Wind Energy Integration into the Grid - Capacity Credit

From energypedia
Revision as of 17:55, 22 July 2011 by ***** (***** | *****)

Power plants using renewable fuels (e.g. power production based on biomass) generally allow scheduling of electricity production, as their primary source of energy can be stored and transported. Within an electricity supply system their use can be planned like any gas- or coal power plant. The process of schedule-development in an electricity supply system is called dispatch. Thus the integration of these dispatchable renewable energy plants does not cause significant changes in the system[1].

In contrast to this, wind turbines or wind parks are non-dispatchable sources of electricity production: Wind velocity and the related amount of electricity generated is only predictable by meteorological methods (with a limited certainty), but of course there is no possibility to influence the availability of the renewable resources. Electricity production by wind turbines is determined by the available wind velocity and the electricity supply system has to adapt to the characteristics of wind energy, in case the potentials of this renewable resource should be used effectively. Efficient integration of wind energy into an existing power system thus requires an advanced management of the conventional power plant[2]. This article focusses on the effects of wind energy integration on the reliability of an electricity supply system. The Capacity Credit is described as a way to quantify the reliability of electricity generation by wind turbines.

Wind variability

The variability of electricity production by wind turbines is generally due to changes in wind speed over time. The wind variability can be described on several different time scales:

  • Variations of wind potentials from year to year
  • Seasonal variations of average wind speeds; in Germany average wind velocity during winter months is usually twice as high as the wind speed in the summer[3].
  • Changes in weather cause wind variability on the time-scale of weeks and several days.
  • Within 24 hours a significant difference of wind speeds between daytime and night can be observed at most sites. Depending very strongly on climatic conditions of the site, variations during daytime could show characteristic patterns
  • Wind speed varies from hour to hour and also from one minute to the next
  • Variations on the time-scale of seconds are described as turbulence[4].

The extent of variations on the listed time-scales differs considerably. The largest share of the total variability of wind speed is contributed by variations within 3-5 days, because during this period of time significant changes in weather can occur. Concerning the operation of an electricity supply system, these changes can be regarded as slow. The second major contribution to overall variability is induced by turbulence, which may cause more serious problem for the management of a supply system. If wind turbines are aggregated in a wind park, this has a balancing effect on turbulence effects on electricity production. 

Within the time-frame of 10 minutes to one hour frequency and extent of variations are relatively small. This so-called spectral gap is a very advantegous characteristic of wind speed distribution: If the the variations within this period of time had been considerable, this would have result in larger complications for wind energy integration into the electricity supply system[5].

Variability of electricity production

The characteristics of electricity production are influenced by the chosen turbine type to a considerable extent. As the description of these influences requires detailled technical explaination of the turbine types, this article abstracts from these differences. The focus is rather set on the general 'transmission' of wind variability into variability of the electricity production.

Variability in electricity production may be described concerning three main characteristics[6]:

  • Full load hours respectively annual energy yield,
  • periodical patterns of electricity generation by wind turbines,
  • Volatility of electricity generation by wind turbines

Hours of full load is a term describing the number of hours the wind turbines have been operated at their rated capacity during one year. The annual energy yield is usually given in MWh. The term 'periodical patterns of electricity generation' describes any patterns in wind variability, which can be observed with regularity irrespective the frequently changes on other time-scales. As an example the regular seasonal changes in wind speed distribution cause similar variations in electricity generation[7]. The term volatility is used for changes within small time-frames from minutes to hours.

The 'translation' of wind variability into the variability of electricity generation, fundamentally depends on the following factors:

  • Characteristics of the extraction of wind energy by modern wind turbine types
  • site selection, geographical distribution of wind turbines 
  • wind turbine operation: in a wind park (aggregated electricity generation) or as a sole application

The power – extractable by a wind turbine at a certain wind speed – is given by the following function:

 

where  is used for air density, A indicates the area swept by the turbine rotor, V describes wind velocity and Cp is the power coefficient describing the efficiency of wind energy conversion by the turbine. As electricity production changes proportional to the cube of wind speed, variability of wind speed results in significantly larger variations in electricity generation.

Depending on the design of modern wind turbines, this is only valid for a certain range of wind speed variations:

Very low wind speeds do not contain sufficient power to operate wind turbines. Typically modern wind turbines have a so-called Cut-in-wind speed Vci of 3,5 m/s and reach their maximum power at a rated wind speed Vn. Many turbines have a Vn-value of 15 m/s. Above this wind speed the operation of the wind turbines are regulated by aerodynamic control mechanisms to limit rotor speed and the related output. As a result variations of power output at wind speeds between 15 m/s und 25 m/s are low. The wind velocity 25 m/s (equivalent to wind force 10 Beaufort) is often determined as Cut-out-Wind speed Vco[8].

Wind conditions on site of course are the foundation for the expected electricity production and its variability. Besides average wind speed and wind speed distribution of the site, the roughness of the surface is essential for expected wind energy production. Any obstacles like forrests, infrastructure or even fences increase friction of wind on the surface as well as turbulence[9]

In many cases wind turbines are integrated into the electricity grid as aggregated wind parks (instead of integrating single wind turbines). Aggregating the output of several wind turbines in a wind park reduces the variability induced into the grid significantly. This effect is based on the spatial distribution of the turbines: If wind turbines are installed distributed over a large area, turbulences affect energy production at the different turbines in an uncorrelated way and variations on a time-scale of seconds are balanced out to a considerable share.

Impact of variability in electricity production on the operation of the electricity supply system

Integration of wind energy into an electricity supply system means introducing some additional uncertainty. In a system containing only convential power plants (e.g. coal, gas) the main sources of variability are misestimations of electricity demand and power plant outages. For both factors a probability can be assumed and is used in calculation of the necessary reserve capacity, which has to be retained to balance out mismatches between electricity demand and supply.
The impact of wind energy integration on the system operation essentialy depends on the policy framework conditions for wind energy in the country. If the policy framework contains a purchase guarantee for electricity generated by wind turbines, wind energy is fed into the grid whenever sufficient wind speeds occur. The impact of grid integration can be described roughly as follows:

As a part of electricity demand is covered by wind energy, the residual demand, which has to be covered by conventional plants, declines. This residual demand is usually described as the residual load. Thus a lower number of conventional plants has to be run to cover this residual load. As the price of electricity is set within the merit-order by the power plant with the highest marginal costs, a decrease of residual load results in a reduction of the electricity price. The impacts of wind energy generation on the electricity supply system can be divided in to main aspects:

  • Impacts on system efficiency
  • Impacts on reliability of the electricity supply system

Impacts on system efficiency

The decline of residual load caused by fed in wind energy depends on the prevailing wind speed and the related electricity production of integrated wind turbines. Besides reduction the variability of the residual load increases considerably, the availability of wind energy varies with wind speed. As a result plants for covering peak loads (e.g. gas turbines in Germany) are used less often and generally power plants are operated in part load more frequently. Output of conventional power plants has to be change more frequently and thus the required flexibility of the plants is increased.

Wind turbines replace conventional plants within the electricity supply and reduce consumption of fossil fuels and the emission of greenhouse-gases. The impacts on operation of conventional plants described above reduce this positive effect, because overall system-efficiency declines: Plants in part load operation have a lower efficiency and frequently start-up and shut down of plants causes additional fuel consumption for the related technical processes[10].
In case a large capacity of wind energy is already integrated into the electricity supply system, surpluses in supply can occur in case of high wind speeds. If the conventional power plants are already operated at minimum load and demand is generally low, it is possible, that electricity generated by wind turbines can not be fed into the grid without causing serious mismatches between demand and supply. In this case electricity production by wind turbines has to be regulated or shut-down[11].

Impacts on reliability of the electricity supply system

The reliability of an electricity supply system can be measured by the probability to cover the annual peak load by use of the installed power plant capacity in the system. More precise the risk of mismatch between supply and demand during peak load is evaluated. To calculate this probability the secured capacity available in a system must be known. To calculate the secured capacity, probability of plant outages is applied and the corresponding capacity is subtracted. Furthermore outages caused by planned revisions are taken into account.

To ensure a reliable supply system, the available secured capacity must be able to cover the annual peak load. Thus the installed capacity must include a surplus corresponding to the difference between installed and secured capacity.
Caused by their significantly higher variability of electricity production, wind turbines are considered as plants with a considerably lower secured capacity compared to conventional fossil fuel plants. Due to this fact, it is important to analyse how much conventional capacity can be replaced by wind energy within the system, without influencing system reliability.

By means of the so-called Capacity Credit (CC) conventional capacity and wind turbine capacity can be compared concerning impacts on system reliability.

