PVGIS optimization vs PVSyst optimization

[Emi]

Goodmorning everyone!

I just tried to use PVGIS "optimize slope and azimuth" and then use PVSyst "Optimization tools", running a simulation to find the best tilt and azimuth using Meteonorm (PVGIS import doesn't work at now).

The results of two tools are significantly different.

The site is in Geneva.

Any idea?

Maybe is it due to different database ? 

Many thanks

Images following:

 

[dtarin]

Batch simulating with different tilt and azimuth will provide more accurate results. 

[Emi]

Thanks for your suggestion,

but I tried for several sites and optmization tools always returns -5° best azimuth...

Why? Anything wrong?

Please, take in account horizon is flat...

With this behaviour optimization tool is useless

Edited January 9 by Emi

[Bruno Wittmer]

There are several possible reasons why a non-zero azimuth could be the optimum. In general, an optimum at -5° means that on average there is more direct beam irradiation in the morning than in the afternoon.

As you mention already, a significant horizon effect can be at the origin of this. Another possibility are systematic differences in cloud cover between morning and afternoon.

In the optimization tool you can choose to display the variables 'GlobInc' and 'GlobEff', to get some hints on the origin of the effect. The GlobInc variable describes only the transposition and will therefore display effects that are already present in the Meteo data (like different direct/diffuse ratios in the morning and afternoon). The GlobEff variable contains shading effects and will therefore add effects coming from the horizon line.

 

[Emi]

Many thanks Bruno, your info are very useful.

I gonna try some other test.

Thanks again