Support CleanTechnica’s work through a Substack subscription or on Stripe.
A major study in Nature Energy, “Global greenhouse gas emissions mitigation potential of existing and planned hydrogen projects” by Terlouw et al, has done something rare in the hydrogen hype bubble that’s slowly deflating. It has gathered thousands of real hydrogen projects around the world, run full life-cycle assessments on them, and drawn clear boundaries between what helps and what wastes effo…
Support CleanTechnica’s work through a Substack subscription or on Stripe.
A major study in Nature Energy, “Global greenhouse gas emissions mitigation potential of existing and planned hydrogen projects” by Terlouw et al, has done something rare in the hydrogen hype bubble that’s slowly deflating. It has gathered thousands of real hydrogen projects around the world, run full life-cycle assessments on them, and drawn clear boundaries between what helps and what wastes effort. It has confirmed what many of us have been arguing for years. Hydrogen makes sense in a few industrial applications and very little else. The evidence matches the engineering logic.
The study examined about 2,000 existing and planned hydrogen projects across 20 years. It found that if every one of them were built, the total output would reach about 110 million tons of hydrogen per year. That would come with about 0.4 gigatons of emissions and at best offset between 0.2 and 1.1 gigatons of carbon dioxide each year. That contribution collapses by more than 80% when the comparison is made to a future world that electrifies instead. The researchers’ conclusion is straightforward. Hydrogen should go where it replaces dirty hydrogen that already exists, not where electrons do the job better.
Most of today’s hydrogen is used for fertilizer, refining, and methanol. That is the world’s real hydrogen problem. It already produces more greenhouse gases than Germany, about the same as all of global aviation. Cleaning that up should be the focus. The hydrogen economy that exists is dirty and old, not futuristic and green. Every new application dreamed up to justify more electrolysis adds cost and emissions that would not exist if we focused on the existing chemical markets.
The study identifies steel, biofuels, and ammonia as the sectors with the best climate returns per kilogram of hydrogen used. That makes sense. Hydrogen is part of chemical reactions that cannot be replaced by electricity alone. But even in steelmaking the picture is more complicated than it appears. Direct reduced iron that depends on hydrogen is expensive and energy hungry. Other options are advancing faster. Electric arc furnaces already recycle over a quarter of the world’s steel, are increasingly powered by renewables and in my projection grow to 75%. Molten oxide electrolysis removes hydrogen entirely. Biomethane based DRI uses existing infrastructure at lower cost with no innovation required except the feedstock. Flash ironmaking uses heat and fine iron ore directly without gas intermediates. All of these pathways are likely to be cheaper and easier to scale than hydrogen DRI.
Ammonia is unavoidable as a feedstock for fertilizer, but that is where its usefulness stops. The temptation to burn it as fuel for ships or power generation ignores its low efficiency and its nitrogen oxide emissions. Using hydrogen to create ammonia only to split it again for energy is a circular exercise that wastes renewable electricity. Hydrogen’s real value in ammonia is chemical, not energetic.
In my analysis I point out that one ton of ammonia fertilizer produced with green hydrogen (roughly 0.18 tons of hydrogen per ton of ammonia) can ultimately support the generation of about 31 tons of biofuel through increased crop yields and subsequent biomass conversion. This multiplier arises because the ammonia drives growth of feedstock crops, which in turn become feedstock for bio-fuels; the hydrogen is indirectly leveraged manyfold. In contrast, manufacturing hydrogen to be used directly as a fuel or via synthetic fuels from captured carbon fails to deliver the same “yield chain” benefit and remains far more energy intensive and costly. The implication is clear: instead of using hydrogen as a direct energy carrier, the logic of fertilizer-enhanced bio-feedstock paths offers far better economics and carbon effectiveness in the near to mid-term.
For methanol and similar products, economics will drive production toward gasified biomass rather than synthetic hydrogen and captured carbon. Biomethanol made directly from waste or residue feedstocks avoids the cost and inefficiency of manufacturing hydrogen and separating CO₂. Gasification produces both the carbon and hydrogen needed in a single process, already balanced for methanol synthesis. In a world constrained by clean energy supply, this route will dominate because it uses existing waste streams and proven chemical pathways instead of expensive hydrogen and carbon capture systems.
The study gives modest credit to hydrogen based synfuels, especially for aviation and shipping, but even there the economics are thin. My view is that biofuels are the better path for long-haul transport. Waste derived hydrotreated vegetable oil and similar second-generation liquids can be produced today at industrial scale without creating a new energy infrastructure. Treating biofuels is the only growth segment for hydrogen, in my opinion. Battery ships and electrified ports are already operating, and they will handle most maritime traffic long before hydrogen combustion or ammonia bunkering make sense. Electrons first, biofuels second, hydrogen only where nothing else fits.
