I wonder where they're getting their average price from, because I've never heard of a new roof and solar panel installation being that expensive.
EDIT: They're comparing it to concrete roof tiles, which a quick search say that they last for ~50 years, twice as long as the warranty for the Tesla solar roof.
Definitely, the solar price in that article is twice what I paid for my 9.4KW system. I wouldn't have gotten Tesla anyway, even if it was cheaper. This is not the sort of place I'm willing to be an early adopter - once they have proven they have decades of experience, like SMA and others, then maybe. Until then, no way.
So, should we expect the Tesla roof, which has historically had so many issues that they’ve done very few installations, to suddenly be able to last 50 years?
What issues? They did very few installations of previous iterations because they took a long time to install, thus were expensive. Doesn't mean there were any quality issues
The author of this article, Linette Lopez, is a personal friend of Jim Chanos, the biggest TSLA short as far as we know. She publicly recognized this friendship and she does not care whether it makes her biased against the company. It's up to you to trust her reporting regarding these accidents and how representative her coverage is. Note that she only seeks negative feedback about Tesla products, she discard positive and neutral ones. That's her filter, an interesting one.
> And we’re actually going to have […] ‘installathons,’ ” Musk said, which will pit two teams against each other to see who can roof one of two similar-sized/designed roofs faster. Musk reiterated later that there’s “quite a bit of R&D just in the installation process itself.”
Not a short seller, but that sounds a lot like: “We don’t yet know how to do this well, but in the meantime we are going to optimize for speed.” This may look good to investors, but homeowners might cringe. Did Musk learn nothing from the pre-Titan installs? Somebody get him a copy of _The Mythical Man-Month_. This quote alone took me from possible early-adopter to wanting to wait 3 to 5 years until they decide to optimize installation for correctness and lowered risk of ignition.
Have you ever seen roofers work? They move like there's a storm coming all day. Total hustle, it's why existing roof installs are relatively cheap. It makes sense that they need to ptiomize for near same amount of time.
Conventional roofs do not have live power and electronic components.
Solar tiles (and panels) add "must not self-ignite" to the usual
install requirements like "must not leak for N years".
Good, fast, cheap - between good and fast, I want good first.
Cheap remains to be seen.
Luckily, you fall in the textbook mainstream part of the adoption "S" curve. I think it's expected by any early adopter that you're basically testing beta products, I know technology taught me that in all regards.
I say luckily because it's a good thing that most people do not jump too quick on all bandwago... oh wait. Well, it's great when they don't, not too fast. Society isn't a startup.
Speed is of course great. But you still have to engineer and execute properly. I was hoping HN knew where he was getting his urgency from other than the obvious. Like is he short on money.
I don't want to turn this into a debate about Tesla's financials, but I think just about everybody can agree Tesla has always operated in a risky position - highly leveraged, relatively low on cash (in terms of months of runway), extremely risky bets (pre-orders on FSD which has yet to come to fruition).
Love or hate Tesla, they are a financial engineering marvel. And it's easy to make an argument that the company needs more cash, but again I don't want to make this into a flame war.
I suspect, from what I've read of his early life, that he is a flow addict and runs his companies to force employees into flow as well. 'Impossible' deadlines are an easy way to hack a team's flow, see e.g. the games industry.
> Tesla’s Solar Roof website now includes a pricing estimator, which lists $42,500 as the total price for the average 2,000 square-foot home, with 10kW solar panels. It also lists $33,950 as the price after an $8,550 federal tax incentive. You can also enter your address and get an updated estimate that takes into account local costs and incentives, and add on any Powerwalls (with three as the default for a 2,000 square-foot roof).
$42k does not sound like the solar roof is cheaper than a traditional roof. What kind of roof are they comparing it to? A mini-mansion with terra-cotta shingles or asphalt shingles?
A quick google search and home advisor suggests it’s about half the price.
I find that really hard to believe. A new roof runs between 5-10k, and a 3kw solar system can be had for ~5k + labor. Even if they are including a power wall in that price it's still pretty steep.
Edit: It looks like they are actually quoting a 10kw system with 3 power walls (~40kwH storage) included, which at that point sounds about right at $40k. Still far beyond the reach of anyone but the wealthy, but not necessarily a bad deal if you're in the market.
If you shop around it's abundantly clear that the price of a new roof varies wildly. Some of it is materials. The guy offering 50 year warranty GAF super-premium install who spots loads of potential problems up front and factors them into the cost is going to run much higher. The guy who is just going to hire some rando non-english speaking guys from the 7-11 and toss on the cheapest possible materials without concern about the underlying structure is going to be like a 1/4 of that price.
$5-10k on a median sized single family home is the latter, especially if you're closer to the $5k mark. It is likely that you will have problems with the roof after it is installed, and the warranty is going to be garbage.
> The guy who is just going to hire some rando non-english speaking guys from the 7-11 and toss on the cheapest possible materials without concern about the underlying structure is going to be like a 1/4 of that price.
You’ve met my guy Serge! He’s actually quite good and I’ve used him on two different houses. He probably has twice the experience of the super premium guy in town.
It’s mostly nailing rocky tar sheets to wood, the guy who has done it a billion times (literally?) sometimes might be the better choice over the guy who bought into an exclusive contract with super premium company and went to their special school for the special product.
They might be good, they might not. It's a luck of the draw when you are hiring the guys just standing on the street corner the morning of the install.
Contractors that have their own dedicated install crews cost more, but are much more reliable.
Rolling my eyes here, bud. You’re missing the forest for the non-PC tree.
I met someone who could have been described as a “non-English-speaking roofer.” (Not Latino, fwiw.) He says he was essentially enslaved by the guy running the contracting company, and he had no recourse as no one spoke his language.
I think a case could be made in some situations that if your employees don't speak English and you don't speak whatever language they speak, you may get poorer construction outcomes. Even if the employees are experiences and smart.
My last house was a 1400sqft cape. Removal of 2 layers of shingles, building a cricket next to the chimney, new lead flashing around the chimney was $7k. I got quote from multiple contractors for roughly the same price. It really depends on the amount of roof.
Assuming you'll be in the same house for 30 years is a big gamble. What if you need to move in 5 or 10 years? You're house value doesn't go up $40k when you put $40k tesla roof on top.
I would bet a tesla roof actually reduces your house value. Imagine the cost of any roof maintenance, or trying to find a roofer who can work on it without just doing more damage.
It's hard, but not impossible. Your costs are higher than a traditional power company for 10 or 15 years until the panels are paid off, but then your costs are much lower than people with traditional grid hookup.
In terms of investment it doesn't beat the stock market, but you have the side benefit of killing the planet slightly less. You are also hedged against rate increases from your power company.
Labor is a significant part of the cost, both for a roof and for a solar installation. I assume only doing it once is one of the gains in cost-efficiency.
I just put in a 12kw system. It was $32k before federal rebate, and it was the cheapest bid by 5 different contractors. Not sure where your getting $15k
FWIW. The panels are $0.86/wh each with a micro-inverter that runs ~$160 per panel and 32 panels. So, ~$5000 for inverters, ~$11k for panels, which brings you to ~$16k and your not even tied into the power of the house, or adding the mandatory shutoff, or the new meters, or the Emphase system that tracks by the minute power consumption and usage of each panel, etc etc. So lets call the rest $4k
Total of parts ~$20k. $12k for construction in the Seattle area seems fair for a complex electrical install that includes a 10 year install and 25 year system warranty.
I think like many things, the last mile of work is way more expensive then people realize. Solar is more than the panels themselves
Note: I could have used cheaper panels, but they take up more roof. I could have also saved money by skipping the per-panel micro-inverters, but then the system is more brittle and less efficient
I have a friend in America who has a fairly large PV set up on part of his property. Listening to the all the obstacles he's had to overcome and all the surprising limitations he's saddled with, I can't help but suspect that the power company in his area that he's selling the power to would rather that these sorts of installations simply not exist.
Coupling that with the fact that off-peak pricing isn't the norm in a lot of markets, I have to wonder if trying to implement smaller build outs is really worth the effort or expense... though I'll be the first to say that I've not researched it thoroughly.
The profit model for utility companies is to build new power plants, power lines, and substations, and charge their rate payers for that build ("rate base" it). Then they get a cut of their expense as profit. If too many people build solar, the utility can't build new infrastructure and charge everyone for it. And, they now have to manage balancing solar production with electricity usage, which means hiring new analysts which can't be rate based.
So, of course they are hostile to it. Personal, distributed solar cuts their profits and increases their costs. But, because their profit is based on charging customers for building new electric infrastructure, it lowers the costs for all of their customers when a group of people preempt that by installing solar on their homes.
It is good for everyone when rich people buy solar; the air gets cleaner, and electricity becomes cheaper. So some states, like California, compel them to facilitate solar via tax credits and other regulations.
The utility companies also have to deal with excess capacity created by solar, typically by selling at a loss to other states. This shortfall is then made up by other customers, which as a demographic, tend to be poorer than solar panel owners.
> If too many people build solar, the utility can't build new infrastructure and charge everyone for it.
Right. Because solar does not play well with existing infrastructure but solar needs the grid to be even close to viable. It needs it because it is a diffuse intermittent power source, where the peak power generation does not actually match peak power use. Solar overproduction cannot be saved for later either, meaning that you have another infrastructure challenge of dealing with overproduction when you don't need it.
In general, solar only really works if it is paired with a a traditional power generator that can ramp up quickly (when there is no solar output) and ramp down quickly (when solar output is high). The only power generators that can do that are ones that burn fossil fuels. That's why natural gas companies are a BIG lobby group for solar. That's why Germany is signing massive, multi-decade deals to ship Russian gas, as they expand their wind and solar infrastructure.
