Questions about Power Optimizers

[julmou]

I've recently been diving into the topic of Power Optimizers, which I thought was pretty simple, but their behaviour is actually more complex than I thought.

So, I just want to clarify something about what I've been reading in the Help Documentation. See screenshot below -> two questions written in bold

[Michele Oliosi]

Hi,

Basically there are two situtations:

• One string per MPPT input. In this case, the optimizers allow for the current to remain the same as Impp, and you "recover" your electrical mismatch losses (i.e. the loss due to shading is mitigated). This is formally equivalent to the case where you connect several strings to the same MPPT input, but they are all shaded in a similar way.
• Several strings per MPPT input, but not all are shaded. This leads to a voltage mismatch between strings, which prevents you from recovering your electrical mismatch losses due to shadings.

I'll try to slightly change the formulation in the help to make things more clear.

 

[julmou]

I also have a question regarding the fixed voltage of a SolarEdge inverter input.

I have a system with one string of 6x DualSun 500Wp modules and 1x SE3000H inverter. This is on a residential house that is quite prone to shading.

From what I observe with the "Output I/V behaviour" tool, in the definitions of a power optimizer, in case of heavy partial shading, the MPP power plateau shortens at the string level -> see screenshot of the P/V curve below (example with 4x optimizers in series)

We can see in this case that the voltage range, at the string level, to keep the Pmpp is between 60V and 190V (not until 340V anymore, as was the case without shading).

If we have 6x modules/optimizers in series (see 2nd screenshot below), with a fixed voltage at 380V, and no shading -> this is okay, the Impp of the string and the various Vout of each optimizer will adjust accordingly to deliver the Pmpp.

But if we have a heavy partial shading (as shown in the previous case with the 4 optimizers in series), we can see that the MPP power plateau could shorten and the Pmpp might only be achieved under a voltage lower than 380V. Hence having a fixed voltage in this case could incur losses.

Help me understand this... Is this correct? It would seem then that under certain circumstances, having an inverter input fixed voltage with optimizers might a disadvantage rather than a benefit?

[André Mermoud]

When connecting SolarEdge optimizers  (Buck/Boost)  in series, you have to add the voltages.

Now if you have different powers (shaded optimizer), the power curves are added down to the voltage limitation of the higher power optimizer. Below this current value you add a portion of hyperbola, and the voltage limitation value: this leads to a little disruption in the curve:

Drawn as power, this becomes:

There is indeed a degradation at high voltages.

This is the reason why SolarEdge imposes a minimum number of optimizers in series  (for attaining the fixed input voltage of the inverter), but also advises to foresee one or two additional optimizers for ensuring a good prevention of the shading effects.  This is mentioned in the help.

 

[julmou]

Thanks so much for the explanation :)