Matthew Whittle 1 , Won Shin 2 , Jon Trevelyan 1 , and Junjie Wu 1 .

1. A Parametric Investigation of the Effect of Generator Misalignment upon Bearing Fatigue Life in Wind Turbines Matthew Whittle1, Won Shin2, Jon Trevelyan1, and Junjie Wu1. Future Reliable Renewable Energy Conversion Systems & Networks (FRENS) www.reliable-renewables.com 14th-17th March EWEA 2011 Brussels, Belgium. School of Engineering and Computing Sciences 1. Durham University, 2. Romax Technology Ltd.

2. Collaboration with Romax Technology • Durham University are collaborating with Romax Technology to understand wind turbine generator failures. • Romax are global experts in drivetrain design and simulation. • To date, Romax has designed 11 different multi-megawatt wind turbine gearbox designs ranging from 1.5MW – 5MW. These are used by over 9 different turbine manufactures worldwide.

3. Contents • Wind Turbine Reliability • Wind Turbine Generator Failure • Drivetrain Misalignment • The Turbine Data • Modelling Methodology • Modelling Software • Results • Conclusions • Further Work Offshore wind farm, near Utgrunden, Sweden GE Energy [www.ecomagination.com]

4. Wind Turbine Reliability Public domain data from Landwirtschaftskammer Wissenschaftliche Mess- und Evaluierungsprogramm, reproduced from Introduction to Wind Turbines and their Reliability & Availability. Feng, Y, and Tavner, P. Warsaw : EWEC 2010, 2010.

5. Wind Turbine Generator Failures • A number of large industrial surveys of electrical machine failures show that in low voltage machines bearing failures dominate. • This is supported in the wind industry by a survey of over 800 failed wind turbine generators in the USA which shows a similar picture. Source: Alewine, K., Chen, W., “Wind Turbine Generator Failure Modes Analysis and Occurrence”, Windrower 2010, Dallas, Texas, May 24-26, 2010.

6. Drivetrain Misalignment • Misalignment is a very common problem for rotating machinery; it may be the root cause of 20-30% of downtime. • In wind turbines the gearbox and generator are mounted on rubber bushings • Under large torque the gearbox torque arm bushing deflect (mm) • Rubber sensitive to environmental conditions • Creep • Fatigue • d • Restoring force of coupling must be reacted by bearings. • Failure Mode: classical rolling contact fatigue. Flexible Linksets (low tilt stiffness) Generator Rotor Parallel Misalignment HSS Pinion Brake disc

7. The Turbine • 750 kW variable speed wind turbine considered (though analysis could easily be scaled to multi-MW turbines). • 3 Stage gearbox: 1 planetary stage and 2 helical stages • Gearbox flexibly coupled to a doubly-fed induction generator (DFIG) • Generator supported on two grease lubricated ball bearings. • High speed stage (HSS) of gearbox has one upwind cylindrical rolling element bearing with back-to-back taper rolling element bearings downwind.

8. Flexible coupling Centa Antriebe Kirschey GMBH [www.centa.info] The Flexible Coupling • Coupling usually comprises two ‘linksets’ which are connected by a hollow shaft • It is the linksets which, having a low tilt stiffness, provide the coupling with flexibility • So in modelling the coupling, the key parameter is the tilt stiffness at each end adjoining the two shafts (gearbox HSS, and generator) Flexible Linksets (low tilt stiffness) Generator Rotor Parallel Misalignment HSS Pinion Brake disc

9. Methodology No. Of cycles Torque (kNm) Data Binned  3-D Histogram (T,w,N) Fatigue Damage Calculated according to Miner’s Principle 20 Years Simulated Load Data (from a WT system model) i Load Cases Solved  Bearing Contact Stresses RomaxWIND i = Number of bins (number of operational points considered)

10. RomaxWIND Software • RomaxWIND is a simulation tool and Virtual Product Development environment for wind turbine drivetrains. • Models include: • Detailed gear contact models • Non-linear bearing models • Flexible housings and planet carriers • Flexible shafts, coupling stiffness, mount stiffness, spline models.

11. RomaxWIND Software • An all-in-one system analysis ensures that all interactions between drivetrain components are considered Bearing loads; bearing stiffness Gear loads considering geometry and micro-geometry Housing and planet carrier flexibility Shaft flexibility

12. RomaxWIND Software • RomaxWIND is the first software certified by GL for gear analysis “RomaxWIND is the first software of its kind to meet the stringent certification requirements of GL, and is the result of a number of years of close working between Romax and GL” - Dr Karl Steingroever from GL Renewables Certification

13. Generator Drive End Results Increasing negative misalignment Increasing positive misalignment

14. Generator Non-Drive End Results Increasing negative misalignment Increasing positive misalignment

15. Gearbox HSS Upwind Results Increasing negative misalignment Increasing positive misalignment

16. Gearbox HSS Downwind Results Increasing negative misalignment Increasing positive misalignment

17. Generator DE Generator NDE Increasing negative misalignment Increasing negative misalignment Increasing positive misalignment Increasing positive misalignment Gearbox HSS Upwind Gearbox HSS Downwind Increasing negative misalignment Increasing negative misalignment Increasing positive misalignment Increasing positive misalignment

18. Conclusions • In line with expectation, as coupling tilt stiffness increased the bearing fatigue damage became more sensitive to misalignment. • The gearbox HSS bearings are very heavily loaded and are expected to fail prematurely. • The reduction in damage to the gearbox upwind bearing with increased misalignment and increased coupling stiffness was at the cost of increased fatigue damage to the gearbox HSS downwind bearing. • The generator DE bearing fatigue life was found to be most sensitive to misalignment. • It is recommended to conduct integrated system analyses in wind turbine design and development as the interaction between assemblies is non-trivial and may have a significant impact upon the wind turbine reliability.

19. Future Work The following aspects should be explored in further work: •Characterise coupling stiffness. • Assess impact of axial misalignment. •Testing to verify modelling.

20. Thanks for listening Any Questions? This work was funded by the EPSRC through the FRENS joint UK-China project (www.reliable-renewables.com). References 1. Faulstich, S. et al. Windenergie Report Deutschland. Kassel : Institut fur solare Energieversorgungstechnik, 2008. 2. Landwirtschaftskammer Windenergie. [Online] http://lwksh.de/cms/index.php?id=2875. 3. Reliawind Design for Reliability. Hendriks, Ben. Warsaw : EWEC 2010, 2010. 4. Reliability of wind turbine subassemblies. Spinato, F., Tavner, P., van Bussel, G., and Koutoulakos, E. 2009, Renewable Power Generation, IET, pp. 287-401. 5. Tavner, P., Ran, L., Penman, J. and Sedding, H. Condition Monitoring of Rotating Electrical Machines. London : IET, 2008. 6. Wind turbine generator failure modes analysis and occurence. Chen, W. and Alewine K. Dallas : s.n., May 24-26 2010. WindPower 2010. 7. Introduction to Wind Turbines and their Reliability & Availability. Feng, Y, and Tavner, P. Warsaw : EWEC 2010, 2010. 8. Vibration analysis of misaligned shaft –ball bearing system. Hariharan, V., and Srinivasan, P. s.l. : Indian Journal of Science and Technology, 2009. 9. The Truth Behind Misalignment Vibration Spectra of Rotating Machinery. Ganeriwala, S., Patel, S., and Hartung, H. s.l. : Proceedings of International Modal Analysis Conference, 1999. 10. RomaxWIND. Nottingham, UK : Romax Technology Ltd. 11. GH Bladed. Bristol, UK : GL Garrad Hassan. Matthew Whittle E: m.w.g.whittle@durham.ac.uk T: +44 (0)7792679431