Comment by drawkbox | Hacker News Reader

Comment by drawkbox | Hacker News Reader

Parent Full thread

Why Teslas don’t—and can’t—have solar roofs [1]

The top of an electric car has maybe 3–5 square meters of flat space.

Solar panels, even at high noon, usually only produce about 200 watt-hours per square meter.

The most efficient production electric vehicles today (probably the Hyundai Ioniq and the Tesla Mod 3) would only be able to travel 2–4 miles on that amount of electricity…in an hour. Most people could walk faster.

Financially, the cost of the panels and electronics, R&D and assembly would never pay for itself in the life of the vehicle, compared to charging from the wall in your garage.

While it might be nice to have for cars that sit there for a while or in the apocalypse. Right now it is better to use home solar panels or charging, as they have more surface area. Or maybe build parking structures with panels that can charge with more area. Maybe one day when they are small enough or can pull in sun into a smaller surface areas it will make sense.

Solar panel cars would probably also really up insurance rates and add more safety work/R&D.

[1] https://qz.com/1482588/why-teslas-dont-and-cant-have-solar-r...

I have issue with this: these panels would still capture enough energy to drive at least 2-3k kilometers annually - that's at least 10% of your typical annual mileage, so nothing to sneer at.

It doesn't get around the fact that it still makes way more sense to put those panels on your house roof and charge from that. The house roof is in direct sunlight for much longer than a car.

That is an excellent point - if you have a house. It might make more sense for people who have to live with streetside parking?

Why not do both?

Because very few people have covered every last spot on their roof so you would be designing a product suitable for very few people

Very few people have electric cars to begin with, and they're really expensive compared to ICE vehicles so you're already talking about a demographic that's demonstrated a willingness to pay a premium for driving an electric car.

If Tesla offered a "Solar Panel Kit" that made their cars even a few percent more efficient it would sell like hot cakes.

Or for people who don't own a roof at all - plenty of such people in cities.

Those people also have their cars in multi level car parks with no sun so that doesn't work for them either. In that case they would be better having the car park have solar and the local government building solar farms. A panel on something almost always undercover makes no sense.

I live in an average residential street in Europe full of 3 story residential houses. Our house with 6 apartments has 3 garages, the majority park on the street.

Sure, the government or the electricity company could just install car chargers and put a solar plant to power that in some rural area. But if I want something now then buying a car is much simpler than lobbying for that.

Because resources are scarce. If solar panels were free, then yeah let's plaster them everywhere. But let's start with where they make the most difference.

And on every parking lots...

> It doesn't get around the fact that it still makes way more sense to put those panels on your house roof and charge from that.

On a car it also acts as a range extender and emergency backup (if you're out of battery and the car is engineered for, you can crawl along at the charge rate or wait a bit for a better charge). It's not great, but being to "walk" with your car might save your ass.

And if the car is designed around this idea (à la lightyear one), you can increase things further (the LY1 folks advertise a solar charge rate of up to 7.5mi/h WLTP with no A/C).

If solar panels are the bottleneck then they shouldn't be wasted on people's rooves, either. They should put them in a big field where the electricity company can point them in the optimal direction, perhaps with a device to redirect them during the day, or mirrors, and look after them properly: clean them, if that helps, or whatever.

Tracking and concentration is not cost effective, so it’s just a tilt advantage.

Solar panels in a remote location need to be converted from DC to AC for transmission, then transmitted long distances with associated losses, then converted to DC again to charge the car, stored in batteries in the car with associated losses. A solar roof could actually be a net gain assuming the energy is used while the car is in motion, and it’s used near the equator so no tilt required.

put the panels on your roof, where they are in the sun more often, and cheaper to install.

I don't drive to work, but 2 miles per hour would be crazy good for me, as I live inside a city and my total commute is exactly 18 miles, which would mean my commute would be "free" (not taking into account the cost of the panels or the vehicle).

It's also unrealistically good as you don't get high noon insolation for 9 hours a day.

You do however get some power for more than 9 hours per day. Also, panels can go well past 200w/m at a theoretically possible 600w/m the numbers look rather different. Cars are only going 12-15k miles a year so hit 40miles a day and you generally stop needing charging stations.

The amount of power incident from the sun is roughly 1000 W/m^2 in a sunny area. Good Silicon PV panels will capture 20% of that power generating 200 W/m^2 at peak. Getting anything more than that is going to be too expensive for a car application.

Pulling 600 W/m^2 from a PV panel may be theoretically possible, but the world record efficiency for a PV device is still less than 50%. To get above 40% efficiency you need to use very expensive multijunction devices and solar concentrators (i.e. not going to work on a car).

I generally agree today, but that says little about the long term.

As solar panels have gotten cheaper installation costs have become more important which has pushed up panel efficiency. Cheap but good panels are up to ~23% efficiency where they where ~18% fairly recently. What’s interesting is these trends seem to be continuing with serious research into various affordable multi junction cells going on.

That means it has to sit directly under the sun all the time.

If you get the plugin Prius, you can already not pay for gas at all with short commutes. And it cost almost half of Tesla Model 3.

Yeah how many people want to park their car in the sun all day long?

The better way is parking it in a garage of small to large size, that has solar on the roof, with a much bigger surface area, that you could plug into.

Another bad part is that since solar is still in heavy innovation phases, you'd need to be able to swap out panels for replacement and newer tech so there would need to be standards there or else it becomes highly cost prohibitive. It would add so much to testing that it would add layers of cost. This would also limit design options and weight.

Even when efficient, you still want to park your car in or under a cover/garage. That is the best place to put the panels because noone wants a hot car.

Isn't it a case of the better killing the good ?

From your excerpt is seems electric panels still give 2-4 miles of range per hour of charge. Even taking into account the added weight, it's a net positive so why dismiss the idea completely ?

Read it again:

> would only be able to travel 2–4 miles on that amount of electricity… in an hour. Most people could walk faster.

I got it wrong the first time, then got it at the second take, thought I had it wrong again reading your comment and turned around again. The writing was weirdly misleading.

> Solar panels, even at high noon, usually only produce about 200 watt-hours per square meter.

> The most efficient production electric vehicles today (probably the Hyundai Ioniq and the Tesla Mod 3) would only be able to travel 2–4 miles on that amount of electricity…in an hour. Most people could walk faster.

From what I get:

- in the best conditions the panels could produce 200W in an hour

- 200W would allow for 2-4miles on the most efficient cars

The author's point seems to be that if the solar panel was the only source of energy the car would be slower that walking. It's different from seeing it as an additional source of electricity, just like regenerative braking is for instance.

> - 200W would allow for 2-4miles on the most efficient cars

Sorry, but that's wrong. They meant 2-4 miles per 1 kWh, not per 200 Wh.

Very good EVs get about 4-4.5 miles / kWh at optimal speed (usually between 25 and 35 mi/hr). So, best case, 200 Wh is good for about 1 mile, assuming you can deliver the 200 Wh over a period of 2 minutes.

However, 200 W can't push the car at 1 mi/hr because of "vampire" losses: Power for accessories and vehicle systems (like power-assisted steering, power-assisted friction brakes, the ABS controller, the airbag monitor, etc). 200 W of power isn't just slower than walking: In a normal-sized car, you will literally go 0 mi/hr.

More importantly, 200 W/m^2 is the best-case scenario at noon, in the tropics, on a cloudless day, with no shadows, using expensive, high-efficiency panels aimed squarely at the sun. As soon as any of those qualifiers is not met, power drops precipitously.

In particular, the atmosphere absorbs a large amount of light: In space, solar insolation is about 1300 W/m^2. At earth's surface in the tropics, solar insolation is about 1000 W/m^2 at noon. At higher latitudes, or other times of the day, solar insolation is lower.

A good rule of thumb is that (for Europe and North America) the total insolation over a full day is about 5000 Wh/m^2. So, a solar panel on a sun-following mount is limited to about 1000 Wh/m^2/day, again assuming no clouds. For a fixed solar panel, such as on the hood of a car, you'll get about half of that: 500 Wh/m^2/day. Remember, you need 1000 Wh to go 4-5 miles.

"Remember, you need 1000 Wh to go 4-5 miles."

That "4-5" implies a lot of accuracy, but I would have thought that the distance you can travel on 1000 Wh would vary by orders of magnitude depending on the speed, terrain and road surface.

However, if you are moving a 1 tonne vehicle up a 5% gradient then you won't do more than 4.5 miles on 1000 Wh, if I've calculated that correctly.

> Even taking into account the added weight, it's a net positive

You just gloss over the most important part of the question right there. At a charging rate that low, even a small amount of added weight may completely destroy the plausibility of this idea.

If your daily commute is short, solar panels could provide enough energy to not worry about charging

Or if you don't have a daily commute at all. Perfect for those who could switch to the occasional taxi or rental but stick to owning a car due to cultural inertia, perfect as that fallback family car that only gets used when the family fleet is fully saturated, perfect for the plugin hybrid that never gets plugged in due to the inconvenience of keeping two kinds of energy storage stocked.

If it can generate enough energy to move the 1 ton vehicle for 2~4 miles, it seems it's pulling its weight, literally.

That actually sounds pretty useful - say the car gets 10 hours of perfect iradiation on a good day, that's 20-40 miles, that could be a sizeable part of a daily commute, not to mention cases where you have the car standing still for a few days.

That would be more like 4 hours. A car solar panel would be horizontal, and it wouldn't track the sun when it's lower on the horizon. The 200W figure is only for noon.

Solar panels in Germany produce about 10% of peak power if you average over the year. Those panels are generally installed in the correct orientation to maximize efficiency. Panels on a car will likely perform much worse.

Honestly it seems like a nice idea. The idea that a car can go _entirely_ on solar on it's roof is clearly off-limits for now, but that doesn't disqualify the idea entirely.

Say I use my car for the daily commute and I'm some 10 miles away from office. I work 8 hours in the office in daytime, and leave my car in the parking. For when I'm going back home, my battery would be topped up.

I'd need to recharge my car much less, once a month maybe.

Also, say you can do 300 miles with a full charge. Depending on how fast you go, you'll do those 300 miles in 5 to 10 hours. So on average, a the solar aid would give you some additional 15 to 30 miles, ie 5%-10%. Which sounds nice to me.

All with the current technology. Obviously there are tradeoffs, in R&D, price and weight.

But it seems like worth pursing, log-term.

The folk building Lightyear One [0], a fully solar powered prototype car, would disagree with the assertion that it is off limits for now. Obviously the car is quite different from a Prius or any other modern car.

[0] https://en.m.wikipedia.org/wiki/Lightyear_One

I would be very surprised if that car could meet crash-test standards to be sold at large scale. If you don't have to meet crash test standards, you can make a much lighter car than you can today.

I drive a gen1 Chevrolet Volt. Plugged into a 110V outlet it adds about 4 miles range in an hour. This is enough to make its benefits worthwhile.

Reply on news.ycombinator.com