e-Fuels

e-Fuels is one of several names for the same idea. You will also see Power-to-X (PtX), Power-to-Liquid (PtL), Power-to-Gas (PtG), electrofuels, synthetic fuels, and, in European regulation, Renewable Fuels of Non-Biological Origin (RFNBOs). The terms overlap and partially conflict. Different policy frameworks and industry bodies use them in slightly different ways.

The defining characteristic, across all the names, is the origin of the energy and the carbon. e-Fuels are made from renewable electricity and a non-fossil carbon source. The electricity provides the energy. The carbon (or, in the case of ammonia, nitrogen) provides the molecular backbone. The product is a fuel that is chemically similar or identical to its fossil counterpart, designed to work in existing engines, pipelines, ships, aircraft, and chemical plants.

This last point matters. e-Fuels exist not because synthesising hydrocarbons from H₂ and CO₂ is novel (the chemistry is decades old, in some cases a century old), but because the world has built trillions of euros of infrastructure that runs on hydrocarbons, and replacing that infrastructure wholesale is slower and more expensive than feeding it with chemically equivalent renewable molecules. e-Fuels are, fundamentally, a transition technology designed to use existing infrastructure during the time when that infrastructure cannot yet be replaced. They are not a long-term ideal solution. They are a pragmatic engineering response to the inertia of the energy system.

It is worth being precise about what e-fuels are not. They are not biofuels. Biofuels use carbon that has been fixed from the atmosphere by biological processes, then converted into fuel by biological or thermochemical processes. e-Fuels use carbon that is captured from the atmosphere or from a point source by industrial processes, then converted by catalytic chemistry. The two are sometimes combined (an e-fuel made from biogenic CO₂ has the practical climate accounting of a biofuel, but the production economics of an e-fuel), but they are conceptually distinct. They are not synthetic fuels in the historical sense either. The synthetic fuels of the 1940s and the 1980s, made from coal via Fischer-Tropsch in Germany and South Africa, used fossil carbon. They were chemically identical to today's e-fuels at the molecular level but very different on every dimension that matters for climate.

Every credible e-fuels project answers four questions, in this order. Getting the order wrong is one of the most common reasons that early-stage projects under-deliver against their announcements. Each question conditions the next. The earlier questions are strategic; the later questions are technical. Skipping the strategic questions because the technical ones are more concrete and more familiar is a recurring failure mode in this industry.

The most consequential question, and the one most often skipped. Direct electrification is more energy-efficient than e-fuels by a factor of three to five in almost every application where both are technically feasible. A battery electric vehicle converts roughly 75 to 80 percent of the renewable electricity at its grid connection into kinetic energy at the wheels. A vehicle running on e-diesel produced from the same electricity converts roughly 13 to 16 percent. The five-fold difference is not marginal. It is structural. It means that displacing one tonne of fossil fuel with an e-fuel requires roughly five times the renewable electricity capacity as displacing the same tonne via direct electrification.

This matters for project rationale. The right time to make an e-fuel is when direct electrification is not technically feasible (aviation at scale, deep-sea shipping, certain high-temperature industrial processes), when the molecule itself is the product rather than the energy (fertiliser, chemicals), or when existing infrastructure can be decarbonised faster by feeding it a green molecule than by replacing it (natural gas networks in some regions, the existing chemicals industry). The wrong time is when direct electrification is straightforward and someone is proposing an e-fuel anyway, usually because they have stranded assets in fossil infrastructure they would prefer to keep.

A useful frame is to ask, for any proposed e-fuel application, what the realistic alternative is. If the alternative is direct electrification, the bar for choosing e-fuels should be high and specific. If the alternative is biofuels with limited feedstock availability, the bar is lower. If the alternative is "no decarbonisation," as is essentially the case for aviation today, the bar is lowest, and e-fuels become essential rather than optional. Honest project framing answers this question before any technology choice gets made.

The end molecule is determined principally by the end-use application. Aviation needs kerosene-range hydrocarbons. Long-distance shipping is converging on e-methanol with e-ammonia as a secondary candidate. The existing chemicals industry already consumes methanol and ammonia at large scale. Natural gas networks want methane. Heavy-duty road transport can run on e-diesel, e-methanol, or e-DME depending on regional infrastructure.

The molecule choice cascades downward through every other decision in the project. The synthesis route follows from the molecule. The reactor design follows from the synthesis route. The downstream upgrading follows from the molecule's specifications. The certification and offtake strategy follows from the end-use market. A project that picks the wrong molecule for the target market commits itself to either an expensive late-stage redesign or a product that has no buyer.

This is the question where the engineering and the commercial strategy must align early. A green hydrogen producer with an offtake agreement for e-methanol for marine fuel is structurally different from one targeting e-kerosene for aviation, even if the upstream H₂ and CO₂ supplies are identical. The molecule choice is not a downstream detail. It is the early strategic anchor of the project.

Three sources, with materially different cost, climate accounting, and regulatory treatment.

Industrial point-source CO₂ is the cheapest in 2025, with capture costs typically in the range of €30 to €150 per tonne depending on the source (ammonia plants, ethanol fermentation, cement, refineries). It is also the most available at scale today. The complication is the underlying source: if the CO₂ comes from an unabated fossil process, then producing an e-fuel from it is, at best, delaying the emission rather than preventing it. The carbon is captured, incorporated into the fuel, and re-emitted when the fuel is burnt. Under European RED III rules, fossil point-source CO₂ qualifies as a source for RFNBO e-fuels only until 2041, and only under significant restrictions.

Biogenic CO₂ comes from sources where the carbon was recently fixed from the atmosphere by biological processes: biogas upgrading, ethanol fermentation, biomass combustion. Cost is comparable to point-source CO₂ (€30 to €100 per tonne), and the climate accounting is genuinely circular: burning a fuel derived from biogenic CO₂ returns to the atmosphere carbon that came from the atmosphere within the recent biological cycle. The constraint is supply. Globally available biogenic CO₂ from concentrated sources is in the hundreds of millions of tonnes per year, which sounds large until it is compared to the gigatonnes of CO₂ that decarbonised aviation, shipping and chemicals would need.

Direct Air Capture (DAC) captures CO₂ directly from ambient air. It is, in principle, infinitely scalable and unambiguously climate-positive when paired with renewable energy. It is also, in 2025, expensive. Capture costs are in the range of €200 to €600 per tonne today, with credible cost reduction trajectories targeting under €200 per tonne by 2030 and €100 per tonne by 2040. For a green e-fuel project, the difference between point-source CO₂ at €50 per tonne and DAC at €400 per tonne is the difference between competing pricing today and prohibitive pricing today. The trajectory matters, and projects designed with long operational lifetimes need to consider it.

The CO₂ source decision is therefore not just a question of cost. It is a question of climate credibility, regulatory eligibility, and long-term project economics. The right answer is project-specific.

The technology choice. Determined by the molecule selected in question two and the engineering trade-offs among the available routes.

For e-methanol, conventional catalytic methanol synthesis is the dominant route, with several vendors offering commercial designs. The chemistry is mature, the reactors are well understood, and the synthesis is single-step. The main engineering question is integration with variable upstream H₂ supply, which is similar to the integration question for Haber-Bosch.

For e-methane, the Sabatier reaction is mature and operates at relatively low temperatures (~300 to 400°C), which gives flexibility on integration. Cost-competitive at small scale, though the price gap to fossil natural gas is large.

For e-Fischer-Tropsch products, the choice of upstream syngas generation (RWGS versus SOEC co-electrolysis) and the choice of FT catalyst (cobalt versus iron) are the consequential decisions, covered in depth on the Fischer-Tropsch pillar page.

