Measured Front POA Equivalence for Bifacial System

[Mschatz]

Hello!  For purposes of a performance test of a bifacial tracking system, we are trying to have an "apples-to-apples" comparison of the front-side POA irradiance measured by pyranometers installed on the tracker (along same plane as the modules), to the "simulated" front-side POA irradiance value from PVsyst.  I understand that GlobInc would be representative of the irradiance received in the collector plane before optical losses are applied (which later results in GlobEff).  However, since this is a bifacial model, there is also the "ReflFrt" factor which is the ground reflection on the front side of the collector plane.  Referencing the PVsyst help documentation and waterfall loss diagram, it appears as if "ReflFrt" is not considered in the transposition model and is therefore a separate contribution of irradiation to the front-side of the collector plane separate from GlobInc, which ultimately contributes to GlobEff.

 

Thus, I am curious if the representation of "simulated" front-side POA based on outputs from PVsyst would more appropriately be GlobInc+ReflFrt when comparing to an equivalent field-installed POA pyranometer?  

[Michele Oliosi]

Just keeping GlobInc is more accurate than GlobInc + ReflFrt. This is because of the albedo contribution present in GlobInc. Therefore you could technically try GlobInc - AlbInc + ReflFrt. But ReflFrt includes the IAM, so GlobInc may still well be your best shot.

Indeed both ReflFrt and AlbInc (included in GlobInc) are albedo contributions. There are slight differences in their calculation (e.g. the IAM factor which is included in ReflFrt), and the albedo values may differ.
Still, in the flow of the simulation, AlbInc is eventually shaded (shadings below the horizontal are extrapolated from the 2°-sun height shadings) before reaching GlobEff. ReflFrt will then usually compensate this shading loss. Some differences may arise but the values can be considered close.