Copper Concentrate Shipment: A Sampling Dispute
A port laboratory is asked to referee a 480 t shipment of copper concentrate split across 24 containers, each about 20 t. The seller's certificate states 25.6% Cu and 8.1% moisture. The buyer's sampler took three scoops from the top of three containers, combined them into a 150 g composite, and reports 23.9% Cu and 6.4% moisture. The contract rejects below 24.5% Cu and applies a moisture penalty above 8.0%. The vessel sails tomorrow. A duplicate split of the buyer's 150 g sample gives 24.3% Cu, a 0.4% absolute difference. It has rained intermittently, and the containers are open. The lab has a 1 m core sampler, a riffle splitter, a 1 kg moisture tin, and a balance readable to 0.1 g. The laboratory must decide what, if anything, can be reported and whether the buyer's rejection is defensible.
1. Critique the buyer's sampling and identify which error dominates the discrepancy. What can the receiving laboratory legitimately report from the existing sample?
Read the worked answer
The buyer's sample is not a probability sample of the shipment. Three top-of-container scoops sample only the surface, where copper concentrate segregates by particle size and density; fines, coarse particles, and moisture do not stay uniformly distributed. Three containers out of 24 also leaves most of the lot unsampled, and 150 g is a spot composite rather than a mass reduced by a proper splitter. The 0.4% absolute difference between duplicate splits is a preparation/heterogeneity indicator, but it cannot validate the primary sample.
Traceability is also weak: container IDs, sampling depth, time, weather exposure, seal numbers, and chain of custody are not described. The laboratory can report only that the buyer's sample is not fit for a contractual determination. The 23.9% Cu and 6.4% moisture values are indicative spot results, not a shipment assay, and they should not support rejection. The defensible action is a joint resample with a documented protocol, not a report based on the existing 150 g composite.
2. Design a defensible sampling and sample-preparation plan for a joint resample under time pressure. Specify increments, mass, moisture handling, splitting, and QC.
Read the worked answer
Take full-depth cores rather than surface scoops, at 3–5 positions per container, from all 24 containers. That gives 72–120 increments. Combine them into a composite of roughly 30 kg, seal it, and record container IDs, depth, time, sampler, and seal number. Take the moisture subsample immediately into a tared, sealed tin because concentrate loses moisture quickly; weigh before and after drying and report moisture on the as-received basis.
Reduce the composite with the riffle splitter in stages, grinding between stages as needed, to a 1–2 kg laboratory sample and retain a sealed umpire portion. Do not attempt to split a 150 g aliquot from unground material. Include a field duplicate from a second set of increments, a method blank, a certified reference material, and duplicate moisture determinations. Cover containers during sampling and record rain exposure. The final report should link every result to the sampling record and to the sealed laboratory sample.
3. The joint resample gives 25.2% Cu and 7.9% moisture. How should the lab report and how should the contractual decision be made?
Read the worked answer
Report the result with its basis and uncertainty, not as a bare number: for example, 25.2% Cu and 7.9% moisture on an as-received basis, with the sampling protocol, analytical method, duplicate data, and estimated expanded uncertainty. The Cu result is above the 24.5% rejection threshold, but the decision should use the contract's decision rule. If the expanded uncertainty is about ±0.3% absolute, the lower bound is about 24.9% Cu, still above 24.5%, so rejection on copper content is not supported.
For moisture, 7.9% is below 8.0%, but an uncertainty of ±0.3% spans the penalty threshold. The laboratory should state that the penalty threshold is not exceeded with confidence and refer the case to the contract's umpire or dispute procedure. The key point is that the original buyer's sample cannot be used to reject the lot; the joint resample, with documented traceability and uncertainty, is the defensible basis.