Transcript TenneT Market Review

Market Review
2015
Electricity market
insights
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday
markets
Market
integration
Balancing
Balancing
Redispatch
Special events
in 2015
Main findings
2
Introduction
Introduction
Price
developments
Flow Based
Consumption
and production
For those with a professional interest in the electricity sector, TenneT publishes this yearly TenneT Market Review.
In this publication we describe the developments in the Western European electricity market, particularly in the Netherlands
and Germany, where, as the transmission system operator (TSO), TenneT has a central role in facilitating the market.
In this publication we give you the highlights of what was a dynamic year for the
electricity market and put the most important developments into perspective,
allowing you to deepen your knowledge of the events driving the market in 2015.
While average wholesale prices rose in the majority of European countries,
including Belgium and France, they decreased in Germany and the Netherlands.
For Germany this was the fourth time in a row and in the Netherlands the third
successive time.
This price development was driven by the developments in the production
capacity. Renewable energy production capacity grew strongly again in Germany,
following the long-term subsidy policy. The new offshore wind farms that were
connected to the TenneT grid contributed significantly. Also, the Netherlands
showed some growth in the installed capacity of renewables, but here the main
development came from the full commercial production of the three new coalfired power plants. Belgium was hit by a number of planned and unplanned
outages, which led to a sharp increase in imports from the Netherlands and
France, resulting in record exports for the Netherlands.
The increased need for imports into Belgium was facilitated by the go-live of
Flow Based market coupling for Germany, the Netherlands, France and Belgium.
This enlarged the possibilities to make capacity available for international trade,
on the border where it is most valuable.
In this fourth edition of the TenneT Market Review, we also want to draw your
attention to the substantial impact that the German energy transition has on
the daily operation of the electricity system. Replacing conventional generation
by renewable generation on a different location and with different production
patterns in 2015 led to a situation where redispatch measures were needed
on 90% of the days. The ongoing grid reinforcements will alleviate this problem.
This TenneT Market Review was created in close cooperation with IAEW from
RWTH Aachen University, and we hope that you will enjoy reading it as much
as we enjoyed creating it.
Intraday markets
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
3
Day-ahead
price developments
Introduction
Price
developments
In 2015, average wholesale prices in Germany and the Netherlands dropped further. Prices also decreased in
Poland and the Nordic countries. Other European countries witnessed rising wholesale prices in 2015.
Figure 1 shows the development of wholesale prices for the Central Western
Europe (CWE) region of Germany, the Netherlands, Belgium and France.
The yearly average price (base price) in Germany went down by 3.6% in
2015 to 31.63 €/MWh and also in the Netherlands by 2.7% to 40.05 €/MWh.
For Germany it was the fourth year in a row of decreasing wholesale prices.
Flow Based
Monthly Average Day-ahead Wholesale Prices in CWE countries
Consumption
and production
€/MWh
60
Intraday markets
50
Wholesale prices in France and Belgium were on average higher than in 2014,
with price peaks in February and October. On average, the French and the
Dutch wholesale prices were almost equal in 2015, where the French price
used to be much lower. Belgian prices were highest in the CWE region, notably
in April, September and October.
40
Market
integration
30
20
Balancing
10
BE
FR
December
October
November
August
2015
September
July
May
June
April
March
January
February
December
October
November
September
July
August
May
June
April
2014
NL
In 2015, prices dropped in Nordic countries, Poland, Germany and the
Netherlands. Conversely, the prices of all other market areas went up
on average.
March
January
0
February
Base prices across Europe are depicted in Figure 2. The price levels for
different European market areas for 2014 and 2015 are visualised by the
colours: the greener the colour of the market area, the lower the price.
In addition, Figure 2 shows the percentage of hours in which a country
had the exact same wholesale price as the Dutch, respectively the German,
market area (i.e. full price convergence). Germany and Austria constitute
one market area, meaning there is full price convergence of 100%.
Redispatch
DE
Figure 1: Monthly average of hourly Day-ahead wholesale prices in CWE countries.
Source: EPEX Spot, energate
Special events
in 2015
Main findings
4
Day-ahead
price developments
Price convergence generally remains at similar levels. The price convergence
of Belgium with the neighbouring countries decreased in 2015. Also the number
Introduction
Price
developments
of hours with full price convergence between France and Germany and between
Denmark and Germany decreased.
Flow Based
European Wholesale Prices and Price Convergence
Consumption
and production
2015
7|7
2|6
2014
2015
9|18
3|11
4|20
14|41
0|0
0|0
100|29
0|0
1|1
100|28
29|100
77|29
44|20
28|100
0|0
Market
integration
0|0
29|100
31|51
Intraday markets
5|20
14|34
28|100
0|0
30|27
0|0
1|0
Balancing
2|1
2|3
4|2
2|3
4|3
Redispatch
Price Convergence to NL (%) | Price Convergence to DE (%)
Price Convergence to NL (%) | Price Convergence to DE (%)
Average Day-ahead spot price (€/MWh)
Average Day-ahead spot price (€/MWh)
20
25
30
35
40
45
50
55
60
20
25
30
35
40
45
50
55
60
Special events
in 2015
Figure 2: Yearly average of hourly Day-ahead prices and % hours full price convergence (in relation to the Dutch and German market area) of different market areas1 in Europe.
Source: energate, APX, EEX, Nordpool Spot, POLPX, OTE, GME, OMIP
1
For countries with multiple market areas, one market area interconnected to one in another country was chosen: Italy: North, Norway: NO2, Great Britain: GB2
Main findings
5
Day-ahead
price developments
Figure 3 shows the development of full price convergence in the CWE region
in more detail. The monthly pattern of the convergence between German and
Dutch prices in 2015 is similar to that of the previous year, but with lower values
in spring and higher values in autumn.
Introduction
Price
developments
Among other CWE countries, greater differences could be observed with
lower full convergence in 2015 compared to 2014 for most months. Between
the Netherlands and Belgium, significantly lower price convergences were
witnessed from April onward. Price convergence was very low between France
Consumption
and production
Price Convergence between CWE countries
20
0
0
%
FR-DE
%
December
0
November
0
October
20
September
20
August
40
July
40
June
60
May
60
April
80
March
80
February
100
January
100
Market
integration
BE-FR
Balancing
Redispatch
December
20
November
40
October
40
September
60
August
60
July
80
June
80
2015
Intraday markets
May
100
2014
April
100
NL-BE
March
%
February
DE-NL
January
%
Flow Based
Special events
in 2015
Figure 3: Shares of hours with full price convergence between CWE countries (2014-2015). Source: EPEX Spot, energate
Main findings
6
Day-ahead
price developments
and Germany with the exception of May and July. Although there were high
levels of price convergence between Belgium and France returned in the first
three months, it was low for the rest of the year, with Belgian prices structurally
above French prices.
Introduction
Price
developments
Volatility of European wholesale prices
Flow Based
2014
€/MWh
2015
25
The standard deviations of hourly Day-ahead prices in Europe are shown in
Figure 4. In general, price movements were less volatile and more evenly
distributed, with the exception of Belgium. The standard deviation in Belgium
is heavily impacted by two days with extreme prices (cf. chapter 9).
Especially prices in Poland and Spain were less volatile, while prices in the
Nordic countries were more volatile, moving slightly in the direction of other
countries on the levels of the European electricity market.
Consumption
and production
20
15
Intraday markets
10
5
0
Prices in the Nordics and the Netherlands were the ones with the lowest price
volatility in 2015. All bidding zones adjacent to the Netherlands (Germany,
Belgium and Great Britain) showed a higher volatility compared to the Netherlands
in 2015.
Market
integration
PL
ES
IT
GB
FR
CH
DE
BE
CZ
NL
DK1
DK2
SE4
NO2
Figure 4: Standard deviation of European Day-ahead wholesale prices in 2014 and 2015.
Sources: EPEX Spot, Energate, APX, EEX, Nordpool Spot, POLPX
Balancing
Redispatch
Special events
in 2015
Main findings
Flow Based
Market Coupling
On May 20th 2015, Flow Based2 Market Coupling (FB) had its go-live in the CWE region. Due to the sophisticated capacity
calculation used in FB, import and export capacities increased. This had a dampening effect on wholesale prices in
the Netherlands and Belgium.
Under FB, a sophisticated capacity calculation procedure enables the use of
the transmission capacity of the grid to be optimized. The previous Available
Transfer Capacity (ATC) based market coupling used fixed values for the
interconnection capacity between the countries of the CWE. Now, under FB,
a domain of possible combinations of flows is used in the market coupling
algorithm. This allows for more degrees of freedom for the algorithm to
optimize. This optimization can lead to flows between countries that are larger
than under ATC. This is clearly shown in Figure 5, where an increase of net
positions3 can be observed after the introduction of FB on May 20th.
There are two possible ways to look at the price developments since the
introduction of the FB. The first one is to compare historical prices and their
convergences before the introduction of FB with the prices after the introduction.
Obviously, that does not take into account changes in other factors that
determine price levels and price convergence discussed in the following
chapters. The second way is to compare the actual outcome with the outcome
of the parallel run4. This has the drawback that changes in bidding behaviour
are not taken into account.
2
In this report we only discuss Flow Based Intuitive Market Coupling (as implemented) which only allows
positive commercial schedules from market areas with lower prices into market areas with higher prices.
3
The net position of a market area is the volume it will export or import in a given hour resulting from the
nominations on long-term capacities and the outcome of the market coupling calculations.
4
Before go-live there was a FB parallel run, after go-live there was an ATC parallel run until September.
Changes of Price Deltas since FB went live
In Figure 1 we saw a decreasing price delta between the Netherlands and
Germany after the introduction of FB. Figure 6 below shows the distribution
of price deltas between selected combinations of CWE market areas in 2014
and 2015. The results show only minor changes in the distribution of price
differences between the Netherlands and Germany as well as Belgium and
France in these two years. In contrast, the price differences between both the
Netherlands and Belgium, and France and Germany show significant changes.
Less price convergence and a greater number of high price differences between
these bidding zones can be observed. This reflects the insight of Figure 6
that less price convergence between the Netherlands and Belgium occurred.
Additionally, a broader range of price differences, in some cases exceeding
20 €/MWh, can be noticed.
7
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
8
Flow Based
Market Coupling
Introduction
Price
developments
CWE Flow based net positions
Flow Based
7,000
6,000
Consumption
and production
5,000
4,000
3,000
2,000
1,000
Intraday markets
0
-1,000
-2,000
-3,000
Market
integration
-4,000
-5,000
DE
FR
NL
16-12-15
30-11-15
14-11-15
29-10-15
13-10-15
27-09-15
11-09-15
26-08-15
10-08-15
25-07-15
09-07-15
23-06-15
07-06-15
22-05-15
06-05-15
20-04-15
04-04-15
18-03-15
02-03-15
14-02-15
29-01-15
13-01-15
28-12-14
12-12-14
26-11-14
10-11-14
25-10-14
09-10-14
23-09-14
07-09-14
22-08-14
06-08-14
21-07-14
05-07-14
19-06-14
03-06-14
17-05-14
01-05-14
-6,000
Balancing
BE
Figure 5: Daily net positions of CWE countries from May 2014 – December 2015. Source: ENTSO-E
Redispatch
Special events
in 2015
Main findings
9
Flow Based
Market Coupling
Introduction
Price
developments
Impact of Flow Based on Hourly Price Difference between CWE countries
Amount
of hours
NL-DE
3,000
8,000
2,400
6,400
1,800
4,800
1,200
3,200
600
1,600
0
0
Amount
of hours
BE-FR
5,000
5,000
4,000
4,000
3,000
3,000
2,000
2,000
1,000
1,000
0
2014
2015
Consumption
and production
Intraday markets
FR-DE
Market
integration
Balancing
->
25
5
-2
20
0
-2
15
5
-1
10
10
5-
5
0
0-
-0
-5
-5
0-1
-1
5-
-1
-1
0
5
0
-2
<-
->
25
5
-2
20
0
-2
15
5
-1
10
10
5-
5
0-
-0
-5
0-
-5
0
-1
-1
5-
-1
-1
0-2
<-
-2
5
0
0
0
0-
Amount
of hours
NL-BE
-2
Amount
of hours
Flow Based
Figure 6: Hourly price difference between the CWE countries divided over clusters of price difference for 2014 and 2015. Source: EPEX Spot, energate
Comparing FB with ATC parallel run
To exclude the impact of other influencing factors, we have to evaluate the
functioning of FB based on the results from the parallel run. Figure 7 shows
the resulting number of price areas for the FB and the parallel ATC run in
the CWE region for May to September 2015. Structural differences between
the market results from FB and the ATC model can be observed, showing that
FB leads to more differentiated results.
Redispatch
Special events
in 2015
Main findings
Flow Based
Market Coupling
On the one hand, the FB approach presents more hours of full price
convergence in all CWE countries. On the other hand, there is a higher share
of hours in which the CWE region is split into four price areas. In contrast, the
ATC methodology mostly results in two or three price areas. So, the full price
convergence increases between two countries, whereas the partial price
convergence in CWE decreases with FB.
This leads to two main findings: firstly, a more efficient and flexible allocation of
available capacities is accomplished; secondly, the existing available capacities
are still restrictive for achieving full price convergence for the whole CWE
region, but monthly levels around 40% were reached in May and July.
Figure 8 plots the Day-ahead prices5 against the net positions6 of the market area
for the FB and the parallel run with an ATC calculation. Looking at the net
positions of Belgium, it is obvious that FB allows for significantly larger imports
(negative net positions) into Belgium. FB is much less restrictive than ATC.
The same is apparent for Germany and France, which have both higher positive
and negative net positions, indicating that FB provides additional room for
the market. Even though FB enables the Netherlands to increase imports, the
bottom right graph in Figure 8 indicates a restriction at around 4 GW. This limit
reflects an external constraint which is added by the TSOs to the FB algorithm
to ensure system reliability. Overall, imports in the CWE region under FB where
43% higher than they would have been under ATC, whereas exports where
30% higher.
10
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
5
Please note that on this scale the graph does not show the extreme price levels that occurred in Belgium;
please refer to Chapter 9 where these situations are discussed
The net position of a country is the amount of electricity this country imports or exports in a given hour
following the outcome of the market coupling calculations
Special events
in 2015
6
Main findings
11
Flow Based
Market Coupling
Comparing the resulting prices of the two methodologies in the period May until
September, it can be noted that the average wholesale prices of the Netherlands
and Belgium decreased considerably as a result of FB, whilst average prices
increased slightly in Germany and France, as shown in the table below.
Introduction
Price
developments
Impact FB on the shares of the number of price areas within CWE countries
%
100
FB
ATC
Difference
BE
44.1
55.1
-20.0%
NL
39.9
46.7
-14.7%
DE
32.2
30.9
4.0%
FR
34.9
32.7
6.9%
Flow Based
Consumption
and production
80
60
Intraday markets
40
Market
integration
20
0
FB
ATC
May
1 price area
FB
ATC
June
2 price areas
FB
ATC
July
3 price areas
FB
ATC
August
FB
ATC
September
Balancing
4 price areas
Figure 7: Shares of the number of price areas within CWE for Flow Based and the parallel run ATC
for May-September. Source: TenneT
Redispatch
Special events
in 2015
Main findings
12
Flow Based
Market Coupling
Introduction
Price
developments
Flow Based effects on wholesale prices and net positions
Flow Based
€/MWh Belgium
€/MWh France
90
90
70
70
50
50
30
30
10
10
-10
-10
-30
-30
-50
Consumption
and production
Intraday markets
-50
-4,000
FB
-3,000
-2,000
-1,000
0
1,000
2,000
-4,000
ATC
FB
€/MWh Netherlands
-2,000
0
2,000
4,000
6,000
8,000
ATC
€/MWh Germany
90
90
70
70
50
50
30
30
10
10
-10
-10
-30
-30
-50
Balancing
Redispatch
-50
-4,000
FB
Market
integration
-3,000
ATC
-2,000
-1,000
0
1,000
2,000
-4,000
FB
-2,000
0
2,000
4,000
6,000
8,000
Special events
in 2015
ATC
Figure 8: Average Day-ahead wholesale prices and net positions for the CWE countries as in FB and in the parallel run ATC from May-September 2015. Source TenneT
Main findings
Flow Based
Market Coupling
Figure 9 underlines the fact that capacity calculation used in FB provides
additional room for the market on the border of the Netherlands and Germany.
The dashed black line represents the exchanges from Germany to the
Netherlands in 2015 in descending order. In about 1,500 hours the import
exceeds the limit that was used in the ATC Market Coupling methodology.
In summary, the investigation of the parallel run of FB and ATC methodology
leads to the conclusion that FB has been implemented and is now working
successfully. In general, the advanced capacity calculation methods applied
in FB provide more cross-border capacity where the value is highest.
13
Introduction
Price
developments
Flow Based
Consumption
and production
Duration Curves Commercial Exchanges between Germany
and the Netherlands
Duration Curves Commercial Exchanges DE - NL
Intraday markets
MWh/h
6,000
Market
integration
5,000
4,000
Balancing
3,000
2,000
Redispatch
1,000
0
2014 DE > NL
2015 DE > NL
MAX NTC 2014
Special events
in 2015
Figure 9: Duration Curves of the Commercial Exchanges between Germany and the Netherlands
in 2014 and 2015 compared with the NTC cap in 2014. Source: ENTSO-E
Main findings
14
Consumption
and production
Introduction
Price
developments
In Germany, conventional power plants are being replaced by wind and solar, leading to a record renewable production of 32.5%
of domestic demand. In the Netherlands, renewables are growing, but at a much lower pace; coal-fired power plants replaced
gas-fired power plants. The development of gas and coal prices brings electricity prices in the two countries closer together.
Consumption
As depicted in Figure 10 a downward trend of electricity consumption in CWE
can be reported for recent years. But this trend was not maintained in 2015.
A slight increase of consumption in Belgium, the Netherlands and France could
be witnessed, whereas the consumption in Germany remains on a downward
trend for the fifth year in a row.
