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 'Aerodynamics and rotor design' taking place on Wednesday, 12 March 2014 at 09:00-10:30. The meet-the-authors will take place in the poster area.

Alexandros Antoniou Fraunhofer Institute for Wind Energy Systems, Germany
Co-authors:
Neelabh Gupta (1) F P Mareike Strach (1) Alexandros Antoniou (1)
(1) Fraunhofer Institute for Wind Energy Systems, Bremerhaven, Germany

Printer friendly version: printer.gif Print

Abstract

Parametric aeroelastic loads analysis of a near shore 5MW wind turbine

Introduction

The International Electro-technical Commission (IEC) 61400–1 [1] design standard defines the design requirements for land-based onshore wind turbines. This design standard requires that an integrated loads analysis is to be performed when a machine is certified. Such analysis is also necessary to obtain cost-effective wind turbines that achieve favorable performance, and maintain structural integrity by analyzing ultimate and fatigue loads. In the current purview, an aero-elastic analysis is performed to gain an insight on performance and loads with respect to rotor mass imbalance, yaw errors in turbine and startup/shutdown sequences.

Approach

The loads analysis is done with the aero-elastic wind turbine simulation tool FAST (Fatigue, Aerodynamics, Structural and Turbulence) by National Renewable Energy Laboratory (NREL) with a model of the NREL 5-MW Reference Wind Turbine according to [2].

Turbulent wind flows were used for power production design load cases and steady wind flow for startup and shutdown conditions during simulation. ‘FAST’ with ‘AeroDyn’ (stands for Aerodynamics) accounts for the applied aerodynamic and gravitational loads, the behavior of the control system, and the structural dynamics of the wind turbine. The latter contribution includes the elasticity of the rotor, drivetrain, and tower, and the dynamic coupling between the motions of the tower and wind turbine. FAST employs a combined modal and multi-body structural dynamics formulation and has been interfaced with AeroDyn to enable the full aero-servo-elastic modeling of wind turbines, see Fig.1.

Fig.1 Integrated Approach to Modeling Onshore Wind Turbine: Aero-Servo-Elastic Simulation

AeroDyn calculates the aerodynamic forces for each of the blades and models almost all aerodynamic aspects of a wind energy conversion system (WECS) including tower effect, vertical wind shear, and yaw error.

All simulations were run with all appropriate and available degrees-of-freedom (DOFs), including FAST’s two flapwise and one edgewise mode DOFs per blade, one drivetrain torsion DOF, one variable generator-speed DOF, one nacelle yaw DOF, two fore-aft and two side-to-side tower mode DOFs. Aerodynamic imbalances can occur through yaw and pitch errors, twist distribution and rotor mass imbalance. To include manufacturing variability, all simulations include rotor mass imbalance which gives 1P (1-per-rev) excitation to the system. This imbalance is achieved by making one blade 3% heavier and the other 3% lighter than the mass of reference blade [3]. Also, the three yaw errors (-8, 0 and +8 degrees) are incorporated along with stochastic seeds and wind speed bins.


Main body of abstract

Figures 2-7 (excluding Fig.4) show results of several output parameters for power production, start up and shutdown design load cases along with depicting the effects of the aerodynamic aspects in wind energy conversion system (wind shears, yaw error, rotor mass imbalance and turbulence) under consideration. These output parameters are defined as follows:

• “Wind Vxi” represents the nominally downwind component of the hub-height wind velocity.
• “GenPwr” represents the electrical power output from generator.
• “GenSpeed” represents the rotational speed of the generator (high-speed-shaft).
• “BldPitch” represents the pitch angle of Blade 1.
• “YawBrFxp” represents tower top/ yaw bearing fore-aft (nonrotating) shear force.
• “RootMyb1” represents blade 1 flapwise moment (i.e., the moment caused by flapwise forces) at the blade root.
A full-field turbulent wind is applied to the wind turbine to study the effect of turbulence when tower influence and vertical shear are taken into consideration. Fig.2 shows how wind speed fluctuations vary the generator power, speed and loads on tower and blade, when the controller tries to compensate these variations.

Fig.2 Time Series System Response for Power Production Operation Showing

Fig.3 depicts the variation of the above mentioned output parameters with respect to imbalances in rotor mass (+3% and -3% variation in two blades with respect to reference blade). This causes approximately, 1.2 % variation in power output and 5 % variation in tower top forces when compared to wind turbine system without any rotor mass imbalances.

