This article is an extract from The Renewable Energy Law Review - Edition 4. Click here for the full guide.

I Introduction

Since January 2016, more than 15 awards related to solar disputes against Spain have been made by tribunals. An interesting divergence in the awards on liability has captured general attention. It is less well known that tribunals also differed significantly on key inputs to the determination of quantum.

The awards reviewed in this article relate to photovoltaic (PV) and concentrating solar power (CSP) technologies. The following table lists the cases by reference to the claimant, the date of the award and the type of technology:

Claimant Date Type Claimant Date Type Claimant Date Type
Charanne Jan 2016 PV Foresight Nov 2018 PV Stadwerke Dec 2019 CSP
Isolux Jul 2016 PV 9REN May 2019 PV RREEF Dec 2019 CSP
Eiser May 2017 CSP NextEra May 2019 CSP PV Investors Feb 2020 PV
Novenergia Feb 2018 PV SolEs Jul 2019 PV Cavalum* Aug 2020 PV
Masdar May 2018 CSP InfraRed Aug 2019 PV      
Antin Jun 2018 CSP OperaFund Sep 2019 PV      
*Cavalum is a decision on jurisdiction, liability and direction on quantum as it has no final award as of the date of writing.

II Background

Spain stepped up its efforts to increase renewable energy production from the mid-1990s. In 1997, the Electric Power Act set out a legal framework for the regulation of the electricity sector and committed to provide reasonable returns to renewable investors. Early support systems offering targeted incentives for renewable support schemes were put in place to attract investments.2 These systems did not succeed in incentivising significant private investment into renewables in Spain.

A decade later, Spain enacted two consecutive renewables support schemes, in 2007 and 2008.3 Both of these offered PV investors a feed-in-tariff (commonly known as FiT), an annual payment per unit of output exceeding the market price of electricity. These regimes significantly accelerated investments into renewable technologies, attracting nearly 10 times the originally envisaged amount.

From 2010, Spain introduced various measures affecting the profitability of solar investments that had already been made. These measures included introducing limits to the duration of the FiTs,4 introducing limits to the annual operating hours entitled to such FiTs,5 imposing additional costs and taxes on electricity generation (an access toll for electricity fed into the grid, and a 7 per cent tax on revenues) and modifying the rules for the annual updating of the FiTs.

Between 2013 and 2014, Spain abolished the above support schemes for renewables plants and established a new support mechanism. Under the new regime, renewables plants are no longer entitled to a FiT for electricity produced. Instead, there is a new remuneration scheme designed to yield a predefined, pre-tax rate of return for a so called 'standard plant'.

III Damages awarded

Investors complained about these changes; and a number of international arbitrations were initiated. As the tribunal in the PV Investors award helpfully summarised, the decisions in these cases were of three main types:

  1. cases that have resulted in a total dismissal of the claim;
  2. cases where the tribunals ruled that claimants were entitled to the FiTs;6 and
  3. cases where the tribunals ruled that claimants were entitled to a reasonable rate of return. 7

The outcomes of these cases are shown in the table below. The two early awards that were in favour of Spain (Charanne and Isolux), were followed by nine awards between 2017 and 2019 that upheld the FiT argument. This wave was interrupted by only one award, NextEra, a special case for reasons described below. From the end of 2019, all decisions (save for Stadtwerke) were based on providing a reasonable rate of return.

  2016 2017 2018 2019 2020
No damage Charanne*Isolux*     Stadtwerke  
Entitled to FiT   Eiser† Novenergia*MasdarAntinForesight* 9REN*SolEs*InfraRed* OperaFund*  
Entitled to reasonable return       NextEraRREEF† PV Investors*Cavalum*

When tribunals concluded that Spain had breached its obligations, this was with respect to the introduction of the new regulatory regime in 2013/2014. No tribunal to date has awarded damages related to the earlier measures. For example, in Foresight, the tribunal concluded that, because the regulations enacted by Spain from 2010 to 2013 'merely modified and did not fundamentally change RD 661/2007', they did not breach the Fair and Equitable Treatment standard.8

IV Entitlement to FiTs

The Eiser tribunal noted that 'ECT Article 10 does not set out a standard for compensation for breaches of its obligations'.9 Nevertheless, in all FiT cases, the standard of compensation was approached as the difference in the fair market value of the investments with and without the disputed measures. This required forecasting the claimants' cash flows had they received the FiT as originally set out by Spain (the counterfactual scenario) and the claimants' cash flows under Spain's new regulatory scheme (the actual scenario).

The assessment turned out to be more straightforward than in the reasonable return cases (which we discuss further below). All tribunals, for example, accepted the discounted cash flow (DCF) methodology as a reliable method (despite the respondent's experts arguing about its speculative nature).

One of the most interesting aspects of the 'FiT awards' is the inclusion or exclusion of a regulatory risk premium in the analysis. When it is included, the regulatory risk premium is added to the cost of capital to reflect additional risks arising from the claimants operating in a regulated market. In essence, the decision to include a regulatory risk premium in the analysis interacts with the decision on the extent to which Spain could legitimately change the regulatory system without breaching its obligations.

In Novenergia, the respondent's expert included a regulatory risk premium in the 'but for' scenario that was higher than in the actual scenario (2.2 per cent in the former compared with 0.5 per cent in the latter). The tribunal found, however, that 'it cannot be correct to assume a higher risk in a scenario where the regulatory framework of the RE sector would have remained stable and RD 661/2007 would have continued to remain in force as originally implemented'.10 It appears that, ultimately, no regulatory risk was included in the quantification of damages.

In 9REN, on the other hand, the tribunal reduced the quantum by 20 per cent to allow, among other factors, for regulatory risk. It put it this way:

A majority of the Tribunal therefore considers that some measure of “regulatory risk” should be reflected in the Award because a prudent and well-informed investor would have been alive to the risk that Spain might reduce the FIT tariff and be held (despite the ECT) to be within its rights under international law to do so without compensation.11

Another element is the illiquidity discount, applied to the present value of the investments (as a percentage reduction) to allow for the difficulty of selling solar plants. In Eiser, the claimant's expert applied an 18 per cent discount for illiquidity. This was based on 'recent published research in corporate finance concerning the effects of liquidity'.12

In Novenergia, the respondent's expert reduced the valuation in the counterfactual scenario by 17 per cent and in the actual scenario by 11.8 per cent.13 The tribunal was 'not convinced that an otherwise healthy business operating under a regulated and stable environment would be more difficult to divest, something which is also confirmed by statistics on RE transactions in Spain showing entirely normal exit times'.14

The claimant's expert in Masdar used an 18 per cent illiquidity discount for both the actual and counterfactual scenarios, but noted that it was a conservative figure for the actual scenario because of the increased regulatory risk as a result of Spain's adjusted measures, which would be expected to increase the discount rate.15 A majority of the tribunal accepted this approach and found that the figure was 'reasonable – indeed, conservative'.16

The claimant's expert in InfraRed used an illiquidity discount of 25 per cent in the actual scenario and of 12 per cent in the counterfactual scenario, noting that the claimant's interests were more difficult to liquidate following Spain's regulatory changes.17 The tribunal agreed, and stated:

. . . the Original Regulatory Framework attracted significant investments in the solar energy industry in Spain and that its dismantlement caused significant turmoil in that industry and market. The Tribunal agrees that these measures served to increase, not decrease, the risk associated with Spain's regulatory environment and the correlative risk of illiquidity for investors such as Claimants as at the valuation date. 18

V Entitlement to Reasonable Return

In cases where the tribunals did not accept that claimants were entitled to the FiTs, the tribunals' decisions on approaches to calculate damages differed more substantially.

RREEF is the first case in which the tribunal concluded that 'the only legitimate expectations of the Claimants was [sic] to receive a reasonable return for its [sic] investment'.19 The majority of the tribunal found this return to be the cost of capital (WACC) plus 1 per cent as of the date the respondent changed the regime (6.86 per cent as of June 2013).20 Further, the tribunal found that the respondent's new regime provided returns below this level to the claimant's CSP plants.21 It therefore directed the experts to construct cash flows that would provide a return of WACC plus 1 per cent in the counterfactual scenario and compare them with the cash flows projected in the actual scenario.

The tribunal explicitly disagreed with the Novenergia tribunal that the regulatory risk in the original regime was lower than in the new regime:

In certain respects, the risk is lower under the new regime considering that the Respondent ensured a better sustainability of the whole system. It is only by considering that the Respondent could not lawfully modify the previous regime that the risk under the previous regime could be considered lower than under the new regime, but the Tribunal already explained this assumption is wrong.22

The NextEra tribunal also rejected the idea that the claimant was entitled to the FiT: 'they just had a legitimate expectation that there would not be a substantial or fundamental change to Regulatory Framework I'. 23 The tribunal, however, did not conclude that the claimant was only entitled to a reasonable return. That the tribunal ended up approaching damages on a reasonable return basis appears to be the result of its rejection of the DCF method (due to a lack of sufficient operational record of the plants) and of the fact that both parties had provided estimations of damages on a reasonable return basis.24

Similarly to RREEF, the NextEra tribunal found that the reasonable return should be based on the claimant's cost of capital plus a premium. However, the tribunal decided that this premium should be 2 per cent, in contrast with a premium of only 1 per cent in NextEra.25 As this rate was post-tax, it needed to be converted into a pre-tax rate. The respondent argued for an effective tax rate, whereas the claimant argued for the 30 per cent nominal tax rate (also citing that the use of a nominal tax rate is 'accepted regulatory practice').26 The tribunal accepted the nominal tax rate.27

Another key issue was whether this return should apply to the costs that the claimant actually incurred or to the standard costs of an efficient plant. The tribunal decided that 'Claimants did not suffer a notional loss based on the “standard costs of cost-effective standard facilities.” They had actual sunk costs and those costs have to be the basis for assessing their loss.'28

In PV Investors, the tribunal concluded that the reasonable return that claimants could have expected was in the range of 7 per cent, post-tax.29 Conceptually, this is a very different rate of return than that found in RREEF where the tribunal was looking for the reasonable return at the time the new regime was introduced, because here the tribunal was looking for the reasonable return 'on offer at the time when the Claimants made their investments'.30

Moreover, in contrast with NextEra and RREEF where the reasonable return was based on what claimants might themselves have expected, the 7 per cent return adopted in PV Investors was based on the costs of the relevant 'marginal plant' (a complicated issue in itself) and be applied to greenfield investment costs.31 By using the costs of the marginal plant, the decision aimed to ensure that investors who managed plants at lower costs could gain from more efficient operations.32

The damages calculations required the construction of an alternative tariff that would provide a 7 per cent return to a marginal plant with greenfield investment costs.33 The counterfactual scenario was then derived by applying this alternative tariff instead of the FiTs in the DCF model.34 Contrary to the decision in NextEra, the PV Investors tribunal found that in order to gross up the 7 per cent post-tax return to a pre-tax one, the claimant's effective tax rate had to be used, taking into account the tax shield from which the plants benefitted as a result of debt financing.35

In summary, substantial differences are observable in the derivation of the reasonable rate of return to which claimants were entitled: (1) the rate that applied when the claimant invested or the rate that applied at the time Spain changed the regime; (2) the size of the premium that should be added to the WACC to reflect Spain's desire to attract investment; (3) whether the post-tax rate should be grossed up using the effective tax rate or the nominal tax rate; and (4) whether the reasonable return should apply to a 'marginal plant' or to the claimant's operations.

VI Date of assessment

There are some other aspects common to both types of cases in which tribunals' decisions differed. Regarding the valuation date, the OperaFund tribunal explained that 'as settled in international jurisprudence, Claimants are entitled to damages valued as of the date of injury or as of the date of the Tribunal's award, whichever is higher'.36 Only two awards rely on an ex-post DCF valuation, those of NextEra (as of June 2016) and of Novenergia (as of September 2016, the date corresponding to the second report of the claimants' quantum experts). All other awards are based on June 2014 (the publication of the ministerial order setting out the details of the new regime).37

In PV Investors, the tribunal considered that using a valuation date linked to the date of the award would not be justified:

In the Tribunal's view, according to the specificities of the dispute, an ex-post valuation may sometimes be preferable because reparation should ideally stand in lieu of restitution and an ex-post valuation allows for the most recent information to be taken into account. This said, a valuation at the time of the breach, i.e. ex-ante, appears particularly appropriate when the consequences of a later evolution of prices, interest rates, or other inputs are unrelated to the impugned measures and the (higher) harm can thus not be deemed to derive from the measures.38

VII Tax gross-up

Several claimants argued for a tax gross-up to account for the tax consequences of the amounts awarded. In Eiser, the requested tax gross-up amounted to €88 million.39 The tribunal found, however, that no evidence was provided on any tax that would be due:40

This substantial claim was predicated on the theory that in order to make Claimants whole, their recovery should be net of taxes. However, Claimants offered no evidence establishing the nature, rate or amount of any tax that might be due.41

On the other hand, in OperaFund, the tribunal declared that its award was made net of all taxes or withholdings, and ordered Spain to indemnify claimants for any tax liability or withholding that might be imposed in Spain, Malta or Switzerland, in relation to the compensation awarded.42

VIII Interest

All awards include decisions on both pre-award and post-award interest rates. In many cases, the claimants argued that a higher post-award interest should be applied to incentivise governments to pay the awards.43 This notion was frequently recognised by the tribunals, although they did not always award higher post-award interest rates. The tribunal in Antin, for example, used the same rate for both, noting:

However, the Tribunal does not agree that it should order post-award interest at a rate higher than 2.07 per cent in the present case. The Respondent has an international obligation to comply with this Award in a timely manner. Imposing a higher post-award interest rate to ensure prompt compliance with the Award would imply that there are reasons to believe that the State will not fulfil its international obligation to comply promptly. In the absence of such reasons, the Tribunal believes that a higher post-award interest rate is not justified.44

On the other hand, the tribunal in Eiser applied different interest rates for the periods before and after the date of the award.45 The tribunal, 'in order to facilitate prompt payment of this Award', set the pre-award interest at 2.07 per cent and the post-award interest at 2.50 per cent.46 In Masdar, the majority adopted a similar approach, in which the post-award interest rate (1.60 per cent) was higher than the pre-award rate (0.906 per cent).47 Similarly, in Foresight, the decision was to exercise a post-award interest higher than the pre-award one.48

In all of the cases noted immediately above, interest was set on a compound basis. OperaFund is an exception: the tribunal awarded 1.59 per cent pre-award interest (based on the relevant Spanish 10-year bond yield) but decided that it would not be compounded.49 The tribunal applied the same figure for post-award interest, but decided to compound it monthly, recognising that '. . . tribunals may determine a different, higher post-award interest to eliminate Respondent's incentive to delay full payment of an award . . .'.50

IX Conclusion

Some of the differences noted above – in the way that appropriate levels of compensation have been derived by different tribunals – will reflect differences in the underlying facts; and others will reflect differences in the evidence that was adduced. Some, however, reflect differences of opinion between tribunals as to how the underlying calculations should be performed.

The opportunity to review decisions across a large number of similar cases is, in our view, a useful one. It helps to bring into view issues on topics that appear in international arbitration cases regularly, on which there is not as yet a settled opinion. Future awards in similar solar cases may provide further opportunities for a convergence of approach.