Jørn Heggset, NOWITECH/SINTEF: New solutions for ... - norcowe
Jørn Heggset, NOWITECH/SINTEF: New solutions for ... - norcowe
Jørn Heggset, NOWITECH/SINTEF: New solutions for ... - norcowe
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>New</strong> <strong>solutions</strong> <strong>for</strong> effective<br />
monitoring of offshore wind farms<br />
(relevant activities in <strong>NOWITECH</strong> and WINDSENSE)<br />
<strong>Jørn</strong> <strong>Heggset</strong><br />
<strong>SINTEF</strong> Energy Research<br />
jorn.heggset@sintef.no<br />
NORCOWE Days, 19 th September 2012 in Bergen<br />
1
Content<br />
► About <strong>NOWITECH</strong><br />
► Condition monitoring framework<br />
► Remote presence<br />
► Technical condition index<br />
► About WINDSENSE (IPN project)<br />
2
<strong>NOWITECH</strong> in brief<br />
► a joint pre-competitive<br />
research ef<strong>for</strong>t<br />
► focus on deep offshore<br />
wind technology (+30 m)<br />
► budget (2009-2017)<br />
EUR 40 millions<br />
► co-financed by the<br />
Research Council of<br />
Norway, industry and<br />
research partners<br />
► 25 PhD/post doc grants<br />
► Vision:<br />
� large scale deployment<br />
� internationally leading<br />
Research partners:<br />
► <strong>SINTEF</strong> (host)<br />
► IFE<br />
► NTNU<br />
Industry partners:<br />
► Devold AMT AS<br />
► Det Norske Veritas<br />
► DONG Energy Power<br />
► EDF R&D<br />
► Fedem Technology AS<br />
► Fugro OCEANOR AS<br />
► GE Wind Power AS<br />
► Kværner Verdal<br />
► Lyse Produksjon AS<br />
► NTE Holding AS<br />
► SmartMotor AS<br />
► Statkraft<br />
► Statnett SF<br />
► Statoil Petroleum AS<br />
► Vestas<br />
► Vestavind Offshore<br />
Associated<br />
research partners:<br />
► DTU Wind Energy<br />
► MIT<br />
► NREL<br />
► Fraunhofer IWES<br />
► Uni. Strathclyde<br />
► TU Delft<br />
► Nanyang TU<br />
Associated<br />
industry partners:<br />
► Wind Cluster Mid-<br />
Norway<br />
► Energy Norway<br />
► Enova<br />
► Innovation Norway<br />
► Navitas Network<br />
► NCEI<br />
► NORWEA<br />
► NVE<br />
3
Multidisciplinary Research Challenges<br />
O&M<br />
Grid<br />
Wind<br />
turbine<br />
Substructure<br />
LPC distribution of<br />
offshore wind farm<br />
(example)<br />
Key issue: Innovations reducing cost of energy from offshore wind<br />
4
Condition monitoring framework<br />
Degradation Process<br />
Wind Turbines<br />
Offshore Wind Turbine<br />
Onshore Wind Turbine<br />
Collapsed Wind Turbine<br />
Maintenance<br />
Management<br />
System<br />
Wireless Data<br />
Collection Networks<br />
Data Acquisition<br />
Fiber Bragg Grating Sensors<br />
Acoustic Emission(AE)Sensors<br />
Ultrasonic Sensors + AE<br />
Vibration sensors<br />
Maintenance Scheduling<br />
/ Maintenance Optimization<br />
Bee Colony Algorithms<br />
(BCA)<br />
Ant Colony Optimization<br />
(ACO)<br />
Particle Swarm<br />
Optimization (PSO)<br />
Gentic Algorithms (GA)<br />
Meta-Heuristic<br />
approaches<br />
Signal Pre-process<br />
Denosing<br />
Compression<br />
Extract Weak Signal<br />
Filter<br />
Amplification<br />
Key Per<strong>for</strong>mance Indicator<br />
(KPI)<br />
Feature Extraction<br />
Time Domain<br />
Time-Frequency<br />
Domain<br />
Frequency Domain<br />
(FFT, DFT)<br />
Wavelet Domain<br />
(WT, WPT)<br />
KPI<br />
Leading<br />
KPI<br />
Logging<br />
Principal Component<br />
Analysis (PCA)<br />
Fault Prognosis<br />
Auto-regressive Moving<br />
Averaging (ARMA)<br />
Fuzzy Logic Prediction<br />
ANN Prediction<br />
Match Matrix Prediction<br />
Fault Diagnosis<br />
Support Support Machine<br />
(SVM)<br />
Data Mining (Decision<br />
Tree & Association rules)<br />
Artificial Neural Network<br />
(SOM & SBP)<br />
Statistical Maching<br />
5
Some important issues<br />
► The key to a successful system is providing in<strong>for</strong>mation<br />
