A lab I audited last spring ran 1,840 injections in a month and reported 1,612 results. That gap, 228 failed or repeated injections, is where most of the money went, and almost none of it showed up in the purchase order. The vials cost $0.31 each. The cost per reported result was $2.47. If you are trying to work out HPLC consumables cost per test calculation for your own lab, the purchase price is the smallest input in the equation, and it is the only one your ERP system tracks.
Price per unit is a catalog number; price per reported result is an output of your failure rate
The unit price on a case of 100 vials tells you what the warehouse paid. It says nothing about how many injections those vials survive, how many sequences they abort, or how much analyst time they consume before a result reaches a client. I have seen two labs buy the identical 2 mL clear glass vial, 12 x 32 mm, 9 mm screw thread, from the same distributor, and end up with a 3.4x spread in cost per reported result. Same part number. Different workflow, different grade of closure, different prep discipline.
The calculation has four terms, and most procurement files capture only the first:
- Consumable cost per injection: vial, cap, septum, filter, pipette tip, solvent volume, column amortization.
- Instrument time per injection: run time plus equilibration plus the autosampler overhead, multiplied by your internal hourly rate.
- Rerun and reject multiplier: the ratio of injections performed to results accepted.
- Prep and review labor per accepted result: sample weighing, dilution, data review, and the second review that a rejected batch forces.
Divide the sum by accepted results, not by injections. That is the only number a client or an internal customer actually pays for.
The rerun multiplier does more damage than any unit price difference
Take a 12-minute run on a 1260 Infinity II with a standard autosampler. Loaded instrument rate, including the analyst and the overhead, runs about $85 to $110 per hour in most contract labs I have worked with. That is $17 to $22 of instrument time per injection. A $0.31 vial is 1.5% of the injection cost. A single rerun of a 20-injection batch because of one bad crimp wipes out the vial savings for the entire batch and then some.
Here is the arithmetic that changes minds in procurement meetings. Suppose grade A vials cost $0.31 and grade B cost $0.19. Grade B has a 4% septum coring rate on your autosampler needle, and grade A has 0.6%. On a 500-injection month:
- Grade B: 20 coring events, most of which trigger a rerun of the affected sample plus a needle or loop cleaning. At $20 instrument time and $6 labor per rerun, that is $520 in avoidable cost, plus the risk of a needle blockage that takes the instrument down for an hour.
- Grade A: 3 coring events, roughly $78.
- Unit price difference on 500 vials: $60 in favor of grade B.
Grade B saves $60 on the invoice and costs $442 more in the lab. The purchase order is not wrong. It is just measuring the wrong thing. When you build the HPLC consumables cost per test calculation, the rerun multiplier has to sit in the numerator, not in a footnote.
What a rerun actually costs, line by line
I break a rerun into five components because each one is negotiable in a different way:
- Instrument time: full run length, plus 8 to 15 minutes of re-equilibration if the failure forced a column flush.
- Sample: often the limiting reagent. If the sample is a single 1.5 mL aliquot and the rerun needs a fresh prep, you may not be able to rerun at all, which turns a rerun into a resample and a client call.
- Analyst time: 4 to 12 minutes to identify the failure, re-inject, and annotate the sequence.
- Data review: the second review of the batch, which is often the largest single line item at 10 to 20 minutes.
- Schedule slip: the batch that was queued behind the failed sequence. This one is invisible but real in a lab running at 80% capacity or more.
At a fully loaded $95 per hour, a single rerun of a 20-minute method with 12 minutes of analyst and review time costs about $50. If your rerun rate is 6%, that is $3 per injection of hidden cost, ten times the price of the vial.
Rejects are a different animal from reruns, and they hit the prep budget hardest
A rerun is an injection that failed and was repeated. A reject is a result that could not be recovered at all: the sample was consumed, the prep was invalid, or the data failed the acceptance criteria and the client was notified. Rejects carry the full prep cost, not just the injection cost. A solid-phase extraction prep with a $14 cartridge, 45 minutes of analyst time, and a $40 internal standard charge is roughly $95 to $120 before the sample ever reaches the autosampler. If 3% of those preps are rejected because of a consumable failure, the cost per reported result climbs by $3 to $4 on every sample, accepted or not.
I track reject rate separately from rerun rate because they respond to different controls. Reruns respond to autosampler and vial closure quality. Rejects respond to filter and syringe quality, and to the sample prep SOP. Blending them into one “failure rate” number hides the lever.
A two-grade comparison, with the numbers that actually move
The table below is built from data I have collected across three labs running the same 18-minute reversed-phase method on 2 mL vials, 9 mm screw thread, with 300 µL sample volume. Grade A is a certified low-extractable vial with a PTFE/silicone septum and a controlled crimp or screw torque. Grade B is a general-purpose vial with a standard silicone septum and looser dimensional control. The instrument rate is $95 per hour, and the prep cost per sample is $105.
| Cost driver | Grade A | Grade B | Delta per 1,000 injections |
|---|---|---|---|
| Vial + cap + septum unit price | $0.31 | $0.19 | -$120 (B cheaper) |
| Septa coring rate (needle punctures before failure) | 0.6% (avg 3 punctures before core) | 4.2% (avg 1.4 punctures) | +36 coring events for B |
| Rerun rate from closure and coring failures | 1.8% | 6.4% | +46 reruns for B |
| Cost per rerun (instrument + analyst + review) | $48 | $48 | +$2,208 for B |
| Reject rate from extractable peaks and filter failures | 0.4% | 1.9% | +15 rejects for B |
| Cost per reject (full prep + instrument + client notification) | $165 | $165 | +$2,475 for B |
| Total cost for 1,000 injections | $5,910 | $9,873 | +$3,963 for B |
| Cost per reported result | $6.02 | $10.69 | +78% for B |
The unit price difference is 12 cents. The cost per reported result difference is $4.67. That is the entire argument, and it is why I refuse to let a purchasing team evaluate HPLC consumables on unit price alone. The same logic applies to filters, syringes, and septa, and it applies with more force as the method gets longer.
