Disappointing Power Generation, and I need help with analysis of pypowerwall output please!

[solarpics]

Hi All,

I've been going back and forth with Tesla about the output of my solar array, and thanks to this board I think I may have enough data to go back to them (again) and convince them to send someone out to check and fix it. But first some background.

My panels went in a couple of months ago, and we received PTO on June 12. I am in San Jose (very sunny), and have a 9.6kW installation with one Powerwall Plus. I requested that Tesla send me the schematics so I can see how the circuit is supposed to be connected, but haven't heard back from them after a week. But the installers told me that I have two arrays, and I recall one had only 6 or 8 of the 24 panels.

In any case, on a clear sunny day, peak production has never exceeded 5.3kW, and daily power generation has never exceeded ~45kWh - even on the clearest of sunny days, even on Tuesday last week when it hit 100F without a cloud in the sky!

I suspect there is something wrong with the installation, so I ran pypowerwall to get the string information and what I got is copied below. Here is my rookie analysis:
It looks to me like strings C and D are connected in parallel (based on their voltages being almost identical). And string B is "stand alone", i.e. one series string. But what is going on with "A"? Is this an actual string that is not functioning properly, or is really unconnected? Or could it be an unused input on the inverter that appears in the status as not connected? Is there anything else wrong here that anyone can discern from this data?

Any details anyone can provide (or corrections to my analysis above) would be appreciated, and I can always get photos of the panels if that will help.

Thanks!!!

*********************************

Battery power level: 37%

Grid Power: 1.02kW
Solar Power: 3.45kW
Battery Power: -1.88kW
Home Power: 2.30kW

{
"A": {
"Connected": false,
"Current": 0.0,
"Power": 0.0,
"State": "PV_Active",
"Voltage": -1.299999999999999
},
"B": {
"Connected": true,
"Current": 9.39,
"Power": 1629.0,
"State": "PV_Active",
"Voltage": 175.9
},
"C": {
"Connected": true,
"Current": 3.8000000000000003,
"Power": 792.0,
"State": "PV_Active",
"Voltage": 209.4
},
"D": {
"Connected": true,
"Current": 3.42,
"Power": 709.0,
"State": "PV_Active_Parallel",
"Voltage": 208.8
}
}

 

[RichardCullip]

What size inverter or details on microinverters if used? Array orientation? Tilt on panels?

These have an influence on your solar production and can limit the power developed from your solar panels. In addition, heat will lower the efficiency of your panels. Panels work best on full sunny but cold days

 

[solarpics]

What size inverter or details on microinverters if used? Array orientation? Tilt on panels?

These have an influence on your solar production and can limit the power developed from your solar panels. In addition, heat will lower the efficiency of your panels. Panels work best on full sunny but cold days

The standalone inverter is 3.8kW; the Powerwall Plus inverter is 7.6kW. This is all Tesla equipment (I thought they didn't use microinverters?).

10 panels are north east facing
4 panels are north west facing
5 panels are south east facing
5 panels are south west facing

The only shaded panels would be the 4 north west panels. These are on the garage roof, which is shaded by the second floor in the afternoon. There is no shading from trees or adjacent structures

By eyeballing the roofline, I would guess all panels are 40 degrees from horizontal.

 

[karkrazy]

Hi All,

I've been going back and forth with Tesla about the output of my solar array, and thanks to this board I think I may have enough data to go back to them (again) and convince them to send someone out to check and fix it. But first some background.

My panels went in a couple of months ago, and we received PTO on June 12. I am in San Jose (very sunny), and have a 9.6kW installation with one Powerwall Plus. I requested that Tesla send me the schematics so I can see how the circuit is supposed to be connected, but haven't heard back from them after a week. But the installers told me that I have two arrays, and I recall one had only 6 or 8 of the 24 panels.

