Electricity Market Coupling in the CWE Region: 1 Market, 4 Price Levels
Despite a range of electricity market coupling initiatives in Europe, there is no trend of price convergence. In the CWE region, the number of hours with price convergence between all four countries dropped remarkably from 68% in 2011 to only 5% in the first 8 months of 2015. Over the same period, the average hourly spread between the Belgian and German electricity markets went up from EUR 4 per MWh in 2011 to EUR 15 in 2015.
Context and scope of the study
The creation of a single internal European electricity market is a major objective of the EU. This “European copper plate” is expected to (i) increase social welfare, (ii) smoothen out the variability of intermittent renewable energy production and (iii) result in a more efficient power system overall.
In order to reach this objective, several market couplings have taken place among neighbouring European markets during the last decade. The CWE region is one of the more mature regional electricity markets in Europe. Belgium, the Netherlands and France have been coupled since 2006 while Germany was added in November 2010. With the Price Coupling of Regions initiative in 2009, the main Power exchanges in Europe moved from an explicit allocation of the interconnection capacity to an implicit auctioning mechanism. More recently, in May 2015, the allocation of cross-border capacity has been further optimised through the implementation of flow-based market coupling. One of the expected and desired outcomes of these measures is price convergence between the involved markets, within the limits of the interconnections.
In this article, Sia Partners investigates to what extent prices converge in the CWE region (Belgium, France, Germany and the Netherlands).
Price convergence in the CWE region
Measuring price convergence
Price convergence is measured using two different indicators. Firstly, the hours of price convergence indicate how often markets in the CWE region had the same price level. For this analysis, price convergence is defined as two markets having a price difference smaller than 1 EUR/MWh, a definition that is also used by the Belgian Regulator (CREG). A second indicator for price convergence is the spread between markets. The spread is defined as the difference in price between two markets.
The input data for this analysis are the hourly spot prices on the Belgian (Belpex), Dutch (APX), French (EPEX, formerly Powernext) and German (EPEX, formerly EEX) day ahead (DA) markets from 2009 until August 2015. These data are made available on the websites of these exchanges on a daily basis.
Hours of price convergence
Within the CWE region, the following scenarios were considered from the perspective of Belgium:
- No price convergence (BEL)
- Price convergence with at least one of the three countries mentioned above (BEL+1)
- Price convergence with at least two of the three countries mentioned above (BEL+2)
- Price convergence between all countries (BEL+3)
The yearly relative occurence of these scenarios is depicted in Figure 1 for the 2009-2014 period. After a marked spike in price converge in 2011 – coinciding with the addition of Germany to the CWE region, there is a clear trend of price divergence. Full price convergence in the region has decreased from 68% in 2011 to 21% in 2014 and only 5% during the first 8 months of 2015.
Figure 1 : Price convergence in the CWE region
A more detailed analysis shows that a decrease in hours of price convergence with Germany is one of the main drivers behind this trend. The occurrence of price convergence between Belgium and Germany dropped from 69% in 2011 to 32% in 2014.
Spread between Belgium and its neighbouring countries
The dynamics of the monthly average day‐ahead spot electricity prices of the countries in the CWE region visualised in Figure 2 indicate that the markets are drifting away. Belgium and the Netherlands have on average the highest prices while Germany is at the lower end of the price spectrum. France oscillates between these two price levels with a tendency to the higher prices in Belgium and the Netherlands during winter and convergence to the lower prices on the German market during summer.
Figure 2 : Average monthly Day Ahead price in the CWE region
By averaging on a monthly basis, information about price convergence at the hourly level is lost. Indeed, it is possible that the average monthly price of two markets is the same while the spread was +5 during half of the time and -5 during the rest of the time. However, in this case, the spread between different electricity markets is usually the consequence of structural trends. As a result, the monthly averages are a good indicator of the average spread at hourly level. This is confirmed by Figure 3 which shows the average hourly spread between Belgium and the other markets in the CWE region. The hourly spread is defined as the absolute value of the price difference between two markets. Note that for 2015, only the first 8 months are taken into account.
Figure 3 : Average hourly spread between Belgium and CWE countries
Flow based market coupling
As of May 2015, allocation of cross-border capacity has been optimized by the introduction of flow- based market coupling. The impact of this solution on price convergence is captured in Figure 4 by comparing the average hourly spread for the second quarter of 2015 (corresponding to the first four months of flow based market coupling) with previous years. It is clear that despite the introduction of flow based market coupling, the average hourly spread between the countries of the CWE region has never been higher.
