Where to sell a wave energy device first
Ninety-nine coastlines screened against three years of global wave data, filtered down to ten places where the power is good enough, steady enough, and expensive enough to replace that someone would actually sign. This is the target list and the reason behind each one.
Ten places to call, in order
Ranked on what the power is worth to the buyer, not on how big the waves are. Each row carries the resource, what that power would replace, who has the authority to sign for it, and the opening move. The reasoning behind the ranking is in the next section.
Read the status column as pipeline temperature. Precedent means this buyer has already signed a marine energy contract. Policy pull means a published government target creates the demand. Procurement means a defense budget line and no utility regulator to satisfy. Reference means it earns a number rather than revenue. Resource figures are computed. The displaced-cost column is the weakest input and the one to challenge first: tariffs are cited where a public figure exists, and remote military generation is a band, not a quote. Nothing here survives without a real load curve and a delivered-cost model.
Conspicuously absent, and worth saying out loud. Tasmania, west Ireland, the Azores, Orkney and Humboldt all rank top ten on raw resource and none of them make this list. Tasmania and Ireland sell into well supplied grids. Iceland runs on geothermal and hydro. Mainland Norway has the single best wave climate in the screen and pays some of the lowest power prices in Europe. Mauritius and Réunion looked strong on both resource and steadiness until the tariff check: Mauritius sells business power at about 13.6 US cents, so the displaced dollars are not there. Good waves in a cheap-power market is not a business.
How a target got onto that list
Three filters, in order. Most wave-energy site lists stop at the first one, which is why they are full of places nobody will ever build.
Enough energy, measured the way a machine feels it
Wave power is not wave height. It scales with the square of height and with the wave period, so long ocean swell carries many times the energy of choppy local sea at the same height. The device modelled here sits below the surface and responds to the pressure difference a passing wave creates, which means short choppy seas are close to invisible to it. Every number on this page is computed that way, every three hours, for three full years.
Steady enough to be worth connecting
A site that averages well but arrives in winter bursts needs storage or a diesel running behind it, and that kills the economics. So the filter is firm power, the output a site beats 90% of the time, not the average. This is where most famous wave sites fall over.
Replacing something expensive, for a buyer who can sign
The same kilowatt-hour is worth four times more on a diesel island than on a hydro grid. Norway has the best waves in this screen and some of the worst economics for selling into. So the last filter is what the power displaces and whether there is a counterparty with a budget and the authority to use it.
What filter two does to a leaderboard
West Ireland is one of the most energetic coastlines on earth. Central Chile is not. On average output Ireland wins comfortably. On the power you can actually count on, it is not close, and the storage bill separating them is the whole difference between a project and a science experiment.
Three ways to look at the same ocean
Every layer is corrected for submergence. A device sitting 15 m down only feels the fraction of surface wave pressure that reaches it, which decays as e−2πz/L. Short wind sea is close to invisible at depth, so enclosed basins fall away under this correction even where the raw flux looks respectable.
What happens to the leaderboard when you ask for firm power
Left: the twelve strongest sites by mean effective flux. Right: where each lands once ranked on the flux it beats 90% of the time. The North Atlantic empties out. Nothing about the physics changed, only the question.
What the full screen carries
This page shows three of the ten targets and one of the five charts. The screen behind it is considerably larger, and it is the part worth paying for.
The other seven targets, named, with the opening move for each
Who signs, what the power replaces, what to say first, and the objection to expect. Two of the seven are places nobody in this industry is currently working.
A defense screen over 37 NATO and US locations
Pacific and high north, sampled windward. It contradicts the obvious answer: the two installations everyone names first are among the weakest in the set, and the strongest is a territory nobody mentions.
The Arctic, where sea ice rather than wave height decides
Solar is zero for four months and demand peaks in the same months. One side of the Arctic doubles its output over exactly that window. The other freezes precisely when it is needed, and its published averages hide it.
The numbers a project actually needs
Firm power, seasonal swing, worst-case lull in hours, design sea, and a month-by-month delivery profile for every site. This is what sizes storage and backup, and it is what a resource atlas will not give you.
The method, its cross-checks, and where it is weakest
Cross-checked against published assessments at three independent sites, with the two data traps that would have corrupted the result written up, and the sensitivity tests that show which conclusions survive changing the assumptions. Including the ones that flatter us least.
Built in a day, from public data
Ninety-nine coastlines, three years of three-hourly hindcast, eight and a half thousand timesteps per site, pulled through Earth Engine and rendered here. If you have a market question shaped like this one, that is the turnaround.