How integrated are European electricity markets?
June 15th, 2010 by Georg Zachmann, Research Fellow, BruegelElectrons passing through international electricity networks follow the complicated physics of meshed alternating-current grids and do not respect price-zone borders. Correspondingly, scheduling power plants on the basis of national prices without taking into account the real physical networks within and between countries makes it necessary for the system operator to take non-market measures to ensure system stability . This is illustrated by two obvious inefficiencies. First, cross-border transmission lines are rarely fully utilised despite persisting price differentials. Second, electricity very often flows from high- to low-price areas . Overall, the current system is light years away from the theoretically optimal use of European power plants and transmission lines that would characterise a fully functioning single energy market.
Despite institutional progress between 2005 and 2009 , we argue, using the case of the systemically important German market , that progress has been limited in the last five years. We base our claim on three observations:
First, prices in the countries that border Germany have not converged significantly with German prices, and high hourly price differentials were common between 2005 and 2009. In the electricity market as currently structured, decreasing differences between national hourly electricity prices could be a sign of greater market efficiency and/or increasing commercially available transmission capacity. However, Nitsche et al (2009) found that German electricity spot prices in 60-90 percent of instances differed by more than five percent from the price in almost all neighbouring countries. In addition, the absolute annual average price differences in 2009 for almost all of these market combinations exceeded their corresponding 2004 value9 . Thus, recent reforms have failed to produce a consistent reduction in price differentials that would point to increasing market integration.
Second, price differentials are almost unrelated to capacity usage. In an integrated market, as long as lines are not fully used, the price differential should be zero while, as soon as congestion occurs (at full capacity), the price differential depends on the difference in marginal cost in both systems. In reality, however, electricity often flows against the price differential. In the German-Dutch case, electricity flowed from the high price area to the low price area for 49 percent of total hours in 2009 (Figure 1). At the German-French border this occurred for 46 percent of total hours. This illustrates the flaws of a system based on coupling large national zones that face internal congestion.
Third, physical interconnections have not developed substantially in recent years. Figure 2 shows the net transfer capacities from and to Germany between 2005 and 2009. While the import capacity remained constant, the export capacity decreased by more than 15 percent. The absence of progress in Germany’s cross-border transmission capacity is representative of the EU-wide picture. Between 2005 and 2009, the average net transfer import capacity of all countries in the European transmission system decreased by more than 15 percent. This is mainly due to the need for increased security margins for cross-border trade because of growing shares of intermittent renewable production. Thus, decreasing availability of cross-border transmission for commercial operations becomes an increasingly limiting factor for market integration.
In conclusion, the remaining international price differentials, the insufficient response of electricity flows to price signals and non-increasing cross-border transmission capacity demonstrate that in the last five years progress towards a single market for electricity has been limited.
Georg Zachman, Research Fellow, Bruegel
P.S. Full document is available here.
June 15th, 2010 at 10:47 pm
Dear Georg,
thanks for focusing on this issue on Energy Policy Blog. I dare to comment on your post because I have just completed a paper on your same evaluation on EU market, it will be published in the next weeks in Italian language on Energia review and if you manage to “chew” some Italian language I will be pleased and glad to send it to you!
Well, let me start from the end: in my research I almost completely agree with your conclusions, namely European markets are not sufficiently integrated yet and since 2005 (end of preliminary EU sector inquiry) few steps forward have been made so far. 2002 Barcelona council targets have not yet been reached in most of the main EU countries (i.e. ITA, DE, FR, ES, PT, GR, PL) and interconnectors capacity is still not sufficient.
Nevertheless, let me point your attention to some steps that have improved the situation if you compare it with 2005. Here I refer mainly to congestion management mechanisms and to implicit auctions. Successful experience in the TLC market coupling is now drawing more interest in market coupling projects and thanks to implicit auctions some of the hinders you mentioned in your entry may be removed (partially and as far as physical capacity is enough), adverse flows may be limited and a better use of existing capacity is being made.
Finally, German example would also deserve some thoguhts on the effect of intermittent RES on import/export flows both with explicit and implicit auctions, wouldn’t it?
Hope I could be of any help and thanks for any feedback on my post as well.
Stefano
June 16th, 2010 at 11:50 am
Dear Stefano,
Thank you for your reply. I fully agree with your comments. Both points were not mentioned in the blog but briefly addressed in the Policy Brief (http://www.bruegel.org/uploads/tx_btbbreugel/1006-Electricity_Single_Market-PB.pdf).
(1) Trilateral Market Coupling had a measurable effect on price convergence (now in 70% of the hours there is no more price difference between F, BEL and NL) and it essentially stopped “internal” flows against the price direction. This success is, however, not easily replicated in larger, more meshed systems. In a nutshell the reason is that electricity flowing from Northern Germany to Southern Germany will always partly pass through the Netherlands, Belgium and France. As in the current system, Southern and Northern Germany have the same electricity price, electricity will flow against the price direction at some border as soon as congestion at any of the borders forces prices to diverge (Flows: N-D -> NL -> BEL -> F -> S-D Prices: N-D ≤ NL ≤ BEL ≤ F ≤ S-D ≤ N-D). Furthermore, it is highly likely that Germany faces internal congestion (difficult to prove without appropriate data). This congestion is dealt with by rescheduling (expensive countertrading) and possibly curtailing international transfer capacities. In the presence of such non-market barriers (artificial uniform price zone and internal congestion) just “merging” the order books of the power exchanges involved will not do the job. Consequently, implicit auctioning alone will not provide better signals for international power plant scheduling – and thus cost reductions. Therefore, a true single market is only possible when the scheduling of power plants (and not only the net export of price zones) is coordinated internationally by regionally differentiated price signals.
(2) RES production in Germany is likely to cause several European TSOs to increase their security margins when calculating net transfer capacities. The academic problem is, that methodologies for calculating NTCs are not published and that it is thus impossible to judge, whether decreasing transfer capacity was entirely due to increasing installed wind capacity. Consequently, creating transparency with respect to the calculation of NTCs is essential to understand the effect of wind power on cross-boder trade.
Best
Georg
June 18th, 2010 at 9:59 am
I think the following: http://www.nera.com/NewsletterIssue/NL_ERI_Issue%2037_Final.pdf
from Graham Shuttleworth of NERA sheds some light on the problems/issues being discussed here and outlines a rational and sensible way forward,