A 40 mL VOA vial that passed every QC check in the lab failed an EPA 8260 audit because the certificate of analysis on file listed the closure as “silicone/PTFE septum” with no lot number, no cure date, and no statement of the septa’s volatile background. The auditor’s finding was not about the vial. It was about the missing paper trail connecting that specific lot of closures to that specific batch of samples. That is the pattern I see repeatedly: the physical consumable is fine, the EPA method requirements for HPLC and GC consumables are met in practice, but the documentation that proves it is not there when the auditor asks. This article is about what SW-846 and the method-specific appendices actually require of your consumables, and the exact paperwork an auditor reaches for first.
SW-846 does not specify a vial. It specifies a container, and the difference matters.
The EPA’s SW-846 test methods for waste and materials are performance-based. That is the sentence auditors keep coming back to. SW-846 Chapter Three (Inorganic Analytes) and Chapter Four (Organic Analytes) describe sample container materials, preservation, and holding times, but they deliberately do not name a part number. Method 5030 (purge and purge-and-trap) and Method 5035 (closed-system purge-and-trap for soils) describe the container as “a glass vial with a septum seal” and leave the septum material, thickness, and closure torque to the analyst. Method 8260 and Method 8270 add the requirement that the container not contribute to the analyte signal.
That last clause is where consumables procurement lives or dies. “Not contribute to the analyte signal” is a performance specification, and it is the one you have to prove you meet. For volatile organics, a silicone septum off-gassing cyclic siloxanes at 60 °C will put D4 and D5 into the blank at concentrations that look like contamination. For semivolatiles, a plastic cap liner with a phthalate plasticizer will put DEHP into every blank. The method does not tell you which septum to buy. It tells you the blank has to be clean, and the burden of showing that is on your documentation.
What the method text actually pins down
Read the container sections closely and you find a short list of things SW-846 does commit to:
- Material compatibility. Glass for volatiles. Amber glass for light-sensitive analytes. The method states that plastic containers are unacceptable for volatile organics because of permeation and adsorption.
- Headspace. Zero headspace for volatile methods. Method 5030 requires the vial be filled so that no air bubble larger than a few millimeters remains. A 40 mL vial filled to the shoulder with a 20 mm crimp closure is the standard geometry.
- Preservation. HCl to pH <2 for most volatiles, and the pH must be checked and documented. The consumable that touches the preservative (the septum, the liner) must be chemically inert to it.
- Holding time. 14 days for most volatiles from collection to analysis, 7 days for some. The clock starts at collection, and the closure has to hold the seal for that entire window without evaporative loss.
Everything else, including septum hardness, closure torque, and lot traceability, is left to you. That is not a loophole. It is a documentation obligation.
The first three documents an auditor asks for, in order
I have sat through enough EPA and state-primacy audits to know the sequence. The auditor does not start with your instrument. They start with the paper that ties the consumable to the result.
1. The certificate of analysis with lot-level traceability
The first request is the CoA for the septa and closures used on the batch under review. Not the catalog page. Not the spec sheet. The lot-specific CoA with a manufacturing date, a cure date for the silicone, and a statement of the septum’s volatile background. If your supplier issues a generic CoA that covers “all lots,” you have a finding waiting to happen. The auditor wants to see that the lot number on the CoA matches the lot number on the box that matches the lot number recorded in your run log. Three numbers, one chain.
For septa, the CoA should state the base polymer (silicone, EPDM, or butyl), the durometer, the PTFE facing thickness (typically 0.003 to 0.005 in for a bonded liner), and the extractables profile. If the supplier cannot give you a cured-date and a lot number, that is your first procurement red flag.
2. The blank and background data
The second request is the method blank from the batch, and the auditor will compare it against the septum lot. If your method blank shows siloxanes at 5 ppb and the previous batch showed 0.5 ppb, the auditor will ask what changed. The answer needs to be a consumable lot change with a CoA and a background test to back it up. This is why I insist on a septum background test at receipt for any new lot: run a 40 mL vial with reagent water through the full purge-and-trap sequence and compare the siloxane and phthalate peaks against the prior lot. Keep that chromatogram with the CoA.
