1. Conceptual research of a downwind turbine, based on a Suzlon 2
Conceptual research of a downwind turbine, based on a Suzlon 2.1MW onshore turbine
Gesine Wanke* & **, Torben J. Larsen**, Morten Hansen***, Thomas Buhl*, Jens I. Madsen*, Leonardo Bergami*
or:
* Suzlon Blade Science Center, Havneparken 1, 7100 Vejle
** DTU Wind Energy, Frederiksborgvej 399, 4000 Roskilde
*** SDU Mads Clausen Institute, Alsion 2, 6400 Sønderborg

2. Why downwind? Potential Risk: Blade – tower clearance
 More flexible blades
 Less material
Passive yaw alignment
 Reduction in LCOE
Risk:
Tower passage
 Noise
 Fatigue loads
Alignment center of gravity to thrust
 Extreme load
Power reduction

3. Research Project How much can be saved on LCOE by a change of conept?
Industrial PhD-project
Start Dec. 2016, End Dec. 2019
Suzlon Blade Science Center & DTU
Conceptual research
Baseline: existing 2.1MW onshore turbine
Changes in design driving loads
Stability of a free yawing turbine
Change of the rotor design
Tower shadow (noise and fatigue load) alleviation
Control strategy
Industrial PhD
IEA Wind task 40
Cooperation between Company & University founded by Innovation Fund
We would like to join the IEA task with our project
 Jump to slide 9 if there is only 10 minutes

4. Change in design driving load cases
wind direction
upwind turbine
S111 class IIIA 2.1MW turbine
1 turbine – 2 configurations
Design load bases (DTU)[1]
HAWC2 simulations
Basic DTU controller [2]
downwind turbine
wind direction
Both Configurations:
Cone and tilt away from the tower
Difference:
Prebend towards Suctions side: more initial tower clearance uw, less initial tower clearance dw
According to IEC:
Inclination angle = 8°, mass and aerodynamic imbalance
DTU controller: adjusted routine for stuck pitch angle 2.2y

5. Tower clearance – Design driving load cases
Normalized tower clearance
Upwind DLC 2.2y
Downwind DLC
Decrease to 62%
wind direction
upwind turbine
downwind turbine
Normalized with the unloaded tower clearance of upwind configuration
Design driving load cases
Highly loaded blades upwind
Unloading situations downwind
DLC 2.2y abnormal yaw error DLC 2.1 grid loss during operation

6. Extreme load – Changes in design driving load cases
Tower bottom flange moment
Longitudinal DLC 4.2 to %
Lateral DLC %
Torsion DLC %
Blade root moment
Flapwise DLC %
Edgewise DLC 2.2y - 5%
Torsion DLC 2.2y - 5%
Main bearing moment
Tilt DLC 1.3 to %
Yaw DLC 7.1 to 2.2y -22%
DLC 1.3, DLC 2.2y, DLC6.2 are cases with highest loads from yaw error  how about a free yawing turbine?
Tower: effect of alignment of thrust and gravity force
Blades: effect of coning downwind
DLC 1.3 extreme turbulence DLC 2.2y abnormal yaw error
DLC 4.2 shut down during gust
DLC 6.2 idling at abnormal yaw error
DLC 7.1 parked in 1y extreme wind

7. DLC 6.2 - Idling at large yaw errors after grid loss
No rotation:
Blade gets stuck behind tower
Yaw range -40° to -50°
Balance of blade forces
Idling:
Change of tower shadow model
Maximum pitch angle
Feasibility?
Tried to :
Take out imbalances mass and aerodynamic
Change the turbulence seed
Change the yaw error slightly
Feasability:
Aoa in ranges where we do not trust the airfoil data
Blade aerodnamically approximated as a line, while
tower shadow model with pure reduction in velocity

8. Fatigue load – Change in life time equivalent load
Tower bottom flange moment
Longitudinal -3%
Lateral %
Torsion %
Blade root moment
Flapwise +5%
Edgewise +7%
Torsion %
Main bearing moment
Tilt +7%
Yaw -2%
98% to 99% of life time equivalent load from DLC 1.2 – normal operation
Effect from tower shadow
Effect from alignment of inclination angle and tilt angle

9. Free yawing concept Would a free yawing concept give a cost advantage?
Highest loads from cases with yaw error
A yaw drive costs about 1 – 2% of capex
Four research questions:
Are all modes positively damped within the operational range?
Do the models agree on the damping?
What would give a stability boundary at lower wind speeds?
Does the system align itself with the wind direction?
Yaw error on DLC 1.3, DLC 2.2y and DLC 6.2
You never get rid of the full yaw drive due to the unwindidng mechanism
Maybe delete question 2 and just mention it, presentation might get too long.

