ناتوانى بازار جفت شدگی بازار و ضد تجارت در اروپا: بحث مبتنی بر مدل گویا
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|16574||2013||14 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Energy Economics, Volume 35, January 2013, Pages 74–87
The horizontal integration of the energy market and the organization of transmission services remain two open issues in the restructured European electricity sector. The coupling of the French, Belgian and Dutch electricity markets (the trilateral market) in November 2006 was a real success. The extension of the system to Germany in November 2010 also proceeded smoothly and the intent is to continue with the same market architecture. But Market Coupling is based on a zonal system which has often failed in meshed grids. This may cast doubts on what will happen in the future when electricity demand picks up again and wind develops. The nodal system has generally been more successful than zonal architectures but its implementation is not currently foreseen in the EU. This paper analyzes versions of Market Coupling that differ by the organization of counter-trading. While underplayed in current discussions, counter-trading could become a key element of Market Coupling as its geographic coverage expands and wind penetrates. We simplify matters by assuming away strategic behavior between the energy and counter-trading markets and conduct the analysis on a stylized six node example taken from the literature. We simulate Market Coupling for different assumptions of zonal decomposition and coordination of Transmission System Operators (TSOs). We show that these assumptions matter: even in the absence of strategic behavior, Market Coupling can be quite vulnerable to the particular situation on hand; counter-trading can work well or completely fail and it is not clear beforehand what will prevail. Our analysis relies on standard economic notions such as social welfare and Generalized Nash equilibrium, but the use of these notions is probably novel. The nodal organization is the reference first best scenario: different zonal decompositions and degrees of coordination are then studied with respect to this first best solution.
Congestion management remains a controversial issue in the restructured European electricity sector. Congestion occurs when the infrastructure constrains transactions in the energy market. Grid congestion evolves with short and long term shifts in generation and consumption patterns and with the development of the grid. Methods of congestion management can be characterized by the integration of energy and transmission that they impose. This determines the extent to which one accounts for the possibilities of the grid when clearing the energy market. Nodal Pricing (see Hogan, 1995 and Hogan, 1998) controls the energy and transmission markets through a single entity and is thus the paradigm of the full integration of these two functions; this results in electricity prices that directly include congestion costs. From a technical point of view the integration of energy and transmission is achieved by solving a welfare maximization problem that involves both functions. With reference to European internal market discussions, the nodal system is the perfect implementation of the “implicit auction” that is becoming a reference option in European Cross Border trade (Article 12 paragraph 2 of Regulation No 714/2009 (European Commission, 2009). Nodal Pricing has now been implemented with success in many regions of the US and in New Zealand1 (see Frontier Economics, 2009, Joskow, 2008 and Sioshansi and Pfaffenberger, 2006). Other architectures separate energy and transmission markets, with this separation taking different forms. Market Coupling considers both an energy market operated by Power Exchanges (PXs) and a transmission system controlled by Transmission System Operators (TSOs). The energy market is subdivided into price zones operated by different PXs. The market clears taking into account limited “transfer capacities” (TC) between zones. TSOs maintain the security of the different control areas of the grid and provide the PXs with the TC linking the price zones. It is also the duty of the TSOs to guarantee that the grid can effectively accommodate the transactions resulting from the clearing of the energy market by the PXs. This organization does not fully integrate the energy and transmission functions, but it creates explicit links between them. The energy market clears on the basis of TCs provided by the TSOs and the resulting power injections and withdrawals are communicated back to TSOs. If the transmission capacities provided by TSOs are small enough compared to the real possibilities of the grid, the energy market does not entail congestion. If not, grid lines are overloaded and TSOs have to restore grid feasibility by reshuffling power flows among energy market participants (producers/consumers). This is the so called counter-trading or re-dispatching function: it relies on increases and decreases of injections/withdrawals in order to restore grid feasibility. TSOs remunerate generators and consumers for these services and socialize these costs that are then charged back to the agents connected to the grid. Market Coupling realizes a spatial arbitrage between different zones of the energy market. Even though electricity is not economically storable today except in hydro reservoirs, the characteristics of the machines also pose problems of inter-temporal arbitrage that will become increasingly important with the penetration of wind. This paper concentrates on spatial arbitrage and leave inter-temporal arbitrage for further research. The following elaborates the description of Market Coupling on the basis of Fig. 1. Consider two markets North (N) and South (S) with supply and demand bids in each of them. Assume that there are two generators in N. Denote them as “gen1” and “gen2”. Gen1 controls a 100 MWh 2 plant whose marginal cost is 5 €/MWh, while gen2 can at maximum run 400 MWh at a marginal cost of 20 €/MWh. Demand in the North is 200 MWh. There are two generators “gen3” and “gen4” in market S. Both generators have an available capacity of 300 MW, but gen3 has a marginal cost of 20 €/MWh, while gen4 operates at a marginal cost of 30 €/MWh. The electricity demand in S amounts to 600 MWh. Note that generation in N is cheaper than in S. Consider first the equilibrium in each market taken in isolation as depicted in Fig. 1. The equilibrium price in market S is higher than the equilibrium price in market N as shown in Fig. 1. This creates an arbitrage opportunity between the two markets: electricity should move from N to S. Suppose now that the two markets are linked by a Transfer Capacity (TC) as depicted in Fig. 2. If this transfer capacity is large enough the usual arbitrage reasoning will imply a flow between the N and S markets that equalizes the prices in both markets: this is shown on Fig. 2a. If this transfer capacity is limited, the arbitrage will be limited to what the TC allows (in our case 250 MWh, as illustrated in Fig. 2b) and the electricity prices will be zonal. This principle can be applied to more complex systems such as depicted in Fig. 3. Fig. 3a is a stylized version of the trilateral Market Coupling including France (F), Belgium (B) and the Netherlands (NL) that went live on 21 November 2006. Fig. 3b represents the extension of that market after the coupling with Germany (G) on 9 November 2010 (referred to as the pentalateral market as Luxembourg has part of its system integrated both with Germany and Belgium3). Leaving aside issues related to the representation of the characteristics of the generators such as block bids the questions raised by Market Coupling boil down to the definition of the price zones and the determination of the TCs linking them and more generally to whether electricity transmission can really be represented by TCs between zones. The relevance and importance of these questions is well acknowledged. Hogan (2005) quotes the US federal electricity regulator stating that transfer capacities are artificial constructs without economic or physical reality that were inherited from the regulatory period when trade was not a key concern. Less assertively, European Transmission System Operators cautiously advise that they do not guarantee the validity of the transmission capacities that they publish. Last but not least, one can observe that TSOs have postponed the publication of their method for computing transmission capacities for several years. At least we know some of the principles that they use (see Rious et al., 2008, for a detailed explanation) and will invoke them later in the discussion. This paper addresses these questions on an example. It is organized as follows. Section 2 gives a brief history of Market Coupling. Section 3 presents a test problem that consists of two configurations of a two-zone system such as those commonly found in the papers of the Transmission System Operators (see Entso papers on ENTSO-E website4). A very brief Section 4 restates the nodal model. Section 5 expands the discussion on Market Coupling and describes different possible organizations of counter-trading to relieve congestion when it occurs. Numerical results of the Nodal Pricing model and the Market Coupling and counter-trading problems applied to these two different market configurations are discussed in 7 and 8 and 9. Finally, Section 10 summarizes the finding of the analysis.