Flow Based
Consumption
and production
Annual electricity consumption compared to the base year 2010
%
Intraday markets
102
100
98
Market
integration
96
94
92
Balancing
90
88
2010
DE
2011
FR
NL
2012
2013
2014
2015
BE
Redispatch
Figure 10: Annual electricity consumption compared to the base year 2010 in CWE countries.
Source: ENTSO-E
Special events
in 2015
Main findings
15
Consumption
and production
Figure 11 shows the monthly demand of CWE countries in 2015 and 2010.
There are no monthly structural changes observable. The demand has slightly
decreased in 2015 throughout CWE since 2010. Besides economic growth,
innovations and efficiency measures, the temperature profile is also a driver
for the amount of electricity consumption. This can be further illustrated with
the example of the consumption in December for France in 2010 and 2015.
The average temperature in Paris in December 2010 was 0° Celsius in contrast
to 9° Celsius in 2015, which led to much higher consumption in December 2010
than in 2015.7
Introduction
Price
developments
Development of Monthly Electricity Consumption in CWE countries
Flow Based
GWh/month
60,000
Consumption
and production
50,000
40,000
30,000
Jan
Feb
DE 2015
Mar
Apr
DE 2010
May
Jun
FR 2015
Jul
Aug
Sep
Oct
Nov
Dec
Intraday markets
FR 2010
12,000
10,000
Market
integration
8,000
6,000
Jan
Feb
BE 2015
Mar
Apr
BE 2010
May
NL 2015
Jun
Jul
Aug
Sep
Oct
Nov
Dec
NL 2010
Balancing
Figure 11: Monthly electricity consumption in CWE countries. Source: ENTSO-E
7
Temperatures from Weather Underground
In order to compare the demand between 2014 and 2015, Figure 12 shows the
differences in demand between 2014 and 2015. Interestingly, French electricity
consumption in February was on average 4 TWh/month higher in 2015 due
to lower temperatures in 2014. A high heating demand leads to an increased
electrical demand due to the widespread electric heating systems in France.
This was one fundamental driver for the observed high prices in February 2015
in France. In contrast, the consumption in December 2015 was almost 5 TWh/
month lower compared to 2014, which can be explained by the previously
Redispatch
Special events
in 2015
Main findings
Consumption
and production
mentioned high average temperature in 2015. Therefore, the low consumption
in this month can be identified as one driver for the low prices in France in
December 2015.
Monthly Load differences in CWE countries
Production
The supply side of the German and Dutch power systems are analyzed below.
The following paragraphs provide an overview of the Dutch and German
electricity production by looking at the three main drivers: generation stack,
generation including renewable generation, and fuel prices.
Dutch and German generation capacity
The trend of increasing generation capacities based on renewable energy
sources in Germany, which could be witnessed over the last years, continued
in 2015. Figure 13 compares the operating capacities between 2014 and 2015,
showing the transition from conventional to renewable.
GWh/month
6,000
4,000
2,000
Power plants fired by oil, natural gas or hard coal and one nuclear power plant
were shut down. Following the German political decision to gradually phase
out nuclear until 2022, the power plant Grafenrheinfeld (1275 MW), located in
the north of Bavaria was, shut down at the end of June 2015.
0
-2,000
In contrast, a high increase of renewable generation was observed. The offshore
capacity in 2015 increased fourfold from 2014, with an absolute increase of
more than 2 GW.
-4,000
-6,000
Jan
Feb
Mar
DE
FR
NL
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
BE
Figure 12: Monthly electricity consumption delta between 2015 and 2014 in CWE countries.
Source: ENTSO-E
16
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Dec
Including generation capacities based on solar power, hydro and biomass, the
addition of renewable generation came to more than 6.8 GW in compared with
the 2.8 GW of nuclear and fossil-fuel power plants that were decommissioned.
Redispatch
Special events
in 2015
Main findings
17
Consumption
and production
German Operating Generation Capacities
Introduction
Price
developments
250
10
200
8
Also in the Netherlands, the installed capacity of wind and solar increased
in 2015. As in 2014, the installed capacity of solar panels grew by 350 MW,
surpassing a total of 1 GW at the end of 2015. The generation capacities
based on wind increased by more than 300 MW last year, reaching a total
of more than 3 GW at the end of 2015. These additional capacities based on
solar, wind and hard-coal substitute nearly 2 GW of gas-fired power plants,
which were mothballed.
150
6
Dutch Operating Generation Capacities
GW
Flow Based
Consumption
and production
Intraday markets
100
4
50
2
GW
-2
Nuclear
Hard Coal
Lignite
Oil
Hydro
Biomass
Wind (offshore)
Solar
Other
2
25
2015
2014
Market
integration
30
0
0
3
35
20
1
Natural Gas
Wind (onshore)
Balancing
15
-4
∆
Figure 13: German Operating Generation Capacities in 2014 and 2015 per fuel type and the
changes in Generation Capacities in 2015. Source: Kraftwerkslisten der Bundesnetzagentur
0
10
5
-1
0
2015
2014
Where hard coal capacity decreased in Germany, it increased in the Netherlands,
as shown in Figure 14. In the last two years, about 3.3 GW of hard coal-fired
power plants came into operation, 1.5 GW of which in 2015. The reverse trend
occurred for gas-fired power plants with a reduction of 2.2 GW, 1.9 GW of which
in 2015.
Redispatch
Nuclear
Coal
Gas
Waste
Biomass/biogas
Wind
(onshore)
Wind
(offshore)
Hydro
-2
Solar
∆
Special events
in 2015
Figure 14: Dutch Operating Generation Capacities in 2014 and 2015 per fuel type and the changes
in Generation Capacities in 2015. Source: TenneT
Main findings
18
Consumption
and production
The observation in the generation stack of the Netherlands with increasing
installed capacities of hardcoal-fired power plants is reflected in the production.
A trend towards more coal- and less gas-fired production in the Netherlands
can be concluded by taking a look at Figure 15.
Introduction
Price
developments
Dutch Coal and Gas Electricity Generation
Flow Based
TWh/
month
8
6
Intraday markets
5
4
Market
integration
3
2
1
Balancing
2013
Coal
2014
November
September
July
May
March
January
November
September
July
May
March
January
November
September
July
0
May
Comparing Figure 16 with Figure 20 further below, the high correlation between
renewable production in the Netherlands and Germany becomes obvious.
Despite the growth of renewable production in the Netherlands, the price
impact is still largely due to high imports from Germany in times of high
renewable production.
7
March
On the other hand, the growth of installed wind and solar capacities is also
observed in Figure 16. Production from solar, especially in the summer months,
becomes more and more relevant. In November we saw an absolute record
for wind energy, with a production of over 1500 GWh. The total share of
renewables in electricity consumption in the Netherlands in 2015 was 11.1%8.
Consumption
and production
January
Since the last month of 2014, the generation by hard coal power plants
increased significantly due to the new installed capacities. The overall
generation of natural gas and hard coal power plants was rather stable
comparing the same months of 2014 and 2015, which supports the perception
of generation based on gas being replaced by a generation based on hard
coal. However, it should be noted that three older coal-fired plants, with a
combined capacity of 1611 MW closed officially on 1 January 2016. In 2015,
these three plants together accounted for 21.7% of coal-fired generation.
Redispatch
2015
Gas
Figure 15: Dutch gross electricity generation from coal and gas plants (>10MW). Source: TenneT
Special events
in 2015
8
CBS Statline: Renewable electricity; production and power. *Preliminary figures for 2015
Main findings
19
Consumption
and production
Introduction
Price
developments
Dutch Monthly Average Solar and Wind Feed-in
Flow Based
MWh/h
€/MWh
1,800
60
1,500
50
1,200
40
900
30
600
20
300
10
0
Wind (MWh/h)
Solar Feed-in (MWh/h)
Intraday markets
Market
integration
Balancing
December
November
October
August
2015
September
July
June
May
April
March
January
February
December
October
November
August
September
2014
July
May
June
April
March
January
February
December
October
November
September
July
2013
August
May
June
April
March
February
January
December
October
November
September
July
2012
August
May
June
April
March
February
January
0
Consumption
and production
Average Dutch Day-ahead Price (€/MWh)
Redispatch
Figure 16: Dutch monthly average wind and solar feed-in and Day-ahead price. Source: CBS, EnTranCe
Also for Germany too, the change of the generation stack is reflected in the
electricity production as shown in Figure 17. The share of renewable generation
– especially wind and solar – continuously increased over recent years and
reached a record of 30% of gross electricity generation. Due to the net
exporting position of Germany, this leads to a share of 32.5% of Germany's
electricity consumption. The German political target of 35% renewable energy
generation of consumption in 2020 seems within reach.
Special events
in 2015
Main findings
20
Consumption
and production
German Power Generation
Introduction
Price
developments
80
Figure 18 compares the distribution of the feed-in from renewables between
different sources in Germany. In 2014, more than one-third of the renewable
feed-in was produced from onshore wind turbines and only 1% from offshore
wind capacities (but the latter increased in 2015 to 4%). The share in production
from offshore wind is larger than its share in capacity. A rough calculation
indicates that 1 MW installed capacity offshore produces 50% more electrical
energy compared to 1 MW installed capacity onshore. Wind energy contributed
a total of 44% of renewable production.
60
Due to the strong growth of wind energy, the share feed-in by solar decreased
to 22% of the total renewable production despite its growth in absolute figures.
%
100
Flow Based
Consumption
and production
Intraday markets
Distribution of Renewables in Germany
Market
integration
40
Wind offshore 1%
Hydro 12%
Wind offshore 4%
Hydro 10%
20
Wind
onshore 34%
Biomass 27%
162.4 TWh
Wind
onshore 40%
Biomass 23%
165.7 TWh
0
Balancing
Waste 3%
2004
2005
2006
2007
2008
2009
Nuclear
Oil
Lignite
Other
Hard Coal
Renewable energy sources
2010
2011
2012
2013
2014
2015*
Natural Gas
Figure 17: Shares of gross electricity generation per generation type in Germany (2004-2015).
Source: AG-Energiebilanzen (*Preliminary values for 2015)
Waste 4%
Solar 22%
Hydro
Biomass
Waste
Solar
Wind onshore
Wind offshore
*preliminary value
Solar 20%
2015*
2014
Figure 18: German distribution of renewables shares feed-in. Source: AG-Energiebilanzen
(*Preliminary values for 2015)
Redispatch
Special events
in 2015
Main findings
21
Consumption
and production
Figure 19 depicts the German average monthly feed-in from solar and wind
over the last four years. A high monthly variation of feed-in – and thereby full
load hours of solar and wind – can be observed. The feed-in of solar and wind
are complementary on a monthly basis. In the winter months from autumn
to spring, the average feed-in of wind turbines is relatively high and the solar
feed-in low, whilst in the summer months it is usually vice versa, but still there
Introduction
Price
developments
is a strong variation. In March and October of 2015, the feed-in by wind
turbines was remarkably low, which was one important driver for higher
average Day-ahead prices in the German market area. In contrast, the average
feed-in increased in November and reached a record in December of 2015.
The high renewable feed-in in these two months reduced the average
Day-ahead prices due to the merit order effect.
Flow Based
Consumption
and production
German Monthly Average Solar and Wind Feed-in
Intraday markets
MWh/h
€/MWh
18,000
60
15,000
50
12,000
40
9,000
30
6,000
20
3,000
10
0
Market
integration
Balancing
Redispatch
Wind (MWh/h)
Solar (MWh/h)
December
October
August
November
2015
September
July
May
June
April
March
January
February
December
October
November
September
July
2014
August
May
June
April
March
January
February
December
October
November
September
July
2013
August
May
June
April
March
February
January
December
October
November
2012
September
July
August
May
June
April
March
February
January
0
Special events
in 2015
Average German Day-ahead Price (€/MWh)
Figure 19: German monthly average wind and solar feed-in and Day-ahead price. Source: EEX, ENTSO-E
Main findings
Consumption
and production
A more general effect of solar energy on the wholesale price is shown in
Figure 20. First, the average feed-in by solar power increases year over year
due to the additional installed capacities. Second, the base price decreased
over the last three years. Especially around noon, feed-in by solar power is
the main driver for decreasing prices.
German Solar Feed-in and Day-ahead prices
GWh/h
€/MWh
15
60
2013
2014
2015
2013
2014
Furthermore, taking a look at the renewable average feed-in per month in
2015, the development of offshore capacities is visible. Although a very small
amount of feed-in by offshore capacities at the beginning of the year can be
observed, a significant increase of the average feed-in through the year can
be reported. As an example, the HelWin2 Project has been in operation since
June 2015 and provides a DC-grid connection of 690 MW for wind farms in
the eastern North Sea. This connection is used to integrate the feed-in such
as the Amrumbank West wind farm with a generation capacity of 288 MW.
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
23-24h
22-23h
21-22h
20-21h
19-20h
18-19h
17-18h
16-17h
15-16h
14-15h
13-14h
12-13h
11-12h
10-11h
8-9h
9-10h
7-8h
6-7h
5-6h
0
4-5h
0
3-4h
20
2-3h
5
1-2h
40
0-1h
10
Figure 21 depicts the distribution of renewable feed-in in the different German
control areas. The major share of solar and wind feed-in is in the TenneT
control area. Due to the long coast line of the German TenneT control area,
especially with the North Sea, most of the offshore wind farms are connected
with the TenneT grid. Apart from high feed-in by onshore and offshore wind
turbines, TenneT is challenged by the integration of the feed-in from photovoltaic
panels in underlying distribution networks, predominantly in Bavaria.
22
2015
Special events
in 2015
Figure 20: German average hourly solar feed-in 2013-2015 and average hourly Day-ahead prices.
Source: EEX, Anlagenregister der Bundesnetzagentur
Main findings
23
Consumption
and production
Introduction
Price
developments
German Wind and Solar Feed-in by TSO and per Month
Flow Based
TWh/a
GW/month
60
20
45
15
Consumption
and production
Intraday markets
30
10
Market
integration
15
5
Balancing
0
Wind Onshore
Wind Offshore
December
November
October
September
August
July
June
May
April
TransnetBW
March
50Hertz
February
Amprion
January
0
TenneT
Redispatch
Solar
Figure 21: Distribution of German wind onshore, offshore and solar feed-in among TSOs and monthly among 2015. Source: ENTSO-E
Special events
in 2015
Main findings
24
Consumption
and production
Marginal costs
In order to explain the developments described, primary fuel prices as well as
prices for CO2 certificates are investigated in more detail. Figure 22 shows the
corresponding graphs for the prices of natural gas at the Title Transfer Facility
(TTF), hard coal (API#2) and CO2 emission allowances.
Introduction
Price
developments
Primary fuel prices continued to fall, whilst CO2 prices increased in 2015.
Due to this contrary development of primary energy prices and CO2 emission
allowances, a distinction between different technologies is necessary.
Flow Based
Consumption
and production
Prices for Natural Gas, Hard Coal and CO2 Certificates
€/MWhth
€/t
50
40
80
10
70
9
8
60
7
50
30
20
40
5
30
4
3
2
10
2014
Natural gas
0
2015
2014
Hard coal
Figure 22: Daily Day-ahead gas prices from EEX TTF Index, monthly hard coal prices from API#2
ASK (CIF ARA) and daily CO2 future price for years 2014/2015 traded through 2013/2014/2015.
Source: Energate9
9
Balancing
1
0
0
Market
integration
6
20
10
Intraday markets
€/tCO2
Assumptions: Gas EEX TTF daily Day-ahead Index (Energate), API#2 hard coal monthly (CIF ARA) (Energate),
CO2 future price daily for years 2013/2014/2015 traded through 2013/2014/2015 (EEX, Energate). API#2
index published in $ and converted here to €, in order to be comparable to the other primary energy prices.
2015
2014
2015
CO2
The outcome of a marginal generation cost calculation for hard coal- and
natural gas-fired power plants is shown in Figure 23. The price movements
of hard coal and CO2 emission allowances compensated each other in 2015.
However the decrease of natural gas prices predominated the increase
of the CO2 emission allowances prices, which resulted in a small reduction
of the marginal generation costs.
Redispatch
Special events
in 2015
Main findings
25
Consumption
and production
The spread between gas and coal generation costs was stable between 2014
and 2015 on average but a decreasing spread over the course of 2015 could
be observed. This contributes strongly to price convergence between the
Netherlands and Germany in the second half of 2015. This is because in 2015
gas remained the price-setting technology in the Netherlands, in contrast to
Introduction
Price
developments
hard coal in Germany for most of the hour, despite the growth of hard coalfired generation in the Netherlands and the growth of renewables in Germany.
Therefore the difference in marginal generation costs shown in Figure 23 is
the most important driver for the difference in the wholesale prices in the
two countries.
Flow Based
Consumption
and production
Marginal Generation Costs Gas and Coal
Intraday markets
€/MWh
60
50
Market
integration
40
30
20
Balancing
10
Hard coal
CO2 certificate costs
December
November
October
September
August
July
May
April
June
2015
Redispatch
Transportation costs
Figure 23: Marginal generation costs of hard coal and natural gas power plants. Source: EPEX, energate10, IAEW
10
March
February
January
December
November
October
September
2014
August
July
June
May
April
March
February
January
December
November
October
September
2013
Natural gas
August
July
June
May
April
March
February
January
0
Special events
in 2015
Assumptions: Gas EEX TTF daily Day-ahead Index (Energate), API#2 hard coal monthly (CIF ARA) (Energate), CO2 future price daily for years 2013/2014/2015 traded
through 2013/2014/2015 (EEX, Energate). API#2 index published in $ and converted here to €, in order to be comparable to the other primary energy prices.
Main findings
26
Intraday markets
Introduction
In 2015, Germany saw a growing importance of Intraday trading. This was supported especially by the development of the intraday
auction and driven by the demand from market parties for quarter-hourly products for shaping, and to be able to act on improving
forecasts during the day to balance their portfolio.
