“Typical” is not a specification, and most sheets blur the line

A customer sent me a lot of 13 mm butyl closures last spring with a datasheet claiming ≤3 EU/mL endotoxin. Their QC lab failed the incoming check at 11 EU/mL, twice, on two separate kits. The sheet was not forged. Buried in a footnote was the phrase “typical value, not routinely tested.” That single line cost them three weeks of production and a rejected shipment. Reading a vial closures datasheet specification line by line, knowing which numbers are guaranteed and which are marketing, is the difference between a clean incoming inspection and a recall. This guide is about that skill specifically: how each value is measured, which are tested, and how to spot a sheet that hides its method.

Open any closure datasheet and count the qualifiers. You will find “typical,” “nominal,” “approximately,” “nominal value,” and “reference only” scattered across the table. In a regulated supply chain, those words mean the number is not a commitment. A specification is a value with a test method, a sample size, and an acceptance criterion attached. Anything else is a design target.

I ask suppliers one question before anything else: which rows on this sheet are release-tested against a written limit, and which are drawn from a design file? A serious supplier can answer in an email within a day. A distributor reselling someone else’s product usually cannot. If the answer is “all of them are typical,” you are buying an unverified product with a decorative table.

Here is the practical test. A tested value has three things attached: a method reference (USP <87>, ISO 8362-2, a specific ASTM method), a sampling plan (for example, n=10 per lot, AQL 1.0), and a limit that a certificate of analysis will actually report against. A nominal value has none of those. Nominal dimensions in particular are almost never what you measure. A “13 mm” closure is a GPI 13 designation, and the actual crown diameter tolerance is a separate line that many sheets omit entirely.

The three tiers you should sort every row into

I sort every datasheet row into one of three tiers. Tier 1 is released: tested per lot, method named, limit stated, CoA reports the result. Tier 2 is qualified: tested once at validation, not per lot, method often named. Tier 3 is nominal: engineering target or marketing figure, no test behind it. Tier 1 rows are the only ones you can build an incoming inspection around. Tier 3 rows are the only ones you should never put in a purchase specification.

How penetration force is actually measured, and why the number on the sheet rarely matches your bench

Penetration force is the row buyers trust most and should trust least. A sheet that says ≤15 N penetration force tells you almost nothing unless it also tells you the needle gauge, the needle bevel, the penetration speed, the stopper thickness at the puncture point, and whether the test was done on a bare stopper or a stoppered vial. Change any one of those and the number moves by 30% or more.

Here is what the measurement actually involves. A motorized test stand drives a needle at a fixed rate (commonly 200 mm/min, sometimes 100 mm/min) into a stoppered vial held in a fixture. A load cell records peak force. The reported figure is usually the peak, but some suppliers report the mean of a run, and a few report the force at first breakthrough only. A 20 mm stopper tested with a 20 G needle at 200 mm/min will read noticeably higher than the same stopper with a 22 G needle. Neither number is wrong. They are different tests.

Ask for the method. If the datasheet cannot tell you the needle gauge and speed, treat the penetration figure as Tier 3. For a pharma fill line running 21 G needles at high speed, a “≤15 N” claim tested with a 20 G needle at 100 mm/min is not the number your operators will feel at 300 vials per minute.

The fragmentation count is the row that matters more

Coring and fragmentation counts are the ones I want released per lot. The method should specify the needle, the number of punctures (commonly 10 per vial, sometimes 5), the liquid used, and the particle size cutoff (usually ≥50 µm, sometimes ≥25 µm). A sheet that reports “low coring” with no number is worthless. A sheet that reports <5 particles ≥50 µm per 10 punctures with the method named is a working specification you can verify.

If the sheet hides the method, the number is usually hiding too. I have seen datasheets where the fragmentation figure was collected on a 20 mm stopper with a 21 G needle at 5 punctures, then applied as a general claim to the whole product line. That is not a lie. It is a test that does not match your use case. For a multi-dose vial punctured 30 times over its in-use life, the number you need is the cumulative count at 30 punctures, not the per-10 figure.

Extractable volume and dimensions are nominal, and you should stop treating them as tested

The extractable volume of a vial closure is a design figure, derived from the stopper geometry and the vial neck. It is not measured per lot in the way a burst test is. When a datasheet says “extractable volume 1.2 mL for a 2 mL vial,” that is a nominal statement about the assembled system, not a tested property of the closure. Your fill volume is set by your filling line and validated by your own gravimetric check.

