Dear @Jason.Jonkman,
Hello! I hope you’re doing well.
I encountered an issue while running a FAST.Farm simulation to evaluate turbine fatigue due to wake effects at a relatively low wind speed with high turbulence intensity (V: 2 m/s, T.I: 37.1). Initially, I received the following error:
T2:FARM_UpdateStates:WD_UpdateStates:ThomasAlgorithm:Tridiagonal matrix is not diagonally
dominant, i.e., abs(b(i)) <= ( abs(a(i))+abs(c(i)) ). Try reducing the FAST.Farm timestep.
In response, I reduced the value of DT_Low in the .fstf file, which resulted in another error:
T1:FARM_UpdateStates:NearWakeCorrection:Wake model is not valid in the propeller-brake region, i.e., Ct > 2.0.
Upon reviewing the FAST.Farm results, it appears that this error occurs when the turbulence intensity is high, causing the wind speed to drop close to 0.3, resulting in a Ct value exceeding 2. According to the FAST.Farm manual, Ct values are notably high at low wind speeds, but I am struggling to find a suitable solution to this issue.
I would greatly appreciate any advice you can provide.
Thank you.
Dear @Young-Ju.Noh,
I would not expect a wind turbine to operate at such a low wind speed with such a high thrust coefficient. I think it is best to avoid this simulation set-up by not simulating with such low wind speeds or changing the wind turbine controller so that the turbine shuts down when the wind speed gets so low.
Best regards,
Dear @Jason.Jonkman,
Thank you for your prompt response and valuable advice.
To address this issue, I will explore either avoiding simulations with such low wind speeds or modifying the wind turbine controller settings to ensure the turbine shuts down when wind speeds drop to such low levels. Your guidance has been very helpful. Thank you once again for your assistance.
Best regards,
Hi @Jason.Jonkman ,
I ran into the same issue for a few cases like [vhub = 7 m/sec, TI = 14%] and [vhub = 8m/sec and TI = 17%]. Lowering DT_Low to 2 m/sec didn’t help. Is this expected?
Simulation were run on FAST.Farm 3.4.1 with NREL 2.8 kW turbine.
Abhineet
Dear @Abhineet.Gupta,
I’m not sure what you mean when you refer to an “NREL 2.8 kW turbine”. Are you using one of the OpenFAST models from the r-test?
How low is the ambient wind speed dipping in these simulations and are the corresponding rotor speed (tip speed ratio) and blade-pitch angles expected for this situation?
Best regards,
Hi @Jason.Jonkman
Thanks for the reply.
I am using this turbine “openfast-turbine-models/IEA-scaled/legacy_models/NREL-2.8-127 at main · NREL/openfast-turbine-models · GitHub”.
The wind speed dips to around 2.5 m/sec but recovers, However, the TSR drops to around 5.6 just before the error. The rotor speed drops to around 5 RPM.
The following plot shows these values for all 12 turbines in the farm
Here’s a a zoomed in plot focusing on data right before the crash.
Regards
Dear @Abhineet.Gupta,
I’m not really familiar with this OpenFAST model (which is a 2.8-MW wind turbine instead of a 2.8-kW turbine), but I would guess the issue is associated with the wind turbines with high tip-speed ratio (TSR, such as the purple or orange), which would correspond to a high thrust coefficient. It doesn’t look like the TSR dwells at a very high TSR for long, so perhaps adjusting (reducing) the corner frequency (f_c) of the low-pass time filter would help FAST.Farm make it through the periods of high tip-speed ratio. Otherwise, it may be necessary to modify the controller so that the TSR does not get so high.
Best regards,
