Dynamic Wind Direction Variation in FAST.Farm Using TurbSim or User-Defined Time Series

Wind & Water Wind Farm Modeling
1

Dear NREL Team,

I hope you are doing well.

I would like to clarify whether it is possible to introduce a dynamically varying wind direction in FAST.Farm using TurbSim or the InflowWind module with a user-defined time series. I came across this 2019 forum thread, which states that TurbSim does not support time-varying wind direction within a single simulation run. However, I wanted to check if there have been any updates or alternative approaches available since then.

Specifically, I would appreciate guidance on:

  1. TurbSim Wind Files: If a dynamically varying wind direction is still unsupported, what are the best practices for approximating this effect? Would generating multiple TurbSim wind fields with different directions and stitching them together be a viable approach?
  2. InflowWind Module: Can the user-defined time series effectively model continuously varying wind direction in FAST.Farm? If so, what format and constraints should be considered when preparing the input file?
  3. Challenges and Limitations: What are the main computational or modeling challenges when implementing time-varying wind direction? Are there potential numerical stability or wake modeling issues in FAST.Farm when using such inputs?
  4. Alternative Approaches: Are there other recommended methodologies or tools (e.g., coupling with external atmospheric models) for incorporating dynamic wind direction into FAST.Farm simulations?

Your insights would be greatly appreciated. Thank you for your time and support.

Best regards,
Abdulbaset

2

Dear @Abdulbaset.Alazhare,

No, TurbSim and InflowWind have not been upgraded to support time-varying wind direction since the conversation you reference. Regarding 1-3, the biggest issue, as referenced in the forum post you cite, is in the Taylor’s frozen turbulence assumption used by InflowWind, which propagates the 2D wind field from TurbSim along a fixed direction to create wind in 3D (effectively relating, the x-direction and time by the mean (propagation) wind speed).

Regarding 4, the alternative is not use TurbSim and InflowWind for inflow to FAST.Farm. Rather, FAST.Farm supports the use of ambient inflow from a high-fidelity precursor (e.g., Large Eddy Simulation), without the use of frozen turbulence assumptions. Effectively, this wind inflow is 4D (varying in 3D space and time), rather than 3D (varying in 2D space and time) as assumed by the TurbSim/InflowWind approach.

Best regards,