How a £7 million claim ended in nothing: exclusion clauses after Trinsic v Procoll
A buyer sues its supplier for more than £7 million in lost profits. It proves that the supplier breached the testing obligations. It proves that the goods were not fit for purpose. It defeats the supplier’s attempt to blame it for the problem. And it walks away with absolutely nothing, having in fact been ordered to account for an unpaid invoice.
That is the outcome in Trinsic Collagen Limited v Procoll Limited [2026] EWHC 1793 (Ch), a judgment of Andrew de Mestre KC sitting as a Deputy High Court Judge, handed down on 17 July 2026 after a five-day trial in the Chancery Division. It is not a case that changes the law. It is a case that shows, with unusual clarity, what the power that a limitation of liability clause in a contract can actually does to a claim, and how much of a commercial dispute is decided long before anyone gets near a courtroom.
The deal
Procoll had developed a novel product: a single alpha chain bovine collagen that dissolves in alkaline water. Trinsic wanted it for high-value nutraceutical drinks. In November 2022 the parties signed a bespoke supply agreement, negotiated with solicitors on both sides, running for at least twenty years. Trinsic had exclusivity in the food, beverage and nutraceutical markets, committed to buy 5,000 grams a month, and paid £6 per gram in year one rising by £1 a year.
Clause 8 of their contract set out a detailed quality regime. The product had to conform to a specification, including a limit of 1,000 colony-forming units per gram for total aerobic bacteria. An independent product tester, agreed between the parties, had to test each batch before delivery. Procoll could not sell, and Trinsic could not buy, product that had not passed testing. If a batch failed, Procoll had to supply a replacement batch for testing.
Clause 17 did the opposite job. It excluded the terms implied by sections 13 to 15 of the Sale of Goods Act 1979, excluded loss of profits, loss of sales and business, and consequential loss outright, and capped Procoll’s total liability at the product price paid in the preceding twelve months. It also carved out deliberate default: neither party could rely on the limitations for liability arising from its own deliberate default.
Between March and November 2023 there were eight monthly batches. One returned a bacterial count of 6,900 against a limit of 1,000. Four returned results reported as “greater than 300”, which told nobody whether the limit had been met. And from July 2023 onwards a substance formed in the product which Trinsic called reformation and which made the collagen impossible to mix into drinks.
The buyer won on breach, and it was worth £3,000
The judge made a series of findings in Trinsic’s favour. The March batch failed testing, clause 8.11 was engaged, and Procoll never supplied the replacement batch it was obliged to supply. The April, August, September and October batches produced ambiguous results which neither passed nor failed, and that ambiguity was itself a breach of the obligation to have the product tested and not to sell untested product. The reformation rendered product from the March to September batches unfit for purpose, and Procoll’s case that Trinsic had mishandled or badly stored the goods was rejected on the evidence.
The total recovery was a late delivery fee of £3,000 on the March batch, capped at 10% of the price. That was then more than cancelled out by Procoll’s set-off of the £30,000 invoice for the November batch, which Trinsic had taken, used and never paid for. Net result: nothing due from Procoll to Trinsic.
Why the money disappeared
Everything Trinsic claimed of any substance was loss of profits, quantified at just over £7 million. Clause 17.7.2.1 excluded loss of profits and clause 17.5 capped liability at the price paid. So Trinsic had to get past clause 17, and it tried two routes. Both failed.
Deliberate default is a very high bar, and an undefined one
The first route was the deliberate default carve-out. Trinsic built a detailed case that Procoll’s founder, Dr Widdowson, knew from April 2023 that the limit was 1,000, understood that the March batch had failed, and then set out to conceal it: searching out a website to manufacture ambiguity, putting a technical argument to a local environmental health officer he knew would not understand it, instructing the laboratory to test only “neat” so that no result could ever exceed 300, withholding the March result, and then lying about all of it in the witness box.
The judge rejected almost every plank. He found that the founder was not intimately familiar with microbiological testing at the relevant time, was genuinely uncertain what the results meant, took real if misplaced comfort from conversations with the environmental health officer and the laboratory, and had a legitimate reason for asking for neat testing. Standing back, he found not a calculating deceiver but “a slightly naïve person who was grappling with translating successful research into a commercial operation for the first time”.
There are two lessons here. The first is strategic. Running dishonesty to escape an exclusion clause is a high-risk approach. It demands cogent evidence, it colours the entire trial, and when it fails the exclusion clause simply bites and the claim is worth nothing. It did not help Trinsic that its own principal witness had to accept in cross-examination that he had said things which were untrue, with the result that the judge approached his evidence with significant caution. Credibility attacks cut both ways.
The second is a drafting point. Nobody had defined “deliberate default”. The judge had no authority on the phrase put before him, took it to mean knowing at the time that the act was a breach, and assumed blind eye knowledge counted only because Procoll did not argue otherwise. If a carve-out is the thing standing between your client and an uncapped claim, define it.
Reasonableness under UCTA, and the myth of the sole supplier
The second route was the Unfair Contract Terms Act 1977. Trinsic argued that the package of clause 17.8 excluding the statutory implied terms, clause 17.7.2.1 excluding loss of profits and clause 17.5 capping liability was unreasonable under section 6(1A), leaving it free to claim lost profits for breach of the implied terms.
