Wildly different production across 3 sets of panels - same location.

[sd70]

HI,

I have a system with 66 panels across 3 sets, each connected to SolarEdge SE7600 inverters. The panels are ground mount in one large array, so all panels have the same angle, though obviously some are closer to the ground than others. There is no localized shading.

The 3 inverters each have different production values -- so far today, inverter #1, representing the top 2 rows of my array, produced 26.43 kwh.
Inverter #2 (middle of array) is at 10.37
Inverter #3 (bottom of array) is at 22.0 kwh

It's unclear to me why the middle would be so different. My installer is extremely slow to respond, and this is my 2nd incident after a problematic install process (the first incident involved an inverter replacement, I think on the #2 inverter) so I don't have much confidence in their ability or willingness to help.

I'm looking for advice on how to troubleshoot this further. In addition to the screencaps below two other things jumped out at me when looking through the SolarEdge reports.

1) In site commissioning, I have 2 different optimizer part numbers -- 56x P400-5NC4ARM-NM21 and 10x P400-5NC4ARM-NA21.
2) also in site commissioning, under my inverters the #1 and #3 inverters show "US 240V" under the Country Name column of the report, but the 'new' #2 inverter (which appears to have an older serial #, if the #'s are sequenced) shows just "USA" in the country name.
3) from the Layout screen, if I go into the errors tab on the Inverter Info, there are a number of V-Line Max errors showing. Inverter #1 has 34 accumulated. Inverter #2 has 824, and Inverter #3 has 175. The error count seems to correlate to total produciton, in that inverter #1 has lowest error count and highest production.
4) the logical layout of the panels is quite different on inverter #2 and than the others. I don't know if this matters, or if its something I can easily fix.


Regarding #3 it appears related to incoming AC voltage from the grid, and SolarEdge notes one cause as distance from grid connection point. The wire run from array to main panel is approx 350 ft, and I do not know what gauge of wire was used.

Thanks in advance
Paul

Screen Shot 2018-09-19 at 1.22.09 PM.png
Screen Shot 2018-09-19 at 1.30.37 PM.png

[J.P.M.]

Get on the installer and make them respond. You have a system that appears to have a problem. It's educated guessing from here as to cause(s). What's needed is on site analysis/inspection/troubleshooting from there. You purchased a system that came with a warranty. Expect the installer to hold up their end.
.
BTW, fool with it yourself and you'll probably risk voiding the warranty.

[bcroe]

if there are no shading issues, I wonder why optimizers are even used? Bruce Roe

[J.P.M.]

if there are no shading issues, I wonder why optimizers are even used? Bruce Roe

If it was still in the design phase, I would too, but now it is what it is.

[Mike90250]

if there are no shading issues, I wonder why optimizers are even used? Bruce Roe

A really good salesperson. only reason. Sell more parts. More parts to go bad or mis-configure

[ButchDeal]

Screen Shot 2018-09-19 at 1.30.37 PM.png

it looks like inverter 2 is miss wired (guess) you can look at individual optimizer outputs and see if they really did one long string like that then you would see a lot of them pretty much shutting down during the peak. I doubt it is really wired like that as you have two of every module in the string 2.0.1 which doesn't make since.
Check the version numbers on the inverters, I bed #2 is a lot lower than the other two.

As for the V line max you can look at graphs for the different inverters line voltage and see what is going on, under charts.

[sd70]

Thanks for the input guys. I'm definitely riding the installer, just like to be educated on this stuff -- if they've misconfigured or miswired I don't want them to just pass it off as a Solar Edge problem, which is what they did when Inverter #2 died the first time.

[sd70]

Regarding versions:
Inverter #1 and #3: Communications board 3.2221.0, DSP1 1.0.451, DSP2 2.0.405
Inverter #2 (replacement): Communications 3.2173.0, DSP1 1.0.350, DSP2 2.0.205

All 3 inverters appear to follow the same AC voltage pattern (see attached), but #2 throws many more errors and apparently is shutting down production each time, which is why it is so much lower.

Tech is coming out today after I chased them again yesterday. Will be interesting to see what he comes up with.

Screen Shot 2018-09-20 at 8.41.11 AM.png

[JSchnee21]

Was the entirety of inverter 2 replaced, or just some of its internal boards? We've seen a number of posts here on the forum where when individual boards are replaced they are frequently not programmed correctly and are set to limit production. Aka the board might be pre-programmed from SE for a 5K inverter when it is installed in a 7.6K unit the production limit (in software) needs to be adjusted.

I wouldn't place too much value on the drawing of the 2nd inverters strings in SE since the entering of this data is completely uncoupled from the wiring reality save for the diligence of the person entering it.

What kinds of errors are you seeing on Inverter 2 during production? Are any errors being reported up to the SE monitoring portal?

You could also contact SE and have them remote connect to the Inverter in question and update the firmware and do some basic troubleshooting.

