On the face of it, hydrogen is "the king of fuels", with a high energy-density (per kg), and with zero carbon dioxide emissions.

In practice, it's not that simple. Storage problems and volumetric energy-density rule out most applications, while compatibility issues means that the natural-gas grid could never safely transition. Using hydrogen is also less energy-efficient than other zero-carbon alternatives.

The key problems for hydrogen, as a fuel, are:

• Volumetric energy density — compressed, or even liquid hydrogen takes up a huge volume, for a given amount of energy.
• Embrittlement and leakage — hydrogen damages pipework, needing special materials to prevent leaks.
• Hydrogen source — there are few natural sources of hydrogen, and if we make it from methane, the CO₂ emissions are just elsewhere.

What is Hydrogen, and why use it?

Hydrogen Properties

Hydrogen is the first element in the periodic table, and the first in the Universe,   of which it comprises about 75%.

On earth, pure hydrogen is rare; it is mostly found as a component of water molecules. As a gas, H₂, has a very low density (1/14th that of air), unless liquefied by chilling to within 14 degrees of near absolute zero.

Hydrogen Chemistry

Production

The manufacture of hydrogen is an energy-intensive process, either by splitting methane (emits CO₂) or by electrolysis of water (requires electricity). Therefore, hydrogen production should really be seen as a way of storing energy, rather than a source of energy.

CH₄   +   2 • H₂O     ⇒     4 • H₂   +   CO₂                 ←  also requires heat, typically by burning even more methane.
2 • H₂O   +   electricity     ⇒     2 • H₂   +   O₂         ←  requires a lot of electricity, 39.4 kWh per kg H₂.

Combustion

When hydrogen is burned, it combines with oxygen to form water vapour,   and nothing else. In particular, there are no CO₂ byproducts to contribute to climate change.

2 • H₂   +   O₂     ⇒     2 • H₂O

You can also react hydrogen in a fuel-cell, to produce electricity directly; this is useful in an electric vehicle, as electric powertrains are more efficient and compact.

Steel-making

Hydrogen is an alternative to coking coal in the steel-making process. The iron ore (iron oxide) must be reduced to pure iron, by something else with a greater affinity for oxygen; traditionally carbon, but hydrogen can also do it. Steelmaking accounts for 7-9% of global CO₂ emissions, which could be avoided by making green steel.

2 • Fe₂O₃   +   3 • C     ⇒     4 • Fe   +   3 • CO₂         ←  traditional, uses coal.
Fe₂O₃   +   3 • H₂     ⇒     2 • Fe   +   3 • H₂O             ←  new, uses hydrogen.

Ammonia

Hydrogen is a component of ammonia, which is essential as fertilizer for global agriculture. The Haber process combines atmospheric nitrogen with hydrogen (usually from methane, CH₄). This also consumes energy, and accounts for about 1.2% of global CO₂ emissions.

N₂   +   3 • H₂     ⇒     2 • NH₃

Applications of Hydrogen

Space exploration: Rockets   

For space exploration, energy density is the critical factor. The very best fuel, if cost is no issue, is the one with the highest specific impulse: Liquid Hydrogen + Liquid Oxygen. The rocket equation shows that most of the fuel needs to move the rest of the fuel: that's why the Saturn V rocket, used for the Apollo 11 moon landing, had to have multiple stages. 

For rocketry, we must accept all the problems of dealing with cryogenic LH₂ (i.e. insulation, handling, safety, bulk, cost, complexity) because of the advantage it gives us in energy density.
But even the Saturn V used RP-1 for its first stage, while SpaceX is using liquid methane, because of the sheer bulk of the LH₂.

Jet Engines   ()

The design of a jet engine makes it flexible in its choice of fuel, so an existing design can easily be modified to adapt from Jet-A (kerosene) to H₂. For aviation, we can use liquid hydrogen, since aircraft don't need to store fuel for long (boil-off is not an issue), and the provision of special facilities to distribute and handle cryogenic liquid hydrogen would practical, given the small number of major airports. This is why Airbus are working on this.

For example, consider the Boeing 787-8 jet. This weighs 227 tonnes (including up to 101 tonnes of fuel, with a volume of 126 m³), to give a maximum range of 13, 530 km.

