A 40-vial batch of 2 mL certified vials came through our receiving dock with a certificate that listed the glass lot, the septum durometer, and the closure torque. It did not list the glass tubing heat number, and when a customer’s ISO 17025 audit asked us to trace one broken vial back to the tubing station, we could not close the chain. That is the failure this article is about: the specific link in a consumable’s traceability chain that most labs and most distributors cannot produce on demand. ISO 17025 traceability for lab vials is not a certificate you file. It is a chain of custody over materials, process parameters, and calibration events, and it breaks at predictable points.
I have built and audited that chain from both sides, as a distributor answering questionnaires and as a supplier writing them. This piece is about what an unbroken chain actually contains for a molded glass or plastic consumable, and where the weakest link usually sits.
ISO 17025 traceability, as it applies to a consumable rather than an instrument, runs through five handoffs. Each one has to be recorded at the time it happens. You cannot reconstruct a tubing heat number after the cullet is melted.
- Raw material lot. For borosilicate glass, the tubing lot and heat number from the drawing station. For resin, the lot number, the certificate of analysis, and the resin grade (for example USP Class VI polypropylene).
- Conversion parameters. Mold cavity, forming temperature, annealing schedule, and dimensional inspection results tied to that lot.
- Assembly and closure. Septum lot, durometer, coating (if any), and the torque applied at capping.
- Calibration of the measuring instruments. Every gauge, micrometer, torque tester, and leak tester used to generate the inspection data must itself be calibrated against a reference traceable to a national standard (NIST in the US).
- Release decision. The certificate of analysis or certificate of conformance that ties the inspection results to the shipped lot, with a signature and a date.
Four of those five links are almost always present in a competent supplier’s document pack. The one that goes missing at a rate I would put above 80 percent is the raw material lot, specifically the glass tubing heat number or the resin lot for the vial body. The certificate will tell you the septum lot. It will tell you the closure torque. It will not tell you which furnace drew the tubing.
Why the tubing heat number is the link that disappears
Glass tubing is drawn in continuous runs. A single heat can feed vial forming for days, and the tubing is stored in bundles that get consumed across multiple vial lots. The manufacturing execution system at the forming station records the vial lot and the machine. It often records the tubing size and the supplier. It rarely records the heat number, because the tubing is treated as a continuous input, not a discrete lot.
That gap matters because the properties that drift between heats are the ones that cause dimensional failure. Tubing ovality, wall thickness tolerance, and the residual stress profile after annealing all vary by heat. If a customer’s vial fails a dimensional check six months later, the heat number is the only way to know whether other vials from the same tubing are at risk. Without it, you quarantine the whole supplier, which is expensive and usually unnecessary.
What “traceable to a national standard” means for a torque value
Here is where people get confused. ISO 17025 traceability does not require the vial to be traceable to a national standard. It requires the measurements that characterize the vial to be traceable. That is a different claim, and it is the one auditors actually test.
Take closure torque. If the certificate states 1.2 N·m ± 0.1 N·m applied at capping, the auditor’s question is: what torque tester, and when was it last calibrated, and against what reference? A torque wrench calibrated by a lab that is itself accredited to ISO 17025 for torque, with a calibration certificate showing the reference weights or transducer traceable to NIST, closes that link. A torque value with no instrument behind it is a number, not a measurement.
The same logic applies to every dimensional callout on the drawing. A 12 x 32 mm vial with a 6 mm neck ID and a 1.0 mm wall is only traceable if the micrometer that measured it was calibrated, and the calibration is documented. The GPI thread finish (the 8-425, 13-425, 15-425, 18-400, 20-400, 22-400, and 24-400 designations) has its own dimensional tolerances, and the gauge that checks them has to be in the calibration system too.
The certificate of analysis is not the traceability record
Procurement teams often treat the CoA as the deliverable. It is not. The CoA is a summary. The traceability record is the set of underlying lot records and calibration certificates that let you reconstruct the CoA if challenged. Ask a supplier for the calibration certificate behind the torque value on their CoA and you will learn quickly which ones have a real system and which ones have a template.
Two field cases where the missing link surfaced
Case one: the dimensional drift that took three months to explain
A pharmaceutical customer running a VOA vial line for purge-and-trap water samples reported that roughly 2 percent of vials were failing the autosampler’s gripper check. The symptom was a vial that would not seat, which the operator logged as a “bad vial” and discarded. Over three months the failure rate crept from 0.5 percent to 2 percent, and the customer’s quality group opened a supplier corrective action.
We pulled the dimensional data. The neck OD on the affected lots measured 0.15 mm over nominal, which was still inside the drawing tolerance but at the edge. The real finding was in the tubing: the affected vials had been formed from a heat with a slightly thicker wall, and the forming process had not been adjusted to compensate. The measurement that identified it was a wall-thickness check at four points around the circumference, which showed an ovality of 0.08 mm against a 0.05 mm internal target. The fix was a tubing incoming inspection that records heat number and ovality, and a forming-set adjustment per heat. The traceability link that had been missing was the heat number, and once it was recorded, the correlation was obvious.
