An auditor asked for the septum lot number on a 40 mL VOC vial and I could not produce it in under four minutes. That single delay turned a routine two-hour audit into a six-hour audit. The vial passed every physical check; the paperwork did not. What I learned that afternoon is that EPA method requirements for a gc vial are mostly about records, and the physical component is the easy part. The auditor’s first question was not “what vial do you use” but “show me the container tracking for the batch that produced this calibration.”

This article is about what SW-846 and the drinking water methods actually say about sample containers and handling, what auditors ask for first, and where the gap between a compliant vial and a defensible record opens up. It is not about why vials matter. It is about what happens when someone with a checklist and a clipboard sits down at your bench.

SW-846 chapter three sets the container rules, and almost nobody reads chapter three

SW-846 is a compendium, not a single method. The individual methods (8260, 8270, 8081, 8082, 8151) tell you the analyte list, the calibration model, and the QC acceptance criteria. The container requirements live in Chapter Three, “Inorganic Analytes,” and Chapter Four, “Organic Analytes,” along with the method-specific tables. When a method says “collect in a 40 mL VOA vial with a PTFE-lined septum and zero headspace,” that sentence comes from the method table, and the chapter explains how to apply it.

The relevant text is short and unforgiving. For volatile organics by 8260, the sample is collected in a 40 mL vial, preserved with HCl to pH less than 2, and there must be no headspace and no air bubbles. The method also states that if effervescence occurs during preservation, the sample is discarded and recollected. Those two sentences generate more audit findings than any other part of the method.

What “zero headspace” means in a physical part

Zero headspace is not a visual judgment call. It is a geometric outcome of the vial’s filled volume versus its rated volume. A standard 40 mL VOA vial holds roughly 43 to 44 mL when filled to the meniscus at the shoulder. If you fill to 40 mL, you have 3 to 4 mL of headspace, which is a 7 to 9 percent loss of the volatile fraction for compounds with Henry’s constants above 0.1. The auditor knows this. When they open a cooler and see a bubble the size of a pea at the top of a vial, they will note it, and they will ask you how many samples in the batch share that defect.

The physical fix is a vial with a rounded shoulder and a fill line at the shoulder, filled until a meniscus domes above the rim before the septum seats. The septum displaces the excess on closure. That is why the septum thickness matters: a 0.125 in (3.2 mm) silicone-backed PTFE septum displaces more volume than a 0.060 in (1.5 mm) one, and the fill target shifts accordingly. If you change septum thickness without changing fill practice, you either trap headspace or overfill and get septum bulging that the autosampler cannot pierce cleanly.

Preservation is a container property, not a bench step

HCl preservation to pH less than 2 assumes the vial is clean and the acid is added before the sample. The auditor’s check is not “did you add acid” but “was the acid in the vial before the sample arrived.” If you add acid after collection, the headspace has already equilibrated with the volatile fraction, and re-closure does not recover it. The record that proves this is the pre-preserved vial certificate, which states the acid volume, the concentration, and the lot. Without it, the auditor has no way to distinguish a compliant collection from a retrofitted one.

I keep pre-preserved vials in a separate cabinet with the certificate taped to the door. It is not elegant. It is fast to produce.

The auditor’s first five requests, in the order they arrive

I have been through this enough times to know the sequence. The auditor does not start with the instrument. They start with the paper that connects the sample to the container.

• Container tracking log for the sample batch, showing lot number, receipt date, and which vials went to which samples. This is the first request in nine out of ten audits I have sat through.
• Certificate of analysis for the vial lot, with the septum material, the liner thickness, and the test method used to verify it. A CoA without a test method is a brochure, and auditors know the difference.
• Preservation record showing acid addition, pH verification, and the analyst initials with a timestamp. For pre-preserved vials, the certificate substitutes for the bench record, but only if the lot matches.
• Holding time log from collection to extraction or analysis, with the method-specific limit (14 days for most 8260 analytes to analysis, 7 days to preservation verification in some state programs).
• Septum puncture count or a statement of single-use policy. If you re-use a vial, the auditor will ask how many injections it has seen, and “a few” is not an answer.

Notice what is not on that list: the instrument log, the calibration curve, the internal standard response. Those come later. The container paperwork is the gate. If it is missing, the auditor assumes the analytical data cannot be tied to a defensible sample, and the batch gets flagged regardless of how clean the chromatography looks.

Why the tracking log is the first ask

Because it is the only document that connects a physical object to a data point. A calibration curve proves the instrument responded linearly. It does not prove the vial that held the sample was the vial you think it was. The tracking log does. If you cannot produce it, the auditor has no chain of custody from container to result, and the entire batch becomes conditional.

The fix is a simple one: a bound logbook at the receiving bench, lot number written on every cooler receipt, and a line in the LIMS that forces the lot entry before the sample can be logged in. I have seen labs try to reconstruct this from purchase orders. It takes a day and it never fully reconciles.

