A rotor 984 feet across leaves just 79 feet of air between its lowest blade tip and the wave crests, and the hub carrying it 19 miles off Fujian stands 571 feet above the water

Everyone quotes the rotor. It is the number that fits in a headline.

The number that actually decides the design is a much smaller one, measured downward.

When the lowest blade passes the bottom of its circle, something has to be underneath it.

The same stretch already holds the first 16 MW machines ever built, which makes this water the place where two records now sit within sight of each other.

That something is open water, moving.

Get the gap wrong and the machine hits a wave once in its life, which is once too often.

So what sets the gap?

The clearance is measured from the surface, not the seabed

Hub height on an offshore turbine is quoted from mean sea level upward.

Subtract the rotor radius from it and you have the air gap: the space between the lowest point of the swept circle and the water. Here that is 571 feet minus 492, which leaves about 79 feet.

What has to fit inside 79 feet is a list of things that all happen at the surface.

The highest wave crest the site can produce, the tide at its top, storm surge pushing the whole sea level up, and a margin on top of all of it.

Depth is not on that list.

Water depth decides how long the foundation is, how the piles go in and which vessel can install the thing, which is a large expensive problem and a different one.

If anything the relationship runs backward. Shallow water limits how tall a wave can get before it breaks, so a shallow site needs less clearance rather than more.

The machine standing in that water

The rotor measures 984 feet across, swinging three blades of 482 feet each.

The swept circle covers roughly ten football fields, and the hub sits 571 feet up, which the developer likens to a 58 story tower.

Rated output is 20 MW from a single machine.

The developer also claims the structure comes in about 20 percent lighter per megawatt than the industry average, which is the figure that matters more than the diameter, because weight per megawatt is what decides whether the next size up is affordable.

It stands more than 19 miles off southern Fujian in over 131 feet of water.

The same stretch already holds the first 16 MW machines ever built, which makes this water the place where two records now sit within sight of each other.

The dates and the numbers behind them

Installation finished in January of 2026 and the turbine synchronized to the grid on February 5.

Lifting it took a self elevating platform with a crane rated near 2,200 tons, holding position on legs driven into the seabed rather than floating, because each blade weighs more than most things ever lifted that high.

The gaps between monsoon swells were measured in hours, and each blade had to be threaded into its root separately.

Expected annual output is over 80 GWh, put at the yearly consumption of about 44,000 households.

The coal displacement is given as roughly 24,000 tons a year, calculated against the carbon intensity of the grid this machine feeds.

It is built and operated by a Chinese state hydropower group with a domestic turbine maker, and it can sense gusts and adjust pitch ahead of them.

What a single turbine cannot show

That 80 GWh implies a capacity factor around 46 percent, which is excellent and is a projection. No operating year exists yet.

The coal figure carries the same caveat, since it depends entirely on which plant sits at the margin when this machine is running.

Typhoon resistance is described as a baseline design condition rather than a demonstrated record, and the difference between those two matters enormously.

One turbine is also not a wind farm. Array cabling, foundations, substations and export capacity at this rating are all still being worked out, and the maintenance question is wide open, because a longer blade means a faster tip and rain erosion is already a live cost on machines half this size.

Bigger machines cut the number of foundations and vessel trips per gigawatt, which is the real argument, and that arithmetic only closes at scale.

The record was set to be broken

The step from 16 MW to 20 MW inside one hardware generation is not incremental.

Western manufacturers spent several years moving from 12 MW to the 14 and 15 MW units now going up off Virginia and Denmark, a pace that reflects supply chain depth as much as engineering ambition.

Meanwhile rivals in the same domestic industry are working on machines in the 18 to 26 MW range, as the grid connection report notes.

So the 20 MW mark is unlikely to stand long, as the operator’s account makes clear it does not expect it to.

What has been proven is narrower and more useful. A rotor wider than three football fields can be assembled in open water, and its blade tip can be held clear of the sea on a fixed foundation.

The question now is whether 44,000 households per machine becomes the floor.