I wrote this last year. I maintain the opinion that a car using hydrogen fuel cells won’t be mass-produced. However, owing to solid oxide fuel cells and an abundance of excess renewable energy it might be useful as energy storage.

The demise of a dream – the end of an ill thought out idea.

To the casual observer, hydrogen is a clean fuel. The only products of its combustion in air are energy and water. No greenhouse gasses emitted, no soot released. Just energy and water.

That portrait of hydrogen is misleading. Owing to the way over 80% of hydrogen is produced, a modern hydrogen fuel cell vehicle emits more CO2 than a typical internal combustion engine. This is because most hydrogen isn’t extracted from water, but from natural gas in a process called steam reformation.

 The principle of steam reformation is this, take methane and water. Put them in a pressure vessel with a nickel catalyst at 900 degrees C and 300 Psi, allow to react. The reactants will form an equilibrium producing either carbon dioxide and water or three hydrogen molecules and one carbon monoxide.

The carbon monoxide can then be hydrolysed with water to produce one further hydrogen molecule and carbon dioxide. The final products are 4H2 molecules and one CO2. The hydrogen is then separated from the CO2. The excess carbon dioxide is released into the atmosphere or used for industrial processes. The whole process produces around 12kg of C02 per Kg of H2 produced.

If this 1kg of hydrogen was used in a fuel cell vehicle it would travel approximately 60 miles. If we assumed that a petrol vehicle does 60mpg (British gallons 4.5L). Then over the 60 miles, the hydrogen vehicle would have released the equivalent of 12.3kg of CO2 into the atmosphere, whereas the petrol car would release 10.35kg of CO2. Not accounting for pressurisation of hydrogen or production of the petrol. The hydrogen produced by steam reformation produces 15-20% more CO2.

There is a theory, that because the carbon dioxide emissions of hydrogen are produced at a single location, rather than out the exhausts of thousands of vehicles, it could be captured and pumped underground. Thus there would be no CO2 emissions from a hydrogen vehicle. It’s feasible, but without grants, it’s economically unviable. Then there is the question of whether humanity should be pumping our problems underground.

There will be some of you thinking, “That doesn’t matter in the future hydrogen will be produced by electrolysis using renewable energy, it will be truly clean. Water and electricity in hydrogen out.” I’ll grant you it’s a very nice idea but it isn’t that simple. Not because it can’t be done, but because the economics of it currently makes no sense.

Making hydrogen by electrolysis is incredibly energy intensive. To make 1kg of hydrogen with a 100% efficient electrolyser requires 39.4kWh of energy. Accounting for inefficiencies that figure is closer to 50kWh. Then to utilise in a car the hydrogen must be pressurised. The Department of Energy calculates that 3kWh of energy per kg is required to compress hydrogen to 700bar (10,000psi). Therefore, to produce, and compress 1kg of hydrogen would require 53kWh of electricity.

Put another way, a Honda Clarity has a 5kg, 10,000psi hydrogen tank, to fill it with hydrogen, made via electrolysis, would require 265kWh of energy. It can travel 68 miles with 1kg of hydrogen which requires 53kWh of electricity to make. On the equivalent electricity, 53kWh, a Tesla Model 3 with an EPA calculated efficiency of 27kWh/100miles can travel 200 miles.  Almost three times further. Logically then the cost of fuelling a fuel cell will always be a minimum of three times greater than for a true EV. This doesn’t factor into account storing and transporting hydrogen.

Clearly, excess renewable power capacity could be used to electrolyse water, creating hydrogen which could be stored and used during times of low renewable output. But this plan rests on the assumption that renewable capacity will be built in considerable excess, and they will sell their energy to the grid for almost nothing.

It fails to account for basic economics – supply and demand. The more energy generating capacity is built the lower the price of electricity to the grid will be. The lower the price of electricity the less profit the generating capacity can make. Less profit means it is less likely that more will be built. Furthermore, the assumption that this would work is based on the idea that no competing storage methods would exist. However, if battery production dramatically increases, or other technologies like molten salt storage mature, then the benefits of hydrogen storage are reduced. Both these storage methods have full cycle efficiencies close to 90%. Whereas using electrolysis to separate hydrogen and oxygen from water, then recombining that in a fuel cell is around 35-40% efficient.

The rough calculation is as follows: 50kWh to separate 1kg of hydrogen by electrolysis. The specific energy of hydrogen is 39kWh/kg. Assuming a fuel cell is 50% efficient then 19.5kWh/kg of hydrogen could be extracted. Giving an efficiency of 50kWh/19.5kWh = 39%

Using hydrogen as an energy storage method requires that there is little competition for purchasing excess electricity produced by the grid. As battery storage operators can pay more for excess electricity and still make a profit. The question that will evolve over the next few years, is just how much battery storage will be connected to energy grids.

Hydrogen has another hurdle, batteries have significant theoretical room for improvement, hydrogen doesn’t. It is a fuel, it is a known quantity and well understood. There aren’t significant efficiency gains to be made in its creation or use, it will always be an energy-intensive fuel. Whereas batteries will always be an efficient way to store electricity.

One situation where hydrogen created by electrolysis might be viable would be if giant solar plants are built in arid regions (deserts near the equator. Here the sun is consistent and intense, leading to predictable energy production. Any excess hydrogen could be shipped around the world in LNG type tankers. This would allow for storage, it would create no CO2 emissions, and it would give a lifeline to the countries whose oil reserves will eventually run out. Plus the hydrogen could be burnt in traditional gas turbines meaning no new major infrastructure is required – other than replacing every gas pipe in existence since hydrogen has a nasty habit of leaking out of everything.

There is an idea floating around that stream reformation could be used to extract hydrogen from natural gas then the carbon dioxide could be pumped underground as a clean way to produce energy in gas power plants. But clearly, this is nonsensical. It creates an extra step where none is needed. Why convert the methane to hydrogen at all. It would be cheaper and easier to build a gas powerplant with a CCS. The carbon is still captured.

There is one way that hydrogen production via steam reformation might make sense. That is if boilers, and cookers, which traditionally run on methane, were retrofitted to run on hydrogen instead. This would mean that CO2 which is produced at individual boilers and cookers all around countries, would instead have a single source. Thus it could be stored. However, due to the leakiness of hydrogen, and it’s proclivity to explode, the cost of retrofitting a national natural gas grid to use hydrogen would be absurdly high. It would probably cost less to build new renewable generating and storage, then transition all the gas cookers and boilers to electric power. Not to mention the fundamental problem with this idea: the advantage of natural gas is that it is cheap – producing hydrogen by steam reformation is not.

There are other theoretical ways of producing hydrogen including using high-temperature nuclear reactors and biological methods, but these are not mature technologies and they may never be. The problem with these methods is they require considerable funding (many billions for nuclear) before they become commercially viable, and even then they may not be competitive cost wise with other technologies.

So who killed the hydrogen economy and the hydrogen car? Nobody, it has never been a viable option and it probably never will be. Storage of energy using hydrogen will always be around three times less efficient than battery storage, and barring electricity grids with massive amounts of excess renewable capacity, that is a hurdle that is incredibly difficult to overcome.