Comparisons between NREL 5MW and CART

Dear all,

I am working on controls to the NREL 5MW reference turbine. I am currently implementing some of the pitch controls of the document “Modern Control Design for Flexible WT” , by Alan Wright (nrel.gov/docs/fy04osti/35816.pdf).
I have found some important differences between the dynamic behavior of NREL 5MW and CART-2 turbine (the CART is the turbine studied in the above cited document). Could you help me to understand them?
For the CART, the first drive-train torsion mode is unstable. So, if we design the full-state feedback control without considering this degree of freedom as a state, the controlled rotor/generator speed becomes unstable (see the figure 1 below).


Simulating NREL 5MW in similar wind conditions and with the same full-state feedback control, first drive-torsion mode doesn’t induce instability (see the attached figure below of rotor speed). The simulation was made using MATLAB/Simulink and FAST.

Even without instability, I’ve tried to apply a control model including first drive-train torsion mode (enhancing this mode’s damping) and the result was surprisingly worst. The regulated speed presents more oscillations than before (see the figure 3).

In summary, why the first drive-train mode causes no instability for the NREL 5MW turbine? And why when I try to enhance damping in this mode, the regulated speed becomes more oscillatory?
Best regards,

Eduardo Menezes
BSc Mechanical Engineering
Master’s Student in Wind Energy
Center for Renewable Energy - Recife, Brazil



Dear Eduardo,

I’m not sure I can really answer your question, but here are a few comments:

  • Instabilities are typically characterized by an exponential increase in response. Your plot of generator speed for the CART-2 turbine shows a linear increase in response, which is more suggestive of a resonance than an instability.
  • Instabilities are physically initiated by negative damping in the system whereas resonances are physically initiated by a loading that excites the system at a frequency close to a natural frequency.
  • I suggest that you study the damping and frequency of the drivetrain-torsion mode to determine what is causing the response you are seeing and the differences between the NREL 5-MW and CART-2 turbine models.

Best regards,

Dear Jason,

Thanks for your response. I think that for doing a comparison between CART-2 and NREL 5-MW I would need the values of damping, stiffness and mass for the CART-2.
Do you luckily know where I can get these values?
It sounds strange for me, because NREL 5-MW is a large-sized turbine and so it certainly should have more problems with resonance and instability (this doesn’t seem to be the case, however).

Best regards,

Eduardo Menezes
BSc Mechanical Engineering
Master’s Student in Wind Energy
Center for Renewable Energy - Recife, Brazil

Dear Eduardo,

The linearization function of FAST can be used to extract the mass, stiffness, damping, natural frequencies, etc. However, while linearization function is available in FAST v7, we are still working to add it to FAST v8, as discussed here: http://forums.nrel.gov/t/natural-frequencies-with-fast-v8/1211/1.

Best regards,

Dear Jason,

The linearization function would really solve the problem, but unfortunately I have no access to CART-2 mode shapes coefficients which were used in http://www.nrel.gov/docs/fy04osti/35816.pdf.
The document contains blades’ and tower’s structural properties, but no information about mode shapes.

Best regards,

Eduardo Menezes
BSc Mechanical Engineering
Master’s Student in Wind Energy
Center for Renewable Energy - Recife, Brazil


Dear Eduardo,

I can’t help with that. The FAST models of the CART wind turbines at NREL are not in the public domain, but are only available to specific collaborators. Based on your earlier posts, I assumed you had access to the complete CART2 FAST model.

Best regards,

Dear Jason,

Ok, no problem. Thanks for trying to help.
I posted the figure with CART-2 generator speed because it is part of the cited NREL Report “Modern Control Design for Flexible WT” (http://www.nrel.gov/docs/fy04osti/35816.pdf).

Best regards,

Eduardo Menezes
BSc Mechanical Engineering
Master’s Student in Wind Energy
Center for Renewable Energy - Recife, Brazil