The word “nominal” on a cod vial datasheet is a tolerance statement, not a

A QC lab in Ohio rejected a 5,000-piece lot of COD vials last March because the batch certificate listed a nominal 25.0 mm OD and the incoming inspection log showed 24.82 mm on one cavity. The supplier pushed back, pointing to the word “nominal.” The lab manager pushed back harder, pointing to the word “datasheet.” Both were right, and both were wrong. Knowing which numbers on a cod vial datasheet are tested and which are nominal is the difference between a smooth method transfer and a CAPA that eats three weeks. This guide covers how each line on the spec sheet is measured, which values are guaranteed versus typical, and how to spot a sheet that hides its test method.

The word “nominal” on a cod vial datasheet is a tolerance statement, not a measurement

Most buyers read a cod vial datasheet the way they read a menu. They see “25 mm OD” and assume the part will be 25 mm. That assumption is wrong. In injection stretch blow molding, the number printed next to a dimension is almost always a nominal value. It is the target the mold was cut to, not the value any individual vial will measure. The actual part will vary around that target within a tolerance band that may or may not be printed on the same line.

I have seen datasheets where the OD is listed as “25.0 mm” with no tolerance, no test method, and no AQL reference. That is not a specification. That is a mold drawing. A real specification attaches three things to every dimensional line: the nominal value, the tolerance band, and the measurement method. If any of those three is missing, the line is decoration.

The tolerance band matters more than the nominal value. A vial with a nominal OD of 25.0 mm and a tolerance of ±0.2 mm will fit a heating block cut for 25.0 mm. A vial with the same nominal OD and a ±0.5 mm tolerance may not. The difference is not academic. Hach and Lovibond digestor blocks are machined to accept a specific OD range, and a vial at the low end of a wide tolerance will sit loose, transfer heat poorly, and give low recovery on the potassium hydrogen phthalate standard. A vial at the high end will bind, and the operator will force it, and the vial will crack at the neck during digestion.

When I audit a supplier, I ask one question: “Show me the cavity-to-cavity dimensional study for this mold.” A 16-cavity mold will produce 16 different ODs. The datasheet shows one number. The study shows the spread. A good supplier will give you the spread if you ask. A bad supplier will tell you the number is nominal and stop talking.

The guaranteed lines are the ones with a test method attached

Every line on a cod vial datasheet falls into one of three categories: guaranteed, typical, and reference. Guaranteed lines have a test method, an acceptance criterion, and a sampling plan. Typical lines are measured on a development lot or a single production run and reported as informational. Reference lines are not measured at all. They are copied from the resin supplier’s datasheet or from a competitor’s sheet.

The problem is that most datasheets do not tell you which category each line belongs to. They print everything in the same font, in the same table, with the same confident decimal places. A buyer cannot tell the difference between a value that was measured on every lot and a value that was measured once in 2019.

Here is the rule I use. If a line has a test method reference (ASTM, ISO, an internal SOP number), it is probably guaranteed. If it has a tolerance, it is probably guaranteed. If it has neither, it is typical or reference. If it has a footnote, read the footnote. That is where suppliers hide the phrase “typical value, not a specification.”

Take the wall thickness line. A datasheet might say “wall thickness: 1.1 mm.” That number is almost always nominal and almost always wrong for at least some cavities. Wall thickness in a blow-molded vial varies by position: the shoulder is thinner than the sidewall, the sidewall is thinner than the base, and the neck is thicker than everything. A single number cannot describe that distribution. A serious datasheet will list wall thickness at three positions (neck, sidewall midpoint, base corner) with a tolerance for each. If your datasheet has one number, the supplier is giving you the mold design value, not a measured value. That is fine if you are buying a mold. It is not fine if you are buying vials for a digestion method where wall thickness controls heat transfer rate and burst pressure.

For Cod Vial products, I always request the cavity study and the lot release test plan before I place a repeat order. The first order is a qualification. The second order is where you find out whether the supplier can hold the numbers they printed.

The burst pressure line is the one most suppliers do not actually test

COD digestion runs at 150 °C in a sealed vial. The internal pressure at that temperature is not trivial. Sulfuric acid and potassium dichromate generate gas, and the headspace expands. A vial that cannot hold pressure will fail at the neck or the base, and a failed vial inside a digestor block is a cleanup problem and a lost sample.

