small array big inverter

[Zee]

What is the energy lost if I design a 2.64 kW system with a 6 kW inverter (SMA 6.0)?
My vague memory is the inverter - when under-loaded - performs low on the efficiency curve......
As opposed to a 3.8 kW.

Want to make it easy to add kW later for customer.

[Zee]

FWIW - My supplier sells the 6 kw for about the same as the 3.8 kw so i don't want to cheap out on the customer.

[ggunn]

What is the energy lost if I design a 2.64 kW system with a 6 kW inverter (SMA 6.0)?
My vague memory is the inverter - when under-loaded - performs low on the efficiency curve......
As opposed to a 3.8 kW.

Want to make it easy to add kW later for customer.

That's a question for the inverter manufacturer. I have seen efficiency vs DC loading curves for inverters and they are fairly flat until you get down to a rolloff point, and my gut feeling is that loading it only 50% is on the flat(ish) part of the curve, but you'd have to get the info for the specific inverter to be sure.

Edit: If you look at the data sheet for that line of inverters there is a plot of what you are looking for. If I am reading it correctly, it says that for 50% DC loading the efficiency is actually a little higher than for 100% loading.

[Zee]

Interesting. So it sounds fine.

Although a 2.64 kW rated system will supply much less. Maybe averaging around just 1.5 kW. So maybe it is 25% loading the 6 kW inverter.

[ggunn]

Interesting. So it sounds fine.

Although a 2.64 kW rated system will supply much less. Maybe averaging around just 1.5 kW. So maybe it is 25% loading the 6 kW inverter.

I'm pretty sure that the chart uses STC module ratings for the DC loading.

[pv_n00b]

Make it easy to add to later by using micro-inverters. I've added a screenshot of the efficiency curve for a 6kW SMA 240V inverter from PVSyst to give you an idea where it rolls off.

[jaggedben]

Make it easy to add to later by using micro-inverters. ...

That would be trusting that the micro manufacturer stays in business and makes things backwards compatible.

[310 BLAZE IT]

If you are under loading, just make sure your string lengths aren't too short or the inverter may trip out on hot days when the voltage is low. I would only consider this cheaping out the customer if the customer actually expects to expand their system later. Download SAM (free) https://sam.nrel.gov/ and model it to see what the production differences are. It's important to model versus estimate because of the amount of time the system produces between 0 and 10 percent of capacity.

Agreed on using enphase (who will be around forever, unlike the others) if looking to expand, especially if the system is this small.

[pv_n00b]

That would be trusting that the micro manufacturer stays in business and makes things backwards compatible.

Well, using an oversized string inverter is trusting that one can get modules in the future that are compatible with the inverter and the existing modules. Got to trust something.

Adding to a micro-inverter system will not require backward compatibility. Maybe you might end up with two monitoring systems if the new micro-inverters can't talk to the existing monitoring system. But any micro-inverter is going to output standard AC that can be combined with any other AC and back-fed to the grid. I would recommend this over the oversizing of a string inverter that may or may not ever be added too.

[Carultch]

Well, using an oversized string inverter is trusting that one can get modules in the future that are compatible with the inverter and the existing modules. Got to trust something.

Adding to a micro-inverter system will not require backward compatibility. Maybe you might end up with two monitoring systems if the new micro-inverters can't talk to the existing monitoring system. But any micro-inverter is going to output standard AC that can be combined with any other AC and back-fed to the grid. I would recommend this over the oversizing of a string inverter that may or may not ever be added too.

For inverters in general, you are correct. Everything on the same power processing zone, requires either the same modules & source circuits, or "electrically close enough" for the second phase of construction.

By contrast, if you have an optimizer system, the second half of your array can have completely unrelated modules as the first half of your array, because the optmiziers localize the power processing. Microinverters offer the same benefit.

Another strategy is to use an inverter with multiple MPPT zones, and only connect one of the zones during construction phase 1. The remaining zone remains empty, until you are ready to come back and build construction phase 2.

[ggunn]

For inverters in general, you are correct. Everything on the same power processing zone, requires either the same modules & source circuits, or "electrically close enough" for the second phase of construction.

[...]

Another strategy is to use an inverter with multiple MPPT zones, and only connect one of the zones during construction phase 1. The remaining zone remains empty, until you are ready to come back and build construction phase 2.

But still, adding modules to an existing string inverter PV system can be a non-trivial task. You have to think about voltage matching between strings sharing an MPPT, maximum string voltages under cold conditions and minimum string voltages when it's hot, keeping the same/similar current capability among modules in a string, keeping modules with the same orientation strung together, DC:AC ratios, et. al. Adding modules to a microinverter PV system is a relative no-brainer.