“The Hype about Hydrogen” by Joseph Romm

This is an unusual book review in the sense that I am NOT going to suggest that you run out to the shop to buy a copy – simply because you can’t as it is no longer in print! It is available, thought, on Amazon Kindle, and second-hand copies like my own can be found (when mine arrived I was delighted to see that it is signed on the inside cover “To Jeff. Keep the Faith, Joseph Romm,  3/10/04”, just a few days after its publication date on the 1st March 2004 - that inscription date is 10 March 2004 btw).

So why am I reviewing this book? For two reasons: first it is a great introduction to many of the reasons why we should be cautious about hydrogen’s role in decarbonisation and secondly because a thorough analysis of the difficulties inherent in a hydrogen economy and what (if anything) has changed over the intervening two decades should provide a very useful insight into current policymaking.

The subtitle of the book is Fact and Fiction in the Race to Save the Climate which I think is an excellent summary of the contents. I should point out here that the main emphasis of the book is an examination of the feasibility of Hydrogen for car transportation, which was being heavily promoted in some quarters in the US around the turn of the century.

It is said that “Those who cannot remember the past are condemned to repeat it”. As we shall see, it seems that lots of folks involved in hydrogen policy have a poor memory, or little inquisitiveness about past lessons. For those folks, or the policymakers, I would suggest that this book is a must read.

First of all, to Joe Romm, Ph.D. 's credentials (I am calling him Joe as that is how he titles himself in his LinkedIn profile). He worked for the US Department of Energy for much of the 1990s, overseeing the work of the Office of Energy Efficiency and Renewable Energy, whose remit included R&D on hydrogen. At the time of writing, he was Executive Director of the Centre for Energy and Climate Solutions. So, it would be reasonable to suggest that he was ideally, if not uniquely, well positioned to form an objective, independent view of hydrogen’s role in decarbonisation.

The 1990’s saw a real upsurge in interest in hydrogen, driven mainly by the growing concern about climate change, and building on technology improvements in fuel cells and methods to produce hydrogen. Indeed around 2000 just about every major US motor manufacturer had a hydrogen vehicle program. Willam Clay Ford Junior, Chairman of the Ford Motor Company said in October 2000:

“I believe fuel cells will finally end the 100-year reign of the internal combustion engine”

In 2003 in his State of the Union address President George W Bush said

“Tonight, I am proposing $1.2 bn in research funding so that America can lead the world in developing clean, hydrogen-powered automobiles.”

So there certainly were high expectations about the potential for hydrogen in cars. But as this book sets out, those expectations were largely hype, unsupported by the very real challenges facing a hydrogen economy and the technology needed to get hydrogen to work in vehicles. In writing this book Joe was trying to expose those challenges and ensure that the US did not invest a large amount of resource in a decarbonisation strategy that would ultimately deliver little, and would potentially distract from other more effective strategies:

Energy efficiency remains the single most cost-effective strategy for minimising CO2 emissions…

Neither government policy nor business investments should be based on the belief that hydrogen cars will have a meaningful commercial success in the near- or medium-term.  

I believe that those statements are as valid today, twenty years on, as when they were written. Hydrogen is, and never will be a panacea for decarbonisation. It just has too many intrinsic challenges to overcome, some of which I will explain in this review.

Fuel cells

The book starts with a broad introduction to the hydrogen hype at the time of publication and a plea for policymakers to be realistic when considering its potential. In Chapter 2 we learn the basics of what makes a fuel cell work. This is the core piece of technology that is used to convert the energy in hydrogen into electrical power – there are others, but we, like the book, will start with fuel cells.

For those who are not familiar with what a fuel cell is, we can think of it as a piece of equipment where two chemicals are combined to produce electricity and heat – in a hydrogen fuel cell the two chemicals are hydrogen and oxygen (or rather air containing 26% oxygen) mix and produce electricity and heat and water. It is essentially reverse electrolysis where electricity is used to produce hydrogen and oxygen (which is usually discarded). Indeed, electrolysis is one of the two principal methods of making hydrogen, which we discuss later. Importantly, we can see that hydrogen isn’t a fuel per se rather it is an energy carrier, made from an energy input.

