What should my Solarglass roof produce...?

[Keoni8888]

Hi all,

I need some advice on whether the output of my systems seems to be performing up to par. It would seem to me that it is not, but I have very little experience in this area and am looking for guidance / help. I recently took the plunge into a Tesla Solar Glass roof (v3) and 2 Powerwalls. In general, I love it. Beautiful roof and keeping me mostly off grid. We maxed out the roof and I was told that the system would rate at 12.98kW. This equates to 222 PV roof tiles. Here are the specs for the tiles:

At 25 degrees C:

Voc (v): 13.34
Vmp (v): 10.99
Isc (A): 5.65
Imp (A): 5.32
Pmax (W): 58.47

Temp coefficient:
Voltage at open circuit: -0.299
Max power: -0.395
Short circuit current: 0.047

Using the Tesla app I observe the following. I live in S. Calif. and summer is kicking in. My roof is not shaded. Tiles are installed on both North and South facing roof angles. At mid-day, and to a couple hours on either side, the sun hits both N and S surfaces and is unobstructed. On hot days (outside temperature of 90+F) I have observed the system producing a maximum instantaneous output of 8.0kW. Hotter days I have seen less. This is at 12:30PM. Total production of about 61kWh per day in hot conditions.

On cool days with clouds that part momentarily to let through the sun, I have observed up to 12kW instantaneous output. But in general on cool days I observe somewhere in the 9.8 to 10.5 kW sustained instantaneous output.

This seems to be a really huge swing. It seems that the temperature coefficient should account for only about a 20% drop. Factor in some general loss in the system itself, and I would think that it should be producing about 9.75kW sustained on a really hot day, at a minimum. But I may be looking at this wrong. That is where I am looking for advice for you.

I understand that certain factors are not optimal, such as the North-facing tiles, but still do not think this accounts for the drop. Also, I have been told that as the tiles are mounted with a relatively small air gap, this impacts productions a lot compared to traditional panels. Tesla quoted 16,444 kWh production per year.

I cannot be 100% certain if they actually put up 222 PV tiles as the roof seems to be too slippery for me to go up and count them

. Any pointers on how to walk on this roof without slipping off?

WHat are others with a Solarglass roof experiencing? Any thoughts, advice, direction would be hugely appreciated!!

John

 

[BrettS]

Given the north facing panels and the fact that the system will not be perfectly efficient I think what your describing is pretty close to what you should expect. If you take that 16,444kWh annual production number and divide it by 365 then that’s only an average of 45kWh per day. Obviously you should expect more production on summer days and less on winter days, but if you’re getting 61kWh on a hot summer day then that certainly sounds in line with your expected annual production numbers.

 

[jboy210]

I see similar lower number on my SolarRoof. We see peaks closer to 8 kW through the roof is rated at 12.6 kW. And like you I see peaks on days that are a little cloudy. Mine will peak at close to 10 kW.

Part of our issue may be trees. We have a 60 foot oak in the back and a row of 35 foot pines to the west. Waiting to see what the roof does after the leaves fall from the oak.

 

[wjgjr]

I see similar lower number on my SolarRoof. We see peaks closer to 8 kW through the roof is rated at 12.6 kW. And like you I see peaks on days that are a little cloudy. Mine will peak at close to 10 kW.

Part of our issue may be trees. We have a 60 foot oak in the back and a row of 35 foot pines to the west. Waiting to see what the roof does after the leaves from from the oak.

Echoing this, we see bigger peaks on cloudy days (cloud edge effect that others have mentioned in this forum). We have also seem somewhat lower peak production than I might have guessed this July. In June, we were more often seeing a sustained midday output of 5.4kW (with an 8.19kW system, exactly half on the south/187 degree face and half on the north face) but are seeing more like 5.0kW. Part of it may be sun angle changing already, but I also think for us the record heat (we set a record for number of 90+ days in a July) has affected performance somewhat both because of the heat and because it has sometimes added just a bit of haze to the sky.

But overall, I believe our performance is in line with - and probably slightly above - what we were promised. The way I am estimating it is to use PVWatts, which provides monthly numbers. Tesla provided me only an annual number, which is about 11% lower than the PVWatts number, so I took all the PVWatts numbers and multiplied by 0.89 to get both a monthly Tesla number and PVWatts number. Even though we only got PTO in mid-July, operating off-grid in June (meaning we didn't capture quite all the possible solar), we exceeded our Tesla number and for July, it looks like we will even exceed the PVWatts number (factoring in that we did not operate the system at all for 5 days.)