If you are sourcing for a multi-site operation, the same table becomes a negotiation tool. Ask the supplier for the coring rate and the extractables profile on the certificate of analysis, and hold them to it. A supplier who cannot produce a puncture count or a lot-specific extractables number is selling you grade B at grade A pricing. I go into the traceability chain in more depth elsewhere, but the short version is that the COA is the only document that lets you put a number in the rerun column.
Prep time is where the per-test calculation usually breaks down
Most labs I have worked with can estimate instrument time to within 10%. Almost none can estimate prep time to within 30%, because prep time is not logged against the sample, it is logged against the analyst’s shift. I have watched a lab spend 22 minutes per sample on a manual liquid-liquid extraction that they believed took 10 minutes. The stopwatch did not agree with the SOP.
To get a usable prep number, time 20 samples end to end with a stopwatch, including the weighing, the dilution, the vortex, the centrifuge, and the transfer. Divide by 20. Add 15% for the interruptions that always happen. That number, multiplied by the loaded hourly rate, is the prep cost per sample. It typically lands between $40 and $130 for a manual prep and between $12 and $30 for an automated one. The gap between those two numbers is larger than the entire consumables budget for most methods, which is why I push labs to automate prep before they optimize the vial price.
Filters and syringes change the prep time, not just the prep cost
A 0.22 µm PTFE filter that clogs at 200 µL of a high-particulate sample forces a second filtration, which is 3 to 5 minutes and a second filter. A 0.45 µm PVDF filter with a larger effective area handles the same sample in one pass. Over 500 samples a month, the second filter costs $0.18 each and the second filtration costs $0.55 in analyst time. The cheaper filter is not cheaper. This is the same failure mode that shows up in the autosampler, just one step earlier in the workflow.
Two field cases where the per-test number exposed a problem the unit price hid
Case 1: the crimp torque that looked fine and cost 11% of the batch
Symptom: a lab running a 96-well plate autosampler on a headspace method saw a 9% rerun rate on one lot of 11 mm crimp vials. No leaks visible to the eye. No instrument alarms. The reruns clustered on the last 20 positions of each plate.
Measurement: I pulled 40 vials from the affected lot and measured crimp diameter with a go/no-go gauge. The mean was 0.4 mm under the target of 7.0 mm, and the standard deviation was 0.3 mm. The crimper had drifted. The seal was marginal, and the last positions in the plate sat longer at temperature before injection, which let the marginal seal fail.
Fix: recalibrated the crimper to 7.0 ± 0.1 mm, added a daily gauge check with a 5-vial sample, and switched to a vial with a tighter neck OD tolerance. The rerun rate dropped to 1.6% within two weeks. The cost per reported result fell from $8.40 to $6.10 on that method. The vial price did not change by a cent.
Case 2: the extractable peak that only appeared on the third injection from a vial
Symptom: a method for a small-molecule API showed an extra peak at 6.8 minutes on roughly 3% of injections, always on the third or fourth injection from the same vial. The peak was below the reporting threshold but above the identification threshold, so the lab was rejecting the results.
Measurement: I ran a blank injection series from a single vial, 10 punctures, and the 6.8-minute peak appeared on puncture 3 and grew through puncture 10. The septum was a standard silicone with no PTFE facing. The needle was coring the septum and the silicone was extracting into the sample. The lot COA listed the septum as “silicone” with no facing specification.
Fix: switched to a PTFE-faced silicone septum with a specified 0.010-inch facing, confirmed the extractable profile on the COA, and re-ran the validation. The peak disappeared. Reject rate on that method went from 3.1% to 0.2%. The septum cost $0.04 more per unit and saved roughly $3.80 per reported result.
The decision rule for a per-test procurement file
Build the file around four numbers per SKU, not one. Unit price, rerun rate, reject rate, and prep time. If a supplier cannot give you the first two in writing, treat the SKU as grade B until proven otherwise. If your own lab cannot give you the last two, run a two-week measurement before you sign the next contract, because you are currently buying on a number that does not predict your cost.
For the calculation itself, use this formula and put it in the purchasing template:
Cost per reported result = (consumable cost + instrument time + prep labor + review labor) × (1 + rerun rate) / (1 – reject rate)
Run it monthly. When the number moves, find out which term moved. In my experience the rerun rate is the term that moves first and the one that procurement is least equipped to see. The HPLC consumables catalog is where the unit price lives. The rerun rate lives in your chromatography data system, and that is where the real negotiation happens. A supplier who will quote a puncture count and a lot-specific extractables profile is worth a 15% premium on the unit price, because the per-test math usually pays it back within the first quarter. If you want a second source to benchmark against, Chromatography Supplier and the technical libraries at Restek both publish the kind of performance data you need to populate the rerun and reject columns, and the Agilent consumables documentation and Thermo Fisher sample handling guides give you the dimensional tolerances to check against. For the vial and closure side specifically, OEM vial specifications and the vial closures reference data are where I send labs that are trying to match septum facing to their needle gauge. And when the method is a regulatory one, check the container requirements against U.S. EPA SW-846 before you assume the vial is the variable.
The rule I give every lab manager: buy on cost per reported result, measure the rerun and reject rates for two weeks before you switch suppliers, and never accept a unit price comparison that does not carry a puncture count and an extractables number next to it. The vial is 31 cents. The result is six dollars. Spend your negotiation time on the six.
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: