The Port as a Power Station: Why Data-Centre Investors Should Reconsider the Power Barge

The latest signal from the US data-centre market came not from a grid operator, but from a shipbuilder.

HD Hyundai Heavy Industries has reportedly agreed a $674 million contract to supply 1GW of medium-speed gas-engine generation to Corban Energy Group for AI data centres. The modular engines are intended to provide continuous prime power, not simply emergency back-up.

It is another sign that large data-centre developers are confronting an uncomfortable reality: if firm grid power cannot be delivered on their timetable, they must develop an alternative power system.

For some UK sites, particularly those near the Thames and other major ports, that system may not need to sit entirely on land.

A forgotten infrastructure option

The power barge is hardly a new idea. Floating power plants — ranging from converted vessels to purpose-built powerships — have long been used where power must be deployed quickly, fuel can arrive by sea and conventional network infrastructure is inadequate or delayed.

What has changed is the customer.

AI and hyperscale data centres now need power at a scale once associated with major industrial developments or utility power stations. A port-located campus may possess an advantage unavailable to most inland sites: access to deep-water or commercial berths, established marine logistics, industrial land, and potentially a route to receive a substantial floating generation asset.

The relevant concept is not necessarily a data centre built on water. It is a land-based, port-adjacent data-centre campus supplied through a dedicated shore connection by a permanently moored power barge or powership.

This is no longer a purely theoretical model. Mitsui O.S.K. Lines and Kinetics have announced an integrated floating data-centre concept that contemplates dedicated power from flexible sources, including powerships, alongside land-based grid and renewable-power options. The model may emerge first in markets with acute power shortages, but its underlying logic is increasingly relevant to UK digital infrastructure.

More than emergency power

A well-structured barge solution can serve three distinct roles.

1. First, it can provide bridge power while a constrained grid connection or network reinforcement is progressed. This may allow a campus to energise early phases years ahead of the ultimate grid timetable.

2. Second, it can provide phased capacity. Generation can be added in modules as individual data halls are commissioned, reducing the need to invest immediately in the full ultimate power requirement.

3. Third, it can become a longer-term firm-power asset within a broader campus energy system: working alongside grid import, battery storage, demand flexibility and future low-carbon power sources.

Its strategic attraction lies in optionality. Unlike a conventional land-based plant, a floating asset can potentially be repowered, converted to lower-carbon fuels, reconfigured or relocated as a project’s needs, fuel economics and grid position evolve.

That does not make it simple. It makes it worth evaluating.

The fuel question cannot be avoided

In the near term, a large floating power asset is most likely to rely on gas engines or gas turbines, supplied by pipeline gas where available or by LNG delivered through the port. The Hyundai–Corban announcement is important precisely because it demonstrates that modular gas-engine prime power is now being assembled at 1GW scale for data-centre use.

For UK projects, an investor case based on gas must be honest about the implications.

A credible scheme needs to address fuel supply and storage, marine bunkering or pipeline interface, emissions control, local air quality, noise, carbon cost, environmental permitting, operating flexibility and the long-term decarbonisation pathway. These are not technical footnotes to be resolved after the site is secured; they are core determinants of consentability, financeability and social licence.

The answer will vary by location. Some sites may support a transitional gas-fired solution that is repowered over time. Others may be better suited to lower-carbon fuels, imported e-fuels, hydrogen derivatives, battery storage or a hybrid system with a growing proportion of grid and renewable supply. The investment case should test those pathways early, not simply assume that a future fuel conversion will be available.

A port does not remove development risk

A berth is not a development consent.

A floating generation asset still requires a robust assessment of port access, berth availability, dredging and mooring requirements, navigation, marine licensing, electrical export infrastructure, private-wire route, protection systems, land rights, planning status and port-operational compatibility.

Environmental regulation must also be considered from the outset. Existing UK regulatory cases involving barge-based generation show that noise, odour and other environmental impacts require formal evaluation — a useful reminder that floating does not mean invisible.

Nor should developers assume that port-related permitted-development rights automatically extend to a substantial power-generation project serving an external data-centre

The point is not to bypass the planning or electricity systems. It is to bring another credible asset class into the development toolbox before an investor commits to a site whose only power plan is an uncertain grid date.

From power barge to power strategy

Earlier in my career at Enron — then a major investor in floating power assets — I worked on a power-barge concept intended to provide contingency power for the City of London around the Y2K transition. That experience covered the realities behind the concept: design, procurement, shipyard construction and the practicalities of securing a UK berth.

Today, the driver is different. It is not millennium contingency planning; it is the need to deliver firm power to rapidly growing digital infrastructure in a grid-constrained market.

But the development disciplines remain recognisable. A successful project requires a credible site and berth, engineering that can be built and operated, secure fuel logistics, a consenting strategy, bankable contracts, clear allocation of risks, and a plan for the asset’s evolution over its life.

For port-located data centres, the power barge should not be viewed as a novelty or a last resort. It should be tested alongside transmission, distribution, on-site generation, BESS and private-wire alternatives as part of an integrated power strategy.

CM Energy Insight helps data-centre developers and investors assess, structure and deliver utility-scale power solutions — including port and marine opportunities, grid strategy, embedded generation, fuel supply chains, permitting, commercial structuring and delivery risk — before grid uncertainty becomes a stranded-site risk.