Single Interconnectors and Congestion Management Challenges-The Legal Framework?

This article explores the impact of an important ruling from the ECJ in Baltic Cable AB v Energimarknadsinspektionen on 11 March 2020 in which that court narrowly avoids giving the referring Swedish court a green light to interpret a key provision of the EU internal energy market legislation contra legem. Invoking instead the principle of non-discrimination, the ECJ relies on a classic remedy to recognise that a company owning and operating an electricity interconnector should be entitled to earn a reasonable profit. Although the interconnection of energy networks is an objective enshrined in Article 194(1) TFEU, the realisation of this objective has spawned a dense and highly technical web of regulation. This article explains the Court’s reasoning and its potential legal as well as economic impact in this complex and evolving regulatory space. We explain that while valuable progress has been made on technical harmonisation, classic fundamental principles of EU law, such as the non-discrimination principle, remain pivotal for resolving modern and central issues of electricity market integration.

1 Introduction * One of the EU's central energy policy aims is creating a European internal market for electricity (IEM) that allows for unrestricted trade of electricity across the continent and facilitates achieving the EU's sectoral policy objectivessustainability/decarbonisation, competitiveness and security of supply. The success of this endeavour critically depends on adequate investment in physical interconnections between the Member States. In other words, a physically interconnected Europe-wide electricity grid is a conditio sine qua non for a genuine IEM. 1 Accordingly, Article 194(1) TFEU stipulates that the "Union policy on energy shall aim [...] to [...] promote the interconnection of energy networks." 2 Interconnectors usually belong to the operators of the high voltage grids in the Member States (transmission system operators -TSOs), but there is a growing number of single interconnectors owned by independent investors. The legal position of these interconnectors has been a controversial subject since the beginning of energy market liberalisation. A particularly fundamental issue concerns the regulation of the revenues generated on these interconnectors. The judgment by the Court of Justice of the European Union (ECJ) in the Baltic Cable case finally sheds light on this and other related issues and thus has wider implications for financing future interconnection investments. 3 While the need for substantial investment in more interconnection capacity is widely recognised, the ways and means to promote such investments are more controversial. Three pillars of effective interconnector investment regulation are frequently discussed. First, the unbundling of network from generation functions to promote adequate levels of network investment. 4 Secondly, the designation of supranational regulatory powers to a specialised European agency, i.e. the Agency for the Cooperation of Energy Regulators (ACER). And last, but by no means least, the earmarking of so called "congestion income" to incentivise new construction and efficient operation. 5 The case at hand is of direct relevance to this last "pillar" and the business model of future investments. A significant feature of the Baltic Cable judgment is that the ECJ comes very close to an interpretation contra legem to avoid a violation of the principle of non-discrimination. The Court based its reasoning on long-standing fundamental principles of EU and energy law. However, as the regulatory regime governing the operation of EU interconnectors has gradually evolved since the adoption of the First Electricity Directive 96/92 in 1996 and the First Electricity Regulation 1228/2003 in 2003 into a dense and technically complex array of detailed legislation, and in particular delegated legislation, the impact of the ruling on the current regulatory architecture will also be considered. * We thank Dan Roberts, director at Frontier Economics, and Tim Schittekatte, Research Associate at the Florence School of Regulation, for helpful comments and discussions on earlier drafts of this article. Any shortcomings are our own responsibility.
For reasons of transparency, we wish to disclose that the research and statements contained in this note could potentially affect a company that we cooperate with in a professional capacity, however on different issues. This note is based on our independent work, so that our conclusions are unaffected by said cooperation.

