Glencore Energy UK Limited v Nis J.S.C. Novi Sad

In this action, the Claimant (“Glencore”) claims reimbursement of the sum of US$2,094,000 which was paid to the Defendant (“NIS”) by Citibank pursuant to a performance bond (“the Performance Bond”) opened at the request of Glencore.

On 28 May 2020, NIS had demanded payment of that sum on the basis that it was due to it because of the delivery by Glencore at Omisalj, Croatia, of a cargo of crude oil pursuant to a contract of sale dated 24 January 2019 (“the Sale Contract”). That cargo was contaminated with organic chlorides, as I will explain, and NIS ended up paying storage fees in respect of the contaminated oil to Janaf Nafteved jsc (“Janaf”), which is the operator of the terminal at Omisalj.

There is no dispute that NIS would have to reimburse Glencore if and to the extent that its demand for US$2,094,000 exceeded what was actually due to NIS from Glencore at the time of the demand. The main issues between the parties concern whether Glencore’s liability for those storage charges was limited by the terms of a settlement agreement which the parties entered into dated 19 March 2020 (“the Settlement Agreement”) and, if so, the amount of that liability.

For the most part, this is an issue of construction of the Settlement Agreement. If Glencore is right and the Settlement Agreement now defines Glencore’s liability for storage charges, then there is an expert issue about the prevailing market rates. There is also an issue – or really a side issue – about whether Glencore negotiated in good faith.

Most of the documentary record I was shown, and the cross-examination of the witnesses, was supposedly directed to the factual matrix against which the Settlement Agreement fell to be construed. Some of this crossed that (blurred) line separating relevant factual matrix from irrelevant evidence of the parties’ subjective intentions. For example, I struggled to see how I could be assisted by examples of different draft terms being proposed and rejected, or evidence about what one party or another had supposedly considered to be a “red line” in the negotiation. But, to be fair to the parties, the cross-examination was kept within a reasonably narrow compass and these excursions into irrelevant materials were only brief.

Although this is not really a case which could be said to turn on the witness evidence, I will make a few short comments about my impression of the witnesses from whom I heard.

Glencore called Mr Michal Wawrzyniuk, the trader responsible for the transaction. He is clearly an intelligent man and has been embedded in this dispute for some time. It would be unrealistic to pretend that this had no effect on the content of his evidence. He was alive to the issues and occasionally strayed into arguing the case or seeking to make what he perceived to be a helpful point. In relation to the meeting on 20 January 2020, and on one or two other points, I have not accepted everything he told me. But overall I found him to be an honest witness.

NIS called two witnesses of fact, namely:

NIS also served two witness statements from Sergey Gvozdev, NIS’s in-house lawyer, who provided NIS with legal support following the emergence of the dispute. Glencore did not require him to be called and did not challenge the content of his statements.

I heard from two experts, the parties having been given permission to adduce such evidence on the issue of the “prevailing market rates for storage” during the relevant period.

Glencore called Charles Daly of Channoil Consulting Limited. NIS’s expert was Kevin Waguespack of Baker & O’Brien Inc. Both experts are very experienced in the oil and gas industry. Both answered the questions which they were asked in a balanced and independent way and I did not doubt that each was providing me with his honestly held opinions. As I will explain, neither had very much hard evidence (as opposed to analysis or commentary) to offer in relation to market rates for storage, but both were straightforward about the limitations to their knowledge. I will discuss their evidence, and its limits, when I come to the issues about market rates in section F below.

I turn next to provide a short account of the factual background. I should perhaps explain that there are only really three factual areas or topics which are relevant. The first is the background, or factual matrix, to the Settlement Agreement. The second is NIS’s allegation that Glencore failed to negotiate in good faith. That concerns events after the signing of the Settlement Agreement. The third concerns NIS’s and Glencore’s respective dealings with Janaf. Some of those dealings are said to shed light on the issue about the prevailing market rate for storage. I am going to address the second and third of these topics in the specific sections of the judgment which concern good faith and the prevailing market rate. I will focus in this section on setting the scene for the Settlement Agreement.

Very little of this factual background was genuinely controversial. For example, an apparent disagreement about the content of the meeting on 20 January 2020 seemed to me more of a difference of perception than a dispute about the specifics of what had been said at that meeting. In the context of identifying the relevant factual matrix for a written contract, such subjective perceptions seem unlikely to be important.

Each party complained that the other was referring to evidence which was inadmissible, or at least irrelevant, cherry picking from communications or negotiations, or treating evidence about their own subjective intentions as part of the factual matrix. There was force in all of these complaints. As I have said, the line between evidence of knowledge which is shared, and evidence of subjective intentions which are communicated, often becomes blurred. In this section, I will focus on telling the story and making findings of fact, before seeking to pull the threads together to describe what seems to me properly to form part of the factual matrix in the next section.

Having said that, I am not going to make findings about every detail canvassed with the witnesses, or refer to every document which I was shown. Instead, I will focus on the points which seemed to me to be of potential relevance to the issues of construction. As I say, much of this was common ground, but to the extent there were differences between the parties, what follows represent my findings of fact.

On 24 January 2019, Glencore and NIS entered into the Sale Contract, by which Glencore agreed to sell and NIS agreed to buy a total of 1,100,000mt (+/- 20% in NIS’s option) Export Blend crude oil of KBT quality, for delivery at Omisalj, Croatia.

KBT crude oil comes from Iraq. It is exported via the Kirkuk-Ceyhan oil pipeline and is loaded into tankers at Ceyhan in Turkey. It was to be delivered at Omisalj in parcels of 80,000-85,000mt +/- 10% in Glencore's option. It could be delivered DES (i.e. ex-ship), or by ITT (i.e. transfer from a tank at Omisalj under the control of Glencore).

My attention was drawn in particular to the following clauses of the Sale Contract:

I observe also that clause 4.7 of the Sale Contract required Glencore to procure a bond by way of security for its performance of its obligations.

On 30 January 2019, Citigroup issued the Performance Bond, up to the amount of USD 12m. The Performance Bond provided that Citibank would pay NIS on its first demand, upon receipt of written confirmation stating that Glencore had not performed its obligations under the Sale Contract. It is a genuine “demand bond”, requiring only a written confirmation in that form.

On about 27 December 2019, the MT Ottoman Equity (“the Vessel”) completed loading of a cargo of around 90,000mts of KBT crude oil (“the Cargo”) at Ceyhan, Turkey. The Vessel then sailed to Omisalj, arriving in late December 2019.

Discharge at Omisalj took place between 31 December 2019 and 1 January 2020 into two tanks at Janaf’s Omisalj terminal, whereupon Janaf commenced the transfer of the Cargo through its pipeline to NIS’ refinery (Pancevo Oil Refinery), which is in Serbia.

I understand that Janaf’s storage tanks are constructed or operated in such a way that a quantity of crude oil will remain in the tank, even when the tank is “emptied”. I assume that this is something to do with the location of the suction header. This remaining quantity is referred to as the “heel” and it seems that the “heel” is agreed to belong to Janaf, not the party putting oil into, or taking it out of, the tank.

I should also note that Janaf publishes Technical Conditions for Access to Transport Capacities which are said to be mandatory for Janaf and its users. Article 6 says that “JANAF will refuse to transport and store any oil whose characteristics could adversely affect other oil being transported or could cause damage to JANAF installations, or to the installations of other users”. Mr Wawrzyniuk agreed that this was the sort of provision which he would expect a facility like Janaf to include in its terms.

On 8 January 2020, a call took place between Mr Wawrzyniuk of Glencore and Mr Zirojevic of NIS, in which Mr Wawrzyniuk informed Mr Zirojevic that other cargoes of crude loaded at Ceyhan at around the same time as the Cargo had been found to have been contaminated with organic chlorides. Mr Wawrzyniuk said he did this to prevent NIS processing the Cargo before they knew if it was OK to do so.

I note in passing that there was a suggestion that, when loading, the Cargo also had a high level of basic sediment and water (BS&W). But this water etc. had largely settled out of the Cargo by the time the Cargo arrived at Omisalj. The BS&W levels were mentioned in some of the correspondence, but my understanding was that it was only the levels of organic chlorides which actually caused a problem with the Cargo.

I understood it to be accepted by Mr Wawrzyniuk that elevated levels of organic chlorides meant that the Cargo fell foul of Janaf’s prohibition in Article 6 of the Technical Conditions: i.e. this was oil which could adversely affect other oil being transported. I am told that processing high levels of organic chlorides is problematic for oil refineries as it promotes rapid corrosion of equipment and also a loss of performance due to the fouling of equipment. It seems that organic chlorides are not often found in crude oil and there was therefore no routine testing for their presence at the loadport.

On 13 January 2020, NIS received test results which confirmed that the Cargo did indeed have a high organic chloride content. It was agreed by the parties that the Cargo was ultimately found to have an organic chloride content of 32 ppm, as compared with an acceptable limit of perhaps 1 ppm.

Mr Zirojevic confirmed to Mr Wawrzyniuk that the cargo had been found to be contaminated with organic chlorides. But, in the interim, around 17,000mts of the Cargo had been transported towards NIS’ refinery (the parties referred to this quantity as the “Transit Portion”). The remainder of the cargo was held in two tanks at Omisalj terminal (the parties called this the “Balance Portion”). As explained above, loading the Cargo into the tanks meant that the heels which had already been in each tank had been contaminated. This was labelled the “Tanks Technical Oil”. Janaf also alleged that the process of transferring the Transit Portion had contaminated the heel in a tank within Janaf’s system at Sisak. This was referred to as the “Sisak Technical Oil”.

On 15 January 2020, NIS wrote to Glencore complaining about the quality of the Cargo and referring in particular to the high organic chloride content. NIS said that Glencore should assume responsibility and take over further handling of the oil and asked its representatives to attend a meeting in Belgrade “to address all the current problems”.

On 20 January 2020, that meeting took place. Mr Wawrzyniuk said that he did not really prepare for the meeting and that he saw it as an opportunity to show that “we are there as partners to work it out”. His evidence was that this was an issue which Glencore “took very seriously and we wanted to rectify it”. He went with his superior, Maxim Kolupaev, Glencore’s Global Head of Crude Oil. They did not take any lawyers. They were therefore surprised when NIS’s lawyers attended. His position was that he and Mr Kolupaev had made helpful noises, but nothing had been formally agreed on behalf of Glencore. He emphasised that he and Mr Kolupaev were not authorised to sign contracts on behalf of Glencore.

NIS relied upon a minute which had been prepared after the meeting and sent to Glencore on 22 January 2020 “for your agreement”. The minute, which was in Russian (which both Mr Kolupaev and Mr Wawrzyniuk speak fluently), suggested, among other things, that:

Mr Wawrzyniuk said that the minute was “factually incorrect”, but a decision was made by Glencore not to comment or seek to amend it, apparently because “We did not want to distract ourselves and devote time to discussions about who said what in Belgrade, since it would not bring us closer to a solution”. I did not entirely follow this explanation, or Mr Wawrzyniuk’s attempt to expand upon it orally.

It seems to me that, if Mr Wawrzyniuk individually, or Glencore more generally, had disagreed in any important way with the content of the minute, they would have said so at the time. At one point in his evidence, Mr Wawrzyniuk seemed to suggest that this might be true if he had been sent a record of an agreement with a counterparty, but he would not bother to respond if there had in fact been no agreement. I do not accept that evidence, which felt to me like an attempt to explain in retrospect the absence of a response. I find that Mr Wawrzyniuk and Glencore did not respond because they broadly agreed with the content of the minute and did not consider this to be an agreement of a type which would require the input of Glencore’s lawyers, or the approval of those with authority to bind Glencore.

Mr Gvozdev took the view that this was a formal decision or agreement by Glencore. However, since we are only concerned with the meeting, and the minute, in the context of the factual matrix to the Settlement Agreement, I doubt that the detail of the participants’ competing perceptions matters. As a matter of the objective impression conveyed by Glencore to NIS, both at the meeting and by its response to the minute, it is clear to me that Glencore was accepting the principle that (assuming further testing confirmed the level of organic chlorides) it would have to resolve this problem, by taking back the contaminated oil and making NIS whole. Perhaps Mr Kolupaev and Mr Wawrzyniuk had not yet, at least in any formal and legally binding sense, agreed to anything specific. But the message which was being conveyed was definitely a conciliatory and encouraging one.

