Batch Sampling Plan: How a Few Vials Represent a Whole Lot
Every purity figure on a certificate describes a handful of vials, not the whole batch. Testing is destructive, so nobody analyses every unit; instead a batch sampling plan decides which units are pulled and how the results are allowed to speak for the rest. For a lab that receives research peptides by the dozen or by the hundred, understanding that plan explains what a supplier certificate really covers, and it gives you a sensible way to design your own incoming checks without testing half your order.
From a few vials to a whole batch
Reporting a sample’s result against a batch is an inference, and it rests on one assumption: that the batch is uniform enough for any unit to stand in for any other. When that holds, a single well-chosen vial is informative. When it does not, the result describes the vial that was tested and very little else.
Uniformity comes from how the material was made, not from the test. Peptide solution that was purified, pooled, filled and freeze-dried in one continuous run is a good candidate for a single population. Material dried in several freeze-dryer loads, or combined from separate syntheses, is less so, and a good sampling plan takes that history into account. Our article on what a batch is covers where those boundaries fall.
Four ways to choose the sample
| Approach | How units are picked | What it guards against |
|---|---|---|
| Simple random | Every unit has an equal chance, usually via a random number list | Unconscious bias in which vials get chosen |
| Systematic | Every nth unit from a filling or packing sequence | Uneven coverage across a long run, provided the run has no repeating pattern |
| Stratified | A set number from each defined subgroup, such as start, middle and end of filling, or each freeze-dryer load | A known source of variation being missed by chance |
| Convenience | Whatever is nearest or easiest | Nothing; it is the weakest design, though very common |
Stratified sampling is the one worth understanding for peptides. If you suspect residual moisture differs between drying loads, or that the first vials filled differ from the last, sampling each stratum is the only way to find out. Random sampling might happen to miss the load that matters.
How many units to sample
There is no universal number. Two conventions show up often:
- The square-root-plus-one rule. Sample the square root of the number of containers, plus one. For 400 containers that means 21. It scales sensibly and is easy to apply, but nobody derived it from probability theory, and it promises no particular level of confidence.
- Acceptance sampling tables. Standards such as ISO 2859-1 give sample sizes and accept or reject numbers for a chosen acceptable quality level. They are built for pass/fail attributes such as seal defects or labelling errors, and they suit inspection of packaging and presentation better than analytical purity.
For analytical testing of research peptides, practical plans tend to be smaller and targeted: enough vials to check that the batch behaves as one population, chosen to cover the most likely sources of variation.
Composite or individual samples
A lab can combine material from several vials into one composite and analyse it once, or analyse each vial separately. The choice changes what the result means.
- A composite gives a good estimate of the batch average at the cost of a single analysis, but it hides any difference between vials.
- Individual results cost more but show the spread. If three vials report 98.4, 98.6 and 96.9 percent, the batch has a problem that a composite averaging to 98.0 would have concealed.
When the question is “is this batch consistent?”, individual analyses are the ones that answer it.
Why sampling usually dominates the uncertainty
A validated HPLC purity method is typically precise to a few tenths of a percent. If a batch is even slightly non-uniform, the differences from one vial to the next can easily be larger. Tightening the method while testing one convenient vial therefore polishes the precision of a number whose weakest link is how the vial was chosen. Our article on measurement uncertainty in peptide purity covers the analytical side; the sampling side is often the larger contributor.
A batch sampling plan for incoming bulk orders
When a large multi-vial shipment arrives, a proportionate incoming plan might look like this:
- Group the vials by compound and by cap and crimp colour, since that pairing is what ties each group to its certificate.
- Inspect every vial visually for seal integrity, cake appearance and labelling. This is cheap and catches handling damage.
- For each group you intend to verify analytically, select vials at random rather than from the top of the box, and record how they were chosen.
- Where a group is large, take vials from different positions in the packing so the sample spans the shipment.
- Analyse selected vials individually if consistency matters to your work, and keep at least one sealed vial per group as a retained sample.
- Log the results against the group, the certificate and your custody record.
Who drew the sample
A testing lab reports on what it received. Who picked those vials, and how, usually sits outside the report. That is a normal boundary rather than a flaw, and it is worth knowing when you read any certificate: the analysis may be excellent while the question of representation rests on the sampling behind it. Documenting that step on your own side is covered in our article on sample chain of custody.
Bulk Peptides has its products third-party tested for HPLC purity, posts certificates for some products on the certificates page, and matches each vial to its certificate by cap and crimp colour so your own sampling can be organised around those groups.
This guide concerns analytical sampling of research materials. Products referred to are for in-vitro laboratory research only and are not for human or animal use.