Dimensions are the same story. A “13 mm” or “20 mm” closure designation is a GPI/ISO reference, not a measurement. The actual critical dimensions (crown diameter, flange thickness, plug diameter, plug length) have tolerances, and those tolerances are what determine whether the closure seals. If the datasheet gives a nominal 13.0 mm crown with no tolerance, you cannot run a capability study against it. Ask for the dimensional drawing with tolerances. This is the one place I will point you to our own vial closures drawing set as an example of what a proper tolerance table looks like, because a nominal-only drawing is a common failure.

Where dimensional drift actually bites

I had a customer whose capper started throwing seal failures at roughly 1 in 400 vials after a closure lot change. The datasheet for both lots said “20 mm, nominal.” The measurement that identified the problem was a crown diameter check with a micrometer across 30 closures from each lot. The old lot ran 20.02 to 20.06 mm. The new lot ran 20.11 to 20.18 mm. The difference was within the GPI tolerance band, but it was outside the capper’s setup window, which had been tuned to the old lot. The fix was to re-set the capper against the new lot’s measured mean and add a crown diameter check to incoming inspection with a limit of 20.10 mm maximum for that line. The datasheet never showed the problem because it never showed a tolerance.

Endotoxin, sterility, and the rows where “tested” and “qualified” get confused

Endotoxin is the row where the difference between tested and nominal has the highest cost. A claim of ≤0.5 EU/mL or ≤3 EU/mL should come with a method reference, usually USP <85> bacterial endotoxins test, a stated LAL format (gel-clot, turbidimetric, chromogenic), and a per-lot release statement. If the sheet says “typical,” you have no release data. If it says “tested per USP <85>,” ask for a sample CoA and confirm the lot number and result appear on it.

Sterility claims for ready-to-use closures should reference a sterilization method and a dose or cycle. For gamma, that is typically a validated dose in the 25 to 40 kGy range with ISO 11137 method 1 or 2 verification. For ethylene oxide, it is a validated cycle with residuals tested per ISO 10993-7. A sheet that says “sterile” with no method is a Tier 3 claim. The same logic applies to USP <661> physicochemical testing: the sheet should report actual results (reducing substances, heavy metals, residue on ignition), not assert compliance.

If your incoming program needs to verify any of these, the lab work is the bottleneck, not the paperwork. That is why I route closure qualification samples through a dedicated analytical testing lab rather than trying to build the capability in-house, and I keep the method references in the purchase specification so the lab knows exactly which test to run.

Moisture vapor transmission, temperature, and the numbers that are qualified once and never again

Moisture vapor transmission (MVTR) for a closure is usually a qualified value, not a released one. The method is gravimetric: fill vials, stopper and cap, weigh, store at controlled temperature and humidity (commonly 40 °C / 75% RH or 25 °C / 60% RH), and reweigh at intervals. The reported figure is a rate, often mg/day per vial or g/100 days. The number depends heavily on the stopper formulation, the cap material, and the crimp or screw torque. A sheet that reports MVTR without stating the conditions is reporting a number you cannot use.

Temperature range is the most abused row on the sheet. A claim of -40 °C to +121 °C usually means the elastomer survives an autoclave cycle and a freezer without visible cracking. It does not mean the closure holds a seal across that range under pressure. For lyophilization, the relevant question is whether the stopper stays seated during the vacuum and re-pressurization cycle, and the test for that is a container closure integrity test (CCIT), typically dye ingress or vacuum decay, not a temperature claim. If you are running a lyo cycle, ask for CCIT data at the low-pressure setpoint, not a temperature range.

Torque and capping force are your numbers, not the supplier’s

No closure datasheet can tell you the right capping torque, because torque is a property of the assembled system: closure, vial, capper, and line speed. For screw-thread closures, a typical validated range on a 13 mm system is 8 to 12 in-oz, and on a 20 mm system 15 to 22 in-oz, but those are starting points. The number that matters is the one your CCIT and vacuum decay tests pass at, and you should establish it with a torque study across at least three torque levels and 30 vials per level. The datasheet can tell you the closure’s recommended seating force. It cannot tell you your line’s torque.

For crimp closures, the equivalent is the crimp height and diameter, measured with a crimp gauge. A datasheet that gives a “recommended crimp” without stating the vial neck specification is only half the information. The GPI/ISO neck finish is the other half, and it is on the vial drawing, not the closure sheet.

