3/30/20101Turbine Condition MonitoringAndrew KusiakIntelligent Systems Laboratory2139 Seamans CenterThe University of Iowa Intelligent Systems LaboratoryThe University of Iowa Iowa City, Iowa 52242 - [email protected]: 319-335-5934 Fax: 319-335-5669http://www.icaen.uiowa.edu/~ankusiakWhy Condition Monitoring of Wind Turbines? Eliminate unscheduled maintenance activities Scheduled maintenance during low or no-windThe University of Iowa Intelligent Systems Laboratoryperiods Reduce the number of site visits for conditionsassessment Reduce overtime expenses and production lossesWhy Condition Monitoring of Wind Turbines? Reduce failure rate of components and systems Reduce the inventory volume of components Reduceunscheduledequipment use, e.g.,cranesThe University of Iowa Intelligent Systems LaboratoryReduceunscheduled equipment use, e.g., cranes Prevent secondary damage Extend component life-time Maximize profitDamages to the Drive TrainThe University of Iowa Intelligent Systems Laboratory3/30/20102Retrofit Experience in Gear TechnologyOverview Upgrade PSC 1002New partsThe University of Iowa Intelligent Systems LaboratoryNearly every manufacturer of wind turbine gears considers retrofit arrangementsStronger bearingspRe-machine partsNew bearing typesExample: Off-line Solutions for Wind Turbine MonitoringThe University of Iowa Intelligent Systems LaboratoryVIBSCANNER® 1-channel FFT datacollector and analyzerOMNITREND® PC softwareVIBXPERT® 2-channel FFT data collector and machine analyzerPortable SolutionsFFT = Fast Fourier TransformExample SpectraFrequencySample sizeNoiseFrequencySample sizeNoiseFrequencySample sizeThe University of Iowa Intelligent Systems Laboratoryhttp://demonstrations.wolfram.com/FrequencySpectrumOfANoisySignal/Sample sizeNoiseExample: On-line Solutions for Wind Turbine MonitoringThe University of Iowa Intelligent Systems LaboratoryVIBROWEB® XPOnline CMS GL certified (European) Machines with variableand constant speed Onshore and offshoreOMNITREND®PC softwareVIBNODE®Online CMS Machines withconstant speedup to 1500 kW Low cost solutionCMS = Condition Monitoring System3/30/20103Types of Vibration Data MonitoredConstant speed machines: High resolution spectraEnvelope spectraThe University of Iowa Intelligent Systems LaboratoryEnvelope spectra Cepstrum Time waveformTypes of Vibration Data MonitoredVariable speed machines: High resolution order spectraThe University of Iowa Intelligent Systems Laboratory Order envelope spectra Order Cepstrum Time waveform High resolution spectraTorque MeasurementSensorsignalamplifierRotatingantennaStrainThe University of Iowa Intelligent Systems LaboratorygaugeStationarypickupantennaAnalysis unit Data analysis: Rain-flow methodRainflow MethodThe rainflow-counting algorithm (method) is used in the analysis of fatigue data in order to reduce a spectrum of varying stress into a set of simple stress reversalsFatigue AnalysisThe University of Iowa Intelligent Systems LaboratoryUniform alternating loading Spectrum loadinghttp://en.wikipedia.org3/30/20104Rainflow Algorithm1. Reduce the time history to a sequence of (tensile) peaks and (compressive) troughs.2. Imagine that the time history is a template for a rigid sheet (pagoda roof).3. Turn the sheet clockwise 90°.4. Each tensile peak is imagined as a source of water that "drips" down the pagoda.5. Count the number of half-cycles by looking for terminations in the flow occurring when either: a) It reaches the end of the time history;b)It merges with a flow that started at an earliertensile peak;orThe University of Iowa Intelligent Systems Laboratoryb)It merges with a flow that started at an earlier tensile peak; orc) It flows opposite a tensile peak of greater magnitude.6. Repeat Step 5 for compressive troughs.7. Assign a magnitude to each half-cycle equal to the stress difference between its start and termination.8. Pair up half-cycles of identical magnitude (but opposite sense) to count the number of complete cycles. Typically, there are some residual half-cycles.http://en.wikipedia.orgRainflow AlgorithmStress (MPa)Whole cyclesHalf cycles10 2 013 0 116 0 2The University of Iowa Intelligent Systems Laboratory17 0 219 1 020 0 122 0 124 0 127 0 1Rainflow analysis of tensile peaksRainflow analysis of compressive troughshttp://en.wikipedia.orgRainflow AlgorithmMethods similar to the rainflow method: Range-pair counting Hayes method The University of Iowa Intelligent Systems Laboratory Range-pair-range counting  Ordered overall range counting  Racetrack counting  Hysteresis loop countingAlignment MonitoringMonitoring of the alignment between the generator and the gearbox due to changing operating loads:The University of Iowa Intelligent Systems Laboratory Permanent or temporary Relative and absolute3/30/20105Basic MeasurementsAccelerationPower-LineDrive - Type - VIB 6.195ICP-Type - VIB 6.172The University of Iowa Intelligent Systems LaboratoryBasic MeasurementsDisplacementInductive sensor - VIB 5.991-DISSensorThe University of Iowa Intelligent Systems LaboratoryBasic MeasurementsNumber of revolutions per minuteInductive sensor - VIB 5.992-BAThe University of Iowa Intelligent Systems LaboratoryVibration MeasurementsParameters: Velocity (peak, RMS) Acceleration (peak, RMS) Displacement Shock pulse (bearing condition) RPMThe University of Iowa Intelligent Systems Laboratory Time waveform Spectrum (amplitude, envelope) Cepstrum (amplitude, envelope) Order analysisRMS = root-mean-squareRPM= revolutions per minute3/30/20106Amplitude TrendingTurbine 1Turbine 2The University of Iowa Intelligent Systems LaboratoryWhy narrow-band trending? Condition diagnosis for the specificmachine component frequenciesWhy differentiate operating states? Identification of 'critical' operatingconditionsAlarm Level Adapts to the Operating StateOp. state 1 Op. state 2500 - 1000 rpm 1001 - 1500 rpmGenerator speedThe University of Iowa Intelligent Systems Laboratorypp5 - 10 m/s 10.5 - 15 m/s100 - 500 kW 501 kW - 2500 kWWind velocity (0-20mA)Generator power (0-20mA)Lightning ProtectionThe University of Iowa Intelligent Systems LaboratoryProtection for up to 5000 AResistorTVS diodeGas discharge arresterTVS = (Transient Voltage Suppression) diodes that safeguard low voltage circuits against damage or latch-up caused by electrostatic discharge (ESD),