Wilmar Oleo Pte Ltd v Vinmar Chemicals and Polymers BV (Rev 1)

In August 2008 the Claimant (“Wilmar”) and the Defendant (“Vinmar”) entered into a contract under which Wilmar agreed to sell and Vinmar agreed to buy a cargo of a particular type of what is known as “biofuel”. The cargo was to be supplied FOB at a port in the Amsterdam/Rotterdam/Antwerp (“ARA”) range in three shipments in January, February and March 2009. I shall refer to this contract as “the August 2008 contract”. A dispute arose in January 2009 which concerned the shipment for that month only. The shipments in February and March were duly performed.

The load port for the January shipment was Vlaardingen, near Rotterdam. On 9 January 2009 Vinmar nominated MT Varkan Ege to lift the cargo. Later that day, in response to a request from Vinmar, Wilmar supplied a certificate of analysis for the biofuel in the shoretank from which Wilmar intended to supply the cargo. The certificate was issued by Saybolt (a division of Core Laboratories). It stated the water content to be 390ppm – within, but only just within, the contractual specification of no more than 400ppm. I shall refer to it as “the Saybolt certificate”. While Varkan Ege was waiting for a berth Wilmar and Vinmar discussed and eventually reached agreement on the terms of a standby letter of credit. Shortly afterwards, at Vinmar’s request, Wilmar issued instructions to the terminal to load the cargo from shoretank 7819 (“ST 7819”), this being the tank identified in the Saybolt certificate. On 14 January 2009 Varkan Ege moved from the anchorage to the berth. At a stage before she had begun loading the Wilmar cargo, Vinmar advised Wilmar that SGS Global (“SGS”) had analysed the biofuel in ST 7819, and had found it to have a water content of 435ppm – in excess of the contractual maximum. I shall refer to SGS’s certificate to that effect as “SGS C04”. At 12.40 Vinmar insisted that by close of business local time that day Wilmar must supply a substitute cargo. Wilmar’s reply was that the cargo in ST 7819 was contractual and ready to load on Varkan Ege, and Wilmar expected Vinmar to load the cargo. At 18.08 local time Vinmar emailed Wilmar noting that close of business had passed, asserting that Wilmar had failed to deliver a cargo for the January shipment, and stating that it would seek an alternative cargo, with all losses incurred by Vinmar to be for Wilmar’s account.

Wilmar says that in breach of contract Vinmar failed to lift the January shipment either on 14 January, or on 15 January (when it says that Vargan Ege would have been ready to load the shipment), or at any time during the remainder of January. It claims damages from Vinmar for non performance in the sum of $643,110. Vinmar accepts that if it is liable for non performance then damages in that sum would be payable. Vinmar says, however, that actual or anticipatory breaches of contract by Wilmar were such as to entitle Vinmar to refuse to perform the January shipment. At hearings in October last year and January this year, and in written submissions thereafter, Wilmar’s case has been presented by Mr Michael Collett of counsel and Vinmar’s case has been presented by Ms Caroline Pounds of counsel. I am grateful to them both for their oral and written submissions.

The structure of this judgment is as follows:

Wilmar was formerly named KemOleo Pte Limited. It forms part of the Wilmar group of companies, which I shall refer to as “the Wilmar Group”. The Wilmar Group also included Wilmar Edible Oils GmbH and Wilmar Europe Trading BV. The group is a producer and trader of biofuels.

The Vinmar group of companies, which I shall refer to as “the Vinmar Group”, included Vinmar, which was the Vinmar Group’s primary vehicle for European business. It also included Vinmar International Limited (“Vinmar International”) and Vinmar Overseas Limited (“Vinmar Overseas”). The Vinmar Group is a trader in biofuels, among other commodities.

Over the period from November 2006 to June 2008 nine contracts concerning particular types of biofuel had been made between Wilmar and companies in the Vinmar Group. Wilmar had been sellers, and the Vinmar Group had been purchasers, in contracts 1, 2, 3, 5 and 6. Conversely contracts 4, 7, 8 and 9 were sales by the Vinmar Group to Wilmar. Contracts 1, 2 and 3 were CFR Houston. Contracts 5 and 6 were CIF Deer Park, Texas. Contract 4 was FOB Houston, and contract 7 was FOB Deer Park, Texas. The last two contracts, 8 and 9, were CIF Rotterdam and Vlaardigen respectively. The August 2008 contract was the first between the two groups on an FOB European port basis.

It is common ground that the agreed contractual specification for the cargo was, so far as material:

Product

FAME meeting EN 14214 specification with the following additional guarantees:

Water content 400ppm max

“FAME” is an abbreviation for fatty acid methyl ester. The process of esterification involves animal or vegetable fats being reacted with alcohol to produce fatty acid methyl esters. EN 14214 is the European standard which specifies requirements and test methods for FAME to be used as an automotive fuel in diesel engines. There was one difference between the contractual specification and the requirements found in EN 14214 in that in the August 2008 contract the water limit was 400ppm. This was 100ppm below the water limit prescribed by EN 14214, which was 500ppm. As to the test method for determining water content in FAME, EN 14214 adopts the test procedure for water content set out in another standard, EN 12937. EN 12937 is a standard devised for petroleum (that is, mineral based) products. Other standards calling for mention are EN 4259 (the European standard for determination and application of precision data in relation to methods of testing), EN 3170 (petroleum liquids – manual sampling) and ISO 17025 (general requirements for the competence of testing and calibration laboratories).

The performance of contracts for the sale and purchase of biofuel will usually involve both quality and quantity surveyors. Among such surveyors were – in addition to Saybolt and SGS - commercial entities which I shall refer to as “Inspectorate” (a division of Bureau Veritas), “Intertek” (Intertek Agri Services, Intertek Caleb Brett and Intertek Group plc), and “Nippon YKK” (Nippon Yuryo Kentei Kyokai).

Starsupply Renewables SA of Nyon, Switzerland (“Starsupply”) is a broker which is used by buyers and sellers of biofuel to conclude trades. As explained below, Starsupply acted as broker in concluding contract 9 and the August 2008 contract between Wilmar and Vinmar. It also acted as broker in concluding a contract (“the INEOS contract”) by which Vinmar sold about 2,700 metric tonnes of FAME to INEOS Manufacturing France SAS (“INEOS”). On 7 January 2009 Vinmar chartered Varkan Ege, intending to fulfil the INEOS contract by blending different parcels on board that vessel at Vlaardingen, one such parcel being the January shipment which Vinmar had purchased under the August 2008 contract.

At Vlaardingen the Vopak terminal includes a number of shore tanks used by the Wilmar Group and the Vinmar Group. In January 2009 those used by the Wilmar Group included ST 1254, ST 7744 and ST 7819.

Surveyors whose services were available at Vlaardingen included Saybolt and SGS. Both were accredited by the National Accreditation Body for the Netherlands as meeting ISO 17025.

A list of issues was prepared in early 2010. It acknowledged that proposed amended particulars of claim would give rise to further issues. Directions concerning the proposed amendments were given at a case management conference. Revisions were subsequently made - or proposed to be made - to the particulars of claim, the defence, and the reply. Regrettably, the parties neither reached agreement on a revised list of issues nor did they seek a pre-trial review so that the court could determine those issues. They did, however, agree that Vinmar bore the burden of justifying its position, and accordingly Vinmar would open its case and adduce its factual evidence before Wilmar’s factual evidence was adduced. Subject only to questions as to costs arising out of the way in which the issues had emerged, the issues identified during Ms Pounds’ oral opening submissions were:

Issue A - Was the vessel nominated by Vinmar ready, and did Wilmar tender or seek to tender a cargo for shipment?

Issue B – If the answer to question A is “Yes”, was Wilmar entitled to rely on the Saybolt Certificate of Analysis as regards the quality of the cargo? In other words, was the determination of the quality of the cargo, as set out in the Saybolt certificate, final and binding upon the parties?

Issue C – If Wilmar is not entitled to rely on the Saybolt Certificate, then was the cargo off-specification as regards its water content? Did the cargo in tank 7819 have more than 400 parts per million of water? [In that context, Vinmar contended that SGS C04 was binding on both parties, with the result that both these questions must be answered, “Yes”.]

Issue D – If the cargo in tank D had more than 400 parts per million of water, and the answer to Issue A is “Yes”, is time of the essence so that Vinmar was automatically released from further obligations in relation to the January shipment? [This issue only arose if Issue A was answered “Yes”, Issue B was answered “No”, and Issue C was answered “Yes”, with the result that Wilmar had committed an actual breach of contract.]

Issue E – If time was not of the essence, as Wilmar says, meaning that the time for Wilmar’s performance did not expire until 31 January 2009, was Wilmar entitled to tender a substitute cargo and did it do so within the necessary time?

Issue F – Whether or not the time for Wilmar’s performance had arisen, and whether or not Wilmar, prima facie, had any right of cure, did Wilmar nevertheless renounce or repudiate the Contract in relation to the January shipment and did Vinmar accept that conduct as relieving Vinmar from the further performance of its obligations as regards the January shipment?

In the course of written and oral closing submissions in January, and subsequent written submissions in February, it was suggested that these issues should be supplemented. On 25 January 2011 Vinmar produced an addendum to its written closing submissions. This clarified its case on repudiatory conduct in a way which was more clearly confined to the January shipment, and which I shall call “the clarified renunciation assertion”:

6.

In short, [Vinmar] was entitled to (and did) treat [Wilmar’s] renunciation of its obligations in relation to the January shipment as discharging [Vinmar] from its own obligations pursuant to the [August 2008 Contract] in relation to that shipment. The effect of [Wilmar’s] renunciation was therefore to strike the January shipment out of the [August 2008 Contract], which was pro tanto discharged.

As regards actual breach, the addendum to Vinmar’s written closing submissions said that Issue A ought to be answered “Yes” pursuant to an analysis which I shall call “the appropriation assertion”. This analysis said, in relation to the January shipment, that there had been an appropriation of the cargo in shore tank 7819 “in the contractual (albeit not the proprietary) sense.” Vinmar said that the effect of this contractual appropriation was to bind Wilmar contractually to deliver the cargo in shore tank 7819, but performance in this way would:

breach an essential condition of the Contract. The Defendant was entitled to reject such a tender of (non-) performance and, moreover, to treat itself as discharged from the further performance of its own obligations in relation to the January shipment…

As to any right of re-tender, Vinmar advanced what I shall call “the re-tender of cargo contentions”. It said that in the present case any right of re-tender was limited to the right immediately to re-tender a conforming cargo upon Vinmar’s rejection of the cargo tended from shore tank 7819, alternatively to do so within a reasonable time (which period of time was in Vinmar’s submission exceeded on the facts of the present case). Wilmar had failed to do so, and Vinmar said that in circumstances where its vessel had been nominated and accepted, and had arrived at the load port, and both parties had embarked upon performance of the shipment, it was both a commercial and legal nonsense to suggest that Wilmar had until 31 January 2009 to re-tender a conforming cargo. Nor would any right of re-tender arise, said Vinmar, in circumstances where Wilmar’s conduct was repudiatory. The addendum to Vinmar’s written closing submissions continued:

3.7.

Even if, which is denied, the (commercially absurd) proposition that [Wilmar] had until 31 January 2009 to re-tender a cargo were correct, at no stage did [Wilmar] ever seek to do so…

3.8.

[Vinmar] is accordingly entitled to judgment in its favour based on [Wilmar]’s actual breach of the Contract, in the form of its failure to tender (or re-tender) a contractual cargo for delivery to [Vinmar], whether within the necessary timeframe or at all.

3.9.

It is no answer to this for [Wilmar] to point to [Vinmar]’s message of 18:08 hours on 14 January 2009 as absolving [Wilmar] from its duty to perform the Contract:

3.9.1.

First, by 18:08 hours on 14 January 2009, it was already too late for [Wilmar] to re-tender performance, either because it had not exercised any right of re-tender in time, or because its conduct as a whole had already amounted to a repudiatory breach which [Vinmar] was entitled to accept as discharging it from the further performance of its own obligations.

3.9.2.

In any event, everything that [Vinmar] said and did after [Wilmar]’s non-contractual tender of performance must be viewed against the background of [Wilmar]’s non-contractual tender and [Wilmar]’s position that it was under no obligation to remedy its non-contractual tender of performance. [Vinmar] could itself be under no obligation to perform the Contract unless and until [Wilmar] indicated that it wished to exercise any right of cure that it may have had, which it never did.

In the course of her oral closing submissions on 27 January 2011 Ms Pounds on behalf of Vinmar asserted that Wilmar’s conduct in providing a certificate of analysis from Saybolt – when according to Vinmar it was not entitled to do so – amounted to either an actual or an anticipatory breach of contract. When it was pointed out that this had not been pleaded, Ms Pounds undertook to prepare a draft pleading in that regard.

Vinmar’s proposed amendment took the form of a document entitled “amended particulars of breach”. This asserted a new case (“the COA breach case”) that:

In written submissions on the proposed amendment Wilmar said that Vinmar’s proposed amendment was hopeless and must fail. It did not make any difference in practice whether the court refused the amendment or permitted it and dismissed the defence based upon it. For that reason, Wilmar did not oppose the amendment on the basis that the costs would be paid by Vinmar in any event.

On 8 February 2011 Vinmar lodged written closing reply submissions, there having been insufficient time on 27 January 2011 to permit oral reply submissions. The written closing reply submissions “vehemently denied” the suggestion that the proposed amendment would be hopeless. Further submissions from Wilmar dated 9 February 2011 responded to Vinmar’s arguments in this regard.

A further topic was discussed in the course of oral closing submissions and in subsequent written submissions. This arose from an observation that I made during the course of oral closing submissions, when I canvassed whether the adoption by the contract of EN 14214 might itself demonstrate agreement that the carrying out of a test in accordance with a test method meeting the specification would have the consequence that the specification was met whenever a conforming result (i.e. a result within the reproducibility range for a particular value) was produced by such a test. After some debate the parties agreed in February that neither of them should be permitted to advance any new case in that regard.

In the remainder of this judgment I shall refer to “Revised Issue A” to mean Issue A with the addition of the question whether Wilmar’s conduct amounted to purported performance as alleged in either or both of the appropriation assertion and the COA breach case. I shall refer to “Revised Issue E” to mean Issue E with the addition of the re-tender of cargo contentions. I shall refer to “Revised Issue F” to mean Issue F with the addition of the clarified renunciation assertion.

Thus at the start of the witness evidence the court was presented with a complex set of issues, which were then revised into an even more complex set of issues. Working one’s way through them is not easy. Lack of clarity as to the issues well in advance of the trial is almost always unsatisfactory. It was particularly unsatisfactory in a case like the present where witnesses had come from abroad expecting to give evidence on the first day of the trial. Any judicial attempt to restructure the issues on the morning of the trial would have thrown the timetable out of kilter and necessitated changes in the travel arrangements for witnesses. If there had been a pre-trial review I am sure that the issues would have been framed in a completely different way.

The extent to which all the issues identified in the preceding section actually arise depends in large part on whether on 14 January 2009 the water content of the product in ST 7819 was within the contractual specification of no more than 400 ppm. I have concluded on the balance of probabilities that it was. My reasons are set out in Annex 1 to this judgment.

My conclusion as to water content does not of itself mean that Wilmar succeeds in its claim. Vinmar’s defence to the claim would prevail if (a) the August 2008 contract contained a term to the effect that the findings of a mutually agreed load port surveyor would be binding, and (b) SGS C04, and only SGS C04, constituted such findings, and (c) Vinmar succeeded in certain other of its contentions.

For reasons explained below I conclude as to (a) that the August 2008 contract did not contain a term to the effect that the findings of a mutually agreed load port surveyor would be binding. That conclusion, along with my conclusion as to water content, means that Wilmar’s claim succeeds. In any event, as will be seen below, if there had been such a term I conclude as to (b) that SGS C04 did not constitute such findings, as the parties had not agreed that SGS should carry out a survey as to quality. Wilmar’s claim accordingly succeeds for this reason too.

For ease of exposition I shall where convenient include comment on these matters, and other potentially relevant issues, in the course of setting out the history of events. At the end of this judgment I return to the issues as revised. Before turning to the history of events I set out some general comments on the factual witnesses. Comments on the expert witnesses are set out in Annex 1.

Wilmar relied on the factual evidence of Mr Bernd Noldt and Ms Christine Schmidt. Ms Schmidt was based in the Hamburg representative office of Wilmar as Operations Manager in the biodiesel department. She did not give oral evidence. Her witness statement was the subject of a hearsay notice under the Civil Evidence Act. What she said as to the course of events was not in dispute. The August 2008 contract was the first contract between the Wilmar Group and the Vinmar Group with which she had been concerned.

Mr Noldt worked until February 2007 for the Cremer group of companies. He joined the Wilmar Group on 1 May 2007, and until 30 March 2010 was based in the Hamburg representative office of Wilmar as European Marketing and Sales Manager for biodiesel and oleochemicals. The August 2008 contract was the third contract between the Wilmar Group and the Vinmar Group with which he had been concerned – he had previously been involved in contracts 8 and 9. Mr Noldt was a careful witness. Although English was not his first language, he had no difficulty in understanding it and speaking it fluently.

Vinmar relied on the factual evidence of Mr Chris Lokey and Ms Pamela Sarwanto. Ms Sarwanto was employed by Vinmar as Logistics Manager, having worked in Vinmar’s Logistics and Scheduling department since January 2003. She is resident in the Netherlands.

Mr Lokey, who is now Global Products Director for the fuels business of the Vinmar Group, was employed from June 2006 until late April 2010 as Vinmar’s Global Products Manager for Fuels. In that role he was responsible for supervising Vinmar’s biodiesel trading activities. He was responsible within the Vinmar Group for negotiating the August 2008 contract and had been involved in all nine of the previous contracts between the Wilmar Group and the Vinmar Group.

For reasons which I explain below, I have concluded that in certain respects Mr Lokey’s and Ms Sarwanto’s recollection of events is not accurate.

Dealings between the Wilmar Group and the Vinmar Group prior to August 2008 concerning biofuels were relied upon by Vinmar as having involved certain terms or understandings which did not need to be set out expressly in order to be incorporated into the August 2008 contract. One of these was admitted by Wilmar. Others were denied.

Such dealings were also relied upon by Vinmar in submissions concerning exchanges between Ms Schmidt and Ms Sarwanto in January 2009.

The issues potentially affected by these matters are Issue B, Issue D, and Revised Issue F.

Vinmar’s defence, with no material revisions in subsequent amendments or proposals for amendment, contained in paragraph 5 the following four assertions in respect of contracts concluded during the course of dealing:

5.1.

Save in respect of those instances where pre-contractual testing of the quality and/or quantity of the relevant shipment had already taken place prior to the conclusion of the contract (and hence where the purchaser agreed to accept the results of the pre-contractual inspection(s) as final and binding), the appointment of an independent surveyor at the load port was mutually agreed between the parties, in relation to both the quality and quantity of the shipment;

5.2.

Where the appointment of the independent surveyor(s) was mutually agreed between the parties, it was further agreed and/or understood by the parties that the findings of the independent surveyor(s), as thus appointed (and no other), were to be final and binding on both parties in relation to the quality and/or quantity of the shipment;

5.3.

Further and/or alternatively, in respect of all the contracts concluded during the course of dealing, the only independent surveyors that were used and/or relied upon were SGS or Intertek Caleb Brett (“ICB”). Vinmar was content to use, and indeed would only use, SGS and ICB for such purposes on the grounds that they are both leading and globally respected companies as regards the inspection of Biodiesel;

5.4.

In relation to none of the contracts concluded during the course of dealing was an independent surveyor unilaterally appointed by one of the parties.

Of these four assertions, only one – paragraph 5.2 – identified a particular contractual term which in my view could be suitable for implication into future contracts. Paragraphs 5.4 and 5.1 respectively simply asserted that in past contracts neither of the parties had unilaterally appointed an independent surveyor, and instead the parties had reached mutual agreement as to an independent surveyor to be appointed to make a final and binding determination at the load port of quality and quantity of the shipment. Paragraph 5.3 merely asserted the identities of the surveyors who had been agreed upon in the past. There was no assertion in paragraphs 5.1, 5.3 or 5.4 that the parties regarded themselves as under a contractual obligation to agree upon such a surveyor. By contrast paragraph 5.2 asserted that a particular event, namely agreement upon the appointment of an independent surveyor, carried with it an implicit further agreement that the findings of that surveyor would constitute a final and binding determination of quality and/or quantity of the shipment.

Wilmar’s reply admitted paragraph 5.2. It was thus common ground between the parties that they had dealt with each other on the basis that agreement upon the appointment of an independent surveyor carried with it an implicit further agreement that the findings of that surveyor would constitute a final and binding determination of quality and/or quantity of the shipment.

Vinmar’s defence at paragraph 9.2 asserted that it was a term of the August 2008 contract that the findings of the independent surveyor(s), as thus appointed [i.e. by mutual agreement in relation to the load port] (and no other), were to be final and binding on both parties in relation to the quality and quantity of the shipment. Wilmar’s reply admitted this, at the same time advancing an additional contention that the same was true for an independent inspector appointed by Wilmar. This additional contention was not founded on any course of dealing. As oral evidence progressed, however, it appeared to be suggested by Vinmar that Wilmar’s additional contention was inconsistent with the course of dealing.

As to what was asserted in paragraphs 5.1, 5.3 and 5.4 of the defence, and assertions that as a consequence there were particular terms in the August 2008 contract, Wilmar made limited admissions about events during the course of dealing and denied that it gave rise to any term other than that at paragraph 9.2.

At the trial there were certain features of the previous contracts to which Vinmar attached particular significance, and which, with one exception, were not significantly disputed by Wilmar:

The one exception concerned contract 5. In relation to this CIF contract Inspectorate’s Singapore office had been appointed by Wilmar as load port surveyor for both quality and quantity and had received the relevant sample some weeks before the contract between Wilmar and Vinmar Overseas was concluded. Wilmar’s draft sale contract dated 25 March 2008 provided for “a mutually agreed independent surveyor to be appointed by Seller”. This was altered in manuscript by Vinmar to “SGS for both”. In the event SGS certificates were produced for samples at the discharge port, and Inspectorate’s certificates were tendered by Wilmar in order to obtain payment under the letter of credit.

Despite the absence of any pleaded assertion that the parties regarded themselves as under a contractual obligation to agree upon a surveyor who would make a final and binding determination, Vinmar sought at trial to demonstrate that this had in fact been the case. In that regard it seems to me important to note that throughout the course of dealing each side had standard terms. Those used by Wilmar, and those used by the Vinmar Group for sales, would, if they applied to the relevant contract, have imposed such a contractual obligation.

Wilmar’s general, but not uniform, practice was to prepare a contract form for each transaction to be signed by the parties. This contract form included a standard clause, sometimes supplemented and sometimes among other provisions concerning determination of quality. I shall refer to the standard clause as “the Wilmar Standard Inspection Clause.” It was as follows:

INSPECTION:

A mutually agreed independent surveyor to be appointed by seller. Their findings shall be final and binding for both parties, with exception for fraud or manifest error. Inspections will be carried out at load port …

During the course of dealing companies in the Vinmar Group sent to Wilmar standard terms and conditions of sale and standard terms and conditions of purchase. Although it was not just Vinmar that used them, for convenience I shall refer to them as ”the Vinmar Standard Sales Terms” and “the Vinmar Standard Purchase Terms”. The Vinmar Standard Sales Terms stated:

5.

QUANTITY/QUALITY DETERMINATION

For bulk liquid product, quantity and quality will be determined by a mutually agreed inspector whose decision is to be final and binding save for fraud or manifest error, based on shore tanks samples at loading port… Each party may have its representative observe all measurements, sample taking and testing… .”

The Vinmar Standard Purchase Terms stated:

3.

INSPECTION

…

B. For FOB transactions: With the cost to be equally shared between Seller and Buyer (unless otherwise agreed in writing) Seller shall appoint independent surveyors (1) to inspect the cargo spaces designated to receive the Product for cleanliness before loading is commenced; (2) to determine quality and quantity of the Product after loading is completed; and (3) to make final determinations of quality and quantity as provided in this contract.

C. For CFR and CIF transactions:

At its own cost (unless otherwise agreed in writing), Seller shall appoint independent surveyors (1) to inspect the cargo spaces designated to receive the Product for cleanliness before loading is commenced; (2) to determine quality and quantity of the Product after loading is completed; and (3) to make final determinations of quality and quantity as provided in this contract.

D. Solely for the purposes of this Contract, independent surveyor quality and quantity determinations shall be made before any of the product leaves, and after all of it is received, in the static shore tanks which are its source and destination, respectively. The determinations shall be final and binding except for fraud or gross error. If the relevant tank is not static, final and binding determinations shall be made on the vessel just after all of the Product is loaded and before any of it is discharged.

The only documentary evidence of the first contract in the course of dealing was a single sheet which, although it was headed with the name of Vinmar International, comprised a message dated 15 November 2006 from Vinmar Overseas as buyer to Wilmar as seller recording the purchase and sale of 4,000 metric tonnes of Palm Methyl Ester (“PME”) CFR Houston at a price of US$795 per metric tonne. No reference was made to the Vinmar Standard Purchase Terms, or to any other standard terms. The other documentary material comprised a draft Intertek certificate of analysis dated 5 January 2007 and a two documents signed by Intertek on 23 January 2007. The first of these was a Certificate of Quality, Quantity and Vessel Tanks Cleanliness. The second was a Certificate of Quality including an analysis, among other things, of water content. Both documents stated that samples had been drawn from ship’s tanks. Mr Lokey explained that Wilmar nominated Intertek as inspectors of record in early January 2007, and provided Vinmar Overseas with a draft certificate of analysis. He continued in paragraph 12 of his witness statement that:

… We were prepared to accept the appointment of Intertek on the basis of their international reputation as a testing company generally. … the international trade in bio-diesel was very young at that stage and we had yet to form a strong view as to who were the most reliable inspectors. The only change that I required to the draft certificate of analysis was the inclusion in the certificate of a Cold Filter Plugging Point maximum of 14 degrees Celsius. This was reflected in the final certificate of analysis dated 23 January 2007 … . The quantity of the cargo was certified in a certificate from Intertek also dated 23 January 2007.

In cross examination Mr Lokey said that he had been personally involved in all the contracts which formed part of the course of dealing. It was suggested to him that on a number of occasions the buyer, whether that was Wilmar or Vinmar, simply accepted the certificate of analysis that was provided by the seller. Mr Lokey denied that this occurred. As to whether he said that this never happened, his reply was, “It was done by mutual agreement.” As regards contract 1, he acknowledged that it was Wilmar that appointed Intertek. When it was suggested that Vinmar accepted that appointment, Mr Lokey replied that he had “agreed” to Intertek. When it was pointed out to him that his witness statement referred to Vinmar being “prepared to accept the appointment of Intertek”, he replied, “Yes, we agreed to it, yes.”

In cross examination Mr Noldt accepted that Ms Schmidt had not had any dealings at all with Vinmar until January 2009. He also accepted that prior to June 2008, if in the past Vinmar had expressed a preference for a particular surveyor, he would not have been aware of that. On specific trades his colleagues who previously dealt with Vinmar would have known about such a preference. As to contract 1, Mr Noldt agreed that the contract said nothing about an inspector being appointed. He had not been involved, but he agreed that it was reasonable to assume that there would have been some discussion between the parties about the appointment of an inspector.

I am prepared to accept that Contract 1 did not involve express contractual provision as to inspection at the load port. By the same token it contained no express provision that a certificate by a load port inspector would be binding as to quality. The evidence concerning Contract 1 does not establish a proposition that, in the absence of express provision, either party considered that such a certificate would be binding or that the parties either were (or if such a certificate were binding, would be) under an obligation to reach mutual agreement as to the identity of the surveyor at the load port. Nor does it establish a proposition that either party considered that, in the absence of express prohibition, it was not open to the seller to appoint a load port inspector unilaterally. When dealing with contract 1 Mr Lokey gave no direct evidence about any of the propositions above. The way in which Mr Lokey expressed the matter in his witness statement, and his failure in oral evidence to give any actual account of discussion with Wilmar about the identity of the inspector, leads me to conclude that - despite his adamant denials in the witness box - he merely accepted Wilmar’s nomination of Intertek. Even if, however, he actively discussed and agreed the identity of the load port inspector, that fact alone would not establish any of these propositions.

In relation to contract 2 an email exchange dated 1 December 2006 was produced. Mr Gopalan Ranganath (“Ranga”) on behalf of Wilmar emailed Vinmar Overseas referring to a “call this morning” and stating:

We are pleased to accept your bid as follows:

Product: Palm Methyl Ester
Specifications: EN 14214 with exception of CFPP/ Cloud Point. I will check and advise you on the typical for the same
Quantity: 1000MT
Origin: Japan
Price USD 795 per MT, CFR Houston
Shipment: First Half Dec 2006

All other terms and conditions as per usual.

Also dated 1 December 2006 was a “purchase order confirmation” prepared by Vinmar Overseas which stated that the Vinmar Standard Purchase Terms were incorporated in the contract. In addition, however, the purchase order confirmation itself stated:

Inspection:
Quality: To be final and based on shore tank analysis at load port
Quantity: To be final and based on static shore tank measurement at load port.

Quality and quantity inspection to be performed by a mutually acceptable independent surveyor at load port appointed by seller and cost to be paid 100% by sellers.

A Nippon YKK certificate of quality dated 13 December 2006 set out an analysis based on a sample taken from the vessel’s No. 7 starboard tank after loading. An Intertek certificate of analysis printed on 28 December 2006 concerned a sample from the same tank and described the results of tests carried out on 15 December 2006.

Mr Lokey said as regards contract 2 in paragraph 13 of his witness statement:

… Given that this was only a small deal done on a quick basis (shipment was due to take place within only a couple of weeks of the trade being agreed), I would have been prepared to accept the Nippon [YKK] certificate in respect of the quantity. I do not specifically recall whether I discussed the appointment of Intertek with [Wilmar]. However, I would not have accepted a binding certificate of analysis from Nippon [YKK], since it is not an inspector with which I am familiar. I would have insisted on a certificate of analysis being provided by an internationally renowned inspector such as SGS or Intertek.

In cross examination Mr Lokey accepted that Wilmar appointed Intertek. He said that if he remembered correctly:

…this was probably either the first or second cargo of biodiesel that we had ever purchased at Vinmar, so we didn’t have a point of view on inspectors at that point on a global basis.

It was put to him that it would be fair to say that, as far as he could recall, he had not objected to the appointment of Intertek. He replied:

A. And as such I agreed to it, correct

Q. Is that the same thing in your view, if someone does not object, they agree?
A. Absolutely.

In cross examination Mr Noldt, when taken to these documents, pointed out that he could not comment on what other terms and conditions on previous business had been agreed upon. He was then simply asked questions about what the documents said. At this point Mr Collett intervened, questioning whether there was any admissible evidence that Mr Noldt could give. I replied that I was waiting to see how this line of cross examination would prove relevant, and that I was sure that Ms Pounds would demonstrate its relevance as she took it forward.

So far as contract 2 is concerned, it seems likely that pursuant to the main body of the purchase order confirmation sent by Vinmar Overseas the parties agreed that a load port shore tank analysis would be binding and that for this purpose there was an obligation on both parties to identify a mutually acceptable independent surveyor. In fact however no shore tank analysis is in evidence for this contract. There having been express provision about load port shore tank analysis, it does not seem to me that contract 2 offers any assistance in relation to the propositions identified under contract 1 above.

Contract 3 was set out in a Vinmar Overseas document dated 8 May 2007 and headed “Vinmar purchase confirmation”. It was signed by Mr Lokey on behalf of Vinmar Overseas and Mr Gopalan Ranganath on behalf of Wilmar. Standard terms and conditions were said to be “as per Vinmar standard terms and conditions of purchase…”. These were annexed, but did not contain the provisions later found in what I have called the Vinmar Standard Purchase Terms. In addition, however, the “Vinmar purchase confirmation” stated:

OTHER REQUIRED DOCUMENTS:

…

CERTIFICATE OF QUANTITY ISSUED BY INTERTEK CALEBRETT

CERTIFICATE OF QUALITY ISSUED BY SGS LABORATORY IN SINGAPORE BASED ON VESSEL COMPOSITE SAMPLES AS PER THE FOLLOWING PRODUCT

…

INSPECTION:
QUALITY: TO BE FINAL AND BASED ON VESSEL COMPOSITE AT LOAD PORT ISSUED BY INDEPENDENT SURVEYOR (I.E. INTERTEC CALEBRETT OR SGS LABORATORY IN SINGAPORE).

QUANTITY: TO BE FINAL AND BASED ON STATIC SHORE TANK MEASUREMENT AT LOAD PORT ISSUED BY INTERTEC CALEBRETT. COST TO BE 100% FOR SELLERS ACCOUNT.

Mr Lokey dealt with contract 3 at paragraphs 14 and 15 of his witness statement as follows:

14.

… In relation to this trade the parties discussed at some length who should be appointed as the inspector under the contract, both after the main elements of the deal had been agreed and after the cargo had been shipped. The main issue was whether Intertek or SGS should be appointed to inspect the quality of the cargo. [Wilmar] preferred Intertek. Ultimately, it was agreed that the quantity of the cargo would be determined by Intertek based on a static shore tank measurement and that the quality would be determined by means of a loadport vessel composite analysis issued by either Intertek or SGS. This was documented in a contract signed by both parties …. The binding certificate of quality under this contract was provided by SGS. …

15.

It was in relation to this contract that Vinmar first expressed a preference for the appointment of SGS. By this time, we had formed the view that SGS were the most reliable inspection company for the testing of bio-diesel. They had developed a reputation as the industry leader and they had become the most trusted inspection company in our own supply chain (and we therefore invariably used them as inspectors upon discharge at Houston).

Mr Lokey acknowledged when cross-examined that it would be fair to say that he did not have a strong recollection of this contract nearly three years after relevant events. Asked whether he remembered that there was a tussle about whether the inspector should be Intertek or SGS, he said he did not recall.

Mr Noldt was asked questions in cross examination about emails between the parties prior to signature of the Vinmar purchase confirmation. He acknowledged that they showed that Vinmar said it would like to make a deal but that the parties needed to agree on which surveyor was to be appointed. Vinmar would prefer it to be SGS, Wilmar preferred Intertek. That remained the position until a late stage when Wilmar’s contacts at Vinmar said they would have to check with their colleagues in Houston, who “seem to insist on certain things.” They then reverted to Wilmar:

We will only accept SGS analysis for letter of credit purposes. There are several reasons including the fact they are considered the industry leader and we use them at this port…

Mr Noldt was then shown manuscript revisions on a draft agreement, before being taken to the provision in the signed agreement referring to the independent surveyor being Intertek or SGS. He was shown an actual certificate of analysis which was produced by SGS, and acknowledged that the parties had specifically discussed and agreed which surveyor was to be appointed.

The evidence shows that in relation to contract 3 the parties considered the identity of the load port inspector sufficiently important to be specified in a detailed and signed formal contract document. There having been express provision both about load port ship’s tank analysis, and about the identity of the inspector, it does not seem to me that contract 3 offers any assistance in relation to the propositions identified under contract 1 above.

This was the first contract in the course of dealing in which a member of the Vinmar Group, in this case Vinmar International, was the seller and Wilmar was the buyer. The documentary material included a number of email exchanges, along with a Wilmar contract form signed Mr Ranganath on behalf of Wilmar but not signed on behalf of Vinmar International.

Mr Lokey dealt with contract 4 at paragraphs 16 and 17 of his witness statement:

16.