Background
Although it concerns what might seem at first glance a highly technical regulatory issue, the Baltic Cable case attracted considerable attention, with the European Commission, the European Council, the European Parliament as well as the Finnish and Spanish governments intervening. In addition, it is worth noting that the President of the ECJ (judge Lenaerts) formed part of the Chamber, which is assumed to indicate that the Court itself considers the judgment important.
The case concerns the EU regulatory framework for the operation of high voltage electricity transmission systems, particularly the use of so-called congestion income. This income is obtained by allocating the scarce transmission capacity on cross-border interconnectors to the international power trade. 6 Congestion income depends on the electricity wholesale prices in the connected markets, since it is calculated as the product of the price differential and the amount of cross-border electricity exchanges. Prices on connected markets align with increased cross-border trade, so that this price difference only persists as long as there is unsatisfied demand for cross-border transmission capacity. This is the case if there is insufficient investment in cross-border lines or if existing lines cannot be fully utilised due to congestion 7 in the "national" network (hence the term congestion income). Most EU borders are affected by both issues simultaneously, and as the ECJ recognises in its ruling, TSOs have no inherent interest to eliminate congestion and provide unlimited cross-border transmission capacity.
At the same time, one of the core aims of EU energy law is the creation of sufficient interconnector capacity to create an IEM characterised by unrestricted cross-border trade. On the one hand, EU energy law therefore tasks TSOs with managing congestion and maximising cross-border capacity in day-today operation using so-called remedial actions, ad hoc measures that are used for maintaining the necessary balance of the transmission grid and to relieve congestion. 8 One particularly relevant example is redispatching, where the responsible TSO orders power producers to reduce power production in the network area "before" congested lines and to increase production "behind" the bottleneck, in order to divert flows and free up transmission capacity. Counter-trading serves the same purpose, but is effectuated on electricity markets instead of bilaterally between TSOs and power producers. Remedial actions can also be applied across borders. 9 In Article 16(4), the Third Electricity Regulation stipulates that "[c]ounter-trading and redispatch, including cross-border redispatch, shall be used to maximise available capacities [and a] coordinated and non-discriminatory process for cross-border remedial actions shall be applied to enable such maximisation, following the implementation of a redispatching and counter-trading cost-sharing methodology".
On the other hand, EU energy law establishes concrete aims regarding investment in interconnector capacity at Member State level. This note focuses on the latter aspect. Whereas several Member States will stay shy of the 10 per cent interconnection target set for 2020, an even higher interconnection target of 15 percent throughout the EU is set for 2030. 10 Achieving these targets thus requires considerable investment in new interconnectors.

Investment in Interconnectors: The Quest for Sufficient Capacity
There are two kinds of interconnectors in the EU. On the one hand, the standard type are cross-border lines that are part of an extended transmission network with connected grid users, owned by "national" TSOs. 11 The TSOs levy tariffs on the national grid users to cover the costs of maintenance and operation, new investments and to retain a reasonable return. On the other hand, there are single interconnectors which connect national systems, but do not form part of the larger network and to which no network users are connected. The latter are owned by separate single interconnector companies (SInCs) and operated on a different commercial basis. SInCs cannot pass on or "socialise" the risks and costs of the operation of their interconnector to network users such as regular TSOs do. Congestion income is their sole source of revenues. Examples of such single interconnectors include BritNed (between Great Britain and the Netherlands) and the NEMO Link (between Great Britain and Belgium). 12 Regardless of their commercial basis, interconnector investments are recognised as highly risky and complex ventures for several reasons: First, in fully liberalised markets, grid and generation investments are decoupled due to unbundling. 13 Grid investments thus face uncertain generation decisions, i.e. uncertainty on the actual use of the infrastructure. In the worst case, an interconnector can become a stranded asset. Second, cross-border projects are subject to high regulatory uncertainty over time. Changing regulatory frameworks, the introduction of new congestion management mechanisms or the review of regulated tariffs might impact significantly on the return on investment. Third, investors also face uncertainty concerning possibly changing market architectures and energy mixes of the interconnected markets as well as volatile fuel and carbon pricesall influencing the price differential, i.e. the base of congestion revenue. Finally, potential interconnector investors are further discouraged by the existence of a regulatory gapif there is no single competent authority that decides on crossborder and regional issues. Investors face an important risk of project failure when the competent national regulatory authorities (NRAs) at each end of the interconnector are unable to agree on key regulatory provisions for a cross-border project, especially if there is no supranational authority to settle the conflict. 14 ACER has only limited powers to adjudicate in case the concerned NRAs request a decision or they cannot agree within the legally specified deadline. 15 To address these hurdles to investment with a view to create a properly functioning IEM, the EU has endeavoured to provide a well-designed, incentive-based regulatory framework over the course of four Robert Schuman Centre for Advanced Studies Working Papers comprehensive legislative packages for the electricity sector. The Third Electricity Package, adopted in 2009, was at issue in the present case. It has recently been superseded by the "Clean Energy Package" (CEP). The regulation of congestion income is a key element in both packages.

Congestion Income Regulation: A Solution?
Due to its natural monopoly characteristics, the transmission sector is a regulated activity in the EU: most transmission investments, including interconnectors, are recovered from revenues generated through regulated transmission tariffs. A debate has raged among economists as to whether transmission investment by SInCs leads to suboptimal results or is an acceptable response to the insufficient level of interconnector investment in the EU. 16 It is a second-best option as the alternativeregulated interconnector investmenthas so far failed to deliver satisfactory results. 17 As mentioned above, congestion indicates that demand for transmission capacity exceeds supply. In the case of a regular TSO considering an (interconnector) investment, the TSO would be required to apply for regulatory approval with regard to its investment plan in order to consider the related investment costs when calculating the regulated tariff. These costs would then be recovered from the revenues of the adjusted regulatory tariff. In addition, investment costs can be covered using congestion income. As we explain below, EU law restricts not only the options available to TSOs but also to the NRAs.
In the case of SInCs, which do not collect tariffs, the investors take the risk. New investments can qualify for a temporary exemption from the congestion revenue rules in return, so that they are allowed to keep their congestion revenue at least for the period necessary to pay back their original investment. 18 In accordance with successive EU regulation, the award of an exemption is decided upon by each NRA on a case-by-case basis and, ultimately, approved or rejected by the European Commission.

