Hacker News Reader

Comment by Animats | Hacker News Reader

Lithium iron phosphate batteries have about 95% round trip efficiency. Some manufacturers say 98%. AC-DC-AC conversion is around 98%-99% efficiency now. So batteries are way ahead on efficiency.

Tesla is switching from lithium-ion to lithium iron phosphate for fixed battery installations.[1] The energy per unit weight is somewhat lower, but that doesn't matter much for fixed installations. The safety is better, too - lithium iron phosphate batteries don't have the thermal runaway problem. BYD, which is the biggest producer of batteries in the world, sells shipping container sized lithium iron phosphate sized battery packs for large scale solar backup.

[1] https://www.utilitydive.com/news/tesla-shifts-battery-chemis...

View on HN

Electricity-to-electricity efficiency should not be compared with solar-to-electricity efficiency.

The typical solar panel is maybe 25% efficient or less. The typical steam engine is maybe 50% efficient. Once things are in the form of electricity, we have a large variety of options for efficient transport and storage.

But sometimes electricity isn't useful. Li-Ion will never be light enough for a serious airplane for example, while hydrogen _IS_ light enough for an airplane. (But maybe too volumetric, as every practical H2 airplane engine relies upon high-pressure storage or even cryogenics to keep H2 at a small enough volume).

EDIT: If we think of it as chemistry... perhaps we can make liquid fuels out of H2 like coal liquefaction turns one fuel source into another. After all, classical fuels are nothing more than carbon + hydrogen + energy... and H2 is a very efficient form of storing energy for that reaction. Or maybe we learn to just use H2 directly? There are fuel cells and engines running on H2 today, but I'm always worried about the volume of H2 (requiring either compression or cyrogenics before you reach practicality).

> perhaps we can make liquid fuels out of H2

You're describing hydrocarbons.

The most-common element in the universe married to the most-common four-bonding element. Using solar panels to synthesize Jet A is probably a better path forward than redesigning every plane in the world to trawl a balloon or carry a pressure vessel.

There's a reason pretty much all of Earth's biology uses hydrocarbons as its fuel of choice.

> You're describing hydrocarbons.

Well yes. But I'm also aware that H2 is a component in the synthetic production of those hydrocarbons. Using the H2 directly could prove beneficial. But if not, then we can always convert it into some combination of C and H and burn that instead.

Either way: producing H2 from clean sources (like this weird chemical solar panel thingy) is a good thing.

Hydrogen also should enable more hydrocarbons to be produced from biomass, by hydrodeoxygenation: (CH2O)n + n H2 --> (CH2)n + n H2O.

IIRC traditional Fischer-Tropsch synthesis of liquid hydrocarbons from coal or biomass is about 30-40% efficient. Partly because the C:H ratio of the source material is not the same as the end product (the excess goes up the smokestack), but also because part of the energy embodied in the source is needed to run the reaction itself.

Adding hydrogen can fix both of these problems, enabling efficiency up to 150% (compared to the energy in the source material, not including the hydrogen). This is kind of a big deal, since for large-scale carbon-neutral production of synthetic hydrocarbons the bottleneck is the source of carbon. Hydrogen can be produced (nearly) carbon-free in unlimited quantities, but barring a breakthrough in DAC technology where is your sustainable source of carbon? And no, we don't want to replace pristine rain forests with biomass plantations.

What's wrong with biomass plantations as a source of renewable carbon, assuming you don't build them on rainforests?

From a biodiversity point of view. Everything.

The generic problem with biomass is the low overall efficiency of photosynthesis. It doesn't work well when it gets too cold or hot, and can use huge amounts of water.

> There's a reason pretty much all of Earth's biology uses hydrocarbons as its fuel of choice.

I feel you are suggesting the reason is because it's an ideal energy transport mechanism, while ignoring the (not insignificant) fact that almost all the hydrocarbons we've used so far have been, effectively, free energy.

Or at the least, incur a cost that we have not yet had to contend with.

> while ignoring the (not insignificant) fact that almost all the hydrocarbons we've used so far have been, effectively, free energy

The reference to biology using hydrocarbons as its preferred energy store is one to fat, not fossil fuels.

That’s far from free energy. And life likely wound up that way because its precursors—hydrogen and carbon—are readily available, it’s energy dense both volumetrically and by mass, it’s safe and it can be converted into various forms of energy directly and easily.

Ultimately, the big unknown is the cost of synthesising Jet A. I think it will be low enough that it blows the hydrogen hypothesis out of the water, but I don’t have more than a hunch to go off.

> And life likely wound up that way because its precursors—hydrogen and carbon—are readily available, it’s energy dense both volumetrically and by mass, it’s safe and it can be converted into various forms of energy directly and easily.

Yes. Most of the available energy is in the hydrogen bonds, but if you want to 'tame' neat hydrogen by binding it to something else, from all the stuff in the periodic table carbon is pretty much the optimal choice.

> Ultimately, the big unknown is the cost of synthesising Jet A. I think it will be low enough that it blows the hydrogen hypothesis out of the water, but I don’t have more than a hunch to go off.

To the extent one of the inputs to synthetic Jet A is hydrogen, it won't be cheaper as such. So the question really is whether the added cost is low enough that the easier logistics of a room-temperature liquid fuel makes the total cost lower. I'm slightly hopeful that this will eventually be the case.

If we can get it be unleaded in the air - and perhaps a little less loud… I live on a river and also a jet path. It’s like a freeway at night and also the residue from the jet exhaust is rather annoying to clean off the boat…

Jet fuel has never contained lead. Lead is still used in aviation gasoline, used by small propeller planes (larger propeller planes typically use turboprop engines that use jet fuel). But consumption of aviation gasoline is a very small fraction of the consumption of jet fuel.

Jets have actually declined significantly in loudness over the past decades.

> There's a reason pretty much all of Earth's biology uses hydrocarbons as its fuel of choice.

Carbohydrate != hydrocarbon.

what’s a little oxygen between friends?

>> perhaps we can make liquid fuels out of H2

> You're describing hydrocarbons.

This is technically correct, but at the right pressure, it becomes politically incorrect.

Right now there is already a huge push against fossil fuels, to the point that countries are setting dates for forbidding internal combustion engines in cars. Note that this will take effect independently of the source of fuel, fossil or otherwise. And methane, which is relatively easy to produce from bio-waste, it is a potent greenhouse gas.

For now, the public is divided. Nobody likes the consequences of global warming but nobody likes to give up the convenience of hydrocarbons.

Give it some time or add some more climate shock, and the public may change opinion to be against all fuels that contain some form of carbon, no matter the origin.

The problem is cost.

Technically it is quite possible to produce hydrocarbons today. In theory a government could mandate, for example, 25% green hydrocarbons in fuel in 2030.

In practice, nobody will do that because that is political suicide.

The good thing about the current BEV cars is that over the lifetime of the car they are cost effective. I will be disruptive for a while. But we now have enough experience with BEV cars to know that they are a practical solution.

Telling people that they can keep driving ICE cars in the future, without telling them that green hydrocarbons are going to cost a fortune and would prevent everybody except some rich people from actually driving ICE cars does not help.

Of course, that could all change if we can find a cheap way to make green hydrogen at scale. With current technology we need to get rid of ICE cars. We can always revisit that if green hydrogen becomes abundant.

Personally I would also be happy if all burning of hydrocarbons gets removed from cities. There is no reason to keep breathing exhaust gasses other than that the fuel is cheap.

ICEs in cars hurt local air quality and cause significant noise pollution. Even if you eliminate the fossil fuel aspect, it's still worth switching to EVs.

Aeroplanes have no viable alternative right now, and the noise/particulates issue aren't as big of a problem unless you live near an airport.

> Nobody likes the consequences of global warming but nobody likes to give up the convenience of hydrocarbons.

Nobody wants to give up the convenience of personal transportation using automobiles. And for the United States, a good chunk of which was designed around that this is a serious problem.

Yes, exactly! Or any other hydrocarbon. This is a carbon-neutral process also, as the carbon will come from CO2 in the atmosphere. And we get the bonus of extra oxygen in the air...

Wouldn't the extra oxygen get burned up as you burn the fuel? It would be a carbon neutral and oxygen neutral process.

I think you're right, the oxygen would be sourced from water and water is also one major product of hydrocarbon combustion, which can be easy to forget.

> And we get the bonus of extra oxygen in the air...

Why would this be desirable?

Lowering the density of surface co2 could result in a reduction of the cognitive impairment caused by the already-increased co2 levels. It would also mean combustible fuels of all sorts would burn more efficiently.

/s, mostly

You could use other elements than C to bind the hydrogen, for example Si. But they are difficult to produce, toxic, unstable and the SiO2 would clog up nozzles

Silanes as Fuel for Aerospace Propulsion

https://ui.adsabs.harvard.edu/abs/2009TrSpT...7.Pa33S/abstra...

There was an article on here the other day about a new approach to generating electricity from ammonia. The downside, though, is that creating ammonia needs H2 feedstock, which is often made by stripping the carbon off of methane obtained from natural gas.

If this is a useful way to get H2 feedstock for producing ammonia, that’s interesting on its own!

We used to make refrigerators using ammonia as the working gas, too, but concentrated ammonia kills.

    Exposure to high concentrations of ammonia in air causes immediate burning of the eyes, nose, throat and respiratory tract and can result in blindness, lung damage or death. Inhalation of lower concentrations can cause coughing, and nose and throat irritation.

Huffing gasoline is really bad, but not ammonia cloud bad.

Also the energy density of liquid ammonia is only about half that of liquid hydrocarbons. In the family of nitrogen fuels there's also hydrazine, which is much spiffier, but if you're worried about the toxicity of ammonia, well.. Yeah, maybe we don't want civilians filling up their cars with hydrazine.

The benefit of nitrogen fuels as a carbon-neutral fuel is that getting nitrogen from the atmosphere is a lot easier than extracting carbon dioxide from the atmosphere.

Fischer-Tropsch. But today, it's the reverse, most hydrogen is produced from methane.

> while hydrogen _IS_ light enough for an airplane

If you do the watt-hours per kilogram calculation the very best lithium ion batteries are something like 255Wh/kg right now.

Hydrogen fuel cell systems for small to medium sized UAS are somewhere in the 700-800Wh/kg range right now, including the weight of the (rather heavy, seriously engineered) ultra high pressure carbon fiber wrapped tanks needed for it, piping, and fuel cell.

Something like a tank of Jet-A fuel is >4000Wh/kg.

Hydrogen itself is light, the tankage systems needed to reliably contain a large volume of it on an aircraft are not.

If you've got H2 on a plane it's probably lighter to just feed it to the jet engine directly rather than try to replace that with a fuel cell/motor combo which will not buy you enough efficiency gain to compensate for its weight.

Or maybe not. It's not the weight of the fuel cell/motor combo that is the limiting factor, but the weight of the H2 storage tanks. An H2 jet engine will probably [0] be less efficient than a fuel cell + motor, so more H2 storage is required in order to reach a comparable performance.

[0] Considering the Carnot efficiency of a jet engine at ~30%, a combination of fuel cell (~60%), motor (~95%) and propeller (~80%) will still beat it by a factor of 1.5.

> Li-Ion will never be light enough for a serious airplane

Not everybody agrees with that. https://en.wikipedia.org/wiki/Electric_aircraft#Commercial_p... says over a hundred electric designs are in development.

I think some of them can be called serious airplanes.

I think what OP meant by "serious airplane" was something that could compete with at least a regional jet for short-distance passenger transport (~50-100 seats with at least 500nm range). Almost all battery-electric aircraft in production or on the drawing board are either proofs-of-concept or geared toward hobby use, flight training, or extremely short-range air taxi services (Harbour Air is a fantastic example of a current company for whom battery-electric aircraft make sense for some commercial missions). They're serious in the sense that some of them will fly well enough to become commercially viable, but not serious in the sense of challenging current turbofan airliners for a significant share of the air transport market.

https://venturi.aero/#highlights claims 50 passengers, 1,000 km, and 40% reduction in operating costs.

The site doesn’t give much more info that that, so I don’t think they’re further than the drawing board, though.

I considered this might be a scam or a “give us money and we’ll see whether this really can be done”, but being a startup with Delft roots (https://en.wikipedia.org/wiki/TU_Delft_Faculty_of_Aerospace_...) gives them some credibility, I think.

Yeah, I have the feeling that they're still very much in the initial design phase, but even so 100% electric doesn't even imply battery-electric (and even if Venturi does plan on doing battery-electric, as their site vaguely implies, there's a good change they're basing their numbers off of an experimental battery chemistry that's being tested to replace lithium ion). Hydrogen fuel cells offer energy densities comparable to liquid fuels, although as mentioned previously they have volumetric issues due to the requirement for high-pressure and/or cryogenic storage of the hydrogen. I think the approach that these guys (https://en.wikipedia.org/wiki/Wright_Electric) are taking is more likely to work out in the near term, since their main problem is finding places in an existing airframe to store a bunch of hydrogen tanks.

I think it would be kind of cool if we all get stuck where we were and couldn't use air travel anymore due to no fossil fuels. Maybe I just like retro stuff too much.

Or restart the zeppelin-age, travel at 150km/h with windows open hanging from a giant hydrogen filled structure (not enough helium to replace all aircraft).

It's already happening, see https://www.gov.uk/government/news/world-record-raf-flight-p...

> perhaps we can make liquid fuels out of H2

We can make methane, which is a lot easier to store than hydrogen.

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

And transport. We already have an existing infrastructure for methane.

The 25% panel efficiency isn't a particularly relevant metric, and doesn't play in storage. A cost or area metric is more relevant economically.

I think that's a very important point. For example if generating fuel from algae was only 0.5% efficient, it may not matter if the process can be conducted in open ocean.

The typical mono-si solar panel is about 20-22% efficiency. For poly-si this drops to about 15%, but are cheaper and therefore still sees a lot of demand. Solar water-splitting at 13.8% is well in the range of feasible competitiveness as long as it is cheap to build.

Mono panels were 75% of the market in 2020, I read.

A concern I'd have with this water splitting is getting the water in and the hydrogen out (and separated from the oxygen). Wires are easier than pipes.

Doing this with ordinary PV also means the hydrogen production can be dispatchable. When PV output is tight, just stop making hydrogen and use the electrical power directly.

Keeping the hydrogen output separate from the water input and oxygen output is easy. You feed water in to the bottom of a U shaped container. You put the hydrogen-producing electrode in one half of the U and the oxygen-producing one in the other. The hydrogen and oxygen bubble up out of the water into separate collectors.

That only works with separate electrode systems; in-panel splitting (where the panel has a sheet of water directly over a catalyst) evolve both gases in the same place.

For larger installations pipes are easier than wires.

Why is that? I would guess solid-state (electrons in wire) is more reliable than moving-parts (fluids in pipes).

And it's easier to connect two wires than it is to connect two pipes.

All the little tubing collecting the hydrogen stream trickling out of these things seems like it would be a nightmare to create and maintain.

Over short distances sure, but once you're talking thousands of km a pipeline is way simpler.

That's nice, but it doesn't help the local collection network in the solar field. Using large separate electrolyzers driven by conventional PV lets you avoid that.

Unless the destination of the electricity is also next to the local collection network the pipeline can be simpler. It's certainly not going to be a worse idea.

You didn't understand what I said. I'm not arguing about a pipeline. I'm arguing about all the little hydrogen carrying piping in the solar field. This is needed to bring together all that hydrogen being generated at the hydrogen-generating solar collectors, before it even gets to a pipeline.

Simpler just to avoid all that and generate the hydrogen at large separate electrolyzers at the side of the utility-scale solar field (say), and then inject into the pipeline.

Should be noted that that will work too. In fact that is the plan for most facilities. However, you will still need a separate electrolysis facility. With this you just pipe hydrogen into bigger pipes. It could be cheaper to implement it this way.

For a household size connection sure, but once you're looking at extremely high voltages the electrical solution becomes hard. It's a totally different kind of cable and has very different problems to deal with. Meanwhile a pipe fundamentally doesn't change whether it's 1in or 48in wide.

The certainty with which you make all these obviously completely wrong statements is baffling.

Really, electricity distribution is a solved problem, and has been for more than a century. And here you are arguing that something that is in very minor use at the moment is 'easier', in spite of many challenges still be to solved before it can operate at scale. I'll just leave this link here and I would very much appreciate it if you stopped making all these assertions without qualification as though possess some kind of oracle because it is bordering on the ridiculous.

https://www.energy.gov/eere/fuelcells/hydrogen-pipelines

Your whole comment history is nothing but an endless stream of assertions without evidence including ones that are 'not even wrong'. That's not the level that I expect for this site and you are not doing us a service with this. I also do not understand why the only subject you are interested in is pushing the Hydrogen angle for more than it is worth.

Piping gases is also a solved problem. It existed even before electricity. I sound certain because I'm certainty right. If anything it's bit amusing to see supposedly smart people have so much trouble accepting the existence of 18-19th century technology.

FYI, coal gas is about 50% hydrogen and is two centuries old: https://en.wikipedia.org/wiki/Coal_gas#Composition

Nice to see ya posting around here. Always a pleasure.

I'm not as pro-Hydrogen as you but I see some potential benefits for sure :-)

Great thing about Jet A, it's actually pretty hard to get burning.

Hydrogen: See Hindenburg.

I would rather not fly in a bomb. I just don't see it being viable for planes.

We're starting to see short distance small planes electrify and then long haul will probably go synthetic fuels over the next 10 years.

I read something that says Hindenburg was more about the mix of the gas and the material of the blimp. Properly mixed H2 doesn’t burn (it explodes quickly, which is what powers an ICE).

I agree though, planes and H2 probably don’t mix. Jet fuel should not be a priority to replace with H2. Priority should be given to static applications before trying to apply it to moving vehicles.

You are right. The issue with the Hindenberg was not the hydrogen. It was the coating.

The main concern for hydrogen as a fuel now is the high pressures. If the pressure vessel is damaged it explodes due to mechanical forces. This is not impossible to work around. Consider, for example, that gas tanks can already be punctured and lead to fires; it takes a fair accounting of the risks to really rule out hydrogen as a fuel.

Right, but consider a punctured gas tank vs... well, have you played "spot the COPVs in the SpaceX explosion"? They spit fire and zoom around like party balloons. Good fun from a distance, but if I had to pick one to ride with every day, it's an easy choice in favor of the gas tank.

Do you know what the fuel is? Rockets are unique in that they carry a significant amount of oxidizer with them. A hydrogen container is unlikely to zoom around spitting fire, there's too much fuel for how much oxygen there is and it rises very quickly.

There are still dangers, but practically I'm not sure they are that much worse than a gasoline fire. There are also systems that reduce the pressures needed, like activated carbon beds. Research on them stalled mostly because there was no good way to generate H2.

I thought the fact that H2 burns invisibly combined with the fact that it goes boom once the burning mixture is stoichiometric is the big risk with H2 over typical hydro-carbon-based fuels, which burn a bright orange and are surprisingly hard to blow up.

The Mythbusters did a show on this exact thing with spraypaint, hairspray, and propane canisters. They didn't explode even when wrapped with flaming rags, but sometimes the released material would combust. In some cases the gas rushing out actually extinguished the flame.

You rarely by accident get a stoichiometric mix. With a slow leak into an enclosed container is usually how it happens. Otherwise there is too much fuel to O2 in the atmosphere.

People really need to stop trying to sound smart by thinking of imagined dangers. That is what stopped nuclear power, it's what's stopping hydrogen as a fuel, it's what's stopping gene therapy research, etc. Get over yourself.

> People really need to stop trying to sound smart

> Get over yourself.

Right back atcha.

Good grief, "stoichiometric mixture" is not a reasonable person's danger threshold. The mechanical energy alone is terrifying, which was the point of bringing up COPVs.

While we're at it, let's put giant flywheels in chemical cars to do regenerative breaking. They'll probably have less mechanical energy than the titanic so they're probably no big deal, right? Right.

Look, this stuff has been studied and you can find it with just a little bit of looking. How this conversation has been going is a bunch of chaff gets thrown up and then people are expected to answer.

I'm doing as best I can. I happened to have a coworker who was involved in alternative fuels research, specifically the H2 cylinders that operated at lower pressures using a carbon substrate. But even if you don't use that you can still construct and secure the container so it splits instead of shatters and contains most or all of the debris. You can install a shield bulkhead into the car to deflect debris to the ground. You can do loads of things.

You are not original, you are not the first person to raise these concerns, and very competent people have been working on them.

https://xkcd.com/793/

https://en.wikipedia.org/wiki/Hindenburg_disaster#Incendiary...

They tested the incendiary paint hypothesis on the TV show Mythbusters.

The MythBusters concluded that the paint may have contributed to the disaster, but that it was not the sole reason for such rapid combustion.

Ah, you're right. I wasn't remembering the details exactly. But the burning material that falls is the skin and can be better fireproofed to let the hydrogen escape upwards.

Those gas tanks are more risky when you have constraints on weight and material of the tank, like in moving vehicles. I agree though, people unfairly disqualify hydrogen on this risk only.

Diffusion is also a problem. A very efficient form to store hydrogen is a solid block of metal, which of course comes with similar restrictions as batteries, if not worse.

> I would rather not fly in a bomb. I just don't see it being viable for planes.

As opposed to the kerosene fueled bombs we're flying in right now?

I'm not sure comparing things to close-to-wartime tech from 90 years ago is a valid approach.

> I'm not sure comparing things to close-to-wartime tech from 90 years ago is a valid approach.

You might be surprised how good WW2-era scientists were at chemistry and science.

The earlier poster has a point. Gasoline, Diesel, and Kerosene have all been studied extensively (especially in the 1930s and 1940s BECAUSE of preparations for WW2). The safety of soldiers was paramount even to the Nazis.

War machines: be they airplanes, tanks, or battleships, were all going to be exposed to enemy fire and explosions. These fuels (Kerosene and Diesel in particular) were chosen because they're extremely safe: high flash points and even higher auto-ignition points.

That means that kerosene __literally__ can't catch on fire at room temperature. You need to warm up kerosene before its appropriate to burn (Of course, once its burning it will warm up the rest of the kerosene and keep burning. But this isn't some super-explosive volatile chemical we're talking about here)

Gasoline is still safe, but not as safe as the other two (-40C Flash Point). Gasoline was safer than a lot of the other petroleum products that were investigated back then, and was still chosen as a fuel for war machines.

---------

The same is not necessarily true for H2. H2 will explode at any temperature (even near absolute zero).

> You might be surprised how good WW2-era scientists were at chemistry and science.

I'm well aware of chemical prowess in that period, especially on the German part. They had a bunch of Nobel prize winners.

However, that's why I said "close-to-wartime". First of all, in case you weren't aware, the Hindenburg was supposed to use helium but due to a lack of helium in Germany (I think it was primarily due to American export restrictions), they had to use hydrogen.

I don't know of other limitations for the Hindenburg, per se, but knowing the overall German shortages of the period, I wouldn't be surprised if they had other structural issues with the Hindenburg itself, especially since it wasn't designed to be used with hydrogen.

On top of this, the Hindenburg design was from the late 20s, 1929, I believe.

If materials science, modeling, etc haven't advanced since 1929, I'll eat a shoe. If we're somehow worse at managing hydrogen after almost 100 years, I'll eat the other shoe.

A comparison with the Hindenburg isn't useful in any case. To put enough hydrogen in a car to get a reasonable number of miles, you need to compress it to 10,000psi. The Hindenburg is basically atmospheric pressure.

10k psi actually makes hydrogen safer (chemically). Consider it a form of positive pressure - there is much less risk of air getting into the tank.

The Hindenburg crew had to regularly test the hydrogen cells for oxygen content.

> War machines: be they airplanes, tanks, or battleships, were all going to be exposed to enemy fire and explosions. These fuels (Kerosene and Diesel in particular) were chosen because they're extremely safe: high flash points and even higher auto-ignition points.

I'm not sure those were really the reasons for the choices that were made in most cases. Airplanes of the era used gasoline engines because Otto engines provided much better power/weight than the diesel engines of that time. For submarines, I'm not sure safety was a bigger factor than being able to use the same(?) fuel oil that the predominantly steam powered surface fleets were using, and that diesel engines provided better fuel economy than gasoline ones.

One place where safety might(?) have been a factor was that the USSR decided to power its tanks with diesel, whereas the other major combatants used gasoline engines in their tanks. I'm sure this prevented a lot of fiery demises for Soviet tank crews.

> The safety of soldiers was paramount even to the Nazis.

I wouldn't say paramount: https://en.wikipedia.org/wiki/Messerschmitt_Me_163_Komet

(the Komet had a tendency to blow itself and its pilots up due to the hypergolic fuel)

This has some interesting comparisons between hydrogen and gasoline as fuel, in terms of historical experimentation as to the effects of rupturing or destroying tanks holding one or the other: https://hydrogen.wsu.edu/2017/03/17/so-just-how-dangerous-is.... Not my field so I'd be interested in if this more or less reflective of current knowledge on the subject.

Even if H2 isn't used directly, there are likely chemical processes that can convert it into some form that is more useful.

EDIT: That's what its called. H2 is a component of Syngas (https://en.wikipedia.org/wiki/Syngas). Syngas can then be further processed into kerosene, gasoline, or other fuels we use.

And where exactly does the energy stored in a battery come from, do you imagine?

This is PV powered production of hydrogen. Yes, the PV electricity could go directly to a battery, granted. But the PV cell is only about 20% efficient in that conversion. (Lab results are up to 30%.)

There are use cases where rather than an electrical transmission line needing to span an ocean, energy could be stored chemically (e.g. in hydrogen, or hydrocarbons produced from them) and transported like LNG across an ocean.

The claim of 95% round trip efficiency is basically relevant if the the energy is to be consumed (or placed on a grid) near the point of production. Otherwise 65% (13%/20%) efficiency for chemical energy that can be physically transported is not too bad.

Hydrogen is a battery from the perspective of a renewable energy system. If it is produced as a byproduct of a fossil fuel (as is commonly done) IMHO, it is not a renewable resource at all. Producing it from solar PV is a completely different kettle of fish.

Other people have responded about circumstances where hydrogen-as-fuel (or fuel component) might still make sense; without commenting on that, it's important to also note that hydrogen is an important industrial feedstock, independent of its potential energy applications (it's critical to producing fertilizer, among other things), and industrial hydrogen pretty much all comes from natural gas at present. Given that we need at least some hydrogen anyway, figuring out better ways to produce it is worthwhile.

In that scenario, total levelized cost is the more important criterion that efficiency -- if these things convert solar energy less efficiently, but the contraption is cheaper than solar panels plus conventional electrolyzers because it's simpler or uses cheaper materials or whatever, it might still be the more economical bet.

There is plenty of demand for millions of tons of hydrogen for uses that batteries are not useful for, notably, nitrogen fertilizer. Thus, advances in hydrogen production are interesting irrespective of ones views of hydrogen vs batteries.

Talking about efficiency with solar seems purely academic to me. Considering it's just there, it doesn't really matter what percentage of the potential energy you're capturing, just the absolute cost per watt of extraction.

There is another dimension at play here: solar is about 1KW/sq meter so efficiency directly translates into less area for the same output.

An interesting bit about the original paper linked -- from the abstract:

> The trimetallic NiFeMo electrocatalyst takes the shape of nanometer-sized flakes anchored to a fully carbon-based current collector comprising a nitrogen-doped carbon nanotube network, which in turn is grown on a carbon fiber paper support. This catalyst electrode contains solely Earth-abundant materials, and the carbon fiber support renders it effective despite a low metal content.

I don't know anything about chemistry so I'll have to take their word on the fact that the elements they use are abundant.

The two concerns for area are:

1) The cost of panels to cover the area

2) The cost of the area itself, and the availability of a ton of empty space in a convenient location

If they are using abundant materials, the first is not as much of a concern, compared to photovoltaics. Since they are producing some sort of fuel rather than directly producing electricity, transmission efficiency is not really a concern*. So maybe we could plop a bunch of these things down in some sunny middle-of-nowhere desert.

* I guess is we consider vehicles to transport fuel, which must themselves burn fuel, to in some sort of abstract sense be part of the transmission efficiency, this computation could be pretty complicated.

To answer your question regarding commonality/ready availability -

Carbon, Nitrogen, and Iron are cheap and easy (you’re likely within body length of a large quantity right now). 32% of the earths mass is estimated to be Iron. $1/lb or less in massive quantities.

Nickel and Molybdenum are slightly harder to find, but not by much - nickel makes up 1.8% of the earth by mass and is a reasonably common metal in everyday manufacturing. It’s currently at $3.97/lb spot price at tonne quantities. Moly is in everything from greases to steels, and while typically not used in large bulk quantities alone, is available for such [https://tradingeconomics.com/commodity/molybden] at looks like $23/lb give or take.

So all commonly available elements, albeit (nanotubes) not necessarily in the form desired just yet.

I think cost of installation / maintenance per sq meter is the real kicker. There's actually a lot of still available space for solar panels (reservoirs being the favourite one I just discovered).

Area isn't a particularly limiting factor for most applications.

That's opposite of my knowledge about this but I'm a decade+ out of date, so possibly this isn't as much a factor as it used to be, but the cost of the ground was a substantial factor in solar installations in days past.

It is one of the most limiting factors for solar, when everything is accounted for.

You either put it really far away (increasing transmission losses and right of way issues for the much longer lines), or you put it closer and then have to deal with expensive land or difficult environmental reviews.

It is not AS BIG of an issue as it could be - for instance 30% efficient cells vs 23% efficient cells, the lower efficiency’s cost vs space usually favors the lower efficiency, but it’s still very strong overall.

As long as we have land for corn as feedstock for Ethanol, land is not an issue. You get a lot more energy from PV than from the equivalent acreage of plants.

All of the world's existing pasturage is prime territory for solar installations. Adding solar to a pasture makes it better pasture, reduces evaporation, and cools the panels vs. desert installation.

Plants generally max out the amount of solar radiation they can absorb early in the day, so are not handicapped at all by partial shade. In fact, most benefit from it in numerous ways.

Installed between rows in active farmland, solar reduces water demand by up to 50%, which is a really huge benefit. Reducing heat and evaporation stress improves crop yields. So, a farm could move from barely getting by to solidly profitable by installing solar, even before selling the power. The panels just need to be placed so as not to interfere with driving a tractor between them.

So, no, there is absolutely no shortage of land for solar installations.

Sure, but you also don't need dedicated land for solar, giving cost savings options vs other plants that need dedicated space. There are so many available installation locations that overlap roofs, parking lots, even projects sharing farmland, reservoirs/canal paths.

All of which drive up costs per installed megawatt. You’re talking about marginal (as in, in the margins), which can produce significant power - but is expensive and has side effects on habitations that many people are currently willing to overlook but won’t be the case forever.

Racking and wiring is one of the dominant costs in any solar installation, and a big flat desert is cheaper both to install and maintain than almost every other option by a pretty hefty margin.

As far as I can tell the “side effects” are beneficial synergies - I love having shade in parking lots for example. You should also be saving on real estate costs if you already had a primary use for the real estate.

Maybe the solar cost efficiency taken alone is reduced, but the system efficiency of the real estate use goes up as does the ownership value for the property owner.

Batteries have much lower capacity, lower cost efficiency of storage, and lower weight efficiency of storage. They’re also slower to charge and slower to discharge.

Batteries are great for storing electricity efficiently as you say, but that only matters when electricity is marginally expensive (sourced from fossil fuels for example). However, if you have a cheap enough marginal cost of electricity, hydrogen is a better option because of the much cheaper fixed cost of hydrogen storage per kWH.

A good way to see this: what option would you choose for storage if electricity on demand was free but only during certain parts of the day? A hydrogen tank or batteries? A hydrogen tank is cheaper. And clean energy is essentially free besides the fixed cost of the windmill/solar panel.

> Batteries are great for storing electricity efficiently

In some cases not even that. Modern Lithium-ion batteries are quite good, but self-discharge still is 2-3%/month. So, if you want to store for months (say charging a battery using the summer sun to heat a house in winter), you easily lose 10% to self-discharge, in addition to what you lose between charging and discharging.

https://en.wikipedia.org/wiki/Self-discharge#Typical_self-di... says NiMH batteries even lose 30% per month.

It also says "Low self-discharge NiMH : As low as 0.25% per month ... introduced in 2005 by Sanyo, branded Eneloop "

70-85% retained after a year is a lot better. Charge in the same charger. I used to use regular NiMH infrequently, the worst way to use them. Low sd way better for that application. Also NiMH doesn't lose capacity delivering high currents like AA alkaline does.

> They’re also slower to charge and slower to discharge.

Yes, but this is changing and there are a number of very interesting options.

https://en.wikipedia.org/wiki/Lithium-titanate_battery

Very cool, there are definitely use cases where batteries are necessary (like moving vehicles) and it’s great to see advancement in battery tech.

EDIT: These batteries seem to be even less energy dense than Li-ion which is one of the weaknesses of batteries in comparison to hydrogen.

Yes, that is their shortcoming for now, but for solar storage and such they are an interesting option.

> The safety is better, too - lithium iron phosphate batteries don't have the thermal runaway problem.

Relatedly, LFP are just rugged as heck batteries. They tend to have double or more the life-cycle count. 5000 cycle count before getting down to 80% capacity? In many cases yes. That's pretty impressive, and a great boon to overall lifecycle/long-term costs. Even if you're getting less energy-density, or even less energy-store/$, that battery is going to live at least 2x the effective lifespan (barring accidents, major defects, &c).

In general, anything built for high current tends but lower energy density tends to have much more ruggedness. Sanyo’s UR18650E is Li(NiMnCo)O2, for example, but in one paper shows 4X the lifecycle count of the extremely well regarded O.G. of LFP, the A123 ANR26650M1-A[1]. Even though it's considered a "lithium ion". And uses some of the rarer/more expensive materials. (Edit: re-reading the paper more closely, I'm less willing to embrace this conclusion. The A123 remains around 2000 cycle count regardless of depth of discharge pattern, while the li-ion drops to 1000 around 50% DoD.)

Not that phosphorous (The P in LFP) is projected to remain cheap/available forever. 2 days ago, talking more about phosphor's availability specifically with regard to agriculture (and rather alarmistly), "Phosphorus is essential to life and the world is running out of it"[2].

[1] http://www.jocet.org/vol7/511-C0056.pdf

[2] https://medium.com/climate-conscious/peak-phosphorus-may-be-... https://news.ycombinator.com/item?id=29244529 (2 days ago, 2 comments)

meta-note: this post has bounced up and down, 1, 2, 1, 2, 3, 1, 2. something like that. i just find that weird. hn has some weird downvoting behaviors that i really don't understand at all. maybe there's something legit downvotable or that people don't like? this seems like an innocuous post to me. i don't see the reason for it to keep getting negged.

apologies. i know discussing meta-topics is against the rules. i just really dont understand why there's an ongoing basis against this content becoming popular. it's so weird to me. i want to believe there's genuine real authentic behavior behind the votes happening here & elsewhere, but it feels so so weird to see this kind of straightforward telling constantly bounded, rising, then flattened back down, again and again. it instills a larger sense of disbelief, feels like there are vulgar reactionary forces about, which is not at all a conspiracy mindset i want to fall into.

LiFePO batteries are also getting a lot of love from the marine community. I recently installed one on my boat to replace three lead acid and the weight savings alone will make up the difference in cost in fuel savings over the course of the year, not to mention the current generation last longer than lead acid in that environment.

Pedantry alert.

Lithium iron phosphate is lithium-ion. What Tesla is doing is changing the cathode material from Lithium nickel cobalt aluminium oxides (NCA) to Lithium iron phosphate (LFP). Other cathodes materials include LMO, LCO and NMC.

NCA has the best energy density, LFP is the best in everything else, including price, NMC is somewhat in the middle and LCO and LMO are becoming obsolete.

That's not a comparison. Hydrogen goes Sun -> storage -> DC. Solar panels go Sun -> DC -> storage -> DC. You're comparing (Sun -> storage) against (DC -> storage -> DC). You need to include hydrogen fuel cell efficiency and solar panel efficiency for a (Sun -> .. storage .. -> DC) comparison.

This method goes Sun -> Hydrogen -> DC

100kwh of Solar energy generates 13.8kWh of Hydrogen, which is about 50% efficent thus generates 7kWh of DC

Going direct to DC is 23%, and the battery layer drops that to 20%, so 20kWh - about 3 times as efficient if you are using storage.

That's not bad, there are many applications where you'll take the extra land/solar take to provide the flexibility and density hydrogen can provide. Sure not for normal grid connected and static usage, but for mobile and temporary usage.

And that's not to mention all the other uses of hydrogen in industry that, without this method, would have to be generated by Sun->DC->Electrolysis

For renewable energy, low efficiency is not a problem per se: you can make as much as you want. The article also mentions that they don’t need rare materials for the electrolyser, which means it will be cheaper. Lithium, that modern batteries require, is quite rare.

> BYD, which is the biggest producer of batteries in the world, sells shipping container sized lithium iron phosphate sized battery packs for large scale solar backup.

Tesla has been playing in that space too with their Megapack.

https://www.tesla.com/en_AU/blog/introducing-megapack-utilit...

There was a pretty bad fire with them in Australia. https://www.reuters.com/technology/tesla-megapack-fire-austr...

It is a stretch to call it a bad fire. One of the battery containers caught fire, and caused a second to be lost for reasons I don't know. The fire didn't spread, did not destroy the facility and did not require extensive downtime for repairs. Which is pretty nice to know as the fear was that one spark would send the whole facility up in a fireball.

> It is a stretch to call it a bad fire. One of the battery containers caught fire,

I suppose it's relative, but it seems it's a very good thing that only one caught because

> A giant Tesla battery pack caught on fire at a building in Australia, and it took 150 firefighters and four days to contain the blaze.

https://www.msn.com/en-us/news/technology/after-tesla-megapa...

150 firefighters taking 4 days to put it out. It certainly could have been worse, but I don't think it's a stretch to say it was pretty bad.

>So batteries are way ahead on efficiency.

Conversion efficiency is a different thing than storage efficiency or transmission efficiency.

Unless your gas tank is leaking, it's efficiency as a portable energy storage device approaches 100%, especially for modern vapor sealed vehicles.

yea, but batteries are still quite expensive (I mean on a grid scale) and the cost does not get better with scale. Storing H2 is much cheaper and will also probably scale sublinearly (simplified: you pay for the wall of the tank and the wall is only volume^(3/2)). Also, h2 can be transported, traded, ... whatever, batteries are impractical for this and cables also have limitations.

However, I see one big down to hydrogen. You can't tell how it was made, you rely on a promise that it's green, while it may actually be from methane... maybe with carbon storage, but you never know when that is gonna leak out. This might be a big loophole for fossil fuel corps / states

You left out the solar to electric efficiency. If that's 30% (which is an upper limit if you produce the best solar cells you can) batteries have twice as efficient a round trip loop from the sun. In return, they are far less mobile and heavier to ship once loaded. Meanwhile, we can talk about moving around H2 via pipelines. Or simply higher energy storage for big uses. (Less energy spent dragging around batteries). Plus, the batteries do require lithium, which is pretty nasty stuff to get.

Solar cells are inefficient, and we should just get power off batteries instead?

That's not an argument for anything, and doesn't have anything to do with the article.

This is electrical to kinetic energy, which is an entirely different story.

The sun basically provides unlimited energy, so efficiency isn't as important.

This is a critical point.

All of the hottest utility-scale storage technologies have really quite poor round-trip efficiency, but it doesn't matter. Lower efficiency requirements so radically reduce the system cost that you add more cheap generating capacity on top, and come out ahead.

There is not now, and never will be, any shortage of available land area for solar panels. Almost every place you can think of to put them improves the place. On a reservoir, crop land, pasture land, canal, they reduce evaporation, and run cooler than in the desert. On crop and pasture land, they reduce heat stress and water demand in the plants, improving yield and irrigation efficiency. On roofs and parking lots, they slow sun damage. On parking lots they even keep the rain off.

> AC-DC-AC conversion is around 98%-99% efficiency now.

I'd really like to see some citations with specifiic BOMs of equipment that can accomplish this, because the datacenter industry would like to have something like that, but it's my understanding that the efficiency is nowhere near that 98-99% figure.

Apparently there are also some major patents that have expired or are expiring soon, which I hope means that costs will drop and we'll start seeing more LiFePO4 battery manufacturers outside of China where the vast majority are made now.

I might be mistaken on storing hydrogen, but all that takes in a high pressure container. I can imagine putting those solar-to-hydrogen plants up at sunny but remote places with plenty of space. Once a month when the containers are full a train comes along, unhooks cars with empty containers and goes off with the full ones. If you were to do that with any sort of battery you'd need two batteries, one be charged, one being used/discharged elsewhere, and your battery (for lack of a better term) efficiency (Total cost of usage?) basically halves.

Likewise the other way, say you need power at a building site for a couple of years, you can do that by delivering hydrogen bottles to replenish your generator, rather than having to take charged batteries and swap them out somehow.

I love LFP batteries and feel they are not given enough importance in the energy transition future. I was wondering by chance if there is a way to estimate the material amount used in these batteries to check if their limiting factor in scaling is lower than typical high Nickel batteries. I was wondering if Phosphate is a concern if not Lithium.

The original technology is properly lithium-nickel dioxide. Elements such as manganese, cobalt and aluminum appear as minor constituents. Both Li-NiO2 and Li-FePO4 are based on conduction of lithium ions in the electrolyte ("Lead-acid" would be sulfate-ion by this naming convention; alkaline batteries are hydroxide-ion).

It's also about energy/kilogram, especially in Aviation.

Reply on news.ycombinator.com