[Feature]: Indicate turbine dependencies for case to case testing

[misi9170]

Description

For large farms, we need a way of determining which (upstream) turbines a given turbine's power depends on.

Should fit naturally in floris_tools.py

[Bartdoekemeijer]

You could look at

https://github.com/NREL/floris/blob/39c466000b1874e06a6f58da9c30bb877fc8d4d2/floris/tools/optimization/yaw_optimization/yaw_optimization_tools.py#L21

and

https://github.com/NREL/floris/blob/6e856e54753f290de83eeeb74b7bc97c35fe22ad/floris/tools/optimization/yaw_optimization/yaw_optimization_tools.py#L20

These functions rely on a simple geometrical wake model to quickly determine which turbines fall inside the wake of another turbine.

[misi9170]

@Bartdoekemeijer thanks for pointing these out! I'll take a look before implementing anything. One 'feature' for this that Paul and I have been discussing is that we'd like to set a threshold for dependency (user defined, but default at 0.1% of rated power), so turbines far downstream (but still technically in the wake) of an upstream turbine are not considered dependent.

[paulf81]

Thanks @Bartdoekemeijer ! These are a good pointer, but do agree with @misi9170 this new method has some novelties. Also it's a bit inverted of these, it's making a list of upstream turbines

[Bartdoekemeijer]

I see! The methods above are really simplified and do not include any kind of quantitative wake loss. You could include a downstream distance threshold here for dependency, e.g., 30 rotor diameters. I can imagine having a threshold for wake loss dependency would require more complicated methods and could therefore be a lot slower but more accurate. Depending on your use case, I can see either work!

[paulf81]

I think that is a nice way to think of it, both are options, and I think the geometric approach will be much faster, but the wake model approach will allow criteria based on relative level of impact

[misi9170]

Closed by PR #70