ChatGPT generated panoramic infographic showing how most CCS concepts fail the purity, proximity, and economics filter, leaving only a few viable industrial niches
One area where the paper and I diverge is carbon capture. The researchers model blue hydrogen and biomass gasification with CCS as credible low-carbon options. The economics are rarely that simple. Carbon capture only makes sense where the CO₂ stream is already pure, the storage formation is close, and the product value can carry the extra cost. That means places like ethanol plants and some cement facilities. It does not make sense for hydrogen made from methane in dispersed plants that would require new pipelines and compression. CCS has a role, but a narrow one.
If all hydrogen projects in the database were built, the world would still see only a small climate dividend. The reason is that half of the emissions come from the way hydrogen is used, not from how it is produced. A low-carbon molecule can still be wasted in a bad application. Making hydrogen for home heating or power generation burns through renewable energy that would cut more emissions if sent directly to the grid. In effect, we are spending two units of clean power to get one back.
In recent commentary I have emphasised that hydrogen is not a benign gas in climate-terms simply because it emits no CO₂ when burnt; when it leaks it behaves as an indirect greenhouse agent, interfering with hydroxyl radicals that clean methane and thereby extending methane’s lifetime in the atmosphere. I noted measurable real-world leak rates in hydrogen systems of 1%–2% even in well-controlled industrial sites, and argued that at full energy-system scale the risk could rise to 5%–10%. In that light I welcomed that the new study explicitly includes the global warming potential (GWP) of hydrogen leakage in its life-cycle assessment: it uses a factor of about 11 kg CO₂e per kg H₂ to account for indirect warming effects, and therefore embeds leak-driven climate burden into the produced molecule’s emissions intensity. That’s conservative as many, including me, argue that GWP20 for methane is the more appropriate measure.
That alignment is important because many hydrogen proponents omit or minimise leakage and indirect effects. Nonetheless, the study does not appear to model detailed leakage scenarios by supply-chain step (production, storage, transport, distribution) or run sensitivity analyses of high-leakage paths, so my caution remains: if hydrogen is scaled beyond tight industrial loops, and if leak rates drift upward, the hidden warming cost will erode much of its claimed climate value.
Adjusted hydrogen demand through 2100 by author
My own projection of future hydrogen demand assumes that alternatives to hydrogen DRI will dominate ironmaking and that electricity will keep taking market share in transport and heat. In that scenario, total hydrogen use in 2050 is much less than what the IEA imagines. The world will still need hydrogen, but mostly as a cleaner version of the same industrial chemical it has always been. Even there, the higher cost of low-carbon hydrogen will drive substitution and efficiencies, and of course the largest use-case, crude oil processing and refining, is going to collapse, with the heaviest and highest-sulfur crudes and hence the highest hydrogen demand segment off the market first due to economics.
Policy should catch up to that reality. Governments should stop chasing hydrogen for every purpose and concentrate on cleaning up the hydrogen we already use. Investors should focus on low-carbon ammonia, industrial pilots for green iron, and the infrastructure that connects those industries safely to renewables. Public funds should not subsidize hydrogen cars, hydrogen heating, or long-distance hydrogen shipping. Those are distractions, not solutions.
When I first read the Nature Energy paper, I felt the tug of confirmation bias. It was saying exactly what I have been arguing for years—that hydrogen’s useful range is narrow and industrial, not universal. That sense of vindication is pleasant but dangerous. I’ve written before about how cognitive biases shape perception of climate solutions, especially the tendency to seek reinforcement instead of contradiction. So I treated this study as a test case. Was I agreeing because it was right, or because it agreed with me? The answer came from its design and its provenance. Nature Energy has an impact factor above 60—incredibly high for a scientific journal when most credible journals are in the two to five range—and is one of the most rigorous peer-reviewed journals in the field. The authors are researchers from the Paul Scherrer Institute, ETH Zürich, Leiden University, and the German Aerospace Center—public institutions with no stake in hydrogen advocacy or opposition. Their funding came from national research agencies, not industry. The analysis used 2,000 real projects, modeled full life cycles, and included indirect warming from hydrogen leakage. It tested hydrogen’s effectiveness against both fossil and fully decarbonized reference systems, avoiding the easy trap of comparing to the worst case. The authors made their data and code public, allowing anyone to check their results. That combination of transparency, academic independence, and statistical discipline—not the fact that it matched my own conclusions—is what made me confident that the findings were solid.
The broader lesson is about discipline. Every decarbonization pathway looks large until someone measures it properly. When a comprehensive, peer-reviewed study lands in the same place as practical engineering and market analysis, that is not confirmation bias. It is convergence. Hydrogen will play a role in the transition, but a limited one. The faster we align ambition with that reality, the faster we cut the emissions that matter.
Sign up for CleanTechnica’s Weekly Substack for Zach and Scott’s in-depth analyses and high level summaries, sign up for our daily newsletter, and follow us on Google News!
Advertisement
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Sign up for our daily newsletter for 15 new cleantech stories a day. Or sign up for our weekly one on top stories of the week if daily is too frequent.
CleanTechnica uses affiliate links. See our policy here.
CleanTechnica’s Comment Policy