On 24 January 2020, NIS wrote to Janaf, telling it that Glencore was accepting responsibility and was prepared to reimburse all costs. It asked Janaf to quantify the damage suffered. It also referred to the fact that the parties (i.e. Janaf and NIS) had not yet signed a contract for the transportation of crude oil in 2020. NIS said that “Signing the contract is a very important issue as a legal basis for further functioning of all mutual operational activities and we expect a draft of it during the next week, as promised to us and in accordance with the agreements at previous meetings”.

Glencore’s submission was that the fact that NIS needed this transportation contract if it was to keep its refinery operational put NIS in a very weak negotiating position as against Janaf. That may be correct, although I could not see any evidence that Glencore knew at the time that NIS and Janaf were seeking in this same period to agree a new transportation contract. It seems to me that the information crossing the line was only that, in a more general way, NIS needed Janaf and would be working closely with them.

On 27 January 2020, Janaf provided a statement of its claim in relation to the contamination, suggesting a figure of $8,667,331.50 for the oil in Janaf’s system which had been contaminated by being mixed with the Cargo and $766,500 for “Ad hoc storage” (up to 27 January 2020). The calculation for the latter referred to a price of $1.50 / cbm / decade.

On 28 January 2020, NIS sent Glencore a formal letter of claim in relation to the contamination, which included a statement that NIS would seek to pass on charges from Janaf in relation to storage of the contaminated cargo, under clause 2.2 of the Sale Contract. That letter of claim attached the letter from Janaf referred to above.

Mr Wawrzyniuk agreed that he had known that Janaf was charging for ad hoc storage at the rate of USD 1.5/m³/decade and that Glencore expected to have to pay storage fees. His objection was to the figure, not the principle that Glencore should pay the fees:

However, his written evidence was that, in his view at the time, “the market rate for storage was zero”. When asked about this, he made clear that he was not intending to suggest that he thought crude oil could be stored for free, but rather that, at that time, there was no interest in buying storage from those who did not have a presence in Omisalj. He insisted that it was never disputed by Glencore that, if you have oil that you need to store, you would have to pay someone to store it in their tanks, and he agreed that the tanks were always worth something to someone who needed to use them. Mr Gvozdev said that Mr Wawrzyniuk had never suggested to NIS that Glencore did not intend to pay anything for the storage. Having myself heard Mr Wawrzyniuk explain Glencore’s position, I am sure Mr Gvozdev is right about that.

On 30 January 2020, however, Glencore responded (via its solicitors), formally disputing the claim on the basis that the Cargo was of the “standard export quality”. Among other provisions, Glencore relied upon clause 13.2 of the Sale Contract, asserting that: “Even if you were to have a quality discrepancy claim against Glencore (which, for the avoidance of any doubt, is denied), it would be premature for you to make a demand under the Performance Bond, in circumstances where Glencore has not made any recovery from its supplier”.

Glencore said that this letter disabused NIS of any notion that Glencore was agreeing to reimburse NIS for any storage fees which were claimed by Janaf. I am not sure that is quite right. Certainly Glencore was making clear that, if NIS did not want to deal with the problem in a collaborative way – if it wanted to call on the Performance Bond – Glencore’s lawyers would weigh in. But Mr Wawrzyniuk’s own email of the same day was more conciliatory. He said “Our position has not changed. Our preference remains to find an amicable resolution”. His evidence was that Glencore wanted to see NIS pushing back on Janaf’s proposed rates, which he was saying were above the market.

From 3 February 2020, Janaf began to send invoices in relation to the storage at Omisalj and Sisak. These invoices referred to “ad hoc” storage and used a rate of $1.50 / cbm / decade.

It is important to note, however, that NIS did not suggest that Janaf had any contractual entitlement to use this, or any other, specific rate. I will explain in more detail in section F.1 below where Janaf’s “default” rate came from. But although Janaf rendered invoices using that rate from an early stage, my understanding of the evidence was that the rate had not been agreed in advance with NIS. Mr Gvozdev said that NIS did not have a contract for storage with Janaf. He referred to the “default” or “ad hoc” rate as having been stated in a 2014 contract which NIS had had with Janaf (see paragraph 177 below). Ms Plese also referred to that rate having been in a 2014 contract and said that it remained “the rate applicable to NIS and generally”. But she made clear that Janaf did not have any contract for the storage of crude oil with NIS.

Also on 3 February 2020, Mr Wawrzyniuk emailed NIS to follow up a telephone call and make a proposal for unwinding the transaction. That proposal envisaged the Cargo being transferred into Glencore’s tanks at Omisalj and then Glencore would provide a replacement cargo.

This began a lengthy series of “without prejudice” exchanges, the detail of which is now irrelevant, which culminated in the parties entering into the Settlement Agreement on 19 March 2020.

There was a mild disagreement between the parties as to the extent to which (and purposes for which) I could even look at the WP exchanges leading up to the Settlement Agreement. There were some issues raised about waiver of privilege and the like. However, it was agreed that I was entitled to look at the messages to the extent that they would otherwise be admissible as part of the factual matrix to construe the agreement (see Oceanbulk Shipping v TMT Asia [2010] UKSC 44 at [36] – [46]). In particular, NIS accepted that the parties could refer to the emails and letters marked “without prejudice” to evidence facts communicated during the course of those negotiations to the extent that they would otherwise be permissible as part of the factual matrix. I could not see any other reason why those negotiations would be relevant and no-one identified any to me.

One message to which both parties referred was the email from Mr Wawrzyniuk dated 14 February 2020 in which he made reference to Janaf’s claims and queried how they had been arrived at. In relation to storage fees, he said: “nor have we received the justification of already provided cost (for example where does the 1.50 USD/m3 per decade storage fee come from – is this from a current JANAF-NIS storage contract or is this a number that JANAF came up with arbitrary?”.

Mr Wawrzyniuk’s evidence was that he wanted NIS to push back against Janaf; that he was consistently telling Mr Zirojevic that the Janaf storage fees were inflated. He accepted that he had not told NIS what Glencore was itself paying for storage at Omisalj, at least before the Settlement Agreement was signed. But he insisted that it was clear that he was saying the rates which Janaf was seeking to charge to NIS were out of line with the market.

Mr Zirojevic for NIS agreed that he knew Glencore was disputing the Janaf rate of $1.50/ cbm/ decade and that Glencore was not happy with that rate and was telling NIS that the market rate was much lower. Mr Gvozdev’s statement also confirmed that Glencore had been concerned that the rate charged by Janaf should be what it (Glencore) regarded as the “prevailing” market rate.

However, on 18 February 2020, representatives of Janaf met with representatives of NIS. The minutes of that meeting record agreement that the storage would be paid for by NIS at $1.5 / cbm / decade. Mr Zirojevic did not attend that meeting, but that matched with his understanding of what had been agreed with Janaf. He did not think that Glencore had been told about the meeting or about NIS’s acceptance of the rate. That fits with the fact that, on 11 March 2020, Glencore was sending an email that there would be a need to agree “the storage fees due to JANAF that would reflect the market conditions and actual loss suffered by JANAF due to the issue”.

Meanwhile, I understand that Glencore had managed to persuade its supplier to take back the Cargo, although there was no specific evidence before me about these dealings between Glencore and its supplier.

By 13 March 2020, the involvement of Glencore’s supplier was generating a degree of time pressure in relation to the negotiations between NIS and Glencore. The vessel sent to lift the contaminated cargo had arrived on 11 March 2020. In an email dated 13 March 2020, Mr Wawrzyniuk pointed out that the laycan was going to expire, but that the tanker could not lift the oil until agreement had been reached and title to the contaminated oil was transferred back to Glencore. Glencore made clear that it could not control how long the ship remained at Omisalj. On 18 March 2020, Glencore passed on a message from its supplier to the effect that the vessel had to load the contaminated cargo “no later than tomorrow March 19^th”.

The Settlement Agreement was agreed on 19 March 2020. I will deal with the contentious parts of the Settlement Agreement in the next section. However, the essential structure of the deal between Glencore and NIS was not controversial.

Following entry into the Settlement Agreement, the parties engaged in initial discussions about various aspects, including the extent to which Glencore was required to reimburse NIS for storage fees paid to Janaf. Given NIS’s allegations that Glencore did not negotiate in good faith, I will deal with these discussions in section E below. For now, it will suffice to say that no agreement was reached in relation to the level of reimbursement for storage fees paid to Janaf.

The dispute was focussed on clauses 33 to 36 of the Settlement Agreement. These provide as follows:

“33.

Subject only to clause 34 below, NIS will procure that Janaf undertakes, represents and warrants to NIS that it will not present any further claims to NIS or Glencore arising out of and/or in connection with the Cargo passing through and/or being stored at the Terminal. Likewise, NIS undertakes, represents and warrants to Glencore that it will not seek an indemnity and/or any compensation whatsoever from Glencore in relation to any liability it incurs to Janaf as a result of the Cargo (or any part of it) passing through and/or being stored at the terminal.

34.

Notwithstanding clause 33 above, the Parties accept that (i) Janaf will be entitled to bring a claim against NIS in respect of storage fees incurred in relation to the Balance Portion and the Sisak Technical Oil from the date of delivery of the Cargo to the Terminal until the Effective Date; and (ii) NIS will be entitled to seek compensation from Glencore for any liability it incurs to Janaf in accordance with 34(i). Glencore will reimburse NIS for such liability up to the extent such liability accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage during the period when the Cargo was stored in the Janaf system. The Parties will discuss in good faith with a view to agreeing the level of reimbursement.

THE OUTSTANDING CLAIMS

35.

This Agreement is in full and final settlement of any and all claims between the Parties arising out of and/or in connection with the delivery of the Cargo, save for:

a.

a claim by JANAF against NIS for storage costs in relation to the Balance Portion;

b.

a claim by NIS against Glencore for any liability that NIS incurs to Janaf for storage of the Balance Portion and the Sisak Technical Oil from 08 January 2020 until the Effective Date (both dates included). Glencore will reimburse NIS for such liability to the extent that such liability accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage. The Parties will discuss in good faith with a view to agreeing the level of reimbursement.

c.

a quality discrepancy claim by N IS against Glencore in respect of the Transit Portion;

d.

a claim by Glencore against NIS in accordance with clause 9 above; and

e.

a claim by Glencore against NIS in relation to clause 31 above.

(the “Outstanding Claims”)

36.

Any and all Outstanding Claims between the Parties will be presented and dealt with in accordance with the terms of the 2019 Contract.”

I will draw attention also to:

In essence, the issue concerns the extent to which clauses 34 and 35 of the Settlement Agreement circumscribe NIS’s right to recover from Glencore the sums it paid to Janaf in respect of storage charges.

Without intending to do any disservice to the careful and detailed submissions made on behalf of each party:

For the avoidance of doubt, I should say that NIS initially contended that this part of clauses 34 and 35b amounted only to an unenforceable agreement to agree, but that argument was (very sensibly) dropped by the time of closing submissions.

If Glencore is right that the sentence “Glencore will reimburse NIS for such liability up to the extent such liability accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage during the period when the Cargo was stored in the Janaf system” represents the full amount of Glencore’s liability for Janaf’s storage charges, there are issues about what is meant by “the actual loss suffered by Janaf” and by “prevailing market rates for storage during the period when the Cargo was stored in the Janaf system”.

Glencore submits that “actual loss” means a loss suffered because Janaf was prevented from renting out the relevant tanks which were being used to store contaminated cargo and that, if Janaf was not financially worse off as a result of storing the Cargo, Glencore has no liability. It argues that the “prevailing market rates for storage during the period when the Cargo was stored in the Janaf system” means typical rates for planned, “take or pay” storage of uncontaminated crude at comparable terminals, including but not limited to Omisalj.

If this stage of the argument is reached, NIS says that, if Janaf’s tanks were being used for storage, Janaf suffered an “actual loss”. It argues that the market rates have to be for storage in the circumstances encountered here, namely the emergency and ad hoc storage of this quantity of contaminated crude oil at Omisalj (and Sisak).

There was no dispute between the parties as to the correct approach to construing contracts. I was reminded of the useful description of the Court’s task provided by Popplewell J. in Lukoil Asia Pacific Pte Ltd v Ocean Tankers (Pte) Ltd (The Ocean Neptune) [2018] EWHC 163 (Comm) at [8]:

“The court’s task is to ascertain the objective meaning of the language which the parties have chosen in which to express their agreement. The court must consider the language used and ascertain what a reasonable person, that is a person who has all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contract, would have understood the parties to have meant. The court must consider the contract as a whole and, depending on the nature, formality and quality of drafting of the contract, give more or less weight to elements of the wider context in reaching its view as to the objective meaning of the language used. If there are two possible constructions, the court is entitled to prefer the construction which is consistent with business common sense and to reject the other. Interpretation is a unitary exercise; in striking a balance between the indications given by the language and the implications of the competing constructions, the court must consider the quality of drafting of the clause and it must also be alive to the possibility that one side may have agreed to something which with hindsight did not serve his interest; similarly, the court must not lose sight of the possibility that a provision may be a negotiated compromise or that the negotiators were not able to agree more precise terms. This unitary exercise involves an iterative process by which each suggested interpretation is checked against the provisions of the contract and its commercial consequences are investigated. It does not matter whether the more detailed analysis commences with the factual background and the implications of rival constructions or a close examination of the relevant language in the contract, so long as the court balances the indications given by each.”

D.4 Relevant factual matrix

I turn then to summarise what seems to me to be the relevant factual matrix, drawing on the facts as found in section C above. Some aspects of this were agreed.

For example, it was agreed that, as they completed their negotiation of the Settlement Agreement, the parties were under time pressure to agree terms which would enable the Cargo to be lifted before the vessel which had been sent by Glencore’s supplier departed. This had the result that various matters were not finally settled and were left over for subsequent discussions and agreement.

NIS invited me to assume that it had negotiated the Settlement Agreement from the starting point that it had unassailable claims for any costs charged to it by Janaf, on the basis that clause 2.2 of the Sale Contract applied. I agree with Mr Hill KC (for Glencore) that such a claim by NIS had the potential to be more complicated, as a matter of law, than Mr Tozzi KC (for NIS) sought to make it appear. First, it seems to me that such a claim would indeed be a quality claim and hence caught by clause 13.2 of the Sale Contract. That mean that NIS’s recovery from Glencore might have depended on Glencore’s recovery from its own supplier. Secondly, the fact that the Sale Contract provided for quality to be conclusively tested at the loadport, when no testing had actually been carried out for organic chlorides, had the potential to make life difficult for NIS.

However, I agree with Mr Tozzi KC that the precise legal analysis here may be less important to the factual matrix than the fact that Glencore was adopting the commercial position that, at least in general terms, if a deal was done whereby the contaminated oil was taken back, Glencore would compensate NIS for these storage fees. I refer, for example, to my findings in relation to the meeting on 20 January 2020 (see paragraphs 31 - 35 above).

Mr Hill KC sought to counter this by pointing (for example) to letters from Ince in which a harder line was taken on behalf of Glencore (e.g. paragraphs 42 - 43 above). However, I see this as Glencore’s lawyers waiving the big stick to encourage cooperation, rather than a change of tack by Glencore. Mr Wawrzyniuk repeatedly accepted that he was not saying to NIS that nothing would be paid by Glencore unless Glencore’s supplier could be persuaded to pay:

That said, it does not follow that the settlement was being agreed against the background of NIS already having a clear legal entitlement to a full indemnity in respect of whatever Janaf sought to charge by way of storage charges, let alone an entitlement to everything which Glencore ended up agreeing to in the Settlement Agreement. In order to put myself in the position of the parties as they finalised the Settlement Agreement, it seems to me that I have to see shades of grey, not just black and white. Glencore was indicating a commercial willingness to resolve the problem and make NIS whole, but it was also taking the formal legal position that there was no breach and that NIS would have to pay for the Cargo.

An aspect of this balancing act was that the parties were already aware how much Janaf was claiming from, and indeed invoicing to, NIS, in respect of storage. They knew the rate which Janaf was using ($1.5 / cbm / decade). Importantly, the parties were also aware that Glencore was saying those charges were excessive and that the market rate was lower. Glencore did not know that NIS had already agreed with Janaf that it would pay at the disputed rates.

The parties were aware that NIS was not in a strong commercial position as against Janaf, because Janaf controlled the sole, or primary, means of the supply of oil to NIS’s refinery. Mr Tozzi KC observed that nothing had changed in this regard since the parties had agreed the indemnity in clause 2.2 of the Sale Contract. That is a valid point so far as it goes, but it does not change the fact that a party in the position of Glencore might reasonably have had a concern, when negotiating the Settlement Agreement, that NIS would end up being pressurised to pay above the market rate for storage, in order to protect its own relationship with Janaf. Glencore might take the view that it should not be required to fund that.

Although there was no positive evidence from either party about the extent to which Janaf’s tanks were being utilised in early 2020, Glencore said that there was no pressure on storage at Omisalj in this period and that everyone knew it. Mr Wawrzyniuk said that, until the oil market changed in mid-March 2020, there was simply no demand for storage tanks.

One uncontroversial aspect of this was that the parties knew that the tanks which Glencore rented from Janaf were empty during this period. Mr Zirojevic agreed that he had been told that Glencore’s tanks at Omisalj were empty. Mr Wawrzyniuk said that Glencore was happy to have the Cargo stored in those tanks. That possibility, or at least the possibility that the contaminated oil would be transferred to Glencore’s tanks, does seem to have been considered at one point, but it looks as if everyone quickly formed that view that moving the Cargo into different tanks would just contaminate more heel and more pipes. Mr Wawrzyniuk described it as “creating more problems than solving”.

However, Mr Wawrzyniuk also suggested that, if Janaf had needed additional storage tanks in Omisalj for third party cargo, it could have used Glencore’s tanks. He said that NIS was “aware of this”. If he was suggesting that this option was actively offered by Glencore to NIS or Janaf at the time, I reject that evidence. I can see no sign that Glencore ever proposed to Janaf that it might “borrow” one or more of Glencore’s tanks to store third party cargo. But perhaps at least part of the reason why that idea never occurred to anyone at the time is that nobody understood that Janaf was struggling to find space to store crude oil in the first part of 2020.

Having regard to that background, it seemed to me that the basic structure of clauses 33-36 is easy to understand.

Clause 33 requires NIS to procure that Janaf give up, and prevents NIS passing on to Glencore, any claim arising out of the Cargo, save as specifically provided in clause 34.

Clause 34 then:

Clause 35 compromises all claims between the parties “arising out of and/or in connection with the delivery of the Cargo”, but preserves 5 specific claims (defined as the “Outstanding Claims”). The claims referred to in clause 35a and 35b match with those described in clause 34. This might be said to be duplicative, but it leaves little room for doubt.

I was not persuaded by NIS’s submission to the effect that clauses 34 and 35 did not finally settle anything in relation to Janaf’s storage charges, because the parties had agreed to preserve NIS’s right make a claim pursuant to the Sale Contract if the discussion in good faith was unsuccessful. It seemed to me that this misunderstood the way in which clause 34 had been put together and gave little or no real meaning to the words: “Glencore will reimburse NIS for such liability up to the extent such liability accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage during the period when the Cargo was stored in the Janaf system”.

NIS pointed out that the first sentence of clause 34 provided that Janaf would be entitled to bring a claim and that NIS would be entitled to seek compensation from Glencore for any liability it incurred to Janaf. That is right, but I would suggest that it is more important that NIS’s entitlement is to “to seek compensation from Glencore” (emphasis added). It does not say that NIS is entitled to be paid compensation for “any liability” (which was what NIS seemed to be saying it meant). To my mind, it is implicit in this formulation that NIS might not actually receive compensation for all the liability it had incurred to Janaf.

That seems to me to fit with the next sentence: “Glencore will reimburse NIS for such liability up to the extent…”. That describes Glencore’s obligation to reimburse NIS for that liability to Janaf, which is subject to an identified limit. It is to be contrasted with the parts of the clause which record what “the parties accept”, because this is Glencore’s obligation.

NIS said that this was identifying only what Glencore had already accepted it would pay; that it functioned as a “floor” for the purposes of the parties’ discussion in good faith. But there are a series of problems with that submission.

First, it would be a peculiar stipulation to include in a settlement agreement: that Glencore agreed to pay at least $x, but failing agreement on how much more than $x was to be paid, a claim would then be brought under the Sale Contract. One might just about be able to see the logic for identifying a “floor” for the level of reimbursement if $x was given as a specific number. But here it represents a level which was (and is) itself controversial. So what useful purpose is it serving? What use is setting a floor when you are arguing about where that floor is? Which party would be seeking the inclusion of this stipulation in the Settlement Agreement?

Second, and further to this problem with identifying what useful purpose the sentence is serving, it was unclear to me what continuing role, if any, this sentence was said to have if the good faith discussions failed. If a claim was brought under the Sale Contract, is Glencore still bound by this agreement to reimburse at least to that extent? If so, how does that fit with NIS’s arguments on clause 36 (see below)? If not, why are the parties bothering to describe one side’s (but not the other’s) position in advance of a good faith negotiation?

Third, it seems to me that the words used are not remotely suggestive of an agreement that this is the minimum which Glencore will pay. It says “Glencore will reimburse” and then identifies the amount in a way that suggests it represents a maximum (“up to the extent”), not a minimum. There is no reference to this being a minimum, or the starting point for the negotiation. If the clause simply said “Notwithstanding clause 33 above, Glencore will reimburse…”, NIS’s argument would not even get off the ground. Yet the prior sentence only preserves Janaf’s right to make a claim and NIS’s right to “seek compensation”. That makes clear that there is no acceptance by NIS that the storage fees claimed by Janaf necessarily exceed what Glencore will reimburse. But it does not alter the meaning of this clear statement, in the context of all other claims being settled, about the level of reimbursement which Glencore is (now) obliged to provide.

Fourth, the final sentence of clause 34 provides that “The Parties will discuss in good faith with a view to agreeing the level of reimbursement”. The “level of reimbursement” must be a reference back to the previous sentence and the words “Glencore will reimburse NIS”. That suggests to me that the good faith discussion is to be about the extent of that reimbursement; i.e. the reimbursement specifically provided for in the prior sentence. Yet, as I understand NIS’s case, the prior sentence is merely identifying what is already on the table. If that were right, one would expect the parties to be discussing what additional amounts are to be paid, in addition to that minimum, not discussing that minimum level of reimbursement.

Another way of expressing all of this is that it is just about possible to read clause 34 in the way suggested by NIS, but it requires one to assume that the parties have not expressed themselves very clearly. I cannot see why, in the context of a settlement agreement, I should be looking for ways of reading the words so as to limit their effect. I do not see why I should be anxious to read a settlement agreement in a way which prevents it from narrowing the scope of the disputes between the parties, or serving any very useful purpose.

To my mind, this is cemented by the way that the Outstanding Claims are defined in clause 35, with the Settlement Agreement being described as otherwise in full and final settlement of any and all claims. NIS submitted that, because clause 35 referred to “Outstanding Claims”, it followed that these were not being settled by the Settlement Agreement. In one sense, that is true. The claim in relation to storage charges was not finally resolved. But that “Outstanding Claim” is then defined in subparagraph (b), and it is again said: “Glencore will reimburse NIS for such liability to the extent that such liability accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage”. It is clear to me that this language is intended to narrow the scope of the dispute over NIS’s entitlement to these sums.

