new member here - looking to learn

[ajonestx]

From looking at the daily graphs (day interval), it looks like the second half of the year track fairly well, and that the bulk of the difference is in the first half ( I assume day # goes from January first (1) till December 31st (365)

The difference between the 2017 (blue) and 2020 (green) could probably be seasonal weather difference.

The fact that during the second half of the year, the production tracks better would indicate that if there was an actual HW problem, it would have fixed itself which is rare, and so also comfort weather as the more likely source.

The fact that each year at times crosses over to be the top production is also encouraging and pointing to seasonal weather rather than HW failure

That said, 2019 (red) looks way low in April May June, So I would make a separate hourly graph of the three years for each of these month.
April only with the 3 years, May only with the three years, and June only with the three years, to see if the offset is uniform or crosses over.

But the next step would really require to have access to the panel level data, first with daily interval during these three month.
If the difference is weather related, all panels will pretty much track, but if the panels do not track it may point to panels or micro inverter failures.
It may also reveal if your shade is evolving for the worst. Trees grow taller each year and could play a big factor.

Troubleshooting is where having panel monitoring is worth it. You have micro inverters, you may want to request or pay for full panel level monitoring.

Good morning scrambler - thank you for the feedback. this exercise is helping me evaluate the available data. I will have to keep pressuring the manufacture to give (sell?) me panel level data... for now, I only have a few snap shots at various times and have to work with what I have on that front. Per your suggestion, I will work on building analysis spreadsheets to evaluate those odd spring months over time. But your comment on HW problems (and the fact that I only cleaned the panels for the first time in March of this year) begs me to think that maybe I can work with something that I do have and removes "some" of those factors - MAX DAILY SYSTEM KW. This morning, I took the time to download and identify daily maximum kW values since I cleaned the panels in March of this year. I am seeing a little uptick, but still way low compared to 2017.

Again, blue is 2017, red is 2019 and green is 2020. New color is orange which represents spring of 2021. Ignore the 2017 references at the bottom - they are January, February, March and April of all 4 years. the light colored lines are the actual values and the large dots are 25 day average trends. Maybe as I collect more data over time and put panel cleaning on my schedule as a yearly or twice a year event, I can further evaluate.

Not sure if I will ever get panel level data - what are your thoughts about how to test the panels or install my own monitoring system? I just feel like I am flying blind.

2021 Spring power compared to 2017.JPG

[scrambler]

If you have Enphase micro inverters, there are two things I believe you can do.

Ask to have Installer privilege on your account so you can access all the data.
From reading people experiences on this forum, it looks like sometimes the installer refuses it (they don't want you to find out problems...), but you can go directly to Enphase to have them give that to you.
In some cases, I have also heard they may want you to pay a subscription for panel level data access.
Enphase users may be able to chime in.

But I would stand firm, you paid a lot of money for your system, and individual panel monitoring is one of the key advantages of microinverters, so you should have access to the full monitoring period!

On another note, it seems the one odd year is 2017, and I am wondering if there are any historical database of Solar radiation in your area that would allow you to see if 2017 was a special year or not. If you can give your county in CA, maybe someone here can provide their daily production graphs for the same years to see if there is a similar trend.
This is not exact as irradiation varies for location to location, and you have shade, but it could be a valuable data point if it shows the same difference between the first half of 2017 and the succeeding years.

Also, can you clarify your shade issue. Need to know what the shade comes from. Is it a fixed structure that does not change overtime, or a variable structure like trees?
If so check when does the shade happen (hours of the day), and what predominant month it happens (shade can be bigger in winter season when the sun is low and the shadows get longer). Then extrapolate how that will change if the trees are growing taller (longer hours periods, and longer day periods...)

[ajonestx]

If you have Enphase micro inverters, there are two things I believe you can do.

Ask to have Installer privilege on your account so you can access all the data.
From reading people experiences on this forum, it looks like sometimes the installer refuses it (they don't want you to find out problems...), but you can go directly to Enphase to have them give that to you.
In some cases, I have also heard they may want you to pay a subscription for panel level data access.
Enphase users may be able to chime in.

