Solar Panel Voc and Isc

[ee753]

I have had solar installed on my roof for 8 years. For the first 6 years the power degradation was about 0.8% per year which approximates the panel specification. For the last 2 years the power degradation has been about 6-15% per year (depending on the panel). I have determined that the microinverters are OK because their efficiency has always been 95-100% (even during the bad 2 years). So the problem must be the panels.

At the request of the panel manufacturer, the installer is going to measure Voc (open circuit voltage) and Isc (short circuit current). Of course these measurements depend on the amount of sunlight hitting the panels so the values will vary.

I already know that the Voc will be normal because the voltage at the microinverter (not open circuit) has dropped about 0.8% per year from day 1  until present (including the 2 bad years) which is normal. But we have not measured Isc yet. I have all the V and I data from the microinverters. But I expect that Isc will be slightly lower than what the microinverter measures (not short circuit).

What information about panel health can be determined from Isc that is different from the I measured by the microinverter (not short circuit)?

Other than placing the panel under a known light source, how can the panel be tested on the roof in sunlight?

The attachments show (1) voltage at the microinverter, (2) current at the microinverter and (3) peak 15-minute period power generated in each month.

[thm_w]

You say 6-15% depending on the panel, so you have the individual panel data?
The worst being 6% drop and the best being 15% drop right?

Seems like an odd coincidence that ALL panels are suddenly dropping off, if it was just one it would make sense. I assume you've checked everything is clean and no possible new sources of shading exist right.

[ee753]

Yes, I have data for each individual panel. The microcontroller company furnished data reports that contained data for every 15-minute period of the day for Power Produced (w-hr), DC Volts and DC Amps. The data for the curves posted were from 15-minute period of the month that had the maximum power produced (all between 12-1pm). This removed concerns of clouds, haze, obstructions, etc.

DC Volts and DC Amps are the voltage and current from the panels into the microinverters. Multiplying these gives us the power into the inverters. The Power Produced is the output of the microinverters. Dividing power in by power out equals microinverter efficiency which is unchanged from the beginning. The microinverter efficiencies haven't degraded at all. That's why I believe the problem must be with the panels.

Correct, the best panel dropped 6% per year and the worst panel dropped 15% per over the last 2 years (12% and 30% total for 2 years).

The panels were last cleaned in July 2023 and data from Aug 2023 is consistent with the other monthly and yearly drops. I found previously that even very dirty panels dropped the production by only about 5%. There are no obstructions.

I agree that it is odd that all panels started degrading about the same time. It's almost as if we have had a permanent solar eclipse for the past 2 years! The date codes for all the panels is the same month and year. Maybe there were defects in a batch of panels?

[nctnico]

Any data on mains / line voltage? Maybe the microinverters are throttling down due to excessive mains voltage on sunny days.

I find it extremely odd that all panels start to degrade at the same time. Only thing I can think of is that the protective layer at the back of the panels may have degraded and now the panels start to corrode from the rear. But this should be happening on all panels.

[ee753]

AC Voltage is about 240 VAC and 60 Hz which seems normal. Since the microinverter efficiency is close to 100% they don't seem to be throttling down.

One possibility is moisture entering the panels due to poor seals. But my best guess is failing connectors. Either corrosion or loose connections. Bad production tooling or contaminated wire might cause this. Since all the panels were built in the same month, there could have been a short-term systemic problem at the factory. But that's only a guess.

I still have no idea what measuring the Isc will do without knowing the intensity of the incoming sunlight. That is the next step as directed by the panel manufacturer.

[bdunham7]

How do the panels appear?  Which model panels and microinverters do you have?  If you have Enphase, do you have full access to Enlighten?

[nctnico]

AC Voltage is about 240 VAC and 60 Hz which seems normal

But is this always the case? Over here mains voltage varies between 223V and 240V depending on how much the sun is shining. In some places it varies even more up to the point solar panel inverters switch off due to the mains voltage getting too high.

[floobydust]

I suggest to borrow/rent an IR thermal imaging camera and inspect the array when it's operating.
This site is dated but full of good troubleshooting: https://photovoltaikbuero.de/en/pv-know-how-blog-en/creeping-losses-in-the-output-of-photovoltaic-installations/
https://photovoltaikbuero.de/en/pv-know-how-blog-en/why-would-the-open-circuit-voltage-in-a-solar-panel-string-be-too-low/

[mtwieg]

So my understanding is that you're observing the peak output power from the panels is lower than in previous years (mainly due to a decrease in peak output current). But I don't see any independent measurements of actual irradiance. Simplest explanation is that irradiance is actually lower, rather than some change in the panels and/or electronics.

I still have no idea what measuring the Isc will do without knowing the intensity of the incoming sunlight. That is the next step as directed by the panel manufacturer.

Agreed, perhaps they will also use a calibrated instrument (or a known-good panel) to measure irradiance during the measurement.

[ejeffrey]

AC Voltage is about 240 VAC and 60 Hz which seems normal. Since the microinverter efficiency is close to 100% they don't seem to be throttling down.

The inverters being near 100% efficiency doesn't tell you anything about whether they are throttling.   Inverters will throttle by drawing less current from the panel than available, not by dissipating extra power internally.

[nctnico]

So my understanding is that you're observing the peak output power from the panels is lower than in previous years (mainly due to a decrease in peak output current). But I don't see any independent measurements of actual irradiance. Simplest explanation is that irradiance is actually lower, rather than some change in the panels and/or electronics.

Sounds like a sensible thing to check. Dave tends to pull up a website that has data on solar panel installations around the world in his solar panel related videos. I think it is a good idea to find such a website and check the output of solar panel installations nearby. See if they have a similar decline.

[ee753]

I made a quick visual inspection of the panels with magnification did not see any obvious anomalies.

LG280S1C-B2 panel and Enphase microinverters.

I have access to the consumer Enlighten app and received full data reports from Enphase. The full data reports contain ACV, AC frequency, DCV, DCA, temperature, energy produced (joules) for every 15-minute period of daylight. That data provided the data for the DCV and DCI graphs.
The AC outputs do vary over the day but seem unrelated to power production or light levels. During one of my sample days the ACV was over 246 VAC mid-morning, approximately 244 midday and about 240 VAC just before sunset.

It is hard to believe that the irradiance during the highest 15-minute period in a given month over 8 years has significantly changed. If that were true, it has been dropping more and more each of the last 2 years. This is NOT a step function.  If decreasing irradiance were the issue, all the panels would likely drop by the same percentage. But some dropped 12% over 2 years and others dropped 30%. That's a BIG difference. And I doubt the sun is dimming

If a calibrated panel is used when the installer measures DCI later this week that would make sense. But I will be very surprised if they do that. The measurements would also be useful if the used a pryanometer or pyrheliometer to measure irradiance. I doubt they will do that either. They didn’t mention anything like that. We will see this coming Thursday . . .

Dave tends to pull up a website that has data on solar panel installations around the world in his solar panel related videos. I think it is a good idea to find such a website and check the output of solar panel installations nearby. See if they have a similar decline.

Who is “Dave” and where can I find these websites? Even if I find them, I doubt they will have the detailed data so cloud conditions can be eliminated. And if is highly unlikely that Enphase will give me a neighbor’s data report.

[nctnico]

'Dave' is the one paying to keep this forum up & running 

Maybe start here for looking at data: https://pvoutput.org/outputs.jsp

[woodchips]

You are in the USA, anywhere near the smoke cloud from Canada in the last 2 years?

It did for the dinosaurs, who knows?

[ee753]

Maybe start here for looking at data: https://pvoutput.org/outputs.jsp

Unfortunately not very helpful.

You are in the USA, anywhere near the smoke cloud from Canada in the last 2 years?

Nope. I'm in SoCal so no smoke. Good guess, though.

[bdunham7]

Who is “Dave” and where can I find these websites? Even if I find them, I doubt they will have the detailed data so cloud conditions can be eliminated. And if is highly unlikely that Enphase will give me a neighbor’s data report.

I don't know if I'm close enough to be relevant, but I'm in Upland, CA.  I'm showing 100% of estimated for Lifetime Production (start date January 2013) and 115% of estimated for the current month-to-date.  If your system was set up correctly in the beginning, the production estimates should account for panel degradation and so forth.  I think there's a way to make my system public so you could see it, but I don't know how helpful that would actually be.

[Someone]

Maybe start here for looking at data: https://pvoutput.org/outputs.jsp

Unfortunately not very helpful.

Why not? it quickly shows other systems nearby and their outputs over months/years for you to compare with.

[tszaboo]

It is probably the weather. The second year I had my panels installed, I got +10%.

[bdunham7]

Here are the detailed results of my #24375 inverter from Aug 1-7 2023.  Let me know if the format works and if this information (and more of it) would be useful.

[ee753]

No. Graphical data doesn't help. Here is a portion of the data report from Enphase

[bdunham7]

The attached chart(1).zip contains a .csv file.  You can click on it and download it.