Two field cases where the datasheet line was technically true but practically useless

Case 1: the “≤3 EU/mL” that was never released

A biotech customer received a lot of 13 mm chlorobutyl closures with a datasheet claiming ≤3 EU/mL endotoxin. Their incoming test by USP <85> chromogenic method returned 11 EU/mL on two kits. The supplier’s response was that the 3 EU/mL figure was a “typical value” from development, not a per-lot release limit, and the CoA did not report endotoxin at all. The measurement that identified the gap was a side-by-side comparison of the datasheet row against the CoA rows. The CoA listed dimensions, penetration force, and fragmentation. It did not list endotoxin. The fix was a revised purchase specification that named USP <85> as a release test with a limit of ≤3 EU/mL, a sampling plan of n=10 per lot, and a CoA requirement. The supplier qualified the test over two lots and the problem did not recur. The lesson: a datasheet row that never appears on a CoA is not a specification.

Case 2: the penetration force that measured the wrong needle

A fill-finish site running 21 G needles at 400 vials per minute kept seeing needle bending and occasional stopper push-through on a new closure lot. The datasheet claimed ≤15 N penetration force. Their bench test with a 21 G needle at 200 mm/min gave a mean of 19.4 N with an RSD of 8.2%. The supplier’s test had used a 20 G needle at 100 mm/min. The measurement that identified the issue was a controlled penetration test at the line’s actual needle gauge and speed, run on 30 closures from the new lot and 30 from the previous lot. The previous lot read 13.8 N under the same conditions. The fix was a specification change that named the needle gauge, the speed, and a limit of ≤16 N at 21 G / 200 mm/min, plus a line trial at 30 vials per minute increments up to 400 to confirm the capper could handle the higher force. The datasheet was not wrong. It was measuring a different test. This is the most common datasheet failure I see, and it is why I insist every penetration force claim carry its needle gauge and speed.

A decision checklist for the next datasheet you open

Before you approve a closure, run this checklist against the sheet. If a row fails, ask for the missing information in writing, and do not put the row in your purchase specification until you get it.

• Method named? Every tested row should name a method (USP, ISO, ASTM) or a written internal method. No method, no specification.
• Limit stated? A number without a limit (“penetration force 15 N”) is ambiguous. Demand “≤15 N” or “13 to 17 N.”
• Sample size and frequency stated? Per lot, per shipment, or once at validation? The difference decides whether you can rely on the CoA.
• CoA match? Every released row should appear on the CoA with a lot-specific result. If the CoA is shorter than the datasheet, the missing rows are Tier 2 or Tier 3.
• Conditions stated? MVTR without temperature and humidity, or temperature range without a test basis, is a Tier 3 number.
• Needle and speed stated? Penetration force and fragmentation rows must name the needle gauge, bevel, and speed. Without them, the number is not transferable to your line.
• Tolerances on dimensions? A nominal dimension with no tolerance cannot be inspected. Ask for the drawing.
• Sterilization method and dose? “Sterile” is not a specification. ISO 11137 dose or EO cycle validation is.

If you are sourcing closures for a chromatography application, the same logic applies to the vial side of the system. The closure and the chromatography vial are qualified together, and a closure sheet that does not reference the vial neck finish is only half the picture. For autosampler work, I also check the HPLC vial drawing against the closure drawing before I approve either, because a 0.1 mm mismatch in the neck finish shows up as a septa push-through at 2 a.m. on a Friday. The method references you will need for the release tests are the same ones your lab already runs; the EPA SW-846 series is a good cross-check for extractables and leachables work in environmental labs, and the vendor technical libraries from Agilent, Thermo Fisher, and Restek are useful for confirming which closure and septa combinations their instruments actually tolerate at temperature.

The decision rule is simple. A datasheet row goes into your purchase specification only if it has a method, a limit, a sampling plan, and a CoA line. Everything else is a design target. Treat “typical” as a question, not an answer, and treat a sheet with no test methods as a sheet with no specifications. The supplier who can answer the method question in a day is the supplier you can build a qualification program around.

References & further reading

Related resources from our lab consumables network:

• Vial Closures at vialclosures.com
• Analytical Testing at labanalyticaltesting.com
• Chromatography Vial at chromvials.com
• Hplc Vial at hplcvialfactory.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