I still remember the purchase order that taught me to stop buying vials on unit price. We switched to a supplier quoting $38 per 100 for 2 mL clear glass vials instead of our usual $52. Six weeks later, our autosampler was jamming twice a shift, the septa were coring after 40 injections, and our QC lab had logged 11 out-of-spec repeat runs. The “savings” of $14 per hundred disappeared in a single afternoon of rework. That is the gap between unit price and landed cost, and it is the number procurement should be negotiating against.
Landed cost, in the chromatography consumables category, means the total cost of getting a compliant result from a part, not the invoice price of the part. That includes freight, duty, inspection, incoming QC, line stoppage, rework, and the cost of a failed analytical batch. I have watched purchasing teams save 18% on the PO and lose 30% on the total cost of ownership. This article is about how to calculate the real number before you sign.
For a typical HPLC or GC lab, consumables are roughly 8 to 15 percent of the operating budget. But the cost of a consumable failure is not proportional to its price. A $0.40 septum that fails mid-sequence can invalidate a 96-well plate of samples worth $4,000 in labor and instrument time. That asymmetry is the whole game.
I break landed cost into five layers. Procurement usually only sees the first one.
- Acquisition cost: unit price, freight, duty, and currency exposure.
- Quality cost: incoming inspection, certificate of analysis (CoA) review, and rejected lots.
- Operational cost: autosampler jams, septum coring, column failures, and repeat injections.
- Compliance cost: traceability, documentation, and audit findings.
- Inventory cost: carrying cost, shelf life, and obsolescence on slow-moving SKUs.
When you add these up, the part that costs $0.40 at the loading dock often costs $2.10 by the time it has done its job. That is a 5x multiplier, and it is why I treat unit price as a starting point, not a decision.
The acquisition layer: freight, duty, and the breakage you did not budget for
Glass is heavy and fragile. A pallet of 2 mL vials in cases of 100 weighs roughly 12 to 14 kg per case, and freight is charged on dimensional weight. If you are buying from an overseas supplier, the freight per case can be 20 to 35 percent of the product cost. That is before duty, which for borosilicate glass labware into the U.S. is typically in the 3 to 7 percent range depending on classification and country of origin.
Breakage is the hidden line item. I budget 1.5 to 3 percent loss on inbound glass shipments from suppliers who pack in single-wall cartons, and under 0.5 percent from suppliers who use partitioned foam and double-wall cartons. Over a year of 40,000 vials, that difference is 600 to 1,000 units, which is real money at any price point.
What to ask before you accept a freight quote
- What is the packing method? Partitioned foam, molded pulp, or loose in a bag?
- Is the carton double-wall corrugated with a burst strength of at least 200 psi?
- Does the supplier ship on a pallet with stretch wrap, or as individual cartons?
- What is the declared value for insurance, and who files the claim on breakage?
I have seen a supplier win a bid on price and then lose the account because they would not file breakage claims. The buyer ate the loss. That is a landed cost the PO never showed.
The quality layer: incoming inspection is not free
Every lot you receive from a new supplier needs incoming inspection until you have a track record. For vials and closures, that means at minimum a dimensional check, a visual check for chips and foreign matter, and a functional check on the autosampler. If you are running USP methods, you also need to verify the CoA against your specification, and you may need to run a suitability injection.
I budget 0.5 to 1.5 labor hours per lot for incoming inspection on a new supplier, dropping to 0.1 hours per lot once a supplier has delivered 20 consecutive conforming lots. That is the real value of a qualified supplier: not the price, but the inspection you can stop doing. A supplier that saves you $300 on a PO but forces you to inspect every lot has cost you more than they saved.
The CoA is a document, not a guarantee
A certificate of analysis is only worth what you can verify. I have received CoAs that listed “pass” for dimensional tolerance without stating the measurement method or the sampling plan. That is not a CoA, it is a wish. Ask for the actual measured values, the sample size, and the acceptance criteria. A real CoA for a 2 mL vial will state the neck OD (typically 11.6 ± 0.2 mm for a 11 mm crimp finish), the height, the inner diameter, and the lot number traceable to the glass batch.
For closures, the CoA should state the septum thickness, the durometer (typically 45 to 55 Shore A for silicone), and the coating or lamination. If the supplier cannot provide that, the “quality” is marketing.
