OPINION OF ADVOCATE GENERAL
WAHL
delivered on 9 March 2017 (1)
Case C‑80/16
ArcelorMittal Atlantique et Lorraine
v
Ministre de l’Écologie, du Développement durable et de l'Énergie
(Request for a preliminary ruling from the Tribunal administratif de Montreuil (Administrative Court, Montreuil, France))
(Environment — Greenhouse gas emission allowance trading scheme — Directive 2003/87/EC — Transitional rules — Decision 2011/278/EU — Validity — Method for allocating allowances free of charge — Steel sector — Benchmarks for hot metal and sintered ore — Production of electricity from waste gases — Use of most accurate and up-to-date data — Most efficient installations — Obligation to state reasons)
1. This case deals with the EU emission trading scheme, which was put in place by Directive 2003/87/EC. (2) Owing, in particular, to the financial interests involved, a myriad of aspects of that scheme have been the subject of cases before the EU Courts. (3) The present request for a preliminary ruling concerns the validity of Decision 2011/278/EU, (4) which was adopted pursuant to that directive. The questions put to the Court deal with the complex issue of the appropriate method to be employed by the Commission in setting EU-wide benchmarks for the free allocation of emission allowances in the steel sector during the period 2013 to 2020.
2. More specifically, the referring court asks the Court to clarify whether, in setting the relevant benchmarks, the Commission could (1) decide not to include the totality of emissions related to the use of recycled waste gases in electricity production in the benchmark for hot metal; (2) base its determination of the benchmark for hot metal on the ‘BREF’ (5) for iron and steel production and Decision 2007/589/EC (6); (3) include a factory producing both sintered ore and pellets in the reference installations for determining the benchmark for sintered ore; and finally (4) proceed in that way without specifically stating the reasons for doing so.
3. On the face of it, the questions referred are highly technical in nature. However, that should not detract from the significance of those questions. On the one hand, in today’s changing climate, the emission trading scheme put in place by Directive 2003/87 constitutes undoubtedly the cornerstone of EU environmental policy. (7) On the other hand, in the wake of the hottest year ever recorded, in which scientists reported, inter alia, all-time lows in the Arctic sea ice levels, the importance of the questions posed can hardly be overstated.
4. In what follows, I will explain why the Commission has determined the benchmarks referred to in Decision 2011/278 in conformity with Directive 2003/87.
I. Legal framework
A. Directive 2003/87
5. Article 10a of Directive 2003/87 provides for the free allocation of allowances for the transitional period 2013 to 2020. Paragraphs 1 to 3, 6 and 12 thereof read:
‘1. By 31 December 2010, the Commission shall adopt Community-wide and fully-harmonised implementing measures for the allocation of the allowances referred to in paragraphs 4, 5, 7 and 12, including any necessary provisions for a harmonised application of paragraph 19.
Those measures, designed to amend non-essential elements of this Directive by supplementing it, shall be adopted in accordance with the regulatory procedure with scrutiny referred to in Article 23(3).
The measures referred to in the first subparagraph shall, to the extent feasible, determine Community-wide ex-ante benchmarks so as to ensure that allocation takes place in a manner that provides incentives for reductions in greenhouse gas emissions and energy efficient techniques, by taking account of the most efficient techniques, substitutes, alternative production processes, high efficiency cogeneration, efficient energy recovery of waste gases, use of biomass and capture and storage of CO2, where such facilities are available, and shall not provide incentives to increase emissions. No free allocation shall be made in respect of any electricity production, except for cases falling within Article 10c and electricity produced from waste gases.
For each sector and subsector, in principle, the benchmark shall be calculated for products rather than for inputs, in order to maximise greenhouse gas emissions reductions and energy efficiency savings throughout each production process of the sector or the subsector concerned.
In defining the principles for setting ex-ante benchmarks in individual sectors and subsectors, the Commission shall consult the relevant stakeholders, including the sectors and subsectors concerned.
The Commission shall, upon the approval by the Community of an international agreement on climate change leading to mandatory reductions of greenhouse gas emissions comparable to those of the Community, review those measures to provide that free allocation is only to take place where this is fully justified in the light of that agreement.
2. In defining the principles for setting ex-ante benchmarks in individual sectors or subsectors, the starting point shall be the average performance of the 10% most efficient installations in a sector or subsector in the Community in the years 2007-2008. The Commission shall consult the relevant stakeholders, including the sectors and subsectors concerned.
The regulations pursuant to Articles 14 and 15 shall provide for harmonised rules on monitoring, reporting and verification of production-related greenhouse gas emissions with a view to determining the ex-ante benchmarks.
3. Subject to paragraphs 4 and 8, and notwithstanding Article 10c, no free allocation shall be given to electricity generators, to installations for the capture of CO2, to pipelines for transport of CO2 or to CO2 storage sites.
…
6. Member States may also adopt financial measures in favour of sectors or subsectors determined to be exposed to a significant risk of carbon leakage due to costs relating to greenhouse gas emissions passed on in electricity prices, in order to compensate for those costs and where such financial measures are in accordance with state aid rules applicable and to be adopted in this area. …
…
12. … in 2013 and in each subsequent year up to 2020, installations in sectors or subsectors which are exposed to a significant risk of carbon leakage shall be allocated, pursuant to paragraph 1, allowances free of charge at 100% of the quantity determined in accordance with the measures referred to in paragraph 1.’
