Audioblog 13: Clean Hydrogen's Missing Trillions

Michael Liebreich

Hello, I'm Michael Liebreich and this is Cleaning Up. In September 2022, the 5th Hydrogen Energy Ministerial in Tokyo, organised by the powerful Japanese Ministry of Economy, Trade and Industry, in cooperation with the International Energy Agency, announced that by 2030, the world would produce and use 90 million tonnes of renewable and low-carbon hydrogen. To put that in perspective, the world currently uses around 100 million tonnes of hydrogen to make fertilisers and petrochemicals, almost all of it made from coal and gas. Then, in September last year, ministers and delegates from 23 countries at the 6th Hydrogen Energy Ministerial not only reiterated the 90 million tonne target, but went further, promising that the overall market for hydrogen would grow to 150 million tonnes by 2030. That would mean replacing 40 million tonnes of existing hydrogen supply with clean hydrogen and adding another 50 million tonnes of new use. All very exciting, and it helped to ensure that hydrogen was one of the hot topics at COP 28 in Dubai a few months later. But someone has to say it: these targets are entirely utterly, ludicrously, unachievable. The issue is simple: it's money. As I wrote for the pre-event brochure for the upcoming 4th Financial Times Hydrogen Summit this June, the biggest challenge facing low emission hydrogen is that it is expensive to produce, expensive to transport, expensive to store, expensive to distribute, and expensive to use. The reality is that whether you're switching existing users to clean hydrogen or pushing hydrogen into sectors where it's not currently used, it takes money - and lots of it. In this week's episode of Cleaning Up, we'll be finding out how much and how much more governments would have to spend for hydrogen to live up to its hype. This audio blog is adapted from a piece I wrote at the end of last year for BloombergNEF, entitled "Clean Hydrogen's Missing Trillions". In it, I estimated that hitting the Hydrogen Energy Ministerial target of 90 million tonnes of clean energy by 2030 would require subsidies of at least $2.3 trillion to be on the table right now, while the actual figure at the end of last year was 1/10th of that. As we shall see, the figures have changed a bit since then: afew more 10s of billions of subsidies have been announced, but the expected costs of clean hydrogen have also gone up. The message remains the same: the subsidy gap remains in the multiple trillions of dollars. It should not therefore be remotely surprising that the news is full of projects being cancelled and delayed. In fact, that will be one of the main hydrogen stories through to 2030 and beyond. 

So let's get started. First of all, where did that hydrogen energy ministerial target come from? As so often with hydrogen targets, it's not at all easy to identify where the 90 million tonne figure originated. It seems to have been based on a 95 million tonne figure quoted in 2022 by the International Energy Agency as the amount required in 2030 for the world to achieve net zero by 2050. In fact, a year before, the IEA is October 2021 Net Zero Roadmap had contained an even more inflated figure: 212 million tonnes of hydrogen overall by 2030 - that's more than double the current global market, of which 150 million needed to be low-carbon. But a year after, the 2023 Roadmap saw the IEA reduce the total figure to 150 million tonnes and the low-carbon figure to 70 million tonnes. Whatever the source, the Hydrogen Ministerial's target of 90 million tonnes of clean hydrogen by 2030 is the sort of figure required to sustain claims that the world is on track to limit warming to one and a half degrees over pre-industrial levels. That's if you buy into the vision of hydrogen delivering 15% or more of emissions reductions. Regular listeners will know that I don't but that's a story for a different episode. But are the targets at all plausible? The 150 million tonnes of overall demand in 2030 would be a 50% increase over 2023's 100 million tonnes in seven years - which is only a 6% compound annual growth rate, so that doesn't raise any red flags. But then you look at the clean hydrogen part: you would need to retrofit almost half of the existing fossil fuel based hydrogen production with carbon capture and storage, build 60x more clean hydrogen production capacity than has been funded to date, and create 50 million tonnes of annual demand in sectors that do not currently use any hydrogen at all. 

