50% round-trip efficiency and 3x the density of li-ion sounds pretty good, actually. Assuming your efficiency number is the entire cost of water -> compressed H2.

Yes, also given that electric motors have an efficiency of 90%+ (or thereabouts) and internal combustion engines only of 20-30% IIRC.

You're comparing apples and oranges. The around 40% efficiency modern ICEs have is thermal. The 90%+ going from battery to electric motor is not thermal. You're just converting high-exergy to high-exergy (kinetic) at that point. The interesting, potentially low-efficiency part happens before the battery, before charging it in the first place. Where does the electricity for that come from, at what efficiency? Perhaps there's a thermal engine involved, at said 40% thermal efficiency. Then you're at 0.4*0.9 total, and the EV looks pretty dull suddenly. (Just an extreme counter-example)

Your comparison reminds me of advertising for electric water kettles. 95%!! 97%!! Wow!! ... or not, because you're converting pure exergy (highest quality energy) in the form of electricity to heat. You're dissipating it (heat), using a resistor. That's child play and the efficiency is meaningless. The other way around is the interesting part.

Modern ICE are hitting 40%.

I don't know why you are being downvoted. You are correct, and some ICEs (large, maritime) are hitting 50% even. For a thermal engine, that's incredibly good (it's not about the gap to 100%, but the gap to Carnot efficiency).

There's a strong contingent here on HN that believes tha continued used of ICEs equates to the imminent death of the human race.

It's also a likely wealthy contingent that imports all of its food, flies everywhere, etc.

On what grounds do you draw that connection?

Previous threads, comments & votes.

... except the cost of the ICE drivetrain, certainly to hit 40% (that takes turbos and other complexities to the basic ICE, correct?) is about to lose to the cost of the EV drivetrain. It's basically at price parity based on what I see out of Tesla and what Lucid/Rivian/VW/GM are targeting. In three years? Less.

And a decade from now? EVs will probably be half the cost of an ICE.

Because of battery energy density, EVs still only work within a certain size&weight range.

ICE work in a much broader range of applications and at wider temperature extremes. They aren't going away.

Toyota's hydrogen tanks looked promising: 300mi range and 3 minute refill time. It's disappointing it never went anywhere.

We really aren't that far from a plausible charging setup that can do 300 miles in 10 minutes. And that would be perfectly adequate. The infrastructure to support it would certainly be less difficult than trying to get hydrogen available everywhere.

I feel like what a lot of people miss is that with an EV if you have a garage or dedicated parking spot you’re leaving home with a “full tank” every single day. You only need to use a public charger if you’re on a road trip.

That is the big 'if' for the majority of population in European cities. Pretty convenient for suburbs though.

Yep, having to charge them at public charging stations is pretty shitty, takes a long time, and a lot of time they're not available to use.

I used to work in a large business building a few years ago, and they had 4 charging stations right next to the main entrance, and having an electric car then was great... the "plebs" would have to park behind the building or in the garage under, and you'd have priority parking infront. Then, the number of electic cars rose to above 4 there, and first 4 that came, parked there, connected the chargers and went to work. If you came 5th, there was nothing you could do, except maybe if you knew one of those 4, politely ask them to move their car during lunch, so you can charge yours too.. but most of the people left their car there until end of work at 3-5h in the afternoon.

Ive charged plenty of times on road trips on both the Tesla and Electrify America networks. Charge time is usually under 20-30 at places and times I’d already be stopping to use the bathroom and stretch my legs.

In regards to the chargers in your work garage: most modern charge systems let you set idle fees. If you install anything worthwhile you could easily configure it so people have to move after their car has been sitting fully charged for an hour.

Plenty of apartment buildings don't have a dedicated powerpoint per parking spot.

This can be fixed but in the meantime it is a significant cost for the building. And you bet that early adopters putting in regular powerpoints will be trumped by later installs of dedicated, secured charging stations.

I am not jumping to an EV any time soon.

> And you bet that early adopters putting in regular powerpoints

A 110V AC outlet is completely inappropriate and inadequate for charging a BEV; for example, it would take three days of nonstop charging to charge my car from a 110V outlet.

BEV automakers include the 110V adapter as an emergency stopgap more than anything else - I suppose they intend it to be used to help you get to a DC charger: but even assuming I'm no more than 100mi from a Supercharger then I still need to be plugged-in to 110V for at least 24 hours to get the 30kWh level I'd need to travel 100mi - and who is going to let someone borrow a days' worth of 1-2kWh electricity like that?

Could an apartment building even support say 50% or 75% of a typical sized apartments buildings parking garage's worth of 'supercharger' like installs? Later on times a whole cities worth of apartment buildings.

I'm not familiar with this field but I'd imagine this would take some special arrangement with the power companies. But so far that seems to be the only solution at the moment.

Without apartment buildings offering this I don't see how it could be widely deployed without changing how cities are built and transportation is deployed. Even if all the buildings offloaded to external Uber-style autonomous fleets they need to be charged somewhere and at a massive volume. And in my city 90% of new housin development for the last two decades seems to be on apartment buildings.

Preface: DC fast-charging is "Level 3". A 240V ("clothes dryer connector" in the US) or hardwired wall-connector is "Level 2" and in the US is typically 15-50A - if you have a modern home with plenty of free 'leccy panel spots then you can do 90A. A slow 110V or low-amperage 240V (for outside North America) is considered "Level 1". The supply of DC fast chargers is not open to the public at all: only Tesla and a tiny number of third-parties maintain Level 3 infrastructure, and they all require outhouse-sized transformer units for each car space, separate from the car-connector part.

---------

Level 3 / DC fast-charging is not intended for long-term and repetitive use: it's only intended for road-trips and that's it - that's because DC fast-charging is bad for your battery.

I primarily charge at home using my 90A wall-connector - but I still rate-limit the charging to ~60A to help conserve the health of my battery-pack. Not that it matters much: since Covid and WFH my driving has dropped drastically, from ~30,000mi/yr to less than 5,000mi in 2020, and less than 1,000mi so far in 2021. (I got my Model X in 2018 and it's still got under 20,000 miles on it, lol).

But you raise an excellent point about apartment complexes and parking-lots not being able to feed dozens or more charging spaces. Yes, it's a problem, and no, I don't see a viable long-term solution.

Other than apartments, cities like London, etc, have built car charger ports into streetlamps so people with street-parking can charge their car at Level 2 speeds using that pre-existing infrastructure - I'm unsure to what extent it's a public-private-partnership vs. public, but the opportunity to profit from car-charging infrastructure means pushing the cost-per-mile up above the cost of gasoline/petrol - which undermines the economic-and-environmental benefits of BEVs (especially when you consider the dreaded fuel-tax on petrol will eventually be extended to electricity if governments don't introduce usage-based road taxes). It'll be mess, methinks.

Also, everything is terrible.

> if you have a modern home with plenty of free 'leccy panel spots then you can do 90A.

How so? J1772 itself maxes out at 80A, and I don't even know of any production EVs that support that: most max out at 48A or lower. And even for Tesla's connector, the highest I've ever seen (which isn't even offered anymore) was 72A.

> And even for Tesla's connector, the highest I've ever seen (which isn't even offered anymore) was 72A.

The wall-connector in my garage has a 90A sticker put on it by the state-licensed electrician who installed it - and I can drag the rate-limit slider in my car's UI up to (and beyond) 90A. I'd take a picture and post it to show you, but HN is famously text-only.

Yes. In the South of the United States most apartments have individual aircon units with a 40-60A 240v draw that run nearly around the clock during the summer. It would not be unreasonable to imagine simply doubling the service to an apartment building and running a large deployment of level 2 destination chargers.

Yes, it would be unreasonable. Electrical equipment doesn't scale proportionally to power levels. It might cost 10 times as much to install a 480V connection than it does to install a 240V connection.

OTOH, plenty of people are getting along fine with 120V. Yes it would take days to charge from zero. But the average person drives 37 miles a day, and a 120V plugin can replenish 37 miles overnight.

I wanted to buy a Tesla and I asked my apartment building management about getting a charging spot. They have 12 spots already (it's a 150 unit building), and they're all taken...and they can't install more unless they run new lines direct from a power station 2 miles away and install $100k of equipment to receive it. So they said no.

Now maybe that is bad planning on their part, but I'd bet they're far from the only building in town that didn't build to accommodate a rapid growth in EVs.

I agree that apartments are CURRENTLY a problem.

But we're talking about a centralized high-density residential building vs sprawled out suburbia. It HAS to be easier to setup electrical charging en masse in an apartment building.

Once your downtown hi-rise loses market competitiveness b/c you DON'T have charging, it will change VERY quickly. The charging infrastructure will be a capital outlay with depreciation and there is about a 99% chance there will be some decent government and utility subsidies that come with it.

I agree none of that (the market force or the subsidies) exist currently, but I'm guessing in three years it will be happening.

We're not though. For the ease of calculation, let's take a 100kWh battery. To fill it up in 10 minutes, you'd need to charge it at 600kW. There isn't even an electrical standard which would allow this, not to mention having to somehow supply over half a megawatt(!!!!!) Into a single charging point, and somehow cool it and the battery. The best of the best charging systems draw "only" 250kW and even that is either impossibly rare in CCS world, and in Tesla world it's shared between chargers so if you have few Teslas charging at once no one is getting 250kW.

I admit 600 doesn't sound like a lot more than 250 but we're nowhere near those charging speeds, not even in a planning phase.

Having said all of the above - the 250kW charging as-is is already stupidly fast as "good enough". And charging at home will always be king even at slower speeds.

> The best of the best charging systems draw "only" 250kW

Actually Electrify America can charge at 350 kW. They have to actively cool the cables to keep them from overheating.

> in Tesla world it's shared between chargers so if you have few Teslas charging at once no one is getting 250kW.

I think with v3 Superchargers, this is no longer true. There is a megawatt allocated per 4 chargers. I'm not sure how much this applies to the really big stations, because it seems they could be using multiple megawatts of power.

Tesla superchargers are delivering 250 kW today, at ~380V. The largest of these is like 40 stalls; they are not shared (you're thinking of the 150 kW v2 chargers that split current between two stalls). Hyundai is about to ship a car that can take current at 800V. (There are 350 kW CCS chargers too, though those seem to be poorly exploited and hard to measure in practice with real cars; but I've watched my Model Y sink 250 kW with my own eyes.)

So just naive application of already-deployed technology can get you to ~500 kW. You don't think someone's going to squeeze out a thicker cable, really? No, this is doable. The real limit is how much current the battery can sink without degradation at the top end of the charge curve. The electrical stuff is kinda trivial.

