Hydrogen can cut global CO₂ by up to 3%—but only if it’s used in the right places. Here’s where hydrogen truly works for climate, and where electrification wins.
Most companies get hydrogen strategy wrong.
Billions are flowing into hydrogen right now, yet a new global analysis of ~2,000 planned projects shows a sobering reality: even if every project is built, they’d only cut about 0.5–3% of today’s global CO₂ emissions by 2043. And a big chunk of those projects are pointed at the wrong uses.
This matters because hydrogen sits at the crossroads of green technology, climate policy, and massive capital allocation. If you’re a policymaker, investor, or industrial leader betting on hydrogen as a climate tool, you don’t just need more hydrogen—you need it pointed at the right problems.
Here’s the thing about hydrogen: it’s a scarce, energy-intensive resource. Treat it like a universal silver bullet, and you waste time, money and carbon budget. Treat it like a precision instrument, and it becomes a powerful part of a net-zero system.
This post unpacks new research on the global greenhouse gas mitigation potential of existing and planned hydrogen projects and translates it into a practical playbook:
- Where hydrogen delivers the biggest climate bang per kilogram
- Where hydrogen use is worse than smart electrification
- How AI, data and life cycle assessment can steer hydrogen investments toward real impact
All within the broader context of green technology and AI-enabled planning that underpins this series.
Hydrogen’s Real Climate Potential: The Numbers
The new analysis looks at about 2,000 hydrogen projects worldwide, using detailed life cycle assessment (LCA) and future energy system modelling out to 2043.
Key headline numbers:
- Planned projects would produce 110 million tonnes of hydrogen per year by 2043.
- Their own production emissions would be about 0.23 GtCO₂e per year.
- Once you include how that hydrogen is used, total life cycle emissions rise to about 0.4 GtCO₂e per year.
- The net emission reduction versus fossil-based business-as-usual is 0.2–1.1 GtCO₂ per year.
- That’s only 0.5–3% of today’s global annual CO₂ emissions.
Two crucial insights come from this:
- Half of hydrogen’s climate impact is in the application, not just the production. Counting only “green hydrogen” production hides a big chunk of emissions.
- Hydrogen’s contribution to global decarbonization is modest unless it’s tightly prioritized. Trying to push it into every sector is a recipe for disappointment.
So the real question isn’t “How do we maximize hydrogen?” It’s “Where should we spend each tonne of low-carbon hydrogen to cut the most CO₂?”
Where Hydrogen Shines: High-Impact Applications
Some hydrogen applications deliver much higher emission reduction per tonne of H₂ because they replace processes that are genuinely hard to decarbonize otherwise.
The research consistently finds four front-runners.
1. Ammonia: Essential for Food, Tough to Decarbonize
Ammonia production (mainly for fertilizer) is extremely carbon intensive today because it uses fossil-based hydrogen.
Switching to low-carbon hydrogen for ammonia:
- Cuts large, concentrated CO₂ emissions from existing plants
- Targets a sector with few easy substitutes at industrial scale
From a climate perspective, low-carbon ammonia is one of the smartest early uses of hydrogen. For countries with big fertilizer industries, this is low-hanging fruit.
2. Steelmaking: Replacing Coal in the Blast Furnace
Steel is responsible for roughly 7–8% of global CO₂ emissions, driven by coal used as both a fuel and a chemical reductant.
Hydrogen can step in as a reductant in direct reduced iron (DRI) processes. In the projects analysed, hydrogen-based steelmaking:
- Delivers strong emission cuts versus coal-based steel
- Remains one of the most climate-effective uses of hydrogen, even in a low-carbon future scenario
If you’re in mining, steel, or heavy industry, hydrogen for DRI isn’t a niche experiment anymore. It’s one of the clearest climate-positive uses of hydrogen we have.
3. Second-Generation Biofuels: Hydrogen as an Enabler
Here “biofuels” means advanced fuels (like hydrotreated vegetable oil from used cooking oil), not food-based biofuels.
Hydrogen is used to upgrade these waste oils into drop-in fuels. In the study:
- These biofuels show exceptionally high emission reduction per tonne of hydrogen used.
- Even when compared to an already low-carbon reference system, they often remain beneficial.
Because the feedstock is a waste stream, you avoid many of the land-use and food-vs-fuel conflicts that plague first-gen biofuels.
4. Synthetic Fuels for Aviation and Shipping (With Caveats)
Hydrogen-derived e-fuels (synthetic kerosene, methanol, ammonia) become particularly interesting in long-distance aviation and shipping:
- Direct electrification is unlikely for long-haul flights or large ocean vessels by 2040.
- Sustainable biomass is limited; it can’t cover all aviation and shipping needs.
When hydrogen-based synfuels are compared against fossil kerosene, the emission reduction per unit of hydrogen can be substantial. Against best-possible biomass-based fuels, the advantage shrinks or disappears—but scaling biomass to cover all global fuel demand is unrealistic.
The honest takeaway: synthetic fuels are not perfect, but they’re probably necessary for deep decarbonization of global mobility.
Where Hydrogen Should Mostly Be Avoided
In many sectors, hydrogen looks attractive politically or on a whiteboard but performs poorly once you factor in physics, infrastructure, and life cycle emissions.
The study strongly suggests that hydrogen use in these areas should generally be discouraged when scalable electrified alternatives exist.
1. Road Transport: Batteries Win Most of the Time
For road vehicles, especially passenger cars and short/medium-haul trucks, the pattern is clear:
- Battery electric vehicles (BEVs) are far more energy efficient.
- When powered by low-carbon electricity, BEVs usually deliver larger emission reductions than fuel cell vehicles.
- Fuel cell trucks also come with extra embodied emissions in tanks and stacks.
Hydrogen trucks might make sense in a few niche cases (remote, cold, long-haul with limited grid), but as a mainstream road solution, they’re a poor climate investment compared to direct electrification.
2. Power Generation: Burning Hydrogen Is a Detour
Using hydrogen in gas turbines to generate electricity is often marketed as “green power.” The numbers say otherwise.
Compared with simply using:
- Low-carbon solar, wind, hydro or nuclear directly on the grid
- Or pairing renewables with storage
Hydrogen power generation usually:
- Wastes energy through conversion losses
- Yields higher life cycle emissions than direct renewables in a decarbonized system
If you’re considering hydrogen for power, treat it as a last-resort flexibility tool, not a default solution.
3. Domestic Heating: Heat Pumps Beat Hydrogen Boilers
Multiple independent reviews now converge on the same conclusion: hydrogen boilers are a bad way to decarbonize home heating.
This new analysis backs that up:
- Heat pumps powered by low-carbon electricity deliver far higher emission reductions.
- Hydrogen in homes is an energy- and infrastructure-heavy detour.
If you’re a utility or policymaker, betting big on hydrogen-ready gas grids for residential heating is likely to lock in an inferior solution.
4. Synthetic Methane for Grid Injection and Passenger Mobility
Converting hydrogen into synthetic methane and pushing it into existing gas grids or using it in passenger vehicles sounds convenient.
In practice:
- You add extra conversion steps and energy losses.
- Electrified alternatives (EVs, electric heat pumps) typically outperform on emissions and efficiency.
The pattern is the same: whenever you can directly electrify, you almost always should.
Hydrogen, AI and Life Cycle Data: A New Decision Toolkit
One of the most valuable aspects of this research isn’t just the conclusions—it’s the method. It combines:
- A global project database (~2,000 hydrogen facilities)
- Location-specific renewable yields (solar, wind, grid mixes)
- Prospective life cycle assessment databases updated along a 2 °C climate scenario
This is exactly where AI and data-driven green technology can change the game.
How AI Can Sharpen Hydrogen Strategy
I’ve found that the organizations making the best hydrogen decisions do three things well:
-
Run scenario-aware LCAs by default
Instead of a static “kg CO₂ per kg H₂” number, they:- Model future grid decarbonization
- Include application-side emissions
- Compare hydrogen against true low-carbon counterfactuals, not today’s fossil baseline
-
Use optimization tools to match hydrogen to its highest-value uses
This is a classic constrained-resource problem:- Limited low-carbon hydrogen
- Multiple potential uses with varying emission reduction per tonne
- Infrastructure, policy and cost constraints
AI optimization and simulation can help decide: “If I have X tonnes of hydrogen in 2035, where do I put it—steel, ammonia, aviation—to maximize CO₂ reduction?”
-
Continuously update decisions as technology and policy evolve
Hydrogen economics and power mixes are changing quickly. AI models that ingest new data (project status, costs, grid carbon intensity) can regularly re-rank priorities.
This is where green technology and AI intersect in a very practical way: not in abstract climate dashboards, but in steering billions in capex toward the most climate-effective hydrogen projects.
Strategic Guidance: If You’re Planning Hydrogen Now
If you’re shaping hydrogen strategy—for a country, region, or large industrial portfolio—here’s a distilled, action-oriented view from the analysis.
1. Prioritize High-Impact Sectors
Focus scarce low-carbon hydrogen where it does the most climate work:
- Ammonia (fertilizers and potentially shipping fuels)
- Steelmaking (hydrogen-based DRI)
- Advanced biofuels (upgrading waste streams)
- Synthetic aviation and shipping fuels (where electrification is not viable)
2. De-Prioritize Where Electrification Wins
Generally avoid building a hydrogen ecosystem around:
- Passenger cars and most road freight (go electric)
- Residential heating (go heat pumps and district energy)
- Bulk power generation (use direct renewables and storage first)
Hydrogen in these spaces might still show up in edge cases, but it shouldn’t be the backbone of your decarbonization plan.
3. Don’t Ignore the Application Side in LCA
When assessing a “green hydrogen” project:
- Always model the full value chain: production + compression + conversion + end use.
- Benchmark against future low-carbon alternatives, not just today’s fossil system.
- Treat any climate claim that ignores the application side as incomplete.
4. Treat the Implementation Gap as an Opportunity
The study highlights a big gap between:
- Announced hydrogen capacity (110 Mt/year by 2043)
- Modelled hydrogen needs in net-zero scenarios (110–610 Mt/year by 2050)
Instead of treating this only as a shortfall, use it as design space:
There’s still time to steer future hydrogen build-out away from low-value uses and toward the applications that truly move the needle.
Hydrogen as a Precision Tool in the Green Technology Toolbox
Hydrogen isn’t the main character of decarbonization. It’s a specialist.
- Electricity, efficiency and demand reduction do most of the heavy lifting.
- Hydrogen steps in where molecules matter: fertilizers, metals, long-distance fuels.
For this Green Technology series, the message is clear: AI, data and rigorous life cycle assessment are now essential to hydrogen strategy. Not as academic add-ons, but as core tools for:
- Screening project pipelines
- Ranking sectors by climate impact per tonne of H₂
- Stress-testing national hydrogen roadmaps
If you’re planning your hydrogen investments for the 2030s and 2040s, the better question isn’t “How big is our hydrogen economy?” It’s:
“For every kilogram of low-carbon hydrogen we produce, are we using it where it avoids the most CO₂?”
Get that right, and hydrogen becomes a powerful ally in a smarter, AI-informed green technology transition.