Fabric Knowledge
What Grade A Means on a Fabric Roll
Grade A is not an industry standard. It is whatever the mill wrote down. What the A/B/C count measures, what it misses, and how to write it into an order.
By the technical team at Shandong Orenda New Materials Co., Ltd. Specification questions: jane@orendascreen.com. Published 13 November 2026.
A purchase order goes out with three words on it: Grade A only. Both sides read it and both sides believe something has been agreed. Then the first container is opened, someone counts marks on a roll, and the argument starts — because the buyer pictured zero defects and the supplier was counting against a threshold nobody wrote down.
Grade A is not an industry standard. There is no international body that defines A, B and C for coated woven fabric. Every mill that uses the letters defines them itself, in its own documents, counting its own unit. Two suppliers can both promise Grade A and be promising different things, and neither is lying.
This article sets out what the letters mean in our specification, what a count can and cannot tell you, what happens when defects cluster, and how to put the definition into an order so that it survives the first delivery.
Key takeaways
- Grade A is a mill's own definition. Ask for the unit, the defect definition and the threshold, in writing, before the order.
- In our packing specification the unit is 30 m and the count is marked defects: three or fewer is Grade A, four is Grade B, five is Grade C.
- A count is auditable. You can unroll the roll and count the same red marks the inspector counted, which an adjective like "first quality" never allows.
- A count does not see severity, distribution or position. Three small knots spread over 30 m and three long crease bands are the same number and very different rolls.
- Where defects cluster inside one metre, the section is cut out and one metre is added free of charge, so a run of bad fabric does not reach you as a short roll.
Three different things called Grade A
The phrase turns up in three unrelated systems, and most disputes come from one side using one and the other side using another.
What the buyer may mean | What it actually measures |
|---|---|
A per-roll defect count against a mill's own threshold | Marks on one roll, in a defined unit of length |
Lot acceptance sampling, such as AQL to ISO 2859-1 | How many defective units are tolerated in a consignment, not how good one roll is |
Appearance against an approved sample or a signed swatch | A visual comparison, with no countable criterion at all |
They answer different questions. A roll can pass a per-roll count and still fail an AQL lot if too many rolls in the container are marginal. A lot can pass an AQL sample and still contain one roll that a count would grade C. And an approved sample settles colour, not defects.
So the useful question is never "is this Grade A". It is: what is your grade, measured on what length, counting what, and what happens to a roll that falls outside it.
How our grading is defined
Defects are marked in red during production, at the point they are found. The marking is not cosmetic. It means the defect cannot be missed further down the line, downstream cutting can avoid it, and the count becomes a record rather than someone's memory of a roll they looked at once.
At final inspection the marked defects are counted per 30 m:
Grade | Marked defects per 30 m |
|---|---|
A | 3 or fewer |
B | 4 |
C | 5 |
Thirty metres is the unit because it matches how the material is sold and handled. Roll lengths run from 30 m to 50 m, so a 30 m unit is one to one-and-a-half units per roll — long enough to average out a single unlucky metre, short enough that one bad patch cannot hide inside a very long roll.
The same construction is used across our 3% range, which is the series most orders specify, and the count applies identically to the 5% weave.
A count can be checked; an adjective cannot
This is the practical reason to prefer a number. If a supplier says "first quality", the only way to test the claim is a judgement call, and judgement calls get settled by whoever is more senior in the meeting. If a supplier says "Grade A, three or fewer marked defects per 30 m", the buyer can open the roll, count the marks and compare.
That changes the shape of the conversation after a bad delivery. Instead of "this fabric is not acceptable", which invites a defence of the whole batch, the claim becomes specific: this roll carries nine marked defects in its first 30 m, which is outside the grade stated on the order. That is a fact with a number attached, and it is much harder to argue with than an impression.
What the count does not see
A count is a good instrument with a known blind spot. It records how many. It does not record what, where or how grouped.
The count records | The count leaves out |
|---|---|
Number of marked defects per 30 m | The size and type of each defect |
Whether the roll is inside a grade threshold | Whether the defects sit together or are spread out |
A number that can be re-checked | Where the defects sit relative to the core and the cut plan |
Three marks could be three short thick yarns that a cutter steps around in ten seconds. They could also be three crease bands running the full width, each one destroying a usable metre. Both are "3" and both are Grade A. The difference only shows up when the roll is unrolled.
This is why the grade is a starting point and not a substitute for looking. If a defect type matters to your process — a crease that will not flatten, a band that will show through a backlit panel — say so in the enquiry. A count threshold will not capture it, and a supplier who only reports the count is not hiding anything; they are reporting the only thing their system measures.
When defects cluster, one metre is added
Defects that bunch together are worse than the same number spread out, because a cluster removes a continuous length from the roll rather than a series of points a cutter can work around. Our packing specification deals with that case directly: where multiple defects are concentrated inside a one-metre section, that section is cut out of the roll, and one metre is added free of charge.
The point of the rule is that nobody should receive a roll that is both defective and short. Removing the bad metre is the quality decision; adding a metre back is what stops the quality decision from becoming a delivery shortfall.