Septum material and the method’s solvent list have to agree

SW-846 methods are solvent-heavy. 8260 uses methanol and purge-and-trap; 8270 uses dichloromethane and acetone; 8081 and 8082 use hexane. The septum has to survive the solvent, the temperature, and the puncture. A silicone septum with a PTFE face handles most of this. A straight silicone septum will swell in dichloromethane and lose its seal within a few hours. A straight PTFE-faced silicone with a thin face (under 0.002 in) will coring-fail after 30 to 50 punctures.

The number I use for planning is 50 punctures for a 0.125 in silicone/PTFE septum on a 23-gauge needle, and 100 punctures for the same septum on a 22-gauge needle with a side-port. Those are not method limits; they are the point at which coring starts showing up in the chromatogram as a ghost peak at the retention time of the previous injection. The method limit is one puncture per sample, because the vial is single-use. The auditor will ask you to confirm single-use. If you cannot, the puncture count becomes the evidence.

A comparison table for the three septum types you will actually see

The sandwich septum costs more per unit and pays for itself the first time an auditor asks why your 8270 batch has a ghost peak at the DCM retention time. I have made that switch twice in my career and never regretted it.

Two field cases where the paperwork, not the vial, was the finding

Case one: the lot number that did not match

Symptom: A state auditor pulled the tracking log for a 40 mL VOC batch and found that the lot number on the log did not match the lot number on the CoA in the binder. The vials were physically fine. The log was off by one digit, a transposition from a handwritten receipt.

Measurement that identified it: A side-by-side comparison of the log entry, the purchase order, and the CoA lot number. Three documents, two matching, one not. The auditor did not care which was right; they cared that the chain was broken.

Fix: We moved to a barcode-on-receipt system, with the lot number scanned into the LIMS at the receiving bench. The handwritten log stayed as a backup but stopped being the primary record. The next audit took 90 minutes instead of six hours. The vials did not change. The record did.

Case two: the headspace that was not visible until the cooler was opened

Symptom: A batch of 8260 samples arrived at the lab with no visible bubbles. The auditor opened three vials at random, inverted them, and found a bubble the size of a small pea in one. The vial was a 40 mL with a 0.060 in septum, filled to the shoulder but not domed.

Measurement that identified it: The auditor measured the headspace by inverting the vial and estimating the bubble diameter against the vial’s inner diameter. A 3 mm bubble in a 27 mm ID vial is roughly 0.5 percent headspace by volume, which is within some method tolerances but not within the “no bubbles” language of 8260. The lab’s fill procedure had been written for a 0.125 in septum and never updated after a switch to 0.060 in.

Fix: We rewrote the fill procedure to specify fill to the meniscus at the shoulder with the septum seated, then verified by inverting three vials per batch and checking for a bubble under 1 mm. We also standardized on the 0.125 in septum for all VOC work and kept the 0.060 in for the semivolatile methods where headspace is less critical. The auditor accepted the corrected procedure and the batch was released.

Holding time, temperature, and the cooler log the auditor will ask for second

After the tracking log and the CoA, the auditor asks for the cooler temperature record. SW-846 methods specify 4 °C ± 2 °C for most volatile and semivolatile samples, and the record has to show the temperature at receipt, not just at some point during transport. A data logger that samples every 15 minutes is the standard. A handwritten log with one entry per cooler is not.

The number that trips people up is the receipt temperature. If the cooler arrives at 8 °C and the lab logs it as 4 °C because the ice was still cold, the auditor will catch it on the logger trace. I have seen a lab lose a batch over a 6 °C receipt temperature that was within the logger’s tolerance but outside the method’s. The fix is a calibration check on the logger every six months and a receipt procedure that logs the actual reading, not the expected one.

The vial itself has a role here. A polypropylene vial with a polyethylene cap will survive a freeze-thaw cycle; a glass vial with a crimp seal will not, because the crimp loosens as the glass and the aluminum contract at different rates. For methods that require freezing (some state programs for 8260), the container has to be rated for it. The auditor may not ask, but the sample integrity depends on it.

What a compliant vial lot looks like on paper

The CoA is the document the auditor reads most closely, and it is the one most manufacturers write loosely. A defensible CoA for a gc vial lot has, at minimum:

• Lot number that matches the tracking log and the purchase order.
• Vial dimensions verified against the drawing: 12 x 32 mm for a 2 mL autosampler vial, 27 x 95 mm for a 40 mL VOA vial, with tolerances of ±0.2 mm on the OD and ±0.5 mm on the height.
• Septum material and thickness, with the durometer (typically 50 to 60 Shore A for silicone) and the PTFE face thickness.
• Test method for each specification. “Visual inspection” is not a test method. “Dimensional check with calibrated calipers, n=30, RSD under 0.5 percent” is.
• Background check for the analytes of interest, typically a GC-MS scan of a solvent blank stored in the vial for 24 hours at 4 °C, with the detection limit stated.
• Closure torque for crimp vials, typically 1.5 to 2.0 in-lb for an 11 mm aluminum seal, verified with a torque tester.