The fourth trade between Vinmar and [Wilmar] was negotiated by Ranga of [Wilmar] and myself, and was confirmed in an email by Ranga to me dated 9 October 2007. The trade was for the purchase by [Wilmar] of 1,000 metric tonnes of PME on a FOB Houston basis at a price of US$680 per metric tonne. Shortly after sending his confirmation, Ranga sent me an email setting out [Wilmar]'s nomination and shipping instructions for the cargo, and attaching a copy of [Wilmar]'s contract document which specified that the inspector of record had to be mutually agreed. In this email, Ranga proposed that Intertek should be appointed as surveyors under the contract. I responded by email a little under an hour later to confirm that we would be appointing SGS as the surveyor, on the basis of tests they had undertaken the previous day when we had taken delivery of the cargo at Houston.

17.

The Respondent accepted the appointment of SGS but there did follow some discussion between Ranga and myself as to whether the specifications I had provided from SGS conformed to [Wilmar]'s requirements. Essentially, I had provided analysis results which conformed to EN (European) specifications and [Wilmar] wanted a certificate to confirm that the material conformed to both the relevant EN and ASTM (North American) specifications. As a result of this discussion, I agreed to arrange for SGS to undertake further quality testing in accordance with [Wilmar]'s request. The binding test results were provided by SGS in certificates dated 15 October 2007.

In cross examination Mr Lokey accepted that Wilmar proposed that the surveyor be Intertek and sought confirmation that Intertek would be appointed at shipper’s cost. Vinmar International’s reply had been that the surveyor would be SGS. When it was suggested to Mr Lokey that Vinmar simply told Wilmar that SGS would be the inspectors, he replied:

I told them that SGS would be the inspectors, and by virtue of their agreement they would be the inspectors.

Mr Lokey was taken through the documents underlying the eventual arrangement for SGS to undertake further quality testing as set out in paragraph 17 of his statement. He acknowledged that Wilmar asserted that the contract required certification of ASTM standards as well as EN standards, that Vinmar’s position had been that Wilmar should pay to arrange their own certificate in that regard from Intertek, but ultimately Vinmar arranged for SGS to undertake further quality testing as requested by Wilmar.

In cross examination Mr Noldt was taken to a passage in the email exchanges indicating that Wilmar confirmed its purchase from Vinmar stating:

Quantity and quality, final at load port as ascertained by an independent surveyor. (Kindly advise surveyor.)

Mr Noldt agreed that Wilmar had then sought confirmation from Vinmar that ITS would be appointed. He also agreed that the Wilmar contract form contained the same inspection clause that had been seen in other Wilmar contracts – a mutually agreed independent surveyor to be appointed by the seller. He was taken to Mr Lokey’s response to Wilmar saying, “surveyor will be SGS”, and was asked to agree that the parties were discussing which inspector should be appointed. He commented that to him it looked like a statement. He was willing to accept that in the end Wilmar agreed to SGS, because the documents included a certificate from SGS.

It being accepted by Mr Lokey that the terms in the Wilmar contract form applied, and those terms expressly providing for load port shore tank analysis to be binding with a mutually agreed independent surveyor to be appointed, it does not seem to me that contract 4 offers any assistance in relation to the propositions identified under contract 1 above.

Mr Lokey dealt with contracts 5, 6, and 7 together at paragraphs 18 and 19 of his witness statement:

18.

The fifth, sixth and seventh trades are all connected and I therefore deal with them together in this statement. The fifth and sixth trades relate to a cargo of 12,000 metric tonnes of PME purchased in early April 2008 by Vinmar on a CIF Houston basis at a price of US$1,150 per metric tonne. My recollection of the background to these trades is that [Wilmar] had shipped a cargo at the end of February 2008 for a deal which subsequently fell through, and found itself towards the end of March 2008 looking for a buyer for a cargo that was already afloat. Vinmar agreed to purchase the cargo on the basis that 7,000 metric tonnes would be immediately sold back to [Wilmar] on a FOB Houston basis at a price of US$925 per metric tonne. This explains why the cargo was purchased under two contracts, rather than just one. The fifth contract concerned the purchase of the 5,000 metric tonnes of material that Vinmar was going to retain and the sixth contract concerned the purchase of the 7,000 metric tonnes of material that Vinmar was going to sell back to [Wilmar]. The seventh contract concerned the sale back of the 7,000 metric tonnes of material to [Wilmar].

19.

In relation to all three contracts, it was agreed that the binding certificates in respect of both quality and quantity would be provided by SGS of the basis of tests undertaken at Houston. This is reflected in the contract documents sent to us by [Wilmar], which indicated that quality and quantity determination would be undertaken at discharge in respect of the fifth and sixth trades and at the loadport in respect of the seventh trade. It is also reflected in the manuscript amendments made by my operator, Theresa Bigg, to [Wilmar]'s contract documents indicating "SGS for both" and in Vinmar's own contract documents which all confirm that SGS would be appointed as inspector of record. In relation to the letter of credit opened by Vinmar for the fifth trade, I did agree that [Wilmar] could present certificates of quality and quantity from Inspectorate … as conforming documents. However, in doing so, it was not my intention that this should have any effect on the underlying contract between Vinmar and [Wilmar]. As far as I was concerned, the binding test results for all three contracts were to be those provided by SGS. Had SGS found the cargo to be off specification upon its arrival at Houston, Vinmar would, in my view, have been entitled to reject the cargo on the basis of those results.

Mr Lokey when cross examined about contracts 5, 6, and 7 acknowledged that as regards the purchase by Vinmar Overseas of the parcel of 5,000 MT the Purchase Order Confirmation provided as regards inspection that quality was to be determined by measurement at ship tank. He also acknowledged that the certificates produced in evidence were all shore tank certificates. He added that if he remembered correctly there was some analysis done on the vessel’s tanks.

Mr Noldt in cross examination agreed that these contracts involved a distress sale, the cargo being already afloat and Wilmar’s intended sale having fallen through. The deal struck was that Vinmar would purchase the cargo in two lots, one batch of 5,000 MT, one batch of 7,000 MT and would sell the 7,000 MT immediately back to Wilmar. Mr Noldt was shown a Vinmar Overseas document described as “purchase order confirmation” dated 4 March 2008 for a quantity of 5,000 MT CIF Deer Park Texas (Vopak terminal). It was accompanied by the Vinmar Standard Purchase Terms (as set out above).

Mr Noldt was then taken to a contract form prepared by Wilmar for the parcel of 5,000 MT. This included:

INSPECTION:
A mutually agreed independent surveyor to be appointed by Seller. Their findings shall be final and binding for both parties, with exception for fraud or manifest error. Inspections will be carried out at load port for quantity and quality. The costs to be shared equally between seller and buyer.

Against this provision a representative of Vinmar – identified by Mr Lokey in the passage quoted above as Ms Bigg - wrote in manuscript, “SGS for both”. Mr Noldt accepted that Vinmar wanted SGS to determine the quality of this cargo for the purposes of the sale contract at the time of discharge.

Contracts 6 and 7, concerning the parcel of 7,000 metric tonnes, waived specifications and quality. Accordingly in my view they offer no assistance on the propositions identified in relation to Contract 1 above. As to Contract 5, the arrangements for documents under the letter of credit (described by Mr Lokey in the passage quoted above) are unusual. Nevertheless, it is clear that Vinmar Overseas was not prepared to accept Wilmar’s Standard Inspection Clause and insisted that the contract identify SGS as inspector of record whose ship tank determinations would be binding. Accordingly it, too, offers no assistance on the propositions identified in relation to Contract 1 above.

Contract 8 was the first contract in the course of dealing in which the contracting company from the Vinmar Group was Vinmar itself. Mr Lokey dealt with contract 8 at paragraph 20 of his witness statement as follows:

20.

The eighth trade between Vinmar and [Wilmar] was for the purchase by [Wilmar] of 4,000 metric tonnes of PME on a CIF Rotterdam basis at a price of US$1,150 per metric tonne. I sent a Sales Order Confirmation to [Wilmar] by email on 25 April 2008 …which indicated that we intended to appoint SGS as the inspector of record. As far as I recall, there was no specific discussion with [Wilmar] about the appointment of SGS. [Wilmar] simply accepted the appointment, which I do not find surprising in light of SGS's market leading position ….

Mr Lokey’s email of 25 April 2008 was addressed to Mr John Cummings and Mr Noldt. It attached a “sales order confirmation” which included the Vinmar Standard Sale Terms, but also stated in the body of the sales order confirmation:

INSPECTION: QUANTITY IS TO BE DETERMINED BY MEASUREMENT AT SHORE TANK. QUALITY TO BE DETERMINED BY MEASUREMENT AT SHORE TANK. INSPECTORS RESULTS TO BE FINAL AND BINDING ON BOTH PARTIES SAVE FOR FRAUD AND/OR MANIFEST ERROR. INSPECTORS COSTS TO BE SPLIT 50/50. SGS TO BE INSPECTOR OF RECORD.

A certificate of quality was issued by SGS dated 1 June 2008. It does not mention any sample or identify the location from which the product analysed was taken.

In cross examination Mr Lokey accepted that for contract 8 Vinmar had been the sellers, Vinmar had selected SGS, and Wilmar simply accepted that.

In cross examination Mr Noldt was shown the email sent to him and to Mr John Cummings by Mr Lokey on 25 April 2008. Mr Noldt said that his only involvement in the negotiations for this trade had been as a recipient of this email. It had been copied to him for information. He acknowledged that if Wilmar had wished to object to the appointment of SGS it could have done so. That had not happened.

On the material before me the contract terms agreed between the parties were those in Vinmar’s sales order confirmation. Those terms made specific provision for SGS to be the inspector of record, and for quality as determined by inspection of shore tank to be binding. There having been express contractual provision both about shore tank analysis, and about the identity of the inspector, it does not seem to me that contract 8 offers any assistance in relation to the propositions identified under contract 1 above.

Contract 9 is the last in the course of dealing between Wilmar and companies in the Vinmar Group prior to the August 2008 contract. Like contract 8, it was a CIF sale by Vinmar itself to Wilmar. It appears to be the first in which Starsupply acted as broker. On 13 June 2008 Starsupply emailed to Mr Lokey and Ms Sarwanto a “Broker Confirmation Note” addressed to Vinmar and Wilmar, giving reference 20806031. I shall refer to it as “BCN 20806031”. It stated, among other things:

We encourage Seller and Buyer to promptly exchange their own documentation.

…

Product/Quality:

PME meeting EN 14214 specification with the following guarantees:

Water ppm 400 max

…

Quantity:

2000 metric tons +/-5%, in seller’s option

Lifting/loading/delivery

CIF Vlaardingen during the period 5th–25th July 2008

Price:

USD 1270.00 / metric tonnes T1

Payment Terms:

Payment at sight. Documentary L/C.

Payment for the Product shall be made in USD by telegraphic transfer in immediately available funds, without any deduction, offset or counterclaim, at the counters of Seller’s designated bank, as stated in Seller’s Invoice 10 days after Bill of Lading date against presentation of Seller’s commercial invoice (telex or fax copy acceptable). Payment secured by standby L/C.

Buyer will remain responsible for payment in the event that payment is not made under the Standby Letter of Credit for any reason.

Determination of Quantity/Quality:

As ascertained by independent inspectors, whose results are final and binding for both parties, whereby inspection costs will be shared 50/50 between buyer and seller. safe fraud or manifest error.

A sales order confirmation dated 13 June 2008 and addressed to Wilmar was signed by Vinmar. It differed from BCN 20806031 in that payment terms were said to be “LC 10 Days”. It expressly incorporated Vinmar’s Standard Sales Terms – and when doing so appears to have twice set out the wording of clause 5 (including the reference to a “mutually agreed independent inspector whose decision is to be final … based on shore tanks samples at loading port”). Against “Other Terms” it stated “Broker Confirm # 20806031 (Starsupply)”.

Two certificates of quality were issued by SGS, one (dated 25 June 2009, presumably by mistake) for a parcel of 1,000 MT and another (dated 25 June 2008) for a parcel of 905.961 MT. Neither mentioned any sample and neither identified the location from which the product analysed was taken.

The sales order confirmation appears to have been attached by Ms Sarwanto to an email sent by her to Mr Noldt on 16 June 2008 giving details of Vinmar’s vessel nomination. This was acknowledged in an email sent by Manish Bhoopal of Wilmar on 19 June 2008 which set out Wilmar’s “document instructions” and requested that Saybolt be appointed as surveyors for 1,000 MT and SGS be surveyor for the balance, asserting that “the appointment of Saybolt as inspectors for 1,000 MTS is imperative.” On the same day Ms Sarwanto emailed Mr Bhoopal and Mr Noldt in reply:

…we cannot agree to have Saybolt as surveyor, since we only work with SGS.

Mr Bhoopal responded insisting on appointment of Saybolt as regards 1,000 MT. This led to an email from Ms Sarwanto on 23 June 2008 saying:

We really insist on using SGS for the survey on this shipment.

You can appoint Saybolt for your account, but the results and certificate of record for Vinmar point of view will be the ones from SGS.

Mr Lokey dealt with contract 9 at paragraph 21 of his witness statement:

21.

The last contract between Vinmar and [Wilmar] before the disputed contract was for the sale to [Wilmar] of 2,000 metric tonnes of PME, CIF Vlaardingen for shipment during the period 5-25 July 2008. The terms of the trade are documented in Vinmar's Sales Order Confirmation dated 13 June 2008, which confirmed that the inspector of record had to be appointed by mutual agreement … . The parties did discuss which inspector should be appointed in relation to this contract and, for the first time, [Wilmar] suggested the appointment of Saybolt. In our response to this proposal, we made it clear that we were not prepared to accept Saybolt as the inspector of record. Pamela Sarwanto, who participated in the relevant email exchanges, deals with this issue at more length in her witness statement.

In cross examination Mr Lokey was taken to the email exchange ending on 23 June 2008. Mr Lokey initially asserted that by this stage the vessel had already sailed. He was then shown documents indicating that berthing was not expected to be until 23 June, and accordingly it was not too late to appoint Saybolt. He then said it would have been duplication because all the rest of the cargo was being tested by SGS, adding that he had never used Saybolt once in the biodiesel business in the US gulf coast, North West Europe or South East Asia. He refused to agree that Vinmar’s approach was that, as seller, it could choose the inspector, even if it was not the inspector that the buyer wanted. He was then shown an email sent by Wilmar to Vinmar on 24 June 2008:

Hi Pamela,

Many thanks for your message. May we please request you to assist us by appointing Saybolt for our account just for 1000 mts (please note that we would need B/L split of 2 x 1000 mts). We agree that results and certificates of record for Vinmar point of view will be the ones from SGS.

Having stated the above, our agreement has always been that surveyors are to be appointed by mutual understanding with costs to be shared 50/50 between buyer and seller. Your insistence of using SGS is really not in the spirit of the contract.

…

Manish

It was suggested to Mr Lokey that this email demonstrated that Wilmar did not regard the appointment of SGS as “mutually agreed,” to which he replied “apparently not.”

In this context Mr Lokey was asked whether Wilmar were entitled to assume that messages from Ms Sarwanto stated the position of Vinmar. His answer to that question was “yes.” He then qualified this by saying that Ms Sarwanto did not have authority for “something to do with nominating an inspector other than SGS or dealing with the contract in itself…”. In response to a question from me, however, Mr Lokey accepted that Wilmar was entitled to proceed on the basis that Ms Sarwanto was the person whom Vinmar had appointed to sort out whatever needed to be sorted out as regards the arrangements for certificates.

In her witness statement Ms Sarwanto described the email exchanges set out above. She added that in response to Mr Bhoopal’s email of 24 June 2008 Vinmar arranged for the bills of lading to be split as requested, but the sole inspector of record under the contract remained SGS, “and we made clear our refusal to appoint Saybolt.”

In cross examination Ms Sarwanto was asked about her role. She said that it was not her job to decide who the inspector should be. If the inspector for biofuel was to be someone other than SGS, as a matter of Vinmar’s internal procedure, she would be expected to consult with Mr Lokey. She agreed that if she communicated with another party in the course of her work, that party could assume that she was stating the position of Vinmar. It was then suggested to her that if she wrote to Saybolt asking them to inspect a biofuel cargo, they would not need to ask Mr Lokey whether she had authority to appoint them. Her response was that she did not know. Ms Sarwanto was not asked questions in cross examination about contract 9.

It appears to me that the contractual terms for contract 9 are to be found in a combination of BCN 20806031, Vinmar’s sales order confirmation, and Vinmar’s Standard Sales Terms, along with Wilmar’s document instructions. There may be scope for argument as to whether Vinmar’s sales order confirmation, by incorporating BCN 20806031, had the effect that Vinmar’s Standard Sales Terms as to quality inspection were overridden by what was said as to determination of quantity/quality by Starsupply in BCN 20806031. There is no evidence, however, that either side focussed on what the position might be under any specific contractual term. The approach taken by Wilmar as buyer was to make a request of Vinmar as seller, and when Vinmar adamantly refused to accommodate that request, Wilmar reluctantly accepted the position. In these circumstances the course of events in relation to contract 9 offers no support for the propositions identified in relation to contract 1 above.

Reviewing the course of dealing generally, Mr Collett asked Mr Lokey about the distinction between whether, on the one hand, as a matter of fact a surveyor was mutually agreed at some point and whether, on the other hand, a contract provided for there to be mutual agreement upon the surveyor. A term to that effect had been put forward in Wilmar’s contract form for contract 4, as to which Mr Lokey replied he did not believe that he signed the contract. He acknowledged that while Vinmar’s Standard Sale Terms provided for a mutually agreed independent inspector, the Starsupply broker confirmation note which featured in contract 9 (BCN 20806031) made no reference to “mutually agreed”. He accepted that there were times when Vinmar contracted on the basis of a term other than that only SGS were to be used.

Mr Noldt accepted in cross examination certain matters put to him by Ms Pounds as follows:

“Q. Let me clarify that. It was not the parties’ practice to draw up a formal contract setting out all the terms of their agreement, which was then signed by both parties?

A. That is correct.

Q. …. you agreed with me earlier this morning, that a clause regarding the appointment of an inspector was an important clause, and what we have seen, by looking at the documents in D2, was that the parties tended not specifically to discuss and agree that, and then set it out in a document that both parties agreed to and signed.

