International shipping produces roughly one billion tonnes of carbon dioxide each year, accounting for nearly three percent of global greenhouse gas emissions. The International Maritime Organization has set a target of net zero by 2050. Yet the deep-sea cargo vessels that carry the bulk of world trade are among the hardest transport systems to decarbonise. Batteries lack the energy density for transoceanic voyages. Hydrogen bunkering infrastructure barely exists. Liquefied natural gas buys time but does not solve the problem. Into this gap, an idea from the mid-twentieth century has returned: nuclear propulsion.
Nuclear power at sea is not a new concept. Russia has operated nuclear-powered icebreakers on Arctic routes for decades. The United States Navy has run nuclear-powered carriers and submarines safely for longer still. The real question has never been whether the technology works. It is whether a commercial shipping fleet can be built around it. When the United States launched the NS Savannah in 1959, followed by Germany’s Otto Hahn and Japan’s Mutsu, all three vessels ultimately retired early — not because their reactors failed, but because ports refused to accept them and operating costs could not be justified commercially. A ship that functions and a shipping system that works are two entirely different things.
The current wave of interest is driven by the maturation of small modular reactor technology. SMRs are compact, factory-manufactured, and designed with greater inherent safety than conventional reactors, making them far more practical for integration into a ship’s engine room. China’s state-owned Jiangnan Shipyard has announced plans for a 25,000-TEU nuclear container vessel powered by a thorium molten-salt reactor — which, if built, would be the first of its kind in commercial history. HD Hyundai in South Korea is collaborating with classification society ABS on a conceptual design for a 16,000-TEU vessel driven by a 100-megawatt SMR. Research by Lloyd’s Register and Seaspan estimated that nuclear-powered containerships could eliminate bunker costs entirely and outperform both conventional and green-fuelled competitors over a vessel’s full operating life.
The obstacle is structural rather than technical. Putting a nuclear-powered merchant ship into service requires far more than a working reactor. Ports need to install specialist facilities and overhaul security protocols. Insurers need a functioning nuclear liability framework. Regulators across multiple jurisdictions need to agree on who approves a vessel flagged in one country and calling at ports in several others. In June 2025, the IMO’s Maritime Safety Committee approved amendments to the SOLAS Convention that for the first time open the door to SMR applications in commercial shipping. That is a step forward, but it remains the beginning of a process rather than a completed framework. No country or international body has yet produced a comprehensive licensing regime for nuclear-propelled merchant vessels.
This is a classic coordination failure. Shipping companies will not order nuclear vessels because no ports will accept them. Ports will not invest in receiving facilities because no nuclear ships are coming. Insurers will not develop products because there is no operating risk data. Regulators will not legislate because there are no applications to process. Every party is waiting for someone else to move first.
There is also an unexpected competitor. The artificial intelligence boom has generated enormous demand for land-based electricity, and the world’s largest technology companies are committing tens of billions of dollars to secure SMR capacity for data centres. The maritime industry cannot match that scale of capital mobilisation, and risks being crowded out of the supply chain for the very technology it needs.
Decarbonising shipping has no single solution. Green methanol, liquid ammonia, and battery power each find a role on shorter or specialised routes. But for the transoceanic trade lanes that carry the majority of global commerce, nothing else approaches nuclear fuel in terms of energy density. The promise of nuclear propulsion is genuine: zero carbon at sea, years of operation without refuelling, and more cargo space freed from the constraints of fuel storage. The distance between that promise and commercial reality, however, is filled by an entire ecosystem of infrastructure, regulation, and public acceptance that does not yet exist. The technology has arrived. The world is not ready for it.
