Should I Use Calibration Weights to Check Production Scales Before Making Large Candle Batches?

What a calibration weight proves, and what it does not

Many workshops already run a quick pre-run check with a check object: an item weighed once and kept for comparison. That check is excellent at catching drift. Its weakness is that it only compares today's scale with an earlier reading of the same object. If the scale was wrong on the day you recorded the object, the check object inherits the error and every daily check passes.

A calibration weight breaks that loop. It is a mass made to a stated tolerance and supplied with a certificate, so its value does not depend on any of your scales. When you place it on a production scale, the difference between the display and the certified mass is the scale's error at that load. That is the evidence you need before trusting a scale with a 25 kg batch.

What a weight does not do is fix anything. If the reading is outside your limit, adjustment follows the scale maker's procedure in its manual, or a service technician. Some inexpensive scales have no user adjustment at all, in which case the weight's job is to tell you to retire or reassign the scale.

Worked example: why two scales matter more than one

Most production benches weigh wax on a large platform scale and fragrance on a smaller, finer bench scale. That arrangement is sensible, but it creates a risk that a single-scale kitchen setup never had: the two scales can disagree.

Suppose a 25 kg batch is planned at 8% of total weight: 23 kg of Luxury Soy Wax Chunks and 2 kg of White Royal Oud. The platform scale under-reads by about 1.5%, so 23,345 g of wax shows as 23,000 g. The oil scale is accurate and delivers a true 2,000 g.

The middle row is the surprise. If a single scale weighs both materials and its error is proportional, the ratio survives: both quantities are 1.5% heavier, so the load stays 8%. The batch is simply bigger than planned, which affects yield and cost but not scent.

The bottom row is the real problem. With the wax scale low and the oil scale right, the load is 2,000 ÷ 25,345 = about 7.89%. Every batch weighed this way is under-scented by the same margin, and a daily check object on each scale would never reveal it, because each scale agrees with its own history.

Offset errors behave differently

Not every scale error is proportional. A scale whose zero has drifted adds or subtracts the same few grams on every reading, whatever the load. On a 23 kg wax weighing, a 5 g offset is irrelevant. On a 16 g fragrance weighing for a single test candle, it is a third of a percentage point of load. That is why the same scale can look perfect on production batches and misleading on small development work, and why verifying at two loads matters: an offset shows the same gram error at both weights, while a proportional error shows a larger gram error at the heavier one.

Knowing which kind you have also tells you what to do next. An offset is often cured by a clean platform, a level bench and a proper zero. A proportional error usually needs the scale maker's adjustment procedure, or a decision to retire the scale from critical weighing.

The cost shows up in two places. The extra 345 g of wax per batch, at ₹5,999.00 per 10 kg (₹0.5999 per gram), is 345 × ₹0.5999 = ₹206.97 of wax per batch that the sheet does not record. And if someone responds to the slightly weaker candles by raising the oil charge, White Royal Oud at ₹8,103.00 per kg (₹8.103 per gram) means every extra 20 g costs 20 × ₹8.103 = ₹162.06 per batch to fix a problem that was never in the fragrance.

Choosing weights: what to look for

CSI does not sell calibration weights, so this is a list of what to ask for from whoever you buy them from, not a product recommendation.

For heavy loads, you do not need a single 20 kg weight. Several smaller certified weights placed together, or a known weight combined with a stable substitute load, are common approaches. Follow the scale maker's guidance on stacking loads on the platform.

A verification schedule that fits a candle workshop

The intervals below are suggested starting points to validate. Tighten them if checks start to fail; relax them only after a long run of passes.

Write each verification result on the scale log with the certificate number of the weight used. After you verify both production scales, re-weigh your check objects on them and update the recorded values. From then on, your daily checks are anchored to a certified reference again.

If verification finds a scale out of limit

Treat it as a finding that reaches back in time. Every batch weighed since the last passing verification is now uncertain.

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