What we know about Onshore & Offshore

Wind Turbine Reliability with particular

reference to Future Offshore Wind Farm
Operational Performance
Peter Tavner
Emeritus Professor, Durham University, UK
Past President, European Academy of Wind Energy

The darkest regions of hell are reserved for those who remain
neutral at times of moral crisis
Dante Alighieri

NTNU, EU FR7 MARE WINT Project

September 2013

Keynotes
Aim to reduce risk, raise turbine Reliability and
Availability, Reduce offshore wind Cost of Energy;
Wind Turbine Reliability from onshore experience;
What we know about WT reliability;
Wind Turbine Availability, what is happening
Offshore?

NTNU, EU FR7 MARE WINT Project

September 2013

Wind Turbine Power Curves

Alstom 1.67 MW,

Variable-speed,
Mitsubishi 1 MW,
Variable-pitch
Fixed-speed,
Stall-regulated,
Variable-pitch

NTNU, EU FR7 MARE WINT Project

September 2013

Wind
Turbine
Operation18 days
WT rating= 1.67MW
Variable-speed, Variable-pitch
Average output=490kW
Capacity factor =500/1670=29%

Actual Wind Power Production

UK & Spain

http://www.bmreports.com/bsp/bsp_home.htm
https://demanda.ree.es/eolicaEng.html

NTNU, EU FR7 MARE WINT Project

September 2013

Trend in Turbine Failure Rates with Time

NTNU, EU FR7 MARE WINT Project

September 2013

WT Reliability and Size, EU

Small, group I

Medium, group II

NTNU, EU FR7 MARE WINT Project

September 2013

Large, group III

Reliability, Downtime and Subassemblies, EU

NTNU, EU FR7 MARE WINT Project

September 2013

Typical WT Generator Failure Intensities

NTNU, EU FR7 MARE WINT Project

September 2013

More detail on WT Generator Failures

NTNU, EU FR7 MARE WINT Project

September 2013

10

Typical WT Gearbox Failure Intensities

NTNU, EU FR7 MARE WINT Project

September 2013

11

WT Reliability-Downtime per
Assembly
Stop Rate and Downtime from Egmond aan Zee Wind Farm, the Netherlands, over 3 Years
-2

-1.5

-1

-0.5

0.5

1.5

2.5

Egmond aan Zee Failure Rate, 108

Turbine Years

Control System
Yaw System

Egmond aan Zee Downtime, 108

Turbine Years

Scheduled Service
Pitch System
Gearbox
Ambient
Generator
Converter
Electrical
Blade System
Structure
Grid
Brake System
100

75

50

25

Annual Stop Frequency

0.5

1.5

2.5

Downtime per Stop (days)

NTNU, EU FR7 MARE WINT Project

September 2013

12

Typical WT Converter Failure Intensities

NTNU, EU FR7 MARE WINT Project

September 2013

13

Reliability & Subassemblies, EU,

Reliawind

Data Source:
Reliawind Deliverable D.1.3 Reliability
Profiles Figures subject to update.

Lighter larger background blocks show sub-systems;

Darker smaller foreground blocks show assemblies;
The line shows the Pareto cumulative contribution.

Downtime & Subassemblies, EU,

Reliawind

Data Source:
Reliawind Deliverable D.1.3
Reliability Profiles

Lighter large background blocks show sub-systems;

Darker smaller foreground blocks show assemblies;
The line shows the Pareto cumulative contribution.

