The European Commission’s proposal is supposed to deliver more cross-border electricity transmission.It is an extension of the current system of national-welfare centred regulations, a system which does not target the optimisation of the EU electricity network, and as such is inconsistent with a truly single market. However, the integrated first-best solution – a single European system operator, regulated by a single regulator, which develops the network in coordination with generators and consumers – appears politically infeasible. To overcome this, we propose a bold blueprint for a European system to fund and incentivise infrastructure development. The approach is fourfold: (1) implement vertical unbundling; (2) add a European system-management layer; (3) establish a stringent planning process; and (4) phase-in European cost-sharing.
Implement vertical unbundling
Léautier and Thelen (2009) find that vertical separation is one key-requirement (the other being a well-designed incentive scheme) for reducing network congestion. It is important that transmission system operators should not be concerned with the interests of affiliated generators. The legal basis for this has already been adopted in the third EU energy sector liberalisation package of 2009. Implementation of the unbundling requirements should have been done by 3 March 2012. The European Commission acknowledges that in most member states the unbundling provisions are not yet fully transposed.
Add a European system management layer
National system operation has major spillovers onto neighbouring countries, but also affects network investment incentives. Uncoordinated system operation increases the incentives for
national operators to close their borders in order to ensure system stability. The straightforward way of escaping this dilemma is to add a European system management layer, in other words, centralising and monitoring electricity system information in real-time. This would enable throughput of electricity through national and international lines to be safely increased without any major investments in infrastructure. This would neither require TSOs to merge or to be expropriated, nor would it substantively infringe on national sovereignty over the security of national electricity systems. A European control centre would complement national operation centres and help them to better exchange information on the status of the system, on expected changes and on planned modifications. The ultimate aim should be to transfer the day-to-day responsibility for the safe operation of the system to the European control centre.
To further increase efficiency, electricity prices should be allowed to differ between all network points across and within countries. That is, electricity in Hamburg might be cheaper than in Munich on the wholesale market if there is a lot of wind in the North Sea, while the sun is not shining on Bavarian solar panels. This would provide the correct incentives for switching off coal-fired power plants in the north and switching on gas turbines in the south in order not to overcharge the network. In addition, investors in generation (or load) will base their location decisions on these locational price signals. This will reduce congestion over time, by creating an incentive for generation/load to move to net electricity
Establish a stringent planning process
Current approaches to network planning suffer from a number of shortcomings: they essentially
reflect the interests of TSOs, which make planning decisions without full information about crossborder impacts; the plans are non-binding, meaning stakeholders are not obliged to comply, and so do not provide the necessary synchronisation of investments in the energy system; the planning process is non-transparent as far as the modelling is concerned; and the planning process is ‘technocratic’ in the sense that it does not a priori take
the concerns of residents into account. Some of these issues have been addressed effectively in
Harmonising national network planning rules is administratively difficult and would take many
years. To avoid this, the European approach is to use the ten-year network development plan
(TYNDP) to ensure the consistency of the results of national planning with European objectives. To achieve this, ACER must provide opinions on the consistency of the individual national ten-year plans with the TYNDP. However, the consistency of national plans with European objectives cannot be enforced by ACER or any EU institution (Commission, Parliament and Council) – Regulation 714/2009 explicitly refers to the “non-binding Community-wide ten-year network development plan”. Hence, to safeguard consistency of individual national network plans and to ensure that they contribute to provide the infrastructure for a functioning single market the role of the TYNDP needs to be upgraded. This could be enacted by obliging national regulators to only approve projects proposed by European planning unless they can prove that deviations are beneficial.
The boosted role of the TYNDP that this would entail would need to be underpinned by resolving
the issues of conflicting interests and information asymmetry. Two approaches to this are conceivable:first, relying on thorough cross-checking of ENTSO-E proposals by the regulator, or, second,shifting the entire planning process to an independent body.
In the first case, ACER should be requested and authorised to thoroughly check that the TYNDP
maximises the welfare of current and future European citizens and that national plans are consistent with the TYNDP. This implies that ACER would not only rely on the modelling results that TSOs use to justify their plans, but would have tools of its own for impartial evaluations. ACER should not resort to consulting proprietary models that are not fully disclosed and that have to be repeatedly procured. Instead, ACER – or another public body – should invest in the capabilities to build, manage and use a European open-source energy
model. Based on a substantial upfront investment in a suitable model, ACER would structure a
process in which all relevant stakeholders can support ACER by updating the assumptions and
the modelling. Individual stakeholders will still have better information on their parts of the electricity system. TSOs will know the network better than any independent network modeller, generators will have a clearer view of their individual plans, large consumers (including distribution system operators) will have more information on their future load, and residents will best be able to evaluate the acceptability of proposed lines. Thereby, ACER’s power to approve the TYNDP based on its own modelling results would shift the burden of proof to the stakeholders (including ENTSO-E) in case they disagree with ACER’s conclusions. This would give the stakeholders an incentive to disclose private information. In addition, the open-source nature of the model would allow inconsistencies to be identified, and improvements to be proposed. Of course, state-ofthe-art could only be ensured by continued
investment in the model’s capabilities.