A. That is correct.

Q. In fact, what tended to happen was that the parties only expressly agreed the core terms that we have discussed: price, quantity, shipment period, and then they proceeded on the basis that their business would be conducted as it had always been conducted before?

A. Correct

Q. In fact it was not necessary, was it, for the parties to specifically agree to a clause saying that the inspector was to be appointed by mutual agreement or was to be mutually acceptable, because both parties understood that there had to be mutual agreement between them as to the inspector to be appointed?

A. Agreed.

Q. We have also seen, have we not, that there were numerous occasions in the previous trade when the identity of a specific inspector, to be appointed on that occasion, was discussed between the parties?

A. Agreed.

Vinmar asserted that there was a course of dealing under which “there would be mutual agreement as to the appointment of an inspector whose determination would be binding.” If there is a contractual obligation to reach mutual agreement on an inspector, how is that to be performed? Plainly it will have been performed if both parties expressly discuss the matter and expressly reach agreement. Equally plainly it will not be performed if one side refuses throughout to accept any inspector proposed by the other – something which did not happen in the present case and the consequences of which I do not need to consider. Mr Lokey said that if one side put forward an inspector and the other did not object, then agreement was reached. In the context of this particular contractual obligation, and of the circumstances in which there is a failure to object, I agree with him. On the footing that the parties are contractually bound to reach agreement, then in my view once one side has put forward an inspector then, if the other party without protest proceeds with arrangements between the two parties for performance that would objectively be expected to be halted if the inspector were unacceptable, that will constitute agreement.

It is, however, both important and elementary to distinguish between what happened in the practical performance of past contracts and what the parties agreed would be terms of those contracts. Moreover, as identified by Mr Collett and not contested by Ms Pounds, there must be consistent inclusion of the term in question in the previous contracts, or the parties must have consistently conducted themselves as though it were included in their previous contracts, and in this regard the court is concerned with the objective appearance derived from the parties’ conduct.

Plainly it was not the case that each previous contract contained a term that the parties would reach “mutual agreement as to the appointment of an inspector whose determination would be binding.” For the reasons given individually above I do not accept that the parties conducted themselves as though such a term were included in each of their previous contracts – or, indeed, in any contract where this was not an express term. While I allowed Ms Pounds to question Mr Noldt about previous contracts, in my view the answers elicited from him did not constitute admissible evidence about contracts 1 to 7, of which he had no personal knowledge. Nor, on analysis, does it appear to me that Mr Noldt was accepting a proposition that the parties conducted themselves as though such a term were included in each of their previous contracts. His response to the questions put to him was consistent with a proposition that where both parties understood that there had to be mutual agreement between them as to the inspector to be appointed then it was not necessary for the parties specifically to agree to a clause saying that the inspector was to be appointed by mutual agreement or was to be mutually acceptable. If Mr Noldt went beyond this, then my analysis of the course of dealing leads me to conclude that he was wrong to do so. Accordingly I find that there was no course of dealing as to the terms of past agreements that could be incorporated into the August 2008 contract.

As to Ms Sarwanto’s authority, I find, in accordance with the answer given to me by Mr Lokey, that Wilmar was entitled to proceed on the basis that Ms Sarwanto was the person whom Vinmar had appointed to sort out whatever needed to be sorted out as regards the arrangements for certificates. In my view that included the identity of the inspector. For the reasons given earlier there was nothing inappropriate in Wilmar’s decision, when Vinmar requested a certificate of analysis, to forward the Saybolt certificate. The course of dealing showed that Vinmar had a strong preference for SGS, and had insisted for contract 9 that SGS be inspector of record and not Saybolt, but it did not follow that Vinmar would not accept Saybolt in the circumstances of the August 2008 contract’s January shipment. Vinmar’s communications to Wilmar in relation to the use of Saybolt for Contract 9 had been sent by Ms Sarwanto. Wilmar was entitled to proceed on the basis that Ms Sarwanto was the person who was to be regarded as acting on behalf of Vinmar in relation to the identity of the inspector as well as other arrangements for certificates.

As noted in section D above, Starsupply acted as broker in concluding the August 2008 contract. It acted for both sides. In his oral evidence Mr Lokey described the mechanics. Starsupply’s brokers sit around a table. Each represents a potential party to transactions under negotiation, and maintains contact by telephone with that party. Where the potential counterparty is represented by another Starsupply broker, the two brokers will shout their clients’ position across the room.

What happened on 27 August 2008 was that Mr Noldt was on the telephone to a broker called Joachim of Starsupply, and Mr Lokey was on the telephone to a colleague of Joachim’s named David. Under discussion was a shipment with the contractual specification set out in section C above. In their oral evidence each of Mr Lokey and Mr Noldt gave an account of what was expressly discussed between the parties. It was common ground that the express discussion between each party and its broker, and between the two brokers shouting their clients’ position across the room, was confined to what Mr Noldt described as “the big points.” They can be grouped under three broad heads, described by Mr Lokey as “price, quantity and timing”. As to each of these:

It was common ground that these were the core terms of the contract and that the parties proceeded upon the basis that the fine details would be worked out later.

On the following day, 28 August 2008, Starsupply sent emails to each of Wilmar and Vinmar. Each was headed, “Starsupply Renewables SA Broker Confirmation Note”. The email to Wilmar noted that a commission of $2 per metric tonne was to be paid by it to Starsupply, while that to Vinmar noted that a similar commission of $1 per metric tonne was to be paid by it. Apart from that, and apart from some minor differences in the introductory words, each email was materially identical. I shall refer to the common provisions of the email as the “Broker Confirmation Note.” It included the following:

“Starsupply Renewables S.A., as brokers only, herewith confirm following transaction …

We encourage Seller and Buyer to promptly exchange their own documentation.

...

Seller:

[Wilmar]

...

Buyer:

[Vinmar]

...

Product/Quality:

FAME meeting EN 14214 specification with the following additional garantees:

Water content 400 ppm max

CFPP degr C 0 max

Virgin vegoil

…

Quantity:

3000 metric tons +/-3%, in buyer’s option

Lifting/loading/delivery

FOB ARA [Amsterdam / Rotterdam / Antwerp], during the period 1^st January 2009 – 31^st March 2009

3 x 1000 mt/per month

Price:

USD 1328.00 / metric tonnes T2

Payment Terms:

Payment for the Product shall be made in USD by telegraphic transfer in immediately available funds, without any deduction, offset or counterclaim, at the counters of Seller’s designated bank, as stated in Seller’s Invoice 5 days after Bill of Lading date against presentation of Seller’s commercial invoice (telex or fax copy acceptable). Payment secured by standby L/C.

Buyer will remain responsible for payment in the event that payment is not made under the Standby Letter of Credit for any reason.

Determination of Quantity/Quality:

As ascertained at loadport by independent inspectors, whose results are final and binding for both parties, whereby inspection costs at loadport will be for seller’s account, safe fraud or manifest error.

Law and Jurisdiction:

This contract shall be governed and interpreted by English law and each party expressly submits to the Jurisdiction of the London High Court. No recourse to arbitration.”

It may be noted that the Broker Confirmation Note encouraged each of Wilmar and Vinmar “to promptly exchange their own documentation.” Vinmar did not respond to this invitation. Indeed, Mr Lokey accepted in cross examination that no one from Vinmar ever suggested to Wilmar that any of the terms in the Broker Confirmation Note were unacceptable to Vinmar or incorrect or incomplete.

By contrast on 29 August 2008 Wilmar emailed Vinmar attaching a draft contract for signature by Vinmar. I shall refer to this document as “the Wilmar Contract Form.” It identified Wilmar as seller, Starsupply as Broker, and Vinmar as buyer, and was headed “CONTRACT #: 1040001207”.

Wilmar’s written closing submissions identified five respects in which the Wilmar Contract Form either differed from the Brokers Confirmation Note or contained additional provisions. I do not understand Vinmar to dispute that these were features which did indeed differ from the Broker Confirmation Note. They were as follows:

(1)

Payment was to be made three New York banking days after receipt of fax or email copies of commercial invoice from Seller.

(2)

Vinmar were to open a standby letter of credit for the full duration and value of the contract, to be provided to Wilmar latest 20 December 2008.

(3)

Upon payment, Wilmar were to provide Vinmar with original: Commercial Invoice, Bills of Lading, Certificates of Quality and Quantity issued by an independent surveyor, and Certificate of Origin.

(4)

There were detailed nomination, laytime, demurrage and force majeure provisions, and Incoterms 2000 for FOB contracts were to apply.

(5)

The Wilmar Contract Form also expressly provided:

10.

quality & quantity

quality final as per independent surveyor analysis at load port basis volumetric composite from nominated shore tank(s). quantity ascertained by an independent surveyor and final at load port and as per b/l. ...

11.

inspection

a mutually agreed independent surveyor to be appointed by seller, their findings shall be final and binding for both parties, with exception for fraud or manifest error. inspections will be carried out at load port for quantity and quality. the costs to be shared equally between seller and buyer.

Vinmar appears neither to have acknowledged nor to have replied to Wilmar’s email of 29 August 2008. No further dealings between the parties appear to have taken place as regards the August 2008 contract during the period between 29 August and 16 December 2008. It is convenient at this point to consider what the contractual position was during this period.

In its particulars of claim Wilmar asserted that the agreement between the parties was contained in the Broker Confirmation Note. Vinmar’s defence accepted that the agreement between the parties was evidenced in part by the Broker Confirmation Note, and that, subject to two qualifications, the terms of the Broker Confirmation Note were terms of the contract. The first qualification was that the Broker Confirmation Note did not contain the entire terms of the agreement between the parties, nor did it constitute the only evidence of such terms. The second qualification followed on from the first. It was set out in paragraph 9 of the defence, and asserted that in addition to the terms in the Broker Confirmation Note there were two further terms agreed between the parties as follows:

9.1

The appointment of the independent surveyor(s) at the load port was to be mutually agreed between the parties. Further and/or alternatively (and strictly without prejudice to the Defendant’s primary case as aforesaid), the only independent surveyor(s) that would be used were either SGS or ICB; and

9.2

The findings of the independent surveyor(s), as thus appointed (and no other), were to be final and binding on both parties in relation to the quality and quantity of the shipment.

As to how these terms came to be agreed between the parties, paragraph 9 identified six different ways in which it was said that this had come about. One of these was by reason of the course of dealing between the parties. I have rejected that contention earlier in this judgment, and say no more about it. The remainder can conveniently be described as contentions that these terms were agreed between the parties because:

Wilmar served a reply which admitted that the results of the “independent inspectors”, either mutually agreed between the parties or selected by Wilmar, were to be final and binding on both parties as to quality and quantity. However it denied that there was any basis for the implication of any term whatsoever in relation to the load port surveyor. In this regard, it noted that the Broker Confirmation Note already provided expressly for “determination of quantity/quality” and required only that the load port surveyors be “independent inspectors”. Wilmar noted that the “determination of quantity/quality” clause in the Broker Confirmation Note provided for the “independent inspectors” to be for Wilmar’s account. Upon a proper construction of that clause, said Wilmar, in the absence of agreement between the parties Wilmar was entitled to select the load port surveyors provided only that they were “independent inspectors”.

By an amendment made to its defence in February 2010, Vinmar inserted two new paragraphs, 9A and 9B. They referred to the Wilmar Contract Form (described as “the Wilmar Contract”), and placed reliance upon it:

9A. As evidence in support of its case as aforesaid, the Defendant will rely (without limitation) upon the terms of a written contract subsequently drawn up by [Wilmar], but in the event not signed by the Defendant, which written contract (“the Wilmar contract”) provided inter alia as follows:

“11.

INSPECTION

A MUTUALLY AGREED INDEPENDENT SURVEYOR TO BE APPOINTED BY SELLER. THEIR FINDINGS SHALL BE FINAL AND BINDING FOR BOTH PARTIES, WITH EXCEPTION FOR FRAUD OR MANIFEST ERROR. INSPECTIONS WILL BE CARRIED OUT AT LOAD PORT FOR QUANTITY AND QUALITY. THE COSTS TO BE SHARED EQUALLY BETWEEN SELLER AND BUYER.”

9B. For the avoidance of doubt, the Defendant does not contend that the Wilmar contract as a whole accurately contains and/or evidences all the terms of the Contract as agreed between the parties, which were as set out in paragraphs 8 and 9 above. The Defendant does say, however, that the provision in clause 11 of the Wilmar contract that the independent surveyor was to be “mutually agreed” evidences an agreed express term of the Contract and was in accordance with the course of dealing between the parties and the terms that are customary in such contracts.

Towards the end of his oral evidence, when responding to questions from me, Mr Lokey said that he did not know what terms may have been discussed between Starsupply and Wilmar. However if he had read the Broker Confirmation Note he would have raised points on it with his broker (“David”). He had not done so, and accordingly his evidence to the court was that he had not read the Broker Confirmation Note. Under further cross examination by Mr Collett, Mr Lokey went so far as to say that he had not appreciated that a Broker Confirmation Note would go beyond price, quantity and timing.

When giving oral evidence Mr Noldt, in the course of cross examination by Ms Pounds, said that it was never intended by either Wilmar or Vinmar that the Broker Confirmation Note was to be the contract. When responding to questions from me he asserted that Wilmar always sent out its own contracts.

It is common ground that it is possible for parties to agree the principal terms of a contract, which will make it a binding contract, while leaving over other terms to be agreed at a later stage. If one party then makes a proposal for a particular term or terms and the other does not object to it when asked if it has objections, that can, in appropriate circumstances, be taken as acceptance of the proposal: see The Harriette N [2008] 2 Lloyds Reports 685 at paragraph 70. Applying this approach, if neither party had objected to the terms in the Broker Confirmation Note (which expressly encouraged the parties to exchange their own documentation), then I would have concluded that, despite what was said in oral evidence by Mr Lokey and Mr Noldt, the conduct of the parties was such as would objectively indicate that both sides accepted its terms.

However, that is not what happened. On 29 August 2008 Wilmar sent Mr Lokey and Ms Sarwanto the Wilmar Contract Form, adding “Please read and sign and return to our Singapore office.”

In my view the Wilmar Contract Form was plainly put forward by Wilmar as a document which would contain the entirety of the contract terms between the parties. In Wilmar’s closing submissions it was urged that the Wilmar Contract Form was merely an offer to contract on the terms that it set out. A contention was advanced that in itself it did not indicate that Wilmar was unwilling to contract on the terms of the Broker Confirmation Note, pending agreement of more detailed terms. I cannot accept this contention: when sending the Wilmar Contract Form no reference whatever was made to the Broker Confirmation Note, and the only inference that can be made in that regard is that Wilmar was not content with the terms in the Broker Confirmation Note. It is accordingly impossible to conclude that there was agreement between the parties on the terms in the Broker Confirmation Note.

If follows that it is not necessary for me to determine contention (f) above (i.e. whether, as Wilmar contended, the quality determination provision in the Broker Confirmation Note permitted Wilmar to appoint the inspector without seeking agreement from Vinmar as to the identity of that inspector). Had it been necessary, I would have upheld Wilmar’s contentions in this regard. The words used in the quality determination provision do not provide for any involvement on the part of the buyer in the selection or appointment of the “independent inspectors”. Vinmar itself had carefully drafted standard provisions which distinguished between mutual agreement on a surveyor in the Vinmar Standard Sales Terms, and, by contrast, made no reference to any such mutual agreement in Vinmar’s Standard Purchase Terms. For the reasons given earlier, the previous course of dealing does not lead to any contrary conclusion.

My analysis of events is inconsistent with contention (a) above. As to contentions (b) to (e), there was no evidence of custom. I am not persuaded that there is any need to imply a requirement for a mutual agreement: on the contrary, a seller may well have good commercial reasons for wishing to be able to specify the independent inspector, and a buyer in the marketplace may be willing to acknowledge those commercial reasons and agree to a contract on those terms.

It was suggested in Vinmar’s written closing submissions that the parties should be taken to have agreed on clause 11, the inspection clause, in the Wilmar Contract Form. I find it impossible to see how Wilmar’s conduct in sending the Wilmar Contract Form to Vinmar can enable the court to say that the parties had agreed on this specific term. In oral submissions there appeared to be a more general contention that the Wilmar Contract Form should be taken to have set out the terms of agreement between the parties. As to that, however, the court must make an objective analysis. What was said by Wilmar in the email of 29 August was that Vinmar was asked, in relation to the Wilmar Contract Form, to read and sign and return it to Wilmar’s Singapore office. Vinmar did not do what was asked. This is not a case of the kind postulated in the Harriette N – Wilmar had not asked Vinmar to state whether it had objections to the Wilmar Contract Form– and accordingly mere failure on the part of Vinmar to object cannot be taken as acceptance.

It will be recalled that Wilmar admitted a part of Vinmar’s case, accepting that when the parties mutually agreed on a named surveyor that carried with it agreement that the certificate of the named surveyor would be binding, and in any event making an assertion that such a provision became part of the August 2008 contract. I do not consider, however, that this can survive my analysis of the course of dealing and other matters relied on by Vinmar. The findings which I have made when rejecting Vinmar’s case are incompatible with Wilmar’s assertions in that regard, and I do not consider that in those circumstances Wilmar can rely on any part of Vinmar’s case as enabling Wilmar to make good this assertion.

Accordingly I conclude that in the period up to 16 December 2008 the parties had agreed only on the core terms of the August 2008 contract. Agreement as to anything else would have to wait until one or other party identified a need to reach agreement on something additional to the core terms.

On 16 December 2008 Ms Schmidt emailed Mr Lokey and Ms Sarwanto stating:

Reference: Our contract number 1040001207 dated 28 Aug 2008 (Enclosed herein for your ready reference)

May we please request your good selves to nominate vessel to load the January 2009 portion of above contract… we write to confirm that cargo will be in readiness to load as from 0001 hours on 01/Jan/2009.

Further, may we request your goodselves to please establish SBLC for the sake of good order.

