Concept for Offshore Wind Turbine Foundation Monitoring

Transcription

1 Concept for Offshore Wind Turbine Foundation Monitoring World of Technology & Science Wim Hendricx Utrecht, October 4-7 th, 2016 Siemens PLM/STS/LMS Engineering

2 Introduction Classical Condition Monitoring Model Based Analyses Page 2

3 LMS Engineering in the Siemens family Siemens Wind power Siemens PLM LMS Engineering Page 3

4 Engineering Services Deployment Services Siemens PLM/STS Simulation and Testing Solutions An unique Portfolio of Engineering Innovation Design - CAD NX CAE, LMS Virtual.Lab, LMS Samtech, CD-adapco 3D Performance Simulation Controls LMS Imagine.Lab Mechatronic System Simulation System Synthesis System Data Management Multi-physics Modeling LMS Test.Lab Test Based Engineering Laboratory Mobile LMS SCADAS Modal NVH Acoustics Durability LMS Test.Xpress Page 4

5 Foundation monitoring Germany: Foundation Monitoring is a legal requirement by the Bundesamt für Seeschiffahrt und Hydrographie (BSH). The monitoring must cover in 10% of the wind turbine foundations in the wind park. There is no guideline how to monitor. The offshore wind park operators are mainly looking at: 1. Acceleration and inclination of the transition piece and tower 2. Loading at transition piece 3. Scour development around the foundation 4. Corrosion of the foundation A permanent Structural Condition Monitoring System (CMS) is most common. Siemens PLM has elaborated a concept to extend the measurements with Model Based Analyses Page 6

6 Introduction Data Classical acquisition Condition Monitoring Model Based Analyses Page 7

7 Monitoring concept Robust system: Siemens automation components Local PLC based acquisition of sensor signals; complemented with wind turbine controller data (SCADA). Central server with WinCC and LMS Test.lab software for automatic analysis, trending, alarm generation, documentation and data storage. Page 8

8 Acquisition systems and server Robust system: Siemens Factory Automation components At each wind turbine: SIMATIC ET200SP PLC Analog signal acquisition Local storage UPS IP67 cabinet. Onshore: central server Industrial PC Redundant capabilities. Digital interface for SCADA system SIMATIC WinCC software for automatic Data analyses & storage Trending Alarm generation Web based app for remote access Page 9

9 Typical system setup at each wind turbine Sensor Shielded Cupper wire Marine UTP cable IP 67 box* SIMATIC ET 200SP PLC + HDD/SDD Interface node acquisition module UPS * Page 10

10 Typical instrumentation at the transition piece for loading at the monopile, vibration & inclination Sensors Inclinometer DC accelerometers. Strain gages Ruggedized version of sensors Strain gages redundancy is foreseen. DC accelerometer 2D Inclinometer Strain gage vertical Strain gage horizontal CMS system Number of measurement channels - Strain gages: 3x2x4=24 - Accelerometers: 2 - Tilt sensor: 1 (2 directions) => 28 channels 1.5m above airtight deck Bottom of TP Page 11

11 Typical instrumentation for differential displacement between transition piece and monopile Displacement sensors Vertical direction: 3 sensors under 120 degrees Tangential direction: 2 sensors under 180 degrees All sensors below the watertight deck in the TP Inductive sensor vertical direction Inductive sensor tangential direction Number of measurement channels - Inductive sensor: = 5 Page 12

12 Scour development around the monopile Ultrasonic sensors 3 sensors under 120degrees; 5 m under LAT Installed in on-shore conditions Analog output signal to synchronize with other sensors Number of measurement channels - Echosounder: 3 Page 13

13 Corrosion and chemical monitoring inside the monopile Sensors commonly used in the chemical industry Electrical Resistance (ER) Probes and reference coupons Digital interface of ER data acquisition system to Siemens ET 200SP PLC system Dissolved oxygen ph Principle of membrane-covered amperometric Temperature PT100 sensor. Water level Ultrasonic sensor. All sensors are installed below the water tight deck. Page 14 Number of measurement channels - ER probe: 2 (digital) - O 2 : 1 - ph: 1 - Temp: 1 - Water level: 1

14 Tracking of resonance frequencies Scouring affects the global bending frequencies Analysis in LMS Test.lab Frequency domain data => Operational Modal Analysis (OMA) Resonance frequency, mode shape & damping Complements WinCC analyses Running in background; seamless integrated in WinCC WinCC visualizes analysis results Trending of WTG resonance frequencies Page 15