Fuel cells all follow the same basic design, with the main differences lying in the central electrolyte. The two most common electrolytes are a polymer proton exchange membranes, PEM, or a ceramic solid oxide fuel cell, SO.

In the PEM fuel cell shown above, the hydrogen flows over the anode where  platinum catalyses its breakdown of the hydrogen into its constituents, one proton and one electron. The proton H+ can flow into the electrolyte, but the membrane prevents the electron e- from entering, which instead pass through an external circuit thus creating the electrical power output. The hydrogen then passes into the cathode where platinum catalyses another reaction, this time with the oxygen to produce water.

H2 + O = H2O* (*note that LI doesn't have subscripts for the numbers)

Strictly speaking the electrons don’t actually flow in the circuit, they simply produce an electric voltage or potential due to the difference in charges between the anode and cathode, which is propagated in the circuit. In a fuel cell that potential is less than one volt, so many fuel cells are stacked on top of each other to produce a more useful voltage.

The first major application for a fuel cell was one that I hadn’t heard of before – the provision of high reliability power without voltage fluctuations. He cites the example of a major bank that purchased one of the very first commercial fuel cells because it offered an availability of “6 to 7 nines”, that is to say 99.9999% or 99.99999% uptime, compared to an typical “uninterruptible power supply” with a 4-nines availability 99.99%. The difference between those reliabilities could make the difference between the probability of failure of 63% in 20 years to less than 1% in 20 years.

The other big stationary use for fuel cells being promoted at this time was in home combined heat and power generation. What was envisaged was that one could combine a steam reformer to convert natural gas to hydrogen, then feed that hydrogen into a fuel cell and so convert fossil gas to electricity to power a home as well as heat to provide warmth and hot water.  

Remember we discussed producing hydrogen from electrolysis (a reverse fuel cells), Well the other way is to combine fossil gas, aka methane, which is a carbon atom and four hydrogen ones CH4, with steam, H2O, at high temperature, in a reformer, which strips the hydrogens from the carbon which forms Carbon Monoxide, CO. Then, in a further step called the water gas shift then mixes the CO is mixed with more water to release more hydrogen and create a CO2 molecule.

CH4 + H2O => CO + 3H2 and CO + H2O => CO2 + H2

The problem with manufacturing the hydrogen in situ in a reformer attached to a fuel cell is that it kills the efficiency of the system and adds a lot of cost and complexity (hence maintenance and reliability issues). By the time the gas is reformed and converted to hydrogen which is then used in the fuel cell the total electricity out of the system is about 35% of the input energy, no better than centralised coal power stations.

For home gas-based fuel cells to provide any value from a decarbonisation perspective compared to centralised generation, all the waste heat would need to be used. Given the typical US home’s heat demand that meant that the fuel cell would need to be rated at 0.73 kW, which is actually a very small system which still had to have all those complex components and so would be almost as expensive to manufacture as a big system. To further weaken the case for stationary fuel cells in homes, a centralised fossil gas combined cycle power plant could achieve an efficiency of 55%, so gas would appear to be better used in these than in in-home CHP systems.

Given all these problems it is easy to see why domestic fuel cells were never adopted in any meaningful way. Yet at the time they were being hyped as the best thing since sliced bread and the answer to our decarbonisation challenges.

Chapter 4 of the book then explores how hydrogen is made, which I have already described above. The primary methods are electrolysis and steam reformation both of which require substation energy inputs leading to an efficiency of about 63% and 70% respectively. This brings us to the first of hydrogen’s many problems:

Hydrogen is an energy carrier and there will be efficiency losses making it which means that it will inevitably cost more per kWh than the input energy.  

Fuel cell cars

in Chapters 5 and 6 we start to get into the core discussion of hydrogen as a transportation fuel. Here it cannot be emphasized enough that the road to a hydrogen economy needs to take into account two things:

1.      The process of making the hydrogen and transporting it to and fuelling the vehicle

2.      The technologies in the vehicle to store and then use the hydrogen to produce motion

Joe quotes a senior scientist at BP at the time:

“One thing we know is that if hydrogen is going to make it in the mass market for transport fuel it has to be available in 30 to 50% of the retail network from the day the first mass manufactured cars hit the showrooms. We know the customers must be able to fill the car in about a minute. Safely, with no leaks will telling the kids to keep quiet in the back to the car”

He then goes on the explain that hydrogen’s very low energy density is the core challenge in developing a hydrogen fueling infrastructure. I am using UK SI units here unlike the book which is in gallons etc. Hydrogen is incredibly light (that is why we fill balloons with it). At normal temperatures and pressures, you get 0.08 kg (8 grams) in a m3 [1]. In energy terms that’s 2.8 kWh (at 35 kWh/kg). You would get 747 kg of Petrol into the same m3 volume [2]. While petrol has only about a third of the energy per kg than hydrogen (12 kWh/kg), you would get almost 9,000 kWh of petrol energy in that m3.

Hydrogen per unit volume at normal pressure has 3200 times less energy than petrol.

So that poses a big challenge from the outset, and Joe quotes the 2003 DOE Fuel Cell report to congress:

"Hydrogen storage systems need to enable a vehicle to travel 300 to 400 miles and fit in an envelope that does not compromise either passenger space or storage space. Current energy storage technologies are insufficient to gain market acceptance because they do not meet these criteria."

The first possible solution would be to turn the hydrogen into a liquid like petrol. But because hydrogen liquefies at -253 °C that involves a huge amount of energy to achieve, not to mention complex, multi-stage refrigeration equipment. You would use 30-40% of the energy in the hydrogen to liquefy it. Joe quotes GM as having developed a 90 kg tank capable of holding 4.6 kg of liquid hydrogen. Another problem is that liquid hydrogen has a tendency to “boil off” due to internally generated heat (because a spin isomer orthohydrogen decays into parahydrogen) or because external heat conducts into the storage vessel.  The GM tank had a boil-off rate of up to 4% per day.

Because of the cost of liquefaction, the energy losses, the weight of the tank and boil-off issues, liquefaction of hydrogen was generally dismissed at the time of Joe's book.  

Another way to carry sufficient hydrogen is to compress it. In practical terms Joe concludes that a tank would need to hold hydrogen at a pressure of 10,000 psi, which would take up about four times the volume of a gasoline car (taking into account a fuel cell is more efficient than a gasoline internal combustion engine).

"Let’s digress from the book for a moment and zoom forwards to today’s specification of the world’s first mass production (but low sales volume) fuel cell car, the Toyota Mirai. This has two hydrogen tanks at 70 MPa (10,000 psi, 700 bar) with a total 122 litre capacity (about 3x times that of a typical gasoline car) that store 5 kg of Hydrogen. The tanks weigh 87 kg [3]."

So Joe got it pretty much spot on!

So, we see that we can get hydrogen into a vehicle, just. But that comes at a cost. First of all, we use about 5kWh to compress that hydrogen to 10,000 psi (15% loss) and we end up carrying around, even with today’s composites technology, an additional 87kg of mass which is an additional overhead.

To carry sufficient hydrogen in a car it needs to be compressed to a pressure of around 700 atmospheres (700 bar; 10,000 psi, 70 MPa). That adds cost and weight to the car and consumes about 15% of the energy in the hydrogen.

I should at this point mention that Joe describes a third storage method. This is to store the hydrogen in a chemical form bound to another material and released by heating or catalysis. The most promising of these at the time were metal hydrides, but they suffered from slow charging and discharging and the fact that they may retain some of the hydrogen they store. They also weigh a lot per unit of hydrogen, a tank carrying 6kg of hydrogen could weigh 300 kg. Joe was relatively unconvinced by these technologies, and nothing has happened in the intervening years to disprove that caution.

Hydrogen’s lightness lies at the heart of the other central challenge of a hydrogen based transportation system – how to make the hydrogen and get it to the car cost-effectively.

Well one solution that we have touched on previously is the notion that the car could have a reformer built into it and simply fill up with petrol (and water) which would be converted to hydrogen in situ. This notion of onboard hydrogen production, whether from gasoline or methanol is examined by Joe but fairly swiftly discounted.

First of all studies at the time show the net emissions in a gasoline-fuel cell vehicle were likely to be higher than existing gasoline vehicles. Then they would be bound to cost a lot more than the existing most efficient vehicles at the time, gasoline-electric hybrids. Finally, it was difficult to see fueling companies investing in supplying methanol when a hydrogen-based system was perceived to be the ultimate goal – why would they build two distribution technologies leaving the first to become a stranded asset.

So the focus was on the two remaining approaches. Make the hydrogen at the filling station or make it centrally and then transport it to the filling station.

But while some folks at the time, like the prestigious Rocky Mountain Institute felt onsite hydrogen production would be the quickest way to deploy a hydrogen fueling infrastructure, Joe disagreed:

"In this scenario, a fueling station would need an SMR system, a hydrogen purification unit, a multistage hydrogen compressor (since a high-pressure tank is the only viable near-term system for storage on board a fuel cell vehicle),  system to fuel cars, a significant amount of on-site high-pressure storage, and various types of control equipment."

All of this would make forecourt production of hydrogen much more expensive than a centralised system. Furthermore, the central plant could buy its gas and electricity much more cheaply than a small forecourt, giving it a further advantage.

It is interesting to note here that for Joe the assumed method of hydrogen production was reformation of fossil gas. That is because at the time of writing the US electricity grid emissions intensity was very high at around 600g CO2 per kWh which would have made hydrogen around 950 gCO2 per kWh, taking into account its 63% efficiency of production. Now petrol combustion emits about 230 gCO2 per kWh [2], and even assuming that a fuel cell is twice as efficient as a petrol engine, that would mean that the electrolytic hydrogen option would emit around twice the emissions as petrol. Of course the narrative today is all focused on the notion that “green” hydrogen will be produced by electrolysis. However that doesn’t particularly make forecourt hydrogen production any more attractive since the electrolyser would inevitably operate intermittently and so the capital cost would be higher than a central plant where the equipment could be better utilized.

Having concluded then that hydrogen will probably be best made centrally and then distributed to the forecourts we then get into the logistics of that. Clearly, for the reasons we have described earlier, the hydrogen would need to be distributed in a compressed form, most likely in a trailer carrying compressed hydrogen canisters. But here again we come up against some pretty awful overheads:

"An analysis in 2003 by Ulf Bossel and Baldur Eliasson of hydrogen delivery in improved high-pressure canisters estimated that a 40 metric ton truck would deliver only about 400 kg of hydrogen into on-site high-pressure storage… .. .from an energy perspective, this delivery approach is exceedingly wasteful. A 40,000 kg truck is being used to deliver a mere 400 kg of hydrogen. More than 39 unused metric tons, a tremendous amount of dead weight, would be hauled around our highways, burning up diesel fuel. Compare that with the fact that the same size of truck carrying gasoline delivers some 26 metric tons of fuel (10,000 gallons), enough to fill perhaps 800 cars.

Bossel and Eliasson draw two striking conclusions from their analysis. First, the energy consumed by tanker truck delivery itself can be a very high fraction of the total hydrogen energy delivered. For a delivery distance of 150 miles, the delivery energy equals nearly 20 percent of the usable energy in the hydrogen delivered. For 300 miles, the energy ratio approaches 40 percent."

The alternative option of piping hydrogen to the fuel stations was also considered but this would require a huge installed based of hydrogen vehicles to ensure sufficient demand for hydrogen. Indeed that option may not have worked even if there was a large number of hydrogen cars on the road as the cost of pipelines is extremely high.

This is the chicken and the egg problem that lies at the heart of any major shift in an energy system – for a cost-effective supply system to be developed there has to be sufficient demand, but the demand won’t happen until the new energy carrier is widely available. Electrification of cars has benefited from the widespread availability of electricity supplies and the relatively cheap incremental cost of installing vehicle charging points.

Hydrogen's incredibly poor energy density also creates inefficiencies in the supply chain, and high fueling equipment cost and low reliability.

If we fast forward to today, we have just two mainstream hydrogen fuel cell electric vehicles, FCEVs, the Toyota Mirai retailing for $50,000 and the Hyundai Nexo at $60,000. These are the only hydrogen cars available around the world outside China. BMW have also released a limited number of a combustion engine BMW7.

“Only 14,451 FCEVs were sold worldwide in 2023, compared to 20,704 in 2022 — a 30.2% slump, largely driven by a 55.2% fall in sales in South Korea” [4] By comparison 14 million electric cars were sold in the same year [5], 1,000 time more [5] and growing very rapidly."

It’s not difficult to see the reason why folks looking for an alternative to the petrol combustion engine might favour battery electric vehicles, their efficiency! Assuming that hydrogen in the future should be produced from electrolysis using renewable electricity we can see that the battery electric options offer much more range per kWh because they don’t have all the energy-destroying overheads that hydrogen has as an energy carrier.

The Tesla model 3 or S, which are comparable to the Mirai retail for $44,000 and $56,000 respectively and their prices are falling in the face of greater competition. From an emissions perspective 100 kWh of electricity delivers 85 kWh of energy into a BEV battery, of which 90% is turned into useful work moving the car forward. By contrast only 57 kWh of hydrogen can be delivered to a FCEV for each 100 kWh of electricity and this is around 50% efficient so we end up with 30 kWh of useful work. What this means from a decarbonisation perspective is that we would need more than twice the electricity generation capacity to decarbonise cars with hydrogen than by using the electricity directly into batteries.

Michael Sura posts regularly on LinkedIn about all sorts of aspects of transportation and he has very kindly agreed for me to use one of his excellent illustrations that make the difference between the technologies quite obvious.

The only real advantage of FCEVs is their potential to be refueled much quicker than battery electric vehicles, in around 5 minutes or so. The gap is, however, reducing with 350 kW charges now able to charge a battery electric vehicle to 80% in around 20 minutes. And, of course, there is the added advantage of BEVs being able to be charged in people’s homes overnight eliminating the need to go to a garage at all.

In fact, it is the lack of a charging infrastructure for FCEVs that is one of biggest barriers to their adoption. It is no real advantage if your vehicle can be refueled quickly if you have to make a long trip to the fueling station. One reason there are so few hydrogen filling stations is that each pump can cost up to $1m due to all the compression equipment needed – that’s one pump not one filling station. Because of the high pressures the compressors have high failure rates which means that they are out of action for service for significant proportions of the time. Then there are all the transportation costs to get the hydrogen to the filling station which adds to the costs of the operators, which coupled with very few customers means that the stations can only survive on very large subsidies. It’s no surprise to learn that operators in the middle tier of the supply chain, in the distribution and retail parts of the market, are withdrawing from the market. Shell has, for example, decided to close all its hydrogen fueling facilities in California citing “market conditions” [6].  All four hydrogen fueling facilities in Denmark were closed around the same time, leaving the country with none [7].

Given that it is very difficult to see the economics of fueling supply chain improve, the fact that twice the electricity infrastructure is needed to decarbonise the same number of vehicles and that the cost and convenience of FCEV are unlikely to never match those of battery electric vehicles it is inconceivable that we should continue to put public money behind the development of this technology.

Precisely as Joe Romm has called it out.

Safety

Before my closing comment on the book I would like to share with you some safety tips:

"Hydrogen has its own major safety issues. It is flammable over a wide range of concentrations and has a minimum ignition energy twenty times smaller than that of natural gas or gasoline: Operation of electronic devices (cell phones) can cause ignition" and "common static (sliding over a car seat) is about ten times what is needed to ignite hydrogen," as explained in a 2002 report by Arthur D. Little Inc." Even "electrical storms several miles away" can generate sufficient static electricity to ignite hydro-gen, noted James Hansel, a senior safety engineer for Air Products and Chemicals Inc., a major global supplier of hydrogen, in a 1998 presentation to Ford Motor Company."

Hydrogen is also odourless and burns with an invisible flame and because it is prone to leaking and can catch fire so easily there is a real danger that people might step into the nearly invisible hydrogen flame and burn themselves.

When I worked around hydrogen at BPs refining facility at Texas City the scariest thing was this invisible flame – the walkways were marked out in often quite convoluted routes, the reason being that they avoided any hydrogen pipework where a leak and flame could occur. Joe Air Products safety document which recommends,

"If a leak is suspected in any part of a hydrogen system, it is good protection to hold a large square of paper before you and approach the suspected leak so that any invisible flame will strike the paper before striking you."

You gotta be kidding!

Needless to say, at BP, I stuck very strictly to the marked out walkways!

Hydrogen and Global Warming

Chapters 7 through 9 then consider hydrogen’s role in a wider decarbonisation of the US economy in the context of Global Warming. I won’t recap much of the details here because this is where the book, understandably, is most dated. A big difference today is the emphasis on a decarbonised electricity production system which is possible because of much greater falls in renewable energy costs than Joe assumed.

One interesting commentary was the acknowledgement, ground-breaking at the time, by BP in Lord Brown’s famous 1997 Stamford speech, that climate change was real and that something needed to be done about it.  It was that statement that actually led indirectly to me selling to BP what might be the largest-ever UK energy efficiency consulting assignment around 2000 and the work I and colleagues were doing at Texas City and other upstream assets (more on this in my book – link at the end).

At around the same time Joe Romm recalls that Shell had also seen the light and had developed a number of scenarios to explore what future pathways the world might take. In the second scenario, grandiosely called  “The Spirit of the Coming Age”, Shell envisaged a new “fuel in a box” which was a 2-litre container carrying enough fuel for a vehicle to travel 40 miles and which would be distributed much as one distributes soft drinks today. The fuel would be hydrogen. This would be produced from gas using carbon capture and storage on a huge scale – by 2025 over 1 billion tonnes of CO2 a year would be stored underground. Being a scenario developed by a fossil fuels company, energy efficiency has no role to play.    

Now you have read all the challenges that hydrogen brings simply because of its immutable physical properties and should by now be smelling a rat. Joe puts it thus:

"You can judge how plausible this all is. From my perspective, this scenario should be called not "The Spirit of the Coming Age" but "Deus ex Machina." Every major obstacle to a hydrogen economy and CO, reductions is eliminated with a magic wand. Breathtaking breakthroughs occur quickly in countless technologies - hydrogen production and storage, fuel cells, solar energy, biofuels, and sequestration. Governments willingly spend hundreds of billions of dollars to bring these technologies to the marketplace. Political obstacles to tripling nuclear power production evaporate. Natural gas supplies are seemingly limitless."

Frankly the same fanciful, fact-free, self-interested ideas still pervade the future visions of Hydrogen lobbyist today. But hopefully, having read this synopsis and, maybe, Joe's book you will be able to detect the BS.   

Conclusion

This just doesn’t feel like a 20 year old book. Apart from a few technical advances in storage tanks in cars, the failure of static fuel cells in homes to take off and the rapid greening of electricity systems in many countries, Joe’s analysis could be true today.

Huge subsidies are still being paid by politicians to the hydrogen vehicle industry, hype abounds, vested interests, mainly in the fossil fuels sector, continue to make outlandish claims for a hydrogen economy “just around the corner”.

Joe is to be congratulated for having stuck his head above the parapet and so thoroughly debunked the hype 20 years ago. Its worth reminding ourselves of his central conclusion:

Neither government policy nor business investments should be based on the belief that hydrogen cars will have a meaningful commercial success in the near- or medium-term.  

If, anything, given the supremacy of battery electric vehicles, maybe we should conclude that hydrogen cars will never achieve a meaningful commercial success.  You simply cannot fight against physics!

I very much enjoyed reading this book because of my own interest in debunking hydrogen hype. I think this book should be mandatory reading for anyone involved in the hydrogen economy because we simply cannot keep going around this circle of hype any longer. Folks need to understand that hydrogen’s properties have not changed ion the last 20 years and will not change in the next.

We have made the investment in a hydrogen economy countless times over. It hasn’t worked, primarily because the physical properties of hydrogen make it a very tough energy carrier to work with and the cumulative costs and losses at every stage of its supply, distribution and use make it very costly, and it is dangerous. Its time to move on. Green hydrogen does have a role, but that is primarily to replace emissions-intense grey hydrogen in industrial processes. It certainly has no role in the home or land transport.

References in [] are shown in the first comment to this review.

Folks - if you have read this book please leave your own thoughts in the comments below - you may have picked out different aspects which others would find useful! You may also find my own textbook on energy and resource efficiency helpful - it's free to download :-)

Found this interesting/helpful? This is a link to all the book reviews so far with a brief summary and evaluation.