How Can Beginning-Middle-End Retained Samples Help Identify Fragrance Distribution, Temperature Drift and Filling Problems in Large Batches?

Three faults, three fingerprints

A large batch can go wrong in many ways, but three account for most of the differences between the first candle and the last: the fragrance was not evenly distributed, the wax temperature drifted during the pour, or the fill quantity drifted. Each of these changes a different property of the finished candle, which is why beginning, middle and end retained candles can tell them apart.

Uneven fragrance shows up in throw. Temperature drift shows up in surface, adhesion and sometimes colour. Filling drift shows up on the scale. The skill is to measure all of these on each retain, in an order that does not destroy evidence, and then look for which measurement changed across the run.

The measurements are not independent, and that is where careful reading matters. A lighter candle throws a little less simply because it contains less scented wax. Cooler wax can pour less evenly and fill differently. So you read the whole set before naming a cause, and you let the measurement that changed most lead the diagnosis.

Read them in the right order

A retained candle can only be burned once, so do the non-destructive measurements first and write each one down before moving on.

Worked example: 20 kg of Juicy Orange in coconut soy

A maker poured 100 candles of 200 g from 20 kg of scented wax at 9% of total weight: 18.2 kg of CSI 468 and 1.8 kg of Juicy Orange. The product page gives a stated load range of 8–10% in soy. Wax at ₹637.00 per kg (the ₹6,370.00 10 kg pack) is 18.2 × ₹637.00 = ₹11,593.40. Oil at ₹7,499.00 per kg is 1.8 × ₹7,499.00 = ₹13,498.20. The batch holds ₹25,091.60 of wax and oil.

Before looking at any numbers, note what the batch record already says: one melt, oil added at a single time, one operator, a pour that ran for the length of the melt with no refill. That rules out a step change from an event and leaves slopes — time-dependent drift — as the likelier explanation for anything the retains show.

Retains were kept at candles 3, 50 and 98. Suppose the readings came out like this:

Fill and surface moved clearly at the end; throw moved only slightly, and roughly in line with the lighter fill. That reads as temperature and filling drift, not a fragrance distribution fault. CSI 468's stated pour is 58–60°C; if the temperature log shows the last pitcher loads went in below that, the story is consistent: wax cooled, thickened, poured short and set rough.

Why each fault is more likely at scale

Fragrance distribution. At 1 kg, a few turns of a stirring stick move the whole mass. At 20 kg, the same effort may leave oil richer near the surface or the point where it was poured in, and the difference can persist if the wax stands before pouring. The signature is throw varying by position with fill and surface steady, and it is the one fault where the retains' throw ranks lead the diagnosis.

Temperature drift. A large mass in the melter holds its heat, but each pitcher load leaves that mass and starts cooling immediately. If the melter is not brought back to the validated window between loads, or the pitcher sits while jars are arranged, later pours can go in noticeably cooler. The retains show it on the surface long before it shows in scent.

Filling drift. Operators pour by eye and by feel. As the run goes on and the wax thickens, the same wrist movement delivers less, and the last pitcher loads often come up short. The scale catches it immediately, which is why fill weight is the first measurement, not an afterthought.

None of these needs new equipment to diagnose. A scale, a thermometer and three retained candles are enough to tell them apart; any spending on mixers or larger melters should wait until the retains show that the process, not the pour technique, is the limit.

When the faults overlap

Real batches often show two signatures at once. A long pour can both cool the wax and hold the fragrance at heat for longer before the last candles are filled. The result is a lighter, rougher end candle that is also slightly flatter in character.

To separate them, normalise throw for fill. If the end candle is about 6.5% lighter, as candle 98 is (13 g ÷ 199 g), and its throw is only a little weaker, fill may explain most of the difference. If it is 2% lighter and clearly flatter, something else happened to the fragrance. This is judgement, not arithmetic, but writing the fill difference next to the throw rank keeps you honest.

Colour can help. Juicy Orange's page notes low discolouration risk, so a darker end candle would suggest heat history rather than the oil's own tendency. With a vanillin-bearing oil, colour change by position is less conclusive, because the oil itself warms in tone over time.

Confirm, then act on the batch and the process

Three retains detect drift; they do not prove its cause. Confirm on a small control: pour two 500 g candles from the same materials, one at the stated pour temperature and one deliberately cooler, and see whether the cooler one reproduces the end retain's surface and fill. If it does, you have your cause.

For the batch, the readings guide release. Candles near the beginning and middle can be released once burn tests pass; candles near the end may need checking individually for fill and finish before they ship. For the process, the fix follows the fault: shorter pour spans per melt, reheating to the validated window between pitcher loads, weighing fills at intervals, or scenting smaller portions in sequence. Keep reading every batch's retains the same way, and the next drift will show up as a small number on a sheet rather than a complaint.

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