The most expensive surprises are underground
Tunnelling has a reputation for overruns, and it is deserved. The reasons are almost always geological: an unexpected fault zone, a band of weak or water-bearing rock, a void, a sudden change in hardness that the boring machine was not specified for. Each surprise stops work, and stopped work underground is ruinously expensive.
The conventional defence is the site investigation borehole. But boreholes along a tunnel alignment are sparse and costly, and the geology between them is exactly where the surprises hide. A few holes cannot characterise kilometres of varied ground.
Imaging the alignment, not sampling it
Geophysics changes the sampling problem into an imaging one. By surveying the corridor non-invasively — mapping conductivity, density and structure continuously rather than at points — engineers see the variation between the boreholes, not just at them. Fault zones, water-bearing layers and voids reveal themselves as anomalies in the field data.
A borehole tells you about a point. The tunnel has to pass through everything in between.
From model to method statement
The resulting 3D ground model lets the project specify the right machine, plan ground treatment where it is genuinely needed, and price the contract against mapped conditions rather than optimistic assumptions. It also reduces disputes: when conditions are characterised in advance, there is far less argument about whether a surprise was foreseeable.
The cheapest tunnelling risk to manage is the one you mapped before the machine arrived.
Through construction and beyond
Because surveys are repeatable, the same approach monitors ground movement and settlement during construction and across the asset's life. The tunnel's ground model becomes a living record — useful for maintenance decades after the ribbon is cut. Risk, in the end, is just uncertainty with consequences; the way to manage it is to remove the uncertainty before the consequences arrive.
