Wake Induced Mixing Gain and turbulence intensity for floating farms

[fbelli997]

Hello everyone,
I was approaching the calculation of the AEP with the wake-induced mixing gain to set the increase in wake amplitude, and I noticed this warning message:
"floris.logging_manager.LoggingManager WARNING Running wake_induced_mixing model with mixing contributions from the atmospheric turbulence intensity has not been vetted. To avoid this warning, set atmospheric_ti_gain=0. in the FLORIS input yaml."

So, is it currently not possible to perform tuning on the wake-induced mixing in the case of floating farms?
Additionally, is there some kind of relationship between the wake-induced mixing gain parameters and the classical turbulence intensity input parameter, or do I need to calibrate the parameter to match the wake deficit as a function of the x-coordinate?

Thanks in advance, everyone!

[misi9170]

Hi @fbelli997 ,

Thanks for you question! If I'm understanding correctly, you're running into this warning because you've set the atmospheric_ti_gain term in the wake_induced_mixing model parameters to something other than zero.

The form of the wake induced mixing (WIM) model is given here in the documentation (you've probably already seen this, but I just wanted to be sure). As you've probably seen, we don't have current documentation for how the atmospheric_ti_gain affects this (and the documented model assumes that atmospheric_ti_gain = 0.0). Setting a nonzero value for atmopsheric_ti_gain does however affect the WIM---see these lines of the Empirical Gaussian (EmG) model solver procedure. However, we have in no way tested that dependency and moreover, we may change the functional form of the dependency in future; finally, the other parameters of the EmG model were tuned with atmospheric_ti_gain set to zero. This is why we raise a warning if users change the value away from zero.

We do have some long-term plans to develop the dependency of the WIM model to TI (see #841). However, there isn't a current timeline for that work, and it is unlikely to happen in the next 6 months or so (and quite possibly not for longer than that).

That being said, you are welcome to try tuning the atmospheric_ti_gain---we'd certainly be interested in what you find! Just be aware that you may need to retune other parameters of the EmG model (likely the wake_expansion_rates and mixing_gain_velocity when you do this). If you are doing this and would like to silence the warning being raised, you can set logging > console > enable to False in the main input yaml file.

Best,
Misha