Development of Wind Turbine Fault Detection Algorithms
The research will focus on the analysis of standard wind turbine SCADA data as well as addition signals including vibration, high frequency monitoring of electrical power output, oil analysis and blade strain data (if available) to determine the principal drivers of damage and failure modes building on the results of past UK and EU projects. There will also be an opportunity to undertake small scale experiments using test rigs at Loughborough University and at partners in the UK and abroad (e.g. Supergen Wind partners and contacts in Korea and China).
The data gathered and analysed will be used to develop condition monitoring algorithms based on damage models, trending, FFTs, wavelets and artificial intelligence techniques (e.g. neural networks). There will be an emphasis on estimating expected reliability to target maintenance in large fleets of machines as it is acknowledged that accurate prognosis is difficult.
The projects objectives are:
To analyse wind turbine SCADA and additional condition monitoring data in order to determine failure signatures.
To relate these failures to the root causes.
To develop algorithms to detect the likelihood of future failures through the use of models such as physical damage accumulation equations, trending, FFTs, wavelets, etc.
To carry out measurements on small-scale machines or test rigs to help inform the algorithm development..
Tentative schedule: March – July 2016
Tentative schedule: March – July 2017
Yürüşen, NY, Tautz-Weinert, J, Watson, SJ, Melero, JJ (2017) The Financial Benefits of Various Catastrophic Failure Prevention Strategies in a Wind Farm: Two market studies (UK-Spain), Journal of Physics: Conference Series, 926, 012014, DOI: 1088/1742-6596/926/1/012014. Full text: http://iopscience.iop.org/article/10.1088/1742-6596/926/1/012014 .
Tautz-Weinert, J and Watson, SJ (2016) Comparison of different modelling approaches of drive train temperature for the purposes of wind turbine failure detection, Journal of Physics: Conference Series, 753(072014), DOI: 10.1088/1742-6596/753/7/072014. Full text: http://iopscience.iop.org/article/10.1088/1742-6596/753/7/072014
Tautz-Weinert, J and Watson, SJ (2016) Using SCADA Data for Wind Turbine Condition Monitoring – a Review, IET Renewable Power Generation, ISSN: 1752-1424. DOI: 10.1049/iet-rpg.2016.0248. Full text: http://digital-library.theiet.org/content/journals/10.1049/iet-rpg.2016.0248 or https://dspace.lboro.ac.uk/dspace-jspui/handle/2134/22713
Colone, L, Reder, M, Tautz-Weinert, J, Melero, JJ, Natarajan, A, Watson, SJ (2017) Optimisation of Data Acquisition in Wind Turbines with Data-Driven Conversion Functions for Sensor Measurements, Energy Procedia, 137(C), pp.571-578. DOI: 10.1016/j.egypro.2017.10.386. Full text: http://www.sciencedirect.com/science/article/pii/S1876610217353717.
Ibrahim, RK, Tautz-Weinert, J, Watson, SJ (2016) Neural Networks for Wind Turbine Fault Detection via Current Signature Analysis. WindEurope Summit 2016, Hamburg. Full text: https://dspace.lboro.ac.uk/2134/23014
Tautz-Weinert, J and Watson, SJ (2017) Condition monitoring of wind turbine drive trains by normal behaviour modelling of temperatures. In Abel, Brecher, De Doncker, Hameyer, Jacobs, Monti, Schröder (ed) Conference for Wind Power Drives (CWD 2017), Aachen, pp.359-372, ISBN: 9783743134560. Full text: https://dspace.lboro.ac.uk/2134/24272
Tautz-Weinert, J and Watson, SJ (2017) Challenges in Using Operational Data for Reliable Wind Turbine Condition Monitoring. In The Twenty-seventh (2017) International Ocean and Polar Engineering Conference, San Francisco, California, USA, ISBN: 978-1-880653-97-5. Full text: https://dspace.lboro.ac.uk/2134/24711
Tautz-Weinert, J and Watson, SJ (2017) Combining Model-based Monitoring and a Physics of Failure Approach for Wind Turbine Failure Detection. In 30th Conference on Condition Monitoring and Diagnostic Engineering Management (COMADEM 2017), University of Central Lancashire, UK. Full text: https://dspace.lboro.ac.uk/2134/25071
Weinert, J and Watson, SJ (2016) Condition monitoring by neural network modelling of drive train temperature, 12th EAWE PhD Seminar on Wind Energy in Europe. Full text: https://dspace.lboro.ac.uk/2134/22437
Weinert, J and Watson, SJ (2015) Wind Turbine Fault Detection by Normal Behaviour Modelling, Midlands Energy Consortium Postgraduate Student Conference. Full text: https://dspace.lboro.ac.uk/2134/22532
L. Colone, M. Reder, J. Tautz-Weinert, J.J. Melero, A. Natarajana, S.J. Watson (2017). Optimisation of Data Acquisition in Wind Turbines with Data-Driven Conversion Functions for Sensor Measurements. IEERA DEEPWIND’2017, 14TH DEEP SEA OFFSHORE WIND R&D CONFERENCE, Trondheim, Norway. Download Poster
MSc Energy Engineering
Host Institution: LBORO