In his oral closing, Mr Tozzi KC placed heavy reliance on clause 36, which he submitted meant that, if there was no agreement, the Outstanding Claims were to be resolved under the Sale Contract. I do not accept that submission. Clause 36 is perhaps not as clearly expressed as clauses 34 and 35, but it makes no sense to suggest that it undermines everything which comes before. NIS said that it was important to read the agreement as a whole, and not to be too heavily influenced by references to dogs wagging tails or tails wagging dogs. I agree, but when reading the Settlement Agreement as a whole, it is important to observe that clause 36 is a short concluding clause, coming after, and utilising, the definition of the Outstanding Claims. One might expect it to supplement, rather than negate, what has gone before.

In addition to its place in the Settlement Agreement, the fact that it uses the words “presented and dealt with” seems to me important. NIS argued that this showed that clause 36 could not be concerned only with procedural matters (e.g. claims machinery regarding documentation, time bar, etc.), because that would only explain the use of “presented” and give no meaning to “and dealt with”. To my mind, the more obvious point is that referring to claims being “presented and dealt with” would be an oddly fussy way of saying that the Sale Contract, not the Settlement Agreement, is going to govern those claims. That phrase “presented and dealt with” does suggest to me that it is directed to procedural considerations, rather than the substance of the dispute.

It is also important that what is to be presented and dealt with are “Outstanding Claims”, which have been defined and thereby limited in scope.

As I have touched on above, I struggled to understand what NIS was saying about the status of the express stipulations in the Settlement Agreement about the Outstanding Claims. In relation to the claim with which we are concerned, I understood that it was being said that the parties’ obligation to discuss in good faith survived, but that if this negotiation failed, the claim was simply to be fought by reference to the Sale Contract. I understood the same to be said about, for example, the claim in clause 9 concerning the discount to be given on the Transit Portion which NIS had retained, to reflect the contamination. For that claim too, it has been agreed that there will be a good faith discussion.

In making this limited concession, NIS recognises that, if it contends that clause 36 means that there is no need for a good faith discussion, it would follow that clause 36 is cutting across numerous other parts of the Settlement Agreement. But, turning that around, if it is right that the obligation to discuss in good faith survives, then (at least to that extent) it must follow that it is not simply open to NIS or Glencore to bring a claim under the Sale Contract. There is no requirement under the Sale Contract to discuss anything in good faith before presenting a claim. That obligation would have to come from the Settlement Agreement.

Nor is there any express recognition in clause 36 of the Settlement Agreement that claims cannot be “presented and dealt with in accordance with the terms of the 2019 Contract” until the process of good faith discussion is completed. NIS is therefore forced to accept that that stipulation is implicit, presumably on the basis that clause 36 is supplementing, not deleting, the earlier provisions of the Settlement Agreement. I agree. However, it seems to me that once one recognises that that is what clause 36 is doing, any uncertainty about its meaning is resolved. Clause 36 is concerned with how a claim might be taken forward, but always subject to what has already been agreed about those Outstanding Claims in the Settlement Agreement. In relation to the Outstanding Claim in clause 35b, as well as agreeing that they will negotiate in good faith, the parties have agreed the extent to which Glencore will provide reimbursement. That agreement is not affected by clause 36.

I should add by way of footnote that it is not clear to me that the parties had fully thought through what it would mean to say that the claims would be “presented and dealt with in accordance with the terms of the 2019 Contract”. In relation, for example, to the claim in clause 9 concerning the discount to be given on the Transit Portion, for example, Mr Tozzi KC asserted that the substantive question as to the amount of that discount would have to be resolved under the Sale Contract, because nothing was said about it in the Settlement Agreement. Perhaps that is right, although it is fair to say that the Sale Contract does not appear to provide any specific mechanism for arriving at such a discount either. However, none of this matters for our purposes. All that matters is that clause 36 is assuming the existence of an outstanding claim which is already defined in clause 35, and making provision for how that claim is to be presented and dealt with. It is not compelling a different construction of clauses 34 and 35b. The tail is not wagging the dog.

I turn, then, to the meaning of those words which I have held define Glencore’s obligation to provide reimbursement: “to the extent that such liability accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage”.

Glencore argues that these words provide two separate hurdles for NIS to clear in order to make a claim. That is not how I read the clause. Having regard to the factual background, it seems clear to me that, when the parties were arguing about Janaf’s storage fees, Glencore was complaining that these were higher than the prevailing market rates. Avoiding getting caught up in the parties’ subjective intentions and taking it all at a purely hypothetical level, I would suggest that, if that was true, there might be at least three explanations for it.

The first, and in some ways most obvious, would be that Janaf was taking advantage of the difficult position in which Glencore and NIS found themselves. I will say rather more about “emergency” rates and the like in due course, but I was interested by Mr Daly’s response to my questions about whether a terminal in the position of Janaf would see that Glencore and NIS had a problem and look to cash in:

That might be a slightly idealistic statement, but the sentiment seemed to me to be honest. In circumstances where clause 34 clearly envisaged that Janaf would in the future bring a claim, and NIS could therefore dispute that claim (not that NIS was contractually committed to pay, or that it had already agreed to pay), it seemed to me interesting that the industry might expect other parties to be supportive, rather than to use the opportunity to maximise their short term profit. There is an analogy with the law of salvage, where there are mechanisms to allow a price to be fixed retrospectively, recognising that an emergency is not always the best setting for such a negotiation. I can see why the parties might take the view that Janaf should not be allowed to profit (in the sense of obtaining more than the market rate for storage) from this unfortunate situation; that NIS ought to push back against any attempt by Janaf to do so.

A second hypothetical reason for Janaf’s claimed fees to exceed market rates would be that Janaf was in a strong negotiating position as against NIS, for reasons which had nothing to do with the current problem and everything to do with Janaf’s control of the pipeline on which NIS depended. Again, I can see why reasonable parties might agree that Glencore should not have to reimburse NIS for the costs of (as it were) assisting NIS’s business relationship with Janaf. Indeed, an obvious reason for fixing reimbursement at the level of market rates is where there is a concern that ordinary market forces will not otherwise operate.

A third hypothetical reason might be that the circumstances meant that Janaf was experiencing losses of its own as a result of the unexpected need to hold the Cargo in its tanks. I rather doubt that the parties had in mind losses in the form of potential storage customers who would be paying market rates. Glencore was taking the position that there was no pressure on storage. Moreover, the loss of a third party storage customer ought to be compensated by payment (by NIS and then Glencore) of the prevailing market rate.

However, it seems to me that reasonable parties might envisage that Janaf would encounter other problems, inconveniences and costs, as a result of the unexpected blockage, and might take the view that Janaf was entitled to be compensated for those. That would be part of making Janaf whole and of unravelling the transaction, as Mr Hill KC himself submitted was the commercial purpose underlying the Settlement Agreement. I note that Ms Plese made clear that “JANAF does not like holding storage because it blocks their system”. To the extent that what she was doing when she explained that was justifying Janaf’s “default” rate (discussed further below) by reference to such inconvenience, it seems to me that the parties might well have taken the view that that was a legitimate justification for an uplift, and a rather different matter from Janaf saying that it had NIS over a barrel (either because of the situation, or because of the relationship more generally) and thus felt it was in a position to maximise its profit.

That commercial logic seems to me to fit perfectly with the words which the parties have used: “(I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage”. If it said “(I) the direct loss suffered by Janaf and (ii) any consequential loss”, nobody would argue that the loss would have to be both direct and consequential, at the same time, in order to be recovered.

The only reason Mr Hill KC felt able to argue that these were two hurdles, both of which needed to be cleared, was that it would be theoretically possible for Janaf’s actual loss to take the form of lost earnings from being unable to rent those tanks to a third party on the market. But it still seemed to me that reading this as involving two hurdles, rather than as two possible ingredients for Janaf’s storage rate, creates a series of problems.

The first is that a result whereby, unless a third party was actually trying to rent those tanks at the relevant time, Glencore pays nothing, would be commercially absurd. It was not a result which Mr Wawrzyniuk seemed willing to defend:

In circumstances where Glencore’s position at the time was that there was no pressure on the available storage at Omisalj, such that the market rates were low, but also that it was willing to pay the market rate for these tanks, for the parties then to agree a deal whereby if Glencore was right that there was no pressure on storage, it paid nothing at all for this storage, would be absurd. The factual matrix which I have discussed above is important here. But it is also clear from the words of the Settlement Agreement. As the final sentence of clause 34 emphasises, the parties were arguing about the level of reimbursement, not about whether Glencore should be paying anything at all.

Nor does it work for Glencore to talk about making Janaf whole, as opposed to obtaining some kind of “windfall”. The first point is that it can hardly be said to be a “windfall” when a terminal providing crude oil storage is paid the market rate for doing so. The second is that, on Glencore’s construction of the provision, it actually prevents Janaf (or at least NIS) from recovering all of its actual losses. If these are two hurdles which must both be cleared, then, if Janaf suffered an actual loss which exceeded the market rate for storage, Glencore would not have to reimburse this ingredient of the storage fee. So, if Janaf had been incurring additional workaround costs, or suffering delays, in addition to losing ordinary rental income, because of those tanks being unexpectedly blocked, it would not in fact be “made whole”. On Glencore’s construction, Janaf (or NIS) loses at both ends of the spectrum. It might perhaps be thought rational to say that Glencore should only reimburse Janaf’s actual loss, or that it should only pay the market price for storage. But, put together, the combination takes on the appearance of “heads I win, tails you lose”.

NIS proposed another way of squaring this circle, which was to say that having cargo in its tanks was, in and of itself, an actual loss for Janaf. It drew an analogy with the law of trespass and what used to be called “mesne profits”, which NIS said was a payment of damages for actual loss. There is something in this and, if there were no other way of reading clause 34 other than as requiring that both hurdles are cleared, I might have felt driven to say that having cargo in tanks which would otherwise be empty is a sufficient detriment to amount to “actual loss”. But the problem with this approach is that it largely deprives the first part of the provision (“(I) the actual loss suffered by Janaf ”) of meaning. At the time of the Settlement Agreement, nobody was debating whether the Cargo had been in Janaf’s tanks. That fact was recorded in recital E to the Settlement Agreement: “The balance, approximately 71KT of the Cargo… is presently in storage tanks A-1503 and A-1515…at Janaf's terminal at Omisalj”. If that meant that, without more, actual loss had been suffered by Janaf, why include that as a condition which supposedly still needed to be satisfied?

Mr Hill KC’s reason for disagreeing with my suggested approach to “actual loss” was that clause 34 envisaged Janaf claiming only storage fees and hence there was no scope for Janaf to make the equivalent of a damages claim for its “actual loss”. That is correct, but perhaps misunderstands my point. By the time of the Settlement Agreement, the parties knew that Janaf was putting forward a claim for storage fees (at a rate of US$1.5 per CBM / decade). It was not (at least in this context) making any separate claim for “actual losses”. What seems to have been less clear to the parties was how that figure for storage fees was made up, or, perhaps it would be more accurate to say, how it was justified. It seems to me that clause 34 is envisaging that Janaf might suggest that it was entitled to charge more than the “ordinary” market rate because these were extraordinary circumstances and that it was suffering actual losses because of those extraordinary circumstances.

One other potential objection to my reading of this part of the clause is that it might be said that, since this envisages “actual loss” being added to the “prevailing market rate”, if Janaf had (contrary to expectations) suffered actual loss in the form of lost rent which would have been earned from a third party, it could make a double-recovery. There are a number of answers to that, given the circumstances and having regard to common sense. First, putting it at its lowest, the parties did not expect that Janaf would actually have lost rental income. Given the state of the market, that was not the kind of “actual loss” they were contemplating. Second, when the reference to “actual loss” sits alongside “prevailing market rates”, it seems to me that it could and would be read in a way which avoided any double-counting (i.e. as “actual loss” being loss other than lost rental income at the market rate). Last, but not least, this was not defining Janaf’s entitlement, but only the extent to which NIS would be reimbursed if liable to Janaf. The parties are unlikely to have imagined NIS allowing Janaf, in effect, to claim the same loss twice.

