Micro Inverter vs Optimizer/Inverter vs String

[solarix]

"Micro inverters will fail. If you have an array installed into your roof, you will have to climb up onto your roof, remove a bunch of panels just to access the bad inverter, replace the inverter, re-install the panels, and climb down from your roof.

Every time you climb onto your roof and start messing with your panels, you risk damaging something other than the bad inverter. You also risk damage to your roof every time you climb up there. On a ground mounted array, micro inverters are easy to replace. But up on a roof of a home, a string inverter is preferred, IMO."

Ditto... and don't believe the marketing hype known as a 25+ year warranty on roof mounted electronics.

[solar_newbie]

I just monitor my systems for couple days and notice majority of panels products the same amount energy, around 6 out of 18 panels are different. 4 generate 2-3% more and 2 generates 2-3% less.

For enphase or solaredge, system will collect entirely all energy. But string inverter will lose 2-3% (as it run with lowest panel.

Note: All 18 panels in my system face the same direction and same condition without any shade.

Is that concluded string inveter will lose 2-3% of total system output ? So I would assume installer must charge 5% cheaper or more for string inverter?

[solarix]

As a early adopter of the SolarEdge distributed inverter system, the turning point back to string inverters came for me when I tested them personally by splitting my 24 Trina panels into two identical arrays and ran one half on SolarEdge and the other half on the old Xantrex GT inverter. There is no shading except very early in the morning when the south facing 15deg tilted racks shade each other a bit.
Drum roll please ...... the Xantrex inverter wins every day. Just by a little bit. I've never seen a single day when the SolarEdge did better. The SolarEdge gets a bit of a head start, but the string inverter does better during the prime hours and the SolarEdge doesn't catch up at the end of the day. Don't believe everything you read about individual optimization being 2% better here and 2% better there and "up to" 25% better in general.

[inetdog]

As a early adopter of the SolarEdge distributed inverter system, the turning point back to string inverters came for me when I tested them personally by splitting my 24 Trina panels into two identical arrays and ran one half on SolarEdge and the other half on the old Xantrex GT inverter. There is no shading except very early in the morning when the south facing 15deg tilted racks shade each other a bit.
Drum roll please ...... the Xantrex inverter wins every day. Just by a little bit. I've never seen a single day when the SolarEdge did better. The SolarEdge gets a bit of a head start, but the string inverter does better during the prime hours and the SolarEdge doesn't catch up at the end of the day. Don't believe everything you read about individual optimization being 2% better here and 2% better there and "up to" 25% better in general.

Do you have any partial shading? If not, then your results are what the "experts" here use to justify recommending a string inverter.

If you have two or more parallel strings going making up your array, the optimizers will work better than the string inverter if any one of the strings is getting partial shade.

[sensij]

As a early adopter of the SolarEdge distributed inverter system, the turning point back to string inverters came for me when I tested them personally by splitting my 24 Trina panels into two identical arrays and ran one half on SolarEdge and the other half on the old Xantrex GT inverter. There is no shading except very early in the morning when the south facing 15deg tilted racks shade each other a bit.
Drum roll please ...... the Xantrex inverter wins every day. Just by a little bit. I've never seen a single day when the SolarEdge did better. The SolarEdge gets a bit of a head start, but the string inverter does better during the prime hours and the SolarEdge doesn't catch up at the end of the day. Don't believe everything you read about individual optimization being 2% better here and 2% better there and "up to" 25% better in general.

Is it safe to assume you would not post a result like this, unless it was based on the output of a revenue grade meter? Do you have public data that others can review to verify this claim?

[solarix]

Concerning my head to head comparison of SolarEdge vs old Zantrex:
There is no shading on this array.
Both arrays were originally one array so all panels were purchased together and were in the same pallet.
They are all mounted the same with the same orientation.
No, I have not tried swapping arrays to see the difference.
No, the readings are from the inverter's display, not from revenue meters.
If the performance advantage is so small that it is difficult to measure then I don't think I want to pay for it.
The resulting observation is over a 3 year period.
I'm just reporting my observation - your milage may vary. Take it for what it is worth.
I'm sorry if this pops your bubble about the advantage of individual optimization, I know it did mine.
I still install SolarEdge systems when shady conditions require it.
So far I've had 2 (out of 12) optimizer failures on this system.
I had the SolarEdge rep come visit one time and I showed him this situation personally. Offered to work with them on using it as a test site, but not interested.

My interest in sharing this is just to spread the benefit of my experience. We install about one house a week, and I want to see solar become widely adopted by the mainstream homeowners so am always looking for better solutions not just what "sounds good".
I fear having a widespread bad premature failure rate that will damage the reputation of solar.

[sensij]

Thanks for clarifying your comments. I definitely agree that output within the measuring error of the inverter can be considered close enough, and that there will be some installations in which string inverters will perform just as well as panel level MPPT. In looking at PVOutput data, it appears to me that there are many more systems that get some amount of shade than those that are truly unshaded, making the unshaded use case in the minority. "Paying for it" is also not clear, since in some installs panel MPPT is the same price or less than the installed cost of a string system, especially if looking at SMA.

The lousy failure rate of the M190 doesn't seem to have done much to hurt PV adoption, and the current generation of SE and Enphase products all seem to be much better than that. In a world where LG can be considered a "premium" brand, you know people must have a short memory.

Don't get me wrong, there are reasons not to like SolarEdge or Enphase, just as there are reasons not to like SMA or any other product. I think for many the rational decision between them is closer than you make it out to be, although panel MPPT does tend to get oversold by most.