For OP, I would suggest looking at PVWatts - it should give you a monthly number and will actually also provide a range for the annual number. Then estimate what those numbers would be assuming the Tesla annual number, which should be lower. Ideally, you would be hitting the PVWatts numbers, but as long as you are exceeding the Tesla-adjusted numbers for the month, it seems like a decent chance things are functioning normally. I don't know what percent of your tiles are north-facing, but with the 50/50 split we have, we are seeing at best about 65% of the rated DC capacity for sustained mid-day production, and it looks like your numbers are above that.

 

[Keoni8888]

Thank you all for sharing. As I said, I really do not know how this works in real life, but some of the research I have done on the temperature coefficient and related calculations led me to believe that the production should not drop as much as I am seeing with the heat. Today, for example, we just hit a peak of 7.8kWh production on the 13kWh system, at 95 degrees and sunny outside. I am definitely in line, and exceeding, what Tesla promised for yearly production and am quite happy with that. It is just peaking my interest about how the system overall seems to be impacted hugely by heat.

I have also been mapping output against heat/weather daily since PTO on May 23, and am seeing a continual drop in total daily production. Some observations are interesting here... In early June (9, 10, 11), on days that were sunny and about 95 degrees outside, the system was peaking at 8.6kWh and producing about 70kWh for the day. Now in July, with a similar heat and sun factor, it is peaking below 8 and producing about 61 kWh for the day. Not a complaint, just trying to figure out what is happening and why (and loving the solar adventure along the way).

I probably should not have done this, but... in a not-too-hot day I sprayed water onto a small section of the roof to see if that made a difference. I'd say that it was an area of about 25% of the total PV up there. The production went up by 0.5 kWh almost immediately. As it dropped back to the former production after the next 20 minutes, I don't think it was the effect of cleaning the tiles, but definitely the heat impacting production.

As a side note, I am also measuring the difference between the TOU plans in SCE in S. California as I am trying to figure out if the 4-9 plan or the Prime plan is better for me. My spreadsheet is doing an interesting job of that and it indicates that for my power profile (no EV car) that the 4 to 9 plan is the best. I am using my Powerwalls on Balanced mode.

 

[jboy210]

Thank you all for sharing. As I said, I really do not know how this works in real life, but some of the research I have done on the temperature coefficient and related calculations led me to believe that the production should not drop as much as I am seeing with the heat. Today, for example, we just hit a peak of 7.8kWh production on the 13kWh system, at 95 degrees and sunny outside. I am definitely in line, and exceeding, what Tesla promised for yearly production and am quite happy with that. It is just peaking my interest about how the system overall seems to be impacted hugely by heat.

I have also been mapping output against heat/weather daily since PTO on May 23, and am seeing a continual drop in total daily production. Some observations are interesting here... In early June (9, 10, 11), on days that were sunny and about 95 degrees outside, the system was peaking at 8.6kWh and producing about 70kWh for the day. Now in July, with a similar heat and sun factor, it is peaking below 8 and producing about 61 kWh for the day. Not a complaint, just trying to figure out what is happening and why (and loving the solar adventure along the way).

I probably should not have done this, but... in a not-too-hot day I sprayed water onto a small section of the roof to see if that made a difference. I'd say that it was an area of about 25% of the total PV up there. The production went up by 0.5 kWh almost immediately. As it dropped back to the former production after the next 20 minutes, I don't think it was the effect of cleaning the tiles, but definitely the heat impacting production.

As a side note, I am also measuring the difference between the TOU plans in SCE in S. California as I am trying to figure out if the 4-9 plan or the Prime plan is better for me. My spreadsheet is doing an interesting job of that and it indicates that for my power profile (no EV car) that the 4 to 9 plan is the best. I am using my Powerwalls on Balanced mode.

We had our solar roof completed in April. They left it on for a few days and we had our highest peak readings in late April. But of course, the solar days (when roof produces) are short, so the net amount of power produced in a day was significantly lower than summer. In July our roof starts production around 6:30AM and continues until around 7 PM when the sun goes behind the hills.

 

[Tk18]

We are in Santa Clara CA, with a V3 glass roof of 8.65kw supposed to produced 10MWh per year. We will not make it...

our installation is
South 40 Tiles 27.0%
East 44 29.7%
West 34 23.0%
North 30 20.3%
Total 148

June was wasted because thread polarity was inverted in several place leading to barely 1.9kw peak on best day of the year (June 21 noon)... still took me a month to get someone to come fix it...
July numbers are better but not what I was expecting either... 6.1 kw peak for 45kwh max per day.... going down fast (we now barely reach 5.5kw peak with 39kwh per day)

we had many issues during the whole project (including zero communication and review of the plan ...) and still do not have any support. So tiles are not maximized on the south side... (I cannot get why) or the East and west.... so I was expecting to not reach as high of a peak than I would have if all tiles were a n the south, but catch up some by getting a fatter bell curve thus getting more energy... but if we barely make 40kwh right now (no pollution, pure sun, relatively cool, Low 80’s) i cannot imagine we will get to the 10MWh which we were told was a lowball number...

Don’t get me wrong: we like the roof Look (although some finishing could easily be improved) and we love to be (for now) off grid and actually putting back a lot on the grid (since we still have only one Model 3 and not commuting), but that won’t be the case if we get another Tesla and commute back to normal...

i wish Tesla could share more concrete data, so we can better understand our system. It is not clear if something is wrong or if this is exactly what the system is supposed to produce.

i wish we had been able to review the project before they went for permits... and if they had listen to what we said we wanted to do in the future (get another electric car, get a charger) and where we wanted batteries and inverters (we wanted in the garage, they decided outside wall)... and we still have no clue who our project advisor is.

i also wonder if cleaning the tiles would help production (would need to do it before sunrise using demineralized water probably...)... because of being the guinea pigs and Covid, it took a good month + to install (plus many long breaks in the middle) they walked so much in the tiles that by the time they were done, there were tons of muddy food marks on the roof ... and through time the dust covered the whole roof which is now more grey than black... still not sure the best way would be to clean this...

 

[Keoni8888]

A couple more observations on this subject... simply to share.

1. I hosed down my roof on a relatively cool day and this seems to have increased overall production by about 0.3kW. My roof was installed in May so there was quite a lot of buildup on it. Need to figure out a way to do a better job at it though. Seems too slippery to go up and clean manually though.

2. It is quite sunny and hot here today (~100*) and my roof maxed out at 7.8kW (for a 13kW installation). It really does seem that heat impacts the PV tiles to a great extent as this is a 40% drop in efficiency. If I were to use the temp coefficient number supplied by Tesla for the tiles, it would mean that they are hitting over 120*C (250*F). I am doubting it is that hot up there.

Overall still quite happy. But what is frustrating is that Tesla does not seem to either know or want to take the time to help me understand the way that their roof works when it comes to heat and production efficiency. They are like a black box.

 

[johnnyS]

Interesting subject, we have had regular solar panels for six years in southern California. Our best month this year was May. About two years ago I mentioned to our neighbor that a queen palm in their yard was shadowing the corner of our array for a couple of hours in the morning. They cut down the palm and our production has been better since then.

I do not have the ability to monitor our system in as much detail as it seem that you can.

 

[Keoni8888]

Interesting subject, we have had regular solar panels for six years in southern California. Our best month this year was May. About two years ago I mentioned to our neighbor that a queen palm in their yard was shadowing the corner of our array for a couple of hours in the morning. They cut down the palm and our production has been better since then.

I do not have the ability to monitor our system in as much detail as it seem that you can.

That Tesla app is pretty interesting on what it shows...

 

[jpizz123]

Does the roof show dust easily? I have only seen pictures of new installs which look amazing, but I always wondered how it would look like a few months down the road.

 

[Keoni8888]

Does the roof show dust easily? I have only seen pictures of new installs which look amazing, but I always wondered how it would look like a few months down the road.

Definitely so... see my photos above. It quickly went from a shiny blackish blue, to a dull gray in about 5 weeks. You can see the difference between where I was able to clean and where I could not. I'll send up the drone for more detail later.

 

[wjgjr]

Definitely so... see my photos above. It quickly went from a shiny blackish blue, to a dull gray in about 5 weeks. You can see the difference between where I was able to clean and where I could not. I'll send up the drone for more detail later.

I suspect a lot of this is location-dependent. Where we are, I feel like there hasn't been any buildup. But we are nowhere near a desert and get relatively frequent rainfall. Plus, we may have a more steeply-sloped roof.

I will agree on the possibility of a heat impact. On some very hot days, we have seen peaks seemingly a bit (maybe 8%) lower than it seems like I would expect. That said, in our area, these hot days also tend to produce more haze, so it is possible for us the effect is more due to that than the heat. Either way, it is something worth watching.

 

[nwdiver]

It really does seem that heat impacts the PV tiles to a great extent as this is a 40% drop in efficiency.

Yep; With solar PV you lose ~0.5%/C. I can EASILY see a solar roof running ~50C hotter than a conventional roof. There's no air flow under the panels. But the purpose of a solar roof isn't to be more efficient than a conventional PV install. It's to look seamless. Complaining about low efficiency in a solar roof is like complaining about the poor mpg of a lambo.

Definitely so... see my photos above. It quickly went from a shiny blackish blue, to a dull gray in about 5 weeks. You can see the difference between where I was able to clean and where I could not. I'll send up the drone for more detail later.

The effect of dirt is a rounding error compared to the effect of heat. The purpose of your roof is to look great and produce a bit of electricity on the side. My advice would be to focus on the first and ignore the second.... you'll be happier for it.....

 

[wjgjr]

The effect of dirt is a rounding error compared to the effect of heat. The purpose of your roof is to look great and produce a bit of electricity on the side. My advice would be to focus on the first and ignore the second.... you'll be happier for it.....

I would only agree with this to the extent the "bit of electricity" is actually what Tesla promised. To my mind, the purpose of the solar roof is to look good and also produce a significant amount of electricity in an integrated product. While it may not match what panels can produce for most roofs, it producing the stated amount of energy is just as important for a solar roof install as a panel install. And I think most who buy the roof only do so because of its dual-use functionality, not just for its looks.

 

[nwdiver]

I would only agree with this to the extent the "bit of electricity" is actually what Tesla promised. To my mind, the purpose of the solar roof is to look good and also produce a significant amount of electricity in an integrated product. While it may not match what panels can produce for most roofs, it producing the stated amount of energy is just as important for a solar roof install as a panel install. And I think most who buy the roof only do so because of its dual-use functionality, not just for its looks.

I'd be interested to see how Tesla is projecting the production from solar roofs. Maybe @Keoni8888 can answer that question. Did they provide a kWh/yr estimate? PV Watts is generally accurate for modeling production but solar roofs isn't part of their model. They even differentiate between roof-mount and ground-mount because ground-mount gets more air flow. I have a roof mount but since it's a flat roof the panels are angled 10deg south which provides more air flow. Even that makes a HUUUGE difference. I get ~10% more kWh/yr, but I will admit it's not nearly as sleek as a solar roof; For me I prefer the 10% increase in production.

With a solar roof you're paying A LOT more to produce A LOT less electricity to have a more attractive roof. It is what it is.

 

[wjgjr]

I'd be interested to see how Tesla is projecting the production from solar roofs. Maybe @Keoni8888 can answer that question. Did they provide a kWh/yr estimate? PV Watts is generally accurate for modeling production but solar roofs isn't part of their model. They even differentiate between roof-mount and ground-mount because ground-mount gets more air flow. I have a roof mount but since it's a flat roof the panels are angled 10deg south which provides more air flow. Even that makes a HUUUGE difference. I get ~10% more kWh/yr, but I will admit it's not nearly as sleek as a solar roof; For me I prefer the 10% increase in production.

With a solar roof you're paying A LOT more to produce A LOT less electricity to have a more attractive roof. It is what it is.

As I do have a solar roof, I can answer that they do provide a kWh/yr estimate, and, at least for me, it was about 12% less than putting the applicable parameters into PV Watts (which does essentially model them as small panels.) For the moment, we are running better than the Tesla numbers, and very close to (maybe 1% below) the PV Watts numbers. Far to early in our system's operation to make any definitive statements, but at least a promising start.

I also very much disagree with the notion that you are paying a lot more to get a lot less, unless you are assuming people are tearing up perfectly good roofs just to install a solar roof. The price comparison should be between a solar roof and a new roof plus solar panels, for homes where a new roof is needed. There are a number of ways to calculate the costs, but for purposes of this discussion, what I would look at is the cost per watt after factoring out the cost of a new roof and also adjusting the ITC appropriately since a shingle install can claim each PV shingle for the full amount of the ITC, which can mean a portion of the roof cost is subsidized.

As I live in a more expensive area, even a basic, 30-year asphalt roof is rather expensive. Based on those estimates, our cost for the solar portion, before the ITC or other incentives (but adjusting for the excess ITC) works out to $2.44 per Watt. While this is more than what Tesla is currently pricing for panels, I would not describe it as a lot more for a lot less (and certainly not with caps.) At the time we were shopping (and even in the past year, a lot has changed), it was essentially comparable to re-roofing plus solar, particularly if you factor in the solar+PW discount we got (but I did not include in my $2.44/W number.)

This math may not work out as well for everybody - I would note that in general, Tesla's pricing is particularly good in expensive areas as they do not charge more for parts or labor, where roofing costs can be much higher. We actually went from an aging slate roof, so the numbers look even better were we to price against slate. And I do agree there is value to the aesthetics of the solar roof, but I would disagree that it is the only reason one would consider a solar roof. Because it really only makes sense when a new roof is needed, the market for the solar roof is certainly much smaller, but in those conditions, the price can be much more competitive than perhaps it seems.

 

[Keoni8888]

Yep; With solar PV you lose ~0.5%/C. I can EASILY see a solar roof running ~50C hotter than a conventional roof. There's no air flow under the panels. But the purpose of a solar roof isn't to be more efficient than a conventional PV install. It's to look seamless. Complaining about low efficiency in a solar roof is like complaining about the poor mpg of a lambo.



The effect of dirt is a rounding error compared to the effect of heat. The purpose of your roof is to look great and produce a bit of electricity on the side. My advice would be to focus on the first and ignore the second.... you'll be happier for it.....

Absolutely respect those statements and you are indeed right - I can sometimes focus on the technical aspects / data rather than the overall result. I am very happy with the roof and have no regrets - it is amazing! It looks great and gives me abundant electricity. However... understanding how it works and if performance is normal, or if there is an undetected issue, is not a bad path. Also - Tesla stated power loss is 0.395%/C.

 

[Keoni8888]

I'd be interested to see how Tesla is projecting the production from solar roofs. Maybe @Keoni8888 can answer that question. Did they provide a kWh/yr estimate? PV Watts is generally accurate for modeling production but solar roofs isn't part of their model. They even differentiate between roof-mount and ground-mount because ground-mount gets more air flow. I have a roof mount but since it's a flat roof the panels are angled 10deg south which provides more air flow. Even that makes a HUUUGE difference. I get ~10% more kWh/yr, but I will admit it's not nearly as sleek as a solar roof; For me I prefer the 10% increase in production.

With a solar roof you're paying A LOT more to produce A LOT less electricity to have a more attractive roof. It is what it is.

Happy to share @nwdiver! Here is my layout and sizing. They changed the layout a bit without telling me, but that basically it minus the small grouping of PV on the left bottom. And the bottom of the layout faces South, top is North. I have only had PTO since May 22 but so far it looks like the roof will produce the quoted 16444 kW/yr or more. But ultimately that will take until next year to figure out. I am positive that the Solarglass cannot have nearly the degree of ventilation as traditional panels. So totally makes sense that it will be hotter than it could otherwise be.

The Tesla app indicates that I have produced 5274 kW so far (in about 3 months). I understand that winter will not produce as much due to cloudiness. But even if the remaining 9 months of my year produce 75% of what I have generated so far, I will still be ahead. And there is a lot of sun where I live.

 

[jpizz123]

Happy to share @nwdiver! Here is my layout and sizing. They changed the layout a bit without telling me, but that basically it minus the small grouping of PV on the left bottom. And the bottom of the layout faces South, top is North. I have only had PTO since May 22 but so far it looks like the roof will produce the quoted 16444 kW/yr or more. But ultimately that will take until next year to figure out. I am positive that the Solarglass cannot have nearly the degree of ventilation as traditional panels. So totally makes sense that it will be hotter than it could otherwise be.

View attachment 576474

Is there a reason why 2 powerwalls would not back up your AC?