Research thesis · est. 2026 · not a company pitch Bristol Channel · Pacific Northwest · Iberian Atlantic

The ocean holds up to 4× humanity's annual electricity demand. The installed base is still 490 MW.

SeaPower is building the engineering, siting and economic case for coastal data centers powered, cooled and stabilized by wave energy. The cold, quiet edge of the ocean is also — by accident — the best place left to put a data center.

Resource ceiling
45,000–130,000
TWh / year (IRENA)
Installed marine energy
490 MW
end of 2025 (IRENA)
What we are

A project-development thesis — not a wave-energy OEM.

SeaPower is
  • A site-screening, engineering, and economic platform for coastal, grid-backed AI data centers.
  • A coordination layer between WEC OEMs, utilities, coastal authorities, fisheries, and compute tenants.
  • An open research thesis — assumptions, models, and environmental data published as we go.
SeaPower is not
  • A wave-energy converter manufacturer.
  • A floating, autonomous offshore data-center platform.
  • A company pitch — we are pre-pilot, currently seeking partners.

Current status · research preview Seeking site-development partners, WEC OEMs, anchor compute tenants, grid & fiber partners, and environmental advisors. First pilot target window: 2027–2028 single-unit POC; Pelagos-1 60 MW campus targeted 2028–2030.

02 The resource

The bottleneck has never been resource. It is survivability, mooring, and the cost of being at sea.

130,000TWh/yr ↳ vs. 30,000 TWh/yr global electricity demand

Wave energy density is concentrated near the surface and decays rapidly with depth. A 1-km stretch of west-facing North Atlantic coastline can carry 40–70 kW per metre of wavefront — comparable to a small wind farm, available on schedules wind cannot match.

Every wave-energy programme since the 1970s has died on the same three problems: a converter that can absorb a 1-metre swell and survive a 15-metre storm; mooring and power-takeoff that hold up for 25 years at sea; and a cost of compute-on-shore that justifies bringing power back.

FIG.01 Annual resource scale vs. installed marine-energy nameplate
Theoretical resource
130,000 TWh
Practical resource (IEA-OES)
≈ 45,000 TWh
Global electricity demand
≈ 30,000 TWh
Offshore wind potential
≈ 8,000 TWh
All marine energy installed (2025)
490 MW
EU/UK wave capacity (2024)
≈ 14 MW

The rust rows are nameplate capacity, not annual generation. At continuous output, 490 MW would equal about 4.3 TWh/yr — less than 0.01% of the low-end practical resource.

03 The wave-energy paradox

A wave converter has to be sensitive enough to absorb a 1-metre swell and strong enough to survive a 15-metre storm — three orders of magnitude in between.

FIG.02 Energy density vs. structural envelope

Three orders of magnitude.

The cost of a wave-energy converter is set not by the average sea state, but by the worst sea state it must survive. Modern designs use active detuning, submergence, feathering and predictive control — tuning the power-takeoff to each incoming wave forecast.

Source: DOE Marine Energy Cost Reduction Pathways, 2025.

Figure 02 — energy density vs. structural envelope Power capture curve rises with sea state; the design point and the survival envelope are separated by approximately one thousand times in energy density. 10⁶ J/m 10⁵ J/m 10⁴ J/m 10³ J/m 10² calm design sea state storm survival DESIGN POINT power optimised SURVIVAL ENVELOPE structure must hold ≈ 1,000×
04 Five mechanisms

Five families of wave-energy converter, each suited to a different sea state and shoreline.

01  /  05
Point absorber
A buoyant body heaves with the wave against a fixed reference — a seabed anchor, a deep submerged plate, or a long-period spar. The relative motion drives a hydraulic or linear-electric power-takeoff.
SiteOffshore, mid-depth
Maturity (TRL)7–8
Rated unit100 kW – 1 MW
Best forModular arrays
heave depth 30–80 m
air ↑ air ↓ Wells turbine · bidirectional flow
≥ 100 m length
surge hydraulic line to onshore plant
head site-specific civil works
01

Point absorber

A buoyant body heaves with the wave against a fixed reference: a seabed anchor, submerged plate, or spar.

Site
Offshore, mid-depth
Rated unit
100 kW - 1 MW
02

Oscillating water column

Waves compress and decompress an air column, driving a bidirectional turbine inside a chamber.

Site
Shore, breakwater
Rated unit
300 kW - 1.5 MW
03

Attenuator

A long segmented body flexes with the swell; power-takeoff sits in the joints between segments.

Site
Offshore, exposed
Rated unit
750 kW - 2 MW
04

Oscillating surge converter

A bottom-hinged flap rocks with nearshore surge and pumps hydraulic fluid to an onshore plant.

