The machine is not the interesting part.
A big cutterhead turning slowly through soft ground is, mechanically, a solved problem.
The hard question is what happens in the few feet directly in front of it, in material that behaves less like ground and more like a very heavy liquid.
That material would like to come inside.
Above it sits a river, and above that, ships.
So the excavation is never actually open.
The face is held up by a liquid the crew makes
Saturated silt and fine sand have almost no cohesion. Cut into them and they flow, and what flows in is replaced by whatever was above it settling down.
So a machine in this ground does not dig a hole and then line it.
It keeps a sealed chamber pressed against the excavation and fills that chamber with a clay based slurry, held at a pressure that answers the ground.
The slurry does two jobs at once. It pushes back with roughly the force the earth and water are pushing forward, and it seeps a short way into the soil and forms a skin there, so the pressure acts on a surface rather than draining away.
The window it has to stay inside is narrow, and it is bounded on both sides by failure.
Too little pressure and the face runs in. Too much and the ground fractures, the slurry escapes upward and a crater opens on the riverbed.
Worse, it is not one number. Water pressure grows with depth, so across a face 54 feet tall the load at the bottom exceeds the load at the top by about 24 pounds per square inch, and the slurry has to answer a gradient rather than a value.
The machine sent in to do it
Excavation diameter comes to just under 55 feet, which puts the cutterhead about as wide as a five story building is tall.
The whole assembly runs 476 feet from the head to the back of the trailing gear and weighs close to 11 million pounds.
The cutterhead alone accounts for about a million pounds of that.
Installed power is 11,300 kilowatts, most of which goes into rotating the head and into pumping, because a slurry machine is as much a fluid handling plant as an excavator.
Spoil leaves as a suspension, gets separated at the surface, and the cleaned slurry goes back down the pipe.
Which means the tunnel is really a closed loop with a cutterhead on one end of it.
The drive, the depth and the dates
The machine started work on April 9 of 2025 at Haimen, on the north bank of the Yangtze estuary in Jiangsu.
Its single drive runs 5.8 miles, part of an underwater section of about 7 miles inside a whole route of roughly 24 miles.
Those three figures get quoted interchangeably, which is where most of the confusion around this project comes from.
The deepest point sits about 246 feet below the river, and the design case for water pressure at the face is 109 pounds per square inch.
The finished bore carries six lanes at a design speed of 62 miles an hour, linking Haimen District in Nantong with Taicang in Suzhou.
A second machine of the same size is driving the parallel tube, because at this diameter one tunnel holds traffic in one direction only.
Numbers worth checking before repeating
Several figures in circulation do not match the project’s own material.
Installed power is given as 11,300 kilowatts, not the higher figure that has appeared in some accounts, and the machine weighs about 11 million pounds rather than nearly twelve.
The record claim also gets rephrased in a way that changes it. What is claimed is the longest underwater highway tunnel, measured across the full route, which is not the same as any claim about diameter or about tunnels bored by machine.
Workforce numbers move between accounts too, and none of them carries an official source.
None of that undermines the engineering, which is documented in detail the way a major crossing or a new concrete method usually is.
Why not simply build a bridge
A bridge across an estuary this busy has to clear the shipping underneath it, and the ships here are large.
Clearance drives tower height, tower height drives span and span drives cost, and the whole structure then sits in the path of the weather.
A tunnel trades all of that for one very expensive machine and a face that has to be held up every hour of the drive, as the launch report sets out.
It also puts nothing in the river, which matters where the river is the reason the region exists, as project coverage described.
The cutting face is the only part of this that is genuinely difficult.
Everything else is logistics.