If any of those six items is missing, the CoA is a marketing document. I have rejected three lots in five years on the background check alone, and each rejection saved a batch of analytical data that would have been flagged in the next audit. The gc vial itself is a commodity; the CoA is the part that has to be specific.

Where the related consumables enter the audit

The auditor will sometimes follow the sample into the autosampler and ask about the HPLC autosampler vial used for the confirmation run, especially if the method has a dual-column confirmation step. The same tracking rules apply. The sample vial used for the standard prep has to be traceable to a lot, even if it never holds a sample. And if the method uses a preserved aqueous standard, the reagent bottle holding the acid or the preservative is part of the record, because the acid lot number goes on the preservation log.

I keep the four consumable families on one tracking sheet. It is one page, it takes five minutes to fill out per batch, and it has never failed an audit.

The autosampler is where the method meets the physical part

SW-846 does not specify the autosampler, but the method’s QC criteria (internal standard response within 50 to 200 percent of the calibration, surrogate recovery within the method window) are measured through the autosampler. If the needle cores the septum, the internal standard response drops and the batch fails QC for a reason that has nothing to do with the sample. The auditor will see the failed QC and ask why. The answer, if you have the puncture count, is a septum that hit its limit.

The physical constraint is the needle gauge and the septum thickness. A 23-gauge needle (0.025 in OD) through a 0.125 in septum is the standard configuration. A 22-gauge needle (0.028 in OD) is faster to pierce but cores more. A 26-gauge needle (0.018 in OD) is gentler but bends on a misaligned vial. The autosampler’s vial holder tolerance is typically ±0.5 mm, and a vial with an OD outside 12.0 ± 0.2 mm will sit off-center and the needle will hit the septum edge. That is a coring failure waiting to happen, and it shows up as a ghost peak at the previous injection’s retention time.

The fix is a dimensional check on incoming lots, 30 vials per lot, calipers on the OD and the height, RSD under 0.5 percent. I have rejected a lot at 12.4 mm OD, which is outside the ±0.2 mm window, and the supplier’s response was that the tolerance was “nominal.” That is the difference between a vial and a compliant vial.

A troubleshooting checklist for the week before an audit

This is the list I run through. It takes about four hours and it has caught a finding every time.

• Pull the tracking log for the last three months. Verify that every lot number matches a CoA in the binder. Flag any transposition.
• Check the CoA against the six-item list above. Any lot missing a test method goes on the reject list.
• Verify the septum material against the method’s solvent list. DCM methods get the sandwich septum; everything else gets the silicone/PTFE.
• Measure 30 vials from the current lot. OD 12.0 ± 0.2 mm, height 32.0 ± 0.5 mm for a 2 mL vial. RSD under 0.5 percent.
• Check the crimp torque on five sealed vials. 1.5 to 2.0 in-lb for an 11 mm aluminum seal. A loose seal is a leak path the auditor will find.
• Confirm the cooler logger calibration is current, and that the receipt temperature is logged as measured, not as expected.
• Walk the holding time log for any batch older than 10 days. Anything past the method limit goes on the exception report before the auditor asks.
• Check the septum puncture policy. Single-use is the method default; if you re-use, the count has to be documented.

For the analytical side, the Agilent GC columns and accessories documentation and the Thermo Fisher chromatography and sample handling resources both publish the autosampler vial dimensions and needle specifications that the method assumes. The Restek technical resources have a useful septum selection guide that maps material to solvent class. The U.S. EPA SW-846 test methods page is where the chapter text lives, and it is worth reading Chapter Three once a year whether or not you have an audit coming.

The decision rule: build the record before the vial

If you are choosing a gc vial for an EPA method, the decision is not which vial to buy. It is whether the lot you buy comes with a CoA that names a test method, a tracking log that ties the lot to the sample, and a septum that matches the method’s solvent list. A vial that meets all three is compliant. A vial that meets only the first is a physical part with a paperwork gap, and the gap is what the auditor finds. Buy the record, then buy the vial, and keep the tracking sheet to one page so the four-minute answer stays four minutes.

References & further reading

Related resources from our lab consumables network:

• Gc Vial at gcvialfactory.com
• Reagent Bottle at reagentbottleforsale.com
• Hplc Autosampler Vial at hplcautosamplervials.com
• Sample Vial at samplevialprice.com

External references and standards:

• Agilent chromatography consumables and columns
• Thermo Fisher chromatography and sample handling
• U.S. EPA SW-846 test methods for waste and materials
• Restek technical resources for chromatography