Improving competition in European electricity markets needs more than “Independent” TSOs
June 4th, 2008 by Jean-Michel Glachant, European University InstituteThe publication of the EU Commission proposal for the 3rd legislative energy market package in September 2007 reopened a much debated issue about the pros and cons of different electricity transmission institutional arrangements. Even if one should not minimize the intrusion of political considerations, economic theory and empirical evidence provide useful insights to the debate.
The EU Commission and Member States are mainly debating between two types of arrangements: the Independent Transmission System Operator (ITSO) and the Legally unbundled Transmission System Operator (LTSO). ITSO arrangement corresponds to a complete Ownership Unbundling of transmission assets from generation and retailing companies. It appears to be the first-best option considered by the EU Commission for ITSO ensures the independence of the transmission company which is critical to guarantee a non-discriminatory access to the network. LTSO arrangement does not go as far as ITSO for transmission assets because they are just legally unbundled. Even if this option was not considered by the Commission in the first proposal, now it appears to be a second-best option supported by the “third way” proposal of several Member States (France, Germany, etc.). The LTSO arrangement considered by the Commission includes additional implementation requirements to address the non-discriminatory access issue. Surprisingly, the Commission seems to pay little attention to a third transmission arrangement: the Independent System Operator (ISO). The ISO is responsible for the system operation. It is independent from generation and retail but does not own the transmission assets which could still belong to the vertically integrated company.
When analyzing the rationale for the Commission’s ranking of institutional arrangements, it seems that the Commission is mainly, if not only, worried about non-discriminatory access. This concern seems to overshadow other criteria. Indeed, ranking institutional arrangements requires to take into account several criteria and to rely on empirical evidence to weight them. A review of the economic literature leads us to identify the five following criteria which are the most relevant to assess the relative pros and cons of each arrangement: (i) non-discriminatory access, (ii) benefits from regional integration, (iii) transaction costs saving (iv) performance based regulation implementation and (v) conflict of interest. The table below presents our view on the relative advantage of each arrangement depending upon the considered criterion.
|
|
ITSOs |
LTSOs |
ISOs |
|
Non-discriminatory access |
+ |
- |
+ |
|
Benefits of market integration |
- |
- |
+ |
|
Transaction costs savings |
+ |
+ |
- |
|
Performance based regulation implementation |
+ |
+ |
- |
|
Conflict of interests |
- |
- |
+ |
Table 1. comparing transmission arrangements
The ranking of these arrangements ultimately depends on the relative weight of each criterion. In this post, we focus on benefits from market integration, even if we are aware that other criteria should also be taken into account. This criterion represents the ability of different institutional arrangements to capture the benefit from markets and networks regional integration. Coordinating regional interconnected power systems generates benefits through the increase in cross-border competition and the internalization of cross-border externalities. We believe that these benefits could be especially valuable in the case of continental Europe which is characterized by a highly meshed network, critical cross-border externalities and high potential for cross-border competition in generation.
Thus, choosing an arrangement that will facilitate regional integration is certainly a major stake of the EU energy policy. Regarding this criterion, ISO appears to do better than ITSO and LTSO. The economic literature indeed explains that the integration of the transmission owner and the system operator may act as a brake to regional coordination (Glachant & Rious 2007, Smeers 2007). ITSOs and LTSOs seem to be more difficult to merge and to coordinate than ISOs. The reason is twofold:
(1) National incentive regulation applied to ITSOs/LTSOs tends to reduce their interest in coordinating with neighboring organizations (Glachant & Pignon 2006, Smeers 2007). Glachant and Pignon (2006) have shown that a TSO which network is connected with an adjacent TSO can distort exchange of information and reduce coordination in order to increase its profit. This raises several difficulties for regional coordination. The problem is worsened when a weak regional multilayer regulatory framework is in place and regulatory power lacks to put national and regional incentives in line. Conversely, ISOs are neutral to coordinate together because they have no incentive to reduce national transmission costs which are pass-through and because they follow some management protocols that can be more easily adapted to integrate the regional perspective.
(2) ISO arrangements facilitate mergers of System Operators to get most of the regional coordination benefits. The ownership of national transmission assets is frequently blocked in the hands of national owners by their respective governments. This complicates a merger between ITSOs or LTSOs because it would imply national transmission assets to be transferred to a multinational (or foreign) owner.
This analytical result is reinforced by empirical evidence:
The introduction of BETTA (British Electricity Transmission and Trading Arrangements) in 2005 considerably improved regional performances. At that time National Grid, already the ITSO of England & Wales, became the ISO in Scotland and the Regional ISO for the Great-Britain. On the one hand, regional coordination increased competition among generators. On the other hand, the introduction of locational signals consistent with Great-Britain wide tariffs generated efficiencies. This is a good example of the relative easiness of enlarging ISO functions to larger geographical zones which lead to significant benefits. The role of two key factors should not be minimized in this success: (i) Great Britain is an island and thus, cross-border issues are not a big deal; (ii) a strong and unique regulator covers the whole zone.
