SolarEdge stopped producing power for 1 of 2 strings

[Tlochra e]

Good Morning Solar Friends,

I moved to a new home in Washington DC in 2016 and first "project" I wanted to do was to add solar panels to my house. I installed my system in Dec. 2016 and its been AWESOME!!!

My system is installed on a flat roof with 5% incline as follows:
17 Panasonic VBHN325 panels
17 SolarEdge P-400 optimizers
1 SolaEdge SE5000 inverter

Originally my system was installed with 15 panels in one string then in Apr. 2017 I had the solar installed comeback to install two more panels since there was room. During this installation, my panels were split into two strings; string #1 has 8 panels and string #2 has 9 panels. Everything was great and during the summer my panels were cranking so much power.

However, in Sept 2017 something went wrong and string #1 stopped producing properly. I attached a photo of my layout with a snapshot of the peak this week so you can see string #1 at the top not producing much.

Back in Sept. when I saw something was wrong, I went to check my inverter and it was in German. I put it back in English and wanted to check to see if there were any fault and everything checked out. I called SolarEdge to see if they could check things on their end and they said everything looked fine. They suggested I reach out to my installer to see if they could do an inspection to find this issue. Finally my installer was able to get to my house and we both tired to figure out what was wrong but still could not pin point the issue.

Here are some hypothesis:
- Optimizer 1.1.8 may need to be replaced as it could be the issue. Since this is the first optimizer in the string, if it was not working could this take the whole string down? Since the minimum optimizers per string is 8 and this string has 8 could this be why its offline?
- Possibly the inverter could be malfunctioning? We are going to test to putting all optimizers on the same string next week to see if this could be a fix. To me it is weird that two strings in series connected in parallel to the inverter that the inverter is causing this... However, the reset to German is suspect.

Here is my SolarEdge Portal if you want to look at my system.

https://monitoringpublic.solaredge.com/solaredge-web/p/site/public?name=ManorParkDC#/dashboard

I also attached snapshots form the days of issue and from this summer to compare. The voltage vs time shows hat something happened and changed the string in September.

Anyone have any similar issues with your systems?

Taylor

[ButchDeal]

There are few failure modes that a failed optimizer could take down a whole string, one would be physical damage to the optimizer
The minimum string length is 8 recommended and 6 actual minimum working optimizers.


you should try re-pairing the inverter to all the optimizers during a sunny day before physically altering the connections.
also you can open a case with SolarEdge to review the logs and update the firmware on the inverter.

[Syberdog]

Could the reduced output be due to winter/cloudy weather back there? I noticed the entire production gradually reduced starting in sept and bottomed in Jan but Feb is looking a little better - although IIRC feb is usually a gloomy month back there...

[Tlochra e]

Update:

ButchDeal Thank you for your comments, we went ahead and followed your suggestions. First we checked the logs with SolarEdge and nothing was showing up as an issue. Next we re-paired them in the light-sun and nothing changed.

We went ahead and rewired the strings from 2 into 1 string and then re-paired the optimizers. So far it looked like it worked as all the panels were producing same power like I saw in summer.

We have rain here in DC for the next few days so we will have to wait until the next sunny day to see the full potential and if any optimizers are acting up.

I still wonder what could have been the issue with the two strings vs the single string.

I will repost in a few days when we have sun to report final results.

Thanks everyone.

Taylor

[Tlochra e]

Update:

After putting all the panels on one string they started to operate a lot better however still having issues with the bottom panels on the back row. My installer said it was due to shadow from the front panels but when I went on roof at noon to see this shadow there was only one row of cells being affected, see photo. Side note, my installer went out of business so looks like I am on my own moving forward.

I did try to repair the panels at noon but that put all the panels went into this same mode of producing and the whole string dropped from producing 3300W down to 300-500W for the rest of the day. Not sure why they all started to not function after repairing. I am hoping today after the night mode they had time to reset and will start up normally. Anyone have similar issues during repairing?

Still don't know what could be affecting those four panels, could it be a hardware issue? Just do see how the shadow is the reason.

[ButchDeal]

That shadow is hitting all the sub strings in the portrait modules taking them completely out. This is a terrible design.

[Tlochra e]

ButchDeal Yeah tell me about it. I have already redesigned the organized the panels on my roof and plan to move those panels to the front group. Not sure why my installers thought this would be an ok design. The string is back to normal this morning from repairing them yesterday. Thanks for your input which confirms my plan to moved these panels.

[ImInPhxAZ]

That Panasonic has 4 bypass diodes I'd have a hard time believing they are oriented all vertically so that one bottom row of shading reduces the panel to essentially full cloud cover output. Is this all the panel or is this something to do with solaredge and optimizers and minimum voltage ?

[ButchDeal]

That Panasonic has 4 bypass diodes I'd have a hard time believing they are oriented all vertically so that one bottom row of shading reduces the panel to essentially full cloud cover output. Is this all the panel or is this something to do with solaredge and optimizers and minimum voltage ?

SolarEdge has one of the lowest possible minimum voltage in the industry. The problem is all a very silly installer that made a pretty stupid design.
Is there any way to adjust the first row to cut out the shadow? Or move the other rows away a bit.
i doubt they can be put into landscape which would also solve the problem.

https://eu-solar.panasonic.net/solar/2016_General_Installation_Manual_VBHNxxxSJ25_SJ40_SJ47_EN.pdf

[Tlochra e]

SolarEdge has one of the lowest possible minimum voltage in the industry. The problem is all a very silly installer that made a pretty stupid design.
Is there any way to adjust the first row to cut out the shadow? Or move the other rows away a bit.
i doubt they can be put into landscape which would also solve the problem.

https://eu-solar.panasonic.net/solar...40_SJ47_EN.pdf

Here is how I plan to address it. I think this will solve the issue and by placing the panels horizontal keeping in mind the the design of the diodes which the installers did not understand.

[ImInPhxAZ]

SolarEdge has one of the lowest possible minimum voltage in the industry. The problem is all a very silly installer that made a pretty stupid design.
Is there any way to adjust the first row to cut out the shadow? Or move the other rows away a bit.
i doubt they can be put into landscape which would also solve the problem.

https://eu-solar.panasonic.net/solar...40_SJ47_EN.pdf

No doubt bad install. Just surprised that little of shading killed the output on a new high tech panel. So much that I went up and tested on my roof and had the same results. Would have thought new panels would handle this better. Maybe only lose a fractional output depending on diode count but seems thats strictly for failed cell modules and they still have zero shade tolerance.

Eye opening for sure.

[ButchDeal]

No doubt bad install. Just surprised that little of shading killed the output on a new high tech panel. So much that I went up and tested on my roof and had the same results. Would have thought new panels would handle this better. Maybe only lose a fractional output depending on diode count but seems thats strictly for failed cell modules and they still have zero shade tolerance.

Eye opening for sure.

These would have good shade tolerance if installed correctly. By putting them in portrait instead of landscape the installer negated all the capabilities for inter row shade tolerance.

It is the same thing as a high performance race car, fill the tank with diesel and it will not go very fast. would you blame the high performance race car for not being smart enough to handle the incorrect fuel?

[clearsky]

Here is how I plan to address it. I think this will solve the issue and by placing the panels horizontal keeping in mind the the design of the diodes which the installers did not understand.

Are you able to provide me images again?" Because the links are broken.