Wave energy developers are positioning ocean power as the missing link for offshore data centers, arguing that co-locating generation with subsea compute infrastructure solves the sector’s toughest constraint: delivering reliable, high-density electricity far from shore without costly cable runs. Mocean Energy’s Cameron McNatt contends that wave converters integrated directly with submerged data modules could bypass the interconnection queues and transmission bottlenecks stalling land-based AI clusters, turning an emerging marine energy niche into a critical enabler for the next wave of compute deployment.
Why offshore data centers need a new power paradigm
Microsoft’s Project Natick proved the technical viability of submerged data centers in 2018, demonstrating lower failure rates and efficient cooling using seawater heat exchange. Yet commercial deployment has stalled because powering these pods remains expensive: running high-voltage cables from shore adds millions in capital cost per kilometer, introduces single-point-of-failure risk, and subjects operators to onshore grid interconnection queues that now stretch three to five years in major markets. Meanwhile, AI workloads are driving rack densities from 10-15 kW to 50-100 kW and beyond, demanding power profiles that dwarf what early subsea prototypes required. A single modern AI training cluster can draw 50-100 MW continuously – roughly the output of a small gas plant – making diesel generators or battery-only solutions impractical for anything beyond backup.
Wave energy enters this gap with a distinct spatial advantage: the resource is strongest where offshore data centers want to sit – on continental shelves in 50-100 meter depths, near population centers but outside shipping lanes. Unlike floating wind, which requires deep water and massive mooring systems, wave converters such as Mocean’s hinged-raft Blue Star or CorPower’s point absorbers can be arrayed in the same footprint as the data modules, sharing mooring infrastructure and subsea cabling. The power profile also aligns: North Atlantic wave climates deliver capacity factors of 30-45%, higher than offshore wind in many regions and far more predictable day-ahead, reducing the storage overbuild needed to firm the supply.
Technology readiness and the economics of co-location
Mocean’s Blue Star 10 kW prototype completed a year-long sea trial at the European Marine Energy Centre in Orkney through 2023, surviving 14-meter significant wave heights and demonstrating survivability – the historic Achilles’ heel of wave energy. The company targets a 250 kW commercial unit by 2026 and a 1 MW “Blue Horizon” model by 2028. CorPower, AW-Energy, and Eco Wave Power are on similar timelines, with pilot arrays planned in Portugal, Spain, and Israel. None have yet reached utility-scale series production, but the convergence of survivability data, supply-chain standardization (marine-grade composites, direct-drive PTOs), and insurance willingness signals the sector is exiting the “valley of death” that stranded earlier developers like Pelamis and Aquamarine Power.
My analysis: if a 1 MW wave unit achieves a installed cost of $4-5 million – Mocean’s stated target – the levelized cost of electricity (LCOE) lands around $120-150/MWh at 35% capacity factor, before any storage. That is above current offshore wind ($50-80/MWh) but competitive with diesel generation ($200-300/MWh) and avoids the $2-4 million/km subsea cable cost to shore. For a 50 MW data center requiring 400 GWh/year, a 150 MW wave array (accounting for capacity factor and storage losses) would need roughly 150 units at 1 MW each, implying $600-750 million in generation capex. The avoided cable cost alone – 50 km at $3 million/km – saves $150 million, closing much of the gap. The economics improve further if the data center’s waste heat (35-40°C seawater return) drives a bottoming cycle or thermal desalination, creating a revenue stack that pure generation cannot match.
Grid integration and the “energy island” model
The most transformative implication is architectural: wave-powered offshore data centers need not connect to the onshore grid at all. They can operate as self-contained “energy islands,” using the data center’s compute load as the primary offtaker and any surplus to produce hydrogen, desalinated water, or charge batteries for firming. This sidesteps the interconnection queue entirely – a structural advantage as FERC Order 2023 reforms only marginally accelerate cluster study timelines. For hyperscalers (Microsoft, Google, AWS) facing 2030 carbon-free energy matching goals, an offshore facility that generates its own renewable electrons 24/7 with minimal curtailment offers a cleaner accounting path than virtual PPAs on congested terrestrial grids.
Cross-sector connection: the same logic applies to green hydrogen projects targeting offshore wind. The European Union’s REPowerEU plan assumes 10 million tonnes of renewable hydrogen imports by 2030, much from North Sea electrolysis islands. Wave energy’s higher capacity factor and lower variability reduce electrolyzer downtime, improving hydrogen LCOE by an estimated 10-15% versus wind-only configurations (my estimate based on NREL H2A modeling parameters). If wave arrays share substations and export cables with wind farms – as envisioned in the UK’s Celtic Sea and Scotland’s Innovation and Targeted Oil & Gas (INTOG) leasing rounds – the marginal cost of adding wave drops further. Data centers and hydrogen plants thus become anchor tenants de-risking the same marine energy infrastructure.
Who this affects
- Hyperscale cloud operators: Evaluate wave-powered offshore sites as a parallel track to land-based campus expansion, especially for inference workloads tolerant of 10-20 ms latency to coastal metros; the avoided interconnection queue may justify a 20-30% power cost premium.
- Marine energy developers: Shift business model from selling kWh to the grid toward selling “compute-ready power” packages – standardized 10-50 MW blocks with integrated storage, subsea switchgear, and thermal interfaces – to de-risk project finance.
- Transmission planners and ISOs: Model offshore data centers as non-wires alternatives that reduce onshore load growth forecasts; a 100 MW subsea cluster powered by wave/wind hybrid avoids equivalent generation and transmission upgrades onshore.
- Infrastructure investors: Treat wave+data center co-location as a new asset class blending digital infrastructure yields (8-10% unlevered) with renewable energy tax equity structures; early movers can secure offtake agreements with creditworthy hyperscalers before utility PPAs materialize.
What to watch next
- Mocean’s 250 kW Blue Star deployment at EMEC in 2025-26: survivability data in winter North Sea conditions will determine insurability and bankability for 1 MW scale-up.
- First commercial offtake agreement between a wave developer and a hyperscaler or colocation provider (e.g., Equinix, Digital Realty) for a pilot offshore data module – likely 5-10 MW scale, 2027-28 timeframe.
- UK Contracts for Difference (CfD) Allocation Round 7 (2025) and subsequent rounds: whether wave energy secures a dedicated pot or competes in the “established technologies” pot against offshore wind, setting the revenue floor for project finance.
- Subsea cable standardization: adoption of 12 kV/33 kV DC collection systems for marine energy arrays, reducing converter station cost and enabling direct connection to data center DC buses (380 VDC or 400 VDC rack-level).
Bottom line: Wave energy’s moment is not about competing with offshore wind on LCOE – it is about solving the spatial and temporal mismatch between where AI compute wants to go and where the grid can serve it. The first 50 MW wave-powered data center will prove the model; the tenth will make it a standard option in hyperscaler site selection.
Read the full report at Renewable Energy World
Note: facts and figures attributed above to Renewable reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply