Ice Profile ClassificationBased on ISO 12494

Presented by:

Matthew Wadham-Gagnon

2012-02-12

WinterWind 2013

2

Co-authors

Dominic Bolduc, TCE, QC, Canada

Bruno Boucher, TCE, QC, Canada

Amélie Camion, Repower Systems Inc, QC, Canada

Jens Petersen, Repower Systems SE, Germany

Hannes Friedrich, Repower Systems SE, Germany

Content

• Introduction to TCE

• Motivation

• Methodology– Measurement Campaign

– Ice profile classification using ISO 12494

• Example of data for one icing event

• Summary for winter 2011-2012

• Conclusions/Further work

TCE infrastructure location

Over 900MW in operation

TCE R&D wind farm

4

TCE office

5

TCE R&D Wind Farm

www.eolien.qc.ca

• Two 2.05 MW RepowerMM92 CCV wind turbines

• Located in Riviere-au-Renard, Québec, Canada

• Icing & complex terrain

• Commissionned in March2010

• Research, developmentand technology transferprojects involvingnorthern climates andcomplex terrain.

6

TCE 126m Met Mast

Compliant withCSA-S37-01

(DLC: 40 mm of iceand 57 m/s)

7

TCE 126m Met Mast• 5 levels of unheated class 1

anemometers

• 3 levels of heated anemometers

• 2 levels of heated and unheatedwind vanes

• Ceilometer to detect cloudheight

• 4 levels of thermometers

• Hygrometer

1Hz data acquisition in Osisoft PI

8

TCE Micro Grid

www.eolien.qc.ca

9

TCE Micro Grid

www.eolien.qc.ca

10

Infrastructure Topographic Layout

projected

projected

11

Objective/MotivationsWhat:• Quantify ice load during different icing events

Why:• To validate and/or improve theoretical ice accretion

models• To correlate with meteorological measurements,

production loss and ice induced vibration• To evaluate the performance of anti-icing and de-icing

technologies

Ice Measurement Campaign

• Ongoing since October 2011• Intended to continue until at least winter 2013-2014

• Observations during icing events• Remote Cameras• Ice throw• Meteorological and production data

12

• Pictures taken from the ground

23 days of observed ice on blades (Winter 2011-2012)

13

Observations - Blades

14

Observations – Nacelle Weather Mast

• Remote Camera on Nacelle – 5 min intervals

800+ hours of icing on nacelle weather mast (Winter 2011-2012)

15

Ice throw

• Shape, size, weight, distance from tower

• 150+ fragments of ice from blades collected to date

16

Estimating Ice Load using ISO 12494

• ISO 12494 – for RIME– Ice Class (IC) for Rime (R) scale 1 to 10 according to kg/m of ice on

structure, shape and dimensions of ice varies depending onstructure size and shape

17

Estimating Ice Load using ISO 12494

• ISO 12494 – for GLAZE (Freezing Rain)– Ice Class (IC) for Glaze (G) scale 1 to 6 according to ice thickness

18

Nacelle Weather Mast Load Classification

• Weather mast ice load estimation– “D” measured,– “L” estimated, higher error but more data

D

19

Blade Ice Load Classification

• Blade ice load estimation– “L” measured– “D” estimated– Limited amount of pictures covering the duration of an icing event.

20

Ice throw Load Classification

• Used to validate blade ice load– More accurate measurement– Precise time of ice shed unknown

D

L

W’h W

21

Ice Class Rime 1 (ICR1) – 0 to 0.5 kg/m

22

Ice Class Rime 2 (ICR2) – 0.5 to 0.9 kg/m

23

Ice Class Rime 3 (ICR3) – 0.9 to 1.6 kg/m

24

Ice Class Rime 4 (ICR4) – 1.6 to 2.8 kg/m

25

Ice Class Rime 5 (ICR5) – 2.8 to 5.0 kg/m

26

Example Icing Event

• April 21 to 23, 2012• Rime, ICR3 to ICR5

27

Meteorological Icing (Ice Sensor)

0.0

0.2

0.4

0.6

0.8

1.0

Met. Icing (ice sensor)

28

Met. Icing (Cloud Height & Temp.)

