Wind turbine participation on Control Reserve Markets

Wind turbine participation onControl Reserve Markets

EWEA ConferenceVienna, 6th February 2013

Malte Jansen (M.Eng.)Fraunhofer IWES, Kasselmalte.jansen@iwes.fraunhofer.de

PlayerManufacturer

RegulatorTSO

European countries

Supplier

Research Institutions

3

Project: „Regelenergie durch Windkraftanlagen“

Project Partners: Fraunhofer IWES (Project Leader) ENERCON GmbH (WT Manufacturer) Energiequelle GmbH (WF Operator) Amprion GmbH (TSO) TenneT GmbH (TSO)Project Period: 05/2012 – 04/2014Aims: Proof Procedure Placement of offers Unit control

10 MW

20 MW

30 MW

40 MW

14:00 16:00 18:00 20:0012:00

Activation Delivery Deactivation

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Time

Pow

er (

norm

aliz

ed)

Actual Feed-InForecast: 95%Forecast: 99%Forecast: 99.9%Forecast: 99.99%Forecast: 99.999%

09-Nov 16:45 11-Nov 15:15 13-Nov 13:45 15-Nov 12:15 17-Nov 10:45 19-Nov 09:15

OFFERING CONTROL RESERVE

Offer (I) - Security level

Reliability of 100%

required

Not possible for

wind farms

Reliability level of 99.994% used in the

project(TSO recommendation)

Use of probabilistic wind power forecasts

Security levels:95 %99 %99.9 % 99.99 %

actual feed-in

Prob

abili

ty

Power

6

Offer (II)- German pool of wind farms at 99.994% reliability

7

Offer (III) - Pooling of wind farms with conventional generation

Use of probabilistic forecast

Probabilistic forecast error is convoluted Wind farms Conventional generation

Offer = Power at a security level of 99.994%

Pool offering control reserve @ 99.994%

8

Offer (IV) - Effects of Pooling

0 20 40 60 80 100 120 140 160 180 2000

100

200

300

400

500

600

700

Time [1/4 hours]

Max

imum

Offe

r [M

W]

Pool Wind FarmsPool GasturbinesSum of both PoolsJoint offering (convoluted)

PROOF MECHANISM

10

0

Proof Method (I) – Balance control mechanism

Time [min]15 30

Pow

er o

utpu

t [M

W]

delivery of negative control

reserve

Currently applied TWENTIES Project Demo#1 (Spain) Regulation in Great Britain and Denmark

Feed-In

ForecastProb. Forecast @ x%

Available active power

Forecast

Feed-In

Prob. Forecast @ x%

11

0

Proof Method (II) – Available active power mechanism

Time [min]15 30

Forecast

Pow

er o

utpu

t [M

W]

Prob. Forecast @ x%

delivery of negative control

reserveForecast

Feed-InProb. Forecast @ x%

Field test in the project „Regelenergie durch Windkraftanlagen“ Regulation in the Irish and Danish Grid Code Considered as proof method in Great Britain

Available active power

Why available active power?

Not opinion of all project partners!

Balancing effects between

balancing groups

Shortterm forecast error of wind farms is balanced by

control reserve

Planned power production =

available active power

0

Forecast

Available active power

Feed-In

Pow

er o

utpu

t [P]

Time [min]15 30

delivery of negative control reserve

Prob. Forecast @ x%Prob. Forecast @ x%

Forecast

Energy Losses with the balance control mechanism – German pool of wind farms

95% 99% 99.8% 99.9% 99.99% 99.994% 99.999%

Lost energy with offers under schedule proof mechanism (German Pool)

Security level

Lost

ene

rgy

per

hour

off

ered

[M

Wh\

h_off

ered

]

0

200

400

600

800

1000

1200

24 Hours4 Hours1 Hour

Capacity prices with the balance control mechanism – German pool of wind farms

95% 99% 99.8% 99.9% 99.99% 99.994% 99.999%0

10

20

30

40

50

60

70

80Capacity Prices under the schedule proof mechanism (German Pool)

24 Hours4 Hours1 Hour

Security level

Capa

city

Pri

ces

[€/M

W/h

]

Capacity costs of control reserve under the available active power mechanism is 0 €/MW/h

15

Comparison of cost saving potentials at the tertiary control reserve market

Cost-saving potential: Available active power

Increased cost efficiency compared to balance control

Cost reduction +212% @ 95% +404% @ 99.999%

Higher market share of wind farms

Security Level

Cost

Red

uctio

n [€

]

95% 99%99.8%

99.9%99.99%

99.994%

99.999%

0

2

4

6

8

10

12

14

16x 106

Available active powerBalance control

1h Product length

POTENTIALS

17

Secure intraday forecastOffer control reserve

Potentials (I) – Potentials under different parameters

Variation of input parameters Security level Block length Bidding strategyTrade-off System security Economic efficiency

Actual feed-in

Secure day-ahead forecast

Powe

r (no

rmal

ized)

00:00

0.05

0.1

0.15

0.2

0.25

0.3

Time04:00 08:00 12:00 16:00 20:00

Potentials (II) – German pool of wind farms, Day-Ahead Tendering (1st July 2010 – 31st Dec 2010)

1h 4h 24h0

2

4

6

8

10

12

14

Product length

Pote

ntia

l [M

Wh]

x 106

95%99%99,8%99,9%99,99%99,994%99,999%

19

Maximum – No additional income for wind farms

Potentials (III) – Ecomonic impact on the tertiary control market of the German wind farm pool (1st July – 31st Dec 2010)

Results for teriary secondary reserve bids / available active power proof

method

95 % 99 % 99.8 % 99.9 % 99.99 % 99.994 % 99.999 %0

4

8

12

16

Security Level

Cos

t Red

uctio

n [€

]

x 106

24 hour product length

4 hour product length

1 hour product length

Minimum – Maximum additional income for wind farms

OUTLOOK

603025201510 5040 45355 55

Field test

1

0

0.5

-1

-0.5

Ener

gy fr

om U

nit [

MW

/ M

w offe

red]

Nega

tive

cont

rol

rese

rve

Posit

ive

cont

rol r

eser

ve

Tertiary controlSecondary control

Protocol Real SignalPro-tocol Real Signal

22

Conclusions

Wind Farm participation in control reserve markets … is economically feasible can substitute fossil fuel fired power plants will increase competition in the market can generate additional revenue for the wind farm operators

Balance control proof mechanism is unfavourable In economic terms In terms of curtailed energyBalance control proof mechanism is favourable In terms of control reserve demand for the TSO

THANK YOU FOR YOUR ATTENTION!VISIT US AT BOOTH B-B77

Malte Jansenmalte.jansen@iwes.fraunhofer.de