Aflatoxin in rice and GAFTA arbitration: whose risk is it?

Rice that has comfortably passed quality control in the country of origin can turn out to be “unfit” in the country of destination — simply because a stricter aflatoxin limit applies there. Who bears that risk: the seller or the buyer? A dispute that went to GAFTA arbitration showed that the answer turns not on common sense, nor on what the goods were actually bought for, but on what the contract actually says.

What aflatoxin is and why standards vary so widely

Aflatoxins are toxins produced by moulds of the Aspergillus genus. They form on grain, nuts and rice when these are improperly stored in warm and humid conditions, they survive ordinary cooking, and they are regarded as carcinogens. That is why almost every country sets a maximum permitted level of aflatoxin in food.

The problem is that these levels vary widely. The very same bag of rice can be a perfectly lawful product in the country of origin and a breach of the limit in the country of destination. To put it in figures — the numbers the dispute revolved around:

The unit “µg/kg” (micrograms per kilogram) is the same as ppb (parts per billion), as is often stated in certificates. The point here is not the numbers themselves but the gap: the Pakistani limit is five times more lenient than the European one. When goods travel from a country with a lenient standard to one with a strict standard, that gap becomes a commercial risk. The only question is who bears it.

The dispute: Pakistani rice for an African brewery

A buyer — a European trading company — bought from a seller a consignment of Pakistani rice (100% Broken White Rice) for delivery to a brewery’s warehouse in Rwanda on DAP terms (Delivered At Place — the seller delivers the goods to the agreed destination). The contract incorporated the GAFTA 88 proforma and the GAFTA 125 arbitration rules; the governing law was English. The quality specification referred to the Pakistani standard for rice (“PSI Standard”). And here is the root of the whole dispute: the contract itself was silent on aflatoxin, and the Pakistani standard it referred to sets no aflatoxin limit at all. The very parameter for which the consignment was later rejected was simply absent from the contract.

The key term was the quality clause: quality was declared final as at shipment by the certificate of a surveyor whom the seller chose and paid for. A certificate duly appeared, and on its face it conformed to the contract: it stated that the goods met the specification and were fit for human consumption. Aflatoxin, however, was not mentioned in that document at all — it certified the goods as fit for human consumption without ever testing them for the carcinogenic mycotoxin.

After arrival in Rwanda the brewery ran a rapid test — positive. The local standards body measured total aflatoxin at around 21 µg/kg — twice the local limit (10 µg/kg). At the seller’s suggestion the retained samples were sent to an independent laboratory, which found around 14.4 µg/kg. That is below the Pakistani limit (20) but still above the Rwandan one (10) and several times above the European one (4).

The buyer rejected the consignment, alleged breach of contract and demanded the price back, plus storage, transport and destruction costs.

The seller’s position

The seller’s logic was simple and rested on the text of the contract:

1. Aflatoxin was not a contract parameter at all: the parties had agreed only the PSI rice specification, with which the goods complied. And turning to the Pakistani food-safety regime, it — on the seller’s argument — permits aflatoxin up to 20 µg/kg, so the 14.4 measured by the independent laboratory fell within that limit.
2. The quality certificate was declared final — it cannot be challenged by later analyses.
3. The pre-contract discussions about delivery “for brewing” and stricter aflatoxin levels did not help the buyer: what mattered was the signed contract, which — on the seller’s logic — superseded all the earlier drafts and understandings, and they never became terms of the contract.

The buyer’s position

The buyer argued the opposite, stressing the real purpose of the deal:

1. Both sides knew from the outset that the rice was going to a brewery — the place of delivery is named in the contract. The goods therefore had to be fit precisely for that purpose (fitness for purpose); rice with aflatoxin of 14–21 µg/kg meets neither the destination standard nor, still less, the European one (4 µg/kg) and is unfit for food production.
2. Even if aflatoxin is not spelled out in the contract, the English Sale of Goods Act 1979 (s. 14) automatically reads into such a contract a term of satisfactory quality: the goods must be fit for the purposes for which goods of that kind are commonly supplied. And rice almost always goes for food: about 90% of the world’s crop is consumed as food — by people or animals — and in either case an aflatoxin test is mandatory. A consignment with this level of toxin was fit for none of those purposes: the authorities in the destination country later ordered the cargo destroyed as unfit for both humans and animals. So it was not of satisfactory quality.
3. A certificate cannot be “final” on a parameter it never tested: the document is silent on aflatoxin yet asserts fitness for human consumption.
4. The shipment certificate was issued by a Pakistani laboratory not on the GAFTA approved register, whereas the independent re-test of the retained samples — which the seller itself proposed — was done by a laboratory accredited by GAFTA, and it was that lab which confirmed the exceedance. Since the seller itself initiated that re-test and the parties agreed to treat its result as decisive, the seller thereby acknowledged that aflatoxin mattered — and lost the right to hide behind the “finality” of the first certificate.

What the tribunal decided

The tribunal sided with the seller and dismissed the buyer’s main claim.

On the standard: the arbitrators read “PSI Standard” broadly — as covering not only the rice specification itself but also the 20 µg/kg limit set by the competent Pakistani food-safety authority. Since the contract referred to the Pakistani standard, that limit, in the tribunal’s view, became part of the contract — and the goods fell within it.

On the certificate: the tribunal held it valid and final. The argument that the certificate was invalid because the laboratory was not on the GAFTA approved register did not succeed. The contract gave the seller the right to choose the surveyor at its own discretion, there was no separate GAFTA-accreditation requirement in it, and the chosen laboratory was officially registered by the state as a quality inspector; the certificate also clearly identified the goods. That was enough for the arbitrators. The fact that the buyer had at the time accepted and paid against the certificate without objection only reinforced the conclusion. The argument that agreeing to the independent re-test deprived the seller of the right to rely on “finality” was also rejected: the tribunal saw no clear waiver of that right in the seller’s conduct. The local standards body’s result (≈21 µg/kg) the arbitrators rejected outright: the samples were not linked to the specific cargo and it was unclear how they had been drawn. And the independent laboratory’s result, even if accepted, did not help the buyer: 14.4 µg/kg was still below the “contractual” limit of 20.

On intended use: the tribunal accepted that the seller knew about the brewing purpose. But it concluded that the seller was entitled to assume that compliance with the Pakistani standard was enough for the buyer’s purposes — precisely because that was the standard the buyer had agreed to in the contract. The buyer’s arguments on implied terms and on section 14 of SOGA the tribunal rejected: the expressly agreed specification prevailed.

That said, the buyer did not walk away with nothing. On a separate lot that was never shipped, it had earlier paid a 20% advance, described in the contract as a “deposit”. The seller refused to return it: there had been, it said, no agreed cancellation, the buyer had wrongfully refused to accept the lot (again over aflatoxin), and so the advance could be retained. But the tribunal sided with the buyer. In the arbitrators’ view, the word “deposit” in this contract did not make the payment non-refundable: nowhere was it stipulated that it would be retained or forfeited if the deal fell through (in English law the question of deposits is subtler, but it is decided by the text of the contract). Since the seller had neither shipped the goods, nor terminated the contract, nor claimed damages, there was nothing on which to retain the advance — and it was ordered returned with interest.

It is worth keeping in mind: an arbitral award is not a judicial precedent. It is the resolution of a particular dispute under a particular contract, and its reasoning can certainly be challenged. For exactly this situation, two-tier GAFTA arbitration provides a second instance — an appeal, where the dispute is heard afresh. But the value of the case lies not in who was ultimately right, but in the practical conclusions worth drawing from it.

Lessons from this case

Whatever one thinks of the outcome, the dispute offers several practical takeaways for anyone in the international trade of food commodities.

Write critical quality parameters straight into the contract. If goods are heading to a country with stricter norms than the country of origin, do not settle for a reference to “the national standard” — specify the exact numerical limit for aflatoxin (or any other parameter). As the case showed, a reference to a standard can mean something quite different from what you assumed.

Lock the intended use into the text of the contract, not into correspondence. The purpose of the goods, promised levels, “rice for the brewery” — under English law it is generally hard to rely on any of that when construing the contract: a court or tribunal looks above all at the text of the contract itself. If the purpose and the related quality requirements matter to you, it is safer to write them straight into the contract than to rely on earlier correspondence.

A quality certificate is a powerful tool — but mark out the limits of its finality. A final certificate strips the other side of the right to re-litigate quality — convenient for the seller and dangerous for the buyer. But final “as to what”? Spell out exactly which parameters the certificate covers. A document silent on a parameter that is critical for you can still tie your hands.

Who issues the certificate — and with what qualifications — matters. In this case the shipment certificate was issued by a Pakistani laboratory not on the GAFTA approved register, whereas the independent re-test was done by a GAFTA-accredited one — but it was the shipment certificate that came first and counted as “final”. If it matters to you that quality be confirmed by a GAFTA-accredited superintendent, write that into the contract expressly. Otherwise you risk being left with a finding you would rather challenge but no longer can.

Conclusion

In the international commodity trade the text of the contract almost always outweighs the history of the relationship. Quality standards diverge from country to country, and only the contract itself can close that gap — by express wording, not by an implied “it goes without saying”. And where the reasoning of a decision remains open to dispute, the outcome depends heavily on how well the case is built — both when drafting the contract and in the arbitration itself. A similar logic of “quality warranties versus reality” was examined in the organic corn case.

If you have a dispute about the quality of goods, an aflatoxin exceedance or another parameter, or need help in GAFTA or FOSFA arbitration, get in touch:

📧 danil@danil-hristich.com 📱 Telegram · WhatsApp