Many datasheets list a burst pressure, often something like “≥ 6 bar” or “tested to 100 psi.” In my experience, fewer than half of the suppliers who print that line actually test it on production lots. Most of them calculate it from the resin’s tensile strength and the nominal wall thickness. That calculation assumes a uniform wall, no stress concentrations, and a perfect seal. None of those assumptions survive contact with a real mold.

The test that matters is a burst test on finished vials, done with a pressure ramp and a failure criterion. A supplier who does this will usually state the test method: “internal hydraulic pressure, ramp 0.5 bar/s, failure at first leak.” A supplier who does not will print a number with no method. Ask for the burst test report. If they do not have one, the number on the datasheet is a calculation, not a guarantee.

I have seen a lot of vials where the datasheet said “≥ 6 bar” and the actual burst pressure on cavity 7 was 3.8 bar. The failure mode was a thin shoulder caused by uneven preform heating. The datasheet did not lie, exactly. The supplier had tested one cavity once and extrapolated. The lesson: a burst pressure line without a cavity-level test plan is a typical value wearing a guarantee’s clothes.

Seal and cap torque values are system specs, not vial specs

A cod vial datasheet will often include a line for “closure torque” or “sealing force.” This is a trap. The vial does not have a torque value. The closure system does. The torque required to seal a vial depends on the thread geometry, the cap material, the liner material, and the temperature at which the cap is applied. A vial supplier who prints a torque value on the vial datasheet is either testing the complete closure system or copying a number from a cap supplier.

The number that matters is the application torque, and it is usually in the range of 1.5 to 2.5 N·m for a 20 mm or 24 mm polypropylene cap with a PTFE-lined septum on a glass COD vial. Below 1.0 N·m, the seal is unreliable during digestion. Above 3.0 N·m, the cap can strip or the vial neck can crack. The exact value depends on the thread finish. A GPI 20-400 thread behaves differently from a GPI 24-400 thread, and a vial with a European DIN 20 finish is not interchangeable with a US GPI 20 finish.

If the datasheet lists a torque value, check two things. First, is it an application torque or a removal torque? Removal torque is always lower, because the cap relaxes after application and the liner compresses. Second, is it tied to a specific cap and liner? A bare torque number without a closure specification is meaningless. I have seen labs apply 2.0 N·m to a vial and cap combination that required 1.2 N·m, and the result was cracked necks on 30% of the lot during the digestion step. The vial datasheet said “torque: 2.0 N·m.” The cap datasheet said “torque: 1.2 N·m.” Both were right for their own part. Neither was right for the system.

When I specify a COD vial for a lab, I treat the vial and the closure as one system and I require a seal integrity test on the assembled unit. The test I use is a dye ingress test with methylene blue at 150 °C for 2 hours, or a helium leak test if the lab has the equipment. The pass criterion is no dye ingress and no visible deformation of the cap or vial. That is the number that matters, and it is not on most vial datasheets.

Chemical resistance lines are copied from resin datasheets, and that is a problem

Every cod vial datasheet has a chemical resistance section. It usually says something like “excellent resistance to acids, bases, and organic solvents.” That line is almost never tested on the finished vial. It is copied from the resin manufacturer’s datasheet for the base polymer. The problem is that a finished vial is not the same as a resin pellet. The molding process introduces orientation, residual stress, and surface defects. A resin that resists sulfuric acid in a lab coupon test may craze or crack in a molded vial under digestion conditions.

The specific failure I see most often is environmental stress cracking in the neck area. The neck is where the mold leaves the highest residual stress, and it is where the cap applies a constant hoop stress. Sulfuric acid vapor attacks the stressed polymer, and the neck cracks. The datasheet said “excellent chemical resistance.” The vial failed in 40 minutes.

A serious chemical resistance line on a cod vial datasheet will specify the test conditions: chemical, concentration, temperature, exposure time, and evaluation criterion. “Resistant to 98% sulfuric acid at 150 °C for 2 hours, no visible change” is a test result. “Excellent chemical resistance” is marketing. If the datasheet does not give you the test conditions, assume the line is a reference value copied from the resin sheet.

