Skip to content
Get On My Space Get On My Space Get My Coverage Plan

Sampling and test flow in food contact testing: part 3

The short answer

Advanced sampling is about splitting correctly: mixed lots into sub-lots, multi-material articles into component sets, and multi-market protocols into separately judged verdicts on shared specimens. Part 3 of this series gives the decision rules for each split, the documentation that keeps complex programs traceable, and the coordination routine that aligns factory, sampler, and laboratory. Done well, one organized sampling round answers every compliance question the shipment will face. Done poorly, one blended bag of specimens answers none of them.

Key takeaways

  1. Mixed lots must be split into defined sub-lots sampled and reported separately, never blended into averaged specimens.
  2. Sampling plans rest on acceptance sampling logic, where random selection and risk-based sizes make lot inferences defensible. [1]
  3. Combined EU and US protocols can share specimens but still need separate verdicts, a transparency discipline reflected in global technical-barriers rules. [2]
  4. Sealed spares, stored properly and documented at sealing time, are the only specimens that can support a credible retest.
  5. The sampling report should let a stranger reconstruct what was tested, how, and why, years after the shipment.

Parts 1 and 2 of this series covered lot definition, sample sizes, and the journey from factory to laboratory bench. Part 3 addresses the situations where textbook sampling meets messy reality: mixed lots, multi-material articles, combined EU and US protocols on shared specimens, and the documentation that keeps complex programs traceable. These are the cases where a rigid plan breaks and a reasoned plan pays.

Complexity in sampling is not sophistication for its own sake. Every additional sub-lot, material, or market in the scope is a separate compliance question that needs its own specimens and its own verdict. Programs fail when buyers try to answer several questions with one bag of specimens, producing averaged results that describe nothing shippable. The art is in splitting correctly, then documenting the splits so the file stays readable.

This final installment gives decision rules for mixed lots, combined protocols, and the sampling report, plus the coordination routine that keeps factory, inspector, and laboratory aligned.

Splitting mixed lots and multi-material articles

A mixed lot is several lots wearing one packing list. Different resins, different colors with different ink loads, different production dates, or different component suppliers each create a sub-lot with its own risk profile, and each deserves its own specimens, identifiers, and verdict. The sampling plan should name the splits before anyone opens a carton, using production records rather than guesswork. Convenience blending, taking a few pieces of each variant into one bag, produces a laboratory average that matches no shippable unit and cannot support any release.

Multi-material articles need the same splitting logic at component level. A bottle is body, cap, gasket, and possibly coating or liner, and each food-contact component needs specimen coverage in the laboratory scope. The sampler collects each component set separately with its own identifier, even when the components arrive assembled, and the laboratory reports each against its applicable limits. UTS Inspection UTS Inspection sampling teams enumerate components from the bill of materials before the visit so nothing small and risky escapes the scope.

The statistical foundation is unchanged by the complexity. Acceptance sampling logic describes how random draws from a defined lot support inferences about that lot, and only about that lot, which is why each sub-lot needs its own random draw; background on that logic explains the operating-characteristic reasoning behind sample sizes. [1] Borrowed specimens across sub-lots borrow credibility that is not theirs.

Running combined EU and US protocols on shared specimens

Combined protocols economize on specimens, not on verdicts. The laboratory selects conditions satisfying the stricter side wherever simulants and temperatures overlap, runs the shared exposures, then judges the results separately against EU limits and FDA expectations. The sampling plan must therefore confirm in advance that the shared specimen count covers the replicate needs of both protocols, including spares. Discovering mid-program that one protocol needed more replicates means a second sampling visit.

Illustration: Running combined EU and US Decorative illustration for the section "Running combined EU and US"; visual only, carries no data.

Coordination is a three-party exercise. The factory confirms which sub-lots are finished and available, the sampler confirms selection method and sealing, and the laboratory confirms the combined condition matrix and intake requirements. A short alignment call before the visit, with the matrix in writing, prevents the familiar failure where specimens arrive but the combined protocol was never agreed. Confirm courier timing against laboratory stability windows at the same time.

Transparency rules for technical requirements on traded goods reinforce the discipline of separate verdicts and open procedures, as described in the overview of the global technical-barriers framework. [2] For the buyer, the file should show one sampling record, one condition matrix, and two clearly separated compliance verdicts. Either market can then follow the reasoning without untangling a blended conclusion.

Sampling documentation that survives scrutiny

Record elementContentsWho relies on it later
Lot definitionProduct, quantities, production dates, sub-lot splitsLaboratory, buyer, customer auditor
Selection recordMethod, carton numbers, random basis, witnessesAnyone questioning representativeness
Sealing recordIdentifiers, seal numbers, photographs, spares locationRetest decision and laboratory intake
Transport recordCourier, dates, storage conditions, receipt confirmationStability and traceability queries
Protocol matrixOrdered tests, conditions, replicates per sub-lotLaboratory and report readers
Deviation logAny departure from plan with reason and approvalRegulators and key accounts

Coordination routine for complex programs

Complex sampling fails on communication, not on technique. Fix a simple routine: a written plan issued before the visit, a start-of-visit confirmation with the factory on lot availability and access, real-time notes on any deviation with factory countersignature, and a same-day summary to the buyer with photographs. Each step takes minutes and each prevents a class of dispute that otherwise consumes days. The sampler should carry the authority to stop and reschedule if the lot is not finished, because sampling unfinished goods wastes the entire laboratory budget.

Illustration: Coordination routine for complex Decorative illustration for the section "Coordination routine for complex"; visual only, carries no data.

Spares discipline closes the routine. Retained sealed spares from the original sampling, stored under agreed conditions with their identifiers logged, are the only specimens that can support a credible challenge retest. Their storage location and condition should appear in the sampling report, and their disposal or expiry should be a conscious decision, not silent neglect. Programs that treat spares as leftovers discover their value only when it is too late.

Finally, file the sampling report with the laboratory reports and the lot documents as one package, and keep it for the life of the product plus the period your customers require. A future question about a lot, whether from an auditor or an authority, is answerable in an afternoon when the file shows what was tested, how, and why. Without that file, the same question restarts the whole investigation from zero.

Split before sampling, judge each split separately, and document every split so the file reads clearly years later. Blended specimens are the most expensive shortcut in testing.

Sources

  1. Wikipedia: Acceptance sampling —— Overview of acceptance sampling including AQL concepts used in product inspection.
  2. WTO: Agreement on Technical Barriers to Trade —— WTO portal for the TBT Agreement covering technical regulations, standards and conformity assessment.

Frequently asked questions

How should sampling handle a mixed lot with several materials or colors?

Split the lot into defined sub-lots by material, color, or production date, and sample each sub-lot under its own identifier. Never blend specimens from different sub-lots into one bag, because the laboratory result then describes an average that matches nothing shippable. Report and release each sub-lot on its own evidence, even when the goods travel in one container.

What changes about sampling during peak-season rush production?

Festival and peak-season production compresses schedules, which tempts factories to offer unsealed or partial lots for sampling. Hold the line: sample only finished, packed, and sealed production, and record any pressure to do otherwise. A test report on unfinished goods describes goods that do not exist yet and cannot support a release decision.

What specimens support a retest if the first report is questioned?

The sealed spares retained at the original sampling, stored under the agreed conditions and never opened. Their value expires if storage conditions were uncontrolled or the seal record is incomplete, so spares discipline is part of sampling discipline. Without usable spares, any retest requires a fresh sampling visit.

What belongs in the sampling report that few buyers include?

The protocol reference, the lot and sub-lot identity, the sampling method and date, the number of specimens and replicates, and any deviations with their reasons. Attach the sampling photographs and the courier record. A future reader, whether a customer auditor or a regulator, should be able to reconstruct exactly what was tested and why.

admin

Writer at Get On My Space