Turning finally to the meaning of “prevailing market rates for storage”, it seems to me that the fact that it is envisaged that Janaf will be entitled to justify claiming storage fees which exceed those prevailing market rates by showing that it has suffered an actual loss, is very important. To my mind, this is the route by which the parties envisaged Janaf being made whole in respect of the various factors which distinguished this from “normal” storage, for which there could sensibly be said to be a prevailing market rate.

The most obvious example is contamination. Ms Plese said that no storage facility would accept contaminated oil willingly. The experts agreed that, in effect, there is no market for storing contaminated oil. Mr Waguespack explained that it “carried risks of cross-contamination to other crude oil inventory, as well as potential operating disruptions and additional costs”. This therefore seems to me to fit with the distinction between market rates and actual loss. To the extent that Janaf experienced an additional cost because the oil it was storing was contaminated (either in terms of handling cost, or supervision, or whatever), it could legitimately allow for that in its storage fee. NIS would be able to recover that uplift as compared with the market price (i.e. for storing ordinary, uncontaminated crude) as Janaf’s “actual cost”.

In my judgment, the same analysis applies to the consequences of this being emergency storage, where the price is (or at least is envisaged by the Settlement Agreement as being) agreed only after the event. A scenario in which the contaminated oil is already in Janaf’s tanks, has been there for a significant period, and is on the point of being removed, is inconsistent with the operation of a market. A market requires at least an element of decision-making: to be a willing buyer and willing seller they must each decide they are willing to enter into the transaction. I am not sure that it is very helpful to add further adjectives, such as “functioning”, as a further qualification. It seems to me that, depending on your context, a monopolistic seller can still constitute one half of a market, so long as the buyer (and indeed the seller) is still able to make a choice about whether or not to do business. Whether that would be described as a “functioning” market might be a different matter. By contrast, if that choice is completely removed: if the buyer must buy, or the seller must sell, regardless of the price, that is simply not a market. Similarly, there is no choice to be made if, in effect, the buyer has already bought and consumed and the seller already sold and is unable to take the product back.

As with contamination, the consequences for Janaf of this being an emergency situation could be catered for by uplifting the market rate for ordinary storage (i.e. agreed in advance in the usual way) to reflect any actual losses which are suffered. For example, Mr Waguespack referred to the possibility that “Short-term use of tanks might disrupt typical operations or otherwise limit the flexibility of the terminal, ultimately costing the terminal lost revenue”. If so, those would be actual losses which could be used by Janaf to justify charging a higher rate.

There does not seem to me anything inherently surprising or uncommercial about that result. Janaf should be made whole, but not secure any windfall profit as a result of this unhappy situation. It is an outcome which also fits with Mr Daly’s view as to how participants in this market expect each other to behave when a problem of this kind needs to be solved (see paragraph 100 above).

It seems to me important to distinguish between “unplanned” storage and “emergency” storage. If by the former, one means storage which is being arranged at relatively short notice – i.e. on what Mr Hill KC referred to as a “spot” basis – it seems clear to me that there is a market for such storage. It is clear that traders do decide, at relatively short notice, to store crude oil (for example, so as to make a profit on the difference between prompt price and future price when the market goes into contango). Mr Daly said that “Short term storage deals are often entered into, for example, to discharge a vessel whose cargo has no immediate home, or to keep oil in store whilst the market is in contango. These types of deals can be for a month with an option to rollover or for three to six months depending on the strength of the contango market”. I accept that evidence.

The important difference between a spot fixture of that kind, and an emergency situation like the present, is that, if the trader approaches the terminal and proposes storing a quantity of crude oil, but is given a price which it considers too high, the trader can say “no thanks”. It can go somewhere else, or do something else. In an emergency situation like the present, NIS could not say “no thanks” to storing the contaminated oil in Janaf’s tanks. There would be no scope for a “market” negotiation of that kind.

I should perhaps deal in this context with floating storage. Mr Waguespack suggested that the safety vale in an emergency situation would be to arrange for floating storage. He called this the “only practical storage alternative other than that at JANAF”. His tentative calculation suggested that the total daily cost of doing so would not have been hugely out of line with the cost of paying Janaf’s “default” rate, although the floating storage would still have been more expensive. NIS relied upon that in support of its case that the Janaf “default” rate was, or was in line with, the market.

This seems to me a false point, given the context in which the Settlement Agreement was being finalised. The parties knew that floating storage had not been arranged. On the contrary, the starting point for investigating the “prevailing market rates for storage” was that the oil had in fact been stored for that period by Janaf in its tanks. If it had been pumped into a vessel, the storage period at Omisalj would, by definition, have come to an end.

In my judgment, the relevant components of this “prevailing market rates for storage” are (a) the quantities actually stored (or, to be more precise, the storage capacity actually used); (b) the period for which the oil was stored; and (c) the location in which it was stored (i.e. Omisalj and Sisak). In simple terms, I would understand the exercise to be to duplicate as closely as possible the actual storage, having first stripped out any features which are not compatible with the identification of a market rate, and which therefore need to be addressed by reference to actual loss (if any).

The last of those (i.e. the location of the storage) was perhaps the only one which was controversial. Glencore argued that the reference to market rates could “look beyond Omisalj”. This was said to be a new argument and Mr Tozzi KC pointed out that the expert issue had been formulated (apparently by Glencore’s legal team) to refer specifically to “storage at Omisalj and Sisak”. The forensic point (i.e. that this is an after-thought on the part of Glencore) may be fair, although if this was a good argument on the construction of the Settlement Agreement, it would not stop being a good argument just because of the way in which the expert issue had been drafted.

However, Mr Tozzi KC was right that this was not a good argument on the construction of the Settlement Agreement. Mr Hill KC argued that clauses 34 and 35b referred to “rates” and argued that the plural meant that the parties cannot have had in mind a single Janaf rate. Mr Tozzi KC responded that clauses 34 and 35b were dealing with storage at both Omisalj (the “Balance Portion”) and Sisak (“the Sisak Technical Oil”), which sufficiently explained the use of the plural. I agree. I also agree with Mr Tozzi KC that the fact that the whole sentence is “prevailing market rates for storage during the period when the Cargo was stored in the Janaf system” is relevant here. It is true to say that the emphasis in the final phrase is on the period, not the location. But there is still a link being drawn here between rate and location.

More generally, it seems to me that location is like quantity and period: you cannot sensibly talk about a market rate without identifying it. It is obvious that the market rate might differ depending on whether one is talking about storing for 5 years or 5 days, and equally obvious that the rate might differ depending on whether one is talking about storing at Omisalj or Rotterdam.

The complexity in the present case comes from the fact (discussed in section F below) that such limited information is available about storage prices generally, let alone storage prices at Omisalj, that it is inevitable that the experts need to look at storage costs more widely. There is nothing inherently wrong with that.

However, if there had been no such evidential difficulty, and the market rate for Omisalj could be confidently established by reference to an index, or a healthy number of publicly available comparators (as with commodity prices, or spot rates for different types of vessel), it does not seem to me that it would make any sense to say that the parties were here contemplating a rate for Fujairah or Rotterdam, or a blended rate across a region. It is the market rates for Omisalj and Sisak which the parties must have had in mind.

The only caveat to that would be if there was not anything which could be described as a market for storage in Omisalj. If Janaf did not offer any storage facilities at all, or offered storage only on a fixed tariff basis, and the parties had known that when agreeing the Settlement Agreement, that might suggest that they must have intended to use a market rate for the region. But it seems to me one would need to be driven to that conclusion by necessity.

That does not seem to me to be the situation here. It may be right to say that Janaf was a monopoly provider for storage at Omisalj. But, as Glencore’s own dealings with Janaf reveal (see further below), there was still a “market” for storage, in the sense that there could still be a negotiation about whether a potential user would pay the price proposed by Janaf, as we can see happened with Glencore.

Drawing the threads together, then, I understand clauses 34 and 35b of the Settlement Agreement to fix Glencore’s liability to reimburse NIS in respect of Janaf’s storage fees to an amount which “accurately reflects (I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage”. The “actual loss” ingredient means any uplift to the storage fees which is charged to reflect that Janaf is suffering an actual loss, due to the contamination of the Cargo, or due to the emergency nature of the storage.

Accordingly, the “prevailing market rates” are to be rates for storage at Omisalj and Sisak, of approximately the volume used, for the period in question, on a “spot” basis.

I turn next to the issues as to whether Glencore breached the obligation to enter into or conduct “discussions in good faith” and, if so, the consequences.

NIS says that Glencore failed to enter into good faith discussions, but instead sought to delay the negotiations, perhaps hoping the Performance Bond would expire before any demand was made.

I will say straight away that, in the present context, it seems to me that NIS’s allegation of breach goes nowhere. NIS argues that, if Glencore had entered into good faith negotiations, it would have agreed to reimburse NIS for Janaf’s storage costs in the sum ultimately demanded by NIS under the Performance Bond. To the extent that what it means is that, if NIS had remained entitled to bring a claim for the storage charges under the Sale Contract (i.e. the case I rejected in the previous section), or if the Janaf “default” rate were the prevailing market rate for storage in Omisalj (discussed in the next section), that would be what was payable, then I understand the logic. But if NIS is right about that sum being its legal entitlement under the Sale Contract or the Settlement Agreement, it does not need to argue about breaches in relation to good faith.

If what NIS is suggesting is that I should hold that, if Glencore had entered into good faith negotiations, it would have agreed to pay more than I find to be NIS’s entitlement under the Settlement Agreement, that seems to me optimistic, to say the least. There is absolutely no evidence here that Glencore might have been willing to reimburse NIS for more than its legal entitlement; quite the contrary.

There have been a number of cases dealing with “good faith” obligations in recent years. I find the most helpful description of the content of that obligation to be that provided by Leggatt J in Al Nehayan v Kent[2018] EWHC 333 (Comm) at [175]:

“the usual content of the obligation of good faith is an obligation to act honestly and with fidelity to the bargain; an obligation not to act dishonestly and not to act to undermine the bargain entered or the substance of the contractual benefit bargained for; and an obligation to act reasonably and with fair dealing having regard to the interests of the parties (which will, inevitably, at times conflict) and to the provisions, aims and purposes of the contract, objectively ascertained. In my view, this summary is also consistent with the English case law as it has so far developed, with the caveat that the obligation of fair dealing is not a demanding one and does no more than require a party to refrain from conduct which in the relevant context would be regarded as commercially unacceptable by reasonable and honest people”.

That final sentence about the obligation not being a demanding one and ultimately only requiring “a party to refrain from conduct which in the relevant context would be regarded as commercially unacceptable by reasonable and honest people” echoes the formulation proposed by the same judge in Astor Management AG v Atalaya Mining plc[2017] EWHC 425 (Comm) at [98]:

“will act honestly towards the other party and will not conduct itself in a way which is calculated to frustrate the purpose of the contract or which would be regarded as commercially unacceptable by reasonable and honest people”.

It appears again in New Balance Athletics, Inc v The Liverpool Football Club and Athletic Grounds Ltd [2019] EWHC 2837 (Comm), per Teare J at [44]:

“In judging whether a party has not been faithful to the parties' bargain it is of course necessary to bear in mind the nature of the bargain, the terms of the contract and the context in which the matter arises. Ultimately, the question for the court is whether reasonable and honest people would regard the challenged conduct as commercially unacceptable…”

It does not seem to me that I need to gloss that description further: I am looking for conduct which would be regarded as commercially unacceptable by reasonable and honest people. It is a classic “jury” question; the type of question where, if the issue were being decided by a jury, one would direct them that they should consider themselves to be reasonable and honest people, and on that basis form a view of what is “commercially unacceptable”, having regard to all of the circumstances and context.