In any case, on a clear sunny day, peak production has never exceeded 5.3kW, and daily power generation has never exceeded ~45kWh - even on the clearest of sunny days, even on Tuesday last week when it hit 100F without a cloud in the sky!

I suspect there is something wrong with the installation, so I ran pypowerwall to get the string information and what I got is copied below. Here is my rookie analysis:
It looks to me like strings C and D are connected in parallel (based on their voltages being almost identical). And string B is "stand alone", i.e. one series string. But what is going on with "A"? Is this an actual string that is not functioning properly, or is really unconnected? Or could it be an unused input on the inverter that appears in the status as not connected? Is there anything else wrong here that anyone can discern from this data?

Any details anyone can provide (or corrections to my analysis above) would be appreciated, and I can always get photos of the panels if that will help.

Thanks!!!

*********************************

Battery power level: 37%

Grid Power: 1.02kW
Solar Power: 3.45kW
Battery Power: -1.88kW
Home Power: 2.30kW

{
"A": {
"Connected": false,
"Current": 0.0,
"Power": 0.0,
"State": "PV_Active",
"Voltage": -1.299999999999999
},
"B": {
"Connected": true,
"Current": 9.39,
"Power": 1629.0,
"State": "PV_Active",
"Voltage": 175.9
},
"C": {
"Connected": true,
"Current": 3.8000000000000003,
"Power": 792.0,
"State": "PV_Active",
"Voltage": 209.4
},
"D": {
"Connected": true,
"Current": 3.42,
"Power": 709.0,
"State": "PV_Active_Parallel",
"Voltage": 208.8
}
}

Do you have your engineering diagrams? All that info is part of the permitting process and Tesla should have provided that to you anyway. Didn’t you sign off on the design/proposal before they even filed for permits? If you can share the design, that will let folks provide better input.

Do you have an solaredge inverter or enphase microinventers? With each string, how are the panels oriented?
Without knowing the orientation and your design (panel sizes, inverter, etc), you may already at peak output

As for optimal performance, the Bay Area’s current heat wave is not the best for solar. I believe ambient temperatures ideally should be 60-70 degrees (as poster above said), as rooftop temperatures are another +20 degrees. Basically the weather now is too hot for peak efficiency, but that only loses a few percentage points.

 

[karkrazy]

The standalone inverter is 3.8kW; the Powerwall Plus inverter is 7.6kW. This is all Tesla equipment (I thought they didn't use microinverters?).

10 panels are north east facing
4 panels are north west facing
5 panels are south east facing
5 panels are south west facing

The only shaded panels would be the 4 north west panels. These are on the garage roof, which is shaded by the second floor in the afternoon. There is no shading from trees or adjacent structures

By eyeballing the roofline, I would guess all panels are 40 degrees from horizontal.

Looks like I hit reply when you posted

What size are the panels? How many watts? Looks like 400w panels?

Unless there’s clipping going on, best guess is that you’re already at optimal performance and your system is working as designed. Tesla probably didn’t do a good job of communicating the system and expectations…

 

[solarpics]

Looks like I hit reply when you posted

What size are the panels? How many watts? Looks like 400w panels?

Unless there’s clipping going on, best guess is that you’re already at optimal performance and your system is working as designed. Tesla probably didn’t do a good job of communicating the system and expectations…

Lol, our messages crossed!
Yes, they are indeed 400W panels.
They provided no engineering drawings. I asked customer service for a schematic last week; they said the drawings should be in the document set in my Tesla account, but they are not. The only drawing I have is the panel layout, but I really want to know how they are connected.

Under what situations would clipping occur? Doesn't this occur only when the power output exceeds the inverter capacity (which sounds like a design error)?

Thanks for your reply!

 

[solarpics]

Do you have your engineering diagrams? All that info is part of the permitting process and Tesla should have provided that to you anyway. Didn’t you sign off on the design/proposal before they even filed for permits? If you can share the design, that will let folks provide better input.

Do you have an solaredge inverter or enphase microinventers? With each string, how are the panels oriented?
Without knowing the orientation and your design (panel sizes, inverter, etc), you may already at peak output

As for optimal performance, the Bay Area’s current heat wave is not the best for solar. I believe ambient temperatures ideally should be 60-70 degrees (as poster above said), as rooftop temperatures are another +20 degrees. Basically the weather now is too hot for peak efficiency, but that only loses a few percentage points.

Thanks for the info regarding heat. I've been comparing my output to my co-worker's. He has almost an identical Tesla system, but fewer panels (8.8k total vs my 9.6kW). His panels are not shaded and are spread on all four sides of the house like mine. He's in San Ramon, I'm in south San Jose so I figure temps and overall climates are very similar. A few weeks ago, on sunny, cloudless days he regularly got ~8.4kW peak and ~65kWh production throughout the day. This is significantly higher than mine (50-60% higher). We would get together and compare outputs on the Tesla apps for specific days. I know I shouldn't be comparing the two since they are different installations, but my thinking is the difference should be 10 or 20%, not 50-60%.

Are you in the Bay Area as well?

Thanks!

 

[SageBrush]

NW and NE panels are not useless but their output is going to be poor -- on the order of 80% annual generation compared to SW -- SE. Moreover, these panels will not add to your peak. You end up with generation through most hours of the day, and trade-off peak generation. Use pvwatts for modeling. This is probably your answer.

In any case, on a clear sunny day, peak production has never exceeded 5.3kW

Peak generation is derated by azimuth and inclination. You do not provide enough information to judge 5.3 kW as reasonable or not.

daily power generation has never exceeded ~45kWh

That sounds pretty darned reasonable to me.

even on Tuesday last week when it hit 100F without a cloud in the sky!

Generation drops as temperature increases, on the order of 0.3 - 0.4% above 20C ambient.

---
If you have calculus skills I can explain how to calculate generation per string. Without those skills, you can still come up with a pretty good estimate with HS level trigonometry. Shall I explain ? You will have to also be able to read a solar map.

 

[jjrandorin]

OP could also just go to PVwatts (PVWatts) and put in their information for their solar install, as accurately as possible, and get a reasonable guesstimate of what their home should produce.

All of the "I am comparing with my co-worker / friend" doesnt matter, but I believe OP already stated they knew that.

 

[solarpics]

I used to have calculus skills in college, but that was a few decades ago!

NW and NE panels are not useless but their output is going to be poor. Use pvwatts for modeling. This is probably your answer.


Peak generation is derated by azimuth and inclination. You do not provide enough information to judge 5.3 kW as reasonable or not.


That sounds pretty darned reasonable to me.


Generation drops as temperature increases, on the order of 0.3 - 0.4% above 20C ambient.

---
If you have calculus skills I can explain how to calculate generation per string. Without those skills, you can still come up with a pretty good estimate with HS level trigonometry. Shall I explain ? You will have to also be able to read a solar map.

I did use pvwatts, but I used the same azimuth (default of 180) for the entire array. I think to get an accurate number, I need to do this four times, for each of the four directions the panels face.

Yes, please explain how to calculate energy generation per string. I haven't used my calculus since college a few decades ago, so maybe the trig explanation is better. I've never read a solar map, but I think I can figure that out.

Thanks for the input!

 

[solarpics]

OP could also just go to PVwatts (PVWatts) and put in their information for their solar install, as accurately as possible, and get a reasonable guesstimate of what their home should produce.

All of the "I am comparing with my co-worker / friend" doesnt matter, but I believe OP already stated they knew that.

I tried PVwatts but will do it again with more accurate input to reflect the true azimuth of all the panels (rather than using the default).

Did you take a look at the string information I printed in the original post, does it look normal? I am really curious what string A's status means when it says "Connected: false". Does this mean there is no string there, and it's an unused input?

Thanks for your feedback!

 

[SageBrush]

I did use pvwatts, but I used the same azimuth (default of 180) for the entire array. I think to get an accurate number, I need to do this four times, for each of the four directions the panels face.

Yes -- PVwatts * 4

so maybe the trig explanation is better

Check one string at a time, when the string azimuth is exactly facing the sun. Then you just have to correct for inclination, which is a right triangle problem. This all presumes no shading, and you do have to correct for temperature as I mentioned above. This exercise is a calculation of power. For energy you would have to review Trig Integration calculus. The sun follows a sinusoidal curve.

Please re-read my last post -- I added an addendum.

 

[jjrandorin]

I tried PVwatts but will do it again with more accurate input to reflect the true azimuth of all the panels (rather than using the default).

Did you take a look at the string information I printed in the original post, does it look normal? I am really curious what string A's status means when it says "Connected: false". Does this mean there is no string there, and it's an unused input?

Thanks for your feedback!

I am not technical enough to advise you on strings, etc (although there are several users here who are). They would also likely want to see the wiring diagram. I do know that you are correct, and tesla does not use micro inverters in any installs, though as you mentioned.

 

[solarpics]

Yes -- PVwatts * 4



Check one string at a time, when the string azimuth is exactly facing the sun. Then you just have to correct for inclination, which is a right triangle problem. This all presumes no shading, and you do have to correct for temperature as I mentioned above. This exercise is a calculation of power. For energy you would have to review Trig Integration calculus. The sun follows a sinusoidal curve.

Please re-read my last post -- I added an addendum.

Will do - I will start with PVwatts *4 before I dive into trig/calculus. Thanks!

 

[solarpics]

I am not technical enough to advise you on strings, etc (although there are several users here who are). They would also likely want to see the wiring diagram. I do know that you are correct, and tesla does not use micro inverters in any installs, though as you mentioned.

I would also like to see the wiring diagram - I'll ping Tesla again for it. Thanks!

 

[SageBrush]

I do know that you are correct, and tesla does not use micro inverters in any installs

The tell-tale sign of microinverters is that each module, string, and trunk is 240V

 

[SageBrush]

Will do - I will start with PVwatts *4

Regarding PVwatts,

It is really useful. Note a couple of things:

1. You can set the STC watts -- just multiply the number of panels in the string * 400 watts/panel (in your case.)
2. Default PVwatts shows average generation each month for your location/azimuth/inclination but you can change the setting to get *by the hour*. It is a lot of data, so have your spreadsheet skills handy.

 

[SageBrush]

Basically the weather now is too hot for peak efficiency, but that only loses a few percentage points.

More like 6 - 10% on a hot day for panels with good ventilation. I'm not sure Tesla panels hugging the roof fit 'good ventilation'.

 

[aesculus]

10 panels are north east facing
4 panels are north west facing
5 panels are south east facing
5 panels are south west facing

FWIW I have a 10.5 kW system and all panels are basically S with 2.5 kW NW. 45 kWh is about as good as it gets. Last June I did have 3 days that got 48, but that was rare.

While the NW panels do better obviously during the summer months, they heat just about negates their input. My peak production is around 6 kW and I don't think I have every seen it at above 7, and that is not often. I have three inverters so I have no clipping.

So if you have clipping then I suspect they could move some panels around at bit to other inverters, but it kind of sounds like with your orientation, you are getting what you should get.

 

[solarpics]

FWIW I have a 10.5 kW system and all panels are basically S with 2.5 kW NW. 45 kWh is about as good as it gets. Last June I did have 3 days that got 48, but that was rare.

While the NW panels do better obviously during the summer months, they heat just about negates their input. My peak production is around 6 kW and I don't think I have every seen it at above 7, and that is not often. I have three inverters so I have no clipping.

So if you have clipping then I suspect they could move some panels around at bit to other inverters, but it kind of sounds like with your orientation, you are getting what you should get.

Great input, thanks so much. Are you in the Bay Area part of Northern California?