It is a good question to put to any supplier, including us: what do you do when defects cluster? An answer in metres is a policy. An answer in reassurance is not.
For the physical causes behind those defects — a band across the width, a crease near the core, and how to tell which is which — see what a delivered roll is telling you. That article covers diagnosis. This one covers the count.
What to check on receipt
The check needs a bench, a light and a tape measure. It takes about twenty minutes for a first roll.
- Open the ends first. Roll out the first three metres and the last three metres. Start and stop effects sit at the ends of a run, so this is where a marginal roll shows itself earliest.
- Count the red marks over a measured 30 m. Do not estimate. Marks are placed for exactly this purpose, and counting them is the one part of grading a buyer can reproduce.
- Note the type of each one before counting the next. A tally of types is what turns a number into a decision.
- Measure the gaps. Marks spaced evenly at a constant pitch mean something different from marks at random intervals, and the distinction is worth recording even if you do not act on it.
- Record the delivered length against the ordered length. If the roll is short, ask whether a clustered section was removed — that is a legitimate reason, and it should be declared rather than discovered.
Writing the grade into the order
The clause does not need to be long. It needs four things, and leaving out any one of them reopens the argument the clause was meant to close.
Clause item | What to write |
|---|---|
Unit and method | Defects counted per 30 m, marked at the point of production |
Threshold per grade | The count that defines each grade you will accept |
Consequence of failure | What happens to a roll outside the accepted grade: replace, credit, or reject |
Clustered defects and length | What happens when a section is cut out, and how the delivered length is treated |
A workable sentence runs like this: defects are marked in red during production and counted per 30 m at final inspection; three or fewer marked defects per 30 m is the accepted grade; a roll outside that count is replaceable on production of photographs and the mark count; where defects cluster within one metre, the section is removed and one metre is added free of charge.
That last element is the one most orders leave out, and it is the one that decides whether a legitimate repair turns into a short delivery.
Questions that come up before the order
Is Grade A the same at every mill?
No. The letters are a local convention, not a standard. Some mills grade per 30 m, some per 100 m, some per roll, and the thresholds differ. A supplier who cannot state the unit and the count in one sentence has not defined the grade at all.
Does Grade A mean zero defects?
No. It means the count is inside a threshold. A Grade A roll can carry three marked defects per 30 m. If your process cannot tolerate any defect in a given panel, that requirement belongs in the order as a separate line, and it needs to be agreed against a cutting plan rather than a grade.
How is this different from AQL inspection?
AQL sampling accepts or rejects a consignment by inspecting a sample of units, and it is a lot-level judgement. Per-roll grading describes each roll. Buyers who already run AQL on incoming goods can use a roll grade as the unit-level detail underneath it; the two are complements, not alternatives.
What if a roll arrives below the stated grade?
Put the count and photographs in the claim, per roll, against the clause. A claim stated as a count is answerable. A claim stated as dissatisfaction usually ends in a replacement for the wrong reason, which is a worse outcome for both sides.
Do we get the inspection record with the roll?
Ask for the mark count per roll to travel with the shipping documents. If the count is recorded at final inspection, reporting it costs nothing; if it is not recorded, that is useful information about the supplier.
Our take
The letter is worth less than the definition behind it. A buyer who asks "is this Grade A" gets a yes, and a yes is not a specification. A buyer who asks "what is your unit, what counts as a defect, what is the threshold, and what happens to a clustered patch" gets four answers that can be written into an order — and an order is the only place a grade has any force.
The second point is that a count is not a verdict. It is the cheapest reliable starting point for one. Read the count, then look at the roll, then decide whether the defect types matter for the panel you are cutting. A Grade A roll with three crease bands is a different problem from a Grade A roll with three thick yarns, and no threshold will tell you which one you have.
Sources
- [S1] Orenda packing specification, Packing Details — defect marking in production, the 30 m grading unit, the A/B/C thresholds, and the clustered-defect rule with one metre added free of charge. Internal document, 2026.
- [S2] ISO 2859-1 — sampling procedures for inspection by attributes, referenced here only to distinguish lot acceptance from per-roll grading. Not used as a claim about our own inspection.
What we do not claim
- We do not claim that A, B and C are recognised grades outside our own specification.
- We do not publish a defect rate, a rejection rate or an AQL level. Those are not measured in a form we can support.
- We do not state that a Grade A roll is free of defects. It is inside a count threshold.
- We do not name or compare another manufacturer's grading system.
- We do not state a warranty term or a service life.
Data notes
Figure | Value · source · type |
|---|---|
Grading unit | 30 m — [S1], stated |
Grade A threshold | 3 or fewer marked defects per 30 m — [S1], stated |
Grade B threshold | 4 marked defects per 30 m — [S1], stated |
Grade C threshold | 5 marked defects per 30 m — [S1], stated |
Clustered defects | Section cut out within a 1 m span, one metre added free of charge — [S1], stated |
Roll length range | 30–50 m — [S1], stated |
Marking method | Red mark applied at the point of discovery in production — [S1], stated |
