The Devolution Of Hydrogen For Energy Over 25 Years Is A Fascinating Tale
About 25 years ago, hydrogen was the solution of choice for climate-aware technocrats and politicians, and with good reason. At the time, there really wasn’t much choice in terms of low-carbon energy carriers. Batteries were good enough for laptops and phones, but clearly no one was going to be running transportation, heating or grid storage with them.
And besides, you could make hydrogen with electricity, something first done in 1800, and a staple of kids’ science classes as early as grade 4. Easy to make, high energy density and you didn’t even have to burn the stuff. You could use fuel cells, and again, those were really old technology with the first one constructed in 1842, and fuel cells used in Gemini spacecraft as far back as 1962. What’s not to love?
The mood was best captured by American economist and social theorist Jeremy Rifkin in his book The Hydrogen Economy: The Creation of the Worldwide Energy Web and the Redistribution of Power on Earth. What’s that? You didn’t remember the rest of the title? No one does.
It’s worth pointing out that Rifkin, while undoubtedly brilliant and a great thinker, has no engineering or science chops to speak of. He never did rigorous cost workups on hydrogen electrolysis facilities, renewable power and the grid costings or worked through the implications of distribution of hydrogen in any depth. He was seduced by the molecule, and was deeply influential in both Europe and America regarding it, for example getting the then President of the European Commission to commit to a multi-billion Euro research and development plan to turn Europe into a green hydrogen superpower.
Is he still as bullish on hydrogen as he was in 2002 when the book was published? Not so much. Comments in the last year have seen him retrenching to a position where only long-haul logistics see any need for hydrogen fuel cell vehicles. He hasn’t quite got the full memo yet.
Even as early as 2005 it was obvious that there were serious problems with this vision. Iceland ran three hydrogen buses for four years, funded by EU money of course to the tune of €12 million in 2023 money. Did they keep the buses on the road when the money ran out? Not a chance. Far too expensive, far too failure prone.
Has anything changed with hydrogen buses since then? Yes, there have been hydrogen bus trials on multiple continents which recreated the Icelandic experience. Lots of governmental money, very high costs to fuel, very high maintenance costs, abandonment when funding is withdrawn.
Recently I stepped through California’s more recent six years of experiences, finding that after several years of operation, their fuel cell buses cost 50% more per year to maintain than their diesel buses, and about twice as much as their battery electric buses. Further, California’s hydrogen refueling stations in the last six month for which data was collected, the first half of 2021 after years of operation when lemons should have been eliminated and maintenance optimized, were out of service 20% more hours than they were actually pumping hydrogen and were costing 30% of capital expenditures per year in maintenance, an order of magnitude above assumed costs.
Meanwhile, electric buses have taken over the world. China has approaching 700,000 on their roads, compared to a couple of thousand fuel cell buses, mostly in Foshan, a city which made an industrial and export policy choice to focus on hydrogen for transportation. Even Foshan has more battery electric buses than fuel cell ones however. Europe is buying thousands of battery electric buses annually, about a third with at least Chinese frames, drive trains, motors and batteries and often the entire bus being delivered from the other side of the world. African countries are buying thousands of electric buses, not hydrogen fuel cell buses. India has thousands on its roads already and is aiming for 50,000 by 2027.
That hasn’t stopped transit agencies, funded by governmental money at various levels, from getting into trouble by buying hydrogen buses. The latest appears to be Spain, where the five Mallorcan buses have been sitting unused and unusable because refrigerant leaked in and destroyed the fuel cells — one of many problems with hydrogen vehicles is that fuel cells demand pure hydrogen and pretty pure air and are deeply intolerant of anything else — and only four of eight Barcelona buses are operational, 18 months or so after they arrived. Despite these failures and the success of their battery electric buses, Barcelona is ordering more hydrogen lemons with EU funding.
