How Do I Monitor Future Wax and Jar Lots so a Change in Glass Adhesion Is Detected Before Thousands of Candles Are Produced?

Three layers: identify every lot, check it before use, and chart every batch

An approved wax and jar are approved as they were on the day you tested them. Every new order is a new lot of wax and, often, a new delivery of glass, and nothing in CSI's listings tells you whether two lots are identical — CSI publishes no lot numbers, certificates of analysis or batch-consistency data for its waxes, and no glass specification for its jars. That is not unusual for small-pack raw materials, but it means the monitoring has to be yours. The goal is modest and achievable: if something about a new lot changes adhesion, you see it within a batch or two, not after a month of production.

The system has three layers. Identify every lot — write the order date, pack size and delivery note on each bag and each jar carton, because that is your lot number. Check each new order before it goes into production, with a small same-day pour against reference wax from the previous approved order. Chart the release check of every production batch, so that a slow drift or a sudden change shows up as a pattern rather than as a feeling. Each layer is cheap; together they turn "the candles seem patchier lately" into a dated, traceable signal.

The worked example is a maker producing a labelled example of 500 candles a month in CSI's Clear Glass Jar with Golden Lid (150 gram) — 12 g of Citrus Lemon and 138 g of Luxury Soy Wax CSI 464 each at 8%. That is about 71.1 kg of wax a month, bought as 7 × 10 kg + 2 × 1 kg (9 bags), with each 10 kg bag filling about 70 candles and each 5 kg bag about 35. CSI 464 is the approved wax only for the example; nothing here says it is more or less consistent than any other wax.

Layer one: identify and retain

Write the order date and pack size on every bag of wax as it arrives, and log it in a simple register with the jar delivery and oil bottles received the same week. When a bag is opened, write its order date on the batch sheet — the locked SOP already has a field for it. Then pour a small retained sample from the bag into a Mini 40 ml jar, label it with the order date, and keep it in your logged storage place. A retained jar is not an adhesion test; it is a physical reference. If a question arises months later — about colour, surface, smell or how the wax sets — you can put the retained jars from two orders side by side.

Why the Mini jar and not your production jar? Cost and space. At ₹21.24 a jar in the pack of 100, with a labelled example fill of 30 g (the Mini's fill is not published — weigh yours), a year of retained samples at 9 bags a month is 108 jars and about ₹1,643.98 of wax. But be clear about what it is: adhesion results in a 40 ml jar do not transfer to a 150 g jar. Adhesion is always checked in the production jar, in layer two.

Layer two: the lot check before a new order is used

Before a new order goes into production, pour a short side-by-side on the same day: 6 production jars from the new order and 3 from your reference wax — the last 1 kg bag of the previous approved order, kept sealed for exactly this purpose. Same oil bottle, same wick pack, same jar delivery, same cleaning, same racks, same pour temperature. CSI 464's page gives an 80–90°C melt/pour range and the Clear 150 g jar's page states hot pour up to 85°C, so pour at 80–85°C as your SOP says. Photograph at 48 hours and 7 days from four sides, wax hidden from the scorer.

What can six jars catch? If the new order pushed the true spotting rate to 20%, six jars show at least one spot 73.8% of the time; at 30%, 88.2%; at 10%, only 46.9%. So the lot check catches a big step change, which is the dangerous kind, and leaves smaller drifts to the release chart. That division of labour is deliberate: a lot check that could catch a 2% change would need hundreds of jars, which the acceptable-percentage post shows in full. The good news is that the lot-check candles are production candles — if they pass, they are sold.

Layer three: chart every batch

The release check from the SOP already photographs a sample of every batch. Plot its result — spotted jars out of 10 — on a simple chart, one point per batch, in the order poured, with the wax order date marked. Then set a trigger from your own pilot. If your pilot's baseline rate was a labelled example of 2%, seeing 2 or more spotted jars out of 10 would happen by chance only 1.6% of the time, so 2 or more is a signal worth acting on. Add one run rule: three batches in a row each with at least one spotted jar, which at a 2% baseline happens by chance only 0.612% of the time.

That last column is the answer to the title. If a new lot or a new jar delivery lifted the true rate to 20%, the chart would typically trip within 1.6 batches — about 64 candles — not thousands. A change to 10% is subtler and takes longer, about 3.8 batches on the trigger alone, which is why the run rule exists: it picks up a persistent small rise sooner than waiting for one bad batch. The thresholds are yours, set from your own baseline; neither the 2% nor the trigger is an industry figure.

When the chart trips, follow the evidence in order. Check the batch sheet first — most trips are process drift: a cooler last jar, a different rack, a new cleaner. Then compare the batch's wax order date with the lot-check result and the retained jar. Then check whether the jar delivery changed. Only when the process is clean and the trip lines up with a new order is it a lot question. Seasonal variation versus raw materials covers the case where the room changed with the season, which looks very like a lot change on a chart.

What monitoring costs, against what it protects

The standing cost is small. At the example volume of 500 candles a month, a year of retained Mini jars is 1 × pack of 100 + 2 × pack of 5 for ₹2,383.60; a 1 litre IPA spray is ₹520.00; and the reference-wax habit means one sealed 1 kg bag of CSI 464 (₹507.40) sitting on the shelf at any time — wax you use later, not wax you lose. The lot-check jars are production candles. The real cost is time: logging deliveries, pouring nine extra jars per order on a separate pour, and plotting one point per batch.

Against that, set what one bad order would cost if nobody noticed. At the example volume, a month's wax is 71.1 kg across 9 bags; each 10 kg bag becomes about 70 candles, each carrying the ₹94.40 jar, the oil, the wick and your labour. Costing cosmetic rejects and the 2,000-a-month hub show how quickly hidden or discounted candles add up. A system that catches a bad order on the first batch instead of the last protects most of that month.

The decision rule

Apply from the top. The rule puts the process before the lot, and the lot before the wax, because that is the order of likelihood and of cost. A lot that fails a repeat check against reference is a conversation with your supplier; a wax that keeps failing across several orders is a reason to reopen the scorecard.

The CSI products in the monitoring system

Where this fits in the qualification

This page owns monitoring after approval. Locking the process supplies the batch sheet and release check; the acceptance plan supplies the baseline; the storage hold explains retained jars over time; and the series capstone places monitoring at the end of the path from wet spots to bulk purchase.

Frequently asked questions