The Capacity Credit

The calculation of the CC (given in MW) is based on the probability of wind turbine availability during peak load periods. In this way the share of wind turbine capacity, which can be considered as secured capacity for covering peak load can be determined. As wind turbines are a non-dispatchable source of electricity production, the resulting CC for wind turbines is subtantially lower than the secured capacity of conventional power plants. Thus to ensure system reliability in a system with a large wind energy share, a large surplus capacity must be installed within the system.
Versorgungssicherheit in einem System mit einem großen Anteil integrierter WEA eine wesentlich
höhere installierte Kapazität im System vorhanden sein muss (Gross et al, 2006,
S.23).
Am besten lässt sich dies an einem Beispiel nachvollziehen: In einem System mit 70 GW
jährlicher Spitzennachfrage und 350 TWh jährlichem Energiebedarf sollen 10% des Stroms
– 35 TWh – durch WEA erzeugt werden. Legt man die CF der Anlagen zu Grunde ergibt
sich folgende Rechnung: Unter der Annahme eines durchschnittlichen CF von 0.3 für
WEA sind 35000 GWh/8760 h * 0.3 = 13 GW Kapazität notwendig, um 35 TWh in einem
Jahr zu erzeugen 10. Soll die gleiche Strommenge durch konventionelle Kraftwerke erzeugt
werden, so wäre dafür – bei einem CF von 0.75 – eine Kapazität von 5,3 GW notwendig
(Freris und Infield, 2008).
Für die Einhaltung der Versorgungssicherheit ist in diesem Beispiel aufgrund von Revisionen
und Ausfallwahrscheinlichkeiten eine installierte Kapazität von 84 GW notwendig,
die die Spitzennachfrage um 20% übersteigt. Nun stellt sich die Frage, ob die Defizitniveaus
eingehalten werden können, wenn 5,3 GW konventioneller Kapazität direkt durch 13
GW WEA-Kapazität ersetzt werden. Dies ist nicht der Fall: Der CF gibt nur die durchschnittlich
Verfügbarkeit der WEA über das Jahr nicht jedoch die Wahrscheinlichkeit der
Verfügbarkeit zur Spitzenlastzeiten an. Diese durch den CC abgebildete Wahrscheinlich-

keit bedingt, dass ein noch geringerer Anteil des konventionellen Kraftwerksparks durch
WEA ersetzt werden kann. Im Beispiel können nur 3 GW direkt ersetzt werden und ein
Anteil von 2,3 GW der konventionellen Kapazität muss als Reservekapazität oder ’Back
up’ für die Gewährleistung der Verlässlichkeit im System verbleiben.11 Der CC der WEA
hat in diesem Beispiel einen Wert von 0.23 (Freris und Infield, 2008, S.79f).
In der Berechnung des CC mittels statistischer Verfahren nehmen folgende äußere Faktoren
Einfluss auf den Wert:
• Zeitpunkt und Dauer der Spitzennachfrage, sowie die damit verknüpfte Korrelation
zwischen Spitzenlast und Output der WEA. Eine positive Korrelation steigert den
CC, während eine negative Korrelation senkend wirkt. Für den Output der WEA
kann für Deutschland allgemein eine positive jahreszeitliche Korrelation mit der
Nachfrage angenommen werden, da die Jahreshöchstlast im Winter auftritt, wenn
durch durchschnittlich höhere Windgeschwindigkeiten auch eine bessere Verfügbarkeit
der Stromerzeugung aus WEA vorliegt.
• Die Variabilität der Nachfrage während Spitzenlastzeiten. Genauer wird die erwartete
Nachfrage, sowie die Spanne möglicher Nachfragelevel während der Spitzenlastzeit
berücksichtigt.
• Der erwartete Output, sowie die Spanne möglicher Outputlevel aus konventionellen
Kraftwerken.
• Die Spanne der möglichen Outputs aus WEA während Spitzenlastzeiten. Prinzipiell
kann der Output aus WEA zwischen Werten nahe 0 bis zu 100% der installierten
Kapazität variieren. Für die Berechnung ist sowohl der erwartete Wert des Outputs
zur Spitzenlastzeit, als auch die Spanne der möglichen Outputs der WEA für die
betreffende Zeit, notwendig. Diese Spanne ist die Varianz der möglichen Outputs.
Allgemein führt eine Verringerung der Varianz zu einer Erhöhung des CC. Wie bereits
oben beschrieben, nimmt die Varianz mit der räumlichen Verteilung der WEA
und der aggregierten Nutzung im System ab und ermöglicht so einen höheren CC.
Der im Beispiel von Gross et al (2006) genannte CC von 0.23 (oder 230 kW pro MW) ist
ein für diesen Parameter sehr hoher Wert. In der Netzstudie I der Dena (2005) wird dieser
Wert für den in Deutschland in 2003 installierten WEA-Anteil mit 7-9% beziffert.
Der CC verringert sich mit zuhnehmender Integration von WEA in ein System. Je höher
der Anteil an Stromerzeugung aus WEA ist, desto geringer fällt der spezifische Beitrag
pro MW installierter WEA-Kapazität aus, den die Anlagen zur Deckung der Spitzenlast
beitragen können (Freris und Infield, 2008, S.74f.).
Die Dena (2005) hat für die Reduzierung des CC eine Modellrechnung durchgeführt, in