0

20

40

60

80

100

120

140

-6-4-202468

101214

Tem

p(o

C)

WEC1 Temp MMV2 78m T MMV2 11m Min CH

Met. Icing (ice sensor) Met. Icing (CH & T)

CH (m

)

29

Instrumental Icing (WEC unheated anemo)

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2)

0

5

10

15

20

25

30

WS

(m/s

)

WS1 WS2

30

Instr. Icing (MMV2 126m unheat anemo)

05

101520253035

WS

(m/s

)

MMV2 126m U MMV2 126m H

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2) Instr. Icing (126m)

31

Instr. Icing (MMV2 80m unheat anemo)

05

101520253035

WS

(m/s

)

MMV2 80m U MMV2 80m H

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2)

Instr. Icing (126m) Instr. Icing (78m)

32

Instr. Icing (MMV2 34m unheat anemo)

05

101520253035

WS

(m/s

)

MMV2 34m U MMV2 34m H

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2)

Instr. Icing (126m) Instr. Icing (78m) Instr. Icing (34m)

33

WEC1 Production Loss

0

500

1000

1500

2000

2500

3000

Pow

er (k

W)

Power CPC

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2) Instr. Icing (126m)

Instr. Icing (78m) Instr. Icing (34m) Prod. Loss (WEC1)

34

Production Loss (WEC2)

0

500

1000

1500

2000

2500

3000

Pow

er (k

W)

Power CPC

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2) Instr. Icing (126m)

Instr. Icing (78m) Instr. Icing (34m) Prod. Loss (WEC1) Prod. Loss (WEC2)

35

Nacelle Weather Mast Ice Load

0

1

2

3

4

5

ICR

Met mast

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2)Instr. Icing (126m) Instr. Icing (78m) Instr. Icing (34m)Prod. Loss (WEC1) Prod. Loss (WEC2) Nacelle WM

36

Blade Ice Load

0

1

2

3

4

5

ICR

Iced Blade Met mast

Met. Icing (ice sensor) Met. Icing (CH & T) Instr. Icing (WEC2) Instr. Icing (126m)Instr. Icing (78m) Instr. Icing (34m) Prod. Loss (WEC1) Prod. Loss (WEC2)Nacelle WM Observ.

37

Example Icing Event

• April 21 to 23, 2012• Rime, ICR3 to ICR5

38

Power Curve

• Production loss during april 20-23 icing event

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30

Pow

er (k

W)

WS (m/s)

Icing Event Data

GPCNon-iced Production

39

Power Curve

• Power curve estimations based on ice classification andproduction losses of all icing events in Winter 2011-2012

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30

Pow

er (k

W)

WS (m/s)

GPC

ICR2&3

ICR4&5

Non-iced Prod.

40

Nacelle Weather Mast Load Classification

• Weather mast occurrence of icing severity over the courseof the winter

41

Winter 2011-2012 in numbers

Meteorological Icing ~200 hoursInstrumental Icing ~400 hoursNacelle WM icing 800+ hoursIce observed on blades 23 daysProduction Loss ~140 hoursProduction Loss (% AEP) ~1.5%Hours of ICR2&3 ~300 hoursHours of ICR4&5 ~100 hoursHours of Glaze ~100 hours

42

Conclusions

• Improved understanding of the severity of icing events

• Measurement Campaign database can be used to validatesimulation models (e.g. weather, ice accretion)

• Methodology can be applied to assessing performance of iceprotection technologies

• Nacelle weather mast ice severity correlates to blade icingseverity

• Duration of weather mast icing generally much longer thanblade icing

• How to make measurement of icing severity on blades couldbe more accurate and more frequent?

43

Checker board tape

• In order to get a better estimation of blade ice load andposition along span of blade

44

Hub Camera

45

Ongoing – Automation of Image Analysis

ICR= 5

In Collaboration with:

46

References

IEA Wind Task 19, Expert Group Study on Recommended Practices: 13. Wind Energy Projects in ColdClimates, 1. Edition 2011

ISO 12494:2001, Atmospehiric Icing of Structures, International Organisation for Standardization, 2001

Thank you, Tack, Danke, Merci!

Matthew Wadham-Gagnon, eng, M.engProject Leadmgagnon@eolien.qc.ca

70, rue Bolduc, Gaspé (Québec) G4X 1G2Canada

Tél. : +1 418 368-6162