String inverters and oversizing of array

[electro7]

Couple questions regarding standard string inverters:

1. Is it true that if one module is shaded it brings down the other modules in that series circuit to the power output of the shaded module? I feel like I am going back to kindergarten about this issue since doing so much with optimizers these days. Do the bypass diodes in the modules allow for the current to flow through and not cause the power output of the rest of the modules in that series to decrease with shading on the string?

2. I have heard, and have done it many times, that oversizing the PV to the inverter is better. I heard from one manufacturer that the inverters are made to work that way and is therefore better for the longevity of the inverters. I understand that modules would rarely hit the stc rating of the modules so maybe oversizing some is okay. But I have had big clipping on Solaredge inverters when I have oversized those in the past, with groundmount arrays facing the same direction. Then I have heard you get more on the shoulders of the IV curve (morning and night) and it far makes up for the clipping that could happen in peak summer time.

So my question here is with standard string inverters, no optimizers, is it better to oversize the PV to the inverter and if so what is the best ratio? And why is that the best ratio? Economics aside.

Thanks!

Sent from my SM-N960U using Tapatalk

 

[ggunn]

So my question here is with standard string inverters, no optimizers, is it better to oversize the PV to the inverter and if so what is the best ratio? And why is that the best ratio? Economics aside.

I always "overload" string inverters, i.e., with DC:AC>1.0, but there is no magic ratio that works for every system. System orientation, system location, stringing configuration, and potential shading are all players. That said, I typically use a rule of thumb of 1.25:1.0 as a starting point.

 

[wwhitney]

On (1), bypass diodes will prevent a single panel being fully or partially shaded from negatively impacting the other panels' power production. Ask if you'd like a detailed explanation.

Cheers, Wayne

 

[electro7]

Thanks guys.

So Wayne, was I brainwashed by Solaredge about optimization? Or do optimizers still monitor the IV curve for each module and "extract" as much power from the module as possible and produce better energy than standard string inverters?

I hooked on to Solaredge early on. Before that I was installing a lot of SMA.

I remember years ago reading an NREL study that did show Solaredge optimizer system did out perform Enphase microinverters and SMA string inverters. I wonder if that has changed now with the new SMA shade fix feature- do you know?

Sent from my SM-N960U using Tapatalk

 

[wwhitney]

Or do optimizers still monitor the IV curve for each module and "extract" as much power from the module as possible and produce better energy than standard string inverters?

The question is how much more energy you get with optimizers and microinverters. For no shading, I think nothing extra. For simple shading, just a little extra. For more complicated shading patterns, the gain may be larger, but how at how many locations and how often does that occur?

Cheers, Wayne

 

[electro7]

Gotcha gotcha.

Sent from my SM-N960U using Tapatalk

 

[jaggedben]

So my question here is with standard string inverters, no optimizers, is it better to oversize the PV to the inverter and if so what is the best ratio? And why is that the best ratio? Economics aside.

There's no 'economics aside', in my opinion. It's all about whether it's more cost effective to pay for more panels or to pay for more inverter power. In residential and small commercial, it generally comes down to 'why pay for more inverter power that will almost never get used'? At all levels, your max inverter output can be limited by existing infrastructure and it rarely pays to upgrade that infrastructure just to be able to install a slightly higher power inverter.

I've never heard that string inverters don't operate as well when arrays are undersized. That is, at least if the start voltages are appropriate and the arrays are within inverter spec. Aside from those small kinds of details, I think the larger question of DC-to-AC ratio is really a totally economic question, not a technical one.

 

[jaggedben]

The question is how much more energy you get with optimizers and microinverters. For no shading, I think nothing extra. For simple shading, just a little extra. For more complicated shading patterns, the gain may be larger, but how at how many locations and how often does that occur?

Cheers, Wayne

For no shading you still have module mismatch and soiling, although arguably MLPE doesn't pay for itself solving those. But it's also not just shading, it's also orientation, string sizing, module mismatch. For residential rooftop systems these were frequently difficult issues to work around even when string inverters started getting two MPPTs. (This all being before rapid shutdown made the point essentially moot.) Micro-inverters and optimizers took all those difficulties away and thus allowed streamlining of business operations even as they generally increased energy output. For larger systems where you have fewer obstacles to putting the modules where you want, maybe not so much.

 

[ggunn]

I've never heard that string inverters don't operate as well when arrays are undersized. That is, at least if the start voltages are appropriate and the arrays are within inverter spec. Aside from those small kinds of details, I think the larger question of DC-to-AC ratio is really a totally economic question, not a technical one.

Many inverters' installation manuals show a plot of inverter efficiency vs. DC loading. The ones I have seen are fairly flat down to a pretty low DC:AC ratio.

 

[tallgirl]

Many inverters' installation manuals show a plot of inverter efficiency vs. DC loading. The ones I have seen are fairly flat down to a pretty low DC:AC ratio.

Inverter efficiency has been pretty flat above small percentages of rated input current for a while. The micros I worked on would charge up the bus capacitors, then produce some power, then wait, then produce. I think we called it "Pulse Mode".

The biggest impact on low power inverter efficiency is the amount of electronics needed to run the power conversion. That's usually just a few (single digit) watts for a micro with a 300+ watt rating. I think on the last micros I worked on it was around 2-3 watts.