Unclear controller behavior – malfunction or model setup issue?

Hi everyone,

I’m currently simulating DLC 1.2 (IEC 61400-1) for a custom wind turbine as part of my master’s thesis, focusing on rotor blade modeling. The wind speed range I’m using spans from 3 m/s (cut-in) to 25 m/s (cut-out) (turbulent wind). However, I’ve observed some unexpected controller behavior in the results.

In several simulations, the controller appears to shut down unexpectedly or fails to actuate pitch control at all. I’m unsure whether this points to an issue with the controller itself, or if the root cause lies within my model setup (e.g., numerical instability or improper initialization).

Unfortunately, I have very limited information about the controller — I only know the rated generator torque and speed (79293 Nm and 739,82 rpm), and that the .dll was compiled from a Simulink model. To further investigate, I ran a simulation with a uniform wind speed ramp from 3 m/s to 30 m/s. These are the results I got:





For initialization, I used a rotor speed of 3.3 rpm and a pitch angle of 2° for all three blades.

Does this look like a controller issue to you, or could this just be a problem with how my model is set up?

Any help or suggestions would be greatly appreciated!

Best regards,
Linus

Dear @Linus.Wermert,

Can you clarify the source of your controller; is this from NREL?

A controller-driven shutdown is not itself indicative of a problem, assuming the conditions warrant a shut down.

It looks like the controller is having problems around 400-s into the simulation, where the response seems to toggle between below and above rated control operation. And it seems that the controller is triggering a turbine shutdown just before 500-s into the simulation.

Best regards,

Dear @Jason.Jonkman,

Thank you for your response.

Unfortunately, I don’t have any further information about the controller. I believe it’s not from NREL — it was provided by a partner institute that uses it for Simpack simulations of the same wind turbine. Do you know if it’s maybe possible that there are compatibility issues with OpenFAST?

Over the weekend, I set up a ROSCO controller, and it appears to be working correctly in the same simulation environment:


This leads me to believe that the issue likely lies with the original controller DLL rather than my model setup.

Best regards,

1 Like