As was acknowledged by Mr Noldt in cross examination, the last sentence of this email referred to the standby letter of credit which the Wilmar Contract Form envisaged would be provided no later than 20 December 2008. Mr Noldt accepted that Wilmar was asking Vinmar to comply with its obligations under the Wilmar Contract Form, adding:

On that date this was our assumption and that the contract would operate.

Vinmar did not reply to the email of 16 December, and Mr Noldt sent a chasing email on 23 December.

On 5 January 2009 Ms Schmidt emailed Saybolt with a copy to Vopak. She asked Saybolt to take a sample from ST 7819 and analyse the water content. If the water was below 400ppm they were to arrange for a full analysis “as per EN 14214.” She instructed Vopak to allow Saybolt to take a sample from ST 7819. In her witness statement Ms Schmidt explained that Wilmar wished to use Saybolt as inspector because the cargo in ST 7819 was to be sold to other buyers as well as Vinmar, and it made sense to have one certificate of analysis that could be used for all onward sale contracts.

In cross examination Mr Noldt accepted that while ST 7819 was a blend of different qualities of biodiesel Wilmar also had available in ST 7744 a cargo of Canola Methyl Ester which had a low water content and, for that and other reasons, was much more valuable than the cargo in ST 7819. When it appeared that the cargo in ST 7819 was within the specifications for the August 2008 contract Wilmar took a business decision that it would ship that cargo to Vinmar and use the more valuable cargo in ST 7744 for other purposes.

Also on 5 January 2009 Mr Noldt chased Mr Lokey and Ms Sarwanto seeking “your feedback on the opening of the SBLC…” Again, there was no reply by Vinmar. A further email was sent, on this occasion by Mr Dario Cavalheiro, on behalf of Wilmar on 7 January 2009. The email was addressed to Ms Sarwanto, stating:

Please find enclosed herein contract 1040001207 for the upcoming delivery of FAME… Please ensure, in addition to the below guidelines, that the SBLC will be opened in accordance to contractual payment clause below.

The remainder of the email set out relevant parts of clause 8 of the Wilmar Contract Form, along with guidelines for use when issuing letters of credit. It prompted a reply from Ms Sarwanto on 8 January stating that she was awaiting “Opening Bank details from our Houston office.”

On the following day, Friday 9 January 2009 Ms Sarwanto emailed Ms Schmidt asking her to advise where the cargo would be available. Ms Schmidt replied the same day that the cargo would be available at the Vopak terminal at Vlaardingen, and asked Ms Sarwanto whether she knew when Vinmar would pick up the cargo. Ms Sarwanto promptly replied nominating the Varkan Ege, setting out various details including a lay/can of 10/11 January 2009 and terms as to lay time and demurrage, adding:

Pls send us your confirmation and COA soonest.

Ms Sarwanto acknowledged in cross examination that this was a request that Wilmar supply a certificate of analysis for the cargo that it intended to load.

As noted earlier, on 7 January 2009 Vinmar had chartered Varkan Ege, intending to fulfil the INEOS contract by blending Wilmar’s January cargo and other parcels of FAME on board that vessel at Vlaardingen. For this purpose on 9 January Ms Sarwanto gave SGS and Vopak, among others, documentary instructions for the entire cargo which was to be sold to INEOS. The instructions referred to SGS as the load port inspector. She sent a further email later that day to SGS among others nominating Varkan Ege under the INEOS contract, and stating that approximately 950 MT were to be trans-shipped from barge Heinz, with the balance of 950 MT “ex-tank Wilmar (sht no TBA)”. Ms Sarwanto accepted in cross examination that neither of these instructions specifically requested SGS to sample any shore tank. Later in cross examination Ms Sarwanto said that since Vinmar had been working with SGS for so long, if she sent them a nomination then SGS would normally take a shore tank sample “as soon as we nominate”.

The next relevant communication was between Wilmar and Vinmar. At 17.54 local time Ms Schmidt emailed Ms Sarwanto:

As per previous…

Please find below analysis report of Saybolt

What was “below” was a forwarded email from Saybolt to Wilmar dated 7 January 2009 and setting out what were described as “results from shore tank 7819”. Among other things, Saybolts’s email stated:

…

Sample: samples drawn by Saybolt…

Sample marked as: shore tank 7819
single tank composite sample.
composite 06/01/2009 – 693362-

…

Water content MG/KG ENISO12937 390

It was common ground that as Ms Schmidt had not been involved in any of the previous trades between Vinmar and Wilmar she would not have known that previously Vinmar had not been prepared to accept a quality certificate from Saybolt. It was suggested to Mr Noldt in cross examination that Wilmar should have been aware that it needed to discuss with Vinmar which inspector was to be appointed. Mr Noldt agreed. I also agree, because on my analysis Wilmar did not have the benefit of the quality determination clause in the Broker Confirmation Note (which would have permitted it to appoint an inspector unilaterally). If, as Wilmar were assuming, the Wilmar Contract Form contained the terms which applied to the August 2008 contract, then there would have been an express contractual obligation to reach mutual agreement on the inspector. I will turn shortly to consider whether such an assumption was correct.

As to discussing which inspector was to be appointed, Mr Noldt’s answer in cross examination was that upon request Wilmar had sent the certificate of analysis from Saybolt. In my view there was nothing improper about this. There were commercial reasons why a certificate from Saybolt was desirable in any event. Such a certificate had been obtained on 7 January 2009 and showed that the FAME in ST 7819 was within the specifications of the August 2008 contract. Vinmar had made no request as to the appointment of a load port surveyor: on the contrary, Ms Sarwanto’s request for “your confirmation and COA soonest” at least envisaged the possibility that Wilmar already had available a certificate of analysis from a surveyor that it had chosen. Given that Ms Sarwanto was nominating Vargan Ege on 9 January with a lay/can of 10/11 January, it is hardly surprising that she did not specify that the certificate of analysis which she sought should be supplied by a named surveyor.

In cross examination Ms Sarwanto said in relation to Ms Schmidt’s email forwarding the Saybolt analysis that she did not “recall actually receiving it or reading it at the time.” She nevertheless accepted that she would have been interested in the email in order to see whether the cargo was on spec, and because it was the first time that Wilmar identified the number of the shore tank, which was an important piece of information. When it was suggested that she would have noticed that the inspector was Saybolt, Ms Sarwanto replied that she “would have read it.” When asked to confirm that she had not made any response to Wilmar, the following exchange occurred:

A.

If I would have actually not overlooked this email and certificate, I would have noticed it was Saybolt and then I would have contacted Chris Lokey in order to get this discussion up and running to see if it was acceptable to Vinmar, yes or no.

Q. Is it your evidence that you now believe that you did completely overlook this email?

A. As I recall it, yes.

Ms Sarwanto also said in cross-examination that she assumed that there had been discussions between Wilmar and Mr Lokey about the identity of the inspector. For reasons which I explain in section M below I am satisfied that Ms Sarwanto did not overlook this email at all.

At 0001 hours on Saturday 10 January 2009 Varkan Ege tendered notice of readiness. She did not, however, proceed to a berth. Also on that day Ms Sarwanto emailed Ms Schmidt attaching the terms of a standby letter of credit dated 9 January 2009 making reference to a Vinmar purchase order number 4302090. Although the Wilmar Contract Form required a standby letter of credit for the full amount payable under the August 2008 contract, the letter of credit forwarded by Ms Sarwanto covered only the amount of the January shipment.

On Monday 12 January Mr Cavalheiro emailed Ms Sarwanto and Ms Schmidt confirming receipt of the standby letter of credit, but stating that Wilmar could not accept the reference to Vinmar purchase order number 4302090. With that exception, he stated that everything else was in order.

Mr Cavalheiro added that if a contract reference was required then Wilmar could only accept a reference to its sales contract [i.e. the Wilmar Contract Form], which he attached. He also requested a copy of Wilmar’s purchase order. None was supplied.

Later that day Ms Sarwanto emailed for Wilmar’s approval a letter of credit amendment application. This stated that in lieu of the reference to Vinmar’s purchase order number 4302090 the words “contractual agreements” should be substituted. At 10.39 local time Mr Cavalheiro emailed Ms Sarwanto and Ms Schmidt confirming acceptance of the amendment application, and instructing Ms Schmidt that the cargo should be released to Vinmar.

Ms Schmidt and Ms Sarwanto used both email and instant messaging to communicate on 12 January. By Instant Messaging, Ms Sarwanto said,

Just received Dario’s message. So you will release the cargo at Vlaardingen now? Tank 7819 right?,

Ms Sarwanto was asked by Mr Collett where she obtained the information that the relevant shore tank was number 7819 if it was not from Ms Schmidt’s email forwarding the Saybolt certificate. She replied that she did not recall where she actually got that information from, at what time or on which date.

Ms Schmidt then sent an email accepting the nomination of Varkan Ege. Aninstant messaging exchange then occurred:

CS: should we use Saybolt or SGS?
PS: SGS pls
CS: okay – did you order them a[l]ready
PS: Yes, I did
CS: okay – thanks
PS: np [no problem]

It will be recalled that on 9 January 2009 Ms Sarwanto had given SGS documentary instructions as load port inspector for the entire cargo which was to be sold to INEOS, and that later that day she emailed SGS with the information that a balance of 950MT was to come from a Wilmar shore tank whose number would be advised. Ms Schmidt in her witness statement pointed out that if the instruction to SGS on 9 January were indeed an instruction to carry out a quality analysis, then that instruction ought to have been copied to Wilmar at the time – without Wilmar’s permission, SGS were not permitted to access and sample from Wilmar’s tanks. Ms Schmidt added that by 12 January she had received no objection to the Saybolt certificate from Vinmar, nor had there otherwise been any discussion of the identity of the independent inspector for the purposes of quality. As far as she was concerned “there was no question of Saybolt not being accepted.” Indeed, the fact that Vinmar requested release of the cargo appeared to her to be further confirmation that Vinmar were prepared to go ahead on the basis of the Saybolt certificate that she had provided. Accordingly, in the instant messaging exchange that occurred on 12 January 2009, when she asked whether Saybolt or SGS should be used, she was referring to use of an inspector for a quantity survey and not a quality survey. That had been done for two reasons. The first was “just to be polite.” The second was that she thought it likely that Vinmar might already have appointed someone to attend loading and determine the quantities of each parcel that was to be loaded on Varkan Ege, in which event it would make sense for the same inspector to produce quantity certificates for all the parcels. She added that it is not unusual to have different surveyors for quantity and quality.

Ms Sarwanto said that she understood Ms Schmidt to have been referring to inspection of the cargo both in terms of quality and quantity. If Ms Schmidt had not proposed SGS as an option, she would have had to refer the matter to Mr Lokey for a decision.

The next event was that Ms Schmidt sent an email to Vopak and SGS attaching the loading advice for the “Varkan Ege”. Neither the email nor the attached loading advice made any reference to quality inspection. The advice stated:

Remarks: Messrs SGS: Please do prepare the loading doc for the vessel along with the quantity survey.

On 13 January 2009 Ms Sarwanto emailed SGS and others giving loading instructions for three parcels to be loaded onto “Varkan Ege”, one of which was the parcel ex shoretank 7819:

Surveyor: SGS
Instructions: Loading and T/S [transhipment] control, Quantity and Vessel Comp after loading MT Varkan Ege. Cargo to be blended onboard Mt Varkan Ege

In cross examination Ms Sarwanto was asked why her full loading instructions to SGS among others were not sent until 13 January. She replied that this was because the instruction dealt with all parcels that were to be loaded on Varkan Ege. In response to further questions from Mr Collett she accepted that this did not explain the delay and she did not know why the delay in question had occurred. She accepted that her instruction to SGS on 13 January 2009 did not make reference to taking a sample from a shore tank. It was in this context that she gave the answer about the regular practice of SGS referred to earlier.

It is common ground that at 0530 on 14 January Varkan Ege left anchorage in order to load the Heinz parcel, which had previously been transhipped into a shore tank. At 0915 Vopak received Notice of Readiness from Varkan Ege for this purpose. That loading operation continued until 0125 on 15 January.

At between 0930 and 1000 on 14 January, SGS informed Ms Sarwanto that the water content of ST 7819 was 435 ppm. In cross-examination Ms Sarwanto said that she contacted Mr Lokey and consulted him straight after she received the call from SGS. Mr Lokey said that he was not aware of the SGS analysis result until after he arrived at his office in the morning (Houston time) of 14 January, and that he only learned of the SGS result from reading Ms Sarwanto’s emails. As Wilmar points out, since the relevant conversations began in the early hours Houston time, it seems unlikely that they had simply slipped Mr Lokey’s mind. However in fairness to Mr Lokey this conflict only emerged when Ms Sarwanto gave oral evidence and after his oral evidence was concluded. I do not need to make any finding as to whose recollection is correct.

At 1022 on 14 January, Ms Sarwanto emailed Ms Schmidt

We received the following from SGS:

Subject: LOADING VARKAN EGE / SHT 7819 / WATER CONTENT

Good morning,

Water content shoretank 7819 before loading mv VARKAN EGE : 435 ppm

...

As you can see the water content is too high, it needs to be less than 400 ppm. Pls check from your side and confirm.

The message caused consternation to Mr Noldt and Ms Schmidt, each of whom responded in that regard.

A further email was sent at 1240 by Ms Sarwanto on the instructions of Mr Lokey:

This morning we have been informed by independent surveyor SGS that the water content of your cargo in tank 7819 at Vopak Vlaardingen has a water content of 435 ppm, which does not meet the specifications as per the contractual agreement.

Therefore Vinmar has not other option than to put Wilmar Oleo Pte Ltd. On notice and hold you fully responsible for all additional charges, including but not limited to Demurrage, which might occur due to this situation.

You will have till cob Rotterdam time today to supply us with substitute cargo, which is in conformance with the contractual agreement.

In case you will not be able to do so, Vinmar holds the right and will consider this contract null and void, due to non performance from your side.

Mr Noldt replied at 1303:

thank you for your below email which contents we do reject.

We do have sent you on the 9^th January the final analysis of tank 7819 basis on the sample which Saybolt did draw on the 7^th January.

We did agree to SGS for the quantity survey.

At no point and time was SGS allowed to draw a sample from tank 7819. Hence the email you are forward to me does not even exist as SGS does not have permission to draw samples from tank 7819.

Basis the coa as issued by Saybolt, we do have fulfilled our contractual obligations and the cargo is by all means ready to be loaded on board of the Varkan Ege.

We do expect Vinmar to load the cargo and not unnecessarily block the berth at Vopak Vlaardingen.

Thank you for your confirmation to load the cargo by return.

At 1658, Mr Noldt chased Vinmar for a reply to his email, adding that, according to SGS, the barge had not even started discharging yet, and it would need 6 hours for the discharge. In fact, at 1658 the “Varkan Ege” was loading the “first foot” of the barge “Heinz” cargo.

At 1808 Mr Lokey emailed Wilmar:

Regarding FAME 0 ex sht 7819 at Vopak Vlaardingen:

...

Wrongfully, and in breach of contract you have tendered for delivery a cargo of FAME 0 with a water content of 435 ppm. The contractual specification for water content is 400 max, and accordingly the cargo you are tendering is not of contractual specification. The surveyors report tendered by you has no validity as at no time did we mutually agree the appointment of Saybolt as the inspection company.

In our message of earlier today we indicated to you that we reserved all of our rights in respect of your breach of contract, and we called upon you to cure your breach of contract by providing us with a substitute cargo by close of business Rotterdam time. Close of business has now passed, and you have failed to cure your breach of contract. Wrongfully and in breach of contract you have therefore failed to deliver to us a cargo for the January shipment. We will therefore not be taking delivery of your off spec cargo, and we will seek to mitigate our loss by seeking an alternative cargo, and by minimising the demurrage we incur on the ship waiting to load the cargo. All our rights our reserved in connection with all losses we incur as a consequence of your breach of contract, including but not limited to any losses we incur to third parties. All and any losses we incur will be for your account.

Please note that the contract for the remaining two shipments of 2000 MT remains ongoing and valid, and we will in due course in accordance with our contractual obligations nominate a performing vessel for the February shipment.

At the time when this email was sent, Varkan Ege was awaiting completion of the analyses of the “first foot” sample from the barge “Heinz” cargo. The analyses were completed and accepted and loading resumed at 1910, continuing until 0110 on 15 January.

Under cross-examination Mr Lokey accepted that after 1808 hours Dutch time on 14 January Vinmar had no intention of performing the January shipment, whatever cargo Wilmar might offer

At 1847 Mr Noldt wrote to Vinmar (D1/122). He asked them “to cooperate” and stated:

We do await your new nomination for the cargo which is available for loading at Vopak latest on the 29^th January for loading latest on the 31^st January. ... I do regret that it is not possible to talk to any representative of vinmar in order to overcome the issues.

Mr John Cummings of Wilmar also had discussion with Mr Friedman of Vinmar, which he reported to Mr Noldt:

Bernd – I did have a long conversation with Doug this evening and I think he heard all the arguments. ...

Basically, the discussion with Doug was:

1)

Vinmar received Saybolt COA on Jan 9 and said nothing in regards to disputing Saybolt until Jan 14 which on the surface is unacceptable. The COA was sent to Pam but it should be Pam’s issue if she did not send to Chris not Wilmars as we sent COA to Vinmar.

2)

If Vinmar wants to contend that it did not mutually agree to [Saybolt], Wilmar can similarly contend it does not mutually agree to SGS as we have had numerous inconsistencies in recent months with various SGS offices. So then the task is to agree a mutually acceptable inspector

3)

There seems to be an option to add 20% lower water PME to the -3 CFPP product to lower water but also increase CFPP. I told Doug that this was possible but there needs to be a dialogue to discuss which was not the case this evening Hamburg time, afternoon Houston. Chris nor Pam returned the calls Bernd made. ...