and decision support, not just data<br />
� Predictive maintenance<br />
► Wireless monitoring systems are an inevitable part<br />
of the future<br />
► <strong>New</strong> sensor types is widely considered to be a key to<br />
future offshore wind energy harvesting<br />
� Examples:<br />
• Fibre-optic sensors<br />
• Micro-electro-mechanical systems (MEMS sensors)<br />
► Monitoring of technical condition without being there
Remote presence reduces O&M costs<br />
► It is costly and sometimes impossible to have<br />
maintenance staff visiting offshore turbines<br />
► Remote presence:<br />
� Remote inspection through<br />
a small robot on a track in the nacelle<br />
equipped with camera / heat sensitive,<br />
various probes, microphone etc.<br />
� Remote maintenance<br />
through robotized<br />
maintenance actions
Remote presence<br />
(PhD study in <strong>NOWITECH</strong>)<br />
8
Degradation process<br />
Technical condition<br />
"As new"- condition<br />
Lower accept limit<br />
Restore by overhauls Upgrade by major overhauls<br />
Life Cycle Cost<br />
Tilstand<br />
Nedre akseptgrense<br />
Levetidskostnader<br />
Tilbakeføring av tapt<br />
tilstand ved hjelp av<br />
overhalinger<br />
<strong>New</strong> price<br />
Life Cycle Cost<br />
Time<br />
9
Technical Condition Index (TCI)<br />
Technical Condition Index<br />
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
64 65 66 67 68 69 70 71 72 73 74 75 76<br />
Temperature<br />
75 °C<br />
Bearing temperature: °C<br />
Variable value<br />
Transfer function<br />
Reference value<br />
Measurement:<br />
Temperature<br />
TCI = 50<br />
Bearing<br />
Measurement<br />
node X<br />
10
Aggregating Technical Condition<br />
TCI = 65<br />
Tower<br />
w = 0.2<br />
TCI = 82<br />
TCI = 96<br />
Armature Rotor<br />
w = 0.2<br />
Turbine A<br />
Generator Hub Gear<br />
w = 0.2<br />
TCI = 95<br />
w = 0.2<br />
TCI = 90<br />
w = 0.2<br />
Etc.<br />
Air cooler Stator Bearing<br />
TCI = 50 w = 0.7 TCI = 100 w = 0.3<br />
Measurement:<br />
Temperature<br />
TCI = 65<br />
w = 0.2<br />
Measurement<br />
node X<br />
11
Input to maintenance and logistics<br />
Component<br />
Turbine A<br />
Turbine B<br />
Turbine C<br />
Turbine D<br />
Turbine E<br />
Hub controller 75 73 96 92 96 95 79 76 72 93 72 74 74 77<br />
Pitch cylinders 83 72 89 75 90 100 76 84 91 88 98 97 71 71<br />
Blade hub 85 81 89 81 71 72 75 96 93 82 75 45 79 90<br />
Main shaft 80 91 50 90 95 93 53 85 77 80 87 72 100 87<br />
Oil cooler 89 93 85 94 99 98 94 87 97 73 91 99 81 72<br />
Gearbox 100 93 74 91 43 80 91 51 90 78 52 81 95 80<br />
Mechanical disc brake 83 75 84 92 97 93 98 82 97 78 73 80 91 91<br />
Service crane 99 76 94 74 74 60 94 77 86 75 87 95 92 85<br />
Top controller 74 87 95 93 74 80 84 78 97 76 98 98 89 97<br />
Wind sensors 93 93 87 70 88 96 78 71 88 86 84 73 87 72<br />
Trans<strong>for</strong>mer 88 95 71 88 76 86 93 100 97 75 77 70 78 72<br />
Blade 92 73 86 71 91 73 89 78 55 85 85 100 93 71<br />
Blade bearing 73 83 49 78 89 72 87 93 97 73 78 97 47 96<br />
Rotor lock system 74 100 96 76 82 94 90 83 71 88 73 80 98 91<br />
Hydraulic unit 70 96 72 77 97 80 98 80 76 85 98 90 85 80<br />
Machine foundation 74 98 95 83 89 80 83 74 91 80 83 96 76 93<br />
Yaw gears 74 78 87 74 57 93 75 93 87 97 87 86 86 74<br />
Disc coupling 99 79 94 70 81 94 90 95 91 85 72 90 86 87<br />
Generator 90 100 89 83 97 91 75 98 94 73 84 70 96 72<br />
Air cooler 79 71 75 96 89 82 75 96 78 93 96 78 70 72<br />
Turbine F<br />
Turbine G<br />
Turbine H<br />
Turbine I<br />
Turbine J<br />
Turbine K<br />
Turbine L<br />
Turbine M<br />
Turbine N<br />
►Many identical systems is an<br />
advantage when using TCI<br />
►Can provide high quality<br />
condition overview at<br />
component, turbine and wind<br />
farm level<br />
►Makes it possible to optimize<br />
