The second role of green hydrogen: how much environmental gain does green gasoline deliver?

Hydrogen breaks out of the fuel cell

When we talk about hydrogen in cars, we usually think of the fuel cell: hydrogen in, electricity out, only water vapour from the tailpipe. But a second role is emerging. That same green hydrogen can, together with CO2 from the air, be turned into liquid fuel: synthetic gasoline, diesel or kerosene that fits into existing engines without any modification. The industry calls this an e-fuel.

Recently four developments came together that all point to the same chain. Toyota, BMW, Bosch and Repsol launched a pilot in Spain with cars on 100% renewable gasoline. Britain's Zero Petroleum makes gasoline from air and water. Shell is building Europe's largest green hydrogen plant near Rotterdam. And Bosch, which we wrote about earlier, is an outspoken advocate of synthetic fuels. So suppose this route succeeds, what does it deliver to society? We dug into the numbers.

How green is a kilometre?

Start with the core question: how much CO2 is released per kilometre driven, across the full life cycle of car and fuel? The authoritative research institute ICCT calculated this for the EU in 2025. An ordinary petrol car works out at around 235 grams of CO2 per kilometre. A battery-electric vehicle (BEV) on the expected European power mix sits at 63 grams, and on purely renewable electricity even at 52 grams: roughly four times cleaner. A fuel-cell car on green hydrogen comes close to the BEV, around 50 grams.

And green gasoline? What is special is that the CO2 leaving the tailpipe was previously taken out of the air to make the fuel. If both that CO2 and the electricity are genuinely green, the net emissions across the chain are low. One honest caveat: a combustion engine on e-fuel still emits nitrogen oxides and particulates. So green gasoline mainly solves the CO2 problem, not local air quality.

Chart 1 — Life-cycle CO2 per kilometre

There is a second yardstick, and this is where it pinches: efficiency from energy to wheel. Of green electricity, roughly 70 to 80 percent ends up as motion in a BEV. In a fuel-cell car that is about 30 percent, and in an e-fuel combustion car just 10 to 15 percent. On pure efficiency the plug wins by a distance. But as we will see, the kilometre is not the whole story.

Chart 1b — Efficiency: green electricity to wheel

What you do not have to build

Efficiency, after all, is not the only sum. The biggest promise of green gasoline lies in what it lets you avoid. It fits the cars that already exist, the pumps already there, and flows through pipelines and tankers that already exist. No new infrastructure needed.

That touches precisely the pain points of large-scale electrification. Think of grid congestion: in the Netherlands the grid is already at its limit in many places. And think of the batteries themselves: they require lithium, cobalt, nickel and graphite, with the mining and geopolitics that come with them. A car on green gasoline needs no large traction battery and therefore far less of those raw materials.

Chart 2 — Critical minerals per car

Raw materials: is there enough for everyone?

Those batteries have to be made of something. Suppose we electrified not part but the entire global fleet of over 1.3 billion cars at once, with today's battery chemistry. Weigh the metals needed against known world reserves and it gives pause: for cobalt you would need more than all known reserves combined, and for lithium and nickel around 40 percent of them each.

Chart 10 — Materials to electrify all cars vs world reserves

That does not mean it is impossible. Reserves grow as prices rise, recycling brings material back into the chain, and battery chemistry is shifting fast toward types with less or no cobalt and nickel (LFP, sodium-ion). Leading analyses, such as the ICCT's, conclude there is enough for the transition. But it does show why you would not bet everything on one card. Green gasoline lets a large part of the fleet drive cleaner without touching a single gram of those scarce metals.

How fast can you decarbonise the fleet?

This may well be the strongest argument. Worldwide there are more than 1.3 billion cars with a combustion engine. In the EU there are some 250 million passenger cars, with an average age of around twelve years. Even if from tomorrow every new car were electric, it would take fifteen to twenty years for the existing fleet to cycle out.

That is where the power of a drop-in fuel lies: it reaches the cars already here, without anyone buying a new car. Where electrification works through the slow replacement of cars, green gasoline works through the tank.

Chart 3 — The existing fleet empties slowly

The gain if you switch now

Enough nuance, let us simply calculate the gain. Suppose the existing fleet switched fully to green gasoline within five years. Because these are cars that already exist, that gain counts immediately for the whole fleet, not only once everyone has bought a new car. For Europe's passenger cars, roughly 250 million of them, the avoided CO2 over five years runs into the order of two thousand megatonnes. Through electrification you reach only the newly sold cars in those same five years, and cumulatively you get no further than a fraction of that.

Chart 8 — Environmental gain over 5 years (cumulative CO2)

That gain is not free, and that belongs in the picture. Making green gasoline takes a lot of green electricity: feeding the whole European passenger fleet with it would require on the order of 2,900 terawatt-hours a year, against some 600 terawatt-hours if those same cars ran on batteries. That is roughly five times as much, and more than all of Europe uses in electricity today. It is a real price, but one you can largely pay with sun and wind in places and at moments where that energy is now thrown away.

Chart 9 — Green electricity both routes need

No holier than the pope

That brings us to an uncomfortable but important question: does everything have to be 100 percent zero-emission? Or are we already a long way there if we cut the whole fleet by, say, 90 percent? The EU itself chose the latter, and with reason. Do the maths. Make 100 percent of cars 90 percent cleaner with a drop-in fuel and you remove 90 percent of fleet emissions. Manage, in the same time, to make roughly a quarter of cars fully zero-emission, and you are stuck at 25 percent.

Chart 11 — 90% of everything vs 100% of a little

Ninety percent of everything beats one hundred percent of a little. A huge, relatively simple step, a cleaner tank for everyone, delivers more climate gain than a perfect solution that reaches only a fraction. That is the sober arithmetic behind green gasoline: not waiting for the ideal car for everyone, but letting the cars that already exist drive a good deal cleaner now.

What is realistic?

And production? It is still at the start. The Haru Oni pilot plant of Porsche and HIF in Chile is designed for 130,000 litres a year; the first plants of more than 500 million litres are now being announced, and the price is falling from around 5 euros per litre at demo plants to an expected 2 euros or so around 2030. Against global gasoline demand of 1,400 billion litres a year, that is small for now.

Chart 4 — Scale gap: production vs demand

But turn it around and simply calculate the gain. Every litre of green gasoline that replaces a litre of fossil gasoline immediately saves the CO2 of that litre. And that holds wherever you use it: avoided CO2 is avoided CO2, whether that litre ends up in a truck, a plane or just the car on the driveway. The often-heard rule that e-fuels should go to industry first changes nothing about that. For the climate, the avoided emissions count, not the sector in which it happens. The faster production grows, the more cars that already exist drive cleaner straight away.

Is it not hugely energy-hungry?

The best-known counterargument, often heard from battery-electric champions, is that e-fuels waste green electricity. And that is partly true: per kilometre an e-fuel car needs roughly five times as much green electricity as a BEV. But two caveats complete the picture.

The first is location. Green electricity is not equally productive everywhere. A wind turbine at the southern tip of Chile (Patagonia) runs roughly 60 to 70 percent of the time, against some 20 to 25 percent for a typical onshore turbine in north-west Europe. You just cannot get that energy to Rotterdam by cable; as a liquid fuel you can. And it is not only far away. In the Netherlands too the waste is growing fast: in 2024 the number of hours with negative electricity prices rose by around 75 percent to a record, moments when sun and wind generate more than the grid can absorb and part is simply switched off (curtailment). Precisely that otherwise-wasted electricity is the ideal feedstock for green hydrogen and green gasoline.

The second caveat: fossil gasoline is far from free in energy either. Crude oil has to be pumped, shipped, refined and distributed. Line up the total energy input per 100 kilometres and you see how misleading the bare efficiency comparison is: a fossil petrol car burns not only its own fuel, but also the energy to make and deliver it.

Chart 5 — Energy input per 100 km (incl. fossil petrol)

The honest conclusion: compare e-fuels with a new BEV and the e-fuel loses on efficiency, plainly. But that is not the right comparison. Green gasoline does not compete with tomorrow's electric car; it competes with the tank of fossil gasoline in the car already on the road today. And against that yardstick the picture is suddenly far less one-sided.

The hidden bill: the energy to build

Because the biggest costs often come before the first kilometre: in the factory. Building an ordinary petrol car takes some 18,000 kilowatt-hours of embodied energy, the energy in the steel, the aluminium, the glass and the assembly. That is roughly what an average Dutch household uses in electricity over six to seven years, before the car has driven a metre.

An electric car costs more to build, and the main culprit is the battery. Producing the cells costs, at modern plants, some 50 to 65 kilowatt-hours per kilowatt-hour of capacity; for a 60 to 75 kWh pack that is 3,000 to 4,500 kilowatt-hours of factory process alone. Count the whole chain, from mining to cell, and estimates rise to 100 to 180 kilowatt-hours per kilowatt-hour, putting the battery quickly at 7,500 to 13,500 kilowatt-hours: close to half of what a complete petrol car costs, in a single component. A car that keeps running on green gasoline does not have to be built again. That build energy is already paid.

Chart 6 — Energy to build a car

To be fair: an electric car earns that extra build energy back over its life, precisely because it is so much more efficient per kilometre, according to the ICCT after about 17,000 kilometres. For a new car it is therefore a one-off investment that pays for itself. But for the car already there, that bill simply is not needed.

Reinforcing the grid costs time and energy

There is another cost that rarely appears in the per-kilometre comparison: the power grid. Mass electric driving demands not just charging points but a fundamentally heavier grid. In the Netherlands the grid operators put that at nearly 200 billion euros of investment towards 2040, about 11,000 euros per inhabitant. And that is not only money: it is copper, steel, transformers and concrete, each with its own embodied energy, and above all time. The queues for grid capacity are growing, and major grid expansions easily take five to ten years. A car that refuels at the existing pump places no extra load on that grid.

Chart 7 — What grid reinforcement costs

Here too the nuance: the grid has to be reinforced anyway for the wider energy transition, since heat pumps, industry, solar and wind all demand it. But less simultaneous charging demand eases that pressure and the lead time, on a grid that is already straining.

What does this mean for society?

Look only at the kilometre and the plug wins on efficiency. Look at the whole system, from building and mining to reinforcing the grid and waiting, and an existing car that can be fuelled cleanly has an edge that is rarely counted: you green the fleet that already exists, without waiting, with less mining and less pressure on the grid. The downside remains: production is still small, the e-fuel is energy-intensive to make, and the local air pollution of a combustion engine does not disappear.

Our verdict

We do not see it as a contest with a single winner. For most new cars, electric driving is the efficient default. For heavy and long-distance transport, the fuel cell has strong credentials. And the unique value of green gasoline lies in what the other two do not reach: the more than a billion combustion engines that already exist. Every car that goes from fossil to near-neutral without being scrapped is a gain, and that gain counts wherever you use the fuel.

Sources

• ICCT, life-cycle greenhouse gas emissions of EU passenger cars (2025 update)
• Transport and Environment; eFUEL-TODAY; eFuel Alliance
• Porsche and HIF, Haru Oni; Toyota Newsroom; Shell; Bosch Mobility
• IEA, minerals in electric cars; Battery Design and Nature Energy, battery production
• Wikipedia and MacKay, embodied energy; NOS and Netbeheer Nederland, power grid
• Solar Magazine and TenneT, negative electricity prices and curtailment 2024; U.S. EIA, gasoline consumption