Its central point was inequality of bargaining power. Procoll was the only supplier in the world of alkaline water soluble collagen, Trinsic could not make its drinks without it, and Trinsic had already invested in equipment. On the face of it that is a reasonable argument. It failed, and the reason is instructive.
At the time the agreement was negotiated in 2022, Trinsic was not a supplicant. It was asserting, through solicitors, both an enforceable existing supply agreement and a claim against Procoll for at least £4.7 million, which would have destroyed the company. Its own founder described Procoll in evidence as having “surrendered”. The agreement that emerged was bespoke and heavily negotiated, and Trinsic won points in it: a better price schedule than the heads of terms, a unilateral right to terminate on notice which Procoll did not have, and one way exclusivity which allowed Trinsic to buy the same product elsewhere. Trinsic had tried to delete the loss of profits exclusion during negotiations and lost that point. And on signature its founder called the deal “amazing news”.
Two further points from the reasonableness analysis are worth thinking about when writing contracts. First, clause 17.8 was defensible precisely because the statutory implied terms were replaced by something rather than by nothing: the parties had built a bespoke quality and fitness regime in clause 8, including an eighteen-month warranty which Trinsic itself had inserted. An exclusion that leaves the buyer with no protection at all is a much harder sell.
Second, and counter-intuitively, the fact that loss of profits was squarely in the parties’ contemplation counted against the buyer rather than for it. Trinsic was anticipating revenue of around £3.5 million a month from collagen costing it £30,000 a month. The judge treated that imbalance between the price of the input and the value of the upside as a reason why allocating the risk to Trinsic was reasonable. If the potential loss is wildly out of proportion to the contract value, expect the court to say that is exactly what the exclusion was for.
The most expensive feature of the case: a contract nobody followed
Clause 8 was premised on testing happening before delivery. From the very first batch, that is not what happened. Trinsic collected product before results were back, and used it. As the judge observed, on the face of it the testing regime was not followed for a single order, and yet neither party wanted to argue the point because it did not suit either of them.
That mismatch produced a genuinely odd result. Trinsic was entitled in principle to an entirely fresh 5kg replacement batch for the March order, most of which it had already used at a profit. The judge described this as surprising and said he reached the conclusion without any great enthusiasm, but Procoll could not identify a coherent legal answer. Waiver and estoppel were floated orally in closing, but neither had been pleaded or evidenced, and the agreement required any waiver to be in writing.
There is a related detail that should make every contract drafter wince. The specification was pegged to a set of Welsh regulations on the production of bovine collagen made in 2005 and repealed in 2006. Neither party noticed, and that phantom regulation ran through the whole dispute, including the obligation on Trinsic to store and transport the product in the manner it specified, which as it turned out said almost nothing.
So: make sure your operations match your contract, or change one of them. Peg specifications to a standard that still exists, and check it periodically. And plead your alternative arguments, because “possibly an estoppel” raised in closing is not an argument.
Restitution is not a way round the exclusions
Trinsic’s fallback was a restitutionary claim to recover the price of the unused product on the basis of a total failure of consideration. It failed on two grounds. The agreement had not been discharged, and indeed both sides were still relying on it, Procoll for its set-off and Trinsic for its entitlement to a replacement batch. And there was no total failure of basis on the facts: Trinsic had received and profitably used a substantial part of what it bought. On the March order that argument was particularly stark, because Trinsic received 6kg, used 5kg profitably in drinks for human consumption, and still said the consideration had wholly failed.
The counterclaim that failed too
Procoll had its own ambitious argument, seeking to set off £845,000 for the minimum purchase amounts Trinsic had not bought from December 2023 onwards. That failed. Trinsic was obliged to buy 5kg a month, but the payment obligation was triggered by delivery and the rendering of an invoice, and Procoll had neither manufactured, delivered nor invoiced. A clause allowing Procoll to delay delivery dates did not silently convert the minimum purchase commitment into a take-or-pay obligation. If you want a customer to pay for volume it does not take, that needs its own clear words and its own payment mechanism.
What I would take from this
If you are a buyer signing a supply agreement for something your business genuinely depends on, the loss of profits exclusion is the clause that decides what your remedy is actually worth. Not the specification, not the warranties, not the testing regime. Those clauses determine whether you have a breach and determine whether the breach is worth anything.
Trinsic proved breach on several fronts, defeated the causation and mitigation defences, and had a judge willing to find that the goods were not fit for purpose. And yet it ended the trial owing money. That is not a criticism of how the case was run. It is the exclusion clause doing exactly what it was negotiated to do, four years earlier, in a paragraph nobody spent long enough on.
Simple lesson: you got to read all the contract – or get me to.
The judgment in Trinsic Collagen Limited v Procoll Limited [2026] EWHC 1793 (Ch) is available on the National Archives Find Case Law service.
I advise owner-managed businesses on supply agreements, distribution and manufacturing contracts, and my colleagues advise on the disputes that follow when they go wrong. If you are about to sign something with a limitation of liability clause in it, or you have just discovered what yours says, get in touch.
