Implementation of mounting system shading in 3D scene (VBPV)

[Luis Zimmermann]

The attached picture shows the rear side of my VBPV system. (the front side isn't shaded by the mounting)
To reflect the shading of the mounting system, I tried to add the mounting structures in the 3D shading scene. 
The limitation is that I have to keep a minimum distance of the 2-3cm from the module to the piles. 
The simulation result stays the same with and without the mounting structure. 
Which setting do I have to choose to reflect it within the bifacial model? (if I put the mounting structure on the front side, it influences the result)

 

[Linda Thoren]

Dear Luis Zimmermann,

The bifacial model in PVsyst does not consider any additional shading objects from the 3D scene for the backside irradiance, only the parameters in the Bifacial system definition is used in the simulation. Thus, to include the impact of the mounting structure, this should be estimated and put in the Structure shading factor in the General Simulation Parameters in the Bifacial system definition window.

 

Kind regards

[Luis Zimmermann]

The issue is the the rear shading factor is applied at all hours of the day and to all irradiance contributions (direct beam, diffuse, reflected). I tried to play around with this factor, but it seems not to work. 
Is there any possibility to simulate the rear side for a vertical system like the front side?
In my opinion the differences between front and rear side are only:

- Bifaciality factor for rear side

- no AR coating on the module glass

- Mounting structure  shading (can be added in 3D shading scene)

[Linda Thoren]

Indeed this is a limitation of the bifacial model today.

A possible workaround, is to modify the .PAN file and place two panels back to back. By creating separate modules for the front and rear—each with a bifaciality factor of 0 (and adjusted properties, e.g., lower power, no AR coating for the rear)—you could place them back-to-back in the 3D scene and simulate the system using the Module layout. This wouldn’t fully replicate real system behavior, and on clear days electrical shading losses might be underestimated due to the fact of having two panels instead of one. This in not necessary a recommendation, but could work as a test to evaluate the shadings for the backside.

Otherwise, for the estimation of the electrical shading losses in general for a vertical EW system, the irradiance distribution on the rear side should be roughly as uniform as on the front. This means that, in principle, the rear mismatch factor should be set to zero, as it only accounts for additional non-uniformity on the rear. However, since PVsyst currently calculates electrical shading losses only for the front side, using the rear mismatch factor as an approximation for rear-side shading losses is reasonable. That said, determining whether 5% is an accurate value is difficult, as it depends heavily on system geometry (row height, spacing, etc.). As a first approximation, it would be best to adjust the rear mismatch factor so that its corresponding losses in the loss diagram match the electrical shading losses multiplied by the bifaciality factor.

[AutoRob]

Hello,

in my understanding, I will get the most reasonable rear side irradiance when I simulate it as front, so for example with unlimited sheds changing 90° to -90° azimuth, while changing the IAM to normal glass. The simulated effective irradiance * bifaciality factor should be the correct effective rear side irradiance for the original simulation.

Is there any way to include this simulated rear side irradiance in the original simulation? Shed transparent fraction and albedo influence the front too and the rear side shading and mismatch factors are only making the irradiance worse, while (at least in my case) the rear side irradiance i want to match is slightly better than the simulated one... Maybe it is possible via changing pitch or number of sheds, but i am not sure if this is reliable and if this will change something else.

Also, I have another question:

The thermal losses are dependent on the effective front side irradiance, this leads to unrealistic thermal gains in vertical simulations. For example, a south oriented plant has 1.1% loss, a vertical system at the same location has 1.1% gain. I dont think there is a way to fix that other than brute forcing the U value. Is that a bug, and if so, when is it planned to be fixed?

Best regards

Robin

 

[Linda Thoren]

Hi,

No it is today not possible to include that in the original simulation. If you choose to simulate both sides as front sides, that can indeed be simulated in the same variant (with the proposed work around of modifying the .PAN files)

You are right, the thermal losses only consider front side irradiance, we will look into this.

Kind regards