When Accredited Labs Disagree: A Dispute Blueprint for UK CBD Wholesale Testing

Two laboratories can examine the same bottle of CBD oil, both accredited to ISO/IEC 17025, and still return potency figures that are not identical. That is not automatically evidence of a bad batch or a careless analyst. It is a normal consequence of analytical measurement, sampling and the way a specification is written. In UK CBD wholesale, however, a difference of a few tenths of a per cent can determine whether a pallet is accepted, discounted or rejected. The trade needs a pre-agreed way to resolve that difference before it becomes a dispute.

Why competent laboratories can reach different results

A certificate of analysis is not a direct photograph of the bottle. It is the output of a method applied to a small sample. Variation enters at several points. The sample itself may not be perfectly homogeneous, particularly with oils that contain natural waxes or have been stored cool. The laboratory may use HPLC-UV for one analyte and LC-MS/MS for another. It may calculate total CBD from CBD plus CBDa using a conversion factor, or report the cannabinoids separately. It may use a different extraction procedure, calibration standard or integration approach. Even the definition of the result matters: total CBD, CBD only, or total cannabinoids expressed as CBD.

Contaminant testing adds further scope for divergence. Pesticide, heavy metal and mycotoxin results near the limit of quantification can move in either direction when method, clean-up and instrument sensitivity differ. Accreditation is important, but it is not a guarantee that two laboratories are measuring the same thing in the same way. Buyers should read the accredited scope and the method statement, not only the logo on the certificate.

Define the question before the laboratory answers it

Many trade disputes begin with a contract that says the product must meet specification and a CoA that says it does or does not. That is too thin. A method-of-analysis clause should identify the analyte, the reporting basis, the method family, the limit of quantification, the uncertainty statement and the calculation convention. It should also require the laboratory to hold accreditation for that specific test and to provide raw data on request. For total CBD, the contract should state whether CBDa is included and which conversion factor applies. For contaminants, it should state the method and the reporting limit.

The specification and the analytical tolerance are separate concepts. A specification might require a minimum total CBD content of 10.0 per cent w/w. The analytical result carries an uncertainty. A sensible contract therefore includes a guard band: the seller targets a slightly higher figure so that a compliant batch is unlikely to fail on measurement alone. The contract should also set an acceptable difference between two laboratories. That might be a relative difference of 10 per cent for potency, or a rule based on expanded uncertainty. The exact figure should reflect the analyte and the method, but it must be written down before the pallet is shipped.

Split-sample retention is the cheap insurance

The most effective dispute mechanism is not a clever clause. It is a sealed sample. At filling or before dispatch, the parties should draw a representative sample according to a documented plan. The sample should be divided into at least three aliquots: one for the seller, one for the buyer and one for a referee. Each aliquot should be tamper-evident, labelled with batch number, date, time and sampler, and stored under conditions that match the product label. A chain-of-custody record should follow the samples from draw to laboratory.

For oils, the sampling plan should address mixing and temperature. A sample taken from the top of a drum may not represent the bottom. A sample that has been frozen and thawed may separate. The retained aliquots should be large enough for full re-testing, not merely for one potency injection. Photographs of seals and labels are useful. So is a joint sampling record signed by both parties. If a dispute later arises, the referee laboratory can test a retained aliquot that has not been opened by either side. That single precaution removes much of the argument about whether the challenged sample was representative.

A practical arbitration protocol

A workable protocol has clear stages and short timelines. First, notice. The buyer should notify the seller in writing within an agreed period of receiving the CoA, identifying the analyte, the result, the specification and the batch. The pallet should be quarantined but not destroyed. Second, exchange. Both parties should provide the full CoA, method statement, uncertainty, accreditation scope and raw data for the challenged test. Third, technical reconciliation. The laboratories should compare sample preparation, calibration, integration and calculation before anyone appoints a referee. Some disputes end here.

Fourth, referee testing. If the difference remains material, the parties should jointly nominate a referee laboratory with the relevant accredited scope. The referee should receive a sealed retained aliquot under chain of custody and should test to the method specified in the contract, or to a method agreed in writing before testing. Fifth, decision. The contract should state whether the referee result is determinative or whether the parties will treat a result within the agreed tolerance as conforming. Costs should be allocated in advance: shared, or borne by the party whose position is not upheld. The protocol should also set a deadline for the referee report and a route for a without-prejudice settlement if the commercial relationship is worth preserving.

The commercial lesson

Testing disputes are predictable. They are also expensive when handled late. A UK CBD wholesale contract that defines the analyte, retains sealed samples, sets tolerance bands and names a referee mechanism will resolve most challenges without lawyers. The aim is not to make every laboratory agree. It is to make disagreement manageable, transparent and proportionate. In a trade where provenance and compliance are part of the sales proposition, that is not bureaucracy. It is product confidence.