Start with the question the sample must answer
A laboratory result can look precise and still tell you little about a shipment if the material sent for testing was not representative. Before sampling begins, agree what decision the result must support: for example, a cargo-property assessment, a quality check or a comparison between loading and discharge. Then identify the consignment or lot being sampled and confirm which cargo-specific, contractual and regulatory procedures apply. Procedures are not automatically interchangeable between commodities. The U.S. Federal Grain Inspection Service (FGIS), for example, publishes procedures for official grain inspection, while the IMSBC Code addresses solid bulk cargoes other than grain; grain has a separate international carriage code. Use these sources as guidance for their stated scopes, not as a substitute for checking the rules applicable to the shipment. [3]
Grain Inspection Handbook, Book I: Sampling · International Maritime Solid Bulk Cargoes (IMSBC) Code
Represent the lot—not just the easiest point to reach
For solid bulk cargoes covered by the IMSBC sampling provisions, physical-property tests are meaningful only when performed before loading on truly representative test samples. The IMO guidance says the method should account for the material type and particle-size distribution, variability in composition, how the cargo is stored and transferred, any chemical hazards, and the property being measured. That is why the sampling plan should be settled before a stockpile is moved or a conveyor starts: the collection method must match both the cargo and its route to the ship. [0]
For grain, FGIS describes systematic sampling from stationary lots or from the moving stream. Its explanation of diverter sampling shows the principle: take repeated portions at intervals while the lot is being conveyed, then combine those portions as a composite. A single convenient scoop may miss differences across a lot; the pattern and procedure should be selected for the actual inspection and cargo rather than improvised at the hatch. The FGIS procedures are for its official grain-inspection system, so they are an example—not a universal rule for every commodity or port. [1]
2023 Amendments to the International Maritime Solid Bulk Cargoes (IMSBC) Code — Resolution MSC.539(107) · Sampling
Treat visible differences as separate questions
A representative average is not the same as a careful examination of a visibly unusual portion. Before sampling, the IMO guidance calls for a visual inspection of the consignment where practicable. Material that appears contaminated or substantially different in character or moisture from the rest should be sampled and analysed separately; results may inform whether that portion is suitable for shipment. In practice, record where the difference was seen and keep its sample identity distinct. Folding an unusual patch into a large composite could conceal the very condition that needs investigation. [0]
This does not mean every variation proves a cargo defect. It means the sampling record should preserve the distinction between the general lot and any suspect area, so the responsible parties can assess each result in context. Agree in advance who can inspect and sample the stockpile, how access will be arranged, and who will receive the observations. For Group A cargoes, the cited IMO provisions specifically call for the shipper to facilitate the ship’s nominated representative’s access to stockpiles for inspection, sampling and testing. [0]
Protect the sample’s identity after collection
Collection is only one link in the evidence chain. Record the lot identity, sampling location or method, time, sampler and any unusual observations; label containers so the sample can be matched to the correct consignment and test request. The FGIS handbook emphasizes accurate records of lot identity and condition, and says official samples must remain secure and under the sampler’s control to guard against manipulation, substitution or careless handling. Those controls are a useful operational model, although its formal requirements apply to the official U.S. grain-inspection service. [2]
At the planning stage, decide who will retain, test and distribute each sample, and how any split or reserve sample will be identified. Keep documented samples tied to the specific lot and purpose they represent; do not assume that a certificate alone explains how the material was selected. If the sampling process, cargo identity or custody becomes uncertain, flag that limitation before treating the result as decisive. A clear record cannot make a poor sample representative, but it can show what was sampled, how it was handled and what conclusions the result can reasonably support. [2]