[ee753]

OK. Yes, that is what I need for the days listed.
Thanks

[ee753]

*** UPDATE ***

Another poster was kind enough to send me data from his system. He is located about 100 mi North of me and has very similar weather conditions. Again I chose the days and times with the highest output (no clouds, no haze) when analyzing his data. It turned out that his DCI dropped about 2.0% over the last 2 years (1%/year) which is normal. My DCI dropped between 12% and 30% (6-15%/year) depending on the panel. Clearly there is something wrong with my system - and it's NOT weather.

The installers measured Voc and Isc on 4 panels in strong sunlight today. Isc is specified as 9.3A. The 4 panels measured were 9.9A, 9.2A, 7.2A and 6.0A. Voc was normal for all 4. Clearly less current is coming from at least 2 panels.

We spoke with Enphase and they are convinced that the microinverters are performing properly (as am I).

There was some discussion that unplugging/replugging the connectors might remove oxidation, corrosion, contaminants, etc. Since 2 of the panels were unplugged/replugged last week I checked the DCI today. The DCI dropped 5-8% on those panels from before they were unplugger/replugged. It got worse. My thoughts are that there are certainly issues with the connectors. Maybe the panels, also.

[nctnico]

Connectors can be the problem. I created a topic about this earlier this year:

Turns out you can't mix random connectors which look like MC4 but are in fact not compatible at all:
https://www.eevblog.com/forum/renewable-energy/fyi-mixing-mc4-connector-brands/

On top of that some installers use pliers instead of the official crimping tools to crimp the connectors.

I'd open up a few connectors to inspect the crimps and see if the connectors which are on the solar panels are compatible with the connectors on the Enphase modules.

[bdunham7]

The 4 panels measured were 9.9A, 9.2A, 7.2A and 6.0A. Voc was normal for all 4. Clearly less current is coming from at least 2 panels.

Presuming that was measured with an ammeter directly across the panel output connectors, you clearly have some bad panels.  If that reduction in current were due to a bad connection, that connection would quite quickly be glowing red hot as it would be dissipating ~150W.  Also, where could there be a bad connection external to the panel if you have it shorted with the meter?

[ee753]

If that reduction in current were due to a bad connection, that connection would quite quickly be glowing red hot as it would be dissipating ~150W.  Also, where could there be a bad connection external to the panel if you have it shorted with the meter?

Isc was measured with a clamp-on meter (not inline).

[bdunham7]

OK, that's fine, but they just shorted the two panel connectors together, right?  So there's no other connectors worry about?  If a panel was capable of 9.9A into a real short but was only putting out 6A due to a bad connection, that connection would get very hot in short order.

[ejeffrey]

That's really weird.  It definitely looks like the panels with 6 and 7.2 amps are not working properly but the other two with 9-10 amps seem OK yet all of your panels are showing abnormally low power? 

[bdunham7]

That's really weird.  It definitely looks like the panels with 6 and 7.2 amps are not working properly but the other two with 9-10 amps seem OK yet all of your panels are showing abnormally low power?

Actually it isn't so weird if you think about it.  Short-circuit current may not be all that useful in finding partial panel failures.  Say you had a the panel that has 3 bypass diodes (typical) for 3 subsections and two of those subsections had one cell with bad busbars that limited its current output to 5 amps.  What do you get when you short the panel?  I think you get the full rated short circuit current or very close to it since you only need enough additional voltage drop to forward bias the 2 bypass diodes, so perhaps 2 volts. All you need is one good subsection to drive full current through that. If you look at a typical MPPT curve, you'll see that the current is fairly constant over a wide voltage range from zero to MPPT.   

[floobydust]

Again, I think a thermal imaging camera can solve the puzzle.
Usually I see cracked cells (1st reason) or corroded interconnects (2nd reason, crappy not even well soldered inside) causing weak output.
At the panel level, any bad connection quickly melts and burns up due to the higher current and dissipation.

[ee753]

I suggest to borrow/rent an IR thermal imaging camera and inspect the array when it's operating.

Are either of these cameras acceptable?

https://www.harborfreight.com/thermal-camerainfrared-digital-imaging-thermometer-58111.html

https://www.homedepot.com/p/VEVOR-Thermal-Imaging-Camera-240x180-IR-Resolution-43200-Pixel-Infrared-Camera-with-16G-SD-Card-LED-Light-Minus-4-F-to-662-F-FYHWRXYWWWIFINPTKV0/325835409

Should the inspection be performed several inches from the panels or several feet?

