My Solaredge system dead. How to figure if it's an optimizer or inverter that's bad?

[kny]

7.2kw system: 22 Sunpower E-327 panels.
13 on west-facing string
9 on south-facing string
P400 optimizers on all
SE6000 inverter

On Friday my system quit producing. Most certainly not coincidentally, but this was a cool May day and when the sun missed the clouds the system was clipping heavily (see attachment from my Curb Energy CT clamp monitoring system - which gives minute level detail), as much as I've ever seen it clip. Which gives an indication this was possibly the peak power ever the system has produced.

The system quit. On the inverter it shows that only 21 of 22 power optimizers are reporting. It also shows Error Code 31 = AC Voltage Too High. The inverter seems to be in a constant state of trying to come online, faulting, and restarting.

The system does try to come back on. On Friday afternoon it had two hiccups of performance, quitting again after each. Saturday and Sunday early mornings and late evenings it also puts out small power before quitting for the day. In the monitoring portal, the inverter seems always to be in a "Starting" state.

My installer has not been responsive. SolarEdge does not deal with homeowners. I just want to know whether we're dealing with a bad PO or a bad inverter. Or both at this point. Everything should be under warranty; it's just 18 months old.

Each of the 13 and 9 string panels should be well within the 5250 watt string limit. So, despite clipping on cool May days, the system should be within operating parameters.

DC Voltage seems to have jumped up from a steady 370 volts to mid 400s when it died. Maybe it goes up to 500, the upper operating limit, and the system shuts down then? See attachment showing the DC voltage going above the prior day steady-state of 370. What would cause system voltage to shoot above the normal operating 370 volts?

Monitoring portal provides "info" on each optimizer on a one-by-one level. I see nothing in any of them that would indicate it is "dead". Though I may not know what to look for. At a minimum all 22 do report in data to the monitoring portal, even if the inverter shows 21 of 22 on its screen.

Help. Thoughts? Optimizer or inverter? The fact that this occurred on a day of probably the greatest peak power produced in the system lifetime (cool mid May) is surely not coincidental. But, as the 13x327 and 9x327 strings are within operating bounds, why?

Help!

[ButchDeal]

Repair the system when all modules are in some light. The system will move the D.C. Voltage up to prevent over current on the AC side. What is the AC voltage graph for this day? What is the power graph ( from solaredge monitoring) for the day?

[kny]

AC Voltage is unremarkable that day. You see some jaggedness in the 1:30pm timeframe when the problem starts, but stays in the 240 range all day. The plummets to 0 in the evening are when I am powering off and on the inverter.

AC Power shows same as the full day power chart from my Curb Energy. Peaks around 6kw and then out. Two hiccups in the afternoon and then out for good.

fwiw, my P400s are not grounded. Based on the installation documentation they are supposed to be installed with a special washer that digs into the aluminum rails, but I took a look at how a few had been installed and that special piercing washer was not used. Probably totally unrelated, but now that I'm aware they aren't grounded properly I of course, gravitate to that....

[bcroe]

Perhaps you are another victim of HIGH AC LINE VOLTAGE. This is most likely to occur when the inverters
are pushing out hard and raising the voltage even higher. Shut down occurs, voltage drops, restart occurs,
voltage goes up, shutdown, etc. Check the AC voltage. Bruce Roe

[ButchDeal]

You may not have enough temporal revolution on the solaredge monitoring but yu can highlight and zoom in to some of those peaks and see if your voltage is getting up there. Maybe your curb has better temporal resolution but it is measuring at a different place. Regardless the high voltage should be apparent. It doesn't look l Ike the SE6000 ever clipped from the power graphs you show.
Which st Inc did you loose the one optimizer from? The one with only 9? Could have set it pretty on edge and might explain the high D.C. Voltages which ar not too unusual for peak production times.

[kny]

Bruce Roe: The error is 31 - AC Voltage Too High, so yes, maybe I am a victim of HIGH AC LINE VOLTAGE. How do I check this?

ButchDeal, the system most definitely was clipping and clipping extensively. You don't see it in the Solaredge Monitoring chart I showed because it aggregates over 15 minutes and it was a mix of clouds and sun so drops bring the 15 minute aggregate under 6kw. Look at the attachment, which shows AC power via CT clamp measurements on each of the two hots at the breaker (the light green is one hot leg, the darker green the other hot leg). You can clearly see flatlining, and the peak is 5904 watts, so the inverter was pushing out 6kw and we lost 96 to line losses or small measurement error on either end. But, you can definitely see significant clipping. Nonetheless, that should not be a problem. Clipping on a cool May day is fine. It's rare, but should be fine.