PJM, the ISO of several states in the Northeast of the USA, is another success story. It succeeded in expanding its control over larger zones and voluntary coordinated with neighboring ISOs. While PJM controls a meshed network zone with important seams effects (e.g., loop flows, etc.), and faces a relative high market power of generators, its expansion over larger zones and its coordination with neighboring ISOs (e.g., Midwest ISO) generate substantial benefits (Krapels and Flemming 2005, Glachant & al. 2005). However an initial lack of transmission investments by the transmission owners (T.O.) had to be fixed by the federal regulator, FERC, giving more power to the ISO in the planning process.
Conversely, the Nordic region is characterized by a regional interconnected power system managed by national ITSOs (Léauthier & Thelen 2008). Even if an association (Nordel) has been created to organize a voluntary cooperation of national ITSOs, evidence tends to suggest that regional integration is not adequately promoted. Indeed, investments appear to be mainly decided on a national basis and lack of an effective mechanism to solve cross-border flows costs. Moreover, each ITSO is incentivized by a national regulator which tends to exacerbate the situation.
To conclude, coordinating national ITSOs to generate regional integration benefits appears to be more difficult than coordinating ISOs. This is even more true in the absence of a strong regional regulator and without strong and credible institutions ensuring cross-border coordination (Glachant & Lévêque 2007). We believe that market integration is a key issue for the future of an interconnected EU power system. Moreover, the implementation of a strong EU wide regulatory framework is certainly a challenging issue. Benefits from market integration could then be facilitated by ISOs which also efficiently ensure non-discrimination access.
Jean-Michel Glachant (Université Paris-Sud 11), François Lévêque (Mines ParisTech) and Marcelo Saguan (Université Paris-Sud 11).
June 4th, 2008 at 11:15 am
The arguments above are very interesting and present a more persuasive case for ISOs in the EU context than I have seen elsewhere. However, there have also been bad experiences with ISOs, most obviously in California.
One issue that arises with ISOs that does not arise with ITSOs or LTSOs is their governance. Establishing effective governance arrangements that ensures adequate investment and operational efficiency can be very difficult. My understanding of the Californian ISO problems is that the board controlling the ISO included representatives of all the players in the industry. In consequence, it was a large, unwieldy body which was continually needing to find ‘political’ compromises. That situation could also well be a risk in EU countries, especially those with publicly owned transmission assets.
Governance issues seem much less of a problem with ISOs. Hence, in practice, ISOs may not be such a great idea for countries with traditions of significant reliance non-market arrangements in public services, as is the case with some EU member states. I think that the authors are right to point out the role of the regulator and the absence of cross-country issues for the apparent success of BETTA – but also, the new UK system seems to have avoided the governance problems that plagued the Californian ISO and might be a concern in some other EU countries.
LTSOs seem to avoid teh governance problems of ISOs, but look to have few other advantages – other than perhaps a transition to full ITSO. Are there any examples that we might learn from? If not, that in itself might suggest that they are not very attractive institutional options.
Note that similar problems of an unwieldy corporate board issues and corporate governance have arisen in practice in the UK with Network Rail, which, in practice, could be considered as a half way house between an ITSO and a (strong) ISO.
Jon Stern
June 4th, 2008 at 4:06 pm
In principle I agree to the arguments made above – although I do not see the major advantage of LTSO in relation to transaction cost savings – IF legal unbundling is properly implemented. Transactions in such a framework are pretty the same is in an ISO model.
Still I would like to make some general comments on the ISO model(s):
1) The ISO model faces two major challenges, how to construct incentives for the ISO, if operation and equity is separated and how to establish a system of liability for major damages caused by the ISO. I think the “minus” for performance based regulation in the table might reflect these problems.
2) The argument that ISO is a better model for regional co-operation, needs a more thorough discussion. In Glachant&Pignon 2006 possible distortions of incentives under special conditions have been shown for the ITSO model. The question is whether this argument holds for all ITSOs in general. If a “for profit company” is able to increase profits by congestions, we might ask whether the appropriate remedy is to set up a “non for profit company” such as an ISO or to change the regulation, that is, that a TSO should not be able to get extra money for congestions. The other problem which is supposed to be solved by the ISO is regulatory failure across borders, as NRAs have national objectives such as to decrease transmission tariffs, which, today, can be achieved by high congestion revenues, which are by the end of the day paid by customers in the high price area. Again the question is, whether the structural decision in favour of an ISO is the appropriate means to overcome this shortcoming. In my view, there is a good argument, that we have to correct the existing distorted incentives for TSOs as well as for NRAs. If we set up a system, where the congestion revenues are not benefitting the accounts of the TSO nor reducing the national transmission tariffs, which in itself isn’t really intuitive, a system of ITSO should have quite good prospects to deliver.