[f4eru]

Do thermal cameras even see anything on panels ?
I inspected thermal solar panels once, and all I could see was the reflection of the sky in the glass.

[ejeffrey]

They probably won't see much from the silicon itself but if an interconnect is heating up locally due to bad contact that can be apparent.

[mzzj]

Do thermal cameras even see anything on panels ?
I inspected thermal solar panels once, and all I could see was the reflection of the sky in the glass.

Haven't tested myself but supposedly it works based on examples I have seen.
https://westerninfrared.com/problems/solar-panels/
https://www.irpod.net/en/news/fault-diagnosis-of-photovoltaic-systems-using-infrared-thermal-imaging-cameras/

If the panels don't have any faulty spots then you won't see any differences and what you see is mostly "low quality" mirror surface.

[nctnico]

Do thermal cameras even see anything on panels ?
I inspected thermal solar panels once, and all I could see was the reflection of the sky in the glass.

AFAIK it will greatly depend on the angle and how reflective the panels are. In theory solar panels shouldn't be super reflective so a thermal camera should pickup the hot spots.

[ee753]

I used a thermal imaging camera to scan 2 of the panels (the best and worst performing panels).I divided each panel into 9 areas so there a 9 scans per panel.

After looking at the images I did not find anything remarkable. There weren't any hotspots on the cell borders (or anywhere else, for that matter). But then, I have no experience looking at thermal images so others may see something that I missed. Others may also recommend that I make additional scans (closer, different angle, or something else).

Attached are images from the worst performing panel (the best performer will be in my next post – limited to 10 images per post).

[ee753]

Here are the images for the best performing panel.

[floobydust]

I think looking for small temperature differences, oddball spots is most important.

Interconnects get hotter if their resistance goes up. Shorted blocking diodes heat up, so that is easier to find.

But solar cells that are putting out more current... run cooler. It's counter intuitive.
About 2-5°C cooler as the incoming energy gets put out into electricity- instead of dissipating in the cell that is doing nothing.
So look for the slightly hotter objects and see if that is a pattern. Cells along the panel outside perimeter get better cooling and are naturally cooler.

pic '708 best' has one cell hot top left? One cell would make that string of say the three, weaker.

The only other way to troubleshoot arrays is treating them like LED's and inject say 3A in to a panel at night to get them to emit infra-red light and again look for inconsistencies.
That one guy is injecting 400VDC 3A to do that!

[ee753]

pic '708 best' has one cell hot top left? One cell would make that string of say the three, weaker.

It looks like 708 (1) mid-left, (2) mid-right, (3) upper-left, (4) upper-mid all have cells that are running hot. From what you are saying these 4 hot cells could be reducing the output. That makes sense.

But I don't see any similar hot cells on panel 809 which had the worst current output of all my panels. That is puzzling.

I am still suspicious of the MC4 connectors, also. I found that the connectors were installed at the panel and inverter factories (not by the installers). That implies that the proper crimping tools were used and the connections should be good. If they were attached by the installers using pliers rather than a crimper the joint would be highly suspect. But that is not the case. I will inspect the connectors if the installer lifts the panels. I should be able to determine whether a crimper or pliers was used.

If my calculations are correct, a 1 ohm resistance in the connector would drop the current from 8.5A to 6.5A. The power dissipated in the connector would be about 40W. That would make the connector quite warm, but not blisteringly hot. When the installers lift the panels I will measure the temperature of the connectors with the thermal imager. With a near-zero resistance I expect their temps to be about the same as the shaded roof temperature. If the temperature is significantly above that the connections may be an issue.

The microinverters have been eliminated as the culprit because their efficiencies are all above 95% and unchanged from Day 1. Currently waiting for the panel manufacturer (LG) to get back to the installer with further instructions.

[nctnico]

The microinverters have been eliminated as the culprit because their efficiencies are all above 95% and unchanged from Day 1. Currently waiting for the panel manufacturer (LG) to get back to the installer with further instructions.

Have you tried to swap the microinverters between the panels? Microinverter efficiency says nothing if they are somehow broken and fail to do MPPT properly.

About the thermal images: you'll be better off making an image from the entire array and look for hot spots. From your images it looks like you also caught a lot of reflection causing too high/too low temperatures. Using a thermal camera on a glossy surface is not a good idea, especially if you want to make comparisons between images.

[ee753]

Good news!