Until receipt of Ms Sarwanto’s email of 10 January 2009 those in Wilmar simply assumed that the Wilmar Contract Form set out the terms of the August 2008 contract. As I explain below, from the time they received Ms Sarwanto’s email of 10 January 2009 they could not and did not simply assume that this was so. In any event, the assumption made up to that time was wrong. For the reasons given earlier as at 16 December 2008 the Wilmar Contract Form did not set out the terms of the August 2008 contract. Nothing had happened from 16 December 2008 until 12 January 2009 to alter that position. At the outset of 12 January 2009, in my view, it remained the case that the only terms agreed between the parties were the core provisions.

On 12 January 2009 a compromise was arrived at for the purposes of the standby letter of credit. Vinmar was aware that Wilmar contended that the Wilmar Contract Form was the governing contract, and a copy of the Wilmar Contract Form had been supplied to Vinmar on several occasions. Wilmar became aware when it received Ms Sarwanto’s email of 10 January 2009 that Vinmar had told its bankers that the contract between the parties involved Vinmar’s purchase order number 4302090, albeit that no copy of this purchase order had been supplied to Wilmar. While Wilmar had said that the only contract reference it could accept was a reference to the Wilmar Contract Form, Vinmar had come back with a proposal that the reference should be simply to “contractual documents.” The choice of those words in my view plainly reflected a recognition on Vinmar’s part that it would not be possible that day to reach agreement on which documents set out the terms of the August 2008 contract. It is equally plain in my view that Wilmar’s acquiescence in the proposed amendment involved a similar recognition.

I conclude that the course of events to the end of 14 January 2009 involved no other change to the contractual position in relation to anything other than the amendment to the standby letter of credit.

In his earlier email on Monday 12 January 2009 Mr Cavalheiro had identified only one discrepancy in the standby letter of credit (the reference to Vinmar purchase order number 4302090) and had said that everything else was in order. If, contrary to my view, the Wilmar Contract Form applied then by this email Wilmar waived the obligation for it to cover all three shipments. On the view which I take, however, all that happened in this regard by the end of 12 January 2009 was that in addition to its rights under the core provisions which comprised the August 2008 contract, Wilmar had gained a promise that an amended standby letter of credit would be issued in the terms discussed.

However, if the contract had contained a provision that the findings of a mutually agreed load port surveyor would be binding as to quality, then in my view the Saybolt certificate would have constituted such findings and SGS C04 would not. Ms Sarwanto said that she overlooked Ms Schmidt’s email forwarding the Saybolt certificate. I am satisfied that Ms Sarwanto did not overlook this email at all. From 9 January onwards she proceeded on the basis that the cargo to be loaded under the August 2008 contract was in ST 7819, and can only have done this because Ms Schmidt’s email forwarding the Saybolt certificate identified this particular shore tank. In cross-examination Ms Sarwanto was unable to identify any other way in which she could have learnt of this number prior to referring to it in the instant messaging exchange on 12 January. Her instructions to SGS on 13 January referred to “Quantity and Vessel Comp” after loading on Varkan Ege, and made no mention of sampling ST 7819. Even if it had been the regular practice of SGS to take quality samples once advised of a nomination, on 13 January 2009 so far as Ms Sarwanto was aware they had not done this, and, given the importance of meeting the quality specification, if there had been a need for a quality inspection by SGS I would have expected this to be picked up in the email of 13 January.

Ms Sarwanto also said in cross-examination that she assumed that there had been discussions between Wilmar and Mr Lokey about the identity of the inspector. Such an assumption would only have been necessary if she knew that the inspector Wilmar had in mind was not SGS.

I am accordingly satisfied that on 12 January 2009 both Ms Sarwanto and Ms Schmidt were proceeding on the basis that quality had been dealt with by the Saybolt certificate, and that their instant messaging discussion about an inspector concerned quantity only.

Vinmar submits that a contractual provision of this kind must be strictly complied with. Assuming, without deciding, that this is right, Vinmar’s conduct following receipt of the Saybolt certificate in my view met the strict requirements postulated for this purpose. I refer in that regard to what was said by Mr Lokey about contract 2 and my comments on it above.

Vinmar also submits that Ms Sarwanto lacked authority to agree to Saybolt. I have explained earlier why in my view Wilmar was entitled to proceed on the basis that Ms Sarwanto was the person who was to be regarded as acting on behalf of Vinmar in relation to the identity of the inspector as well as other arrangements for certificates. In any event it was Vinmar’s conduct as a whole, not just the conduct of Ms Sarwanto, which demonstrated that Saybolt was agreed.

There is no factual dispute as to events after 14 January. In what follows I adopt the account of those events given in Wilmar’s written closing submissions.

On 15 January, Mr Noldt asked SGS to prepare a lab blend of 80% “sample ex tank 7819” and 20% “sample ex tank 1254” and check the CFPP and water content. Later on 15 January he spoke to SGS and asked them to re-analyse the sample they had previously taken and sampled at the request of Vinmar. SGS informed Mr Noldt that the water value was 398 ppm. Mr Noldt forwarded this information to Vinmar, saying:

We do have arranged for another analysis ex 7819 and SGS has confirmed that the value of the water is at 398 ppm. The vessel is still on berth at Vopak hence loading can take place. As before, and now also confirmed by SGS, the cargo in Rotterdam is on spec. Please do let us know what your plans are

Mr Noldt also forwarded an SGS certificate (“SGS D10”)by a later email.

In a further email, Mr Noldt reported to Vinmar that the “handblend” (of samples from tanks 7819 and 1254) was analysed by SGS as 392 ppm water. He said the components would need to be loaded directly on board Varkan Ege.

Mr Noldt sent an email to Vinmar at 1817 on 20 January. It read in part:

Given that the shipment period does not end until 31 January ... there is ample time for us to deliver and for Vinmar to accept a cargo for January shipment.

Vinmar is obliged, under the Contract, to nominate a vessel to lift a cargo of 1,000 mt FAME within the shipment period. We therefore call upon you to nominate a vessel without further delay and in accordance with the Contract. Once we have received your vessel nomination, we will supply you with a certificate of analysis issued by an independent inspector should the existing certificates issued by Saybolt and SGS not suffice for you. ... We therefore invite you to co-operate with us in arranging a shipment of 1,000 mt FAME before 31 January 2009 and in accordance with the Contract. ... We confirm that we are willing and able and fully intend to perform the February and March shipments in accordance with the Contract, as well as the January shipment.

Vinmar responded the following day, saying that:

as a consequence of your breach of contract, we will not be retendering the ship.

Mr Noldt wrote to Vinmar again on 22 January, stating:

We are, as we have already made clear, ready and willing and able to deliver a cargo within the January shipment period which does not expire until 31 January 2009. Does Vinmar have any intention of nominating a vessel to lift this cargo and taking delivery of the January cargo?”

Mr Lokey replied to Vinmar on 22 January stating,

We have made it clear in our messages that [it is] Wilmar who are in breach of contract in respect of the first shipment. As a result of this breach of contract we will not be renominating or retendering a ship for the January parcel.

Wilmar’s solicitors, Clyde & Co, wrote to Vinmar on 29 January stating:

Our clients are, as they always have been, willing and able to provide a cargo within January, in accordance with the contract.

There was no response within the shipment period.

Issue A originally concerned whether the vessel nominated by Vinmar was ready, and whether Wilmar tendered or sought to tender a cargo for shipment. By the end of oral submissions it was apparent that as at 14 January 2009 the answer must be “No” on both counts. Varkan Ege was still loading the Heinz cargo, and would not be ready to load from ST 7819 until the following day. Until then Wilmar – who undoubtedly intended to load cargo from ST 7819 and had made this clear to Vinmar - could not in fact tender cargo or seek to do so. As to Revised Issue A, the appropriation assertion and the COA breach case both involve a novel analysis of relevant legal principles which in my view is best left to a case where that analysis may be decisive of the outcome.

On Issue B, if the answer to question A were “Yes”, then because there was no contract provision for a certificate to be binding, the answer must be “No.” Had there been such a provision, however, for the reasons given above I would have held that the determination of the quality of the cargo, as set out in the Saybolt certificate, was final and binding upon the parties.

As to Issue C, for the reasons given in Annex 1 I hold that the cargo was not off-specification as regards its water content. On the balance of probability the cargo in ST 7819 had no more than 400 parts per million of water. I add that for the reasons given above Vinmar fails in its contention that SGS C04 was binding on both parties.

Issue D concerns whether time is of the essence. I consider that this question is best left over for decision in a case where it will be determinative of the outcome.

I turn to Issues E and F. Issue E asked whether Wilmar was entitled to tender a substitute cargo and did it do so within the necessary time. Revised Issue E raised the same question with the addition of the re-tender of cargo contentions. Issue F concerned whether Wilmar renounced or repudiated the Contract in relation to the January shipment. Revised Issue F clarified Vinmar’s assertions in this regard. All these matters, too, are in my view best left over for decision in a case where they will be determinative of the outcome.

I conclude that Wilmar’s claim succeeds.

In the event that SGS C04 is not conclusive under the August 2008 contract, Vinmar seeks a finding from the court that the FAME in ST7819 contained more than 400ppm of water. This question has been the subject of expert evidence from Mr Revell on behalf of Wilmar and Mr Barden on behalf of Vinmar. Initial reports were exchanged in June 2010. Mr Revell and Mr Barden met on 30 September 2010 and a joint memorandum of the meeting was prepared. Mr Barden prepared a supplementary report dated 12 October 2010. Both Mr Barden and Mr Revell gave oral evidence at the trial.

Mr Roland Revell is employed by Minton Treharne & Davies Ltd. He is a chartered chemist, a member of the Royal Society of Chemistry, a fellow of the Institute of Petroleum, and has long experience of oil, chemical and liquefied petroleum gas cargos. He provided a report dated 3 June 2010 on behalf of Wilmar.

Mr Peter Barden is employed by General Physics (UK) Ltd. He has been involved in the biofuel industry “for the last 2/3 years”, during which period he has done work concerning the UK Renewable Transport Fuels Obligation and the EU Renewable Energy Directive. He has over ten years of experience working with oil sector clients including many of the major UK oil refineries. He holds a Master of Arts degree in Agriculture and Forest Sciences, and is a fellow of the Institute of Environmental Management and Assessment and a chartered environmentalist. He provided an initial report dated 16 June 2010 for Vinmar and a supplemental report dated 12 October 2010.

By the conclusion of cross examination there were few issues on which Mr Revell and Mr Barden disagreed. The most significant of these was whether I should assess the water content in ST 7819 by taking an arithmetical mean of various results.

Mr Revell noted that while EN 14214 has a maximum water limit of 500mg/kg (500ppm weight), the August 2008 contract provided for a maximum water content of 400ppm. Mr Revell commented that FAME is hygroscopic and can absorb moisture during storage and transportation.

Mr Revell then explained the test procedure under EN 12937, noting that while the test procedure stated it was to be used for the determination of the water content of petroleum products, it had thereafter been adopted in EN 14214 for FAME. His report said this:

2.4

In this test procedure the water content of a FAME sample is determined using a Coulometric Karl Fischer titration method. A weighed portion of the thoroughly homogenized FAME sample is injected into the sealed titration vessel of a Coulometric Karl Fischer apparatus in which iodine for the Karl Fischer reaction is generated coulometrically at the anode. When all the water in the sample has been consumed/titrated against the iodine, excess iodine is detected by an electrometric end point detector and the titration is terminated. The water content of the FAME is calculated from the amount of iodine generated and the weight of the sample used in the test with the result being reported to the nearest 0.001% mass (i.e to the nearest 10mg/kg). The water limit for the “Varkan Ege” FAME cargo was 400mg/kg maximum, which is equivalent to 400ppm weight maximum.

2.5

Prior to analysis the sample must be thoroughly homogenised. If the sample is not clear and bright, or water droplets or particulate matter are observed on swirling the sample, a portion of a solution of sodium dioctylsulfosuccinate is added prior to homogenizing with a mixer. After homogenization a sample is taken using a clean dry syringe and discarded to waste, this is repeated two more times. A 1ml to 2ml test portion of the sample is then taken using the same syringe, which is cleaned on the outside with a tissue and weighed to the nearest 0.1mg. The syringe needle is then passed through the septum on the sealed Karl Fischer Titration vessel and with the tip of the needle just below the titration cell solution an aliquot of the sample is injected and the titration is started. The syringe is withdrawn, the needle wiped again and reweighed. At this stage using an Automatic Apparatus the weight of sample that has been injected is then entered into the machine. At the end point the apparatus will report the water content result, which as stated above is based upon the iodine generated coulometrically and the weight of sample injected.

2.6

The time taken to perform a water content determination using an Automatic Coulometric apparatus, which was the apparatus used by the Saybolt laboratory, would be less than 5 minutes. If the apparatus is calibrated as required by the test method the results reported should be reliable as human error is significantly reduced, although not eliminated, by the use of automatic equipment.

Measurement is no more exact than many other sciences. Mr Revell pointed out that no test method will give an absolute (or “true”) value. The true value is defined as the average of the single results obtained by an infinite number of laboratories. Test method EN 12937, recognising that carrying out the same test using the same equipment may produce a range of results, and that carrying out the same test on similar but not the very same equipment may produce a wider range of results, gives values for repeatability and reproducibility, so as to indicate the possible variation in two results even if the two analyses have been performed correctly. Mr Revell explained at paragraphs 2.8 and 2.9 of his report how these values would apply to the contractual maximum of 400ppm:

2.8

‘Repeatability’ of Test Method EN ISO 12937. The term ‘Repeatability’ refers simplistically to the difference in the results that can be expected if an identical sample was tested by the same operator using the same equipment in the normal and correct operation of the test method. The maximum specification limit for the water content of the “Varkan Ege” cargo was 400mg/kg and as per the precision statement of EN ISO 12937 the ‘Repeatability’ at this level calculates to be approximately 37mg/kg. Therefore if an identical sample, which had an average water content of 400mg/kg, was analysed in duplicate by the same operator using the same equipment then the two results should not differ by more than 37mg/kg (i.e 382 and 419mg/kg).

2.9

‘Reproducibility’. The term ‘Reproducibility’ refers simplistically to the difference in results that can be expected if the same sample was tested in two separate laboratories, in the normal and correct operation of the test method. In the case of the EN ISO 12937 Test the ‘Reproducibility’ for a sample with an average water content of 400mg/kg, again the specification maximum limit, would be approximately 138mg/kg. Therefore if an identical sample, which was found to have an average water content of 400mg/kg, were to be analysed correctly in two separate laboratories then the two results should not differ by more than 138mg/kg (i.e 331 to 469mg/kg). This precision statement is used to compare differences in quality as determined in different laboratories.

Turning to the Saybolt certificate, Mr Revell produced an email sent to him by Saybolt on 26 May 2010. Paragraph 2.10 of Mr Revell’s report continued:

The e-mail explains the water contents of the three samples were as follows:-

Upper Sample (T) 392ppm

Middle Sample (M) 393ppm

Lower Sample (B) 391ppm

The three results corroborate the water content of shore tank 7819 was within the contractual limits of 400mg/kg. As per the test method the water results are to be reported to the nearest 0.001% mass (10ppm) therefore the Saybolt result was reported correctly as 390ppm on the Saybolt full quality report attached as Appendix 2A.

At paragraph 2.11 Mr Revell stated that in his opinion the Saybolt certificate confirmed that the quality of the FAME in ST 7819 met the contractual quality specification.

Turning to SGS C04, Mr Revell in paragraph 2.12 of his report noted that it referred to a “running sample before loading”. In his opinion that would not be as representative as testing the individual upper, middle and lower samples. Saybolt had determined water content for such samples, and had thus obtained three results corroborating each other. In his opinion those results would be more reliable.

At paragraph 2.13 of his report Mr Revell noted that SGS D10 reported the water content as 398ppm. Mr Revell understood that this was a re-test of the retained running sample used for SGS C04, and said that his comments made in paragraph 2.12 applied.

At paragraphs 2.14 to 2.16 of his report Mr Revell noted that the difference between the result for water content in the Saybolt certificate and that in SGS C04 was 45ppm. If both certificates had been in respect of the same sample, their results would have been well within the reproducibility range of 138ppm. There was thus no significant difference between the two results. Indeed, as they concerned analyses of two different samples, the difference in result was even less significant. On the footing that SGS C04 and SGS D10 involved the same sample, Mr Revell noted that the difference of 37ppm was exactly the same as the range permitted for repeatability under EN 12937. This again suggested that there was no significant difference in the two SGS results.

Mr Revell summarised his views at paragraphs 2.16 and 2.17 as follows:

2.16

The Saybolt result of 390ppm is based upon three separate water determinations on three individual shore tank samples, which have then been averaged and correctly reported. The water results for the three individual determinations gave excellent agreement with results of 392/393/391, which, in my opinion, indicates these results are very reliable. At this stage it would appear that the SGS figures of 435 and 398ppm are individual results of duplicate tests on apparently the same running sample. It should be noted that in ISO 3170 “Petroleum liquids – Manual Sampling” it states the following:-

(i)

Upper, middle and lower are normally taken from shore tanks, which can be used to demonstrate that the tank contents are homogeneous.

(ii)

A running sample can be taken of a shore tank, however, this will result in only a single sample being taken which cannot be used to assess the homogeneity of the tank contents.

Therefore, in my opinion, the testing of individual upper, middle and lower samples of shore tank 7819, such as that performed by Saybolt, would give a more reliable assessment of the water content of the shore tank contents than a single analysis, as apparently performed by SGS, on a running sample.

Did the Quality of the FAME in Shore Tank 7819 meet the Contract Specification?

2.17

The Saybolt analysis has confirmed that the quality, including the water content, of the FAME in shore tank 7819 met the contractual specification. In my opinion there is no reason to doubt the Saybolt results were correct for the following reasons:-

(i)

The Saybolt e-mail has confirmed that they performed three water determinations on upper, middle and lower samples with the results being were 392, 393 and 391ppm, which have then been averaged and reported correctly at 390ppm. These three results corroborate each other and are therefore a reliable indication that the water content of shore tank 7819 was within the specification limit.

(ii)

The Saybolt analysis included water determinations on upper, middle and lower samples, which in my opinion would give a more representative assessment of the water content of the contents of shore tank 7819 than the testing of a running sample as tested by SGS.

(iii)

One of the SGS results of 398ppm is virtually the same as the three Saybolt results of 392/393/391 and all four results confirmed the quality of shore tank 7819 met the contract specification for water content. Therefore if any result is suspect it would be, in my opinion, the SGS result of 435ppm, the only off specification result.

Section 4 of Mr Barden’s first report noted the accreditation of both Saybolt and SGS as meeting ISO 17025. He commented, however, that Saybolt’s accreditation schedule made no reference to tests for FAME, nor for water content against EN 12937. By contrast, the accreditation schedule for SGS referred to “plant based materials and moisture content.”

Mr Barden then turned to the Saybolt certificate and the email of 26 May 2010 explaining the results achieved for each of the upper, middle and lower samples. Having set out the results for each sample, Mr Barden continued at paragraph 4.9:

4.9

This would give a mathematical average result of 392 mg/kg. This result has been rounded to 390 mg/kg. Clause 11 “Expression of results” of BS EN ISO 12937 states that results should be reported to the nearest 0.001% (m/m), which would require the result to be reported as 392 mg/kg.

Mr Barden then explained that there had been an error in SGS C04:

4.11

Clarification of the analysis report C.04, was sought from SGS, and it was confirmed that the description on their report should have read "composite based on upper, middle and lower samples" not "running sample before loading".  This was confirmed by SGS as an error.  The water content test is, therefore, based on the composite sample, rather than the results of individual tests on each constituent element of the sample arithmetically averaged to give the test result.

By contrast, however, SGS D10 had indeed been a test of a running sample:

4.12

Likewise, clarification of the analysis report D.10 was sought, and it was confirmed that the description "running sample as requested" indicates the manner in which the sample was taken.  Essentially, it is understood that this is a method by which the testing company draws a sample from the upper, middle and lower sections of the tank in one go, a methodology which relies upon the skill of the tester.  The single resultant sample is then tested to obtain a result for water content.

In the remainder of section 4 of his report Mr Barden made comments as follows:

4.14

The test methodology applied in BS EN ISO 12937:2001 is a generally accepted method used for petroleum products. That said, its use for testing of bio-diesel products is in its infancy and has yet to be fully developed. As a consequence the level of uncertainty associated with the methodology for this purpose is not fully understood, and no documented uncertainty (margin for error) for the methodology for this purpose has yet been defined.

4.15

Consequently, there is a significant amount of uncertainty associated with the use of the BS EN ISO 12937:2001 methodology for this purpose, associated with a difficulty in obtaining reproduceable results.

4.16

Therefore, whilst the 435 mg/kg result appears on the face of it to be potentially an outlier, it is considered to be within the currently understood uncertainty associated with the test methodology.

4.12

It should be noted that FAME is hygroscopic and will absorb atmospheric moisture. The degree and extent to which this occurs is dependent on the storage and handling conditions.

In section 5 of his report Mr Barden set out his conclusions as follows:

5.1

My conclusion is that the lack of formal accreditation by Saybolt Nederlands BV to conduct the test required by EN ISO 12937 places the test result in question.

5.2

The uncertainty associated with the test methodology in terms of water content testing for FAME is significant, suggesting that an averaging of the two remaining test results be appropriate for this purpose, giving an average water content figure of 416 mg/kg for this cargo.

5.3

Therefore, on the basis of the above consideration it is my opinion that on the balance of probability the cargo in question was off-specification for water content (the contractual maximum being 400mg/kg).

The joint memorandum dealt with six points. Mr Barden and Mr Revell agreed on three of these points. At point 1 they agreed that Saybolt had correctly reported the result of 390ppm to the nearest 0.001%. The SGS results of 397 and 435ppm should, however, have been reported as 400 and 440ppm. At point 5 they had been asked to consider whether when averaging results all results from all laboratories should not be included. They agreed that when assessing the water quality of FAME they should consider all results. Point 6 asked which would be more representative, a “running sample”, or upper, middle and lower samples? Both agreed that upper, middle and lower samples would be more representative of shore tank contents.

As to points 2, 3 and 4, the joint memorandum stated:

2.

Is the testing of water content of FAME by EN ISO 12937 covered by the quality system ISO 9001 in both the Saybolt and SGS laboratories?

We are both agreed that the quality management systems of the two laboratories are ISO 9001 Certified. With regards the ISO 17025 Accreditation we need further clarification to establish if the SGS schedule, which is in Dutch, covers EN ISO 12937 test method. The Saybolt ISO 17025 Accreditation does not cover EN ISO 12937.

3.

SGS have now issued four separate analytical reports for the testing of samples representing shore tank 7819. Does this indicate the SGS quality system was operating in a satisfactory manner?

RR states that the SGS quality system was not satisfactory.

PB states that the Saybolt analytical report also contained factual errors.

4.

Of the three Saybolt water results and the two SGS results for shore tank 7819 would the result of 435ppm be the outlier?

PB states that all the results are within the Reproducibility of the test method and therefore there are no outliers.

RR states that although it is agreed that all results are within the test method Reproducibility the SGS result of 435 is the one result that indicates any real difference in the water content.

Mr Barden said that he had been asked to provide this supplementary report in order to address points raised by Mr Revell’s report, the second witness statement of Mr Noldt, and further developments since the date of his first report. Section 2 of Mr Barden’s second report dealt with growth of the biodiesel industry as follows:

2.1

Over recent years there has been a marked drive to increase the use of Renewable Energy globally. In Europe, various regulatory drivers have been introduced by the European Commission to support this drive, most notably the Renewable Energy Directive (RED) and its sister Fuel Quality Directive. The RED comes into force at the end of this year.

2.2

In the UK the RED has been preceded by the domestic Renewable Transport Fuel Obligation (RTFO), which was brought into force by the Renewable Transport Fuel Obligation Order 2007 / 3072 (implementing provisions of the Energy Act 2004) and has since been amended by the Renewable Transport Fuel Obligation (Amendment) Order 2009 / 843. The RTFO requires suppliers of fossil fuels to ensure that a specified percentage (currently 5%) of the road fuels they supply in the UK is made up of renewable fuels. The RTFO is now in its third year of operation.

2.3

Both of these moves have meant a significant increase in the international trade of biofuels in the last three years, most notably of Fatty Acid Methyl Esters (FAME) and bio-ethanol. FAME is the generic name given to vegetable oils used as raw materials for the manufacture of Biodiesel. Up until about 3 years ago, the international trade in FAME was limited.

Section 3 of Mr Barden’s second report was entitled, “prominence of SGS vs Saybolt,” and said that in general terms four test houses were predominant in the oil and gas field and in the biofuels markets in particular. In alphabetical order they were Inspectorate, Intertek, Saybolt and SGS. Mr Barden continued:

3.2

Since the introduction of the RTFO, fossil fuel suppliers have been required to provide independently verified / audited reports to the Renewable Fuels Agency (an executive agency of the UK Government) demonstrating compliance with their renewable fuel obligations. Over the past two years, I have audited the reports of a wide range of fossil fuel suppliers and, in doing so, have reviewed hundreds of sample shipping files and contract documents relating to the international trade of bio-diesel. Of the test certificates and out-turn certificates I have seen, by far the largest number have been issued by SGS and Intertek. Only a minority have been issued by Saybolt and Inspectorate. Whilst the number of certificates issued by SGS and Intertek have been broadly similar, SGS have the more established global reputation in the fuels sector.

3.3

As a consequence of both my involvement with RTFO audits and my general industry experience, I concur with the observation made by Christopher Lokey in his witness statement that SGS has the best established international profile of all the major test houses. I have read the Witness Statement of Bernd Noldt and, in particular, his comments regarding the profile of Saybolt in Europe. In my experience, few of those active in the European FAME market appoint Saybolt as their inspectors of choice.

Section 4 of Mr Barden’s second report concerned errors in documents. In relation to SGS his first report had noted the incorrect description of the sample as a running sample, whereas in fact it was a composite sample; as to this, Mr Barden pointed out that the body of SGS C04 referred to upper, middle and lower samples having been tested individually for density, something which would not be possible in relation to a running sample. A second error by SGS had come to light: D10 gave a water content result of 398ppm but a corrected certificate showed that the result was 397ppm; had SGS rounded to the nearest 10ppm, this discrepancy would not have been identified.

At paragraph 4.3 of his second report Mr Barden identified factual errors in the documents provided by Saybolt:

(a)

The Saybolt certificate describes their sample as being date marked 6 January 2009. However, in the email dated 26 May 2010 from Gertjan Hendriks to Mr Revell, the sampling and testing date is give as 5 January 2009.

(b)

The Saybolt certificate describes their sample as being a single tank composite sample. However, the 26 May 2010 email indicates that three (upper, middle and lower) samples were tested for water. The water result cannot properly be described as being derived from the testing of a composite sample. That would require a single sample prepared by combining numerous smaller samples from throughout the lot (in this case the FAME in tank 7819).

Mr Barden commented on all these errors generally that there was always scope for human error in the production of certificates, and that in his opinion these errors did not materially affect the underlying test results and did not provide a reason for discounting any of the results contained in the certificates. He added:

I also note from documentation recently disclosed by Wilmar that concerns arose within Saybolt concerning the initial test results which necessitated the tests being re-run. I understand that the original test results were published on the strength of the retests. However, no proper explanation has been given as to why Saybolt were concerned with the original results and only two of the retest results (undated) have been made available.

In section 5 of his second report Mr Barden commented that it was logical and common commercial practice to test for moisture content first in order to reduce the costs of further unnecessary analyses on materials if the water content specification were not met. Mr Barden continued in section 6, under the heading “suitability of test methodology”:

6.1

The Coulometric Karl Fischer Titration Method ("CKFTM") applied in this case is the generally accepted methodology used for petroleum products. The regulatory drive to promote renewable fuels created a pressing need for an accepted test method for biodiesel and the CKFTM represented the logical choice. It has since garnered support from various bodies, including the European Association of National Metrology Institutes which believes it to be an accurate method, but not one which is suited to on-site measurements. However, the reality is that the method's use for testing bio-diesel is in its infancy and has yet to be fully developed.

6.2

The level of uncertainty (margin for error) associated with the use of the CKFTM for testing bio-diesel is not fully understood and remains undocumented. This ultimately means that individual test results can only be regarded as indicative of the water content of the particular cargo under inspection. As Mr Revell explains in his report, no test method will give an absolute/true value. This can prove problematic since water content is the most critical characteristic of FAME in terms of its marketability for biodiesel manufacture. The water content of FAME can have a significant impact on the manufacturing process and the quality of the end product. FAME is also extremely hygroscopic which means that it will readily absorb atmospheric moisture until it reaches equilibrium with the prevailing atmospheric conditions in terms of relative humidity.

6.3

Bearing these factors in mind, I find it surprising that Wilmar tendered a cargo to Vinmar which, even on the basis of the Saybolt certificate relied on by Wilmar, was only 8ppm within the contractual limit. In my experience parties tend to trade cargoes which are certified to be comfortably within the contractual limit for water, recognising the inherent difficulties of establishing the true water content of FAME.

6.4

As indicated in my earlier report, it is my firm view that none of the reported samples should be discounted as an outlier since they are all within the acceptable bounds of reproducibility. In my opinion, the fact that the reported results range from 392ppm to 435ppm does not indicate that any of the samples are unrepresentative of the cargo in tank 7819, but rather that the science of testing the water content of biodiesel has yet to catch up with industry practice. As I agreed with Mr Revell at our meeting on 30 September 2010, all the test results are equally valid and must be taken into account when assessing the water content of the cargo tendered by Wilmar in this case.

In section 7 of his second report Mr Barden noted that SGS C04 had involved individual testing of upper, middle and lower samples for density, but not for water content. Mr Barden commented:

7.6

As Mr Revell indicates in his report the individual test results on the Saybolt samples corroborate one another, and this suggests that the averaged result is reliable. However, that does not mean that the SGS certificate results can be discounted. The SGS sample was tested for homogeneity by a method permitted under BS EN ISO 3170:2004 and there is no reason to doubt the validity of the reported results. In my opinion, any assessment of the water content of the cargo tendered by Wilmar must give weight to the SGS result.

Section 8 of Mr Barden’s second report dealt with Wilmar’s instruction to SGS for the analysis on 15 January 2009 of a hand-blended sample composed of 80% FAME from ST 7819 and 20% palm methyl ester (“PME”) from ST 1254. Mr Barden commented:

SGS subsequently reported the water content of the hand-blend to be 392ppm. Wilmar have not disclosed any certificates of analysis indicating the water content of the PME from Tank 1254. However, the reported result of 392ppm indicates that the 80% FAME within the sample had a water content greater than 392ppm. As a matter of arithmetic, it the water content of the PME was 356ppm or lower, then the water content of the FAME from Tank 7819 will have exceeded 400ppm (401ppm x 80% = 320.8; 356 x 20% = 71.2; 320.8 + 71.2 = 392ppm).

Section 9 of Mr Barden’s second report was entitled, “conclusion” and stated:

Wilmar's best case is that the cargo had a water content of 392ppm, which on any view is remarkably close to the contractual limit of 400ppm. As Mr Revell notes, the individual test results for the constituent samples do corroborate one another and this suggests that the result is reliable. However, there are three certificates which indicate that the water content was higher than 392ppm; namely, the two SGS certificates relating to Tank 7819 and the SGS certificate for the hand-blend. Of these, one indicates that the water content was 435ppm, substantially above the contractual limit. None of the results should be discounted but rather they must all be taken into account when forming a view as to the water content of the cargo. I therefore consider that in order to determine the most likely actual water content one should take an arithmetical average of the three certificated results (i.e. 392, 397 and 435). This would indicate that the water content was most likely 408ppm. Given (1) the uncertain nature of the testing methodology, (2) how close the Saybolt result of 392ppm is to the contractual maximum of 400 ppm, (3) the SGS result of 435ppm, (4) that the water content of the hand-blend (which must have been lower than the FAME) was also apparently 392ppm, it is my opinion that when looking at the evidence in the round that the water content of the cargo exceeded 400ppm.

In his evidence in chief Mr Barden said that having looked at the accreditation schedules for both SGS and Saybolt he considered that they were both of equal status. As regards his second report, he added that it was now known that the moisture content for the PME element of the hand-blend was 290ppm. He calculated that this showed that the FAME from ST 7819 forming 80% of the hand-blend had had a water content of 417ppm.

In cross examination Mr Barden explained that he had studied chemistry in order to achieve his qualification in agriculture and forest sciences. He had not worked in a laboratory, but had reviewed and assessed laboratory work. He had never taken a sample from a shore tank, but had had responsibility for testing of such samples.

The first question addressed in Mr Barden’s second report was the prominence of SGS as opposed to Saybolt. Mr Barden acknowledged that this not arisen in any of the matters which had been identified at the outset of the second report as giving rise to that report. He said initially that the first report had discussed the equivalence of SGS and Saybolt, “so this follows on from that report.” He then accepted that the first report was concerned with accreditation, and prominence was not the same. He said he thought it was of relevance, but accepted that it was not of relevance to helping the court to decide the water content of the FAME on the balance of probabilities. He acknowledged that what he had done could be construed as attempting to shore up Mr Lokey’s evidence about why Vinmar preferred SGS and would not have agreed to Saybolt. Mr Barden said that over the last three years he had been reviewing biodiesel contract files under the renewable transport fuel obligation. The majority had been prepared by SGS and Intertek, a much smaller proportion by Saybolt, and an even smaller proportion by Inspectorate. He maintained that this provided a proper basis for him to say whether Saybolt were or were not used in a great deal of the trades relating to FAME.

Mr Barden was then asked about paragraph 6.3 of his second statement, quoted above. When asked why he had said he found the tender of the cargo surprising, he replied that because FAME is extremely hygroscopic it would quickly continue to take up moisture and surpass the 400ppm even if it was only 392ppm at the time of the certificate. He accepted that this observation was not relevant to the question whether the cargo met the specification, and that the point he had made could be construed as a point about the commercial elements of the deal. It was suggested to him that again it appeared he had included material in his report which was simply designed to advocate the position of one of the parties. Mr Barden replied, “no, it is an observation with regard to the specification.” As to paragraph 9.1 of his second report, Mr Barden denied that he had approached the matter on the basis that Wilmar had a “best case” which it was his task to disprove.

Mr Barden then accepted that the second sentence of paragraph 2.3 of his second report was inaccurate. The correct position was that FAME did not refer to vegetable oils used as raw materials, but instead to fatty acid methyl esters that were derived from vegetable oils and also from animal fats. His second sentence had been an over simplification.

Turning to relevant standards, Mr Barden accepted that EN 14214 adopts EN 12937, and that this showed that the European Committee for Standardisation considered that EN 12937 was an appropriate standard for determining water content in FAME. He was asked whether the European Committee for Standards would have had regard to EN 4259 before adopting EN 12937 as the standard for water determination in FAME, and replied that he would assume so. Mr Barden then said he did not dispute that it was reasonable to assume that EN 12937 was an appropriate method for determining water in FAME. As to paragraph 4.14 of his first report, he said it was not disputed that it was the appropriate standard. What he had said about the level of uncertainty not being fully understood and there being “no documented uncertainty, margin for error” were statements of fact. Mr Barden then stressed that precision, uncertainty and accuracy were three different things.

