BP Gas Marketing Ltd v La Societe Sonatrach & Anor

INTRODUCTION AND BACKGROUND TO THE DISPUTE

This action is concerned with the proper construction of clause D2.2.2(ii)(b) of a Joint Shipper's Agreement ("JSA") entered into on 28 June 2005 between the Claimant, BP Gas Marketing Limited ("BP"), and the First Defendant La Societe Sonatrach (Footnote: ¹). Clause D2.2.2(ii)(b) of the JSA provides that nitrogen costs (which includes power used to generate nitrogen) allocated by Grain LNG Limited ("Grain") to BP and Sonatrach (together defined as the Shipper), in respect of nitrogen added to cargoes of Liquefied Natural Gas ("LNG") (Footnote: ²) imported through Grain's terminal, shall be allocated between BP and Sonatrach according to a formula as there set out.

It is the proper construction of that formula, and one element within it, namely a figure of 51.41, that gives rise to the issues that arise for determination in this action. In order to understand the issues that arise (which are defined in section B below) it is first necessary to set out something of the background to the dispute.

The Isle Of Grain, which is situated on the Medway Estuary, is home to a gas terminal (the "Terminal") operated by Grain, a National Grid company. The Terminal was, until 2002, a peak shaving terminal, used to supply gas to the national pipeline system during peak use periods, and storing gas as LNG when not required. It was not an import terminal and did not have facilities for berthing or unloading of vessels. The gas which the Terminal stored was North Sea gas which came to it through Transco's National Transmission System (the "NTS").

In 2002, Grain started converting the existing plant into a Terminal for discharging LNG from ocean-going tankers, storing it in tanks on-site, and then converting it back into gas to be sent it out into pipelines for distribution to domestic and commercial consumers. The Terminal came into operation in 2005. At that time, it was the only import and regasification terminal in the UK. There was accordingly no track record as to how the Terminal would be operated (including, for present purposes, in relation to the Terminal's practices concerning the addition of nitrogen to LNG to alter its composition), which provides the backdrop to the dispute between BP and Sonatrach, as to how costs charged by Grain in relation to nitrogen are to be allocated between BP and Sonatrach.

In October 2003, BP and Sonatrach jointly contracted for the use of the Terminal's initial discharge, storage, re-gasification and send-out capacity under a 20-year Specific Terms Agreement ("STA"). "STA" will be used in the judgment to include subsequent amendments and additions thereto (the STA is also called the "Services Agreement" in some documents). BP and Sonatrach had to tender (and thereafter contract) jointly because the Phase 1 capacity was not sufficient to split between the two shippers - the capacity was around 192,000m³ and the LNG tankers available at the time had a capacity of around 155,000m³ with the result that a right to half the capacity (100,000m³) would not have allowed either shipper to unload a full cargo. In their relations with Grain, BP and Sonatrach are accordingly treated as a single shipper.

Grain later twice expanded the Terminal. The original capacity, which is covered by the STA, is known as Phase 1. The additional capacity is divided into Phases 2 and 3. Sonatrach contracted for some Phase 2 capacity. BP is not involved in Phases 2 or 3.

Although BP and Sonatrach contracted jointly with Grain, they act severally in sourcing LNG and import from different countries. Like oil, LNG is a blend of hydrocarbons. The blend differs from source to source (in terms of the precise amounts of methane, propane, nitrogen and other gases it contains). So LNGs from different sources have different properties.

The LNG property which is most relevant to this case is the Wobbe number (or index). In simplistic terms, Wobbe is a measure of an LNG's energy by volume. It is expressed in megajoules per standard cubic metre (MJ/m³). More precisely, Wobbe is an LNG's higher heating value divided by the square root of the LNG's specific gravity, the specific gravity being the ratio of the density of the LNG to the density of air: Wobbe = HHV/√SG.

The UK gas transportation infrastructure is built on two separate, but connected, gas transmission (pipeline) systems, namely Transco's National Transmission System (the NTS), and Local Distribution Zones ("LDZ"). The NTS is a high-pressure system for transmitting large volumes of gas over long distances to large users, e.g. power stations. The LDZ are lower-pressure systems for transmitting gas to users in the vicinity of a plant. The Terminal is connected to both systems.

There are a number of parameters which apply to gas entering the NTS and the LDZ. These are largely set out in the Gas Safety (Management) Regulations 1996 (the "GSMR”). These include that the Wobbe index value of the gas cannot be lower than 47.2MJ/m3 or higher than 51.41MJ/m3. The UK infrastructure was developed around North Sea gas which is at the very lean end of the spectrum of gas produced worldwide.

Under the STA, BP and Sonatrach contracted that LNG would comply with the regulatory limits at the point of discharge into the Terminal, and Grain contracted that gas (i.e., regasified LNG) would comply at the point of send-out from the Terminal. In this way, the 51.41 Wobbe limit was written into the STA.

The upper value of 51.41MJ/m³ is very low by international standards, and was potentially problematic for BP and Sonatrach as most international sources produce LNG with a Wobbe above 51.41. However, there is a solution, namely to add nitrogen to the gas. Adding nitrogen increases the density of the gas but lowers its energy value since nitrogen has no heating value, and in consequence the Wobbe of the LNG/nitrogen blend is lower than that of the original LNG.

By a First Letter Agreement ("FLA") concluded in November 2004, Grain agreed to arrange nitrogen facilities at the Terminal to ensure that LNG discharged under the STA met the 51.41 limit on send-out. BP and Sonatrach agreed to bear the costs. This arrangement meant that the Wobbe limit at point of discharge into the Terminal could be relaxed.

Following the Award of the Phase 1 capacity in October 2003 BP and Sonatrach discussed the necessary arrangements between themselves. BP and Sonatrach originally contemplated creating a joint venture entity for their joint use of Phase1, but ultimately in June 2005, when the Terminal was about to become operational, BP and Sonatrach concluded a Joint Shipper's Agreement (the JSA), under which this dispute arises.

The JSA regulates the use of the Phase 1 capacity as between BP and Sonatrach (as opposed to between BP/Sonatrach and Grain, which is regulated by the STA). Amongst other matters, it regulates how costs and charges which Grain passes on under the STA are allocated between BP and Sonatrach. One such cost is nitrogen (including power used to generate nitrogen).

The JSA deals with the monthly allocation of nitrogen costs at D2.2.2(ii). The JSA also provides for annual reconciliations of nitrogen costs by the same basic method, but using annual values instead of monthly values. D.2.2.2(ii)(a) allocates fixed nitrogen costs 50/50. The dispute that has arisen between BP and Sonatrach concerns the proper construction of clause D2.2.2(ii)(b) which calculates a party's allocation of variable nitrogen costs according to a formula.

Clause D2.2.2(ii) is in the following terms:

“Nitrogen Costs shall be allocated as follows, by reference to all relevant terms as they are defined in, and by measuring all relevant units in accordance with, the Services Agreement and the document entitled "GLNG - Agreed Network Entry Provisions" dated 20 June 2003 and initialled on behalf of GLNG and the Shipper:

(a)

the fixed component of nitrogen costs allocated by Grain to the Shipper shall be borne equally by the Co-Shippers; and

(b)

subject to Section D2.2.2(ii)(c) the variable component of nitrogen costs allocated by Grain to the Shipper shall be allocated in accordance with the quantity of gas Sent Out and the quality of the LNG delivered by each Co-Shipper (the delivering Co-Shipper) as follows:

X_MA = TC_M x TN_2A

TN_2A + TN_2B

Where:

X_MA is the cost of nitrogen to be borne by a delivering Co-Shipper with respect to a month (M);

TC_M is the total cost of nitrogen with respect to the month (M)

TN_2A is the quantity of nitrogen used by the delivering Co-Shipper in the month (M)

TN_2B is the quantity of nitrogen used by the other Co-Shipper in the month (M)

where each of TN_2A and TN_2B is calculated as the product of:

(i)

gas Sent out in Gwh in the month by the delivering Co-Shipper (in the case of TN_2A) or by the other Co-Shipper (in the case of TN_2B); and

(ii)

The greater of X or Y, where:

X is the correction factor for incomplete combustion factor (tonnes of nitrogen N2 per GWh) calculated as follows:

((weighted average higher heating value of LNG as delivered by the delivering Co-Shipper (in the case of TN_2A) or the other Co-Shipper (in the case of TN_2B) in the month (M) x 6.35) – 250.01)

5.83

Y is the correction factor for Wobbe (tonnes of nitrogen N₂ per GWh) calculated as follows:

((weighted average higher heating value of LNG as delivered by the delivering Co-Shipper (in the case of TN_2A) or the other Co-Shipper (in the case of TN_2B) in the month (M) / (LNG molar weight) ^ 0.5) – 51.41)

0.62381

D2.2.2(ii)(b) accordingly provides that the variable component of nitrogen costs allocated by Grain to the Shipper is to be allocated in accordance with (1) the quantity of the gas sent out and (2) the quality of LNG delivered by each of Co-Shipper applying the formula that then follows. Although this formula appears somewhat complex at first sight, once all the variables have been identified, the formula can be applied with the figures contained in the formula, to produce a calculated result.

Part of the formula multiplies the quantity of gas which the party sends out in the month by the greater of Factor X or Factor Y. Factor X and Factor Y are themselves determined by formulae. Factor Y involves (in simplified terms) calculating the difference between the Wobbe of the LNG which the party discharges in the month and a figure of "51.41". This figure of 51.41 is at the heart of the issue of construction.

BP say the figure of "51.41", like other figures in the formula is, as a numerical value, a constant, that constant of 51.41 (Mj/m3) being the maximum Wobbe index value of gas entering the NTS and the LDZ as specified in the GSMR. In contrast Sonatrach submits that 51.41 is not a constant at all, but instead is to be understood as a variable, namely "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time, alternatively was a proxy for the same" ((Re-Amended Defence and Counterclaim para 38A).

In practice, if both parties discharge LNG in a given month, one party's LNG will have a lower Wobbe than the other's. Broadly, BP's LNG has had a lower Wobbe than Sonatrach's. Factor Y allocates a lower proportion of nitrogen costs to the party whose Wobbe is lower. This is because the arithmetical effect of Factor Y is that a party's allocation becomes smaller the closer that party's Wobbe gets – relative to the other party's Wobbe – to 51.41.

BP's claim is for nitrogen costs which it says should be borne by Sonatrach under D2.2.2(ii)(b), based on the respective Wobbes of the parties' cargo relative to 51.41. BP and Sonatrach have paid between them 100% of the nitrogen costs invoiced by Grain. So the claim relates to the allocation, as between BP and Sonatrach under the JSA, of costs already incurred and paid under the STA. The claim is for £3,362,216.16 principal, plus contractual interest at LIBOR+3%. Sonatrach denies that it is liable in such sums, or that the invoices issued by BP are valid and enforceable.

The Terminal was not operational when the formula in D2.2.2(ii)(b) was contractually agreed. The formula had been drafted by Mr Winstanley of BP (one of the witnesses called by BP at the trial). It appears that there was little or no discussion between BP and Sonatrach about the content of the formula before it was agreed. Equally it appears that Mr Winstanley did little in the way of modelling using the formula.

The initial conversion work at the Terminal was completed in 2005 and the Terminal was commissioned in July of that year. In April 2006 the Agent (jointly appointed by the Co-Shippers) had approached BP about "strange results" with the nitrogen cost allocation spreadsheet that had been prepared for the Co-Shippers. Mr Winstanley identified that he had omitted the 28.9 constant (the "28.9 constant") for the molecular weight of air in the formula (which involved a calculation of specific gravity relative to air in Factor Y). I address this error in due course, but it was essentially a typographical error - Mr Winstanley had simply omitted it from his formula. In due course, and as will appear, the parties agreed to add the molar weight of air into the formula.

It was also identified in May 2006 that the formula sometimes produced allocations in excess of 100%, and Mr Wood as Agent, referred to this in a letter to the parties on 18 May 2006 (“Cost allocation calculation: Co-Shipper A: 125%, co-shipper B -25%”). It was not immediately clear whether this was intentional given the apparent possibility that a lean cargo might dilute a richer cargo’s nitrogen requirements, generating a credit. However, it transpires that this would only be so if the lean cargo’s Wobbe was below Grain’s operational Wobbe (and it appears that that has not happened in practice). In due course, and as will appear, the parties agreed to a capping of allocations to 100% of costs.

Once the Terminal became operational it became apparent that the level of nitrogen being used by Grain was significantly above the parties’ expectations. By December 2005 BP was already in touch with Grain to investigate the use of nitrogen. In due course Grain in its March 2006 Monthly Report indicated that Grain targeted an "optimum" Wobbe of 51.1 and often injected more nitrogen than the "optimum" quantity.

BP's case is that "51.41" is plain and unambiguous, and means "51.41", which is a constant. Factor Y says what it means, and means what it says. BP submits that the purpose of the clause is to produce an allocation of variable nitrogen costs in accordance with the quantity of gas sent out and the quality of LNG delivered, and that it does so by reference to the contractually agreed formula, which is to be applied in accordance with its terms.

Sonatrach submits that if "51.41" is treated as a constant, the formula produces what it says are arbitrary and absurd outcomes that it says no reasonable third party would have understood in June 2005 to be the effect of the formula in D2.2.2(ii)(b). Sonatrach submits that the purpose of clause is to allocate variable nitrogen costs based upon the amount of nitrogen used in relation to the parties' respective cargoes, judged by reference to their Wobbe values at unloading, and that it will only do this if "51.41" is construed not as a constant but as meaning (as has already been quoted above), "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" alternatively is "a proxy for the same."

Following amendments to its then Defence in January and June 2016, Sonatrach in its Amended Defence and Counterclaim (the "Defence"), advances its case under five limbs:-

BP's riposte to these arguments is, in summary, as follows. "51.41" is a constant, and cannot possibly mean "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" nor is it capable of being construed as a proxy for such. The JSA is a sophisticated legal document running to 180 pages and not notable for concision. If the parties had intended to refer to an actual operational Wobbe rather than "51.41" as a constant, then they would have said so, especially as they use both averages and constants in the formula. The alleged implied terms are contrary to the express term of "51.41", are not necessary, reasonable or go without saying, and the second alleged implied term, in particular, is also vague and uncertain. Whilst the parties identified and discussed a number of issues with the D.2.2.2(ii)(b) formula, (and agreed to resolve the 28.9 constant and 100% cap points), there was no agreement to replace "51.41" with operational Wobbe in either the September 2006 or February 2007 Steering Committee meetings, and there was no breach of any obligation of good faith on BP's part at either meeting. The invoices were validly issued by BP in November 2008, and BP is entitled to the sums claimed with interest from November 2008 in accordance with the terms of the JSA.

THE ISSUES IN DISPUTE

Accordingly, there are five issues for determination:

THE WITNESSES

I heard oral evidence from seven witnesses who had involvement in the relevant events, four for BP and three for Sonatrach. BP also served some evidence under CEA Notices as well as a statement from a Ms Midwinter of BP's solicitors Holman Fenwick and Willan explaining why some potential witnesses were not giving evidence.

For BP the following witnesses were called and cross-examined:

For Sonatrach the following witnesses were called and cross-examined:

It will be apparent from the issues as identified that this is not a case which turns on disputed issue of fact on which there is conflicting factual evidence, other than what was or was not discussed and agreed in the September 2006 and February 2007 Steering Committee meetings (in the context of issues 3 and 4). Indeed, the only witness called who attended the September 2006 Steering Committee meeting is Mr Stranks of Sonatrach, and the only witnesses called who attended the February 2007 Steering Committee meeting are Mr Duncan and Mr Wood. Various witnesses also gave evidence as to events after the execution of the JSA, subsequent developments, and events before and after the September 2006 and February 2007 Steering Committee meetings, as well as evidence from Mr Way as to the work of the Agent, and the preparation of various nitrogen reconciliations.

Both Mr Eaton for BP, and Mr Harris for Sonatrach, made submissions in closing as to how I should treat the evidence of their respective witnesses and particular witnesses in that regard, which I have borne in mind. Where necessary I have commented on the evidence of particular witnesses as it arises in the consideration of particular issues. Generally speaking, however, I consider that all the witnesses were doing their best to assist the Court in relation to events which occurred up to ten years ago, and which can hardly have been fresh in their minds, absent refreshment by reference to the documents.

Mr Winstanley and Mr Stranks gave evidence, which is relied upon by both parties, that is said to go to the factual matrix in which the JSA, and D2.2.2(ii) of the JSA are to be construed, and what a reasonable person having all the background knowledge which would have been available to the parties would have understood the parties to be using the language of the contract to mean. I will have more to say about that in the context of the issues of construction that arise, and the associated applicable legal principles, but suffice it to say at this point that, as is often the case, both witnesses, on occasions, strayed into expressing views as to what they thought provisions meant or were intended to mean, and such evidence, being subjective in nature, has always been recognised as inadmissible (from Prenn v Simmons [1971] 1 WLR 1381 and Reardon Smith Line Ltd v Yngvar Hansen-Tangen [1976] 1 WLR 989 through Investors Compensation Scheme v West Bromwich Building Society [1998] 1 WLR 896, to Arnold v Britton [2015] AC 1619).

THE CONTRACTUAL BACKGROUND

As already identified BP and Sonatrach's Phase 1 rights were formalised in the Specific Term Agreement (STA) dated 27 October 2003 which sets out the detailed terms as between Grain as operator, and BP and Sonatrach as a single shipper. Whilst the STA consists of a two-page agreement it incorporates 30 pages of Schedules and 200 pages of Version 1 of Grain's General Terms and Conditions ("GTCs"). The STA was subsequently subject to various amendments including a new Version 2 of the GTCs. The rights were effective from 15 July 2005 and are valid for a period of 20 years from that date. Together, BP and Sonatrach have a contractual right to 55 berthing slots per contract year, 192,000m³ of LNG storage capacity and 140 GWhrs per day of send out capacity.

The GTCs are intended as a set of uniform terms to which anyone who wants to use the Terminal (a "Shipper") must adhere (A1.1.1). They are divided into six Parts, directed to specific topics such as LNG discharge and LNG storage. The GTCs are the foundation for a contractual scheme by which a Shipper is entitled to discharge LNG into the Terminal, store it there, and nominate quantities to be regasified and sent-out into the distribution pipelines. In return, the Shipper contracts to make specified payments to Grain. The core payment is an Annual Capacity Charge, which is the basic fee for the contractual service (A6). Shippers must also reimburse Grain for certain costs (for example as set out at C10.2 & C12.1), which are to be determined in accordance with Costs Allocation Principles ("CAPs") (see D2).

More specifically, Clause C10.2.1 of the GTCs deals with Grain taking blending measures for the purpose of blending regasified LNG in order to ensure that gas delivered to the LDZ System Entry Point(s) and/or the NTS System Entry Point(s) complies with the applicable Gas Entry Conditions. Clause C.10.2.2 provides that Grain is entitled to recover from Shippers (in accordance with Section C10.2.3) the capital and operating costs incurred by Grain in taking any blending measures to the extent of the costs incurred by Grain acting as a Reasonable and Prudent Operator ("RPO"), such costs being determined and allocated in accordance with the CAPs, in accordance with section D2.2.

The GTCs contemplate that each Shipper would conclude a Specific Terms Agreement with Grain. This covers matters personal to the Shipper, such as capacity levels and charges (A1.2). For BP and Sonatrach, these were set out in STA Schedule 1, which essentially reserved all of the Phase 1 capacity to BP and Sonatrach.

Shippers warrant under the GTCs that their LNG will comply with a "Specification" at the point of discharge into the Terminal (B8.2.1). Grain is entitled to reject LNG which does not comply (subject to a "best endeavours" obligation to implement measures which will allow for discharge) and to compensation for any losses caused (B8.3.1 & 8.3.5-6).

By B8.1.1(b), the Specification matches the "Gas Entry Conditions... in... the NTS Network Entry Agreement". This refers to the gas quality requirements in the contract between Grain and Transco under which Grain sends gas out from the Terminal and into Transco's National Transmission System (the NTS that has already been referred to).

As has already been foreshadowed, the NTS is one of the UK's gas transmission (pipeline) systems. It is a high-pressure system for transmitting large volumes of gas over long distances to large users, such as power stations. Another system consists of the Local Distribution Zones (the LDZ). These are lower-pressure systems for transmitting gas to users in the vicinity of a plant. As previously noted, the Terminal is connected to both systems (B.2.1).

Anyone who wants to send gas into either system must have a Network Entry Agreement ("NEA") with Transco. Grain's NEAs with Transco were still in draft when the STA was signed. Clause 8 of STA Schedule 2 provided that the finalised NTS NEA would match a document which had been initialled in June 2003. The maximum Wobbe in that document was 51.41. This value was written into the STA Specification through B8.1.1(b).

Grain's finalised NEAs for both the NTS and the LDZ set the same 51.41 maximum. They also set a minimum of 47.2. These values are from Part I of Schedule 3 to the Gas Safety Management Regulations 1996 (i.e. the GSMR). GSMR Section 8(1) provides that "no person shall... convey gas in a network unless the gas conforms with the requirements specified in Part I of Schedule 3"). Sections 8(2)-(4) make an exception for "network emergencies".

Corresponding to the Shipper's obligation under B8.1.2 to meet the Specification when LNG is discharged into the Terminal, Grain is obliged to ensure that LNG which is regasified and sent-out pursuant to Shipper nominations complies with the NEA when it is delivered into the NTS or LDZ. So the Shipper is obliged to ensure a maximum Wobbe of 51.41 at the point where LNG enters the Terminal, and Grain is obliged to ensure a maximum Wobbe of 51.41 at the point where it leaves the Terminal (as gas).

BP in its submissions (and in its witness evidence) draws attention to the fact that compliance on arrival is no guarantee of compliance on departure. LNG can change during storage in-tank. This is called "weathering". In particular, some LNG will vaporise back into gas ("boil-off"). Because some hydrocarbons vaporise more readily than others (notably those high in nitrogen), boil-off can change the composition – and properties – of the LNG which is left behind. C10.2.1 recognises this by acknowledging that Grain may have to blend regasified LNG with other gases before send-out in order to meet the NEA, even if the LNG complies with the Specification on delivery.

As mentioned above, Grain is entitled to reimbursement of any such blending costs which it incurs acting as a "Reasonable and Prudent Operator" ("RPO") C10.2.2. The variable component of nitrogen costs in the nitrogen allocation formula in the JSA did not take into account factors such as "weathering" and "boil-off", it simply allocated in accordance with the quantity of gas sent out and the quality of the LNG delivered by each Co-Shipper (together the "Shipper" or "Co-Shippers", individually a "Co-Shipper"). This may have been to the potential benefit of Sonatrach, as some of its cargoes were more liable to "boil off" (due to a high nitrogen content).

It will be necessary to say rather more about the RPO provision in due course, as an issue has arisen between BP and Sonatrach in the context of how the parties might, at the time of contracting, have reasonably expected Grain to approach nitrogen blending for Wobbe.

By Clause 1 of STA Schedule 3, BP's and Sonatrach's STA rights and obligations are generally joint and indivisible, and they are "Co-Shippers". By Clause 2, this does not apply to financial obligations (whether in debt or in damages), where the default is a 50/50 split.

LNG from most sources has a Wobbe higher than 51.41. So the 51.41 maximum Wobbe in the Specification potentially restricted the Co-Shippers' use of the Terminal. In particular, Sonatrach is based in Algeria, and Sonatrach contemplated shipping Algerian cargoes, which have high Wobbe values.

Clause 4 of STA Schedule 2 gave the Co-Shippers a right to increase the Specification Wobbe, conditional on their paying for nitrogen blending at the Terminal so that Grain could still comply with the 51.41 Wobbe limit in the NEAs at point of send-out. The consequence of this was that Co-Shippers would be allowed to discharge LNG which exceeded the 51.41 limit on arrival, provided that they paid for treatment to ensure that it met that limit on departure. If the Co-Shippers requested an increase, Grain was to prepare, for discussion and agreement with them, a proposal detailing "Specific Blending Measures" (Clause 4.8 of STA Schedule 2).

The Co-Shippers invoked Clause 4. The resulting agreed proposal was brought into contractual effect by the FLA in November 2004. The FLA is founded upon, and repeatedly cross-refers to, a Nitrogen Supply Agreement ("NSA") between Grain and Air Products (BR) Ltd ("AP").

By Clauses 2 to 3 of the NSA, AP agrees:

Clause 2.2 of FLA Annex A, and Clause 1 of Appendix 1 to that Schedule, state that the NSA represents agreed Specific Blending Measures for the purposes of Clause 4 of STA Schedule 2. Clause 8.4 of FLA Annex A confirms that the Specific Blending Measures shall be deemed to comply with Clause 4.6 of STA Schedule 2.

Based on AP's commitment under the NSA to supply nitrogen up to specified levels (the "Firm Nitrogen Commitment"), as set out in Clause 4.3 of the FLA Annex A, Clause 4.4 of FLA Annex A, revised (amongst other matters) the Specification Wobbe (i.e. the maximum permitted Wobbe) to:

Clause 3 of Appendix 1 to FLA Annex A states the categories of cost which Grain is entitled to recover from the Co-Shippers in relation to the Specific Blending Measures. They include power (to run the GAN plant), which is to be metered separately from other Terminal power costs (Clause 4).

Annex B to the FLA introduced Version 2 of the GTCs. One change was a new C12.2.2(a) which effectively provided that power costs for the Specific Blending Measures would be recovered under the FLA, as opposed to the general provisions in Section C12. It will be necessary to say more about Specific Blending Measures and Grain's recovery of those costs in due course, in the context of how the parties might, at the time of contracting, have reasonably expected Grain to approach nitrogen blending for Wobbe. In this regard BP submits that the RPO provision in the GTCs does not apply to nitrogen blending under the FLA as this does not constitute blending measures within GTCs C10.2.1. It is said that such nitrogen blending constitutes shipper-specific blending measures within C10.2.4, and as such Grain's right to recover costs is not subject to the C10.2 RPO limitation. I will address this point in the context of a consideration of issue 1.

The JSA contains some 120 pages, with 60 pages of Schedules (1-10). The Clauses are divided into 6 Parts, A-F, in a structure which follow that of the STA. Recital H sets out the purpose of the JSA:

Thus the JSA regulates the division, as between BP and Sonatrach, of rights and obligations which, as between the two collectively and Grain, are generally joint and indivisible under the STA. In this regard Grain treated BP and Sonatrach as a single shipper and did not apply the CAP to them, but rather invoiced a single sum, which was then to be allocated between BP and Sonatrach in accordance with the provisions of the JSA. In this regard Sonatrach rightly notes that the relationship between BP and Sonatrach is not an adversarial commercial relationship under which BP and Sonatrach were seeking a commercial advantage over the other. Ultimately, of course, what is to be interpreted are the terms of the JSA and the allocation formulae contained therein, because those are the terms that the parties have chosen and agreed to allocate sums between them.

Part A provides for the establishment of two entities, a Co-Shippers Steering Committee (the "Steering Committee") and the Agent. The Co-Shippers Steering Committee is of relevance in relation to issue 3 (and whether there was any agreement at the September 2006 Steering Committee meeting to amend the JSA to replace 51.41 in D2.2.2(ii)(b) with the monthly weighted average Wobbe value in fact blended to by Grain) and issue 4 (and whether if BP did not so agree at the September 2006 or February 2007 Steering Committee meetings, whether BP thereby breached a duty of good faith under the JSA). The Agent is relevant because one of its functions is to calculate the allocation of costs between BP and Sonatrach under the JSA, and BP's claim is based upon the Agent's calculations. The relevant provisions of the JSA in relation to the Steering Committee and the Agent are addressed in more detail in the context of issues 3 and 4 below.

The Steering Committee's general role is to "be a forum for discussion and agreement on all commercial, technical and operational matters relating to" the STA (A1.1.1). Its more specific functions include to:

BP and Sonatrach are each obliged to appoint to the Steering Committee two employees with authority to make binding decisions with respect to matters within the Committee's remit (A1.2). In performing this appointment obligation, and in performing their other obligations under the JSA, BP and Sonatrach are obliged to act as a Reasonable and Prudent Operator ("RPO") (A1.7.4). An RPO means "a person seeking in good faith to perform its contractual obligations" (Part F1.1).

The Agent also consists of representatives appointed by BP and Sonatrach (A1.5 and Schedule 1, Clause 1.1). It has various duties under the JSA, including, “to determine how costs are shared between the Co-Shippers.". The Co-Shippers acknowledge at A1.6.2 that the Agent is to carry out its duties in each case, "efficiently and effectively, acting as a Reasonable Prudent Operator and in accordance with, and subject to, the provisions of this Agreement or as otherwise agreed between the Co-Shippers" (A1.6.2). Schedule 1 and Clause 1.7 thereof provides that the representatives designated by the Co-Shippers "shall carry out the duties of the Agent under this Agreement in accordance with the provisions of this Agreement, and shall do so in a fair and equitable way, for the benefit of both Co-Shippers."

The JSA's Financial Provisions, and the Agent's functions in relation to invoicing and allocation of costs, are set out in Part D.

