Decentralised Cross-Border Interconnection.

• Author: Crampes, Claude

1. INTRODUCTION

In Europe, as well as in most other parts of the world, cross-border interconnection is typically based on bilateral agreements between the operators of the systems linked by the interconnector. While such agreements may cover both the design of the interconnector and the sharing of its costs, they generally do not extend to the reinforcements in domestic transmission systems that would be warranted to achieve the full benefits of interconnection. As a result, such projects tend to be suboptimal, or they are not undertaken at all.

An example from France and Spain may illustrate the difficulties. In 2008, the electricity transmission and system operators of France and Spain, Reseau de Transport d'Electricite (RTE) and Red Electrica de Espana (REE), created Inelfe, a corporation jointly-owned in equal shares, with the aim of constructing a new electrical interconnection through the Eastern Pyrenees that would effectively double the exchange capacity from 1,400 MW to 2,800 MW. (1) In a report published by the French regulator in November 2015, (2) shortly after the line was inaugurated, it appeared that the commercial capacity effectively made available to the market could not reach the level initially expected:

'In the Spain-to-France direction, the delay in the installation of a phase-shifting transformer in Spain limits capacity that can be allocated to the market to 2,300 MW. This equipment is set to be put into operation in 2017. Moreover, the interconnection capacity effectively available in both directions is currently limited by constraints in the Spanish domestic network. In particular, due to problems with local acceptance, the construction of two separate lines downstream of the new link did not go as scheduled, with a portion of the route finally being built with one line only. As a consequence, this part of the route is the cause of stricter capacity limits, in compliance with Spanish operating rules. Interconnection capacity between France and Spain is therefore limited to an average of 2,000 MW in both directions, for the greater part of the year 2016.' As a result, the benefits of the interconnection were reduced. (3)

The difficulties arising from decentralised decision-making in an integrated network have not gone unnoticed. In Europe, the Agency for Cooperation of Energy Regulators (ACER) was set up in 2010 as an Agency of the European Union by the Third Energy Package to further the completion of the internal energy market both for electricity and natural gas (European Council, 2009a); its aims include 'an efficient energy infrastructure guaranteeing the free movement of energy across borders and the transportation of new energy sources, thus enhancing security of supply for EU businesses and consumers' (www.acer.europa.eu). European transmission system operators cooperate in the European Network for Transmission Operators for Electricity (ENTSO-E) and European Network for Transmission Operators for Gas (ENTSO-G); among their tasks is to produce Ten-Year Network Development Plans (TYNDPs) to provide a consistent view of the pan-European infrastructure and signal potential gaps in future investment--these plans form the basis for the European Commission's selection of so-called Projects of Common Interest. In the 2016 Winter Package (European Commission, 2016), the European Commission foresaw the establishment of regional entities which would take over functions and responsibilities from national transmission system operators. (4) Nevertheless, even though much has happend to coordinate decisions on energy infrastructure in Europe, it is still the case that, within their jurisdictions, national regulators and system operators have discretion. (5)

From a purely technical point of view, building a new line between two network nodes causes costs and benefits that do not depend on in which jurisdiction nodes are located. The basic economic models of electricity transmission developed for building and operating domestic lines may therefore be applied to the study of interconnectors. (6) Interconnectors generate revenue based on price arbitrage between nodes. If the price differential between two nodes is sufficiently large, the discounted revenue stream is larger than the cost of building and operating a connecting line, and private investors would be willing to bid for the right to install a new link between these nodes. However, when the two nodes are in different jurisdictions, they are typically subject to different sets of rules and controlled by decision-makers with potentially divergent interests. It is this heterogeneity that makes the economics of interconnectors different. For example, depending on whether markets on the two sides of a border are coupled or related through a system of coordinated auctions, the way to manage cross-border trade may be different, and so is the (private) value of an electric link. (7) The prospects and problems of transmission investment also vary depending on whether it is purely merchant or under tight regulation. (8) Similarly, the organization and regulation of the markets at the two ends of the line have an effect on the incentives to reduce congestion costs. (9)

In the literature on the economics of energy markets, there is a variety of works related to interconnectors. Keppler and Meunier (2018) use cost-benefit analysis to determine the socially optimal increase in interconnection capacity. Hoffler and Wittmann (2007) investigate capacity auctioning. Newbery and Grubb (2015) defend the idea that the contribution from interconnectors should be included to determine the amount to procure in capacity mechanisms and Hagspiel et al (2018) consider the role of interconnectors for reliability assessments. Turvey (2006) explains why the utilisation of some interconnectors is sub-optimal. Debia et al. (2018) and Massol and Banal-Estanol (2018) analyse the impact of market power on the use of electricity and gas interconnectors. On the regulation side, Mountain and Carstairs (2018) explain why self-assessed proposals by transmission companies for interconnector development do not provide appropriate incentives. None of these papers explicitly takes externalities from domestic investment and the related regulatory issues into consideration.

The issue is addressed in some case studies. In de Jong et al. (2007), one finds three case studies of European interconnector investment: NorNed (between Norway and The Netherlands), Estlink (between Estonia and Finland) and BritNed (between United Kingdom and The Netherlands). The authors describe the regulatory assessments of the three interconnector projects. At that time, ACER did not exist so that only pairs of national regulators were involved. Crampes and Rives (2011) analyse the hierarchical regulatory structure created by the Third Energy Package through a study of the powers attributed to each actor and a modeling of the actors' relationships. (10) Both national and European regulators scrutinize transmission system operators' activities and each organization has powers that affect the transmission system operators' decisions on interconnection. The main conclusion of Crampes and Rives is that it is always optimal to decentralise part or all of the provision of incentive policies. The authors also consider the possibility of mergers between national transmission system operators and the subsequent likely development of international transmission system operators with stakes in several countries under separated regulation mechanisms, discussing how the regulatory structure should evolve and how the relationships between an international transmission system operator and its regulator(s) could be altered. (11)

In this paper, we abstract from many technical and institutional details considered in previous studies and concentrate instead on the interaction between cross-border interconnectors and national infrastructure, a topic that has so far received relatively little attention in the literature on networks and interconnectors. We demonstrate that such interaction inevitably creates inefficiencies, even when the countries involved are able to reach an efficient agreement on interconnection; so long as investments in national infrastructure are not coordinated, neither interconnector capacity nor domestic capacities are optimal. For this reason support to interconnectors--along the lines currently being followed in Europe--cannot restore optimality; indeed, under reasonable assumptions such support should be restricted, in order not to encourage the building of interconnectors that will not be efficiently utilised.

Our analysis is closely related to the literature on local provision of public goods, starting with Williams (1966). (12) A recent contribution to this literature is Bloch and Zenginobuz (2007), who consider the impact of spillovers on the supply of public goods in a non-cooperative game between different governments in which spillovers may be symmetric or asymmetric and jurisdictions may differ in size; they conclude that the complexity of interactions will plague the design of institutions for multijurisdictional local public good economies with spillovers. A distinguishing feature of our model is that we assume that governments may be able to reach an efficient solution for the public good itself (the interconnector), but that overall optimality is not achieved because of the interaction (spillovers) between local networks and the interconnector. We point out that this result implies that moving the decision on the interconnector to a supranational level does not solve the problem, unless that authority can also control national investments (either directly, or indirectly through financial transfers). We also consider the possibility that the public good (interconnector) is provided by a third party (merchant line) and that national transmission operators have ownership...