TL;DR: A COA that lists caliper and GSM without burst strength, moisture content, and Sheffield smoothness is incomplete — and an incomplete COA from a new board supplier is a qualification rejection in our incoming process.
TL;DR: In our 2024 incoming inspection log, 11 of 47 new substrate lots from unqualified suppliers failed on moisture content alone — all were between 9.2% and 11.8%, well above our 8.5% ceiling for offset-printed folding carton board.
What a Complete COA Must Declare — and What Suppliers Routinely Omit #
When we onboard a new board or substrate supplier, the first document we request is their Certificate of Analysis for each grade and caliper they’re proposing. Not a product brochure. Not a mill spec sheet with nominal values. An actual lot-level COA tied to a production run date and roll or sheet number.
A compliant COA for folding carton board in our qualification process must declare: grammage (GSM) ±4% tolerance per ISO 536, caliper (µm or mm) ±5% tolerance per ISO 534, Bendtsen or Sheffield smoothness on both print face and reverse, moisture content (%), burst strength (kPa) per ISO 2758, and for coated grades, coat weight per side in g/m².
What suppliers routinely omit: moisture content, reverse-side smoothness, and coat weight per side. These three omissions are not accidental. Moisture content above 8.5% in stored sheet stock causes dimensional instability on press — the sheets absorb or release moisture relative to pressroom conditions and the register shifts. Reverse-side smoothness matters because creasing performance on a folding carton blank is sensitive to the friction and compression behaviour of the back liner. Coat weight per side matters because a 70 g/m² coated board with 8 g/m² coating on one side and 4 g/m² on the other will print differently on each side — a problem we’ve seen cause visible ink density variation on two-sided cartons.
| COA Parameter | Minimum Acceptable Declaration | Our Rejection Threshold |
|---|---|---|
| Grammage (GSM) | Nominal ± 4% per ISO 536 | >±5% from stated nominal |
| Caliper | Nominal ± 5% per ISO 534 | >±6% or missing entirely |
| Moisture Content | % by weight, lot-specific | >8.5% for FBB/SBS/CUK |
| Burst Strength | kPa per ISO 2758 | <350 kPa for 300 gsm SBS |
| Sheffield Smoothness | Both sides, ml/min | >250 ml/min on print face |
| Coat Weight per Side | g/m², coated grades only | Absent on any coated grade |
A supplier who cannot or will not provide all six parameters on a lot-level COA gets flagged in what we call our QM-S02 Supplier Readiness Assessment. If the gap isn’t resolved within two revision cycles, the qualification does not proceed. We don’t treat this as a negotiation.
Where Incoming Inspection Catches What COAs Miss #
A COA is a supplier’s self-declaration. Our incoming inspection is independent verification, and the two rarely agree perfectly — which is the point.
Our protocol for all substrate and board lots entering production uses a 10-sheet random sample per pallet (minimum 3 pallets per lot) and tests against the same six parameters in the table above. We use our internal procedure code QC-07 for this sampling method, which aligns AQL 1.5 per ISO 2859-1 for critical defects and AQL 2.5 for major defects.
The failure scenarios we encounter most often fall into three categories.
The first is caliper compression from improper pallet stacking. A 400 gsm greyboard lot arrives COA-declared at 1.9 mm caliper. Our incoming measurement on the bottom sheets of a pallet that was double-stacked in transit reads 1.72 mm. That 9.5% reduction doesn’t recover — compressed greyboard does not spring back to original caliper under normal storage conditions. When that board goes to a rigid box line, the magnetic closure geometry is calibrated to a specific panel thickness. A 1.72 mm panel where 1.9 mm was specified will cause the lid to sit 0.18 mm low, making the closure feel loose and causing hinge crease splitting within roughly 40 open-close cycles. What to check: inspect bottom-pallet sheets separately and compare to top-pallet sheets in the same lot.
The second is moisture stratification within a roll or skid. Average moisture across a lot can pass at 8.1% while individual sheets at the skid perimeter — the ones that have been exposed to warehouse humidity variation — test at 10.4%. We run moisture readings on perimeter sheets explicitly because suppliers sample from the core. This is an important source of disagreement between COA moisture values and our incoming data. Sheets above 9.0% go into a conditioned hold room at 23°C / 50% RH for 48 hours before we retest. If they don’t come down to ≤8.5%, they’re rejected.
The third scenario is coating adhesion failure that doesn’t show on any standard COA parameter. A coated SBS lot with all COA values in range — including coat weight and smoothness — can still exhibit coating pick during printing if the coating binder has aged or the coat was applied at below-spec temperature. We catch this with a tape adhesion test (3M 610 tape, 180° peel) on press-conditioned sheets before releasing a lot to the print floor. This test is not part of ISO 536 or ISO 534 — it’s our own addition to the QC-07 protocol, and it’s caught four lots in 18 months that would otherwise have made it to press.
The consequences of releasing a failed lot to production are disproportionate. A substrate failure at the print stage means reprinting the job, but a substrate failure discovered post-converting (during erection or during the brand partner’s own quality check) means scrap at the finished-goods stage — labour, print, die-cutting, and finishing costs all absorbed with no recovery path.