The operational layer: where the money actually leaves
This is the layer that eats budgets. A failed injection is not a failed part; it is a failed sequence, and sequences are expensive.
Septa coring and the puncture count that matters
A septum is a wear part. The question is not whether it cores, but when. For a standard silicone/PTFE septum on a 11 mm crimp vial, I expect 40 to 60 punctures before coring becomes a risk with a 23-gauge needle. With a 22-gauge needle, expect 25 to 40. If you are running a 96-well plate with multiple injections per well, that math changes fast.
The failure mode is specific: a cored septum drops a silicone plug into the vial, the autosampler needle pulls it up, and the injection port or the column frit catches it. The result is a partial blockage, a rising backpressure, and a chromatogram with a splitting peak. I have traced a 14% RSD on a replicate standard to a single cored septum in a 60-injection sequence.
If you are buying closures, buy from a supplier who states the puncture count on the CoA and who can tell you the needle gauge they tested with. The vial closures specification sheet should list durometer, thickness, and the recommended crimp or torque. If it does not, ask.
Crimp torque and the seal that leaks slowly
For 11 mm crimp vials, the crimp should be set so the seal is compressed but the septum is not deformed. I target a crimp height of 5.5 to 5.9 mm measured at the crimp ring, and I verify with a crimp gauge, not by feel. Under-crimping gives you evaporation and a drifting internal standard. Over-crimping gives you a deformed septum and a coring risk.
For screw-thread vials, the closure torque matters. A 9 mm screw cap should be tightened to roughly 0.8 to 1.2 N·m. I have seen labs hand-tighten to 2 N·m and split the cap, and I have seen labs under-tighten and lose 3% of the sample to evaporation over a 24-hour run. Both are landed costs that never appear on the PO.
Column failures and the consumables that cause them
A column is the most expensive single consumable in the lab. A typical analytical column runs $400 to $700, and a guard column runs $30 to $60. The guard column exists to protect the analytical column, and it works only if you change it on schedule. I change guard columns every 100 to 200 injections for dirty matrices, and every 500 injections for clean ones. A lab that skips guard changes to save $40 will replace a $600 column twice as often. That is a landed cost decision disguised as a consumable decision.
If you are sourcing columns, the Agilent chromatography consumables and columns catalog is a useful reference for phase chemistry and dimension options. The Restek technical resources are equally useful for understanding retention and bleed. Use them to specify, then buy on landed cost.
The compliance layer: traceability has a price
If you are running methods under U.S. EPA SW-846 or USP <621>, the consumable is part of the method. That means the vial, the septum, and the column are documented in your SOP, and a substitution requires a change control. A cheaper vial that is not on your approved list is not cheaper; it is a compliance project.
I have seen an audit finding triggered by a vial substitution that was never documented. The lab saved $900 on the PO and spent three weeks on a corrective action. That is the extreme case, but the principle holds: the landed cost of a consumable includes the cost of the documentation that lets you use it.
For labs that outsource method development or validation, the analytical testing partner’s approved consumables list is part of your contract. Buying outside that list to save on unit price can void the validation. Check before you switch.
A landed cost comparison you can actually run
Here is the checklist I use when comparing two suppliers for the same SKU. The numbers are from a real comparison I ran for 2 mL clear glass vials with 11 mm crimp closures, 10,000 units per year.
| Cost line | Supplier A (low unit price) | Supplier B (qualified) |
|---|---|---|
| Unit price per 100 | $38 | $52 |
| Freight per 100 | $9 | $6 |
| Duty per 100 | $1.90 | $2.60 |
| Breakage loss (percent) | 2.5% | 0.4% |
| Incoming inspection hours per lot | 1.2 | 0.1 |
| Autosampler jams per 1,000 injections | 6 | 1 |
| Repeat sequences per year | 9 | 2 |
| Cost per repeat sequence | $420 | $420 |
| Estimated annual landed cost | $6,940 | $6,180 |
Supplier A looks 27% cheaper on unit price and is 12% more expensive on landed cost. The gap is entirely in the operational and quality layers. That is the number to negotiate against, and it is the number most POs never calculate.
How to build this table for your own lab
- Pull 12 months of repeat injection logs and autosampler error logs.
- Assign a fully loaded labor rate (salary, benefits, overhead) to each hour of rework.
- Count rejected lots and the labor to inspect and return them.