In terms of consequences of breach, in Petromec Inc v Petroleo Brasileiro SA Petrobras (No. 3)[2005] EWCA Civ 891, Longmore LJ explained how causation and loss would work in the context of an obligation to seek in good faith to agree the reasonable cost of an upgrade (at [118]):

“…If the court is able to conduct the exercise of finding the reasonable cost to Petromec of the upgrade, there should be no difficulty in deciding what the result of good faith negotiations is likely to have been. Unless there are special factors present, it is likely to be the same as the reasonable cost. No doubt there could be argument in the present case as to whether, if negotiations did not proceed (but should have proceeded) in good faith, they would have embraced an uplift and whether, in that event, the uplift would have been in any particular amount, but it is not uncommon for courts to have to assess, by way of calculating damages, whether a claim against a third party was good or not and for how much it might have been settled. Any exercise in relation to uplift would raise similar (but not insurmountable) problems…”.

I understand this to mean that, absent “special factors”, the product of the good faith negotiation will match the basic entitlement (e.g. to be paid the reasonable cost). There may be arguments, on the facts of the particular case, about whether an uplift might have been paid. But the essence of the exercise would be to decide, on the evidence, what final figure would have been arrived at in the negotiation.

On 23 March 2020, Mr Wawrzyniuk emailed NIS to thank them for their cooperation in the removal of the contaminated oil from Omisalj. He said: “...There are two points that still need to be closed in parallel as we continue our cooperation under 2020 contract – quality discount and Janaf claims. ... we remain fully committed to this partnership and we trust that the signed Agreement is the best proof of this commitment. Once we receive from you the fully documented claims we will work without delay to review them and to negotiate in good faith the final settlement value”.

On 30 March 2020, NIS sent Glencore a formal letter, attaching Janaf’s invoices in respect of storage fees. These totalled $1,900,889.14. NIS said “We trust that figure can be agreed since the charges are justified in terms of the invoices provided the times for storage are accurate and we have no suggestion the rates would be considered unreasonable rates in the market”. NIS also attached a board declaration from Janaf dated 30 December 2019, setting the “default” rate for 2020 (see paragraph 181 below).

Glencore did not respond to this message.

On 11 April 2020, NIS wrote again. It referred to a “2020 Contract”, which I understand was a new contract for the sale of crude to NIS which either had been, or was in the process of being, agreed. NIS said that it was suspending negotiations because of the COVID pandemic, which was disrupting all aspects of its business. It then referred to the Sale Contract and the Settlement Agreement, reminding Glencore that the Janaf charges had not been reimbursed, but Glencore had not challenged any of the information provided.

Mr Wawrzyniuk said that the pandemic was causing huge problems – what he called “demand destruction” – and the Glencore crude oil team was “run on a very lean structure” (i.e. him and one colleague). He said that he was swamped with urgent matters and simply did not consider this to be urgent. He described it as an “important matter for us, but it was not urgent” and said that he had to prioritise.

Glencore then responded on 14 April 2020. In relation to the claim for storage fees, it asked some questions and commented on Janaf’s charges:

“…Therefore, in order to assess the claim please provide us with the following additional documentation:

1.

Inquiries (emails/letters of request) from market participants for crude oil storage in Janaf system in Omisalj for the period between 8th Jan and 19th March 2020

2.

Average storage fees in Omisalj charged by Janaf under the existing contracts with market participants for the period between 8th Jan and 19th March 2020

3.

Inquiries (emails/letters of request) from market participants for crude oil storage in Janaf system in Sisak for the period between 8th Jan and 19th March 2020

4.

Average storage fees charged by Janaf under the existing contracts with market participants for crude oil storage in Janaf system in Sisak for the period between 8th Jan and 19th March 2020

Just so you know at the time of the issue we had approx. 240k cbm of storage capacity in Omisalj which was totally empty because it was uneconomical to store crude at the time... Also, though we can’t share with you due to confidentiality the exact storage fees we paid at the time to Janaf for our tanks I can assure you that this was noticeably lower than what Janaf is trying to charge you. We are of course happy to share that storage fees level with you with the full confidentiality assuming Janaf is ok with that...”

Mr Wawrzyniuk confirmed that he had been involved in drafting this message. It was put to him that Glencore knew that NIS would not have any of the requested information and knew that there was no way that Janaf would hand it over to NIS. He did not agree with this, although he did accept that, in general, information about storage rates was kept confidential. He suggested that inquiries from market participants would not be confidential, and that giving a range of average fees was not confidential. He said that he had envisaged this all being discussed together between Janaf, NIS and Glencore.

Although Mr Gvozdev said that he had feared at the time that Glencore was just trying to give excuses for non-payment, it seems NIS did in fact pass on these requests to Janaf on 15 April 2020, which might be said to suggest that NIS did not consider doing so to be entirely pointless. However, Janaf immediately replied that the information was “JANAF’s commercial secret”. Moreover, it said that the amount of the reimbursement owed by NIS “is undisputed and set by item 4 of the Minutes of the Meeting held in Belgrade on February 18, 2020, signed by authorized representatives of both NIS and JANAF”. Ms Plese agreed that Janaf’s attitude was that NIS had agreed to pay at the meeting on 18 February 2020 (see paragraph 52 above) and there was nothing to talk about in this regard.

On 16 April 2020, NIS informed Glencore that it had approached Janaf, and received the reply that “such data represents business secret of JANAF”. It contended that “the amounts claimed by JANAF from NIS are all duly documented with internal local acts of JANAF, as well as invoices which should be duly paid”.

On 17 April 2020, Mr Wawrzyniuk indicated that Glencore was willing to provide new security to replace the original Performance Bond and suggested all other matters could be discussed on a conference call after the Orthodox Easter Holidays. It looks as if there was a conference call, or remote meeting, on 27 April 2020, at which the parties made clear to each other their respective positions, but did not move any closer together. There was a dispute about whether Mr Wawrzyniuk told NIS what Glencore was paying Janaf for storage, but I cannot see why it matters whether he ever gave a precise figure.

On 27 April 2020, NIS implied that it was working on the open issues and “discussing this with Janaf intensively”. It asked for a partial payment under clause 35b of the Settlement Agreement. Mr Wawrzyniuk replied that, if a fully documented claim for the Omisalj storage was received and agreed, it could be paid, even if Glencore had not received a fully documented claim for the Sisak storage. I do not think that is quite what NIS had had in mind. It was put to Mr Wawrzyniuk that Glencore already had a fully documented claim. He made clear that he did not accept that invoices from Janaf, without explanation of how the numbers were derived, were enough. He pointed out that this had not been enough when the parties had been negotiating the Settlement Agreement.

On 7 May 2020, NIS wrote again, providing some examples of Janaf charging its default rate for storage (discussed further below). It said: “From NIS point of view we do not understand there to be any dispute that the storage was required on both an ad hoc and an emergency basis and JANAF were put to a lot of trouble to accommodate the additional volumes”.

Glencore did not respond to this. Mr Wawrzyniuk said that Glencore was dealing with a number of issues and it must have slipped through the net.

On 19 May 2020, NIS chased for a response. It chased again on 28 May 2020, complaining that: “The Agreement requires you to engage with NIS in Good faith to reach agreement and make payment. You have not challenged the fact that JANAF has incurred actual loss and you have asked for and been provided with evidence of the prevailing market rates for the required storage which you have not challenged. You have not responded to our recent correspondence”.

Later on 28 May 2020, Glencore responded in an email which also referred to some complaints which Glencore had; specifically about the quality discount which had been applied to the Transit Portion and about NIS’s statement that it would not be going ahead with the 2020 Contract. In relation to the Janaf storage fees, Glencore asserted that “We have sought to engage with you, with a view to determining both the “actual loss suffered by Janaf” and “the prevailing market rates for storage during the period when the Cargo was stored in the Janaf system””. It referred to the documentation which had been requested on 14 April 2020 and said that: “To date, you have not provided any of the requested information. Instead, you have provided documentation which you say, “demonstrates that JANAF consistently charged at the rate of 1.50 $/m3 for storage on an ad hoc basis”. However, this does not demonstrate either the “actual loss suffered by Janaf” or “the prevailing market rates for storage during the period when the Cargo was stored in the Janaf system”. It is, therefore, irrelevant”.

Mr Wawrzyniuk insisted that Glencore remained willing to pay at a market rate and was ready to have a conversation with Janaf about this. It was suggested to him that Glencore was seeking to drag things out and waiting for the Performance Bond to expire. He denied this and insisted that Glencore was ready to open a new bond so that NIS remained secured for its claim.

The same day, NIS made the claim on the Performance Bond in the sum of $2,094,000, which demand was the jumping off point for the current dispute. There were some further exchanges between the parties, but neither side suggested that they were relevant to the issue as to whether Glencore breached the obligation to negotiate in good faith. Once a demand on the Performance Bond had been made and met, there was limited scope for any further negotiation.

Having regard to those facts, I must decide whether Glencore failed to negotiate in good faith. The issue is inevitably somewhat impressionistic. I have to put myself in the position of a reasonable and honest person and ask whether Glencore’s conduct was, from that perspective, commercially unacceptable. I am not persuaded that it was.

There seem to me to be two themes to NIS’s complaints. The first is that Glencore was dilatory in its responses to NIS’s letters. That is true. But just being slow to respond is not commercially unacceptable.

I accept Mr Wawrzyniuk’s evidence about the extent of the impact of the pandemic on traders such as Glencore. In March and April 2020, lockdowns and the move to home working was causing delays for many businesses. But it went further than that for oil traders. As the market for oil products collapsed, their workload must have exploded. Mr Tozzi KC invited me to speculate that Glencore was actually making huge sums of money as a result of market volatility. I have no way of knowing whether that is right or wrong, but it is obvious that, whether Glencore saw the pandemic as a crisis or as an opportunity, its traders and lawyers would have been very busy dealing with the consequences.

Mr Tozzi KC said that it was not acceptable for Glencore to prioritise its own business. Again, whether or not this is a fair comment seems to me a question of degree. I accept that there must come a point where merely leaving the negotiation at the bottom of Glencore’s list of priorities would, of itself, amount to a failure to conduct that negotiation in good faith. But it would take a more extreme case in order for that point to be reached and I see nothing extreme about what happened here. Glencore was slow responding to some letters during the pandemic. It was not alone.

What NIS really wanted me to conclude was that Glencore was deliberately slowing down these communications with an ulterior motive: namely to run down the clock and enable the Performance Bond to expire. But there was absolutely no evidence of that; no hint in any communication that that was what was “really” going on. Mr Wawrzyniuk denied the allegation in the clearest terms. Mr Tozzi KC said that the Court should not be naïve, but it does not seem to me naïve to ask for some evidence before making a finding of that kind.

The second theme to NIS’s complaints was that, when it did respond, Glencore asked NIS to provide documents which it knew NIS could not obtain from Janaf. It pointed, in particular, to the request made on 14 April 2020. NIS called these “unrealistic demands”.

I am not sure that it is right to say that it was obvious that NIS could not get any of these documents from Janaf. I accept that Janaf did not in fact provide them, describing them as “business secret of JANAF”. But, unbeknownst to Glencore, NIS had already agreed to pay Janaf using the “default” rates. That being so, Janaf had no incentive to provide any documents to NIS. If Janaf had wanted or needed to provide some or all of them, it does not seem to me that they would necessarily be prevented from doing so by confidentiality obligations owed to third parties. It is not obvious that inquiries from market participants would be covered by any confidentiality undertaking, nor that Janaf would be prevented from giving average storage rates.

NIS say, of course, that Glencore’s requests were misconceived: that there was no need for it to prove that Janaf was turning away other customers. For the reasons explained in section D.5 above, I agree. But it does not follow that Glencore could not, in good faith, take the position that the reference to “actual loss” required NIS to show that Janaf had lost business in order to be able to recover any storage fees at all. Mr Hill KC advanced exactly that position, with his usual vigour and skill, before me. The fact that NIS took, or I take, a different view of the meaning of the provision does not mean that Glencore was not negotiating in good faith.

NIS also made some other complaints, such as about the way that, in its letter of 28 May 2020, Glencore referred to other disputes between the parties and sought to tie them together with the negotiation about storage fees. This does not seem to me to add very much. Again, I can see that, if Glencore had consistently refused to engage with storage fees unless NIS made a concession on the new contract, there might come a point at which this amounted to a failure to negotiate about those storage fees in good faith. But it never got anywhere near that stage.