[ButchDeal]

I wouldn't place too much value on the drawing of the 2nd inverters strings in SE since the entering of this data is completely uncoupled from the wiring reality save for the diligence of the person entering it.

The installer doesn't label the optimizers though. OP has double of every optimizer label which shouldn't happen. I don't think that is wiring but something is hinky there.

[sd70]

Was the entirety of inverter 2 replaced, or just some of its internal boards? We've seen a number of posts here on the forum where when individual boards are replaced they are frequently not programmed correctly and are set to limit production. Aka the board might be pre-programmed from SE for a 5K inverter when it is installed in a 7.6K unit the production limit (in software) needs to be adjusted.

I'm not sure to be honest -- but a technician came out yesterday and spent a bunch of time on the phone with Solar Edge. They decided the country code setting was wrong ("USA" vs "US 240V" on the other 2) and changed that. Production was much better and there were no errors after that change, but it might be too early to draw conclusions. Clearly whoever swapped the inverter (or boards, as case may be) was sloppy though.

I wouldn't place too much value on the drawing of the 2nd inverters strings in SE since the entering of this data is completely uncoupled from the wiring reality save for the diligence of the person entering it.

That's what the tech said too, but I agree with ButchDeal that it's a sign of someone sloppy and I'm going to double check the physical mappings. Sloppy is an apt description for the challenges I've dealt with from the start of the install on -- so not really surprised at this point.

What kinds of errors are you seeing on Inverter 2 during production? Are any errors being reported up to the SE monitoring portal?

Inverter #2 has 800+ V-Line Max errors, the other two had nominal quantities of them. I'm told this is something else in my house coming on or off and causing an AC voltage spike, but that doesn't explain why #2 is so much more sensitive. Seems like there were no more errors after the country code change though, so maybe "USA" vs "US 240V" changes its sensitivity. V-Line max causes the inverter to take production offline whenever it happens.

You could also contact SE and have them remote connect to the Inverter in question and update the firmware and do some basic troubleshooting.

Good idea, thanks.

[ButchDeal]

Did they correct the labeling on Inverter #2 ?

Inverter #2 has 800+ V-Line Max errors, the other two had nominal quantities of them. I'm told this is something else in my house coming on or off and causing an AC voltage spike, but that doesn't explain why #2 is so much more sensitive. Seems like there were no more errors after the country code change though, so maybe "USA" vs "US 240V" changes its sensitivity. V-Line max causes the inverter to take production offline whenever it happens.

I am not sure of the settings but USA 240V might have had Inverters #1 and #3 pushing up the voltage a bit out of Inverter #2s "USA" range.

HEre is a document on the settings. it looks like "USA" is the main catagory and defaults to "auto" or "USA auto" and the other two were set to USA and 240V


Page 27: https://www.solaredge.com/sites/defa..._countries.pdf

[JSchnee21]

Was it an error like this?

70 8x25 VLL Max
Voltage too high
o Verify that the inverter is set to the correct country.
o Consult the grid operator.
o If permitted by local authorities, change the grid protection values.

It could be that your PoCo's line voltage is too close to the max voltage threshold and under high PV production scenarios the local increase in line voltage was causing the unit to trip off. This can also occur if there is any resistance to current flow from the inverter back to your utility service (aka bad termination / line side tap / corrosion / cracked conductor) The increased local resistance in this case restricts the flow of current coming from the inverter during production and can cause a localized increase in apparent line voltage as measured in the inverter.

There are multiple "USA" country codes with different sets of min, max, slew, etc. properties to "negotiate" with different grid conditions / requirements around the US. Mine says USA1 (or possibly USA3 -- I forget). I've not seen a USA 240V before. Does this install have normal residential power? (aka 240V split phase?) How many amps is the service connection? How far away are the inverters from the meter / electric pole? I've not been able to find a table of the different USA specific country codes and the various tolerance values they take. This seems to be something SE doesn't share too readily and not all installers know about.

Hopefully this is just a minor issue of an incorrect country code. If not, you may need to have an electrician check and measure the resistance of all of the connections / conductors and perform thermal imaging under load to look for hot spots.

[sd70]

My error code was 55 for V-Line Max, but didn't have the descriptive part about verifying country. Since the tech changed to USA 240V the production graph looks much better. I'm still seeing low numbers on the layout screen, so I need to find more time to explore reporting and see if perhaps layout just has lag or something. There have been no VLL Max errors on any of the inverters since he made the change.

Our house has 600amp service with a very large panel (I'd call it commercial style, but only because I've never seen one this large on a house before), so definitely not "normal". Inverters are 350 feet from the panel by the way the wire routes.

Thanks for your help -- I'll update the thread in a week or so when I've confirmed things are stable.

[tyab]

For the inverters to be 350 feet away from the service entrance - I would be very curious what gauge they used for that long AC run.