Hacker News Reader

Comment by yazaddaruvala | Hacker News Reader

One thing to keep in mind is that the energy density of gasoline is a chemistry measurement.

The energy density of a gasoline power train is a different measurement taking into account chemistry and physics.

Specifically car engines can only turn 33% of gasoline’s stored energy into motion. Which still puts gasoline ahead in terms of efficiency relative to hydrogen or battery powertrains but reduces the difference a lot.

View on HN

An important point EV advocates often gloss over is that you don't get to burn battery. In aviation, especiaslly, this really matters - over half an aircraft's weight at takeout is often fuel. The later in the flight it is, the more fuel has burned off, the lighter the plane is, and the more efficient the rest of the flight will be. You have to lug that empty battery around at full weight.

Technically the battery gets a teeny bit lighter because of E=mc².

Over long journeys, those 3 nanograms really add up.

https://physics.stackexchange.com/questions/34421/does-the-m...

It's about a microgram.

So, in other words, you gained more mass from bugs splating on the windshield than you lost in the battery.

Don’t worry, we’ve been working on that:

https://www.science.org/content/article/where-have-all-insec...

But in the winter, as the cabin warms up, you have fewer air molecules to cart around.

The simple sum is to use half the weight of the fuel as though you have it with you all the time. That's not quite accurate but close enough.

Well now you've just made me an engineering scenario of aircraft equiped with drop tanks full of heavy batteries used to extend range. Drop tanks that they let go mid flight 'bombs away!' as they discharge.

So all we need then is parachutes, some kind of GPS based glide guidance control, and a network of aviation battery collection stations :)

I guess overseas flights will just have to absorb the cost, unless it happens to be economical somehow to operate floating stations for the purpose of collecting dropped batteries.

Rocket Lab's Electron rocket actually does this with the batteries that power its electric turbo pumps.

> The electric pumps are powered by lithium-polymer batteries. The second stage uses three batteries which are "hot swapped", two of the batteries are jettisoned once depleted to shed mass.

https://en.wikipedia.org/wiki/Rocket_Lab_Electron

Another interesting dynamic is that mid-air refueling is possible (though rare), where as for batteries that's impractical with current tech

Simple. Get hit by lightning.

Ah, the Back to Future scenario!

Solar panels on the outside of the plane? They're above the clouds for the most part.

> can only turn 33%

Energy density of gasoline is literally 12900 Wh/kg and 9500 Wh/L, Li-ion(not combusted) is ~250 Wh/kg and ~700 Wh/L. That's 50x/14x raw and 17x/13x at 33% efficiency without hybrid/ICE optimization gains.

(23222915000 Wh/kg and 370000000000 Wh/L for Pu239 btw, source[1])

1: https://en.wikipedia.org/wiki/Energy_density

Had a double take on reading Li-ion (not combusted) and then immediately went looking at EV battery fires. A rabbit hole of solid fuel and it's considerations followed. I have become a trained monkey. Well done internet.

And thank you for the 'x' factor. Had never seen that, outside of everyone stating ICE is only 33% efficient.

Thank you for the math. But I don’t think you really tried to understand my point.

Firstly, the newest solid state batteries (relevant because that is what I was responding about) are 2x the density so that is quite good (8.5x/6.5x). Which is already pretty close. They are also going to get better through iteration likely to 3-5x or so (but let’s ignore that for now).

Secondly, the power train also includes regenerative breaking, reduced weight from transmission, in hub motors (reducing axcel weight), reduced center of gravity allowing better drag profiles, no external coolant systems, no engine oil compartments, no fan belts, etc.

I’m happy for you to keep doing the math, but the reality is XXX km or miles per “full tank” is the best measurement because it takes all those factors into consideration.

Oh I'm not so against EV, I'm just being pro-nuke as I personally always has been. Honestly I don't get why some sort of EV supremacy theory has to be defended at all costs. Those are just facts that density and recyclability of Li-ion is just atrocious. Gasoline particles tho NOx bad for health yeah I get that part.

Seriously, if we could just build 2-3 air-cooled Plutonium fission per each megalopolis core and use e-Fuel on existing hybrids we'd solve everything in modern society ever. The reactor could double as cityscale thermojet ventilation system as well, eliminating heat and atmosphere problems once and for all. It will of course work as a training exercise for coming real Space Age in which humanity builds bunches after bunches of cylindrical habitats fabricated and populated in space. The only two roadblocks towards that path are some NIMBYism/NPT fears and cost.

Fair enough! I’m just on board with the electric ground transportation thing.

I’m very happy for there to be really nice and safe fission reactors everywhere for grid scale generation. Including efuel generators for air and space transportation.

If you dont like him doing the math why do you bring up the thermodynamic efficiency instead of the max range? It sounds like you just like math tidbits that make EVs sound better than they are. And dislike math that shows gasoline is superior from a purely practical standpoint (its far worse environmently)

I remember people saying that hybrid vehicle used ICE not to generate car scale torque, thus being build differently to improve the efficiency (mostly due to better control of thermal constraints ensuring better chemical reaction, higher conversion and less pollution). Do you know how true that is ?

They might be referring to the Atkinson cycle engine, which is really just adjusting the valve timing. Lower power but higher efficiency.

https://www.autoweek.com/news/green-cars/a1832081/atkinson-c...

No, it means that you can leverage a more efficient part of an engine’s torque curve by supplementing with electric power at the low and high end (since motors have a flatter torque curve).

Not just that. It does require Atkinson cycle to reach the efficiencies they have. It just means that the compression stroke is shorter than the power stroke.

Basically with that cycle they can get massive compression ratios. As an example newer Toyotas are at 14:1 ratio with 41% thermal efficiency. That’s close to diesel territory on a gasoline car.

Naturally when you floor it it goes to Otto cycle and efficiency drops.

And, in the case of Toyotas eCVTs, keeping the ICE in the most-efficient-rpm sweet spot longer by allowing the ICE RPM to be pretty much independent of the wheel RPM. It's somewhat strange because RPM noise barely correlats with speed or acceleration anymore.

Also: start-stop-automatic for free ... I am still somewhat spooked when the motor just turns off going 70 km/h because the ECU decides that the ICE is not needed anymore.

Not true at all. Torque is basically just a number you can manipulate with gearing.

This is the “series hybrid” or “range extender” approach and the issue is that all of the engine’s power needs to be converted into electrical power and then back into mechanical power, even when running the traction motor directly off the generator output. This causes significant losses in efficiency which more or less cancel out the gains from being able to optimise the engine for a specific constant speed.

As of this year, and limited to three models exclusive to Japan: yes this happens.

The new 2023 model year Serena from Nissan uses a generator optimized engine in a series hybrid setup. The engine runs at a fixed rpm (except in extreme situations like accelerating a bunch on the highway going from 110 to 140) then it switches to a second higher rpm mode. I bought one, it's a lot of fun to manage the engines heat and battery room. My record is 38km per liter using local driving up and down our hilly community. Pretty incredible when you consider hills are the natural enemy to none hybrid millage.

The car accelerates like an EV. And with Japan's expensive electricity I'm paying only about double the millage cost of an EV. Despite the full size minivan costing us about 20k USD new.

Most hybrids, and all Toyota ones, are not using a special fixed rpm engines. They use a regular motor with different timings. The Nissan fixed rpm engine is actually less powerful than the electric motor powering the wheels. Thus at max acceleration you can deplete the battery faster than it is charged.

If we are just using it as a generator, a turbine engine makes more sense. Very efficient at fixed rpm. Many trains use that sort of hybrid setup. But if you put in enough battery to warm up a small proane fuel cell to warm up a larger propane fuel cell you could just fill up at the propane pump and still get battery efficiencies but with a lightening tank. And they think they have propane generation from environmental co2 over 90% efficient now too.

Turbine in a car is completely impractical. Some tanks, like M1 Abrams use a turbine and they are painfully loud and the exhaust is so hot it can literally set on fire anything too close to it. Imagine sitting in a traffic jam and lighting up a car behind you :)

oh right, I remember some guy mentioning fixed RPM helping a lot.

The Chevrolet Volt works like that.

the heat itself is useful in winter to keep the system regulated and heat the passengers

Power train don't pour.

I personally wonder how much more the 33% would be if we had continued the focus on ICE. I believe there is no business case anymore to invest significantly in advancing technology, if everyone wants a new shiny toy.

It’s been 200 years and untold amounts of R&D money thrown at this incredibly lucrative, huge industry. It’s not budging much beyond what it is.

There's not much more you can get out of current fuel blends. Methanol is potent but very expensive to mix, the even more expensive nitromethane is so powerful it strains the science of metallurgy and thus is some of the most dangerous stuff to handle, and hydrazine is so energy dense but so unstable it's used as rocket fuel. Porsche's E-Fuel is just barely more efficient than gasoline because of the density at point six megajoules a kilogram, versus point five four for one hundred octane gasoline.

The major issue is parasitic drivetrain loss from having to push the cylinder which pushes the cam which turns the flywheel which friction fits against the clutch which friction fits against the transmission flywheel which turns dozens of metal gears which rotates the driveshafts. Direct drive implementations where the engine's output shaft is mated directly to the axle are much more efficient and have been used in massive ships, but are entirely impractical for use in passenger vehicles. There's the Otto, Atkinson, Wankel, Scuderi and diesel cycles for liquid fuel combustion, and all have tradeoffs leaving them competitive with eachother instead of having a clear best solution thanks to everything else required to transfer the power from the initial combustion.

In short, we've hit the mechanical engineering dead end for massive efficiency gains with ICEs. It's all chemical from here, and thus gets much more expensive to run.

ICE is kind of a dead end, you convert energy from chemical to thermal to mechanical and every conversion is lossy. You want to use chemical energy directly instead, like some kind of "hydrocarbon/air cell" and we don't know how to do it. At least not with mixed hydrocarbons.

The theoretical limit is 50% according to Wikipedia, but there's other factors being optimized for as well aside from efficiency:

https://en.wikipedia.org/wiki/Engine_efficiency#:~:text=Mode....

Modern ICEs are already insanely sophisticated, in pushing the boundaries with all kinds of tiny tricks, just to get a small percent extra. There is only that much you can get by changing the valve timing, freevalve, compression ratio, air boosters, electric turbos, fuel injection methods, mixture ratio, how the fuel get sprayed most efficiently into the chamber, predictive hill cruise control, balancing rods and whatnot. Everything has been tried and the improvements for each new trick is only marginal.

Something more radical has to be done about the entire crankshaft and a 4 piston design. Believe me, lots of smart people have tried that as well, for many decades. Because the payoffs if you did find something are huge.

There are lots of YouTube channels needing into all of this. Engineering Explained is a good one if you want to understand the level of how advanced modern engines really are.

That’s where competitive motorsports development helps:

Mercedes has devised an engine that achieves more than 50 percent efficiency.

https://www.roadandtrack.com/motorsports/a15049580/mercedes-...

Reply on news.ycombinator.com