Intraday trading is becoming increasingly important due to two factors. First,
the forecasts of renewable feed-in are updated during the day and through the
ongoing expansion of renewable generation, Intraday trading is more and more
relevant. Second, whilst in the Day-ahead market the tradable products are
based on hourly granularity, the possibility to trade quarter-hourly products
in the Intraday market provides the participants with a better trading product
for ramp rates. This is very helpful for trading the feed-in of solar power,
which is always associated with high – and predictable to a high degree –
gradients. Furthermore, average hourly products cannot be sufficient for
managing balancing groups in quarterly-hour granularity, especially in
respect to ramp rates.
In Germany there were two ways to participate in the Intraday market. On the
one hand, there is the continuous trading which allows the participants to trade
hourly and quarter-hourly products from 4 pm the previous day up to the gate
closure time of 30 minutes before delivery. This continuous trading option is
based on an open order book. On the other hand, since December 2014 there
is the Intraday auction clearing at 3 pm, shortly after the Day-ahead auction.
This auction allows trading quarter-hourly products one day before delivery
with a market clearing price system.
the introduction of the Intraday auction, which accounts for 4 TWh/a.
Summing up the quarter-hourly products of the auction and the continuous
trading and comparing it to 2014, an increase of 60% is achieved.
Price
developments
Flow Based
Consumption
and production
German Intraday trading volumes
TWh/a
Intraday markets
40
30
Market
integration
20
10
0
2013
ID Hourly
ID Quarterly
2014
2015
Balancing
ID auction
Figure 24: Intraday Auction, Quarterly and Hourly trading volumes in Germany. Source: EPEX
Figure 24 depicts the traded volumes of the products on the respective
Intraday market. An overall growth rate from 2013 to 2015 underlines the
increasing relevance of the Intraday market. The traded hourly product in
the continuous trading increased by about 7 TWh/a, reaching a new record
of 26 TWh/a in 2015, whereas a small decrease of continuous traded
quarterly-hour products can be reported. This small decrease is due to
Redispatch
As already mentioned, one reason to provide tradable products to the market
are ramp rates. Such ramp rates can increase with additional installed generation
capacity based on solar due to the natural predefined gradient over the day.
Due to the quarterly settlement, balance-responsible parties will want to match
the predicted production in the market on a quarterly basis.
Special events
in 2015
Main findings
27
Intraday markets
Figure 25 differentiates the traded volume between the quarter-hours. It shows
that significantly higher volumes in the first quarter-hour and in the fourth
quarter-hour were traded than in the second or third quarter-hour. This shows
how quarter-hour products are used to fulfill ramp rates because hourly
Introduction
products can only trade the hourly average, which results in a 'block profile'
that does not match the continuous developments in load and renewable
production and does not match the balancing responsibility of market parties
on a quarter-hourly basis.
Price
developments
Flow Based
German Quarterly traded Intraday volumes
GWh/a
Consumption
and production
180
160
140
Intraday markets
120
100
80
Market
integration
60
40
Balancing
20
0
1
5
Vol quarter-hour 1
9
13
17
Vol quarter-hour 2
21
25
29
Vol quarter-hour 3
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
Vol quarter-hour 4
Redispatch
Figure 25: Quarterly traded Continuous and Auction intraday volumes in 2015 in Germany. Source: EPEX
To combine this insight with the knowledge of the natural solar gradient over
the day, Figure 26 depicts the weighted average price (WAP) of the continuous
traded quarter-hour over the day and the average feed-in by generation
capacities based on solar in 2015. It is observed that the WAP of the first
quarter-hour is higher than the WAP of the fourth quarter-hour in the morning,
and vice versa in the afternoon. One explanation is the ramp rate of solar feedin. If market participants trade the expected hourly average on the Day-ahead
market – even without any forecast errors – trading on the Intraday market
Special events
in 2015
Main findings
28
Intraday markets
or a new dispatch of the own portfolio is necessary to be balanced within
a quarter-hour. In the morning with a positive slope of the solar feed-in, one
would be short in the first quarter of the hour and long in the fourth quarter.
This is reflected in the prices. Although one should not neglect other drivers
Introduction
for Intraday prices like residual load, forecast errors and bidding strategies,
it points out the relationship between renewable generation, especially solar,
and the Intraday market due to ramp rates of solar is one driver for systematic
quarterly ramp demand.
Price
developments
Flow Based
German average Quarterly Intraday prices and average Solar Feed-in
GWh/h
€/MWh
60
18
50
15
40
12
30
9
20
6
10
3
0
Consumption
and production
Intraday markets
Market
integration
Balancing
0
00
10
15
WAP quarter-hour 1 €/MWh
20
25
30
WAP quarter-hour 4 €/MWh
35
40
45
50
55
60
65
70
75
Average Solar Feed-In GWh/h
80
85
90
95
Redispatch
Figure 26: Yearly Weighted Average Prices of Quarterly traded products during the day in 2015 in Germany and Yearly Average solar feed-in. Source: EPEX
Special events
in 2015
Main findings
Market integration and
interconnection flows
Figure 27 shows the yearly aggregated commercial scheduled flows (Day-ahead)
for the CWE region and at the German borders in 2014 and 2015.
In 2015, the exports from Germany to southern countries (France, Switzerland,
Austria and the Czech Republic) significantly increased in comparison to 2014.
At the same time, exports from Germany to Scandinavian countries decreased.
In total, the German net commercial exports were 17 TWh/a higher in 2015.
In 2015, the exports from the Netherlands increased, most significantly the
exports towards Belgium increased by 69% to 6.1 TWh. On the other hand,
imports into the Netherlands decreased, most significantly from Belgium,
decreasing 58% to 1.4 TWh, but also from Germany, with a decrease of 7.8%
to 16.6 TWh. Patterns on the DC interconnectors remained stable with
structural imports from Norway and structural exports to Great Britain. Despite
the overall decrease of imports and the increase of exports, the Netherlands
remained a net importer in 2015.
Such imports and exports reflect the net positions that result from the market
coupling calculations. In 2014, the market coupling algorithm was based on
ATCs. In 2015, this was changed to the Flow Based method, which has a
more sophisticated way of taking available transportation capacity into account
(cf. chapter 3).
In contrast to Figure 27, Figure 28 shows the yearly aggregated physical flows
for the CWE region and at the German borders in 2014 and 2015. In 2015,
more exports from Germany to Switzerland and Austria could be observed.
Exports to the Netherlands decreased slightly. Imports from France,
Switzerland and the Czech Republic to Germany also decreased.
The commercial schedules reflect the transactions that are needed for
settlement between market areas with a positive net position and those with
a negative net position in any given hour. The physical cross border flows –
shown in Figure 28 - that result from these net positions and the subsequent
intraday trade will depend on where in a market area consumption and
production take place and the capacities in the transmission grid, irrespective
of country borders and also include transit flows and loop flows.
29
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
Market integration and
interconnection flows
European Commercial Cross-Border schedules
2014 TWh/year
2015 TWh/year
0
5.8
GB
0.1
8.2
NL
3.2 3.6
DKW
DKE
1.9 2.1
1.3 1.9
0.9 3.6
0.01
18
0.3
FR
PL
GB
0.7
9.7
DE
BE
16.4 0.1
0.1 1.6
DKE
0.1
0.8
7.7
0.2
16.6
0.9
17.5 1.4
FR
5.4
8.6
Market
integration
CZ
3.8
Balancing
13.3
13.6
AT
5.8 4.2
CH
PL
DE
CZ
9.9
39.1
NL
1.4 6.1
Intraday markets
0.5
BE
4.1
16.9
Consumption
and production
SE
0.8 1.8
2 2.4
Price
developments
NO
SE
DKW
Introduction
Flow Based
NO
0.1 5.5
30
44.9
AT
5.6 10.4
CH
Redispatch
Figure 27:Annual total of commercial Day-ahead cross-border schedules in CWE region and at the German borders in TWh. Source: TenneT, ENTSO-E, BritNed, Swissgrid
Special events
in 2015
Main findings
Market integration and
interconnection flows
European Physical Cross-Border Flows
2014 TWh/year
2015 TWh/year
0.1 5.4
0.01 5.8
SE
GB
0.01
8
NL
1.4 7.9
FR
2
0.3
9.2
24.3
0.01
PL
GB
0.01
8.2
DE
BE
10.7 0.5
1.8
3.6
6
DKW
DKE
1.7 2
0.8 3.1
NL
0.3 12.7
10.7
23.7
0.001
1
FR
0.5
2.5
2.3
Market
integration
CZ
11.5
Balancing
0.8
0.2 14.5
AT
2.7 7.6
CH
PL
DE
CZ
9.3
Intraday markets
0.03
BE
14.4
0.6 10.7
Consumption
and production
SE
0.1 1.1
DKE
2.3
Price
developments
NO
1 1.1
2
Introduction
Flow Based
NO
DKW
31
AT
1.1 12.1
CH
Redispatch
Figure 28:Annual total of physical cross-border flows in CWE region and at the German borders in TWh. Source: TenneT, ENTSO-E, Swissgrid
Special events
in 2015
Main findings
Market integration and
interconnection flows
Belgium net Imports
Particularly Belgium showed an increasing commercial import dependency.