After Enphase agreed that the microinverters were functioning properly and after providing data to LG they will be replacing all of my panels. Before receiving LG's response I checked the inverter-to-panel connectors and found them to be at ambient temperature proving that they were not the problem.

LG will not reclaim my original panels. So what do I do with them? Buy a few more inverters and put them on the roof? (Already over-producing so I don't need them unless I purchase an EV.) Sell them at a VERY REDUCED price on eBay as "functioning, but not at 100%"?

[nctnico]

Nice that LG respects the warranty even though they pulled out of the solar panel business. Maybe you'll get panels from a different brand.

FYI: It looks like more people have problems with LG panels:
https://www.classaction.org/lg-solar-panel-problems-lawsuit

[ee753]

FYI: It looks like more people have problems with LG panels:
https://www.classaction.org/lg-solar-panel-problems-lawsuit

Yes, I am aware of that. Clearly I am not the only one. But most people are not geeks like me who analyze the data monthly. That's the only way that I found the problem. There are probably hundreds or thousands of homeowners who have the same problem but don't know it.

Unfortunately lawyers will get involved through classaction.org which will cost everyone more and the owners will get pennies from the class action suit. Better to provide LG with hard data as I did. They have already shipped a complete new set of panels to me and are making me good. They told me to keep the original panels. Not sure what I will do with them. Maybe a recycler will pay me for them. Otherwise, the installer will cart them away.

[boB]

Since power output goes down with cell temperature,  maybe it is progressively warmer now than it has been in past years ?

[mzzj]

They told me to keep the original panels. Not sure what I will do with them. Maybe a recycler will pay me for them. Otherwise, the installer will cart them away.

Based on diysolarforum.com there is  active market in US even for partially broken & used solar panels. SanTan solar is even doing business with them but they are probably not interested in small quantities.
https://www.santansolar.com/product-category/solar-panels/used-solar-panels/

List at 100usd/kWh to Graigslist or something like that?

[ee753]

Good news!

LG replaced all of my 280W panels with 315W panels under warranty. After the originals were removed I placed each in full sunlight (no clouds or other obstructions), pointed them directly at the Sun and measured the open circuit voltage (Voc) and short circuit current (Isc). After correcting for solar irradiance for my location and date I found that all of the panels were within 5% of the nominal specification. Obviously that was a big surprise. I didn’t want to interfere with the installers so I did not measure the new panels.

However, when the new panels were in full sunlight (no clouds or other obstructions) the maximum 15-minute output for the day was 878 W-Hr. Several days earlier under the very similar conditions the original panels output was 616 W-Hr. That is a 42.5% increase in production. Changing from 280W to 315 W panels accounts for a 12.5% increase. That leaves an additional 30% increase. That just happens to be the degradation that I measured on the original panels. So clearly the original panels have significantly degraded
But if they degraded, why is Voc and Isc still within specifications? What other factors affect output?

[nctnico]

That just happens to be the degradation that I measured on the original panels. So clearly the original panels have significantly degraded
But if they degraded, why is Voc and Isc still within specifications? What other factors affect output?

As noted before, measuring a panel at short-circuit is not an operating condition so the result tells you very little. You should have used a DC load and find the maximum power point.

[bdunham7]

If you have one section (typically a third) that has a cell that is cracked or something and it therefore has reduced current carrying capacity, the panel will still produce full voltage when open circuit.  And if you short it, the bypass diodes will allow the full current from the other two sections to flow through unimpeded.  You'll only notice the problem when you try to extract power from the panel and when you do, you'll find that the MPPT voltage will be about 2/3 of what it should be because one section is bypassed.

[ee753]

Now it is starting to make sense. bdunham7's posted a typical MPPT curve on Nov 3, 2023. When Isc is measured V = 0 so there is no power generated. What is important is the MPP (not Isc). In my case the MPPT had probably decreased. Why did both Enphase and LG ask for Voc and Isc? That was worthless information and mislead me. It seems that every installer should have an MPPT tester.

How can I measure MPP without an MPPT tester? I would need a variable load of about 0-5 ohms @ 500+ watts. It looks like an MPPT tester might be the only practical way to measure it.

BTW - My output degraded on every single panel between 8-15% during year 6. They all decreased another 8-15% during year 7. I wonder if they would have continued to degrade at that rate and eventually reach 0 output? Why would they all start losing cells and continue to do so?

[uer166]

If you have an E-load with constant voltage mode, you can sweep Vin and measure power at each setpoint. That's how MPPT controllers work more or less.

If course a dedicated MPPT tester would likely be cheaper, so this only makes sense of you have or can borrow an E-load.