The Regulatory Framework for Congestion Income
Article 16(6) of the Second Electricity Regulation 19which was at issue in the Baltic Cable caserestricts what TSOs can do with congestion revenues generated on an electricity interconnector. This provision sets out a menu of options: "Any revenues resulting from the allocation of interconnection shall be used for the following purposes: (a) guaranteeing the actual availability of the allocated capacity; and/or (b) maintaining or increasing interconnection capacities through network investments, in particular in new interconnectors. 16 A frequent argumentation against investment by SInCs is that private investors will aim for an amount of capacity that provides maximum congestion revenues, which may stay short of the capacity required for maximising social welfare, cf.

Single Interconnectors and Congestion Management Challenges -The Legal Framework?
European University Institute 5 If the revenues cannot be efficiently used for the purposes set out in points (a) and/or (b) of the first subparagraph, they may be used, subject to approval by the regulatory authorities of the Member States concerned, up to a maximum amount to be decided by those regulatory authorities, as income to be taken into account by the regulatory authorities when approving the methodology for calculating network tariffs and/or fixing network tariffs.
The rest of revenues shall be placed on a separate internal account line until such time as it can be spent on the purposes set out in points (a) and/or (b) of the first subparagraph." We refer to this as the "earmarking regime". The rules in question are sensible in relation to TSOs that also run a more extensive transmission network, as they avoid that congestion revenues become windfall profits for the regular TSO and ensure that network users, that pay for all costs and risks associated to the interconnector through the network tariffs, benefit from the congestion revenues. However, they create a problem for SInCs, which do not charge any network tariffs, and the calculation of such tariffs would be futile as no network users are connected to the interconnector. At the same time, an interconnector is a vast one-time investment that amortises over decades. 20 Therefore, even if building additional cross-border capacity would be economically efficient, the original investors cannot be expected to make investments of similar magnitude into an existing single interconnector to increase its capacity. Yet it was unclear whether SIncS could cover their costs of operation and maintenance with congestion income under the earmarking regime. As mentioned before, new single interconnector projects may qualify for an exemption from the earmarking regime, inter alia because they cannot socialise costs. However, this path is only available to interconnectors not completed by 4 August 2003. 21

The Baltic Cable
The Baltic Cable is an electricity interconnector between Sweden and Germany. It is owned by Baltic Cable AB (BCAB), a Swedish SInC whose only business is the operation of this interconnector, which is not an extension of either the adjacent Swedish or the German transmission grids. The Baltic Cable became operative in 1994 and as such it could not qualify for an exemption as "new" infrastructure. Up to the events leading to the ECJ's judgment in the present case, the earmarking rules described above were not applied to BCAB, who used its congestion income to cover the maintenance and operation costs of the interconnector and regarded any surplus as regular revenues.
The NRA competent to supervise BCAB's compliance with the earmarking rulesthe Swedish Energy Markets Inspectorate (Energimarknadsinspektionen -"EI")faced a dilemma. A strict application of the earmarking regime would require that BCAB keep all its income on a special "escrow" accounteventually dooming it to bankruptcy, as BCAB had to cover operation and maintenance costs. At the same time, it was reluctant to deviate from the earmarking regime by applying it contra legem, 22 which would open up the possibility of the Commission initiating infringement proceedings against Sweden. 23 On 9 June 2016, EI ordered BCAB to place its congestion revenues on a separate internal account until they could be used in line with the purposes specified in EU energy law. BCAB reacted by requesting that its congestion income be taken into account in conjunction with calculating and/or fixing the tariffs for accessing the Baltic Cable. 24 24 In the absence of network users to apply such tariffs to and given the prohibition in art.14(5) of the Second Electricity Regulation to charge network tariffs on cross-border transmissions already subject to congestion income ("pancaking"), it is unclear how this request could have been implemented in practice.
BCAB challenged both decisions before the Administrative Court in Linköping, Sweden (Forvaltningsrätten i Linköping), claiming that the earmarking regime in EU energy law did not apply to SInCs, or that it had to be interpreted as allowing them to cover the maintenance and operation costs of the Baltic Cable and make a reasonable return with the congestion revenues. The Forvaltningsrätten submitted the case to the ECJ for a preliminary ruling, in essence requesting the Court to clarify whether the earmarking regime applied to SInCs and whether a SInC's particular (income) situation warranted a modification of the earmarking regime. The five questions referred to the ECJ can be paraphrased as follows: 1. Does the earmarking regime apply in all cases where a person obtains congestion revenues, or does it apply only to TSOs?
2. If the earmarking regime applies only to TSOs, is a SInC a TSO?
3. If the earmarking regime applies to SInCs, can the costs relating to the operation and maintenance of an interconnector in any event be regarded as network investments to maintain or increase transmission capacities?
4. If the earmarking regime applies to SInCs, and costs relating to the operation and maintenance of an interconnector are considered as eligible network investments, can the NRA approve that a SInC, which has a methodology for fixing tariffs but does not have customers paying network tariffs which can be reduced, uses congestion revenues to make a return or, if operation and maintenance costs do not constitute eligible network investments, to operate and maintain its interconnector?
5. If the earmarking regime applies to SInCs, and SInCs may neither use congestion income to cover costs relating to the operation and maintenance of their interconnector nor to make a return, or if SInCs may use congestion income to cover costs relating to the operation and maintenance, but not to make a return, is application of the earmarking regime to a SInC contrary to the EU law principle of proportionality or any other applicable principle?
The Judgment

The Proceedings and the Advocate General's Opinion
In general, the intervening parties argued that the earmarking regime must apply (in full) to SInCs. Only the Finnish government argued in favour of allowing SInCs to use congestion income to cover their maintenance and operation costs and making a reasonable profit. In particular, the Commission advocated a strict application of the earmarking regime and considered that "general tariffs paid by all users of the national transmission system" may be used for covering interconnector costs, naming the German Offshore-Umlage as an example, a surcharge levied on German power consumers to cover the connection cost of offshore wind parks. 25 In his opinion, Advocate General (AG) Tanchev argued that the earmarking regime applies to SInCs, regardless of whether they are considered as TSOs. 26 He merely scrutinised whether SinCs are TSOs for the sake of completeness, answering the second referred question in the affirmative. 27 Since the earmarking regime specified that congestion income could be used to cover "in particular" investments in new interconnectors, AG Tanchev reasoned it was not excluded to use such income for covering maintenance costs as investments in existing interconnectors. 28  the third question, the AG also discussed the other option to spend congestion income foreseen by the earmarking regime, i.e. for guaranteeing the actual availability of the interconnector capacity. He concluded that costs relating to the operation and maintenance are incurred to guarantee the availability of cross-border capacity and could thus be covered with congestion income, also referring to subsequent clarifications in the wording of the revised earmarking regime under the CEP. 29 Furthermore, he reasoned that the earmarking regime allows for making a reasonable return from congestion income, subject to the approval and supervision of the NRAs, in order to encourage investment in interconnectors. 30 As an exception, AG Tanchev reasoned that where SInCs dispose of other sources of income intended to cover costs relating to the operation and maintenancesuch as specific tariffs or fees available under the national regulatory frameworkcongestion income may only be used to cover maintenance and operation costs in the second order. 31 In addressing the last question referredon the proportionality of the earmarking rules -the AG highlighted the EU's broad discretion in areas involving complex assessments and evaluations of political, social and economic choices. Since he argued that the earmarking regime allows using congestion income both for covering costs relating to the operation and maintenance and for making a reasonable profit, AG Tanchev concluded that it did not go beyond what was necessary for the achievement of the underlying objective to create a functioning IEM. Finally, while the earmarking rules limit any TSO's right to property as guaranteed in Article 17(1) of the Charter of Fundamental Rights of the European Union, he considered this limitation as justified for the same reasons. 32