3) The valid argument is the one put forward under (2), that ISOs will facilitate mergers and that one European SO is preferable to a cooperative model. This is certainly quite appealing as long as the European ISO is clearly defined. The European ISO would have to be able to operate the system, to manage congestions, to invest and to maintain the grid or at least decide on equity issues without facing major governance problems. Although merging equity is certainly more difficult than merging functions, a realistic development of an ITSO model already entails cross border investment, be that by investment funds, grid companies or some other investors. It would be interesting to analyse what such a development might lead to.
June 5th, 2008 at 9:43 am
I couldn´t agree more. In the Eurelectric RIO model all the benefits Glachant is discussing in the paper can be found. The goal is of course market integration if we ever are going to see a European market. Then any national TSO solution will only be half full our half empty. The reason for the unwillingness to face this real problem are different for the different involved interests. The TSO position is simple, who wants to be forced to be reorganized and maybe loose his influence. The governments position is more difficult, they believe that they have control over security of supply by having the TSO under national control. In a meched network of course a blackout can start in an other network and spread regardless of any government control. Governments also tend to believe that TSOs can be used to support renewables but support systems becomes more stable if they involve a larger area. Finally governments can see it important where the head office of a RIO is located, forgetting that we here are not talking of a workplace with thousands of people. One important difficulty to set up a RIO is that it needs a Regulation covering the same geographical area. To this I would say , What kind of regulation would we need if we shall achieve the same result without a Regional System Operator? Do anyone believe that that would be a very simple regulation? So Regulators shall not ne afraid of regional regulation, they will not loose anything compared with today because that work doesn´t exist today. On the other hand implementing regional regulation would be an enlargement of the tasks for regulators.
Governments in a region, for example the nordic, should develop a regional regulation with the starting point in the EU Crossborder Regulation, implement it in national law, agree on a regional regulatory function to execute the regulation, set up a Regional System Operator that have all responsibilities including grid planning and investment decissions ( to avoid the negative side of PJM ), the RIO should also be the owner of the PX-spot, transmission assets will stay in todays TSOs that becomes TOs.
If the third pachage now fails because of wrong focus in the unbundling debate I believe that we should start to prepare a development in this direction so that in 2010 when the time is there for a new pachage we can focus on the right things to improve the European Market.
June 16th, 2008 at 3:23 pm
Jean-Michel et al bring useful clarity to a sometimes confused debate. The contribution is especially relevant given the June 6 decision at the Energy Council to (i) adopt a compromise on the Third Package, involving all three solutions (ITSOs, ISOs and LTSOs); and (ii) review the performance of the LTSO (which in effect will require a comparative review of the performance of the different models adopted across the Union) within a couple of years of implementation. The set of criteria that Jean-Michel has specified would be a good starting point for that review.
I and my colleagues have recently undertaken analysis (for Endesa SA) of the performance of different arrangements for transmission governance, focusing on evidence from international experience. Our work shows inter alia that:
1. There is indeed a “conflict of interest” issue, i.e., in a liberalised market the TSO has an incentive to use its role as System Operator (SO) to increase its profits as Transmission Owner (TO), for example by undertaking expensive balancing actions (cost typically passed through to system users) rather than undertaking new investments (where it would typically bear part of the cost). This issue can be resolved by having a non-profit TSO (although this raises issues of effective incentivisation), or through a combination of incentive schemes and detailed rules, as in Great Britain. The latter approach requires however a strong and independent regulator. The ISO model may provide another solution, especially if the regulator lacks sufficient independence or powers.
2. Evidence from the US does not support the claim that “investment is the Achilles’ heel of the ISO model”. In some parts of the US utilities in ISO regions have invested more than those outside ISOs, in other parts less. Moreover the picture varies across time, and depends on the starting point. There is no simple pattern.
3. Although the introduction of an ISO brings new costs associated with co-ordinating the activities of the SO and TO, these costs do not appear material enough to be determinative in choosing between different systems.
4. Voluntary or quasi-voluntary co-ordination between national TSOs (as foreseen in the original Third Package) does not appear to be effective in addressing barriers to regional integration, when the solution would create winners and losers (actual or perceived) across national boundaries.
Our work is summarised in two papers:
1. “Regulating Unbundled TSOs: Rules, Incentives or an ISO?” (The Brattle Group, 2007) available at http://www.brattle.com/_documents/uploadlibrary/upload688.pdf
2. “Independent System Operators for Power Transmission: Evidence-Based Assessment” (The Brattle Group, 2008) available at http://www.brattle.com/_documents/uploadlibrary/upload687.pdf