There are inherent limits on battery energy density.[1] Lithium-ion is nearing the limits of that chemistry. Magnesium-ion could potentially have about 30% more energy density, but so far, nobody can get it to work. Lithium-sulfur may be the next big step. Those actually exist.[2] However, they present a major hazardous waste problem, much worse than lithium-ion batteries.[3] The ultimate limits on energy density are probably 2x or 3x above the best current products. Basic problem: higher energy density comes from elements that are highly reactive.

Cost can be brought down through manufacturing economies. We'll have to see if Tesla meets their price point with their Nevada factory.

[1] http://www.tedsanders.com/graphs/batteries-over-time [2] http://www.sionpower.com/ [3] https://yosemite.epa.gov/OSW/rcra.nsf/Documents/CC7D81DF3070...

Ramez Naam has done lots of research/writing about "learning rates" for power technology. Here's a writeup from late last year on lithium ion batteries: http://rameznaam.com/2015/10/14/how-cheap-can-energy-storage...

There are two different studies, one which shows a 15% and one which shows a 21% learning rate (decreased price due to volume).

One thing that's worth noting is that lithium ion energy density (Wh/kg) has not increased appreciably in the past few years (~200 Wh/kg) - we're probably maxed out w/o switching chemistries. Oxis and Sony are targeting 500 Wh/kg Li-S batteries for 2019/2020, but I'll believe it when I see it - most of the Li-S research is super gee-whiz (lots of coated nanotubes, nanocomposites, etc) but that stuff never tends to make it outside of the labs, and if it does, will be quite expensive.

> if batteries continue to improve over the next four years as they have over the past four years

If CPUs had improved similarly, we would now have machines running at 100GHz. Just saying that extrapolating technology is a tricky business.

Fuel engine are improving too and oil price doesn't seem to want to go down any time soon. I'd say we have an interesting 10 years ahead. Electric is probably going to win, but time estimates are very hard to do. Don't forget that replacing a car fleet is expensive and takes years. A modern car can easily last 15-20 years, so even if we started replacing fuel with electric today (which we are not), replacing the whole car fleet would require 15-20 years. I don't see the 35% so far fetched in 2025...

And don't forget that there are millions of people who have a car right now, but who wouldn't have a place to charge an electric car. People living in apartments, terraced houses, especially in the UK it's very common to see rows upon rows of houses, none of them with their own driveway.

I know what you're getting at but I don't think the above comparison is entirely fair. Those houses don't have petrol pumps outside them either

Yes, but I go to a petrol station once a month, takes me 10 minutes to fill up and the whole tank lasts me for the whole month. With electric cars there are two ways out of this, either build enough electric charging points that you can charge while you are at work, or make every single car supercharger-compatible. I personally don't think the first option is likely, I don't see why my work(or any other) would have literally every single parking spot wired for charging(there's over 1000 parking spots at our office), and as for the second option, even 30 minute charging is excessive, it's fine o long journeys so you can go an have a coffee and stretch your legs, but if you want to quickly "fill up" after work 30 minutes is not acceptable. That's why I said that for majority of people charging at home is ideal - but a lot of people don't have private place to charge at home.

Yeah I understand what you mean and I think you're right. 1000 charge points is unacceptable but do we really need that? Given the current generation Tesla can already self navigate pretty well what's stopping the cars navigate the car park in your absence and charge themselves at a few central charge points.

https://m.youtube.com/watch?v=uMM0lRfX6YI

Sure, but we are having a conversation about whether it's possible that electric cars will have 35% market share by 2025. Considering that apart from a few first-world countries(and even then, only in big cities) there is very little infrastructure to fully support a large fleet of electric vehicles, and the battery swapping technology is not available anywhere apart from a couple tesla testing stations, I don't think that's possible. It would require a huge investment, and even then, cars easily last 15-20 years, so expecting 35% of the market to be electric in just 9 years is crazy.

35% Market Share generally means in this context: "35% of the cars sold in the year 2025 will be electric".

If you want to talk about the number of cars in use that are electric, then "installed base" may be the more traditional term.

Market share generally preceeds, installed base, especially in markets with durable goods.

I understand the skeptical point of view and its fair enough, we'll all just have to wait and see how this all plays out. I'm more optimistic bearing in mind that nine years ago the electric car industry barely even existed and its ramping up at an astonishing pace over the last three years.

To avoid making assumptions, I'm thinking of the cars that you can get right now. I live in northern Europe where the "charging stations" are actually abundant (in the form of electric feeds to parking places, currently needed for the convenience and benefits of engine block and cabin heaters in the winter) but charging an electric vehicle still takes quite some time.

But if we allow for technical development, then both electric and gasoline/diesel/lpg cars will also move ahead.

I'm getting down voted for giving my point of view and trying to be cordial at the same time but cest la vie.

I agree there is progress to be made on fossil power cars but it's still unsustainable in any analysis. We're also not seeing anything like the kind of gains that are needed medium to long term. Progress through one technique yields a regression in other areas (emissions vs efficiency)

I too live in Northern Europe (Ireland) and we're really badly setup for electric car adoption as it stands now though the situation is slowly improving. Policy and infrastructure here (as I imagine in other countries) always lags demand.

Fwiw, I didn't downvote you. It could even be just someone's mis-click. It happens.

Over here (Finland), you can also buy almost completely renewable fuel for flexifuel cars (85 % alcohol made of food waste, not sure how the 15 % of gasoline is produced.)

BTW, one thing I didn't know until just googling around now: Ford model T was also a flexifuel car (i.e. it could run on ethanol).

We would probably solve that by putting up more charging points at the roadside (there are quite a few of these around already in the UK), with a card reader for payment or perhaps a payment chip within the car's charging connector.

I can't see people liking that. Charging at home works because you can sleep while it happens. Road side charging takes time away from some other part of your waking day. Wikipedia is saying 40 minutes for an 80% super charge for a Tesla. That's a sizable chunk of time to be loitering at some random location.

I think it would work fine. Parking meters exist for a reason - a lot of people park their cars on the street and go do whatever they came to do. You could upgrade the meters to also support payment for charging.

Installing roadside charging points is the only approach that I can see working for inner cities - maybe create "charging only" zones where you can only park and charge cars - parking a normal car there getting a ticket.

NB I live in central Edinburgh in a flat and there is no way we could have an electric car without some changes to the parking/charging infrastructure.

At least here in Oslo, there are charging stations on every other street in the city center ( https://m.ladestasjoner.no/map ). You need to park anyways, and parking beside a charging pole will not take any more time than just street parking in general, as the number of electric vehicles increase I would expect there to be poles everywhere.

And if you live outside the center. Where there are no poles, can have a driveway with charging or use your block's parking garage.

On the other hand,car ownership in dense urban areas shouldn't be the norm. The future is also in active transport (walking, cycling) and transit, not only electric cars.

As someone who comes from a place in the world that can see -30C in winter, I disagree. For 2-3 months of the year cars(and in cities buses) are the only way to travel anywhere.

So if you want a car for hobbies (e.g. getting to backcountry) or transporting your kids around, you shouldn't live in an urban area?

When I was living in Montreal for a couple of years I signed up to a community car share scheme. No insurance, no maintenance, no road taxes so it worked out cheaper to "rent a car" just when I needed it: https://www.communauto.com/index_en.html

With a good, cheap, reliable public transport system with this as backup I could easily get rid of my car if I lived in an urban area but still needed to get around at the weekends.

We have a system like this in my hometown, Bergen. I live 50 meters away from two parking garages containing four share cars each. I can reserve any of these at any time via a web app, given that no one have reserved them already.

Payment is $10 per usage plus a surcharge per kilometer driven. Key management is handled either with an electronic card key which checks the current reservation status, or key drop boxes to which all members have access. It's both cheaper and more convenient than owning a car, since I only use a car once per week on average.

That's what I don't get about 'car sharing'. Everybody wants the car on the weekend. If the service has to have enough cars for everybody, what's the difference between that and leasing/owning? It doesn't save on cars that's for sure.

In my experience of using the service, and Montreal is a much smaller city than London (where I am from), you don't want a car every weekend. I was usually able to find a car for hire as long as I booked in advance (esp for Bank Holiday type weekends) but sometimes it meant going to a different location than the one right next to me. As the system get's more popular and more people use it then they are able to purchase more cars.

> It doesn't save on cars that's for sure

I disagree. Several of my friends also used the service, just between us that saved having several cars. If you need a car all weekend, every weekend then the service is probably not for you. But if, like me, you needed it every so often then it saved a lot of money over the course of a year.

I doubt that everybody needs a car on the weekend: People that want the car to go to work are actually less likely to need a car on the weekend. However, even if we concede the point that everybody want a car on the weekend, there's still the question of whether everybody needs a car on the weekend at exactly the same time. Given that most parking spots on my street are still full at any given time during the weekend, I doubt that.

Hah yes. In the early days of Zipcar in NYC (not sure if this is still allowed as I moved to the country and now drive a gas guzzling truck), I reserved every weekend for 48 hours for a full year in advance. Then on the Friday before the reservation started I would decide what I really needed, as there was no cancellation penalty. I probably only used it every second or third weekend, but noticed it really improved my "get out there and do something" motivation, as I set the default for "I have a car."

Probably wasn't good for Zipcar though.

Do you really want a car every day of most weekends? What for?

(I suspect the "good, cheap, reliable public transport" thing the parent mentioned is important)

There's no reason you can't have a car in an urban area, but we should work towards a transport model that doesn't require you to own one. Public transport in combination with cycling for short distances plus a solid carsharing infrastructure should be sufficient for a large chunk of the population. Even removing 30% of the cars from the streets would go a long way to improve the situation.

If you want a car for hobbies, you should figure out where to store it when you're not using it. Maybe rent a spot in an underground car park somewhere in the city. Public transport is just fine for transporting your kids around, and will be much better when there're less cars on the city roads.

If you only use it a couple of times a month, it would be cheaper to rent one than to own one.

If you are over 26 then maybe....at least in the UK if you are younger than 26 you have to pay a MASSIVE young driver surcharge(a car can cost 30 pounds/day to rent, but a young driver surcharge can be 100 pounds/day). It's literally cheaper to lease, a brand new Citroen C1 will cost you 100 pounds/month + ~80/month in insurance. The car is new so whatever goes wrong with it you are covered, you just have to put petrol in it.

Where do the normal cars park?

Here is an example of such a street: https://www.google.de/maps/@51.2310008,7.0720156,3a,75y,127....

You bascially park your car somewhere half on the street, half on the curb. The sidewalk is just wide enough for people to walk there.

By the side of the street. In between the street and pedestrians passage. Or between street and bicycle path. Sometimes diagonally.

The only possible charging would be some stuff that's underground and comes out of the ground right next to the car. Hopefully attaches to the bottom of the car, so there're no hoses/wires flinging around.

And some cars are parked at grey-legal spots blocking other cars. People just leave their phone numbers if someone needs to get out earlier. That is super common when you know people around and can expect to drive away sooner than your neighbour.

Or overhead?