Odd choice to feature Project Loon in the banner image. I designed the solar array for Loon, it's built from plain old mono crystalline silicon cells. Sure, they're high efficiency cells, in a lightweight plastic stack up (no glass), but there's no exotic thin film cell technology in there. The options (so far) are either too expensive, or too inefficient.

Are these the space-grade multi-junction panels? Are the Loon panels size-constrained or cost-constrained ?

The article itself is odd - most PV will go to utility and commercial scale. Even on rooftop, aesthetics are possibly down the list after durability, reliability, efficiency and cost.

From the photos I've seen, no, they are not using triple junction GaAs cells. Boeing/spectrolab cells are incredibly expensive. They used high efficiency 156mm monocrystalline silicon cells. Just not in a glass + aluminum frame like you'd see on a roof.

edit: at least in 2013 they were using thin film amorphous silicon, or what appears to be CdTe thin film flexible, not rigid cells encapsulated in clear lightweight membrane.

https://www.wired.com/images_blogs/business/2013/06/20130609...

This appears to be the source paper: https://www.nature.com/articles/s41560-018-0258-1

... and NREL press release: https://www.nrel.gov/news/press/2018/nrel-identifies-where-n...

nrel is full of super interesting reports, such as EV efficiency and projections

This article seems to make the point that lightweight thin film photovoltaics are the future. I'll offer a counterpoint:

Massive economies of scale and automation are the future for PV. Using standard 60 and 72 cell polycrystalline and monocrystalline silicon cells. It's already not uncommon to see 5 to 10 megawatt sized ground mount PV systems.

Getting the $ per STC watt lower for panels is key. Having quick to build, low cost and efficient ground mount systems is an important part of it.

Things will be very different economically when a 360W, 72 cell panel I can buy now at $0.58/W is more like $0.22/W.

Battery storage is the other important part. The technological problem of generating enough watthours in one day is SOLVED. Storing and using it consistently 24x7 is now the hard part.

> Storing and using it consistently 24x7 is now the hard part.

Is it really? https://pv-magazine-usa.com/2018/03/01/12-hours-energy-stora...

edit: added quote from parent

It's not so much that it's hard technologically, as the Tesla massive battery system in Australia has recently proven, but to do on a REALLY huge scale at an economical $ per kWh cost, and the lifecycle cost of the batteries. When compared to current non-carbon-producing energy sources like big hydroelectric dams, big pumped-storage hydroelectric.

https://reneweconomy.com.au/tesla-big-battery-turns-one-cele...

Every few months I see some big press release and tech press media hype about liquid metal batteries and flow batteries, I've yet to see something useful that has a market price beyond "contact us for more details".

It seems to mostly be all test and evaluation installs on a small scale so far.

https://www.seattletimes.com/seattle-news/pullman-project-te...

If you look at homes in America the solar panels are these big heavy lumbersome things. They're unsightly to look at and devalues the leafy suburb view. These panels are really v1 of the solar panels for homes. You need solar shingles as the true means to maximize solar in homes. Hopefully those will come within the next decade.

For now I will avoid solar power.

> solar panels are these big heavy lumbersome things

Not really, compared to other systems in your house. Take HVAC... big metal furnaces burning gas within the conditioned space with tubes/duscts strung through the walls, attics, and crawl spaces. This system is way more cumbersome. Bolting some panels on the roof is probably one of the easiest retrofits you can do. Source - I redo houses.

> They're unsightly to look at and devalues the leafy suburb view

To me the leafy suburb view gets a lot uglier if you think about the massive consumption it requires compared to only slightly denser options (eg streetcar suburbs). It only works because of massive gov't subsidy (basically by not putting a price on carbon emissions). We Americans need to get over single occupancy car dependency. It is an environmental disaster. It is also a health (drive everywhere, don't walk/get implicit exercise) and social disaster (socio-economic stratification).

Tesla is already manufacturing solar shingles, they use tempered glass and are more than three times stronger than standard roofing tiles https://www.tesla.com/solarroof

Wiring up their solar shingles isn't cost effective due to labor. Their solar roofs is a negative income stream for Tesla. When is the last time you heard the Solar Roof grow into 'record breaking growth'? It hasn't and it won't because Tesla is still trying to make the labor expense cheaper.

Look into the labor it takes to install an entire Solar Roof vs installing a new roof plus traditional solar installation. On a dollar for dollar basis, it may look competitive but Tesla is selling their Solar Roof's at a loss.

Tesla is selling their Solar Roof at a loss. Not at a profit. It's labor intensive and they aren't growing at a fast rate because of that hidden expense.

Disclosure: I installed around 200kw of solar panels and did a decent amount of research into Tesla's Solar Roof since it seemed so novel and innovative. The wiring of the Solar Roof is much more complicated than what is publicly understood. Tesla's good at PR because most of the public are superficial in their understanding. Tesla most likely saves a lot of money in the manufacturing process of the solar shingles but they lose money due to their complicated wire installation process. How are those shingles connected on the underside? How are those wires protected from the elements, without affecting the plywood on the roof? Etc. These are simple questions that are often overlooked.

There's plenty of companies other than Tesla offering solar roof tiles who aren't selling at a loss. Wiring seems to be a simple case of plugging them in. How do Tesla manage to make this complex or labour intensive?

First one that cropped up in search: http://www.reuk.co.uk/wordpress/solar/solar-roof-tiles/

In the UK, mounting solar on top of the tiles is far more common as roofs are rarely replaced.

Wow. The UK link you provide is running at about ~$7.65 per watt. That's extreme/exorbitant plus their tiles look ugly and I doubt the resell value on the home will be as high as a Tesla Solar Roof (since aesthetics are king on homes).

Currently, even in Cali a solar system ranges at about $4-5 per watt. The cost to redo a roof depends on the size but almost certainly it would be cheaper than the extra $2.65-3.65 watt left over on using a conventional system and less long-term liability to homeowner to replace a roof and do conventional solar.

Also, let me further point out this system was a 2kw system which isn't enough to fully be energy independent as a home owner. Sizes range in the 5kw to 8kw, at least in the USA. So, that would add additional money into the equation and only further emphasize my statement on 'that's exorbitant'.

UK homes use far less electric than the US, and air conditioning is almost unheard of.

Commonest UK size for solar is currently 3-4kWp, which is plenty for household independence, and above 5 or 6 rare. Earlier systems were smaller as panels were costly. I think somewhere around 5 is the point an installation is considered commercial rather than domestic making larger even less attractive.

Exorbitant or not it pays based on UK electric prices and feed in tariffs.

Price-wise I think that may be a decade, perhaps more, out of date as they link to a long replaced grant scheme.

No idea of current prices, and Tesla are available here too. The couple of times I've seen solar tiles in situ (like I say, not common here), have been far more like the last photo than the first which stands out.

I question why wiring should be labor intensive?

Why not lay wires across the sub roof first that the panels clip into?

Or another idea is have the panels pass power between themselves to a single output connection point?

Great questions.

Tesla's problem isn't by intention. It's by NEC (and local) code requirement.

For the second question, why not lay it across the sub roof: Water can be very damaging. What insures those wires (over 20-30 years) will not be damaged? For a conventional solar installation the wires are elevated and protected due to the cavity of each panel. Local codes generally require that no wires touch the rooftop, so that prevents the daily ware/tare.

As for the latter idea, I do believe that's Tesla main goal. But to practically do this is an engineering feat that hasn't been solved yet. Can Tesla do it? Sure. But I haven't seen them invest in a solution yet. Heck, Elon is focusing on too much and not fixing this issue. It's okay, he's got bigger fish to fry. But since it's not solved, they will continue to live with this problem.

Tesla means massive expense.

But not a lot more than regular roof + panels. Tesla tiles are the roof surface.

> They're unsightly to look at and devalues the leafy suburb view.

This is highly subjective, though. Where I live, most houses have solar panels and they're just part of the appearance of a house. Unless your installer botched the job, solar panels just look like lower power bills and environmental responsibility.

Panasonic has been selling "low profile" solar panels for a while, here's Tesla's page for them: https://www.tesla.com/solarpanels

These aren't the Tesla shingles, and aren't that much more expensive than conventional panels.

What are you even talking about ? Integrated solar panels have at least 10 years of existence

I'm talking about solar shingles. The technology might be there but the cost is not which is why it'll take at least 10 years to be energy efficient.