Comment by anovikov | Hacker News Reader

Comment by anovikov | Hacker News Reader

Parent Full thread

That's now, this is because they are not massively produced yet.

Every high tech product follows a law that it gets 20% cheaper when volumes of production double, and electric cars make up about 1.2% of the market now. So, they are going to fell 3x by the time production volumes reach 36% of the current car production, or about 25% of then-current car production.

But, they will actually own the market before then because cost of owning and operating an electric car is much lower than comparable gasoline powered one.

Easy to see that this curve converges to 100% of market share with quite a bit of margin. At least 2x margin.

There is nothing inherently costly in producing an electric car. Transmission is simpler because it doesn't need a gearbox. It has much lower center of gravity due to heavy battery so little mechanical trickery is needed to make it stable on the road. Electric motor is much cheaper than a complex ICE which also requires a lot of service. Only stumbling block is the battery - and they are getting cheaper every month. Most of the existing cost of electric cars is due to low production volumes as they eat away market share - there is a limit on speed of market adoption as well as factory retooling and business acceptance of the fact that electric cars are the future - but these are just technicalities, there is absolutely nothing preventing making a $15,000 mass market Ford Focus-like electric car, if built in Ford Focus-like volumes, that is.

Nothing but all those things just mentioned - the weight and cost of the battery, complexity of electronics and software, resistance to market adoption.

The price has a large intractable stone in it - the battery cost. It won't go down very fast until new technologies are adopted. And it take a giga-factory to adopt a new technology. Rule of thumb: battery tech has to double the charge and halve the weight or nobody is interested in building the factory.

The center of gravity argument depends upon the current heavier batteries - but e-cars can't take off until something like metal-air batteries are working. And they weigh just about as much as a gas tank does.

As for complexity, sure compared to the mechanical engine it's simpler - 100's of parts vs a dozen. But the software has just exploded by orders of magnitude. Millions of lines of code instead of thousands. We mustn't underestimate this.

Then, there's a core of car buyers that have to grow old and die, and be replaced by e-car enthusiasts before the market moves significantly. Beta-Max was better than VHS (for the old videotape guys out there) but it died simply by being too late.

No, we won't see a $15,000 focus for decades.

You should pick a start date for your rule of thumb.

The advances you are apparently dismissing as incremental have been enormous.

http://www.nature.com/nclimate/journal/v5/n4/abs/nclimate256...

Tesla hit $190 per kwh last year and some speculators claim that they are now at $125 per kwh, while that is debatable (i don't buy that claim). Anyway, $100 per kwh after 5 years should be easy and that already makes battery+drivetrain for an average electric car cheaper than ICE engine+transmission, at about $9000.

>But the software has just exploded by orders of magnitude. Millions of lines of code instead of thousands.

Maybe its only one order of magnitude -- if I remember correctly, there were about 300,000 lines of code running on the main processor of the Toyotas that were investigated for unintended acceleration, with functions like "throttle angle control" that took several thousand lines.

This is plausible, but I still have a few concerns. EV will need battery replacement on some routine basis. How will that effect the TCO and lifetime of the vehicle? One of the advantages of gas vehicles is that I can buy a vehicle very cheap and operate it for many many years. Furthermore, I can park it outside (it doesn't need to be in a garage). I expect and usually see lifespan of 20 years and 200,000 miles on gas vehicles.

It seems like there are still too few EVs to fully understand their lifespan and long-term maintenance costs (including battery replacement).

I also want to see how fast "fast charging" will be and what kind of adverse impact it has on the battery. Is it acceptable to the usability and lifespan of the vehicle to routinely charge on fast chargers? I don't want to sit at the charging station for 45 minutes when a gas car can be fully fueled in 5 or less.

At home, my expectation is that a vehicle does not need a garage, so charging at home may present a logistical challenge (OR, weather-proof outside charging stations might also be deployed). Likewise for apartments and condos with shared parking.

I think all these issues can be overcome, but they will need some thoughtful approaches to ensure that EVs are reasonable for as many people as possible to own and use, not just the wealthy.

Tesla's existing batteries are expected to be more or less good for 500,000 miles, some already made it to 190,000 miles - average lifespan of a car - with under 10% degradation. 200K miles - easily, nearly every battery will do that with at worst 20% degradation. No, almost no car will ever need a battery replacement unless battery was bricked by gross mishandling (equivalent of fueling a gasoline car with diesel or something like that).

https://electrek.co/2016/11/01/tesla-battery-degradation/

My 2015 LEAF has about 10% degradation in under 11K miles. I doubt Tesla is 20x better.

It looks like the 2015 Leaf has a 24 kWh battery. The current entry-level Model S has a 75 kWh battery and it can have up to a 100 kWh battery.

That matters because battery capacity fade increases non-linearly with depth of cycling. If both a Leaf owner and a Model S owner drive 40 miles per day, I'd expect the Leaf battery capacity to fade more than 3x as fast as the Model S capacity, because of the deeper battery cycling. That's true even of the Model S 60/60D that had a 75 kWh battery software-locked to 60 kWh of usable capacity. The capacity fade is linked to the physical characteristics of the battery; the software unlock to a full 75 kWh just bought the option to discharge the battery more deeply for longer uninterrupted driving.

> It looks like the 2015 Leaf has a 24 kWh battery. The current entry-level Model S has a 75 kWh battery and it can have up to a 100 kWh battery.

So, a net price of about $1K/kWh of battery seems to be the going rate.

Correct. And when the Roadster was released, it was the price of just the battery.

> No, almost no car will ever need a battery replacement unless battery was bricked by gross mishandling

How would that be possible? Charge and discharge cycles are closely monitored and controlled by computers.

A real Soviet hero, with good training and some vodka, can do almost everything comrade... No fool proof is a real proof, people manage to break all kind of things. But, that is not a wear and tear - it is about as easy to brick a new battery as a 10 year old one.

Weatherproof outdoor charging stations are already here. The mall near me has a couple with no cover whatsoever.

The charging protocol is pretty well-designed for electrical safety already.

We are constrained by electricity production and the cost of materials needed to make batteries. Electric cars ain't happening anytime soon.

Not at all. There are about 250M cars in the U.S., each drives about 12K miles per year. Tesla makes 400 miles on a 100kwh charge, add 86% charger efficiency, 115kwh. That makes 860 TWh per year to replace all U.S. cars and light trucks with electric cars. U.S. produces 4060 TWh per year, so that's just about 21%. Sizeable increase, but that's fine: most of the charging will be by night when the car is parked at home, and by night, there is a big drop in electricity consumption so grid won't be overloaded at all. It will be the cheapest electricity in the world - just adding the variable cost.

Same thing now, with about 24 average mpg an existing car makes, eats 10 million barrels of oil per day (counting 81.5% average oil refining efficiency), or more than 20% of entire U.S. energy use - all kinds of energy, that is 7x more energy (ok, about 3.5x more given average efficiency of electricity production).

Even if that electricity was produced from oil on old-style single-stage thermal plants (a crazy idea - that would be the most expensive electricity in the world and no one is going to do that for sure), it would mean 2.7x savings. In reality, because most of that will be cheap coal and natural gas and renewables, there will be more than 10x savings.

> Every high tech product follows a law that it gets 20% cheaper when volumes of production double

Where to read more about this law?

https://en.wikipedia.org/wiki/Swanson%27s_law

Reply on news.ycombinator.com