Inverter loss due to Voltage Threshold

[kamomeno68]

Quick question about Voltage Threshold. I assume this item reflects inverter losses incurred when the input voltage to the inverter drops below its minimum dc operating voltage. In a real-world installation, I can see how this scenario would arise. Object shading could knock out a number of modules in a string, for example, which would drop the string voltage below the min inverter operating voltage, or you could have row-to-row shading on modules in landscape which would engage bypass diodes which would also drop the string voltage below the min inverter operating voltage. In the PVsyst model, however, I'm having trouble seeing when this will occur. For extremely high temperatures, maybe, but not as a result of shading. I have run two models which are identical in every respect other than the row-to-row spacing. Flat roof, single rows in landscape, near shadings defined "according to strings, 100% electrical effect". The wider row-to-row spacing shows 0% "loss due to voltage threshold", while the narrower spacing shows 0.2% "loss due to voltage threshold". With this near shading definition, it seems to me object or row-to-row shading would never result in a low string voltage - as soon as a shadow hits the string, it is considered 100% lost. I've tried running a number of different test models, just trying to better understand this loss factor, but the results aren't intuitive to me. Can you offer some insight into this loss factor? Thank you.

[André Mermoud]

In the "Module Layout" part, you can analyse the electrical effect of a shade by clear sky conditions (when the modulelayout configuration is well defined, using the "Shading 3D" sweep, => "I/V curve").

This is indeed not simple !

You will see that with several strings in parallel, when one string is shaded the voltage of the MPP usually doesn't drop.

As a contrary, with one only string per MPPT, the voltage drop due to beam shading may be important according to the shaded cells.

However with a high diffuse ratio the "best" MPP will usually stay on the curve representing the diffuse, i.e. without significant voltage drop.

[chpavan87]

Hi André Mermoud,

I also see this " Loss due to voltage threshold" factor very significantly from PVsyst 6.4.3 Update.

Correct me if am wrong, this value is highly dependent on temperature. Apart from analysis done by kamomeno68 . I also have done small exercise on this factor with respect to temperature which is as follows.

While defining thermal loss factors in detail system losses, when I select "Free mounted modules with air circulation", the loss factor due to voltage threshold is around 1.6%. But when i select "Semi integrated with air duct behind" the loss value is surging to 7.4%. All other parameters are kept identical in both simulations. To give a detail, I used Fronius Eco series inverter with narrow Mppt Voltage (580 - 800V). Also string voltage exceeds inverter mppt limits at 60degs in either case.

In earlier versions of PVsyst this loss was almost zero with same conditions I stated above.

I just want to understand this loss factor more clearly and factors that will effect this factor. Can you help me with that?

Just out of curiosity, can we consider not to include this factor in reports or show it as zero percent just like in previous versions 8-)

Thank You

[André Mermoud]

I don's see any reason for such an increase of this loss as a function of the temperature. If there is a depency it should be almost unsignificant.

This may be an error of the simulation, that we have to analyze in detail. Please send your full project (using "Files > Export project" in the main menu) to support@pvsyst.com.