15 SCADA data Turbine specific information (rpm, torque, power, yaw angle, pitch angles, wind speed, wind direction, ) Synchronized acquisition with other measurement signals. The WinCC software communicates with the turbine s controller (Siemens or third party). Correlation between the operating conditions of the turbine and the dynamic responses of the installation. E.g. - Inclination versus wind speed - Resonance frequency versus water level (tide) and temperature. Page 16

16 Trending, alarming, visualization & archiving in WinCC Display of current or historical data. Process values in tabular form or trend display. Real-time statistics max/min, average (weighted) mean Integral, total, RMS. Messages display Freely configurable WinCC Alarm Control Accessibility Control and monitoring of plant processes via internet or an intranet using mobile devices (tablet PCs or smartphones). Historical process information Process value archives in SQL database. Time series are stored and available for detailed analysis. Page 17

17 Introduction Classical Condition Monitoring Model Based Analyses Model Based Analyses Page 18

18 Extension: Model Based Analyses Page 19 Classical CMS algorithms based on Experience, mainly with rotating machinery (bearings) Reference measurements in a normal operating condition. Identification of deviations from an initial condition. Applicable to offshore wind turbine foundation? E.g. shift in resonance frequency! What is underlying mechanical change in the structure? Is this a problem? Is action required? Deeper insight in the global condition by Model Based Condition Monitoring. Measurements completed by model-based data. Information at locations where no sensors are installed. Prerequisite: Validated model Comparison of measurements and predicted values can identify a degrading component. Algorithms for remaining lifetime

19 Extension: Model based monitoring 1. SWT model creation and creation & updating LMS SAMCEF for Wind Turbine (SWT) Jacket/monopile + Tower + Drivetrain Wind loads and wave loads. Input data: Detailed FE models and/or template for components Validation of model by measurements Fast & easy All Methods, models and analysis encapsulated into a user-friendly platform Page 20 Open User components integration Platform Customizable Cost effective Most efficient process Reliable & accurate Fully Coupled Approach thanks to FMBD Consistency with test results

20 Concept for Model Based condition monitoring of the structures. 1. SWT model creation and creation & updating Basic SWT model Operational measurements turbine Example Offshore SWT model Correlation & updating General dynamic information (resonance frequencies, damping, ) OMA* Validated SWT model *OMA: Operational Modal Analysis Page 21

21 Concept for Model Based condition monitoring of the structures. 2. Condition Monitoring Algorithm Development Damage scenarios Validated SWT model Environmental parameters Correlation Variation & dynamic updating properties Variation dynamic properties Discrimination Analysis Page 22 Robust CMS algorithm

22 Concept for Model Based condition monitoring of the structures. 3. Experienced fatigue loads Measured responses/loads (partial information) Validated SWT model Complemented macroscopic loads Loads at (sub) component level Rainflow counting Design loads Experienced fatigue loads Consumed lifetime Page 23 Degree of aging

23 Concept for Model Based condition monitoring of the structures. 4. Remaining lifetime estimation Measured responses/loads (partial information) Validated SWT model Complemented macroscopic loads Detailed FE model critical zones Example local stresses around a WT door Responses/loads History (consumed lifetime) Expected lifetime Remaining lifetime Page 24

24 Concept for Model Based condition monitoring of the structures. 5. Damage scenarios: Scouring Periodic sonar scans around the 4 piles of a jacket Change of bending mode frequencies as a function of scour depth Scour detection methods Page 25

25 Concept for Model Based condition monitoring of the structures. 5. Damage scenarios: Grout monitoring of monopiles or jackets Jacket Grout connection under water: no sensors Monitoring of general dynamic behavior: tracking resonance frequencies and mode shapes Monopile Local sensor monitoring is possible Relative displacement of monopole/transition piece. Strain links over the connection Monitoring of general dynamic behavior: tracking resonance frequencies and mode shapes Page 26

26 Concept for Model Based condition monitoring of the structures. 5. Damage scenarios: Monitoring of bolts Most critical location: Interfaces to transition piece (tower monopole/jacket) Challenge: Monitoring of the preload in 72 bolts. Instrumentation at all bolts is not realistic! Model Based approach: FE simulations to identify pre-load sensitive parameters (frequencies, mode shapes, levels,..) Preload loss effect only detected at higher modes. Dedicated instrumentation remains necessary Higher order modes of the monopole/transition piece/tower Mode (1,0) Mode (2,0) Mode (3,0) 16 additional sensors are proposed to monitor changes in higher order circumferential modes of the transition piece Mode (4,0) Page 27

27 Thank You World of Technology & Science Wim Hendricx Utrecht, October 4-7 th, 2016 Siemens PLM/STS/LMS Engineering