Solar Roof

Starting March 15th, Tesla Forums will become read only. To continue the conversation with the Tesla community visit

12.8 kWh Tesla Solar Glass roof only outputs max of 7.9kwh

It's been about 7 weeks since the completion of our 12.8 kWh roof and (3) Powerwalls. We got the OK to power within a week. The Max kWh the roof will put out is between 7-7.9kWh . I've been watch carefully and reviewing the daily Performance on the app. We've had plenty of cloud free days and no haze. What's concerning is that when it peaks at 7-7.9 it starts to immediately decline to about 5kWh. It won't hold at higher rates for more than 10-15 min. Very easy to see the declines on the app. The daily result is only about 28% Self Powered. The (3) Powerwalls never charge to more than about 20% each day and that get used fairly quickly. Would appreciate any thoughts...

I have been on the phone with Tesla and they also agree that something is off as they can read it remotely. My first guess is dip switches or general wiring issue oil the inverters.


  • Best advice i could give is that you are handling it properly, and Tesla needs to come out and inspect what might be going wrong. Seems like a large portion of tiles are not working. But i also do not know what your layout looks like in relation to the sun.
  • Temperature is a real factor for me (not solar roof - but the physics is the same). My best output is usually in June with the longest days, full sunshine and moderate temperatures (generally along with a gentle cooling breeze). As I progress in to true summer and temps exceeding 100 degrees (had a temp of 111 last week) - the panels lose a surprising amount of efficiency when they get really hot. Have not attempted to measure surface temperature, but have no doubt the black glass (SunPower) easily exceeds 140. In those conditions, I lose as much as 20% off the top. So my output in heat is not unlike what you describe...hit the peak production at or about 1pm and start ramping down even though the sun exposure does not fall off.
    Record your temps and see if there is any correlation...
  • This is the sort of real-world experience that has me holding off of pulling the trigger. I am quite surprised that there is so little content/discussion on this channel of the forum. Is there someplace else that is the "go-to" for gaining insight into the real world results of the solar roof? I am concerned because (1) in Houston - very hot for a good part of the year (2) my roof is 8/12 pitch and I have heard that steep roofs are not ideal for solar panel efficiency and (3) Tesla is hiring and training installers . . . that is scary because normally I avoid tradespeople that do not have 5+ years experience in their field . . . afraid that I will have a well-designed system that is installed by an amateur, (4) it is nice that the tiles are guaranteed for a long period, but cannot find any data (fine-print) of what the labor to make a repair will cost or if one failed tile in the middle can even be replaced without removing and reinstalling half of the roof, and (5) can these roofs even be insured? So far the insurance companies that I have contacted are inclined not to insure the Tesla Roof. And unfortunately, the folks manning the sales lines know so little that speaking with them is not worth the time. Is it generally an accurate statement that you must pay the $100 in order to be able to engage with someone that knows things?
  • Interesting article attached regarding the issue of heat. In this case - look at the operating temperature of the micro-inverter. The max temp is easily exceeded on the back side of a panel on a hot day. This is one of the reasons I prefer central inverters. Then the only heat issue is the panel itself and the easily predictable decline in output.
  • I have found that on string wired systems, all of the 19 active tiles, needed to fully power the string grid tied inverter, must be facing the same direction to work properly. My lay out came with both South and East panels series together and that is a miss match and will not put out the full value of watts ever. To cover the house with more active panels, the sales people will just pile on panels and get you a bigger tax credit but not give you the full paid for output, even facing them North, where the sun's angle is to far down that a 58 watt tile panel will only put out 20 to 30 watts max during the summer and nothing in the Winter. They also tried to sell me 5,000 watts facing North but having 7,000 watts of existing off grid solar, I knew it was a bad investment so I am laying out exactly the surfaces I want and paying only for what will work facing South by South West and East by South East. Inactive tiles do not get a tax credit but if the active panels only put out 25% of their rated value, it still is not a good deal for the home owner to load up on active tiles and inverters. larger systems also need to have electrical panel upgrades over 8,000watts and that is also an extra cost if you do not have the south facing roof space to make it pay off.
  • The balanced output strings headed towards the inverter as characterized by @BetazedEd is true of older inverters. Several years ago, inverters (especially for larger installations) allowed unbalanced strings - it will depend on the inverter. Mine conform to the balanced string requirement and all of the panels are on the same exact plane. My son had a 10kW system installed with panels in three different orientations on his roof - single inverter and clearly not balanced - total panel rating similar to mine and his output matches mine pretty closely. It depends on the inverter...
    My examples are for panels - not tiles - but the principle holds - depends on the inverter.
    All of that assumes that each individual string has a single orientation. If the installer managed to tie individual panels or tiles with different orientations into the same string - he/she has fallen off the roof, landing on their head waaaay too many times. Clearly a case of brain damage.
  • Yeah, depends on the inverter. The problem with a string inverter that can isolate multiple strings from each other (allowing different panel orientations), is that the isolation circuits add some losses to the system, effecting efficiency. Micro-inverters solve the problem, but add *many* more possibilities for potential failures. One inverter, versus twenty (or more), with twice as many connections, to boot.
  • For my micro-inverter string, one panel was under-performing, so they came out to replace it under warranty. Probably the inverter. Odd thing is that it required replacement of all 10 micro-inverters (3kW system) for some obscure problem that was never satisfactorily explained.
    Not sure if that is normal or not.
    While peeling up all the panels, one numbskull in the crew was setting them on the pointy end of the racks (against the back). They were also on a low pitch asphalt roof and decided safety harnesses were required. Another idiot in the crew kept dragging his safety gear over the panels. Not sure which idiot to blame - but they managed to shatter the cover glass on one panel in the process. They weren't going to tell me. I went up to inspect prior to them leaving and discovered it. They then started to say it was broken when they arrived (I had washed them the prior day - and inspected that morning - with pics). We had a very short and intense "discussion". They brought back a different crew of five people (including two supervisors) the next week to replace the broken panel. Always something.
    This was *NOT* Tesla...SunPower.
Sign In or Register to comment.