Of all the documents that travel with a mineral shipment, one decides how much money changes hands: the Certificate of Analysis. The bill of lading moves the cargo; the COA prices it. Yet buyers routinely file COAs unread, missing both the red flags that signal trouble and the details that protect them in a dispute.
Here is how to read one properly - field by field, with the professional's eye.
What a COA Is (and Is Not)
A Certificate of Analysis is a laboratory's formal statement of what a specific sample contained when tested by specific methods on a specific date. Three implications follow:
- It certifies the sample, not the cargo. The chain connecting sample to lot - the sampling procedure - is what makes the COA meaningful. Weak sampling makes a perfect COA worthless
- It is only as good as its laboratory. An accredited, independent lab's COA is evidence; a seller's in-house table on letterhead is marketing
- It is a settlement instrument. In a well-drafted contract, the COA (or the umpire lab's version of it) mathematically determines the invoice
Field-by-Field: Reading the Document
1. The Laboratory Block
Check the issuing laboratory's name, address, accreditation (ISO/IEC 17025 is the standard to look for) and report number. Then do what almost nobody does: verify the certificate with the lab directly. Fabricated COAs bearing real laboratories' logos circulate constantly in the minerals gray market; every major lab operates a verification channel precisely because of it.
2. Sample Identification
The COA must tie unambiguously to your cargo:
- Lot/parcel number matching the packing list
- Sampling date, location and method (e.g., "incremental sampling per ISO 12743, mechanically divided")
- Who drew the sample - seller's staff, or an independent surveyor (SGS, Bureau Veritas, Intertek)?
- Seal numbers, if samples were sealed
A pristine assay from an unidentified sample is a decoration. For meaningful lots, insist on surveyor-witnessed sampling - as covered in our testing methods guide.
3. The Analytical Method
Legitimate COAs state the technique per analyte: ICP-OES/ICP-MS for elemental content (with digestion method - critical for refractory minerals like monazite), gravimetric oven method for moisture, gamma spectrometry for radionuclides. Vague entries ("chemical analysis") or a single method implausibly covering everything are warning signs.
4. The Results Table - Where the Money Is
For a rare earth concentrate, demand the full element distribution, and read it on three levels:
- The headline - Total REO (or TREO). Necessary, wildly insufficient
- The value lines - Nd₂O₃ and Pr₆O₁₁ individually. These typically carry 60–85% of basket value (see rare earth pricing explained); heavy elements (Dy, Tb, Y) deserve their own check
- The ballast and penalties - cerium/lanthanum share, plus impurities: Fe₂O₃, SiO₂, P₂O₅, F, and always ThO₂ and U₃O₈
Check the basis. Results may be reported as oxide or element, on a dry or as-received basis. Oxide-vs-element confusion alone can shift apparent content by 15–20% (Nd₂O₃ is 85.7% Nd). Any professional dispute begins by reconciling bases.
5. Moisture
Reported separately because settlement uses it directly: payable weight = wet weight × (1 − moisture). On bulk parcels, tenths of a percent are real money - and for sulfide concentrates, moisture versus TML also governs whether the cargo may legally sail (IMSBC Code).
6. Radiological Values
Th and U content (percent or ppm) and, for NORM-relevant cargoes, activity concentrations in Bq/g. These must be consistent with the shipment's declared transport classification - a COA showing 5% ThO₂ attached to cargo booked as ordinary minerals is a contradiction a port authority will eventually notice, with you as consignee. The full framework is in our monazite compliance guide.
7. Date, Signature, Caveats
Recent (lot-specific, not a year-old "typical" analysis), signed by an identified analyst or authorized signatory, with standard-form caveats. Labs' boilerplate ("results relate only to the sample tested") is normal - it is precisely why sampling procedure matters.
Red Flags: The Fast Screen
- ☐ Round numbers everywhere (60.0%, 15.0%) - real assays have decimals and scatter
- ☐ Total REO given with no element breakdown
- ☐ No Th/U line at all on a monazite or mixed concentrate
- ☐ Lab you cannot find, or a real lab that cannot verify the report number
- ☐ "Typical analysis" presented as lot-specific
- ☐ Assay predates the material's stated production
- ☐ Numbers identical to the seller's brochure specification, digit for digit
Any one of these justifies pausing; two justify walking away.
