Some thoughts on the hydrogen economy

| March 31, 2021

There are many applications for hydrogen, the world currently produces somewhere between 70 and 160m tonnes of it per year (depending on the information source) and it is handled with little drama. Most of it is used in oil refining.

However it is expensive and energy intensive because some 90% of it is made by steam reforming of natural gas with high CO2 emissions a) because gas is used to heat the boiler make the steam and b) the byproduct of steam reformed natural gas is CO2. This is known as brown or grey hydrogen. Current price is around A$1.50 – 2.50/kg which works out at around $10-20/GJ of energy content. However that price is directly linked to the natural gas price

So-called green hydrogen is made by electrolysis of clean water. This can can be low carbon if the electricity used to make it is low carbon solar, wind, hydro, geothermal or even nuclear. However at current costs it is about three times more expensive than using natural gas.

There are substantial efforts being made to reduce the cost of electrolysis and there is a reasonable chance that it will match the cost of brown hydrogen by 2030. Already there are tiny amounts of supply available at less than double the cost of grey hydrogen

In the net zero carbon economy, hydrogen has five main applications

 Power generation at times of low renewable supply
 Replacement of gas heating
 Iron production
 Transport particularly long distances
 Feedstock for some chemicals and plastics

Water consumption

Many people worry about where the water is coming from for electrolytic or green hydrogen. Water consumption is actually not such a big consideration. To convert all Australia’s steel production to hydrogen based “Green Steel” requires about 2.5 GL/y. Supplying 10% of all electricity (probably an significant overestimate, see below) water usage is only about 14 GL

The Yallourn brown coal plant alone uses about 14 GL/y not including water used in mining to supply 4% of Australia’s electricity. Fresh water cooled coal power plants and coal mining consume around 500 GL per year. By some estimates the Coal Seam gas industry uses another 200 GL/y At the highest possible use for hydrogen for domestic purposes including heating, metal refining, electricity, fertilisers and ammonia, hydrogen production would need less than 100 GL of water

Power Generation

In power generation it is claimed that we have wind droughts and that battery storage is too expensive and pumped hydro is too difficult. However while these statements have an element of truth, with modern technology in most countries they are more or less irrelevant.

In pre wind and solar grids we usually had sufficient nameplate capacity to supply 40-50% more peak power than was needed and similarly 40-50% more energy per year than the system required.

In Australia, renewables including rooftop solar supply about 28% of electricity, varying between 20% and 34% on any day. Using the same ratios of minimum to average supply, if we built enough capacity to supply 140% of demand, i.e. 5 times as much as we have now, then there would still be sufficient on the worst day to supply 100% of demand for the day.

That does not mean that we don’t need substantial backup for hours at a time particularly in the evening. It means that we need little or no long duration backup that we can’t get from existing hydro. So other than a few niche cases 4-8 hours of battery storage, flexible demand and perhaps a little more pumped hydro would do the job.

Even in the best case, hydrogen supplied electricity will be expensive. If the price of green hydrogen gets down from about A$6/kg now to $2, running that through a gas turbine will produce power at over $150-170/MWh, the fuel cost alone would be around $135. Hydrogen fuel cells and combined cycle gas plants will be cheaper to run but involve much more investment.

Hydrogen production/generation has one big advantage for a power system. If there is a sudden shortage of power, the electrolyser can be quickly ramped down and the stored hydrogen fed into a gas turbine, a fuel cell or even a “reversible” electrolyser. While this rapid response is very valuable it will account for a very small share of the system’s energy over a year

In Japan, Korea, Taiwan and the UK where energy demand density is much greater, there may be a case for hydrogen in winter but in most of the world there is sufficient space for wind / solar / hydro / biomass / geothermal and some 4-6 hour batteries to provide all the power needed at less cost than hydrogen.

Gas Heating

The old town gas systems had about 40-50% hydrogen in them and it is argued that it would be much cheaper and quicker to feed at least some hydrogen into the existing gas networks, than to convert all gas heating to electric.

As most gas appliances can work at 10-15% hydrogen concentration without modification that sound like a good idea, but lets say at a still optimistic $3/kg, that is a hydrogen price of $21/GJ vs natural gas at $6/GJ so a 90/10 NG/H2 mixture is still a 25% rise in fuel price.

On the other hand a customer spending $1,500 per year on gas hot water and space heating could save $6-800/y by replacing the gas appliances with heat pumps for about $7-12,000.

A heat pump driven from a 40% efficient gas generator still accounts for half the GHGs of direct gas heating. That gives the home owner an 5-10% return on investment and a 50-100% reduction in GHGs vs a $300 cost per year for the NG/H2 mix for a 10% reduction in emissions.

In industrial settings once temperatures required are above 120-150 C heat pumps are more difficult but industrial customers pay significantly less for power. Recently industrial power prices have fallen below 6c/kWh. That is equivalent to $16/GJ. They can buy natural gas for $6-8/GJ and hydrogen at $3/kg delivered is $21/GJ.

In some cases a reducing atmosphere is required so even if hydrogen was more expensive than electricity it would still be required in a zero carbon world. But in many cases resistive, induction or microwave heating is more efficient and requires less ventilation so in those cases gas heating will be replaced by electricity rather than hydrogen.

Iron Production

Hydrogen reduction of iron ore is a promising application for GHG reduction, but even at $2kg for hydrogen the resulting steel will be 10-20% more expensive than the coke route but there will be about two tonnes of CO2 reduction per tonne of steel. Iron ore reduction with hydrogen requires about 55 kg of hydrogen per tonne of iron.

Australia currently produces about 5 million tonnes of raw steel, that would require about 275,000 tonnes of hydrogen per year. On a worldwide basis by about 2035 it is expected that there will be around 1,100 m tonnes of raw steel made, the resist will come from recycling.