How do I find out which power optimizer is lost? The only indication I have lost one is that the inverter reports "21 of 22 reporting" How do I make the inverter tell me which is missing (or which 21 are reporting?). The SolarEdge Monitoring portal does provide info on each of the 22, if I click on each and click on Info, one at a time for each. So, if one is dead, it is still calling home to SolarEdge.

It does seem there was some acute condition. Earlier in the week the system has put out 6kw or very near. Yet DC voltage stayed pegged at 370. But then something happened and DC voltage spiked to 450 or so and then system died. To me it does indicate that something "broke", and probably caused by the extreme condition (all panels putting out 300+) despite system design being totally within specifications.

One attachment here shows how the system was definitely clipping.
The second attachment shows how the DC voltage is always 370 volts, even on prior days when the system was working hard and putting out 6kw. But then something happens and voltage skyrockets and system dies.

[ButchDeal]

I don't see this clipping and it shouldn't be clipping at 5904w with higher voltage anyway. Even your definition of a "peak" belies clipping.
to see the optimizers go to the solaredge monitoring web site and click on layout view, then weekly and playback. You should see one wink out.
that is if the installer gave you access to the layout view.

btw, D.C. Voltage goes up pricipitusly with AC voltage increases.
Also are you sure you have p400 optimizers? Some installers always install the p400 but p370 is better for high power modules though it is pretty new.

[kny]

Between 12:30pm and 1:40pm I count 6 horizontal flat lines at the ~5900 watt level. There is no chance that is not clipping.

I have the playback and no single panel obviously drops out. They all ratchet down to 0 between 1:30 and 2:00pm when the system shuts down, but there is no indication that any single panel looks different from the rest in any kind of analysis I can do on the portal.

My invoice says P400 optimizers so I assume that is what was installed.

I am curious the "Perhaps you are another victim of HIGH AC LINE VOLTAGE." from Bruce Roe. Can please describe in more detail what others have experienced here? If I have high line voltage and my system pushes me over the limit, would I not expect to be operable the next day, a cloudy affair?

[ButchDeal]

Between 12:30pm and 1:40pm I count 6 horizontal flat lines at the ~5900 watt level. There is no chance that is not clipping.

I have the playback and no single panel obviously drops out. They all ratchet down to 0 between 1:30 and 2:00pm when the system shuts down, but there is no indication that any single panel looks different from the rest in any kind of analysis I can do on the portal.

My invoice says P400 optimizers so I assume that is what was installed.

I am curious the "Perhaps you are another victim of HIGH AC LINE VOLTAGE." from Bruce Roe. Can please describe in more detail what others have experienced here? If I have high line voltage and my system pushes me over the limit, would I not expect to be operable the next day, a cloudy affair?

Clipping would be at a set voltage the exact same voltage for an hour or more. You have roughly flat but not exactly flat at the time of day it should be at the peak of the parabolic curve. Your inverter should not clip st 5904 watts anyway. If it is limiting to 5994 watts there is something wrong.

What bruce is talking about is when the grid his high voltage and/or wire sizes are marginal then when inverters are pushing high amps the voltage at the inverter climbs too high and they cut out as they are required, Voltage drops because inverter isn't pushing high amps on marginal wire with marginal grid voltage, so the inverter turns back on, ramps up, repeats....
it is possible that is happening. You do seem to have high grid voltage at times at the inverter. What voltage does your other equipment show for AC voltage?
dies solaredge monitoring show the optimizers as p400?
are they grounded with ground wire sinner they did not use grounding washers?

did you turn off and back on the inverter! You can also go into the menus ( when sun is up) and have it re-pair the optimizers.

[kny]

re: grounding. The POs are not grounded with wire or grounding washers. The rails are grounded with wire, but no grounding washers for the POs.

Yes, I did turn inverter off and on. Twice! I have not tried to learn how to use the menus. There's just a single button at my disposal, so don't know how to scroll. But, I can learn all that I'm sure.

Don't know the AC voltage for other equipment in my house.

SE-6000 inverter, about 100 feet of 10AWG wire running from inverter to panel. 30 amp breaker. 200 amp panel.

If issue were high grid voltage (and Error Code 31 - AC VOLTAGE TOO HIGH is good indication of this) exacerbated by inverter pushing high amps, would this not go away when inverter is not pushing high amps? Like yesterday and today, cloudy days? All indications are I am down and out for the count, although the system does make feeble attempts to put out power in the mornings and evenings.

Watching my system like the solar nut I am, I can promise you those flat lines are clipping. And, I've never seen >5904 watts via the CT clamp measurement at the breaker box, 100' downstream from the inverter. I have no doubt that either those clamps measure low or there is line loss between inverter and panel or the inverter clips low, but I'm certain that is clipping. And, it does not matter, other than to show it was a high production time when the system went bad. Lastly, as an aside, I need to take 3.5% off my inverter reads in order to reconcile with my revenue grade utility meter, so 5904 vs 6000 does not surprise me in the least. But, again, I don't want to get distracted by clipping or not. You are being great help and I want to pursue this potential grid voltage issue.

[kny]

Ok, look at this attachment. This is today. The system worked from sunup until about 8:10 am and then from 6:45pm until sundown.

Just after 8am when the power output got to about 1.2kw, the AC voltage jumped from 244 volts to 252 volts. DC voltage jumped from 370 to 435 volts. System shut down presumably due to high AC volts.

For the remainder of the day the ac volts stayed in the 244-245 range. DC volts stayed jaggedly in the 400s. The two times in the afternoon the DC volts dropped (briefly) back to the 370 range and the system briefly produced power.

Then, starting at 6:45pm until after sunset the DC volts were at 370 volts and the system operated normally.

So, what the heck is happening while the sun is shining with strength (actually today was cloudy so not much real strength) that causes DC volts up in the 400s? AC volts are 244-245 throughout that period.

[ButchDeal]

Lastly, as an aside, I need to take 3.5% off my inverter reads in order to reconcile with my revenue grade utility meter, so 5904 vs 6000 does not surprise me in the least.

that 5904 watts wasn't measured at the inverter but your curb. If you take off 3.5% it gets even more unlikely as well.

what voltage did the curb record, 100 feet from the inverter?


Ok well 100 feet of 10 awg is 5 volts drop or rise in your case. A bit much for such a run and will cause problem
that and the fact that your optimizers are not grounded show that your installer is not up to par.

How did you shout down the inverter? Turn off the D.C. Disconnect then the inverter on/off switch? Or inverter on/off switch? Improper inverter shutdown at high load can cause problems. Another poster posted about that recently.

[ButchDeal]

Did all your optimizers come on today when the inverter was on? What is that AC voltage spike going up to? I really think you have some bad connections on that AC cable and the cable itself is undersized You shouln't see that kind of voltage increase at 1kw
was your installer certified?

[kny]

Optimizers operated today. Pathetically, but they all report as contributing their tiny bit to the tiny total of the day. See attachment.

My installer was certified. They are the dominant player in Washington, DC. That doesn't mean that they don't take shortcuts. Are you saying that due to the length of the run they should have used 8 AWG?

When I turned off inverter it was already in fault mode and it was late evening. So, not under high load. But, I turned off DC Disconnect and then the power on/off switch.

Don't worry about the whole 3.5% thing. I have learned empirically that my SolarEdge inverter over-reports production by 3.5%. Or there are 3.5% of line losses between inverter and the utility meter. But, doesn't matter. Irrelevant here.

[ButchDeal]

Before you mentioned strings of 13 and 9 but that map shows strings of 12 and 10.

your previous post you talked about a voltage drop rise in your case of 8 volts. That is due to poor connections and the 10 awg wire.
There may well be other issue as well. You already mentioned no grounding on the optimizers. Nothing is irrelevant when trying to diagnose complex systems. The SE6000 shouldn't be clipping and your graphs are not really showing clipping that is no to say it is looking right though.
You should definitely get the installer out there and likely management to review. I would look at getting a different master electrician familiar with solar to review it.
optimizer 1.2.5 seems low compared to the rest. Unless there is a shadow on it.

you should turn the inverter off first, then after the D.C. Drops you can turn off the D.C. Disconnect. You don't need to turn off the D.C. Disconnect though if you are just trying to reset the inverter just the inverter off switch.