Long-term Performance of Wind Farms - · PDF fileLong-term Performance of Wind Farms 1 EWEA...

DNV GL © 2014 9 December 2014 SAFER, SMARTER, GREENERDNV GL © 2014

9 December 2014

Michael Wilkinson

ENERGY

Long-term Performance of Wind Farms

1

EWEA Technical Workshop, Malmo, December 2014

DNV GL © 2014 9 December 2014

Does wind turbine performance degrade over time?

Recent studies from academic institutes in the UK suggest that wind farm

production degrades by a significant margin each year

Resulting in a loss of confidence from wind energy operators and the investment

community

DNV GL draws on extensive experience and high quality, high resolution data to

test these claims and perform a root cause analysis

2

Experience matters: >50 GW of operational analysis; forecasting for >40 GW; Inspections for >700 turbines per year; Largest independent provider of wind farm SCADA


Factors Affecting Turbine Performance

3

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation



4

Environmental

effects

Wind conditions

Sub-optimal

performance

Wind speed

pow

er

Wind speed

pow

er

pow

er

Wind speed

Icing

High impact on some sites

Insects

High impact for short periods

Dirty blades

Subtle impact but persistent

Drive train

degradation



5

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation

Typical energy loss -0.5% (site specific)

Annual Trend? Flat



6

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation

Parameters that impact the power curve:

Flow inclination

Turbulence intensity (TI)

Shear profile

Air density

Conditions that influence these parameters:

Atmospheric stability (TI, Shear, density)

Topography (flow inclination, TI, and shear)

Forestry (TI and shear)



7

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation



8

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation

Typical energy loss -1% (site specific)

Annual Trend? Flat


Annual Trend? Flat



9

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation

De-rating

Usually intermittent effects:

De-rating

Lower power curve

Power curve scatter

Resulting from:

Maintenance

Control modes

Sensor errors

Component misalignment



10

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation

Typical energy loss -1% (operator specific)

Annual Trend? Flat


Annual Trend? Flat


Annual Trend? Flat



11

Gearbox Generator

Powerconverter

Transformer

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation



12

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation



13

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation

Typical energy loss ???

Annual Trend? ???


Annual Trend? Flat


Annual Trend? Flat


Annual Trend? Flat


Drive Train Degradation Study - Method

Category

Control Mode

ChangeConsistent

Anemometry

1 NO YES

2 YES YES

3 NO NO

4 YES NO

14

Identify suitable

projects

Remove

intermittent

effects

Make binned

‘clean’ power

curve

Calculate energy

loss

31 wind farms

235 turbines

Across Europe



15

Identify suitable

projects

Remove

intermittent

effects

Make binned

‘clean’ power

curve

Calculate energy

loss



16

Identify suitable

projects

Remove

intermittent

effects

Make binned

‘clean’ power

curve

Calculate energy

loss



17

Identify suitable

projects

Remove

intermittent

effects

Make binned

‘clean’ power

curve

Calculate energy

loss


Drive Train Degradation Study - Results

18



19

31 wind farms

235 turbines



20

31 wind farms

235 turbines



21

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation


Annual Trend? Flat


Annual Trend? Flat


Annual Trend? Flat

Typical energy loss ???

Annual Trend? ???



22

Environmental

effects

Wind conditions

Sub-optimal

performance

Drive train

degradation


Annual Trend? Flat


Annual Trend? Flat


Annual Trend? Flat

Typical energy loss -0.2% / year for the first 5 years

Annual Trend? Downward trend


Conclusions and Further Work

Factors affecting turbine performance have been reviewed

Long-Term Degradation Study

Drive train degradation has been investigated

35 wind farms with 235 turbines from across Europe included in the study

Individual sites show differences, but overall a small downward trend is identified

-0.2% / year for the first 5 years

Further work

Can we quantify degradation in individual components?

What happens after year 5/6?

23


SAFER, SMARTER, GREENER

www.dnvgl.com

24

Michael Wilkinson

[email protected]

+44 117 972 9900


What could cause production degradation?

3 possible elements that could cause degradation of wind farm production over

time:

25

Wind resource Availability Turbine performance

Wind speed

pow

er


Wind resource

DNV GL Wind Index

Historical wind data from 1996 to 2014

50 meteorological stations spread

across the mainland of the UK

Each station has measured consistent

wind speed data for 10 years or more

26


Availability

27

DefinitionDatabase• Over 300 wind farms across Europe• Between 1 and 15 years of operation• >1,500 wind farm years• All major turbine suppliers to UK represented

0

50

100

150

200

250

300

350

90%

91%

92%

93%

94%

95%

96%

97%

98%

99%

100%

1 2 3 4 5 6 7 8 9 10

Num

ber

of w

ind f

arm

s

Syste

m A

vailability

number of wind farms

50% exceeding thisvalue (median)

25% exceeding thisvalue

75% exceeding thisvalue

Years of operation

𝐒𝐲𝐬𝐭𝐞𝐦 𝐀𝐯𝐚𝐢𝐥𝐚𝐛𝐢𝐥𝐢𝐭𝐲 =

𝐓𝐢𝐦𝐞 𝐫𝐞𝐚𝐝𝐲 𝐭𝐨 𝐨𝐩𝐞𝐫𝐚𝐭𝐞,𝐨𝐫 𝐨𝐩𝐞𝐫𝐚𝐭𝐢𝐧𝐠

𝐓𝐨𝐭𝐚𝐥 𝐭𝐢𝐦𝐞

Source: Optimising the UK operational fleet in a mature market (Harman et al., Renewable UK 2013)

Long-term Performance of Wind Farms - · PDF fileLong-term Performance of Wind Farms 1 EWEA...

Documents

Transcript of Long-term Performance of Wind Farms - · PDF fileLong-term Performance of Wind Farms 1 EWEA...