Differences In Panels

[DrLumen]

I just had a 8.4kw PV system installed in June and finally got the PTO from the utility company on July 5th. I am curious as to how much power output difference can be expected between panels. I have some that I'm guessing are good (maybe too good for all I know) and then some in the same general area that output about 10% less. The installer said that is normal but I keep thinking about the ~2% first year warranty that the panel manufacturer claims. Is 10% difference really 'normal'?

I would add more details if I knew what would be relevant.

[Murby]

I don't think 10% is normal.... I smell Chinese panels... What kind are you using?

[max2k]

I just had a 8.4kw PV system installed in June and finally got the PTO from the utility company on July 5th. I am curious as to how much power output difference can be expected between panels. I have some that I'm guessing are good (maybe too good for all I know) and then some in the same general area that output about 10% less. The installer said that is normal but I keep thinking about the ~2% first year warranty that the panel manufacturer claims. Is 10% difference really 'normal'?

I would add more details if I knew what would be relevant.

where do you see the difference in power (kW) or energy (kWh)? Any shading, difference in orientation / location? Without all those details which actually matter 2 panels from the same batch should not be that different but that would mean they're next to each other in the middle of the roof without any sun or air flow obstructions.

[J.P.M.]

What methods are you using to determine output differences ?
Are your findings on day long or instantaneous measurements or longer ?
Do all the panels have the same orientation ?
Is there any shading of the panels ?

A 10 % difference in output is outside of any reliable and quality panel specs. Are you certain that's what the installer is telling you. If so, and it's not shading or different orientations, it sounds like the B.S. you'd get from a "Larry with a ladder" type installer.

Or as Murby notes, in effect, crappy equipment.

Did you spend any time doing research on equipment or at least equally important, vendor quality/experience/integrity ?

[DrLumen]

Murby, they are Silfab 300B watt panels connected to a SolarEdge SE7600 and optimizers.

max2k, there are no shading issues. The differences are in the daily readings and accumulate over time. They are roof mounted and on the same plane. Here is the southern array layout showing the differences over about a 6 week time span. The west array (12 panels) shows some differences but not as drastic as these.


Before settling on a contractor I did some research and compared equipment across the bidders and these appeared to be the highest quality of the bids submitted. The others were quoting HuaWei (something like that) panels and enphase or solarcity inverters.

The installer also mentioned that something in the attic might be causing them to be hotter but there is nothing in the attic or house that would generate heat in these specific places.

[ButchDeal]

The differences are in the daily readings and accumulate over time.

On the web site you can click on the Daily button and then click on show playback.
you can scroll through the day.

Also you can compare it to the modules right next to it in the charts tab for voltage as well as watts.

you can't really compare module 1.0.13 to 1.0.21 directly like this . 1.0.21 is on the end and gets more air flow.
you cn compare 1.0.13 to 1.0.22 & 1.0.24 where it is not that far off.

Also modules are binned and come with -0,+5 watts so 1.0.21 could be on the high side there while some of the others are on the low side.

[Murby]

Is your array on the ground or on your roof?

[cebury]

I had an intermittent problem with an SE module which showed lower kwh like this, or you may not have seen a problem, depending on whether it had issues on a day in that range. I eventually ran custom reports showing Power for the offending panel/module pair and ran it through a graph over a month and voila you coild see the days which it was acting up. Sometimes it could produce 100% normally until 3pm then shutoff, all other modules and panels were fine. Other times it shutoff at noon, it was random. Which was why it just seemed a bit low on energy production. It got worss and then I called the installer.

Assuming no very small intermittent issue, yours really doesnt look that bad. +-5kwh from the neighbor panels may be normal. And the fact the other low producer is right underneath it makes me suspect there is a heat issue or a very small shade issue, like a roof jack, kids tetherball pole? Did they install the array over any attic vents (probably wouldnt be that easy).

Take Butchdeals advice here over mine, I've only helped a dozen or so folks troubleshoot their system issues like this.

[DrLumen]

Due to the way that SolarEdge does their reporting, it is impossible to get any type of real time data to compare panels. For example, here is the data for about a 1 hour period:

Seconds would be helpful as well but no go (see line 210). Either they only poll the optimizers about every 18 minutes or they only retain some type of 15-18 minute increments on their server. The inverter only communicates with the SolarEdge servers every 4 hours. Either way, it is almost impossible to get very granular with the data - hence the 6 wk summary as above.

What I am ultimately gathering is that, under these conditions, it is almost impossible to try to force any type of panel performance warranty replacement. Is that correct?

I am exploring some type of modbus logging. At this point, I am not a fan of SolarEdge. I hate all this cloud crap (solar and internet). I am seriously considering ripping out a 6 wk old inverter and optimizers and going to a different system.

Sorry for the rant. If this is normal then so be it. What bothers me is that there seems to be no way to know for sure what is "normal".

[ButchDeal]

Due to the way that SolarEdge does their reporting, it is impossible to get any type of real time data to compare panels.

Why do you need realtime?

Seconds would be helpful as well but no go (see line 210). Either they only poll the optimizers about every 18 minutes or they only retain some type of 15-18 minute increments on their server. The inverter only communicates with the SolarEdge servers every 4 hours. Either way, it is almost impossible to get very granular with the data - hence the 6 wk summary as above.

If your inverter is only reporting every 4 hours then that is because it is not connected via ethernet, most likely GSM.
Also the optimizers are not polled, they report in at random intervals to avoid collisions.

What I am ultimately gathering is that, under these conditions, it is almost impossible to try to force any type of panel performance warranty replacement. Is that correct?

We have several times. you look at the live data on the charts section of the solaredge monitoring cloud interface on a clear day and you can compare two live feeds (yes they will not be at the same second, thus the relatively clear day).

I am exploring some type of modbus logging. At this point, I am not a fan of SolarEdge. I hate all this cloud crap (solar and internet). I am seriously considering ripping out a 6 wk old inverter and optimizers and going to a different system.

Sorry for the rant. If this is normal then so be it. What bothers me is that there seems to be no way to know for sure what is "normal".

I think you are drastically overthinking things here but your equipment. Maybe you would be happier with a string inverter, they do say ignorance is bliss....

[sensij]

Due to the way that SolarEdge does their reporting, it is impossible to get any type of real time data to compare panels. For example, here is the data for about a 1 hour period:



Seconds would be helpful as well but no go (see line 210). Either they only poll the optimizers about every 18 minutes or they only retain some type of 15-18 minute increments on their server. The inverter only communicates with the SolarEdge servers every 4 hours. Either way, it is almost impossible to get very granular with the data - hence the 6 wk summary as above.

What I am ultimately gathering is that, under these conditions, it is almost impossible to try to force any type of panel performance warranty replacement. Is that correct?

I am exploring some type of modbus logging. At this point, I am not a fan of SolarEdge. I hate all this cloud crap (solar and internet). I am seriously considering ripping out a 6 wk old inverter and optimizers and going to a different system.

Sorry for the rant. If this is normal then so be it. What bothers me is that there seems to be no way to know for sure what is "normal".

The optimizers have a randomizing component to the timing for each report. Based on the information in SolarEdge's patents, this is to help prevent message collisions. The effect is as you've seen... non-deterministic timing that makes it very hard to compare real-time data. Around solar noon on a clear day, the 15-20 min difference should have the smallest effect, but you've still got thermal changes to consider.

I have a 10% difference on the system I just installed this spring. Here is some production data I captured a couple of weeks ago:
production.JPG



I attribute the variation to a combination of panel variation and temperature, but the overall effect on my system's production is so small that it doesn't bother me.

Let's try to pencil out what that panel is costing you. The average of the other 15 panels (ignoring the low one) is 81.9 kWh. The low one was 76.4 kWh. The difference is 5.5 kWh, over 6 weeks. How much is 5.5 kWh of electricity at your current rates?

Before ripping out the entire SolarEdge system in frustration, you should at least consider swapping the positions of the low and high panels, to see how much of the difference follows. Thermal gradients on the roof can explain a lot, and operate in ways that aren't always intuitive.

[cebury]

It is most likely fine and normal.

Supposedly Enphase has stated the main reason they stopped giving free access for customers to see their own data was the type of frustration from the OP generating so many support calls and complaints. People want to know why they arent exactly performing the same. More reasonable people who understand there are numerous factors involved at least want to know the exact "range of what normal is". That is impossible to give because again it depends on installation factors and moving a panel three feet either direction can drastically change its production.


What about just swapping 1.0.1 with 1.0.13 ? Either the installeer or OP himself (if installer doesnt care) can read up on doing it safely and are physically fit enough to do the roof work.

[sdold]

A 10% difference doesn't seem too bad. I see at least that on mine. I don't know what the accuracy of the power measurement is, but if it's +/- 5%, you'd probably see a difference of around 10%. Remember that the number you see is what the inverter thinks it is, not what the panel is actually producing. For that you'd need to measure the voltage and current at the input to the inverter. I think it's better to look at the average and decide if that is a good number.

[DrLumen]

I will probably try swapping panels around to see if it makes a difference. As to heat, I can't see how there would be anything in that specific area to cause more heat. The panels are over the living room and a master bedroom which have high ceilings. If anything, I would have expected the western facing panels to show some heat effects as they are over the kitchen, a water heater and a non air-conditioned garage.

Granted 5.5 kw is not a lot but it will add up. While the utility power cost is not that much to replace the loss, $3.00 per watt for install is to be considered too.

I have looked into getting an infra-red camera to see if it is actually a heating issue. However, I could just about replace the panels for the cost of a good FLIR camera or to have a heat survey done.

I guess I'll continue to watch it for a while... Thanks for the insight and suggestions!

[SunEagle]

I went to post a response and got an "Unapproved Post" message. I'm not going to type it again.

Thanks for the responses and insights!

I am not sure why your other post went to Unapproved. I approved it so that others can now read it.