A shorter version of this article appeared in New Scientist magazine on 26 November 2008.

Whatever happened to the hydrogen economy? At the turn of the century it was the next big thing, promising a Jetsons-style future of infinite clean energy and deliverance from climate change. Hydrogen, it was claimed, would transform the entire energy infrastructure – revolutionizing everything from heat and power to transport. Enthusiasts confidently predicted the breakthrough was just five-to-ten years away. But today, despite ever-worsening news on global warming and the looming threat of peak oil, the hydrogen economy seems as distant as ever.

Even in Iceland, whose grand ambitions for a renewable hydrogen economy once earned it the title Bahrain of the North, visible progress has been modest. After years of research, the country now boasts one hydrogen filling station, a handful of hydrogen cars, and one whale watching boat with a fuel cell for auxiliary power. A four year trial of three hydrogen powered buses ended in 2007, when two were broken up for parts and a third consigned to a transport museum. Yet more trials are planned, but that was before the meltdown of the country’s banking system.

In California, where governor Arnold Schwarzenegger promised a ‘Hydrogen Highway’ of 200 H2 filling stations by 2010, today there are precisely five open to the general public. In London, where 10 hydrogen buses are due to come into service in 2010, a second tender to provide 60 smaller hydrogen vehicles was recently scrapped by Transport for London.

But although the timetable seems to slip constantly, there is still enormous effort going into hydrogen research and development. “Fuel cells have been a rollercoaster of hype and disillusionment”, says Martin Green, Strategic Development Director for Johnson Matthey, which makes fuel cell components for the car industry, “but I am more confident now that the hydrogen economy is going to happen than ever before”.

Real products are now inching closer to market. Honda claims to be the first company with a fuel cell car, the FCX Clarity, in commercial production, although its definition of commercial is something of a stretch. Honda will make just 200 of the cars over three years, leasing them to customers for $600 per month, at the end of which they will have to give them back. So far Honda has shifted three.

Meanwhile GM, whose share price recently plumbed its lowest level since 1950, is testing the first 100 of its Equinox fuel cell model in a free trial for potential customers around the world. The company claims to have spent €1 billion on hydrogen R&D, and its research boss Larry Burns recently predicted the emergence of a small market for fuel cell vehicles by 2012-14. So could hydrogen finally be ready for take off, or will the mirage continue to recede?

Enthusiasts claim the remaining hurdles to a hydrogen economy are not so much technical as financial, and that mass production will bring costs down dramatically. But so far the hydrogen fuel cell – the device at the heart of the entire hydrogen project – has remained stubbornly expensive, in part because of its reliance on precious metals. An automotive fuel cell stack currently contains up to 100 grams of platinum – against 5g in a catalytic converter – costing around $3,250 at today’s prices. So for the hydrogen economy to happen, there is an urgent need to reduce the amount of platinum required in fuel cells.

But if Daihatsu is to be believed, that problem has already been cracked. The Japanese carmaker claims to have found a way to eliminate platinum entirely, by turning the fuel cell concept on its head (see box below).

Click to enlarge. Source: Daihatsu.

A conventional fuel cell needs platinum catalysts because other potential candidates such as nickel would be corroded by the acidic chemistry of the proton exchange membrane (PEM). To solve this problem, Daihatsu and Japan’s National Institute of Advanced Industrial Science and Technology, have developed a fuel cell in which the PEM is replaced with an anion exchange membrane, which is alkaline, meaning the catalysts can be made from common metals that are much cheaper.

A conventional hydrogen fuel cell works by exploiting the mutual attraction of hydrogen and oxygen to produce electricity and water. The cell consists of two chambers separated by a membrane that is permeable to hydrogen protons but not electrons, the proton exchange membrane, or PEM. On either side is a catalyst layer and an electrode. When hydrogen and oxygen are fed into the opposing chambers, the catalyst separates the hydrogen atoms into protons and electrons. The protons pass through the membrane to the oxygen cell, and the electrons travel through the electrode and an external circuit – so creating the electric current – to the other side of the cell, where they combine with the protons and oxygen to form water. Because the PEM is highly acidic, the catalyst must be made of corrosion resistant metals such as platinum.

The Daihatsu cell exploits the same basic attraction but in reverse. Oxygen and water are fed into one side of the cell, and hydrazine hydrate, made of hydrogen and nitrogen, into the other, and the two chambers are separated by an anion exchange membrane. Instead of hydrogen protons passing from the cathode to anode, negatively charged hydroxide ions go the other way, producing an electric current, water and nitrogen. Because the anion exchange membrane is alkaline, the catalysts can be made of common metals such nickel, which are much cheaper.

There is nothing new about alkaline fuel cells per se; NASA has used them in its spacecraft for decades. But traditional alkaline cells have used a liquid electrolyte – rather than a membrane – which is irreversibly poisoned by atmospheric CO2, and this tends to stop the cells from working within a few weeks. Daihatsu’s innovation has been to develop an alkaline membrane which has apparently solved that problem.

In another contrast to existing fuel cells, Daihatsu’s will run not on hydrogen gas, but a liquid fuel, hydrazine hydrate (N2H4.H2O), which is far more energy dense and, the company claims, easier to handle. Better still, says Daihatsu, its new cell produces power output “comparable” to hydrogen fuel cells, and since all that comes out of the exhaust is water and nitrogen, the system is “safe” and “environmentally friendly”. But competitors and experts in the field are deeply skeptical.

Like alkaline fuel cells, there is nothing new about hydrazine, which was first used as a solid rocket fuel for Messerschmitt fighter planes in WWII. A derivative of ammonia, it is also used to make spandex, polymer foams and explosives. In its liquid form, hydrazine hydrate, it is more energy dense than hydrogen gas – which has to be highly compressed to get enough range from even a reasonably small tank – but then hydrazine takes much more energy to produce and so must always be more expensive. Worse still, it is caustic, highly toxic and a suspected carcinogen.

To allay these fears Daihatsu has been forced to devise a complicated fuelling system, in which the fuel tank contains a granular polymer that converts the liquid fuel into a solid, hydrazone, for safer storage. When the fuel is needed, warm water is circulated around the tank to rehydrate the solid and release liquid fuel for the cell. The company claims this arrangement should “[minimize] the adverse effects that any dispersed fuel could have on humans or the environment should the fuel tank be damaged during a collision”. But Martin Green of Johnson Matthey, for one, is not convinced. “Daihatsu seems to be suggesting that this fuel is comparable to most industrial chemicals, but that stretches PR artistic licence beyond breaking point. “It’s an interesting idea”, he concludes diplomatically, “but I can’t see it leaving the lab as a real alternative”.

So it looks as if fuel cells will continue to need platinum for the time being, which raises the question of whether there is remotely enough of the element to permit a large scale shift to hydrogen fuel cells. Green dismisses such concerns because he is convinced that carmakers will be able to slash the amount of platinum needed to just 20 grams per car by the time the technology is commercialized, which he foresees in the middle of the next decade, and because platinum can be recycled. Yet still the numbers still look daunting.

Global car production in 2007 was just over 71 million, and at 20g per car a wholesale shift to hydrogen fuel cells would need 1,420 tonnes of platinum per year, six times current platinum production. At that rate – if it could remotely be achieved – the world’s total resource of platinum group metals would be gone in 70 years, although of course output would peak and decline long before the resource was exhausted. And that calculation makes no allowance for any growth in car production.

“Platinum is really scarce, and only produced in five mines around the world”, says Professor Armin Reller of the University of Augsburg, who advised the Swiss government on hydrogen for almost twenty years. Having recently made a study of the potential resource constraints on a range of new technologies, Reller is convinced that such bottlenecks mean hydrogen can only be a partial solution at best: “when you introduce new technologies the dynamics are such that even if you have the reserves you can’t produce them in time”. So it looks as if the quest to replace platinum will have to continue.

For the hydrogen economy to happen, industry must not only come up with cheap technologies to exploit the fuel, but also cheap ways of producing it. Most hydrogen is currently made in refineries by steam reforming of natural gas, but that gives off carbon dioxide, meaning that the climate benefits of fuel cell vehicles are scarcely better than those achieved by petrol hybrids such as the Prius, according to a 2003 study led by Malcolm Weiss at MIT. To produce hydrogen cleanly and in bulk will almost certainly mean using low-CO2 power generation to electrolyse water, and electrolysis is both energy-intensive and expensive.

Here too there is enormous research effort going on, and no shortage of claimed breakthroughs – although some seem to have more to do with academic fundraising than fundamental advances in science. But now a small British company, ITM Power, says it has found a way to slash the costs of critical components in electrolysis, allowing it to produce a small-scale electrolyzer that will eventually be so cheap that every home could have one. This in turn would solve the stubborn problem of how to distribute hydrogen, without the need for expensive new pipelines, by opening the way for decentralized production of the fuel close to where it will be consumed. And all this because the company has invented a new material that looks like a large, thick contact lens, which it says solves a long standing conundrum of electrolysis.

Industrial electrolysis has long been based on enormous cells containing a liquid electrolyte such as potassium hydroxide. The advantage is that since the chemistry is alkaline, cheaper nickel catalysts can be used. But the disadvantage is that with a liquid electrolyte it is harder to keep the hydrogen and oxygen separate – which is important since they can make an explosive combination. The degassification equipment needed to make the cells safe also makes them bulky and costly.

During the 1960s space programme, NASA developed fuel cells using proton exchange membranes, and the same technology was then applied to electrolysers. The advantage was the oxygen and hydrogen were kept safely separate by the membrane without the need for bulky degasification equipment, but the disadvantage was the membranes were acidic – so again, the catalysts had to be platinum – and the membranes themselves remain furiously expensive. “An alkaline membrane is the holy grail of both electrolysis and fuel cell technologies” says Jim Heathcote, chief executive of ITM Power.

Now, using little more than “bucket chemistry”, ITM Power claims to have found it. After four years and 27 thousand attempts, the company has discovered a way to combine half a dozen commonly available hydrocarbons to produce clear liquid, which under UV light cures into a solid but flexible polymer gel that is three times more ionically conductive than existing proton exchange membranes. Because of its simplicity, and because it is made from hydrocarbons rather than fluorocarbons, it should also be massively cheaper.

The company claims that under mass production its membrane would cost just $5/m2 against $500/m2 for existing PEMs. And because the polymer can made either acidic or alkaline, it also dispenses with the need for platinum in small-scale electrolysis. As a result the company claims its electrolyser stack – the collection of cells that forms the working heart of the machine – would cost $164/kW capacity against a current average of $2000/kW.

Unfortunately, that’s not all there is to an electrolyser, because the company’s “Green Boxes”, about the size of an old-fashioned telephone box, also need pressure pumps, heat exchangers, safety systems and other bits a pieces, collectively known as ‘balance of plant’ – which also increase cost. Jim Heathcote refuses to reveal the price of the first ten units that are currently being produced – though it is certainly tens of thousands of pounds each – but claims that mass production will bring it down to “less than the cost of an Aga – £5,000 to £10,000”.

This is the key to the company’s vision of decentralized hydrogen production – with no need for an expensive new hydrogen pipeline system. Home electrolyzers would be connected to the water mains and at least partially driven by solar panels on the roof or perhaps a micro wind turbine. This would produce hydrogen for cooking, heating, and to drive a generator or fuel cell to produce electricity for light and power. It could also produce fuel for a car, powered either by a converted internal combustion engine or a fuel cell. All of these applications are demonstrated in a showroom at the company’s factory in Sheffield, and Heathcote argues this kind of setup would not only be low carbon, but would also reduce household reliance on gas and electricity grids, which he expects to become increasingly unreliable.

But the extent to which households could ever become meaningfully energy independent using ITM Power’s electrolyser is highly questionable, a point illustrated by Heathcote’s own home, where he has installed one of the largest domestic solar arrays in Britain on the roof of a large out-house. The panels cover about 60 square metres, which is more than twice the average size, generate about 10,000 kilowatt hours per year. If they were connected to an electrolyser that was 60% efficient (ITM Power claims a range of 50%-70%), they would produce hydrogen containing 6,000 kWh annually, whereas the average house in Britain currently consumes almost four times that much energy in gas and electricity. The panels alone would cost around £50,000, never mind the electrolyser, fuel cell and other equipment you would need.

If that same hydrogen were used to power ITM’s converted Ford Focus, the results would be scarcely better. The car does about 25 miles per kilogram of hydrogen, meaning you could drive about 4,500 miles a year, around half the annual mileage of the average British car. And there are other drawbacks. The fuel tank can only hold 1 kilogram of hydrogen at the pressure delivered by the electrolyser (75 bar, or 1125psi), and it takes eight hours to refuel. That means the daily range of the vehicle is also 25 miles. To extend it would mean adding a compressor to the system to push more hydrogen into the tank, which would itself consume energy and further reduce the overall mileage, not to mention the additional cost.

“For the average person it does sound absurd” concedes Heathcote, “but that’s how every technology starts, there are early adopters, and then mass production brings costs down hugely”. He accepts that it is very unlikely that many homes will ever go completely off-grid, but remains convinced that with excellent insulation many could use the ITM Power approach to provide most of their household energy. He also admits that hydrogen cars will probably never be powered by the roof of the house, but maintains the fuel could still be produced by a home electrolyser running on off-peak nuclear power.

If ITM Power has found a way to slash the costs of electrolysis, nobody has yet solved the more fundamental problem of the spectacular inefficiency of the hydrogen fuel chain, with profound implications for the amount of energy that would be required – a point made forcefully by Dr Gary Kendall in a recent report for the World Wildlife Fund for nature called ‘Plugged In’.

In the report Dr Kendall, a chemist who previously spent almost a decade working for ExxonMobil, highlights how the energy losses of the hydrogen fuel chain – from electrolysis, compression and the fuel cell itself – mean that less than 25% of the original energy does any useful work on the tarmac (see graphic). By contrast, the fuel chain for battery electric vehicles and plug-in hybrids – with no electrolysis or compression to worry about – suffers much smaller losses, meaning almost 70% of the original energy is used productively. “Cars running on hydrogen would need three times the energy of those running directly on electricity, and that would force us to build many more wind turbines”, says Kendall. “The developed world needs to completely decarbonize electricity generation by 2050, and that in itself is a monumental challenge, so we can’t afford to just throw away three quarters of the primary energy turning it into hydrogen”. The point is supported by the results of another study, conducted by the consultancy E4Tech for the Department of Transport, which found that a if Britain were to switch to battery electric vehicles, electricity demand would rise by 16%, whereas switching to hydrogen fuel cell cars would raise it by 34%.

Click to enlarge. Source: ‘Plugged In: The End of the Oil Age’, Dr Gary Kendall, WWF, March 2008

Of course, battery electric vehicles also have their drawbacks: limited range, long recharge times and – like hydrogen fuel cell vehicles – the absence of a widespread recharging infrastructure. But rapid progress is being made on all these fronts, and smart money is now pouring into the sector – so long the poor relation to hydrogen.

In California, Tesla is now producing a battery electric sports car with a range of 220 miles on a three and a half hour charge, whereas the average American drives around 40 miles per day (or at least she did before the credit crunch). In Japan, Tokyo’s local electricity company, Tepco, has developed a roadside recharger that it says will deliver 60km range in ten minutes; Mitsubishi will release a five-door battery electric hatchback next year, with Nissan-Renault and Subaru close behind; and one local authority has promised to install 150 of the new rechargers, and to have 3000 electric vehicles on the road within five years. And in countries as far flung as Israel and Denmark, the Better Place company founded by former software executive Shai Agassi is developing plans for an all-electric car infrastructure using quick rechargers and battery exchange stations.

Battery electric vehicles have also won the interest of the legendary investor Warren Buffett, whose MidAmerican Energy subsidiary recently bought a $230m stake in the Chinese battery electric carmaker BYD. MidAmerican’s chairman David Sokol said he was not opposed to hydrogen in principle, but stressed that only battery electric transport would deliver the necessary emissions cuts in time and cheaply enough. “Battery electric technology is critical to achieving the major CO2 reductions that the world is looking for over the next 10-15 years” said Sokol, “as the economics pencil out today, hydrogen still has a ways to go”.

It’s true that carmakers continue to invest millions in hydrogen R&D, but some, like GM, are hedging their bets by developing both fuel cell and battery electric models. Lars Peter Thiesen, GM’s director in charge of hydrogen strategy, acknowledges the greater efficiency of batteries, but insists that hydrogen, with superior range, will be the eventual winner. Yet he remains cautious about when: “If there’s enough money, if the technical development continues as it has for the past few years, and if the stakeholders – not just car companies, but in politics and energy – are all on the same page, then it really could happen in the middle of the next decade”. In other words, the hydrogen economy is still 5-10 years away, which somehow has a familiar ring.

Other industry observers are far more equivocal. “The jury is out on battery versus fuel cell”, says Richard Wenham, a director at the car industry consultancy Ricardo, “that’s why everybody is researching everything”.

But for all the research into hydrogen, fuel cells remain dependent on platinum, the fuel chain is still punitively inefficient, and battery electric technologies are making big strides all the while. So the jury may not be out for very much longer. According to Gary Kendall, “hydrogen has always been the fuel of the future, and it looks like it always will be”.


  • Your conclusions are spot on, except that the Jury will stay out much longer than you think. The vested interests in hydrogen, particularly at the universities, are immense. So expect many $ billions will still be spent, many air-miles traveled and many conferences in agreeable locations will be heavily subscribed by attendees paid for by unknowing tax-payers.

    The uncomfortable facts are clear. Hydrocarbons, now at near peak production flows, have delivered immense benefits while they were/are still around. We are no where near “decarbonising” the Global economy because we still have no serious technical solutions, let alone any that are near economic.

    Enthusiasm, even if it is totally genuine, is no substitute for analysis. And your analysis is excellent.

  • Lucas

    I do agree with most of your remarks and to my opinion, hydrogen is at least premature (you can better start with solving electricity and heating).

    However, there are a few things missing in your report.

    Making hydrogen with electrolysis is the most inefficient way. There are some alternatives.

    – First of all, if you have a lot of renewables for electritiy (not yet the case :-), you will have some surplus at certain moments. Use that to generate hydrogen. Hydrogen generation with peak power is of course horrible (and use it for heating or cooking????)
    – Second, you can use heat to make hydrogen. This can be with a closed chemical loop or High Temperature Electrolysis. The last one uses a SOFC, which does not contain platinum. For the heat you need 800C, which can come from a nuclear source (you need a high temperature reactor) or concentrated solar (point concentration). The Austalian company Solar Systems is busy with a demonstration project.
    – Third, much more experimental, direct use of solar rays, photolysis (see http://www.nanoptek.com).

    Also another observation. If oil realy starts to deplete we have certain options:
    – Go electric, we need more power stations.
    – Free up other carbon usage, for the car. Use natural gas or methanol from coal for the car. Implies, replacing power stations with renewables.
    – Third use hydrogen.

    But where do we get the hydrogen? We have two options:
    – Generate from renewables, no change in power stations.
    – Generate electricity from renewables and use the carbon for generating hydrogen.

    The second option is cheaper. For natural gas (CH4) it is just stripping the carbon atom from the molecule. For coal it is more complicated. Use C with H20 to generate CO and H2. The C steals the O from the water.

    So, when we start using hydrogen now, we still have to free up other carbon sources in other sectors. Why not use natural gas? As in the pickensplan?

    If hydrogen is viable, it is better to wait until carbon sources are so expensive that we can directly start with hydrogen from renewables (which is not before 2030 to my estimate).

    There is one niche sector which might be interested. Off-grid fully renewable. The cheapest way for electricity is a combination of batteries and hydrogen. Full battery usage is more expensive.

    ITM power has a good product, but their story is nonsense. For heating use heat pumps, solar thermal or deep geothermic. Hydrogen is way too espensive. For cooking, just use direct electricity. Cooking is not very much energy.

  • Stephen Watson

    … and not forgetting that all those new cars – either battery or hydrogen powered – need huge energy in their body construction and oil for interior plastics (everything!), tyres, greases and lubricants, paintwork, cabling and of the course the continuing road network investment.

    It’s about a lot more than just changing the engine so we really need to rethink the way we travel instead of trying to continue with cars for as long as possible.

Leave a Reply

Your email address will not be published. Required fields are marked *