For methanol-to-jet, the multi-step upgrading sequence (methanol-to-olefins, oligomerisation, hydrogenation) is more complex than FT but uses more individually-mature processes. The trade-off between MtJ and FT-PtL for SAF is one of the most important open questions in the synthetic fuels industry today.

The technology choice should follow from the strategic decisions, not lead them. A project that picks the synthesis route first and then works backward to find applications is structurally weaker than one that picks the application first, the molecule second, the carbon source third, and the synthesis route fourth.

The energy story of e-fuels is the same as the energy story of green hydrogen, with additional steps added at the synthesis stage. Every step costs efficiency. The cumulative cascade is long enough that the well-to-wheels efficiency of e-fuels in mobility applications is consistently in the 10 to 20 percent range, compared to 70 to 80 percent for direct battery electrification. This is the structural reason that direct electrification, where feasible, is always preferred.

Begin with 100 units of renewable electricity at the grid connection. After electrolysis at 65 to 70 percent efficiency, 65 to 70 units of hydrogen energy remain. The carbon source contributes about 5 to 15 additional units (the CO₂ capture itself consumes energy, particularly for DAC). The synthesis step varies by molecule: methanol synthesis is relatively efficient (~85 percent energy retention), Fischer-Tropsch is moderate (~75 percent), methanation is lower (~80 percent but with significant heat losses). Each synthesis step is exothermic, and the heat is usually low-grade and difficult to recover usefully. Downstream upgrading (hydrocracking for FT products, oligomerisation for MtJ) costs another 5 to 10 percent.

The end-use combustion is then the largest single loss. An internal combustion engine running on e-diesel converts roughly 25 to 30 percent of the fuel's energy into useful work. A modern marine engine on e-methanol does slightly better. A gas turbine on e-methane is similar. A fuel cell on e-hydrogen does notably better than combustion (around 50 to 60 percent), but fuel cells in mobility applications are rare today.

The cumulative effect, well-to-wheels, sits in the ranges noted above. For a passenger vehicle running on e-diesel, the cumulative efficiency is around 13 percent. For an aircraft running on e-kerosene, the cumulative efficiency to thrust is similar. For a ship running on e-methanol, modestly better. These figures are not arguments against e-fuels in the applications where they are needed. They are arguments for being specific about where e-fuels are needed, and for not deploying them in applications where electrification is feasible.

The gap between announced e-fuels capacity and operational e-fuels capacity is, in 2025, very large.

ReFuelEU Aviation requires 1.2 percent of European aviation fuel to be synthetic e-fuel by 2030. At current jet fuel consumption, this implies approximately 750 thousand tonnes of synthetic SAF per year in 2030. Operational synthetic SAF production in Europe in 2025 is in the low tens of thousands of tonnes per year, across a handful of pilot and demonstration plants. The gap is roughly two orders of magnitude, to be closed in five years.

Similar gaps exist elsewhere. The IMO's revised greenhouse gas strategy implies hundreds of millions of tonnes of zero-emission marine fuel demand by 2050. Operational green ammonia and e-methanol marine fuel production in 2025 is a small fraction of that. Multiple national hydrogen strategies imply gigawatt-scale electrolyzer capacity in five years' time. Operational capacity today is in the tens to low hundreds of megawatts.

These gaps are not unique to e-fuels. Almost every decarbonisation technology shows similar patterns. What is specific to e-fuels is the integrated nature of the project: an e-fuels plant needs gigawatts of dedicated renewable generation, hundreds of megawatts of electrolyzer capacity, a CO₂ supply, a synthesis train, and downstream upgrading, all permitted and built and commissioned together. The integration is the hard part, and the timelines for first-of-a-kind projects are consistently slipping. Most flagship projects announced in 2022 to be operational by 2026 are now targeting 2028 or 2029. Some have been quietly cancelled.

The honest framing is that the late 2020s and early 2030s will be the period in which the e-fuels industry establishes whether it can deliver at scale, or whether the policy ambitions need to be recalibrated against the engineering and project finance reality. Both outcomes are still genuinely possible. The next few years will be decisive.