I think Porsche also has an 800V car.

I think the power-delivery aspect is solvable. From the perspective of a power company, EV chargers seem like ideal customers: you can run high-capacity power lines to one big customer instead of running little power lines to every residential customer. Though on the other hand, if you have a fast charger that's seldom used, that's a lot of infrastructure in place that's sitting idle most of the time.

Coming up with batteries that charge quickly is a different matter. Batteries are getting better, but it seems that part of the solution so far is to just have batteries that are so big that they don't need to be charged very often. Charging a 30kwh battery in an hour only gets you half the range in the same time as charging a 60kwh battery in an hour.

Maybe if we keep increasing the power of the chargers, we'll eventually ditch the cables and move to something like parking a car over a pair of big copper plates that extend to make contact with a matching pair of copper plates on the bottom of the car. I could also see some kind of system where the car has a fixture for attaching a coolant hose, and the charger pumps cold water through the battery while it's charging.

At some level, though, I think this is a trying to engineer around a problem that shouldn't exist in the first place. If we had electrified roads that are capable of charging vehicles while they're moving, then battery sizes and charge times would be less of an issue, and we wouldn't need to have cars that devote a quarter of their mass or so to batteries. It would only really need to be on major roads, and probably only in short sections at regular intervals.

> I admit 600 doesn't sound like a lot more than 250 but we're nowhere near those charging speeds, not even in a planning phase.

Chargers substantailly bigger than that are definitely in or past the planning phase, for example: https://www.teslarati.com/tesla-semi-megacharger-charging-po...

1MW chargers exist today (for electric buses)

4.5MW systems are being planned

https://cleancities.energy.gov/files/u/news_events/document/...

I think anything within a factor of two is just a mild engineering challenge away. The latest chargers are 350 kW. So use two of them.

With current battery tech we'd probably need a 600 mile pack to make those charging speeds feasible for 300 miles in 10 minutes. That's a big pack, but again, just a bit of an engineering challenge, not revolutionary.

Though I otherwise agree, home charging is where it's at. Starting every day with a full charge and never going to the gas station is really nice. I hope we can quickly scale up the infrastructure so everyone can enjoy it. Right now maybe 50% of the US population is able to easily charge at home.

What does all that fast charging do to batteries though?

I'm surprisingly ignorant on this. ICE engine is fast-recharge as baseline; feels like batteries will forever have slow charge as baseline, fast charge as emergency occasional compromise that you know is killing your durability. This is where hydrogen seemed an exciting alternative. Am I wrong?

> What does all that fast charging do to batteries though?

Jury's still out on that. Some manufacturers (GM comes to mind) are claiming their next generation batteries won't degrade from constant fast-charging.

I have been warning my apartment building committee for a while now they need to set aside money for a "every carpark has it's own powerpoint" project.

Charging at home can happen. But this will be a hassle for a lot of people.

That’s not adequate at all! The lines at gas stations would be massive. Charging via electric at home or in a parking spot seems a lot more convenient, even if it takes a few hours.

10 minutes isn't that far off what people routinely spend today to fill a tank of gas. Especially people with trucks, which is a large part of the US.

Over half the population probably has the infrastructure to home charge right now, so this does not need to replace gas station pumps 1 for 1. I was not suggesting that all of us charging at home need to start using public chargers.

> 10 minutes isn't that far off what people routinely spend today to fill a tank of gas.

I don't want to "uhm, acktually", but it's less than 5 minutes: https://www.quora.com/How-much-time-does-the-average-person-... - and that's including time spent faffing around with the awful pay-at-pump machines' UX.

I admit my experience is non-typical, as the only gasoline vehicle I own is a truck with a big tank, but 10 minutes is well within normal. For cars with 15 gallon tanks, 5 minutes is probably reasonable.

But in any case, 10 minutes is fine no matter what. If you are willing to spend 4x as much for fuel in order to save less than 10 minutes per week when you stop for fuel, then the problem isn't the charge time.

Filling your tank with fossil fuel is 5 minutes wall clock but also 5 minutes where you can't do anything else. If you need to pee or eat that's a few more minutes.

Filling your battery might be longer 10-15 minutes, but (at least for Tesla superchargers) about 20 seconds of your time, for the remaining 10-15 minutes you're free to do whatever you want.

Worth noting that "filling a gas tank" is something that makes sense to do every time you fill up at a gas station since that's your obly option to get energy.

It doesn't make sense to do the equivalent of charging an EV to 100% basically ever, so its a weird thing to compare against. Even very long distance trips will aim to bounce between 20-80% to minimize total charging time.

The big difference being that you don't have a small gas pump at home, in your office or at the supermarket that will fill your tank cheaply for you, and so you only need enough fuel to ensure you get to the next time that is available to you.

I think going forward, EV charging locations aren't going to be like gas stations: they'll be on bigger lots, maybe on the edge of town or between towns where land is cheaper. Or they'll co-locate in shopping center parking lots, which you see now.

Converting an existing gas station to EV charging just wouldn't work in most cases because they don't have the space to charge, say, a hundred cars at the same time.

> Converting an existing gas station to EV charging just wouldn't work in most cases because they don't have the space to charge, say, a hundred cars at the same time.

On the contrary - have you ever noticed how gas-stations don't have an _upstairs_? Without the massive fire-and-explosion risk from pumping gasoline, a typical urban filling-station could be converted to a multi-storey charging plaza at a a reasonable expense, imo.

I remember there was a brief fad in the 1990s to quickly build additional storeys of parking spaces on bare concrete/asphalt lots using metal superstructure: it worked and it was relatively cheap, but the results weren't long-lasting (my university tore-down their metal-based parking lot after it had been up for about 10-15 years IIRC) - I assume something similar could be done in this case.

The issue is that for a BEV, charging speed is directly coupled to energy demand from the grid. Using nice numbers, if you want to charge a BEV with a 100kWh battery in 6 minutes, you'd need a 1MW charger, which is a non-trivial amount of power that needs to be dispatched. Multiply this by the number of cars charging at any given time, and you have a recipe for huge demand volatility, which is a problem if our grid is going to be powered by non-dispatchable sources like wind and solar. The alternative is having chargers be so ubiquitous that everywhere you park you have some form of trickle charging, so that people living in apartments don't have to sit at a fast charger once a week.

Hydrogen gets around this, because eventually it's going to be a clean source of consistent demand that takes advantage of the fact that renewables require overproduction. A country like Denmark meets 40% of its electricity demand from wind, but produces up to 140% during periods of oversupply. If Denmark wanted to produce 80% of its electricity demand with wind, and do so with 90% confidence, it will generate more than 100% of its needs on average. Even if you wanted to look at the EU instead of Denmark, the same will still apply, because in order for two locations A and B to able to cover for each other when one has excess and the other doesn't, it must also be true that there will be times where no-one has a shortage. Since we want to guarantee our ability to provide X units of power at any given time, we need the grid as a whole to almost never have a shortage in aggregate, which means we will generally overproduce.

To give you an idea of the scale of this; California already curtails around 2.5TWh a year, which is enough to charge anywhere between 25 and 40 million BEVs. And that's one state, today, at relatively low-moderate renewable penetrations. It doesn't really matter if the round trip efficiency of hydrogen is lower than Li-ion, because we aren't going to have a shortage of electricity by any means. We're already throwing the stuff away.

Combination of needing lots of infrastructure, and lack of political will.

If US decided that instead of waiting for green tech to emerge, we would just simply stop burning fossil fuels and revert to an Amish standard of living until clean fuels are developed, we could have converted to Hydrogen already

We don’t have to convert to an armish standard. We could have just massively reduced car transport and improve electric rail and walkability.

I’m always amazed at how Americans consider driving to be a simple fact of life where you either drive daily or go back to throwing spears in the jungle.

You are regurgitating a clichéd argument.

> massively reduced car transport

How the hell could enough people be convinced to spend say 10% of their income on public transport instead of their car and petrol/charging? https://www.statista.com/statistics/748911/us-average-per-ca...

> improve electric rail and walkability.

So click your fingers and you have cheaply and immediately magically redesigned the distribution of residential and commercial property to suit public transport and/or pedestrian/bike access?

Finally, ignoring the above, building out that infrastructure and other changes would cost trillions, which indirectly causes trillions of dollars worth of extra CO2…

Your argument is as trite as saying “why don’t we just stop producing CO2?”.

Edit: disclaimer: yes I own a car and I also use public transport when convenient (I was even 100% carless for a few years, by living close to work and bus route was accessible).

Well it seems you can switch a lot of people to work from home vs commute quite easily.

As long as Jeff Bezos gets to keep his new half a billion dollars yacht, I’m sure the elite would be happy to see the rest of us finally return to serfdom where we belong.

That's what's coming now.

Could be

It's 3 minutes assuming the fuelling station is at peak pressure - 700 bars. After the first fill up it needs several minutes to pressurize. Also such a facility costs at least $1mln and is unlikely to ever get much cheaper.

> Also such a facility costs at least $1mln and is unlikely to ever get much cheaper.

How much does a conventional gas station cost? I would expect several hundred thousand, at least.

A more relevant question is how much does a single gas pump or EV charging station cost. A gas station can easily have a dozen pumps for ~500k.

> EV charging station cost.

A commercial-grade Level 1 charger (US domestic 110V 15 amp), if they even exist anymore, is probably on the order of $1000 or less - if it breaks buy a no-name replacement from Alibaba until it catches fire.

A professionally-installed Level 2 charger without any payment system infrastructure - so one that's basically free-to-use and taps directly off your AC service-panel at 240V 50-90A will cost about $750-2000 including installation by an electrician - most states offer generous tax-deductions for this btw.

A Level 2 charger that's part of some charger network, like Chargepoint... I've no idea. The hardware looks nice (i.e. expensive) but I'm unsure who pays the cost of the hardware and installation (is it the landowner or is there a cost-sharing deal with the network operator?)

Level 3 chargers... we're talking a minimum 5-figure amount for even a single-stall set-up: as you need a huge (literally: easily 3x3x6 foot) DC transformer appliance for each pair of charging stalls (hidden behind a fence around the corner), connections to seriously high-power mains electricity (imagine a 16-stall Supercharger: that's 8 DC transformers doing ~120kW each: that's 2MW!) in addition to the aesthetically designed car-connector party. According to this article from 2013 (so it's already 8 years out-of-date) Tesla pays $100-200k towards a station, I'm unsure what the cost-split is, but I'd be surprised if the total cost overall was under $250k imo - I also wouldn't be surprised (still) if simply land-use-rights was the majority share: https://techcrunch.com/2013/07/26/inside-teslas-supercharger...

Hydrogen tanks can also have multiple outlets (I get that they are more expensive than gas outlets, but the tank is presumably a large part of the hydrogen station cost).

If a 1 million dollar instillation can’t keep up with continuous 3 minute fill ups then it’s going to need significant extra hardware to have enough throughput for multiple pumps.

Why would you assume that? Electrochemical compressors, as well as high pressure vessels, are rapidly declining in cost. Fuel station grade electrochemical compressors were ~$1M ten years ago, $300k five years ago, and about $150k today.

Compare that to hooking up to a grid, costwise. All the H2 equipment is 20x more I'd guess than a charging station for an EV.

All the other H2 issues aside, you're competing with a good, established logistic network in EVs: the grid.

H2 is like any new nuke: you're 10 years out with a ton of subsidy and investment from being a player against the EVs of TODAY. Your price targets to compete will have to anticipate 5-10 years of accumulated economies of scale, tech improvements, etc in EV/grid/alternative energy/storage.

I'd be shocked if a bank would float capital to support that given the uncertainty. Just like I'd be shocked if a bank would float financing to a new major fossil fuel development. It might be BANNED in under a decade.

You can't just "use the grid"...existing grid connections can't handle the load. You have to have high voltage lines run direct. Just the equipment to receive and distribute a high voltage connection, you're talking ~$150k. Then another $60k per charger. And before you can even consider installing anything, you have to have lines run specifically for your station, which can often be several miles. That's definitely not cheap.

So no...hydrogen infrastructure is not 20x more. It's probably slightly more expensive than an urban charging station, but significantly cheaper than a rural charging station. And remember: electrical infrastructure is mature...it's not going to get significantly cheaper any time soon. Batteries are the innovation driving EV growth, not the grid.

The only thing that is unambiguously more expensive is the fuel. And it always will be more expensive...but prices are going down constantly. If it costs $100 to drive a truck across the US, why would it matter if it costs $200 to do the same with hydrogen, especially if it means not having to stop for hours at a time to recharge?

You might be right about the grid enhancement, but the rural stations will generally be served by alternative energy that can only be placed in ... rural areas.

The midwest of the US is generally wind-rich. The south and southeast will be solar-rich.

>It's disappointing it never went anywhere.

There's probably a sound reason for that

I'm not an expert in fuel cell efficiency, I'd read that number somewhere and a quick google confirms it: https://www.sciencedirect.com/topics/engineering/round-trip-...

The energy density situation looks less attractive when you include the beefy 350-bar storage tanks, and the fuel cell itself. A battery just has two wires sticking out and that's it.

> A battery just has two wires sticking out and that's it.

What about thermal management, the outer shell that can hold the battery cells and modules, and the battery management system?

Sure, but a hydrogen fuel cell system has most of that too. My point was just that hydrogen can't be directly compared to batteries because it's just the fuel, while batteries are the complete system. I guess I worded it badly.

> A battery just has two wires sticking out and that's it.

Considering the heating/cooling mechanisms required for EV batteries this is a pretty disingenuous statement. Not even considering the highly-toxic fire risk with Li-ion.

Well okay, but if we're bringing those things into it, hydrogen requires cryogenic storage and is, ah, famously flammable, even when it's not pressurized to 350 bar. I wasn't saying Li-ion is perfect - all I was trying to say was that if you want to compare like with like, you can't compare batteries with just hydrogen - the hydrogen is just the fuel, while the battery is both the fuel and the generator.

You need to do some more research. Hydrogen is 57 times lighter than air, which means in the case of a leak, it disperses almost instantaneously...almost always completely dissipating before having the chance to ignite at all. Rapid depressurization will freeze nearly everything in it's path, minimizing your chances of a heat induced ignition. And unless you're giving it liquid oxygen, you have almost no chance of an explosion.

https://hydrogen.wsu.edu/2017/03/17/so-just-how-dangerous-is...

> hydrogen requires cryogenic storage

Not for grid-scale stationary storage, it doesn't. You store it underground as compressed gas, just as many millions of tonnes of natural gas are stored today.

Gas cars are 11x more likely to catch fire than an EV.

Also it's odorless, is flammable in air from 4% to 74%, and has an invisible flame during the day when it does. It's a nightmare.

Don't forget low ignition energy and hydrogen embrittlement

We will know that "green" hydrogen is nearly feasible when they stop talking about hydrogen and just admit that ammonia, which doesn't cause embrittlement and stores energy more densely than hydrogen, is what they meant all along.

But hydrogen is the prerequisite for ammonia, and ammonia still needs cryogenic infrastructure?

No it just has to be pressurized in order to stay in liquid state. The pressures involved at normal ambient temperatures are less than 300 psi, which is not extreme. Scuba tanks, for example, handle pressures of 3000 psi or sometimes more.

Yet it's absolutely required for synthesizing hydrocarbons from CO2 so, there's that.

I mean, pick any fuel you want, their utility as stores of energy are all based on hydrogen content. So we're going to kick the fossil habit and still have fuel of _any_ sort, we're going to have to learn to produce and utilize hydrogen at scale.

Grey and blue hydrogen are transition fuels. They build a market for green hydrogen to fill. It's going to be a little messy along the way.

Sure, but when people say "hydrogen fuel" they usually mean pure H2. Trouble is, those carbon atoms everyone's so keen to get rid of are damn useful, chemically speaking. I actually think there's a lot more mileage in synthesizing hydrocarbons from biowaste. There's nothing inherently wrong with burning hydrocarbons - the problem is that we're lazily digging historical supplies out of the ground, instead of keeping the carbon cycle closed.

Really, we should just put massive tax levies on fossil fuel extraction operations and call it a day. The market will sort the rest out.

> There's nothing inherently wrong with burning hydrocarbons

...if and when you burn them with pure oxygen like done in rockets. Mostly they're burnt using using atmospheric oxygen though, leading to the production of nitrogen oxides (NOx) which come with their own problems [1]. It is possible to reduce the emission of these gases [2] at the cost of increased complexity and cost.

[1] https://en.wikipedia.org/wiki/NOx#Health_and_environment_eff...

[2] https://en.wikipedia.org/wiki/NOx#Regulation_and_emission_co...

NOx (and PM) is a local, temporary consequence. Catalysts exist to remove NOx out of exhaust gasses. They are perfectly serviceable.

Carbon dioxide is a permanent, global issue, where no (large-scale) capturing/cleaning methods exist.

NOx removes itself out of the atmosphere through natural means, CO2 does not.

You can close the cycle, hence 'ignore' CO2 emissions since they cancel out, and deal with NOx using today's existing technology. Same thing for PM and sulfur.

Tell that to the Dutch authorities who are using NOx emissions to restrict industry and agriculture...

> NOx removes itself out of the atmosphere through natural means

...because of this [1], among others.

[1] https://en.wikipedia.org/wiki/Human_impact_on_the_nitrogen_c...

CO2 is a greenhouse gas, and combustion of hydrocarbons emits it.

And creation of hydrocarbons consumes it, so it all evens out in the long run. The problem is that we're digging up deposits hundreds of millions of years old, so evening out over the long run means atmospheric CO2 goes back to what it was back then, which is much higher. But if you burn vegetable oil that you only grew this year, that doesn't emit any net CO2.

We don't need an alternative to petrochemicals for rare use cases we decide are worth the cost-benefit ratio. I expect there will be jerrycans of synthesized gasoline on the McKendree cylinders riding fusion engines to other stars. For the odd use case where you need a stable high density fuel with minimal weight-costs on the engine, and you can tolerate the atmospheric effects it will make sense to use it so we will.

Fossil fuel is hopefully going away, but we will have petroleum in the toolbox forever and we can still use it when it makes sense to.

> I mean, pick any fuel you want, their utility as stores of energy are all based on hydrogen content.

OK, I pick black coal. No hydrogen content. What do I win?

Actually it does have hydrogen...

https://en.wikipedia.org/wiki/Bituminous_coal >The carbon content of bituminous coal is around 45–86%;[1] the rest is composed of water, air, hydrogen, and sulfur, which have not been driven off from the macerals.

>Actually it does have hydrogen...

Shitty coals like lignite and bituminous coal do. The real chad coal, anthracite, is up to 98% carbon.[1]

[1]https://en.wikipedia.org/wiki/Anthracite

Using hydrogen as an ingredient is not at all the same as using it as a fuel.

Plus you don't need hydrogen molecules to synthesize hydrocarbons, and when people say hydrogen they mean molecules.

Last time I looked at hydrogen-electric, a huge limitation was the lifetime of the electrolyzer membrane. They are expensive and you have to replace regularly.

It's not a stupid fuel in cases where you need massive amounts of energy and can't spare the weight... Like rockets and... rockets

Not even rockets. Lose too much from the low density leading to a larger structure and isolation required to handle it as a liquid, and not freeze the oxygen. Meaning worse performance than the pure numbers would imply. Not to mention cost of handling it.

Case in point, SpaceX going larger rocket with less efficient kerosene first and now methane to increase efficiency and reduce soot, while still keeping everything simpler than hydrogen.

Yes, I concede it's a good engineering solution there. Especially when you also need water!

If you’re burning hydrogen to power a rocket, you don’t get to keep the exhaust. If, on the other hand, you use hydrogen and oxygen to power a fuel cell, keep the water, and use the power output to power an ion thruster or similar device, I suspect you end up with worse overall results than simply using a rocket and carrying some water.

Apollo used fuel cells to good effect. Their constraints lined up perfectly - they already needed oxygen tanks, and a supply of water, and they didn't have li-ion batteries back then either. Also, the weight cost of the fuel cells themselves amortizes nicely when you need power for an entire week - and hydrogen is light. And once it had been used, the water was flushed, saving further weight. The big downside to batteries is they weigh the same empty and full.

More importantly, aircraft.

Maybe airplanes and boats?

Hydrogen energy storage capacity underground: $1/kWh

Li-ion energy storage capacity: somewhere around $200/kWh?

So, no, hydrogen is not "hard to store". It's vastly cheaper to store it at scale than it is to use shorter-term storage technologies. The disadvantage is low round trip storage efficiency (maybe 40% if used with a combined cycle power plant) but THAT'S OK for long term storage, which will have many fewer such iterations than diurnal storage.

If you want to level grid energy use and demand over a year, batteries would be insane. Hydrogen, however, could be quite practical.

This is true, but probably could be stated better.

Increased Green Hydrogen production is an essential and inevitable part of the response to climate change.

But the people pushing it for personal transport (often putting more effort into talking down their battery equivalents than actually building and selling vehicles) and anyone suggesting blue hydrogen, as a stopgap because green hydrogen is so far away and expensive, is more pro-fossil than pro-hydrogen.

If oil and gas majors aren't annoyed by your hydrogen strategy, then you're not doing it right.

It has much better energy density than batteries, which is useful when mass is at a premium (in specific circumstances because it also requires heavy apparatus), or when you need some backup power supply to supplement a worst case scenario that would require a prohibitively large battery to handle.

I'm pretty sure it's not true when you account for the wight of the hydrogen tank (as opposed to just looking at hydrogen).

Toyota Mirai is heavier, more expensive, has less range and cargo space than Tesla Model 3.

So it's definitely not true when we're talking about batteries vs hydrogen for passenger cars.

Maybe when you scale the tank to the size of a semi it might edge out batteries, but at the moment there's no data to support this.

Hydrogen is worse than gasoline / diesel.

The hydrogen that is only slightly more expensive than diesel is just as much (if not more) damaging to environment. That's the hydrogen made from natural gas.

The "clean" hydrogen i.e. the one made by electrolysis, is way more expensive than gasoline which is why it's almost non-existent, commercially speaking.

Not to mention all the "but electric cars are not clean because coal plants!" arguments apply 3x as much to hydrogen, because due to losses you need 3x time electricity (and therefore cost) for hydrogen compared to just putting it in a battery.

A Mirai is much bigger car than a Model 3. The better comparison would be the Model S, which has almost identical dimensions to the Mirai; when you do that comparison, the Mirai is around 300kg lighter and goes further.

It doesn't matter which way you cut it, even with all of its disadvantages, hydrogen with a tank still has 3-5 times the energy density of lithium ion, and it scales better with larger vehicles. This is also the reason why industrial drones are moving to hydrogen. They've been using batteries for more than a decade now, and it's been the big limitation holding back a number of applications. Hydrogen gets you double-to-triple the endurance, even at relatively low pressurizations, which makes things like surveying easier, and delivery services possible. Military-grade platforms are supposedly already looking into using liquid hydrogen, with a number of platforms having already been converted over the past decade.

Any idea about the round trip efficiency for other synthetic fuels like methane or propane? It's probably awful but may be worth the increased density over hydrogen.

You can synthesize gasoline, for that matter. You can't really do much better than that for a liquid fuel. I'm surprised there's not more of a market for "green gasoline". It's not that hard to make.

IIRC, wasn't the price point for a bio-gas (made from atmospheric CO2) something like $4/gallon?

If that's true then it's revolutionary. I'm sure that price point makes Americans clutch their pearls, but in Europe gas already costs a lot more than that.

This article [1] says the cost point is closer to $34 per gallon (back in 2009).

[1] https://www.greentechmedia.com/articles/read/algae-biodiesel...

I mean... vegetable oil is 5 bucks a gallon at Wal-Mart and you can run diesel vehicles on it almost unmodified. I'm struggling to see the problem here.

> I’m struggling to see the problem here.

The demand curve would push out dramatically.

The supply isn't fixed. We already know it's possible to economically produce vegetable oil at 5 bucks a gallon. If demand rose, supply would rise to meet it. If anything, increasing scale would reduce the price.

Also, we have a ton of supply for free to start with. All the oil we use right now is still usable for fuel even after it's used up for cooking. Food grade oil is a higher grade product than fuel grade. Most of it is just wasted now.

I'm not saying we should build our entire energy infrastructure off vegetable oil. There are probably cheaper, better ways. But the market will sort that out. The point is we don't have to give up oil - we just have to give up digging it up. Biofuel isn't some remote and impractical future tech, it's something we throw away millions of gallons of right now.

Mass produced food has a pretty elastic supply.

https://wimflyc.blogspot.com/2021/02/powerpaste-fuel-of-futu... sounds interesting.

Musk/SpaceX is most probably working on efficient production of methane using water, CO2 and electricity for Mars refueling, and one can wonder whether it can get downstream-ed into the Earth economy giving Musk's energy related (solar/storage/etc.) business here.

Dude are you lost in ideology. "downstream-ed [from Mars] into Earth economy" In which world do you live? A Screen Fantasy?

i meant downstream as in feature commits propagation between source repos. As in from SpaceX Mars R&D into Tesla's solar/storage Earth business.

Or liquid fuels like gasoline or kerosene that don't just boil off if you don't keep them compressed. I'd like to see that analysis.

One of the dirty secrets of li-ion is that in order to use it for cars, you need high voltage batteries, and if your charging source isn't similarly high voltage, your charging efficiency tanks because you spend 30-40% of your energy stepping your voltage up to a level that will charge your batteries. For example, if you charge your Tesla with a 120v AC power supply, you're looking at 58% charge efficiency, which will put your end to end efficiency at the same level: ~50%.

The numbers from my friends' Leaf point to a charging efficiency loss of 110V vs. higher rates at around 10%. FWIW.

[edit] From https://www.mynissanleaf.com/viewtopic.php?f=31&p=545335#p54...

>Level 1 (120v, 12 amps): 78% efficient

>Level 2 (240v, 16 amps): 91% efficient

>Level 2 (208v, 30 amps): 91% efficient

>CHAdeMO (500v DC, ~100 amps): 93% efficient

Yeah, that's because of the lower nominal voltage of the battery: 350v vs Tesla's 450v. So you get higher charge efficiency, but you'll get lower motor efficency with it. Not sure if it comes out ahead in end to end efficiency.

Why is it so hard to step up voltage efficiently? 120VAC is already AC, so a simple transformer can boost it to e.g. 480VAC easily, and then you have the same AC to DC problem that a native 480VAC charger would have.

At least for li-ion, the infrastructure problem isn't too terrible, as most cars will have an endpoint with high-voltage, i.e. the workplace.

To reduce cost ( improve volumetric efficiency of transformers) such voltage conversion are done at much higher frequency. Side effect is lots of switching losses. Diamond based semiconductor are coming but that will take time.

The size of a high-power 60 Hz transformer is prohibitive in weight, both directly (where are you going to put it?) and in impact on price (shipping, amount of metal).

It ends up lighter, cheaper, and not much less efficient to do a high-frequency conversion instead, either as a boost directly from the 120V AC or as a DC-DC in the secondary.

Green hydrogen is the only realistic alternative to nuclear in areas which rely on wind power for renewable energy. Li-ion batteries can not economically create weeks long capacity of stored energy with a charge cycle of a few times a year. Wind does not have a day cycle as solar has.

What green hydrogen need to do is become economical, and to my knowledge it is even more expensive per w/h then nuclear is.

For context, I'm a software developer who works with a lot of battery researchers.

My understanding from them is that while Li-ion batteries cannot create weeks long capacity, rust batteries may be able to in the future.

Check out Form Energy[0] for an early version of this, they're claiming ~100 hours of storage right now and if they're right about the science and can manage the engineering it could grow past a month. These batteries are challenging to run, a bit like a mini-chemical plant, but they do actually seem plausible.

(No personal association with the company, some of my coworkers know the founders though)

[0] https://formenergy.com/

Maybe you can answer a few question since the public available information is a bit limited. With a solar + Li-ion battery solution currently in use I often hear about 4hrs capacity at 80% output. How many years does it take for such installation to get a complete return of investment on the batteries, and how on average how low do they need to discharge every day?

With those number we can then look at rust batteries. If we take a country like Germany we can estimate the longest period of low wind conditions in order to estimate how long capacity is required to have the country exclusively run on wind+batteries, and how often such periods occur to calculate the charge cycle. Can rust batteries create an return of investment within a reasonable time frame (with nuclear energy prices as the economical limit for what the market will bear).

It would be very interesting to read a reply.

I can't give you dollar values, but I think there are some ideas that might be informative:

1. The cost of these iron flow batteries is almost entirely in operations, the cost of battery material is negligible. It's early days so it's really hard to say how low ongoing costs for things like valves, pumps, control systems, etc. are going to be.

2. In principle, flow batteries shouldn't need to discharge at all. This is why researchers are confident they can create batteries with multi-month capacity.

I think there is a misunderstanding. If a battery solution is going to produce revenue outside of subsidizes it need to generate power. The number I would look towards is how many time the battery need to be charged and then discharged in order to repay the operational and initial construction costs. No discharge would mean no revenue and so either I am misunderstanding you or you me.

If the cost can be reduced so much that a single discharge is all that is required to both repay the production cost and charge costs (assuming reasonable sell price for the energy) then it would be like a fuel and could be stockpiled as much as could be expected to be used.

If the technology is so much in early development that there aren't much concrete numbers, could you at least give a suggestion about where the costs are right now? How far the economics of nuclear is the technology?

For context of why I am asking this, my country has set the date 2030 as the year when no more IC cars might be sold. We do not have a date for IC power plants and we really should have. It is wrong to continue to burn fossil fuels just because the alternatives are more expensive.

Interesting. Is the idea just that they're optimizing for cost, rather than optimizing for watt-hours per kilogram? I.e. if it doesn't matter how much it weighs, you can use a simple, cheap sort of battery that wouldn't be practical in a phone or a car?

The price point is attractive but this is more than a clever insight into optimizing for different use cases.

Previous attempts at these batteries have run into a lot of operational problems and have had trouble keeping the batteries up and running long enough for them to be useful. There's a lot of novel engineering and chemistry that's gone into operationalizing the system.

"only realistic alternative" isn't quite right since it's one of several grid-scale storage options none of which have yet proven themselves to be economical.

Dunno how it goes for other countries, but I spoke with someone who did professional analysis on green hydrogen from renewables for Poland... and it's not a good outlook.

Essentially, in order to produce it in any reasonable quantity, assuming highest available efficiencies, required such massive overbuilding of renewables that economics started to take second position to just plain "zomg" effect.

It got better if you combined it with nuclear, but there the free market that EU insists on in energy actually made it harder to achieve anything useful, so you end up with massive overbuilding still. It had a little chance if you overbuild wind & solar while also making it so you had nuclear providing close to 100% of the "base load" (defined as "floor in daily electricity consumption") while possibly providing green hydrogen as extra sink for renewables to recoup costs... but it still was problematic economically :(

Methane is a better fuel for that than hydrogen, you can make it from air plus energy (although not easily - there's not enough CO2 in the air for that). Or, more practically, coal and water and energy.

And yet hydrogen is the only realistic solution we have for 0-emission mid/long range aircraft.

what about synthetic kerosene? Net-zero.

You need to make hydrogen as a precursor to making synthetic kerosene.

Also it can damage your ears pretty badly if it burns in open air.