[rrwsandiego]

So are string inverters actually cheaper? The quotes I got it was actually more. What is the draw to each?

[solar_newbie]

So are string inverters actually cheaper? The quotes I got it was actually more. What is the draw to each?

Is it the same installer ? If not, it is not apple to Apple.

[J.P.M.]

Concerning my head to head comparison of SolarEdge vs old Zantrex:
There is no shading on this array.
Both arrays were originally one array so all panels were purchased together and were in the same pallet.
They are all mounted the same with the same orientation.
No, I have not tried swapping arrays to see the difference.
No, the readings are from the inverter's display, not from revenue meters.
If the performance advantage is so small that it is difficult to measure then I don't think I want to pay for it.
The resulting observation is over a 3 year period.
I'm just reporting my observation - your milage may vary. Take it for what it is worth.
I'm sorry if this pops your bubble about the advantage of individual optimization, I know it did mine.
I still install SolarEdge systems when shady conditions require it.
So far I've had 2 (out of 12) optimizer failures on this system.
I had the SolarEdge rep come visit one time and I showed him this situation personally. Offered to work with them on using it as a test site, but not interested.

My interest in sharing this is just to spread the benefit of my experience. We install about one house a week, and I want to see solar become widely adopted by the mainstream homeowners so am always looking for better solutions not just what "sounds good".
I fear having a widespread bad premature failure rate that will damage the reputation of solar.

I appreciate your monitoring efforts. Lots of fun huh ?

On the differences: Assuming the per panel variance in performance for the entire array is small and also tends to even out as a total per half of the array, I'd suggest at least some of the difference is due to the tolerance of the reporting equipment as well as some likely slight differences in how the different inverters handle the input(s). A few % difference on a fairly consistent basis, particularly on clear days would lead me to suspect slight instrument differences. Any quasi "instantaneous" readings ?

[solar pete]

Thanks for clarifying your comments. I definitely agree that output within the measuring error of the inverter can be considered close enough, and that there will be some installations in which string inverters will perform just as well as panel level MPPT. In looking at PVOutput data, it appears to me that there are many more systems that get some amount of shade than those that are truly unshaded, making the unshaded use case in the minority. "Paying for it" is also not clear, since in some installs panel MPPT is the same price or less than the installed cost of a string system, especially if looking at SMA.

The lousy failure rate of the M190 doesn't seem to have done much to hurt PV adoption, and the current generation of SE and Enphase products all seem to be much better than that. In a world where LG can be considered a "premium" brand, you know people must have a short memory.

Don't get me wrong, there are reasons not to like SolarEdge or Enphase, just as there are reasons not to like SMA or any other product. I think for many the rational decision between them is closer than you make it out to be, although panel MPPT does tend to get oversold by most.

Just wanted to chime in re the LG being a premium brand ?, I am stumped to why they seem to get the status they do, my first flat screen tv was LG it was crap, it crapped out about 2 months after the 2 year warranty expired and they said to bad, it was too expensive to fix so just go and buy a new one. OK enough of my rant

[solar_newbie]

Just wanted to chime in re the LG being a premium brand ?, I am stumped to why they seem to get the status they do, my first flat screen tv was LG it was crap, it crapped out about 2 months after the 2 year warranty expired and they said to bad, it was too expensive to fix so just go and buy a new one. OK enough of my rant

I think Korean will promote LG.
If someone is American, try to promote something made in America... Or at least companies create American Jobs.

[solarix]

At least LG is one of the few TV brands that makes their own LCD panels. Most brands buy the panel from Samsung, LG, (or fine makers AUO, CheMei, CPT) and add their own electronics. The typical lifespan of TV's these days is getting worse averaging about 5 years - maybe. (despite the supposed inherent reliability of LCD and LED) So we solar guys are all convinced that a hard working AC converter mounted on a hot roof is going to last 25 years????

[maximizese]

Just wanted to chime in re the LG being a premium brand ?, I am stumped to why they seem to get the status they do, my first flat screen tv was LG it was crap, it crapped out about 2 months after the 2 year warranty expired and they said to bad, it was too expensive to fix so just go and buy a new one. OK enough of my rant

I think you're TV experience is anecdotal and is a broad assumption that their television division has anything to do with their solar energy production division. I could share the same experience with a 1992 Samsung television we had. It was absolute crap and was repaired 3 times in a 12 month period before we junked it. However, Samsung, LG (Lucky Goldstar), and other Korean brands have propelled themselves into much better consumer, commercial, and industrial brands. You can thank the International Monetary Fund for the huge injection of capital in 1997. Now Samsung/LG make some of the best consumer appliances and Hyundai/Kia are miles ahead of where they were in the early 2000s. I'm not trying to puff up Korea…I do my best to avoid flying on a Korean operated airline (Asiana), but I wouldn't rule out their PV panels.

[solarix]

My experience with electronics is more than just anecdotal. After 40 years working in the electronic trenches (quality instrumentation, military avionics, consumer repair, industrial controls, and now solar) , I have a healthy skepticism about long term reliability. I agree that PV panel reliability is the exception in routinely lasting decades and decades. That characteristic is the foundation of the solar industry. But the reliability of inverters is a different story. Getting a consumer quality inverter to last 5 years is very comparable to getting a TV to last 5 years. Once in a while, a TV will go 10 years maybe even 20 years - but they are the exception not the rule. Putting a 25 year warranty on inverters is wishful thinking.