For labs running EPA methods, the relevant reference is U.S. EPA SW-846 test methods for waste and materials. Method 410.4 and Method 5220 specify the digestion conditions. The vial datasheet should be read against those conditions, not against a generic chemical compatibility chart.

Transparency and optical clarity are measured, but the method matters

COD vials are read in a spectrophotometer or a colorimeter, so optical clarity matters. The datasheet will list a transparency value, usually as percent transmittance at a specific wavelength. The question is: measured how?

The standard method for measuring light transmittance of a vial is to place the empty vial in the light path and measure the transmittance at a specified wavelength, often 600 nm or 420 nm. A clean, well-molded glass COD vial should transmit at least 85% at 600 nm. A plastic vial will be lower, typically 80 to 85%. But the number depends on the wall thickness, the surface finish, and the presence of any mold release or surface contamination.

The trap is that some suppliers measure transmittance on a flat plaque molded from the same resin, not on the finished vial. A flat plaque has uniform thickness and no curvature. A vial has variable wall thickness and a curved surface that scatters light. The plaque number will be higher than the vial number. If the datasheet does not state that the measurement was made on the finished vial, assume it was made on a plaque and discount the value by 5 to 10 percentage points.

I have also seen datasheets that list “transparent” as the optical specification. That is not a measurement. That is an adjective. A real optical spec will include the wavelength, the path length, the instrument type, and the acceptance criterion. Without those, the line is useless for method validation.

The dimensional lines that actually affect your method

A cod vial datasheet will list a dozen dimensions. Most of them do not matter for your method. The ones that do are the outer diameter, the overall height, and the neck finish. Everything else is secondary.

The outer diameter determines whether the vial fits in the digestor block. Hach blocks are cut for a specific OD, typically 16 mm for the low-range vials and 25 mm for the high-range vials. A vial that is 0.3 mm over the block diameter will not seat. A vial that is 0.3 mm under will sit loose and heat unevenly. The tolerance on OD should be no wider than ±0.2 mm for a glass vial and ±0.3 mm for a plastic vial. If the datasheet shows a wider tolerance, the supplier is telling you they cannot hold the mold.

The overall height matters because the spectrophotometer or colorimeter has a fixed light path. The vial must place the sample at the correct height in the beam. A vial that is 1 mm short will give a low reading because the light passes through less sample. A vial that is 1 mm tall will give a high reading or will not fit in the cell holder. The height tolerance should be no wider than ±0.5 mm.

The neck finish is the dimension that causes the most field failures. A COD vial with a GPI 20-400 finish will accept a cap with a GPI 20-400 thread. A vial with a GPI 24-400 finish will not. The datasheet should state the finish in GPI or ISO terms, not in vague language like “20 mm neck.” A “20 mm neck” can refer to the OD of the neck, the thread pitch, or the cap size, and those are three different things. The GPI system is unambiguous. If the datasheet does not use it, ask for a thread drawing.

For labs that also run HPLC, the same dimensional discipline applies to Vial Insert selection and to 2Ml Vial purchases. A 2 mL HPLC vial with a 9 mm neck is not the same as a 2 mL vial with an 8-425 neck, and the datasheet must say which one it is.

How to spot a datasheet that hides its test method

Suppliers hide missing test data in predictable ways. Here is the checklist I use when I review a cod vial datasheet:

• No tolerance on a dimensional line. A number without a tolerance is a mold target, not a specification.
• No test method reference on a performance line. Burst pressure, chemical resistance, and seal integrity without a method are calculations or copies.
• No sampling plan. A guarantee without a sampling plan is a wish. The plan should state the lot size, the sample size, and the AQL.
• The word “typical” in a footnote. Typical values are not guaranteed. They are measured once and reported forever.
• Values copied from a resin datasheet. If the chemical resistance section matches the resin supplier’s wording, it was not tested on the finished vial.
• No cavity identification. A multi-cavity mold produces variation. A supplier who does not track cavity-level data cannot guarantee lot-to-lot consistency.
• A single wall thickness number. Wall thickness varies by position. A single number is a mold design value.