Part D2 is headed "Cost Allocation Principles" and Sections D2.1 and D2.2 refer to "Cost Allocation Principles under GTCs" and "Co-Shippers Cost Allocation Principles" (respectively). D2.2.2. provides:

Six types of costs are then addressed (Electrical Power Charges, Nitrogen Costs, Propane Costs, Recondenser Costs, Blending Costs and Other Costs). Notwithstanding the references to "principles” the cost allocation principles are either expressly stated or are formulaic (as in the case, for example, of Nitrogen Costs and Propane Costs).

D2.2.2(i) relates to Electricity Power Charges. The fixed component is borne equally, and the variable component is borne by the Co-Shippers, "pro-rata to their Send Out Quantities in that month”. However, D2.2.2(i) expressly provides that "power costs in relation to nitrogen blending will be separately measured and recovered by [Grain] as nitrogen blending costs and not as power costs" (which reflects the terms of the FLA and the new Clause C12.2.2(a) of the GTCs, which provide for power for nitrogen blending to be treated as a nitrogen cost).

D2.2.2(ii) relates to Nitrogen Costs (including power for nitrogen blending) and has already been quoted in full above, being the provision at the heart of the dispute. As already identified, D2.2.2(ii) divides nitrogen costs into fixed and variable. D2.2.2(ii)(a) allocates fixed costs (for example, AP's annual charge under the NSA) 50/50.

D2.2.2(ii)(b) allocates variable costs (for power for the GAN plant, trucked in LIN) according to a formula. The variables are:

The "quantity of nitrogen used" is itself determined by another formula. This multiplies the gas which each party sends out in the month (in GWh) by the greater of Factor X and Factor Y. Factor X relates to the Incomplete Combustion Factor ("ICF") of the LNG which each party discharges in the month. ICF is another LNG property limited by the GSMR and NEAs.

Factor Y relates to the Wobbe of the LNG which each party discharges in the month. In simplified terms, and as already identified, Factor Y represents the difference between the weighted average Wobbe of the LNG which a party discharges in the month and 51.41.

If a reconciliation under D2.2.2(ii) shows an amount owing between BP and Sonatrach, the creditor is entitled to issue an invoice (D2.2.3, D1.6). Invoices are payable 10 days after receipt and bear interest at LIBOR+3% if unpaid (D1.6.3, D1.6.8). It will be necessary to consider the provisions as to invoicing in more detail in the context of issue 5 (the validity of the invoices issued by BP).

Under D1.2, D2.2.2(ii)(c), and D2.2.3, the Agent is to calculate the allocation of nitrogen costs monthly, with annual reconciliations. But the calculations depend upon information from Grain, and Grain did not provide sufficient information in time for the Agent to perform calculations according to plan. In the event, the Agent did not produce any complete nitrogen cost reconciliations until September 2008. Although the Agent was unable to calculate allocations, Grain invoiced nitrogen costs to BP and Sonatrach, and BP and Sonatrach have paid Grain. It is in that context that the issue arises (as subsumed within issues 1 and 2), as to the proper allocation of nitrogen costs, which have already been incurred and paid under the STA, as between BP and Sonatrach under D2.2.2(ii) of the JSA.

D2.2.2(iii) provides that fixed propane costs relating to the additional propane facility "shall be borne by the Co-Shipper who has requested use of the facility, i.e. Sonatrach. If BP utilises the additional propane facility, BP shall pay a portion of such fixed costs which is to be agreed." Variable propane costs are to be borne by the Co-Shippers in accordance with "the quantity and quality of the LNG delivered" in accordance with provisions set out in D2.2.2(iii)(c).

After addressing "Recondenser Costs" and "Blending Costs" at D2.2.2(iv) and (v), D2.2.2(vi) provides in relation to Other Costs (for which a method of allocation between the Co-Shippers is not already set out), that fixed costs incurred for the benefit of both Co-Shippers are borne evenly, and in all other circumstances, "in accordance with the appropriate method of allocation as agreed between the Co-Shippers and notified to the Agent by the Co-Shippers. The appropriate method of allocation agreed between the Co-Shipper shall take into account the relevant basis of usage of the service or commodity or cause of the cost as the case may be."

Section E15 of the JSA provides that, "Except as expressly provided in this Section E15 or otherwise elsewhere in this Agreement, no amendment or variation of this Agreement shall be effective unless in writing and signed by or on behalf of each Co-Shipper". On established legal principles, a provision such as E15 does not preclude the parties from amending the JSA other than in signed writing - see, for example, MWB v Rock [2016] EWCA Civ 553.

In November 2004, Grain issued draft Cost Allocation Principles ("CAPs") for the purposes of:

Section 3.1 referred to the fact that the GTCs define the basis for further allocation principles, and stated, “In general, fixed costs are to be allocated pro rata Shippers’ Delivery Capacity, and variable costs are to be allocated pro rata Shippers’ Delivered Quantity...”.

The draft CAPs also referred, in Section 7, to the GAN plant. This was beyond the scope of the CAPs which were required under the GTCs, because nitrogen costs were already covered by the FLA (as recognised in the new GTCs Clause C12.2.2(a)). Grain acknowledged this, but (looking forward to the future and a multi-shipper environment e.g. in Phases 2 and 3) stated that it was possible that two or more shippers may share the benefit of a single nitrogen generating facility or share liquid nitrogen which had been transported to the terminal (trucked-in LIN), and Grain had given consideration to the basis on which the related monthly cost might be allocated to each shipper. Grain stated that its current approach (to such a future situation) would be that the related monthly costs to be allocated to each shipper would be calculated by reference to a formula.

The formula (in simplified terms) was as follows:-

BP agreed with Grain's proposal in March 2005. There were no new Shippers until Phase 2 came into operation in December 2008. Before Phase 2, there was only one Shipper - i.e. BP and Sonatrach as joint Shipper under the STA with Grain passing on all nitrogen costs to BP and Sonatrach. After 2008, Grain began allocating nitrogen costs between BP and Sonatrach (as the single Phase 1 Shipper) and the Phase 2 Shippers (using, for at least some of the time, a figure of 50.9 rather than 51.41 as proposed in the draft CAPs). The further sub-allocation of Phase 1's share of nitrogen costs as between BP and Sonatrach continued to be governed by D2.2.2(ii).

In January 2010 (when Phase 2 was in operation) Grain proposed a new approach to nitrogen cost allocation between Shippers in a Consultation Paper as to Nitrogen Allocation dated 4 January 2010. It recognised that in a multi-shipper terminal such as the Terminal it was necessary to comingle and manage LNG stocks from different suppliers to satisfy future storage and sendout requirements. Consequently, it was not possible to isolate and regasify an individual shippers LNG and to directly measure the nitrogen ballasting requirements for allocation of the nitrogen costs to that shipper. It identified that weaknesses in that methodology were that it took no account of deterioration of in-gas quality during storage and the resulting increase in nitrogen demand on send out and of other gas interchangeability parameters (i.e. "weathering" and the like).

The proposal made was to take into consideration all gas interchangeability parameters, including both the Incomplete Combustion Factor (ICF) and Sooting Index (SI), when calculating nitrogen requirements, the energy held by each shipper at any particular time in the tanks and the LNG ageing. This proposal was not well received by shippers, and Grain did not proceed with it.

Instead Grain decided in May 2011 (some five and a half years after the Terminal started operating), that it would allocate nitrogen costs between Shippers by reference to the outturn monthly flow weighted average Wobbe (i.e. the average monthly operational Wobbe), notifying Shippers of the same in a letter dated 11 May 2011. All Shippers supported this, and in a letter dated 16 June 2011 Grain notified all Shippers that it would adopt this approach with effect from 1 July 2011.

Once the Terminal became operational and nitrogen costs began being allocated by reference to contractual formulae it became apparent that the level of nitrogen being used by Grain was significantly above the parties’ expectations. By December 2005 both Mr Martinez and Mr Winstanley were looking at the issue and BP was already in contact with Grain to investigate the use of nitrogen. As I have already identified, Grain's March 2006 Monthly Report indicated that Grain targeted an "optimum" Wobbe of 51.1 and often injected more nitrogen than the "optimum" quantity. In a meeting in May 2006 between Grain, BP and Sonatrach, Grain explained that levels of nitrogen consumption were driven by the trigger points for the LDZ, which were set to 51.26, that being the target of 51.3 less an allowance for inaccuracies.

I have already referred to the fact that the Agent also approached BP about "strange results" with its nitrogen cost allocation spreadsheet in April 2006. In an email to Mr Wood on 28 April 2006, and having reviewed the spreadsheet, Mr Winstanley identified and explained that he had omitted the 28.9 constant for the molecular weight of air in the formula (which involved a calculation of specific gravity relative to air in Factor Y). This was characterised by Mr Winstanley in his evidence, rightly in my view, as a “typographical error”, it was simply a constant that he had inadvertently omitted when setting out the formulae. The failure to include the 28.9 coefficient for the molecular weight of air was referred to by Mr Wood in his email to the Co-Shippers on 18 May 2006. On 3 July 2016 the parties agreed to add the molar weight of air into the formula.

In May 2006 it was identified that the formula sometimes produced allocations in excess of 100%, and Mr Wood referred to this in his letter to the parties on 18 May (“Cost allocation calculation: Co-Shipper A: 125%, co-shipper B -25%”). As I have already noted, it was not immediately clear whether this was intentional given the apparent possibility that a lean cargo might dilute a richer cargo’s nitrogen requirements, generating a credit. However, it transpired that this would only be so if the lean cargo’s Wobbe was below Grain’s operational Wobbe (and it appears that that has not happened in practice). In any event the parties agreed to a capping of allocations to 100% of costs in August 2006 (the "cap" or "100% cap"). This limited the sum which could be allocated under D2.2.2(ii)(b) to 100% of the costs charged by Grain. Whilst there was some cross-examination of Mr Bax as to whether there was any agreement as to how this was to be implemented (which is not a pleaded point), it does not appear that there has ever been any challenge to how the Agent implemented the cap in its calculations.

There were also discussions between BP and Sonatrach in the summer of 2006 as regards the use of 51.41 in the formula, and the possibility of 51.41 being changed to the monthly weighted average of the Wobbe of the gas sent out from the terminal as advised by Grain. These discussions are addressed in due course below in the context of issue 3, and whether there was any agreement in the subsequent September 2006 Steering Committee meeting to replace 51.41 in D2.2.2(ii)(b) with the monthly weighted average Wobbe value in fact blended to by Grain. It suffices for present purposes to note that it is not suggested that those discussions had led to any agreement in that regard in advance of the September 2006 Steering Committee meeting (the subject matter of issues 3 and 4).

In addition to the dispute between themselves, BP and Sonatrach challenged Grain about nitrogen consumption and the level of costs which Grain was passing on to them as Co-Shippers under the STA/FLA and CAPs. Sonatrach relies on this in relation to its good faith argument under issue 4, as addressed below. In August 2013, BP wrote to Grain jointly with Sonatrach complaining that Grain was using excessive nitrogen and referring specifically to the gap between the 51.41 limit and Grain's operational Wobbe. In particular it was noted that the Terminal typically blended gas to a Wobbe index of 51.1 or lower and it was submitted that this was not a reasonable level of blending, and that Grain was not permitted to recover such costs under the STA. In the event, shippers' attempts to persuade Grain to use less nitrogen have been unsuccessful.

Sonatrach point out that if the figure of "51.41" is a constant, in circumstances where Grain did not target a Wobbe of 51.41, but rather targeted a Wobbe of less than that (targeting an "optimum" Wobbe of 51.1 and often injecting more nitrogen than the "optimum" quantity), then the effect of the formula, applying 51.41 as a constant, is that one Co-Shipper could end up paying for nitrogen used to blend the other Co-Shipper's cargo of LNG. It is said that that is an absurd outcome, and BP 's construction is an uncommercial construction which must yield to business commonsense. BP's riposte is that the allocation produced by the formula is simply a consequence of the application of the agreed contractual formula in circumstances where Grain, in the event, when the Terminal was operational, subsequently chose to blend to a Wobbe of less than 51.41, and to a level less than the Wobbe of cargoes being delivered to the Terminal. BP submits that Sonatrach is asking the court to rewrite the parties' contractual bargain by ignoring the agreed constant of 51.41 and substituting a variable of the monthly weighted average Wobbe value in fact blended to by Grain, having regard to events subsequent to the entering into the JSA.

In support of its submissions on commercial construction, and its submission that BP’s construction of the formula produced what it characterised as absurd results, Sonatrach relied upon a number of scenarios which it put to Mr Winstanley in cross-examination. Whilst the volumes of gas outflow were simplified and the cost of nitrogen was fixed at £100,000 for convenience, the other data is derived from real experience at the Terminal. The scenarios do not, however, take into account the effect of the 100% cap.

The first and second scenarios assumed that BP brought in a cargo from Trinidad with a Wobbe of 51.37 and Sonatrach brought a cargo from Algeria with a Wobbe of 51.68. It also assumes that both parties’ send-outs (of 200 GWh) were equal for that month, that Grain blends to a target Wobbe of 51.1 and that Grain’s final invoice for the month was of £100,000. On these assumptions:

The third and fourth scenarios use the same assumptions as above, but assume an unequal ratio of send out (350 GWh to BP to 50 GWh to Sonatrach, which is approximately the same as the ratio in June 2011. On these assumptions:

The fifth and sixth scenarios assume that BP brings a cargo from Trinidad with a Wobbe of 51.37 but that Sonatrach brings a cargo with a Wobbe of 51.40, from Qatar (as reflected by a particular Sonatrach cargo). They also assume equal send-outs of 200 GWh. On these scenarios:

BP accepts that these worked scenarios are broadly correct, but points out that they do not take into account the 100% cap agreed between the parties (which meant that neither party would pay more than 100% of the cost charged by Grain). In any event BP say that the cost allocation is simply a consequence of the application of the formula with the contractual constant of 51.41 in circumstances where Grain, in the event, targeted a lower Wobbe when blending and that Sonatrach’s case on construction does not reflect what was contractually agreed, and Sonatrach is effectively inviting the Court to re-write the contractual bargain in the light of subsequent events that were outside the control of either party. That is simply impermissible.

Sonatrach also refers to calculations for the 2008 Reconciliation in relation to which it makes similar points. During that year, BP sent out 5,690,925 GWh of gas and Sonatrach sent out 3,234,181 GWh of gas, and they were together invoiced £550,553.64 by Grain. Applying the contractual formula (corrected for the molar weight of air), this results in an allocation of nitrogen of approximately -1,024,000 tonnes to BP and 2,784,900 tonnes to Sonatrach (in circumstances where every cargo required nitrogen based on the level Grain were blending to), and an allocation of approximately £876,6100 to Sonatrach (on an invoice of £550,553.64) and of -£326,610 to BP. BP makes the same points in response - the calculations do not take into account the 100% cap, and the outcome is the result of applying the contractual formula with a constant of 51.41 in circumstances where Grain, in the event, targeted a lower Wobbe when blending. Sonatrach’s construction (whether by reference to the express terms of the JSA or the implied terms alleged) is not an available construction, and would be to re-write the parties’ contractual bargain, in circumstances where it is not clear what the parties would have agreed, had it been known what levels Grain would blend to at the time of contracting.

PRINCIPLES OF CONSTRUCTION AND IMPLIED TERMS

The proper approach to contractual construction was common ground between the parties, and the main area of debate between the parties was the application of those principles to the facts, and (to adopt the words of Lord Neuberger PSC in Arnold v Britton [2015] AC 1619 at [15] citing Lord Hoffman in Chartbrook v Persimmon [2009] 1 AC 1101), identifying the intention of the parties by reference to, "what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language in the contract to mean."

Both parties referred me to, and relied upon, the judgment of Lord Neuberger in Arnold v Britton, supra, as containing a helpful summary of the applicable principles on contractual construction, although they disagreed on the extent to which there were factual similarities between that case and those of the present case.

For its part BP also referred me to Lord Hoffman's well-known speech in Investors Compensation Scheme Ltd v West Bromwich [1998] 1 WLR 896 at 912F-913E, and Lewison, The Interpretation of Contracts, 6th edn at paragraph 1.01, and identified the Court's task as to ascertain the meaning which the document would convey to a reasonable person seised of the background "matrix" which would reasonably have been available to the parties in the situation in which they were at the time of contracting.

BP rightly pointed out that negotiations and drafts of the relevant contracts are not part of the matrix - Lewison at paragraphs 3.07 & 3.09; Chartbrook, supra at [42] where Lord Hoffmann said: -

In the event the inadmissibility of negotiations and pre-contractual drafts was recognised by the parties at trial, although it is regrettable that the trial bundles included no less than eight lever arch files full of what were described, somewhat unpromisingly, as "Draft Contractual Documents". Their contents were clearly inadmissible, and did not feature in the trial.

Additionally, whilst Sonatrach accepted that evidence of the parties' actions after the conclusion of a contract is not admissible as evidence of the parties' intention at the time of contracting, it submitted that it can nevertheless be admissible if it would be relevant to understand what a reasonable third party would have understood by the clause (referring by analogy to what was said by Lord Hoffman in Chartbrook as quoted above). Whilst I agree that evidence of parties' actions after the conclusion of a contract may, potentially, shed light on what would have been known by the parties or a reasonable third party at the time of contracting - the stance a particular party adopts after the event (for example, in the present case, BP's submission to Grain which it is said is inconsistent with BP's stance on construction of D2.2.2(ii)(b)) may or may not shed light on what a reasonable third party would have understood by the clause at the time of contracting. There may be all sorts of reasons why a party adopts a particular stance (especially with third parties) after the time of contracting, whether or not it is consistent with its stance on construction, or indeed what the parties' knew (or a reasonable third party would have known) at the time of contracting.

In addition to the points made on admissibility, BP also submitted that the focus is on the words which the parties have actually used, that it is not the Court's function to search for ambiguities to justify a departure from the natural meaning of those words; and that the Court should be very slow to reject the natural meaning simply because it may result in an improvident bargain for a party, referring to Arnold v Britton @ 14ff.

For its part Sonatrach submitted that it was "clear that something has gone wrong with the language" (Lord Hoffman in Chartbrook at [25]), and that whilst it is well established on the authorities that although commercial common sense and surrounding circumstances "should not be invoked to undervalue the importance of the language of the provision which is to be construed" (Arnold at [17]) there will be "unusual cases" (ibid) and the ordinary meaning of a word or phrase in a contractual term is not always determinative, referring to Re Sigma Finance Corporation (in administrative receivership) [2010] 1 All ER 571 in which Lord Mance, at [12] expressed the opinion that:

“the conclusion reached below attaches too much weight to what the courts perceived as the natural meaning of the words of the third sentence of clause 7.6, and too little weight to the context in which that sentence appears and to the scheme of the Security Trust Deed as a whole … Even the most skilled drafters sometimes fail to see the wood for the trees, and the present document on any view contains certain infelicities, as those in the majority below acknowledged” (at [12]).

Particular passages from the speech of Lord Hoffman in Chartbrook are relied upon by both BP and Sonatrach, who each find something to rely upon in support of their arguments on construction in paragraphs [20] and [24]-[25]:-

BP relies upon the opening words of paragraph 20, and the final sentence of paragraph 25 (submitting that such final sentence is not apt to describe D2.2.2(ii)(b) and the issue of construction in this case). Sonatrach relies on the remainder of paragraph 20 and paragraphs 24 and 25, as referred to above, (submitting that the final sentence of paragraph 25 is apt to describe D2.2.2(ii)(b) and the issue of construction in this case).

Sonatrach also referred, in the context of its submission that the literal meaning of a word (or here a number) would, in Sonatrach's submission, lead to what it categorised as a "commercial nonsense" to the well-known, and often quoted, words of Lord Diplock in Antaios Compania Naviera S.A. v Salen Rederierna A.B. ("the Antaios") [1985] A.C. 191, 201, referred to with approval by Lord Hoffmann in ICS v West Bromwich at p. 913E, that

“if detailed semantic and syntactical analysis of words in a commercial contract is going to lead to a conclusion that flouts business commonsense, it must be made to yield to business commonsense”

In this regard Sonatrach also referred to the Supreme Court decision in BNY Mellon Corporate Trustee v LBG Capital No 1 Plc [2016] UKSC 29 in which Lord Neuberger agreed that a clause in a financial instrument that defined a credit default event by reference to the bank’s “Core Tier 1 Capital” should be read as including the proviso “or its then equivalent” (at [35] to [38]).

BP's riposte to assertions of commercial nonsense was two-fold. Firstly, from a factual perspective, to submit that the formula in D2.2.2(ii)(b) was clear and deliberate, and intended to mean what it stated on its face, producing an allocation of variable nitrogen costs in accordance with the quantity of gas sent out and the quality of the LNG delivered by each Co-Shipper - such a construction did not, BP submitted, lead to a conclusion which flouted business commonsense. Secondly, from a legal perspective, by reliance upon the first of the seven factors identified by Lord Neuberger in Arnold v Britton, at [17] (which I set out in full below when addressing Arnold v Britton), in which he emphasised that the reliance placed in some cases on commercial common sense should not be invoked to undervalue the importance of the language of the provision which is to be construed, and for the reasons he gives.

Lord Carnwath JSC in his dissenting judgment in Arnold v Britton referred to the similar sentiments expressed in a rider to the well-known quotation from the judgment of Lord Diplock in The Antaios referred to by Lord Clarke in Rainy Sky SA v Kookmin Bank [2011] 1 WLR 2900 at [23] in which he quoted the words of Hoffmann LJ in Co-operative Wholesale Society Ltd v National Westminster Bank plc [1995] 1 EGLR 97 at p. 99:-

Sonatrach submitted that the position is stronger if the situation is one where it can be said that it was clear how the parties would have understood and/or drafted the contract had they been aware of an event subsequent to the contract which was “plainly not intended or contemplated by the parties” (referring to Lord Neuberger at [22] of Arnold v Britton, and Lloyds TSB Foundation for Scotland v Lloyds Banking Group Plc [2013] 1 WLR 366 at [23]). Sonatrach submits that this was the case here, and that had the parties known what subsequently transpired in terms of Grain's actions concerning blending, the parties would have substituted the number "51.41" with “the average monthly operational Wobbe value used for LNG blending and regasification at the Terminal from time to time”.

BP's riposte to this submission was that this was not such a case, whether in terms of events subsequent to the JSA, or as to how the parties would have understood and/or drafted the contract had they been aware of events subsequent to the contract, at the time of contracting. BP submitted that subsequent events showed Grain (subsequently) blending to a variety of different targets and/or operational Wobbe numbers, and it was far from clear how the parties would have understood or drafted the contract had they been aware, at the time of contracting, of what Grain subsequently did in relation to blending, in terms of what the parties might or might not have done to the figure of 51.41 given the various differing possibilities that existed (for example change the number to another number (e.g. 51.1), or to a reference to the Wobbe targeted from time to time, or to a reference to the actual Wobbe value that Grain would blend to, or a reference to an average operational Wobbe (monthly or otherwise), or indeed to do nothing).

BP submitted that such permutations illustrate that Sonatrach was effectively seeking to re-write the contractual bargain between the parties. In this regard BP referred to what Lord Neuberger PSC said in Arnold v Britton at [41]:-

I bear in mind all the authorities referred to by the parties, and the principles established by those authorities. Ultimately I consider that the most recent guidance from the Supreme Court, in Arnold v Britton, and in particular the judgment of Lord Neuberger which draws together the applicable principles on contractual construction and emphasises factors that may be of importance in particular cases, is of the greatest assistance when it comes to addressing the issues of construction that arise in this case.

At paragraphs 14 to 23 of his judgment in Arnold v Britton (with whom Lords Sumption and Lord Hughes JJSC agreed) Lord Neuberger PSC said as follows:

Lord Hodge JSC (in agreeing with the majority) stated in relation to the conclusion that he had reached as the proper construction of the leases, as follows [76] to [79]:-

Arnold v Britton concerned the proper construction of service charge contribution provisions in the leases of a number of chalets in a caravan park in South Wales. A first group of leases included at clause 3(2), a lessee's covenant “to pay … a proportionate part of the expenses and outgoings incurred by the lessors in the repair, maintenance, renewal and the provision of services [as set out in the lease] in the yearly sum of £90”, increasing thereafter at a three-yearly compound rate of 10%. Subsequently, leases in respect of further chalets were granted in which clause 3(2) provided for an annual 10% compound increase. In 2011 the tenants of those chalets whose annual service charge had risen to over £2,700, as opposed to £282 for the chalets subject to the earlier leases, and who claimed that an interpretation of the clause which required a fixed sum payment resulted in such an absurdly high annual service charge that it could not be right (the charge would reach over £550,000 by 2072), claimed that clause 3(2) should be read as requiring them to pay a variable sum, being a fair proportion of the cost of providing the services, with the specified sum being no more than a cap on the maximum sum payable.

The landlord commenced Part 8 proceedings in the county court seeking a declaration that clause 3(2) required the payment of the fixed sum and not any lesser variable amount. The judge accepted the construction contended for by the tenants and so dismissed the claim. Allowing an appeal by the landlord the High Court judge held that, on a natural reading of the clause, the object of the verb “to pay” was the fixed sum of £90 as escalated, whereas the construction contended for by the tenants would involve rewriting the bargain which the parties had made, which, given the high levels of inflation in the United Kingdom when those leases had been granted, could not be said at that time to have lacked commercial purpose. Dismissing the tenants' appeal, the Court of Appeal affirmed the High Court judge's reasoning and held, additionally, that the words “a proportionate part” were not inconsistent with a fixed service charge in circumstances where other lessees were contributing to the overall service charge, which in consequence was to be apportioned between them. The Supreme Court dismissed the appeal.

Lord Neuberger rejected the construction proposed by the tenants notwithstanding the consequences of the construction proposed by the lessors which he described as “unattractive, indeed alarming” to the lessees (at [30]). He identified that “the natural meaning of the words used, at least until one considers the commercial consequences, seems clear” (para [24]). He identified that the advantage of a fixed charge formula was certainty, and that the leases were made at a time of inflation which the parties objectively, and commercially, could be expected to want to confront. “They chose to do so by this particular formula of increase.” (para [26]).

Lord Neuberger then addressed the various arguments of the lessees at paragraphs [29] to [32] of his judgment:-

“[29] ….given the way things have turned out, it is tempting to latch onto the absence of words such as “quantified in the sum of”, and to see the two halves of clause 3(2) as mutually inconsistent in their effect. This would be on the ground that the first half of the clause requires the lessee to pay a “proportionate part” of the cost to the lessor of providing services, whereas the latter half requires the lessee to pay a sum which could exceed the whole of that cost. On that basis, it might be said that the court can reject or modify one half to give effect to the real intention of the parties:  eg Walker v Giles (1848) 6 CB 662 . However, as explained in para 24 and 25 above, this argument would, in my view, involve the court inventing a lack of clarity in the clause as an excuse for departing from its natural meaning, in the light of subsequent developments.

30 Were it not for the percentage increases of 10% per annum specified in the 25 service charge clauses which are being considered on this appeal, coupled with the subsequent history of inflation in the United Kingdom, that would be the end of it…. However, the consequences of the annual sum of £90 being increased annually by 10% on a compound basis are plainly unattractive, indeed alarming, to a lessee holding a chalet under one of the 25 leases. If one assumes a lease granted in 1980, the service charge would be over £2,500 this year, 2015, and over £550,000 by 2072. This appears to be an alarming outcome for the lessees, at least judging by how things look in 2015, because annual inflation in the last 15 years has hardly ever been above 4%, indeed has been under 3% for ten of those years, and has notoriously been falling recently almost to the point of turning negative, whereas the service charge over that period has increased, and will continue to increase, by 10% per annum.

31 The appellants argue that these figures illustrate the extreme unlikelihood of the parties to the 21 leases (or to the four subsequent deeds of variation), and in particular the lessees, having intended to agree that the original £90 service charge would be automatically increased by 10% annually on a compound basis. Accordingly, they contend, the latter half of clause 3(2) should be interpreted as imposing a maximum on the annual service charge recoverable by the lessor. In other words, the effect of the clause is said to be that the lessor is entitled to an appropriate percentage of the annual cost of providing the contracted services, subject to a maximum—which was initially £90, but which increases by 10% compound annually.

32 Despite the unattractive consequences, particularly for a lessee holding a chalet under one of the 25 leases, I am unconvinced by this argument. It involves departing from the natural meaning of clause 3(2) in each of those leases, and it involves inserting words which are not there.

33 Further, the appellants' argument involves attributing to the parties to the 25 leases an intention that there should be a varying service charge and that the lessor (or some other unspecified person) should assess the total costs of the services and determine the appropriate proportion of the cost of the contractual services to allocate to each chalet. Although I accept that it has an element of circularity, it appears to me that the average reader of clause 3(2) would have thought that those are exercises which the clause seems to have been designed to avoid.”

Lord Neuberger also stated at paragraph [41]:-

BP relies, in particular, on the passage highlighted above submitting, by way of analogy, that there is no principle of interpretation that would entitle the court to re-write the Y formula substituting for the figure of 51.41, a variable in the form of, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” simply because of the allocation that subsequently resulted because of the Wobbe level that Grain, in the event, chose to blend to (a matter outside the control of either party – like inflation says BP).

