My Tesla Solar Roof is underperforming by at least 20%

[jimm01]

Sorry to see all of the issues that you documented... it's now time to lawyer up!

I've about reached that point—I just filed a complaint with SEIA. To make matters worse the roof is underperforming by at least 20%. I put in a service request and Tesla said they will do nothing since, in a nutshell, my usage is too low and we've basically had nice weather. The roof was generating 102%. The thing is I’ve yet to turn on my air conditioning and unlike many because of the pandemic I’ve been using less electricity not more. I was told by the Tesla representative he cannot justify a service call, he told me: “I cannot ‘in good conscience’ put in the request”. In good conscience? My guess is Tesla knows it’s a design issues and want to touch, it because underperformance is sure sign of service and could be a safety issue.

 

[h2ofun]

I've about reached that point—I just filed a complaint with SEIA. To make matters worse the roof is underperforming by at least 20%. I put in a service request and Tesla said they will do nothing since, in a nutshell, my usage is too low and we've basically had nice weather. The roof was generating 102%. The thing is I’ve yet to turn on my air conditioning and unlike many because of the pandemic I’ve been using less electricity not more. I was told by the Tesla representative he cannot justify a service call, he told me: “I cannot ‘in good conscience’ put in the request”. In good conscience? My guess is Tesla knows it’s a design issues and want to touch, it because underperformance is sure sign of service and could be a safety issue.

What is your data point to prove under at least 20%? I never ever have gotten close to my panel specs, and never will. I just think too many folks do not understand that solar does not give as much as folks think, no matter what any program says. I have noticed that as it has gotten warmer, as an example, even though I have more sun, and longer, my peak output has dropped, just like folks who understand solar have told me.

 

[MorrisonHiker]

What is your data point to prove under at least 20%? I never ever have gotten close to my panel specs, and never will. I just think too many folks do not understand that solar does not give as much as folks think, no matter what any program says. I have noticed that as it has gotten warmer, as an example, even though I have more sun, and longer, my peak output has dropped, just like folks who understand solar have told me.

Our 20.5 kW system is regularly producing at up to 22.3 kW thanks to our cool spring and the cloud edge effect. On cloudless days, they are currently peaking at around 18.5 kW but should get closer to 20.5 kW over the coming weeks.

 

[h2ofun]

Our 20.5 kW system is regularly producing at up to 22.3 kW thanks to our cool spring and the cloud edge effect. On cloudless days, they are currently peaking at around 18.5 kW but should get closer to 20.5 kW over the coming weeks.

You might be an exception.

 

[jimm01]

What is your data point to prove under at least 20%? I never ever have gotten close to my panel specs, and never will. I just think too many folks do not understand that solar does not give as much as folks think, no matter what any program says. I have noticed that as it has gotten warmer, as an example, even though I have more sun, and longer, my peak output has dropped, just like folks who understand solar have told me.

On a sunny day the data gets within 75% and is clipping at 8.3kW by 1100 hours; starts losing efficiency at about 1200 hours due to heat; and steadily rolls off until it drops off significantly after 1330 hours. Never once in any 5 minute interval has it ever reached more than 8.3kW. It’s a 10.6kW system 80% is 8.48. I’m not expecting two digit resolution, but come on.

 

[h2ofun]

On a sunny day the data gets within 75% and is clipping at 8.3kW by 1100 hours; starts losing efficiency at about 1200 hours due to heat; and steadily rolls off until it drops off significantly after 1330 hours. Never once in any 5 minute interval has it ever reached more than 8.3kW. It’s a 10.6kW system 80% is 8.48. I’m not expecting two digit resolution, but come on.

i have 14.5k. most i ever saw like twice was 11.4 clip. now with 2 inverters and 15k. most i have seen is 12 twice with no clipping. but my panels are not perfect directions. oh well when this first happened i was emotional. now i understand its normal and will get worse over time

 

[jimm01]

You might be an exception.

MorrisonHiker and everyone else I know.

 

[wwhitney]

I would think it depends alot on whether all the panels are on a single (equatorial facing) roof plane. If they are, then at some point during the year, the sun should be close enough to perpendicular to the panels to fully illuminate them and yield close to the STC rating (adjusted down for temperature).

But if you have multiple roof angles with PV, then they are never going to all be fully illuminated at the same time. So their power output will never match the simple sum of their DC rating. The inverter sizing should reflect this.

When it comes to evaluating system performance, the benchmark should be the annual energy production estimate provided, not whether the daily peak power matches the nominal DC size of the array.

Cheers, Wayne

 

[jimm01]

I would think it depends alot on whether all the panels are on a single (equatorial facing) roof plane. If they are, then at some point during the year, the sun should be close enough to perpendicular to the panels to fully illuminate them and yield close to the STC rating (adjusted down for temperature).

But if you have multiple roof angles with PV, then they are never going to all be fully illuminated at the same time. So their power output will never match the simple sum of their DC rating. The inverter sizing should reflect this.

When it comes to evaluating system performance, the benchmark should be the annual energy production estimate provided, not whether the daily peak power matches the nominal DC size of the array.

Cheers, Wayne

When a solar system is designed a model is used to determine where and how many tiles to install. Originally my roof was sized at 11.3kW but was downsized by the local electricity distributor. If Tesla needed to install more tiles to achieve 10.6kW there's plenty of room for more tiles. However, there are other factors that could cause underperformance like a faulty rapid shutdown device or inverter—but I think that would report an error, hopefully. In any case, if not a PV issues than whatever the cause is it could cause further damage or even a fire. By the way, there’s no reason to accept an underperforming solar roof—I paid, or at sometime will pay, for 10.6kW—I should get that.

 

[wwhitney]

By the way, there’s no reason to accept an underperforming solar roof—I paid, or at sometime will pay, for 10.6kW—I should get that.

That sounds like a marketing misunderstanding--what it makes sense to pay for is a certain number of kWh/year, not a certain peak power output. Unless you designed your system around the ability to operate certain loads while off grid during peak insolation, the peak power output is immaterial.

Does your roof have active tiles on multiple planes? And what is the AC size of the inverter(s)?

Cheers, Wayne

 

[wjgjr]

When a solar system is designed a model is used to determine where and how many tiles to install. Originally my roof was sized at 11.3kW but was downsized by the local electricity distributor. If Tesla needed to install more tiles to achieve 10.6kW there's plenty of room for more tiles. However, there are other factors that could cause underperformance like a faulty rapid shutdown device or inverter—but I think that would report an error, hopefully. In any case, if not a PV issues than whatever the cause is it could cause further damage or even a fire. By the way, there’s no reason to accept an underperforming solar roof—I paid, or at sometime will pay, for 10.6kW—I should get that.

If you paid for a 10.6 kW DC solar roof, then you should have received a roof with about 180 active tiles on it. If so, you got what you paid for. Nowhere does Tesla state, nor is it expected, that you will see continuous power output of 10.6 kW AC from the system. As @wwhitney noted, there are a number of variables, including orientation, that matter.

As a specific example, our solar roof is ~8.2 kW DC system (140 tiles.) It is half north-facing and half south-facing. 9.5 months after PTO, ~5.5 kW AC is the peak power I have seen on the system, excluding spikes due to cloud edge effects. And yet, the numbers suggest the system will meet the estimated annual output number Tesla provided - currently trending towards 3% above. Certainly if we only had solar on our south-facing roof, we would get closer to the nameplate rating, but even then, we would not hit it. Nothing wrong with that as long as we are getting the expected annual production.

I would suggest PVWatts as a place to get an estimate of your monthly production to get a better sense of whether your system is really under-performing.

 

[holeydonut]

My SEIA agreement has an estimated annual solar output... but, it also has a box checked where the installer has no requirement to guarantee that production. Without knowing the specifics of your contract with Tesla, it's hard to tell if you indemnified Tesla for your production figures when you signed the contract.

It's kind of a catch 22 circular logic for someone who is trying to finance their own solar without doing a lease or power purchase agreement (PPA).

If a customer does a PPA, then the installer has to make sure the system generates a certain minimum power each year the system operates, or else the installer has to pay the difference. This is because the homeowner needs that production to make the system economically viable.

However, with straight up self-financed purchases, the system's ability to perform is not guaranteed. So a owner needs to identify when a component clearly fails and recoup on the back end as a warranty item. But the pressure is now on you to prove something failed.

Hopefully Tesla can sit down with you and show you how that annual solar output (that is hopefully on your contract) is broken down by month. Then they can show you how your previous few months production are on track with that annual forecast.

