New 14kwh system not producing what I expect

[Jriell]

Hi all, I am new to this forum and new to solar but am hoping I can gain some good information here.

I have a 14kwh system with 36 400w panels (solarever) and a SolarEdge inverter.
I live in the Dallas, Texas area and was hoping to be able to produce around 57 kwh per day in October.
I have found that I am actually only producing 40 kwh on sunny days.
I have some shade on my roof at times but it is limited to a small number of panels at certain times of the day.

I wanted to get some information from the knowledge and experience on this forum to better understand what the normal amount of production per panel should be under my conditions?
Could the low numbers relate to the panels or possibly how the strings are arranged?
My output per panel is strange because some panels produce fairly well but others that are in full sunlight are not producing near as much.

I have attached a sample output on a peak part of a sunny day.

Thanks for any and all feedback

[Mike 134]

When you use the PVwatts program what does it show for your expected KWH for the month of October? You can even download by the hour, you'd have to do some math, but you'd see what their average shows for your best possible output.

[oregon_phil]

Hi all, I am new to this forum and new to solar but am hoping I can gain some good information here.

I have a 14kwh system with 36 400w panels (solarever) and a SolarEdge inverter.
I live in the Dallas, Texas area and was hoping to be able to produce around 57 kwh per day in October.
I have found that I am actually only producing 40 kwh on sunny days.
I have some shade on my roof at times but it is limited to a small number of panels at certain times of the day.

I wanted to get some information from the knowledge and experience on this forum to better understand what the normal amount of production per panel should be under my conditions?
Could the low numbers relate to the panels or possibly how the strings are arranged?
My output per panel is strange because some panels produce fairly well but others that are in full sunlight are not producing near as much.

I have attached a sample output on a peak part of a sunny day.

Thanks for any and all feedback

PV Watts is a good idea. Didn't your installer use aurora or helioscope (or equivalent) to predict your solar performance?

What inverter model number do you have?

The SolarEdge image you provided doesn't really show much. I bet you have a pretty complex roof. Did your installer provide you with a set of drawings showing how the solar panels were going to be installed on your roof?

[Mike 134]

Just a follow-up was that layout image taken at least 1/2 hour after sunset? I've found with my SolarEdge some of the data can lag about a 1/2hour.
Here's what this month looks like

[Jriell]

Thanks all for the feedback. Here are some answers:

Inverter Model: SE11400H
PV Watts shows that I should produce 1,786 for the month of October which is inline with what the installer mentioned.
Expected Average per day in October: 57.5
My current production on sunny days: 40-41 kwh

I do have some shading on some of the panels so I do not expect may setup to be perfect but was really thinking it should be higher than 40 per day on full sun days.

My roof is complicated and the diagram I uploaded shows east facing and south facing panels. The south panels get a bit of shade on the bottom row at times but the higher panels do not.
I have uploaded the design here.

[OCJ]

Where did you get your production wishtemite? I did a PVWatts run at 14.4 kW, premium,180 degree South, 1.26 DC/AC ratio and got 1,734 kWh for October, which is 55.9 kWh per day. And your system is not perfectly South.

Cut that it half, 7.2 kW South and 7.2 kW East, that drops to 1,556 kWh for October, or 50.2 kWh per day. So the question in my mind is does that shading account for 20% losses or is there a problem with the system?

The layout of your reporting tool does not seem to match the layout of the panels for your roof, except for that bottom set of panels which in that satellite picture you posted looks like that whole bottom row being taken out by shadows from trees.

To debug this I think you need 2 things:

1. Fix your layout in the reporting tool
2. Start capturing time-series data for your panels

[oregon_phil]

For curiosity sake, for Solar Edge users, I noticed that panel naming convention in OP's first picture goes from 1.0.1 to 1.0.36. Where Mike 134 layout has 1.2.1 to 1.2.9 and 1.11.1 to 1.11.9. To me, this implies Mike 134 layout is 2 strings and OP's is 1 string. I understand logical layout is different than physical layout. Adjacent panel numbers are quite different in OP's layout.

In OP's photo, I counted 21 panels South and 15 panels East.

[Jriell]

Thanks all, the feedback is very helpful. I am still very new to this and did not do much technical homework ahead of time to understand what I should expect out of the production.
I frankly had no idea the amount of power loss that occurs when the the sun is not angled perfectly above the panel. From the feedback, it sounds like the power capture dramatically drops and that my production is not completely out of scope for what I should expect.

OCJ, you are correct that the bottom row of panels on the south side suffer the most from shading. In the morning up until 1 pm, that row gets a decent amount of shade.
I wish my installer explained how the system is strung together as I have no idea. I was not even sure if panel number were supposed to make some sort of sense as it relates to the layout. It certainly makes sense that it should and certain numbers I would think should be grouped together on a string but I am not sure at all how they set this up.

I do believe the bottom row of panels on the south side can be re-arranged and placed a bit higher on the roof to limit shading.
What is strange is that, as I understand it, the level of production is limited to the least common denominator of the string the panels are on. If that is the case, I would expect much less deviation of some panels that are near each other.

What is also strange is that panels 18 and 6 on the upper left facing south produce such low numbers. Those are on the highest part of the roof and not blocked by shading at all. I would think they would be some of the best producers.