The COA in Your Contract
The document only protects you if the contract makes it operative:
- Name the laboratory (or class of laboratories) whose COA governs provisional settlement
- Define sampling - standard, witnessing, splits, sealed umpire samples
- Set splitting limits and umpire procedure - if buyer's and seller's analyses differ beyond a defined tolerance, a named third laboratory analyzes the sealed umpire sample, and its result binds both parties (usually with the "loser pays" convention)
- Tie the pricing formula to COA fields explicitly - which element lines, which basis, which moisture
This architecture - provisional COA, final umpire - is how copper, zinc and now rare earth trades have settled fairly for a century.
COA Fraud Patterns: A Field Guide
The gray minerals market produces document fraud in recognizable species, and buyers who can name them screen faster. The borrowed certificate - a genuine COA from a real historical lot, presented for different material; defeated by lot-number reconciliation and lab verification of issue date. The template forgery - real laboratory branding around invented numbers; defeated by direct verification through the lab's official channels (never the phone number printed on the suspect document). The typical-analysis switch - a producer's generic brochure specification dressed as lot-specific assay; defeated by demanding sampling dates, methods and sample IDs. The accomplice laboratory - a real but obscure lab issuing accommodating numbers; defeated by insisting on internationally accredited institutions from a pre-agreed list. And the basis shuffle - numbers honest in themselves, presented on the basis (wet, element, as-received) that flatters; defeated by the reconciliation arithmetic every professional runs before pricing.
None of these survive the standard discipline: pre-named laboratories, surveyor-witnessed sampling, direct verification and umpire provisions. Fraud in this market is a filter that catches only the unprepared - which is precisely why sellers of genuine material embrace buyer diligence rather than resisting it.
Frequently Asked Questions About Certificates of Analysis
Who should issue the COA - the seller's lab or an independent one? For settlement purposes: an independent, accredited laboratory, per contract. Sellers' internal labs legitimately produce working analyses, but money should move on third-party numbers - a principle honest sellers propose before buyers ask.
Is a COA legally binding? By itself it is evidence, not obligation. Bound into a contract - as the defined settlement basis with umpire provisions - its numbers become contractually operative. The drafting, not the document, creates the force.
How recent must a COA be? Lot-specific by definition: issued from samples drawn from the actual shipment lot, typically days to weeks before shipment. Any certificate predating the lot's assembly describes something else.
What if my own analysis on arrival differs from the COA? Invoke the contract's mechanism: compare against splitting limits, and beyond tolerance, trigger umpire analysis on the sealed sample. This is routine commerce, not scandal - which is why sealed umpire retention at loading is non-negotiable diligence.
Do certificates differ for radiological values? Radiological data (activity concentrations, Bq/g) often arrives as a distinct gamma spectrometry report alongside the chemical COA, from appropriately equipped laboratories. For NORM-relevant cargo both documents travel together, and both should reconcile with the transport classification claimed.
Key Takeaways on Reading COAs
- The COA certifies a sample, not a cargo - sampling procedure is what connects the two, making surveyor-witnessed, sealed-split sampling the foundation every other check rests on.
- Verify with the laboratory directly: accreditation (ISO/IEC 17025), report-number confirmation through official channels - forged certificates bearing real logos are the gray market's standard tooling.
- Read the distribution, not the headline: NdPr lines carry 60–85% of basket value; Th/U lines govern transport classification; basis (dry/wet, oxide/element) reconciliation prevents 15–20% phantom swings.
- Round numbers are confessions: real assays scatter in decimals; brochure-perfect figures presented as lot-specific analysis end conversations.
- The contract makes it operative: named laboratories, splitting limits, sealed umpire samples and loser-pays provisions convert a certificate into an enforceable settlement instrument.
- Fraud follows five patterns - borrowed, forged, typical-switched, accomplice-issued, basis-shuffled - and standard discipline defeats all five.
- Moisture is money: wet-to-dry conversion moves real value on every bulk parcel, and for sulfides, governs whether cargo may legally sail.
Our Standard
Every CriticalOre shipment carries a lot-specific COA from an accredited independent laboratory: full 17-element distribution, moisture, impurities and radiological values, with sealed umpire samples retained and umpire terms standard in our contracts. Buyers are welcome to verify every certificate with the issuing lab - we would worry about any supplier who felt otherwise.
See our certificates page for the complete document set, or request a quote and read a real one for yourself.