How much does it cost to transmit electricity from a nuclear plant close to Lyon to a car manufacturer located in Stuttgart? What price should a system operator in Scandinavia charge a pulp and paper mill that uses the high voltage grid for only a few weeks per year? To what extent can a trader be assured that there will be transmission capacity available for the first quarter of 2004 at an affordable price between Spain and France? Where is it more profitable in Italy to reinforce or build a new high-voltage line? What is the best place to locate a gas-fired station in England to pay the lowest transmission cost?
Those questions are new and crucial for Europe. They are however difficult to answer. That is why this book has been written.
Grid access: a key issue for the success of the European liberalisation of electricity
The liberalisation of the electricity sector was initiated much later than the opening to competition for other network industries (e.g., airlines). The process started in 1996 with the issuance of Directive 96/92 that had to be implemented into national legislation two years later. It establishes common rules to achieve a gradual introduction of competition in all member states, including the conditions of access to the grid. Those are of utmost importance because the transmission network can no longer be exclusively used by the historical operator. Being an infrastructure too costly to duplicate, competitors must have the right to access to it on a nondiscriminatory basis. A grid tariff has therefore to be set and published. This is a new issue because, under the old system of vertically integrated monopolies, there was no trade of transmission services and therefore no needs to elaborate specific methods for pricing them. A benefit of the opening-up of the electricity sector to competition is to achieve cost-efficiency, that is, that consumers’ demand will be cleared by firms that produce at least cost. For instance, suppose that there is a lowcost generator with 120 MW of capacity and a high-cost generator with 20 MW of capacity. An efficient supply to serve a 130 MW demand requires that the low-cost generator will operate at full capacity whereas the high-cost generator will serve the residual demand by operating at mid capacity. The economic role of the transmission grid appears here. Users are no longer dependent on local power generation. They can purchase electricity from other places if it is cheaper. In other terms, the more open the access to the grid, the fiercer competition is between producers and the lower the price for users. Transmission, however, is not free. As for any other good, an efficient supply of electricity requires the joint minimisation of production costs and transportation costs. In the above example, if the low-cost generator is located far away and if its cost advantage over the local highcost producer is lower than the cost of transport, the efficient supply will be achieved with the high-cost generator operating at full capacity. Here is the key role of the grid tariff for the success of liberalisation: if transmission costs are over-estimated, more efficient producers will not be able to compete with incumbents on their local markets. The attempt to liberalise electricity will be nipped in the bud. If transmission costs are underestimated, liberalisation will also fail for it will result in facilitating the entry of inefficient producers.
Electricity liberalisation is part of the founding fathers’ vision to abolish internal economic frontiers within Europe. The ultimate objective of the Electricity Directives and Regulations is the creation of an integrated single market. It is not to create fifteen liberalised but largely national electricity markets. The setting of transmission tariffs is a major element to achieve this objective. In the absence of common principles to design tariffs and collective committees to ensure their compatibility, cross-borders exchanges of electricity between Members States are unlikely to develop. If a Scandinavian electricity exporter wants to sell electricity to Baden- Wurtemberg and has to pay for the use of the grid in Sweden, in Denmark and in Germany, it will no longer be competitive. Such pancaking of transmission tariffs would be a serious obstacle to achieve the internal electricity market.
Transmission pricing: a non trivial problem
Setting the price to use an infrastructure that is too costly to duplicate is an old and well-known theoretical problem in economics. It was at first addressed one hundred and fifty years ago by a French economist, Jules Dupuit (1849), for calculating bridge tolls. In economic jargon, the conundrum designates the problem of tariff-setting applied to a natural monopoly. It raised a long controversy in the first half of the twentieth century between theorists. Some were advocating pricing at marginal cost in association with a public subsidy whereas others were supporting average cost pricing to make the users pay for the full costs of the infrastructure. Railways rather than bridges were then at the core of the economists’ thinking. After 1950, the optimal pricing analysis of natural monopolies was extended to other infrastructures, such as urban roads, airports or local telecommunication networks. The case of the electricity grid is one of the last to have been studied. In fact, it is only at the end of the 1980’s that the optimal transmission tariff was theoretically solved by Schweppe et al. (1988). Since this seminal step, economic literature on the topic has flourished. It remains, however, difficult to access for non specialists. Papers are dispersed in a large number of different economic and electrical engineering journals. Moreover, their contents are very technical. It is true that the application of the optimal tariff theory to the transmission grid is far from trivial. It is much more difficult than for a bridge or a motorway. This may be intuitively understood in pointing out two differences between road transportation and electricity transmission. Unlike transportation, where the contract path, say for 1 ton of oranges between Valencia and London, actually dictates the movement of the commodity, electric power moves along all the available paths between the supplier and the customer. An injection of current in the Centre of Spain to serve a client in the South of England may use transmission lines in Switzerland. Moreover, like roads, the grid may be congested, however there is a major difference: queues are not possible in electricity with the consequence that the network risks collapse whenever there is a traffic jam.
Contents of the book
The book aims at providing a methodological and practical guide to transmission tariffs for those who are involved in the electricity business as managers, engineers, lawyers, economists, regulators, policy-makers, but are not specialists in electricity transport, nor in tariff-setting for public utilities. It attempts to offer a synthesis of the recent economic research on the subject. The book is divided into three major parts, each presenting one general aspect of transmission pricing: its legal and accounting background, its basic theory, and its implementation, successively.
Part I shows how the economics of transmission tariffs is located between law and accounting. The law sets principles that constrain the choice of tariffs whereas accounting helps economics to compute costs. It opens with an introduction to the juridical framework that features the EU liberalisation process of electricity. Legal guidelines related to access charges are analysed by François Lévêque from an economic perspective. For sure, the law does not associate economic objectives such as cost-efficiency with the setting of transmission tariff. It influences, however, their structure and components. For instance, the legal obligation of non-discrimination impedes the use of non-linear tariffs whereas the costrecovery constraint excludes marginal cost pricing associated with a public subsidy. Accounting is the other companion to economic tariff-setting. The approach of economists is based on the identification of the cost function and the notion of opportunity cost. It is compared by Nicolas Curien to the approach carried out by accountants. Usually, cost assessments via econometrics and via accounting methods are viewed as rival rather than complementary approaches. However, as pointed out by Nicolas Curien, there is a bridge between the cost function and cost accounting methods. A second link between economics and accounting is that they both provide tools to allocate common costs between different business lines. Relying on co-operative game theory, in particular the notions of nucleolus and Shapley value, Nicolas Curien shows how economics casts light on the conflicting objectives the allocation is supposed to achieve. For instance, it is impossible to find a method that both ensures the industry stability and provides incentives for cost reduction. Here is an advantage of the economic approach over accounting methods: the economic approach makes explicit the distributive and equity effects of different rules of cost allocation.
Part II of the volume presents the building blocks of the economic theory of pricing in electricity transmission. It starts with a presentation by Claude Crampes of the so-called nodal prices. Nodal prices are an application to electricity of the general method of marginal-cost pricing in order to implement the socially optimal allocation of goods. They take into account two specific features of electric transport. Firstly, a fraction of the kWh injected into the grid is lost because of energy dissipation. Secondly, at some time periods the line is congested and therefore the use of the grid by some users obliges others to find an alternative (e.g., get electricity at a higher price from a local generator). The cost of losses and the opportunity cost owing to congestion are the two variable parts of the cost to use the grid and therefore the components of the marginal cost price. Because the increasing returns to scale are a feature of the transmission system, as of any other natural monopoly, such a proportional tariff will not balance the budget of the grid operator. If a public subsidy is not allowed by law, nodal tariffs have to be adjusted to cover all the costs, either by adding a mark-up to pay for the fixed cost or by designing two-part tariffs. Because the size of the network is unlikely to be optimal, nodal prices send the right signal to use the grid but not the right signal to invest. Richard Green’s chapter shows how long-run marginal costs can diverge from short-run marginal costs and how investment signals can be sent to grid users based on long-run marginal costs. This is another key role of transmission tariffs. Were such signals lacking, producers might wrongly decide the location of new generators and industrial consumers might inefficiently locate their new plants. As pointed out by Anne Perrot, two other factors require one to adjust and complement a tariff based on nodal prices. Firstly, power generation at the nodes of the network is likely to be dominated by a small number of firms. Facing such an imperfect competition, the regulatory agency may use the transmission tariff to eliminate profit margins due to market power or to facilitate entry. Secondly, the regulatory agency is likely not to know the costs of the grid operator exactly. The tariff must therefore allow a choice between a price-cap oriented scheme that provides incentives to reduce costs and a cost-plus oriented scheme that minimises the rents left to regulated firms.
Part III is devoted to implementation issues. The textbook case, as presented in the previous part, cannot be implemented like magic. As with any piece of theory, the application of normative economics related to transmission pricing requires time, adjustment and modesty. It also requires us to enlarge the picture and consider technical and data constraints. Ignacio Perez Arriaga and Yves Smeers propose on this basis a three-tiered transmission tariff. The first layer, preferably based on nodal prices, reflects variable costs. It aims at sending the right short-term signals to achieve the most efficient use of the grid as it exists. The second and third layers are used to recover the full cost of the system operator. The second layer gathers the remaining elements of costs that may be ascribed to specific uses or activities on the basis of a technical analysis. They may reflect, for instance, costs due to some geographic localisation or time period of use. The ultimate layer is the part of the cost that cannot be allocated in using the cost-causality principle but is necessary to meet the budget constraint. In so far as it cannot be said after several iterations that a particular expenditure is caused by a given use and will benefit known users, Ignacio Perez Arriaga and Yves Smeers recommend to adopt the principle of taxing more the users who are less sensitive to price changes. The study by Jean-Michel Glachant of actual tariffs implemented in different member states shows that there is no resemblance between tariffs as they are prescribed and as they are used. The structures, principles and levels of transmission tariffs largely vary within EU. They vary much more than what would have been expected from taking into account the different size of the national transmission systems.
There is a long way before discrepancies between textbook tariffs and practised tariffs will be reduced. It is expected that the book will contribute to this rapprochement in synthesising economic literature on the subject. The latter is very technical because the matter is. Providing the access to a transmission grid is not as simple as for a bridge. Therefore it would have been irrelevant to write a book without any equations, formulas or curves. The reader would not be able to use and apply for his own sake the ideas summarised in the book if they were not expressed, when necessary, in mathematical terms. On the other hand, formulation with words is essential to explain the concepts and to understand the findings. The equilibrium between maths and English is as follows. Half of the chapters of the volume (Chapters 1, 5, 6 and 7) are written without algebra. The other half rely on mathematical results and use both languages jointly (i.e., hypotheses and results are discussed in words and algebra is used for demonstration) or separately (equations are provided in appendices).