Simulation With 3D Shading Not Running For Full Year

[ndiorio]

Hello,

 

I am trying to run a simulation for a system where I've set up a 3D shading scene.  When I click to generate results, two simulations occur.  The first is the calculation of shading losses.  This appears to run for the full 365 days.  The next simulation says "Simulating", but only runs until day 239 before stopping.  Then, the "Hourly Values" csv file only has data until August 28.

 

Is there a setting somewhere that would govern this behavior?  

 

Thanks 

[developer_mh]

Hi ndiorio,

 

Does your system contain battery storage or are you simulating an offgrid system? That could be a reason as the simulation stops automatically if the batteries are below a certain state of charge too long. But then you should see corresponding messages after the simulation.

 

Other then that, there are actually no reasons for the simulation to stop before day 365.

 

So, if you use batteries in your system, check for the charging parameters, especially SOC min and the three preservation mode levels (for offgrid). If you are not using batteries, it is hard to say what causes this behaviour, which is why we would be glad if you could send us the project file (post it here, or send it via pm if it contains confidential data)

 

Kind regards,

Martin

[ndiorio]

Hi Martin,

Here is an example file where the simulation only appears to run for about 54 days.  

 

Thanks,

Nick

RunningShort.pvprj

[developer_mh]

Hi Nick,

 

thanks a lot for the project file. The reason for the simulation to stop is a bug that is already fixed and will be released in the next bug fix release, 2016 R5.

 

The bug (actually a rounding error) was in the calcution of the two-diode model which we use to model the electrical behaviour of the pv modules. It happened only at extreme low light conditions, for diffuse irradiance values below 1 W/m², and only if the global irradiance was above 1 W/m². Thus a really, really, rare case that you encountered there Sorry for the inconvenience caused, and thanks for bringing this up to us.

 

So you'd have to wait for the R5 version of PV*SOL premium 2016 to be able to model your pv module with the two diode model.

 

In the meanwhile you could replace the two-diode model with another low light model (we call it PV*SOL model since it was invented here at Valentin). I prepared a project for you where I did just that:

RunningShort PV Module Copy.pvprj

 

Funny pv module that you have there, by the way...

 

Hope that helps and sorry again for this bug,

Martin

[ndiorio]

Hi Martin,

 

Great, thanks for the information, I'm constantly debugging programs myself so I understand  

 

Thanks for posting the file, I'm still learning about the options available in PV*SOL so that will be helpful.

 

With the PV module I was trying to evaluate primarily the shading behavior alone, and sized it to some unrealistic dimensions.  I agree it is not a PV module which one would ever see in reality...

 

Thanks again for the detailed explanation, very helpful!

 

Nick