B. Decision 2011/278
6. The Commission has set the product benchmarks referred to in Article 10a of Directive 2003/87 in Decision 2011/278. To understand the method applied for determining the contested benchmarks for sintered ore and hot metal, the following recitals are of particular relevance. They read (footnotes excluded):
‘(2) In defining the principles for setting ex-ante benchmarks in individual sectors or sub-sectors, the starting point should be the average performance of the 10 % most efficient installations in a sector or sub-sector in the EU in the years 2007-2008. …
…
(4) To the extent feasible, the Commission has developed benchmarks for products, as well as intermediate products that are traded between installations, produced from activities listed in Annex I to Directive 2003/87/EC. … Where a product is a direct substitute of another product, both should be covered by the same product benchmark and the related product definition.
…
(6) The benchmark values should cover all production-related direct emissions, including emissions related to the production of measurable heat used for production, regardless of whether the measurable heat was produced on-site or by another installation. Emissions related to the production of electricity and to the export of measurable heat, including avoided emissions of alternative heat or electricity production in cases of exothermic processes or the production of electricity without direct emissions, were deducted when setting the benchmark values. ...
(7) In order to ensure that benchmarks lead to reductions in greenhouse gas emissions, for some production processes in which direct emissions eligible for the free allocation of emission allowances and indirect emissions from electricity production not eligible for free allocation on the basis of Directive 2003/87/EC are to a certain extent interchangeable, the total emissions including indirect emissions related to the production of electricity have been considered for the determination of the benchmark values to ensure a level playing field for fuel and electro-intensive installations. For the purpose of the allocation of emission allowances on the basis of the benchmarks concerned, only the share of the direct emissions in the total emissions should be taken into account in order to avoid providing free allocation of emission allowances for emissions related to electricity.
(8) For the determination of benchmark values, the Commission has used as a starting point the arithmetic average of the greenhouse gas performance of the 10% most greenhouse gas efficient installations in 2007 and 2008 for which data has been collected. In addition, the Commission has in accordance with Article 10a(1) of Directive 2003/87/EC analysed for all sectors for which a product benchmark is provided for in Annex I, on the basis of additional information received from several sources and on the basis of a dedicated study analysing most efficient techniques and reduction potentials at European and international level, whether these starting points sufficiently reflect the most efficient techniques, substitutes, alternative production processes, high efficiency cogeneration, efficient energy recovery of waste gases, use of biomass and capture and storage of carbon dioxide, where such facilities are available. Data used for determining the benchmark values has been collected from a wide range of sources in order to cover a maximum of installations producing a benchmarked product in the years 2007 and 2008. First, data on the greenhouse gas performance of ETS installations producing benchmarked products has been collected by or on behalf of the respective European sector associations based on defined rules, so-called ‘sector rule books’. As reference for these rule books, the Commission provided guidance on quality and verification criteria for benchmarking data for the EU-ETS. Second, to complement the data collection by European sector associations, consultants on behalf of the European Commission collected data from installations not covered by industry’s data and also competent authorities of Member States provided data and analyses.
…
(10) Where several products are produced in one installation and an assignment of emissions to the individual products has not been regarded feasible, only single product installations have been covered by the data collection and included in the benchmark setting. This concerns the product benchmarks for lime, dolime, bottles and jars of colourless glass, bottles and jars of coloured glass, facing bricks, pavers, spray-dried powder, uncoated fine paper, tissue, testliner and fluting, uncoated carton board as well as coated carton board. To increase the significance and check the plausibility of the results, the values for the average performance of the 10% most efficient installations have been compared against literature on most efficient techniques.
(11) In case no data or no data collected in compliance with the benchmarking methodology has been available, information on present levels of emissions and consumptions and on most efficient techniques, mainly derived from the [BREF] established in accordance with Directive 2008/1/EC … has been used to derive benchmark values. In particular, due to a lack of data on the treatment of waste gases, heat exports and electricity production, the values for the product benchmarks for coke and hot metal have been derived from calculations of direct and indirect emissions based on information on relevant energy flows provided by the relevant BREF and default emission factors set out in Commission Decision 2007/589/EC … For the product benchmark for sintered ore, data has also been corrected based on relevant energy flows provided by the relevant BREF, taking into account the combustion of waste gases in the sector.
(12) Where deriving a product benchmark was not feasible, but greenhouse gases eligible for the free allocation of emission allowances occur, those allowances should be allocated on the basis of generic fallback approaches. A hierarchy of three fallback approaches has been developed in order to maximise greenhouse gas emission reductions and energy savings for at least parts of the production processes concerned. The heat benchmark is applicable for heat consumption processes where a measurable heat carrier is used. The fuel benchmark is applicable where non-measurable heat is consumed. The heat and fuel benchmark values have been derived based upon the principles of transparency and simplicity, using the reference efficiency of a widely available fuel that can be regarded as second-best in terms of greenhouse gas efficiency, considering energy efficient techniques. For process emissions, emission allowances should be allocated on the basis of historical emissions. ...
