Shadings: Electrical loss v.6.40

[j.meier]

Hi all,

thanks for the new v6.40 and all the improvements I have seen so far, but:

I have noticed that since I have updated to 6.40 the shadings: electrical loss acc to stings are twice as high as for PVsyst 6.39.

I used the same simulation variant, same system same losses, just recalculated the shading table and noticed that these losses increased from around 1% to around 2%.

Was the loss calculation underestimated in the older versions or is there another special reason for this?

Can you please explain this change?

Looking forward for your reply.

[Bruno Wittmer]

The electrical loss according to strings depends strongly on the specific details of the simulation. Could you send your project (or a minimal example where you observe the behavior) to support@pvsyst.com, so we can have a look at it?

We did not introduce any fundamental change in this calculation, so we do not expect major differences in the results of V6.39 and 6.40.

[deblynn]

I just observed the same change as well from 6.39 to 6.40.

[j.meier]

I did send an example project.

Do you have any update regarding this topic?

I noticed that the difference still occurs in PVsyst 6.41.

[Jen H]

I am having the same problem too. When I export a project from 6.39 into version 6.41 and recompute the shading table, sometimes there is a small difference in linear shading but bigger difference in electrical shading, which can translate to a significant difference in electrical shading loss between the two versions.

I've attached a project to this post as a reference. The differences in electrical shading factor (shading factor according to strings) between version 6.39 and 6.41 is below:

 

I did not change anything in the simulation other than recomputing the shading table.

P.S. In 6.41 the shading table is calculated much more quickly than for previous versions, so thank you!

Bedford_Project.zip

[Bruno Wittmer]

When doing the calculation according to strings, there is indeed a bug in V6.40 and V6.41, causing some of the shaded areas to be counted twice, resulting in too large electrical shading factors.

The fix is being worked on.

[Bruno Wittmer]

The bug has been fixed for PVsyst V6.42, to be released in the coming days.

[Jen H]

Many thanks for looking into this. Looking forward to the update!

[masheets]

Was this bug also in V6.39? I noticed a sharp increase in production estimates when I upgraded from V6.39 to V6.43 when using detailed electrical calculations due to module layout with optimizers. I ran an old variant in the new version to compare, and saw "Near Shadings: irradiance loss" decrease from -17% to -9.4%, and "Shadings: Electrical Loss detailed module calc" decrease from -6.7% to -2% (this project is in a very shady area). The resulting production estimate difference is %15, and I would like to make sure the new result is accurate before I finalize my recommendation.

Thanks.