At this stage I became concerned that the evidence was not focussing on the matters that I had to decide. I stated that it would help me to have an understanding of the process once a sample had been taken – precisely what processes then took place and what it was that those processes did. Mr Barden was then taken to the part of EN 12937 stating that once a sample has been visually inspected and found appropriate, a weighed proportion is injected into the titration vessel of a CKFA. His explanation of titration was:

Titration, in its simplest form, is you inject an agent into a substance, you put something else into that substance. At the point where the reagent changes colour, that gives you the switch point and from that piece of information you can calculate the… water content…

This did not entirely accord with what I understood to be meant by titration. Mr Barden then agreed with me that titration is a method or process of determining the concentration of a dissolved substance in terms of the smallest amount of reagent of known concentration required to bring about a given effect in reaction with a known volume of the test solution. He said that the issue then was, “how you spot the point where the change occurs.” There were various parameters that could be used for that, and in this case it was an electrical change. Mr Barden agreed with Mr Collett that what happened was that when all the water had been titrated excess iodine was detected by an electrometric end point detector. In order to understand how this was done in a coulometric apparatus, I asked Mr Barden if he could explain to me Coulomb’s Law. He was unable to do so. As to whether that law assisted in understanding the process of working out how much water there was in a particular sample, Mr Barden replied that his detailed knowledge of the test in question was not to that depth, his experience was in the interpretation of test results. If an approved test methodology had been followed then that would be taken at face value. I explained to Mr Barden that in a case where different tests might, perfectly properly, produce a result which is more than 100ppm different, I thought I ought to have some understanding of what chemical processes were taking place which might give rise to those differences. Mr Barden replied that there were issues around the physical way that the operatives took the sample, the manner in which the sample was then handled and stored, the manner in which a portion of the sample is taken for testing, issues associated with the way the test sample was then weighed, and the way that the apparatus was used by the operative to produce a result. Mr Barden continued that the standard gave guidance on the bounds of the results to be expected in the section looking a repeatability and reproducibility. Reproducibility, he said, concerned the boundaries to be expected where the same test was run on the same piece of equipment by the same operative, while repeatability was where the sample was run by another operative on another machine, or another sample.

Mr Barden then added that in addition to the physical matters he had described there were elements introduced by features of the chemistry. As to that all that Mr Barden could say was that it was recognised that pieces of equipment could not “deliver 100% results so that is defined as the uncertainty of the test methodology, or of the piece of metering equipment.”

At this point I indicated that if it were possible for the chemistry to be explained to me before further cross examination of Mr Barden then that might assist me. The parties discussed the matter and concluded that it would be appropriate for Mr Revell to give evidence on chemical aspects at this point.

Mr Revell explained that the test was performed using a sealed reaction cell. The reaction cell contained a reaction solution, an anode, a cathode and a detector. The cell was sealed with a septum to ensure that there was no moisture available within it prior to introduction of the test sample which could affect the result. The sample itself is homogenised before a test sample is drawn into a syringe, which is then weighed, before the needle is passed into the septum and the test sample injected into the reaction cell. The syringe is then placed back on the balance, and the difference in the weights is entered into the machine. The machine is then activated so that at the anode iodine is generated from the reaction solution. The iodine reacts with any water present in the sample. When excess iodine is detected – i.e. iodine which is not reacting with water – the machine automatically stops on the basis that at that point no water is present. The machine then automatically calculates the amount of electricity required to generate sufficient iodine to react with all water present in the sample. The amount of iodine that is produced is in line with the amount of coulombs of electricity necessary to produce it.

Mr Revell was then asked questions by Mr Collett concerning EN 12937. He noted that it covered diesel or kerosene type products, and that it was said that it might be applicable to lubricating base oils, although “the precision has not been established for these materials.” It was necessary to be on the look out for the presence of substances or compounds which might independently of the iodine be involved in a condensation or oxidation reduction reaction. He would not expect such substances to be present in FAME.

Mr Barden was asked further questions about his statement that the use of EN 12937 for biodiesel products “has yet to be fully developed.” He said that his point was that the precision and the uncertainty associated with that test methodology for biodiesel had not been formally assessed. There then followed a question and answer as follows:

Q. But are you saying from your own knowledge you can positively say that there has been no investigation of the precision of that test method when used on FAME?

A. No, I’m saying that it has not been documented in the standard, that its application is appropriate to FAME – whilst it is being used for that purpose as the best available methodology, that standard 12937 is for petroleum products specifically, not for bio-products. It is just a – I wouldn’t read too much into that. We are agreeing it is the accepted methodology.

It then emerged that Mr Barden had investigated a number of learned papers which he had not referred to in his report but which were relevant. He had not referred to them in his report because he agreed that the standard methodology in EN 12937 was appropriate for the FAME standard. There then followed a question and answer:

Q. But you are making a positive assertion in the next sentence that:
“The level of uncertainty associated with the methodology for this purpose is not fully understood.”

A. There is no documentation in the standard that lays out the uncertainty associated with that particular test for the use of measuring FAME.

Q. Let’s just pause on what you actually said in your report:
“The level of uncertainty associated with the methodology for this purpose is not fully understood.”
So you are saying that when the European Committee for Standardisation adopted the test, they didn’t understand what the uncertainty would be when used on FAME?

A.

I’m saying that they determined that 12937 was the best applicable test available for that purpose and that the uncertainty associated with that test for petroleum products is understood, but there is no documented process to list that uncertainty for the use of FAME because it is not in the standard.

Mr Barden acknowledged that he had been wrong in his first report to say that reporting to the nearest 0.001% meant to 1ppm, he should have said 10ppm. He also acknowledged that the answer he gave to me earlier about repeatability and reproducibility had got the two concepts the wrong way round.

As to the concepts of repeatability and reproducibility, Mr Barden was taken to the passage in the standard which explained that a formula was used for a particular test value to identify for each of repeatability and reproducibility a difference between the two test results obtained in those circumstances which would, in the long run, in the normal and correct operation of the test method, only be exceeded in one case in twenty. Mr Barden agreed that the concept thus sought to identify a difference for which one could have 95% confidence. He agreed with the calculations that had been performed by Mr Revell. As to the true value, he agreed that no test on its own would give that result other than by chance – it always had a boundary of uncertainty associated with it. He accepted that this precision information was relevant to FAME and it was the only precision information that we have in relation to this test. Mr Barden then, in response to questions from Mr Collett, accepted that the essence of the question he had considered in his report was whether the court could have more than 50% confidence that the water content of the FAME in ST 7819 was more than 400ppm.

Mr Barden was then taken to worksheets provided by Saybolt comprising the printouts generated by the coulometer. Certain of the printouts stated “work conditions not OK”; they were said by Saybolt to have been retested, and although they were undated he had no reason to doubt what Saybolt said.

As to the date of the Saybolt tests, Mr Barden accepted that Saybolt had been asked to take a sample on 5 January, and that the worksheets stated that samples had been taken and tested on 5 January. He also accepted that normal practice would be to test first for water, and that the full EN 14214 test could take 18-20 hours. He accepted that it was likely that the water test was performed on 5 January, a composite sample was then formed and the other EN 14214 tests were performed so that the entire process was completed on 6 January. In that case, Mr Barden accepted, there was nothing in the point he had made in paragraph 4.3(a) of his second report.

Mr Barden accepted that, similarly, there was nothing in the point he had made at paragraph 4.3(b), as it was likely that the water testing had been based in the individual upper, middle and lower samples, and that after that the three samples were made into a composite and the other two tests were done on that composite sample. He did not suggest that the Saybolt results should be discounted.

Mr Barden was then taken through the course of events concerning the SGS certificates. SGS C04 had referred to a sample number 101, described as “running sample before loading” with the date of sampling given as 13 January 2009. SGS had effectively withdrawn that certificate and issued a new certificate (“SGS C04 corrected”). This stated that sample 101 was a tank composite upper middle lower before loading rather than a running sample, but kept other results the same.

As to certificate D10, it concerned a sample 105, described as “running sample as requested” with the date of sampling given as 13 January 2009 and a water result of 398. It had been “avoided” by a further certificate (“SGS D14”) which still referred to sample 105 described as “running sample as requested” but gave a date of sampling as 15 January 2009, and the result as 397.

Mr Barden was asked to give an account of how a running sample was taken, and said that a sampling device was lowered at a controlled rate down the height of the tank, taking the sample as it went so that in theory it represented the entire cross section of the tank. Under cross examination he then corrected this, accepting that the sampling bucket was first lowered and then raised.

Mr Barden accepted that the anomalies in these certificates were anomalies that would not be expected. A running sample was acceptable for use under EN 12937. A composite was preferable because it enabled a check to be made on the homogeneity of the cargo, but in this case that had been verified both by the Saybolt certificate and SGS C04 with good results. Mr Barden made an additional point that if the results in SGS D10 and SGS D14 had been rounded as required to the nearest 10ppm then both 397 and 398 would have been rounded to 400.

Turning to the hand-blend, Mr Barden accepted that SGS D08, while it referred to “sample 102” was a composite which involved sample 601 taken from ST 1254 on 7 December 2008. The conditions in which the sample was stored were not known, and if it had been stored in a plastic bottle it is conceivable that the FAME would be able to take up water through its hygroscopic nature. It was not clear which sample from ST 7819 was used to make the hand-blend, but it could have been the same sample used for SGS D14. Mr Barden accepted that there were greater uncertainties associated with the hand-blend than with SGS D14 because the true value of ST 1254 was unknown, the result for that shore tank was itself subject to standards of repeatability and reproducibility, and it was not known whether it was in the same condition as when it was analysed in December, nor was the true value of ST 7819 known, that result itself was subject to standards of repeatability and reproducibility, and no individual result could be taken as more than indicative. As to whether a back calculation from the hand-blend was a futile exercise, however, Mr Barden replied that it gave an indication and was no more futile than taking a single sample in isolation.

Under further cross examination Mr Barden accepted an “outlier” had the technical meaning, “a result that is far enough in magnitude from other results to be considered not part of a set.” It was then suggested to him that the three results he relied upon, Saybolt at 392, SGS C04 at 435 and SGS D14 at 397, were outside the criteria in EN 12937 for repeatability and reproducibility in the sense that they did not concern identical samples, they were taken at different times, and in different ways, and in relation to the SGS results it was not known whether the same operator performed the tests. Even so, Mr Barden said that the court could have more than 50% confidence that the true value of the water in ST 7819 was above 400ppm because “those are the only bounds that we have to work with…”.

Mr Barden acknowledged that in his second report what he had done was to take the three results, add them, and divide them by three to arrive at a mean average. He maintained that this had statistical validity: the only way to get common ground from three samples of purportedly the same material was to take a mean average. The more samples one had, the nearer one got to the true mean. Accordingly three samples gave a stronger statistical result than taking one in isolation. In response to a further question from Mr Collett, however, Mr Barden accepted that the average of three results could not give more than 50% confidence that the true value of the water content was the same as the mean of those three results. However, it gave more confidence than any single result would in isolation. He accepted that he had not referred the court to any statistical information which would enable him to state that averaging the three results gave more than 50% confidence that the water was over 400ppm.

In response to questions from me, Mr Barden said that if a result was outside the bounds of reproducibility or repeatability it will be taken as an outlier. If there are only two results, one could not put any stronger weight on one or the other. In relation to each, even if it were an outlier, there was a 1 in 20 chance that there was no problem.

Mr Barden added that it was fair to assume that it was data concerning tests on the determination of water in petroleum products boiling below 390 degrees centigrade which had enabled the relevant standards committee to arrive at the formulae for repeatability and reproducibility, that this may be a factor in the comment that international standard had not been established with precision for lubricating base oils, and the same comment might be applicable to FAME. Under further cross examination Mr Barden agreed that when establishing precision data the relevant committee would proceed in accordance with EN 4259, this being a standard for petroleum products concerning determination and application of precision data in relation to methods of test.

In cross-examination Mr Revell explained that he is a chartered chemist. In order to explain exactly what the test for water content involved, and how it was carried out, he had drawn on his chemistry background and knowledge. For his report, however, he had not had to do this.

In response to questions about paragraph 2.7 of his report, Mr Revell commented that two tests might produce the same result, but that does not mean it is the true result. In an ideal world you would have an infinite number, of results. When trying to work out a true value, if one had 100 results, one would add them all up and divide by 100.

It was suggested to Mr Revell that the same approach could be adopted for 50 or 20 results. He did not agree: at 20 there would not be sufficient results even to start considering what a true value was. He could not say that a mean average would be as close as one could get to true value, for one did not know what the other results would be.

Mr Revell acknowledged that the water content recorded in the Saybolt certificate had most likely been tested on 5 January 2009. As to what was said in SGS C04, at the time of his report he had not seen SGS C04 corrected. He had thought the water content was done on a running sample. There were density readings for separate upper, middle and lower samples, but his understanding from the certificate had been that the water content on its own had been done on a running sample.

Mr Revell accepted that if the figure of 435 had been reported to the nearest 10ppm it would be rounded up to 440. Returning to SGS C04 corrected, whether it was a running or composite sample in this case would not have mattered as the tank was homogeneous.

As regards SGS D14, which was based on a running sample, it was suggested to Mr Revell that a composite sample would have been likely to give a more accurate result. He did not accept this. What was known from SGS C04 was that the density was homogeneous, but it did not follow that the water content was homogeneous. He did not know why SGS had not tested each of the upper, middle and lower samples for water content.

Mr Revell was taken to paragraph 2.14 of his report, where he had said that the difference of 45mg/kg between the Saybolt and SGS results was well within the reproducibility of the test method, which meant there was no significant difference. He accepted that the results involved different samples, and so were not within the reproducibility criteria, but this should only mean that one could expect greater variation. A difference of 45 is not significant, and one could not choose between the two results, 390 and 435, at that stage.

When it was suggested that his report could have said that SGS C04 had confirmed that the water content of the FAME in ST 7819 was outside the contractual specification, Mr Revell replied that the Saybolt result was far superior on the information he had. He had results on 3 samples analysed by Saybolt, and he had the printouts for those results. In addition SGS D14 had produced a result of 397. Of the four results only one was above 400ppm, and on that basis he concluded that the product in ST 7819 met the specification.

As to the Saybolt worksheets, Mr Revell said he had noticed that the printouts said “work.conditions not OK”. He did not know what it meant. He had used many of these machines, and it depended on how the machine had been programmed. He had not seen such a statement on a printout before. Saybolt had responded that the solution in the cell may be at too low a level, and that might or might not affect the results. He had not mentioned it in his report, as by that time he had the check results which verified the results in the printouts.

Under further questioning Mr Revell said that Saybolt’s three water content results for separate samples confirmed that the product was homogeneous for water. While it was correct that proper testing by SGS might produce a result of 430, he had a second SGS result which verified the three Saybolt results. He was not saying that the true value must be closer to the Saybolt result, just that SGS C04 did not show the Saybolt result to be incorrect. He was comfortable with the result in SGS D14 of 397ppm, but found it unusual that SGS C04 showed 435ppm. As he had 4 results below 400 and one above, the probability was that the cargo was on spec. While it would be within the reproducibility range, if it had a difference of 138ppm he would repeat the test. Personally, while the specified method said it was not a significant difference, he would say it merited investigation. There was no reason to say the Saybolt results were incorrect.

It was put to Mr Revell that the total of the three Saybolt analyses along with SGS C04 and SGS D14 was 2008, which when divided by 5 produced a mean of 401.6. Mr Revell acknowledged that at the joint meeting he had said that all results should be considered. When rounded down to the nearest 10ppm the figure of 401.6 would be 400. If one took the total of the Saybolt certificate along with SGS C04 and SGS D14 and divided by 3 this produced a mean of 408. The total of the results for the Saybolt certificate and SGS C04, both being composite samples, when divided by 2 produced a mean of 413.5. He had not said, however, that the results should be averaged. In his opinion where he had four results that were very similar and one was higher within reproducibility, he would tend to say that the four results were more indicative of water quality.

When asked about the hand blend, Mr Revell agreed with Mr Barden’s mathematics., but even then the result was not as high as 435. In Mr Revell’s view, however, the water content of the PME would have been much higher than 290 after a month’s storage.

In response to questions from me Mr Revell said that the potential for deterioration after 5 January 2009 depended on how the product was stored to keep moisture off. This might be achieved by keeping the tank full, or by a nitrogen blanket. Over that short period there should not be much change. It could depend on the location in the tank. There was a need to ensure that no delivery in or transfer out took place between the taking of the sample and shipment. As to averaging and whether one took a mean, mode or median, there were guidelines on how to assess samples but he had not looked at them as relevant conditions would not be met in the present case.

Vinmar accepts that it bears the burden of proving that it is more likely than not that the water content of the product in ST 7819 exceeded 400 ppm. In my view it faces insuperable problems in that task.

Mr Barden rightly accepted that the relevant standard for testing is EN 12937, even though that standard was originally devised for petroleum products. Under that standard it is recognised that water content cannot be measured exactly. All that can be said is that if a product is to have a water content of 400 ppm then a correct measurement may well produce a result within the reproducibility range. In the present case SGS C04 produced a result of 435 ppm, which is well within the reproducibility range. On its own, therefore, SGS C04 does not show on the balance of probabilities that the water content of the product in ST 7819 exceeded 400 ppm.

Vinmar seeks to answer this problem by relying on an arithmetical mean of SGS C04 and other, lower, readings. This approach to my mind suffers from the same problem. Each of the additional readings may well reflect a water content of substantially less than the reading. In relation to each of them, one cannot say on the balance of probabilities that the water content is likely to be no lower than the reading. I do not consider that taking the arithmetical mean of a number of readings whose precision is uncertain gives rise to a likelihood on the balance of probabilities that the water content is higher than 400 ppm when the levels of uncertainty for each are such that each of them is consistent with a reading well below 400 ppm.

Vinmar sought also to rely on back-calculation from the hand blend. This, too, suffers from the same problem. Moreover I find persuasive Mr Revell’s reasons for casting doubt on the validity of the back calculation when more than a month had passed since one of the relevant samples had been taken and analysed.

Standing back, it seems to me that Mr Revell must be right when he says that greater confidence in a water content value is obtained when there are a number of readings at or below that value. In this case four out of five readings were at 400 ppm or less. While that does not meet the scientific criteria for a “true” value, it does in my view enable the court to say on the balance of probabilities that the water content was no greater than 400 ppm.

In reaching these conclusions I have not taken account of a number of matters which redound to the credit of Mr Revell, who displayed impressive knowledge of the subject, including the chemistry. Nor have I taken account of a number of criticisms of Mr Barden, although the criticisms which he accepted more or less readily in cross-examination might well be regarded as troubling.