…
(32) It is also appropriate that the product benchmarks take account of the efficient energy recovery of waste gases and emissions related to their use. To this end, for the determination of the benchmark values for products of which the production generates waste gases, the carbon content of these waste gases has been taken into account to a large extent. Where waste gases are exported from the production process outside the system boundaries of the relevant product benchmark and combusted for the production of heat outside the system boundaries of a benchmarked process as defined in Annex I, related emissions should be taken into account by means of allocating additional emission allowances on the basis of the heat or fuel benchmark. In the light of the general principle that no emission allowances should be allocated for free in respect of any electricity production, to avoid undue distortions of competition on the markets for electricity supplied to industrial installations and taking into account the inherent carbon price in electricity, it is appropriate that, where waste gases are exported from the production process outside the system boundaries of the relevant product benchmark and combusted for the production of electricity, no additional allowances are allocated beyond the share of the carbon content of the waste gas accounted for in the relevant product benchmark.’
7. The benchmarks for sintered ore and hot metal are laid down in Annex I to the decision. The values are set at 0.171 and 1.328 allowances/tonnes respectively.
II. Facts, procedure and the questions referred
8. In the main proceedings before the Tribunal administratif de Montreuil (Administrative Court, Montreuil), ArcelorMittal Atlantique et Lorraine (‘ArcelorMittal’) seeks the annulment of the Decree of the Minister for Ecology, Sustainable Development and Energy of 24 January 2014 laying down a list of the operators to whom greenhouse gas emissions allowances have been allocated and the amount of allowances allocated free of charge for the period 2013-2020. (8) In addition, it has asked that court to annul the Minister’s decision of 11 June 2014 dismissing its application for that decree to be withdrawn.
9. The decree at issue was adopted in accordance with Decision 2011/278. ArcelorMittal claims, in essence, that the decree and the Minister’s decision are illegal in that they are based on Decision 2011/278, which in turn is not compatible with Directive 2003/87. That is so, in its submission, because, on the one hand, the Commission did not take account of emissions related to waste gases used for electricity production in the calculation of the benchmark for hot metal and that, on the other hand, the best available data was not employed in determining that benchmark. ArcelorMittal further submits that the benchmark for sintered ore is not correct because the emissions of an installation which also produces pellets were included in the calculation of that benchmark.
10. The referring court acknowledges that under Article 10a of Directive 2003/87 the Commission enjoys broad discretion in determining the benchmarks for the allocation of emission allowances free of charge. Nevertheless, it entertains doubts as to whether the benchmarks for hot metal and sintered ore are compatible with that directive. That is why it decided to stay the proceedings and to refer the following questions to the Court for a preliminary ruling:
‘(1) In its [Decision 2011/278], did the [Commission], by excluding emissions from recycled waste gases used in the production of electricity from the benchmark value for hot metal, breach Article 10a(1) of [Directive 2003/87] concerning the rules for establishing ex-ante benchmarks, and in particular the objective of efficient energy recovery of waste gases and the option of allocating allowances free of charge in the case of electricity produced from waste gases?
(2) By basing its determination of the benchmark for hot metal in that decision on the data in [the BREF for the production of iron and steel] and [Decision 2007/589], did the Commission infringe the obligation to use the most [accurate] and up-to-date scientific data available and/or the principle of sound administration?
(3) In [Decision 2011/278], is the [Commission’s] inclusion, if proven, of a factory producing both sintered ore and pellets in the reference installations for determining the benchmark for sintered ore such as to vitiate the value of that benchmark on grounds of illegality?
(4) Did the Commission, by failing to state specifically the reasons for proceeding in that way in [Decision 2011/278], infringe the obligation to state reasons laid down in Article 296 [TFEU]?’
11. Written observations were submitted by ArcelorMittal Atlantique et Lorraine, the French and German Governments and the Commission. Oral argument was presented at the hearing on 26 January 2017 by ArcelorMittal, the French and Swedish Governments and the Commission.
III. Assessment
A. Introduction
12. To understand the context of the present request for a preliminary ruling, it is useful to begin by charting the basic tenets of the EU emission trading scheme.
13. Unprecedented in terms of ambition and reach, the EU emission trading scheme became operational in January 2005. It is designed to promote the reduction of greenhouse gas emissions in a cost-effective and economically efficient manner. From the outset, the stated objective was to ensure that the joint greenhouse gas emission commitments of the EU Member States stemming from the Kyoto Protocol were realised. (9) Those commitments have evolved over time. So has the EU emission trading scheme.
14. During the first two phases (1 January 2005 to 31 December 2007 and 1 January 2008 to 31 December 2012) of the emission trading scheme, the Member States were responsible for the issuing of emission allowances in accordance with national allocation plans. During that period, the majority of emission allowances were allocated free of charge.
15. As a result of the entry into force of Directive 2009/29, Directive 2003/87 underwent considerable changes. Those changes concern the third phase (from 1 January 2013 to 31 December 2020) of the emission trading scheme. (10) The extension of the scheme to new areas of the economy and other substantial amendments were introduced to adapt that scheme to international commitments based on contemporary climate change research and evolving reduction targets reflecting that research. (11)
16. Contrary to what occurred in the first two phases, the process for allocating emission allowances is now harmonised at the EU level. The Commission sets the total amount of available allowances in the EU. That amount will decrease by 1.74% each year from 2010 onwards. Although full auctioning is to be achieved in the long term, a transitional system was put in place for sectors other than power generation. Under that system, the free allocation of allowances in the manufacturing industry is to decrease gradually from 80% in 2013 to 30% in 2020. The stated goal is that no free allocation will take place in 2027.
17. As far as sectors deemed to suffer from a high risk of carbon leakage (12) are concerned, such as the steel industry, those sectors receive allowances at 100% of the quantity set for free allocation. That quantity is determined, in principle, on the basis of the installation’s production output (measured in tonnes of product) multiplied by the benchmark value of the relevant product.
18. Under the transitional system, the Commission assumes a key role.
19. At the beginning of the third phase, the Commission verified that the lists provided by each Member State of the installations covered by Directive 2003/87 and of the number of free allowances to be allocated to those installations were complete and that those lists complied with the relevant legal framework. (13) On that basis, the Member States took final allocation decisions for the period from 2013 to 2020. Given that the requested allocations for all installations in the EU exceeded the total amount available for free allocation, the allocation per installation was reduced by the Commission. That is known as the ‘cross-sectoral correction factor’, which has also been the subject of recent cases before the Court. (14)
20. In order to attain the objectives of the emission trading scheme, the rules for free allocation are of particular significance. Indeed, it is of fundamental importance that free allocation does not exceed what is necessary. The Commission has laid down those rules, which are applicable across the European Union, in Decision 2011/278.
21. In that decision, the Commission determined the benchmarks that function as the basis for the free allocation for each installation. Generally speaking, a product benchmark is based on the average greenhouse gas emissions of the 10% most efficient installations producing the product in the EU. The rationale is that installations thatdo not reach those benchmarks will receive fewer allowances at no cost than they actually need.
22. The present request for a preliminary reference addresses from a number of perspectives the lawfulness of the method applied by the Commission in determining the relevant product benchmarks. I shall now consider the questions referred in more detail.
B. The validity of Decision 2011/278
23. As a preliminary point, it should be observed that the Commission enjoys broad discretion in determining product benchmarks in the context of the EU emission trading scheme. It follows from Article 10a(1) of Directive 2003/87 that the Commission is to consider a wide array of parameters in setting the product benchmarks that form the basis for free allocation of emission allowances. That same paragraph specifies that those parameters are to be taken into account ‘to the extent feasible’. Accordingly, given the technical nature of benchmarking, the judicial review undertaken by the Court will be limited to checking that the measure is not manifestly inappropriate. (15)
24. It is in that light that the questions put to the Court must be assessed.
1. The benchmark for hot metal: electricity production from waste gases
25. By its first question, the referring court asks whether the method employed by the Commission to set the benchmark for hot metal pursuant to Decision 2011/278 is compatible with the third subparagraph of Article 10a(1) of Directive 2003/87.
26. The third subparagraph of Article 10a(1) of Directive 2003/87 provides, inter alia, that efficient energy recovery of waste gases is to be taken into account in benchmarking. That is to ensure that allocation takes place in a manner that provides incentives for reductions in greenhouse gas emissions and for energy efficient techniques. That same provision also states that no free allocation shall be made in respect of electricity production, except for electricity produced from waste gases.
27. In the light of that provision, ArcelorMittal complains that in setting the benchmark for hot metal, all emissions pertaining to the use of waste gases for the production of electricity ought to have been taken into account by the Commission.
28. It is because of that complaint that the referring court entertains doubts regarding the compatibility of Decision 2011/278 with Directive 2003/87. It asks whether the method applied by the Commission to set the benchmark for hot metal can be reconciled with the objective of efficient energy recovery of waste gases, on the one hand, and with the exception concerning the allocation of free allowances for electricity produced from waste gases, on the other.
29. First of all, the parties that have submitted observations agree that, as a rule, allowances are not to be allocated free of charge for electricity production. That emerges with clarity from Article 10a(1) and (3) of Directive 2003/87. Nevertheless, given the exception regarding free allocation for electricity produced from waste gases laid down in Article 10a(1) of the directive, disagreement persists as to how electricity produced from waste gases ought to be dealt with and whether the choice made by the Commission contributes to efficient recovery of those gases.
30. Before moving on to consider that issue, it might be helpful to point out that waste gases constitute an inevitable by-product of the manufacture of coke and steel. Those gases are recovered and used mainly for electricity production, but also to a lesser extent for the stoves in blast furnaces, for the under-firing of coke oven plants, ignition of sinter stands and reheating of furnaces. Electricity is produced either within the installation itself or outside the installation by a third-party. Both economically and environmentally speaking, the recovery and reuse of waste gases as fuel is far more sensible than discharging those gases or flaring them off to no useful purpose. (16)
(a) Efficient recovery of waste gases
31. The use of waste gases as fuel helps explain why Article 10a(1) of Directive 2003/87 includes the promotion of the use of waste gases among the incentives to reduce greenhouse gas emissions and employ energy efficient techniques. It also explains why that provision creates an exception to the exclusion of electricity generation from the free allocation of allowances for electricity produced from waste gases. (17)
32. Contrary to what ArcelorMittal argues, however, I find no support in the directive for the view that all the greenhouse gas emissions resulting from the combustion of waste gases should automatically give rise to free allocation of emission allowances on the basis of benchmarking.
33. More specifically, there is nothing in Article 10a(1) of Directive 2003/87 that indicates to what extent electricity production from waste gases ought to be taken into consideration in benchmarking. Admittedly, the third subparagraph of Article 10a(1) of Directive 2003/87, read as a whole, allows electricity production from waste gases to be taken into account. Yet it does so only to the extent that that is feasible in view of the need to promote the efficient recovery of waste gases and the overall objective of that directive to reduce greenhouse emissions. (18)
34. That brings me to the issue of efficient recovery of waste gases and the overall objective of the directive.
35. Recital 32 of Decision 2011/278 explains that, in devising the benchmark for hot metal, the Commission did indeed take efficient recovery of waste gases into consideration. The Commission took into account the fact that waste gases are combusted to generate electricity (or heat) in the steel sector. For that reason, the benchmark for, inter alia, hot metal was adjusted upwards, so that it included not only the emissions related to the production of hot metal, but also those related to the use ofwaste gases.
36. However, the benchmark for hot metal does not take into account emissions related to the use of waste gases for electricity production in full (but rather, to a level of approximately 75 to 80%). (19)
37. The reason for that becomes apparent on examination of the case-file. In fact, the use of waste gases for electricity production means that those gases are used as a substitute for another fuel. Globally, that saves greenhouse gas emissions, because instead of using coal for steel-making and a different fuel (that is to say, natural gas) for electricity production, only the coal is needed to produce both electricity and steel.
38. In order to determine the extent to which the product benchmark for hot metal should take account of the carbon content of waste gases in electricity production, the Commission used natural gas as the reference fuel. Recycling waste gases and using them as fuel for electricity production, instead of having recourse to natural gas, means that the installation in question emits more greenhouse gases. As the parties having submitted observations have explained, recycling waste gases results in approximately 75% more greenhouse gases, as compared with a situation in which natural gas is used as fuel for electricity production. To promote the recovery and use of waste gases, the free allocation must take account of that increase in order not to penalise the reuse of those gases. Since emission allowances are allocated to the producer on the basis of the difference in emission intensity between the waste gas and natural gas, electricity production from waste gases is placed neither at an advantage nor at a disadvantage in comparison with an installation producing electricity from natural gas.
39. The free allocation of allowances on the basis of a benchmark which takes account of the increased emissions resulting from the use of waste gases for electricity production would seem to encourage the efficient recovery of such gases.
40. In Borealis Polyolefine and Others, (20) which dealt with the cross-sectoral correction factor, the Court acknowledged that. In that case, it observed that the Commission’s approach indeed took account of efficient recovery of waste gases. The Court explained that by taking into consideration the use of waste gases to a large extent in devising the benchmark for inter alia hot metal, the Commission sought to encourage undertakings to reuse or sell the waste gases generated during a manufacturing process. (21) The Court has reiterated that position in a subsequent case. (22)
(b) Dichotomy between heat and electricity production and the issue of carbon leakage
41. It is true that an asymmetry exists in the methodology chosen by the Commission: additional free allowances can be allocated for the use of waste gases where those gases are exported and combusted for the production of heat. In that case, additional allowances are allocated on the basis of the heat or fuel benchmark to the heat consumer. No such additional allowances are available for electricity produced from waste gases.
42. The fact that the Commission did not consider it appropriate to extend the allocation of additional free allowances to the production of electricity by an electricity generator reflects the general rule laid down in Article 10a(3) of Directive 2003/87. As noted above, that provision states that electricity generation is not to receive free allowances. In that regard, in recital 32 of Decision 2011/278, the Commission explains that such a rule is also necessary in order to avoid undue distortions of competition on the markets for electricity supplied to industrial installations and to take account of the inherent carbon price in electricity.
43. On that point, I must confess to having some sympathy for ArcelorMittal’s argument. In fact, I am not convinced that asymmetrical treatment of heat and electricity generation is necessary for avoiding distortions of competition in the electricity market. The need for such asymmetrical treatment appears questionable inasmuch as the steel sector is a net consumer of electricity and the production of electricity from waste gases apparently represents no more than 1% of total electricity production in the European Union. Nonetheless, the Commission’s choice appears to be wholly in keeping with Article 10a(3) of the Directive and cannot therefore affect the validity of Decision 2011/278.
44. Here, it is necessary to address also ArcelorMittal’s argument concerning the risk of carbon leakage in the steel sector. It essentially argues that the way the Commission set the benchmark for hot metal increases the risk of leakage in an industry that is a heavy consumer of energy contrary to the objective of Directive 2003/87.
45. It might well be that another benchmarking methodology could have addressed that issue (23) in a more appropriate manner, as suggested by ArcelorMittal. Nevertheless, it should not be forgotten that Directive 2003/87 addresses the risk of leakage through a compensatory mechanism laid down in Article 10a(6) thereof. Accordingly, Member States can adopt financial measures in favour of sectors or subsectors concerned. Aid may be granted in accordance with EU State aid rules where, owing to costs relating to greenhouse gas emissions passed on in electricity prices, there is a significant risk of carbon leakage. In other words, in addition to receiving allowances free of charge, the steel sector may be eligible for aid under that rule.
46. More importantly still, it is not readily apparent from ArcelorMittal’s submissions (or, from the request for a preliminary ruling) in what way the method applied by the Commission might increase the risk of carbon leakage.
47. Article 10a of Directive 2003/87 is of no assistance to ArcelorMittal here. Article 10a(12) simply states that installations in sectors or subsectors which are exposed to a significant risk of carbon leakage are to be allocated allowances free of charge at 100% of the quantity determined in accordance with the measures referred to in paragraph 1. It is clear that the directive does not state that sectors deemed to be at a significant risk of carbon leakage should, in all circumstances, receive free emission allowances that correspond to all greenhouse gas emissions they emit.
48. On the basis of the above, I conclude that consideration of the first question referred has disclosed nothing that affects the validity of Decision 2011/278 regarding the method employed by the Commission to set the benchmark for hot metal.
2. The benchmark for hot metal: the use of the most accurate and up-to-date data
49. The second question referred turns on the quality of the data employed by the Commission in setting the benchmark for hot metal pursuant to Decision 2011/278. More specifically, this question asks whether the Commission could employ data based on the BREF for iron and steel production and Decision 2007/589 in calculating that benchmark without breaching Article 10a(2) of Directive 2003/87 and the principle of good administration.
50. The Commission’s decision to use such data should not, in my view, be criticised.
51. It is true, as ArcelorMittal argues, that under Article 10a(2) of Directive 2003/87, the Commission is to consult the industry concerned for the purposes of defining the principles for setting the benchmarks in individual sectors or subsectors. It is equally true that Article 14(2) of Directive 2003/87 requires the Commission to employ ‘the most accurate and up-to-date scientific evidence available, in particular from the [International Panel on Climate Change (IPCC)]’ in adopting a regulation for the monitoring and reporting of emissions.
52. Nevertheless, it cannot be inferred from those provisions that the Commission is also under an obligation to employ the data provided by the industry concerned. No such obligation exists in Directive 2003/87.
53. As observed above, the process of setting benchmark values is a complex and technical one. Thus, the EU legislature has merely defined the main objectives for the setting of benchmarks and left the Commission broad discretion in how to achieve those objectives. By extension, the Commission must enjoy a degree of discretion also in the choice of data used in defining the benchmark values as long as that data is not manifestly inappropriate for ensuring that the objectives of Article 10a of Directive 2003/87 (reductions in greenhouse gas emissions and the use of energy efficient techniques) are achieved.
54. Conversely, that data cannot be chosen arbitrarily. That is undoubtedly why, with regard to monitoring and reporting, Directive 2003/87 requires the use of the most accurate and up-to-date scientific evidence available.
55. Recital 11 of Decision 2011/278 explains why, in setting the benchmark for hot metal, the Commission employed information based on the BREF and Decision 2007/589, rather than the data provided by the industry concerned. The documents concerned were used to derive benchmark values where no data, or no data collected in compliance with the benchmarking methodology, was available. The recital further explains the specific problems encountered in setting the benchmark for hot metal: due to a lack of data on the treatment of waste gases, heat exports and electricity production, the benchmark value for hot metal was derived from calculations of direct and indirect emissions based on information on relevant energy flows provided by the BREF documents and default emission factors set out in Decision 2007/589.
56. The Commission further explained in its written submissions that the data provided by the association representing the European steel industry (the European Steel Association (Eurofer)) did not contain data relating, so far as concerns hot metal, to the treatment of waste gases, to heat exports or to electricity production. Even though the data provided by the industry may have been more comprehensive as regards the number of installations considered, that data could not be used in accordance with the Commission methodology.
57. In those circumstances, it is, in my view, warranted for the Commission to have recourse to data from other sources.
58. The Court has already held in Borealis and Others and Yara Suomi and Others that, in relying on the sources at issue here in order to determine the benchmarks in accordance with Article 10a of Directive 2007/83, the Commission did not exceed the limits of its discretion. (24)
59. Indeed, the choice made by the Commission to rely on the BREF for iron and steel production and Decision 2007/589 does not appear to be manifestly inappropriate for the purposes of reducing greenhouse gas emissions and for promoting the use of efficient techniques. Both documents constitute sources of information on emissions and consumption, and on the most efficient techniques in the field.
60. On the one hand, in accordance with (what is now) Article 13 of Directive 2010/75, (25) BREF documents are drawn up and published for each sector on the basis of an exchange of information between the Commission, Member States, the industries concerned and environmental organisations. As their name suggests, they constitute reference documents for the best available techniques in a given sector. The BREF report regarding iron and steel for the year 2001 employed by the Commission was the most up-to-date reference document available in the field. (26)
61. On the other hand, default emission factors referred to in Decision 2007/589 are based, to a large extent, on IPCC guidelines. Those factors are used for establishing national inventories of greenhouse gas emissions under the United Nations Framework Convention on Climate Change.
62. On that basis, I conclude that consideration of the second question referred has disclosed nothing that affects the validity of Decision 2011/278 regarding the data employed by the Commission for the purposes of setting the benchmark for hot metal.
3. The benchmark for sintered ore: the determination of the most efficient installations
63. The third question referred deals with data collection. It concerns the discretion the Commission has in choosing the installations for the purposes of devising the relevant product benchmark. More specifically, the question is whether the Commission could take into account an installation that produces both sintered ore and pellets as one of the 10% most efficient installations in order to set the benchmark for sintered ore without breaching Article 10a of Directive 2003/87.
64. Article 10a(2) of Directive 2003/87 states that in benchmarking, the starting point is to be the average performance of the 10% most efficient installations in a sector or subsector in the EU in the years 2007-2008. That same idea is reiterated in recital 2 of Decision 2011/278. On the other hand, recital 4 of Decision 2011/278 sets out the principle that where a product is a direct substitute of another product, both should be covered by the same product benchmark and the related product definition.
65. It transpires from the case-file and the arguments presented at the hearing that pellets and sintered ore are not direct substitutes and, accordingly, are not covered by the same product benchmark. That is essentially because the composition and product characteristics of pellets differ significantly from those of sintered ore.
66. The complaint put forward by ArcelorMittal, and echoed by the referring court, is based on the assumption that the Commission has, in devising the benchmark for sintered ore, taken into account an installation that produces both pellets and sintered ore, contrary to Article 10a(1) of Directive 2003/87 and recital 4 of Decision 2011/278. Accordingly, the benchmark for sintered ore is, in ArcelorMittal’s submission, distorted due to the use of data pertaining to the production of pellets.
67. At the hearing, important clarifications were made in that regard. Indeed, even though the written observations of the Commission are by no means a model of clarity on that point, the oral argument presented by that institution has left me satisfied that the choice of installations cannot affect the validity of Decision 2011/278 as regards the benchmark for sintered ore.
68. The Commission explained that the installation in question is unique in the European Union. It produces a mixture of pellets and sintered ore, which, as regards its properties and usage, is employed as a direct substitute for sintered ore in the blast furnaces in the integrated steel plant. Even though the integrated steel plant in question includes a pellet unit and a sinter production unit, those units are connected and run simultaneously to produce a mixture that is directly fed into the blast furnace. In that regard, the Commission also noted that the pellet plant is directly linked to that of the sinter unit in accordance with the definition of sintered ore in Annex I to Decision 2011/278. (27) In the opinion of the relevant stakeholders and experts consulted by the Commission, the production process of that mixture can be regarded as similar to that of sintered ore. In other words, the final product has similar properties to sintered ore and it is used as a direct substitute for it in the blast furnace. (28)
69. Despite those clarifications, ArcelorMittal maintained its position at the hearing. It noted, in particular, that all official documents mention that the installation in question has separate production units for pellets and sintered ore. It also observed that it is in no way unusual to mix pellets with sintered ore in a blast furnace.
70. There are two points to be made here. Firstly, EU acts are presumed to be lawful. (29) Secondly, as observed above, we are dealing with complex technical assessments with regard to which the Commission enjoys broad discretion.
71. From the information available, it is not obvious that in taking the installation in question into account, the Commission made a manifest error of assessment. On the contrary, it could, in accordance with the margin of discretion it enjoys, decide that the installation in question may be regarded as one of the reference installations for the purposes of setting the benchmark for sintered ore. Indeed, it appears that, despite the special characteristics of the installation (in particular as regards the existence of a pellet unit on-site for the purposes of the production of the mixture of sintered ore and pellets described above), the product thus arrived at constitutes a direct substitute for sintered ore.
72. It is clear that determining whether the special characteristics of a particular installation or rather, the uniqueness of those characteristics in the EU, have a bearing on its inclusion as a reference installation falls within the sphere of complex technical assessments. The Commission is far better placed than the Court to make such an assessment.
73. Overall, the Commission’s approach seems to me to be warranted. If the installation in question were not taken into consideration as one of the reference installations, because of its special characteristics, that would artificially limit the data collection and thus exclude one of the most efficient installations in the sector. As a result, the benchmark arrived at would be considerably higher. (30) That would clearly not reflect the overall objective of Directive 2003/87 to reduce greenhouse gas emissions in an economically efficient manner. Indeed, it seems that precisely because of those characteristics, that installation is capable of optimising its emissions.
74. As the Commission has pointed out, the taking into account of the installation producing a substitute for sintered ore is in line with the aim of Article 10a(1) of Directive 2003/87, which is to encourage the use of procedures generating the least emissions. Benchmarking should be based on the most efficient installations in the sector and should not take account of the technology used, mix of fuel employed, and other factors such as meteorological conditions or the raw materials used. If not, the objective of reducing greenhouse gas emissions would be considerably watered down.
75. On that basis, I conclude that consideration of the third question referred has disclosed nothing that affects the validity of Decision 2011/278 regarding the choice of reference installations for the purposes of setting the benchmark for sintered ore.
4. The benchmark for sintered ore: the obligation to state reasons
76. The fourth question referred asks whether, in Decision 2011/278, the Commission complied with its obligation to state reasons in accordance with Article 296 TFEU with regard to the setting of the benchmark for sintered ore. In particular, the referring court asks whether that obligation has been met as regards the choice of reference installations for the purposes of setting the benchmark for sintered ore.
77. On a general note, it is settled law that the statement of reasons required by Article 296 TFEU must disclose in a clear and unequivocal fashion the reasoning followed by the institution which adopted the measure in such a way as to enable the person concerned to ascertain the reasons for the measures and to enable the court having jurisdiction to exercise its power of review. The EU institution which adopted the contested measure is not, however, required to go into every relevant point of fact and law. (31)
78. The obligation to state reasons is highly context dependent: the question whether a statement of reasons satisfies the requirements of Article 296 TFEU must be assessed with reference not only to the wording of the measure but also to its context and to the whole body of legal rules governing the matter in question. The Court has held that if the contested measure clearly discloses the essential objective pursued by the institution, it would be excessive to require a specific statement of reasons for each of the technical choices made by the institution. (32) In particular, as opposed to individual measures, in the case of a measure intended to have general application, as is the case of Decision 2011/278, the statement of reasons may be limited to indicating, first, the general situation which led to its adoption and, second, the general objectives which it is intended to achieve. (33)
79. As that case-law makes clear, the Commission is not under an obligation to explain the specific reasons for the various technical choices it has made. Since the Commission enjoys a broad discretion and that discretion is open to judicial review only within narrow limits, the statement of reasons need contain only the elements necessary to support such limited review. (34)
80. Nevertheless, the Court must be able to conduct its review, failing which the statement of reasons cannot be considered sufficient.
81. In the present case, it is my view that the reasons regarding the choice of reference installations are stated with sufficient clarity in Decision 2011/278. The reasons for adopting Decision 2011/278 and the objective thereof emerge from the recitals of the decision. In addition, a number of technical details are included in the recitals, which further clarify the method employed by the Commission to set benchmarks, including the choice of installations.
82. True, the Commission has not explained in detail the technical reasons that led it to consider, for example, that an installation producing both pellets and sintered ore can be taken into account in calculating the benchmark for sintered ore. However, requiring it to do so would, in essence, extend its obligation to state reasons to include, in the preamble, an account of the complex technical choices that it has had to make.
83. Bearing in mind the type of measure we are dealing with and the explanations offered in the preamble regarding both factual and legal considerations pertaining to the setting of benchmarks, the Commission is under no obligation to provide such an account. In particular, it is apparent from the recitals that Decision 2011/278 was adopted pursuant to Directive 2003/87 in order to set EU-wide rules for the free allocation of emission allowances during a transitional period. It is clear that the Commission has sought to devise those rules in a way that ensures that greenhouse gas emissions are reduced.
84. Most importantly, in recital 2, the Commission explains that the starting-point for calculating a benchmark is the average performance of the 10% most efficient installations in a sector or subsector. In addition, recital 4 explains that where a product is a direct substitute of another product, both should be covered by the same product benchmark and the related product definition. Recitals 6 to 8 explain what kind of information has been taken into consideration in calculating the benchmark values. Recital 11 further illustrates the Commission’s approach in cases where no data or insufficient data has been available, and recital 12 explains how the Commission has proceeded where it has not been possible to set a benchmark.
85. For the Court to conduct its review, the information contained in the preamble is sufficient. Not least for practical reasons, the Commission cannot be required to explain the characteristics of each reference installation and make explicit the technical reasons why certain installations were considered suitable for benchmarking and others not. To determine whether the relevant product benchmark is vitiated by a manifest error of assessment, it is sufficient to understand the objective of the measure in question and the general method employed by the Commission to set those benchmarks (including the general rule applicable to the choice of reference installations and the treatment of substitutes). Those considerations emerge with clarity from the recitals of Decision 2011/278.
86. On that basis, I conclude that consideration of the fourth question referred has disclosed nothing that affects the validity of Decision 2011/278 regarding the Commission’s obligation to state reasons in setting the benchmark for sintered ore.
IV. Conclusion
87. In the light of the arguments presented, I propose that the Court should answer the questions referred by the Tribunal administratif de Montreuil (Administrative Court, Montreuil, France) as follows:
Consideration of the questions referred has disclosed nothing that affects the validity of Commission Decision 2011/278/EU of 27 April 2011 determining transitional Union-wide rules for harmonised free allocation of emission allowances pursuant to Article 10a of Directive 2003/87/EC of the European Parliament and of the Council, as regards the setting of benchmarks for hot metal and sintered ore.