Two kilometres off the Kona coast on an August morning in 1979, a converted US Navy barge with a plastic pipe dangling 670 metres into the dark made ten kilowatts of electricity out of nothing but the temperature of the water.
No fuel, no combustion. Warm water on one side, cold water on the other, and a gap of a bit over twenty degrees between them doing the entire job.
How a 20 degree gap turns a turbine
Every heat engine needs a hot side and a cold side, and the tropical ocean stocks both at the same address. Surface water off Hawaii sits in the mid twenties year round. Drop a pipe to 670 metres and you are into water in the single digits.
By power station standards that gap is pitiful. A coal boiler works with hundreds of degrees. The tropics are the only place the ocean offers even this much, which is why ocean thermal power has never found a market outside the belt around the equator. But thermodynamics does not ask for a big difference, only a real one, and the barge had exactly that.
The engineering trick is choosing a fluid that boils inside such a narrow window. Ammonia does. Warm seawater runs through a heat exchanger and boils the ammonia; the vapour spins a turbine; cold seawater condenses it back to liquid; a pump sends it around again; and the same ammonia lives in the loop forever. Mark Betancourt, writing in Eos, put the conversion rate of this arrangement at roughly two to three per cent of the energy in the seawater, which sounds damning until you remember the fuel arrives by itself and never runs out.
The pumps eat most of it
Everything about ocean thermal power comes down to one subtraction: gross output minus the electricity needed to shift all that seawater. Makai Ocean Engineering, the Hawaiian firm that engineered several parts of the barge, records that the plant produced 50 kilowatts that morning and consumed 40 of them, leaving ten. Across the whole program it averaged about 50 kilowatts gross and 15 net, at an average seawater temperature difference of 21°C, per the operational results published in the Journal of Solar Energy Engineering. Cold water came up through a 0.61 metre polyethylene pipe that doubled as part of the mooring system.
Fifteen kilowatts is a couple of household kettles. It was also the first time anyone had pulled a positive number out of an ocean thermal plant at sea.
Georges Claude, the neon millionaire who tried it first, never managed that. He built a plant in Matanzas Bay, Cuba, in 1930, sank a fortune into cold water pipes that kept failing, and eventually admitted the machine was swallowing about four times the power it produced. Betancourt reports that a hurricane finished off the one pipe that worked.
Ten days, then the program ended
The 1979 barge, known as Mini-OTEC, was a demonstration and nothing more. The US Department of Energy’s marine energy projects database records first power on 2 August, shutdown on 18 November, and a longest unbroken run of ten days. Japanese engineers followed with a similar unit on Nauru in 1981 and 1982 that reached 35 kilowatts net, as Britannica documents.
Then oil prices settled, and the funding went with them.
Still waiting on a pipe
Forty-seven years on, the largest grid-connected ocean thermal plant in the world is a few kilometres up the same coast, and it is still a test rig. Makai’s 105 kilowatt unit at the Natural Energy Laboratory of Hawaii Authority went live in August 2015, and the federal listing for the device says it plainly: the plant is not commercial and does not run on a continuous basis.
Japan has the other one. A 100 kilowatt demonstration unit has been running on Kume Island in Okinawa since 2013, and the ambition there is a full megawatt. That requires a 1.5 metre intake pipe, which the island expects to cost between US$60 and US$80 million, a figure Betancourt reported when the switch-on date was still being measured in years.
The schedule has slipped accordingly. Mitsui O.S.K. Lines, the shipping company backing the project, spent years pointing at “around 2026” for the world’s first commercial megawatt-scale plant. Its October 2025 partnership agreement with Kumejima Town now aims at operations inside the 2031 financial year, and the job listed first is designing the intake.
Nobody in the field is still arguing about the thermodynamics. That argument was settled decades ago. What’s left is the plumbing, and for ninety-six years the plumbing has been winning.
