Describe the bug
A user reported finding negative POA shading loss values when using the non-linear self-shading model for vertical arrays.

Forum post: https://sam.nrel.gov/forum/forum-general/4443-negative-pao-front-side-shading-losses.html#12722

Email context: I’d have to dig a bit further but at first glance it appears this diffuse_reduce function has a cosd(180 – tilt) that would return a 0 when tilt = 90, leading to a divide by zero issue for the factors F2 and F3. This function seems to be applied for both the non-linear (https://github.com/NREL/ssc/blob/develop/shared/lib_pvshade.cpp#L567) and linear (https://github.com/NREL/ssc/blob/develop/shared/lib_pvshade.cpp#L511) options, although Janine can correct me if I am wrong. We might need to add in a limit of 89.9 for that equation or something similar.

https://github.com/NREL/ssc/blob/develop/shared/lib_pvshade.cpp#L213

Expected behavior
There should be a warning or an equation limit to prevent this negative shading loss from happening

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version [e.g. 22]

Additional context
Add any other context about the problem here.