Case two: the septum that was not the septum on the certificate
A contract lab doing residual solvent testing under SW-846 methods had a run of failed system suitability checks. The internal standard response was drifting low by about 8 percent across a sequence. They suspected the instrument and spent two weeks on the inlet. The actual cause was a closure substitution: a purchasing change had moved to a lower-cost septa product with a different silicone formulation, and the certificate of analysis on file still referenced the original supplier’s lot.
The measurement that found it was a septum extractables screen: soak a sample of septa in the method solvent, inject, and compare the chromatogram to the historical baseline. The new septa showed an additional siloxane peak at the retention time that was co-eluting with the internal standard. The traceability failure was not the septum itself. It was that the certificate on file did not match the physical lot in the cabinet, which is the assembly-link gap in its purest form. The fix was a receiving inspection that compares the septum lot number on the bag to the certificate, every time, with a signature.
A decision checklist for auditing a vial supplier’s traceability chain
This is the checklist I use when a customer asks whether our chain is complete, and it is the same one I would apply to any supplier of lab vials. Score each line as present, partial, or absent. Two or more absences and the chain cannot be closed for a regulated application.
| Link | Record required | Concrete acceptance criterion |
|---|---|---|
| Raw material (glass) | Tubing heat number, supplier, drawing date | Heat number appears on the vial lot record, not just the tubing PO |
| Raw material (resin) | Resin lot, CoA, grade | Grade matches drawing (e.g. USP Class VI PP), lot matches CoA |
| Conversion | Mold cavity, forming temp, anneal schedule | Inspection data tied to the lot, with sample size stated |
| Closure assembly | Septum lot, durometer, torque applied | Torque within 1.2 ± 0.1 N·m, septum lot matches bag |
| Calibration | Calibration certs for every gauge used | Cert traceable to NIST, within calibration interval at time of use |
| Release | Signed CoA/CoC with date | Signature is a named person, date is before shipment |
Note what is not on this list: the certificate itself. The certificate is the output. The chain is the input. A supplier that can produce the heat number and the calibration certificate behind the torque value has a real system. A supplier that can only produce a signed PDF has a template.
Where the calibration interval quietly breaks the chain
Even a supplier with all five links documented can fail an audit on one technicality: the calibration interval. If a micrometer was last calibrated 13 months ago and the interval is 12 months, every measurement taken in that 13th month is out of the traceability system. The vial lot is still shippable, but the inspection data behind it is not defensible.
I have seen this trip up otherwise well-run operations. The fix is boring and effective: put every gauge used for release inspection on a calibration schedule with a 30-day warning, and record the calibration due date on the inspection record itself. Then the auditor can see at a glance that the gauge was in interval on the day of measurement. For a torque tester, the interval is usually 12 months. For micrometers and pin gauges, 12 months is common, though some labs use 6 months for high-use gauges. The number matters less than the discipline of recording it against each measurement.
The instrument side of the chain
There is a second traceability chain that touches the vial but is not the vial’s: the analytical instrument. When a customer runs a method on an Agilent GC or a Thermo Fisher HPLC, the instrument’s calibration and the column’s lot are part of the method’s traceability. The vial’s contribution is that it must not introduce a variable the method did not account for. That is why the septum lot matters: a closure change is a method change unless you have data showing it is not. The Restek technical library has published extractables data that is useful for setting that baseline, and I would build the comparison from published data before running my own screen.
How to close the chain when your supplier cannot
If you are a lab manager and your current supplier cannot produce the heat number, you have three options, and only one of them is good.
Option one: switch suppliers. This is the clean answer if the application is regulated and the supplier cannot close the chain. Ask for the heat number on a sample CoA before you place the order. A supplier with the system will send it in a day. A supplier without it will send a revised template and hope you do not notice.
Option two: build the link yourself at receiving. If switching is not practical, add a receiving inspection that records what the supplier does not. Photograph the bag label, record the septum lot, and run a dimensional spot check on a sample of vials per lot with your own calibrated gauge. That does not give you the heat number, but it gives you a defensible record of what you received, which is often enough to satisfy the auditor’s question about incoming goods.
Option three: accept the gap in writing. Document the limitation in your quality system, state the risk, and move on. This is legitimate for non-regulated applications. It is not legitimate for a method that will be audited under ISO 17025, because the auditor will ask and the answer will be “we do not have it.”
For suppliers, the fix is upstream. Record the tubing heat number at the forming station, even if it means a manual entry per bundle. The cost is a few seconds per lot. The benefit is that you can answer the audit question that your competitors cannot. If you are sourcing through a distributor, ask whether the distributor holds the lot records or passes them through. The Aijiren line, for example, ships with lot-level documentation that includes the tubing source, which is the difference between a compliant pack and a decorative one.
The decision rule
ISO 17025 traceability for lab vials closes when you can name, for any vial in the cabinet, the tubing or resin lot, the conversion parameters, the septum lot, the torque value with its calibration certificate, and the release signature, all tied to the same lot number and all dated before shipment. The link that breaks most often is the raw material lot, and the reason is structural: it is the only link that originates at a different facility from the one that molds the vial. Before you sign a supply agreement, ask for the heat number. If the answer is a template instead of a number, the chain is open, and no amount of filing will close it after the fact.
References & further reading
Related resources from our lab consumables network:
- Lab Vials at labvials.com
- Aijiren Brand at aijirenvials.com
- Voa Vial at voavial.com
- Septa at septacaps.com
External references and standards: