US Bets on Nuclear Cargo Ships to Break China's Shipbuilding Grip

US Transportation Secretary Sean Duffy used a post on X on 28 August 2026 to promise a new fleet of nuclear-powered cargo ships. He claimed vessels that would carry more cargo because they need no internal fuel tanks, cross oceans up to 75 percent faster, run for two decades without refuelling and emit nothing at sea. He also said the first nuclear-powered vessel is set to be built in 2028. The post is the loudest public signal so far of a policy push that has been building inside the Department of Transportation since spring.

The programme started on 7 May 2026, when Duffy and the Maritime Administration launched an initiative to develop small modular reactors, known as SMRs, for commercial shipping. The first concrete step was a Request for Information asking shipbuilders, reactor designers and operators how a US-built, scalable and commercially viable marine reactor could work. Industry comments were due by 5 August 2026.

Three agreements have followed. In June the Port of Long Beach became the first US seaport to sign with the Maritime Administration, creating what the department calls the nation's first testing area for nuclear-powered vessels. That framework brings in the Coast Guard, the Department of Energy and the Nuclear Regulatory Commission on safety standards, inspections and operating protocols. On 19 August the Port of Corpus Christi signed a similar memorandum covering reactors, port microgrids, shoreside power architecture and workforce training. On 24 August the Maritime Administration signed a third memorandum, this time with CORE POWER (US) Inc., a developer of maritime reactor technology, to work through regulatory acceptance, port operations, crew credentialing, nuclear fuel and lifecycle services, insurance and finance. The company says it is targeting first construction starts from 2028, subject to regulatory, commercial and industrial readiness.

None of these documents authorises the construction of a nuclear merchant ship or commits the federal government to a reactor design. They are exploratory frameworks, which matters for anyone modelling capacity that might arrive in the 2030s.

Nuclear propulsion at sea is not new. Submarines, aircraft carriers and Russian icebreakers have used it for decades. Civil cargo shipping is a different story, because only four nuclear merchant ships have ever been built. The American NS Savannah, launched in 1959 under the Atoms for Peace programme, carried passengers and then cargo before being retired in the early 1970s. Every port call required lengthy negotiations over berthing, emergency procedures and public reassurance, which made commercial scheduling close to impossible. West Germany's Otto Hahn carried ore for nine years and covered around 1.2 million kilometres, visiting 33 ports in 22 countries, most of them only once and each time under a special permission. Its reactor was removed and replaced by a diesel plant at the end of the 1970s. Japan's Mutsu leaked neutrons through a shielding fault during reactor testing in 1974, never carried commercial cargo and was later converted to conventional power. Russia's Sevmorput, delivered in 1988, is the last of the four still on the books, but it has spent much of the past two decades idle or under repair and Russian authorities have moved to retire it.

The commercial case for a new attempt rests on energy density. A compact reactor can deliver power for years, so the space and weight normally taken by heavy fuel oil tanks becomes revenue-earning capacity. Bunkering stops disappear, fuel price volatility stops driving voyage economics, and higher sustained speed becomes affordable because fuel burn no longer rises steeply with speed. For long ocean legs that are hard to decarbonise, the emissions argument is real, since a reactor burns nothing while operating.

The drawbacks are just as concrete. Capital cost per ship is far above a conventional newbuild. Insurance is the sharpest constraint, because nuclear liability is potentially enormous and existing marine cover was never written for reactors. Crews need nuclear qualifications that the merchant labour market does not produce at scale today. Fuel supply, refuelling, spent fuel handling and eventual decommissioning all need dedicated shoreside infrastructure that no commercial port currently has. CORE POWER says it intends to build a US nuclear-grade integration and lifecycle shipyard covering installation, first fuelling, testing, commissioning, refuelling and through-life support. Public acceptance is also unresolved, because a reactor calling at a busy urban port is a political question as much as a technical one.

The hardest operational question is geography. Container and bulk trades depend on chokepoints, and chokepoints are exactly where a nuclear ship loses freedom of movement. Otto Hahn was never allowed through either the Panama Canal or the Suez Canal, which forced longer routes and ruined its schedule economics. A canal transit means hours at low speed in a narrow, shallow and heavily used channel with very little room to manoeuvre, and both waterways are single points of failure for global trade. A grounding, a collision or a fire involving a reactor vessel would raise a contamination question in confined shallow water, and the canal authority would have to weigh that risk against the cost of closing the route. The Ever Given grounding in 2021 showed how fast a blocked canal spreads through supply chains, and a nuclear casualty would stretch the closure well beyond the salvage phase.

Straits raise a parallel problem. Malacca, Hormuz, Bab el-Mandeb, Gibraltar and the Danish straits are narrow coastal-state waters where transit rights coexist with national environmental and nuclear safety rules. The Turkish Straits are governed by the Montreux Convention and by Turkish traffic regulations that already impose strict conditions on hazardous cargo. Under Chapter 8 of the SOLAS convention, nuclear ships need arrangements between the flag state and the countries whose ports and waters they use, which turns each service loop into a diplomatic negotiation rather than a routing decision. Southeast Asian coastal states have long been sensitive about radioactive material moving through their waters, and attacks on merchant shipping in the Red Sea in recent years have added a security dimension that did not exist when the Otto Hahn was refused passage.

Washington is not moving into an empty field. CORE POWER frames the push as a response to China, which is developing nuclear designs for large commercial vessels and already dominates global shipbuilding. South Korea's HD Hyundai has presented an SMR-powered container ship concept in the 15,000 TEU class. The International Maritime Organization has begun updating its 1981 safety code for nuclear merchant ships, the Nuclear Energy Maritime Organisation was created in 2025 to coordinate standards, and the International Atomic Energy Agency is launching its ATLAS initiative on nuclear applications at sea.

No nuclear cargo ship will call at a US port before the 2030s, and the current agreements fund studies and frameworks rather than steel. The signals worth tracking are whether a reactor design is actually selected, whether the Nuclear Regulatory Commission and the Coast Guard publish a licensing pathway, whether insurers settle on a liability framework, and above all whether canal and strait authorities indicate they would accept a nuclear-powered box ship. If those answers stay negative, a US nuclear fleet would be confined to routes that avoid chokepoints, a far smaller market than the one described in the announcement.