10. Stability of a free yawing turbine
Free yawing model without bearing friction
2 DOF analytical model for parameter study
HAWCStab2 for comparison and extension of DOFs
HAWC2 for alignment study
wind direction
Motivation for 2DOF model:
Analytical expressions in the 2 by 2 matrices
”See” the parameters influencing the solution
Study the change in damping due to varyation of parameter

11. Yaw stability - analytical 2DOF modelLs
Lcg
Wind direction gc y x
q(t)
Side view
Top view
2DOFs: ux(t) and q(t)
BEM for Aerodynamic forces
Linearization around 0° yaw
Parameter study
gc /-10°
Ls +/- 50%
Lcg /-0.5
Damping from eigenanalysis
ux(t)
Assumption:
No tilt (-> (+stiff in tilting) no gravity influence)
No structural damping
No bearing friction
Aerodynamically straight blades  Quasi steady aerodynamics
 The steady state is at 0°
Limitations:
It does NOT predict the correct values of the damping for the full turbine
It DOES show the influence of parameter variation on damping
Straight blades
No tilt
No structural damping

12. Yaw frequency and damping - Comparison with HAWCStab2
Comparison between analytical 2DOF model and HAWCStab2
Reasonable agreement in frequency and eigenvalue
Stable yaw mode
Yaw frequency for v>44m/s
This slight could as well be deleted for time reasons and the Questions could be reduced to 3
Yes, the models agree on the damping and I can use it for the parameter study.
If f=0Hz than damping = 100% or -100%

13. Stability boundary for yaw mode - Parameter study
Cone angle
Stabilizing effect
Instability for negative cone angles
Length parameter
Minor influence from Lcg and Ls
Minimum damping 75%
No instability
Positive cone away from the tower
Negative cone towards the tower
Sharp change in damping because f=0Hz
A negative or too low cone makes the yaw mode instable

14. Stability boundary- HAWCStab2 extension of 2DOF model
Including blade geometry and rotor flexibility
Instable yaw mode v>23m/s
Including full turbine flexibility and structural damping
Instable yaw mode at v>18m/s
To calculate the stability boundary you NEED to include the rotor flexibility
And better:
Full turbine flexibility
 The flexibility of the rotor will give a stability boundary at low wind speeds

15. Aligning with the wind direction HAWC2
Released yaw in constant wind field after 300 seconds
Misalignment due to tilt
Minor change with full model
Controller reduces misalignment
Low yaw frequency v>17m/s
Power loss below rated wind speed
Stability vs alignment
Stable= positively damped motion, not necessary at 0°
Without controller = predescribed pitch and rotational speed
No load assesment done for these high yaw misalignments
But we would expect very high loads for the operation at large yaw errors

16. Summary- free yawing turbine
Stabilizing effect from cone
2DOF model is insufficient
Instability for high wind speeds
Misalignment with wind direction
Controller can reduce misalignment
Loss in power
Yaw mechanism for cable unwind
Future work with yaw controlled turbine in normal operation.

17. Future work and cooperation with IEA wind task 40
Cost out on existing rotor
New rotor design
Tower shadow alleviation
Control strategy
Cooperation proposal
Tower shadow/ noise modeling
Testing models in simulation environment
PhD-exchange
What are our next steps?
And how would we like to cooperate with the IEA task? – Let’s read that document about what they will do... 
And let’s discuss it BEFORE I go 
They might ask if we would test on a real turbine. And the answer is: maybe.

18. Thank you, for your attention.

19. References
[1] Hansen, M. H., Thomsen, K., Natarajan, A., & Barlas, A. (2015). Design Load Basis for onshore turbines -Revision 00 [2] BasicDTUController,