More generally, BP submits that Arnold v Britton is an analogous case factually – a case where the literal words of the contract are clear, yet they produce a result which it is submitted is so uncommercial that the parties cannot have intended that, notwithstanding which the court has applied the ordinary and natural meaning of the words used, and declined to re-write the contractual bargain to produce what might be perceived to be a more commercial result in the light of subsequent events. For its part Sonatrach denies that there is an analogy to be made, and points out that it would have been possible, at the time of entering into the leases, for the parties in that case to have appreciated (indeed calculated) the consequences of the formula in the lease applying different levels of inflation (even though the level of future inflation, in fact, was unknown). By the same token, and by way of riposte, BP submits that BP and Sonatrach could have run the formula using figures different to 51.41, but they did not do so. They chose the formula that is in clause D.2.2.2(ii)(b) and that is to be given effect to in accordance with its express terms.

In my opinion the real worth of Arnold v Britton is the definitive guidance given in the judgment of Lord Neuberger as to the principles of contractual interpretation and the guidance he has given in that regard in the course of his judgment. To the extent that these assist in relation to the exercise of construction in this case (which they clearly do), it is in relation to the application of those principles to the clause under consideration that is key, rather than any attempt to consider factual analogies. Ultimately it is clause D2.2.2(ii)(b) which has to be construed applying the applicable principles as to contractual construction.

Sonatrach submits that if it is wrong as a matter of construction of D2.2.2(ii)(b), then the two terms it alleges fall to be implied into D2.2.2(ii)(b) as a matter of necessity. BP denies that they meet the requirements for the implication of terms. It submits that the alleged terms are neither necessary to give business efficacy to the JSA, nor are they so obvious that “it goes without saying”, nor are they sufficiently certain. Even more fundamentally, BP submits that in any event such terms are inconsistent with, and contradict, the express terms of D2.2.2(ii)(b), and the figure of 51.41, and as such cannot be implied on established principles. BP also submits that the alleged implied terms are based on what it says are the same incorrect premises as Sonatrach’s construction argument and, as such, lend no support for Sonatrach's implied term argument either.

The applicable principles on implication of terms are well-known, and not in dispute, although there was a difference between the parties as to whether the proposed implied terms were inconsistent with the express terms, and as to the utility or otherwise of considering cases in which the courts have been willing at least to modify existing phrases in contracts by an implied term, in the context of the express terms of D2.2.2(ii)(b).

The applicable principles in relation to the implication of terms were reviewed and summarised by Lord Neuberger PSC in Marks and Spencer plc v BNP Paribas Securities Services Trust Company (Jersey) Limited and another [2016] A.C. 742 at [16] – [20]:

Lord Neuberger then went on to add six comments of his own (at [21]):

“It could be dangerous to reformulate the principles, but I would add six comments on the summary given by Lord Simon in the BP Refinery case 180 CLR 266, 283 as extended by Bingham MR in the Philips case [1995] EMLR 472 and exemplified in The APJ Priti [1987] 2 Lloyd's Rep 37. First, in Equitable Life Assurance Society v Hyman [2002] 1 AC 408, 459, Lord Steyn rightly observed that the implication of a term was “not critically dependent on proof of an actual intention of the parties” when negotiating the contract. If one approaches the question by reference to what the parties would have agreed, one is not strictly concerned with the hypothetical answer of the actual parties, but with that of notional reasonable people in the position of the parties at the time at which they were contracting. Secondly, a term should not be implied into a detailed commercial contract merely because it appears fair or merely because one considers that the parties would have agreed it if it had been suggested to them. Those are necessary but not sufficient grounds for including a term. However, and thirdly, it is questionable whether Lord Simon's first requirement, reasonableness and equitableness, will usually, if ever, add anything: if a term satisfies the other requirements, it is hard to think that it would not be reasonable and equitable. Fourthly, as Lord Hoffmann I think suggested in Attorney General of Belize v Belize Telecom Ltd [2009] 1 WLR 1988, para 27, although Lord Simon's requirements are otherwise cumulative, I would accept that business necessity and obviousness, his second and third requirements, can be alternatives in the sense that only one of them needs to be satisfied, although I suspect that in practice it would be a rare case where only one of those two requirements would be satisfied. Fifthly, if one approaches the issue by reference to the officious bystander, it is “vital to formulate the question to be posed by [him] with the utmost care”, to quote from Lewison, The Interpretation of Contracts 5th ed (2011), p 300, para 6.09. Sixthly, necessity for business efficacy involves a value judgment. It is rightly common ground on this appeal that the test is not one of “absolute necessity”, not least because the necessity is judged by reference to business efficacy. It may well be that a more helpful way of putting Lord Simon's second requirement is, as suggested by Lord Sumption JSC in argument, that a term can only be implied if, without the term, the contract would lack commercial or practical coherence.”

An issue that arose in cases subsequent to the Belize Telecom case was whether Lord Hoffmann, in that case, had suggested that the traditional “business efficacy” and “officious bystander” tests were no longer central to the implication of terms, and also in relation to his suggestion that the process of implying a term is part of the exercise of interpretation. Both these matters are of potential relevance in the present case. As to the former BP submits that Sonatrach’s proposed implied terms fail the “business efficacy” and “officious bystander” tests, whilst as to the latter Sonatrach’s case is put both on the basis of construction of the terms of D2.2.2(ii)(b) and on the basis of implied terms. In this regard, Lord Hoffman had stated in Belize Telecom at paragraph [21], that, “There is only one question: is that what the instrument, read as a whole against the relevant background, would reasonably be understood to mean?”.

Lord Hoffmann's observation, was addressed by Lord Neuberger in Marks and Spencer at [22] -[31] where he reiterated that there had been no dilution of the requirements which have to be satisfied before a term will be implied and that in most, possibly all, disputes about whether a term should be implied into a contract, it is only after the process of construing the express words is complete that the issue of an implied term falls to be considered (also noting that the process of implication involves a rather different exercise from that of construction):-

In terms of the “officious bystander”, Sonatrach submits that the officious bystander would be aware of the potential contingencies which might arise - in so far as they could have been foreseen at the time - even if those were not at the time contemplated by the parties. In this regard Sonatrach rely on what was said by Lord Hoffmann in Belize Telecom at [25]:-

“The need for an implied term not infrequently arises when the draftsman of a complicated instrument has omitted to make express provision for some event because he has not fully thought through the contingencies which might arise, even though it is obvious after a careful consideration of the express terms and the background that only one answer would be consistent with the rest of the instrument.”

(emphasis added)

BP, for its part, emphasises the words highlighted above, submitting that it is not obvious that there is in this case only one answer that would be consistent with the rest of the JSA. Indeed, BP submits that there are many possible answers, and the alleged terms are not so obvious as to go without saying or to be necessary for business efficacy. Developing this submission BP asserts that it is far from clear how the parties would have addressed matters, in terms of what the parties might or might not have done to the figure of 51.41, had they been aware, at the time of contracting, of what Grain subsequently did in relation to blending, given the various differing possibilities that existed, such as changing the number to another number (e.g. 51.1), or to a reference to the Wobbe targeted from time to time, or to a reference to the actual Wobbe value that Grain would blend to, or to a reference to an average operational Wobbe (monthly or otherwise) or, indeed, leaving matters as they were.

Whilst Sonatrach maintained that the appropriate term to be implied if the Terminal blended to a Wobbe other than 51.41 was that, “the Co-Shippers would allocate the nitrogen costs between themselves under section D2.2.2(ii)(b) of the JSA on the basis of the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time", on occasions Sonatrach expressed itself in terms which would not appear to be identical. For example, at paragraph 83 of its Written Closing Submissions Sonatrach submitted that “it would have been evident to any officious bystander that, had the question been asked, the parties would have agreed that insofar as Grain were to blend at a level below 51.41, that lower Wobbe value would be used in the second formula in order to allocate nitrogen according to usage".

As to the fifth principle identified by Lord Simon in BP Refinery (Westernport) Pty Ltd v Shire of Hastings (1977) 180 CLR 266, 283 that an implied term must not contradict any express term of the contract, BP directed me to Lewison at 6.11. This represents a convenient summary of the applicable principles and the reasons why a term will not be implied if it is inconsistent with an express term or its general tenor:-

Sonatrach characterises its first implied term as a "modification to a phrase in an existing term of the contract" rather than (as BP submits) a contradiction of the express term (51.41). In this regard Sonatrach refers to cases in which the court has been willing to imply terms to modify existing terms of a contract, referring to the cases of Aberdeen City Council v Stewart Milne Group Limited [2012] SLT 205 at [31] to [32], Rathbone Brothers v Novae Corporate Underwriting Limited [2014] 2 CLC 818 at [85] and Spencer and another v Secretary of State for Defence [2012] 2 All ER (Comm) 480 at [91].

BP's riposte is that such cases are not in point. Cases such as Aberdeen v Stewart Milne and Rathbone v Novae illustrate the familiar concept that a provision which does not contain an explicit restriction on scope may be impliedly subject to a restriction, whilst the decision of Vos J in Spencer v Secretary of State turned on its own facts, and the carrying into the memorandum in that case of a process of arbitral review which had been referred to in the existing lease. In contrast, submits BP, the terms alleged by Sonatrach are simply inconsistent with the express constant of 51.41 in D2.2.2(ii)(b). It is not a question of modification but of inconsistency. On Sonatrach's implied terms the express term, and use of the constant 51.41 in the formula, has to be disregarded, and it would be no answer, say BP, to submit that "51.41" is to be read as "51.41 or the average monthly operational Wobbe if different" for, in such a formula ,"51.41" is not being used a constant at all - the overall element is simply based on the average monthly operational Wobbe (and the reality is that it is inherently unlikely that it will ever be precisely 51.41 in any event).

ISSUE 1: THE PROPER CONSTRUCTION OF CLAUSE D2.2.2(ii)(b)

The court identifies the intention of the parties by reference to what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language in the contract to mean, and the court does so by focussing on the meaning of the relevant words, in this case clause D2.2.2, and in particular clause D2.2.2(ii)(b), and that meaning is to be assessed in the light of (i) the natural and ordinary meaning of the clause, (ii) any other relevant provisions of the contract, (iii) the overall purpose of the clause and the contract, (iv) the facts and circumstances known or assumed by the parties at the time that the document was executed, and (v) commercial common sense, but (vi) disregarding subjective evidence of any party's intentions - see generally Arnold v Britton, supra, at [15].

The starting point, albeit only the starting point, is accordingly the natural and ordinary meaning of clause D2.2.2(ii)(b) of the JSA and its constituent elements. Sonatrach points out that no one part of that clause (specifically the figure of 51.41) is to be viewed in isolation and the clause is to be construed as a whole, as is clearly right. As recognised above it is, of course, also necessary to consider the clause in context set against the backdrop of the admissible factual matrix and the purpose of the clause so far as this can be established. However, when doing so, one starts by looking at what the clause actually states, and the normal and ordinary meaning of the clause and its constituent elements.

For ease of reference I will repeat Clause D2.2.2(ii) at this point:-

“Nitrogen Costs shall be allocated as follows, by reference to all relevant terms as they are defined in, and by measuring all relevant units in accordance with, the Services Agreement and the document entitled "GLNG - Agreed Network Entry Provisions" dated 20 June 2003 and initialled on behalf of GLNG and the Shipper:

(c)

the fixed component of nitrogen costs allocated by Grain to the Shipper shall be borne equally by the Co-Shippers; and

(d)

subject to Section D2.2.2(ii)(c) the variable component of nitrogen costs allocated by Grain to the Shipper shall be allocated in accordance with the quantity of gas Sent Out and the quality of the LNG delivered by each Co-Shipper (the delivering Co-Shipper) as follows:

X_MA = TC_M x TN_2A

TN_2A + TN_2B

Where:

X_MA is the cost of nitrogen to be borne by a delivering Co-Shipper with respect to a month (M);

TC_M is the total cost of nitrogen with respect to the month (M)

TN_2A is the quantity of nitrogen used by the delivering Co-Shipper in the month (M)

TN_2B is the quantity of nitrogen used by the other Co-Shipper in the month (M)

where each of TN_2A and TN_2B is calculated as the product of:

(iii)

gas Sent out in Gwh in the month by the delivering Co-Shipper (in the case of TN_2A) or by the other Co-Shipper (in the case of TN_2B); and

(iv)

The greater of X or Y, where:

X is the correction factor for incomplete combustion factor (tonnes of nitrogen N2 per GWh) calculated as follows:

((weighted average higher heating value of LNG as delivered by the delivering Co-Shipper (in the case of TN_2A) or the other Co-Shipper (in the case of TN_2B) in the month (M) x 6.35) – 250.01)

5.83

Y is the correction factor for Wobbe (tonnes of nitrogen N₂ per GWh) calculated as follows:

((weighted average higher heating value of LNG as delivered by the delivering Co-Shipper (in the case of TN_2A) or the other Co-Shipper (in the case of TN_2B) in the month (M) / (LNG molar weight) ^ 0.5) – 51.41)

0.62381

(emphasis added)

Applying the express language of the clause, the purpose of the clause is accordingly expressly stated to be to allocate the variable component of nitrogen costs allocated by Grain to the Shipper in accordance with (1) the quantity of the gas sent out and (2) the quality of LNG delivered by each of Co-Shipper applying the formula that then follows. The formula does allocate the variable component of nitrogen costs allocated by Grain to the Shipper in accordance with (1) the quantity of the gas sent out and (2) the quality of LNG delivered by each of Co-Shipper applying the formula that then follows. Sonatrach submits, and BP denies, that a wider commercial purpose is to derived from clause D2.2.2(ii)(b), and the JSA as a whole, namely to divide up nitrogen costs based on usage. Sonatrach's submission is considered in due course below. I would note at this point, however, that this is not stated on the face of D2.2.2(ii)(b), and the only reference to the quantity of nitrogen used appears in the definition of TN_2A and TN_2B within the formula.

D2.2.2(ii)(b) states a formula for calculating X_MA, the cost of nitrogen to be borne by a Co-Shipper. As a formula, it requires inputs. There are 3 inputs: TC_M, TN_2A, and TN_2B.

The gas sent out and X and Y are all variables. The gas sent out is simply a given (measured) number. X and Y are calculated numbers. They are calculated according to sub-sub-formulae. The Factor X and Factor Y formulae share a similar structure. The 1^st input for each is "the weighted average higher heating value of LNG as delivered...". That is a variable. In Factor Y (but not Factor X), the second input is also a variable ("LNG molar weight".) Then, in both Factor X and Factor Y, the three remaining inputs are "6.35", "250.01", and "5.83" in Factor X; "0.5", "51.41", and "0.62381” in Factor Y. On their face none of these are variables, they are each a numerical figure or value (i.e. a constant).

Sonatrach has to say, on its case as to construction, that one (and only one) of the numerical figures in one (and only one) of the formulae for Factor X and Factor Y, "means, and would have been understood by a reasonable third party to mean" a (complex) variable namely "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" or is a "proxy" for the same (see Defence para 38A).

As to the former, it is with the greatest of respect to Mr Harris, and Sonatrach, impossible to suggest (as a matter of language) that "51.41" means "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time". If there is anything to Sonatrach's submission it must be by reference to the latter i.e. that "51.41" is "proxy" for, or is some form of short-hand, or dictionary definition for, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time". At first blush, at least, that would appear to be a challenging submission based on the express language of clause D2.2.2(ii)(b).

Of course, and as will be addressed, one does not stop with a consideration of the language of the clause, and the clause is to be construed set against the backdrop of the admissible factual matrix and its commercial purpose. However, as Lord Neuberger made clear in Arnold v Britton at [17] reliance placed on commercial common sense and surrounding circumstances should not be invoked to undervalue the importance of the language of the provision which is being construed. As Lord Neuberger pointed out, the exercise of interpreting a provision involves identifying what the parties meant through the eyes of a reasonable reader, and, save perhaps in a very unusual case, that meaning is most obviously to be gleaned from the language of the provision. Unlike commercial common sense and the surrounding circumstances, the parties have control over the language they use in a contract. And, again save perhaps in a very unusual case, the parties must have been specifically focussing on the issue covered by the provision when agreeing the wording of that provision.

At least on the face of section D2.2.2(ii)(b), the figure of "51.41", like all other figures in the formula, is a numerical value, and as such is a constant. The meaning of a constant, as a noun, is well known as a matter of English language, and is defined by the Shorter Oxford English Dictionary in these terms, "MATH A numerical quantity which does not vary or is assumed not to vary. SCIENCE a number expressing a relation, property, etc. and remaining the same in all circumstances or for the same substance in the same conditions ". I consider either of these definitions (reflecting the ordinary and natural use of the word "constant") is apt to describe the numerical values to be found on the face of D2.2.2(ii)(b).

In testing whether "51.41" is a variable or a constant it is relevant to consider why "51.41" is used in the formula. As Lord Neuberger noted in Arnold v Britton the parties have control over the language they use in a contract, and the JSA is no exception, and unless this is that "very unusual case" (and I do not consider it is) the parties must have been specifically focussing on the issue covered by the provision when agreeing the wording of the provision. At one level at least, the reason why "51.41" (as a number) was chosen is common ground between BP and Sonatrach, as it would be to the reasonable person having all the background knowledge that would have been available to the parties at the time of contracting. Such a person would have known that 51.41 (Mj/m3) is the maximum Wobbe index value of gas entering the NTS and the LDZ as specified in the GSMR. Thus it appears from the face of the clause, viewed against the background knowledge that would have been available to the parties at the time of contracting, that the parties have deliberately chosen a constant of 51.41 because that is the maximum Wobbe index value of gas entering the NTS and the LDZ as specified in the GSMR.

A particular difficulty facing Sonatrach, it seems to me, addressing matters as a matter of construction, is that "51.41" is quite clearly a figure, and it is equally clear that the objective common intention of the parties is to include that figure into this part of the formula, and to do so because that constant is the maximum Wobbe index value of gas entering the NTS and the LDZ as specified in the GSMR.

If "51.41" is, as it appears to be on its face, a constant it is not "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" nor is it a proxy for that - it is quite simply a constant. Viewing clause D2.2.2(ii)(b) as a whole, it is clear that the parties have used a defined variable when they wish to use a variable (for example "the weighted average higher heating value of LNG as delivered) and a figure when they wished to use a constant. To substitute a variable for a constant is, at least at first blush, to do violence to the natural and ordinary meaning of the words used in the clause (and that is viewing the clause as a whole, and not focussing on individual uses of words such as "51.41" within that clause), and to change the contractual bargain between the parties (because the parties have chosen figures when they intend to use constants, and a variable when they wished to introduce a variable).

If the parties had wished to use a formula they could have stated such a formula. A similar generic submission is often made when considering competing contractual constructions. It is not always helpful as the court is construing the language of the contract and not what parties might have stated instead. However, I consider the submission has more force where, as here, the parties have used various formulae within the very same clause. It can be said with some force that the parties used a figure when they intended to include a constant, and included a variable where they intended that element to be a variable. Viewing matters objectively, the parties must have put their mind to whether constants or variables were appropriate and have drafted accordingly.

Additionally, this is not a case where it can be said that there are two alternative meanings to the clause and the word "51.41" used within it, nor can it be said that clause D2.2.2(ii)(b) is ambiguous. Clause D2.2.2(ii)(b) is clear on its face as to what it says. Thus Sonatrach must demonstrate that clause D2.2.2(ii)(b), and in particular "51.41" in the formula, means something other than a constant, and is the proxy for the precise formula alleged by Sonatrach. I say proxy for the precise formula alleged by Sonatrach as there are, on examination, a number of different possible formulae that it could be alleged "51.41" is proxy for (as addressed in due course below) and conceivably there are also other possible figures that 51.41 could be proxy for (for example 51.30 or 51.1). That is not an auspicious start to Sonatrach's case on construction.

It will be recalled that the sixth factor identified by Lord Neuberger in Arnold v Britton (at [22]) concerned a situation where an event subsequently occurs which was plainly not intended or contemplated by the parties judging from the language of the contract. In such a case "if it is clear what the parties would have intended the court will give effect to that intention" (emphasis added). Whilst Sonatrach submits that what occurred in terms of Grain’s Wobbe blending was not contemplated by the parties (although BP disagrees as to what would have been within reasonable contemplation), even if such a situation arose, it will be necessary to consider whether it is clear what the parties would have intended. Sonatrach has to say that it is clear that the parties would have intended the 51.41 element in the formula to mean "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time", whilst BP's riposte is to say that that was not clear at all and there are various possibilities as to how the parties might (or might not) have addressed such a situation had that been contemplated (ranging from doing nothing, to changing the constant to another constant, to inserting a variety of potential variables).

In opening Sonatrach submitted that the language of D2.2.2(ii)(b) had gone wrong, and the question for the court was "how wrong" referring to what it said were errors in the formula. It will be recalled that the second factor identified Lord Neuberger in Arnold v Britton [at 18] was that when considering the centrally relevant words to be interpreted, the less clear they are, or to put it another way, the worse their drafting, the more ready the court can be to depart from their natural meaning. He pointed out, however, that this does not justify the court embarking on an exercise of searching for, let alone constructing, drafting infelicities in order to facilitate a departure from the natural meaning - as he stated, if there is a specific error in the drafting it may have no relevance to the issue of interpretation which the court has to resolve.

The failure to include the 28.9 coefficient for the molecular weight of air is a good example of this for two reasons. First I do not consider it is properly to be characterised as reflecting poor drafting in terms of D2.2.2(ii)(b) or the JSA as a whole - it was, as I have already found, properly to be regarded as a simple typographical error. Secondly that error has no relevance to the issue of interpretation that the court has to resolve.

The second alleged "error" that Sonatrach relies upon arises out of the fact that the formula sometimes produced allocations in excess of 100%, and as has already been referred to, in August 2006 the parties agreed to a capping of allocations to 100% of costs in this context. I do not consider this allocation by the formula reflects an "error" in the drafting of the formula, or is indicative of bad drafting - it is a function of the operation of the formula when the Terminal operated at the level of Wobbe blending that it did. The parties' reaction (and agreement) as a consequence of the operation of the Terminal after the JSA was entered into, does not support a submission that D2.2.2(ii)(b) was badly drafted or that the court should more readily depart from the natural meaning of the words.

Sonatrach has also identified two other situations that it suggests reflect poor drafting in the formulae. The first is a suggestion that if a party imports more than one cargo in a month there is a problem with calculating the LNG molar weight in Factor Y. In fact, on analysis, what Sonatrach proposes, namely to use the weighted average LNG molar weight, simply reflects what the parties must have intended, and can be reached as a matter of construction (and is the approach the Agent has adopted in practice).

The second is a situation in which a party does not import any cargoes in a year. Whilst it is true the formula does not work in this situation, it is simply a scenario that the parties did not foresee and did not cater for. It is not an error in the drafting, nor does it suggest that D2.2.2(ii)(b) was badly drafted.

In the above circumstances I do not consider that Sonatrach can pray in aid any aspect of the drafting of clause D2.2.2(ii)(b) to submit that the clause is badly drafted and the court should more readily depart from the natural meaning that "51.41" is a constant and not a variable.

For the reasons I have given above I consider that "51.41" viewed as part of D2.2.2(ii)(b) as a whole, appears, on its face, to be a constant, and not a variable and does not (on its face) mean, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time”.

However, as stressed by Lord Clarke in Rainy Sky at [21] as referred to by Lord Hodge in Arnold v Britton at [76], the exercise of construction is essentially one unitary exercise in which 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.

In this regard the clause has to be viewed in the context of the JSA as a whole (and its other terms), the facts and circumstances known or assumed by the parties at the time the contract was executed and commercial common sense, but disregarding subjective evidence of any party's intentions.

There are two main limbs to Sonatrach's case as to the parties' intention by reference to the facts and circumstances known and assumed by the parties at the time the contract was executed and "what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language in the contract to mean" (Lord Hoffmann in Chartbrook at [14], as quoted with approval by Lord Neuberger in Arnold v Britton at [15]), and commercial common sense.

First that "the parties did not know precisely to what Wobbe value Grain would blend, but believed and operated for the purposes of drafting their contract on the basis that it would be close to 51.41" (Sonatrach Closing Submissions para 28). Secondly that it was the Co-Shippers, "common commercial purpose that variable nitrogen costs would be allocated between them on the basis of the amount of nitrogen used in relation their cargoes, judged by reference to their Wobbe value at the time of unloading." (Sonatrach's Closing Submissions para 32).

In relation to the first limb, and facts and circumstances known to the parties at the time of contracting, Sonatrach also rely upon the nature of the JSA (that it was not adversarial) and the fact that the JSA was a long term agreement requiring some flexibility to adapt to changing circumstances.

As quoted above Sonatrach's case in closing was that, "the parties did not know precisely to what Wobbe value Grain would blend, but believed and operated for the purposes of drafting their contract on the basis that it would be close to 51.41" (Sonatrach Closing Submissions para 28). In contrast BP submits that at the time of contracting, and having regard to all facts and circumstances known at the time of contracting, the parties would not have expected, or operated for the purpose of drafting their contract, on the basis that Grain would blend "close to 51.41" or "target 51.41" (as Sonatrach put it at para 20 of their Defence).

When Sonatrach originally pleaded its case it was put in absolute terms: "the Terminal would blend regasified LNG to the upper limit Wobbe value of 51.41". This was an assertion that the parties would have expected Grain to blend to a Wobbe value of 51.41 on a continuous basis. The first difficulty with that is that, as BP submit, it is inherently implausible and the parties would have known that. Blending is not so precise. Grain could not be expected constantly to hit a target with two decimal place precision even had it wished to do so (which it would not as a consequence of its obligations in terms of maximum Wobbe value in relation to putting gas into the national transmission network). The second (and even more fundamental) difficulty is that if Grain did blend to the upper Wobbe limit it was at risk of the limits in the GSMR being exceeded with the commercial (and potential criminal) (Footnote: ³) consequences that would follow - and any reasonable person in the parties' position would have known that, and that therefore Grain would not blend to 51.41 or "target" 51.41 (as it was put at paragraph 20 of the Defence), as is addressed further below in terms of the parties' reasonable expectations.

If Sonatrach had been right that the reasonable expectation of the parties at the time of contracting was that Grain could be expected to blend to 51.41 all of the time, this would have lent some support to the submission that "51.41" was synonymous with (a "proxy" for) an operational Wobbe (of 51.41). However, that clearly was not the reasonable expectation of the parties at the time - for the reasons already identified (and considered further below) any reasonable person in the party's position would have known that the Terminal would not aim to blend to the maximum limit itself - quite apart from the difficulties of doing so that would have been to risk commercial (and potential criminal) consequences.

Sonatrach's amended case was that Grain would blend LNG to "51.41, or as close thereto as possible" (Defence para 38B(b)(iv), and that Grain "did have some discretion to blend LNG to a Wobbe index value slightly lower than 51.41" (Defence para 38B(b)(iv), which mutated in Sonatrach's written Closing to " Grain would blend...close to 51.41". One difficulty with this amended case is it does not lie well with the submission that "51.41" is a proxy for "operational Wobbe", still less as pleaded with "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" (which suggests a changing variable over time).

There is another difficulty. If in fact it was the reasonable expectation of the parties at the time of contracting that Grain would blend to 51.41 or "close to 51.41" or "slightly lower than 51.41” (and BP says it was not) then there was every reason to contractually agree a constant of 51.41 - that would be simple, easy, and clear, and would obviate the need to attempt to ascertain what Grain was in fact blending to (generally or at any particular time). If Grain was blending to "51.41" or "close to 51.41" the parties might well consider that a figure of 51.41 would do perfectly well, especially absent any better information. It is to be borne in mind in this regard that, as Mr Winstanley confirmed, Grain were not forthcoming about what level they would blend to, and there was no track record as to how they would approach it (the Terminal was not operational at the time and there was no historic experience). It should be noted, however, that there is no correspondence that suggests there was any discussion between BP and Sonatrach about the use of 51.41, and Mr Stranks' evidence was that there was no discussion that it would be best or easiest to use a fixed number, or that this would be a simple and efficient mechanism for allocating costs.

An additional point is that if it transpired subsequently (post the signing of the JSA and the Terminal going operational) that Grain was in fact targeting a different level (such as 51.1) I do not consider that it follows that had the parties been aware of that at the time of contracting this would have caused them to insert "operational Wobbe". Indeed, Mr Winstanley said when cross-examined that, " If Grain had written to us and told us that they were going to be blending to a different number, then we would have had a conversation with Grain at the time about the basis of that number, and then whatever number we agreed, that would have gone into the cost allocation principles".

If such evidence is admissible (which is questionable as it is essentially subjective in nature as to what BP would have done), and leaving aside the rather hypothetical nature of the question (given that such conversation did not take place) this rather suggests that the parties would (or might) have gone for a different fixed value rather than a variable based on some form of operational Wobbe (though this, of course, depends on what they would have been told by Grain on this hypothetical). However even if they had gone for some form of variable there would have been various (different) possibilities - such as the Wobbe targeted from time to time, or a reference to average operational Wobbe or the like.

Indeed, in their Written Closing at paragraph 48, Sonatrach submitted that, "the number 51.41 in the formula simply reflected what both parties assumed (and what any reasonable third party would have assumed) was the blending target by Grain" (emphasis added). That is not the same as "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time”, which is no doubt why in their oral closing Mr Harris said that what he had put in paragraph 48 of Sonatrach's Written Closing was not his case. However, the fact that there are different permutations as to what the parties might have agreed (from fixed numbers to different variables) does rather suggest that it is not clear what the parties would have intended in this scenario (for the purpose of Lord Neuberger's sixth factor in Arnold v Britton), as I address below in due course.

