Example uses for infinite sheds

[mikofski]

I can't wait until this module is available, and was thinking how can we showcases it's capabilities with lots of gallery examples?

1. infinite sheds can give hourly, daily, monthly, and annual beam shading from row-to-row (r2r), although this doesn't translate into electrical losses, if we assume linear, 70% electrical effect, or a constant shade impact factor we could summarize the effects of r2r on output - I think this would be a first for pvlib, right?
2. obviously BIFACIAL! duh! (head slap)
3. comparison of existing diffuse masking module vs. pvfactors vs. infinte sheds
   • total diffuse shade for bifacial (both sides)
   • total diffuse shade for monofacial (front side only)
4. develop a loss tree to compare with PVsyst and other software?

[cwhanse]

An example showing how to compute annual "bifacial gain", e.g., bifacial single-axis tracker energy / monofacial SAT energy, would be a good start.

[williamhobbs]

Has there been any more discussion on infinite_sheds gallery examples? I'm interested in one that uses the new diffuse fraction and shaded fraction outputs and @kandersolar's simple shade model for row-to-row self-shade losses.

Here's a gist of (most of) what I had in mind, https://gist.github.com/williamhobbs/9d1fb23a64f1921400938250af6a67f0, resulting in one or more plots like this one:

It could also be interesting to illustrate:

1. a linear model for thin film fixed and tracking (I already have those accounted for in the gist above, but don't actually run through it)
2. c-Si tracking w/ backtracking, using infinite_sheds to get diffuse shade losses
3. c-Si tracking w/ backtracking and tracker clock drift to show power loss impacts of misaligned trackers

1 and 2 could be easy to add in the same example if more of the calculations were wrapped in functions, so they didn't have to be repeated 3-4 times. But then it might be harder to explain intermediate steps cleanly in the gallery?

[cwhanse]

I think what's in the gist tends to a more complex example, not in the code but in the question it addresses: how much power/energy is given up due to row-to-row shading? That requires two calculations +1 more for each shding model. Once the analysis is explained, and the calculations are presented, it will be a long-ish example. My reaction is to put the tracking/backtracking comparisons into other examples.

[williamhobbs]

Thanks for the feedback, @cwhanse. I realized it's easy to add a linear shade (think film) case to the gist example (same gist link). What do you think about a cleaned-up and explained version of this as a standalone gallery example?

[cwhanse]

The harder part is writing the explanations :)

[williamhobbs]

Feedback on the updated gist would be great, https://gist.github.com/williamhobbs/9d1fb23a64f1921400938250af6a67f0. I tried copying Mark's illustration style from the tracker_shaded_fraction docs when explaining how I got creative with GCR for the top strings of a 2P system.

[williamhobbs]

@mikofski, how did you make this pvlib.shading.tracker_shaded_fraction illustration?

(https://github.com/mikofski/pvlib-python/blob/linear_shade_gh1690/docs/sphinx/source/_images/FSLR_irrad_shade_loss_slope_terrain.png)

It looks like it could have been PowerPoint (my drawing tool of choice, don't tell anyone 😬), but I figured I'd check in case there is some great PV geometry sketch tool I should know about...

[mikofski]

I admit it was pptx, but Kevin turned me on to asymptote, it’s a pretty nifty tool for sketches

[williamhobbs]

Thanks, I’ll have to try out asymptote. And nice to know I’m not the only one using PowerPoint for technical drawings.