نتیجه گیری انگلیسی
This paper only offers an illustrative analysis, but the observed phenomena may usefully be studied on real networks. The discussion is conducted on a simple six node example taken from the literature. The market is split into a two zone system of which we consider two versions. Both take the form of a North–South decomposition to which we refer as the (3–3) and (4–2) configurations (the first and second figures are respectively the number of nodes in the “Northern” and “Southern” zones). We then look at different organizations of cross border trade in these two configurations. Even with this simple set up we find diverse and interesting phenomena that are also valid for a multi-zone market that better represents a real power system. In fact, the application to a market with multiple zones further degrades the results as it implies a decrease in capacity leading to an increase of congestion costs. We first apply Nodal Pricing that we take as the first best solution with respect to which we measure welfare losses incurred in other organizations. By definition, Nodal Pricing gives the same results for the two zonal configurations and maximizes welfare. This is well known and needs no additional comment. Various considerations hamper the introduction of Nodal Pricing in Europe and the so-called Market Coupling is currently the state of the art in the continent. This architecture relies on a partial integration of the energy and transmission markets. Energy clears on the basis of a simplified representation of the grid that currently relies on “transfer capacities” but is apparently evolving to a “flow-based” (close to the flowgate in US parlance) type approach where the grid is represented by a set of “critical infrastructures”. Whatever the final description of the grid, the inevitable simplifications implied by the “transfer capacity” and “flow based” approaches may require counter-trading in order to match the clearing of the energy market with the real capacities of the grid. Regulation 714/2009 requires TSOs to maximize transmission capacities. This obligation is not implementable in general because increasing the transmission capacity between two zones in a meshed grid may at the same time reduce transmission capacity between other zones. The example used in this paper bypasses this logical impossibility by addressing a simple two zone system for which it is then effectively possible to “maximize” the transfer capacity. Even with this restriction, Market Coupling still raises questions. Counter-trading can be organized in different ways depending on the degree of coordination of the TSOs operating at zonal level. We first consider an ideal situation where TSOs are perfectly integrated in order to minimize counter-trading costs. Our simulations show that the choice of a “right TC” can make Market Coupling with proper counter-trading relatively efficient compared to Nodal Pricing. Conversely, a too high or too low TC degrades welfare. The problem is obviously to select the right TC, a question that remains, and will probably remain, unsolved. Suppose that the TC has been properly derived in the sense that an integrated counter-trading minimizes welfare losses compared to Nodal Pricing. Numerical experiments suggest that the creation of an internal market of counter-trading resources can sometimes perfectly substitute the horizontal integration of counter-trading operations as the arbitrage taking place to access the counter-trading resources can lead to an implicit coordination among TSOs. This is not a general result though: the implicit coordination can easily be replaced by less satisfactory equilibrium (with losses of welfare) as a result of small changes of assumptions. Moreover, the conditions where implicit coordination hold may be quite difficult to achieve in practice: experience has shown that implementing an internal market, whether for energy or ancillary services such as reserve or balancing is difficult in the European electricity market. The same is likely to be true for access to counter-trading resources as some TSOs may want to reserve some resources for their own actions. We show that limitations upon the access to these counter-trading resources can effectively jeopardize the implicit coupling of the actions of the TSOs. One should also recall that reliability, which is not discussed in this paper, would considerably degrade the case for implicit coordination. We also consider another instrument for enhancing the coordination of TSOs in counter-trading, namely the creation of a more or less complete market of line capacity. This market can embed all constrained lines (a complete market) or only some of them (an incomplete market). This market may also be completely missing but we do not test this latter case. A complete market significantly enhances the efficiency of counter-trading, whatever the assumption made on access to counter-trading resources. Conversely an incomplete market of line services coupled with a limited access to counter-trading resources progressively degrades welfare, possibly leading to absence of equilibrium solutions or to a multiplicity of solutions. The limiting case occurs when each TSO has access only to counter-trading resources in its own control area. Welfare losses when there is an equilibrium become huge and the equilibrium is not guaranteed to exist. High counter-trading costs have been observed in the market as well as the impossibility to counter-trade because of lack of resources. It remains to see how the absence of equilibrium would be translated in practice. Note that these very bad outcomes only occur when the clearing of the energy market leads to congestion. It can thus be expected that TSO will ex ante reduce transmission capacities so as to avoid resorting to counter-trading. The analysis was conducted under two scenarios of counter-trading resources: TSOs can procure these resources from generators and consumers in one case or with only generators in the second case. Enlarging access to counter-trading resources is beneficial when the market functions well (good access to cross border counter-trading resources and a market for line capacities). The outcome becomes unpredictable otherwise. This is not altogether surprising in a market that restricts quantity coordination (limited joint optimization) and price coordination (incomplete market). All this analysis can be summarized in a simple recommendation: if one persists in not installing a proper transmission market at the energy level, at least one should install it at the counter-trading level. The compounding of inadequate TC computations and inefficient counter-trading can create havoc in the market and large welfare losses.