In the 2000s, Formula Student, a global initiative where mechanical engineering students at different schools build small cars and compete on about fifteen different criteria including acceleration, handling, economic viability and the like, had a new category for zero emission vehicles. For the decade, a bunch of schools chose hydrogen and a bunch chose batteries. At the end of the decade, battery electric cars were threatening the conventional internal combustion engines. Early in the 2010s, one won overall. Last year, a battery electric entrant accelerated from zero to 100 km/h in 0.956 seconds within a distance of 12.3 meters. As of last year, there was exactly one hydrogen entrant left, and it likely won’t last.
In 2003, a couple of things are worthy of note. Amory Lovins, who did amazing work and founded the RMI, thought hydrogen was the answer, including for his Hypercar, one of his few failures of analysis, as the Hypercar was a hyperefficient hairshirt in vehicular form. Meanwhile, on the other side of the Rockies, Tesla was founded with the premise that they were going to build fully electric cars that drivers would salivate over.
The original Roadster run of 2,500 was sold out as rapidly as they could get them off the factory floor, most pre-sold in fact. The Tesla Model S’ arrival in 2012 was the death knell for the internal combustion car. Toyota clearly didn’t understand this, introducing the hydrogen-powered Mirai in 2014. Over half of the 23,000 delivered were sold in the USA, specifically in California where there were the often-not-working hydrogen refueling stations mentioned earlier. The sales of hydrogen from those stations make it clear that California’s hydrogen vehicles travel an average of 15 miles per day, well under the 37 miles per day that Americans typically drive.
Now we’re at a point where every legacy manufacturer and a bunch of new ones, especially from China but also Vietnam, have multiple electric cars in their lineups and only a couple still hold onto the clean-burning torch for hydrogen. Some, like Toyota, are leaning into hybrids, but most are going fully electric. It’s hard to find a serious analyst who thinks that there’s a place for hydrogen in light vehicles, although firms like Ballard and Plug Power continue to pretend otherwise.
One of the great hopes of gas utilities is that they would be pumping hydrogen through their pipelines into homes and buildings for heating and cooking in the coming decades. They want to start by blending hydrogen into pipelines with natural gas in what would have to be homeopathic amounts for a few obvious technical reasons. They’ve invested a lot of money into trying to convince politicians that this is a reasonable idea, when it really isn’t.
Many hydrogen enthusiasts were looking forward to continue to cook with gas, albeit with a paler blue flame. They weren’t thinking through the implications of having a gas inside their homes that is much more flammable over a much broader range of concentrations that would ignite with lower temperature sparks and that couldn’t have odorants added to it. Safety studies make it clear it’s four times more likely to cause explosions and harm than natural gas, something that destroys about 4,000 buildings a year in the USA at present.
Even hydrogen-focused Japan realized that this was foolish, hence the reason that some of the biggest heat pump companies in the world are Japanese, Mitsubishi and Daikin among them. Traveling around many parts of the world that aren’t Europe or North America, you are much more likely to find induction stove tops than anything else.
While low-carbon hydrogen would either require throwing away half of the energy in natural gas at great expense, tripling or more the cost of heat, or having green hydrogen made from low-carbon electricity at even greater expense, heat pumps get three units of heat from the environment for one unit of electricity. On average, they are four times more efficient than natural gas furnaces, lowering heating bills in many places and providing air conditioning as well. And induction stove tops heat up only the pot or pan, providing the instant heat of gas stoves with none of the risks and more efficiency than older electric stove tops.
There are now 54 independent studies which make it clear that hydrogen has no place in homes or commercial buildings. And if it has no place in buildings, there’s no future for gas utilities and there will be no convenient and inexpensive hydrogen network used for a multitude of purposes running through cities. Dreams of using that non-existent network to bring hydrogen to truck stops are just that, dreams. District heating and cooling systems that increasingly use massive ground or water source heat pumps are a much bigger growth segment.
There’s a strong trend emerging, isn’t there. That widespread hydrogen for energy economy is getting narrower and more constrained with each passing year.
