La demanderesse et les défenderesses avaient conclu un accord de coentreprise en vue de la prospection et de l'exploitation de gisements de pétrole. Le demanderesse a fait valoir qu'elle avait été privée du droit de développer et d'exploiter deux des gisements, A et B, car ce droit avait été accordé en exclusivité à la défenderesse n° 2 par la défenderesse n° 1 (Etat X). La demanderesse réclamait en conséquence des dommages-intérêts pour son préjudice passé et futur. Le tribunal arbitral a considéré à la majorité que, conformément à la loi applicable, des dommages-intérêts pour un manque à gagner futur pouvaient être alloués, à condition qu'il y ait eu une réelle probabilité de perception des revenus en question par la demanderesse en l'absence de rupture de contrat.

'Future production (volume) and discount factor

The Claimant relies upon production profiles prepared by [its petroleum engineering expert, E] in relation to the [A] fields. [Expert E] calculated the future production of this field using a probabilistic analysis at P50 and P90 levels of probability. The P50 profile was 1,497,703 barrels (bbl), with production ceasing in 2016. The P90 profile was 790,007 bbl with production ceasing in 2010. [Expert E] did not conduct a probabilistic assessment in relation to the [B] field. For this field, the Claimant accepts the Second Respondent's production profile but maintains that production will continue up to 2016 totalling 4,116,566 bbl. For both fields together, the claim in respect of lost future production is 5,605,269 bbl based on the P50 forecast for [field A] and 4,904,573 bbl based on the P90 forecast for [field A].

The Second Respondent's profiles for [field A] and [field B] are 513,319 bbl and 3,112,538 bbl respectively based on production up to 2008. The total forecast by the Second Respondent for both fields is 3,625,857 bbl.

The difference between the Claimant's profile for both fields (based on P90 for [field A]) and the Second Respondent's profile for both fields is therefore 1,278,716 bbl. The same difference but using the Claimant's P50 forecast for [field A] is 1,979,412 bbl.

The Tribunal has considered the three production forecasts presented to it in relation to the [A] field: [Expert E]'s P50 and P90 forecasts and the Second Respondent's forecasts. The Tribunal has to determine which is the most reasonable, and whether that forecast is sufficiently conservative that there is a real probability that it would have been achieved.

On this basis, the Tribunal has preferred [Expert E]'s P90 forecast. Although the P50 forecast represents the most likely forecast of production, the P90 forecast reflects a higher level of confidence such that there is a real probability that it would be met. The primary differences between the three production profiles is that the Second Respondent's profile assumes that no horizontal wells will be drilled in the [A] field, whereas [Expert E]'s P90 forecast assumes one such well, and his P50 forecast assumes two horizontal wells.

[Petroleum engineering experts F and G, for Respondent No. 2] did not accept that it would be prudent to drill any horizontal wells in the [A] field having regard to the risks. [Expert F] said he thought that it would be too risky to drill even a pilot well, with a view to extending it into a horizontal well, until more was known as to the reasons for the high water cut in the existing [field A] wells. However, he did not rule out a horizontal well once more information became available.

[Expert G] also doubted that a horizontal well drilled in the [A] field would be successful because it would be close to the fault and also close to the existing [field A] wells.

It was accepted by the experts that the probabilistic approach to risking a production profile into categories of P50 and P90 was commonly used in the industry. This was the approach adopted by [Expert E] in relation to the [A] field. . . . Neither [of Respondents No. 2's experts] had conducted such a probabilistic analysis. They accepted the production profiles put forward by the Second Respondent.

It is unclear to what extent the Second Respondent's production profiles have been "risked". In paragraph 63 of his report, [Expert F] said he thought that the Second Respondent's forecasts were both "reasonable in light of knowledge of the fields". In paragraph 64, he went on to explain that these forecasts had been "risked". Specifically, his report states:

Fourthly, there is an element of risk in any forecast of production, caused by unrecoverable reserves, technical problems with wells, and economic limits. Typically proven reserves are given a 90% probability of being recovered, so the figures supplied by [Respondent No. 2] have been adjusted to 90% of their forecast.

It therefore appears that both the Claimant's and the Second Respondent's profiles have been risked. However, the basis upon which the Second Respondent's profile has been prepared is less clear.

As the probabilistic approach used by [Expert E] is commonly accepted in the industry, the Tribunal is persuaded to adopt that approach. This approach is both reasonable and provides a forecast, which has a high probability of being met from a statistical point of view. Indeed it is probable that it will be exceeded.

The remaining question on the [field A] production profile is whether it was right for [Expert E] to allow for one horizontal well in his P90 forecast. [Expert G]'s view on this was as follows:

My guess is that the impact of the horizontal well would be negative on the economics, but would not necessarily destroy the economics, and therefore it would still be reasonable to say it was a 90% case.

It appears from this that, although there was disagreement between the experts as to whether a horizontal well would be economically successful, it was reasonable for [Expert E] to include one such well in his P90 case. The Tribunal has included the P90 case with one horizontal well in the spreadsheet.

In relation to [field B], the only production profile put forward by any of the parties is the production profile prepared by the Second Respondent. There was some doubt, however, as to whether this profile has been appropriately "risked" or not.

Initially, [a drilling engineer and former director of Respondent No. 2] gave evidence that the [field B] production profile had been fully risked, and that only political and economic risks were not accounted for. Subsequently, however, the Tribunal was given a letter signed by . . . of the Second Respondent stating that the production profiles were generated using certain simulation software. They said this software assumed that the wells would be shut in when the water cut reaches 95%. The Tribunal does not know what other risks, if any, are provided for by the software.

As noted above, in paragraph 64 of his report, [Expert F] explained that the [field B] forecast had been "risked". Specifically, his report stated that, to take account of "technical risks," the figures supplied by [Respondent No. 2] have been adjusted to 90% of their forecast." [sic]

In his evidence, [Expert F] said that the discount rate which he thought should be applied dealt with commercial risks rather than technical risks. It therefore appears from [Expert F]'s evidence that he regarded the [field B] production profile as fully risked for technical purposes. The letter [referred to above] did not state what production profile they were referring to, or whether it had been reduced by 10% to reflect technical risks, as had the profile considered by [Expert F].

At the hearing, it was also put to [Expert F] that the Second Respondent's forecasts were "equivalent" to the P90 case, and he was asked whether he thought that [Expert E]'s P50 case was reasonable as such. Although he did not answer the question, and referred the Tribunal to [Expert G], [Expert F] did not disagree that the Second Respondent's forecasts were equivalent to a P90 forecast.

There was also an issue as to whether the [B] field will cease production in 2008 (as the Second Respondent maintains) or in 2016 (as the Claimant maintains). The Second Respondent argued that, after the [C] and [A] fields cease production, which it believes will be in 2008, it would be uneconomic to continue production from [field B]. However, the Tribunal accepts the Claimant's assessment that it would have been economic for the [B] and [C] fields to continue production until 2016 if it were operated by the JV-Company.

Oil price

It was recognised by all the oil price experts that the "best predictor" of future oil prices is the Futures Market. These are real prices obtainable in the market for oil delivered in the future. As such, they are generally discounted. In fact it can be seen that the prices predicted from 2003 are likely to be significantly lower than current oil prices. As these future prices are obtainable in the market today, they satisfy the real probability test for damages under [applicable] law. The Tribunal has therefore included the Futures Market prices, as set out in the graph submitted to us [by Claimant's expert on oil price forecasts and discount rates, (Expert H)].

These prices have been used for the years 2003 to 2008. No Futures prices are available after 2008. For 2009 to 2016, the Tribunal has taken the prices shown in Table 12 attached to the EIA Report, which was given to the Tribunal by the Second Respondent, but referred to extensively by [Expert H]. This shows a reference case price of USD 22.73 in 2005, USD 23.36 in 2010, USD 24.00 in 2015 and USD 24.68 in 2020. This compares to the average of all price forecasts shown in [report of Respondent No. 2's expert on oil price forecasts] of USD 25.84 in 2010, USD 30.11 in 2015 and USD 36.07 in 2020. The EIA reference case therefore appears to be a reasonably conservative price compared to the average price forecasts. These prices are also comfortably within the OPEC target range of USD 22 to USD 28 (reported in [Expert H]'s Report). The Tribunal has therefore concluded, that there is a real probability that the oil price will on average be at least as high as the reference case in the EIA report.

In computing the price for the spreadsheet, a further USD 0.74 per barrel has been deducted from the EIA prices because they reflect projections for average landed imports to the United States. [Expert H] gave evidence that the appropriate differential between such oil and Brent is USD 0.74 per barrel.

In addition, a further differential between the Brent price and the sales price of USD 1.83 has been applied. This reflects the price differential achievable for [State X] oil at [town] in [country], where the oil has since 1 January 2002 been sold by the JV-Company.

. . . . . . . . .

Discount factor

All the experts are in agreement that future income from production should be discounted to reflect the weighted average cost of capital and that this factor would be between 5% and 6%.

The experts also broadly agreed that the production profiles should be properly "risked" to allow for the technical risks. Technical risks predominantly encompass the mechanical difficulties of lifting the oil in place, and the uncertainties as to the volume of the oil in place, and the performance of the reservoir in terms of such matters as pressure and water cut.

The experts also agreed that a P90 form of probabilistic analysis accounts for technical risks, and is a method commonly used by oil companies to risk their production profiles. Indeed, as [Expert F] accepted, it is a more conservative approach than a P50 analysis. However, [Expert E] had only conducted a P90 analysis for the [A] field and had not performed such an analysis for the [B] field.

In relation to the [B] field, [Expert F] confirmed that the Second Respondent's production profile had been discounted by 10% in order to reflect technical risk. In his view, some further discount factor (in addition to the cost of capital) was necessary in order to reflect commercial risks (economic and political).

By contrast, [Expert G] said that he did not think it necessary to discount the production profile for economic or political risks.

The Tribunal nevertheless does find it appropriate to add some factor to account for commercial risk. [A witness] gave evidence that the Claimant applies a discount factor of 8% to its proven reserves in [another location]. The Tribunal sees no reason why the commercial risks in [State X] should be regarded as greater than the commercial risks in [that other location]. The Tribunal has therefore adopted a discount rate of 8%.'