Nuclear-Powered Merchant Ships Under a US Flag: Reading the MARAD–Core Power Memorandum of Cooperation

On 24 August 2026 the US Maritime Administration (MARAD) and Core Power (US) Inc announced a non-exclusive Memorandum of Cooperation (MoC) to develop the regulatory, industrial, commercial and financing framework for future US-flagged nuclear-powered merchant ships. That is the whole of the publicly confirmed substance. There is no ship order, no disclosed fleet size, no engineering, procurement and construction contractor, no reactor vendor selection, no site, no capital stack and no final investment decision.

For an industry conditioned to treat "announced" as a synonym for "financed," that sounds like thin gruel. It is not. The binding constraint on nuclear merchant shipping has never been the reactor — pressurised water reactors have put to sea in warships for seven decades and in icebreakers since 1959. The constraint is the institutional scaffolding: who certifies a reactor that enters foreign ports, who insures the hull, who trains the crew, who classifies the containment, and who signs the mortgage. A memorandum of cooperation is exactly the instrument through which governments and developers attack scaffolding problems. The correct question is not "how many ships?" but "which of those five barriers does this document actually begin to move?"

This article sets out what is confirmed, what the MoC likely covers given the four domains named, and what steel and shipbuilding professionals should watch over the next 12–24 months.

What was actually announced

The event, as reported by World Nuclear News, is a non-exclusive memorandum between MARAD — an agency of the US Department of Transportation — and Core Power (US) Inc, the American arm of the UK-headquartered Core Power group founded by Mikal Bøe.

Three features of that description deserve emphasis:

• "Non-exclusive" means MARAD has not anointed a single developer. Other consortia can and presumably will seek parallel arrangements. This is deliberate: for a pre-competitive framework problem, exclusivity would be counterproductive.
• "Memorandum of Cooperation", not a contract, not a cooperative agreement with appropriated funds, and not a public-private partnership. It creates no procurement obligation and, on the public record, no money changes hands.
• "Framework", not "design" or "construction." The deliverable is the set of rules, standards, commercial structures and supply-chain capabilities that would allow someone to build and operate such ships later.

The four named domains — regulatory, industrial, commercial and financing — are effectively a map of the obstacles. Each maps onto a distinct set of institutions and each has a different clock speed.

The ordering matters. Financing leads nowhere until regulatory and commercial structures exist, which is why the MoC's breadth is a signal of seriousness rather than of vagueness.

Why MARAD is the right counterparty

MARAD is not a nuclear regulator and does not license reactors. Its leverage is elsewhere, and it is substantial:

1. Title XI federal ship financing guarantees. MARAD administers the programme that backstops US shipyard construction. Its eligibility rules, appraisal methodology and credit criteria determine whether nuclear-powered vessels can be financed on terms comparable to conventional tonnage.
2. Flag-state administration. The US-flagged requirement in the announcement is not incidental. A US flag brings US Coast Guard inspection, US manning rules and access to US cabotage trades under the Jones Act — a market where fuel cost and emissions compliance weigh heavily and where foreign-flag competition is excluded.
3. Federal convening power. MARAD can bring the Coast Guard, the Department of Energy, the Nuclear Regulatory Commission and the Department of Defense to the same table in a way no private developer can.
4. National security framing. The US Navy operates nuclear surface combatants and submarines; the industrial base, workforce and training pipelines overlap directly with what a commercial programme needs. This is the strongest political argument available in Washington.

The non-exclusive structure suggests MARAD is positioning itself as a neutral framework-builder rather than a project sponsor — creating the conditions under which multiple developer–shipyard–utility teams could eventually compete.

Core Power's position: first mover, not yet incumbent

Core Power's strategic bet is that the vessel architecture is the right unit of analysis, not the reactor. The company has consistently argued for modular, mass-manufacturable reactor plants integrated into conventional hull designs, and for floating nuclear power as the bridge application — a platform that generates revenue and builds regulatory precedent before propulsion applications are approved.

That sequencing is commercially intelligent for three reasons:

• A floating power plant is stationary, so it encounters a far simpler regulatory surface than a ship that calls at dozens of jurisdictions.
• It can be classed and insured under existing offshore conventions with adaptation, rather than requiring wholly new instruments.
• It forces the supply chain — forgings, pressure vessels, nuclear-grade welding, qualified inspectors — into existence ahead of propulsion demand.

The MoC's jurisdictions — United States, Japan, South Korea — track Core Power's existing relationships with Asian shipbuilding and heavy industry. Japan and South Korea are where the world's most capable commercial shipyards sit, and where the fabrication capacity for large nuclear-grade modules has historically been concentrated in the civilian supply chain.

The port-access problem is the hard one

Any US-flagged nuclear merchant ship that trades internationally must be accepted in foreign ports. This is where most nuclear shipping concepts die, and it is worth being precise about why.

Under the Convention on the Safety of Life at Sea (SOLAS), a nuclear-powered ship must hold a flag-state-issued Nuclear Ship Safety Certificate, and — critically — a port state's acceptance is not automatic. That certificate regime was written in the 1960s for a handful of experimental vessels. It presumes a bilateral or multilateral understanding that in practice does not exist for a commercial fleet.