If Tesla refuses to help you understand your system production, you can try to punch in your system specifics into that PVWatts calculator that wjgjr linked. You'll need to know the exact pitch/azimuth of each of your roof panes though. I'm not sure if Tesla puts this data in their design proposal, but hopefully your build permit has the roof slope and shingle information so you can accurately model what each of your roof facets should be producing. PVWatts can then provide you with a report that explains what production you would expect on an "average" basis with "average irradiance".

 

[h2ofun]

My SEIA agreement has an estimated annual solar output... but, it also has a box checked where the installer has no requirement to guarantee that production. Without knowing the specifics of your contract with Tesla, it's hard to tell if you indemnified Tesla for your production figures when you signed the contract.

It's kind of a catch 22 circular logic for someone who is trying to finance their own solar without doing a lease or power purchase agreement (PPA).

If a customer does a PPA, then the installer has to make sure the system generates a certain minimum power each year the system operates, or else the installer has to pay the difference. This is because the homeowner needs that production to make the system economically viable.

However, with straight up self-financed purchases, the system's ability to perform is not guaranteed. So a owner needs to identify when a component clearly fails and recoup on the back end as a warranty item. But the pressure is now on you to prove something failed.

Hopefully Tesla can sit down with you and show you how that annual solar output (that is hopefully on your contract) is broken down by month. Then they can show you how your previous few months production are on track with that annual forecast.

If Tesla refuses to help you understand your system production, you can try to punch in your system specifics into that PVWatts calculator that wjgjr linked. You'll need to know the exact pitch/azimuth of each of your roof panes though. I'm not sure if Tesla puts this data in their design proposal, but hopefully your build permit has the roof slope and shingle information so you can accurately model what each of your roof facets should be producing. PVWatts can then provide you with a report that explains what production you would expect on an "average" basis with "average irradiance".

The more and more I read these folks with Tesla stuff and how unhappy they are, I just say I am SO happy I went with my installer, even though they were more money! They have allowed me to be involved in every step of the process, which most, like Tesla or I hear even sunrun do not. I can call a guy daily and he picks up the phone, and deals with any questions. So again, one gets what they pay for. I wanted better service, and was willing to pay for it!!!

 

[holeydonut]

The more and more I read these folks with Tesla stuff and how unhappy they are, I just say I am SO happy I went with my installer, even though they were more money! They have allowed me to be involved in every step of the process, which most, like Tesla or I hear even sunrun do not. I can call a guy daily and he picks up the phone, and deals with any questions. So again, one gets what they pay for. I wanted better service, and was willing to pay for it!!!

Yeah, I agree it seems the people who pay a premium to get a local shop to do their install typically have a better experience (like I hate your guts and how easy it was for you to get your system and PTO).

But it's kind of like rubbing salt in jimm01 since Tesla was the only option for Tesla Solar Shingles and he can't really un-do a previous decision. Hopefully we can help jimm01 to the best he can expect now that he is dealing with Tesla and their world-class* service.




* C's get degrees too

 

[Laketime]

By the way, there’s no reason to accept an underperforming solar roof—I paid, or at sometime will pay, for 10.6kW—I should get that.

That's not the way solar works. You still haven't answered the most basic questions- does your roof face due south and are all tiles facing that same direction? I'm at a 20 degree axis off due south and 2/3rds of my panels face 20 off south and the other 1/3rd face due west (22 degree roof pitch). For my 16.32kW system I have maxed out at 14.4kW and that's for a brief time at that. I'm still very impressed with total output. YTD I'm at 112% production/consumption (granted peak usage is coming up). That's much better than the 64% Tesla estimated.

 

[h2ofun]

Yeah, I agree it seems the people who pay a premium to get a local shop to do their install typically have a better experience (like I hate your guts and how easy it was for you to get your system and PTO).

But it's kind of like rubbing salt in jimm01 since Tesla was the only option for Tesla Solar Shingles and he can't really un-do a previous decision. Hopefully we can help jimm01 to the best he can expect now that he is dealing with Tesla and their world-class* service.




* C's get degrees too

It just gets old hearing folks blame Tesla for everything under the moon, but they knew these issues and went with low price anyways.

No one ever, basically is going to get real time max panel output! I never have and never will. So to hear someone say but I am some how not getting
what they paid for, well, ....

I guess my better looks helped my easy pass to fast PTO. And I have no insurance stuff on my policy

 

[wjgjr]

It just gets old hearing folks blame Tesla for everything under the moon, but they knew these issues and went with low price anyways.

Who knew what issues, and when? There are certainly some who came to these forums or other sites and caught wind of some of the Tesla service issues, but the majority of people going for solar are likely not doing that level of research. Tesla is a big name and it is offering great technology for a great price. Most people probably consider that sufficient.

Additionally, this board is not a representative subset of Tesla (or non-Tesla) solar owners. It is almost certainly over-represented by people who are very interested in the technology and by people who have run into problems with their install and are looking to others for assistance. As to how overall customer satisfaction is for Tesla compared to others, I do not know, but it is worth recognizing that there exists a substantial number of people who contracted with Tesla for solar, got an acceptable install, and just moved on with their lives.

So, this forum will have an outsized number of people with issues with Tesla, that they likely (reasonably) would not anticipate, even if we have heard it before. And it will have detailed discussions of minutiae like how to manage loads to squeeze every penny out of the system that the average customer likely just doesn't bother with. I don't see anything wrong with either type of post, and it should be expected on this or any similar type of product forum.

 

[wjgjr]

My SEIA agreement has an estimated annual solar output... but, it also has a box checked where the installer has no requirement to guarantee that production. Without knowing the specifics of your contract with Tesla, it's hard to tell if you indemnified Tesla for your production figures when you signed the contract.

If my contract is typical, all Tesla guarantees is the "power output capacity", which, for solar roof, will be 95% of rating after 5 years, and declining no more than 0.5% annually after. So you are correct that annual output is not guaranteed, and the customer would probably not have a legal remedy for shortages due to component failures that temporarily knock the system partially or completely off-line.

But, the number is certainly a good indication of what to expect and likely could be used to demonstrate liability in the event Tesla failed to promptly address a reported issue (perhaps asserting it wasn't a problem.) I do think Tesla tends to be pretty conservative with that number since we haven't seen waves of posts from people complaining about not getting close to their annual numbers. More often it is discussions like this regarding newer systems where it is a matter of properly setting expectations around peak power as well as expectations for time of year.

 

[jimm01]

That's not the way solar works. You still haven't answered the most basic questions- does your roof face due south and are all tiles facing that same direction? I'm at a 20 degree axis off due south and 2/3rds of my panels face 20 off south and the other 1/3rd face due west (22 degree roof pitch). For my 16.32kW system I have maxed out at 14.4kW and that's for a brief time at that. I'm still very impressed with total output. YTD I'm at 112% production/consumption (granted peak usage is coming up). That's much better than the 64% Tesla estimated.

There are 229 PV modules all facing 138 SE, There are basically 3 levels, 2 on the house and 1 on the garage. All roofs are 36 degrees. The garage has 141 tiles and the house 88. One thing that concerned me is even this time of year the garage starts getting shade from the house before noon. On the house, the roof with 42 tiles slightly shades the roof with 46 tiles in the am. There are no obstructions and the home is at 39.37° N Latitude.

 

[holeydonut]

If my contract is typical, all Tesla guarantees is the "power output capacity", which, for solar roof, will be 95% of rating after 5 years, and declining no more than 0.5% annually after. So you are correct that annual output is not guaranteed, and the customer would probably not have a legal remedy for shortages due to component failures that temporarily knock the system partially or completely off-line.

But, the number is certainly a good indication of what to expect and likely could be used to demonstrate liability in the event Tesla failed to promptly address a reported issue (perhaps asserting it wasn't a problem.) I do think Tesla tends to be pretty conservative with that number since we haven't seen waves of posts from people complaining about not getting close to their annual numbers. More often it is discussions like this regarding newer systems where it is a matter of properly setting expectations around peak power as well as expectations for time of year.

Yeah, but I think jimm01 has a solar roof, so his issue is probably compounded by the fact he had only one company to choose from and I'm not really sure if the solar roof has the same conservatism baked in that you may expect from a normal PV array.

Tesla could have easily over-promised (or poorly communicated their marketing/sales) to make homeowners think their solar roof would be comparable to a PV array... but people aren't seeing it. I remember @jboy210 having some issues understanding the production on his solar roof, and he's like a tech/whiz.

If someone with above average interest and knowledge has difficulty figuring out if the system is behaving as normal, I would expect the average homeowner to also have challenges deciphering the issue.

One would assume that a company selling a very innovative product would set aside separate well trained people to provide advanced education for their early adopters. But we know Tesla doesn't work that way.