der der CC in Abhängigkeit von der Erhöhung der WEA-Kapazität von 1-50GW im EVS
abgebildet wird. Dabei werden Effekte berücksichtigt, die der Minderung des CC entgegen
wirken: Nach 2003 setzt eine Verbesserung der räumlichen Verteilung und gleichzeitig eine
verstärkte Aggregierung der WEA ein, die die Varianz des Outputs des WEA-Kollektivs
verringert und den CC erhöht. Im Besonderen gewinnt die – durch konstantere Windverhältnisse
bedingte – höhere Auslastung der Offshore-Windenergieanlagen an Bedeutung.
Die Darstellung (Abbildung 12) der Dena enthält daher einen Vergleich des Verlaufs des
CC unter der Annahme zweier unterschiedlicher Verteilungen/Aggregierungen für die Jahre
2003 und 2020. Die Darstellung lässt deutlich erkennen, dass der positive Effekt der
Abbildung 12: Abnahme des CC in Abhängigkeit von der installierten WEA-Kapazität
für zwei unterschiedliche räumliche Verteilungen des WEA-Kollektivs in 2003 und 2020.
Quelle: Dena (2005)
verbesserten Verteilung und Aggregierung durch die generelle Abnahme des CC dominiert
wird und für große Mengen installierter WEA-Kapazität mit einem deutlich gesenktem
CC gerechnet werden muss (Dena, 2005). Die notwendige Backup-Kapazität im EVS steigt
daher mit zunehmender Integration überproportional an.

References

  1. Gatzen C (2008) The Economics of Power Storage - Theory and Empirical Analysis for Central Europe, Schriften des Energiewirtschaftlichen Instituts zu Köln, vol 63. Oldenbourg Industrieverlag
  2. Gatzen C (2008) The Economics of Power Storage - Theory and Empirical Analysis for Central Europe, Schriften des Energiewirtschaftlichen Instituts zu Köln, vol 63. Oldenbourg Industrieverlag
  3. Jarass L, Obermair G, Voigt W (2009) Windenergie – Zuverlässige Integration in die Energieversorgung. Springer
  4. Freris L, Infield D (2008) Renewable energy in power systems. John Wiley & Sons, Ltd
  5. Freris L, Infield D (2008) Renewable energy in power systems. John Wiley & Sons, Ltd
  6. Dena (2005) Konzept für eine stufenweise Entwicklung des Stromnetzes in DeutschlandfckLRzur Anbindung und Integration von Windkraftanlagen Onshore und Offshore unter BerücksichtigungfckLRder Erzeugungs- und Kraftwerksentwicklungen sowie der erforderlichenfckLRRegelleistung. In: Energiewirtschaftliche Planung für die Netzintegration von WindenergiefckLRin Deutschland an Land und Offshore bis zum Jahr 2020 - Netzstudie I, DeutschefckLREnergie Agentur
  7. Dena (2005) Konzept für eine stufenweise Entwicklung des Stromnetzes in DeutschlandfckLRzur Anbindung und Integration von Windkraftanlagen Onshore und Offshore unter BerücksichtigungfckLRder Erzeugungs- und Kraftwerksentwicklungen sowie der erforderlichenfckLRRegelleistung. In: Energiewirtschaftliche Planung für die Netzintegration von WindenergiefckLRin Deutschland an Land und Offshore bis zum Jahr 2020 - Netzstudie I, DeutschefckLREnergie Agentur
  8. Freris L, Infield D (2008) Renewable energy in power systems. John Wiley & Sons, Ltd
  9. Dena (2005) Konzept für eine stufenweise Entwicklung des Stromnetzes in DeutschlandfckLRzur Anbindung und Integration von Windkraftanlagen Onshore und Offshore unter BerücksichtigungfckLRder Erzeugungs- und Kraftwerksentwicklungen sowie der erforderlichenfckLRRegelleistung. In: Energiewirtschaftliche Planung für die Netzintegration von WindenergiefckLRin Deutschland an Land und Offshore bis zum Jahr 2020 - Netzstudie I, DeutschefckLREnergie Agentur
  10. Gross R, Heptonstall P, Anderson D, Green T, Leach M, Skea J (2006) The costs andfckLRimpacts of intermittency: An assessment of the evidence on the costs and impacts offckLRintermittent generation on the British electricity network. UK Energy Research Center
  11. Gross R, Heptonstall P, Anderson D, Green T, Leach M, Skea J (2006) The costs andfckLRimpacts of intermittency: An assessment of the evidence on the costs and impacts offckLRintermittent generation on the British electricity network. UK Energy Research Center