Figure 29 shows the monthly net imports into Belgium. During the first eight
months, a significant import increase can also be seen to be supported
by high production unavailability. Chapter 9 discusses the price peaks
that occurred in Belgium in 2015 in more detail).
TWh
3
32
Introduction
Price
developments
Flow Based
Consumption
and production
2
Intraday markets
1
Market
integration
2014
December
November
October
September
August
July
June
May
April
March
February
January
0
Balancing
2015
Figure 29: Monthly commercial net imports (import – export) from Belgium in 2014 and 2015.
Source: TenneT, ENTSO-E
Redispatch
Special events
in 2015
Main findings
33
Balancing
Introduction
Germany saw a further improvement of the balancing performance of balance-responsible parties in 2015, as is testified by the
reduced volumes of net imbalances; smaller volumes of reserves were activated significantly more often than in the two years
previous, as the number of periods with a small net imbalance increased and the volumes of undersupply decreased. However,
both the Netherlands and Germany saw an increase in the occurrence of more extreme imbalance prices in 2015.
The difference between the imbalance price and the market price can be seen
as the incentive for market parties to be balanced. Typically, one wants this
incentive to increase when the system shows a larger imbalance. Figures 30
and 31 therefore show this price difference11 in relation to the net imbalance in
2015, for the Netherlands and Germany respectively. The graphs also show a
count of the number of imbalance settlement periods12 (ISPs) in which the net
imbalance volume fell within a certain range.
In both countries we see that the differences between the spot prices and
the imbalance prices increase when the net imbalance is larger, reflecting
the scarcity or abundance of energy in the system. However, in Germany
there are also significant differences between the spot and imbalance prices
in situations without significant net volumes of activated reserves
(between -100 MW and +100 MW), this was discussed in more detail
in the TenneT Market Review 2014.
Furthermore, in both countries we see that the net imbalance is more often
positive than negative. Apparently there is a tendency for market parties
to oversupply the system. Figures 32 and 33 show that this tendency has
developed in recent years.
Another effect we can see in Figures 32 and 33 is that the number of ISPs
in which the system was well balanced has increased significantly again in
Germany in 2015, whereas for the Netherlands this has somewhat decreased.
The significant improvement (decrease) of the volumes of reserve activation
can most likely be attributed to further improvement of the actions of balanceresponsible parties, both in forecasting and in real time. The incentive for
balance-responsible parties to improve their actions can likely be attributed to
the improvements in the calculation of the German imbalance price that have
been applied since 1 December 2012 to improve the consistency between
the imbalance price and the market value of energy in the (intraday) spot price.
In Figure 30, the 90th percentile of the price delta in the Netherlands reaches
similar values for a range of different volumes of imbalances. This reflects
bidding behaviour on the merit order list and could be an indication of
limited liquidity.
11
For Germany the Intraday auction price was taken as the relevant market price, whereas for the Netherlands the Day-ahead Market price was used, in the absence
of a liquid intraday market.
12
The imbalance settlement period (ISP) is the period in which imbalances are settled with balance-responsible parties. This period is equal to 15 minutes for both
Germany and the Netherlands.
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
34
Balancing
Introduction
Price
developments
Net Imbalance and Balance Incentives for the Netherlands
€/MWh
Number of ISPs
400
6,000
Flow Based
350
300
4,800
250
200
3,600
Consumption
and production
150
100
2,400
50
Intraday markets
0
1,200
-50
-100
650 <
600 - 650
550 - 600
500 - 550
450 - 500
400 - 450
350 - 400
300 - 350
250 - 300
200 - 250
150 - 200
100 - 150
50 - 100
0 - 50
-50 - 0
-100 - -50
-150 - -100
-200 - -150
-250 - -200
-300 - -250
-350 - -300
-400 - -350
-450 - -400
-500 - -450
-550 - -500
-600 - -550
-650 - -600
< -650
0
Market
integration
< short long >
90% quantile
75% quantile
Balancing
Median
25% quantile
10% quantile
Number of ISPs
Figure 30: Net imbalance clusters (in MW) and differences between the imbalance price and the Day-ahead price in the Netherlands. Source: TenneT, EPEX Spot
Redispatch
Special events
in 2015
Main findings
35
Balancing
Introduction
Price
developments
Net Imbalance and Balance Incentives for Germany
€/MWh
Number of ISPs
400
6,000
Flow Based
350
300
4,800
250
200
3,600
Consumption
and production
150
100
2,400
50
Intraday markets
0
1,200
-50
-100
1400 - =>
1300 - 1400
1200 - 1300
1100 - 1200
1000 - 1100
900 - 1000
800 - 900
700 - 800
600 - 700
500 - 600
400 - 500
300 - 400
200 - 300
100 - 200
0 - 100
-100 - 0
-200 - -100
-300 - -200
-400 - -300
-500 - -400
-600 - -500
-700 - -600
-800 - -700
-900 - -800
-1000 - -900
-1100 - -1000
-1200 - -1100
-1300 - -1200
-1400 - -1300
<= - -1400
0
Market
integration
< short long >
90% quantile
75% quantile
Balancing
Median
25% quantile
10% quantile
Number of ISPs
Figure 31: Net imbalance clusters (in MW) and differences between the imbalance price and the Intraday price in Germany. Source: Regelleistung.net, Energate, TenneT
Redispatch
Special events
in 2015
Main findings
36
Balancing
Introduction
Dutch Imbalance volume distribution development
Price
developments
German Imbalance volume distribution development
6,000
6,000
5,000
5,000
4,000
4,000
Consumption
and production
3,000
2013
2014
2015
Figure 32: Net imbalance volume distribution in the Netherlands by Net imbalance clusters (in MW).
Source: TenneT
2013
2014
1400 - =>
1300 - 1400
1200 - 1300
1100 - 1200
1000 - 1100
800 - 900
900 - 1000
700 - 800
600 - 700
500 - 600
400 - 500
300 - 400
200 - 300
0 - 100
100 - 200
-100 - 0
-200 - -100
-300 - -200
-400 - -300
-500 - -400
-600 - -500
-700 - -600
-800 - -700
-900 - -800
-1000 - -900
-1100 - -1000
-1200 - -1100
-1300 - -1200
< short long >
Volume ranges (MW)
Intraday markets
<= - -1400
650 <
600 - 650
550 - 600
500 - 550
450 - 500
400 - 450
350 - 400
300 - 350
250 - 300
200 - 250
150 - 200
100 - 150
0 - 50
50 - 100
-50 - 0
-100 - -50
-150 - -100
-200 - -150
-250 - -200
-300 - -250
-350 - -300
-400 - -350
0
-450 - -400
0
-500 - -450
1,000
-550 - -500
1,000
-600 - -550
2,000
< -650
2,000
-1400 - -1300
Number of ISPs
3,000
-650 - -600
Number of ISPs
Flow Based
< short long >
Volume ranges (MW)
Market
integration
Balancing
2015
Figure 33: Net imbalance volume distribution in the Netherlands by Net imbalance clusters (in MW).
Source: regelleistung.net
Redispatch
Special events
in 2015
Main findings
37
Balancing
Figures 34 and 35 give insight in the distribution and the development of
difference between the imbalance prices and the market prices in the Netherlands
and Germany respectively. In both countries we see that the number of ISPs
with a large price difference increases. The price differences below -50 €/MWh
and above 50 €/MWh are highlighted as the striped surface in Figures 34 and
35. In the Netherlands the frequency of price differences of below -50 €/MWh
increased for injection and withdrawal with 41.3% and 49.0% respectively, as
for above 50 €/MWh increased with 59.9% and 51.6% respectively. In Germany
the price differences below -50 €/MWh decreased with 13%, whereas the price
differences above 50€/MWh increased with 34%. This means that, for both
Germany and the Netherlands, high differences become more frequent.
Introduction
This is true for both sides: the imbalance prices are more often much lower,
but also more often much higher than the market prices. This trend is shown
in the top 20 highest price differences on both the negative and positive side.
For the Netherlands, the top 20 maximum prices on the negative side clearly
increased by a maximum price of -697 €/MWh and on the positive side
remained the same. For Germany, the top 20 maximum prices clearly increased
on both sides to 6328 €/MWh and -6027 €/MWh.
Finally, for the Netherlands it is relevant to note that the number of ISPs with
two imbalance prices13 increased to 14%. In 2014, this number decreased,
but in 2015 it was back at 2013 levels.
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
Special events
in 2015
13
The Dutch balancing regime applies a different imbalance price for settlement of positive and negative
imbalance for those ISPs where TenneT has dispatched upward regulating bids as well as downward
regulating bids.