Oregon_phil, you are right, the numbering seems much more logic from Mike 134's setup. I am going to ask more questions of the installer to see if my system is on more than one string. I have no idea why they would put it all on one as that does not seem ideal for my type of roof. Any idea how many strings can be run on a system and what the drawbacks are for running multiple?

[foggysail]

Question because I am unfamiliar with PVWatts. The OP appears to have used his advertised panel ratings of 400W for his calculations. Now we all understand (or should) that the best we should expect from a 400W panel is the NMOP ratings which most likely are around 300W, not 400. Is my assumption correct?

[Mike 134]

Thanks all, the feedback is very helpful. I am still very new to this and did not do much technical homework ahead of time to understand what I should expect out of the production.
I frankly had no idea the amount of power loss that occurs when the the sun is not angled perfectly above the panel. From the feedback, it sounds like the power capture dramatically drops and that my production is not completely out of scope for what I should expect.

OCJ, you are correct that the bottom row of panels on the south side suffer the most from shading. In the morning up until 1 pm, that row gets a decent amount of shade.
I wish my installer explained how the system is strung together as I have no idea. I was not even sure if panel number were supposed to make some sort of sense as it relates to the layout. It certainly makes sense that it should and certain numbers I would think should be grouped together on a string but I am not sure at all how they set this up.

I do believe the bottom row of panels on the south side can be re-arranged and placed a bit higher on the roof to limit shading.
What is strange is that, as I understand it, the level of production is limited to the least common denominator of the string the panels are on. That is not true with your Solar Edge system because you have an optimizer under each solar panel to prevent that. If that is the case, I would expect much less deviation of some panels that are near each other.

What is also strange is that panels 18 and 6 on the upper left facing south produce such low numbers. If those low numbers still exist at least one hour after sunset over several days, then you have a warranty problem. Don't worry about weird numbers during the day, look for a trend over a week or more. Those are on the highest part of the roof and not blocked by shading at all. I would think they would be some of the best producers.

Oregon_phil, you are right, the numbering seems much more logic from Mike 134's setup. I am going to ask more questions of the installer to see if my system is on more than one string. I have no idea why they would put it all on one as that does not seem ideal for my type of roof. Any idea how many strings can be run on a system and what the drawbacks are for running multiple?

You have 2 strings because you're limited to 25? optimizers on one string. Looks like your installer was lazy when he used the Solar edge design site and just put them on one string.
I have 2 strings of 9 each.

[oregon_phil]

The SE 11400 has three string inputs and has a limit of 6000 watts per string. Therefore, I assume the installer used three strings. An as built schematic would be nice.

[SolTex]

Question because I am unfamiliar with PVWatts. The OP appears to have used his advertised panel ratings of 400W for his calculations. Now we all understand (or should) that the best we should expect from a 400W panel is the NMOP ratings which most likely are around 300W, not 400. Is my assumption correct?

If you click on the "Help" button on the PVWatts Calculator page, you will find a detailed explanation of what information is required for each entry in the calculator. Specifically, it wants the STC rated DC output of the panel, which is also typically the "nominal" rating (the advertised rating). The calculator makes adjustments for the difference between the STC rating and what happens in the real world. See below the excerpt from the Help page:

DC System Size

The DC system size is the DC (direct current) power rating of the photovoltaic array in kilowatts (kW) at standard test conditions (STC). PVWatts^® can model any size of array, from small residential rooftop systems to large ground-mounted power generation systems.

[Jriell]

The SE 11400 has three string inputs and has a limit of 6000 watts per string. Therefore, I assume the installer used three strings. An as built schematic would be nice.

thanks for the reference to the inverter specs. I am not sure where the max of 6k per string is stated on those docs? I have seen my system go up to 7.3 kw and even over 9 kw on the first day measuring.

I will try and get more info on the build out.

[foggysail]

If you click on the "Help" button on the PVWatts Calculator page, you will find a detailed explanation of what information is required for each entry in the calculator. Specifically, it wants the STC rated DC output of the panel, which is also typically the "nominal" rating (the advertised rating). The calculator makes adjustments for the difference between the STC rating and what happens in the real world. See below the excerpt from the Help page:

DC System Size

The DC system size is the DC (direct current) power rating of the photovoltaic array in kilowatts (kW) at standard test conditions (STC). PVWatts^® can model any size of array, from small residential rooftop systems to large ground-mounted power generation systems.

Thanks, that helps me a lot. I did not spend much time on it but just enough to get estimates of what I might be lucky enough to get with my 11.2 system. AND THINGS ARE GETTING CLOSSER FOR ME!!! Eversource has approved my installation and I am now waiting for meters. Eversource requires a separate meter to monitor the solar output that feeds back into my load center and from there of course, into the grid.

Anxiety is an understatement for me. I have great expectations to see grid tie within the next week or so although Eversource gives themselves 10 days to do so after approval. More later--

[oregon_phil]

Search for this file on the SolarEdge website

se-P5-series-add-on-power-optimizer-datasheet-na.pdf

to find out power limitations per string for your inverter.

Your attached file shows your system ENERGY for the day, kWh. This is different than the 6000 watts maximum POWER per string.

I have highlighted the power limitation cited in the .pdf file I mentioned above.