Site
Nearshore, 10-15 m
Rated unit
500 kW - 1 MW
05

Overtopping device

Waves run up a ramp into a raised reservoir, then drain through low-head turbines back to the sea.

Site
Shore or floating
Rated unit
0.5 - 5 MW
05 The data center concept

From wavefront to GPU rack, in seven hops.

A SeaPower facility is not a floating autonomous platform — that is the bet of operators like Panthalassa. Ours is the inverse: a coastal data center with the wave farm as primary generation, the deep ocean as heat sink, and the grid as backup. Tier-1 fiber under 5 ms, maintenance accessible by road, no satellite-only egress; satellite is a bolt-on resilience layer for telemetry, control-plane continuity, and disaster failover. We co-locate compute with energy to skip the most expensive part of the marine-energy stack: the export cable and the substation.
Local module bay

Run the core campus, then bolt on satellite resilience where the site needs it.

Core mode keeps production traffic on terrestrial fiber, with local controls, battery dispatch, and cooling loops able to run from the coastal campus.

SeaPower architecture — sea-to-rack schematic Seven-hop diagram from offshore wave array through subsea hub, cold-water intake, substation with battery storage, data hall, seawater cooling loop, grid tie, and an optional satellite telemetry layer. 1 — WAVE ARRAY 60 × 1 MW point absorbers · 4 shown 33 kV subsea export 2 — HUB 3 — COLD-WATER INTAKE SHORELINE 4 — SUBSTATION + BESS 33 → 11 kV 60 MWh battery 5 — DATA HALL ≈ 60 MW IT load · 18,000 GPUs 6 — seawater liquid-cooling loop · ΔT 8–12°C 7 — GRID TIE (firming + spill) ↻ REPLAY SCHEMATIC · NOT TO SCALE
01·02

Co-located generation

60-unit point-absorber array (60 MW rated) delivers ≈ 19 MW average through a single subsea hub. The data hall sits within 2 km of the lead converter, skipping the long export cable that dominates marine-energy capex.

03·06

Ocean as heat sink

Direct seawater cooling at 8–14 °C eliminates evaporative cooling towers, drops PUE below 1.10, and saves ≈ 1.4 ML / day of freshwater versus a comparable inland facility.

04

Buffered, not islanded

Onshore battery (60 MWh / 30 MW) absorbs minute-to-hour wave variance. The grid is the firming layer — it covers seasonal lows and absorbs winter swell maxima as exported surplus. Wave supplies ≈ 29 % of annual facility load; grid handles the rest.

05·07

Built like a substation, run like a DC

Modular 6 MW pods, liquid-to-chip cooling, two-tier security, latency < 4 ms to nearest tier-1 metro. Designed for AI training workloads where 24/7 uptime is shaped by SLA, not loss-of-grid.

A Fiber handoff

Primary data plane

Dual terrestrial routes carry workload traffic to the nearest tier-1 metro. This remains the latency and capacity path for AI training.

B Satcom edge

Out-of-band resilience

A local satellite terminal keeps site telemetry, dispatch instructions, and emergency command reachable when terrestrial backhaul is degraded.

C Remote ops cache

Local-first control

Forecasts, safety envelopes, and device set-points are cached on campus so the wave array can ride through network loss without manual intervention.

06 Reference design — Pelagos-1

Pelagos-1 — sized for a single AI-training tenant.

Pelagos-1 · 60 MW IT

Site (target)Cornwall · Wave Hub footprint
Wave resource42 kW / m
Mean significant wave height (Hs)2.4 m
Array (rated)60 × 1 MW point absorbers · 60 MW
Annual capacity factor0.31
Wave generation≈ 19 MW avg · 60 MW peak · 163 GWh / yr
Onshore battery60 MWh / 30 MW (≈ 2 h smoothing)
IT load60 MW
Facility load (PUE 1.08)64.8 MW · 568 GWh / yr
Cooling architectureDirect seawater · liquid-to-chip
PUE (annual)1.08
WUE (annual)0.02 L / kWh
Latency to London / Paris3.8 / 5.2 ms
Energy mix (annual)≈ 29 % wave · 71 % grid · BESS for shifting

Model v0.5 · illustrative Cornwall Hs from UKHO/PML buoy data (2018–2023). CF 0.31 follows NREL's published wave convention (Kilcher et al., NREL TP-5700-78773, 2021) and current 1 MW point-absorber operating data (DOE Marine Energy Cost Reduction Pathways, 2025). Energy mix assumes PUE 1.08 facility load with BESS roundtrip losses absorbed in the storage line. Site-specific bathymetry, fisheries consultation, and grid impedance studies pending. Full assumptions on request: hello@seapower.ai.

Why wave, not wind, for compute

Wave power lags wind by 6–18 hours and storms by 12–24 hours, so a coastal pairing smooths variance dramatically. Where wind alone gives 32 % capacity factor, wave + wind + battery reaches 71 % at our reference site.

