Ground object shadow casting - high loss problem

[Luca Vignoni]

Hello,

When doing a 3D scene with a ground object, what it is more accurate to do regarding the shadow casting option? Should I always left the ground object enabled?

In some simulation if I disable the ground object I get around 4% loss (summing shading irradiance loss and electr. shading), whereas if I leave the ground enable I get a loss of 10%, which seems very high.

Can you support on what would be more correct to do with the ground?

[Michele Oliosi]

It is more accurate with the ground. If the ground blocks the sunlight in some way for your PV field, that should be included in the shading objects.

[Ander]

Hi Luca,

In case you are including ground for the shadows, you have to check the layout carefully. If you are using the natural ground with the structures placen on top of it, it can happen that in some cases modules are under ground. Also in trackers, it can happen that they look fine, but when the tracker is tilted, modules get partialy underground.

You should create a new surface that looks similar to how it will be after all the earth movements.

 

Best regards,

Ander.

[dtarin]

And to extend what Ander provided, what I have done in the past is if there are ground obstructions/collisions with the tables or trackers, lower the ground image in the Z coordinate just enough so that it clears everything once the tables have been set to the terrain-specific elevations. It will then still capture most of the impact of the terrain. You can run the check for interpenetrations tool to see if it has been resolved. 

It is often the case though that there may be irregularities in the terrain data, for example if you are looking at existing terrain conditions for a tracker system, the tracker may extend over odd and irregular areas of ground terrain since they are so long. If your site is small the number of instances are few, you can edit the ground terrain directly. If you have a large site and many instances, lowering the entire ground image may be most efficient.