3. The preservation and holding-time record
The third request is the pH check and the collection-to-analysis timestamps. The consumable link here is the closure’s ability to hold the seal. A crimp cap that was under-torqued will show evaporative loss over 14 days, and the auditor will ask whether the loss is preservation failure or closure failure. If your closure torque is not documented, you cannot answer. This is where the physical spec becomes an audit item.
Septum background is the failure mode that shows up as a data quality finding
The most common consumable-driven audit finding I have seen is not a leak and not a puncture. It is septum bleed. A silicone septum with a thin or poorly bonded PTFE facing will off-gas at the temperatures used in headspace and purge-and-trap. The siloxane peaks land in the same retention window as target analytes and push the method blank above the reporting limit.
The fix is a specification, not a brand. For volatile methods at 40 to 60 °C, I specify a silicone septum with a bonded PTFE facing of at least 0.005 in, a durometer of 45 to 55 Shore A, and a cure that is fully post-cured (not just molded). The PTFE facing is the barrier. A thin or mechanically bonded facing will delaminate at the puncture site after 50 to 100 injections, and once it delaminates the silicone is exposed to the sample and the headspace. That is the moment the blank goes bad.
Field case: the siloxane blank that tracked a lot change
A lab running Method 8260 had method blanks creeping from 0.4 ppb to 6 ppb for octamethylcyclotetrasiloxane (D4) over two weeks. The instrument was clean, the trap was new, and the reagent water was verified. The measurement that identified it was a lot-to-lot comparison: the new septum lot, run against reagent water, showed D4 at 8 ppb versus 0.3 ppb for the prior lot. The supplier had switched from a post-cured silicone to a faster-curing formulation to meet a lead time. The fix was a purchase specification requiring post-cure and a receipt test for every lot. The finding was closed by adding the receipt chromatogram to the CoA file.
Closure torque is a documented number, not a feel
Auditors increasingly ask for the torque value used to seal crimp and screw closures, because torque controls the seal and the seal controls preservation. For a 20 mm crimp vial, the working range is 0.5 to 1.0 in-lb (about 5.6 to 11.3 N·cm). Below 0.5 in-lb the seal is incomplete and you get evaporative loss. Above 1.2 in-lb you crack the neck or deform the septum, and a deformed septum leaks at the puncture. For 9 mm screw-thread vials, the range is 0.8 to 1.2 in-lb with a PTFE-lined silicone septum.
The audit item is not the number itself. It is whether you have a torque setting on the crimper and a record that it was verified. A torque wrench or a calibrated crimper with a set stop, checked at the start of each batch, is what the auditor wants to see. If your lab seals by hand, you have no record, and the holding-time question becomes unanswerable.
Field case: the 14-day loss that was a crimp problem
A lab reported low recoveries for a volatile analyte on day-14 reanalysis but acceptable recoveries on day 0. The measurement that identified it was gravimetric: vials sealed at the start of the holding period and weighed at day 14 showed a mean mass loss of 12 mg, versus 2 mg for a control set sealed with a calibrated crimper at 0.8 in-lb. The hand-crimped vials had a mean torque of 0.3 in-lb. The fix was a calibrated crimper and a documented torque check, and the holding-time finding was closed with the gravimetric record. This is the kind of paper trail that turns a finding into a corrective action instead of a repeat observation.
What the auditor checks on the HPLC side of the same file
SW-846 covers the sample prep and the GC side heavily, but many labs run the same extracts on HPLC for methods like 8330 (nitroaromatics and nitramines), 8310 (PAHs), and the carbamate methods. The consumable audit items there are different, and they are usually about the column and the mobile phase, not the vial.
Column lot and performance documentation
For HPLC methods under SW-846, the auditor asks for the column’s performance test and the lot number. Method 8330, for example, requires baseline separation of the nitroaromatic analytes, and the auditor will check that your column’s performance test at receipt showed the required resolution. If you swapped columns mid-batch, the auditor wants the new column’s test and a system suitability run on the day of the swap. A column without a performance test on file is a documentation gap even if the chromatography looks fine.
The same logic applies to the guard column and the inline filter. A guard column that was changed without a record is an uncontrolled variable. I keep a column log with the install date, the lot, the injection count, and the system suitability result for every change. That log is what the auditor reads when they ask whether the batch was run on a qualified column.
