How does the real world performance of wind … · How does the real world performance of wind turbines compare ... • Post-construction energy forecasts ... • Flow inclination

How does the real world performance of wind turbines compare

with sales power curves?

EWEA: Lyon, July 2012 – Keir Harman

Background

Asset Management and Optimisation Services

(AMOS)

• Turbine performance monitoring

• SCADA-based condition monitoring

• Fault diagnosis and forensic analysis of SCADA data

• Post-construction energy forecasts

• Warranty calculations

• End of warranty inspection analyses

• O&M advice

• Reliability profiling and benchmarking

Over 30 GW of operating wind farms assessed to date

What do we typically see in operating data?

• Power curves rarely lie on the sales power curve

0 5 10 15 20

Nacelle anemometer wind speed [m/s]

Avera

ge P

ow

er

[kW

]

WTG 1

WTG 2

WTG 3

WTG 4

WTG 5

WTG 6

WTG 7

WTG 8

WTG 9

WTG 10

WTG 11

WTG 12

WTG 13

WTG 14

WTG 15

WTG 16

Warranted power

Category Typical range of

loss/gain

(nominal energy %)

Most likely

(nominal energy %)

1) Generic power curve performance

2) Mechanical sub-optimal performance

3) Environmental: icing and dirty blades

4) Wind conditions: turbulence intensity, shear

and flow inclination

Real world power curve losses/gains categorised

???

Category 1: Generic power curve performance

• 115 project power curve tests using IEC guidelines [61400 pt 12-1]

• Average of results = 99%

• IEC measurement uncertainty typically 5%

< 92% 92% to 94%

94% to 96%

96% to 98%

98% to 100%

100% to 102%

102% to 104%

104% to 106%

106% to 108%

>108%

Pro

po

rtio

n o

f d

atab

ase

Percentage of warranted power curve energy achieved (NME/NWE)

Category 2: Mechanical sub-optimal performance – common causes

1) De-rating 3) Component misalignment /

Sensor error 2) Non-optimal controller settings

Power vs.

Wind speed

Power vs.

Rotor speed

Rotor speed [rpm]

Avera

ge P

ow

er

[kW

]

0

10

20

30

40

50

608

0%

81

%

82

%

83

%

84

%

85

%

86

%

87

%

88

%

89

%

90

%

91

%

92

%

93

%

94

%

95

%

96

%

97

%

98

%

99

%

10

0%

10

1%

Nu

mb

er o

f w

inf fa

rm y

ea

rs

Annual Operating Efficiency

Category 2: Mechanical sub-optimal performance - What can be expected?

Mean = 98%

50% of database with OE of

99% or higher (Median)

Database

62 wind farms across Europe

Between 1 and 6 years of operation

134 wind farm years

Definition

Energy produced

Operating Efficiency (OE) =

Energy expected with ‘normal’ power curve

Target

Operating Efficiency =100%

Category 3: Environmental - causes Icing

High impact on some sites

Bugs

High impact for short periods

Dirty blades

Subtle impact but persistent

Wind speed Wind speed Wind speed

pow

er

pow

er

pow

er

• Typical range -3% to -0.2% and very region specific

Category 4: Wind conditions

The power curve is impacted by:

• Flow inclination

• Turbulence intensity (TI)

• Shear profile

• Air density

Influenced by:

• Atmospheric stability (TI, shear, density)

• Complex terrain (flow inclination, TI, and shear)

• Forestry (TI and shear)

1

2

3

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

0 2 4 6 8 10 12 14 16 18 20 22 24 26

Wind speed [m/s]

Avera

ge p

ow

er

[kW

]

0

30

60

90

120

150

180

210

240

270

300

330

Category 4: Wind conditions:

Flow inclination impact on power curve (Extremes) GLGH validation of Madsen/Pederson research for MW-scale wind turbines

Yaw error observations for MW-scale turbines (GLGH)

Category 4: Wind conditions

Turbulence Intensity (TI) and Shear impact on power curve (Extremes)

0

200

400

600

800

1000

1200

1400

1600

0 5 10 15 20 25

Wind speed [m/s]

Po

wer

[kW

]

0

100

200

300

400

500

600

700

800

900

Distribution

14%

15%

16%

17%

18%

19%

20%

21%

22%

23%

24%

25%

High TI case:

• 2% drop in nominal energy between TI of

14% and 20% due to ‘rounded knee’ for a

high wind speed site

20%

30%

40%

50%

60%

70%

80%

90%

7 8 9 10 11

Po

we

r [%

of

Ra

ted

]

Wind Speed [m/s]

High TI

Mid TI

Low TI

Low TI case:

• 3% drop in nominal energy during periods of

low TI (<8%), which corresponds to stable

atmospheric conditions

Category Typical range of

loss(-ve)/gain(+ve)

(nominal energy %)

Median

(nominal energy %)

1. Generic power curve performance -5% to +3% -1% (model specific)

2. Mechanical sub-optimal performance -5% to +0% -1% (operator specific)

3. Environmental -3% to -0.2% -0.5% (region specific)

4. Wind conditions – turbulence intensity,

shear and flow inclination -5% to +1% -1% (site specific)

Conclusions

• Real world turbine performance does generally deviate from sales power curves

• Causes can be grouped and quantified based on observations from operational analyses

Questions?

Keir Harman

[email protected]