Theses›Energy›Offshore power islands

Offshore power islands

A sibling of space-based-datacentre in the 'escape the grid + land bottleneck' family, but far cheaper to reach: instead of a spaceport, deploy from a dry dock.

  • Bear
  • Early
  • $772M by 2035

The callFloating pumped hydro skips grid queues on paper, and every cost number is a model until a unit is fielded; audited efficiency and cost per MWh come first.

D1D2R1R2R3C1C2foragainst D1D2R1R2R3C1C2foragainst
Fig. 1The balance. Drivers and catalysts against restraints, each block sized by its weight; catalysts hang dashed because they have not happened yet, sized by their odds. It tips toward the restraints.
  1. D1Grids cannot connect renewables fast enough, with 5-7 year interconnection queues, and the world needs up to 8 TW of long-duration storage by 2040.
  2. D2Onshore pumped hydro needs mountains and decade-long permits and batteries cover hours not days, while a shipyard can build floating reservoirs in series.
  3. R1Every economic number is a model until the 2027 demo; nothing floating has run commercially.
  4. R2The storage category has no €1B independent outcome: Powin in Chapter 11, Wärtsilä's storage arm into a JV at a loss, Energy Vault from $1.07B to about $618M.
  5. R3Value may pool with shipyards or offtake counterparties, and if it prices as marine construction rather than energy equipment the outcome shrinks again.
  6. C1Terna's MACSE storage auction in Italy, 16 GWh for 2029 delivery, the first dated procurement offshore pumped hydro could bid into.Nov 2026
  7. C2Sizable's 1MW demo, the first fielded unit with a measurable round-trip efficiency and cost per MWh.2027
Rests onthe beliefs underneath
Supported by · confidence
Threatened by · confidence
What would prove itdated tests, open and resolved
20272028202920302031today123456predictionswatch 20272028202920302031today123456predictionswatch
Fig. 2Each ring is a dated test: dashed while open, struck through when overdue, ticked or crossed once resolved.
  1. 1Terna's MACSE auction: whether any offshore storage bids or wins.Dec 2026 · open
  2. 2Sizable Energy announces its demonstration unit deployed at sea between 6 Nov 2026 and 31 Jan 2027.Jan 2027 · open
  3. 3Sizable's 1MW demo and its measured efficiency, then the 10MW investment decision in 2028.Dec 2027 · open
  4. 4Sizable Energy does not deliver a costed, audited cost/MWh within 1.3x of grid+BESS from a fielded unit before end-2028; the 1MW demo (2027) and 10MW FID (2028) slip or the unit economics come in above that band.Dec 2028 · open
  5. 5The call that Sizable does not produce an audited cost per MWh within 1.3x of grid plus batteries from a fielded unit.Dec 2028 · open
  6. 6No offshore floating pumped-hydro or floating power+compute island reaches a >=100MW commercial deployment before end-2030; the category stays at demo / single-digit-MW pilot scale this decade.Dec 2030 · open
Expressionslisted, private, technologies
  • ENR.DESiemens Energy
    • cheap
     Sells the BlueVault lithium-ion storage system for drilling rigs, FPSOs and vessels, plus offshore grid and electrification equipment (spun out of Siemens in 2020).
  • ELI.BRElia Group
    • not screened
     Belgian transmission operator building Princess Elisabeth Island, the only offshore energy island under construction (cost up from EUR2.2B to ~EUR8B, completion targeted 2028).
  • 010140.KSSamsung Heavy Industries
    • not screened
     Shipbuilder developing a 50 MW floating datacentre with its own power generation (ABS and Lloyd's Register approval in principle), targeting first deployment by Q2 2028.
Private · 5 private companies, 2 of them in the US.

Sizable Energy · Ocean Grazer · FLASC · Sperra · Panthalassa

Technologies
Who has already bet on itprivate rounds, from public sources

$150M raised by 5 companies, $158M of it in the last three years. Wave building: Money is arriving: recent rounds and tier-1 funds are in, but no runaway leader yet.

Backed by a top-tier venture firm: Panthalassa, Sizable Energy.

What changedthe record, newest first
  • 1 Jul 2026The interesting bet under 'floating datacentres' isn't the datacentre at all — it's offshore floating pumped-hydro storage, with the compute as a premium offtake that lifts value-per-platform. It escapes the same grid + land bottleneck as space-based datacentres but deploys from a dry dock instead of a spaceport. Whether it works comes down to one unproven number, cost/MWh at scale, which isn't testable until the 2027-28 demos.