If you care about getting rid of carbon emissions, solar is a disaster. Not to mention energy use in building panels and the number of rare earth materials each one needs, the problems with recycling them, and their land-use requirements and the environmental impact of that.
>It is good for everyone when rich people buy solar
This was true 20 years ago and will be laughable 20 years from now.
> The only power generators that can do that are ones that burn fossil fuels.
Fossil fuel generators do NOT ramp up and down quickly. BTM and FTM battery storage solutions ramp up and down faster and are rapidly becoming cheaper and more environmentally friendly alternatives.
> energy use in building panels and the number of rare earth materials each one needs, the problems with recycling them, and their land-use requirements and the environmental impact of that.
Every one of these supposed drawbacks of solar also apply to generating power with fossil fuel.
Fracking causes earthquakes in Oklahoma and wastelands in southeast New Mexico. Mountaintop removal for coal has probably permanently destroyed parts of PA, WV, VA, and KY. I won't argue that the issues you bring up aren't bad, they just lack context. In my view, human addiction to electricity is bad in all forms, but fossil fuels are clearly worse.
>This was true 20 years ago and will be laughable 20 years from now.
Will the sun start shining at night in 20 years?
>Fossil fuel generators do NOT ramp up and down quickly.
Natural gas ramps as quickly as it can. But let's assume that's true, that's not a point against me, it's a point against solar because then it means that solar is completely unworkable.
>Every one of these supposed drawbacks of solar also apply to generating power with fossil fuel.
No. Because we know we can power an economy with fossil fuels, and do it in any weather, day or night.
>Fracking causes earthquakes in Oklahoma and wastelands in southeast New Mexico.
Yes. Natural gas isn't great.
>In my view, human addiction to electricity is bad in all forms, but fossil fuels are clearly worse.
Right. That's why we need to get off fossil fuels. But you can't do that with solar or wind.
Here's what works: Nuclear, Hydro, Geothermal - that's basically it and the latter two need particular geography to be viable.
Storage costs are dropping rapidly. New batteries are nearly there for cost competitiveness. As the first generation of BEV batteries become 'spent', they will be reused in grid-scale applications. Space and weight are not a concern, and most will have ~70% of their original capacity. That accelerates quickly once Gen 2 batteries have hit end of life.
Read around, nobody actually believes Li-ion batteries can bridge the intermittency gaps of solar and wind. The production capacity may never be there and cost aren't there today and may never get there (cost for US to store 10-15% of excess summer energy to use in winter with Li-ion batteries would run tens of trillions of dollars). Li-ion batteries leak energy if you want to use them to store excess power for weeks (as you would have to).
That's the problem. There isn't a battery technology now, or forthcoming.
Who in their right mind ever suggested such an obviously overpriced option when you can just build enough for peak and then shut them off during the off-season?
> Natural gas ramps as quickly as it can. But let's assume that's true, that's not a point against me, it's a point against solar because then it means that solar is completely unworkable.
You're thinking about this wrong. If you want 100% generation from solar then you need a ton of storage, period. But who says it has to be 100% solar or 0?
Take a look at the daily electricity demand curve. Lower at night, higher during the day. The (large) difference between the nighttime demand and the daytime demand? That's what you use solar for. Even if batteries are completely off the table, solar is still useful for that. And in that case your natural gas plants only have to run for a couple of hours between sunset and bedtime instead of all day long.
We don't know if storage technologies are ever going to be cost effective enough to use 100% solar day and night. They're not right now, but they're also declining in price. If they get there, well, that's the end of the matter. If not, you use nuclear at night, supplement it with solar during the day and, in the long term, suffer the high cost of batteries for a couple hours around dusk, which is still a whole lot cheaper than needing enough of them to get you through the whole night.
But you have a significant amount of solar in every case. And it can also take on close to 100% of the new load from charging electric vehicles simply by charging them while the sun is out.
>And in that case your natural gas plants only have to run for a couple of hours between sunset and bedtime instead of all day long.
Right. So we agree, solar cannot remove carbon emissions. And I thought the entire point is to remove fossil fuels entirely from energy generation, not just reduce it.
>you use nuclear at night
It doesn't work like that. Nuclear can take days to ramp up and ramp down. Solar does not play well with nuclear.
That's one problem.
A more major issue is...if you're using nuclear, why the heck do you even need solar? This is the energy mix of my home province[1] - we're essentially fossil fuel-free because our energy mix is nuclear and hydro. There is no point in increasing investment in solar and wind. But we're doing it, not because it's good for reducing carbon emission, but because it's a fad.
> So we agree, solar cannot remove carbon emissions.
Except for all the carbon emissions it does remove while the sun is shining and you didn't have to run that natural gas plant for 14 hours instead of 2.
> It doesn't work like that. Nuclear can take days to ramp up and ramp down. Solar does not play well with nuclear.
You don't ramp it up and down at all. You have 100GW of demand at night and 200GW during the day, so you have 100GW of nuclear and 100GW of solar. Nuclear generates 100GW night and day and the solar generates an additional 100GW during the day.
> A more major issue is...if you're using nuclear, why the heck do you even need solar?
Because it's cheaper during the day and there is more demand during the day.
There is no battery technology (now or forthcoming) that can store enough energy to power a moderately sized city for a few hours, much less days or weeks.
>..and irreparably destroy the planet by doing so
I wasn't making a case for natural gas. I was a case that solar needs to be paired with natural gas.
>But I think we are perfectly capable of utilizing these AND solar, wind, and storage?
Why? If you're using nuclear and hydro, why bother with solar and wind?
Here's a live view of the energy mix of my home province [1]. Almost all clean. Why bother with solar and wind?
Nuclear is great in your part of Canada, but where I live in California there are several advantages to solar...
1) solar is distributed. In the past couple years, California has had massive fires caused by transmission lines. Right now, millions of people are without power because the utility can't safely deliver electricity. Distributed solar obviates the need for huge transmission lines, and is more durable in a disaster.
2) utility scale solar is cheaper than nuclear here.
3) There are large earthquakes here that make it less obvious that we should run 70% of our economy on nuclear.
The only fossil fuel generators that ramp up and down quickly are the ones burning natural gas powering turbines. The rest is rather sluggish. And Nuclear is terrible at this too, especially at load shedding, which can cause a plant to be shut down after which it will take up to 72 hours to power back up.
Utilities generally use what is called spinning reserve. That is a plant, or a few plants, that are sitting at idle power with the turbine generator up to speed but may or may not be connected to the grid. The old coal and oil plants were good at this and could quickly meet peak demand. Gas is much better at it though. Nuclear is always base loaded as it's usually the cheapest kilowatt maker in the system. It's never taken me 72 hours to start the nuclear plant I worked at except after an outage where a valve lineup check of the entire plant is performed and mountains of paper work to do.
Your practical knowledge obviously trumps my second hand one. The 72 hour figure came from a guy I knew that worked on the CL&P nukes in Pickering. We talked about it in the context of the cascading failure that caused the widespread grid outage, he said that 'once the nukes shut down we will be down for at least three days because that's how long it will take to bring them back up'.
Or you use batteries which is in the default option. From the article "You can also enter your address and get an updated estimate that takes into account local costs and incentives, and add on any Powerwalls (with three as the default for a 2,000 square-foot roof)"
>Or you use batteries which is in the default option.
What batteries? Hate to break it to you, but there is no battery technology capable of storing enough energy to bridge solar intermittency gaps at even the scale of a city.
This solar roof is for single family homes, not city scale energy. You can power a single family home by building extra solar capacity so you have a surplus of energy during the day to charge batteries which you use at night. Just search for "off grid homes" to see examples. It'll most likely be more expensive at todays prices but the technology absolutely exists.
The problem with hydro is that it doesn't scale. You can build it where there is a suitable location, but most of the suitable locations already have it.
Batteries double the cost of solar for a 24 hour storage cycle. If you want to store energy for 48 hours, the cost doubles again. 96 hours? Doubles again. Want to store a mere 10% of US electrical generation from summer to use in winter? Literally 52 trillion (yes, trillion) dollars if you use batteries. A bit expensive, even if you spread the cost over 25 years. Pumped hydro, CAES and hydrogen stored in salt caverns are much cheaper at that scale.
Only if you’re failing to build enough solar. There is zero reason to ramp up storage rather than reduce the need for it. Over production is cheaper than any other option and reduces the need for longer term storage as daily solar output never drops to zero.
Solar is cheap enough it’s still cost competitive if you assume significant power production is wasted. Further, battery systems are designed to avoid full discharge, design for 99.5% of the time and you have spare capacity for the last 0.5%.
True, but it is also almost never 'rated power'. Typically a solar installation will deliver somewhere between 30 and 50% of rated power on average, depending on your latitude it might be even lower. The times when solar panels work at their best is during clear winter days and when they are capable of two axis adjustment. Because of the cost of such an installation the typical choice is to install more panels running at lower efficiency.
And you'd need to size your batteries for those worst case scenarios of obscured sub in which case your panels might only output between 5 and 10% of their nominal capacity.
For reference, my 1600W array is outputting 15.9 Watts right now, earlier, at the best time of the day it was making about 300W due to the sky being overcast. If I had to rely on battery power to make up the difference I would not be able to make it on solar alone.
A healthy mix uses both solar and windpower, hydro if you have access to it (it is by far the most stable of the three and capable of providing baseline power without any tricks).
The economics of residential solar is heavily tied to local conditions. But, for grid solar there is little reasons to build it in the north rather than just move electricity from the south. We are already below 2c/kWh for solar power in some areas, the average consumer price is 13.27 cents per kilowatt hour in the US. That’s a lot of wiggle room.
Let’s assume 50% over production that’s 4c/kWh, backed up by 60h worth of batteries and transmission and your still cheaper than Nuclear while covering base load and peaking power needs. Further at National grid scale 50% over production an 60h of batteries is crazy overkill.
> But, for grid solar there is little reasons to build it in the north rather than just move electricity from the south.
That's true, and if you're going to ship it then HVDC lines running East/West can add a couple of hours overlap as well. All this focus on rooftop solar is nice because it is decentralized but quite a few parts of a structural solution will not come from rooftop solar but from very large grid scale installations.
> We are already below 2c/kWh for solar power in some areas
Yes, the power is very cheap exactly when nobody needs it.
> Let’s assume 50% over production that’s 4c/kWh, backed up by 60h worth of batteries and transmission and your still cheaper than Nuclear while covering base load and peaking power needs.
Can I see your calculations? Last time I did the math,
building enough batteries to store 48 hours worth of US energy use was simply not viable.
Note as this is crazy overkill I am ignoring battery and lithium shortages of actually trying to do this. Any such transition would aim to minimize such costs. This is also not the kind of thing you do in a weekend long term pricing is required.
Batteries are predicted to cost 62$/kWh or 62 Billion$ per tWh in 2030 assuming no supply shortages. We can’t build anywhere near that many batteries by then, so it seems like a reasonable baseline long term. Further, by having vastly more batteries than required they end up with fewer and longer discharge cycles which extends lifespan. Current system are also deigned for much shorter lifespans because of rapid battery price drops. Based on that and further tech progress a 20-30 year lifespan seems reasonable I will use 25. Capacity is also going to drop over time and installation and maintenance costs are > 0 so I am going to add 10% ‘other’ and ignore 7% of US grid electricity generation being hydroelectric.
US annual electricity usage is ~4,000TWh / 365.24 days per year / 24h * 60h = 1.7 trillion dollars that’s a lot of money but the US can easily borrow on that scale for cheap. Further over 25 years that’s 4,000 TWH * 25 = 96,000 TWH for 1.7 Trillion dollars or 1.70$ for 96kWh or ~1.78c/kWh * 1.1 so call it 2c/kWh. Most of this power would of course come directly from solar generation, but having that many batteries is what transforms intermittent solar into peaking power.
Add 4c/kWh for crazy overkill on solar and you’re at 6c/kWh to have both base load and peaking power power. Though you still need to add significant new transmission and would have significant losses associated with charging and discharging batteries etc. Still add another 2c/kWh for all the costs I am not including and your still cheaper than Nuclear though not by as much.
PS: At grid scale wind and hydro further reduce battery storage requirements. Combined with excess though more reasonable generation say 1.5x and averaging across huge geographic areas 15h of battery power is probably significant overkill.
Who predicts that? National Renewable Energy Laboratory[1] summarizes the predictions, and the average of predictions is $200/kWh in 2030, which is already more than triple of your estimate.
> or 62 Billion$ per tWh in 2030 assuming no supply shortages.
It is convenient to assume perfectly elastic supply, yes.
> US annual electricity usage is ~4,000TWh
Why do you consider only electricity? It's only just above a third of US energy use. US uses about 101 quads[2] of energy annually, which is just under 30,000 TWh (BTW, US Energy Information Administration believes that we're using 11000 TWH electricity annually, which is again triple of what your numbers are).
So, by the numbers of US government agencies, we're at (30,000 TWh / 365 days) * 3 days = 246 TWh, which, again, using government estimates, is 246 TWh * 200 dollars/kWh, which is $50 trillion, which is 250% of US GDP. So, assuming that the lifetime of the battery installation is 20 years, it means that 12.5% of US GDP will need to be spent on maintaining battery infrastructure, in perpetuity. That's already 2.5 times larger than current energy sector, and that's only cost of battery construction, we haven't even started talking about generation or grid maintenance!
In comparison, two new US nuclear power plants in Georgia, which are already extremely expensive due to cost overruns (US cannot build stuff, and you should also expect cost overruns for battery plants too), are astimated to cost $23 billion, and planned to have 2.2 GW power, and, unlike renewables, they'll actually produce as much day and night. 30,000 TWh/year is 3500 GW of power, so we'd need about 2000 nuclear plants like the one in Georgia. 2000*$23 billion is $46 trillion, amortized over 50 years is less than trillion a year. But, of course, if we could build reactors as cheaply as, say, South Korea can today, we'd only need spend fifth of the cost, so with $200B/year of construction costs, we could satisfy all our energy use with nuclear, not even needing solar/wind for peaks.
23 billion for 2.2GW is just a building and some equipment you still need workers, fuel, repairs, decommissioning etc. Further Nuclear only has a ~90% capacity factor they can’t operate 24/7 365 for 50 years without turning off. Nuclear also has the opposite problem, it needs to scale up production in the day to cover peak demand, and even more to reach peak seasonal demand. Run the numbers assuming exactly the right amount of power based on an annual average and your shifting energy across months. At a 90% capacity factor nuclear costs ~12c/kWh, at a 45% capacity factor that jumps to about 24c/kWh though as peak instantaneous demand is over twice average demand you would still need batteries.
Also, pretending we suddenly need to replace all sources of energy from PV power that’s then stored in batteries is crackpot territory. Part of those 101 quads are fuel used to generate electricity at sub 40% thermal efficiency. We don’t need generate electricity to replace the energy from fuel used to generate electricity.
Further, batteries only store electricity and it’s silly to pretend they are part of the loop for jet fuel which is also part of that 101 quads estimate in these calculations. Hell even using electricity rather than fuel oil to heat your house uses less energy via a heat pump.
As to prices, I think we are comparing apples to oranges here. If you want to design a system that takes AC power from the grid coverts it to DC for storage on a battery and then back to AC your adding far more than just batteries. But, if you’re producing PV power on site that’s DC so it would need it’s own DC to AC converter as well as equipment to connect to the electric grid and regulate production. This all adds up to significantly reduced prices when packaged together. So when I say battery prices are already under 200$/mWh I mean Battery prices.
“In March, an analysis of more than 7000 global storage projects by Bloomberg New Energy Finance reported that the cost of utility-scale lithium-ion batteries had fallen by 76% since 2012, and by 35% in just the past 18 months, to $187 per MWh” So even today we are beating your 2030 estimate. https://www.sciencemag.org/news/2019/07/giant-batteries-and-...
PS: The article makes it’s own estimate at 900GWh of batteries being sufficient for 100% rentable electricity generation using some other set of assumptions.
> 23 billion for 2.2GW is just a building you still need workers, fuel, repairs, decommissioning etc.
Workers, fuel, repairs, decomissioning etc. are very cheap, much cheaper than equivalent coal fired plant, for example. Levelised cost of electricity, which includes all of these plus construction costs, is still the lowest for nuclear energy, much lower than wind and solar even if you don't build any batteries at all.
> Further Nuclear only has a ~90% capacity factor they can’t operate 24/7 365 for 50 years without turning off.
??? How is this relevant for the discussion at all? If you have 2000 plants, then with 90% capacity factor, you'll have 10% of plants down at any given time, thus you simply need to have 10% more plants than what you need. The crucial thing is that with nuclear, down time can be scheduled, while with wind and solar, overcast sky and windless days are completely beyond your control.
> Nuclear also has the opposite problem, it needs to scale up production in the day to cover peak demand, and even more to reach peak seasonal demand.
You simply need to build enough for peak seasonal demand. The nuclear plants can scale up and down as needed, albeit slower than gas plants, so you just scale up to cover peaks, and have aluminium plants enjoy cheap electricity when you overshoot.
> Run the numbers assuming exactly the right amount of power based on an annual average and your shifting energy across months.
My 2000 plants for $200B/year was already 12% above average use, but sure, I could double my numbers to 4000 plants and still be below 10% of US federal government spending. Imagine how big of a boon would such plentiful and cheap off-peak energy be for US heavy industry or data centers.
> Also, pretending we suddenly need to replace all sources of energy from PV power that’s then stored in batteries is crackpot territory. Part of those 101 quads are fuel used to generate electricity at sub 40% thermal efficiency. We don’t need generate electricity to replace the energy from fuel used to generate electricity.
Sure, but note that the nuclear numbers also scale just as well, so even if the increased efficiency might make the whole batteries business a bit closer to the realm of feasible, it will make the nuclear approach even cheaper too. The inefficiencies you mention will probably amount to something like 30% of current energy use, so they don't change the calculations substantially: utility scale batteries still infeasible.
> PS: The article makes it’s own estimate at 2.25 trillion for 900GWh of batteries being sufficient for 100% rentable electricity generation using some other set of assumptions.
Yes, their estimates are 10 times higher than mine. That should make you make think really hard about feasibility of large scale batteries. For 2.25 trillion, South Koreans can build 900 GW worth of nuclear generating capacity -- which one you think is more sensible choice, one hour of 900 GW from batteries, or continuous 900 GW from nuclear plants, if your goal is "decarbonizing the grid"?
However, that’s including loan financing which inflates the construction costs further. Similarly, decommissioned might cost 1 Billion, but as it takes place so long after construction it’s easy to set money aside which can then compound for decades. https://www.reuters.com/article/idUS178883596820110613
Looking at pure construction costs in a steady state environment without loans it’s a significantly smaller fraction which continues to increase with age. This is why several viable nuclear power plants have been decommissioned early, their operating costs are to expensive even after construction has been paid off.
> ??? How is this relevant for the discussion at all?
Rather than a 2.2 GW power plant producing 2.2 * 24 * 365 GWh it gives you 90% as much energy. So if you want 2.2 TW 24/7 you need to build ~1,112 of them not 1,000 of them. Further this also multiples every other cost.
> down time can be scheduled
Up to a point, your talking days of downtime not hours so you still need peaking power. The cheapest approach is for a steady state of workers smoothly moving from one project to the next. That’s hard to pull off, and you end up needing even more generation to cover scheduling issues and unexpected problems.
> “Areva, the French nuclear plant operator, offers that 70% of the cost of a kWh of nuclear electricity is accounted for by the fixed costs from the construction process.”
Yes, this exactly means that operating costs are very cheap: if 70% of the cost per kWh is amortization of construction costs, and only 30% is operating costs, and the kWh from nuclear plants are still the cheapest, or exactly means that the operating costs of running a nuclear power plant are very, very low, much lower than coal plants.
> Looking at pure construction costs in a steady state environment without loans it’s a significantly smaller fraction which continues to increase with age. This is why several viable nuclear power plants have been decommissioned early, their operating costs are to expensive even after construction has been paid off.
Which ones? What were their operating costs?
> So if you want 2.2 TW 24/7 you need to build ~1,112 of them not 1,000 of them.
Sure, but as I noted, we can build twice the amount we need rather easily, while building enough of utility scale batteries to ride out the variations in solar/wind production is just infeasible. Of course nuclear isn’t perfect, but I thought your alternative is solar/wind, which has all the same problem, but worse - instead of predictable and controllable 90%, you get 40-50% of installed capacity on average, and it can be all down at times beyond your control.
> Up to a point, your talking days of downtime not hours so you still need peaking power. The cheapest approach is for a steady state of workers smoothly moving from one project to the next. That’s hard to pull off, and you end up needing even more generation to cover scheduling issues and unexpected problems.
That’s okay, since we can build twice our peaks and still be firmly in the realm of feasible.
I note, however, that you gave up on the idea of utility scale batteries being feasible approach to work around the inherent unpredictability of solar and wind. Good.
“Each nuclear power plant employs 500 to 1,000 workers. Nuclear worker salaries are 20 percent higher on average than those of other electricity generation sources.” 40 mill x 50 years = 2 Billion dollars just for wages. https://www.nei.org/advantages/jobs
As giant complex mechanical systems over time moving components like pumps and turbines need to be replaced. Due to contamination some of these costs are crazy high.
Insurance is a big one. They need normal insurance for stuff like fires and workplace accidents, plus government subsidized insurance for the rare major accidents that could be horrifically expensive.
While cheap relative to coal they still need fuel which adds up to 14% of operating costs. While U3O8 is $68 by the time it’s ready to be used the costs increase to $1390 per kg (https://www.world-nuclear.org/information-library/economic-a...) and a 1GW reactor goes through ~25,000kg of enriched uranium per year. https://www.nuclear-power.net/nuclear-power-plant/nuclear-fu... That’s 35 million dollars per year and 1.75 billion over 50 years. (As a sanity check. If it’s 14% of total operating costs then it’s operating costs are 12.5 billion for a 1GW nuclear power plant operating for 50 years. If 12.5 billion operating costs = 30% of total costs then a 1GW Nuclear reactor costs 41.5 billion over 50 years or 830 million per year and at 90% capacity factor it’s 10.6c/kWh which is close to the 12.5c/kWh I have seen quoted frequently.) With reprocessing it’s about only 3kg for 1GW of electrical power per day at 33% thermal efficiency, but without repressing they need significantly more fuel. Unfortunately repressing is not currently cost effective.
They also need the basics like office equipment like computers and supplies, as well as the normal costs associated with buildings like lightbulbs, roof repair, water and sewage etc. Plus the normal taxes etc.
They also need to set aside money for decommissioning.
PS: Digging into those fuel costs where really interesting it’s interesting to see how 68$/kg and 3kg per day which I have seen quoted before gets transformed in a much larger expense.
“Each nuclear power plant employs 500 to 1,000 workers. Nuclear worker salaries are 20 percent higher on average than those of other electricity generation sources.” 40 mill x 50 years = 2 Billion dollars just for wages. https://www.nei.org/advantages/jobs"
Yes, if you multiply things by 50, you get large numbers. Note, however, that $40M per year for a plant generating 2200 MW day in day out (well, let's say 90% of the time) results in $40M / (2200 MW * 0.9 * 1 year) = $0.002 per kWh, that is, 1/5th of 1 cent per kWh in employees cost. I say that's really cheap.
> As giant complex mechanical systems over time moving components like pumps and turbines need to be replaced.
Same is true for all kinds of power plants. You need to compare the costs between all for a sensible comparison. It's fun to make all these calculations by yourself, but fortunately, there's already a metric for that, called levelised cost of electricity (LCOE). It's basically the total lifetime cost of a facility (construction, operation and decomissioning) divided by total amount of energy produced. As you can see in [1], LCOE of nuclear easily beats solar, and matches wind plants, which suffer from the disadvantage of only producing electricity when the wind blows. If you didn't have option of base load fossil or nuclear plants, you'd have to build storage for wind and solar, which would make their LCOE jump through the roof.
Look, it's clear that nuclear isn't free, so yeah, you really need to spend a billion on fuel and another on wages over 50 years. More importantly, though, over those 50 years you'll produce $50 billion worth of electricity. Just look at the published LCOEs, and consider how much they'd have to go up for solar & wind if we couldn't fall back on fossils and nuclear.
In future, if we ever manage to get off fossil fuels, the way we do it will be nuclear for base load along with some amount of wind and solar (though not much, due to environmental concerns that will make it very hard to build significant amounts of wind and solar). As I hope is clear for you, it most definitely will not be wind & solar + batteries, and no nuclear.
Some utility companies in the U.S. are quite solar-hostile. A major Alabama utility charges a monthly “backup fee” of $5 per kW of installed solar, regardless of usage:
I’d bet the unsubsidized economics of residential solar for the utility actually make this likely a pretty reasonable number.
I have solar myself, but we have to be honest that residential installs really destroy the economics of traditional power generation. And net metering regulation makes that even more the case, which is why net metering can’t last and/or storage will be required along with residential solar.
Where does he live? Some states are surprisingly anti-progressive. California actually has measures to help people install solar (since 1979!). Things like HOAs can't deny your installation based on aesthetics, and anything under 10KW is allowed with very little oversight.
I had a professor who worked for the electrical utility in Ontario for years before teaching. They definitely would prefer if they didn't exist. Not so much for the competition with their own generation, but their distribution infrastructure gets more complex as more small generators tie into it, hence more work for them.
which lists $42,500 as the total price for the average 2,000 square-foot home, with 10kW solar panels. It also lists $33,950 as the price after an $8,550 federal tax incentive.
I had a roof put on my 1600sq ft ranch and it was only $8,000
It's comparing apples and oranges. Presumably the OP is not using premium concrete (terracotta) tiles on his roof. Tesla is cheaper than a premium roof with solar panels and the details of said comparison are laid out on their site.
The comparison on Tesla's site still looks misleading.
"comparable price of a typical roof + solar panels"
Having premium concrete shingles isn't a "typical roof". They also use small greyed out text to say they are comparing it to concrete shingles, but use big black text for the higher price, and hide the fact that the Solar Roof is only cheaper because of tax breaks behind a "See More" button.
For a ~10kW (AC) system you're going to pay about that much, I just had it done.
It's misleading to look at the price of the panels alone. You also need (micro) inverters, you need to have it hooked up to the the grid, you need several inspections (construction + electrical). Getting all those panels installed on your roof and wired up is quite labor intensive as well. The installation cost is significant.
> It also lists $33,950 as the price after an $8,550 federal tax incentive.
There is something absolutely astonishing about the fact that in America the way we have go about being eco-friendly is to give huge tax breaks to rich people to buy solar panels and electric vehicles. To think with all that money we could build better infrastructure and denser housing. Really boggles the mind.
Incentiving rich people to buy expansive tech (and still at lost for the riches) help the technologies to ride the cost curve and bring it to the masses.
Let rich people get the buggy version 1.0 stuff and help the technology mature for everyone! I'll wait a few years and get the debugged version 2.0 or 3.0 for 1/10th the price.
4K resolution, but only support for HDMI input at 30hz, or having to use some incredibly fiddly solutions where two cables had to be used at once. Internal image processing was too slow, so you would see 100ms delay which made them too slow for playing games. And then the quality of 4K panels was frequently far below good 1080p, but for more money.
No no. That's not how it works. Rich people have money, so they are perfect early adopters since they subsidize new technology until the technology gains economies of scale. But that's the key thing, they are using their own money to do that. It's not the same when you have the government subsidize their purchase.
That represents an income of at least $70,000/year, assuming only the standard deduction and single. If you are married, it's closer to $100,000 income. That puts you at the 62th and 77th percentile, at a minimum to get the full tax deduction.
Because that disconnects the company from actual demand.
Production of early versions is always more expensive because of the research & development, creation/configuration of the machines, etc, etc but the marginal cost goes down drastically. The numbers work out even better as you can amortize those early costs over a larger and larger production run. This means v2 and later editions are cheaper.
I would also say that Rich people pave the way for adaption.
For example, look at prices of flights. Before there was no "first class" everything was first class, now there are cabin options.
While that example is a service, more proliferation of service == wider adaption == increased scale of economies == eventually prices come down for everyone.
> This is exactly what is meant by trickle down economics.
This is not true. Trickle down economics is the theory that money given to rich people will enrich poor people because the rich people will use the money to buy goods and services the production of which will require the labor of poor people. That is, this extra money at the disposal of rich people will create an economic domino effect such that money will ultimately end up at the disposal of poor people. In a trivial sense this is true. The theory is only interesting if this is an efficient way to increase economic activity generally without deleterious side effects.
But it never was the theory that giving money to rich people would spur research and development such that services ultimately were cheaper.
As to whether it's a scam, that is another matter. The allegation then is that the boosters of trickle down economics do not believe their own claims as to the efficiency of this process. That is, they believe, because it is tautologically true, that this will enrich those who are already rich, and they desire this, and the story about the efficiency of this process is a tall tale to salve consciences or win over rubes.
Except that time and time again, it has been shown that it absolutely does not work.
"[I]f the income share of the top 20 percent (the rich) increases, then GDP growth actually declines over the medium term, suggesting that the benefits do not trickle down. In contrast, an increase in the income share of the bottom 20 percent (the poor) is associated with higher GDP growth. [1]"
That would be true if solar roofs were some novel technology that the rich would adopt first, and then the rest. This is a remix of existing technology, much like the iphone. The rich don't have roofs, they have domes or yachts. The poor don't own roofs. This is directed at the middle class squarely and only. To the extent that richer people adopt this, it's mostly for signaling reasons and not because it affords them some luxury they didn't have (like flights did).
Your analogy is more appropriate for space tourism.
This presumes rich people never have good motives and any appearance of morality or civic-mindedness they might show is a self-aggrandizing sham. This seems unlikely.
Why not just do unit tests instead of integration tests? Because you want to create a system that provides feedback on the entire end to end process. You subsidize purchases because The thing you want to incentivize is not making solar panels, but bringing solar panels all the way to market. You are trying to get the market off the ground, not just the technology.
>Incentiving rich people to buy expansive tech (and still at lost for the riches) help the technologies to ride the cost curve and bring it to the masses.
The masses don't own houses to put these roofs on. Because of concentration of wealth and return on capital exacerbating that concentration. Because of rent seeking pyramid schemes. Which I imagine artificially restricts demand for improvements like solar panel roofs: who has more inventive to invest in things like solar panel roofs? Owner occupants who are in it for the long term, or rent seeking landlords just trying to make their mortgage payments on time and turn a buck?
Of course it is, and unfortunately you see it a lot on HN - apparently no one in any country can afford houses anymore. It's like this Futurama joke - "no one drives in Manhattan anymore - there's too much traffic". On this side of the pond it's the London residents saying that no one except for the ultra rich can own houses, forgetting that in other parts of the country you can buy a (new!) house for as little as £100-120k.
Yeah it's true inside London. I was there a few years ago looking at house prices. 500k pounds gets you a two bedroom studio flat in the London and a 4 or 5 bedroom fully furnished family home 1hr outside of London. That same 4 bedroom house inside London was between 3 and 5 million pounds.
But for the working class/lower middle saving up 10k pounds for a deposit is real work.
No? Because I live in one of those shitty parts of the country and since minimum wage laws still apply here turns out that saving £5k for a mortgage deposit isn't as impossible as saving £50k down in London is.
This is vastly moving the goalposts of the discussion. The point is 64% of Americans are dealing with issues like deciding when to replace their roof, and what material to replace it with.
Yes, for purposes of the discussion about installing roofs, you are correct. I am just bemused by this statistic. The vast majority of “homeowners” are paying a lot of interest to lenders.
Outside some strips of land on the coasts much of the masses do own homes. Some people get them at 30, some people get them at 50 but for the most part everyone who isn't just bouncing around minimum wage jobs will be a homeowner. That's not to say that the home isn't just a trailer on land and that the location and/or upkeep of the home won't make most upper middle class HN-ers recoil in horror but the fact of the matter is that a heck of a lot of people are homeowners and most of them aren't rich by any stretch.
Even if everything you say is true, it still doesn't negate the fact that it's a new technology which is expensive and needs investment in order to bring down the cost.
When computers were new, only rich organizations like governments, banks, and large corporations could afford them. Now you can buy a raspberry pi for close to nothing.
Photovoltaics (PV) is not a new technology. It’s been in production-state since the 50’s and powering percentages of some countries for at least the past ~5 years.
Adopt the existing, proven and cost-effective implementation of PV (~3’ x ~5’ aluminum framed panels).
Waiting on Elon to figure out boutique solar so the affluent can all install it first and drive mass adoption (decade?) stalls PV development as much as the US tariff on Chinese-made solar!
>To think with all that money we could build better infrastructure and denser housing.
Denser housing isn't built because our local land use policymaking paradigms give the power to block it to the people who don't want it, not because of a lack of federal subsidies.
If anything, the infrastructure built with that money would be road infrastructure to support further sprawl.
Really? Isn't that how all successful technology starts off - highly priced, effectively experimental, but then becoming more and more available to the general public as the price goes down and production increases?
The trouble is that you're necessarily picking winners and losers in order to design the incentive scheme.
It makes no sense to pay somebody $7500 to buy a new car to drive to work, but not to switch to cycling or telecommuting.
In this case, it's not necessarily clear that this is a product that needs to happen. Grid scale solar is significantly cheaper due to the economies of scale.
Were there a carbon tax, there'd be the same encouragement of environmental solutions while leaving the market some flexibility in achieving that goal.
With that said - I'll take the politically feasible solution rather than nothing.
>Grid scale solar is significantly cheaper due to the economies of scale.
Utility scale solar has met the same ardent opposition in land use decisionmaking fora as denser housing, including from 'environmental' organizations.
Summary of 5 links: 2 projects cancelled due to NIMBY, 1 project being appealed due to NIMBY, 1 project cancelled due to ecological concerns (unclear if just NIMBY), 1 irrelevant link.
fredericksburg.com - application cancelled - "He, like other neighbors of the project and another larger one south of Stevensburg, have consistently expressed concerns about impact to their property values ... Orye said he didn’t think solar fit anywhere in Culpeper County, emphasizing its rich historic resources related to the Civil War and early American history."
lakegastongazette-observer.com "Opponents of the project ... turned out in droves at last week’s community meeting", "Many residents expressed concerns about the project’s effects on the ecology and their view", "the commission’s reasons for the denial were fivefold. They said the proposed facility develops more than 500 acres, is removing more than 1,100 acres of agricultural property from production, is proposed to be located in a populated residential area, is located on real estate that is located on the Virginia Scenic Byway and that setbacks and buffers on the proposed project do not adequately address the need to visually block the project from the view of its neighbors"
richmond.com - "a solar project that had significant community opposition", "plans to build a solar energy farm on 1090 acres of a 2998 acre property", "400 acres of the project would [become] a conservation easement" , "The planning commission determined that the proposed use is not in accord with recommendations made in the 2010 Long-Range Comprehensive Plan, in particular since part of the project would fall into the area designated Priority Conservation Area and Protected Land.", "Among the citizen comments that were made, most were in stringent opposition because of the negative ecological impact the project would have on protected wetlands in the area, the only remaining wildlife corridor connecting the James and Appomattox rivers, and on rare and endangered species."
starexponent.com - "Culpeper County will consider placing a cap on 'utility scale solar sprawl' equal to the amount of the renewable energy its residents could actually consume.", "In October, the board approved the county’s first-ever solar farm on 1,000 acres near Stevensburg to the dismay of neighbors, who have since filed a civil suit in the matter asking the circuit court to reverse the approval."
www.pecva.org - looks like last link just shows planning details and doesn't show opposition.
>www.pecva.org - looks like last link just shows planning details and doesn't show opposition.
They oppose utility-scale solar anywhere except existing industrial land, in favor of the "low hanging fruit" of rooftop solar.
>PEC is an advocate for solar energy, especially distributed solar power generation - small scale solar (usually rooftop) primarily designed to meet the immediate demands of the property in which it is located. In comparison, the size and nature of USS Facilities create challenges for any locality to protect important resources and the public health, safety, and welfare of the community. Virginia has thousands of acres of rooftops, parking lots, and landfills devoid of solar panels in areas of moderate to high energy demand, in addition to contaminated and/or underutilized industrial sites. It is PEC’s belief that we should be looking to these developed areas as the low hanging fruit of future solar sites.
>Utility-Scale Solar (USS) Facilities are industrial facilities and should not be allowed by-right on
agriculturally zoned lands. Given their size and nature, they should be:
>● Subject to a Conditional or Special Use Permit in agriculturally zoned areas with maximum acreage allowed per project and possibly per the entire County; and/or
>● Limited to existing industrial zoned areas.
> ○ Agricultural land should not be spot zoned to create additional industrial zoned areas for USS Facilities.
That's exactly what regulation is and does. It picks winners, picks harmful phenomenon, and acts accordingly, with a short-to-medium timeframe based on what would get the most votes.
I, too, believe that taxing actual atmospheric carbon impact would be great, but how would you measure such a thing fairly?
Go as much to the source of production as possible : forbidding unconventional fossil fuel extraction, heavily taxing fossil fuel imports long range plane landings, methane-producing livestock, &c.
>That's exactly what regulation is and does. It picks winners, picks harmful phenomenon, and acts accordingly, with a short-to-medium timeframe based on what would get the most votes.
The subsidies are terrible. They are way too narrow and don’t address the root cause of the issue. Climate change is not happening because we don’t have enough renewables, climate change is happening because our GHG emissions are too high.
Tax the emissions and all activities which generate emissions become less appealing, while all opportunities to reduce the emissions become more lucrative. The example of riding a bike instead of buying an EV fits into this as well. With a carbon tax, using a bike would be equally accounted for.
By using subsidies, the government gets to pick the winner and doesn’t let people think. Renewables especially as they are now are a terrible way to tackle the problem. If we used taxes instead of subsidies, I’m afraid that not many renewable power plants would be built, because other alternatives are way more effective at curbing emissions.
Solutions like rooftop solar – and by extension, these Tesla tiles - wouldn’t make a lot of financial sense if people were billed for their electricity appropriately based on costs. The cost of electricity distribution is primarily a fixed cost in the infrastructure and its upkeep. The variable cost of generating the electricity is small in comparison. People, however, are generally billed more in variable cost based on use of electricity rather than with a larger, fixed fee. It’s set up in this way because it makes sense; the poor can afford connecting themselves to the grid and keep the necessities powered, while reckless use of energy i.e. having AC on 24/7 is disincentivized by the cost.
But once people start pushing rooftop solar on their grid and selling the energy back to the grid at the retail price, not the wholesale price, they are being massively subsidized in their electricity generation, as part of the fixed cost of operating the grid was baked into the variable cost of electricity prices. The pushback from utilities comes from exactly this perspective; if more and more people start net-metering their own bills lower with the rooftop solar, then it means that fewer and fewer people will be responsible for appropriately funding the grid.
In an extreme scenario where everyone net-metered their bills down to zero, it would mean that nobody would be paying for the grid. Obviously, such a situation is unsustainable. And even though there’d be plenty of solar to go by in such a situation, the grid would still be a necessity, as is all the generation which functions as a back-up. And somehow, it would all still need to be paid for.
But there’s more to it, too. Net-metering creates a reverse-incentive. Electricity prices in the wholesale market can even go negative if there is overgeneration, because overgeneration is harmful, and it needs to be dealt with. If net-metering practices are in place, and people start doing more and more rooftop solar to the point of more and more frequent overgeneration, the costs of operating the grid go up. Consequently, so would price of electricity. With a net-metering scheme still in place, everyone else is further incentivized to build solar panels on their roof and do the same to avoid paying the high prices.
If the compensation for the electricity fed back into the grid was based on wholesale price instead, then, once average price (and value) of solar-generated electricity goes down, so do the incentives to build more solar panels.
Adding insult to injury, the back-up issue is generally solved by using gas-fired power generation, as batteries cannot address the issue at this scale. Batteries are useful for bridging the gaps in between switching power generation sources, but the kind of quantity of capacity you’d need for a battery-only backup solution is not feasible.
For back-up, gas is the primary back-up solution in California as well as in Australia, and the Russians are building a pipe to supply Germany. Ironically this makes solar generation - when the whole power generation solution is inspected at the grid-level - not all that green. And while this aspect of solar is ignored by subsidies, a carbon tax would still take it into account.
The point being, subsidies can not only be ineffective, they can also be counterproductive in addition to ignoring all the alternatives.
> I, too, believe that taxing actual atmospheric carbon impact would be great, but how would you measure such a thing fairly?
You can measure emissions at sources and impose heavy fines on anyone who tries to circumvent them. The common problem in globalization, though, is defining and applying these standards universally and fairly. If you can’t trust the Germans to not fumble their numbers, what you’d think would happen in some other countries? Still, with a coordinated effort and systems to ensure transparency, it would be possible.
Another method would be to subsidize more effective alternatives or for the government to build the more effective solutions by themselves. While this wouldn’t have the direction-changing impact of the carbon tax, it would be more straightforward to implement without the requirement of tracking pollution. Targets of subsidies could be selected by their impact on emissions per $ invested. Of course, the numbers could be fumbled here as well, but at least the calculations could be made public and be subject to scrutiny.
I pretty much agree on all points. This is a tragedy of the commons on a global scale. I’m not sure what the answer is.
Subsidies are woefully inaccurate, but can be applied locally on most political climates. Carbon emission taxes would be better, but global competition makes that difficult.
Man, if they gave me $7500 to ride a bike, I'd ride the fuck out of that bike. I'd buy like 5 and attach them to each other and take my friends with me everywhere.
This effectively multiplies the amount of money invested in climate-friendly technology and infrastructure by a factor of 5 with 80% coming from rich people, while generating jobs and more taxes. I find the classist outrage detrimental and disappointing.
Denser human scale development is the real answer. Electric cars and rooftop solar is not going to solve the massive suburban sprawl crisis. You can thank the federal interstate system for that (effectively subsidizing land).
I agree we should put the breaks on urban sprawl, but the US is already spread out. I mean what’s more feasible: Have all current suburbanites sell their house, move downtown, and have their old property leveled OR get them driving an EV?
We don't have enough housing or mass transit for the first option, and building more batteries has proven significantly easier than building either more housing or more transit.
This is where we'll have to disagree - the limitations to building housing and mass transit are mostly economical, while for batteries they are mostly physical.
Would definitely disagree on that. The limitations of building housing and mass transit are mostly political, while for batteries they're both economical and physical. I'd take physical problems over political ones any day.
We know we can't quickly replace all gas combustion engines on the road today with batteries because of material supply constraints alone. However if the goal is to reduce carbon emissions by vehicles by 90%, we can probably get there by replacing 95% of all gas-only vehicles with plug-in hybrids, which require a small fraction of the amount of battery material than fully electric cars require. I'm probably off with my estimates, but I believe they're in the ballpark.
Hmm, that's interesting - I thought that hybrids were less efficient, but maybe if they also require much less batteries they could be a great stepping stone ?
You don’t really have to be “rich” to afford this roof. It’s a totally sensible investment that actually pays for itself between 10-20 years and you can use a loan to finance it. It’s also very sensible for new home construction which is also traditionally financed via a mortgage.
>It’s a totally sensible investment that actually pays for itself between 10-20 years
You can't say that!!!
That may be sort of true for traditional solar panels. There is scant data on whether a solar roof will ever be cost-effective. They are an unproven technology at this point.
There are several differences between solar panels and solar roofs:
1) Solar roof is more expensive and less efficient. You're trading efficiency for something that looks nice. Is that something a government should subsidize? Rich people getting subsidies for aesthetic reasons?
2) Solar roof panels are not as durable as solar panels, are not as scalable, and the market for them is tiny - limited to people who want to get solar panels and a new roof. V1 and V2 versions of Tesla roof was a disaster.
3) Installation is much much harder and requires specialized professionals. Installing solar panels is already fraught with challenges (safety and otherwise), and all those need to do is collect sunshine. A solar roof also needs to take into account protection from the elements and therefore needs increased durability.
Total subsidies for this stuff are in the low $billions. Really it's not that much money overall (the EV market is still small, and the other stuff is all still noise), and wouldn't go far if spent for infrastructure.
FWIW: newly built dense housing is actually cheaper than the more typical sprawl. In fact it's much cheaper if you factor in (heh) infrastructure savings due to the reduced road and utility lengths.
Building denser housing is problematic not due to lack of money.
Local non-monetary regulations mandate low-rises or single-family houses, huge parking lots, etc. If money could solve that, it may be by spending it on political campaigning, public education, or may be even bribing, to get the regulations out of the way.
Americans on average do not want denser housing. It's not part of their culture, they don't see the point, they want a backyard, and they don't mind having to drive all the time.
I'm pretty sure that even the people benefitting for the wealth cities produce would prefer not to live there, and I think the development of rich people's houses along the CT coast attest to that, they try to balance the wealth of the city and low density housing.
I think you are right that a lot of Americans do not want dense housing. One thing we are missing, especially on the west coast, is good public transportation. Its not that I don't mind driving as much as that I don't have a choice. I'd love to take a train/subway to work or to the next town over or to the big city nearby - but those trains do not exist.
I just moved back to France after living 2 years in Phoenix. You can't rely on public transportation, I don't want to wait for a freaking bus to go get milk. I want density enough that there will be a shop I can walk to. public transportation, is good for important planed travel like going to work, but not for small stuff like going to the pharmacy or the gym or going from shop to shop to find the couch of your dreams.
Without dense housing, trains are just unfeasible. It'd take an absolutely massive network to spiderweb out into the vast suburbs of most cities and the costs would be beyond belief.
I used to live in Palo Alto. There’s VTA (bus) and Caltrain. Stops are few and far between so getting to and from the stops are in and of itself a problem. Buses tend to arrive every 15 minutes (or longer) I believe? Pretty infrequent. And outside peak hours there’s usually like one Caltrain an hour, so missing one really sucks.
I had a colleague who had to travel frequently between Stanford and Berkeley (slightly unusual); it’s a sub-hour drive, but easily takes ~3 hours or more via public transit.
P.S. At least the Valley is fairly bike-friendly by U.S. standard.
Yup - and everything is 2 stories or less. Nothing gets built, not because of demand, but because zoning laws. There’s no reason other than the city of Palo Alto won’t allow tall buildings to be built - the less that is built the more wealth homeowners and real estate investors get.
In town, yes. Between towns, maybe/sort of/infrequently.
I can drive to work in 15 minutes. The bus takes over an hour including 30 minutes of walking. There are no bike lanes between home and work, its possible to bike it but really not safe - I'd guess the bike would take about 40 minutes.
I'd imagine if a larger percentage of people took the bus, there could be much more frequent buses and dramatically less traffic on the roads making for faster busses.
The same thing happens in Europe, most of the eco-related additional costs are actually regressive taxes that will mostly fall on the less better off. There are not that many media outlets openly discussing this, and the few people who actually do talk about it are painted as “wanting to destroy the planet” or worse.
UK solar subsidies were particularly daft in this regard - because we're further north and have terrible weather, capacity factor is pretty awful compared to somewhere like California, plus solar is guaranteed to contribute a big fat zero to peak power demand (which is after sunset in winter). It basically just acted as a hand-out to wealthy home owners paid in part out of the power bills of the less well off. Somehow this didn't stop our nominally left-wing party attacking the government for ending it, though...
...and yet it is particularly common in the UK for the home solar to be installed by a commercial provider (who don't get that subsidy, just feed-in tariffs), that provide the panels in exchange for the feed-in tariff from your roof. You get mostly free electric and sometimes some portion of feed-in tariff, they own the panels.
If it was that pointless, this sort of deal could never add up, for either party.
Feed-in tariffs are a subsidy - they pay out more than the actual market value of the electricity, and that money comes from the bills paid by other consumers. This is especially true of the ones on older UK solar installations, which are higher than the retail cost of the electricity. There might be some justification for arguing that the ones in other places with much greater reliance on air conditioning and a summertime peak in usage aren't, because they could act as a form of peak shaving, but that doesn't apply to the UK.
This didn't happen because the numbers added up. The government mandated it for political reasons and all parties had to go along.
The sad thing is, the feed-in tariff subsidy would be less than UK's existing fossil fuel subsidies for homeowners. Yet solar is worth doing today in the UK, even with no subsidy if you project to stay in the house long enough to pay back capital cost.
With current solar pricing, it would be better for everyone if they simply moved a requirement for solar into the building regs. Like they've steadily done with insulation levels- though these are not yet adequate.
A large chunk of the UK's existing "fossil fuel subsidies" for homeowners consist of the government charging a lower 5% VAT rate on domestic electricity and heating, much of which of course comes from fossil fuels. Take that away and spend the money on feed-in-tariffs, and you're again transferrirng money from the less well off to the wealthy.
And all we get out of it is less carbon in the atmosphere!
Installing solar panels can pay off in the long run, but the big draw is to reduce your carbon footprint. Even if you are a rich asshole it's not a bad thing for you to be less polluting.
Those same incentives put a new Model 3 at $31,000 [0]. The entry level Toyota Camry is $24,000. You'll save $700-$1,200/year on 'energy' with the Tesla, save on oil changes and brakes and the Model 3 costs less to own than the Camry in maybe 7 years. Reduce pollution, avoid gas stations, have a lot more fun driving the safer Model 3. Electric cars are not just for the rich any more.
Depends a lot of locale. Where I live, in local dollars, an entry level Camry is is $24,295 (or $28,250 for a hybrid) while a Model 3 is $50,990 after incentives.
Also not sure about a Model 3 being safer than a Camry, they weigh a similar amount and are both IIHS TSP+. (And a hybrid Camry weighs close to 1000 lb more than a Model 3.)
> To think with all that money we could build better infrastructure and denser housing
How will an 8K subsidy per home help increase housing supply?
Will that subsidy also automatically take care of global warming?
64% if Americans own a roof, FYI.
What is truly detrimental to society is not the promotion of carbon neutrality, but this kind of crab mentality where some people benefiting more hurts more than the fact that this is a positive step for the planet as a whole.
Indeed. The electric vehicle incentives are very much designed that way. The rich get more money than the poor. The higher your income, the more you get back. I'd have to earn an extra $30K to get the full $7,500 (and switch from high deductible to traditional healthcare/cancel my HSA), whereas someone wealthier buying the same car gets every cent.
Not quite. That $7500 is a full on tax credit. Not a deduction. So as long as you pay more than $7500 in taxes, you get it all back. And if for some reason you pay less, you’ll get a refund.
I didn't realize it wasn't refundable. However you don't need to make much to pay $7500 in taxes. I don't make enough to buy an electric car and I pay three times that.
> However you don't need to make much to pay $7500 in taxes.
You do if you're a family, with a 24K standard deduction, a family HSA ($7K), and any number of 401Ks for retirement. A group of people who would benefit significantly from an electric vehicle.
As I said, I don't qualify for the full amount and I earn well above the state's average.
Maybe it seems like this isn’t the best of all possible worlds, but are you sure there isn’t some rationalization that will let me continue to believe that better things really aren’t possible?
Reframe that question: Why should poor people pay more because they’re not rich enough for the incentives?
OP is not saying rich people should pay more; He’s saying get rid of the tax incentives. Yes, they end up paying more, but then poor people are paying the same as the rich.
One of the problems is ambivalence. For many, PV is "nifty" but unnecessary- if I go to a fast food joint and I have a choice between a pretty good hamburger for 1/3 the price of a pretty good hamburger made from beef substitute, which do you think I'm gonna buy? But if I'm told that I can get the beef substitute version for just a few dollars more than the regular version because of a tax break, you can bet I'm far more willing to try it out.
"Rich" (which is a relative term anyway) people are, by and large, the same. Get rid of the incentive and you'll get rid of the progress that comes with it. Saying "oh take away the incentive because not everyone pays the same amount" is a great way to get people to stop investing in it at all. It's not like poor people are going to say "oh wow, the rich people won't get a tax break for investing in this, now I'll definitely invest!" right? So poor people still won't be able to invest and now rich people will have less of a reason to overcome their ambivalence. Shockingly enough, taking away incentives will deincentivize people. Who knew?
That doesn't answer my question. Why someone has to paid more just because it has more money in the bank, what makes you morally superior to get better tax cuts just because you have less dollars in it?
I think a major challenge that is yet to be addressed is the fact that solar panels lose their efficiency over time [1] [2]. Seems like you're looking at 80% of rated efficiency after 20 years - and I'm unsure this includes failures.
In the UK a 50+ year old roof is very common - most people expect a house to outlast them without anything but minor repairs. If I were building a house I would want something where it's primary power source can be repaired without larger than necessary costs.
UK roofs are made and ceramic and slate. US roofs are made of asphalt. Mostly because asphalt is inexpensive (particularly due to economies of scale/lack of them on ceramic). Ceramic lasts a lifetime, asphalt can be taken out by one bad hail storm.
UK homes are made of brick. US homes are made of fiberboard and inexpensive woods. There's a reason why UK homes have lasted over one hundred years, and most US homes won't last 50. It is a consumer culture applied to homes. They're designed to be replaced completely.
Very much depends on how it is constructed. As a Swede 100% used to sturdy, well insulated wooden homes having lived in Texas for a while it felt like they were made of paper in comparison.
Then Sweden is also known for ridiculously high building standards, even compared to other European countries.
New builds have lower life expectancies. Thing about your fridge from the 40s, compressor built to last. How long has your stuff been lasting recently?
A tip someone told me here in the UK is that if you are buying an inexpensive property try and get one that was built and owned by social housing operators - they actually have much higher construction standards and will have been maintained properly.
I've lived in 11 properties in the UK (owning 4 of them) and most of them were made of stone and were 150+ years old - only the first one (a 1950s council house) was brick.
Well, it would have been a huge waste to build houses to that standard here because they all would have had lead paint! Nobody knew that then, but still. I recently considered buying an old house and ended up with a post-1978 one so I don't have to worry about that.
Also, people mention again and again that people replace houses in Japan even more, and yet it doesn't seem to get the jingoistic juices flowing like criticizing the US.
Of course, since Tesla's cost comparison only works out if you compare their solar roof to one of those more expensive ceramic/slate roofs, it seems fair to compare the lifespan to those roofs too.
50 years is almost new. Most 100 year old terracotta tile are still on the original set, and slate often lasts 200.
Where roofs are replaced, aside from storm damage, it's most common for that to be down to having got rot in the woodwork than fault with the tile. I suspect a fair few are down to iffy sales tactics from dodgy builders. (i.e. get someone out to fix the chimney flashing "oh, major problem you need a whole new roof...")
A relatively routine roof replacement required because it is now expected to serve more than one purpose (simultaneous weather
- proofed shelter and PV surface area) could be justified, especially if the entire neighborhood utilizes its existing local grid because the entire street of individually owned Powerwalls is one large shared, connected, metered (people with high energy needs could still run welders and power tools that pull from other Powerwalls), software driven reservoir:
Each existing home (even ones built in newly minted neighborhoods) would still be connected in parallel. It's the nearby 70-150kV subtransmission relay station historically feeding that parallel bank of homes that would become inactive - ironically bypassing the Nikola Tesla prompted long distance HV amperage flow from remote generation sources. AC implies spinning dynamos (loud flywheel momentum), and PV-to-battery is solid state, with high frequency silicon carbide switching (quiet). Classical home 13kV transformer bypassed or replaced with a small DC boost to launch amperage across the hundreds of feet of a neighborhood's acreage. Typical loads could be balanced even if a few people want to weld or smelt metals during a bright sunny day (especially summer solstice).
As a comment above pointed out, the solar capacity would decay over time, but since there is a local grid reservoir buffer - the reduced capacity PV over time can mean the high quality glass protection can be kept even when cells under them are down below 50%. Such a roof would be contributing less to the reservoir than neighbors feeding the same grid with newer solar roof replacement. The user can be paid for their generation if they use less than they produce, incentivizing low noise, subtle generation sources (charging your powerwall with an exercise bike like Jaimie Mantzel, adding your electric vehicle(s) charge reservoir to the neighborhood grid if next day commute will mean you're parked near other solar roofs that will top you off until ready to return home - transporting and trading energy).
50 years? I have noticed that slate roofs are much more common in the UK than the US but still I see majority asphalt shingles, curious what y'all call them, on single family homes built within the last couple hundred years.
I wish y'all had some more thatch roofs but I suppose given the history in London, driving to the Cottswolds to see a thatched roof is probably fine ;)
In my local area we did, but at one point there was a push to demolish them all as they were "unsafe". I have a few stories about this for another time...
The problems with thatch is it's relatively hands on, you have to swap it out occasionally as the stuff rots.
The manufacturer states 25-30 years for shingles, but it's rarely the case they actually last that long unless you're paying for premium shingles. I'm in Canada though, so perhaps the harsher weather has something to do with that. Regardless, in Canada at least, 25-30 years for shingles is not common; 15-20 is more realistic from what I can tell in my area.
GAF has a 50 year warranty on its shingles. People who are talking about failures after 10-20 years are buying cut rate product and having it installed shabbily.
Why do they degrade so quickly? I also live in UK and replacing an entire roof is not a common thing to do, every 50 years sounds about right, my roof is 35 years old without any leaks.
UK roofs typically aren't asphalt shingles like most homes in the US and Canada. Asphalt shingles are closer to sand paper than the tiles you're accustom too.
We had to replace ours after only 15 years (and we should have done so earlier). The manufacturer's claims are suspect. Expansion/contraction due to the hot/cold cycle caused ours to crack (literally apart in some cases).
> Seems like you're looking at 80% of rated efficiency after 20 years
If you are living off grid, that might be an issue although it could be fixed by just getting some more solar panels later or oversupplying at first. However, for people still connected to the grid, they can just use a little more later on from the grid so I don't see what the big deal is.
This claim came up a couple years ago, and it is only true when considering very expensive/fancy roof types. It isn't at all true compared to normal shingles.
I am glad to see they are making progress here. This roof looks like it has a much better chance of being acceptable to various home owner associations (HOAs) than putting panels up on a "regular" roof.
The interesting parts of the call yesterday for me were talking about the challenges they have with v2, although we didn’t really get to hear much about how those issues were addressed.
First, that basically everyone was pulled off working on Energy to deal with Model 3 issues because they were do or die for the company. Now with Model 3 under control those resources were put back into Energy - solar, roof, and storage.
Second, there was not a fundamental manufacturing issue with the solar glass roof, but rather it was a product which fundamentally was not worth scaling until the economics made sense.
The economic driver Tesla is targeting is that anyone getting a new premium roof with solar would pay less for the solar glass roof than to get the new premium roof and then add panels separately. For someone with an existing premium roof with significant life remaining, the solar glass roof may still be something you choose, but it’s not literally going to be cheaper.
The major issue with v2 was what they call “edge conditions”. It is funny that this sounds like software guys talking about roofs, not probably the way that roofing guys would talk about it. But they are talking about anywhere that two planes meet on a roof where tiles need to create a waterproof barrier. Flashing, ice & water shield, etc. With their v2 product this required essentially custom fabrication in the field - what Elon called an artisanal roof, like building the airplane out on the runway. This required several days (even weeks) of work on-site to fabricate and install the product, and so they were never going to try to scale that.
Version 3 is supposed to address the fundamental issues with edge conditions. I didn't catch a lot of detail on the call of exactly what they did to address this, but the end result is that they do not expect to have to do any custom fabrication on-site, and they are targeting a process where a largish crew can compete an average roof install in 8 hours, ultimately they want the tiles to go down as fast as asphalt (which is quite fast if you’ve ever watched it).
Lastly, they talked about reducing the number of sub-assemblies in the tiles by 50%, and the expected difficulty in ramping up production. Elon tried to be very clear that predicting the exactly volume output on the steep part of the S-curve is basically impossible, but they want to be producing enough for 1,000 roofs per week, and manufacturing is already under way for V3.
They will be running install-athon competitions internally in Tesla on test roofs that they have to test different methods, and they also want to invite 3rd party contractors to compete as well. Their intent is to have both Tesla internal as well as 3rd party certified installers installing the glass roof. They will leverage their existing presence in 25 states where they do Solar today, and ultimately want the product available across 50 states.
If I was able to ask a question on the call I would have asked about winterization. I want to know that they’ve tested their solution against ice & snow, and particularly that it won’t ice dam during a New England winter.
Financially, I wonder how the margins compare to Model 3. ASP is in the same ballpark, so ramping up to 1,000 installs a week would be total revenue just barely on the same order of magnitude as Model 3 revenue, which was a serious reality check for me. Elon said on the Q3 call that they think Energy will be as big or bigger than EVs overall and that he thinks people are significantly undervaluing their generation and storage LOBs.
>If I was able to ask a question on the call I would have asked about winterizarion. I want to know that they’ve tested their solution against ice & snow, and particularly that it won’t ice dam during a New England winter.
This is the magic question that I hope gets answered.
I quite like this product and think it'll be a big thing.
I do wonder about a comparison to what I've seen recently, which is standard roofs with solar built in almost like skylights. Just big panels at the same level as the tiles, not placed on top in a retrofit.
Seems like it's a good compromise (as long as you can get panels that blend with the tiles, and that generate enough that you don't need the whole roof covered).
Yeah, I did actually. Similar to electric cars or open source it's possible to spot the potential of an idea when it's still in its infancy and not really ready for prime time in every circumstance.
Roofs that generate make a lot of sense and many jurisdictions will mandate some kind of net zero build for new houses. I think there's a lot of room at the top of the market for this kind of product and it will just be standard at some point in the future.
How much of that market Tesla gets and keeps is a harder question, but combining roofing with solar is too obviously sensible to fail I think.
One of the things I've always wondered about Tesla's Solar Roof is how is it to walk on? Do you have to be extra careful because it's glass? Is it more slippery?
I get up on my roof several times a year (inspections, etc.) and know how to move around on a traditional asphalt shingle roof. Will I slide off or damage the roof while putting up Christmas lights?
Super scary if it's flashed wrong or damaged.. What if it starts leaking and/or needs repair. How much is THAT gonna cost and are the wait times shorter than for repairing a Tesla?!
EEVBlog would like to have a word with you-- are these systems efficient? There are concerns about cleaning, repair, and angle of insolation compared to traditional panels.
This looks like any other BIPV roof product -- what makes it different? More importantly: how do they plan to avoid the problems that have plagued BIPV for decades?
I'm not familiar with the existing issues, but likely thorough testing and quick scale. I'd expect tile durability and cable corrosion to be the main issues, which seem solveable.
They don’t seem to have addressed the two worst problems. One is interconnect — which drives the cost of installation and of maintenance too. The second is the bonding of the tile to the roof. Typically panels are mounted away from a surface so there is cooling airflow behind them. These are flush on the roof so heat up which is not only as for your house but dramatically reduces efficiency AND lifetime — which will be much shorter than the life of the tiles themselves.
I haven’t seen any mention that this product brings anything new to the table and I’m sure they would have trumpeted it if so. Basically 2012’s BIPV roof solution AFAICT.
This is very good news. It's going to lead to a great many new solar installations that otherwise would not have happened, including in new home construction. And if we are lucky there will be other companies that will develop similar products that are good enough to compete with Tesla.
Does anyone know the situation for putting these tiles on one side of the roof but not the other? I live in northern Indiana and would gladly put these on my southern facing roof, but my northern roof is particularly complex (a lot of joints/edges) and obviously not really ever going to get any sun.
You can do that with normal solar panels, or with these
I recommend using project sunroof to see if doing so would make you money. They analyze your roof size and tilt, trees around your property, etc. For free.
I believe they have "dud" panels which are not power generating, as well as for edges and joints where a full solar tile wouldn't fit. I also imagine the cost they show is not for 2000sqft of generating panels.
2 questions:
- If you opt-out of the powerwalls can the system switch to the regular grid when needed?
- Can you sell energy back to the regular grid in case you produce more than you need?
Yes you can switch to the regular grid and yes you can sell back your power in most states ("net metering"), though that often just manifests as a credit for your night time electricity usage.
Last I heard from PSEG, anyone who produces extra energy is just given a credit towards their next bill, there is no redeemable cash value. Not bad though :)
- 100ft.2 of solar panels (Qty: 5, off the shelf, ~3’ x ~5’ panels) = ~1kW of AC electricity (post-inverter, in low-sun Seattle)
- At the roof, this represents 10 wires and ~8 roof penetrations (racking lags through flashing into rafters)
AKA: simple setup, proven implementation
Questions:
- How many wires (points of failure) are in 100ft.2 of solar shingles?
- How many roof penetrations?
- How many roofer-electricians do you know?
- Why can we not just embrace existing panel technology and put them everywhere, yesterday?
Rant:
- We don’t need boutique renewable energy, we need ubiquitous RE.
- Elon, get your priorities straight. In the three years you’ve been trying to figure out the perfect solar status symbol, you could have installed megawatts of photovoltaics.
> Elon, get your priorities straight. In the three years you’ve been trying to figure out the perfect solar status symbol, you could have installed megawatts of photovoltaics.
They already do regular solar installs and so do a ton of other competitors. This is a boutique product for people who care about aesthetics and are willing to pay a premium for it.
EDIT: They're comparing it to concrete roof tiles, which a quick search say that they last for ~50 years, twice as long as the warranty for the Tesla solar roof.
https://www.tesla.com/solarroof/design