The 60 MW step

60 MW is the sweet spot for a single hyperscaler training tenant: enough to host one frontier-class run end-to-end, small enough to permit and finance against a single coastal lease.

Built where data centers won't go

West-facing coasts have low industrial competition for land, abundant grid-injection capacity, and natural environmental envelopes (wind, salt, fog) that a marine-grade facility can absorb without exotic hardening.

07 Where we'd build

A short list of coastlines where wave resource, grid, latency and permitting align.

SeaPower priority sites — world map Six candidate coastlines: Cornwall UK, North Iberia, Oregon USA, Central Chile, Southwest Australia, and Wairarapa New Zealand. Loading map… FIG.03 — SeaPower priority sites · Natural Earth projection · screened against wave resource, grid HV access, latency, permitting
01

Cornwall, UK

Bristol Channel and Wave Hub footprint. Existing 33 kV grid takeoff, deep marine industrial heritage, ≤ 6 ms to London-Slough.

Hs 2.4 m · 42 kW/m · TRL 8
02

North Iberia

Asturias to Galicia. Bimep test site, EU Innovation Fund eligible, complementary to Spanish solar via winter peaking.

Hs 2.7 m · 50 kW/m · TRL 7
03

Oregon, USA

PacWave South leasehold (PMEC). Wave resource is 1.5× the state's annual electricity generation — Oregon could export wave power. Tier-1 fiber to Hillsboro-Portland.

Hs 2.9 m · 56 kW/m · TRL 7
04

Central Chile

Bío-Bío to Valdivia. Among the highest year-round wave power in the world; HVDC corridor under development.

Hs 3.1 m · 62 kW/m · TRL 6
05

SW Australia

Albany / Augusta. Carnegie test legacy, AEMO grid spare capacity, sub-sea cable to Singapore in planning.

Hs 2.6 m · 47 kW/m · TRL 7
06

Wairarapa, NZ

Cook Strait corridor. Combines tidal-stream and wave; Transpower 110 kV interconnect within 8 km of coast.

Hs 2.3 m · 38 kW/m · TRL 6
08 Economics

Wave power doesn't beat solar on cost. Co-located AI compute changes the question.

0
200
400
600
800
€55–95
Solar PVEU 2024
€60–110
Onshore windEU 2024
€80–170
Offshore windEU 2024
€110–480
Tidal streamEU 2025 BER
€160–750
Wave (today)EU 2025 BER
€95–195
Wave + DC offtakeSeaPower 2030E
Solar PV€55-95
EU 2024
Onshore wind€60-110
EU 2024
Offshore wind€80-170
EU 2024
Tidal stream€110-480
EU 2025 BER
Wave today€160-750
EU 2025 BER
Wave + DC offtake€95-195
SeaPower 2030E
FIG.04 Levelised cost of electricity · €/MWh

Three levers move wave from outlier to competitive.

1. Captive offtake. A co-located DC pays a power price benchmarked against grid + curtailment, not against wholesale.

2. Capex amortisation. Sharing substation, cooling intake, civil works and permitting between energy and compute halves balance-of-plant.

3. Cooling credit. Direct seawater cooling at PUE 1.08 vs inland PUE 1.40 yields ~ 30 % more IT energy per delivered facility MWh — equivalently, ~ 23 % less facility energy for the same IT load.

The industry has not received any subsidies in any shape or form in a similar way that the wind or solar industry have received in the early stage. Rémi Gruet · CEO, Ocean Energy Europe

Sources: IRENA Renewable Power Generation Costs 2024 · EU Blue Economy Report 2025 · OES Annual Report 2024 · NREL TP-5700-78773 (Kilcher et al. 2021) · SeaPower internal model.

09 Roadmap

Four phases. Modular by design, so each one funds the next.

2026 — 2027

Resource & site

  • GIS screening across 6 candidate coastlines
  • Two priority sites under exclusive lease
  • 1-year wave buoy + bathymetry campaigns
  • Reference DC tenant LOI signed
2027 — 2028

Single-unit pilot

  • 1 × 1 MW point absorber + onshore POC
  • 4 MW container DC bolted to substation
  • 12-month survival & availability proof
  • Environmental monitoring under OES protocols
2028 — 2030

Pelagos-1 · 60 MW

  • 60-unit array · subsea hub · 33 kV export
  • 60 MW IT data center, single tenant
  • 60 MWh BESS for minute-to-hour smoothing
  • First commercial offtake at < €120 / MWh
2030 +

Pelagos series

  • 3 × 60 MW campuses across Atlantic basins
  • Hybrid wave + offshore wind topologies
  • Open-source siting & survival framework
  • 1 GW order book by 2033