Even once the parties knew that Grain was blending to lower levels the parties did not have information of exact levels blended to and, initially at least, only had information about levels being targeted. As addressed in the context of issues 3 and 4, as late as the Agent's 1 February 2007 reconciliation sent to the parties shortly before the 6 February 2007 Steering Committee Meeting, the Agent was only using fixed numbers (51.41, 51.1 and 51.25), and even at the time of a subsequent reconciliation on 14 July 2008 the Agent was using a figure of 51.1, and in its reconciliation on 8 September 2008 scenario 4 did not use the actual operational Wobbe as that had not been advised by the Terminal. Set against such backdrop it is not clear what the parties would have intended, had they been aware of what they subsequently became aware, at the time of contracting. Certainly I do not consider it to be clear that the parties would have substituted “the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” for the figure of 51.41, not least given the other possible constants or formulae that they could have agreed.

A further difficulty with Sonatrach's submission as to what was reasonably expected at the time of contracting is what is meant (or would have been understood) by, for example, "close to 51.41” when set against the backdrop of the minimum and maximum Wobbe values under the GSMR of 47.2MJ/m3 to 51.41MJ/m3. Mention had been made of Grain blending to 51.31 in the context of Mr Winstanley's discussions with the terminal manager Mr Belmore, albeit that figure was mentioned in early 2004 (and Mr Winstanley's subsequently focussed on the figure of 51.41). Subsequent to the Terminal going operational the Co-Shippers became aware that Grain was in fact aiming for a Wobbe of 51.1 for gas sent to the NTS and 50.9 for gas sent to the LDZ. These are all figures below 51.41 but very much towards the upper end of the range. It does not follow that the parties would necessarily have changed "51.41" in the formula to any particular figure or formula had they been aware at the time of contracting of the level(s) that Grain subsequently targeted at the upper end of the range. It has transpired after the JSA was entered into that nitrogen consumption is very sensitive to the Wobbe value and even small differences in Wobbe inserted in the formula make a large difference in terms of nitrogen consumption, but it cannot be assumed that the parties (or indeed a reasonable person having all the background knowledge the parties had) would necessarily have realised this at the time. Mr Winstanley had not, for example, run a working example through the formula - if he had he would have spotted the error in relation to the omitted 28.9 constant.

When cross-examined Mr Winstanley accepted the proposition that, "Grain didn't know what they would be targeting, but [he]thought the best available number was 51.41" and that he was, "hoping that [Grain] would get as close ... to the 51.41 as they could reasonably attain", though he "knew that they would be achieving a number beneath 51.41”. Mr Winstanley stated as follows in response to a question I asked:-

For his part, Mr Stranks said in his second statement that, "My understanding, as set out in Mr Winstanley's presentation given in January 2005 (by which time JSA negotiations had begun) was that the Terminal would blend gas efficiently targetting the upper permitted Wobbe limit contained in the network entry agreement (i.e. 51.41). I did not anticipate that GLNG would significantly over-inject nitrogen by targeting numbers as low as 51.1 and 50.9, nor do I believe that anybody else attending the JSA negotiations did."

Mr Stranks also asserted that, "Sonatrach agreed to the use of the formula on the basis that it would allocate costs in a manner that would reflect actual operations at the Terminal”. This latter evidence is clearly inadmissible as it looks to one parties' subjective intention - see, for example, the observations of Lord Neuberger in Arnold v Britton at [12] where he pointed out that even if evidence shows what both parties intended this is probably only admissible if there is a claim for rectification - see also Lewison at para 2.03 and the reference to what Lord Hoffman said in Chartbrook at [48], "English law...mixes up the ascertainment of intention with the rules of law by depersonalising the contracting parties and asking, not what their intentions actually were, but what a reasonable outside observer would have taken them to be."

To the extent that the understanding of particular witnesses as to what it was anticipated Grain would do in terms of nitrogen blending is admissible as an aid to construction, and it could only be to the extent that this may shed light on what a reasonable person having all the background knowledge which would have been available to the parties would have known and understood, such evidence has to be judged against the entirety of the evidence as to what the reasonable expectations of the parties were at the time of contracting. BP submits that the evidence taken as a whole does not support the proposition that parties would reasonably have expected Grain to blend to (precisely) 51.41 or "aim to” blend to "51.41", or to blend (or aim to blend) to "slightly below" 51.41 or "as close as possible" to 51.41. I will now turn to address the arguments they make in this regard.

First, I agree with BP's submission that no such expectation could reasonably have been based on operational experience, because there was no relevant operational experience. The Terminal was the first modern LNG importation, storage and regasification plant in the UK as Mr Winstanley and Mr Stranks confirmed at paragraphs 11 and 15 of their respective statements. The Terminal was not commissioned until July 2005, and did not become operational until September 2005. When the JSA was signed the Terminal was not running, and Grain had no recent practical experience of blending imported LNG to reduce the Wobbe to meet the specification in the GSMR, and Grain had no established operational practice as to how they were going to do it, as Mr Stranks accepted when cross-examined. As Mr Stranks put it, "When the JSA was being negotiated, it was all theoretical as we did not know how the Terminal would actually operate in practice" and "As the Terminal was to be the first modern LNG importation terminal in the UK, it was not possible, prior to start up, accurately to predict all operational realities over the twenty year duration of the [JSA]” (paragraphs 48 and 22 of his statement).

This uncertainty applied as much to Wobbe as to anything else. All that Mr Winstanley (who had a good relationship with Mr Belmore) could find out was that there would be an operational margin which might change over time, and that the highest Wobbe in Grain's contemplation was 51.31 (paragraph 63 of his statement) which he confirmed in his oral evidence that Grain would not commit to writing. Mr Winstanley incorporated this figure of "51.31" into a nitrogen requirements calculation which he sent to Sonatrach in March 2004, the table showing, "..the nitrogen required to inject into the NTS to meet the maximum Wobbe of 51.4MJ/Sm3 (actual nitrogen flowrate calculated based on achieving 51.31 MJ/Sm3)". Although Mr Way had no recollection (when he was cross-examined), I accept that the most likely interpretation is that this is the "nitrogen requirements study" referred to at a progress meeting between BP, Sonatrach and Grain on 15 May 2006. Subsequent modelling analyses carried out by Mr Winstanley (for the purposes of specifying the requirements of the GAN plant), prior to the JSA being agreed, were based on Grain blending to a Wobbe of 51.41. There is, however, no suggestion that Grain had changed its intentions (whatever they might be, and the parties were singularly lacking in information about that).

The second point made by BP is that there was no reasonable technological basis for expecting Grain to "aim at" 51.41 or achieve an operational Wobbe particularly close to it. For anything approaching precision in nitrogen blending for Wobbe, Grain would need to be able, with a good degree of certainty, to know the Wobbe of the gas being sent out (before adding any nitrogen), calculate the quantity of nitrogen needed to reduce that Wobbe to (or close to) 51.41 and limit the quantity of nitrogen injected to the calculated quantity.

Mr Winstanley explained when he gave his evidence that:

Mr Winstanley's evidence was that the Terminal is so set-up that the nitrogen injection points are upstream of the analysers which capture the composition of the send-out gas. So nitrogen is injected before the composition of the send-out gas is captured and before the Wobbe is calculated. The Wobbe is unknown at the moment of injection. Grain does not obtain the (estimated) Wobbe of the send-out gas and then calculate how much nitrogen to inject, rather Grain estimates the Wobbe of gas into which nitrogen has already been injected and then assesses whether or not to decrease (or increase) the injection rate. It is likely to err on the side of caution (towards over-injection) if there appears to be any risk of missing the GSMR specification.

Mr Winstanley also explained that Grain controls nitrogen injection manually, not automatically. And, in deciding whether to increase or decrease the rate of injection (and by how much) the operators consult a chart which states the effect on the Wobbe for each additional 0.1mt of nitrogen injected. The process is not based on very precise calculations. BP submits that there is no reason to think that this is something which could not reasonably have been anticipated in June 2005, and Mr Winstanley described it as "quite common".

Whilst I do not consider that a reasonable person having all the background knowledge that would have been available to the parties at the time of contracting would know the precise detail of the above matters (not least given that the Terminal was not in operation at the time), I do accept that there was no reasonable technological basis for expecting Grain to "aim at" 51.41 or achieve an operational Wobbe particularly close to it, and a reasonable person having all the background knowledge available to the parties would have known that Grain would seek to avoid any risk of breaching the GSMR specification, and would err on the side of caution in blending levels given the circumstances in which blending would be undertaken and the lack of any technological basis to allow the obtaining of a particular Wobbe level with precision.

It is common ground that the STA (October 2003) and the FLA (November 2004) pre-dated the JSA and that Grain's contractual obligations are relevant to how the parties might, at the time of contracting, have reasonably expected Grain to approach nitrogen blending for Wobbe. There is no provision in the STA which requires Grain to target 51.41 or to blend as close to it as possible. Instead, Grain's basic STA obligation was to ensure that send-out gas complied with the NEAs: GTC's C10.1.1. The NTS, NEA and the LDZ NEA both included the GSMR maximum Wobbe of 51.41. The minimum was 48.14 for the NTS or 47.2 for the LDZ. On the face of C10.1.1, Grain was free to select a Wobbe anywhere within the range.

Sonatrach relies upon the RPO provision in GTCs C10.2.2 and (amended) GTCs C12.1.1, arguing that an RPO would operate the Terminal in a reasonable and economically efficient manner, which would mean blending to 51.41 "or as close thereto as possible". This argument has led to a difference of construction between Sonatrach and BP as to whether the RPO provision applies to nitrogen blending. BP submits that, upon proper analysis it does not - and the proper construction is what a reasonable person having all the background knowledge which would have been available to the parties would have understood Grain's obligations to be. Accordingly, an entity in the position of BP and Sonatrach (who are to be taken to understand the existing contractual position) would not assume that Grain had an obligation to operate the Terminal as a RPO or have any obligation to use as little nitrogen as possible, and blend to 51.41 or as close thereto as possible.

In short BP submits that nitrogen blending, and Grain's right to nitrogen costs (including the cost of power to run the nitrogen plant) is governed by the FLA. The RPO provision in C10.2.2 and C12.1.1 does not apply, and the FLA itself does not contain any RPO provision or anything like it. I address this issue in due course below.

However, BP also submits that even if the RPO provision applied (contrary to its submissions on construction) it would not provide any reasonable basis for an expectation that Grain would target 51.41.

A consideration of these issues necessarily involves a consideration of the relevant contractual principles in some detail. It is necessary to undertake this exercise because Sonatrach places some considerable emphasis on its argument that an RPO would operate the Terminal in a reasonable and economically efficient manner, which would mean blending to 51.41 "or as close thereto as possible", and that a reasonable person having all the background knowledge which would have been available to the parties would have assumed that Grain was doing so.

Before turning to the relevant contractual provisions, I would note that Sonatrach put to Mr Winstanley that he knew that Grain had an RPO obligation which he confirmed, and that it was then put to him that it means Grain had to act reasonably and could not do what it liked in theory, which Mr Winstanley also agreed. I do not find such (abstract) evidence of what was encompassed by Grain's RPO obligation of any real assistance as it is not directed at the particular contractual provisions, and in any event a witness' views on contractual construction and what he (subjectively) understood a contract to mean would not be admissible even if this had been explored with him (which it was not).

The following exchange then occurred:

I bear in mind Mr Winstanley's evidence that he hoped that Grain would get as close as possible to 51.41 (though he did not elaborate upon his reasons for such hope), but I do not understand his answer to link his hope to Grain acting as a RPO, or that he had this in mind at the time of contracting. In any event it is ultimately a matter of contractual construction what the contractual position was.

I will first consider the situation in which (contrary to BP's submissions) the RPO obligation does apply when Grain is nitrogen blending. The first point BP (correctly) makes is that GTCs C10.2.2 and C12.1.1 do not impose an enforceable contractual obligation upon Grain to act as an RPO. The RPO provision operates by limiting the costs which Grain is entitled to recover. It qualifies Grain's C10.2.2 and C12.1.1 rights, rather than imposing any obligation which sounds in damages or is subject to specific performance. Sonatrach's argument must therefore be that the limitation would operate as a financial inducement for Grain to act like an RPO when blending. But as BP also point out, Grain would balance any such inducement against the severe potential detriment involved in exceeding 51.41. Equally there is no reason to assume that Grain would go out of its way to behave like an RPO simply in order to eliminate any risk that it might not be able to recover a portion of its blending costs under C10.2.2 and/or C12.2.1.

Secondly, even if it were assumed that Grain would have to act as an RPO when blending, the GTC definition is a flexible one in which regard is to be had to Grain's contractual obligations:-

BP submits that in acting as an RPO Grain would reasonably be expected to take into account:

I consider that in the context of the above matters (which I accept), an RPO could reasonably be expected to adopt a cautious approach to blending in which it avoided the upper end of the permitted Wobbe range in order to ensure that send-out gas never exceeded 51.41. It is to be noted that the variable costs of nitrogen are a comparatively small part of the overall Phase 1 costs. I accept BP's submission that an RPO could reasonably be expected to conclude that the balance between (i) a marginal increase in variable nitrogen costs and (ii) the potential downside of exceeding 51.41 should be struck in favour of incurring the former in order to minimise the risk of the latter. As Mr Stranks confirmed when cross-examined, an RPO would try to eliminate the risk of putting off-spec gas into the networks. I also agree that a cautious approach would be expected in the context of the first UK plant of its type, and the fact that Grain had no operational experience in relation to nitrogen blending, such that it would err on the side of caution in terms of the amount of nitrogen injected, particularly as an RPO could reasonably be expected to be subject to, and to recognise and take into account, the technological limitations in relation to nitrogen blending to ensure that the maximum Wobbe was not exceeded.

For the above reasons I do not consider that an RPO would reasonably be expected to target 51.41, or to consistently hit 51.41 or anything very close to it. However, and for the reasons which I will now give, I consider that properly construed, Grain was not under an obligation to act as an RPO in any event in relation to nitrogen blending concerning cargoes delivered into the Terminal by BP and Sonatrach. This requires a consideration and analysis of the relevant contractual terms in some detail. Before doing so I would note that I do not consider it assists Sonatrach to pray in aid the stance adopted by BP and Sonatrach in its submissions to Grain in 2013 (i.e. that Grain was obliged to act as an RPO and was injecting too much nitrogen) - ultimately the contractual provisions have to be construed to establish what they mean - and the Co-Shippers stance against Grain (and Grain's response thereto) do not assist in that regard.

Grain's basic C10.1.1 obligation is to ensure that send-out gas complied with the NEAs, meaning a Wobbe range of 48.14 to 51.41 for the NTS or 47.2 to 51.41 for the LDZ. Under the STA as executed in 2003, BP and Sonatrach were under a corresponding obligation to ensure that LNG had a Wobbe of 51.41 or less when it was discharged into the Terminal - see GTCs B8.1.1(b) (under which the Base Specification matched the NTS NEA), GTCs B8.1.2 (under which the contractual Specification was the Base Specification subject to any agreed variations) and GTCs B8.2.1 (under which LNG which BP and Sonatrach discharged into the Terminal had to comply with the Specification).

BP point out, as is the case, that compliance by BP and Sonatrach with their B8.1.2 obligation would not automatically enable Grain to comply with its C10.1.1 obligation. Boil-off/weathering meant that, even if LNG had a Wobbe of 51.41 when it was discharged, the Wobbe might have increased by the time the LNG was regasified into send-out gas. So in-tank changes to the properties of LNG which had been on-Specification when discharged might prejudice Grain's ability to comply with C10.1.1. This is recognised in GTCs C10.2. C10.2.1 acknowledges that, even if LNG was on-Specification when discharged, it might be necessary for Grain to take "blending measures" in order to ensure that send-out gas complied with the NEAs.

Under C10.2.2, Grain is entitled to recover the costs of "blending measures". But this was subject to a limitation: Grain was only entitled to recover "to the extent of the costs incurred (or which would have been incurred) by [Grain] acting as a Reasonable and Prudent Operator in taking such measures". By C10.2.2, the RPO limitation relates to the costs of blending measures, i.e. measures which Grain takes under C10.2.1.

C10.2.4(a) introduces the concept of "shipper-specific blending measures". Shipper-specific blending measures are measures which are taken pursuant to an arrangement for a variation to the Base Specification, i.e. to such arrangements as referred to in B8.2.1.

C10.2 therefore refers to two sorts of measures, and they are distinct, first blending measures, which are taken under C10.2.1 and shipper-specific blending measures, which are taken under B8.2.1 arrangements. The fact that the two are distinct is reinforced by C10.2.6(b), which provides that the recovery of costs under C10.2 is "without prejudice to and in addition to" any amounts for which a Shipper may be liable pursuant to any B8.2.1 arrangements.

The costs of blending measures are recovered under C10.2 (and so are subject to the RPO limitation in C10.2.2) while the costs of shipper-specific blending measures are recovered under the relevant B8.2.1 arrangements (and so are not subject to the RPO limitation). When the STA was executed, there were no B8.2.1 arrangements varying the Base Specification and no agreement for shipper-specific blending measures. So Grain could only have recovered costs of blending with nitrogen for Wobbe under C10.2, subject to the RPO limitation.

But the contractual position changed with the FLA. STA Schedule 2 Clause 4 set out a mechanism for BP and Sonatrach to request a revised Specification which varied from the Base Specification. The FLA set out the basis on which this mechanism was implemented (Clause 1.1-1.2). That basis includes an increased maximum Wobbe of 52.08 during an interim period and 53.01 thereafter. The FLA is an arrangement within GTCs C8.2.1, as referred to in C10.2.4(a). The FLA also includes shipper-specific blending measures, as referred to in C10.2.4(a). Annexe A, Clause 2 is headed "Specific Blending Measures". It cross-refers to Appendix 1, which in turn cross-refers to the NSA.

The FLA includes shipper-specific blending measures. The increased maximum Wobbe only applies to BP and Sonatrach's obligation under GTCs B8.2.1, which attaches to LNG discharged into the Terminal. It does not affect Grain's C10.1.1 obligation in relation to send-out gas. The 51.41 maximum Wobbe for send-out gas is imposed by the NEAs, which are contracts between Grain and the networks, and by the GSMR. The maximum Wobbe for send-out gas cannot be relaxed - if BP and Sonatrach discharged LNG into the Terminal with a Wobbe above 51.41 and Grain regasified it and sent it out without reducing the Wobbe, Grain would be in breach of the NEA and the GSMR (as well as GTCs C10.1.1) and at risk of gas being shut-in and the Terminal tripped as Mr Winstanley explained in his witness statement and when giving evidence. So if BP and Sonatrach were going to be entitled to discharge LNG with a Wobbe up to 53.01 (before the impact of in-tank boil-off/weathering) Grain would need the capacity to blend, in order to avoid trips and being in breach of contract and criminal law.

Whilst BP and Sonatrach are (by virtue of the FLA) permitted to deliver LNG into the Terminal with an increased maximum Wobbe, STA Schedule 2, Clause 4.3 provided that Grain would be entitled to recover the costs of any measures taken in relation to an increase in Wobbe. This is carried through into FLA Annexe A, Clause 6 which refers to costs which are recoverable pursuant to STA Schedule 2, Clause 4.3. The relevant costs are described in more detail in Appendix 1, Clauses 3-4 and Appendix 3. They include the cost of power for running the nitrogen plant, which is to be metered separately from other power.

Applying the above contractual provisions, I consider that BP is right to say that nitrogen blending under the FLA does not constitute blending measures within GTCs C10.2.1. It constitutes shipper-specific blending measures within C10.2.4. As such, Grain's right to recover costs is not subject to the C10.2. RPO limitation.

The fact that nitrogen blending under the FLA is outside C10.2.1 is also confirmed elsewhere in the FLA. By Clause 3.1, BP and Sonatrach acknowledge and accept three Notices and agree that they comply with the GTCs. Notice 2004/01 is headed "Blending Measures (Section C10.2)”. It states that the purpose of the Notice is to set out "certain details relating to the application (as at the date of this Notice) of Section C10.2". Paragraph 1 then states that, "for the purposes of Section C10.2.1, the prevailing blending measures taken by GNLG are as described in Part 1 of the Schedule below". Paragraph 2 refers to costs recoverable under C10.2.2, again by cross-reference to the Schedule. The "Prevailing Blending Measures" and associated costs described in the Schedule relate to a recondenser and to propane enrichment. (Mr Bax's evidence - at paragraph 27 of his statement - is that boil-off is sometimes blended with propane before injection into the LDZ). The Notice does not say anything at all about nitrogen blending.

However, nitrogen blending does appear in Notice 2004/002. This Notice refers to C10.3. C10.3 is a new provision which was inserted when the GTCs were amended by the FLA (see FLA Clause 2). The new C10.3.1 states that Grain may contract with a third party for the provision of either blending measures (pursuant to C10.2.1) or shipper-specific blending measures (pursuant to C10.2.4(a)). Clause 1 of Notice 2004/002 states that "For the purposes of Section C10.3.1, [Grain] has arranged shipper-specific measures with a third party". Clause 2 states that the arrangements are described in the Schedule below. The Schedule refers to the NSA.

I accept that the Notices embody the distinction between blending measures and shipper-specific blending measures. Propane blending is a blending measure within C10.1. Costs recovery is subject to the RPO limit in C10.2.2. Nitrogen blending pursuant to the FLA and NSA is a shipper-specific blending measure within C10.2.4(a). Costs recovery is not subject to the RPO limit.

In addition to the new C10.3, the original C12 entitled Grain to recover power costs. In the amended C12.1.1, recovery of power costs is subject to an RPO limit (as in C10.2.1). However, by C12.2.2(a), costs which are recoverable under an agreement within B8.1.2 are excluded from C12. This mirrors the "without prejudice" provision in C12.2.6(b). The FLA is an agreement within B8.1.2 and the FLA specifically entitles Grain to recover the cost of power for running the nitrogen plant. This means that, under C12.2.2(a), nitrogen power costs are excluded from C12, so that they are not subject to the RPO limit in C12.1.1.

So the RPO limits in the GTCs do not apply to the recovery of nitrogen costs (including power) under the FLA. And the FLA itself does not impose any RPO or similar limit or obligation. Accordingly, Grain is not subject to any RPO obligation when it performs nitrogen blending, and its right to recover nitrogen costs is not subject to any RPO limit. Provided that it does not blend below the NEA minimum (which would be a breach of C10.1.1), it has a discretion, within the parameters of the NSA, as to how much nitrogen it chooses to inject, and is entitled to pass the costs on to BP and Sonatrach. Although this very much weighs the contractual arrangements in Grain's favour, this was no doubt the price to pay for increasing the maximum Wobbe limit for LNG discharged into the Terminal (in circumstances where Grain generally had the stronger bargaining position).

Sonatrach submits that shipper-specific blending measures relate to blending "off-spec cargoes" with a Wobbe above 51.41, and that Grain has "no discretion" in relation to shipper-specific blending measures because it is only entitled blend down from the actual Wobbe to 51.41. It appears to be submitted that once Grain started blending to a level below 51.41, any shipper-specific blending measures become "spent", so that blending below 51.41 falls within C10.2. I agree with the points made by BP in response to these points:-

Sonatrach also submits that, "in circumstances where neither Grain nor BP nor Sonatrach nor any body else has ever raised this point, it is also irrelevant... because plainly nobody ever thought about this distinction at the time of contracting". If Sonatrach is there submitting that this is because BP's interpretation is wrong then it is open to Sonatrach to argue that - although for the reasons I have given I reject Sonatrach's submission. However I consider that Sonatrach is wrong in its submission that the proper construction of the relevant provisions is irrelevant when considering the construction of D2.2.2.(ii)(b), which is an objective not subjective exercise. The relevant factual matrix consists of the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of contracting and that includes the surrounding contractual framework properly construed (which is itself to be construed objectively). Whether or not either BP or Sonatrach raised the point is not, in any event, in point. A reasonable person having all the background knowledge which was available to the parties would be aware of the terms of the agreements that had already been entered into.

I therefore conclude that, for the above reasons, nitrogen blending, and Grain's right to nitrogen costs (including the cost of power to run the nitrogen plant) is governed by the FLA, the RPO provision in C10.2.2 and C12.1.1 does not apply, and the FLA itself does not contain any RPO provision or anything like it.

However whether or not the RPO obligation exists in relation to nitrogen blending, and whether or not that would have been known to a reasonable person having all the background knowledge that would have been available to the parties, for the reasons I have found I do not consider that an RPO would, in any event, reasonably be expected to target 51.41, or to consistently hit 51.41 or anything very close to it, rendering the issue of construction as to whether the RPO obligation exists, somewhat academic.

Sonatrach relies upon the fact that the parties were aware of Grain's November 2004 draft CAPs (directed at a future multi-shipper environment) at the time of contracting. It had identified that it would allocate variable nitrogen costs between Shippers on the basis of the extent to which their cargoes' Wobbe values on unloading exceeded 51.41. From this Sonatrach submits that "Grain thereby communicated to both Co-Shippers that it would blend LNG cargoes received with a Wobbe index in excess of 51.41 to 51.41" (Defence paragraph 38B(b)(ii). BP submits that this simply does not follow - the CAPs were concerned with allocation of nitrogen costs, not any statement as to the level to which it would blend.

I have already identified Grain's Cost Allocation Principles (CAPs) in Section D above. The contractual background to these CAPs is that, under the revised GTCs, Grain was to establish General CAPs, applicable to all of the costs which it was entitled to recover under the GTCs (D2.2), Specific "further CAPs" in relation to costs of blending measures under C10.2.1 and Specific "further CAPs" in relation to power costs: C12.2.1.

Grain circulated draft CAPs in November 2004. The draft was structured in 3 parts, an Introduction and Parts A, B and C. The introduction stated that Part A contained general CAPs pursuant to D2.2, Part B contained "further CAPs" pursuant to C10.2.1 and C12.2.1 and Part C "sets out certain other related matters". Part C referred to the GAN plant and the possibility that two or more Shippers might share the benefit of a single plant or of the same trucked-in LIN. This was a hypothetical scenario in November 2004, as the project was still at Phase 1, the only Phase 1 shippers were BP and Sonatrach (treated as one shipper) and the plant was not yet operational.

The draft CAPs went on to state that the scenario in which two or more Shippers shared nitrogen facilities did not fall within the "further CAPs" which were required under the GTCs. But Grain was minded to cater for the scenario anyway. Its "current approach" was that monthly costs of shared nitrogen facilities would be allocated among multiple Shippers according to a formula. I have already quoted the formula in Section D above, but in summary, one element of the formula, "V", was "the volume of nitrogen required by each Shipper". This variable was to be calculated by a sub-formula. The sub-formula multiplied a Shipper's monthly send out by (i) a constant, and (ii) a variable, "W". "W" was "the extent to which to the wobbe of the shippers last unloaded LNG exceeds the maximum permitted wobbe in the Gas Quality Specification" (i.e. 51.41 the maximum in the NEAs and GSMR).

The formula for V in the draft CAPs is simpler than the formula for TN_2A/ TN_2B in D2.2.2(ii)(b) (as Mr Stranks confirmed when cross-examined). It refers only to Wobbe, not to ICF. In D2.2.2(ii)(b) terms, it allows for Factor Y but not for Factor X, so there is no "greater of X or Y" element, and it simply refers to the Wobbe of a Shipper's last cargo whereas D2.2.2(ii)(b) involves a more detailed calculation, based upon the weighted average higher heating value of all of the LNG delivered by a Shipper over the relevant month (or year). It is right, however, that both formulae involve comparing the variable quality of a Shipper's LNG with the constant 51.41.

BP agreed with the proposal in Part C of the draft CAPs. In agreeing, BP noted that the proposal involved "allocating nitrogen costs in proportion to the amount of gas sent out by a Shipper in the month and the quality of the last cargo delivered (measured by the extent to which the wobbe exceeds the maximum permitted wobbe". That was an accurate summary of what the formula did.

At paragraph 37 of his witness statement Mr Stranks suggested that BP and Sonatrach "were in agreement that the cost allocation principles in the JSA should mirror the draft cost allocation principles issued by GLNG for use in Phase II (i.e. the future multi shipper environment at the Terminal.” However there is no evidence of such an agreement, the provisions are not the same (as has been identified), and D2.2.2(ii)(b) is a more sophisticated formula.

The purpose of the CAPs is to allocate, between different Shippers, the nitrogen costs which Grain has incurred. The first variable which goes into the overall formula is "TC", the total monthly nitrogen costs. That is a given figure. It depends upon what blending Grain has in fact performed, which determines the quantity of nitrogen consumed and the cost. The CAPs formula simply divides that figure, whatever it happens to be, between different Shippers. In doing so, the formula uses a sub-formula which contains, as a constant, the 51.41 maximum Wobbe.

However, the fact that the formula uses a sub-formula which contains, as a constant, the 51.41 maximum Wobbe does not mean that Grain will only blend LNG down to 51.41 (or as close as possible, or will target 51.41, or anything of the like). I do not consider that the CAPs can be construed as being directed at defining or limiting what nitrogen costs Grain will incur. They have a different purpose which is to divide up whatever nitrogen costs Grain has in fact incurred (as Mr Stranks agreed in cross-examination). As Sonatrach put it in cross-examining Mr Winstanley, Grain uses the quantity of nitrogen that it uses. Changing the number in a costs allocation formula does not change that quantity, or the quantum of the costs which Grain passes on to Shippers under CAPs and/or STAs.

Grain did not undertake in the STA to target 51.41. I accept that it is inherently unlikely that in a document concerning a hypothetical future scenario Grain was thereby representing that it would blend to 51.41, and I do not consider a reasonable person having the background knowledge of the STA and the draft CAPs would have understood Grain to be indicating that it would only blend LNG down to 51.41, still less that it would do so in a Phase 1 environment. Of course in the event (as subsequently transpired) Grain did not use 51.41 in its calculations but a different constant, referring in its letter of 11 May 2011, to 50.9 as, "the benchmark value used within the current nitrogen cost allocation rules."

In the above circumstances I do not consider that the terms of the draft CAPs were such that "Grain thereby communicated to both Co-Shippers that it would blend LNG cargoes received with a Wobbe index in excess of 51.41 to 51.41”, nor would that have reflected what a reasonable person having all the background knowledge which would have been available to the parties have thought having regard to all the matters that I have already identified above.

I do not consider that events after the Terminal became operational, and the parties learnt that Grain was blending to levels below 51.41 are of any real assistance in assessing what the parties believed Grain would be targeting at the time they entered into the JSA. It is clear that following receipt of Grain's March 2006 Monthly Report the parties were aware that Grain targeted an "optimum" Wobbe of 51.1 and often injected more nitrogen than the "optimum" quantity. As will be addressed in the context of Issue 3, there were also discussions between BP and Sonatrach in the summer of 2006 as regards the use of 51.41 in the formula and that in this regard Mr Way produced a memorandum on 25 July 2006 for Sonatrach on nitrogen allocation matters in which he stated that, “BP are proposing the current wobbe correction factor of 51.41 in the JSA be changed to be a monthly weighted average of the Wobbe of the gas sent out from the terminal as advised by Grain”. However, there is no suggestion that the parties immediately required that Grain must blend to 51.41 or considered (or suggested) that 51.41 in D2.2.2(ii)(b) meant, or was proxy for, "monthly weighted average of the Wobbe". Indeed, the memorandum itself acknowledges that what was being proposed would be a "change" i.e. that it does not reflect what the formula in D2.2.2(ii)(b) provided for and a contractual amendment would be required to the formula. In the event there is no evidence those discussions were progressed in advance of the September Steering Committee meeting (as is addressed in the context of Issues 3 and 4).

If anything (and as BP submits) the parties reaction to discovering, from Grain's March 2006 Report, that Grain regarded a nitrogen injection rate which targeted 51.41 as a "minimum" and a rate which targeted 51.1 as the "optimum", and that the quantity of nitrogen which Grain injected quite often exceeded the "optimum", was distinctly muted and it did not, at that time, lead to calls that Grain target "51.41" and nothing less (or reflect any suggestion that the parties had understood that Grain blended to 51.41).

The flavour of the parties' initial response can be seen from the following points (highlighted by BP in their closing):

Thereafter it is clear that BP and Sonatrach continued to be dissatisfied with the amount of nitrogen that Grain was using, and in due course in its 11 May 2011 letter Grain indicated that it would move to a variable factor based on the outturn monthly flow weighted average Wobbe number. In this letter Grain also noted that "the adoption of a fixed factor of 51.41mJ/m³ for allocating nitrogen costs was not appropriate in that, whilst it would be aligned with the upper limit specified in the Gas Safety (Management) Regulations 1996, it would not be possible for Grain LNG to operationally ballast to this level due to measurement errors and the limitations imposed by the alarm levels set by NGG for gas delivered to the NTS". All Shippers agreed Grain's proposal to modify the existing CAPs to use an operational Wobbe value. But Shippers do not appear to have asserted that the operational Wobbe should be much closer to 51.41; or to have complained that they had signed STAs on the understanding or in the expectation that Grain would target 51.41; or to have complained that the alarm levels set by the NTS off-takers were an entirely unexpected constraint; or to have challenged Grain's statement that it would not be possible to operationally ballast to 51.41.

In due course BP and Sonatrach did make a formal contractual complaint to Grain in their letter of 6 August 2013 stating that the Terminal blending to 51.1 was not a reasonable level of blending and that Grain was not entitled to recover the costs of blending below 51.41 (though without suggesting how 51.41 might be achievable notwithstanding the measurement and NTS alarm level constraints). In the event in a subsequent presentation in December 2013 Grain repeated that the Terminal operated to targets which were based on the NTS alarm, and as Mr Wood confirmed when he gave his evidence Grain continued to blend below 50.9. Grain has never changed it stance in relation to nitrogen blending.

I do not consider that BP and Sonatrach's immediate reaction lends any support to the suggestion that anyone could reasonably have expected, at the time the JSA was entered into, that Grain could, and would, blend to 51.41 or target 51.41.

In all the circumstances, and for the reasons I have given above, I do not consider that a reasonable person having all the background knowledge which would have been available to the parties would have understood that Grain would blend to "51.41" or "close to 51.41" at the time the JSA and the formulae in clause D2.2.2(ii)(b) were agreed. The most powerful factor in this regard is, in my opinion, the fact that any such person (and BP and Sonatrach) would know that 51.41 was the maximum value for gas entering the gas transmission system, and Grain (whether acting as an RPO or not), would not risk exceeding this value and so would err on the side of caution and therefore blend to a level below this (not least given the technical difficulties of blending to a precise level). What that level would be was not known at the time as the Terminal was not operational and Grain was not forthcoming about its blending intentions. However, I base my conclusion on all the matters that I have addressed above.

In reaching this conclusion I have borne well in mind the evidence of Mr Winstanley and Mr Stranks that I have referred to as to their beliefs at the time of contracting, but ultimately regard is to be had to all the background knowledge that a reasonable person would have had which would have been available to the parties, and on the basis of that knowledge, such a person would not have understood that Grain would blend to "51.41" or "close to 51.41" at the time the JSA and the formulae in clause D2.2.2.(ii)(b) were agreed.

In such circumstances what a reasonable person having all the background knowledge which would have been available to the parties would have understood that Grain would blend to, lends no support for the submission that "51.41" either means, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” or is proxy for the same.

Furthermore, and for the reasons already identified, even if a reasonable person had understood that Grain would blend "close to 51.41” (and what is "close" to 51.41 would itself have been unclear contemporaneously), or indeed had been made aware of what Grain subsequently targeted, it is not at all clear what the parties would have intended, or what (if anything) different they would have agreed, whether by reference to another constant or a variable.

Before leaving the facts and circumstances known to the parties at the time of contracting I will briefly address the two further points made by Sonatrach under this heading in closing. The first is to note that the JSA is not adversarial in nature, in the sense of one party seeking to extract a commercial advantage at the other's expense (as Mr Winstanley accepted when cross-examined). Whilst it is right that BP and Sonatrach were not in an adversarial relationship (but rather a relationship of Co-Shippers viz Grain), I do not consider that this assists in relation to the issue of construction that arises. The JSA set out the terms for sharing capacity and (amongst other matters) how costs for which they were liable to Grain were to be allocated between themselves. Ultimately the court is concerned with the construction of a particular clause which allocates variable nitrogen costs in accordance with the quantity of the gas sent out and quality of the LNG delivered by each Co-Shipper in accordance with a formula which has to be construed. It is a question of what the clause means. The level that Grain blends to could have led to a benefit to either party (the precise nature of the cargoes they would supply was not known with certainty at the time of contracting), and it is not a question of one party seeking to extract a commercial advantage from the other but rather what the clause, properly construed, means. It is not, in any event, the role of the court to re-write the parties' contractual bargain or to substitute for that which was agreed something that was not agreed or intended, even if that would be regarded by the court as reasonable or fair in the context of a non-adversarial relationship (if that is what Sonatrach is inviting the court to do).

The second point made by Sonatrach relates to the nature of the JSA. It is said that the JSA, as a long-term agreement, required some flexibility to adapt to changing circumstances, especially in circumstances where the Terminal was the first of the new import terminals and there was no operational history on which the parties could rely when planning their relations. Reference is also made to the fact that Clause D.2 of the JSA, dealing with cost allocation, expressly refers to the provisions it contains as “principles” (see, for example, D.2.2.1) which it is said reflects their flexibility and also the dominance of the principle over the mechanics of the particular formula. Sonatrach then submits that, against that background, “a court [may] take into account that, by reason of the changing conditions affecting such a contract, a flexible approach may best match the reasonable expectations of the parties” (per Lord Steyn in Total Gas Marketing Ltd v Arco British Ltd & Ors [1998] CLC 1275, also see Leggatt J in Yam Seng PTE v International Trade Corp [2013] 1 All ER (Comm) 1321 at [142]).

It is correct that the JSA is a long term agreement but it must be construed in accordance with its terms. Equally, and notwithstanding the use of the word "Principles", the detailed provisions of D2.2 define how costs are to be allocated in some detail and by reference (in many cases) to formulae, and as such are to be applied. There is no warrant to disregard the wording of a particular formula (or as Sonatrach puts it the mechanics of a particular formula) in favour of some wider principle. Whilst in some cases a flexible approach may best match the reasonable expectation of the parties, the JSA had its own mechanisms for discussion (through the Steering Committee) and agreement (through variation), and any flexibility in approach is in any event not a warrant for substituting for that which was agreed something that was not agreed or intended, especially in circumstances where it is not clear what the parties would have intended or agreed, had they contemplated the approach to blending that was in fact adopted by Grain.

Sonatrach makes three main submissions in relation to what it says was the commercial purpose of the JSA and D2.2.2(ii)(b) and commercial common sense when construing D2.2.2(ii)(b):-

It says that BP's construction of D2.2.2(ii)(B) does not give effect to the parties’ common commercial purpose, and that the formula, as construed by BP, will or may have the effect identified in the second and third points.

BP says that D2.2.2(ii)(b) makes its purpose clear in its openings words, “the variable component of nitrogen costs... shall be allocated in accordance with the quantity of gas Sent Out and the quality of the LNG delivered by each Co-Shipper" [C1/1/74], and the objective is an allocation by reference to the quantity of each shipper's gas, and the quality of each Co-Shipper's LNG. The formula does precisely that applying the elements of the formula that the parties contractually agreed. The consequences that arise when the contractually agreed formula is applied (including the potential consequences identified in Sonatrach's second and third points) are not the result of any error in the formula, or an uncommercial construction, but simply the fact that (in the event) Grain blended not only to a Wobbe level some decimal points below 51.41, but to a Wobbe level below that of both Sonatrach cargoes and BP's cargoes on discharge into the Terminal.

BP submits that Sonatrach is simply wrong when it asserts that both as a matter of construction, and as a matter of commercial purpose, that the parties’ contractual intention is that variable nitrogen costs would be allocated pro rata between them on the basis of, "the amount of nitrogen used in relation to their cargoes" i.e. according to usage. There is accordingly a fundamental difference between the parties as to what the purpose of D2.2.2(ii)(b) is, and whether the purpose of D2.2.2(ii)(b) is to allocate nitrogen costs pro rata according to use as Sonatrach alleges, or whether the purpose is to allocate the variable nitrogen costs allocated by Grain to the Co-Shippers in accordance with the quantity of the gas sent out and the quality of the LNG delivered by each Co-Shipper in accordance with the formula that follows.

Turning first to the language of D2.2.2, Sonatrach relies upon the opening words of D2.2.1, that "the following principles shall apply, and the Agent shall apply the following principles, in determining how any cost... payable to... [Grain]... in relation to the usage of the Terminal by the Co-Shippers are to be borne, or allocated between, the Co-Shipper” (Defence para 42) (emphasis added). The reference to "usage" is clearly a reference back to "any cost... payable". It does not relate forward to "borne or allocated". The purpose of the emphasised text is to identify those costs which are subject to allocation under D2.2.2. The meaning is that costs which are payable in relation to the usage of the Terminal by the Co-Shippers shall be allocated in accordance with D2.2.2. The wording does not mean that D2.2.2. is intended to allocate (unspecified) costs pro rata according to each Co-Shipper's use of the Terminal.

That the words relate back and not forward, and that there is no over-arching principle of allocation according to usage is also shown by the provisions of D2.2.2(i)(a) (electricity) and D2.2.2(ii)(a) (nitrogen) which allocate the fixed costs equally between the parties, which has nothing to do with usage (it is to be borne in mind that some cargoes require more nitrogen than others - for example the lean Trinidadian and Egyptian LNG which BP was importing in 2006 needed less nitrogen than Sonatrach's richer Algerian cargoes (as Mr Way noted in his memorandum of 25 July 2006)). It was open to the parties to allocate fixed nitrogen costs (principally the annual fee due to AP under the NSA) according to usage, as occurred in relation to propane costs (D2.2.2(iii)(a)).

Sonatrach also refers to the allocation of costs in relation to other variable costs (electricity, propane and other (unspecified) costs) in D2.2.2. I do not consider that any real assistance is provided by a consideration of these separate provisions in relation to the proper construction of D2.2.2(ii)(b) and the formulae contained therein, and its commercial purpose, so far as variable nitrogen costs are concerned. The variable component in relation to electricity costs is to be borne by Co-Shippers "pro-rata to their Sent Out Qualities in that month", the variable nitrogen costs are to be allocated in accordance with the formulae and sub-formulae contained in D2.2.2(ii)(b), the variable propane costs are to be allocated in accordance with D2.2.2(iii)(c)(I) or the formulae and sub-formulae contained in D2.2.2(iii)(c)(II) (as applicable), whilst other fixed costs are borne equally, and "in other circumstances, in accordance with the appropriate method of allocation as agreed between the Co-Shippers and notified to the Agent by the Co-Shippers. The appropriate method of allocation agreed between the Co-Shipper shall take into account the relevant basis of usage of the service or commodity or cause of the cost as the case may be" (D2.2.2vi). Whilst the provision in relation to "other costs" does provide that the appropriate method of allocation agreed between the Co-Shippers shall take into account the relevant basis of usage of the service - so there is here a link to usage, in relation to nitrogen and propane the parties have set out detailed formulae to allocate costs, and in the case of nitrogen this is based on the quantity of gas sent out and the quality of the LNG delivered.

The JSA is a very detailed agreement, and the provisions as to the allocation of variable nitrogen costs contain detailed principles, with allocation to be in accordance with the quantity of gas sent out and the quality of the LNG delivered. If the parties had wished to agree that variable nitrogen costs be allocated between the parties based on the amount of nitrogen actually used by a particular cargo they could have done so (albeit that would not have been a straight-forward exercise given the effects of weathering, boil-off and the like). Instead they chose the allocation, and associated formulae, that is set out in clause D.2.2.2(b)(ii).

Turning then to the language of D2.2.2(ii)(b), it does not state that variable costs are to be allocated according to the nitrogen used, and indeed it does not purport to undertake such a calculation either. As BP submitted, if you want to know how much gas is being used on a particular cargo of regasified LNG at the time that it is being sent out, then what you need to know is what is the Wobbe of the LNG at the time that it is regasified and before any nitrogen is added to it. The other thing you need to know is what is the Wobbe of the send-out gas. That is because the amount of nitrogen required for that cargo is the amount of nitrogen required to bring a Wobbe of the regasified LNG, whatever that is, down to the level of the Wobbe of the send-out gas, and the formula, does not refer to either of those. So whatever the formula is dealing with, it is not calculating the amount of nitrogen actually used on a particular cargo.

In the context of whether D2.2.2(ii)(b) is concerned with usage of nitrogen, Sonatrach relies upon the fact that D2.2.2(ii)(b) defines TN_2A and TN_2B as "the quantity of nitrogen used by the... Co-Shipper in the month". However, TN_2A and TN_2B are simply calculated values, determined by a fixed formula.

Equally, the submission that allocation according to usage is embedded in D2.2.2(ii)(b) (and specifically in Factor Y) does not fit with what Factor Y actually does. BP illustrates this with a simplified example (utilising whole numbers for ease of illustration):-

In fact, the position is more complex than BP's simple example, as BP pointed out in its Written Closing. Shipper A's LNG changes for reasons beyond boil-off/ weathering. It is co-mingled with other Shipper's LNG (even in Phase 1, to some extent). Title passes to Grain on discharge (GTCs B7.1) and Shipper A's only right is to the delivery of an equivalent quantity as part of a co-mingled stream: C1.3.4 & 2.2.2. This means that the gas sent out on Shipper A's nominations is not the regasified form of the LNG which Shipper A discharged. That makes a calculation of nitrogen usage by reference to the quality of Co-Shipper A's send out gas impossible in literal terms. Sonatrach might say that the parties recognised that they could not know the Wobbe of a Co-Shipper's regasified LNG, so they decided to use the Wobbe of the Co-Shipper's discharged LNG instead. But then it could be said in return that the parties also recognised that, at the time of contracting, they could not know what the eventual Wobbe of a Co-Shipper's send-out gas would be (for the reasons that have been identified in relation to reasonable expectations), so they decided to use 51.41 instead. If they did not consider the maximum Wobbe of gas entering the distribution system appropriate they could have used either some other figure, or a variable, but they chose to do neither, specifying instead the figure of 51.41.

If the parties had intended to allocate nitrogen costs based on usage they could also have chosen to compensate for the fact that the Wobbe of a Co-Shipper's LNG as discharged is not an appropriate value for calculating the quantity of nitrogen used by that Co-Shipper's cargo. It will be recalled that this is what Grain was trying to do when it circulated its CAPs consultation paper in January 2010. The thinking behind Grain's consultation paper was that the approach in the CAPs did not take account of the effects of boil-off/weathering after discharge. Depending upon the composition of the LNG, ignoring boil-off/weathering might mean that an allocation by reference to the Wobbe of the LNG as delivered would allocate less nitrogen to a Shipper than that Shipper's cargoes would actually require. What Grain had in mind was a calculation to estimate the Wobbe of each Shipper's LNG in-store on each day and use that value, instead of the Wobbe of the LNG on discharge, to allocate nitrogen costs.

It will be recalled that this suggestion did not win support, and was not pursued. It may well have been thought to be over complicated (a view held by Sonatrach as reflected in an email from Mr Stranks to Mr Way of 28 January 2010), but equally Sonatrach noted that allocating nitrogen costs by reference to the Wobbe of the LNG delivered is favourable to cargoes with high boil-of rates, because less nitrogen will be allocated to them than they need, and Sonatrach perceived that its relatively rich, high nitrogen cargoes were like this and did not welcome Grain's proposal.

The point can be illustrated by a (simplified) example given by BP. Consider a rich cargo with a Wobbe on discharge of 52 and a high boil-off rate. By send-out, the Wobbe has increased to 53. If the operational Wobbe at send-out is 51, this cargo requires sufficient nitrogen to reduce the Wobbe by 2 points. But allocating nitrogen costs by reference to the Wobbe of LNG as delivered means that costs are allocated to this cargo as though it only required sufficient nitrogen to reduce the Wobbe by 1 point.

Factor Y, as written, uses the Wobbe of LNG discharged, not the Wobbe of regasified LNG. This favours nitrogen rich cargoes like those which Sonatrach imported at relevant times, because it allocates less nitrogen to them than they require. Factor Y, as written, uses 51.41, not the Wobbe of gas sent out. That favours lean cargoes, like the ones which BP imported at relevant times, because it allocates less nitrogen to them than they really need. Quite simply the formula does not allocate nitrogen in accordance to usage in a manner that reflects how much nitrogen a cargo actually needs.

What it does do, and what I consider the parties intended it to do, is allocate variable nitrogen costs by reference to the quantity of each Co-Shipper's gas and the quality of each Co-Shipper's LNG, which are the two components of TN_2A and TN_2B. This is the allocation that the parties contractually agreed, as can be seen from the opening words of D2.2.2(ii)(b), "the variable component of nitrogen costs... shall be allocated in accordance with the quantity of gas Sent Out and the quality of the LNG delivered by each Co-Shipper" (emphasis added).

Factors X and Y deal with the quality component (with the greater of X or Y being fed back into the main formula). As highlighted in the opening words, this quality component relates to the Co-Shipper's LNG, not its gas: the first variable in each of X and Y is "the weighted average higher heating value of LNG as delivered". That is a quality of the LNG. And it is a component which, in Factor Y, is used to calculate another quality, viz the Wobbe of the LNG. And that quality, the Wobbe, is the quality which Factor Y uses as the basis for a calculation which can be fed back into the main formula so that costs can be allocated "in accordance with the quantity of gas Sent Out and the quality of the LNG delivered by each Co-Shipper" as D2.2.2(ii) says.

D2.2.2(ii)(b) allocates costs according to quantity and quality in broadly the same way that Grain's original CAPS did. When BP saw the draft CAPs, it commented that they allocated costs in proportion to quantity and quality. Whilst the precise method differs between the two documents, that is correct. Wobbe was the only quality in the CAPs, whereas D2.2.2(ii)(b) also covers the ICF (in Factor X), and the CAPs looked at the Wobbe of the last delivered cargo, rather than an average over a period. This shows that there are various different approaches, but both the CAPs and D2.2.2(ii)(b) compare the relevant quality, the Wobbe, with a constant rather than with a variable operational Wobbe. In doing so they do allocate costs according to quality.

Mr Williamson of BP asked Mr Winstanley in an email of 10 November 2004 about how easy it would be to "come up with a simple formula relation to N2 and C3 consumption to the quality of the cargo". D2.2.2(ii)(b) states that in terms. The fact that Grain operates to a value below the 51.41 constant in the formula does not mean that the formula fails to allocate costs by reference to the quality of the cargo, as Mr Winstanley confirmed when he gave his evidence.

Mr Winstanley was also asked for a "simple formula" in the email of 10 November 2004. Mr Winstanley's evidence was that he regarded it as a simple formula - certainly it would have been possible to have a more complicated formula to take into account actual usage of nitrogen by particular cargoes (assuming that Grain provided the requisite information) or indeed the parties could have chosen a complex variable instead of a 51.41 constant (whether that be the weighted monthly average Wobbe of the gas sent out or some other variable). But that would have been more complicated than the use of a constant of 51.41. For example, the weighted monthly average Wobbe of gas sent out from the Terminal would either have to be obtained from Grain or calculated from information supplied by Grain, and Mr Way's evidence was that it was, "challenging at times" to get information from Grain when he did his subsequent reconciliation exercise. In consequence the parties would be dependent on Grain for information in relation to any such formula, and at the time of contracting the parties could not be sure they would be able to obtain reliable information (certainly Grain was not forthcoming, even to Mr Winstanley) as to the level it would blend to. The parties might well have been reluctant to introduce a variable element into the formula that they could not be sure would be forthcoming from Grain (and even if it was they might need to conduct checks of their own as to accuracy). As BP submits (and I accept), the parties might well have been wary about creating information dependency (as a variable in relation to Wobbe would create) given the lack of information coming from Grain, and the formula had the merit of simplicity.

In the email to Mr Winstanley of 10 November 2004, Mr Williamson stated that, "I don't think that we have to be terribly precise, just want something that should be reasonably fair over time". Time, in this context, means over the life of the contract i.e. a 20 year term. Of course at any point in time one party might perceive that it was doing better than the other party. When giving his evidence Mr Winstanley explained his understanding of "reasonably fair over time" in these terms,

As Mr Winstanley also said in his evidence in a response to a question from Mr Harris,

It is important to bear in mind, as Mr Winstanley confirmed in his statement and in his evidence, that the source and composition of BP's future cargoes was not known at the time of contracting, so that how the allocation would pan out over the long term was not known at the time of contracting. What was known was that the maximum Wobbe of regasified LNG under the GSMR was 51.41, and that gave an indication of the minimum amount of nitrogen which a cargo would require (not necessarily a precise one, because D2.2.2(ii)(b) does not take account of boil-off/weathering), and what was also known was that Grain (whether acting as an RPO or not), would not risk exceeding this value and so would err on the side of caution and therefore blend to a level below this as I have already identified.

Sonatrach asked various witnesses about the effect of the formula in circumstances where Grain blended to the level that it did and that this could result in Sonatrach paying for nitrogen used to treat BP's cargoes. I did not find this line of questioning of particular assistance. The effect of the application of D2.2.2(ii)(b) based on the level to which Grain blended is in general terms accepted by BP, as are the scenarios put by Sonatrach to Mr Winstanley. It is no doubt right that using the actual blending level in the formula would have given a more accurate allocation of nitrogen usage than is achieved by the formula (as Mr Winstanley accepted in an answer to a question from the Court), but little is achieved by getting witnesses to agree such matters (or criticising them for being unwilling to agree this), as the point can equally be made by way of submission, and getting witnesses to agree the point does not give the submission greater weight. In any event, the fact is that Clause D2.2.2(ii)(b) does not contain a different constant, nor does it refer to the variable that Sonatrach is contending for, so whilst a different contractual provision might have resulted in a more accurate allocation of nitrogen that is not what was agreed (unless, of course Sonatrach is right in its case on construction).

Equally Sonatrach's submissions on commercial commonsense and the consequences of the application of the formula to the levels to which Grain blended, can be made perfectly well by way of submission rather than being put to the witnesses. Equally I did not find questioning of witnesses as to what Clause D2.2.2(ii)(b) was designed to achieve of particular assistance either. The witnesses' subjective views as to what was intended by Clause D2.2.2(ii) are inadmissible, and to the extent their evidence can be characterised as reflective of what a reasonable person having all the background knowledge which would have been available to the parties would have understood the parties to be using the language of the contract to mean (which I rather doubt in any event) I do not consider that their evidence leads to the conclusion that Sonatrach is correct that "51.41" means anything other than 51.41.

In the above circumstances, I conclude that the formula does not, and is not intended to, allocate the nitrogen costs according to the amount of nitrogen used by each of BP's and Sonatrach's respective cargoes but rather does, and is intended to, allocate the variable nitrogen costs allocated by Grain to the Co-Shippers in accordance with the quantity of the gas sent out and the quality of the LNG delivered by each Co-Shipper in accordance with the formula that follows. That construction follows from the express language of what the parties agreed in D2.2.2(ii)(b), and the matters that I have identified above, and that is also the commercial purpose of the clause - namely to allocate variable nitrogen costs in accordance with the quantity of the gas sent out and the quality of the LNG delivered by each Co-Shipper, which the formula does.

For the reasons that I will come onto, I do not consider that this construction flouts business common sense, and there is no justification for re-writing the contract by substituting a new variable “the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” for the figure of "51.41" that is expressly stated in the formula.

The consequences that arise when the contractually agreed formula is applied (including the potential consequences identified in Sonatrach's second and third points as identified above) are not the result of any error in the formula, or an uncommercial construction, but simply the fact that (in the event) Grain blended not only to a Wobbe level some decimal points below 51.41, but to a Wobbe level below that of both Sonatrach cargoes and BP's cargoes on discharge into the Terminal.

Sonatrach submits that treating "51.41" as a constant (as I consider it is for the reasons already stated) would lead to what it categorised as a "commercial nonsense" relying on the words of Lord Diplock in Antaios Compania Naviera S.A. v Salen Rederierna A.B. ("the Antaios") [1985] A.C. 191, 201, referred to with approval by Lord Hoffmann in ICS v West Bromwich at p. 913E, that “if detailed semantic and syntactical analysis of words in a commercial contract is going to lead to a conclusion that flouts business commonsense, it must be made to yield to business commonsense” (referring also in this regard to the Supreme Court decision in BNY Mellon Corporate Trustee v LBG Capital No 1 Plc [2016] UKSC 29 which I have already referred to above).

I do not consider those sentiments to be apposite in the present case. The construction advocated by BP does not lead to a conclusion which flouts business commonsense. However even if that had been the case, that would not be, and has never been, a justification to re-write the parties' contractual bargain in a manner that the court perceives to be "fairer" or (in this case) to result in a "fairer" allocation of nitrogen costs - that would be to change the parties' contractual bargain in circumstances where it is not clear what the parties would have intended had they been aware of the level to which Grain (subsequently) chose to blend (as I address further below). Sonatrach's case on construction would require the Court to re-write the contractual bargain and substitute, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" for the figure of "51.41". That is not giving effect to the parties' intentions but re-writing the contractual bargain.

In this regard the guidance given by Lord Neuberger about commercial common sense and surrounding circumstances in Arnold v Britton at [17] and [19]-[20] is of particular relevance. First that reliance placed in some cases on commercial common sense and surrounding circumstances (e.g. in Chartbrook [2009] AC 1101, paras 16-26) should not be invoked to undervalue the importance of the language of the provision which is to be construed. Secondly, that commercial common sense is not to be invoked retrospectively. The mere fact that a contractual arrangement, if interpreted according to its natural language, has worked out badly, or even disastrously, for one of the parties is not a reason for departing from the natural language. Commercial common sense is only relevant to the extent of how matters would or could have been perceived by the parties, or by reasonable people in the position of the parties, as at the date that the contract was made. Thirdly, whilst commercial common sense is a very important factor to take into account when interpreting a contract, a court should be very slow to reject the natural meaning of a provision as correct simply because it appears to be a very imprudent term for one of the parties to have agreed, even ignoring the benefit of wisdom of hindsight. The purpose of interpretation is to identify what the parties have agreed, not what the court thinks that they should have agreed.

I consider these points to be apposite to the present case. First, it is important to have regard to the language of clause D2.2 (and D2.2.2.2(ii)(b) in particular), and the meaning is most obviously to be gleaned from the language of the provision. Clause D2.2.2(ii)(b) is a detailed provision containing detailed calculations and formula. For reasons that have already been identified this is not a case where there are two alternative meanings or constructions nor can it be suggested that the clause is badly drafted. On its face the clause is to be taken to mean what it says – namely to utilise a constant of 51.41. Secondly, commercial common sense is not to be invoked retrospectively - the mere fact that the effect of the formula results in particular allocations of nitrogen as a result of the level that Grain chose to blend to (as relied upon by Sonatrach to suggest the clause produces absurd results), is not a reason to depart from the natural language. Commercial common sense is only relevant to the extent of how matters would or could have been perceived by the parties, or by reasonable people in the position of the parties, as at the date that the contract was made - at the time the JSA was entered into it no doubt made sense to include a value of 51.41. Thirdly, a court should be very slow to reject the natural meaning of a D2.2.2(ii)(b) as correct simply because it appears (in the event) to have been an imprudent term to have agreed, even ignoring the benefit of wisdom of hindsight (exposing the parties, as it did, to the risk of paying for nitrogen that their cargoes had not in fact used).

Sonatrach submits that allocating nitrogen costs on the basis of the quality of a Co-Shipper's LNG compared to a constant of 51.41 is irrational or absurd - relying on the scenarios that I have already referred to. In fact, the scenarios that Sonatrach relies upon (as already identified above) are simply a consequence of the application of the formula set against the backdrop of the amount of nitrogen used by reason of the level to which Grain subsequently chose to blend (and the scenarios also ignore the effect of the agreed 100% cap which, once applied, allows for an allocation between the parties of the sum charged by Grain).

It is important to bear in mind that at the time of contracting there were various unknowns yet it was necessary for there to be a contractual formula for the allocating of nitrogen costs. What the parties chose was a formula which allocates based on the quantity of gas sent out and quality of LNG delivered, and the associated formula including a constant of 51.41, no doubt recognising that there were various unknowns. These included that no one knew the precise characteristics of the cargoes that each would import and deliver (still less knew what they would be over a 20-year term). Nor were Grain's operating practices settled. Nor was Grain forthcoming about what Wobbe value Grain would target (though the maximum Wobbe of regasified LNG entering the networks was a known maximum figure of 51.41, and a reasonable person would also have had the background knowledge that I have found). Set against such a backdrop allocating nitrogen costs on the basis of the quality of a Co-Shipper's LNG compared to a constant of 51.41 cannot be said to be either irrational or absurd. It was part of a simple formula that the parties agreed. It avoided the complexities of additional variables (which might also require information that the parties could not be certain they would obtain).

The formula does, and is intended to, allocate the variable nitrogen costs allocated by Grain to the Co-Shippers in accordance with the quantity of the gas sent out and the quality of the LNG delivered by each Co-Shipper in accordance with the formula that follows. That construction follows from the express language of what the parties agreed in D2.2.2(ii)(b). That construction does not flout business common sense, and there is no justification for re-writing the contract by substituting a new variable “the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” for the figure of "51.41" that is expressly stated in the formula.

It will be recalled that Lord Neuberger in Arnold v Britton at [21] identified that:-

In my view the words I have highlighted are crucial. If there is an event that subsequently occurs which was plainly not intended or contemplated by the parties, judging from the language of the contract, it is only where it is clear what the parties would have intended that the court will give effect to that intention - otherwise the court would not be giving effect to the parties' intentions but would be re-writing the contractual bargain based on speculation as to what the parties might or might not have agreed, or based on its perception of what might have been a fairer or more appropriate contractual term each of which is impermissible.

Equally if the inclusion of "51.41" was to be construed as a mistake (and I do not consider that would be an apposite characterisation of the use of the constant), the words of Lord Hodge in Arnold v Britton at [78] would need to be borne in mind:-

It is clear (as Mr Winstanley accepted) that Mr Winstanley had not run tests on the formula with various gas compositions before contracting, nor had he given thought to the effect of the clause. Equally there is no evidence that anyone else within BP or Sonatrach had either. I have already addressed what might have been within the reasonable contemplation of the parties in terms of blending. What is clear is that they did not know the level that Grain was planning to blend to, and as the plant was not operational there was no operational experience.

What in my view is unclear is what the parties would have intended if they had been aware at the time of contracting that Grain would blend to a level lower than that which had been contemplated and the effect of the formula in such circumstances. This is fatal to the suggestion that the "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” should be substituted for the figure of "51.41".

There are a number of possibilities as to what the parties might have agreed in this scenario but it is not possible to identify one common intention that can be given effect to, even assuming what occurred is to be characterised as an event subsequently occurring which was plainly not intended or contemplated by the parties (and it could be said that the parties did contemplate that Grain might blend to below 51.41 albeit not to the level it did, and yet contracted on the terms they did).

Subsequent events showed Grain (subsequently) blending to a variety of different targets and/or operational Wobbe numbers, and it is not clear how the parties would have understood or drafted the contract had they been aware, at the time of contracting, of what Grain subsequently did in relation to blending. There are various possibilities as to what the parties might or might not have done to the figure of 51.41 - for example change the number to another number (e.g. 51.1), or to a reference to the Wobbe targeted from time to time, or to a reference to the actual Wobbe value that Grain would blend to, or a reference to an average operational Wobbe (monthly or otherwise), or indeed to do nothing.

It is to be borne in mind that the parties had chosen a fixed number, a constant of 51.41. Some variation in the number Grain was targeting might not have caused them to change the number of 51.41 at all (for example Grain's previous reference to a figure of 51.3 was not something that caused Mr Winstanley to insert anything other than a fixed value of 51.41 in the formula for the parties' consideration). However, if the parties had been aware that Grain targeted another number e.g. 51.1 (as appeared in the March 2006 Report as "optimum" Wobbe) they might have chosen to insert a different number (for example 51.1) - certainly it appears that is something Mr Winstanley would have proposed to them - though there is no evidence that they would have agreed such fixed number/constant in that situation, and I do not so find.

They might have considered changing the constant to some form of variable but it does not follow that they would necessarily have done so, indeed that would have made the formula more complex (which might have been regarded as undesirable), and they would have been aware that Grain was not particularly forthcoming about what it intended to do (indeed that seems to have continued throughout the life of the contract as Mr Way's evidence was that it was, "challenging at times" to get information from Grain when he did his subsequent reconciliation exercise).

It is notable that even when Mr Way did his reconciliations he used a figure of 51.41 and other fixed numbers (51.1 and 51.25). Indeed, these were used in both the reconciliations sent to Sonatrach (in late 2006 and early 2007), and those circulated to BP and Sonatrach on 1 February 2007 (as addressed in more detail when considering issues 3 and 4 below). As late as 2008, in his reconciliations of 14 July 2008 and 8 September 2008 (as addressed under issue 5 below), he used a number which he described as "close to operational Wobbe" adding, "Please note that this wobbe number is not as advised by the terminal it is only provided as an indicative example." So even in reconciliations years after the event fixed numbers were being used in reconciliations, and it appears there was difficulty in obtaining information from Grain.

Even if the parties might have considered agreeing a variable in substitution for 51.41 there were various possibilities including " the Wobbe targeted from time to time", or "the actual Wobbe value that Grain would blend to", or to "an average operational Wobbe" (monthly or otherwise). These are all different, and all have the vice of being dependant on information being supplied by Grain (and on an ongoing basis) that might or might not be forthcoming, and might have been less desirable than (some) fixed number (even if that meant less precise allocations).

Sonatrach have not themselves been consistent as to what they are advocating which is not an auspicious start to a submission that it is clear what the parties would have intended. For example at paragraph 48 of their Written Closing they say that, "The formula for "Y" would be understood as intended to identify how much nitrogen was used by Grain to bring cargoes down to the Wobbe level which Grain was targeting, and the number 51.41 in the formula simply reflected what both parties assumed (and what any reasonable third party would have assumed) was the blending target by Grain" (emphasis added) which focuses on a target Wobbe (in oral closing Mr Harris said that this was not his case). At paragraph 60 of Sonatrach's Written Closing it is submitted that the court should seek to give meaning to what it is submitted was the evident intention of the parties, "which is to measure nitrogen consumption against the level at which Grain blended" (which itself sounds more like a target level or a fixed value, and certainly not “the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time." Even Sonatrach's pleaded case bears some ambiguity - a Wobbe value used by Grain sounds rather like a target (or "optimum" as it was put in the March 2006 report), and this is not the same thing as the actual Wobbe of send-out gas, or as "the monthly weighted average of the Wobbe of the gas sent out from the Terminal as advised by [Grain]” which is how Sonatrach puts its case on the September 2006 meeting (as addressed under issue 3) (Defence para 52c).

In the above circumstances it is not clear what the parties would have intended if they had been aware at the time of contracting that Grain would blend to a level lower than that which had been contemplated and the effect of the formula in such circumstances. There are various changes they might have agreed and that might have involved anything from no change to a particular fixed number to a variety of possible variables, but there is no mandate for the court to intervene and choose one of those possibilities, and re-write clause D2.2.2(ii)(b) accordingly. On any view it is not clear that the parties would have agreed, “the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time” and there is no basis for substituting this for "51.41" in D2.2.2(ii)(b).

For the reasons that I have identified, and having regard to the meaning of the words in clause D2.2.2(ii)(b) in their documentary, factual and commercial context, I find that "51.41" is used as a constant in the formula, and is to applied as a constant. I have reached this meaning by reference to the natural and ordinary meaning of the figure used and of the clause as a whole, the overall purpose of the clause, the facts and circumstances known or assumed by the parties at the time the JSA was executed and commercial commonsense - all as addressed in the preceding paragraphs of this judgment. In doing so I have applied the principles identified at paragraphs [14] to [22] in the judgment of Lord Neuberger in Arnold v Britton and the other authorities I have referred to, and have had regard to.

Ultimately "51.41" means what it says and says what it means. There is no warrant for construing "51.41" as meaning, or as a proxy for, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time", and I reject such a construction. To adopt such a construction would not reflect the intention of the parties by reference to what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language of the contract to be. It would be to re-write the parties' contractual bargain, and to do so in a manner for which there is no warrant on the applicable principles of contractual construction that I have identified.

ISSUE 2: IMPLIED TERMS

Sonatrach submits that if (as I have found) it is wrong as a matter of construction of D2.2.2(ii)(b), two terms fall to be implied into D2.2.2(ii)(b) as a matter of necessity. It will be recalled that the two alleged implied terms are as follows:-

I do not consider that either of these terms falls to be implied, as I address below. However, in summary, neither is necessary to give business efficacy to the JSA, nor are they so obvious that “it goes without saying”, nor are they sufficiently certain. Even more fundamentally, such terms are inconsistent with, and contradict, the express terms of D2.2.2(ii)(b), and the figure of 51.41, and as such cannot be implied on established principles. It will also be apparent that the alleged implied terms are based on what I have found to be the same incorrect premises as Sonatrach’s construction argument and, as such, do not support the implication of the terms alleged. I address each of these matters below.

Dealing with the last point first. Sonatrach's implied terms case is based on incorrect premises. It is based on the same premises as Sonatrach's construction argument (see the particulars in the Defence at paragraph 38B). Those premises are incorrect, for the reasons stated above in relation to construction. They can provide no more support for Sonatrach's implied terms argument.

Sonatrach's implied terms argument is pleaded in the alternative to Sonatrach's construction argument. It therefore only arises if Sonatrach's construction argument is wrong. (If Sonatrach is right on construction, it does not need to rely on implied terms). This means that the construction argument proceeds on the basis that "51.41" is a constant - this is why, as addressed in more detail below, Sonatrach's implied terms are inconsistent with the express terms, and the constant of "51.41".

I have already set out the applicable principles in relation to the implication of terms as reviewed and summarised by Lord Neuberger PSC in Marks and Spencer plc v BNP Paribas Securities Services Trust Company (Jersey) Limited and another [2016] A.C. 742 at [16] – [20].

So far as business efficacy is concerned I take the starting point from the three classic statements that form part of the "clear, consistent and principled approach" identified by Lord Neuberger at paragraph 16 of his judgment, and endorsed by him at paragraph [20] of his judgment:-

In the present case neither of the alleged terms is necessary to give the JSA and Clause D2.2.2.2(ii)(b) such efficacy as both parties must have intended that at all events it should have (The Moorcock (1889) 14 PD 64, 68). For the reasons that I have already identified in the context of a consideration of the construction arguments that arise, Clause D2.2.2(ii)(b), as I have construed it, gives meaning and effect to the intentions of the parties, and provides an allocation of variable nitrogen costs in accordance with the provisions of Clause D2.2.2(ii)(b). Business efficacy does not require that either of the two alleged terms be implied.

As Scrutton LJ stated in Reigate v Union Manufacturing Co (Ramsbottom) Ltd [1918] 1 KB 592 , 605 (whereby a term can only be applied if it is necessary in the business sense to give efficacy to the contract), a term will only be implied if it is such a term that it can confidently be said that if at the time the contract was being negotiated the parties had been asked what would happen in a certain event, they would both have replied: “‘Of course, so and so will happen; we did not trouble to say that; it is too clear.’” That is not this case, and Sonatrach cannot possibly contend that it is. As I have already identified, if at the time the JSA was being negotiated the parties had been asked what would happen if Grain blended to a level below 51.41 (or targeted a level below 51.41) it is not at all clear what the parties would have said or agreed (there being many different possibilities). They most certainly would not have said “‘Of course, we will allocate the nitrogen costs between [ourselves] under section D2.2.2(ii)(b) of the JSA on the basis of the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time" or "we will apportion costs under section D2.2.2(ii)(b) of the JSA would result in each Co-Shipper paying an amount as close as possible to the actual costs attributable to the nitrogen used to blend the LNG they had respectively delivered by the Terminal" continuing, "we did not trouble to say that; it is too clear.”

Equally neither of the alleged terms is so obvious it goes without saying (Shirlaw v Southern Foundries (1926) Ltd [1939] 2 KB 206, 227), and each of the alleged terms fails the "officious bystander" test. A term will only be implied, “if, while the parties were making their bargain, an officious bystander were to suggest some express provision for it in their agreement, they would testily suppress him with a common ‘Oh, of course!’”.

Lord Neuberger in the Marks and Spencer case at [21] noted that if one approaches the issue of implication by reference to the "officious bystander" it is "vital to formulate the question to be posed by [him] with the utmost care" and that is clearly right (as reflected in the fact that in many cases each party will formulate the question in a manner designed to give them the right answer). However, in the present case, there is no great difficulty in the starting point for the formulation, as it is reflected in the implied term. If the officious bystander had asked BP and Sonatrach, "what if the Terminal in fact blended to a Wobbe index value other than 51.41" (the opening words of the first alleged implied term) and suggested they then use, "the average monthly operational Wobbe index value used for LNG blending and regasification at the Terminal from time to time", the parties would most certainly not have testily suppressed him with a common, "Oh, of course". On the contrary I consider that they would have thought about it, and might have considered various possibilities (ranging from doing nothing, to changing the number, to considering changing to a variable, to considering what variables might or might not be appropriate), and each party's view as to how they might consider it appropriate to deal with the matter might well differ - that is no basis for the implication of a term, and illustrates why the first alleged implied terms is not to be implied.

It is notable that when referring to the "officious bystander" test at paragraph 83 of their Written Closing Submissions Sonatrach submits that, "it would have been evident to any officious bystander that, had the question been asked, the parties would have agreed that insofar as Grain were to blend at a level below 51.41, that lower Wobbe value would be used in the second formula in order to ensure the allocation of variable nitrogen costs according to usage." For the reasons I have already given I do not agree. However, it will be noted that this formulation (by reference to "that lower Wobbe value would be used") is different to the implied term alleged (and indeed sounds like a constant not a variable figure). This goes to illustrate the various permutations that might or might not have been each party's separate response.

The second alleged implied term does not meet the "officious bystander" test either. If the officious bystander had said, "should the apportionment of costs under section D2.2.2(ii)(b) of the JSA result in each Co-Shipper paying an amount as close as possible to the actual costs attributable to the nitrogen used to blend the LNG they had respectively delivered by the Terminal” the parties again would most certainly not have testily suppressed him with a common, "Oh, of course". It is not clear on the evidence whether they would have a joint or common response, and it might well have simply led to debate as to what they were seeking to achieve in clause D2.2.2(ii)(b).

One response might have been that it was intended to be a simple formula to provide an allocation of variable costs based on the quantity of gas sent out and the quality of the LNG delivered, and was not intended to lead to each Co-Shipper paying an amount as close as possible to the actual costs attributable to the nitrogen used to blend the LNG they had respectively delivered to the Terminal - not least because the formula was not designed to calculate the nitrogen used by a particular cargo (and that was no easy calculation in any event given the effects of co-mingling, weathering and boil off - all matters that would have been in the reasonable contemplation of the parties). Whether that would or would not have been either party's response it cannot be seriously suggested that they would have testily suppressed the officious bystander in the manner alleged.

Reliance has been placed on what Lord Hoffmann said in Belize Telecom at [25] in the context of the implication of any term:-

“The need for an implied term not infrequently arises when the draftsman of a complicated instrument has omitted to make express provision for some event because he has not fully thought through the contingencies which might arise, even though it is obvious after a careful consideration of the express terms and the background that only one answer would be consistent with the rest of the instrument.”

(emphasis added)

Even assuming that it could be said that the draftsman has failed to make express provision for some event because he has not fully thought through the contingencies which might arise, it is most certainly not obvious that there is in this case only one answer that would be consistent with the rest of the JSA. For the reasons I have given there are many possible answers, and the alleged terms are not so obvious as to go without saying or to be necessary for business efficacy. As I have found, it is far from clear how the parties would have addressed matters, in terms of what the parties might or might not have done to the figure of 51.41, had they been aware, at the time of contracting, of what Grain subsequently did in relation to blending, given the various differing possibilities that existed, such as changing the number to another number (e.g. 51.1), or to a reference to the Wobbe targeted from time to time, or to a reference to the actual Wobbe value that Grain would blend to, or a reference to an average operational Wobbe (monthly or otherwise) or, indeed, leaving matters as they were.

In the context of Lord Simon's formulation in BP Refinery (Westernport) Pty Ltd v Shire of Hastings (1977) 180 CLR 266, 283 the five conditions identified were (1) it must be reasonable and equitable; (2) it must be necessary to give business efficacy to the contract, so that no term will be implied if the contract is effective without it; (3) it must be so obvious that "it goes without saying"; (4) it must be capable of clear expression; (5) it must not contradict any express term of the contract.

As to the first condition that a term be reasonable and equitable, there are no doubt various possible terms which might have been inserted which might have been regarded as reasonable and equitable, but they differ, and it would be to re-write the parties contractual bargain to substitute a particular wording for what they have (expressly) agreed. Whilst accepting that they are not unreasonable or unfair, I do not consider the proposed implied terms to be any more reasonable or fair, than any other possibility. But as Lord Neuberger pointed out in the Marks and Spencer case at [21], a term should not be implied into a detailed commercial contract merely because it appears fair or merely because one considers that the parties would have agreed it if it had been suggested to them (though I do not consider that to be this case). As he also pointed out, those are necessary but not sufficient grounds for including a term. In any event he also recognised (rightly in my view) that it is questionable whether Lord Simon's first requirement, reasonableness and equitableness will usually, if ever, add anything.

In the present case, and for the reasons I have given, neither of the proposed terms is necessary to give business efficacy to the contract (condition 2) or so obvious that “it goes without saying" (condition 3). I recognise that business necessity and obviousness can be alternatives, though here neither is satisfied and it is apt to recognise that the terms are neither necessary to give business efficacy nor obvious.

Nor do I consider that the terms to be capable of clear expression either (condition 4). The former alleged implied term is but one possible formulation each of which differs in meaning and effect (even if one goes down the route of a "variable" there are various possible variables all with differing meanings), and the latter implied term referring to "an amount as close as possible" to the actual costs attributable to the nitrogen used to blend the LNG they had respectively delivered by the Terminal, is inherently vague.

The fifth condition is that the proposed implied term must not "contradict any express term of the contract". I have already identified why this condition exists by reference to paragraph 6.11 in Lewison and the cases there cited which are in point. Ultimately, "the reason that it is not possible to imply a term which contradicts an express term is because the parties cannot have intended that” (in the words of Toulson LJ in Anders & Kern UK Ltd v CGU Insurance Plc^, supra).

This is a further problem, and indeed a fundamental, and insurmountable, problem in relation to each of the alleged implied terms. They are simply inconsistent with an express term of the JSA, namely the constant of “51.41" in the formula.

Sonatrach seeks to characterise its first implied term as a "modification to a phrase in an existing term of the contract" rather than a contradiction of the express term (51.41). I disagree. On Sonatrach's implied terms the express term and use of the constant 51.41 in the formula, has to be disregarded. It is no answer to submit that the clause is to be read as if it provided, "51.41 or the average monthly operational Wobbe if different" . If the average operational Wobbe for a particular month did happen to be 51.41, then that value would be used for the calculation on the basis of the alleged implied term. But that is not because 51.41 = 51.41 (the constant in the formula). It is because 51.41 = the average monthly Wobbe. It is the monthly average, not the constant which actually appears in the contractual formula, which would drive such a calculation. So 51.41 (the contractually agreed figure), would not be used as a constant at all. It would also be purely coincidental if the average monthly operational Wobbe was 51.41. What is more it is inherently improbable that the average monthly operational Wobbe would ever be 51.41 (to two decimal places) not least given that Grain would be blending to below that level, and the actual Wobbe of the send-out gas changes constantly, depending on the composition of the gas and the quantity of nitrogen being injected.

The reality is that the proposed implied terms are simply inconsistent with the express terms of the JSA and the contractually agreed use of a constant, in the figure of "51.41". Sonatrach is not assisted by reference to cases in which the court has been willing to imply terms to modify existing terms of a contract - Aberdeen City Council v Stewart Milne Group Limited [2012] SLT 205 at [31] to [32], Rathbone Brothers v Novae Corporate Underwriting Limited [2014] 2 CLC 818 at [85] and Spencer and another v Secretary of State for Defence [2012] 2 All ER (Comm) 480 at [91].

Such cases are not in point. Cases such as Aberdeen v Stewart Milne and Rathbone v Novae illustrate the concept that a provision which does not contain an explicit restriction on scope may be impliedly subject to a restriction, whilst the decision of Vos J (as he then was) in Spencer v Secretary of State turned on its own facts, and the carrying into the memorandum in that case of a process of arbitral review which had been referred to in the existing lease. In contrast the terms alleged by Sonatrach are simply inconsistent with the express constant of 51.41 in D2.2.2(ii)(b). It is not a question of modification but of inconsistency. I would only add that little assistance is to be gained by seeking to apply, by analogy, the facts of other cases on implied terms, as in each case one has to construe the particular contract, and its express terms, and apply the principles already identified as to the implication of terms. I do not consider the facts of any other case assist in this particular case. It is the applicable principles, not the facts of other cases, which assist. In the present case each of the alleged implied terms requires that the contractually agreed constant be ignored and be substituted by something else. There is no warrant to do so, and to do so would be to re-write the contractual bargain in a manner that is contrary to an express term of the contract, which the parties cannot have intended.

Sonatrach submits that what is being proposed is that a single word or phrase "is being modified to reflect the purpose of the clause and indeed the agreement within which it sits". I disagree. The constant of 51.41 is not being modified, it is being ignored or contradicted. Further Sonatrach's alleged implied terms do not reflect the purpose of the clause (as I have found it to be). In any events the requirements for the implication of a term are not met in the multiple other respects I have identified.

I find that neither of the alleged implied terms falls to be applied. Neither is necessary to give business efficacy to the JSA, nor are they so obvious that “it goes without saying”, nor are they sufficiently certain. Even more fundamentally, such terms are inconsistent with, and contradict, the express terms of D2.2.2(ii)(b), and the figure of 51.41, and as such cannot be implied.

In the above circumstances Sonatrach's case on construction, whether by reference to the express terms, or any alleged implied terms, fails.

Sonatrach allege in its re-Amended Defence and Counterclaim (at paragraph 44), as reiterated at paragraphs 8, 49.4 and 104 of Sonatrach's Skeleton Argument for Trial, that at the Steering Committee Meeting on 28 September 2006 in Paris the parties agreed to amend the JSA to replace 51.41 in D2.2(ii)(b) with the monthly weighted average Wobbe in fact blended to by Grain. Sonatrach asserts that the minutes of the meeting support this as does the statement of Mr Stranks. BP denies that there was any such agreement, and submits that the evidence of pre-meeting events, the minutes of the meeting, the evidence of the meeting (as given by Mr Stranks when cross-examined) and events after the meeting all demonstrate that there was no agreement at the meeting to change "51.41" to an operational Wobbe.

The meeting itself was attended by Claire Torrance for BP and Mohand Bouadi for Sonatrach as Steering Committee members. It was also attended by Mr Stranks, as well as Omar Bensalem and Martin Stewart-Smith (of the law firm Simmons & Simmons) for Sonatrach, and Jonathan Lunn for BP. The only witness called to give evidence in relation to the meeting was Mr Stranks. Sonatrach did not call any of the other Sonatrach attendees (including the drafter of the minutes Mr Stewart-Smith). Ms Midwinter (an associate at BP's solicitors Holman Fenwick Willan) gave a short statement explaining why Ms Torrance and Mr Lunn had not provided statements. In short Ms Torrance no longer works in the energy industry and was unable to recall any relevant matters, and Mr Lunn, whilst recalling his attendance at the meeting, was unable to recall anything in addition to what is set out in the contemporary documents.

The meeting, and the minutes of the meeting, need to be understood in context. It is accordingly relevant to have regard to the events that preceded the meeting and set the scene for the meeting, as addressed below.

As for the minutes themselves, these provided amongst other matters, as follows:

So far as Mr Stranks' witness statement is concerned, this provided at paragraphs 57 and 58 as follows:-

I would make two preliminary observations about Mr Stranks' statement. First he does not state expressly that at the meeting there was an agreement to replace 51.41 with the monthly weighted average Wobbe value (i.e. an operational Wobbe), which might be thought surprising if indeed that was his evidence given that this was Sonatrach's pleaded case, and one of the five issues in the case. Secondly it will be noted that there is an apparent inconsistency between paragraph 57 of his statement (which refers to "the monthly weighted average of the Wobbe of the gas sent out"), and paragraph 58 (which refers to "the actual Wobbe target level being used by GLNG”) which did not auger well for Sonatrach's case that there was an agreement, in the meeting, to replace 51.41 with the monthly weighted average Wobbe value.

In the event, when cross-examined, it became plain from a number of the answers Mr Stranks gave (in which he referred to a "discussion" (itself challenged by BP) but not an agreement) that it was not his evidence that there was an agreement in the meeting to replace 51.41 in D2.2.2(ii)(b) with the monthly weighted average Wobbe value in fact blended to by Grain. Whilst this was also apparent from a number of answers he gave in response to questions asked by Mr Eaton, it suffices to quote from the following passages in Mr Stranks' evidence in response to questions I asked him:

In the light of the totality of Mr Stranks' evidence Mr Harris in closing, on behalf of Sonatrach, realistically and rightly, conceded that he could not make good Sonatrach's pleaded case, on the evidence, that there was an agreement to change to any particular number at that meeting:-

That there was no agreement in the September meeting to amend the JSA to replace 51.41 with the monthly weighted average Wobbe value in fact blended to by Grain is also demonstrated, and corroborated, by a consideration of the evidence concerning events before the meeting, and after the meeting to which I will turn. BP also submits that this evidence demonstrates that there was not, in fact, any discussion about Wobbe values in the meeting still less any agreement (in principle or otherwise) to change the constant of 51.41. What was agreed (say BP) was that there was an agreement to instruct the Agent to perform a reconciliation (which Sonatrach wanted carefully checked), to take account of the missing 28.9 and to get a definitive figure of sums owing rather than BP's £1.5 million estimate. It will be necessary to have regard to the evidence concerning events before the meeting, the minutes of the meeting, what Mr Stranks says about this, and evidence after the meeting in order to reach findings on this point.

It has already been noted that the Agent had approached BP about "strange results" with its spreadsheet in April 2006, which led to Mr Winstanley identifying that he had omitted the 28.9 constant for the molecular weight of air in the formula. As I have found Mr Winstanley was right to characterise this as a “typographical error”, it was simply a constant that he had inadvertently omitted when setting out the formulae. It could equally be described as an “error” or a “manifest error” in my view (in the context of the subsequent language of the minutes of the September meeting). The failure to include the 28.9 coefficient for the molecular weight of air was referred to by Mr Wood in his letter to the Co-Shippers on 18 May 2006. On 3 July 2016 the parties agreed to add the molar weight of air into the formula. Obviously any reconciliation (as envisaged in the September meeting) would need to take the inclusion of the 28.9 constant into account.

Equally in May 2006 it was identified that the formula sometimes produced allocations in excess of 100%, and as has already been referred to, in August 2006 the parties agreed to a capping of allocations to 100% of costs. Whilst the 100% cap point was not obviously describable as a “manifest error” (in contrast to the omission of the coefficient for the molar weight of air), as it was a function of the operation of the formula when the Terminal operated to the Wobbe level that it did, any reconciliation (as envisaged in the September meeting) would need to take the 100% cap into account.

In addition, following receipt of Grain's March 2006 Monthly Report the parties were aware that Grain targeted an "optimum" Wobbe of 51.1 and often injected more nitrogen than the "optimum" quantity.

There were discussions between BP and Sonatrach in the summer of 2006 as regards the use of 51.41 in the formula. In this regard Mr Way produced a memorandum on 25 July 2006 for Sonatrach on nitrogen allocation matters (addressed to the attention of Martin Stewart-Smith, the Simmons & Simmons solicitor working for Sonatrach) in which he stated that, “BP are proposing the current wobbe correction factor of 51.41 in the JSA be changed to be a monthly weighted average of the Wobbe of the gas sent out from the terminal as advised by Grain.”

There is an issue between the parties as to whether such a proposal had originated from BP (as this memo suggested) or from Sonatrach. Mr Martinez in his witness statement for BP (nearly ten years after the event) stated that he thought that changing the Wobbe number in the formula was Sonatrach’s suggestion. He accepted that there was discussion around changing the Wobbe number to an alternative fixed number of 51.1, but his evidence was that it was never taken up during his involvement. In the event Mr Martinez did not give evidence (it is said in his second statement that this was due to a potential conflict of interest with his current employer Repsol). Mr Winstanley in his first witness statement states that he does not know when a change to the Wobbe number was first discussed, but he believed that it was Sonatrach’s suggestion (as opposed to one from the Agent or BP). He notes that he took notes of all meetings with the Agent and he has no note of a proposal being made by BP.

What does exist is a contemporary email of 2 August 2006 from Mr Martinez to Mr Lunn, Mr Winstanley and others within BP in which he stated, “we did offer Andy to modify the formulas for ICF 0.3 and Wobbe 51.1, but he wanted to stick to the original ones.”. On 4 August 2016 Mr Winstanley emailed Mr Martinez (in an internal BP message, copied to Mr Lunn) stating that the modification for the costs cap and the 28.9 figure should be incorporated into the Agent’s spreadsheet to reflect the JSA. Mr Martinez also sent a message to the Agent, to Mr Lunn and Mr Winstanley on 4 August 2006, in which he said, “The only modification the formula in the JSA needs is the 28.9 factor. Once that is done the formula will reflect what the JSA says”, and also referred to the cost cap.

When cross-examined, Mr Winstanley accepted that he and Mr Martinez probably had discussions with Mr Way about the impact of a change in Wobbe to 51.1. In this regard the following exchange occurred:

It has to be said that the evidence as to precisely what was discussed, between whom, and who instigated any such discussion in relation to the Wobbe number, is somewhat scant even after all the evidence was heard. On reviewing all the documentary and witness testimony I am satisfied, and find, that there was discussion between Mr Way and Mr Martinez about the Wobbe number and the possibility of a change to the Wobbe number, from 51.41 either to 51.1 (which seems the more likely given the testimony quoted above and the fact that Grain were stating that they were targeting an “optimum” Wobbe of 51.1) or (possibly) a monthly weighted average of the Wobbe of the gas sent out (as stated in the memo).

Either way the evidence does not support the conclusion that there was any agreement at this time to change the Wobbe (nor did Sonatrach suggest that in closing). If there had been any such suggestion it would have been necessary to consider whether any BP personnel involved would have had authority to enter into any such agreement. In this regard the evidence was that neither Mr Martinez nor Mr Winstanley would have had any such authority, and Mr Way would not himself have been authorised to accept any proposal (he would have simply communicated the proposal on).

Ultimately I do not consider it matters as to who first raised any possible change to the Wobbe or any proposal in that regard. Either way there was no agreement to change the Wobbe in the summer of 2006. At first blush it does appear less likely that it would have been BP (due to the impact of any change in the Wobbe value downwards), but at the same time (as Mr Winstanley recognised in cross-examination) I can see no particular reason why Mr Way would mis-state matters in his memo. If relevant to anything, and it does not appear to me to be relevant to the particular issue under discussion, I would find that it is more likely than not that the possibility of changing the Wobbe number originated from BP.

Returning to Mr Way’s memorandum of 25 July 2006, it was recognised that an incorrect interpretation of the formula by the Agent had led to variable nitrogen costs being allocated approximately equally between the Co-Shippers. That was recognised as being wrong, because BP's leaner Egyptian and Trinidadian cargoes needed less nitrogen than Sonatrach's richer Algerian cargoes. So the allocation should have been weighted towards Sonatrach's cargoes, rather than approximately equal. The view at the time was that BP had been paying for nitrogen which had in fact been used for the richer Algerian cargoes. However, as was stated, “BP have identified a missing component in the cost allocation formula which when correct would allow the Agents to assess the amount of miscalculation that has occurred” (i.e. the 28.9 constant). Mr Way estimated that the amount due to BP (after rectification of the incorrect interpretation of the formula and adding in the missing constant) would be in the region of approximately £1.0 million-£1.5 million, the sense at the time being that BP had paid too much and had effectively paid for nitrogen which had been used for Sonatrach's cargoes. Mr Way’s memo also recorded that the effect of Grain targeting a Wobbe below 51.41 was that BP's cargoes required more nitrogen than was currently being allocated to them. The consequence of this is that if the formula had used 51.1 (or actual operational Wobbe) instead of 51.41, more nitrogen would have been allocated to BP.

Mr Way confirmed in cross-examination that he could not remember receiving any response from any of the recipients to his memo within Sonatrach at the time, and none is documented. There is no evidence that anyone took forward an idea of changing 51.41 to any operational Wobbe at this time. Equally there is no indication that anyone planned to table a proposal at the September meeting to change from 51.41 to an operational Wobbe. Mr Wood circulated a draft Agenda within Sonatrach on 13^th September. It said nothing about JSA costs allocation. Mr Stranks in his email to Mr Wood of 22 September suggested that BP would want to add Costs allocation to the Agenda "in light of the discussion on Power and Propane". Whilst "Power and Propane” was not defined, I agree with BP’s submission that it probably refers to the allocation of power connection costs, and power and propane in the context of Grain's November 2005 CAPs. It is not a reference to nitrogen and Wobbe. Mr Stranks did not raise any suggestion that Sonatrach ought to raise a change from 51.41 to an operational Wobbe (such an amendment would logically favour Sonatrach as it would allocate a greater proportion of nitrogen costs to BP as recorded in Mr Way’s memo).

The day before the meeting, BP sent Sonatrach a "modified and amended list of JSA amendments", in relation to which Point 21 was "D2 cost allocation of N2 and C3". There was no elaboration as to what was meant by “D2 cost allocation of N2 and C3”, but I consider that it was, and would have been understood, to be a reference to the 28.9 constant and 100% cap points, rather than any discussions about Wobbe given that there is no evidence that such discussions had been followed up or progressed. This is consistent with the subsequent minutes of the meeting (and indeed subsequent actions of the Agent in terms of the Agent’s initial circulated reconciliations). This entry was not highlighted in red signifying that it was an amendment that had not yet been instigated (as was the case given that there had been no follow-up on the agreement on the 28.9 constant and 100% cap).

Mr Stewart-Smith circulated a first draft of the Minutes of the meeting to the other Sonatrach attendees (Mr Stranks, Mr Bensalem, and Mr Benaoum) and to BP the day after the meeting asking for any comments. In the context of the fact that Mr Stewart-Smith is a solicitor, who would be used to taking an accurate note of a meeting it is reasonable to assume that he took care in the preparation of the minutes and that they are an accurate record (albeit inevitably a summary) of the meeting. As he circulated the draft Minutes the day after the meeting his memory, and those of all who attended, would no doubt have been fresh. Mr Stranks made some comments. BP submit, rightly in my view, that if he thought that the Minutes were incomplete, inaccurate, or unclear, he would have said so. He did not do so. As will appear this is of relevance in the context of certain answers he gave in cross-examination concerning what was discussed in the meeting. Mr Bensalem and Mr Benaoum, who also attended the meeting, do not appear to have made any comments. Again one would have expected them to do so had they regarded the Minutes to be incomplete, inaccurate, or unclear. Ms Torrance made some comments on 2 October.

After collating the various comments Mr Stewart-Smith circulated a second draft on 2 October. He stated in his covering e-mail that he had had taken on board everyone’s comments where he could, and had sought to produce "a simple factual record of what was discussed at the meeting and have carefully correlated [the Minutes] with my notes" (if that is a reference to manuscript notes, no such notes have been located, though it is possible he made notes on his laptop and then worked them up into the draft Minutes, subsuming any original notes). Ms Torrance made 3 further (minor) comments on 3 October. Given that both Mr Stranks and Ms Torrance made comments I consider that it is reasonable to assume that they reviewed the Minutes carefully. In all the circumstances the Minutes have the hallmarks of an accurate and complete record of the Meeting (always subject to the fact that as with any minutes, they are a record, rather than a verbatim account, of what was discussed).

It is apparent from a consideration of the Minutes, as quoted above, that BP raised the allocation of nitrogen costs in the meeting in the context of a settlement with Air Products, BP’s proposal as to settlement being conditional upon a reconciliation of nitrogen costs for the period July 05 to July 06 under the terms of the annual reconciliation procedures under the JSA but using the new nitrogen costs allocation formula (as Mr Stranks confirmed when cross-examined).

The Minutes record that "BP raised concerns about the fact that due to a manifest error in the JSA cost allocation formula, BP had been paying for nitrogen which BP believes is being used by Sonatrach and that a new formula for sharing nitrogen costs under the Joint Shippers Agreement (JSA) has been discussed on several occasions at the operating level and agreed in principle between Sonatrach and BP."

The question arises as to what the “manifest error” was that was being referred to. I am satisfied that the “manifest error” is a reference to the missing constant of 28.9. This conclusion is supported by a number of matters. First and foremost, and as I have already noted, “manifest error” is entirely apt to describe the typographic error on Mr Winstanley’s part as a result of which the 28.9 constant for the molecular weight of air was omitted from Factor Y in the formula. This had indeed been discussed “on several occasions at the operational level” and “agreed in principle” (and in fact) as I have identified. It is less apt to describe the 100% cap (given that that is not, strictly, an “error” but rather a consequence of the formula applied to the facts), though it is possible that the 100% cap was also being referred to in combination with the 28.9 point, given that the cap had also been discussed and agreed, as already identified.

However, “manifest error” in the JSA cost allocation formula is not apt to describe the consequences of the fact that it had transpired that Grain was targeting 51.1 not 51.41. The product of the formula was the result of the application of the formula not an error in the formula. Equally the operational approach which Grain had implemented after the JSA was signed (the Terminal was not commissioned until July 2005) cannot be described as an error in the contractual formula (other than in "loose language" as Mr Way put it in his evidence), still less a “manifest error”. More fundamentally, whilst a change in Wobbe value had previously been discussed (as identified above) it had not been agreed at all whether as an “agreement in principle” or otherwise. Equally, the result of using an operational Wobbe instead of 51.41 would be to increase the share of the nitrogen costs which were allocated to BP, and so to increase BP's costs (as recognised in Mr Way's Memo). So BP cannot have been referring to Wobbe when they said that due to a manifest error BP had been paying for nitrogen which BP believed was being used by Sonatrach.

Equally the Minutes in the next bullet point record that, "BP's proposal to share in the settlement of the Air Products claim is on the basis that a reconciliation of nitrogen cost allocation is carried out by BP and Sonatrach for last year (i.e. for the period July 05 to July 06) under the terms of the annual reconciliation procedures in the JSA but using the new nitrogen cost allocation formula. Since July 06 it appears that the Agent has been applying the new formula notwithstanding the terms of the JSA for the allocation of the nitrogen costs. BP have not calculated this amount but have an expectation that the reconciliation could be as much as £1.5 million".

This further confirms that what was being discussed was the constant of 28.9 (possibly in combination with the 100% cap). The reason that the proposed reconciliation would be for the period July 2005-July 2006 was that the Agent had already been using the "new formula" since July 2006. The "new formula" was the same new formula as referred to in the previous bullet point (which, as discussed above, was a reference to the addition of the 28.9 constant) and the Agent had been adding in the 28.9 constant since July 2006. The Agent had not been using an operational Wobbe since July 2005 (so the new formula cannot have related to that). The reference to BP not yet having calculated the amount, “but have an expectation that the reconciliation could be as much as £1.5 million” chimes with Mr Way's estimate in his July Memo that the reconciliation amount might be in the region of approximately £1.0-£1.5 million (and that was based on factoring in the missing 28.9 constant not any change to take account of operational Wobbe). In addition, any change from 51.41 to an operational Wobbe would increase BP's nitrogen allocation and share of costs. It would not result in a £1.5 million reconciliation payment due to BP. In context, the reference to the new formula being applied "notwithstanding" the terms of the JSA for the allocation of the nitrogen terms relates to the fact that the Agent had been adding in the missing 28.9 since July notwithstanding that a formal amendment to the JSA had not been progressed.

In the next paragraph of the Minutes it is recorded that Sonatrach agreed in principle with a historic reconciliation for the period July 2005 – July 2006, “as it was accepted that there was a manifest error in the formulae in the JSA, but stated that this calculation should be thoroughly checked first”. For the reasons already identified the “manifest error” being referred to is the missing 28.9 constant. The Minutes record that it was agreed that the Agent be instructed to carry out “such a calculation” and revert to BP and Sonatrach with its calculations and justification. The calculation, in context, was the correction of the error the formula, the adding in of the 28.9 constant and the implementation of the 100% cap (the latter two having been agreed between BP and Sonatrach, and the error having been recognised by the Agent as identified in Mr Way’s Memo). It was not an agreement to perform a reconciliation by reference to a different Wobbe number to that in the JSA formula (as also evidenced by the fact that the initial subsequent reconciliations done by the Agent and supplied to Sonatrach and BP, took account of the 28.9 constant and 100% cap, but did not utilise differing Wobbe values, as addressed below).

There was then an agreement to instruct the Agent to perform a reconciliation (which Sonatrach wanted carefully checked) and report to the parties. That was a reconciliation to take account of the missing 28.9 to get to a definite figure, rather than BP's £1.5 million estimate. After discussion of how the reconciliation amount would be paid once the Agent had calculated it, the Minutes record that someone "noted" that it would be inappropriate to use a new formula "ad hoc" until the JSA had been formally amended. BP commented that the new formula was being applied by the Agent despite the terms of the JSA (i.e. absent a formal amendment) so as to avoid building up a greater liability between BP and Sonatrach (i.e. in terms of the size of the reconciliation payment to BP). Again this confirms that the point under consideration was the 28.9, which generated reconciliation amounts in BP's favour, not operational Wobbe, which would have worked the other way.

The Minutes then record that Mr Stranks stated that the new formula was applied to the last month’s invoice (August) but approval had not yet been given by Sonatrach to this application, reference being made to Ian MacLeod at the Agency believing that Mr Stranks had given verbal approval but it was said that this was not the case. In fact, Mr Way had agreed the addition of the 28.9 in writing, and the Agent had been adding the 28.9 since July (with effect from the August invoice) which again evidences that what was being addressed was the 28.9 constant not Wobbe (the Agent had not been using an operational Wobbe since July).

Set against the background of the preceding paragraphs of the Minutes the sentence that, “IT WAS AGREED that the JSA would be modified so as to deal with this issue of the appropriate allocation of nitrogen costs”, evidenced the fact that the parties had agreed to amend the JSA formula to include the 28.9 constant and the 100% cap. There was no agreement in the meeting to amend the JSA to replace 51.41 with the monthly weighted average Wobbe value in fact blended to by Grain, as emerged from Mr Strank’s cross-examination, and as I have already found above, and for the reasons I have given.

That leaves the suggestion, advanced by Sonatrach in its oral closing submissions, that in the Meeting there was nevertheless “an agreement in principle to fix the formula for the allocation of the variable nitrogen costs, which included the Wobbe number issue.” (my emphasis). There is no support that there was any agreement, in principle or otherwise, in relation to the Wobbe issue in the Minutes of the Meeting. Indeed, I consider it is clear, for the reasons I have identified, that what was discussed and agreed was that the Agency would perform a reconciliation in a new formula that would take into account what had been agreed (i.e. the inclusion of the 28.9 constant, and the 100% cap). Nor do the Minutes support the submission that there was any discussion of the Wobbe issue at the meeting (Mr Stranks accepted in cross-examination that there was nothing in the Minutes about operational Wobbe). Nor is there anything in Mr Stranks’ notes to support that.

However, Mr Stranks maintained in his oral evidence that there was a wider discussion that included discussion of Wobbe, He asserted this in a number of passages in his evidence. For example, there were the following series of exchanges:

I found Mr Stranks’ evidence on this point to be lacking in detail and lacking in corroboration. The language of “I would contend” is not the language of clear recollection, and Mr Stranks himself expressed limits to his recollection as can be seen above. Later in his evidence he also stated that, “I have to say. I cannot recall the exact specifics”. If there had indeed been a discussion of Wobbe in the meeting, even more so if there had been any agreement in principle that the formula in the contemplated reconciliation to be undertaken be the Agent would extend to the Wobbe issue then it would surely have been recorded in the Minutes. It would have been an important discussion that Mr Stewart-Smith would have noted and recorded, and that others from both BP and Sonatrach would have remembered, and proposed as alterations to the Minutes. It would have been a glaring omission in the Minutes that would surely have been corrected. For the same reason Mr Stranks’ evidence about his note-taking, that, “I freely admit when I’m notetaking, when I’m actually engaged in conversation I don’t write notes”, is not an explanation for the absence of any reference to such a discussion in the Minutes, given the other persons present, and the opportunity for everyone (including Mr Stranks) to propose amendments to the Minutes.

If, as Sonatrach alleges, there was not only a discussion that extended to the Wobbe value used (despite that not being recorded in the Minutes), but it was agreed in principle that the formula in the contemplated reconciliation to be undertaken by the Agent would extend to the Wobbe issue then the Wobbe issue would surely have featured in the initial subsequent reconciliations supplied by the agent to Sonatrach and BP. It did not.

Mr Way circulated Version 1 to Sonatrach and BP on 20 October 2006. Version 1 added the missing 28.9 into the formula and used 51.41 ("Constant A" and "Constant B” in the box under "correction factor for Wobbe"). The reconciliation figure in Version 1 was £1,373,380.03 payable by Sonatrach to BP (within the £1.0-£1.5 million estimate in Mr Way's Memo, and reasonably consistent with the £1.5 million which BP mentioned in the Minutes). Mr Way circulated Version 2 (incorporating additional data) to Sonatrach and BP on 6 November. The figure had increased towards £1.5 million. Again, Version 2 added in the missing 28.9 and used 51.41.

Mr Way confirmed that when he produced Version 1 and 2 adding in the missing 28.9 and using 51.41 he thought that he was doing what the Agent had been instructed to do (“I would assume so, yes”). This is consistent with him having been instructed to produce a nitrogen reconciliation study following the 28 September meeting to add in the missing 28.9 but not instructed to use an operational Wobbe. In his witness statement Mr Stranks acknowledged that it was clear that neither Mr Way nor Mr Wood knew, in November/December 2006, about any agreement in September to use operational Wobbe. Indeed, Mr Wood apparently did not know anything about it until he read Mr Stranks' statement, whilst Mr Way made no reference to ever finding out. It is now clear that there was no such agreement – but even if there had been an agreement in principle, or contemplation that the reconciliation would address Wobbe values, then Mr Way would surely have been instructed to address such matters.

Since Mr Way (reporting to Mr Wood) was performing the very reconciliation exercise which had been agreed at the September meeting, it would obviously have been relevant for him to know about any agreement (or agreement in principle) to use operational Wobbe instead of 51.41. I agree with BP’s submission that the only plausible explanation for why he was not told is that there was no agreement (in principle or otherwise). Furthermore, had there been any agreement in principle to address operational Wobbe (or indeed any Wobbe value, other than the contractual constant of 51.41) then versions 1 and 2 plainly did not do so, and both BP and Sonatrach would surely have responded asking why Wobbe values had not been addressed. There is no such correspondence.

Operational Wobbe only came into the reconciliation exercise after an internal Sonatrach meeting on 4 December, when Mr Way was instructed (as ECC, working for Sonatrach, not as Agent working for the Co-Shippers) to investigate the effect of operational Wobbe. BP was not told that Mr Way was going to do this. Instead, it was agreed at this internal meeting that Mr Way would write to both BP and Sonatrach stating that the Agent would commence a 2005 annual reconciliation using "the formula currently in the JSA". Mr Way sent such a message the next day (5 December 2006).

Mr Way sent Version 3 to Sonatrach (but not BP) on 15 December. Version 3 enabled the user to switch between 51.41 and alternative constants of 51.1 (Grain's "optimum" or target Wobbe, according to, e.g., the March 2006 Monthly Report) and 51.25 (midway between the other 2 values). Using 51.1 brought the reconciliation amount down towards £0.5 million. This illustrated the point, indicated in Mr Way's Memo that using an operational Wobbe would allocate more nitrogen to BP's cargoes and so increase BP's share of the costs. Mr Way sent Version 4 to Sonatrach (but not BP) on 17 January. This version again allowed the user to choose between 51.41, 51.1, and 51.25. Mr Way also provided a "user guide".

By this time, BP was chasing for the reconciliation which Mr Way had indicated back in December was in-hand. Mr Way had in fact produced 3 new documents since then (Version 3, Version 4, and the user guide). But BP did not know that, because they had only been sent to Sonatrach. (Mr Way could not say why this was so when he gave evidence). Mr Way promised his best, and sent Version 5 to both Sonatrach and BP on 1 February shortly before the February meeting. Version 5 built on Version 4, with the same user guide and the ability to switch between 3 different values. It also contained a summary-sheet of 4 different scenarios. This aspect is addressed further when considering the next issue and BP’s alleged breach of a duty of good faith in relation to the February meeting. On any view this material was only supplied to BP shortly before that meeting.

At this point it suffices to note that in Mr Martinez’s email to his colleagues of 2 February 2007, he noted, “We were expecting SH to raise at some point that 51.1 should be used instead of 51.41. I think this is a fair request from them, and we should show an open mind and willing to work with SH if we accept this change”. This is consistent with Sonatrach not having raised the use of a Wobbe of 51.1 subsequent to the discussions in the summer of 2006 which have been referred to above. This comment is not consistent with Sonatrach having raised the use of operational Wobbe in the September meeting or at any time thereafter prior to version 5 of the reconciliation.

Other events subsequent to the meeting also support the conclusion that not only was there no agreement in the meeting to amend the JSA to replace 51.41 with the monthly weighted average Wobbe value in fact blended to by Grain, but that that there was no discussion, or agreement in principle, to address the Wobbe value as part of the reconciliation exercise.

On 4 October, Mr Lunn (one of BP's attendees on 28 September) had a meeting with Sonatrach. His notes record:"N2 Error in JSA Formula – JSA needs rectifying.

Some of the detail (e.g., the optimistic reference to the Agent producing a report within a matter of days) may possibly be from a meeting on 4 October rather than 28 September. However, the basic content of this note reflects the Minutes, and is indicative of an agreement that there should be a reconciliation – and a formal amendment – to take account of the omission of the 28.9 constant. There is no mention of any discussion of Wobbe values or that the reconciliation would extend to that.

Mr Lunn then circulated an e-mail within BP of "action points" from the 4 October meeting. He recorded: "N2 – error in formula, accepted by SH, SH waiting to receive report before approaching BP on how to resolve reconciliation and correct allocation going forward (AW). I agreed to the formula being kept as is now (wrong) until SH report issued and SH come back to us. Keeping ICF and Wobbe in formula as are (not as plant operates) which benefits BP." Mr Lunn's reference to keeping ICF and Wobbe as they were indicates that he did not understand that there had been any agreement to change either of these (or to address these in the reconciliation).

Mr Duncan was taking over from Ms Torrance around this time. She briefed him at handover meetings. At one meeting, she told him about the AP dispute and possible settlement, in terms which seem reminiscent of the Minutes. In that context, she referred to "wrong formula" and to Sonatrach owing BP about £1.5 million. That was the same estimated reconciliation referred to in the Minutes, and it referred to a reconciliation to take account of the missing 28.9. At the next meeting, she said that the parties had agreed in principle that the formula was wrong and that this was "noted in committee notes", which, in context, must mean the Minutes. There was no reference to Wobbe values, and Ms Torrance did not tell Mr Duncan that there had been an agreement on 28 September to change to operational Wobbe. This is hardly surprising as there was no such agreement (as emerged when Mr Stranks was cross-examined), but even if there had been a discussion on Wobbe or an agreement in principle to address Wobbe values as part of the reconciliation, Ms Torrance would surely have mentioned that during the hand over (particularly as it is clear that Ms Torrance was telling Mr Duncan about points which had been discussed at the meeting).

Having had regard to the events before the meeting, the Minutes of the meeting, the evidence that I heard as to what was discussed at the meeting and events after the meeting, I conclude, for the reasons given above, that Mr Way was not instructed following the September meeting to undertake a reconciliation that extended to the Wobbe issue and that the reason for this was that there was no agreement in principle at that meeting that the reconciliation should extend to a consideration of the Wobbe issue. To the extent that Mr Stranks’ evidence was that there was any such agreement in principle in the Meeting I consider that Mr Stranks is mistaken in his recollection. Equally I consider that Mr Stranks is mistaken in his recollection that there was any discussion of Wobbe values in the meeting. If there had been any such discussion it would in my view have been recorded in the Minutes, or if not recorded initially, then added in when the draft was circulated.

In the above circumstances issue 3 is to be answered in the negative. BP and Sonatrach did not agree at the September 2006 Steering Committee meeting to amend the JSA to replace 51.41 with the monthly weighted average Wobbe value in fact blended to by Grain. Equally I find that there was no “agreement in principle to fix the formula for the allocation of the variable nitrogen costs, which included the Wobbe number issue”, nor (if it be relevant to any issue) was there any discussion of the Wobbe value or operational Wobbe in the meeting. It is clear from the totality of the evidence that there was no agreement at the September 2006 Steering Committee meeting to use an operational Wobbe.

Sonatrach defines issue 4 in these terms in its Written Closing Submissions (consistent with its pleaded case):

It will be seen therefore that there are two specific alleged breaches of the obligation of good faith, on two specific dates, and by reference to specific conduct (or lack of it) on those dates, namely at the September 2006 Steering Committee meeting and at the February 2007 Steering Committee meeting.

It will be recalled that in the context of the Co-Shippers Steering Committee, Clause 1.1.5 of the JSA provides that the Co-Shippers' representatives must, "consider and propose for consideration by the Co-Shippers any amendments to [the JSA] that may be considered necessary (i) for commercial, technical and operational reasons from time to time."

Article 1.7.4 of the JSA provides that "Each Co-Shipper shall act as a Reasonable and Prudent Operator in performing its obligations under this Agreement." Under Part F1.1 "Reasonable and Prudent Operator" is defined in these terms:-

The obligation on each party under the JSA is to act in good faith when performing obligations under the JSA. As it is Sonatrach that is alleging breach of this term it is for Sonatrach to plead, and prove, breach of an obligation, which would involve identification of one or more obligations under the JSA being performed by BP and proof that BP had breached that term by not acting in good faith, by not acting in a particular way at a particular time or times.

There was some debate before me as to whether there can be a breach of a contractual duty of good faith without "bad faith". In this regard I was referred by BP to the judgment of Vos J (as he then was) in CPC Group Ltd v Qatari Diar Real Estate Investment Co [2010] EWHC 1535 (Ch) at [246]. In that case Vos J did not require, "to decide whether this obligation could only be broken if [the defendant or the claimant] acted in bad faith". He considered that "it might be hard to understand, as Lord Scott said in Manifest Shipping, how, without bad faith, there can be a breach of a 'duty of good faith, utmost or otherwise'". On the facts of that case he found that there had been no breach of the obligation of utmost good faith (the express term in that case) as so interpreted.

In response to BP's submission that the obligation to act in good faith to perform obligations under the JSA can only be broken if BP has acted in "bad faith", Sonatrach accepted in their Written Closing Submissions that "that may be so", but submitted that "bad faith" as "good faith" must be understood by reference to the case law on express terms concerning, "good faith", relying amongst other dicta, upon what was said by Morgan J in Berkeley Community Villages Ltd v Pullen [2007] EWHC 1330 (Ch) at [97], in relation to the express term there under consideration, which he found imposed "... a contractual obligation to observe reasonable commercial standards of fair dealing in accordance with their actions which related to the Agreement and also [required] faithfulness to the agreed common purpose and consistency with the justified expectations" of the parties.

This is not a case in which it will be necessary to grapple with the precise boundaries of what is required by a contractual duty to act in good faith. As appears below, Sonatrach simply does not begin to make out its pleaded case on breach of the contractual term on the facts of this case in relation to either of the two pleaded alleged breaches and dates in question.

I would, however, make the following point. Whatever the precise boundaries of a contractual duty of good faith, an allegation of breach of such a duty, put at its lowest, involves an assertion that the other party has not acted in good faith. This is a serious allegation. In such circumstances the party making such allegation should plead its case with proper particularity so that the other party knows the case it has to face as to what breach is alleged to have occurred and when, and the party making such allegation will generally be held to its pleaded case as to what breach of duty is said to have occurred, and when, due to the nature of the allegation being made, and the fact that the evidence that will be called, will have been called to rebut that specific allegation, and only that allegation.

As I have already recounted, in closing, and in the light of the totality of Mr Stranks' evidence, Mr Harris conceded on behalf of Sonatrach, that he could not make good Sonatrach's pleaded case, on the evidence, that there was an agreement to change to any particular Wobbe number at the September 2006 Steering Committee Meeting. He submitted, however, that at the meeting there was nevertheless “an agreement in principle to fix the formula for the allocation of the variable nitrogen costs, which included the Wobbe number issue.”

In the light of his concession, and the way Sonatrach was now putting its case, the pleaded case as to breach of the good faith obligation in relation to the September 2006 Steering Committee Meeting must also fall away. This was accepted by Mr Harris in closing:-

Sonatrach's pleaded case on good faith in relation to the September 2006 Steering Committee Meeting (at paragraph 88A of the Re-Amended Defence and Counterclaim) was that a proposal to change to operational Wobbe was considered at the meeting but that in not agreeing it BP breached the obligation of good faith in the JSA. Yet on either Sonatrach's case in closing about the 28 September meeting and what was allegedly agreed in principle, or upon the findings of fact I have made in relation to the meeting, there can have been no breach of a duty of good faith by BP at the meeting on 28 September. As to the former, in advocating an agreement in principle at the meeting to fix the formula for the allocation of variable nitrogen costs including the Wobbe number issue, Sonatrach’s case that there was a breach of duty by BP in rejecting an agreement at the meeting to change to operational Wobbe necessarily falls away (as Mr Harris accepted). As to the latter, I have found that not only was there not an agreement to change to operational Wobbe, but that there was neither an agreement in principle to fix the formula for the allocation of the variable nitrogen costs including the Wobbe number issue, nor any discussion of operational Wobbe at all. In such circumstances there is in any event no evidential basis for any breach of good faith by BP at the 28 September meeting. In short the evidence simply does not support the allegation of breach of a duty of good faith by BP at that meeting.

In the above circumstances I find that BP did not breach any obligation of good faith in relation to the September 2006 Steering Committee Meeting.

Sonatrach's alternative pleaded case (at paragraph 88B of the Re-Amended Defence and Counterclaim) is that there was an agreement to consider a change to operational Wobbe at the September 2006 Steering Committee meeting, that at the next meeting of the Steering Committee on 6 February 2007 the issue of the amendment of the Wobbe index value in the JSA arose for discussion in the context of Agent's calculations, and that the BP's representative at the meeting (Mr Duncan), "failed and refused to consider that amendment in good faith (or at all) and agree it, and [BP] accordingly was in breach of its obligations under the JSA."

I have, of course, found that there was no such agreement (in principle or otherwise) to consider a change to operational Wobbe at the September 2006 Steering Committee meeting, but even if there had been I consider the allegation of breach of a duty of good faith in relation to the subsequent 6 February 2007 meeting fails on the facts in terms of what happened at that meeting, which has to be viewed against the backdrop of events leading up to that meeting, and the timing of the provision to BP of a reconciliation raising a change in the Wobbe number.

It will be recalled that Mr Way circulated Versions 1 and 2 of the reconciliation to BP on 20 October and 6 November 2006 respectively and these both added the missing constant of 28.9 and used a Wobbe of 51.41. Different Wobbe values initially only came into the reconciliation internally at Sonatrach, after an internal Sonatrach meeting on 4 December, when Mr Way was instructed (as ECC, working for Sonatrach, not as Agent working for the Co-Shippers) to investigate the effect of operational Wobbe. BP was not told that Mr Way was going to do this. Instead, it was agreed at this internal meeting that Mr Way would write to both BP and Sonatrach stating that the Agent would commence a 2005 annual reconciliation using "the formula currently in the JSA". Mr Way sent such a message the next day (5 December) stating, "Please note that the annual nitrogen reconciliation for both 2005 and 2006 will be performed using the current formula in the JSA".

Mr Way sent Version 3 to Sonatrach (but not BP) on 15 December 2006. Version 3 enabled the user to switch between 51.41 and alternative constants of 51.1 (Grain's "optimum" or target Wobbe, according to the March 2006 Monthly Report) and 51.25 (midway between the other 2 values). Mr Way sent Version 4, which again allowed the user to switch between 51.41, 51.1 and 51.25 to Sonatrach (but not BP) on 17 January.

Rupal Patel of BP chased Mr Way on 23 January 2007, stating that BP should have received something for the 2006 reconciliation by now, and asking for an update. Mr Way had, of course, produced three new documents since his email to BP on 5 December 2006, namely Version 3, Version 4, and the accompanying user guide, but BP did not know of these documents (and so had not had chance to consider them) as Mr Way had only sent them to Sonatrach. When cross-examined Mr Way said he couldn't answer why they were not sent to BP. The Agenda for the 7 February Steering Committee meeting was circulated on 26 January 2007. It contained a number of items including agent outsourcing and secondary capacity issues. Under "Joint Shippers Agreement" was the item “Annual reconciliation under the JSA (including allocation of historic nitrogen use)". No reconciliation had yet been received from Mr Way.

Finally, at 5.52pm on Thursday 1 February 2007, Mr Way sent a reconciliation (that is Version 5) to Sonatrach, and this time, BP. Like Version 4 it allowed the user to switch between 51.41, 51.1 and 51.25, and was accompanied by the same user guide as accompanied Version 4 (and which BP had not seen). It also contained a summary-sheet of 4 different scenarios [E8/2267]. In his covering email Mr Way stated:

In his internal email the next day, Friday, at 5.54pm, Mr Martinez sent an email to Mr Cockcroft cc'd to Mr Duncan with his comments on the reconciliation. I have already referred to this email earlier in my judgment. In it Mr Martinez stated amongst other matters:

Mr Duncan confirmed in his evidence that the "we" did not refer to him. Mr Duncan's evidence, which I accept, was that he had not been expecting the Agent to start altering Wobbe values, and that he felt rather ambushed by the material. I consider that such sentiments were entirely understandable. There is nothing to suggest that BP (still less Mr Duncan) was aware of the work that Mr Way had been doing for Sonatrach using different Wobbe numbers, and it was BP who had chased for a reconciliation, and very shortly before the meeting one was finally provided.

The February 2007 Steering Committee meeting was to be held on Tuesday 7 February 2007 in Algiers. Mr Duncan had to prepare for that meeting, and travel to Algiers in advance of the meeting. His evidence was that the reconciliation was simply not on his radar as something that was as urgent as secondary capacity mechanism and agent outsourcing, and he could not absolutely confirm whether he would have looked into the detail of the spreadsheet on the Friday or on the flight on the Monday. His evidence was that he had not had an opportunity really to form an opinion on a change to operational Wobbe, and that when he was going into the meeting this wasn't something that he was expecting to be agreed or even possible to take a view on at that time. I accept Mr Duncan's evidence as an accurate reflection as to his state of mind at the time, and one that I consider was understandable. The reconciliation was received very late, and the possibility of a change to a different Wobbe or an operational Wobbe was clearly a matter of large potential financial significance in relation to a long term agreement which it would take time to consider.

The breadth of subjects discussed at the meeting on 7 February 2007 is apparent from the fact that the minutes run to some eleven pages, with secondary capacity being discussed first and at length. The annual reconciliation is not reached until the end of page six of the minutes, and it is clear that the discussion was a brief one. The minutes record that BP (i.e. its attendee Mr Duncan) had not had chance to review the Agent's work product yet, given that they had only received it on 1 February. The action item was that BP and Sonatrach would revert to the Agent and each other with comments on the nitrogen spreadsheets and to arrange meetings with the Agent for clarification if necessary.

Sonatrach's pleaded allegation is that at the 7 February Steering Committee meeting Mr Duncan, "failed and refused to consider [the] amendment in good faith (or at all) and agree it and the Claimant accordingly was in breach of its obligations under the JSA” (sub-paragraphs 88(c) and (d) of the Re-Amended Defence and Counterclaim.

This allegation jars with what occurred at the meeting, and with the agreed action item. There was no such failure or refusal. The stance which I find was adopted, and which reflected the factual position as to what Mr Duncan had considered, was simply that BP had not had chance to review the Agent's work product yet given that BP had only received it on 1 February 2007, and BP and Sonatrach had agreed to revert to each other and the Agent with their comments. When he gave evidence on behalf of Sonatrach, Mr Wood expressed the view that if BP felt that at the meeting that they had not had enough time because it was only a few days before the meeting to have received the document, "then that's perfectly reasonable". Whilst Mr Wood's personal views (as his own personal views) are neither here nor there, I consider that his views reflect those that would be held by a reasonable counter-party in the factual situation that I have found to exist. I accept that what Mr Duncan said accurately reflects what he felt at the time (i.e. that he had not had sufficient time to consider the reconciliation), and I consider that then to adopt such a stance in the meeting (coupled with the agreed action item) was a perfectly reasonable stance and does not begin to give rise to any suggestion of a breach of a duty of good faith under the JSA.

That then is the factual reason why there was no breach of the obligation of good faith under the JSA at the February 2007 Steering Committee meeting. As the claim fails on the facts, wider issues concerning the duty of good faith do not arise for consideration. However as they were argued before me I will briefly express my views in relation to such matters.

The allegations of breach of a duty of good faith are only of potential relevance to the outcome of the dispute if Sonatrach has lost its arguments on construction, implied terms, and the September 2006 Steering Committee meeting, which I have found it has. On this scenario BP is correct that 51.41 is a constant and BP has a, prima facie, vested contractual right to a nitrogen allocation calculated on the basis of a fixed number. In this scenario Sonatrach has to place its argument very high - namely that the contractual obligation of good faith in the JSA means that BP is obliged to relinquish its contractual right for Sonatrach's benefit, and to its own financial detriment.

Good faith does not normally require a party to surrender contractual rights - see the judgment of Stephen Furst QC (sitting as a Deputy High Court Judge) in Gold Group Properties Ltd v BDW Trading Ltd [2010] EWHC 1632 (TCC). In that case he referred to Automasters Australia Pty Ltd v Brunbess Pty Ltd [2002] WASC 286 in which Hasluck J cites the observations of Barrett J in the Supreme Court of New South Wales in Overlook v Foxtel, (2002) Aust Contract R 90-143 at [65-67]:

“It must be accepted that the party subject to the obligation is not required to subordinate the party's own interests, so long as pursuit of those interests does not entail unreasonable interference with the enjoyment of a benefit conferred by the express contractual terms so that the enjoyment becomes (or could become) … ‘nugatory, worthless or, perhaps, seriously undermined’ … the implied obligation of good faith underwrites the spirit of the contract and supports the integrity of its character. A party is precluded from cynical resort to the black letter. But no party is fixed with the duty to subordinate self- interest entirely which is the lot of the fiduciary … The duty is not a duty to prefer the interests of the other contracting party. It is, rather, a duty to recognise and to have due regard to the legitimate interests of both the parties in the enjoyment of the fruits of the contract as delineated by its terms.”

If BP had a vested contractual right to a calculation in accordance with a 51.41 constant then I do not consider that Article A1.7.4 read together with Clause 1.1.5 of the JSA required BP, as a Reasonable and Prudent Operator, to give up that right in the discharge of its duties under Clause 1.1.5 whereby the Co-Shippers representatives must "consider and propose for consideration by the Shippers any amendments to [the JSA] that may be considered necessary for commercial, technical and operational reasons from time to time."

Article 1.7.4, and the definition of a Reasonable Prudent Operator, only imposes an obligation on a party to act in good faith when performing JSA obligations. In other words, there is no free-standing obligation of good faith. It is therefore for Sonatrach to identify a relevant obligation that BP is called upon to perform under the JSA. It appears that Sonatrach alleges that BP is contractually obliged to consider any proposed amendment which was raised before the Steering Committee (and indeed determine it in a particular way). However, the JSA does not say that (expressly or as a matter of implication). Furthermore, Sonatrach would be obliged to go so far in its argument as to say not only would BP have to consider it, but would have to agree an amendment even if it was contrary to its existing contractual entitlement. In my view far clearer words would be needed in the JSA if the parties were to be found to have contracted out of their right at common law to accept or reject any proposed amendment to a contract.

The parties agreed to the establishment of the Co-Shippers Steering Committee in Article 1.1, which provided that that Committee shall, subject to Section E15 (i.e. no amendment or variation of the JSA unless in writing and signed by or on behalf of each Co-Shipper), "consider and propose for consideration by the Co-Shippers any amendments to the JSA that may be considered necessary (i) for commercial, technical and operational reasons from time to time; and/or (ii) as a result of any amendments that may be made from time to time to the Services Agreement, the TSA and the Network Code." Thus the role of the Steering Committee was to propose contractual amendments to the parties themselves if they were "necessary" for commercial, technical or operational reasons. The power to decide upon an amendment remained with the Co-Shippers, not the Steering Committee. There was no fetter in the JSA on the contractual freedom of the parties to accept or reject proposals made by the Steering Committee. Put another way, the terms relied upon by Sonatrach do not amount to a contacting out by the parties of their common law rights and an assumption of an obligation to consider or accept a proposal made by the Steering Committee.

I would also add that in the present case, a change to operational Wobbe was not, in circumstances where BP had a contractual entitlement to a nitrogen allocation on the basis of a fixed number, "necessary" for commercial, technical or operational reasons in any event.

Sonatrach also makes other allegations in the context of any alleged wider breach of duty of good faith which do not arise for consideration given the findings of fact that I have made in relation to the September 2006 and February 2007 Steering Committee meetings. For example, it says that BP has acted inconsistently in standing on its contractual rights under the JSA whilst supporting with other shippers, including Sonatrach, Grain's 2011 decision to switch from 51.41 to operational Wobbe in its CAPs, and complaining with Sonatrach to Grain (but without any success) about its level of nitrogen consumption, including in their joint letter in June 2013 to Grain.

I do not consider that there would have been anything in either of these points. As to the former there is nothing wrong in principle in BP adopting inconsistent stances - it is arguing matters in relation to different parties and different contracts. The JSA and the STA (including CAPs which Grain issues under the STA) are separate contractual arrangements, even if they are linked in the sense of relating to the same Terminal. D2.2.2(ii)(b) and the CAPs were never identical in any event (for example, because the CAPs made no allowance for ICF – Factor X – and used the Wobbe of the last unloaded cargo, rather than a monthly average). The fact that a change was made to one of these contracts did not mean that BP was bound, in good faith, to agree to change the other.

As to the latter, both BP and Sonatrach were troubled by Grain's nitrogen consumption, and would have liked Grain to use less. If Grain had agreed, then the total nitrogen costs which Grain would have passed to the Co-Shippers as joint Phase 1 Shipper would have reduced. That would have reduced the quantum of Sonatrach's allocation. I do not consider that BP can be criticised for endeavouring, together with Sonatrach, to achieve an outcome with Grain that would have been beneficial to both BP and Sonatrach even if it adopted a different approach in its relationship with Sonatrach. Equally the fact that BP's efforts were unsuccessful does not mean that BP was bound, in good faith, to amend the JSA to provide for operational Wobbe.

The above observations on the scope of any duty of good faith are academic, because whatever the scope of such duty, Sonatrach has failed to prove any breach of such obligation on the facts in relation to either the September 2006 Steering Committee meeting or the February 2007 Steering Committee meeting for the reasons I have already identified. Quite simply BP did not breach any duty of good faith as a matter of fact on either date, and for the reasons I have given.

In closing, and no doubt recognising the difficulties it faced in relation to its pleaded case concerning the September 2006 and February 2007 Steering Committee meetings, Sonatrach endeavoured to widen its case submitting that "in flagrant disregard of its obligations in relation to the Steering Committee it agreed to establish to address precisely these kinds of operational issues, BP has steadfastly refused to consider at all - let alone in good faith - whether an amendment to the 51.41 Wobbe value in the second formula is necessary for operational reasons" (paragraph 6.9 of Sonatrach's Written Closing Submissions). There are at least two problems with this submission, quite apart from the factual investigations that would have been necessary had such a case been admitted.

First, that is not Sonatrach's pleaded case, and is not the case that the witnesses were called to meet. Sonatrach did not, in the event, apply to amend to plead such a case. If it had done so I do not consider it would have been an appropriate case for the granting of permission at such a late stage, and having regard to the nature of the allegation as well as its generality, and Sonatrach's failure to identify when any alleged breach is said to have occurred. For the reasons I have already identified, and whatever the precise scope of the obligation of good faith, an allegation of breach of such an obligation is an allegation that a party has not acted in good faith. Any such allegation is a serious allegation that should be specifically pleaded with proper particularisation, so that evidence directed at that specific allegation can be adduced. Any new case would be far more wide ranging over an extended time period, and the evidence was not directed at that nor should cross-examination be directed at that (as opposed to the pleaded issues). In addition, any such plea suffers from a lack of particularity, and the vice of a failure to identify and plead when any alleged breach occurred. In any event any application, after conclusion of the evidence, would have been far too late depending, as it does, on factual allegations.

Secondly, and fundamentally, Sonatrach did not identify when the alleged breach (or breaches) are alleged to have occurred. When pressed in oral closing Mr Harris rightly said, "My Lord, the criticism that might properly be made of my pleading I suppose, as you put to me earlier, is I haven't said when the breach crystallises. Well at the latest it crystallised when BP sends the invoice." The difficulty Sonatrach had in identifying the moment or moments in time when there was any such breach simply highlights the vagueness of its unpleaded case. I cannot in any event see how the presentation of an invoice based on one of the reconciliations set out by the Agent, could itself be characterised as a breach of the duty of good faith.

I accordingly would not have allowed any such amendment, but had such amendment been allowed it would have been necessary to consider events after the February 2007 Steering Committee meeting. That evidence does not support the conclusion that there was any breach of a duty of good faith on BP's part after the February 2007 Steering Committee meeting. The meeting agreed an action point for both parties to revert with comments on Version 5. Yet it would appear that neither party pressed the topic in the following months, it seems because the secondary capacity mechanism arrangements were beginning to come to the fore. The Agent continued to try to get complete and accurate data from Grain, with BP trying to prioritise the reconciliation process from early 2008. For example, on 13 February 2008, BP wrote to Sonatrach requesting that the completion of any outstanding reconciliations under Section D2.2.3 of the JSA be given high priority and completed. The Agent produced a 2005-2007 reconciliation in July 2008 and then the September 2008 "4 scenarios" document on the basis of which BP invoiced Sonatrach in November 2008, and which is the subject matter of the next issue. I do not consider that there is anything in this period which could be characterised as a breach of any duty of good faith on BP's part.

I accordingly answer issue 4 in the negative. BP and Sonatrach did not agree at the September 2006 or February 2007 Steering Committee meetings to amend the JSA to replace 51.41 in D2.2.2(ii)(b) with the monthly weighted average Wobbe value in fact blended to by Grain, and BP did not thereby breach any obligation of good faith under the JSA. By the same token, and in relation to any wider allegation that BP breached any obligation of good faith in not agreeing to so amend the JSA at the time of the meetings or after the February 2007 Steering Committee meeting, such allegation is not pleaded, but would have failed on the facts in any event.

On the basis of the findings I have made in relation to Issues 1 and 2, and as Mr Harris rightly acknowledged in closing on Sonatrach's behalf, this issue is only concerned with whether BP is entitled to interest on the sums invoiced by BP in November 2008, as BP would be able to invoice Sonatrach, at the present time, in respect of the sums claimed based on the construction of the JSA I have found. This essentially turns on whether BP was entitled to invoice Sonatrach, under the JSA, in November 2008.

BP invoiced Sonatrach on 12 November 2008. Section D1.6.3 of the JSA provides that, "The due date in respect of an invoice issued pursuant to section D1.6.1 above is the tenth (10th) day after such invoice is deemed to be received under Section E13".

Section D1.6.8 provides, "Where any amount payable under an invoice issued by a Co-Shipper under this Agreement is not paid to the Invoicing Co-Shipper on or before the due date, the paying Co-Shipper shall pay interest, before and after judgment, at a default interest rate of LIBOR plus three (3) per cent., on the unpaid amount from the due date until the day on which payment is made ("Default Interest")".

In relation to invoicing Section D1.6 (entitled "Invoices between Co-Shippers and Payment") provides at Section D1.6.1:

In relation to the role of the Agent Sections D2.2.3 and D2.2.4 of the JSA provide as follows:

So far as the Agent's duties are concerned Section A. 1.6.2 of the JSA provides, amongst other matters:

The Agent produced a reconciliation spreadsheet to the parties on 14 July 2008. Sonatrach commented on 19 August 2008, and whilst acknowledging that, "the Agent has correctly determined the reconciliation amount between BP and Sonatrach in respect of nitrogen costs, in line with the provisions of the JSA" it is clear that this was not an acknowledgment of the correctness of the calculations as Sonatrach went on to propose that the Agent repeat the calculation using 51.1, and suggested that the Agent make an analysis of the actual operational Wobbe as reported by Grain in their monthly reports to determine whether operational Wobbe was a constant value of 51.1 or whether the value varies over time, Sonatrach stating that in that case the correct monthly operational Wobbe should be used.

On 8 September 2008 the Agent emailed BP and Sonatrach with a 2005-2007 reconciliation spreadsheet which incorporated suggested changes, error corrections and sensitivities and other changes indicated in Mr Way's covering email.

Under the heading "LIN Costs reconciliation" Mr Way stated as follows in his covering email:

BP's invoice of 12 November 2008 invoiced on the basis of reconciliation 3 (which accords with my findings in relation to issues 1 and 2 - i.e. using a constant of 51.41 as per the JSA) coupled with the correction for the molar weight of air and the cap (which had been agreed between the Co-Shippers as I have found above). The invoice provided, under "Description":

Sonatrach submits that the Agent never confirmed the excess amount due under Section D2.2.4, and that BP's invoice is invalid. It says that there is a simple framework which requires the Agent to calculate the amounts due as between the Co-Shippers, confirm this amount, and this then allows the Co-Shippers to invoice each other (as the case may be). It says that this avoids disputes between Co-Shippers as to the amounts due, by allowing an independent agent to act as an intermediary. Sonatrach points out that once it became apparent that there were difficulties with the application of the formula, the Agent raised the issue with the Co-Shippers. The Agent also provided assistance to the parties in dealing with the dispute and produced a "first draft" of the reconciliation to the parties on 27 June 2008, a further draft of an annual reconciliation on 14 July 2008, and the 2005-2007 spreadsheet on 8 September 2008, quoted above, with the "four different reconciliations". Sonatrach submits that at no point in time, in this correspondence, was the Agent confirming to the parties the precise amount to be invoiced. Rather, the Agent was giving different options to the parties. Despite this, BP sought – unilaterally – to rely on its preferred scenario in the Agent’s spreadsheet alleging that this constituted a determination by the Agent. In such circumstances Sonatrach submits that the invoices issued by BP were without any contractual right to do so, and without any confirmation by the Agent, with the result that they are invalid.

In the alternative, Sonatrach submits that if the Agent had purported to issue four contradictory determinations in the spreadsheet, with an unfettered right for BP to choose the most advantageous determination, then the determination on which BP relied to issue invoices would in any event be invalid. First, because such a determination would be based on an incorrect construction of, or in breach of an implied term in, the variable nitrogen formula (this argument falls away in the light of my findings on issues 1 and 2). Secondly, because in issuing such a determination (effectively requiring Sonatrach to pay for BP’s nitrogen), the Agent would not be acting in a “fair and equitable way, for the benefit of both Co-Shippers”, as it is said a reconciliation requiring Sonatrach to pay for BP’s nitrogen is neither equitable nor fair. It is said that it is also inconsistent with D2.2.4 of the JSA which requires amounts payable to be linked to the “use of the Terminal by the Co-Shippers”.

I consider that BP was entitled to invoice Sonatrach on 12 November 2008 in the amounts that it did and that its invoice (and subsequent invoices identified below) were valid for reasons which I will now identify.

First, I consider that BP is correct in its submission that BP's invoice was for amounts due under annual reconciliations under Section D2.2.3 (and not Section D2.2.4). The former is dealing with annual reconciliation. The latter is dealing with excess amounts due to a Co-Shipper by the other Co-Shipper. The reconciliations in the 8 September 2008 spreadsheet are annual reconciliations, and Section D1.6.1 provides that an Invoicing Shipper may deliver to a Paying Shipper an invoice of the amount which may be invoiced by such Invoicing Shipper as notified by the Agent in a reconciliation statement under Section D2.2.3. I consider that the 8 September 2008 and accompanying spreadsheet are a reconciliation statement. Section D2.2.3 does not require a "confirmation of the excess amount due" (which only appears in D2.2.4). Equally I do not consider the fact that the Agent included four reconciliations means that any invoice based on one of those reconciliations which is calculated in accordance with the JSA (or "otherwise agreed between the Co-Shippers" - i.e. as to the 28.9 constant and the cap) is not a "reconciliation of all the costs or amounts owed" for the purposes of Section D2.2.3 and D1.6. Additionally, what is owed is ultimately a question of the proper construction of the JSA - that is not a matter for the Agent, or indeed the Steering Committee. Reconciliation 3 does reflect the amount owed by Sonatrach to BP under the JSA.

Secondly, if Section D.2.2.4 were, contrary to the above, to be in point, then I consider that the Agent's reconciliations in the 8 September spreadsheet are to be regarded as confirming the amount due. The Agent's role and function was to perform the calculations which it did correctly, but only one of them (Reconciliation 3) accords with the terms of the JSA (as I have found them to be). As identified above it was not the role or function of the Agent (or the Steering Committee) to determine what the JSA meant. Ultimately it is a matter of construction for the Court. In circumstances where reconciliation 3 does accord with the terms of the JSA as I have found them, I consider BP was entitled to invoice Sonatrach based on that reconciliation which does correctly state the amount due. In this regard I do not consider it relevant that the Agent was unsure itself as to which reconciliation accorded with the terms of the JSA and sought guidance from the Steering Committee. Which reconciliation complies with the terms of the JSA is a question for the court. It would be a strange result if the reconciliation provided by the Agent did comply with the JSA, and yet BP was not entitled to invoice on it.

Thirdly if, contrary to the above, Section D2.2.4 were to be in point and required the Agent to construe the JSA, determine its proper construction, and issue one reconciliation based on the same, then the contractual machinery has broken down as the Agent was either unwilling or unable to do so. In circumstances where an entity that is to make a calculation that is required under a contract fails or refuses to do so in accordance with the contractual procedure, and the court is in a position to do so based on its construction of the contract, then the court will substitute a calculation that complies with the terms of the contract - here a reconciliation conforming to D2.2.2(ii)(b) and the agreed 28.9 constant and 100% cap - see the discussion in Chitty on Contracts 32dn Edition at 2-138, and the cases there cited, including Sudbrook Trading Estate Ltd v Eggleton [1982] 1 AC 493.

As for the other points made by Sonatrach against BP, if reconciliation 3 is a reconciliation which complies with the JSA (which I have found it is) then the Agent cannot have been in breach of its JSA Schedule 1.7 obligation to carry out its duties, "in accordance with the provisions of this Agreement" and "in a fair and equitable way, for the benefit of both Co-Shippers" when it performs its duty, "in accordance with, and subject to, the provisions of this Agreement or as otherwise agreed between the Co-Shippers" (A1.6.2). As reconciliation 3 complies with D2.2.2(ii)(b), as amended by the parties by the addition of the 28.9 constant and the 100% cap, such compliance with A1.6.2 cannot amount to a breach of Schedule 1.7. Put another way an Agent will be carrying out its duties in a "fair and equitable way" if it produces a reconciliation which accords with the JSA and further agreements in relation to the 28.9 constant and 100% cap. In any event any breach by an Agent of its duties cannot invalidate a reconciliation which does undertake a calculation in accordance with D2.2.2(ii)(b). Finally, all nitrogen payments to GLNG (which are the subject matter of D2.2.4) are "in relation to the use of the Terminal by the Co-Shippers" construing such words in accordance with their normal and natural meaning.

After receipt of BP's 12 November 2008 invoice Sonatrach rejected that invoice, asserting that actual operational Wobbe should be taken into account. Correspondence ensued, without progress being made save that Sonatrach agreed that £128,404.69 was due and paid it. BP subsequently re-issued the November 2008 invoice for the 2005-2007 reconciliation to take account of Sonatrach's £128,404.69 payment and the Second Defendant's accession to the JSA, splitting the reissued invoice into two, covering nitrogen and non-nitrogen (e.g. power) costs in invoices dated 26 March 2010 in an amount of £1,855,920.50 and 7 September 2010 in an amount of £830,786.84. Each invoice bore a similar description to that quoted above in relation to the 12 November 2008 invoice. The Agent produced further nitrogen reconciliations for the years 2008-2011, each with the same four reconciliations. On 20 June 2013 BP issued an invoice for the reconciliation 3 amount in the sum of £675,508.82.

BP's claim is based on the invoices dated 26 March 2010, 17 September 2010, and 28 June 2013, in the total amount of £3,362,216.16 (including VAT), as set out at paragraph 6.3 of the Particulars of Claim.

For the reasons set out above, in relation to issue 5, I find that these invoices issued by BP are valid and enforceable, and that BP is entitled to interest in accordance with D1.6.3 and D1.6.8 of the JSA.

I would hope that the parties will be in a position to agree an Order reflecting my findings, interest to judgment, and the incidence of costs in the light of my findings, but if any issues remain outstanding I will hear argument on such matters from the parties.