We’re now at a point where there are only a few places where hope is being held out for hydrogen. While buses are an answered question where hydrogen is only being kept alive by long-running governmental bureaucracies created to dole out money for hydrogen transportation, many consider freight trucking to be an obvious fit. It’s not nearly so obvious once you start looking at the data more clearly.
There are now multiple heavy truck manufacturers delivering battery electric semi tractors, including Tesla. And, of course, Tesla’s performs far above the rest of them for a lower price point, leveraging the massive economies of scale for batteries, electric motors, power management systems and high-powered fast charging its dominance in light electric vehicles provides. It also avoided the trap other manufacturers fell into trying to economize by reusing existing semi ladder frames instead of building the vehicle from the ground up to be electric.
That’s why in September 2023’s NACFE Run on Less, the Tesla trucks were able to put in full days of work covering over 1,000 miles (1,600 km) with a couple of recharging breaks. Tesla is making the megawatt scale charging standard for its Supercharger network, and remember that in North America, Tesla’s plug is now the standard. Legacy truck manufacturers have a choice. They can either follow Tesla down the pure electric road, or continue to faff about with batteries, hydrogen or diesel in the frame optimized only for the latter energy carrier. If they persist, their trucks will be more expensive and less capable than not only Tesla’s, but also Chinese vehicles.
Recent total cost of ownership studies including the fatally flawed International Council on Clean Transportation’s November 2023 effort put every thumb on the scale possible for hydrogen trucking and still found that in every single category of trucking from lightest to heaviest, battery electric trucks have the best TCO, lower than diesel and lower than hydrogen even leaving the thumbs intact. Those studies have also been using 3% and 4% of capital expenditure for refueling stations as the annual cost of maintenance, based on guesses from the mid-2010s that became enshrined in report after report because no one bothered to look at the real maintenance data that was available from Europe and California. As noted, 30% is California’s experience, adding over US$9 per kilogram dispensed by itself.
Organizations not realizing that battery electric has already won pretend trucks can’t drive far enough or that batteries are too heavy. The ICCT again thinks batteries won’t achieve 500 watt hour per kilogram energy densities, double Tesla’s, until 2050. Meanwhile, the world’s biggest battery manufacturer, CATL, released a 500 Wh/kg battery in 2023. That enables a Tesla Semi to to travel 750 miles between recharges while weighing a lot less, or travel 1,000 miles with the same completely reasonable 2% to 3% extra weight allowance. And silicon chemistries already commercializing promise double to quintuple CATL’s energy density and hence longer ranges, lower weights or both.
I’m engaged in a review group of a total cost of ownership study for European freight trucking now, the reason I went and looked at California’s bus and refueling station maintenance. One of my comments was that certain mixes of vehicles, such as hydrogen and light vehicles, should be excluded as that debate is over, but that hydrogen for trucks should be kept in because some people refuse to accept the reality and so driving it home with another study is still useful.
And then there’s heavy rail. That question has been answered as well. Outside of North America, the world is simply getting on with electrifying the parts of the rail network that don’t already have overhead wires. India expects to be 100% electrified in 2024, leading the world. All of China’s massive and growing network of high-speed rail is electrified, as are the high-speed rail systems in Indonesia and Morocco. China is delivering freight to Europe along fully electrified rail.
And total cost of ownership studies are clear here too. Baden-Würtemberg did it right. They put a gimlet-eyed spreadsheet jockey on the job, told them to compare grid-tied, battery electric and hydrogen rail systems, and hybrids of them. The answer came back that where they couldn’t put in overhead wires because it was too expensive, mostly bridges and tunnels built without overhead wires and the lines that have a lot of them, batteries were almost as cheap as overhead wires, while hydrogen was three times as expensive. Meanwhile, next door in Lower Saxony, they took some of that lovely governmental money, about €14 million per train, to buy some hydrogen trains and a year after starting service with them announced they would not buy any hydrogen trains ever again.
The USA will get to electrifying rail too, and drag Mexico and Canada along with it into the 21st Century of rail, but not until it’s wasted as much time as possible scraping out what Wall Street analysts want to hear every three months.
