Diffuse from meteonorm vs. shading

[mhschneider]

Hi,

usually when importing a meteofile containing diffuse insolation, PV Syst will use it for the sim and note it on the report as "imported".

When importing a meteonorm *.dat file, the report says under diffuse "Erbs, Meteonorm". Does this mean PV Syst is then not using the diffuse part for the simulation

but is calculating the diffuse with Erbs?

The other possible explanation would be that Pv Syst knows that Meteonorm uses Erbs to calculate the diffuse part, where I couldn't find any hint for in Meteonorm documentation.

This does have a huge impact on shading losses here in Europe.

[André Mermoud]

When the diffuse is available on an hourly file, PVsyst will always use it. This is the case for your Meteonorm external file.

If a dataset is imported from Meteonorm (internal to PVsyst), the diffuse is also calculated by Meteonorm.

Now there is a mistake in the denomination in the PVsyst report: the model for diffuse used by Meteonorm is indeed the Perez model, not the Erbs model.

I just got the confirmation from Jan Remund last week. We will correct that in the next version.

However I don't see well how the fact to know which model is used could have so "huge" effects on the shading calculations. The effect of diffuse is more important on the transposition.

[Jen H]

Hi

We have seen changes in the diffuse model in PVsyst projects in V6.42 which we believe is due to the change mentioned in this post.

This raises a few questions for us:

 

• Will we be able to select between Perez and Erbs in future updates?
• What is the difference between the two e.g. rough % difference seen
  
• As far as I am aware we can’t actually test this as we also have several other changes e.g. update in PVsyst shading engine, which makes a like-for-like test impossible, is this correct?