I reject NIS’s allegation that Glencore failed to negotiate in good faith.

As I said at the beginning of this section, I struggle to see how NIS could, on the facts of the present case, ever have been entitled to damages exceeding its legal entitlement under the Settlement Agreement. In my judgment, therefore, Mr Hill KC was right to say that the question of breach of good faith ended up being irrelevant, however it was decided.

There was no evidence that Glencore might have been willing in a negotiation to pay more than NIS’s legal entitlement. Indeed, the whole thrust of NIS’s complaint about the negotiation was that Glencore was delaying, making impossible requests, and generally being obstructive. As ever with quantifying damages for breach, if I had accepted NIS’s case on liability, I would have had to use a counterfactual in which there was no breach: i.e. where Glencore was (just) on the right side of that line. I struggle to see how doing so would result in a conclusion that Glencore would have been throwing money at NIS.

I suppose it might be said that, in a negotiation, anything could happen: e.g. Glencore might be persuaded to pay more in order to dispose of the problem and get back to trading. But the problem with that approach is that, absent specific evidence, it becomes entirely unstructured, or so speculative as to be unworkable. I would also have to factor in the possibility that NIS would take less than its entitlement, and the possibility that there would be no deal at all, and the issue would come before the Court. Indeed, if one sees this as akin to a loss of a chance, perhaps it would be logical to say that the chance of a deal whereby Glencore pays more than is due, broadly matches the chance of a deal in which Glencore pays less than is due, with the remainder of the outcomes (including no deal, followed by Court proceedings) resulting in Glencore paying exactly what is due.

Absent some evidential basis for applying an uplift (or a discount), it seems to me that the only safe assumption I could make would be that, if there was a deal, it would involve Glencore paying what was actually due. Mr Tozzi KC suggested that the problem with that approach was that it effectively deprived the obligation to negotiate in good faith of any value. I do not agree. If NIS was able to show that a deal would or might have been done if the parties had negotiated in good faith, for example, it might have a claim for the legal costs wasted litigating to the same outcome. No doubt there are any number of other scenarios in which a loss might result from a failure to negotiate in good faith. My conclusion is that, in the present case, there is no evidence of any loss, over and above NIS’s legal entitlement to reimbursement, and additional loss cannot simply be assumed. I am not suggesting that it is conceptually impossible for such a loss to be suffered.

I turn, then, to the amount of NIS’s legal entitlement to reimbursement.

I should start by observing that, save for its submission that the use of its tanks amounts to an “actual loss”, NIS has not advanced any case that Janaf suffered any “actual loss”. I have already explained (at paragraphs 98-112 above) why that does not have the result for which Glencore contended (i.e. that NIS is reimbursed nothing because Janaf suffered no “actual loss”). But it does mean that, when the current stage in the argument is reached, NIS is left seeking to justify the Janaf storage fees purely on the basis that they reflected the prevailing market rates.

NIS did not shrink from this contention. It said, in summary, that the Janaf “default” rates were the market rates for storage of this kind at Omisalj and Sisak and hence that the prevailing market rate was $1.5/ cbm / decade.

For its part, Glencore relied primarily upon the rates it had negotiated with Janaf in 2020. Its primary case was that the best evidence of the prevailing market rate was what it agreed with Janaf in Annex 1: $1.5 / cbm / month.

I was shown that, in 2014, NIS and Janaf had had a contract (identified as P-01/14) for storage of crude oil at the Omišalj Terminal. Article 9 provided for a storage price of $1.50 / cbm / decade for 70,000 m³. Glencore made a number of points about this and about what was included in this price (i.e. Mr Daly suggested it included some transportation, although Mr Waguespack disagreed), but probably the most important was that this had been agreed in 2014 and hence was very distant from the period with which we are concerned.

Ms Plese explained that the transportation contracts entered into with NIS did not envisage any storage as such. NIS needed storage only as a transit point for onward transportation to its refinery. The storage required for that purpose was included in the transportation rate. However: “If storage beyond the expected minimum is required there is a specific rate for what is called ad hoc storage”. This “ad hoc” storage might happen if refinery tanks were full or for some other logistical reason. Mr Plese said that “JANAF does not like holding storage because it blocks their system and they make most money from their transport charges which in turn needs to maximise volume of throughput”.

Ms Plese also explained that the “default” rate is set annually and has not changed for years. She made clear that “It is not open to negotiation because JANAF do not want ad hoc storage as part of their business”.

Consistently with this, Janaf provided to NIS (and NIS provided to Glencore) two examples of third parties being required to pay the “default” rate:

I have also seen a record of a decision by the management board of Janaf that what it called the “ad hoc” price for crude oil storage for 1 January 2020 – 31 December 2020 would be $1.5/ cbm/ decade. I am not sure that adds very much to the discussion.

In the first example (i.e. that from March 2016), it was not explained why the oil had needed to be stored. I infer that there was a short-term problem which caused the permitted 20-day period to be exceeded, probably in the context of a transportation contract. The second was part of a proposal for a transportation contract.

Mr Waguespack described this rate as “in effect, a default rate used for unplanned circumstances”. Mr Daly said that it “acts as a short circuit safety valve for unforeseen temporary hold ups in the normal operation”. I agree with both descriptions. I would equate the rate with a “penalty” charge in a car park: the price of outstaying the 20 day period which is priced into Janaf’s rate for transporting the oil.

In January 2017, Janaf offered Glencore either 160,500m³ or 240,500m³ of storage for 3 years at a price of either $3/ cbm / 30 days or $2.55/ cbm / 30 days, depending on which option for volume was selected. There were also some other costs for discharging operations and quality and quantity control. This offer seems to have resulted in a contract (“the 2017 Contract”) using the same volumes, but with improved rates of $2.90 / cbm / 30 days or $2.45/ cbm / 30 days.

On 5 December 2019, Janaf made a proposal for the renewal of Glencore’s storage facilities. The volume offered was 239.856 m³ and Janaf’s offered rate was $2 / cbm / month. On 13 February 2020, the parties exchanged emails about an interim solution so as to enable the agreement that would otherwise expire that day to continue. Glencore was also asking Janaf to repeat the proposal for 1 year at $2 / cbm / month and keep it open for acceptance until the end of March 2020. Janaf did so.

There was then a document dated 13 March 2020 and referred to as “Annex 1” to the 2017 Contract which embodies the interim solution, extending the 2017 Contract from 14 February 2020 until 31 March 2020. Annex 1 included a specific limit for organic chloride content (max 1ppm in crude oil). In relation to price, Article 18 of the 2017 Contract was amended to provide for a price of $1.50 / cbm / month, with an uplift for high BS&W (basic sediment and water) content.

NIS complained that it and its expert had believed until shortly before the trial that Annex 1 had been signed on 13 March 2020, because that was the date it bore. In fact, late disclosure revealed that its terms had actually been agreed on 13 February 2020 and Annex 1 was signed shortly thereafter.

Mr Wawrzyniuk was asked about one of the exchanges in this context and gave a long (and not entirely responsive) answer, which included the following explanation of how the rate of $1.50 / cbm / month had been arrived at:

Mr Tozzi KC was critical of this evidence, which he said (correctly) was not foreshadowed in Mr Wawrzyniuk’s witness statement. He described it as a prepared speech. I am not sure that is quite fair. Mr Wawrzyniuk had plainly been aware of the issues about the further disclosure which had been made in relation to Annex 1 and seemed to have gone back to his old emails. I certainly did not get the impression that he was making up this narrative about exchanges with Janaf leading to Annex 1 and, as I will explain, I am not sure that it is all helpful to Glencore’s case.

I accept, however, that Glencore’s decision not to disclose the relevant emails, while not involving any breach of any Court order (since disclosure was on a model B basis), made it very difficult for NIS to verify what was being said. It is also a little unsatisfactory that Glencore had not adduced this evidence in chief, if it says the negotiation of Annex A is relevant. I therefore treat this evidence with some caution.

In any event, it is clear from Mr Wawrzyniuk’s evidence above that the rate of $1.5 /cbm/ month was only being put in place for a bridging period, as a compromise figure, in circumstances where Glencore was not actually using these tanks, and it seemed unlikely that it would need to do so before 31 March 2020. Mr Wawrzyniuk agreed that this was “essentially just a holding position”. That seems a fair description to me.

On 30 March 2020, Glencore and Janaf entered into a new storage contract for the period from 1 April 2020 until 31 March 2021. This was still for a capacity of 239.856 m³, at a price of $2 / cbm / month. There was an uplift of $1 / cbm for an elevated BS&W content, but no other charges for discharge etc.

Glencore pointed out that this was the same rate which Janaf had offered on 5 December 2019, and then agreed to keep open in February 2020.

On 5 May 2020, Janaf put forward some further offers, for the period from 1 April 2021. These offers suggested that the rate per CBM would be higher for a shorter period (e.g. $2.80 / cbm/ month for 3 months vs $2.40/ cbm/ month for 1 year) and higher for a reduced volume (e.g. $2.80/ cbm/ month for 239,856m³ vs $2.85/ cbm / month for 123,000m³). It was agreed that the market for storage was very different by May 2020 (as the pandemic created a worldwide glut of crude oil), so these figures were mostly relied upon (by NIS) for the purposes of drawing those comparisons in relation to periods and volumes.

Much of the expert evidence consisted of (perfectly legitimate) comment on that material about the rates actually charged by Janaf to NIS and Glencore and others. There was only a small quantity of further information which each expert was able to provide about storage rates. They agreed that storage contracts are often kept secret, which was why there is so little publicly available information about this “market”.

Indeed, the experts agreed on a number of important points.

It was agreed that crude oil storage was generally priced on the capacity of the tankage, not the amount of oil stored. Indeed, it is normally provided on a monthly “take or pay” basis (i.e. you pay for the whole capacity made available for the whole month, regardless of whether all of that capacity was used, or whether it was used for the whole month). It was agreed that 10-day periods are not common. Mr Waguespack commented on the potential attraction for a very short-term user of a 10-day period (as used in Janaf’s default rate), but I understood him to accept that he had never encountered a 10-day rate himself.

It was agreed that the structure of the oil market price drives demand for discretionary crude oil storage for trading and investment opportunities. In very simple terms, when the crude oil market was in “contango”, such that the price for crude oil for prompt delivery was lower than the price for delivery in one or more months’ time, there might be scope for a trader to buy, store, and sell at a profit which exceeded the cost of the storage. As such, demand for crude oil storage is likely to be stronger when the market is in contango and weaker when it is in “backwardation” (i.e. when the prompt price exceeds the future price).

It was agreed that the market was mostly in backwardation in the first quarter of 2020, swinging round in mid-March 2020 – around 9 or 12 March 2020 – as a result of the pandemic (as demand reduced and refineries were unable to receive cargoes, resulting in a glut of oil).

There was a mild disagreement between the experts as to the extent to which the market structure would affect demand for storage at Omisalj. Mr Waguespack suggested that the effect would be limited, because providing storage for such “contango plays” by traders was only a very small part of Janaf’s business. Perhaps importantly, however, he suggested that the market structure would have little or no effect on prices for medium and longer term storage at Omisalj, and more effect on “spot” (more immediate, short term) storage arrangements. He agreed that, during periods of contango, there is an elevated demand for storage and rates increase, and that a market which was in backwardation would put a dampening pressure on rates:

In the same way, Mr Daly suggested that rates for storage were low in early 2020 because of the market structure (i.e. the market being in backwardation) and thought this explained the low rate agreed between Janaf and Glencore in Annex 1.

Mr Waguespack suggested that a short-term “spot” agreement for storage might be expected to command a premium of 15-25% as against rates for longer term storage arrangement (i.e. 1 year or more), at least when storage availability was limited. At first sight, this suggestion did not fit with his evidence about the different effect of market structure on spot agreements as compared with longer term storage (and indeed he acknowledged that spot rates might be higher or lower than longer-term rates).