Figure 4 : Evolution of the average hourly spread in Q2 for the 2009-2015 period
Drivers behind the increasing spreads
Lack of interconnection capacity
One obvious culprit for the increasing spread between the different national electricity markets would be the lack of interconnection capacity. However, the relation between the average hourly spread and the saturation on the interconnector lines is not a straightforward one. The reader is referred to Appendix 1 for a case study on the Belgian-French border. Only a very weak correlation between the difference in Day Ahead electricity prices and saturation on the interconnector lines was found.
Unavailibility of nuclear power in Belgium
Another factor that might have contributed to the decrease in price convergence in the region is the unavailibility of nuclear power in Belgium. This tale is explored in more detail in appendix 2. The analysis reveals that when available nuclear capacity was low, prices in Belgium and France were not only diverging more often, they were also diverging more strongly.
Finally, a third factor that was studied is the increasingly different energy mix in Germany compared to the rest of the region (see Appendix 3). Indeed, the Energiewende in Germany caused a shift to an energy mix with more renewables and coal-fired power generation. Meanwhile France and Belgium remain largely dependent on nuclear power. The Energiewende’s impact on spot prices is profound: electricity produced from renewable energy sources has a marginal cost of nearly zero. As a result, during the day (when solar reaches its peak) and especially during sunny and windy days, there is a structural price difference between Germany and the rest of the region.
The anticipated increase in price convergence that was expected from recent market coupling initatives is largely overshadowed by other trends in the European energy system. As this system has become extremely complex, it is nearly impossible to quantify the exact drivers behind the diverging prices in the CWE region. Nevertheless, this study identified key drivers such as the unavailibility of nuclear power in Belgium and the increasingly different energy mix in Germany in a qualitative way.
Appendix 1: Relation between spread and physical electricity flows over the BE-FR border
A case study of cross border flows between Belgium and France shows that interconnection capacity is not always used in an efficient way or that congestion occurs at other areas. Figure 5 shows the correlation between the spread and the utilisation of the interconnection capacity on the French-Belgian border. It is clear that congestion on the interconnectors does not necessarily translate into a high spread in economic terms, i.e. the way the available interconnection capacity is calculated does not always stroke with the physical flows. For example, the highest price differential of 171 EUR/MWh was realised when only 743 MW of the more than 2,000 MW available interconnection capacity was used.
Figure 5 : Spread BE-FR in function of physical flows on the French-Belgian border in 2014 [Sources: Elia, Belpex, EEX]
Appendix 2 Unavailibility of nuclear power plants in Belgium
The Belgian nuclear electricity generation park consists of 7 reactors with a combined capacity of approximately 6 gigawatt of electric power. In 2012, Doel 3 and Tihange 2 were shut down due to concern about flaws in the reactor pressure vessels due to hydrogen flakes from their manufacture. The two units, representing 1/3rd of Belgium’s nuclear capacity, were restarted in May 2013 but shut down again in March 2014. 2014 turned out to be an annus horribilis for nuclear electricity production in Belgium: Doel 4 also had to be shut down from August to December due to sabotage of the steam turbine. The low availability of nuclear power in Belgium in 2014 had an important impact on the spread between Belgium and its neighbouring countries. Instances of price divergence between the spot markets in Belgium and France were more frequently when available nuclear capacity was low (see table 1).
Table 1 : Available Nuclear Capacity versus Price Divergence BE-FR
When available nuclear capacity was low, prices were not only diverging more often, they were also diverging stronger. Figure 6 shows the spread between Belgium and France in function of the total nuclear power production in 2014 for the instances were the spread was higher than 1 EUR/MWh. It is clear that when nuclear power generation was low, there was a higher probability of a larger spread between Belgium and France.Table 1 Available nuclear capacity versus price divergence BE-FR 2
Figure 6 Spread BE-FR in function of the nuclear power production in Belgium in 2014 [Sources: Elia, Belpex, EEX]
Appendix 3 Germany’s Energiewende
Since the Fukushima Daiichi nuclear disaster in March 2011, Germany decided to move away from nuclear power (uranium in figure 7) and shift towards an energy production park that is dominated by renewable energy sources.
Figure 7 Net installed generation capacity in Germany [source: Fraunhofer ISE]
On a sunny and windy day, solar PV and wind can produce more than half of the electricity in Germany (figure 8).
Figure 8 German electricity production per source at the end of August 2015 [source: Fraunhofer ISE]