If fuelled by cryogenic liquid hydrogen rather than kerosene, it would need 2.8× less mass, which saves 65 tonnes. However, it would need 4.1× more volume for the same energy.
Currently, that would consume all the existing fuel-tank space, all the existing baggage/cargo capacity (136 m³) and 84% of the cabin capacity (302 m³). So:

• This is achievable, but with almost no room for passengers or bags!
• However, the reduction in weight is 25% of the total. Even with extra insulation, there's less fuel needed to carry the fuel itself, so we can have half the cabin space back.
• Or, we would have to re-engineer the airframe with much bigger tanks, which increases drag, and fuel-consumption.
• Or, we could use these for short-haul: if we used a 787, but only for 4-hour flights up to ~ 2500 km (e.g. London/Athens, or New York/Denver), then it will fit within the available tank capacity.

However, the advantages of liquid hydrocarbons (density, compatibility, simplicity) are so great that it's arguably a better use of energy to synthesise e-kerosene, rather than use LH₂ itself on the aircraft.

Transport: Hydrogen Fuel-Cell Electric Vehicles   

Vehicles can run on hydrogen. It's possible to burn H₂ in a slightly-modified internal combustion engine, but more efficient to use a fuel cell to convert the chemical energy to electricity, then use an electric powertrain: this also allows for a small battery for regenerative braking.

The biggest problem is the need to carry sufficient hydrogen: this means compressing it to high pressure, and using a very heavy steel tank to contain it.

The compression/decompression cycle (to ∼700 bar), and the electrolysis/recombination cycle are both quite inefficient.

Here are some detailed comparisons, with numbers, between hydrogen cars vs. battery-electric vehicles. In summary:

• Range: both types of vehicles have similar range. H₂ wins slightly, with a potential advantage at-scale, especially for large trucks.
• Refuelling: H₂ can refuel in 5 minutes (once you've driven to the filling station — and H₂ filling-stations are scarce). EVs take at least 30 minutes to fast-charge, but can charge anywhere, if super-fast charging isn't needed.
• Energy Efficiency H₂ is much worse, needing far more electrical energy end-to-end: 25-30% vs. 70-90% efficient.

Energy Storage   

Hydrogen could be used as a way to store surplus electricity (to balance supply and demand from renewables). This works, but it is expensive and inefficient (especially at grid scale, wasting 50%+ of the energy won't do). Furthermore, storing grid-scale quantities of hydrogen requires a large volume, and there are significant hazards.
In very remote environments and hard-to-reach locations, far from the grid this may be an acceptable tradeoff. It's also reasonable if the surplus renewable energy is conveted to hydrogen needed by the chemical industry.

Off-Grid Electrical Generators   ()

If you need a large amount of power, off-grid (e.g. at a construction site, or a mine), the conventional approach is diesel-powered generators. These can be replaced with hydrogen, brought onsite in large amounts. This is a practical, and superficially green solution.

The problem is, you didn't fix your CO₂ problem, you moved it. Unless your H₂ was derived only from renewable-energy at off-peak surplus times, your own greener footprint has made somewhere else correspondingly less green.
i.e. environmentally it would have been better if you had stuck with diesel-generator, and let someone else make better use of electricity (that you inefficiently turned into H₂). This calculus changes after we have removed all fossil-fuels from our baseload.

Heating   

It's not fundamentally impossible to modify the design of a given appliance: a boiler or gas cooker to change it from natural gas (methane) to hydrogen, or hydrogen-blend.
However, the differences in the physics, particularly flame-speed (affecting flame-stability), higher flame-temperature , and flame-visibility mean that this is not straightforward. It requires alterations in flow-rates, different fixtures (entirely stainless-steel pipework, better gas seals), and new types of sensors and safety-systems.

Using a 20% hydrogen-blend (by volume) would mitigate these challenges - and indeed most boilers sold now [2024] are hydrogen-blend ready; but there are no 100%-hydrogen-ready boilers currently available.
But hydrogen-blend would only reduce CO₂ emissions by 7%, even if the H₂ were entirely green.