But I would stand firm, you paid a lot of money for your system, and individual panel monitoring is one of the key advantages of microinverters, so you should have access to the full monitoring period!

On another note, it seems the one odd year is 2017, and I am wondering if there are any historical database of Solar radiation in your area that would allow you to see if 2017 was a special year or not. If you can give your county in CA, maybe someone here can provide their daily production graphs for the same years to see if there is a similar trend.
This is not exact as irradiation varies for location to location, and you have shade, but it could be a valuable data point if it shows the same difference between the first half of 2017 and the succeeding years.

Also, can you clarify your shade issue. Need to know what the shade comes from. Is it a fixed structure that does not change overtime, or a variable structure like trees?
If so check when does the shade happen (hours of the day), and what predominant month it happens (shade can be bigger in winter season when the sun is low and the shadows get longer). Then extrapolate how that will change if the trees are growing taller (longer hours periods, and longer day periods...)

interesting... I had not thought of going to the manufacturer of the micro inverters as an option. I might get up on the roof over the weekend and see if I can capture some name plate or manufacturer data, then research further.

I appreciate the encouragement on insisting that panel level monitoring is a requirement. I guess I was just ignorant at time of purchase and could not imagine that this would not be provided after seeing other people's system and the ability to see this deep into the system. Shame on me.

2017 being an odd year? Huh, had not thought of that either. I would be interested if anyone knew of a local database (or comparative system) that would shed some light on this idea for reference. This is all so "shades of grey". I think this is one more reason to stay focused on max kW per day or per week or other... (and not kWh), right?

I also wonder if maybe when the inverters were replaced in 2018, that maybe I got a bad batch and they are not performing up to specs at 320W in subsequent years...

[ajonestx]

If you have Enphase micro inverters, there are two things I believe you can do.

Also, can you clarify your shade issue. Need to know what the shade comes from. Is it a fixed structure that does not change overtime, or a variable structure like trees?
If so check when does the shade happen (hours of the day), and what predominant month it happens (shade can be bigger in winter season when the sun is low and the shadows get longer). Then extrapolate how that will change if the trees are growing taller (longer hours periods, and longer day periods...)

I missed this one question about trees in my last response. Yes, our shade is variable - trees. We have trimmed them once (December of 2018) since install in 2017. Since then, one has gotten shorter (rot at the top) and one has gotten taller. Trees are deciduous and during the winter, sun passes mostly through them and in the Summer, sun passes over them. Based on "sun eye" shade analysis, originally estimated production levels were discounted by 16% for the shade. it is probably time to trim them again. Every panel sees full sun during the day, so that is why I am thinking that kW is the right focus (I think).

[RichardCullip]

... Ialso wonder if maybe when the inverters were replaced in 2018, that maybe I got a bad batch and they are not performing up to specs at 320W in subsequent years...

well that’s a decent reason why 2017 is the anomaly in your performance. Seems like 2019,2020 and 2021 are performing within a tight range. Any idea on what the inverter change was?

[ajonestx]

well that’s a decent reason why 2017 is the anomaly in your performance. Seems like 2019,2020 and 2021 are performing within a tight range. Any idea on what the inverter change was?

no clue what the inverters were changed to, only that the original ones failed and they replaced them under warranty.

[scrambler]

Well this is a very important data point, you need to get the make and exact model of the original micro inverters as well as the one for the new ones.
(you should not have to go on the rood for that, your installer should provide that information!)
You also need to tells us if All of them were replaced or just some of them
You also need to tell when they were replaced (all at once on a specific date, or across what period of time)
Finally you need to say why they were replaced, did they all fail to zero? that would be strange for micro inverter to all fail at the same time.

All that said, there is still the discrepancy between first and second half of year....

[ajonestx]

Well this is a very important data point, you need to get the make and exact model of the original micro inverters as well as the one for the new ones.
(you should not have to go on the rood for that, your installer should provide that information!)
You also need to tells us if All of them were replaced or just some of them
You also need to tell when they were replaced (all at once on a specific date, or across what period of time)
Finally you need to say why they were replaced, did they all fail to zero? that would be strange for micro inverter to all fail at the same time.

