Where does hydrogen make sense, and where is it a waste?

Hydrogen belongs where electricity cannot go. The Energy Transitions Commission's Hydrogen Ladder sorts uses into four tiers: essential (steel reduction, ammonia for fertiliser), beneficial (heavy trucking, deep-sea shipping), useful (synthetic aviation fuel, seasonal storage) and wasteful (home heating). A hydrogen boiler delivers roughly 25%–30% efficiency end to end, while the same renewable electricity in a heat pump delivers 300%–400%.

The ranking rests on the arithmetic of the efficiency paradox. Trace the full chain from solar panel to wheels: the solar-to-electric-vehicle route passes through AC conversion, charging losses, battery storage, motor and drivetrain for an overall efficiency of about 77%–80%. The solar-to-fuel-cell-vehicle route passes through electrolysis (about 70%), hydrogen compression (about 90%), transport and storage (about 95%), a PEM fuel cell (about 60%), then motor and drivetrain, for an overall efficiency of only about 25%–35%. A hydrogen car therefore needs roughly three times the solar panels to deliver the same energy at the wheels. For family cars, where mileage is predictable and recharging is frequent, that gap leaves fuel cells almost no room to compete.

Where electricity cannot substitute at all, the same arithmetic reads differently. In steelmaking, the conventional blast furnace route uses coke as the reducing agent and emits about 2 tonnes of CO₂ per tonne of steel; replacing coke with hydrogen via direct reduced iron (H₂-DRI) plus an electric arc furnace cuts steelmaking emissions by more than 95%. Sweden's HYBRIT project — SSAB, LKAB and Vattenfall — delivered the first automotive parts made from hydrogen-based steel in 2021, with orders from Toyota and Mercedes-Benz. Fertiliser is the same story: Haber-Bosch ammonia synthesis consumes about 50% of world hydrogen production, and here hydrogen is a feedstock rather than a fuel, so electricity cannot replace it. Green ammonia displacing grey-hydrogen-based ammonia is green hydrogen's most certain near-term bulk demand.

The middle tier is driven by physics rather than efficiency. A 40-tonne long-haul truck would need a 5–8 tonne battery for 1,000 km of range, crushing its payload; the same 1,000 km comes from 80 kg of hydrogen at 700 bar, refuelled in about 15 minutes, matching diesel's rhythm. Deep-sea shipping still runs mainly on heavy fuel oil and is among the hardest transport modes to electrify, and green ammonia was named a key option in the International Maritime Organization's 2023 strategy. The 'useful' tier includes synthetic aviation fuel made from green hydrogen and captured CO₂ via Fischer-Tropsch synthesis — fully compatible with existing aircraft and airports but currently four to five times the cost of fossil kerosene — and seasonal storage in underground salt caverns, whose 25%–40% round-trip efficiency struggles against pumped hydro and compressed air but may be the only option where terrain rules those out.

The top tier is wasteful precisely because a clearly better option exists. Heating a home with a hydrogen boiler runs the long chain of renewable electricity to electrolysis to hydrogen to boiler to heat, at roughly 25%–30% overall efficiency, against 300%–400% for the same electricity in a heat pump. The UK's two hydrogen village trials, at Whitby and Ellesmere Port, were both cancelled in 2023 for lack of economic viability. Infrastructure widens the gap: the world's roughly 3 million kilometres of natural gas pipeline are specified in steels designed for methane molecules, and atomic hydrogen penetrates the steel lattice to cause hydrogen embrittlement, so pure hydrogen cannot simply be pushed through them. The storage penalty is unavoidable too — compressing to 700 bar consumes 10%–15% of the hydrogen's own energy, liquefaction to −253°C consumes 30%–35%, and liquid hydrogen tanks boil off about 1% per day. Carrier options each carry a price: liquid ammonia liquefies at −33°C and over 120 ports worldwide can already unload and store it, but it is acutely toxic (LC₅₀ around 1,000 ppm) and poorly controlled combustion produces N₂O, a greenhouse gas with 300 times the warming potential of CO₂; liquid organic hydrogen carriers such as dibenzyltoluene are non-toxic and compatible with existing oil logistics, but hydrogenation and dehydrogenation each consume 15%–25% of the hydrogen's energy, for a round-trip efficiency of only about 60%–70%.

Burning hydrogen in a domestic boiler has been likened by critics to using a Bentley to deliver takeaway — technically possible, economically indefensible.

This question is covered in depth in The Full Spectrum Every Energy Source Explained — A Singapore Perspective，第十四章 14.5–14.6、14.8