Hmmm… no ground transportation demand for hydrogen. At all. And with renewables providing all of the energy, not much room at all for hydrogen in the electricity system. While it’s not on the infographic, as room did not permit, let’s talk about that. All of the coal, natural gas and oil burned for electricity today must go away if we are to solve climate change. The idea of carbon capture on power generation plants has been tried, and it’s vastly more expensive and less effective than any rational actors are willing to consider.
We’re now into the end game of a quarter century of unfortunately misguided energy policy, kicked off in large part by Rifkin’s book and proselytizing in Europe and North America.
But wait, the hydrogen advocates say. We’ll need hydrogen for electricity storage! Well, no. That’s actually a solved problem with the exception of continent-scale, weeks-long lulls in sunshine and wind which only come along every few decades. Even in the archipelago nation of the United Kingdom, modeling only finds that is a significant problem every decade or so.
For fast-response, short-duration storage, the kind that’s great for peaks and smoothing spiky power, as well as moving solar power a small handful of hours into the future, the same increasingly cheap batteries in an increasing variety of chemistries are completely adequate. For longer duration storage, the same pumped hydro that was built to give nuclear plants something to do at night and that represents 93% of grid energy storage today covers from four hours to 24 hours of storage quite easily.
China gets this. It’s built 58 GW of pumped hydro already, with probably 600 GWh to more than a TWh of energy capacity. It’s building or has in plan another 365 GW of capacity representing 4 to 8 TWh of energy storage by 2030. The rest of the world has woken up to this old solution as well. The Australia National University’s greenfield off-river, closed-loop pumped hydro atlas is undoubtedly getting a big work out. They did the study a few years ago of all locations where small area upper and lower reservoirs could be placed reasonably close together, avoiding streams and rivers, with over 400 meters of head height to provide lots of energy, near transmission and off of protected lands. A bunch of the blank areas in the map like Siberia just didn’t have any good data, but undoubtedly have lots of resources.
The ANU estimates that there are 100 times the energy storage resources available for the end state of full electrification in these identified sites, and 200 times in North America. And with HVDC transmission, storage doesn’t have to be right beside either the generation or the demand center. Hong Kong’s pumped hydro backup system of 25 GWh is a couple of hundred kilometers away in mainland China, as an example.
Then there is the emerging technology of redox flow batteries, where you can scale up big tanks of chemicals on either side of the two-way fuel-cell equivalent technology in the middle and store a lot of energy. There are a handful of commercialized solutions already and more in development.
It’s really only that 10 to 50 years dunkeflaute where inefficient, hard to store, hard to use hydrogen still has any opportunity to play, and even there, if the question is “What molecule should we store as a strategic energy reserve?” it’s hard not to find better alternatives like capturing a bunch of the methane emitted from human biomass waste in landfills and the like and put it into existing natural gas strategic reserves. You really have to start with finding a use case for hydrogen, as Sir Chris Llewellyn’s UK study that threw out pumped hydro without consideration and downplayed HVDC interconnect potential did, and then announce that green hydrogen was the answer there, so by definition it would be an answer for shorter duration storage too.
So there’s very little for hydrogen to do for energy on land, where the vast majority of energy is consumed. We’re down into the really short strokes now, with only industrial heat, maritime shipping and aviation left as potential markets.
But 45% of industrial heat is under 200° Celsius and heat pumps can do that now. There are an awful lot of organizations that should know better claiming that temperatures above 200° require something to burn, but this just isn’t the case. Resistance heating up to 600° is already commercialized. 70% of US steel comes from scrap fed through electric arc furnaces that can generate 1,500° to 3,000° heat. There are microwave, infrared and plasma solutions. There are a relatively tiny handful number of industrial heating requirements that need the characteristics of an open flame, and once again, that biological methane which is a big climate problem right now is a more reasonable choice than hard to make, hard to store, hard to distribute and expensive to use hydrogen.
But surely ships need fuel, right? Not as much as you would think. There are two 700 unit container ships running 1,000 kilometer (600 mile) routes on the Yangtze right now. There’s a 1,000 passenger cruise ship doing three hour tours in the Three Gorges. There are innumerable battery electric ferries, utility craft and tug boats already quietly churning through inland and port waters. All inland shipping and about two-thirds of short sea shipping like routes between Germany and Norway are completely viable with batteries and where something can be electrified, it will be, simply because the operational and maintenance costs are so low that they blow the rest of the total cost of ownership elements out of the water.
But that still leaves the big ships that cross the ocean. Surely hydrogen has a play there? Well, first the good news. About 55% of bulk shipping is going to diminish radically. Most of that is bulk coal, oil and gas, and that’s going away in any rational world. The rest is raw iron ore, which using green electricity and hydrogen as way to remove excess oxygen — derusting it — from the iron instead of coal will allow much more of it to be processed close to mines. Container shipping will go up, but not nearly as much as bulks go down.
Between shorter haul electrification and dropping long haul shipping, there’s not nearly as much energy required. My estimate is that about 70 million tons of diesel or equivalent will be required in 2100. And right now we already manufacture 70 million tons of biodiesel, we just waste most of it on ground transportation that will be electrifying.
This is, however, clearly a case where the maritime industry and the hydrogen for energy types have not figured out the inevitable. Organizations like the ICCT are still doing hydrogen for shipping studies, including liquid hydrogen, something so difficult to work with that the rocket industry is moving away from it to liquid methane. And the maritime industry has been seduced by ammonia and methanol lobbyists with claims that their current products are as cheap as existing maritime fuels — they aren’t —, that they are clean burning — that much is true —, that they are low-carbon — that ignores the very high upstream carbon debt of making them — and that actually low-carbon well-to-wake versions will be cheap in the future — a complete and utter shameless lie.
As a result, big shipping concerns like A.P. Moller-Maersk are wasting money on dual fuel ships that can run on methanol or ammonia, and LNG ships are a quite short-sighted growth industry. And Maersk is mostly contracting for biomethanol, not synthetic methanol, so that’s actually bad news for hydrogen advocates, not good news.The economics will play out as they play out. The International Energy Agency (IEA) chimed in with a rather stunning report in late 2023 on e-fuels. Even with their very optimistic electricity costs from completely new wind and solar farms devoted to brand new integrated industrial facilities that generated carbon dioxide in one process that was used with green hydrogen in another, green fuels were 4-6 times the cost of current maritime fuels and twice as expensive as biofuels.
I’ve done bottom up cost workups with the currently cheapest electrolyzers, standard balance of plant figures, and the cheapest, most available electricity on the planet, Quebec’s US$49 per MWh, 24/7/365, amortized hydroelectric and transmission, and come in at synthetic fuels in the same range as the IEA.
But electrolyzers are going to get cheaper, shout frustrated hydrogen proponents. It doesn’t matter. They are perhaps 8% of the cost case, so even if you make them free, synthetic fuels won’t get significantly cheaper. But electricity will be free, shout proponents. No, it still requires transmission, distribution and firming, and the organizations that make it need to make profits. There will be cheaper electricity available a portion of the time, but when capital costs are so high for e-fuels plants, they need to be operated at much higher capacity factors.
So, no hydrogen for shipping. But surely in the skies? Remember that the space industry that’s part of aerospace is trying to move away from liquid hydrogen because it’s so hard to deal with. Occasional rockets surrounded by highly trained specialists is not a replicable solution for commercial aviation. Pure hydrogen in gaseous form can’t get enough energy on board aircraft. In liquid form it requires bulbous tanks inside the fuselage with passengers who prefer temperatures about 273° warmer, and that’s in Celsius. Those bulbous tanks have to be at the back of the air craft for safety of passengers and crew, so as they empty, the plane would become nose heavy and fall out of the sky. There is no path to the certification for hydrogen aviation. Alternative aircraft designs won’t fit into any current airports. And airports would have extraordinary costs and challenges dealing with liquid hydrogen.
