Hydrogen Demand Shrinks When It Is Counted Properly

Support CleanTechnica’s work through a Substack subscription or on Stripe.

Hydrogen demand is usually discussed as if the world is about to discover a giant new fuel market. That is the wrong starting point. Hydrogen is already a very large industrial molecule, but most of that demand is tied to the fossil-fuel system, fertilizer, methanol and process chemistry. It’s a climate change problem on the scale of all global aviation today. When those sectors are counted properly, the transition story is not explosive growth. It is contraction, cleanup and concentration.

That distinction matters because hydrogen has been sold for years as a future energy carrier for road transport, home heating, grid storage, shipping, aviation, steel, backup power and industrial heat. Many of those claims treat hydrogen demand as if every hard-looking energy problem automatically becomes a hydrogen market. The better test is narrower: where is hydrogen already used, which of those uses survive decarbonization, and which new uses can beat direct electrification or better molecules?

I’ve formalized and normalized my hydrogen demand projection through 2100 as part of the series of decade by decade decarbonization projections across industry and transportation I now maintain at TFIE Strategy Briefing.

Start with today’s hydrogen demand. It is barely an energy carrier, with homeopathic amounts used in the very small number of hydrogen vehicles that exist. It is almost entirely a feedstock and process input. The largest categories are oil refining, ammonia for fertilizer, methanol and a set of smaller industrial uses. Most hydrogen today is also made from fossil fuels, primarily natural gas and coal, so cleaning existing hydrogen is already a major climate task before adding speculative new demand.

Refining is the first large denominator problem. Refineries use hydrogen to remove sulfur and upgrade heavier hydrocarbons. That demand exists because the world processes enormous amounts of crude oil. As road transport electrifies, oil demand declines. As internal combustion vehicle sales fall, gasoline and diesel volumes fall. As petrochemical growth slows and carbon constraints tighten, the refining system shrinks and changes shape. Hydrogen used to make fossil fuels cleaner at the point of combustion does not survive unchanged in a world using fewer fossil fuels.

There will still be some residual refining. Petrochemicals do not disappear. Some biofuels require hydrotreating. But the idea that refining hydrogen demand simply becomes a base on which a giant new hydrogen economy is built is backwards. A major existing source of demand erodes as the fossil-fuel system erodes.

Ammonia is the next large category. Fertilizer is real hydrogen demand, and it is not going away. The world still needs nitrogen fertilizers, especially where soils, yields and food security depend on them. But that does not mean ammonia demand grows without limit. Better application, precision agriculture, reduced losses, manure management, nitrification inhibitors, crop rotation, biological nitrogen fixation and shifts in diet can all moderate nitrogen demand. The climate task is to clean the remaining ammonia supply, not to pretend that fertilizer becomes proof of hydrogen as a general-purpose fuel.

Methanol and process chemistry are different again. Some methanol remains useful as a chemical feedstock. Some industrial processes need hydrogen as a reducing agent or chemical input. These are not the same as burning hydrogen for heat or pushing it through urban gas networks. They are molecule uses, not generic energy uses. That is where hydrogen is strongest: where the hydrogen atom is needed, not merely where energy is needed. And there are biomass to methanol processes that create the precursors for methanol without manufacturing hydrogen.

Steel is often presented as a major future hydrogen market, but it is more bounded than the hype suggests. Scrap electric arc furnaces expand as more steel becomes available for recycling. Direct electrochemical ironmaking and other emerging pathways compete for new iron production. Some direct reduced iron using hydrogen may make sense in regions with excellent renewable resources, high-grade ore access, suitable infrastructure and strong carbon constraints. That is a real segment, but not a universal hydrogen destiny for steel.

Transport is where most hydrogen-demand projections fail the simplest adoption test. Battery-electric vehicles have won light-duty road transport. Electric buses are winning against hydrogen buses on cost, reliability and infrastructure. Battery-electric trucks are moving into return-to-base, port, regional and increasingly heavy freight applications faster than hydrogen truck ecosystems can mature. Rail electrification and batteries beat hydrogen in all rail use cases. Hydrogen vehicles require not just vehicles, but a parallel fuel system with production, compression or liquefaction, distribution, storage, dispensing, maintenance and utilization high enough to pay for all of it.

Shipping and aviation do not rescue the story. Shipping fuel demand falls as fossil cargoes decline, batteries take inland and short-sea routes, efficiency improves and residual liquid fuels are allocated carefully. Hydrogen, ammonia and synthetic fuels remain expensive, lossy and infrastructure-heavy, with mostly biomethanol and ethanol supplying energy in my projections. Aviation will still need liquid fuels, but that points toward biomass derived sustainable aviation fuels, not large-scale hydrogen aircraft. Hydrogen aviation adds storage, volume, airport logistics, certification and aircraft-design burdens that make it a poor general answer.

Grid storage is another category where hydrogen is repeatedly overcounted. Storage is not one thing. Seconds, minutes, hours, days, weeks and rare strategic reserve events are different grid services. Batteries, pumped hydro, transmission, demand response, thermal storage, hydro reservoirs, existing dispatchable assets and overbuilt renewables cover much of the useful work before hydrogen-to-power enters the frame. Hydrogen may have niche reserve roles in some systems, but low utilization and round-trip losses make it a weak general storage solution.

Industrial heat follows the same pattern. Electricity, heat pumps, resistance heating, induction, electric arc furnaces, thermal storage and better process integration cover far more of the heat market than hydrogen advocates usually admit. Hydrogen combustion can produce high temperatures, but so can electricity in most industrial contexts. The fact that hydrogen can burn does not make it the lowest-cost, lowest-infrastructure or best-controlled way to deliver heat.

The resulting pathway is not complicated. Clean up the hydrogen we already use where the molecule is genuinely needed. Expect refining demand to fall. Expect fertilizer demand to moderate and diminish slightly rather than explode. Keep methanol and process chemistry in the surviving molecule bucket. Add some hydrogen for selected new industrial uses, some biofuel hydrotreating and some residual chemical pathways. Remove the fantasy uses in road transport, home heating, most grid storage, most shipping, most aviation and most industrial heat.

That produces a much smaller hydrogen future than the one sold in many corporate decks. It is also a more useful one. Hydrogen remains important, but as an industrial molecule to clean up and allocate carefully, not as a replacement for oil and gas across the energy system.

The policy implication is straightforward. Stop measuring hydrogen by project announcements, electrolyzer aspirations or generic “hard-to-abate” slogans. Measure it by surviving demand. Which sector? Which molecule requirement? Which incumbent hydrogen demand declines? Which electrification pathway competes? Which infrastructure has to be built? Which utilization rate pays for it? Which buyer can afford the green premium?

Once those questions are applied, the hydrogen boom becomes a hydrogen triage. Existing industrial demand has to be decarbonized. Some new industrial uses survive. Most energy uses fail against direct electrification, batteries, transmission, efficiency or better liquid-fuel allocation. Hydrogen demand through 2100 is much more likely to shrink than grow. This used to be a radically heterodox position, but it’s becoming much less of one as reality catches up with even the most ardent proponents.

For the full TFIE Strategy Briefing pathway review, read Hydrogen Demand Through 2100: Transition Pathway Review.

The Briefing version includes the sector-by-sector demand model, decadal pathway, emissions framing and survival filters behind the conclusion: hydrogen remains an important industrial molecule, but it does not become the new oil.