So, how are we doing in building clean hydrogen capacity? Well, as of the end of 2023, there were just 700 megawatts of operating electrolyzers, producing around 110,000 tonnes of green hydrogen annually. There are still no operating blue hydrogen plants (blue hydrogen, of course, being that made from natural gas with a carbon captured and permanently sequestered). Expectations have raced far ahead of reality. Over the past two years, electrolyzer manufacturers have expanded manufacturing capacity from 6.8 gigawatts to 33.7 gigawatts, and by the end of 2024 BloombergNEF is expecting capacity of no less than 52.6 gigawatts. Orders in 2023, by contrast, struggled to hit just two gigawatts. Share prices of electrolyzer manufacturers have plummeted up to 95% from their 2021 peaks and industry bellwether Plug Power, after 24 straight years without profits, has warned investors that it may not survive to see its 25th birthday without new funding. It's not that there's a shortage of hydrogen projects in the pipeline. At the end of last year, BloombergNEF's clean hydrogen database contained projects that could produce over 170 million tonnes of hydrogen, of which 47 million could in theory begin production by 2030. The IEA 2023 Global Hydrogen Review reported a slightly lower figure of 38 million tonnes that could be working by 2030. But whichever dataset you prefer, by the end of 2023, just 1.5 million tonnes worth of projects had reached final investment decision or were already producing. That's less than 1% of all announced projects by volume, and only 3% of those that were targeting a 2030 start date. BloombergnNEF also maintains a database of clean hydrogen offtake agreements, the client contracts without which developers can't raise finance. It turns out there were, at the end of last year, identified offtakers for just 8 million tonnes of hydrogen per year. But of this, only 1 million tonnes were covered by binding contracts rather than non-binding letters of intent, or just press statements. Looking at Europe, last month consultancy PwC released a report pointing out that the EU would need 120 gigawatts of electrolyzers in order to meet its 10 million tonne target for 2030. Although they reported 205 gigawatts of projects that had been announced, only three gigawatts had so far reached final investment decision or begun construction. That's enough to hit just 2.5% of the EU's target for 2030. 

So, what is the reason for this huge gap between the ambitions for clean hydrogen and what the market is actually delivering? Well, it's all about costs. Every one of the 52 National Hydrogen Strategies around the world is predicated on one fundamental assumption: that the cost of clean hydrogen will fall substantially and quickly. The EU's 2020 Hydrogen Strategy stated that, 'in regions where renewable electricity is cheap, electrolyzers are expected to be able to compete with fossil-based hydrogen in 2030.' The US' Hydrogen Shot Programme, launched in June 2021, aimed for what it called "the Three Ones": clean hydrogen at $1 for 1 kilo in 1 decade. Japan's updated Hydrogen Strategy, released in 2023, aimed for domestic green hydrogen production at a cost equivalent to $2.20 per kilo by 2030. There are however three important reasons why these expectations are proving hopelessly optimistic. First, while the relentless logic of the experience curve will indeed deliver very cheap electrolyzers in the end, the time it takes to get there depends on how quickly experience is accumulated.  The EU Hydrogen Strategy posited five gigawatts of European electrolysis capacity by 2024, but BloombergNEFs Hydrogen Project Tracker suggests the number will barely reach a 10th of that. Second, electrolyzer stacks account for only around 40% of the capital cost of green hydrogen plants. The rest is stuff like compressors, tanks, valves, pipes, power supplies, and site infrastructure, whose costs are not about to fall off a cliff. Furthermore, capital costs only account for around 1/3rd of the cost of green hydrogen. The biggest element is green electricity which is already far down its own experience curve. And there's an inescapable tradeoff between the cheapest green electricity, which is variable wind and solar, and the need to get high utilisation rates on your hydrogen plant. The third reason why clean hydrogen costs are going to stay way higher than expected for way longer is the current spike in inflation and interest rates. Over the past two years, the European Central Bank and the Federal Reserve have hiked their main rates from 0 to 4.5% and 5.5%, respectively. It's not just the interest cost of the green hydrogen plants that's the problem; wind and solar are all about higher upfront capex with low running costs, so they are very sensitive to interest rates. The wind industry has been hit particularly hard as the business model that drove its profits for a decade - bidding low and letting equipment cost reductions deliver profits - has fallen apart. The rise in interest rates has caught many developers unawares, and manufacturers cannot pick up the slack. Inflation, product proliferation, technical problems and extreme cost pressure have driven them to unprecedented losses. 

So, taking all of this into account, what does clean hydrogen really cost? While there's a wealth of literature on what clean hydrogen is supposed to cost at some point in the future, there's a dearth of data on the prices actually paid to the small number of projects that have actually secured finance. A recent report by consultancy BCG estimates the figure at between €5 and €8 per kilo - that's between $5.40 and $8.60 per kilo. BloombergNEF's July 2023 Levelized Cost of Hydrogen report showed green hydrogen costs of between $1.80 and $4 per kilo in most countries of the world by 2030, but that simply doesn't reflect financing conditions in the second half of last year, which really deteriorated. The IEA's 2023 Global Hydrogen Review still maintained that green hydrogen could be produced in Europe for €1.60 per kilo by 2030, but the assumptions needed to achieve that, such as a cost of renewable electricity €13 per megawatt-hour - or a 3% cost of capital - are simply not credible. Use more reasonable assumptions and the cost of European green hydrogen comes out in the €5 to €7 price range - that's $5.50 to $7.70 per kilo. Now, prices in the US, Saudi Arabia and Australia - places with tremendous renewable resources and low costs of capital - may come in lower, but even they look unlikely to deliver hydrogen much below $3 per kilo by 2030, and that's excluding subsidies and the US inflation Reduction Act tax rebates. Inflation and higher interest rates have also hit blue hydrogen projects. Once more, there's a dearth of real data, but at current interest rates, it's hard to see a price much below $3 per kilo even in the US or the Gulf region. So, for any clean hydrogen project operating by 2030, you need to count on production costs being in the $3 to $8 per kilo range, depending on location. 

But what about imported hydrogen? If you think importing green hydrogen from placed where it can be made particularly cheaply - like Australia, Namibia, Chile, the Gulf region or North Africa - can deliver hydrogen at prices below $3 per kilo, I have bad news for you. In my December 2022 piece The Unbearable Lightness of Hydrogen, I highlighted the McKinsey and Company Hydrogen Flows Report on behalf of the Hydrogen Council, which promised 400 million tonnes of internationally-traded hydrogen by 2040. They've just halved that forecast for 2040 - not their last downgrade, I'd be prepared to wager - but are still forecasting 20 million tonnes of hydrogen imports as early as 2030. It is, of course, technically feasible to liquefy hydrogen and put it on a ship, but the cost of doing so will be somewhere between $3 and $8 a kilo even by 2040, which at that point will be more than double the cost of production. Liquid hydrogen on a ship is frankly even less plausible than Snakes on a Plane. But what about shipping ammonia instead? Today 16.5 million tonnes of ammonia are transported by ship. According to shipping services provider Clarkson's, setting aside the need to build new ammonia plants in faraway lands, 20 million tonnes of hydrogen translates into 113 million tonnes of ammonia - that's just chemistry - so sending it by ship would constitute a seven-fold increase in shipped volume. A two-fold increase by 2030, plus a few ships powered with ammonia: I could believe. A seven-fold increase by 2030 is just not going to happen. As for pipelines, the gas industry is promising that they can repurpose existing pipelines. But by the time you account for hydrogen's extremely low density, the reduction in pressure required to avoid embrittlement and the increase in power requirement for big new compressors and more of them, their energy carrying capacity could easily be reduced by 75%. Most future hydrogen pipelines will be purpose-built, and a sensible cost estimate would be between $0.4 and $0.8 per kilo of hydrogen transported. But that's irrelevant. It's hard to see any commission this side of 2030. So whichever way you look at it, you're back at hydrogen costs of $3 to $8 per kilo. 