If 70% of the raw steel was made with hydrogen, that would require about 45 million tonnes of hydrogen. If oil use has fallen by 40% by then the hydrogen freed up from oil refineries roughly balance hydrogen for steel.

To produce the hydrogen required for Australian steel production needs about 13,000 GWh/y, as much electricity as SA produces from all sources today or about as much as all rooftop solar panels in Australia produce. However it can be done.

Using modern wind turbines and solar systems a 50/50 wind solar mix would need about 400 wind turbines and a set of dispersed solar farms about 30% larger in total area than the old Hazelwood mine and power station site.


There is a big debate about hydrogen in transport, it is difficult to see it making much headway in road transport. Advances in energy density and cost falls in batteries have now pushed the economical range of battery powered heavy vehicles out to about 600 km. Outside Japan no-one is pursuing light vehicles seriously and even in Japan, plug in vehicles are significantly outselling hydrogen powered vehicles.

A battery powered heavy vehicle at a “mega-charger”can add 400km during the driver’s compulsory rest stops so within 6-7 years there will be very few journey’s where a hydrogen vehicle will be cheaper or faster to operate than a BEV. The key reason is that at best half the energy used to make hydrogen is converted into power at the wheels

There are two good opportunities, long distance trains and shipping, hydrogen regional passenger trains are starting to appear in Europe but as far as I am aware there are no commercial long distance locos yet. Some experiments are occurring with shunting locos.

At the moment the cost and storage space on the loco for a long fuel load is probably uneconomical and the value in terms of emission reduction per dollar of expense is quite low. In shipping liquid ammonia is looking more promising than gaseous hydrogen at the moment, but either will involve significant investment.

At $2/kg hydrogen is actually about 1.4c/MJ of energy content, about the same price as low sulphur bunker fuel so if the rest of the system can be built economically then hydrogen is a real opportunity for long distance transport depending on progress with batteries.

There are four technical breakthroughs required for hydrogen to be widely as a transport fuel.

• Hydrogen at or below $2/kg

• Cheaper lighter storage. Rather than store hydrogen as a liquid at very low temperatures or a compressed gas at very high pressures, hydrogen can be stored in a sort of “chemical sponge” called metal oxide frameworks (MOFs). If storage of hydrogen in MOFS becomes practical, it could transform the technical aspects of hydrogen as a transport fuel.

• Lighter and cheaper fuel cells, fuel cells can require about 50% less fuel per kWh than current jet engines but at the moment are much heavier and more expensive per kW.

• Lighter motors, power converters and cabling particularly for aircraft. These components are now competitive for trucks, locos and seagoing vessels,

Fortunately progress is being made in all these areas but there is still a long way to go to make hydrogen viable

In air travel hydrogen and batteries are not yet viable for long distances although the flexibility of electric motors can make aircraft design much more efficient so it is possible that over the next twenty years a hydrogen powered long distance aeroplane may become practical even if the weight of the fuel and fuel tanks is greater than those for conventional jet fuel.

Feedstock for chemical production

This is potentially a large opportunity for hydrogen. World production of ammonia is about 175m tonnes per year which requires 30 million tonnes of hydrogen. If Australia could capture 5% of that market. 1.5 million tonnes of hydrogen at $2/kg is $3bn per year. However current ammonia prices are around US$240/ton.

To compete with that assuming that hydrogen is 2/3rds of the cost would mean that hydrogen would need to be about $900/ton or around $1/kg. Without a significant worldwide carbon price it is hard to see green hydrogen being used for large scale ammonia production.

Exports to Asia

Even if costs are brought down to A$2/kg, current estimates for liquefaction transport and re-gasification are in the order of A$2.20- $3.00/kg, i.e. a total of $5,000/tonne landed at a Japanese or Korean port.

Run through a high efficiency combined cycle plant (62% peak efficiency, 55% annual average) that works out at around $230/MWh just for fuel alone, say $280/MWh including maintenance and operation of the plant and amortisation of capital. However Korea and Japan have almost unlimited opportunities for floating offshore wind that is generating power for about A$80-110/MWh.

They can have two approaches, build more wind than they need and curtail some of it so that combined with widespread solar, pumped hydro, existing nuclear and batteries even on low wind days they still have most of the energy they need. The second is to use wind and solar to make their own hydrogen.

Even if their average power cost is triple ours at $60/MWh for renewables, if we can make hydrogen for $2/kg with $20/MWh renewables they can make hydrogen for $3.5-4/kg which is still much cheaper than shipping it from Australia.

As Australia is much further advanced in renewable power generation than Japan or Korea, there may be a medium term opportunity to ship some hydrogen which would probably used to displace some emissions from coal in the 5-15 year time horizon, but it is hard to see that long term hydrogen exports could be a substitute for natural gas or coal revenues.

Extrapolating from the German situation it is not hard to see Japan achieving domestic energy self sufficiency in the next 20 years (not including marine and air fuels). It will be much more difficult for South Korea if it maintains its current economic structure, however the combination of declining population and a movement towards advanced less industrial economic structure, should see Korea’s energy use fall significantly over the next 25 years.

For example Korea’s 51m people use 580 TWh of electricity (11.3 MWh/ person/y). Germany’s 82 million people use 480 TWh (5.8 MWh/p/y) In the UK per capita use is 4 MWh/person across the whole economy.


In summary, while hydrogen will almost certainly fill a vital role in the Australian and world low carbon energy scene, it will be smaller than some of its proponents expect and it will take longer to expand, because in most cases it will remain more expensive than wind/solar/flexible demand which will capture most of the easy to decarbonise markets like electricity production and road transport storage