• No se han encontrado resultados

In order to allow for a realistic economic assessment of the different charging coordination objectives, empirical price data from the German energy wholesale market, the European Energy Exchange (EEX)2was collected for 2007 and 2009.

The price data includes all hourly reference (i.e. mean) prices of the intraday market of the respective period. Since the simulation employs a 15 minute time resolution the prices during four intervals were set to the corresponding empir-ical value, adapted by a scaling factor which will be explained in more detail in the respective sections (cf. Sections 4.2.1, 4.3.2). If the optimization objective is not only determined by the economic implications, but by the higher utilization of fluctuating sources, the prices provide the basis to compare the costs resulting

2In the meantime European Power Exchange (EPEX):https://www.epexspot.com/en/

Table 3.3: Summary statistics of the employed intraday wholesale energy price time se-ries.

[EUR / MWh] Min. 1st Qu. Median Mean CV 3rd Qu. Max.

EEX (2007) 3.69 22.00 31.69 38.21 0.73 46.62 601.10 EEX (2009) -648.60 28.77 37.86 39.02 0.63 48.34 173.70

from the different strategies.

Table 3.3 provides an overview of the descriptive statistics of the empirical price data. One of the main differences between 2007 and 2009 data is the fact that starting from 2008, negative prices were introduced in the market in order to allow for correct economic signals at times of high renewable generation and low load. Negative prices occurred in particular in winter and transition weeks when wind-power contributed a high total share, and load was low due to holidays or week-ends. Due to the grid topology and increasingly problematic transmission grid bottlenecks (Ilg et al., 2012), so called must-run units that locally stabilize the power grid have to bid negative prices in order to be allocated and allow for a secure operation of the power system from a technical perspective. These negative prices provide incentives for flexible demand to shift its consumption and even be compensated for it. For storage devices that operate on an arbitrage strategy this instrument can be very profitable.

The price data was clustered in different data sets based on the TRY (Test Ref-erence Year) climatic day type conditions in order to allow for a better charac-terization of general patterns in the price levels. The TRY day types are dis-tinguished mainly by the average temperature, leading to three main groups encompassing winter, summer and transition days and are also employed to as-sess the demand for thermal energy requirements (DWD, 2004). Winter days have an average temperature of less than 5C, transition days between 5−15C and summer days above 15C. This classification enables a better detection of seasonal price patterns that depend on intermittent generation and in particular wind-power.

Figure 3.7 shows the price variation for summer, winter and transition weeks for the year 2007. It can be observed, that the variation, and in particular the extreme values, are less prominent in the summer weeks. Transition weeks have more outliers than the other week and day types. Winter weeks in turn have an overall higher price level and the highest range of price variation, often in between 25 - 100 EUR/MWh for the year 2007. In extreme cases the wholesale energy price reached the level of 600 EUR/MWh, a value more than ten times

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(a) Hourly price variation in summer weeks

●●

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(b) Hourly price variation in transition weeks

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(c) Hourly price variation in winter weeks

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(d) Hourly price variation in all weeks

Figure 3.7: Variation of wholesale intraday electricity prices for the year 2007, distin-guished by TRY day type, (EEX, 2007).

higher than even expensive hours served by peak generators. When compared to 2009 the overall price level is slightly lower, but the tendency of the outliers is less strong in the positive price direction. Negative prices in turn also reached levels of more than 600 EUR/MWh in 2009, which favors flexible loads that have the possibility to take advantage of these situations.

Figure 3.8 shows the different week types for 2009 and the respective hourly price variation. Overall a similar general price pattern can be observed as for 2007, summer weeks have a slightly lower price level, whereas transition weeks have the highest variation and winter weeks have a higher overall price level with a notable variation bandwidth. Negative prices mostly occur on transition and winter weekends which reflects the low load - high wind generation

situa-●

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(a) Hourly price variation in summer weeks

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(b) Hourly price variation in transition weeks

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(c) Hourly price variation in winter weeks

Day of the week

[EUR / MWh]

Mon Tue Wed Thu Fri Sat Sun Mon

(d) Hourly price variation in all weeks

Figure 3.8: Variation of wholesale intraday electricity prices for the year 2009, distin-guished by TRY day type, (EEX, 2009).

tions sketched above. In the analyses performed in Chapters 4 and 5 the prices will partly be scaled to correspond to the end-consumer price level. This proce-dure has implicit assumptions about the possible dimension of the variable price and the absolute spread which will be addressed in the respective context of the analysis.

Documento similar