If a datasheet triggers three or more of these flags, I treat every number on it as a typical value until proven otherwise. That does not mean the supplier is bad. It means the datasheet is incomplete, and incomplete datasheets cause qualification failures.

Field case: the lot that passed the datasheet and failed the method

A contract lab in New Jersey ran 200 COD samples per day for a wastewater client. They bought vials from a supplier whose datasheet listed “OD 25.0 mm, tolerance ±0.2 mm” and “burst pressure ≥ 6 bar.” The lot passed incoming inspection on OD. The first 50 samples digested normally. Sample 51 cracked at the shoulder during the 150 °C hold. Sulfuric acid leaked into the block. The lab shut down for two days to clean the block and recalibrate.

The measurement that identified the problem was a wall thickness profile on the failed vials. The shoulder wall measured 0.62 mm on the failed cavity, compared to 0.95 mm on a passing cavity from the same lot. The datasheet said “wall thickness: 1.1 mm” with no position and no tolerance. The supplier had never measured the shoulder. The burst pressure line was calculated from the nominal 1.1 mm wall, so it predicted a failure pressure that the actual shoulder could not survive.

The fix was a supplier change and a new incoming inspection step. The lab now measures wall thickness at the shoulder on 10 vials per lot and rejects any lot with a shoulder wall below 0.80 mm. The new supplier’s datasheet lists wall thickness at three positions with tolerances and a burst test method on finished vials. The lab has not had a digestion failure in 18 months.

Field case: the cap torque number that belonged to a different closure

A municipal lab in Texas switched cap suppliers to save cost. The new cap was a polypropylene 20-400 with a PTFE-lined septum, same as the old cap. The vial datasheet said “torque: 2.0 N·m.” The lab applied 2.0 N·m to the new caps. During digestion, 12% of the vials leaked at the seal. The method recovery dropped below the acceptance limit, and the lab failed a proficiency test.

The measurement that identified the problem was a seal integrity test on the assembled vial and cap. The new cap’s liner was 0.2 mm thinner than the old liner, and the cap’s thread engagement was shallower. At 2.0 N·m, the liner did not fully compress, and the seal failed at 150 °C. The old cap sealed at 1.8 N·m. The new cap required 2.6 N·m to achieve the same seal, but the vial datasheet still said 2.0 N·m.

The fix was to treat the vial and cap as a system and to qualify the combined assembly, not the individual parts. The lab now specifies a seal integrity test on every new lot of vials and caps together. The torque value is set per system, not per part. The vial datasheet is read as a component spec, not a system spec.

Decision checklist: what to require before you buy

Before you place a COD vial order, get the following in writing from the supplier. If they cannot provide it, walk away or price the risk into the purchase.

This checklist is not exhaustive, but it covers the lines that cause the most field failures. If a supplier’s datasheet cannot answer these lines, the datasheet is a brochure, not a specification.

For labs that run both COD and HPLC methods, the same discipline applies to the broader consumables budget. A Hplc Consumables purchase should be evaluated the same way: tested values with methods, not nominal values with adjectives. The suppliers who provide real data tend to be the same across both product lines. The ones who hide behind “nominal” do it everywhere.

External resources that help are the Agilent chromatography consumables and columns library for method-specific guidance, the Thermo Fisher chromatography and sample handling technical notes for vial and closure compatibility, and the Restek technical resources for chromatography for dimensional and material data that is actually measured. These resources do not replace a supplier’s datasheet, but they give you a benchmark for what a complete specification looks like.

The decision rule is simple. A cod vial datasheet is a contract, not a menu. Every line that affects your method must have a test method, a tolerance, and a sampling plan. Every line that does not is a typical value, and typical values do not survive contact with a 150 °C digestion block. Buy from suppliers who give you the cavity study, the burst test report, and the seal integrity data. Pay less attention to the number itself and more attention to whether the number was earned. A nominal value is a promise. A tested value is a fact. Your method runs on facts.

References & further reading

Related resources from our lab consumables network:

• Cod Vial at codvial.com
• Vial Insert at hplcinsert.com
• 2Ml Vial at 2mlsamplevial.com
• Hplc Consumables at hplcconsumable.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