Simulation with vertical panels (tilt 90º)

[raul.martin]

Hi, I’m working in a project with vertical panels (tilt = 90º) and an orientation E-W (-90º). I realized some problems when working with that configuration in PVsyst:

- The performance ratio is over 100 % : I find performance ratios around 120-145 % which should be impossible as the theorical maximum performance ratio of a panel is 100 %.

- I run a test with a near shading scene where I have only one tree and only one table of panels. If I put my table at the west of the tree, I have near shading losses of around 30 %, but if I put it at the same distance but at the east I have only 2 % of irradiance losses, which doesn’t make any sense. I think that the losses should be quite similar in both cases. My system is a bifacial one, so the panel is supposed to produce electricity from both sides, and I think that PVsyst is only considering the irradiance arriving on one side.

Thank you for your answer.

Raúl.

[Michele Oliosi]

Hi,

The bifacial model only considers mutual shadings due to the geometric arrangement of the PV tables, and not the shadings extracted from the 3D scene. The 3D scene only affects the front side. This is due to the analytic nature of the bifacial model, that works under the assumption of regular rows of tables or trackers.

The performance ratio used by PVsyst is the standard one, i.e. normalized to the frontside POA irradiance, please see https://www.pvsyst.com/help/performance_ratio.htm for its definition. This is why it can be above 100% for bifacial systems.

[dtarin]

Correct me if I am wrong, but you can manually calculate the correct PR by including the rear-side irradiance. For a vertical system, this is necessary since the rear-irradiance contribution is significant, and you will likely have irradiance instrumentation on both sides of the module in a constructed facility.

[Michele Oliosi]

@dtarin yes indeed, the norm IEC 61724-1:2021 https://webstore.iec.ch/publication/65561 explains how to compute a "bifacial" PR.

This is however not included in the results of PVsyst yet, so at the moment you have to do the calculation by yourself.