But the aviation industry too has not received and understood the memo on hydrogen and derivatives made from hydrogen. As a reminder, the IEA found that synthetic fuels would be a lot more expensive than current fuels. But that was in the best case scenario. The current reality is that the 25 European synthetic sustainable aviation fuel proposals haven’t been able to find any airlines willing to pay the current 10 times cost of e-kerosene. As a result, those proposals haven’t reached final investment decision, something that they share in common with virtually all green hydrogen proposals that aren’t for ammonia for fertilizers, a clear and pressing requirement for decarbonization.
So what will aircraft run on? Unsurprisingly, the airline industry is already buying millions of tons of sustainable aviation biofuels and capacity is increasing regularly. Once again, while 2 to 2.5 times more expensive than historic costs for fossil kerosene, they are half the best case costs for synthetic fuels, and a quarter of the current costs in proposals that have been rigorously costed out.
And once again, batteries are vastly more fit for purpose than the industry appears willing to admit or even realize, especially with hybrid models where divert and reserve energy can be in the form of biofuels with an onboard generator. In 2023, a hybrid aircraft flew for twelve hours before landing, and had divert and reserve left over. Heart Aerospace has hundreds of orders for its 30 passenger hybrid turboprop with 400 kilometers of usable range. Studies in 2023 started showing the reality that up to 100 passenger turboprops could have ranges of a thousand kilometers with current battery technologies.
The potential exists for ranges of 3,000 kilometers with divert and reserve provided by biofuels with silicon chemistries. Even if those chemistries only double or triple CATL’s 500 Wh/kg, vast amounts of within-continent aviation can run 99% of the time on electrons. Aviation will be slower in this model, but vastly quieter, cheaper and more efficient, so it will end up winning economically, just as with every other mode of transportation.
In my projection of aviation through 2100, with much more realistic, lower projections of aviation growth, only 110 million tons of biofuels will be required in 2100.
And to be clear, biofuels can be made from waste biomass. Every ton of dried biomass can make about 0.4 tons of biofuels. Do we have enough waste biomass? We certainly do, 2.5 billion tons of food waste alone globally, and 1.5 billion tons of livestock dung in just Europe. The first alcohol to jet fuel plant just opened in Georgia in the USA, one with a target capacity of 9 million tons a year. That alcohol is made from biomass that’s fermented and distilled from pretty much any biomass, but sourced mostly from midwestern corn at present. That an alcohol to jet fuel plant made final investment decision and no hydrogen to jet fuel plant can achieve that milestone should be a key indicator of which way the industry will go.
And so, the end of the hydrogen for energy road. It’s already lost light vehicles, with only relative fanatics pretending otherwise. Buses are a foregone conclusion. The tiny number of forklifts, 50,000 in total, are dwarfed by battery electric sales exceeding a million a year. Trucking total cost of ownership studies and tests like Run on Less make it clear that there will be no trucking hydrogen, but some people haven’t received the memo yet. Grid storage is a dead end for the molecule. And even in the two last and diminishing markets for burnable fuels, aviation and shipping, hydrogen has no real hope of being part of the solution.
Will the hydrogen for energy story end in 2024? No, of course not. 25 years of bureaucracies and reputations and investment have inertia that will keep it going long past the time when it’s obvious that the entire idea is fatally flawed in every segment it’s considered compared to alternatives that already work better, are cheaper today and will remain cheaper.
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Latest CleanTechnica TV Video
I don’t like paywalls. You don’t like paywalls. Who likes paywalls? Here at CleanTechnica, we implemented a limited paywall for a while, but it always felt wrong — and it was always tough to decide what we should put behind there. In theory, your most exclusive and best content goes behind a paywall. But then fewer people read it!! So, we’ve decided to completely nix paywalls here at CleanTechnica. But…
CleanTechnica uses affiliate links. See our policy here.