Grid-tie inverter sizing ?

[garybeck]

I have noticed that a lot of grid-tie packages (kits) include an array that outputs more power than the maximum output of the inverter.

for example, a 3400 watt array packaged with a Fronius 3000IG inverter. You can even go on to the Fronius website and use their online configurator tool to size your array strings and it actually recommends as "optimum" to use an array that is larger than the inverter.

I called Fronius and they said that if the array is in full (peak) sun and capable of putting out 3400 watts, the inverter will only put the maximum rating to the grid (or your appliances), and in this case 400 watts is going out the window.

I understand the inverters operate most efficiently under higher loads, but when I look at the efficiency curves it looks like they're all pretty good at 50% or greater. So I'm confused as to why you would size an array that is larger than the inverter output. One would hope to have significant time when the sun is shining full.

Is this all because of the discrepancy between standard test conditions and real life?

thanks

[Mike90250]

........... Is this all because of the discrepancy between standard test conditions and real life?

Yep. There is about a 20% loss, as panels heat up in the sun (that navy blue color sure gets hot) Panels get tested at 72F with a flash strobe, so they don't heat up. Otherwise folks buy a 5KW system, and only see 4.3KW sold.

And as the sun moves, your roof does not track, so there is only a 20 minute window per day, where you get full power anyway. All the GT inverters throttle themselves back (supposedly safely) when hot, or overloaded.
I've installed 4.5KW, on a west facing roof, and only max out my 3.8Kw Xantrex a few days a year.

[garybeck]

And as the sun moves, your roof does not track, so there is only a 20 minute window per day, where you get full power anyway. All the GT inverters throttle themselves back (supposedly safely) when hot, or overloaded.
I've installed 4.5KW, on a west facing roof, and only max out my 3.8Kw Xantrex a few days a year.

well some people do have trackers... and I thought these inverters are so good at tracking the maximum power point.... plus i normally like to leave extra room on the inverter so people can add panels. I guess it just seems strange to spec out an inverter that's smaller than the array. I'm on the Fronius website now and their online config tool is letting me put a 4.4K array on a 3.8k inverter!

[Mike90250]

You are confusing a system where the panels follow the sun as it moves across the sky (tracker) with
MPPT electronics, where the electronic package adjusts itself, to harvest the most power from the panel, regardless how it is aimed. While it will maximize the harvest, the panel will never output it's full power, unless it is precisely aimed at the sun, noon, cloudless.

[garybeck]

You are confusing a system where the panels follow the sun as it moves across the sky (tracker) with
MPPT electronics, where the electronic package adjusts itself, to harvest the most power from the panel, regardless how it is aimed. While it will maximize the harvest, the panel will never output it's full power, unless it is precisely aimed at the sun, noon, cloudless.

hey Mike, i'm not confusing trackers with MPPT controllers... it might have sounded that way, but believe me I know the difference. my reference to array trackers was because someone said that there's only a 20 minunte window when you could be getting close too full sun... I think the window is bigger than that with a tracker. yes "sun, noon, cloudless" is required to get the full output, "sun, noon, cloudless" is something that happens every day in many places. It is true, STC is not a real life scenario... I'm guess I'm just having a hard time spec'ing an array that's bigger than the inverter... I like to tell the customer they can add panels in the future and even with a 20% reduction due to STC, that aint gonna happen if they have a 3.4k array and a 3k inverter.

[solaree]

You can generally size the array (using STC values) 10% to 25% over the rating of the inverter. To justify doing this you need to add up all of the losses in the system.

• Array tilt and orientation. Are you directly South and tilted at Latitude?
• Module power variation (5%)
• Mismatch
• Soiling
• Shading
• DC Wiring losses
• Inverter Efficiency
• Etc (See derate factors in PV Watts for typical values)

By the time you add up all of these factors, you can be derated as much as 25% from the ideal system at STC.

[garybeck]

nice list solaree... i think you're missing the biggest factor which is temperature. STC is at 25C and solar panelsare almost always significantly hotter than that when the sun is shining on them. I noticed that some manufacturers like Evergreen now provide specs on more than just STC, reflecting more "real life" scenarios. They also give the Voc, Vpp, Isc, and Ipp at higher temperatures. It's amazing what affect the temperature has on PV cells. IMHO they should change the temperature of STC.

Gary

[solar_geoff]

All valid points mentioned above... The main point being the DC wattage of the array doesn't equal the AC wattage of the system. Remember the DC wattage uses the nameplate wattage of the system, which is produced under STC laboratory conditions. The formula we use to calculate the AC wattage of a system is as follows:

(Module PTC wattage)x(#modules)x(inverter efficiency)

An example using 40 Evergreen ES-A-200 fa3 modules on a SMA7000, DC wattage is obviously 8.0 kw with a 7000W inverter. 2 strings of 20 are obviously within the voltage window of the inverter:

(180.6w)x(40)x(.955)=6898.92 W

As you can see, by this method, this system is perfectly configured.

Also, it is actually desireable to oversize the DC array to some degree, because as we all know, the output from the array is going to decrease somewhat over time (~4-10% depending on who you talk to. By oversizing the DC array, you ensure that your inverter is outputting as close to its maximum capacity as possible as long as possible into the future. As I have mentioned in several other posts, as long your array is configured with the correct size string which provides the correct voltage for the inverter's voltage window, you can add as many strings in parallel as you want without damaging the inverter. It would be possible to string 1MW on this same SMA7000 as long as the string size didn't exceed 22 or so, all of the excess power would just not be created... This is all that happens when the DC wattage exceeds the inverter AC wattage.

[Foreverenergy]

Hi Geoff. Can you tell me how you can 'string' up to 1 MW on the SB7000 when the maximum output amps on this inverter is only 34A?
Have I missed something?

[Black6spdZ]

Hi Geoff. Can you tell me how you can 'string' up to 1 MW on the SB7000 when the maximum output amps on this inverter is only 34A?
Have I missed something?

Easily, just parallel endless amounts of strings.. voltage wont increase and the inverter will only draw up to its maximum rated wattage/amps from the panels. So if you had 10x the amount of parallel strings necessary you would only draw 1/10th the needed load on each string.

[russ]

Easily, just parallel endless amounts of strings.. voltage wont increase and the inverter will only draw up to its maximum rated wattage/amps from the panels. So if you had 10x the amount of parallel strings necessary you would only draw 1/10th the needed load on each string.

Pardon?

[Black6spdZ]

Not that anyone would wire 1MW worth of panels to a 3KW inverter.. Just saying electrically it could be done.

[russ]

Not that anyone would wire 1MW worth of panels to a 3KW inverter.. Just saying electrically it could be done.

Let us stick to the real world.