Embed Size (px)
Reliawind Field StudyReliawind Field Study
Michael Wilkinson, Garrad Hassan & Partners Ltd
2009 Wind Turbine Reliability Workshop2009 Wind Turbine Reliability WorkshopAlbuquerque, NM, 17-18 June 2009
Garrad Hassan Around the World• Founded in 1984 in UK• Now have offices worldwide
f• Local understanding informs global perspective
350 professionals25 Offices
Asset Management and Optimisation Services (AMOS)Services (AMOS)
Performance MonitoringFault diagnostics and forensic analysis of SCADA dataPost-construction energy forecastsO&M adviceWarranty calculationsEnd of warranty inspections
20 GW of operating plant assessed worldwide
Introduction to Reliawind
Summary of Reliawind•EC funded R&D project•€7.7M total funding
•10 organisations involved•3 years
•€5.5M EC funding
Aims:•Improve general understanding of wind turbine and farm reliability•Develop reliability models specific to wind turbines
•Increase MTBF •Increase availability•Decrease MTTR •Decrease CoE
Important Onshore Critical OffshoreImportant Onshore Critical Offshore
Summary of ReliawindMain project work packages:1. Field data reliability analysis2. Design for reliability3. Algorithms for condition monitoring4. Proof of concept4. Proof of concept
Previous Work – Availability
140“Standard” long-term availability
i 97% 50% of wind farm years with
Distribution of annual availability
100
120
year
s
assumption ≈ 97%
Mean = 96 4%
50% of wind farm years with availability 97.5% or higher
60
80
umbe
r of w
ind
farm
y Mean 96.4%
90% of wind farm years with availability of 92.5% or higher
99% of wind farm years with availability of
80.0% or hi h
20
40
Nu higher
0
<75% 76
%
77%
78%
79%
80%
81%
82%
83%
84%
85%
86%
87%
88%
89%
90%
91%
92%
93%
94%
95%
96%
97%
98%
99%
100%
Annual System Availability [%]
Database:Database:• Over 250 wind farms worldwide• All major manufacturers
• Over 1000 wind farm operating years• Operating from 1 to 15 years
Source GH, presented at EWEC 2008
Previous Work – Reliability
Courtesy of Durham University
Previous Work
• Previous availability work limited to high level availability and fault information
• Previous reliability work limited by quality of datay y y
• Need better quality data to get better results…
Reliawind Field Study
Reliawind Field Study Methods
Data Available from a Wind Farm
• 10 minute average turbine and substation SCADA databases
• Fault logs per turbine and substation• Manual records• Work orders• Monthly operational reports compiled by the operator
and or manufacturer
Reliawind Common Formats
Turbine taxonomyA common approach to describe wind turbines from different manufacturers
DatabaseA common method to store downtime events
d th i i iand their origin
Reliawind Initial Results
Results – The Story so FarData sources:• Manufacturers – Consortium members• Owners & Operators – Users’ Working Group
Data added so far:• 290 turbines
240 ind farm months• 240 wind-farm months
Results – Failure Rate
Converter
Generator
Pitch System
G bGearbox
Percentage contribution to overall failure rateData source: turbines from multiple manufacturers
Results – Downtime
CConverter
Generator
Pitch System
Gearbox
Percentage contribution to overall failure rateData source: turbines from multiple manufacturers
Results – Reliability Profile
35.0%
40.0%
year] 100%Power Module
15 0%
20.0%
25.0%
30.0%
35.0%
te [failures / turbine /
40%
60%
80%
Rotor Module
Control UnknownS l M d l
YourSite
Pitch System
0.0%
5.0%
10.0%
15.0%
Failure Rate
0%
20%
NacelleControl Unknown
Drive Train
AuxiliaryEquip
Other
Structural Module
Generator
25.0%
30.0%
35.0%
40.0%
/ Turbine / Year]
60%
80%
100%
Power Module
Rotor Module
Generator
5.0%
10.0%
15.0%
20.0%
Hours Lost [hrs
20%
40%
Nacelle Control
Unknown
Drive T i
AuxiliaryEquip
OtherStructuralMod le
Pitch System
0.0% 0%
Concluding Remarks• Reliawind addresses the need for a quantitative
measurement and understanding of turbine reliability
• Common standards defined – turbine taxonomy and data structure
• Database populated with data from 290 turbines
• Database growing!