Three credible mitigations are circulating, none of them yet settled:

1. Bilateral agreements. Ship-by-ship or route-by-route state consent, similar to the arrangements that governed the NS Savannah in the 1960s. Slow, political, but proven.
2. Bounded trades. Restrict early vessels to routes between consenting states or to US domestic cabotage where a US flag is already required and port-state questions collapse into flag-state ones. This is the closest thing to a low-risk commercialization path.
3. Refuelling at sea or at a single certified hub. Concentrates nuclear handling at one approved facility, reducing the number of jurisdictions that must develop nuclear port capability.

A fourth possibility — insurance — is the sleeper issue. Protection and indemnity (P&I) clubs have no standard product for a commercial nuclear merchant hull. Until the insurance market prices this risk, charterers will not sign, and without charters there is no revenue line for a lender to underwrite.

What this means for steel and the industrial supply chain

For MetalMate's core readership, the near-term information value of this announcement is close to zero on tonnage and high on qualification. Nothing in the MoC commits a single tonne of plate.

But the medium-term implications are worth thinking through now, because nuclear-grade demand is a different demand from conventional shipbuilding demand.

1. Reactor pressure vessels and primary circuit

The dominant candidate architectures for maritime use are light-water SMRs and molten-salt or liquid-metal-cooled concepts. Light-water designs require heavy-section reactor pressure vessel forgings in low-alloy steel grades (SA-508 Cl.3, SA-533, similar) with very tight chemistry and toughness control, plus extensive cladding with stainless or nickel-based alloys. There are a very small number of global suppliers capable of forging these at scale. If maritime nuclear demand arrives, it arrives into a supply chain that is already constrained by naval and land-based SMR programmes.

This is the single most likely pinch point, and it is a forging constraint, not a plate constraint.

2. Containment and safety-class structural steel

A commercial nuclear vessel needs a containment boundary, shielding structures and safety-class supports. These are typically stainless, carbon-manganese or specialty grades with full traceability, qualified welding procedures and third-party inspection. Volumes are modest — thousands of tonnes, not hundreds of thousands — but the cost is in qualification, not material.

For a commercial shipyard, this is a business-model question as much as a metallurgical one. Nuclear-grade fabrication requires a documented quality management system, weld qualification records, and a permanent labour force that can hold those qualifications. That is fundamentally incompatible with the project-based staffing model most commercial yards use.

3. Conventional hull structures: the volume play, deferred

If a nuclear merchant fleet ever scales, the conventional hull — the majority of the steel — would be built to standard classification rules in standard grades (AH36, EH36, and similar). The nuclear part is a small, expensive, highly controlled island within a normal ship.

That distinction matters for how a yard would organize itself: a nuclear-capable shipyard is really a conventional yard plus a nuclear module hall plus a quality system. The capital question is whether the volume justifies the hall.

4. Regulatory-driven material change is the realistic near-term driver

Separate from nuclear propulsion, the shipping industry faces tightening carbon intensity requirements. Nuclear propulsion is one of a small number of options that eliminates rather than reduces voyage emissions. If carbon pricing bites hard enough, the economics flip from "nuclear is too expensive" to "nuclear is the only compliant option for certain routes" — and that, not a technology breakthrough, is what would move steel demand.

Reading the timeline sceptically

The gap between an MoC and a delivered US-flagged nuclear merchant ship is long and passes through several gates. A reasonable, explicitly analytical sequencing — not a forecast from any participant — looks like this:

Core Power's own public statements have suggested first commercial deployment could target roughly the early-to-mid 2030s. That is consistent with the table above, but it assumes the regulatory and port-access problems are solved in parallel rather than sequentially. Historically, they have been sequential — which is a polite way of saying slower.

What to watch next

Five observable signals would materially change the assessment:

1. A funded instrument. Appearance of Title XI eligibility guidance, DOE loan authority, or EXIM/DFC engagement specific to nuclear vessel construction. Money converts a memorandum into a programme.
2. A named reactor vendor and class society. Without a selected reactor technology and a classification society writing rules, "industrial framework" remains abstract.
3. A named shipyard. Hull construction is where steel demand becomes real. Until a yard is identified, no material forecast is defensible.
4. Port-state announcements. Any bilateral arrangement permitting a nuclear merchant ship to call at a foreign port is a bigger milestone than any shipyard contract.
5. Insurance capacity. A P&I club or syndicate announcing a nuclear merchant product would be the clearest evidence that the commercial framework is closing.

Assessment

Judged as a shipbuilding project, the MARAD–Core Power MoC is pre-announcement. There is no procurement, no design freeze, no contract, no tonnage. Companies looking for order-book signals should look elsewhere.

Judged as an institutional project, it is more consequential than it appears. Nuclear merchant shipping has failed to launch repeatedly — Savannah, Otto Hahn, Mutsu — for reasons that were never primarily technical. Each failed on economics, public acceptance and port access. The MoC explicitly targets the first and third of those, and MARAD is one of the few institutions that can convene the rest.

The realistic read: this is the opening move in a framework-creation process that will take most of a decade to produce a first hull, and will require several more announcements — vendor, yard, class society, insurer, lender — before any of it becomes steel. The task for industry is not to endorse or dismiss it, but to track which of those five signals appears, and in what order.

Sources

See structured source list.