Technical resource
Most quality disagreements are not disagreements about chemistry. They are disagreements about who sampled what, at which point, and on which basis — and they surface after the material has moved, when nothing can be re-drawn. A protocol written before loading settles those questions while they are still cheap to settle; a certificate issued afterwards cannot.
Where the argument comes from
A disputed cargo figure is usually not the product of error. It is the product of two parties having measured different things on different bases, each according to a procedure nobody wrote down in advance.
Loading and discharge are different positions in the same journey. Both may be sampled, both may be right, and they answer different questions. Without the protocol naming which one governs, both parties will reasonably believe theirs is the one that counts.
The quality basis and the weight basis are separate mechanisms, and they have to describe the same material. Where they diverge, the parties can fully agree on the assay and still dispute the invoice.
Once the lot has moved, a disputed result can only be revisited through a sample that was set aside for the purpose. If no referee portion was held, the dispute has no material basis whatever the parties agree afterwards.
It does not make the figures agree. It makes them comparable — which is what allows two parties who disagree to resolve it rather than simply hold different numbers.
When the lot, the sampling point, the basis, the method and the retained portion are all recorded in advance, a disagreement becomes a question with an answer: re-run the retained sample, compare against a stated method, and apply the rule the contract already set. When those are absent, the same disagreement has no route to resolution and tends to be settled by whoever is more willing to absorb the loss.
Decision aid
Work down the protocol as you would write it. Every stage is a point at which responsibility changes hands, which is why a protocol that quietly skips one leaves a moment nobody owns. The tool tracks what is agreed and what is still open.
Before the material moves is the only time the parties can agree who attends and under what authority. Nominating after loading means the inspector documents what happened rather than witnessing it.
Which party nominates, the inspector's authority, and the notice period the other party gets.
A protocol governs a lot, and a lot is a defined quantity. Cargo, parcel, sub-lot and a day's production are different objects, and evidence drawn against one does not describe another.
What the lot is, its boundaries, and which document identifies it.
Tonnage and quality are settled on different evidence, and the two have to describe the same material. Where the weight basis and the sampling basis diverge, the parties can agree on quality and still disagree on quantity.
The weighing method, the basis it reports on, and who witnesses it.
Loading and discharge sit at opposite ends of the same movement. Both may be sampled, but the two samples answer different questions, and the protocol has to say which one the settlement rests on.
The sampling point, the mechanical or manual approach, and the plan the sampler works to.
A sample is divided into portions, and one of those portions is the only fallback if the result is disputed. Deciding afterwards which portion is the referee is deciding after the material has gone.
How the sample is divided, which portion is retained, how it is sealed, and who holds it.
A figure means little without the method that produced it. Two laboratories can both report total iron under the same parent standard while having run different procedures.
Which determinations, by which method and edition, on which portion, and by which laboratory.
Moisture is both a quality figure and the basis on which tonnage settles. Where it is determined, and on what basis it is reported, changes what the same material is worth.
The moisture method, the determination point, and whether figures are expressed wet or dry.
A sample is only as good as the record of who held it. Sealing and labelling at each handover are what make a later check possible at all — and what make a chain of custody provable rather than asserted.
The seal identifiers, the labelling convention, and who signs at each handover.
Deciding what happens after a disputed result is too late once the material has moved. The window and the fallback have to be written before the cargo sails, not argued when the figures arrive.
The claims period, what triggers a check, which result binds, and which party pays for the referee.
A count of the nine stages, based only on what you have selected. Nothing here supplies a tolerance, an increment count, a mass or an acceptance value.
Nothing reviewed yet. Start with the lot definition, because every later stage depends on what the lot actually is.
The tool reports coverage only. It does not tell you what a compliant protocol should contain, because that depends on the lot, the material and the terms your contract sets.
Reconciling the evidence
Each of these answers a different question, and each is held by a different party. Asking for them by name is more useful than asking for 'inspection documents', which tends to produce whichever one the other party happens to hold.
| Record | The question it settles | Normally held by |
|---|---|---|
| Nomination and authority | Who attends, on whose behalf, and with what power to act at the point of transfer. | Both parties, exchanged before the movement |
| Sampling plan or procedure | How the sample was to be taken, from what lot and at which points. | The party commissioning the sampling |
| Weighing or draught survey record | The tonnage, its basis, and who witnessed the measurement. | The weighing party and the attending inspector |
| Sampling and preparation record | What was actually done, when, and by whom — as distinct from what the plan said. | The sampler or the appointed inspector |
| Chain-of-custody record | Who held the sample between drawing and testing, and whether the seal was intact. | Whoever signs at each handover |
| Determination report and certificate | The figures measured, the methods applied, and the party that issued them. | The testing laboratory |
Holding every document is not the same as having a reconciled set. A sampling record that names a lot the weighing record describes differently leaves the two describing different cargoes, and no amount of completeness in either will bridge that.
The same applies down the chain of custody: where the seal identifier on the handover contradicts the one on the preparation record, the question is no longer what the sample contained but whether it is the same sample — and that question is usually unanswerable after the fact. Reconciling the identifiers at the time is the only cheap moment to do it.
The sample itself
A procedure can be followed exactly and still produce a sample that cannot be relied on. The failures happen in the handling, between the point of drawing and the point of testing.
A reference sample has to be divided, sealed and traceably held while the sample is intact. Once testing has consumed the material, no later arrangement can create a fallback that was not set aside. This is the single most consequential act in the whole protocol, and it costs nothing to include.
Drying, crushing, dividing and reducing a sample to laboratory size is the stage at which a sample most easily stops representing the lot. It has to be done to a stated procedure by a named party, and the record of it has to travel with the result.
Chain of custody is only meaningful if it is recorded: a seal identifier, a labelling convention, and a signature at each handover. Without those, custody is an assumption, and an assumption cannot be checked when it is questioned.
No increment count, mass, tolerance or acceptance limit is published on this page. Those are set by the agreed plan for a specific lot, and they belong in the contract and the sampler's procedure rather than in a guide.
Where this sits
This guide describes the protocol around the sampling. The documents it points to describe the methods themselves — and none of them substitutes for the others.
| Scope | What it answers | The limitation to hold in mind |
|---|---|---|
| This guide — the protocol | Who samples what, at which point, on which basis, and how a disagreement is resolved. | It governs the workflow. It performs no measurement and reproduces no method's procedures. |
| The sampling method | How a lot is sampled and how the sample is prepared and reduced before determination. | It produces the sample. It reports no figure, and it does not decide which party's result binds. |
| Chemical composition | How the constituent figures are established, and what each determination covers. | It describes the determinations taken on the sample. It is only as representative as the sampling behind it. |
This page publishes no certificate, no test result and no analysis for any material, and makes no claim about any inspection party, laboratory or accreditation.
Procurement navigator
A protocol is settled against the destination as well as the material, because the receiving end is where the second set of figures is produced. Choose the receiving market to carry the requirement into the right context.
Each market page covers import planning for that destination. This guide covers the sampling and inspection protocol a shipment into any of them has to settle.
Questions
No, and treating one as the other is the failure this page exists to prevent. A certificate records what a laboratory found on a sample it was given. It says nothing about how that sample was drawn from the lot, whether the sample was representative, or whether the parties agreed to the procedure. A protocol establishes those things in advance; a certificate is one output of it.
No. Inspection attendance is a witnessing function — it puts a qualified party at the point of transfer and creates an independent record. The sampling plan is a separate instrument that defines the lot, the approach, the increments and the preparation. Attendance without an agreed plan documents a procedure nobody specified.
Because it only has value if it was set aside while the sample was still intact. A reference portion has to be properly divided, sealed and traceably held — all of which are acts taken at the time of sampling. Once the lot has moved and the sample has been consumed by testing, no arrangement made afterwards can create a fallback that did not exist.
Yes, and this is more common than outright error. Where the two parties sample at different points, apply different methods, or report on different bases, two defensible results can disagree without either being wrong. That is why the protocol has to state which basis governs, rather than leaving the parties to compare figures that were never measuring the same thing.
The lot definition and its boundaries; the nomination and authority of any inspector; the weighing method and basis; the sampling point and the plan the sampler works to; how the sample is divided and which portion is retained as referee; the determinations required, each with its method and edition; the moisture method, determination point and reporting basis; the sealing and labelling convention with signatures at each handover; and the claims window with the route to be followed if a result is disputed.
No. It describes the workflow a protocol is built from and the stages at which responsibility changes hands. No tolerance, increment count, mass, assay value or acceptance limit is published here, because every one of them is agreed between the parties for a specific lot and contract.
Source
The sampling and preparation standard is published by the International Organization for Standardization. Check the current edition and status there before relying on any particular clause.
Qualified enquiry
Tell the commercial desk which material you are buying, which sampling and inspection basis the contract will use, and which records you expect with the shipment. The enquiry reaches the same desk that answers the specification, so the offer comes back against the protocol you set.
Price on request — no figure is published on this page.
Commercial desk reachable Monday to Saturday, 8.00 AM to 6.00 PM.