And now you have to get customers. Let's just suppose you've managed to produce or import your 90 million tonnes of clean hydrogen by 2030. Now, you need to figure out how to get customers to use it. If they're producers of fertilisers and petrochemicals, that's fairly straightforward because they already use hydrogen. While a few of their customers may be prepared to pay a so-called "green premium" - perhaps some Western food brands that want to be seen using a bit of zero-carbon fertiliser -most won't. This means you'll have to match what they currently pay for fossil-based hydrogen, plus any carbon price to which they're subject, and that means selling clean hydrogen for $2 to $4 per kilo below your production or import cost. What about sectors which are not currently using hydrogen, such as steel heating and transport? Getting them to use hydrogen will take much more than just matching the current cost of grey hydrogen - after all, they could have switched to grey hydrogen at anytime they wanted, but they didn't. For heat-related use cases, even if you ignore the cost of converting equipment, you'll have to sell clean hydrogen at the same price on a heat-adjusted basis as natural gas. Right now, that means $0.3 per kilo in the US, $1.40 in Japan, and $1.70 per kilo in Europe. Or, in sectors covered by the European Union's Emissions Trading System, $2.45 per kilo, but that means selling clean hydrogen again somewhere between $3 and $5 below its production or import cost. For the steel industry, getting manufacturers to switch will require each of them investing billions of dollars to retire and replace existing coking coal-based plants, and then hundreds of millions of dollars more per year in higher running costs. Sweden's H2 Green Steel has raised €5.3 billion euros - that's $5.6 billion - for its first plant, but that does not include running costs. Turning the levels of support required across the industry into a cost per kilo of hydrogen is hard, but conversion is so staggeringly expensive that even providing free hydrogen (you know, geological, white hydrogen anyone?) would probably be insufficient to drive uptake on the economics alone. Getting hydrogen used for land transportation is even harder. First, you have to get people to buy vehicles that cost significantly more than petrol or diesel equivalent. Then, you have to retail hydrogen at a cost equivalent to diesel or petrol - around $5 per kilo - even though hydrogen filling stations cost $1-2 million each, and it takes 16 or 18 hydrogen tube trailers to replace each diesel or petrol fuel distribution tanker. Then, you have to do this in competition with electric vehicles, which are faster, simpler, more roomy, and already offer broadly the same total cost of ownership as their fossil-fueled equivalents. It shouldn't be remotely surprising that the news is filled with stories of cancelled hydrogen taxi, hydrogen bus and hydrogen train projects. In many ways, what's surprising is the hope still being invested in heavy freight, although last week, Patrick Pouyanné - testifying under oath before the French Senate - did say, 'we decided to invest in a European network of hydrogen filling stations for HGVs, but I'm not sure we got it right.' 80% of HGVs in Europe do less than 500 kilometres per day, and as soon as you say that, it seems fairly obvious they will go electric. 

Right. So how much subsidy are we talking about in total? With all this information on costs, we can now produce a ballpark estimate for the total volume of subsidy or support that would be needed in order to hit the hydrogen energy ministerial target of 90 million tonnes of clean hydrogen per year by 2030. Of those 90 million tonnes, you will recall that 40 million tonnes are meant to go to existing use cases, and 50 million to new ones. Multiplying the $2 per kilo to $4 per kilo cost penalty for existing use cases by the 40 million tonnes used there, and you'll get to $80-160 billion per year. For new use cases, let's just be incredibly generous and make the maths a bit simpler, and assume that on average, they require a $3 to $5 per kilo subsidy to drive uptake. (The reality is it'll be more than that.) Multiply those figures by 50 million tonnes required in new use cases, and you get an additional $150-250 billion of support required per year, bringing the total subsidy requirement to somewhere between $230 billion and $410 billion per year. But actually, it's worse than that - much worse - because of the way project finance works. To get to a final investment decision, one year of contracted revenue is not enough for a project developer. The median offtake agreement in the BloombergNEF hydrogen offtake database is 15 years, with a range from 7 to 30 years. Again, let's be incredibly generous and make the maths easy, and assume that an average term of 10 years will be sufficient for projects to get built. So you have to multiply that annual $230-410 billion figure by 10. For the Hydrogen Energy Ministerial target to be hit, finance ministries of the world need to put between $2.3-4.1 trillion on the table. And they need to do it right now if it's to be spent in time for the project to be operating by 2030. In fact, here's a quick heuristic you can use to check if targets being thrown around make any sense. 1 million tonnes is 1 billion kilos. For each dollar of subsidy or green premium that has to be found per kilo of green hydrogen, that translates into a billion dollars per year. Turn that into a 10 year offtake agreement and you have $10 billion needing to be committed upfront by a government or by some other bankable entity. So remember this: for each dollar of subsidy required for each million tonnes per year of hydrogen, someone has to find $10 billion. And frankly there are far too many politicians today boasting of support in the millions for hydrogen volumes whose costs will be in the billions. 

So if we're looking for $2.3-4.1 trillion, how much has been committed so far? Well, the biggest provider of support for hydrogen projects today is the US, thanks to its 2022 Inflation Reduction Act. At the time it was passed, the US Congressional Budget Office estimated its total cost at $390 billion, of which just $13.2 billion was expected to go to hydrogen production. However, that figure was uncapped and the real number will clearly be many times higher. BloombergNEF estimates a hydrogen figure of $137 billion, although that is for disbursement over the next 10 years, not just until 2030. By the end of 2023, BloombergNEF was estimating the figure for Europe at $125 billion of subsidy. The EU Hydrogen Bank was funded to the tune of €3 billion, Germany was set to commit €3.6 billion to its H2 Global plan for disbursement over 10 years and is working on a €50 billion 15-year contract-for-difference plan for industrial decarbonisation. The EU was earmarking €100 billion, the UK: £900 million and so on. Japan was planning to spend ¥7 trillion (that's $47 billion dollars), China: $10 billion, and India: 2 billion. BloombergNEF estimated at the time that the total committed subsidy for Asia was $12 billion, bringing the global total at the end of 2023 to just under $280 billion. This month, BloombergNEF released an update showing that the overall figure had risen to $360 billion. However, not all of this can be disbursed in time to move the needle for 2030. My estimate for the total flow of subsidies that may actually see the light of day in time to produce clean hydrogen for 2030 would be around $200 billion. That's a lot of money, clearly enough to keep the hydrogen souffle somewhat inflated, but it will leave the world $2.1-3.9 trillion short of the amount that would be needed to deliver the Hydrogen Energy Ministerial's 90 million tonnes of clean hydrogen target - a shortfall of 90 or 95%. So it's hardly surprising that only a few percent of the projects in the BloombergNEF database have reached final investment decision and are being built. 

Now a quick aside: before you shoot off a dismissive post telling me I'm obviously wrong because the European Hydrogen Bank's first auction cleared at a subsidy of just 48 cents per kilo, hold your horses. The actual cost of hydrogen for the winning project is set to be between €5.30 and €13.50 per kilo. The support from the Hydrogen Bank is not going to cover anywhere near 100% of the cost differential; it's just to kind of bankable amuse bouche. As Rogaciano Rebelo, CEO of winning bidder Madoqua Renewables explained to Portugal's journal economical, 'the buyer [will pay the rest] there has to be a premium.' And, he's also betting on electrolyzer costs being cut in half over the two years before he has to actually build the projects. So, will the winning projects get built? Maybe. Should you wager your pension on it? I would suggest not. 