Retrofitting the existing system (both equipment and pipework) for hydrogen is far more complex, on account of leakage and hydrogen-embrittlement.
As a result, converting the entire gas network to hydrogen (or even hydrogen-blend) would be extremely difficult: so impractical as to be nearly impossible.

Chemical Industry   ()

Hydrogen is already widely used in the chemical industry, mostly derived from methane.

Environmentally-speaking, the most exciting new application is manufacturing green steel, where hydrogen replaces coking-coal. This reaction normally requires carbon, whereas the hydrogen could be potentially green.

The Problems with Hydrogen

No natural sources of H₂

There are (almost) no natural sources of H₂. Because the molecule is so light, it always escapes the atmosphere into space.   So, we have to make it ourselves, usually either by:

• splitting natural gas (methane, CH₄), which results in carbon emissions,
• or by electrolysis of water (H₂O), which requires huge amounts of electricity.

Depending on the source, we have the so-called "colours" of hydrogen.

• grey hydrogen – made from methane, or coal gasification; useless for climate-change purposes.
• blue hydrogen – like grey hydrogen, but "potentially" with carbon capture. 
• turquoise hydrogen – methane pyrolysis (yielding solid carbon, easy to capture); this is a very new approach.
• green hydrogen – made from renewable-sourced electricity: but if we already have electricity, why turn it into H₂?
• white hydrogen – rare natural sources of hydrogen; some have recently been discovered.

This means that we are either:

• emitting carbon to make the hydrogen, or
• using renewable electricity inefficiently:   electricity  ⇒  H₂  ⇒  store/transport ⇒  less-electricity.

Both of these are pointless, unless we want the hydrogen specifically for its chemical properties, rather than as a fuel.

  Be particularly skeptical about the claims of blue hydrogen. Energetically, converting the fossil fuel to hydrogen is less efficient than consuming the fossil fuel directly, while carbon capture and storage is expensive, incomplete, inefficient, and does not currently operate at meaningful scale.

Leakage and Embrittlement

Leakage

Hydrogen is a really tiny molecule. This means that it leaks through tiny gaps in pipework, permeates through seals and gaskets, and can diffuse through plastic HDPE pipe.

So the gas infrastructure that is OK for methane is unsafe for hydrogen. We wouldn't just need to upgrade appliances; we'd need to replace almost the entire national infrastructure.
This diffusion through materials also means that H₂ can escape where mercaptan  cannot, so a leak would be harder to detect. Hydrogen fires have flames that are almost invisible in daylight: NASA resorted to brooms to detect them.

Embrittlement

Hydrogen has an affinity for metal, diffusing into the crystal structure and weakening it. This means that it damages the metal that makes up the pipes that transfer it, and the tanks that store it.

This is hydrogen embrittlement, and hydrogen-accelerated fatigue cracking . As a result, existing natural gas pipelines (whether steel or HDPE) cannot be safely repurposed for hydrogen transport. It also means that dedicated hydrogen tanks are expensive and heavy.

Technical problems with changing the gas supply

There are great complexities in every aspect of changing the gas supply infrastructure over to hydrogen.

These are explained in detail in: Energy Science & Engineering: A review of challenges with using the natural gas system for hydrogen:

Hydrogen has fundamentally different physical and chemical properties to natural gas, with major consequences for safety, energy supply, climate, and cost. …
… a transition to pure hydrogen is not possible without significant retrofits and replacements. Even if technical and economic barriers are overcome, serious safety and environmental risks remain.

Practical problems with changing the gas supply

Before anyone can switch, everyone must upgrade, and virtually everything needs to be replaced.   In any area, a single non-H₂-compliant appliance would be an explosion hazard. This means sending gas-engineers door-to-door to check an entire region a time, with no room for errors. If a single overlooked, non-upgraded appliance remained connected to the network when 100%-hydrogen was switched on, it would probably cause a fatal accident.

The UK did it once before, so why not just switch again?
The entire UK switched gas supply before, converting the entire nation from town-gas to natural-gas in 1968-1976. This was a major engineering and logistical exercise .
However, that switch was driven by huge economic and technical advantages of methane, and it was easier and much safer in case of errors, than a methane-to-hydrogen transition.