The ECJ's Judgment
In its judgment handed down a mere four months later, the ECJ reached the same conclusions as AG Tanchev, albeit pursuing a different line of reasoning. The Court came remarkably close to interpreting the earmarking regime contra legem to avoid, on the one hand, discrimination between standard TSOs and SInCs and, on the other hand, a negative impact on the EU policy aim of increasing the interconnection between the Member States.
In reply to the first and second questions, the ECJ stated that SInCs are to be regarded as TSOs, to which the earmarking regime applies in principle. The ECJ thus followed the Commission's opinions on the certification of BCAB as a TSO by the German NRA, 33 and also the German Supreme Court (Bundesgerichtshof). 34 Moreover, the ECJ recalled its earlier VEMW judgment that EU energy law provisions have to be interpreted according to their context and in the light of the aims they pursue. 35 Putting the conditions for the use of congestion income in conjunction with their wider context reveals that they apply to TSOs only. 36 Likewise, while the ECJ recognised that the legal framework does not explicitly cover the case of SInCs, the underlying premise is that SInCs are TSOs. 37 The ECJ then addressed the third question referred to itwhether costs relating to the operation and maintenance constitute "network investments to maintain or increase interconnection capacities". 38  this context, the Court highlighted that congestion income only accrues where congestion persists and reasoned that TSOs therefore have no economic incentive to reduce congestion, which EU law seeks to counteract through the earmarking regime. 39 However, the ECJ stressed that according to other closely related provisions, investments are to be "assigned to specific predefined projects", 40 whichcontrary to AG Tanchev's conclusionsexcludes considering operation and maintenance costs as investment costs. 41 Strictly confining its scrutiny to the wording of the third question, the Court did not discuss the AG's statement that congestion income could be used to cover costs relating to the operation and maintenance of an interconnector as costs for guaranteeing the actual availability of cross-border capacity.
In contrast to its restrictive interpretation of the wording of earmarking regime, the Court applied a more liberal argumentation in its reply to the fourth question, and reasoned that in the case of SInCs, the earmarking regime must be applied so as to allow SInCs to operate and maintain their interconnector and make a reasonable return. The Court emphasised that whereas regular TSOs earn revenues from their transmission tariffs, which allow them to cover costs relating to the operation and maintenance and make a reasonable return, this is not the case for SInCs. 42 EU energy law reiterates the fundamental principle of non-discrimination in Article 16(1) of the Second Electricity Regulation and opens up for differential treatment of single interconnectors in a variety of contexts, including in the context of congestion income regulation "due to the specific nature of [an] interconnector". 43 Hence, the ECJ considered that the application of the earmarking regime in the case of SInCs has to account for their particular situation. 44 Given that all TSOs are to fulfil their tasks "in financially acceptable conditions", the earmarking rules must not deprive SInCs of their entire congestion income, which would lead to their bankruptcy. 45 The ECJ concluded that the NRAs "must, [when applying the earmarking regime to a SInC,] put [that SInC] in a position in which it is able to carry out its activity in financially acceptable conditions […] which includes making an appropriate profit […] to prevent it being discriminated against by comparison with the other TSOs concerned." 46 Given its response to this fourth question, the ECJ saw no need to address the fifth question referred to it on the proportionality of the earmarking rules.

Discussion
The ECJ's decision in Baltic Cable brings about a welcome clarification of the legal situation of SInCs and outlines an application of the EU earmarking regime to SInCs that does not compromise the financial viability of single interconnector projects. However, the Court has only solved part of the puzzle and left the details to the NRAs, allowing them considerable discretion with regard to implementation of the regulatory provisions. This section will discuss the main elements of the ECJ's ruling, also in the light of the revisions introduced subsequently in the CEP, as well as possible implications of the judgment for the Methodology on the Use of Congestion Income currently being drafted under auspices of the CEP. 47

On the TSO Status of SInCs
The Court rightly qualified SInCs as TSOs in the meaning of EU energy law. However, the ECJ's discussion of the first and second questions is surprisingly brief, particularly given the Court's emphasis on the systematic context of the relevant provisions. For instance, further support for the ECJ's reasoning can be derived from considering the underlying aim of EU energy law to create an IEM. To achieve this goal, TSOs are subject to numerous obligationssuch as the congestion income and unbundling regimes. If SInCs were not considered as TSOs (and thus not subject to these strict sector-specific rules), TSOs could avoid the rigorous regulatory framework for transmission systems by transferring ownership of interconnectors to separate undertakings. This would create a dangerous regulatory gap, putting the aims of EU energy policy into jeopardy.

Implications for the Implementation of the Clean Energy Package
The judgment is based on Article 16(6) of the Second Electricity Regulation. The earmarking regime has been revised with the recent adoption of the CEP. Under the Third Electricity Regulation, derogations from the earmarking regime due to the specific nature of an interconnector are still possible, so that the ECJ's statements continue to apply in principle. 48 There are, however, some changes to the wording of the relevant provisions. The earmarking regime is set out in Article 19 (2) and (3) of the Third Electricity Regulation as follows (substantial changes are emphasised): "(2) The following objectives shall have priority with the respect to the allocation of any revenues resulting from the allocation of cross-zonal capacity: (a) guaranteeing the actual availability of the allocated capacity including firmness compensation; or (b) maintaining or increasing cross-zonal capacities through optimisation of the usage of existing interconnectors by means of coordinated remedial actions, where applicable, or covering costs resulting from network investments that are relevant to reduce interconnector congestion.
(3) Where the priority objectives set out in paragraph 2 have been adequately fulfilled, the revenues may be used as income to be taken into account by the regulatory authorities when approving the methodology for calculating network tariffs or fixing network tariffs, or both. The residual revenues shall be placed on a separate internal account line until such a time as it can be spent for the purposes set out in paragraph 2." First, the revised earmarking regime establishes that using congestion revenues for the objectives defined therein "shall have priority", which explicitly renders using congestion income for lowering tariffs a subordinate option. Already under the old earmarking rules, only the part of congestion income could be used for lowering tariffs that "cannot be efficiently used for the purposes" of guaranteeing, maintaining or increasing cross-border capacity. 49 Yet in practice, the amount of congestion revenues used for reducing tariffs exceeds the amount used for the other purposes. One reason might be that NRAs favour immediate resultsreduced network tariffs in their own Member State -over (uncertain) longrun welfare gains and therefore do not sufficiently insist on interconnector investment. In this regard, it 47 Third Electricity Regulation art.19(4). 48 Recital 38 in the Third Electricity Regulation. 49 Second Electricity Regulation art.16(6). was problematic that the assessment what precisely constituted an efficient use of congestion income contained a strong subjective element that opened for differing interpretations. The revised and stricter wording of the earmarking regime clarifies that fostering cross-border connectionswhich is essential for achieving the IEMenjoys priority over national considerations.
Second, the list of priority uses for congestion income has been extended, as TSOs may now use congestion revenues for "maintaining or increasing cross-zonal capacities through optimisation of the usage of existing interconnectors by means of coordinated remedial actions". The previous earmarking rules did not explicitly mention remedial actions. For SInCs, this change is quite significant, since they can theoretically coordinate the use of remedial actions with the neighbouring regular TSOs and might be, under the new earmarking regime, obliged to bear associated costs. However, such an understanding clashes with the EU law principle of non-discrimination, which also prohibits treating different situations in the same way. 50 To begin with, the ECJ recognises that SInCs cannot invest in their "network" to relieve congestion, in contrast to regular TSOs. 51 Moreover, there are no (re-)dispatchable generators or loads connected to single interconnectors. Accordingly, there is no congestion on single interconnectors that has to be resolved through coordinated remedial actions. For the same reasons, regular TSOs dispose of numerous options to deal with congestion (including measures without costs, such as switching operations), whereas SInCs cannot influence the topography of other TSOs' networks or the production of power in these grids through redispatching. At best, they can use countertrading, which is not always the most efficient remedial action available. In order to avoid a discrimination between SInCs and regular TSOs, these significant differences need to be taken into account in connection with congestion income regulation. An obligation to participate in the congestion management costs in adjacent grids seems disproportionate and could jeopardise the financial viability of SInCs. Therefore, SInCs cannot be obliged under the revised earmarking regime to spend congestion revenue to cover the remedial action costs in adjacent grids due to the "specific nature of the [ir] interconnector". 52 Third, the revised earmarking regime newly specifies that investments must be "relevant to reduce interconnector congestion". 53 From the wording, it is unclear whether the phrases "maintaining or increasing cross-zonal capacities" and "relevant to reduce interconnector congestion" are cumulative, or if they denote two distinct alternatives for spending congestion income. In particular, increasing crosszonal capacities and reducing interconnector congestion seem to be two sides of the same medal, whereas an investment that merely maintains cross-zonal capacities while also reducing interconnector congestion appears oxymoronic at first glance. However, a practically relevant example concerns network reinforcements that avoid additional congestion caused by the deployment of new generators in the vicinity of an interconnector. Moreover, allowing only investments that reduce congestion and thereby increase cross-zonal capacities to be funded with congestion income would limit the scope of eligible investments in comparison to the previous regime, which clearly included investments that merely maintained interconnection capacity. Most importantly, it is evidently undesirable from the perspective of market integration to forego investments that avoid a future increase of interconnector congestion only because they do not also reduce current congestion levels. The main argument in favour of reducing the scope of eligible investments would be to avoid overinvestment, yet this appears as a subordinate issue given the EU's concerted action to promote important infrastructure reinforcements and its ambitious interconnection targets. 54 The newly introduced focus on investments that are relevant to reduce interconnector congestion thus only serves to exclude purely internal investments without any

Single Interconnectors and Congestion Management Challenges -The Legal Framework?
European University Institute 11 impact on cross-zonal capacity. It is for the NRAs to ensure that this is respected in practice, keeping in mind that investments by SInCs in their interconnector are relevant to reduce cross-zonal congestion by default. It should further be noted that the revised earmarking regime has relinquished the preference that "[t]he use of congestion income for investments […] shall preferably be assigned to specific predefined projects". 55 According to the ECJ, it was particularly this explicit preference for specific predefined projects that excluded considering maintenance costs as eligible investments. 56 Its deletion renders a wider interpretation possible, and one may ask whether the understanding of AG Tanchev that maintenance costs constitute investments that can be covered with congestion revenueswhich the Court did not endorsecould be possible under the revised earmarking regime.

The NRAs as "Good Shepherds"
In answering the fourth question, the ECJ establishes a far-reaching mandate for the NRAs with regard to congestion income regulation of SInCs. The Court emphasises the contribution of single interconnectors to the energy goals of the EU, in particular enhancing competition on the internal market. 57 This, in conjunction with the general prohibition to discriminate and the specific situation of and challenges faced by SInCs, leads the Court to a weighty normative conclusion. According to the ECJ, the NRAs are mandated to apply the earmarking regime to SInCs in a way that does not jeopardise their financial viability, in order to facilitate the continuing integration of the European electricity markets. However, the ECJ's statements only define fundamental conditions, leaving considerable discretion to the NRAs to determine the details. This leads to three observations.
First, the Court's conclusion rests on a recital to the Electricity Regulation that allows to deviate from the earmarking regime where "the specific nature of the interconnector concerned justifies an exemption". 58 This wording establishes a relatively low threshold. Deviating from the earmarking rules seems possible whenever the circumstances of an interconnector justify itthere is no need for the owner to demonstrate compelling reasons. In the case of SInCs, NRAs appear to have no other choice than applying the earmarking rules in accordance with the ECJ's statements, yet other interconnectors regardless of their ownershipmight potentially be eligible, too. It will be highly interesting to see if and on what grounds regular TSOs will endeavour to obtain a "special treatment" for their cross-border connections. This raises the question of the legal nature of such a special treatment, particularly with a view to the two models for treating congestion incomeregulated or through an exemption. The relevant recital speaks of an exemption, while in the words of the ECJ, it is "possible to derogate from the rules on the use of congestion revenues". 59 Nevertheless, it is important to stress that according to the unambiguous wording of Article 17 of the Second Electricity Regulation, the granting of a formal exemption from the earmarking rules is only possible for new single interconnectors. 60 Thus, the ECJ rather implies a more flexible application of the earmarking regime in a regulated context that allows the owner to retain some of the congestion income as "an appropriate profit". 61 This can be construed as a tacit approval of congestion income regimes based on a more liberal understanding of the earmarking rules, such as the British "cap and floor" regime for interconnectors which are owned by SInCs without a formal exemption. 62 This regime, a variant of which is also considered for the planned 55 Second Electricity Regulation art.16 (6) NorthConnect cable between Norway and Scotland, 63 ensures the viability of SInCs by guaranteeing a minimum income for covering the operating costs and a reasonable return on investment (the "floor"), while revenues above a certain threshold are used to cover the operating cost of the national grid (the "cap"). This way, both costs and revenues are socialised to some extent.
Second, the NRAs of Member States connected by a single interconnector have to agree on conditions for the concerned SInC's use of congestion income that are in line with the requirements identified by the ECJ and issue a formal decision under domestic law. When national entities autonomously decide on deviations from binding EU law, they have to be particularly careful to consider the aims and context of the provisions in question, as the Court emphasised in VEMW. 64 Therefore, NRAs will have to give the aims of EU energy law and the fostering of the IEM particular weight when fleshing out the mandate established by the Court and deciding on derogations from the earmarking regime due to the specific nature of interconnectors.
Third, electricity transmission tariffs differ considerably between Member Statesin contrast to the harmonised gas transmission tariffs. NRAs hence retain a considerable margin of discretion when establishing the details of such derogations. This is partially justified due to the expertise of the NRAs and their role as national authorities under the EU law principle of subsidiarity. Yet the absence of a "standard frame of reference" could result in very heterogeneous investment conditions for SInCs in different Member States. This is problematic, given that the need for additional investment in interconnectors rather necessitates fostering investor certainty through predictable and uniform conditions. Moreover, decisions on congestion income regulation concerning single interconnectors will involve at least two NRAs and two jurisdictions. In absence of direct recourse to a cross-border resolution mechanism, negatively affected stakeholders will have to proceed against each decision of the involved NRAs separatelya general issue in cross-border constellations. The ensuing procedural risks and effort might compromise effective judicial protection against decisions establishing bespoke congestion income regulation, creating further obstacles to interconnector investment.

Role of the Network Codes and the Upcoming Congestion Income Methodology
The judgment also touches upon a current methodological issue of EU energy law: the degree of contribution of the detailed technical legal rules in the network codes to resolving the obstacles to a European IEM. These network codesof which eight have been adopted to dateendeavour to provide harmonised and detailed rules on the operation of European power grids and markets. While the Court included in its scope of scrutiny several provisions not explicitly covered in the referred questionssuch as the annex to the Second Electricity Regulation or the Third Electricity Directive 2009/72and despite emphasising the importance of considering the provisions at the heart of the litigation in their wider context, the ECJ did not consider the network codes at all. Apparently, they were not relevant for the matter at hand.
This provides an example that the network codesan exceedingly complex and comprehensive body of law together encompassing over 600 provisionsstill do not address some of the most contentious barriers to the EU's energy policy goals. 65 On the one hand, this may be due to their purpose, which is limited to providing harmonised rules on preponderantly technical matters. 66 Yet on the other hand, the "one size fits all" approach at the heart of the problems in the Baltic Cable case can also be appreciated in the network codesjust on a different level of detail. Despite their impressive volume, these codes just provide ready-made solutions for standard issues, without regard for special cases, such as the situation of SInCs. As argued above, the lack of a clear-cut framework for congestion income regulation on single interconnectors poses an obstacle to much-needed investment in the interconnection of the European electricity markets.
However, the resulting gap might be closed by the Third Electricity Regulation in the near future, which mandates the creation of a detailed "methodology" on the use of congestion revenues applicable throughout the EU. This methodology is to be developed by the TSOs and to be approved by ACER. 67 While an extensive scrutiny of the TSOs' recently published proposal for this methodology is outside the scope of this note, it is important to recall that the ECJ's reasoning in principle also applies to the revised earmarking regime under the CEP and thus informs this methodology.
The proposal explicitly establishes that the methodology applies "to interconnectors which are owned by TSOs or by other legal entities", ie to regular TSOs and to SInCs. 68 In this context, the proposal appears to follow the same line of reasoning as the Court: the explanatory document accompanying the proposal elaborates on the differences between regular TSOs and SInCs and states that "[i]t is important to recognise these differences". 69 Yet the explanatory document is not binding and merely serves "to further explain the options and main criteria considered by ENTSO-E when drafting" the proposal. 70 In order to further the goal of increasing interconnection and to provide legal certainty for potential investors, the proposed methodology on the use of congestion income could be enhanced to fully reflect the ECJ's statements. In this sense, if the establishment of concrete rules applicable to SInCs is not feasible at present, the methodology could at least clarify that its application to SinCs must ensure that no discrimination in relation to regular TSOs takes place and that the SInCs' financial viability (including a reasonable return) is ensured. ACER has the opportunity to align the final methodology with the ECJ's statements and thus provide sufficient clarity on the regulation of congestion income in the case of SInCs while ensuring no discrimination takes place. This would be a small, but significant step towards the achievement of the EU's energy policy goals.

Final Remarks
In summary, the ECJ's judgment contributes to legal certainty in an important segment of the internal market for electricity. The Court's interpretation comes within a hair's breadth of interpreting the earmarking regime contra legem in order to avoid a violation of the non-discrimination principle. Consequently, NRAs have the option and the obligation to deviate from the strict earmarking regime where these rules are not adequate for the constellation at hand. With numerous investment projects in single interconnectors under consideration, 71 authoritative statements on the business case of SInCs without a formal exemption was urgently required.
The Court's discussion of the earmarking rules highlights that a "one size fits all" approach couched in in widely-worded, general provisions fails to capture all facets of complex economic areas, such as 67 Third Electricity Regulation art.19(4). 68 See "All TSOs' Proposal for Use of Congestion Income Methodology in Accordance with Article 19 (4)  the electricity sector. However, the much more detailed provisions in the European network codes apparently provide no additional guidance on the question at hand, either. This shows that while the network codes provide valuable harmonisation for technical issues, the sensible application of classic fundamental principles of EU law, such as the non-discrimination principle, is still pivotal for resolving modern and central issues of electricity market integration. However, tangible guidance on the application of these broad principles would be required to ensure that said issues are resolved predictably and in line with EU policy aims.
That said, the ECJ's normative mandate to the NRAs when applying the earmarking regime to SInCs is far from clear-cut. While some flexibility is undoubtedly necessary to account for the particularities of each case and interconnector, the ECJ's reasoning leaves the NRAs with a very wide margin of discretion. This makes it unlikely that a "standard" European model for regulating congestion revenue on single interconnectors will emerge anytime soon, despite the significance of such interconnectors for the achievement of the EU's energy policy goals. In this vein, the adoption of a methodology on the use of congestion income under the CEP offers an opportunity for aligning the regulation of congestion revenues for SInCs and for outlining them more precisely. However, so far the proposal only mentions the case of SInCs in a non-binding explanatory document, raising concerns that this opportunity might be missed. ACER, as the entity responsible for approving this methodology, has the opportunity to provide guidance on the proper application of the earmarking regime to regular TSOs and SInCs and to align the content of the methodology with the ECJ's statements. A suitable blueprint for congestion income regulation would facilitate the application and enforcement of the related rules, provide potential investors with the required certainty and thus contribute to the realisation of a European IEM and the achievement of the Union-wide interconnection targets.
We recall the three pillars of effective interconnector investment regulation mentioned in the introduction: unbundling, supranational regulatory oversight and congestion income regulation. The Baltic Cable judgment clarifies the scope of two of these pillars for SInCs without a formal exemption: First, it determines that SInCs are TSOs and thus subject to unbundling. Second, it defines the perimeter for congestion income regulation in such cases. Given that the ECJ's statements leave the NRAs considerable leeway, their impact in practice will depend on the remaining pillar, as ACER is now in the position to flesh out the outline provided by the ECJ. In carrying out that task, ACER will have to consider the fundamental principles of EU and energy law, as well as the EU's policy aims for the electricity sector.