Should You Specify Board Grade or Just a Performance Requirement? #
Specify both — but understand which one the supplier is actually held to.
Giving a supplier only a grade designation (“please supply FBB 280 gsm”) without performance thresholds means they can technically deliver compliant material that performs poorly in your application. Board from two different mills at the same stated GSM can differ in stiffness by 15–20% depending on furnish composition and formation, and neither is “wrong” by grade alone. Our qualification briefs always include minimum Taber stiffness values (MD and CD separately per TAPPI T566) alongside the grade designation. For a standard 350 gsm SBS folding carton, we specify ≥4.0 mN·m in MD and ≥2.5 mN·m in CD.
That said, performance-only specs with no grade reference create sourcing complexity. Suppliers need the grade anchor for procurement — the performance thresholds are the gate.
Specification Notes for Brand Partners #
When you brief us on a new packaging line requiring board sourcing, we need more than the grade and weight. The most useful information you can provide upfront is: finished carton dimensions, fill weight (for crush resistance calculation), surface finishing specification (lamination, UV, aqueous — each changes the substrate requirement), any food-contact requirement (FDA 21 CFR or EU 10/2011 compliance, which restricts certain board grades), and target cost tier.
The brief gap that causes the most sample iterations is fill weight without compression load context. We regularly receive briefs that state carton dimensions and board grade but don’t specify how the cartons will be stacked in retail or logistics — palletised 8-high, shelf-stacked 4-deep, or single-unit retail. The board selection for a candle carton shipped loose in a mailer is materially different from the same carton shipped 48-count on a pallet. Getting that context in the first brief cuts sample rounds from an average of 2.8 down to 1.5 in our project log.
Our standard substrate qualification timeline for a new project is 15–18 working days from confirmed brief to first structural samples, assuming the board specification is agreed and the supplier is on our approved vendor list. If we’re qualifying a new substrate supplier alongside the project, add 10–15 working days for the QM-S02 assessment and incoming inspection of the qualification lot.
Frequently Asked Questions #
How many COA fields are actually checked on incoming inspection — or do you just file the COA?
Every field in the COA is physically verified on incoming for any first-time lot from a supplier, and for any lot where the COA shows a value within 10% of our rejection threshold. For qualified suppliers with a 12-month clean record, we move to a reduced inspection frequency — testing 5 parameters on a 5-sheet sample per lot rather than the full protocol. The COA is never filed without at least a visual check and a spot caliper measurement.
Does FSC-certified board require different incoming inspection steps?
The FSC Chain of Custody standard (FSC-STD-40-004) requires that certified materials be received with a valid FSC transaction certificate from the supplier, not just a mill certification. We verify the transaction certificate number against the FSC certificate database before the lot enters our certified material storage zone. Physically, the incoming inspection protocol is identical — FSC status is a custody documentation requirement, not a performance override.
Our board supplier offers the same grade at two price points — is there a detectable quality difference?
It depends on where the cost difference originates. If the lower-cost option has a different furnish (recycled fibre proportion, pulp source), the difference will likely show in Taber stiffness and burst strength. If it reflects packaging or logistics cost only, physical performance may be identical. We always request COAs for both options before quoting, and if the stiffness delta exceeds 10% in either MD or CD, we consider them different materials for structural calculation purposes — not interchangeable variants of the same grade.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
On the reverse-side smoothness requirement — are you specifying a Sheffield range for the back liner, or just flagging suppliers who omit it entirely? We’ve had mills quote Bendtsen on face only and argue the reverse “isn’t a print surface,” which doesn’t hold once you’re running the blank through a high-speed folder-gluer with vacuum transport.
The moisture content point is the one that keeps biting us. Had a lot of 350gsm CUK come in at 10.4% from a new mill in Łódź and the register drift on the second pass was bad enough that we scrapped the first 600 sheets before catching it.
Coat weight per side is the one we added to our incoming rejection checklist after a bad run on 280gsm FBB for a watch sleeve — the supplier had declared total coat weight only, and we didn’t catch the asymmetry until we were seeing gloss differential on the tuck flap versus the front panel under store lighting.
Switching to lot-level COA requirements on incoming board added a qualification rejection step that cost us roughly 3-4 weeks of extra lead time per new supplier in the first year, so we built a 6-week buffer into any new substrate onboarding. The harder cost was the two carton runs we reprinted before we tightened the process — around $4,200 in waste on a 280gsm SBS job where the supplier’s coat weight declaration turned out to be face-only and the reverse printed noticeably flat on a two-color back panel.
Burst strength is the one we nearly skipped treating as a hard rejection criterion until a run of 300gsm SBS for a praline gift box completely fell apart on us — supplier COA listed burst at 310 kPa but the lot we received tested at 231 kPa in our incoming lab. Boxes were auto-erected on a Kliklok and the base panels were buckling under the chocolate weight before they even hit the wrapping line. We’d never have caught it without lot-level testing because the nominal spec looked fine on paper.
The coat weight asymmetry point tracks — we caught a 350gsm GC2 lot where the supplier had declared 12 g/m² total coat weight and it wasn’t until we ran gloss meter checks on both sides that the face/back split came back 8/4, which explained the ink trap issues we’d been chasing on the inner panel for two weeks.