All that said, there is still the discrepancy between first and second half of year....

ok. I have the model number of the originally installed micro inverters from my agreement with the original installer. The replacements were installed by a 3rd party installer - I'll look into what I can find on the replacement model numbers were.

ALL of the micro inverters were replaced at the same time and turned back on June 4, 2018. I am pretty sure they ALL failed to zero over a period of time - since I do not have access to inverter data, I can only look at system drops and they appear to begin dropping down about six months prior to 100% failure. On the last day before the system was turned off (May 23, 2018), the 5,760 W system maxed out at 660 W.

Let me take another look at the data and think about the discrepancy between the first and second half of the year.

I really appreciate all the help here people!

[ajonestx]

Well this is a very important data point, you need to get the make and exact model of the original micro inverters as well as the one for the new ones.

All that said, there is still the discrepancy between first and second half of year....

Scrambler - I have a working theory on the discrepancy between the first half of the year and the second half of the year. I looked back into old emails and there were some odd micro-inverter errors reporting in June of 2017. My theory is that maybe they started failing in June of 2017 with their ultimate replacement one year later in May of 2018. I added max kW to the chart for 2018 below in cyan with a 25 day average trend line.

[scrambler]

That is a better graph, having the real vertical scale from zero gives a better perspective, the drop in the first half being less pronounced.
Can you share what micro inverters were initially installed, and did you find which ones are installed now.

Did they, or can they tell you what lead to ALL of them failing, because it is very unlikely you were unlucky enough to get a bad batch, and it would seem like may there was a flaw in the installation itself leading to their death. If so it would be nice to know what and what was done to remedy the issue so they would not fail again.

[ajonestx]

That is a better graph, having the real vertical scale from zero gives a better perspective, the drop in the first half being less pronounced.
Can you share what micro inverters were initially installed, and did you find which ones are installed now.

Did they, or can they tell you what lead to ALL of them failing, because it is very unlikely you were unlucky enough to get a bad batch, and it would seem like may there was a flaw in the installation itself leading to their death. If so it would be nice to know what and what was done to remedy the issue so they would not fail again.

My graph is getting better as I add more data and tweak the presentation - I might add notes and a better index next time! I am an average excel user and creating graphs and plots are not my expertise. Practice makes perfect.

In my agreement as well as the engineering documents, it states that the micro inverts are 320's. I have made lots of phone calls this morning and sent lots of emails, but have yet to confirm what model of micro inverters are installed now. Still digging, but still do not have the new micro inverter information yet - I will get it somehow.

I was never told what lead the original micro inverters to fail. Actually, I don't think they ever admitted that they did fail... only that after 6 months of emails and phone calls did a truck show up one day to replace them. I should have been more in their business and documented what the replacement equipment was. I did hear that there were other customers who had their micro inverters replaced at about the same time, so I am inclined to think that there was a problem with the micro inverter and not the installation.

[oregon_phil]

327 watt panels with 320 watt microinverters seems an odd enough combination to be googleable. Is this the combination that you have? If yes, is there some history on SunPower Microinverters (other than Sunpower now uses Enphase)? If this is not relevant, sorry for the diversion.

[scrambler]

it states that the micro inverts are 320's..

This is not much to go on
What brand and hopefully actual Model Numbers (should be in all your warranty / manuals given at end of installation)

[Mike90250]

I heard of a failure mode being traced to the "installer" gathering up all the gear, hanging by the cables (so as to carry more per trip) and after being swung around the water seals on the cables failed, couple years later, the gear ( I can't recall if they were micro's or optimizers ) had a +80% failure rate

[ajonestx]

This is not much to go on
What brand and hopefully actual Model Numbers (should be in all your warranty / manuals given at end of installation)

No manuals received at end of installation. Let me do a little homework this week and give you more data (actual model numbers). Having difficulty getting data...