Subsea tunnel depth profiles: how deep the world’s tunnels go

• 44entries listed

The cross-sections

Cross-sections drawn to a common scale from the depths recorded in this archive. The low point of each profile is the published figure; the shape between the portals is schematic.

How this archive measures: lengths, bores, dates and ordering

How this table is compiled

Round figures in an old record are usually estimates that hardened into fact. A disused rail bore is the commonest cycleway in some countries, which is a second life a length ranking cannot show. Where a crossing has two bores of different lengths, the longer governs the comparison and both are recorded on the source table.

Escape provision has changed more over the past forty years than excavation method has. A length quoted to the metre implies a survey that was actually carried out, and not every published figure carries one. An intermediate shaft turns one heading into two and leaves a structure on the surface for the life of the tunnel.

Linings are inspected on a cycle of years and repaired in windows of hours. Transverse ventilation gives each section its own supply and extract, which is why the long two-lane crossings carry a false ceiling. Length stands in for terrain more than for difficulty, and on a head-to-head the two crossings rarely faced the same terrain.

This page plots the 44 subaqueous road tunnels for which this archive records both a length and a lowest point. Published lengths for the same structure differ by a few metres between sources, depending on whether cut-and-cover approaches are counted. Water management is the maintenance problem that never ends, and drainage design decides its cost.

The tail of a record is more representative of what gets built than its head is. Trade reporting is faster than official publication and less reliable, which is why opening dates sometimes move after the event. Toll revenue is usually pledged against the debt that built the crossing, and the toll ends when the debt does.

Reading the figures

The first long bores were driven by hand and gunpowder, and their alignments still show it. Two-way traffic in one bore is what modern regulation avoids and what most older road crossings were built with. Base tunnels replaced summit lines that climbed by spiral and switchback, and several of those older routes survive above them.

A ruling gradient set by the heaviest train decides whether a rail crossing needs a tunnel at all. A comparison of two crossings is a comparison of two ground conditions, and the length column records neither. Water under pressure delays a heading more reliably than distance does, and it appears in no ranking.

A pilot bore driven ahead of the main tunnel is usually kept afterwards as the escape route. Environmental consent now outlasts construction on many schemes, and the dates in a record rarely show it. Reconstruction after a fire is recorded because it usually changes both the profile and the ventilation arrangement.

Two independent sources back every value. A short bore beneath a city can cost more than a long one through a mountain. The most useful sources are the least accessible: inspection records and commissioning documents rather than announcements. Pressure waves in a long rail tunnel are severe enough to require relief shafts or sealed carriages.

Opening year means the year traffic was admitted. Figures are checked against one another where they differ. Sources are named on each structure's own table. Blank cells mean the value is not on record. Sequential excavation is slower per metre and far more tolerant of ground that changes without warning.