Delegates are invited to meet and discuss with the poster presenters during the poster presentation sessions between 10:30-11:30 and 16:00-17:00 on Thursday, 19 November 2015.
Lead Session Chair:
Stephan Barth, ForWind - Center for Wind Energy Research, Germany
Kun Marhadi (1) F Carsten Andersson (1) Alexandros Skrimpas (1)
(1) Brüel & Kjær Vibro A/S, Nærum, Denmark
Printer friendly version: Print
Presenter's biographyBiographies are supplied directly by presenters at EWEA 2015 and are published here unedited
Kun Marhadi is a development engineer in the Remote Monitoring Group at Brüel & Kjær Vibro. He joined Brüel & Kjær Vibro in 2012. Previously, he worked as a postdoctoral fellow in the Department of Mathematics at the Technical University of Denmark (DTU). He received his PhD in computational science in 2010 from San Diego State University and Claremont Graduate University. He has M.S. and B.S. in aerospace engineering from Texas A&M University. Dr. Marhadi's expertise is in structural vibration and analyses, probabilistic methods, and design optimization.
PosterDownload poster (8.40 MB)
Lessons Learned and Experience from Condition Monitoring of 6000 Turbines
Condition monitoring system (CMS) has become an important part of wind turbines operation and maintenance. It helps detect incipient faults early so that maintenance can be well planned in order to minimize downtime. Wind turbine CMS typically involves monitoring large number of the same type of turbines over the same or different geographical areas. Thus best practice of monitoring a particular type of turbine can be learned from experience of monitoring a large number of its type. Currently Brüel & Kjær Vibro is monitoring around 6000 turbines all over the world from different wind turbine manufacturers. Monitoring this large number of various types of turbines poses many challenges. Over the years, Brüel & Kjær Vibro has learned to overcome the challenges and to work toward improving wind turbine CMS.
With accelerometers installed at different components of wind turbines, various vibration measurements are trended over time. Alarms will be generated to indicate possible incipient faults when vibration trends cross pre-defined limits. Different turbine type requires different monitoring strategy. To monitor large number of turbines and to avoid potential alarm flooding, Brüel & Kjær Vibro has developed a system that allows storing all important vibration data (including time waveforms), a platform for engineers monitoring the turbines to share analyses of potential turbines’ faults, and a simple reporting system that produces actionable recommendation for turbine maintenance.
Main body of abstract
Monitoring large number of turbines necessitates a system that helps avoid alarm flooding (automatic alarm management) and to not miss any incipient faults at the same time. Diagnostic engineers play a very important role in evaluating incoming alarms, performing faults diagnosis, and giving actionable recommendation for turbine maintenance. Their experience and expertise not only help detect common faults detected by CMS but also unusual faults that can potentially cause costly downtime. Such examples will be presented. Feedback from technicians is important in improving fault diagnosis because they help corroborate analysis with actual findings in the turbines. Brüel & Kjær Vibro has developed a system that integrates all previously mentioned factors so that monitoring large number of turbines is effective and can be improved over time. Along with important vibration data and relevant information from turbine controller, all monitoring activities (e.g. incipient fault reports, observation, and analyses of the engineers) are kept in a database. That way, engineers have easy access to learn the likelihood of certain faults in certain turbine types and how the faults may develop to the point they may cause secondary damages. Comparison of faults development between similar turbines can easily be done. Having all relevant data and associated faults (i.e. availability of relevant statistics) provides an essential foundation for machine learning in order to achieve more intelligent monitoring.
Monitoring a large number of turbines allows gaining valuable experience in understanding the likelihood of certain faults to develop in certain turbine types, the development of the faults, and in detecting other faults that may not commonly detected by CMS. It is essential to store all relevant information pertaining to faults detection and diagnosis in a database. The information can be used for learning purposes and as a reference for future improvement.
This paper shows the importance of knowledge sharing in monitoring wind turbines in order to detect incipient faults. Knowledge sharing is important not only among diagnostic engineers who analyze the turbines remotely but also between diagnostic engineers and field technicians who provide important feedback to the engineers regarding actual turbine condition. The information is essential for better diagnostic and recommendation for turbine maintenance.