The idea of extracting energy from ocean waves and turning it into electricity is an alluring one. The first serious attempt to do so dates back to 1974, when Stephen Salter of Edinburgh University came up with the idea of “ducks”: house-sized buoys tethered to the sea floor that would convert the swell into rotational motion to drive generators. It failed, as have many subsequent efforts to perform the trick. But the idea of wave power will not go away, and the latest attempt—the brainchild of researchers at Oscilla Power, a firm based in Seattle—is trying to address head-on the reason why previous efforts have foundered.

This reason, according to Rahul Shendure, the firm’s boss, is that those efforts took technologies developed for landlubbers (often as components of wind turbines) and tried to modify them for marine use. The consequence was kit too complicated and sensitive for the rough-and-tumble of life on the ocean waves, and also too vulnerable to corrosion. Better, he reckons, to start from scratch.

Instead of generators with lots of moving parts, Oscilla is developing ones that barely move at all. These employ a little-explored phenomenon called magnetostriction, in which ferromagnetic materials (things like iron, that can be magnetised strongly) change their shape slightly in the presence of a magnetic field. Like many physical processes, this also works in reverse. Apply stresses or strains to such a material and its magnetic characteristics alter. Do this in the presence of permanent magnets and a coil of wire, such as are found in conventional generators, and it will generate electricity.

The core of Oscilla’s design is a bar made from an alloy of iron and aluminium, a mixture that is strongly ferromagnetic. Such bars need be compressed by only one part in 10,000 to have the desired effect. This means, to all intents and purposes, that the generator has no internal moving parts that can go wrong. But compressing a solid metal bar by even this tiny amount requires the application of a huge force. Fortunately, ocean waves are powerful enough to generate this force. Oscilla’s design, as the firm’s name suggests, does it by oscillation.

Its oscillating generators consist of two large objects connected by cables (see diagram). At one end of these cables, floating on the surface, is a buoy that contains the generating apparatus of alloy bars, magnets and coils, together with sets of hydraulic rams which can squeeze the bars as desired. At the cables’ other ends hangs a structure called a heave plate, which is kept stationary by a combination of inertia and the drag of the surrounding water. This arrangement means that, as the buoy rises and falls with the waves at the surface while the heave plate stays more or less put, the tension on the cables increases and decreases. That changing tension drives the rams. The whole system is kept in place by a second set of cables that moor it to the seabed.

A full-scale device, which Oscilla hopes to build by 2018, will be a foam-filled steel buoy 27 metres in diameter, six metres high and weighing 1,000 tonnes, tethered to a toroidal concrete heave plate 70 metres below the surface. It will carry 12 magnetostrictive generators within. Mr Shendure says that a single such buoy, placed a few kilometres offshore, should deliver an average of 600 kilowatts—about the same as an onshore wind turbine. A prototype four metres in diameter underwent a brief but successful open-ocean trial off the Atlantic coast of America last year.

Oscilla’s generators will, Dr Shendure acknowledges, be expensive to build and install. But their simple design, he says, should allow them to operate for decades with no more maintenance than an occasional scrub to remove accumulated barnacles. He calculates that the cost of producing electricity from them will be around ten cents a kilowatt hour. That compares with 16 cents a kilowatt hour for offshore wind farms and six cents for the onshore variety. A grid-connected fossil-fuel power station would be cheaper still—five cents or less. But ten cents represents a decent start for such a novel way of generating electricity.