Fig.3 Time Series System Response Showing Rotor Mass Imbalance (+/-3 %) Effects for Steady Hub-Height Wind Speed of 9 m/s

A yaw error can significantly increase the generator power and loads fluctuations in the system. The effect has been depicted in Fig.4 (δ − γ). The wind turbine yaw error is simulated taking no control action for conpensating the wind direction changes.

Fig.4 Horizontal wind shear and Yaw Error (positive δ − γ) [6]

As can be seen in Fig.5, the yaw error decreases the average output power and increases the load oscillations (“YawBrFxp” and “RootMyb1”).

Fig.5 Time Series System Response Showing Yaw Error Effects for Steady Hub-Height Wind Speed of 9 m/s

This figure depicts that the power and load oscillation amplitudes are not symmetrical with respect to yaw error due to blade rotation direction and dynamics of machine. This is mainly due to the application of positive and negative torque about the torque axis (which lies in the plane of rotor) for two yaw errors. This is mainly attributed to asymmetry of flow caused due to yaw error and wind shear. The effect of vertical wind shear and tower influence is not analyzed separately as it is something that will always occur and is not under user’s control at present.

The frequency of oscillations seen in the “GenPwr”, “GenSpeed and “YawBrFxp” plots (Fig.3) around 0.17 Hz which is the 1P frequency for a rotor speed of 10.3 rpm and is around 0.52 Hz in Fig.5, which is the 3P frequency for a rotor speed of 10.3 RPM (wind speed = 9 m/s).

For the design load situation of startup and shutdown cases, cut-in, rated and cut-out velocities have been used as they are assumed to represent the major range of transient loads. For tower top force (YawBrFxp) and root flap wise bending moment (RootMyb1), the first observation that can be made is, they are larger for 11.4 m/s when compared to 25 m/s. This response characteristic (Fig.6) is the result of peak in rotor thrust at rated speed due to the change in the control logic at this speed.

Fig.6 Time Series System Response for Start-up Operation (Hub-Height Wind Speed of 3 m/s, 11.4 m/s & 25 m/s)

The perceivable fluctuations in ‘YawBrFxp’ and ‘RootMyb1’ are only observable after the start of the generator for startup design load cases and the frequency of these fluctuations matches tower fore aft motion and rotor angular motion (tower dam, gravity and wind shear effects) respectively. The smaller fluctuations at around 200s give way to steady and larger ones for ‘YawBrFxp’ due to computational transients at turbine start up. Similar reasoning can be applied to fluctuations in ‘YawBrFxp’ and ‘RootMyb1’ for shut down cases (Fig.7) with the behavior being opposite of startup cases.

Fig.7 Time Series System Response for Shutdown Operation (Hub-Height Wind Speed of 3 m/s, 11.4 m/s & 25 m/s)

Conclusion

Comparing the results, for yaw error analysis, it can be concluded that -8° yaw error has more impact than +8° due to larger loads and power fluctuations for the considered NREL 5MW turbine (more in-depth analysis is required to assess this nature of variation). It has to be noted here that this is the case for clockwise rotation of turbine. For anticlockwise rotation, the positive yaw errors would be more damaging than negative ones in terms of loads. Again, rotor mass imbalances (+3%, 0% and -3% in three blades) seem to have a worse effect with respect to yaw errors (+/- 8°) in terms of power and load fluctuations. But both these effects are subject to the amount of rotor mass imbalance and yaw errors induced during the study. They can change by the variation of these parameters itself. Start up and shutdown operations have a worst influence on loads at rated wind speed when compared to cut-in and cut-out speed. A fundamental conclusion can be drawn about the asymmetry of flow caused due to yaw errors, wind shear and tower influence: they cause 1P excitation in the rotating frame (rotor loads) which is translated to 3P excitations in fixed frame. On the other hand, rotor mass imbalance causes 1P excitation in fixed reference frame (generator speed, generator power and tower loads).

To study the effect of stochastic load on the blade structure, a fatigue analysis would give more insight into structural integrity, as it is difficult to comment anything, due to the inherent randomness of these turbulent induced loads.



Learning objectives
This project will help the readers to understand the importance of manufacturing variation (mass imbalance), control issue (yaw misalignment) and transient operations (startup and shutdown of turbine) on loads acting on a wind turbine.


References
1. Wind turbines - Part 1: Design requirements, IEC 61400-1, Geneva, 3.0 edition, 2005-08.

2. Jonkman, J., S. Butterfield, W. Musial, and G. Scott (2009, February). Definition of a 5-MW Reference Wind Turbine for Offshore System Development. Technical Report NREL/TP-500-38060, National Renewable Energy Laboratory.

3. IEC. Wind turbines - Part 5: Wind Turbine Blades. IEC 61400-5, Revison H (Draft Version).