Unit price is a numerator; the denominator is reported results, not vials

Last quarter a 96-well plate run on a GC-MS headspace method came back with 11 of 96 vials showing siloxane peaks at 4.2 minutes, and the analyst reran the whole plate. That rerun cost 3.5 hours of instrument time, 96 septa, 96 vials, and a 0.5-day slip on a client report. The invoice for those 96 vials was $38. The rerun was $610 in loaded labor alone. That gap, between what the vial costs and what a reported result costs, is the entire subject of this piece. If you are building a chromatography supplier cost per test calculation, you cannot stop at the unit price on the quote. You have to carry the rerun rate, the reject rate, and the prep labor through to a per-result number, and then compare grades on that basis.

Unit price is a numerator; the denominator is reported results, not vials consumed

Most procurement exercises I see compare suppliers on price per vial, price per septum, price per column. That is a numerator. The number that matters is the denominator: how many reported results you got for the money. Two suppliers can quote $0.42 and $0.68 per 2 mL clear vial with a 9 mm screw thread, and the $0.68 vial can be the cheaper one by a wide margin once you divide by completed tests.

The formula I use:

Cost per reported result = (consumable spend + prep labor + instrument time on reruns + reject investigation labor) / (total tests submitted − reruns − rejects)

Every input is measurable. Most labs measure only the first one. The other three are where the spread between grades lives, and they are the reason a “cheap” consumable can cost 2.4x more per result than the one you rejected on price.

You do not need a data system to run this. A spreadsheet with four columns (lot, tests submitted, reruns, rejects) and a monthly pull from the LIMS is enough. Do it for one quarter before you negotiate anything.

Reruns dominate the equation, and they cluster around specific failure modes

Across the labs I have worked with, rerun rates on a clean method run between 1.5% and 3% on a good consumable and 6% to 11% on a marginal one. The difference is not random. Reruns cluster around four consumable-driven failure modes:

• Septum coring. A needle punching a 0.5 mm-bore septum at 40 psi inlet pressure will core if the septum durometer is below 45 Shore A or the pre-slit geometry is wrong. Cores land in the liner and produce ghost peaks or a split peak. Typical rerun trigger: 1 in 20 injections.
• Vial-to-vial volume drift. A 2 mL vial with a body ID tolerance of ±0.15 mm changes the headspace volume enough to shift partition equilibrium by 2% to 4%, which shows up as RSD creeping from 1.8% to 4.1% across a plate.
• Cap torque loss. A screw cap torqued to 1.2 N·m instead of the 1.5 N·m the closure was designed for will back off during a 40-minute GC run at 60 °C, and the vial leaks. You see it as a low response on one well and a rerun on the whole plate.
• Particulate from the septum. Silicone or PTFE laminate sheds if the laminate bond is poor. Particles land in the needle and cause a partial injection, which the software flags as a short injection and the analyst reruns.

The pattern is the same every time: the rerun is triggered by a single well, but the analyst reruns the plate. That is the multiplier that kills you. One bad vial in 96 costs you 96 vials of instrument time.

Reject rate is a different animal from rerun rate, and it is often invisible

A rerun is a result you did not get the first time. A reject is a result you cannot report at all: QC out of control, a failed system suitability check, a sample that has to be re-collected because the vial leaked before it reached the instrument. Rejects are worse because they carry a re-collection or re-prep cost on top of the instrument time.

In my experience, reject rates on consumable-driven failures sit at 0.4% to 0.9% on a good lot and 2% to 5% on a bad one. The bad ones almost always trace to a lot-level variable: a septum lot with a durometer shift, a vial lot with a body ID drift, a closure lot with a liner delamination. That is why lot-level incoming inspection is worth the hour it costs. A 30-vial dimensional check with a go/no-go gauge on the neck OD and a caliper on the body ID catches most of them before they reach the bench.

If you want a reference on what “in control” looks like for the analytical method itself, the U.S. EPA SW-846 test methods spell out the QC acceptance criteria that a rerun or reject has to fail before you can legitimately call it one. Use those criteria, not a gut feel, when you classify a failed result.

Prep time is the largest line item nobody puts in the cost model

Here is a number that surprises people: on a typical headspace or LC-MS prep, the consumable is 4% to 7% of the fully loaded cost per test, and hands-on prep labor is 35% to 50%. The rest is instrument time, overhead, and QC.

That means a consumable that adds 30 seconds of prep per sample (a cap that does not thread cleanly, a septum that needs to be seated by hand, a vial that does not fit the autosampler tray without a shim) costs more than a consumable that is 40% cheaper per unit. Do the arithmetic: 30 seconds × 96 samples = 48 minutes per plate. At a $55/hour loaded rate, that is $44 per plate. If you run 20 plates a week, that is $880 a week, or roughly $45,000 a year, for a “saving” of a few cents per vial.

This is where the OEM vial specification matters. A vial that is manufactured to the same dimensional drawing as the autosampler manufacturer’s own (body OD, neck OD, neck height, crimp or thread pitch) will load and seat without intervention. A vial that is “close” will not. The difference shows up as prep time, not as a line on the invoice.

Two grades, same dimensions, different cost per reported result

The table below is built from a real comparison I ran on a 96-well headspace method, 20 plates per week, loaded labor at $55/hour, instrument time at $18/hour. Both grades meet the dimensional spec on paper. The difference is in the lot-to-lot consistency and the septum construction.

