Issues with shading models and vertical bifacial module assumptions

[mjprilliman]

Describe the bug
There are a few issues when applying self-shading corrections to bifacial systems with high tilt angles, such as vertical (tilt=90deg) systems. These are important to model correctly for agrivoltaic applications. The issues have been investigated in some detail, but not completely resolved. Areas of investigation include:

1. Bifacial rear-side gain is always applying a linear correction for row-to-row shading of the back side of the panel, regardless of whether the front-side self-shading is being calculated using linear or non-linear approach.
2. The non-linear self-shading model in SAM, originally intended for monofacial Si modules, applies corrections based on beam shading to DC power, not beam irradiance. So this correction is being applied to total DC power from front-side and rear-side POA irradiance, not just front-side.
3. There is asymmetry in results for vertical bifaciality with assumed perfect bifaciality. Further investigation is needed to identify if this is due to morning vs afternoon differences, or differences in modeling approaches for front vs rear side. Assumptions of infinited sheds models may be a factor. One suggested workaround from @shirubana is to flip the axis convention after solar noon, where the 'rear' surface becomes the front after solar noon for the calculations. This would require re-working spatial albedo definitions to keep different surfaces in the correct location relative to the front and rear surface definitions.

To Reproduce
Steps to reproduce the behavior:

1. Go to Detailed PV - No Financial
2. Check 'Module is bifacial', set 'Bifaciality' to 1.0
3. Set Tracking to 'Fixed', Tilt to 90, Azimuth to 90 (front-side facing East) on System Design page
4. Set 'Self Shading' to 'Standard (non-linear)' on Shading and Layout page
5. Simulate
6. Duplicate case (drop down on case name)
7. On duplicate case, set Azimuth to 270 (front-side facing West) on System Design page
8. Simulate duplicate cases
9. See differences in results

Desktop (please complete the following information):

• OS: [e.g. iOS]
• Browser [e.g. chrome, safari]
• Version 2025.4.16.r1

Additional context
-Investigated as part of efforts for InSpire projects
-Also motivated by Erin Tonita paper 'Vertical Bifacial Photovoltaic System Model Validation: Study With Field Data, Various Orientations, and Latitudes' https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10985871&tag=1

[shirubana]

Want to point out the work by Erin 10.1109/JPHOTOV.2025.3561395
We followed our method of 'switching' the front and back due to the different ways the front and rear irradiance are calculated
S.