Main findings
38
Balancing
Dutch price duration curves imbalance prices
€/MWh
200
Top 20 max
Top 20 max
500
150
Price
developments
German price duration curves imbalance prices
€/MWh
200
Introduction
Flow Based
7,000
6,000
150
450
100
Imbalance - Day ahead price delta’s
Imbalance - Day ahead price delta’s
5,000
400
350
300
250
50
2014
Withdrawal
2014
Injection
2015
Withdrawal
2015
Injection
0
4,000
Consumption
and production
3,000
100
2,000
1,000
0
50
2013
2014
2015
Intraday markets
0
Top 20 min
-50
-350
0
-400
-1,000
-450
-100
-2,000
-100
-500
-3,000
-550
-5,000
-150
-650
-6,000
-700
-7,000
2014
Withdrawal
-200
2014
Injection
2015
Withdrawal
2015
Injection
2013
2014 Injection
2015 Withdrawal
2014
2015
Redispatch
-200
Number of ISPs
2014 Withdrawal
Balancing
-4,000
-600
-150
Market
integration
Top 20 min
-50
Number of ISPs
2015 Injection
Figure 34: Price duration curve difference between imbalance prices and Day-ahead price
for the Netherlands
2013
2014
2015
Figure 35: Price duration curves differences imbalance prices and Day-ahead prices for Germany
Special events
in 2015
Main findings
39
Redispatch
Introduction
The energy transition has a major impact on the system operation of the TSOs. The new renewable generation has different
production patterns and is produced at different locations than the conventional generation it replaces. Redispatch measures,
whereby generators modify their production schedule on request of the TSO, were needed on 90% of the days in 2015,
so have become standard practice in Germany.
Figure 36 shows the development of redispatch volumes per month since 2014
for the German TenneT control zone by type of redispatch measure. The bars
point out how much redispatch has increased in 2015 compared to a year
before, but also the variation in the need for redispatch measures from month
to month.
Also, it shows significant increases in 2015 in the use of multilateral remedial
action (MRA) and countertrade, especially in August and September, and of
the power plants in the grid reserve, especially in November and December.
Both aspects are explained below.
Price
developments
Flow Based
Monthly redispatch TenneT control zone in Germany
Consumption
and production
GWh
3,000
Intraday markets
2,500
2,000
Market
integration
1,500
1,000
Balancing
500
0
1
2
3
4
5
6
7
8
9
10 11 12
1
2
Redispatch
Countertrade
Multilateral Remedial Action
3
4
5
6
7
8
9
10 11 12
Redispatch
2015
2014
Compensation for EisMan
Netzreserve
Special events
in 2015
Figure 36: Monthly redispatch volume by German TenneT control zone in 2014 and 2015
by different types of redispatch measures. Source: TenneT
Main findings
40
Redispatch
Due to wind power representing the highest share of renewable power production
in Germany, and its connection to high north-south transits, redispatch measures
were taken in 90% of the days and therefore became a daily business for the
German transmission system operators. The high level of measures that had to
be taken in November and December can be explained by the high production
of wind energy in these months. Figure 37 shows the correlation between
the level of the wind energy production and the total capacity of the measures
TenneT has to take to ensure security of supply. It clearly shows that higher
wind feed-in leads to a higher need for measures for a given network.
MW
6,000
Flow Based
4,000
2,000
0
Consumption
and production
-2,000
-4,000
-6,000
Intraday markets
-8,000
0
Frequently, the need for measures is so high that virtually no additional measures
would remain, and grid reserves have to be called upon. The network expansions
that are needed to reduce the number of measures are progressing. Still, there is
a mismatch between the pace at which production locations change as a result
of the energy transition and that at which the necessary infrastructure can
be realized.
Price
developments
Redispatch for Wind feed-in power adjustment
Average adjustment of power
The increased need for MRA in August and September is due to the Polish grid
situation in that period. During August, the prolonged heatwave caused a gradual
increase of unscheduled outages of conventional power plants due to the
reduced availability of cooling water. The unavailabilities of conventional power
plants in the Polish control area led to MRA activations instead of redispatch.
Introduction
1
2
3
4
5
6
7
8
9
10
11
12
Wind feed-in
Redispatch (Sink)
Redispatch (Source)
Netzreserve (Source)
Figure 37: Average redispatch versus wind feed-in in the TenneT Germany control area.
Source: TenneT
13
14
GW
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
41
Special events
in 2015
Introduction
Price
developments
Besides the general trends in the Dutch and the German electricity market, we want to highlight three special events that took
place in 2015. These are the solar eclipse, the price peaks witnessed in the Belgian market, and the day on which a record share
of demand was covered by renewable energy.
Consumption
and production
Solar eclipse solar Feed-in Germany March 20th
MW
Intraday markets
25,000
20,000
Market
integration
15,000
10,000
Balancing
5,000
Figure 38: German solar feed-in on March 20th of 2015. Source: EPEX, German TSOs
23:45 - 24:00
22:30 - 22:45
21:15 - 21:30
20:00 - 20:15
18:45 - 19:00
17:30 - 17:45
16:15 - 16:30
15:00 - 15:15
13:45 - 14:00
12:30 - 12:45
11:15 - 11:30
10:00 - 10:15
08:45 - 09:00
07:30 - 07:45
06:15 - 06:30
05:00 - 05:15
03:45 - 04:00
02:45 - 03:00
02:30 - 02:45
01:15 - 01:30
0
00:00 - 00:15
Solar eclipse
This year was marked by a rare event that proved to be a special test case for
the power markets and power system – a solar eclipse on the 20th of March 2015.
Due to the high solar capacities already installed in the German generation stack,
the eclipse led to predictably high gradients of the solar feed-in. Hence, other
power plants had to provide the flexibility to cover these gradients, coordinated
by the power markets. The solar feed-in on that day is depicted in Figure 38,
also showing the maximal gradient, a positive slope of 4.26 GW/15min that
appeared at 11 AM.
Flow Based
Redispatch
Special events
in 2015
Main findings
42
Special events
in 2015
Without a solar eclipse, such gradients could only occur during normal
operation if Germany had an installed solar capacity of 54 GW. In the German
Grid development plan this is foreseen for the year 2025.
This follows from the analysis of historic positive gradients between 2011
and 2013, which were below the gradient observed during the solar eclipse,
and scaling to 4.26 GW per 15 minutes. Using only the 99.9%-quantile of
the historically observed gradients, solar capacities of 77 GW are possible
until such a positive slope can be reached during normal system operations.
This leads to the conclusion that this was an extraordinary event which will
not become a feature of normal operation in the coming few years.
Introduction
Price
developments
Installed Solar
Capacities
Foreseeable
realization
Max
54 GW
2025 (NEP)
99.9%-quantile
77 GW
2050 (EU-Trends)
Nonetheless, the plannability and anticipation of the upcoming solar eclipse
led to much preparation on the part of the system operators and market
participants. Hence, as Figure 39, the expected high ramps led especially to
high prices on the Intraday auction the day before. The lower weighted average
prices during continuous trading on the Intraday market indicate conservative
strategies or lower than expected ramp rates. The solar eclipse was handled
with the available trading options on the intraday market.
Flow Based
Consumption
and production
Intraday markets
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
43
Special events
in 2015
Introduction
Price
developments
Day-ahead and Intraday markets during solar eclipse
Flow Based
GW
€/MWh
25
500
Consumption
and production
400
20
300
Intraday markets
15
200
100
10
0
Market
integration
5
-100
-200
Day-ahead (€/MWh)
ID Auction price (€/MWh)
ID hourly WAP (€/MWh)
ID Quarterly WAP (€/MWh)
13:45 - 14:00
13:30 - 13:45
13:15 - 13:30
13:00 - 13:15
12:45 - 13:00
12:30 - 12:45
12:15 - 12:30
12:00 - 12:15
11:45 - 12:00
11:30 - 11:45
11:15 - 11:30
11:00 - 11:15
10:45 - 11:00
10:30 - 10:45
10:15 - 10:30
10:00 - 10:15
09:45 - 10:00
09:30 - 09:45
09:15 - 09:30
09:00 - 09:15
08:45 - 09:00
08:30 - 08:45
08:15 - 08:30
08:00 - 08:15
07:45 - 08:00
07:30 - 07:45
07:15 - 07:30
07:00 - 07:15
06:45 - 07:00
06:30 - 06:45
06:15 - 06:30
06:00 - 06:15
0
Balancing
Redispatch
Solar feed-in (GW)
Figure 39: Solar Feed-in and Intraday market volumes during the solar eclipse on March 20th. Source: EPEX, German TSOs
Special events
in 2015
Main findings
44
Special events
in 2015
In preparation and to be able to ensure system stability, the German TSOs
contracted additional reserve capacities ahead of the solar eclipse, as shown
in Figure 40. These included higher secondary control reserve capacities for
Introduction
Price
developments
the whole day due to the product length. Additional tertiary control reserve
capacities were auctioned only for the critical product from 8 to 12 AM and
a minor surplus for the product from 12 to 16 PM.