In the end, however, I understood him only to be saying that, all other things being equal, a terminal would prefer to have a longer term commitment to a shorter term commitment and would be willing to incentivise the former with a more favourable rate. I accept that evidence. Mr Daly agreed that “When the market is in contango, and storage space is in high demand, I can see that the rates for ad hoc storage will climb and are likely to climb above longterm storage rates. It is simply the application of the principle of supply and demand”.

In a similar way, Mr Daly acknowledged that, if a party was willing to commit to a larger volume, it might often be able to negotiate a lower rate. He described this as “self-evident”: “if you are going to get more cash flow, you can actually discount the rate”.

In terms of other information about rates, Mr Daly referred to “current” contracts ranging from $0.80 / cbm / month up to $2.6/ cbm / month. However, the only detail he was able to offer was that the latter was a 10-year contract, which he acknowledged was “not really representative of the spot rates in December 2019”. I agree that this is of limited relevance. I do not accept his suggestion that it can be treated as a “cap” on market rates. If a ten-year rate was $2.6/ cbm/ month, there is no reason why the market rate for a shorter period could not be higher than that.

Mr Waguespack suggested (on the basis of information from confidential sources) that the rate for storage in NW Europe in 2019 might be $3.0 – 3.2 / cbm / month and the rate for Fujairah might be $2.50-2.75 / cbm / month. He suggested that the rate for the Mediterranean might be the same as the Fujairah rate, although he did not seem very confident about the analogy when pressed in cross-examination. He made clear that he had limited knowledge about operations at either Fujairah or the Mediterranean and that he was relying on the judgement of an independent consultant. I understood that to be a reference to Mr van Stralen, who Mr Daly had previously told me was not really an expert on storage in the Mediterranean (as opposed to storage in Northern Europe). I did not feel able to place much weight on this, at least as evidence about the rate payable in the Mediterranean.

Mr Waguespack concluded that Janaf’s default rate was the market rate for what he called “unplanned, uncontracted, and short-term storage”. He took the view that it was the requirement for “short -term, emergency, or uncontracted storage” which was the dominant characteristic.

He explained that terminals are impacted by unplanned events. He gave an example of how it might have an actual impact if tanks are unexpectedly tied up: “there are usually a minimum number of “working” tanks required to ensure that it can continually receive and transport crude oil. Unplanned or emergency storage can tie-up the tanks used by the operator (and not contracted out for third-party storage) and potentially delay the receipt of crude oil”. He suggested that a terminal would want to charge for the additional cost and inconvenience of providing storage on that basis. Viewed in that light, the only rate which could be said to fit that description was Janaf’s default rate, which was the rate charged for “emergency storage”. Indeed, he said that he was unaware of any other rate for “emergency crude oil storage”.

For his part, Mr Daly suggested that the Annex 1 rate agreed between Glencore and Janaf for a short period into March 2020 was the best guide to the spot price at that time.

Mr Waguespack identified a number of reasons why, in his view, the Annex 1 rate was not a suitable benchmark. He pointed out that it was very short-term, bridging a period in which the tanks in question were actually empty. It was for a greater storage capacity than was needed for the Balance Cargo: 240,000m³ vs 157,000m³. It seems to me that these are valid points.

I have already explained how I understand the interaction between “(I) the actual loss suffered by Janaf and (ii) prevailing market rates for storage” in clauses 34 and 35(b) of the Settlement Agreement (see paragraphs 98-112 above). It seems to me that my conclusions in this regard mean that the Janaf “default” rate does not reveal the “prevailing market rates for storage”. If it is a market rate for storage at all, it is not the type of market rate which was envisaged by the Settlement Agreement. It was an emergency rate, fixed in the context of the transportation of oil, mostly as a disincentive to delay. Treating this as a market rate for storage is like treating the penalty charge imposed by a car park as the daily rate for parking there. In one sense, that penalty charge might be said to be the “cost” to you of parking for a day, at least if you fail to buy a ticket. But it is certainly not the “market” rate for parking your car in that car park for a day.

An important part of the justification for the “default” rate was that Janaf might be inconvenienced by unexpected/ unplanned delays in the transportation process: by tanks which were intended to be used as part of the process of transporting oil being blocked up. But, if that had actually happened, Janaf would have suffered an actual loss, which could then have been included in addition to the market rate for the purposes of arriving at Janaf’s storage fees. As it turned out, either Janaf did not suffer any actual loss of that type, or Janaf had no incentive to provide evidence of such actual losses, perhaps because (unbeknownst to Glencore) NIS had already agreed to pay the “default” rate before the Settlement Agreement was signed.

Having put the Janaf “default” rate to one side, Glencore’s submission that the rate of $1.50 / cbm / month which it agreed in Annex 1 was the best evidence of a “spot” rate appeared at first sight to have much to commend it. The timing of this agreement (around 13 February 2020) was in the middle of the period with which we are concerned, and this was a short-term arrangement. However, in the end, I felt uncomfortable about treating that rate as if it were a spot rate which had been freely negotiated between market participants. Part of that was the fact that Glencore’s disclosure and factual witness evidence about Annex 1 was a little unsatisfactory (see paragraphs 187 - 190 above). But much more important was that the limited evidence which I did have made clear that this was not a freestanding agreement. It was a compromise born out of an existing relationship.

Annex 1 was, in effect, a bridging agreement between the end of a three-year storage contract and the envisaged commencement of a further one-year contract, in relation to the same tanks. Glencore was not actually using the tanks and may well have been aware that it was unlikely to need them in the short term. In some ways, the simplistic nature of the negotiation was emphasised by Mr Wawrzyniuk’s “speech” (as NIS described it) about how the price of $1.50/ cbm/ month was arrived at. Janaf had been proposing a figure of $2 and Glencore was (apparently) at $1. So they split the difference.

I find it very hard to equate an arrangement like this, where parties in an existing and long-term relationship find a way to avoid arguing about what to do with tanks which Glencore does not actually need in the short term, with a spot storage contract, in which a party wanting to store some oil for a short period approaches Janaf and a market price is agreed. This is more akin to friends agreeing to split the bill for dinner, rather than an arm’s length negotiation about the price for any particular individual’s chosen meal.

By comparison, the Glencore contract dated 30 March 2020 has the advantage of providing a genuinely negotiated rate for storage at Omisalj ($2 / cbm / month). The date reveals that it was finally agreed after the period with which we are concerned, but that rate had been offered by Janaf back in December 2019, and Glencore had asked in February 2020 for that offer to be repeated and kept open. This might be said to be good evidence that the market rate for longer term (1 year) storage for a capacity (take or pay) of 239.856 m³ was around $2 / cbm / month throughout the relevant period.

There was some discussion of relative bargaining power. I was not persuaded that Glencore was in some special or different position from an “ordinary” market participant when negotiating with Janaf for storage space. It is obviously a very large and successful oil trader. But that did not give it any special hold over Janaf. No doubt Janaf wanted the business and knew that Glencore could store elsewhere. But, equally, Glencore had a contract with NIS and having storage facilities at Omisalj gave it flexibility. Mr Wawrzyniuk had explained that Glencore wanted the storage at Omisalj more for operational reasons than for contango plays.

I do not see any evidence that Glencore’s market presence enabled it to obtain a discount on the “market” price from Janaf. I can see that the fact that there was an existing relationship may have played a role in the negotiation, but find it difficult to tell whether that would have resulted in a discount (on the basis that it would be a relationship that Janaf would be keen to preserve) or a premium (on the basis that Glencore could be seen to have a continuing need). Perhaps neither.

It is right to observe that the rate offered to Glencore was for more capacity (239.856 m³ vs 157,000m³) and for a longer period (1 year vs 3 months) than our notional market “spot” rate. I accept that, all other things being equal, both of those factors might be expected to bring down the rate; that the market rate (per cbm per month) for a smaller volume and for a shorter period might be expected to be a little higher.

However, it also seems to me, on the basis of the expert evidence, that as one moves from a longer-term rate (for a year) to a “spot” rate (for 3 months), the market structure is likely to have more influence. There is no doubt that, for most of the period with which we are concerned, the crude oil futures market was in backwardation. It seems to me that would have played a role in depressing the market rate. My sense from the limited evidence is that that role would have been important in the context of a 3 month “spot” arrangement, even if not outweighing all other factors.

In a market in which there was more reported information about rates, I might feel able to calibrate the effect of these competing influences more precisely. But here I was provided with very few comparables to work with. As I have explained, I did not find Mr Waguespack’s reference to rates in Fujairah, or Mr Daly’s range of other rates which he had come across during this period, of any real assistance.

I did consider whether there was any role in this assessment for the burden of proof: whether I should be allowing for uncertainties in favour of one party or the other. When I raised the topic in the course of closing submissions, each party suggested that the other shouldered the burden (Glencore said it was NIS’s claim and NIS said Glencore was in effect relying upon a limitation clause), but neither submitted with any real enthusiasm that this was the solution to the evidential problem.

On reflection, it seems to me that they are right about that. The problem is not that either side finds itself unable to prove a necessary ingredient of its case. It is that the available evidence, while undoubtedly enabling the Court to arrive at a conclusion as to the market rate (see below), does not permit as precise an adjustment for the factors which have been identified as relevant as might be possible in the context of a different market. It seems to me that I must simply accept the limitations of the material with which I am working, and do the best I can, rather than look for a way of avoiding answering the question.

Approaching it in that way, therefore, it seems to me on balance that the factors which I would expect to move the market rate up from the starting point which I have identified of $2 / cbm / month (i.e. difference in volume and period) are likely to have been approximately cancelled out by the depressing effect on spot rates for storage of the market structure at the time.

For these reasons, I find that the prevailing market rate for storage at Omisalj, as envisaged by clause 34 and 35(b) of the Settlement Agreement, was $2 / cbm / month. I have no evidential basis for using any different figure for Sisak than for Omisalj (and neither party suggested that I should do so).

On that basis, it seems to me that I apply that rate to the total storage capacity being used for the Cargo at Omisalj and Sisak (i.e. 177,000m³) for a 3-month period, which gives a total reimbursement due from Glencore of $1,062,000.

For the reasons I have explained in section D above, I reject NIS’s primary case that it was entitled to bring a claim pursuant to or for breach of the Sale Contract in respect of Janaf’s storage fees. It had settled that claim, save only for its entitlement to be reimbursed in clauses 34 and 35b of the Settlement Agreement (which entitlement I have now quantified).

I made clear at the outset of the trial that, since Glencore had not pleaded any case to the effect that Glencore did not have to pay this claim by NIS under the Sale Contract because it had not made a recovery from its own supplier, that defence was not open to it. As it turns out, that does not matter.

For completeness, I note that Glencore continued to dispute that the Janaf storage fees fell within the words “JANAF penalises the Buyer because of that as a result” in clause 2.2 of the Sale Contract. I was not very taken with arguments about whether it could be said that the additional storage costs were unrelated to the contamination, which seemed a little unrealistic (to use Mr Tozzi KC’s word). However, I did have my doubts as to whether a non-contractual “default” charge, which NIS ends up agreeing to pay to Janaf, is what was intended by this provision, which seems to me directed to contractual “penalties” which can be identified and invoiced without any scope for dispute. However, my instinct was that, if it had mattered, NIS would have been able to show that that payment to Janaf was a consequence of Glencore’s breach in delivering off-spec cargo. It was not suggested that it was unreasonable (in the sense of involving a failure to mitigate) for NIS to agree to pay, given the commercial difficulties of its position.

Given my conclusions on the other issues, however, it is not necessary to explore these issues any further.

On the basis that NIS was actually only entitled to reimbursement of $1,062,000 in respect of Janaf’s storage fees, it must return $1,032,000 out of the $2,094,000 which was paid to it pursuant to the Performance Bond. Glencore is entitled to judgment in the sum of $1,032,000.

I will hear the parties on interest and costs and any other consequential matters.

I will conclude by expressing my gratitude to Counsel for the quality of their written and oral advocacy, to the solicitors for the care with which the case had been prepared, and the legal teams as a whole for the sensible and cooperative way in which the trial was conducted.