In the second case, resolving the issues of conflicting interests and information asymmetry in network planning could also be achieved by building on the significant effort that ENTSO-E has made in developing the TYNDPs. Using the TYNDP expertise would require that its governance structure be made independent from the interests of TSOs. Hence, a dedicated TYNDP governance structure should be developed that is representative of all electricity sector stakeholders (in a membership committee). An executive board that is independentfrom industry interest should have full operational control. Finally, the by-laws of the institution governing the TYNDP would need to ensure that the model used for planning is made fully transparent and open source. Irrespective of the model chosen (‘cross-checking’ or ‘independent planning’) it is essential to make both stakeholders’ inputs and the final plan binding in order to improve the synchronisation of investment. That is, stakeholders which, for strategic or other reasons, deviate ex post from their predictions (eg building a power plant or consuming electricity at a certain point of the network) will be liable to claims for damages from other stakeholders.
Finally, planning will not be able to make all stakeholders equally happy. And certain choices that do not affect overall welfare might have substantial redistributive effects. To rectify the distributional consequences, an ultimate political decision by the European Parliament on the entire plan could open a negotiation process around selecting alternatives and agreeing compensation. This need for democratic approval ensures that all stakeholders have an interest in ensuring a maximum degree of balance of interest in the earlier stages. In fact, transparent planning, early stakeholder involvement and democratic legitimisation are well suited for minimising as much as possible local opposition to new lines.
The delivery of the plan would then be left to the TSOs or any other investor willing to deliver individual lines according to the regulated conditions. In case of multiple interests, the national regulator might choose the best value offer.
Phase in European cost-benefit sharing
A critical element in the discussions about EU electricity networks is cost and benefit sharing. Different stakeholders have diverging interests, and it sounds unnatural to require stakeholders to pay for a transmission line that actually reduces their profits. On the other hand, stakeholders that are the major beneficiaries of a new line should not be able to pass all the cost onto society.
Hence, all easily attributable cost should be levied on the responsible party. If new generation requires grid reinforcements, the reinforcements should be largely paid for by the generator. In this way, the investor has the right incentives to tradeoff high locational prices in one place (eg close to consumption centres), with cheap network access in another place (eg in a zone where an old power plant has recently been shut down), and good access to resources in a third place (eg for a wind turbine, a zone with high constant wind).
For all remaining network extension the question is how to share the cost between network users in different regions. Having all line extensions in Sweden being equally financed by Bulgarian network users seems difficult. Having a Belgian line that is required to accommodate loop-flows caused by inner-German imbalances being paid for only by Belgian network users is not reasonable either.
Based on the assumption that the outlined network development plan delivers an efficient
proposal, and that new generators have to pay deep connection charges, we suggest that some
redistribution is unavoidable. The reason is that, so far, even the most sophisticated cost-benefit analysis models have been unable to identify the individual long-term net benefit in an uncertain environment. For all infrastructure (eg rail and road) there is some socialisation of the costs of individual projects within the different regions of a country. Hence, we propose that consumers in all nodes that are predicted to receive more imports through a line extension should be obliged to pay a certain share (eg half) of the line extension through their network charges, while the rest of the cost is socialised to all consumers. Such a cost-distribution scheme will involve some intra-European redistribution from the (infrastructure-wise) well-developed countries to the laggards. However, such a scheme would perform this redistribution in a much more efficient way than ad-hoc disbursements by the Connecting Europe Facility to politically chosen projects, because it would provide the infrastructure that is really needed.
Implementation of this proposal will deliver the infrastructure needed to achieve the European
energy policy targets in the field of electricity. It will increase the reliability of the network, enable a truly borderless European electricity market, and facilitate the integration of renewables. If the EU decides to wait for the results of the nonbinding plan to materialise in the 2020s, valuable time will have been lost. All approaches involving throwing money at the problem to achieve flagship projects will fail to resolve the complex underlying issues. After three energy sector packages and 20 years of work, the EU possesses many of the key institutions and laws necessary for achieving the single electricity market. In the past, the benefits of a more coordinated system have not been great enough to outweigh the significant political and transaction costs required to achieve such a system. However, recent developments (unbundling, renewables, more trade) have substantially increased the value of greater coordination. Thus, it is the right time for the EU to take a bold step towards a borderless electricity infrastructure.
Georg Zachmann, Research Fellow, Bruegel
P.S. A longer version of this paper has been originally posted at www.bruegel.org