Mobile phase and consumable compatibility
For methods using buffered mobile phases (phosphate at pH 2 to 7, or ion-pair reagents), the consumable that matters is the seal and the check valve, not the vial. A seal that sheds particles will show up as a pressure ripple and a baseline spike. The audit item is the maintenance record: when the seal was changed, and whether the change was tied to a pressure trend. If you cannot show the trend, the auditor cannot tell whether the batch ran on a qualified pump.
Receipt testing: the procedure that closes the loop
The gap between “the supplier says it is compliant” and “the auditor accepts it” is a documented receipt test. I use a short, repeatable procedure for every incoming lot of septa and closures, and it takes about 30 minutes.
| Test | Acceptance criterion | Record kept |
|---|---|---|
| Septum background (reagent water, full purge-and-trap) | Siloxane and phthalate peaks at or below prior lot; no target analyte above reporting limit | Chromatogram + lot number |
| PTFE facing thickness (micrometer, 10 septa) | 0.005 in ± 0.001 in, no delamination visible at 10x | Measurement log |
| Durometer (Shore A, 5 septa) | 45 to 55 Shore A | Measurement log |
| Closure torque (calibrated crimper, 10 vials) | 0.5 to 1.0 in-lb for 20 mm crimp; 0.8 to 1.2 in-lb for 9 mm screw | Torque wrench reading |
| Puncture count to failure (same septum, repeated injection) | ≥ 100 punctures before leak or coring | Puncture log |
| Dimensional check (vial OD, height, neck ID) | Within supplier drawing; 22.5 mm OD for 20 mm crimp, 11.6 mm for 9 mm screw | Caliper log |
The puncture count is the one most labs skip and the one auditors ask about when they see re-injection. A silicone septum with a bonded PTFE facing typically holds 100 to 150 punctures at 22 gauge before the facing cores. Below 100, you are re-injecting through a compromised seal. Document the count and replace the septum on schedule, not on failure.
For the physical consumables themselves, the dimensional match is what keeps the autosampler and the crimper working. A 20 mm crimp vial with a 22.5 mm ± 0.2 mm outside diameter and a 75.5 mm ± 0.5 mm height is the GPI standard geometry; a closure that is out of tolerance will not seat, and the torque record becomes meaningless. I source vials and closures through vial closures and OEM sample vials when I need the drawing and the CoA in the same document, because the audit trail starts at the drawing.
The paperwork packet I hand an auditor, in order
When the auditor sits down, I hand them a single packet. It is not a binder of everything. It is the five documents that answer the consumable question before they ask it.
- Purchase specification for each consumable class (septa, closures, vials, columns), with the dimensional and material numbers from the table above.
- Lot-specific CoA for each consumable lot used on the batch, with manufacturing and cure dates.
- Receipt test records (background chromatogram, thickness, durometer, torque, puncture count) for each lot.
- Consumable log tied to the batch: which lot of septum, which lot of closure, which column, and the date of each change.
- Corrective action record for any lot that failed receipt testing, with the disposition (returned, quarantined, or accepted with justification).
That packet is the difference between a finding and a clean audit. It is also the reason I treat consumables procurement as a documentation function, not a purchasing function. The supplier’s catalog is not the record. The lot number is the record.
The decision rule
Buy consumables against a written specification that names the material, the dimensions, the cure, and the receipt test, and require a lot-specific CoA with every shipment. If a supplier cannot give you a lot number, a cure date, and a background chromatogram for septa, do not put that lot on a compliance batch, no matter how the price looks. The EPA method requirements for HPLC and GC consumables are met in the physical product almost every time; they are failed in the paperwork. The auditor is not checking your septum. They are checking whether you can prove which septum it was, when it was made, and what it contributed to the blank. Build the packet before the audit, and the consumable question answers itself.
For the broader consumable catalog and the flow-path side of the same file, see HPLC consumables and the supplier network at chromatography supplier. For column and instrument-side reference data, the technical libraries at Agilent chromatography consumables and columns, Thermo Fisher chromatography and sample handling, and Restek technical resources for chromatography are where I check a specification before I write it into a purchase order.
References & further reading
Related resources from our lab consumables network:
- Hplc Consumables at hplcconsumable.com
- Chromatography Supplier at chromatographysupplier.com
- Oem Vial at oemsamplevials.com
- Vial Closures at vialclosures.com
External references and standards: