In a near-future world we have stopped directly fueling private, public, and mass transit and we have migrated to a hydrogen economy as the standard energy carrier.

Consider the following factors:

  • Global energy consumption has grown 15% beyond today.

  • Fuel resource availability

  • Ease of resource distribution by land and sea (technically advanced beyond today).

  • Efficiency of energy transfer

  • Renewability of the fuel

Neglect / Hand-wave the following considerations:

  • Technical difficulty of the generating process

  • handling and containment material bulk or weight

  • Storage difficulties

What would the most practical primary fuel be to sustain a First-world power on a hydrogen economy?

Note, in technical terms hydrogen is sometimes called a “secondary fuel” but it is never itself a source of energy. According to Wikipedia:

An energy carrier does not produce energy; it simply contains energy imbued by another system.


Chemical fuels are divided in two ways. First, by their physical properties, as a solid, liquid or gas. Secondly, on the basis of their occurrence: primary (natural fuel) and secondary (artificial fuel).

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – Monty Wild
    Commented Nov 28, 2019 at 12:32

9 Answers 9



If I'm understanding your question correctly, it's essentially asking 'what the most effective means of producing hydrogen gas', as hydrogen gas is the innate backbone of a hydrogen economy. Which means you're essentially asking 'what's the most practical fuel source for energy', as a hydrogen economy just changes the carrier of the energy from hydrocarbons to hydrogen.

So it's nuclear. Nuclear energy is a great source of power, and if it weren't for the inherent risks involved with using small chunks of plutonium to power everything, I'd be for nuclear powering everything. But nuclear energy is just the most efficient source of energy, so all you'd need is a few nuclear power plants to produce your canisters of condensed hydrogen gas (presumably using something like high-temperature electrolysis). Relevant xkcd below on why nuclear power is great.

required XKCD reference

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – Monty Wild
    Commented Nov 28, 2019 at 12:34

There are two good ways to generate hydrogen, and several good ways to power them.

To generate hydrogen you either want to use gas reformation(of which several types exist) or electrolysis. these have high enough efficacy to to be practical. Which you use does not actually matter much since they all rely on an outside energy source, usually electricity (the ones that rely on fossil fuel combustion can be skipped due to the requirements for renewability since they are not more efficient). Although all forms of reformation rely on hydrocarbons as a reactant biologically produced hydrocarbons work nearly as well. likely you will see all in use depending on an areas resource availability. Once hydrogen is produced the transportation and distribution is the the same regardless of source, so it is moot.

Generating the electricity will be based on local conditions, where available hydro-electric is almost always the best, but wind, solar, and nuclear also work, as do a number of other sources. Electricity is electricity regardless of where it comes from. each source has its own advantages and disadvantages, hydro is by far the most efficient and highly consistent too boot, but also very location specific. Solar is limited by climate, time, and latitude but in the right location is very efficient provided you have a storage system. Wind is constrained by local wind patterns and has technological longevity limitations currently which drive the price up, but there are a lot of usable locations. Nuclear works anywhere, can be supplied on demand, is safer, and has about the same efficiency as wind, but has high setup costs and PR problems. Honestly you are unlikely to see only one in use, more likely ALL of them will be used, they each have locations in which they work best, and like I said hydrogen does not care what the electricity comes from, and it will likely be produced in many places. Thus whatever works best for that location is what will be used.

  • 2
    $\begingroup$ Nuclear is also non-sustainable: it will slowly increase background radiation over decades, the biggest contributor being nuclear accidents (roughly one per decade at the current number of nuclear plants, modern reactors are safer but that safety is so expensive that solar and wind plus storage are much cheaper, and offsetting safer reactors with higher reactor counts means you stay at comparable accident rates). So it's not mere PR problems here; the problem is that the optimists ignore low-probability, high-damage risks. $\endgroup$
    – toolforger
    Commented Nov 15, 2019 at 9:07
  • $\begingroup$ One of the attractions of a hydrogen economy is that one can use solar/wind power for electrolysis when there is too much, and just use stored hydrogen when there isn't. $\endgroup$ Commented Nov 15, 2019 at 13:08
  • $\begingroup$ @toolforger nuclear kills fewer people than wind, even including meltdowns and other catastrophic accidents, also if every nuclear plant on the planet failed the global increase in radiation would be negligible,and converting fossil fuel hydrocarbons into hydrogen will release more radiation than nuclear plants, fossil fuel are notorious carriers of radioactive materials. Also if you look at my sources the cost wind is almost identical to nuclear and solar is worse than both so saying it costs more in nonsense. statista.com/statistics/494425/… $\endgroup$
    – John
    Commented Nov 15, 2019 at 14:57
  • $\begingroup$ @MartinBonnersupportsMonica hydrogen is a rather poor way to store energy for large scale production, of the portable means it is the best which is why we will likely use it for transportation, but at a static location pumping water uphill or just compressing air is a way better way to store overproduction. $\endgroup$
    – John
    Commented Nov 15, 2019 at 15:19
  • $\begingroup$ @John Producing hydrogen just as a way to store energy is a bad idea, but using excess energy (essentially "free" energy) to produce hydrogen for transport seems sensible. $\endgroup$ Commented Nov 15, 2019 at 15:32

Nuclear can certainly power our civilization for some centuries at least. Moving from Uranium to Thorium would mean the fuel easily available is adequate to give everybody reasonable amounts of energy for at least a couple thousand years. That without any seriously different designs of reactor, just tweaks to known-to-work designs.

Then, depending on how far in the "near" future, there is a pretty solution. You could allow "real soon now" to be now, and use fusion as your energy source.

The "easy" reaction for fusion is D-T. You get the D from processing water to extract the D's, and you get the tritium from putting lithium blankets around your fusion reactor. The primary fuels would be then be: lithium and deuterium.

  • $\begingroup$ How efficient would Thorium-to-hydrogen be? And the fuel for a fusion reactor is already hydrogen - how efficient is this going to be? Is this going through electricity and electrolysis first? $\endgroup$
    – Vogon Poet
    Commented Nov 14, 2019 at 21:59
  • $\begingroup$ @VogonPoet fuel for fusion reactor is not a thorium-to-hydrogen, but Deuterim -Tritium. Wich require quite a lot amount of energy to refine in form of heavy water.And electorolysis enegy is negligable to it. But even then working termonuclear reactor would have a , probably, a huge positive output. And if you wander for "primer" electicity source - hydrodams are perfect for that. $\endgroup$
    – ksbes
    Commented Nov 15, 2019 at 7:19
  • $\begingroup$ @VogonPoet the fuel for a fusion reactor is deuterium; this can be separated from $^1\mathrm{H}$ when they're in the gas phase... its less than 1 part-per-thousand, so there still plenty left, and you'd have to go to all the effort of deuterium separation and extraction anyway. Thorium reactors should be no less efficient than uranium fuelled ones. $\endgroup$ Commented Nov 15, 2019 at 11:02

If you really want to have a "hydrogen economy" there should be no wasteful electrical power grid. If so - the economy would be purely "electrical", without need of any other energy medium to transfer. It means that primary power sources should be in places were electrical line are a bad option, but road/tanker/tube transport is a good option.

So my prepositions:

  • Geothermal fully automatic hydrogen producing plants: They are placed at kilometers depth, produce electricity for themselves only (for electrolysis) and outputs hydrogen (and oxygen) at the surface. They have to be placed in very special areas, maybe even under the sea (+rare metals and gold production from sea water).

  • Single dangerous, but enormously powerful thermonuclear power plant. Also mostly automatic. With power generation totally covering all the humanity demands and more but not less. It can't produce small amounts of power - that's will make it impossible to "split" to lesser power plants. Since it is dangerous - it could be placed only far form any civilization. And since it's huge power - it is hard to build an electrical power grid that transfer such a power world wide.

  • Orbital/space based energy production. Like, say turning the Moon into one big solar farm or putting all nuclear reactors on Moon's orbit. In both ways wires are not an option and hydrogen is a good way of energy transfer (and rocket fuel required for it).


Nuclear Fusion

While we still haven't completely figured out how to do it efficiently, Nuclear Fusion is without a doubt the most potent source of energy that we have available on earth that we can fuel for longer than we can anticipate. So if you can handwave the specific production process (the only remaining hurdle), then it's definitely the way to go.

Nuclear Fission

Second candidate. Current day, for some reason the nuclear power is being demonized, despite it being:

  1. More (money) efficient than nearly anything else with the possible exception of oil (if OPEC didn't inflate prices)
  2. Can be done without environmental effects (unlike any other source, waste products are solid can can be contained in for example old mine shafts in geologically stable area's.) This includes the current-day panic about CO2 and other greenhouse gasses, the production of which due to artificial means will drop hard if we'd start to use it today.
  3. Quite abundant. You can literally harvest it from ocean water. This source claims the oceans hold 500 times as much uranium as land-based ores, but I haven't bothered to double-check it. It's US gov, make of that what you will.
  • $\begingroup$ I do think nuclear fission is the best choice for our power, for the reasons you point out - but don't agree it's the best choice to support a Hydrogen economy, with inefficiencies of the process. $\endgroup$
    – Vogon Poet
    Commented Nov 15, 2019 at 20:06
  • $\begingroup$ I'm assuming that for every type, you create electricity and then use that with electrolysis to create hydrogen. That would result in equal relative inefficiencies as we have now. How and why would other fuels be more efficient ? $\endgroup$
    – Gloweye
    Commented Nov 16, 2019 at 8:38
  • $\begingroup$ Because that's why we don't use electrolysis today. Hot high pressure steam and methane is much more efficient. Methane is also much easier to replace than platinum electrodes. $\endgroup$
    – Vogon Poet
    Commented Nov 16, 2019 at 15:00

I'm going to suggest solar.

There are a whole bunch of ways to power an electrolytic water-splitting scheme, but they all require some initial electrical source and a load of electrical hardware and the additional (albeit not too serious) inefficiencies of water electrolysis itself.

You can cut out the middleman by photocatalytic water splitting. Then what you have is a load of comparatively dumb photocells immersed in water exposed to sunlight, which then evolve gas that can be separated and handled as you would for any other hydrogen source. The current efficiencies are not particularly high, but there's no reason that they should not increase as time goes on. Even with their lower efficiencies, if the cost is low enough the greater simplicity of the plants might well tip the balance in their favour.

The problem of course is that you need both a good supply of water and a good supply of sunlight. There are big chunks of the world that have both of these things, but there are a lot of population centres far from places with really high and reliable levels of insolation.

Work has also been done on photochemical carbon dioxide reduction. The CO2 source for this might be the atmosphere, or it could be dissolved carbonates in sea water. The end product of the reaction is carbon monoxide. Given hydrogen and carbon monoxide, you have syngas from which you can synthesise a whole range of useful hydrocarbons (such as methanol) which are then much more easily stored, pumped or otherwise transported than hydrogen, solving the whole "how do you get hydrogen to central Canada" problem raised above... it gets there in the same way they get oil, petrol and diesel right now. These hydrocarbons can also be used to run fuel cells directly, meaning that vehicle refuelling operations can continue as they do now, or be used to synthesis more complex chemicals such as resins for glues or plastics.

This also has the happy side effect of consuming atmospheric CO2, if the end-products are not combusted.

Electrolysis will still have its place of course... there are places which have convenient supplies of clean and renewable electricity (such as iceland) which might tip the economic balance against artificial photosynthesis, and there will be places that might be happy to use nuclear power, the economics of which will depend very much on local politics. It may be possible to phase it out as efficiencies and costs of artificial photosynthesis drop over time.

With regards to your follow-up question, you can consider that anywhere capable of running photovoltaic cells right now might reasonably be able to run photosynthetic cells too, albeit at lower yields. The possibility exists to have small-scale distributed hydrogen and hydrocarbon plants suitable for community use. The very far north and south would not be able to take advantage of this, but they are already highly dependent on modern infrastructure for their ongoing survival, so it isn't like they end up more vulnerable by a move to a hydrogen-based fuel economy.


Laser Transmitting satellites

They are satellites that collect sunlight, and literally beam it to Earth in a concentrated form. We aren't even that far from being able to develop it if we really wanted to invest in it.


Laser transmitting satellites, as described by our friends at LLNL, orbit in low Earth orbit (LEO) at about 400 km above the Earth’s surface. Weighing in in at less than 10 metric tons, this satellite is a fraction of the weight of its microwave counterpart. This design is cheaper too; some predict that a laser-equipped SBSP satellite would cost nearly $500 million to launch and operate. It would be possible to launch the entire self-assembling satellite in a single rocket, drastically reducing the cost and time to production. Also, by using a laser transmitter, the beam will only be about 2 meters in diameter, instead of several km, a drastic and important reduction.

To make this possible, the satellite’s solar power beaming system employs a diode-pumped alkali laser. First demonstrated at LLNL in 2002 -- and currently still under development there -- this laser would be about the size of a kitchen table, and powerful enough to beam power to Earth at an extremely high efficiency, over 50 percent.

While this satellite is far lighter, cheaper and easier to deploy than its microwave counterpart, serious challenges remain. The idea of high-powered lasers in space could draw on fears of the militarization of space. This challenge could be remedied by limiting the direction that which the laser system could transmit its power.

At its smaller size, there is a correspondingly lower capacity of about 1 to 10 megawatts per satellite. Therefore, this satellite would be best as part of a fleet of similar satellites, used together.

You could say SBSP is a long way off or pie in the sky (puns intended) -- and you'd largely correct. But many technologies already exist to make this feasible, and many aren't far behind. While the Energy Department isn't currently developing any SBSP technologies specifically, many of the remaining technologies needed for SBSP could be developed independently in the years to come. And while we don't know the future of power harvested from space, we are excited to see ideas like this take flight (okay last pun, I promise).

The same site lists microwave transmitters as a possibility, but they would be huge by comparison.

This has some huge advantages over nuclear because it's actually renewable, whereas nuclear power is us digging up dangerous elements, using them, and then tossing them in a pool for a 100 years and hoping future generations have an idea of what to do with them.

The one big flaw, that would make a good book or movie, is that these are swarms of satellite weapons aimed at Earth. Every country would have to be ok with these WMDs floating around under likely a foreign governments control.


Well, heck, if we're hand-waving both the difficulty of generating and the difficulty of storing, the answer is obvious: Antimatter.

The reason Nuclear Power has such a huge Energy/FuelWeight ratio is because it's not using a chemical reaction - its actually losing 0.1% of its fuel mass in the process. Which might not sound like much - a tenth of a percent - but that's enough to generate a massive amount of power from a little tiny bit of fuel. If you plug in a kg of mass into the famous e=mc^2 equation, you'll see that even 1 kg of mass "lost" generates a crazy-stupid number of joules (~90 quadrillion)

Fusion? That number dwarfs Fission by a power of 7. A hydrogen-helium fusion process involves the loss of 0.7% of its mass.

... but you might notice that these numbers are still... well... low. Under 1%. What if you could get that number to 100%? Convert all the energy in a lump of mass to energy? You'd have something literally 1,000 times as powerful as nuclear. Literally 142 times as powerful as a fusion reaction.

Which is exactly what a Matter-Antimatter reaction is.

The reason we don't have Antimatter reactors powering our grid is: we can't easily get antimatter (so far, mankind has only produced a total of a few nanograms of it.) And if we did have it, it'd be difficult to store - it's not like you can contain it in any container made of matter, requiring something like a Penning Trap to hold it.

But if we handwave those two problems away? We can easily convert matter into its anti-matter version, and can effectively store it? Then Antimatter Plants for everyone!

  • $\begingroup$ OK, doesn't this somewhat go against the "fuel availability" goal of the question? From a practical sense? Also, how is antimatter fitting into a primary fuel category? If you have to "make" the fuel, it's a secondary fuel by very definition? (this is not really snarky, if you can say "antimatter is naturally abundant with the right conditions" - then well, it's correct. I just don't see it.) $\endgroup$
    – Vogon Poet
    Commented Nov 15, 2019 at 17:33
  • $\begingroup$ I'll add that I do know that the sun radiates antimatter particles - so it does exist in nature. Is there enough to harvest and can that be done efficiently? That seems to be important for your answer - where the natural fuel source is coming from. $\endgroup$
    – Vogon Poet
    Commented Nov 15, 2019 at 17:39

This is a question about energy sources, not about hydrogen

Since you are not interested in all the points that differentiate hydrogen from other energy carriers, what remains is the question about the best energy generator.

It will be a mix.

Nuclear (though this is contested).

As long as it generates electricity, or enough heat to run a generator, you can use it to generate hydrogen.

Large-scale hydrogen generation would satisfy the energy storage requirements for wind and solar, so all that mattes is cost for generation.

Nuclear is a bit special here, as it has a lot of unique impediments.
Let me expand on them a bit:

  • It is slow to research, just because you have to eliminate the risk of an experimental plant leaking radioactivity in quantities. This means that the other technologies have a research speed advantage. Wind and solar have been profiting immensely from material sciences advances, and right now it looks like there is more to come (better solar harvesting, ligher rotor material and better generator technology for wind); how long that will last is anybody's guess, but I'd say there's at least a decade. (There is research for gas as well, but not the change-the-orders-of-magnitude advances in solar and wind.)
  • Risk acceptance in population and regulators varies greatly, depending on how long ago the last disaster was. Given the low-probability, high-damage nature of nuclear power, there's a bit of a cycle at work: accident anywhere in the world - strengthened regulation - nothing happens - regulation becomes sloppy - safety sinks - another accident anywhere in the world, back to hysteria. (The hysteria is a PR problem, but the base cause is regulatory sloppiness, and I believe it is deeply rooted in human nature.)
    Cycle time seems to be about a decade, with overall stricter and stricter regulation, imposing cost increments; currently, EPR costs are higher than solar and wind cost when calculating per expected Joule generated over the lifetime.
  • The capital requirement is immense, and requires at least a decade for amortization. A decade is enough for interest rates to rise again, having invested in a nuclear plant means that this capital won't be available for another investment with better ROI.
  • Nuclear requires specialized skills, and specialized factories (last time I checked there was only a single forge worldwide that can actually create a pressurized vessel that is known and certified for that use). Ramping up the production capacity will take about a decade until you can have everything in quantities, then another decade to plan, build and regulate the first generation of mass-produced nuclear power plants.

All of these factors make nuclear less interesting for investors.

  • $\begingroup$ Is the biogas for reformation? $\endgroup$
    – Vogon Poet
    Commented Nov 20, 2019 at 2:08
  • $\begingroup$ @VogonPoet Either reformation, or a heat engine, a generator, and electrolysis. The question is assuming some changes that take one or more decades, it is not that helpful to decide: Maybe some nanoscale or catalytic technique will make reformation efficient enough to rival the heat->electrolysis path, maybe not. I guess that's why the question is handwaving efficiency details away, so I tried to limit my answer to what we can make educated guesses about. $\endgroup$
    – toolforger
    Commented Nov 20, 2019 at 15:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .