Hacker News Reader

> For fuel-cell cars to work, you would need to have energy some place that consumes water and splits it into hydrogen and oxygen, releasing the oxygen into the air, and the hydrogen into containers that can then go into some sort of infrastructure such that individual cars can be filled up with it as necessary.

Yeah, sure. We know how to do this, its called electrolysis; we even know how to build fuel cells that can both generate energy and act as electrolysis systems.

> And if you think about the amount of gasoline infrastructure that exists, we'd have to pretty much duplicate that, or convert that, into hydrogen shipping infrastructure, which is a pretty tall order.

No, we don't. Because -- as many currently existing hydrogen stations, like the public one in Emeryville, do -- you can very easily do the electrolysis at the filling station. So you just need water distribution infrastructure (which is one of the basic necessities of civilization) and electricity distribution infrastructure (which, as you note, is likewise a solved problem), and then you don't need any additional hydrogen distribution infrastructure.

> So the story for electric car adoption is "You don't have to take trips to the gas station anymore!", but the story for fuel cell car adoption is "You have to take trips to a new type of gas stations that don't exist yet!".

Well, except that they do exist now.

There is absolutely no way that Hydrogen would be produced via electrolysis. The process is horribly inefficient but more importantly the cheapest way to produce hydrogen is through steam reformation of coal (or organic material to a way lesser extent). You spray coal into a hot stream, methane decomposes from the mixture, then the methane is converted to Hydrogen. It's by far the cheapest way to produce hydrogen and would easily cost a fraction of hydrogen produced from electrolysis. This process is also extremely CO2 intense.

I really have to wonder if this is just a show put on by Toyota. The viability of H2 production is dubious at best, and at worst if you consider our drive towards a low carbon economy.

And what about something like this to 'collect' hydrogen?

http://www.manchester.ac.uk/discover/news/article/?id=13372

When I see proof of concept I will be less skeptical. As well this only solves conversion efficiency, batteries are already +85% for conversion efficiency.

> There is absolutely no way that Hydrogen would be produced via electrolysis.

You mean, other than the fact that it absolutely is produced that way in existing fueling stations, and much of the drive for hydrogen as part of a clean energy solution relies on the use of electrolysis (or, possibly, splitting water with nuclear reactor heat) as the generation mechanism.

> The process is horribly inefficient but more importantly the cheapest way to produce hydrogen is through steam reformation of coal (or organic material to a way lesser extent).

Its certainly a cheaper way to produce it, but its not necessarily cheaper as an overall process, because then you do need an expensive distribution infrastructure. More, the necessary distribution infrastructure, even if end-to-end efficiency including distribution costs is better, has a much higher startup cost than distributed electrolysis.

Also, distributed electrolysis buys you directly in to any improvements in efficiency, greenhouse gas reduction, etc., for large scale energy production, whereas relying on steam reformation of coal, natural gas, etc., leaves you not only maintaining more dedicated distribution infrastructure, but with a separate carbon-intensive generation infrastructure that needs its own improvements.

Again how many stations exist? I'm not arguing the concept, I am arguing the feasibility. What is the cost for fueling at these stations? Solar panels generate electricity yet we don't use them to power industrial operations because solar expensive! Just like H2 from electrolysis won't be able to compete with industrial H2 production from coal. Also, what you're suggesting is using predominantly coal to generate electricity from a traditional plant (~40% eff), then transmitting the electricity to a station (90%), then creating H2(60%). Compare that to coal steam reformation (75%) and transportation (80%, being conservative), the latter is far more efficient and cheaper.

And what infrastructure do you need? A gas truck with a pressurized tank on it. The cost of hydrocarbon based h2 will displace electrolysis in large scale production.

And your last point about distributed load efficiency gains, electric cars are already able to balance load and feed into a grid. It's always going to come down to cost, I'm going to buy the cheapest h2 around and it's going to come from steam reformed coal.

> And if you think about the amount of gasoline infrastructure that exists, we'd have to pretty much duplicate that, or convert that, into hydrogen shipping infrastructure, which is a pretty tall order.

On the other hand, electrolysis can be significantly more distributed than gasoline, you don't need a limited number of refineries anymore. Hydrogen can be generated city-wise, at the station, or even (with further efficiency hit) at-home.

You need to pressurize the hydrogen to 5-10 kpsi which requires a compressor in every home, this cuts efficiency. Electrolysis is also very inefficient (%65 is usually cities as the upper efficiency[1]). Most importantly the cheapest way to produce hydrogen is through stream reformation of Carbon based fuel (high CO2 emissions req), making electrolysis cost prohibitive.

Hydrogen is not a great way to go, especially when you consider that batteries and electric motors are already very efficient. Working on energy density in batteries would be more cost effective then chasing after H2.

[1]http://gcep.stanford.edu/pdfs/hydrogen_workshop/MacQueen.pdf

The only game-changer I could see there would be if we could make some sort of in-your-home hydrolyzer, so that you just connect water and electricity to it, and it connects to your car to fill it up. That's on-par with the convenience of an electric car, if we ignore the safety issues of hydrogen gas leaks in urban neighbourhoods...

Regenerative fuel cells are cells with an in-cell hydrolysis system; if you could incorporate them, a fuel cell vehicle would have pretty much the same recharging UX as any other EV.

Even with stationary at-home electrolysis systems, if you don't store more than one fill-up in the system, there's marginal additional risk compared to having a hydrogen-fueled vehicle in the first place.

Great, that allows you to very easily compare that vehicle to one with a chemical battery, and evaluate it on cost, range, engine power, "battery replacement", etc! And then we can avoid all this "but it runs on water" nonsense and look at the actual capabilities.

>And if you think about the amount of gasoline infrastructure that exists, we'd have to pretty much duplicate that, or convert that, into hydrogen shipping infrastructure, which is a pretty tall order.

Imho, the only way to do this would be to have solar assisted charging stations in parking lots across the country.

We appear to already have the infrastructure for electric cars, and at low adoption rates that's true. But according to the utilities, that infrastructure is in no way ready for sweeping adoption of electric vehicles, so then too we would need to overhaul our existing infrastructure.

The difference there lies in that the electrical infrastructure can be overhauled gradually, as electric car adoption increases. We already have something that is useful to the current EV owners, and it can be made better. I'm also not convinced that the utilities are correct, current peak use is probably a lot higher than what a fleet of slow-charging vehicles would consume at night.

In the case of hydrogen infrastructure, we have nothing. If you're the first guy to buy a hydrogen car, you can't fill it up anywhere, there's an enormous amount of investment and work needed to get to that first hydrogen gas station, and I don't see that happening. Who would put up the money to make that bet?

> In the case of hydrogen infrastructure, we have nothing.

Well, more than nothing, but not a lot just now. Though more is in development.

> If you're the first guy to buy a hydrogen car, you can't fill it up anywhere

I'm going to guess that the first several tens of thousands of people to buy consumer electric vehicles in the US will be clustered around the places where the existing infrastructure is [1], so that this won't be even approximately true (there'll be an interesting trade off until infrastructure builds out in that FCVs will have a longer single charge/tank range than EVs, but a much more limited area of operations since there'll be limited refueling locations especially outside of the "home" areas.)

[1] e.g., in California, http://www.cafcp.org/stationmap

Oh wow, I did not know this existed! That's great, nothing's better than healthy competition!

We could take the gasoline, burn it to generate electricity and use that to charge electric vehicles. If you consider the massive efficiency advantage, we'd probably still come ahead over using internal combustion engines.

This is in other words a complete non-issue. But I'd love to hear plans on safeguarding inevitable hydrogen leaks in residential areas. Should we tell people to get rid of their garages?

I think you are confused. Gasoline drive is perhaps 30% efficient. Electric drive is perhaps 80%. Charging batteries I don't remember, but let's call it 80%.

So what you've proposed is:

    Burn gasoline (30%) * Charge batteries (80%) * Drive e-car (80%) = 19% efficient

It works better if you have a very high efficiency gasoline-powered generator (a 60% efficient generator gets you 38% efficient) but that can't really be done at the homeowner scale.

I guess you are underestimating just how terrible internal combustion engines are and overestimating charging + driving losses.

Consider that hybrids or HEVs can improve substantially on the efficiency of just a pure ICE. I guess the problem with using an ICE for driving directly is that you are often forced to run it at RPMs and states where it is nowhere near that hypothetical 30% efficiency, simply to get the acceleration the user demands.