Flow Based
Contracted Control Reserves during solar eclipse
Consumption
and production
MW
8,000
6,000
Intraday markets
4,000
2,000
Market
integration
0
2,000
Balancing
4,000
19-03-2015
FCR pos
aFFR pos (HT)
mFFR pos
20-03-15
FCR neg
aFFR neg (HT)
mFFR neg
20-24
16-20
12-16
08-12
04-08
00-04
20-24
16-20
12-16
08-12
04-08
00-04
20-24
16-20
12-16
08-12
04-08
8,000
00-04
6,000
Redispatch
21-03-15
Special events
in 2015
Figure 40: Contracted control reserve capacities during solar eclipse in Germany on March 20th. Source: regelleistung.net
Main findings
Special events
in 2015
The higher than usual auction volumes also led to higher prices as depicted
in Figure 41. The highest impact could be observed on the prices for negative
secondary control and positive tertiary control reserve.
Prices for Contracted Control Reserves during solar eclipse
Belgian price peaks
Throughout 2015, Belgium faced high rates of unavailability, both planned
and unplanned, of its conventional generation stack, as can be observed
in Figure 42.
45
Introduction
Price
developments
Flow Based
Consumption
and production
€/MW
600
Intraday markets
500
400
Market
integration
300
200
100
Balancing
19-03-15
aFRR neg (HT)
20-03-15
aFRR pos (HT)
mFRR neg
20_24
16_20
12_16
08_12
04_08
00_04
20_24
16_20
12_16
08_12
04_08
00_04
20_24
16_20
12_16
08_12
04_08
00_04
0
21-03-15
mFRR pos
Redispatch
Figure 41: Prices for contracted control reserve prices during solar eclipse in Germany on
March 20th. Source: regelleistung.net
Special events
in 2015
Main findings
46
Special events
in 2015
Introduction
Price
developments
Unavailability Belgium Power Plants
Flow Based
GW/day
6
Consumption
and production
5
Intraday markets
4
Market
integration
3
2
Balancing
1
Redispatch
Planned nuclear
Planned natural gas
Planned waste
Planned other
Unplanned nuclear
Unplanned natural gas
Unplanned waste
Unplanned other
365
358
351
344
337
330
323
316
309
302
295
288
281
274
267
260
253
246
239
232
225
218
211
204
197
190
183
176
169
162
155
148
141
134
127
120
113
99
106
92
85
78
71
64
57
50
43
36
29
22
8
15
1
0
Special events
in 2015
Figure 42: Planned and unplanned unavailabilities of Belgian power plants. Sources: ELIA, Federal agency for nuclear
Main findings
Special events
in 2015
This led to some serious adequacy concerns for the Belgian system, which
in turn also led to price peaks. Especially during the end of September and
middle of October, several planned power plant outages plus planned
infrastructure work on the Belgian high-voltage grid led to a high unavailability
of generation and transmission capacities and a high volume of international
flows through the Belgian system. International redispatch measures were
needed on several days to relieve the transmission grid and to maintain
system security.
An exemplary situation is illustrated in Figure 43 for two days in the middle of
October in 2015. During those two days the spot price was consistently very
high during the day, with the highest peaks reaching 450 €/MWh. This led to
most of the thermal generation stack being utilized, even oil-fired power plants,
whilst importing at maximum capacity throughout the period. The very high
prices represented a generation adequacy challenge in Belgium for those days,
but in cooperation with its neighbouring countries, Belgium managed to
prevent any acute adequacy issues from materializing.
Analysis by the TSOs showed that these exceptional price spikes would
not have occurred under Flow Based Plain Market Coupling. Unlike the
implemented Flow Based Intuitive this method does not have the restriction
that only allows an outcome in which commercial flows only go from a market
area with lower prices to market areas with higher prices. The reason for the
introduction of this restriction was that it was considered counter-intuitive
by stakeholders to have commercial flows from higher to lower priced areas.
In this particular event this restricted further imports into Belgium, leading
to higher prices.
47
Introduction
Price
developments
Flow Based
Consumption
and production
Intraday markets
This exemplary situation, as well as the price levels and high degree of
unavailability of generation capacities throughout the year, shows how much
planned and unplanned outages are an impact factor on prices in Belgium
in the current tight situation.
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
48
Special events
in 2015
Introduction
Price
developments
Price peaks in Belgium on October 15th and 16th
Flow Based
€/MWh
GWh/h
500
12
Consumption
and production
10
400
8
Intraday markets
300
6
200
4
100
Market
integration
2
0
0
15-10-15
00:00
15-10-15
04:00
15-10-15
08:00
15-10-15
12:00
15-10-15
16:00
15-10-15
20:00
Nuclear
Net Imports
Coal
Other
Gas
Fuel
Hydro
Wind
16-10-15
00:00
16-10-15
04:00
16-10-15
08:00
Day-ahead price
16-10-15
12:00
16-10-15
16:00
Balancing
16-10-15
20:00
Redispatch
Figure 43: Day-ahead generation schedule and Day-ahead prices during 15th and 15th of October 2015 in Belgium. Sources: Elia, TenneT
Special events
in 2015
Main findings
Main findings
Average wholesale prices in Germany and the Netherlands decreased in 2015 following downward commodity prices
for coal and gas and increased generation from solar and wind.
In the course of 2015, the marginal generation costs of gas- and coal-fired
generation converged. Typically this contributes to the convergence
of the wholesale prices for electricity in Germany and the Netherlands.
Additionally, the increasing share of coal-fired generation at the expense of
gas-fired generation in the Netherlands contributes to price convergence in
hours where coal-fired generation becomes price-setting in both countries.
The introduction of Flow Based market coupling on May 20th, further increased
the interconnection capacity that can be made available if the CWE price
differences are large.
However, strong growth of generation from wind energy in Germany and
sharply increased exports from the Netherlands to Belgium caused the
percentage of hours of full price convergence to remain similar to that in 2014.
In Germany the generation stack and the generation itself is under a long-term
transition. So in 2015, the ongoing trends led again to the shutdown of one
nuclear and some conventional units, whereas additional capacities of
renewable generation, especially offshore wind turbines, entered the market.
Additionally, the increase of solar capacities amplified the merit order effect
around noon. Thus, the pressure on income for conventional power plants
remains unabated.
49
Introduction
Price
developments
Flow Based
In the Netherlands, the generation from renewables also grew. The production
from solar panels, although still small, now becomes significant at the system
level. Despite its growth, the gap between the Netherlands and Germany grew
wider. Where the German targets for 2020 seem within reach, the Netherlands
will need a very steep increase to reach its goals.
Consumption
and production
Intraday markets are becoming increasingly important for balancing ramp rates
and forecasts in Germany. Supported by the introduction of Intraday auctions,
intraday volumes grew significantly with the growing need to balance renewables.
Intraday markets
The balancing performance in Germany improved, especially the negative
net imbalances (where the system is 'short') decreased. Imbalance prices
in the Netherlands and Germany in 2015 were more extreme than in 2014.
The number and volume of necessary redispatch measures in Germany
increased sharply during 2015. The need for redispatch is strongly correlated
with the amount of wind energy production for which there are not yet sufficient
north-south transmission capacities. Redispatch to ensure system stability has
become a daily necessity in Germany.
Market
integration
Balancing
Redispatch
Special events
in 2015
Main findings
50
Colophon
This Market Review is a publication of TenneT
Contact
Address for visitors
TenneT Holding B.V. and TenneT TSO B.V.
Utrechtseweg 310, Arnhem, the Netherlands
T: +31 (0)26 - 373 11 11
Postal address
P.O. Box 718
6800 AS Arnhem, the Netherlands
Corporate Communications Department
T: +31 (0)26 - 373 26 00
E: [email protected]
W: www.tennet.eu
Project management
Erik van der Hoofd (TenneT)
Disclaimer
Under no circumstances shall TenneT Holding BV nor its subsidiaries, in
particular TenneT TSO B.V. and/or TenneT TSO GmbH, hereinafter “TenneT”,
be liable for any claims, penalties, losses or damages resulting from,
or connected to the use of (the information in) this publication. The information
in this publication is presented “as is”. TenneT makes no warranties or
representations, whether express or implied, about the information contained
in this publication. In particular, TenneT is not liable for information that is not
accurate, up-to-date, comprehensive, verified, or complete. TenneT expressly
disclaims all liability for claims, penalties, losses or damages (whether arising
in contract, tort, or otherwise), that arise after the use of, or reliance upon,
any information and material in this publication.
Price
developments
Flow Based
Consumption
and production
Intraday markets
Market
integration
Scientific Supervision
Prof. Albert Moser (IAEW)
Project team
TenneT
Esther Bos
Tobias Frohmajer
Anton Tijdink
Berno Veldkamp
Anne Martin van der Wal
Introduction
Balancing
IAEW
Fabian Grote
Mihail Ketov
Denis vom Stein
Redispatch
Special events
in 2015
Main findings
TenneT is a leading European electricity transmission system operator (TSO)
with its main activities in the Netherlands and Germany. With approximately
22,000 kilometres of high-voltage connections we ensure a secure supply of
electricity to 41 million end-users. We employ approximately 3,000 people,
have a turnover of EUR 3.3 billion and an asset value totalling EUR 15.4 billion.
TenneT is one of Europe’s major investors in national and cross-border grid
connections on land and at sea, bringing together the Northwest European
energy markets and enabling the energy transition. We take every effort to
meet the needs of society by being responsible, engaged and connected.
Taking power further.
© TenneT – April 2016