Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Advanced drive trains technologies' taking place on Tuesday, 11 March 2014 at 16:30-18:00. The meet-the-authors will take place in the poster area.

Ralf Schelenz RWTH Aachen University, Germany
Co-authors:
Ralf Schelenz (1) P Stefan Franzen (1) F Dominik Radner (1)
(1) RWTH Aachen University, Aachen, Germany

Printer friendly version: printer.gif Print

Abstract

Hardware in the loop operating mode for full size nacelle testing

Introduction

Dynamic loads on wind turbine drivetrains caused by wind and grid effects lead to poorly predictable deformations and complex vibrations in the entire drivetrain and must be taken into account regarding reasonable load calculations. The steadily increasing importance for system tests of wind turbine na-celles and drivetrain components are recognised by the OEMs like Vestas [1] or Areva [2], since this approach can lead to a higher quality and reliability. Beside the existing system test benches in the USA, Spain [3] and Germany, additional test benches are being built up in Denmark [4], England [5], USA [6] and Germany [7].

Approach

For the purpose of testing a Wind Turbine Generator (WTG) under real load conditions, a demonstrator test bench of 1 MW nominal power has been developed at RWTH Aachen University (Fig. 1) [8]. The test bench uses a combination of a high speed motor and a modified WTG gearbox to apply high torques to the WTG rotor shaft. Therefore the load capability of the prime mover attains of the nominal input torque of the WTG.

Fig. 1: 1 MW WTG system test bench
To provide realistic loads at the rotor flange of the nacelle, an additional four degree of freedom load application system (LAS) has been developed (Fig. 2). It consists of four servo hydraulic actuators with a nominal force of 160 kN each. While the LAS is capable of emulating the thrust, tilt and yaw bending loads of the synthetic wind field such as the static rotor weight, these loads are related to the virtual point of load application where all three blade coordinate systems intersect. As the rotor hub and the blades are not existent at the test bench, this point is located on the middle axis of the shaft flange about 200 mm left of the red marked triangular structure.

Fig. 2: Hydraulic load application system
In detail, the wind field and force calculation is done at the rotor system level. An entire 3D wind field is simulated which based on the manual input parameters of average wind speed, turbulence intensity and oblique-flow. Including the structural and aerodynamic properties of the rotor blades, these loads at the load application point are calculated using the blade element momentum theory (BET). As a result of the validation measurements, it has been identified, that the LAS performs best at operation points with high values of average load and that in most cases dynamical excitations greater than 3 Hz can be applied without violating an error criterion of 2 % [9].


Main body of abstract

At this moment we little know about the overall system behaviour and the complex interaction of the drivetrain at uncharted circumstances. The lack of knowledge requires particular realistic ground tests of nacelles to clarify reliability issues. Beside the accuracy and performance of the LAS, the 1 MW nacelle test bench focuses on the verification, that a full nacelle can be tested inside a multi physical hardware in the loop (HIL) environment. This special feature offers great advantages since the HIL control system provides both mechanical and electrical loops.

The nacelle is completely integrated into the HIL system which consists of four interactive units (Fig 3) - the HIL wind load calculation, the test bench interface, the mechanical test bench and the nacelle includ-ing the WTG controller. Both prime mover and LAS are controlled by the test bench. The load applica-tion is capable to fulfil the dynamic requirements of wind load application using the HIL operation mode. Within this mode, the input parameters on the rotor side define the wind field in which the nacelle is operating while the independent real-time grid simulation emulates the grid behaviour. The target val-ues for the grid-side converter can be provided by the test bench interface as well as for “Fault Ride Through” (FRT). Because the implementation of all these models and calculation steps are done in real time, it is possible to run the nacelle with its own controller as used on the tower.

Fig. 3: Hardware in the loop nacelle integration in the test bench
Especially the transient effects like gusts as well as critical operating points like grid connection, the generator configuration switching and the emergency stop are significant load cases in the certification. These effects on the drive train deviate significantly. In order to investigate the system behaviour of the drivetrain, those effects have been analyzed in the HIL test bench operation and can be emulated quite accurately.

Some of the successful measurement campaigns demonstrate the analytical capability of the HIL oper-ating mode. During various wind field simulations, the pitching strategy of the WTG controller was ob-served as well as the switching of the generator configuration. In order to increase and optimise electric power output, the generator switches from star to delta connection. Therefore the controller pitches the blades out of wind, disconnects, synchronises and reconnects the generator while the sensors record all mechanical effects related to these events. Especially the torque and bending moments of su-perposed wind turbulences are measured and analysed at critical drivetrain components.

One of the most interesting issues of overload testing is the mechanical system respond to a 100% voltage dip of the grid [10]. The measurement results in figure 4 show the critical torque increase. Depending on the drop speed of the voltage dip, the amplitudes of torque and amperage rise quite differently.

Fig. 4: Torque respond at different drop speeds of 100 % voltage dip
The upper diagrams show the different drop speeds of the voltage dip and the lower diagrams show the reaction of the torque at the high speed shaft. Slow voltage loss cause little rise of amperage and torque whereas the torque increase at higher drop speed can reach about 70 %. This conclusion has great impact on the drivetrain design considering that a voltage dip might occur in less than 10 ms. If the WTG does not have adequate FRT ability, the torque overload might harm the mechanical compo-nents.
For all shown events the behaviour of the nacelle correlates to the documentation of the control strategy of the controller. Even the deviation from the simulated progress is negligible. The evaluation of several system tests under different conditions using the HIL operating mode shows, that the dynamic performance of the LAS is adequate to apply realistic dynamic wind loads within the relevant frequency range of 5 Hz [11].



Conclusion

The experience shows, that size does matter. While several influences are unaffected by the size, most of the relevant effects like global deformation on local contacts or slipping have to be analysed within full size tests. Unlike most of the other testing facilities, the unique RWTH test bench is capable of doing flexible and especially realistic test campaigns using the multi physical HIL operating mode.

Beside the advantages of the LAS like adjustable, definable and reproducible loads, all synthetic test cycles are mechanically and electrically independent from wind and grid states. Time scaling tests come into consideration and allow a deep insight into the overall system behaviour and the relevant load cases for gear and bearing components within the drivetrain. One of the main objective is to analyse several damage mechanisms with high accuracy in order to understand them and give advice how to improve existing and future designs. That is why the test bench focuses on comprehensive functionality with good observability under real conditions.

Using the original WTG controller of the specimen offers the opportunity to analyse strengths and weaknesses of the control strategy while the WTG behaves unaffectedly. The controller itself has a major influence on the loads of the nacelle by controlling the generator, pitch and yaw system. On the other side, the grid simulator provides advanced FRT test functionality within the HIL environment to examine the FRT performance in many conceivable ways.

Based on the results of the measurement campaign on the 1 MW demonstrator, a 4 MW WTG system test bench has been designed [12]. This new improved test bench is located at the campus of the RWTH Aachen University and will be operated by the Center for Wind Power Drives.



Learning objectives
In addition to the HIL system for the aerodynamic loads, a HIL system for the real-time calculation of grid loads and system perturbation has been developed and tested at the 1 MW demonstrator. Both simulation environments will be ported to the 4 MW system test bench and integrated into the control system of the test bench to create the most advanced testing facility for on-shore wind turbines.


References
[1] Vestas “Verification Testing.” available at
http://worldofwind.vestas.com/en/verification-testing, last access October 2013.
[2] Areva “Offshore wind turbines - AREVA’s 5 Megawatt full load test bench in operation since October 2011” available at
http://www.areva.com/EN/news-9108/offshore-wind-turbines-arevas-5-megawatt-full-load-test-benchin-operation-since-october-2011.html, last access October 2013
[3] Pascual, P. A. “Operational experience in wind turbine development test benches.” 4th International Conference Drivetrain Concepts for Wind Turbines, 14-16 October 2013, Bremen.
[4] Quitter, J. “Vestas picks 10MW Lorc centre for V164 test” Wind Power Monthly, Oct. 2013.
[5] Narec “Testing - Turbine Drive Trains.” available at
http://www.narec.co.uk/testing/turbine-drive-trains, last access October 2013.
[6] Mander, A. “Clemson WTDTF: focus on testing and product validation.” Presented at Conference for Wind Power Drives CWD, 19-20 March 2013, Aachen.
[7] Pilas, M. “Wind turbine component vs. full nacelle testing.” 4th International Conference Drivetrain Concepts for Wind Turbines, 14-16 October 2013, Bremen.
[8] Bosse, D., Barenhorst, F., Radner, D. and Schelenz, R. “Analysis and Application of IEC 61400 orientated Wind Loads for Full Scale Ground Testing.” Conference for Wind Power Drives CWD, pp. 103-123, 19-20 March 2013, Aachen.
[9] Schelenz, R., Bosse, D., Radner, D. and Jacobs, G. “Demands on dynamics of LAS (Load Application System) for full scale ground testing of 1 MW Wind Turbines.” DEWEK, 7-8 November 2012, Bremen.
[10] Vollmer, R. “Test Systems.” Presented at Conference for Wind Power Drives CWD, 19-20 March 2013, Aachen.
[11] Schelenz, R., Barenhorst, F. and Berroth, J. “F&E-Modelle für WEA-Antriebstechnik.” 16. Fachkongress Zukunftsenergien Windenergie, E-World Energy and Water, 7 Februar 2012, Essen.
[12] Bosse, D., Radner, D., Schelenz, R. and Jacobs, G. “Analysis and Application of Hardware in the Loop Wind Loads for Full Scale Nacelle Ground Testing.” DEWI Magazin, 43, pp. 65-70, 2013.