supply support <strong>for</strong> many<br />
identical systems<br />
►Provides necessary input to<br />
maintenance prioritization<br />
and planning<br />
►Enables cost efficient<br />
operations and maintenance<br />
12
TeCoMan Software (by MARINTEK)<br />
Hierarchy<br />
of the<br />
system<br />
Bad actor:<br />
Fuel pump<br />
TCI values of fuel pump<br />
TCI graph of fuel pump<br />
Benchmark graph<br />
of fuel pump<br />
Benchmark of<br />
the fleet<br />
13
WINDSENSE (IPN project 2012-2014)<br />
► Objective<br />
� To introduce a flexible and independent add-on instrumentation<br />
system with advanced use of in<strong>for</strong>mation from sensors and<br />
condition monitoring equipment<br />
► Secondary objectives<br />
� To develop a cost-efficient add-on instrumentation system <strong>for</strong><br />
wind turbines, with focus on pilot turbines<br />
� To test the add-on instrumentation system in pilot turbines<br />
� To provide input to a predictive operation and maintenance<br />
regime through advanced methods <strong>for</strong> data analyses<br />
14
Windsense partners<br />
/15/
Windsense GAP analysis<br />
► Failure modes in the drive train<br />
� with high frequency and/or consequence<br />
� and possibilities <strong>for</strong> improvements in condition assessment<br />
� Prepared by experts associated with the Windsense project<br />
► Failure modes that are candidates <strong>for</strong> further work (examples)<br />
� Generator<br />
• Loose windings, Ground Fault, Internal short circuit, Moisture/dirt<br />
• Loose magnets<br />
• Insufficient cooling<br />
� Converter<br />
• Trip<br />
• Insufficient cooling<br />
� Blade adjustment<br />
• Blade angle signal error<br />
� Rotor system<br />
• Degradation of leading edge<br />
• Cracks in trailing edge<br />
� Main bearing<br />
• Bearing vibration<br />
� Main pumping station<br />
• hydraulic leakage 16
Closing the loop is essential<br />
Degradation Process<br />
Wind Turbines<br />
Offshore Wind Turbine<br />
Onshore Wind Turbine<br />
Collapsed Wind Turbine<br />
Maintenance<br />
Management<br />
System<br />
Wireless Data<br />
Collection Networks<br />
Data Acquisition<br />
Fiber Bragg Grating Sensors<br />
Acoustic Emission(AE)Sensors<br />
Ultrasonic Sensors + AE<br />
Vibration sensors<br />
Maintenance Scheduling<br />
/ Maintenance Optimization<br />
Bee Colony Algorithms<br />
(BCA)<br />
Ant Colony Optimization<br />
(ACO)<br />
Particle Swarm<br />
Optimization (PSO)<br />
Gentic Algorithms (GA)<br />
Meta-Heuristic<br />
approaches<br />
Signal Pre-process<br />
Denosing<br />
Compression<br />
Extract Weak Signal<br />
Filter<br />
Amplification<br />
Key Per<strong>for</strong>mance Indicator<br />
(KPI)<br />
Feature Extraction<br />
Time Domain<br />
Time-Frequency<br />
Domain<br />
Frequency Domain<br />
(FFT, DFT)<br />
Wavelet Domain<br />
(WT, WPT)<br />
KPI<br />
Leading<br />
KPI<br />
Logging<br />
Principal Component<br />
Analysis (PCA)<br />
Fault Prognosis<br />
Auto-regressive Moving<br />
Averaging (ARMA)<br />
Fuzzy Logic Prediction<br />
ANN Prediction<br />
Match Matrix Prediction<br />
Fault Diagnosis<br />
Support Support Machine<br />
(SVM)<br />
Data Mining (Decision<br />
Tree & Association rules)<br />
Artificial Neural Network<br />
(SOM & SBP)<br />
Statistical Maching<br />
17
We make it possible<br />
<strong>NOWITECH</strong> is a joint 40M€<br />
research ef<strong>for</strong>t on offshore<br />
wind technology.<br />
� Integrated numerical<br />
design tools<br />
� <strong>New</strong> materials <strong>for</strong><br />
blades and generators.<br />
� Novel substructures<br />
(bottom-fixed and<br />
floaters)<br />
� Grid connection and<br />
system integration<br />
� Operation and<br />
maintenance<br />
� Assessment of novel<br />
concepts<br />
www.<strong>NOWITECH</strong>.no<br />
18
Change language