Hacker News Reader

Comment by SigmundA | Hacker News Reader

Definitely something I missed when I was more skeptical about BEV's and Tesla a few years ago.

It's not about some new breakthrough to make batteries as cheap or energy dense as gas, it's a steady YoY improvement and steady decline in prices.

Lithium batteries where developed in the 70's didn't get released until the 90's and has steadily improved since then with a big cost decline in the last few years.

They are currently up around 300 wh/kg with a theoretical peak of around 2600 wh/kg with sulfur cathodes there is still plenty of headroom for further improvement. At around 2000 wh/kg it roughly equivalent to a diesel drivetrain power density if my math is right.

View on HN

Years ago I went to a presentation by Tesla in the Bay Area because they were going to let people sit in one of the early prototype Roadsters.

They basically explained that they were betting the company on a "Moore's Law" of batteries. The presentation included a graph of battery capacity doubling every 7 years.

If you look at the chart in this article, it looks like energy density has been on a linear rather than exponential path for the last three decades.

I think that moore's law was a huge contributor to the success of silicon valley.

It let people predict the future 2 or 4 or 10 years out, from the conservative people in charge of finances to the wildcats who are creating new technology that will need something cheap or fast or just viable to match up with when they ship.

> developed in the 70's didn't get released until the 90's

Patents last for 20 years.

Oddly enough it doesn't prevent semiconductor technology from advancing - don't know if it's from cross-licensing or something else?

Cross licensing definitely had a part in it. Some of that was necessity, either practical or contractual. In the 70s, before Projection Masking as a fabrication technique came into play, more complex designs could have yields as bad as 10%.

As a result, 'Second Sourcing' was more frequent, and while Intel and AMD had some Cross Licensing before x86, the main reason the agreement expanded to include x86 was because IBM was still requiring Second Sources in the 80s.

I know at one point, Chevron had it's tentacles on some NIMH patents via their control of Ovonics. Not sure if there are/were any similar situations for Lithium Ion technology.

Also the price fixing done by Lithium companies between 2000-2011. That likely hampered adoption and perhaps improvements.

When I've tried to work out estimates, currently EV's suffer a 20-25% weight penalty over gas/diesel. Part of the weight penalty is due to cascading effects. A battery plus and electric motor weighs more than a gas tank plus engine/transmission. Which means the structural elements have to be heavier.

I think if batteries energy density increases by 50% the weight penalty goes away.

All true but on the good side the weight distribution of the BEV batteries is much better than the weight distribution of the ICE power train. So it's not all just about the amount of weight.

Is weight distribution really helpful for anything other than handling?

This is also why I've become a bit more skeptical about Nuclear. Renewables + batteries are pretty mediocre right now but if I'm to bet on something for the next twenty years, it's that.

I think nuclear is still really important and has a lot of room for improvement. It's hard to beat its output stability and energy density while emitting no carbon.

Just have to figure out how to keep it safe.

Just have to figure out how to make it cost less than other energy sources.

Small modular reactors are probably our best bet here. Centrally manufactured, fail-safe designs, short on-site construction times and time-to-first-power-output. It all depends at this point on if the regulatory environment can be improved soon enough for it to matter.

Nuclear is already incredibly safe. Installing solar has probably killed more people than nuclear ever has.

Nuclear is ok for district heating because there aren't many alternatives. Burning gas 20% of the time as backup is not as big of a problem as many people make it out to be. The atmosphere doesn't care when you pollute, just how much, so intermittency is one of the biggest red herrings of all time.

If we had grids with 80% renewables you can start thinking about grid storage but right now it doesn't matter that much.

2000 wh/kg (crazy units, but I guess this is conventional in some fields?) is 7.2MJ/kg. Diesel fuel and gasoline are something like 40MJ/kg. So same ballpark.

Also, electric motors can easily reach 80%-90% efficiency, while ICE engines in cars typically get 20%-35%. Even in principle ICE engines are thermodynamicly required to be inefficient as combustion itself produces heat, and there is a limit to how much useful work you can extract from a thermal gradient.

You still need to account for the discharge efficiency of the battery (which I cannot find numbers for).

For overall energy usage, electric still still has whatever inefficiencies exist in the generator (and charging inefficiency), but those don't effect power density

The inefficiency inclusion from the generator for electric is somewhat unrelated to an ICE. No ones including the inefficiency of extracting and refining the oil, the huge amounts of time/energy required within the earths crust to create the oil, nor the energy conversion from the sun into organic matter in the first place. It seems a lot of the details on the fossil fuel side get left out.

Discharge efficiency varies from battery to battery. Generally speaking, it's a 90+% efficient process (as is charging). However, there are certainly exceptions which are more or less efficient.

Transmission costs probably eat a bigger portion of EV total efficiency. Those depend on how far you are from your power source and what voltage your power is transmitted at.

That said, a thing EVs do that ICEs can't is accept power from any source. That means that only a fraction of the fuel is actually generated from a fossil fuel.

One other thing to consider is that as power density goes up for batteries, the amount of required batteries go down, ultimately reducing vehicle weight and improving range.

Assuming batteries were ever 10x their current power density, you'd probably not see cars with 10x range, rather, you'd see something like 400mile range cars with the same total power capacity of yesteryear cars. Pretty much all because you've cut the pack weight down by 1/10th (probably less because you need a lot less infrastructure around maintaining temperature for 1/10th the number of batteries).

Also the ability to recover kinetic energy during braking which is lost with ICE engines.

I wouldn't say you would really count battery discharge efficiency, usually a batteries energy density will be how much useful energy you get out on discharge so it already accounts for charge/discharge efficiency.

Same with electrical generation with generators or whatever. One of the best things about BEV even if run off fossil fuels is that power plants don't care to much about space or weight, they can do multiple levels of heat recovery and whatever emissions controls and get higher thermal efficiency with lower emissions and all that complicated gear is not being carried around in your car nor maintained by you. On top of that now you are decoupled from energy source so changing fuel oil to natural gas or mixing solar, wind or nuclear doesn't change your vehicle.

Most of the energy of combustion is wasted through heat loss. One way to increase efficiency would be a secondary system that harnesses the heat output.

Batteries are usually wh/kg since wh is what you use for battery capacity.

You have to watch it with fuel energy density since it normally doesn't take into account the efficiencies of converting it to useful work and the equipment weight needed, its just the heat energy released when burned. Electrons probably have a really high wh/kg.

A battery has power leads coming out of it to give you electricity and modern electric motors are above 90% efficiency at converting to mechanical energy.

Diesel needs fuel tanks and a heavy engine and more complicated and heavy transmission as well. The engine itself is probably under 40% efficient at converting those MJ to horsepower.

Same goes for fuel cells they are similar in efficiency to ICE and have weight and say with hydrogen storage you need heavy tanks and protective structure to hold safely. Also with hydrogen it has low volumetric density so even though it has good MJ/kg it has low MJ/L relative to say Diesel. All current fuel cell cars are similar in weight to a BEV and seem to have just a slight range advantage with big power disadvantage. Now look at the efficiency of creating hydrogen at I am very skeptical of it becoming mainstream anytime soon.

Tesla's 50% reduction in cost of their new batteries is made up of 50 different things

Reply on news.ycombinator.com