Short answer
A weighing tolerance is the largest difference you accept between the target weight of a line in a batch and what is actually weighed. A practical rule combines a percentage of the target with an absolute floor set by your balance — for example ±1% of the target or ±0.02 g, whichever is larger, on a 0.01 g balance — so large lines are held to a percentage and small lines are not held to a precision the balance cannot deliver. Tighten it for materials close to a restriction limit, bring tiny lines up to weighable amounts with dilutions, and record target, actual and pass or fail for every line.
Why a tolerance, and not "exact"
No weighing is exact. The balance reads in steps, the last drop off a pipette is unpredictable, and evaporating materials lose weight on the pan. Without a stated tolerance, two people will make different decisions about the same 0.3 g overshoot — one ignores it, one remakes the batch. A written tolerance turns that into a rule, makes batches comparable (repeatable perfume batches), and tells you when a deviation needs recording.
A rule with two parts
tolerance for a line = the larger of (r × target) and a
- r is a relative tolerance — for example 1% — that governs large lines.
- a is an absolute floor tied to the balance — for example 0.02 g on a balance reading to 0.01 g — below which you cannot reasonably hold a line.
The values are a judgment for your lab, not a standard. The floor should reflect the balance's readability and your technique; why small weights carry large relative errors is shown in how to weigh perfume ingredients.
| Line | Target (g) | Tolerance (g) | Acceptable range (g) | Actual (g) | Result |
|---|---|---|---|---|---|
| Material A | 25.00 | ±0.25 | 24.75–25.25 | 25.12 | Pass |
| Material B | 12.00 | ±0.12 | 11.88–12.12 | 12.15 | Fail (over) |
| Material C | 1.50 | ±0.02 | 1.48–1.52 | 1.51 | Pass |
| Material D (10% dilution) | 0.80 | ±0.02 | 0.78–0.82 | 0.79 | Pass |
For Material C, 1% of 1.50 g is 0.015 g, smaller than the 0.02 g floor, so the floor applies. Material B's failure is an overshoot; what to do about one is below.
What an error does to the percentages
Overshooting one line changes the whole formula, but less than people fear. Suppose a 100.00 g batch has Material A at 25.00 g, and 25.25 g goes in (+1%):
- New total: 100.25 g.
- Material A's share: 25.25 ÷ 100.25 = 25.19%, against 25.00% intended — up by 0.75% of its own value.
- Every other line's share falls by 0.25% of its own value (100 ÷ 100.25 = 0.9975).
So a 1% tolerance on each line usually changes each line's share by less than 1%. The question is whether that matters for the smell and for compliance.
Tighter tolerances where they matter
Restricted materials near a limit
If a material sits close to its IFRA limit, its tolerance must keep the worst case below the limit. An invented example: a concentrate contains 0.400% of a restricted material; the finished perfume is 20% concentrate, so the material is 0.0800% of the product, against an invented limit of 0.0850%. The headroom is 0.0050 ÷ 0.0800 = 6.25% of the line. A tolerance of 1% with a 0.02 g floor on a 100 g batch allows up to 0.42 g instead of 0.40 g, which gives 0.42 ÷ 100.02 × 20 = 0.0840% — still under 0.0850%. If the headroom were smaller, you would weigh that line as a dilution, make a bigger batch, or reformulate for more margin. The compliance calculation itself is in how to check IFRA limits.
Powerful materials
A trace material with a large odor effect — some sulfur compounds, strong aldehydes, natural absolutes — may need a tighter tolerance than its weight suggests. The fix is the same: weigh it as a dilution so the line becomes large enough to hold. See how to dilute aroma chemicals.
Under or over
- Under: add more, slowly, until inside the range. Record the final actual.
- Over, inside tolerance: accept and record.
- Over, outside tolerance: stop. Either scale up the other lines to restore the proportions — the method, and when it is impossible, are in added too much of an ingredient — or set the batch aside as a deviation. Never quietly continue.
Batch-level checks
- The sum of actual weights should match the batch's gross weight on the scale, within the sum of the tolerances. A larger gap means a line was missed, doubled or recorded wrongly.
- Compare the finished batch with a retained sample of the last good batch by smell, and by density if you measure it (measuring density).
Recording it
A batch record should carry, for each line: target, tolerance, actual, pass or fail, and initials; for any failure, what was done and why. Put the tolerance rule in the compounding procedure so it does not change from batch to batch (perfume lab SOPs), and the record template in perfume batch records. More batch arithmetic is in the perfume dilution and calculations guide.
Frequently asked questions
What tolerance do professional labs use?
It varies with balance, batch size and product; there is no single industry figure. Set yours from your balance's performance and your products' sensitivity, write it down, and tighten it where the table above shows it matters.
Should the tolerance be the same for trial and production batches?
The rule can be the same, but trial batches are small, so more lines hit the absolute floor. That is one reason trials use dilutions for small lines.
Is a line exactly on the limit of the range a pass?
Decide and write it down. Most procedures treat the limits as inclusive.
Written and reviewed by the RUŌOD Lab team. This article is general education about perfume formulation and record-keeping; it is not legal, regulatory or safety advice, and the examples are illustrations, not validated commercial formulas. How we write and check these guides.