- Add the cost of any failed analytical batch (instrument time, standards, labor).
- Divide by annual unit volume to get a landed cost per 100 units.
If you do not have the logs, start keeping them. The first year is the hardest; after that, the comparison takes an afternoon.
The wholesale and OEM dimension
If you are buying in wholesale volumes, the landed cost calculation changes because the failure modes scale. A single bad lot of 50,000 vials is not a rework problem; it is a production stoppage. I have seen a wholesale buyer reject an entire pallet because the neck OD was out of tolerance by 0.4 mm, which was enough to cause autosampler gripper failures on 3% of vials.
OEM buyers have a different problem: the consumable is sold under their brand, so the failure is theirs. That is why I recommend OEM vial sourcing only from suppliers who will share the glass batch records and the dimensional data, not just the CoA. The landed cost of an OEM program includes the cost of a recall, and that is not a line item you want to discover.
For wholesale buyers, the chromatography supplier relationship is a supply-chain relationship. Ask for the production schedule, the lead time on a reorder, and the minimum order quantity. A supplier who cannot commit to a lead time is a supplier who will cost you a line stoppage.
Inventory and shelf life: the cost of holding
Consumables have a shelf life. Septa degrade, especially silicone, and adhesive-backed septa lose tack after 18 to 24 months. Vials are stable, but the packaging is not; a carton that has been on a shelf for three years has absorbed moisture and the vials may need drying before use.
I budget a carrying cost of 20 to 25 percent per year on inventory. That means a $10,000 stock of vials costs $2,000 to $2,500 a year to hold. If you are buying in bulk to get a price break, make sure the break exceeds the carrying cost. A 10% discount on a 12-month supply is not a saving; it is a 10% discount against a 22% carrying cost.
Reorder point math that respects lead time
For a consumable with a 6-week lead time and a 4-week usage cycle, the reorder point is 1.5 times the usage cycle, not 1.0. That buffer is the cost of lead time variability, and it is cheaper than a line stoppage. I have seen labs run to zero because the reorder point was set at the usage cycle, and the supplier’s lead time slipped by two weeks.
Troubleshooting checklist: when the landed cost goes wrong
When a consumable problem shows up, the first question is not “which supplier” but “which layer.” Use this checklist to isolate the failure.
- Peak splitting or tailing: check the guard column and the column frit before blaming the vial. A cored septum can block the frit.
- Rising backpressure: check the inlet liner and the guard column. A liner with a glass wool plug can shed fibers.
- Autosampler gripper errors: measure the vial neck OD and the cap OD. A 0.3 mm deviation is enough to cause intermittent failures.
- Evaporation or drifting internal standard: check the crimp height or the screw torque. Under-crimping is the usual cause.
- Out-of-spec repeat injections: check the septum puncture count and the needle gauge. A 22-gauge needle on a thin septum cores early.
- Foreign matter in the vial: check the supplier’s wash and inspection process. A vial that is not washed before packing can carry glass dust.
Each of these has a landed cost. The checklist is how you find it before it finds you.
Recommendation: negotiate on landed cost, not unit price
My recommendation is to build the landed cost table for your top 10 consumable SKUs, then use it in every supplier negotiation. The specific parameters I would put in the specification are:
- Vial neck OD: 11.6 ± 0.2 mm for 11 mm crimp finish; 9.0 ± 0.2 mm for 9 mm screw finish.
- Septa durometer: 45 to 55 Shore A for silicone; state the puncture count at a specified needle gauge.
- Crimp height: 5.5 to 5.9 mm, verified with a gauge, not by feel.
- Screw torque: 0.8 to 1.2 N·m for 9 mm closures.
- Guard column change interval: 100 to 200 injections for dirty matrices; 500 for clean.
- CoA content: measured values, sample size, acceptance criteria, and lot traceability to the glass batch.
- Lead time: committed in writing, with a reorder point of 1.5 times the usage cycle.
Put those in the RFQ and the landed cost comparison becomes a conversation about data, not opinion. The supplier who can meet them will usually be within 10 to 15 percent on unit price and 10 to 20 percent better on landed cost. That is the supplier worth keeping.
The point is not that cheap consumables are bad. The point is that the invoice price is a starting bid, and the landed cost is the final score. Procurement teams that learn to calculate the second number stop paying for the first one twice.
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: