Hello!

While designing a PV system over a 3D OBJ (captured in drone deploy and exported on ReCap) the result of the simulation was 15000kwh/year. It is a system surrounded by 2 buildings, with lots of shading.

After the system was installed, the efficiency was aroung 30% of the simulated values. Reviewing the 3D simulation, I noticed that the shading simulation was incorrect, with values around 3% where it should be higher.

So, I designed walls to simulate the building walls, and turns out that the simulation result became very close to the actual generation. Se attached pictures with and without the walls.

It seems like PvSol cogeneratedBy_react-csv (2).csvnsidered the actual 3D model as "transparent" or something like that. What is the reason and correction for this behaviour?

Tested again with the same OBJ but exported with Blender and the shading was correct (same as the model with the walls).

 

Now, why can`t pvsol measure the shading of the recap model?

Hello DMA,

thank you for your request and testing the shading frequency.

Unfortunately there is a bug in the calculation of the shading frequency values.
Even small movements of the mounting surfaces result in big deviations of the values.

Please have a look into the following thread:

Kind regards,
Frederik

Hello Frederik

Thank you for your support in advance.

In the linked thread you state that this only affects the shading frequency and does not affect the yield. In my scenario it does affect the generation as well.

Hi DMA

Oh, I got that wrong, my bad.

Could you send the two project files via private message here in the forum so I can have a look into it?

Thank you in advance and kind regards
Frederik

Hello dear Frederik

Have you analyzed this issue? As precaution, I am now using a solid block anytime I have a significant shadow, and the result is always a higher shading loss.