Recently, even the UK's small scale hydrogen-infrastructure pilot, in Redcar, Teeside, in 2023, was scrapped.
For the same investment, we could achieve better CO₂ results with insulation and heat-pumps.

Low volumetric energy density

While hydrogen has the best energy density per kilogram, it has a very low energy density per litre.

Reducing the volume, for transport applications, is a highly energy-intensive process. These are the options; see appendix I for data:

• Unpressurised gas — via pipelines, used for fixed installations, such as heating. The density is far too low for portable applications.
• High pressure gas — in steel tanks for vehicles. Much higher density, but requires very heavy steel tanks to contain it. It also wastes ∼ 20% of the energy to compress to 700 bar.
• Cryogenic liquid — low pressure and high density, but very cold (-253 °C) so very difficult to handle and store. Used for aerospace. It consumes ∼ 40% of the energy to cool and liquefy.

This makes hydrogen a very poor choice of fuel: less convenient than gasoline, and less energy-efficient (more CO₂ emissions) than battery-electric-vehicles.
The inefficiencies from cooling/compression are mostly due to fundamental physical and thermodynamic properties of hydrogen, so there is little scope to improve them.

Conclusion

Summary

• Cryogenic H₂ is the most-energy dense (per kg) fuel that exists, outweighing the disadvantages in certain special cases: rockets now, and potentially commercial aviation in the future.
• Compressed H₂ for vehicles has little to recommend it, compared to batteries, and there is no environmental benefit.
• Piped H₂ for heating will never replace methane, because we could never practically make a safe transition (and it's better to invest the money on heat-pumps and insulation).
• Chemical H₂, i.e. using hydrogen as a reagent in the chemical industry, has benefits, especially to manufacture green steel without coal.

What's better?

We have to move as fast as we can to electrify our infrastructure, including electric vehicles - and the best options we currently have are offshore-wind (large scale) and nuclear.
Nuclear fission is the safest form of energy ever invented, especially with modern generation IV designs, such as the LFTR.

Solar panels are plummeting in cost, but still best in sunny climates, during the daytime. 
Meanwhile, nuclear fusion remains urgent, but we need it now, so given the ongoing delays in ITER, it's worth considering alternative approaches, such as first light.

For aviation, the energy density requirements mean there is no alternative to a liquid chemical fuel. Thus, bio-kerosene, synthetic-fuel, cryogenic LH₂, or allocate our 10%-emissions budget entirely to jet-fuel .

Therefore…

  Hydrogen, as a fuel is not, fundamentally can not be, and never will be, the answer to our energy problems.
The use in large-scale rockets (where the complexity of cryo-fuels is justified) does not extend to earth-bound applications.

Other perspectives

The UK Parliament's House of Commons Science and Technology Committee report: "The role of hydrogen in achieving net zero", 19th December 2022 [summary] states:

… In our view multiple changes will be needed to the way we obtain, use and store energy if we are to reach Net Zero emissions by 2050. Hydrogen will have its place in this portfolio. But we do not believe that it will be the panacea to our problems that might sometimes be inferred from the hopes placed on it. …

Sabine Hossenfelder explains: Hydrogen will not save us. Here's why.

The Guardian, FOE, and Financial Times describe the hydrogen hype, greenwash, and fizzle.

David MacKay FRS compares energy options in Sustainable Energy Without the Hot Air.

Appendix I.   Thermochemical Data

Here is a comparison of selected mass and volumetric energy densities for burning various fuels:

For comparison, the density of air is ~ 1.2 g/L while the density of water is 1000 g/L. Note that combustion releases its energy as heat; only about 40-50% of this is available to do useful work. 

Note how good gasoline, and other liquid hydrocarbons, (kerosene, bio-kerosene, and diesel), are as fuels. If not for the CO₂, they are near perfect: energy-dense; easy to use, store, transport, process, handle; safe (with care); cheap; ubiquitous — it's no surprise that the modern world thrived on oil. While bio-diesel can never replace crude-oil wholesale (because it requires so much land), we could make synthetic-fuels from the same energy input that we would have made the hydrogen from.

Appendix II.   Heat Pumps

A heat pump works by using high-grade energy (in the form of electricity), to do work to force heat to flow the "wrong" way (against a temperature gradient). So we move a given amount of heat from a cold-source (outside air) to warmer-sink (inside the house). Even better, the energy we put in to do this also appears as heat, in the output. The result can be 400% "efficient", or more.

η   =   T_hot   /   (   T_hot − T_cold   )

For typical houses, this is 20 °C vs. -5 °C, but thermodynamics works in Kelvin, i.e. 293 vs. 268. So the potential efficiency could be theoretically 11× better than a combustion-system, making heat pumps even cheaper than gas.
Note that a domestic boiler that uses radiators will actually circulate water at 50 °C, so the potential efficiency here is reduced to about 5×. Air-conditioners, used in reverse, can do better.

In brief, this works because of the 2nd Law of Thermodynamics, which normally restricts the efficiency of heat-engines; but in reverse, the Universe works in our favour. Electricity is a higher-grade of energy than almost any other form: it has no associated entropy, and so all of it can potentially be converted into useful work. Here are some more details.

As a result, an electrically driven heat-pump will always be better than a conventional combustion boiler.
This remains the case, even if that boiler were driven by hydrogen — which itself is highly unlikely, given the problems with switching.

Appendix III.   Bad Arguments

There are also a number of bad arguments, both for, and against hydrogen. Here are the most common, and their rebuttals.

• Use of existing filling stations. Hydrogen vehicles can be quickly refueled via the existing network of filling stations, already set up to handle gasoline, with minor modifications.
  Yes, that's great if you're the oil industry, seeking to maintain dependency on your highly profitable distribution network!
  For everyone else, one of the great things about electric vehicles is that we don't have to go to a petrol station every time to refuel. Transmitting electricity is far easier; if you need small quantities of hydrogen, it's easier to make it locally.
• The Hindenburg exploded! Isn't hydrogen dangerous and explosive?
  Yes, any improperly handled fuel is hazardous, and liquid hydrogen is particularly tricky, but we've now learned how to handle this safely, and to prevent and detect leaks.
  Furthermore, it probably wasn't just the hydrogen at fault . Because hydrogen is so buoyant, an H₂ fire is less hazardous than other fuels, as it rises while it burns; furthermore, pure hydrogen (unless pre-mixed with pure oxygen) merely burns and does not explode.
• Hydrogen is a great way to store energy during long gaps in renewable energy supply (a dunkelflaute).
  Not really - the amount of energy storage required to make a difference here is truly vast, and energy is lost in the storage-cycle (electrolysis, compression, expansion, fuel-cell). Other technologies, such as large scale battery storage, or molten-salt thermal storage are cheaper and more efficient.
• Burning (any) fuel directly for heat is a good idea.
  Joule-for-joule, heating a building with combustion (gas boilers) and heating with electricity (resistive heaters) are both highly efficient, any wasted energy from the heating process would also appear as heat .
  Gas heating is about 4x cheaper than electricity, simply because of supply/demand and the price of natural gas. But, that's really wasteful. Heat pumps are much smarter; see above.
• Hydrogen is cheap(er).
  Many such arguments, comparing the cost of H₂ with petrol, omit the tax. For now, there are tax incentives for reducing CO₂ emissions (which is a good idea, and should be substantially increased), but remember the fundamentals.
• Natural gas is "as bad" as coal.
  All fossil fuels are hydrocarbons. However, methane (CH₄) has less carbon than oil (typically C₈H₁₆) which in turn has less than coal (mostly C).
  For a given amount of energy, the CO₂ emissions of natural gas are 56% that of coal: 202 kg/MWh and 364 kg/MWh respectively. For more data, see: emissions factor and global, by fuel.
  Therefore, anything that gets us off coal, even building new gas-fired power-stations (in the short term), is a win.

Telos Digital: Insight

Solving the Climate Catastrophe is the great challenge of our time. Among the many good ideas, there are a lot that won't work, and that distract from things that will.

We don't have time for ineffective measures: solving the problem means 50% CO₂-reduction in the next decade, all while increasing energy-consumption 5-fold (as the developing world increases its standards of living).

So we need to be 10× better at energy-vs-CO₂. There is no role for hydrogen fuel in any of these solutions, that doesn't have a technically and environmentally better alternative. We can offer advice to evaluate green ideas, products, and processes.