Shipping’s Fuel Transition Starts With Less Fuel, Not Better Molecules
The serious maritime pathway is denominator-led: less fossil cargo, less raw iron ore, more short-route electrification, and scarce liquids only where ships still need fuel.
Shipping decarbonization keeps being framed as a contest among alternative fuels. The useful split is simpler: cargo that fades, electrons where routes fit, scarce residual liquids where ships still burn fuel, and fuel pathways that have not yet shown operating evidence. Ammonia, hydrogen, LNG and synthetic fuels all have advocates, pilots, procurement stories and policy narratives, but those signals do not answer the first question: how much fuel does shipping still need after the cargo and route filters have done their work?
Shipping’s core problem is not which new fuel replaces today’s oil products on a one-for-one basis. The core problem is how much liquid fuel shipping still needs after fossil cargo declines in a serious energy transition, after raw iron ore trade is structurally reduced by changes in steelmaking, after the routes structurally suited to electrification have shifted to batteries, shore power or battery-dominant hybrid operation, and after operational efficiency, routing, slow steaming, hull management and hybridization have reduced fuel demand where they can. A fuel-substitution model asks how to make today’s marine fuel stack greener. A transition pathway asks how much of today’s marine fuel stack should still exist.
The first denominator is not the ship. It is the cargo. Roughly 40% of current maritime tonnage is fossil fuels, and another roughly 15% is raw iron ore. In a decarbonizing world, coal, oil and gas cargo decline sharply. Raw iron ore is also structurally exposed as steel demand shifts, scrap use rises, electric arc furnaces take more share, and more iron reduction can occur near renewables-rich mining regions instead of near distant coal-based blast furnaces. That does not eliminate shipping, but it means the future fuel problem is not the same size or shape as today’s bunker-fuel problem.
More than half of today’s maritime tonnage is therefore structurally exposed before the fuel debate even begins. That is the key difference between this reference scenario and many International Maritime Organization (IMO)-style demand projections. IMO-linked demand projections tend to preserve more shipping growth through 2050 than looks plausible once roughly 40% fossil-fuel cargo and roughly 15% raw iron ore exposure are taken seriously. I dug into the provenance of the IMO projection, and it traces back to a late-2010s modeling exercise that does not adequately account for fossil-fuel cargo decline, raw iron ore exposure, slower Chinese infrastructure growth, scrap growth, electric arc furnace expansion or renewables-based iron reduction closer to ore bodies. In my reference scenario, those cargo shifts come first. Fuel substitution is solving the smaller shipping system that remains.
That is why the TFIE shipping fuels pathway starts with the denominator. It does not begin by asking whether ammonia, methanol, hydrogen, biofuels or synthetic fuels will “win.” It begins by asking what work remains fuel-burning after the energy transition has done its first-order work elsewhere. A smaller residual liquid-fuel pool makes constrained low-carbon liquids more plausible, while a one-for-one replacement of today’s oil demand makes most alternative-fuel stories look much weaker.
The workbook behind this article formalizes my existing shipping fuels projection as a reference scenario. A defensible shipping pathway has to be tied to useful work, which in shipping means cargo moved over distance. The Transition Pathway Initiative’s (TPI) international-shipping methodology uses emissions intensity in grams of CO₂ per tonne-kilometre, a useful discipline because it keeps vessel announcements, fuel options and policy targets from being mistaken for delivered decarbonization. Its 1.5°C pathway moves from 6.54 gCO₂/t-km in 2022 to 4.55 in 2030, 1.58 in 2040 and 0.40 in 2050.
The IMO strategy adds the policy checkpoint layer. The 2023 IMO strategy targets at least a 40% reduction in CO₂ emissions per transport work by 2030, at least 5% and striving for 10% zero or near-zero fuels or energy by 2030, net-zero GHG emissions by or around 2050, and absolute GHG reductions of at least 20% by 2030 and 70% by 2040, compared with 2008, with stronger “striving” levels of 30% and 80%. It also points toward well-to-wake accounting, which matters because marine fuels are not clean if emissions have merely been moved upstream.
Both methods push the analysis back to transport work before fuel choice. The relevant questions are how much cargo is moving, how far it moves, how much energy is required per unit of work, how much of that work disappears with fossil-fuel and raw-iron-ore decline, and how much of the remainder can be electrified directly.
Shipping still starts almost entirely with oil products. The International Energy Agency (IEA) says oil products have historically supplied over 99% of international shipping energy, while biofuels supplied less than 0.5% in 2022. In the IEA Net Zero Scenario, low-emission fuels reach almost 15% of international shipping energy by 2030, which would be a steep climb from a very small base.
A sector that is still more than 99% dependent on oil products will not decarbonize because ships are ordered with alternative-fuel capability or because a port hosts a demonstration bunkering event. Those are early signals. The evidence that matters comes later: delivered fuel, operating routes, verified lifecycle emissions, repeat procurement and economics that survive outside demonstration conditions.
Trade growth and route disruption complicate the picture further. UN Trade and Development’s (UNCTAD) 2025 maritime review reports 2.2% maritime trade growth in 2024, a slowdown to 0.5% in 2025, and an expected 2% annual average over 2026 to 2030. I doubt those projections as well given the current Strait of Hormuz-related demand destruction that’s ongoing and leading to substantial changes in national strategies related to energy. That does not defeat the denominator thesis, but it does mean the near-term pathway has to contend with cargo growth, rerouting, geopolitics and longer average hauls before the structural decline in fossil cargo and iron ore work becomes decisive.
The public conclusion is straightforward. Shipping is not going to decarbonize by finding a universal alternative fuel and applying it to the existing oil-shaped system. It has to reduce the amount of liquid fuel required, recognize that fossil-fuel cargo and raw iron ore are structurally exposed, electrify the routes where batteries and shore power are structurally advantaged, and reserve constrained low-carbon liquids such as biomethanol, biodiesel, hydrotreated vegetable oil (HVO) and potentially ethanol for the smaller pool of vessels and voyages that still need fuel.
The workbook’s central chart is a reference scenario, not a prediction. It asks what a plausible denominator-led pathway looks like if fossil cargo work declines, raw iron ore movement is structurally reduced, short-route electrification scales where routes suit it, operational efficiency keeps improving, and constrained low-carbon liquids are reserved for the residual work that still needs fuel.
In that reference scenario, residual liquid fuel falls from 215.6 Mt diesel-equivalent in 2020 to 196.3 Mt in 2030, 81.3 Mt in 2050 and 25.7 Mt by 2100. Fossil liquid fuel falls from 215.6 Mt in 2020 to 186.5 Mt in 2030, 32.5 Mt in 2050 and zero by 2100. Electric displacement rises from effectively nothing in 2020 to 6.3 Mt diesel-equivalent in 2030, 42.5 Mt in 2050 and 64.3 Mt in 2100.
Those numbers should be read as scenario anchors, not forecast points. They say what has to be true for this pathway to be plausible: fossil cargo work has to decline, raw iron ore shipping has to be structurally reduced, inland and short-sea routes have to electrify where they are structurally suited to electrification, deep-water vessels have to use less energy per unit of work, and residual low-carbon liquids have to be reserved for the work that remains liquid-fuel dependent.
The right public reading of the pathway is simple enough. Count fuel only where ships remain fuel-burning. Discount the parts of shipping where cargo disappears, where electrons win, and where batteries, shore power and route structure reduce the fuel pool. Then argue about residual liquids, not the entire legacy marine fuel pool.
Below the paywall is the professional layer: the transition-pathway framing, cargo-denominator logic, transport-work and IMO checkpoint caveats, residual-fuel split, comparator analysis, update triggers, decision implications and the scorecard I’ll use to judge whether shipping fuel pathways are scaling, progressing, niche-valid, stalled or merely generating activity.