The economy grade is 38% cheaper per unit and 104% more expensive per reported result. That is the whole argument. The vial price is a rounding error next to the rerun and prep line items. If you are comparing suppliers on unit price alone, you are comparing the wrong column.

I will add one caveat: the premium grade only wins if your method is tight enough to notice the difference. On a rough screening method with a 15% RSD acceptance criterion, both grades pass and the economy one is genuinely cheaper. The calculation has to be run against your method, not a generic one.

What breaks the cost model: three failure modes and the fix

Field case 1: the drifting RSD that was a vial body, not a method

A contract lab running a headspace GC-FID method for residual solvents saw the RSD on the internal standard climb from 2.1% to 5.8% over six weeks. The method had not changed. The instrument had been serviced. The analyst had not changed. What had changed was the vial lot: the incoming lot had a body ID at the top of the tolerance band, which changed the headspace-to-sample ratio and shifted the partition coefficient enough to widen the RSD.

The measurement that identified it: a 30-vial dimensional check with a bore gauge on the body ID, plotted against the RSD from each of the 30 wells spiked at the same concentration. The correlation was 0.81. The fix was to add a body ID check to incoming inspection (go/no-go at ±0.15 mm) and to reject the lot. RSD came back to 2.2% on the next lot. The cost of the six-week drift, in reruns, was about $4,100.

Field case 2: the cap torque that leaked at temperature, not at ambient

An environmental lab running volatiles by purge-and-trap saw a 4% reject rate on a batch that had passed leak testing at ambient. The leak test was a visual and a weight check at room temperature. The method ran at 40 °C for 11 minutes. At temperature, the cap backed off on about 1 in 25 vials, and the low response showed up on the internal standard.

The measurement that identified it: a torque audit with a calibrated cap torque tester on 50 vials, before and after a 40 °C hold. The vials torqued to 1.2 N·m lost 0.3 N·m over the hold; the vials torqued to 1.5 N·m lost 0.05 N·m. The fix was to set the capper to 1.5 N·m and to add a post-hold torque check to the QC plan. Reject rate dropped to 0.6%. This is a closure problem, and the vial closures supplier should be able to tell you the design torque for the closure, not just the thread pitch.

Building the calculation into your supplier comparison

Here is the decision checklist I use when comparing two chromatography suppliers on per-test cost. It is designed to be run on a spreadsheet in an afternoon with data you already have.

• Pull three months of rerun and reject data from the LIMS, split by consumable lot. If you cannot split by lot, start splitting today; without it the calculation is a guess.
• Measure prep time per plate with a stopwatch on five plates per grade. Do not use the SOP time; use the observed time.
• Get the design torque for the closure from the supplier, and verify it with a torque tester on 20 vials. Anything outside ±0.1 N·m is a flag.
• Run a 30-vial dimensional check on the body ID, neck OD, and neck height against the autosampler drawing. A Cpk below 1.33 on any of the three is a flag.
• Ask for the septum durometer and the lot-to-lot range. A range wider than ±5 Shore A is a flag for coring.
• Calculate cost per reported result for both grades using the formula above, and compare.
• Re-run the calculation after one quarter on the winning grade. The number moves; do not treat it as fixed.

For a reference on what a good supplier should be able to give you on the technical side (durometer, dimensional drawings, lot traceability), the Restek technical resources and the Agilent chromatography consumables documentation are useful benchmarks for what “complete” looks like. If your supplier cannot match that level of specification detail, that is itself a cost input, because you will spend prep time and rerun time discovering the gaps.

One more input that belongs in the model: the analytical testing lab’s own QC acceptance criteria. If your analytical testing partner applies a tighter RSD or a stricter system suitability criterion than your in-house method, the cost per result shifts because the rerun threshold moves. Run the calculation against the acceptance criteria that actually govern the report, not the ones in the SOP.

When the premium grade is the wrong answer

I want to be clear that the premium grade does not always win. It wins when the method is tight, the volume is high, and the rerun cost is real. It loses when:

• The method has a wide acceptance criterion (RSD > 10%) and the economy grade passes it consistently.
• The volume is low (under 200 tests a month), so the fixed cost of qualifying a new grade exceeds the rerun savings.
• The samples are dirty or the matrix is harsh, and the vial is a consumable in the literal sense (it gets thrown away after one injection regardless of grade).

In those cases the economy grade is the right call, and the calculation will show it. The point is not “buy premium.” The point is “buy the grade that minimizes cost per reported result for your method and your volume,” and that requires the calculation.

The decision rule

Buy on cost per reported result, not cost per unit, and compute it as (consumable spend + prep labor + rerun instrument time + reject investigation) divided by reported results. Run it for one quarter on your current grade, then run it on the alternative. If the alternative is not at least 15% cheaper per reported result, the switching cost (requalification, retraining, incoming inspection changes) will eat the difference. If it is more than 30% cheaper, switch and re-run the calculation after a quarter. The chromatography supplier that gives you the lot-level data to run this calculation is worth more than the one that quotes the lowest unit price, because the unit price is the smallest number in the equation.

References & further reading

Related resources from our lab consumables network:

• Chromatography Supplier at chromatographysupplier.com
• Oem Vial at oemsamplevials.com
• Vial Closures at vialclosures.com
• Analytical Testing at labanalyticaltesting.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