Summary of least reliable sub-assemblies

& their failure modes
Sub-system /
Assembly
Electrical
(5 out of
13)

Failure
Mode 1

Failure
Mode 2

Failure
Mode 3

Failure
Mode 4

Failure
Mode 5

Battery Failure

Pitch Motor
Failure

Pitch Motor
Converter Failure

Pitch Bearing
Failure

Temperature or
Humidity Sensor
Failure

Internal leakage
of proportional
valve

Internal leakage
of solenoid valve

Hydraulic
cylinder leakage

Position sensor
degraded or no
signal

Pressure control
valve sensor
degraded signal

Frequency Converter
(5 out of 18)

Generator-side or
Grid-side Inverter
Failure

Loss of
Generator Speed
Signal

Crowbar Failure

Converter
Cooling Failure

Control Board
Failure

Yaw System
(5 out of 5)

Yaw gearbox &

pinion lubrication
out of
specification

Degraded wind
direction signal

Degraded
guiding element
function

Degraded
hydraulic
cylinder
function

Brake operation
valve does not
operate

Temperature
sensor modules
malfunction

PLC analogue
input malfunction

PLC analogue
output
malfunction

PLC digital input

malfunction

PLC In Line
Controller
malfunction

Pitch
System

Hydraulic
(5 out of 5)

Control System
(5 out of 5)
Generator Assembly
(5 out of 11)

Worn slip ring

brushes

Stator winding
temperature
sensor failure

Encoder failure

Bearing failure

External fan
failure

Gearbox Assembly
(5 out of 5)

Planetary Gear
Failure

High Speed Shaft

Bearing Failure

Intermediate
Shaft Bearing
Failure

Planetary
Bearing Failure

Lubrication
System
Malfunction

Studies on Different WTs

WT Make A
153x1.5-2MW WTs with 3-blades, electric pitchregulation, geared-drive, DFIG generator and
partially-rated converter
6 Wind Farms
Over 2 years
WT Make B
366x2.5MW WTs with 3-blades, electric pitchregulation, geared-drive, synchronous generator
and fully-rated converter
1 year
NTNU, EU FR7 MARE WINT Project
September 2013

17

SCADA Alarm System

Performance Evaluation -KPIs
* Reference : Alarm systems, a guide to design, management and
procurement No. 191 Engineering Equipment and Materials Users
Association 1999 ISBN 0 8593 1076 0

KPIs: Key Performance Indices *

KPI 1, Average Alarm Rate: Long term average of the number
of alarm triggers occurring within a 10 min SCADA interval.

KPI 2, Maximum Alarm Rate: Maximum number of alarm

triggers occurring within a 10 min SCADA interval.

Additional Definitions *
Alarm Shower: A single fault causing a large number of alarm
triggers, in this work an Alarm Shower consists of > 10 alarm
triggers

NTNU, EU FR7 MARE WINT Project

September 2013

18

Alarm KPIs from 7 Wind Farms

NTNU, EU FR7 MARE WINT Project

September 2013

19

1000
100

Wind Farms
3&6

Reactive Alarms

Wind Farm 4

10

Stable Alarms

Wind Farms
1&5

0.1

KPI 1, Average AlarmTriggers per 10 min

WT Make A Alarm System Performance

Recommendation
for Alarm presentation
to humans

Wind Farm 2

Presentation
to Operators
Alarm Rates
must be preprocessed

10

100

1000

10000

100000

KPI 2, Maximum AlarmTriggers per 10 min

Reactive- peak alarm rate during upset is unmanageable and alarm system
will continue to present an unhelpful distraction to the operator for long period.
Stable- Alarms have been well defined for normal operation, but the system is
less useful during plant upset.
NTNU, EU FR7 MARE WINT Project
September 2013

20

WT Pitch Mechanism Taxonomy

NTNU, EU FR7 MARE WINT Project

September 2013

21

WT Pitch Alarm,
Relationships over 2 years,
Qiu et al

NTNU, EU FR7 MARE WINT Project

September 2013

22

WT Converter Taxonomy

NTNU, EU FR7 MARE WINT Project

September 2013

23

WT Converter Alarms,
Statistical Relationships over 2 years,
Qiu et al
337
Grid-side
Inverter Overcurrent

338/343
Rotor-side Inverter Overcurrent/
Over- temperature
345
DC Overvoltage

349
Grid Voltage Dip

369 - Pitch
372-374 - Blade1-3 Emergency

322 - Inverter

263
Main Switch

WT Converter Alarm
Showers,
2 years,
Qiu et al

1st grid
incident

2 years

2nd grid
incident

Power Electronics Reliability,

Measured, Wolfgang

NTNU, EU FR7 MARE WINT Project

September 2013

26

WT Unreliability & Importance of Pre-Testing

Root Causes

Wind condition
Weather
Faulty design
Faulty materials
Poor maintenance

Condition
Monitoring
Signals

SCADA
Signal
Analysis

Failure Modes
And Effects
Analysis,
FMEA

Failure Location

How?
Pre-Testing during Prototype Development
Or In-Service SCADA & CMS Analysis & Diagnosis

Why?
Root Cause Analysis

NTNU, EU FR7 MARE WINT Project

September 2013

27

Onshore
Availability and Wind Speed
Brazos, Texas, USA, 160 MW, 160 x Mitsubishi MWT1000, 1 MW
100.0

Availability, %

80.0

60.0

40.0

20.0

0.0
0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

Wind Speed, m/s

NTNU, EU FR7 MARE WINT Project

September 2013

28

Offshore
Availability and Wind Speed
Barrow, UK, 90 MW, 30 x Vestas V90, 3 MW
100.0

Scroby Sands, UK, 60 MW, 30 x Vestas V80, 2 MW

40.0

80.0

0.0
0.0

100.0

60.0

80.0

40.0
2.0

4.0

20.0

Availability, %

20.0

Kentish Flats, UK,90 MW, 30 x Vestas V90, 3 MW

6.0

8.0
60.0
Wind Speed, m/s

0.0
0.0

2.0

10.0

12.0

100.0

40.0

4.020.0

Egmond
aan Zee, Netherlands, 108 MW, 36 x Vestas V90,
14.0
3 MW

6.0

8.0

Wind Speed, m/s

0.0
0.0

2.0

80.0
10.0

12.0

14.0

60.0
40.0
4.0
20.0

North Hoyle, UK, 60 MW, 30 x Vestas V80, 2 MW

6.0

8.0
100.0
Wind Speed, m/s

10.0

12.0

14.0

80.0

0.0
0.0

Availability,%

100.0

Availability, %

60.0

Availability, %

Availability, %

80.0

2.0

60.04.0

6.0

8.0

10.0

12.0

6.0

8.0

14.0

Wind Speed, m/s

40.0
20.0

0.0
0.0

2.0

4.0

Wind Speed, m/s

10.0

12.0

2914.0

Onshore Availability and Wind

Speed, World
40%
energy
produced
at wind
speeds
>11m/s

NTNU, EU FR7 MARE WINT Project

September 2013

North European Offshore Wind Farm Performance

Three North European Offshore Wind Farms
Scroby Sands North Hoyle Egmond aan Zee
100

Wind Speed, m/s; Capacity Factor, %; Availability, %

90

Availability
80

Capacity
Factor

70
60

50
40

Wind
Speed

30
20

10
0

NTNU, EU FR7 MARE WINT Project

September 2013

31

Relationship between Failures & Weather

NTNU, EU FR7 MARE WINT Project

September 2013

32

Relationship between Failures & Weather

NTNU, EU FR7 MARE WINT Project

September 2013

33

Power Curves & Turbulence from SCADA

1101

1087

1114

Power curves
Red SCADA real power
curve

Blue Manufacturers
theoretical power curve
NTNU, EU FR7 MARE WINT Project
September 2013

34

Wind Turbulence from Fast Data

Distribution of
wind velocity
Is this difference
driving failures?
Gaussian
Distribution

Probability density function of spatial transversal wind

velocity increments over a distance of 10 m, for =4 s
compared to a Gaussian distribution.
NTNU, EU FR7 MARE WINT Project
September 2013

35

Wind Turbulence from Slow Data

WT SCADA

Distribution of
wind velocity
from SCADA
Gaussian
Distribution

Probability density function of wind velocity

from 10 min SCADA
NTNU, EU FR7 MARE WINT Project
September 2013

36

Wind Turbulence from Slow Data

WT SCADA

NTNU, EU FR7 MARE WINT Project

September 2013

37

Turbulence in
context
But turbulence of this
dimension could affect
drive train
Turbulence of this
dimension will not
affect drive train

NTNU, EU FR7 MARE WINT Project

September 2013

38

SCADA Load Changes, 18 different WTs

Percentage load change occurrences

2.5%

1097
1209

2.0%

Most based on 2.2 years data

Use bin size of 100 N
High variety between 0-6000N
WTs are from 5 different sites

1.5%

1221
1222
1096

1217
1202
1219
1098
1103

1.0%

1114
1120
1557
1558

0.5%

1560
1339
1341
1342

0.0%
0

5000

10000

15000

20000

25000

Radial load change on left HSS gearbox bearing every 10 mins (N)
Courtesy: Dr Hui Long NTNU, EU FR7 MARE WINT Project
September 2013

39

Difference between sites, Site 1

Percentage of load change occurrences

2.5%

Healthy WTs with relatively low load change

WTs from same site have similar load change
distributions

2.0%

1.5%

1097

1096
1.0%

1098

0.5%

0.0%
0

1000

2000

3000

4000

5000

6000

Radial load change on left HSS gearbox bearing every 10 mins (N)
Courtesy: Dr Hui Long NTNU, EU FR7 MARE WINT Project
September 2013

40

Difference between sites, Site 4

Percentage of load change occurrences

2.5%

Two WTs on this site had

serious failures

2.0%

1.5%
1209
1221

1222
1217

1.0%

1202
1219

0.5%

0.0%
0

1000

2000

3000

4000

5000

6000

Radial load change on left HSS gearbox bearing every 10 mins (N)
Courtesy: Dr Hui Long NTNU, EU FR7 MARE WINT Project
September 2013

41

PDF of HSS Bearing Radial Load Increments from SCADA

Compared with Gaussian distribution the evident different tail

corresponds to large load increments;
Suggesting large load increments observed more frequently than
expected;

Courtesy: Dr Hui Long

Two WTs have different

bearing load increment
occurrence frequencies.

Conclusions

For onshore WTs 75% faults cause 5% downtime, 25% faults cause
95% downtime
Pitch and Main Converters suffer from many faults but with low
downtimes.
Pitch and Main Converters represent a significant part of the 75%.
This behaviour could be problematic offshore.
WT reliability is affecting offshore performance
Sub-assemblies with high failure rates are consistent
Turbulence seems to be causing failures
Analysis of SCADA wind speeds, torques & shaft speeds is showing
ample evidence of turbulent effects
Link between turbulence and failures is difficult to prove
The mathematical tools to be used are not yet clear
NTNU, EU FR7 MARE WINT Project
September 2013

43

References

Tavner PJ, Xiang JP and Spinato F, Reliability analysis for wind

turbines. Wind Energy 10(1): 1-18, 2006.
Spinato F, Tavner PJ, van Bussel GJW and Koutoulakos E, Reliability
of wind turbine subassemblies. IET Renewable, Power Generation
3(4): 115, 2009.
Wilkinson M, Hendriks B, Spinato F, Gomez E, Bulacio H, Roca J,
Tavner P J, Feng Y, Long H, Methodology and results of the
ReliaWind reliability field study, Scientific Track, European Wind
Energy Conference, Warsaw, 2010.
Feng Y, Tavner PJ and Long H, Early experiences with UK round 1
offshore wind farms, Proc Institution of Civil Engineers Energy,
163(EN4): 167181, 2010.
Tavner, P J, Faulstich, S, Hahn, B, van Bussel, G J W, Reliability &
availability of wind turbine electrical & electronic components,
European Power Electronics Journal, 20(4): 1-6, 2010.
Faulstich, S., Hahn B., Tavner, P. J., Wind turbine downtime and its
importance for offshore deployment, Wind Energy,
DOI: 10.1002/we.421, published on early view.

NTNU, EU FR7 MARE WINT Project

September 2013

44