melt in small proportion, or set aside, never re-dosed.

Commercial protection (practical judgement, not legal advice): the client approved a candle, not a recipe. If more than a handful of candles end up off-formula, tell the client before delivery rather than after, and offer a remake or a split delivery; a written order that names the approved sample and an approval process for changes makes that conversation simple. Keep the retained sample from the approval; it is your reference.

The four that are measurement and handling

Pour-over and tare habits

Pouring a viscous oil to an exact figure is hard, and most people overshoot and leave it. Six grams a melt is 72 g a week. Weighing in a jug that was tared while wet, or not re-tared between pours, adds more. The fix is to weigh by difference: put the bottle on the scale, tare, pour into the pot until the display reads −720 g, stop. The oil goes straight into the wax, and the jug, its film and its tare problems disappear.

A scale reading low

If the scale reads 1% low, every charge is 1% heavier than the sheet says: 86 g a week here, and the log looks perfect because the recorded figure is the target. Only the stock count catches it. A known weight every morning and a second check at midday close it.

Oil left in bottles and jugs

Thick, sweet oils such as this one cling. An "empty" 1 kg bottle in the worked audit still held about 8 g (80 g over ten bottles), and the weighing jug kept about 5 g a melt (60 g). Drain bottles upside down into the next bottle of the same lot overnight, and weigh by difference so there is no jug. Do not mix residues from different lots into a bottle you will use for the order without noting it.

Heat and hold time

The practice most often blamed, and the hardest to measure. Adding oil very hot or holding the pot hot for a long time is commonly said to drive off lighter notes, and it may affect how the scent comes out. Whether it measurably reduces the weight of oil in your pot is a question for the scale, not for a rule of thumb: in the worked audit, a 9 kg pot weighed before and after a 30-minute hold showed no difference the scale could see. So heat was not where the 760 g went. Control it anyway, for quality: keep oil addition and pouring at the wax page's figures (CSI 464 states 80–90°C; in the 150 g jar, rated "Safe for hot pour up to 85°C", pour at 80–85°C), and check the thermometer itself. What temperature does to scent is covered in the post on adding fragrance to very hot wax.

The worked check and what it means for the rest of the order

The decision rule, in order: candles touched by a load-changing practice are held (240 from four melts, 15 from the reject pot). The reject-pot candles are above CSI 464's 10% and are not shipped. The others are burn-checked against a retained approved sample after a cure; many makers allow at least 48–72 hours for soy before judging a burn (common practice; neither the Cinnamon Vanilla nor the CSI 464 page states a cure time in the text we captured). That is three to four calendar days, so it runs alongside week 2's production, not after it, and the end-of-October delivery still holds. Handling fixes start the same day. Re-audit after a week: if the unexplained part stays under about 1% of the standard (here 0.6%), stop auditing and keep the daily scale check.

Fixing the practices saves ₹2,737.60 on the top-up, but that is not the main win. The main win is that the second 510 candles match the first good ones and the approved sample, and that the order does not run short in the last week before Diwali. How to calculate a remaining requirement and when to order it is the post on running short mid-production; tracking each material's variance week by week so the next overrun is caught in days is the post on tracking variance separately for each material.

Argued against our own interest: do not answer an oil overrun by buying more oil "to be safe" before the audit. In the worked order that would mean a kilo or more of a client's Diwali scent, bought to feed habits that change the candle. Audit first, fix, and buy the 750 g the corrected rate needs. And a top-up bottle is probably a new lot: compare it with the old lot before it goes in, as in the post on a new lot of the same fragrance mid-run.

The tools and materials behind the audit

Chosen on page statements, price and stock on 7 October 2026. The scale and thermometer are the audit's instruments; the oil and wax are the worked order's, and their stated limits are what make three of the practices urgent.

The honest summary: the audit costs a production day and needs a scale you have proved and a thermometer you have checked. It does not need new materials, and the biggest finding is usually a habit, not a leak.

Frequently asked questions