To finish off, let's look at how much clean hydrogen we might actually see in 2030. It turns out that the leaders of the world's biggest economies were not taking the Hydrogen Energy Ministerial 's 90 million tonne target particularly seriously, even as they were signing the summit communiqué. BloombergNEF's Hydrogen Strategy Tracker shows that the governments of the 52 countries with hydrogen strategies are targeting total production of no more than 40 million tonnes by 2030. The US is hoping to produce 10 million tonnes, Japan: 3 million, and South Korea: 4 million. The UK recently upped its target to 10 gigawatts of production - equivalent to about 1.5 million tonnes. Then there's the EU's stated target of 20 million tonnes - 10 million of local production and 10 million of imports from outside the EU. Even when you add those and ignore double counting, you're still 50 million tonnes short of the Hydrogen Energy Ministerial figure. But even 40 million tonnes globally is an absurdly high target. In fact, BloombergNEF has just published a Hydrogen Supply Outlook for 2030, forecasting just 16.4 million tonnes. But I think even that is optimistic: take our $200 billion of committed and likely to be dispersed support. Divide into 10-year offtake agreements, divide it again by the $2-5 per kilo of subsidy required to drive clean hydrogen into the market and you get around 6 million tonnes. If inflation and interest rates continue to ease and things pick up, we might see a bit more than that. 10 million tonnes by 2030 -equivalent to 10% of the current global hydrogen market? Maybe 16.4 million tonnes? I'll take the under on that. 40 million tonnes? Forget it. 90 million tonnes? Delusional. Now, reality is dawning on the politicians. There's a live debate in the corridors of the EU about how to scrap the 20 million tonne target for 2030 - since it's obviously going to be missed by at least 75% - and what to replace it with. My money is on a face saving claim that the target was spot on; it will just take longer - to 2038, say. But the cost problem is not going away, not by 2030, not by 2040, not even by 2050. Last month, the US National Petroleum Council, a federally chartered committee established in 1946 at the request of President Harry S. Truman, published a major report entitled: Harnessing Hydrogen. Led by our old friends McKinsey and oil company Chevron, with modelling by MIT, it found that even in 2050, clean hydrogen for refineries will cost $2-5 per kilo against the market price of just $1. For heavy duty trucking, clean hydrogen will cost $6-10 per kilo at the pump, when $2-4 per kilo is needed to compete with diesel or electrification. The report finishes with 21 recommendations, but the short version is: show us the money or hydrogen doesn't happen. In fact, it's the same point as I made in my recent Five Horsemen of the Transition Piece, that some sectors will need a carbon price for the very, very long term in order to decarbonize. 

So, in conclusion, what's the takeaway? Next time you hear a politician or promoter waxing lyrical about the hydrogen economy, the hydrogen world, hydrogen - the gas of the future, please remember this reality check: there will be no 90 million tonnes of clean hydrogen in 2030. There might be 10 million, or maybe 16 million, enough to decarbonise less than 20% of the current hydrogen demand. Sorry to burst their bubble. In fact, fewer than 1 in 10 of the projects currently in BloombergNEF's Hydrogen Project Database will have been built by 2030. We'll be able to look back in that year and score my predictions against those of the hydrogen boosters. Until then, let's just be realistic about how much to rely on hydrogen to do the heavy and difficult work of decarbonisation. 

As always, we'll put links to the resources mentioned in this episode in the shownotes, that's: my December 2023 Bloomberg NEF essay: Hydrogen's Missing Trillions from which this audio blog was adapted, the Chair's Summary for the September 2023 6th Hydrogen Energy Ministerial, my December 2022 piece for BloombergNEF entitled: The Unbearable Lightness of Hydrogen and the Cleaning Up episode based on it, my September 2023: Five Horsemen of the Apocalypse piece for BloombergNEF, and the Cleaning Up episode based on that, the US National Petroleum Council's: Harnessing Hydrogen report, and PwC's April 2024 report on Navigating the Global Hydrogen Ecosystem. To access the BloombergNEF Hydrogen Projects Database and the Hydrogen Strategy Tracker, you will need a Bloomberg Terminal or BloombergNEF subscription. If that's what you're after, please make sure you contact BloombergNEF- they'll be happy to help you.

Now, on June the 12th, I'll be the opening keynote interviewee at the 2024 4th Financial Times Hydrogen Summit. Last year's interviewee was Dr. Andrew Forrest, executive chair and founder of Fortescue Future Industries, and of course, one of the leading hydrogen promoters in the world. I'm hoping to inject a little realism, physics and economics into this year's event. If you're planning to attend, do grab me and say hello - unless, of course, you're one of my trolls, in which case please don't. If you've enjoyed today's episode, please subscribe, give it a like and leave a review on your favourite podcast platform, or a thumbs-up and a comment on YouTube. And do please tell others about Cleaning Up - we would love a bigger audience. You can subscribe to our free newsletter on cleaninguppod.substack.com, that's cleaninguppod.substack.com, and follow us on Twitter - because I refuse to call it X - Instagram or LinkedIn for news, highlights and Bryony's and my hot-takes. Finally, you'll find our archive of over 160 hours of conversations with extraordinary climate leaders on cleaningup.live, that's cleaningup.live.