← Tunnel Engineering & Rock Mechanics

Tunnel Setting-Out, Monitoring, QA/QC & Safety

Surveying for tunnels — control networks, transfer of line and level underground, gyro and laser guidance, breakthrough accuracy; instrumentation and monitoring — convergence, extensometers, load cells, settlement and piezometers with alert–alarm–action trigger levels; quality assurance of shotcrete, bolts, grouting and lining; and safety — ventilation, gases, blasting, ground falls, fire and emergency escape.

📑 Contents (5 sections)

Last reviewed 30 Sept 2026 · 7 min read

Tunnel surveying and setting-out

A tunnel is driven from one or more portals or shafts, often from both ends, and the headings must meet within a tolerance (the breakthrough accuracy, typically a few centimetres over kilometres). The line and level are transferred and carried underground by careful surveying.

Surface control network

  • A network of control points is established around the portals and along the surface alignment with GNSS (GPS) and precise total stations and levelling, tied to the national grid and datum.
  • The alignment is set out in coordinates; redundant observations and adjustment ensure accuracy.

Transfer of line and level underground

  • Through a portal or adit — the line is carried by a traverse of total-station observations into the tunnel.
  • Through a vertical shaft — the surface coordinates are transferred by plumb lines (two wires hung in the shaft and observed at both the top and bottom), by optical plummets or laser plummets, or by gyro-theodolite orientation.
  • Level is transferred by precise levelling and steel tape or by electronic distance meter down a shaft, with corrections for tape temperature and tension.

Underground control

  • Traverse stations are placed on the tunnel roof or walls (forced-centring brackets) and observed regularly; the traverse is closed where possible, and long straight tunnels use a gyro-theodolite to check the azimuth, as the traverse accumulates lateral (refraction) error in narrow tunnels.
  • Laser guidance — a laser beam mounted on the wall defines the line; the drilling jumbo or TBM navigation system uses it for direction and grade; TBM guidance systems show the position relative to the design line in real time.
  • Grade and level by laser level or precise level, with benchmarks on the walls.
  • Centreline pegs and profile control: the face position is marked by the surveyor, and the excavated profile is checked by a laser scanner or profiler against the design line to measure overbreak and underbreak, and the shotcrete thickness.

Error in breakthrough

The combined error of the surface network, the shaft transfer and the underground traverse decides the breakthrough error. It is estimated during planning by error propagation, and the survey method is chosen to meet the tolerance (a longer tunnel needs gyro control). A breakthrough check compares the coordinates from both headings before the final round.

Monitoring during construction

Monitoring measures the response of the ground and the support, so that the design can be verified and safety maintained. It is fundamental in NATM and in tunnels near buildings.

FormulaTypical monitoring measurements
  • Convergence — the change in distance between points on the tunnel walls or between the crown and invert, measured with a tape extensometer or by total-station readings of reflective targets (three-dimensional displacement of each target).
  • Crown settlement and invert heave by levelling the targets.
  • Extensometers — multipoint borehole extensometers (MPBX) measure the movement of the rock at several depths from the tunnel wall.
  • Load and pressure cells — on rock bolts (load cells), and pressure cells in or behind the shotcrete or lining; strain gauges on steel ribs.
  • Surface and building monitoring — settlement points, tilt meters on structures, inclinometers and piezometers for ground water.
  • Vibration from blasting (PPV) and noise.

Reading and interpretation

  • Readings are taken at intervals — daily near the face, then less often as the movement stabilises — with the distance from the face and time recorded.
  • The results are plotted as displacement vs time and vs distance to the face. A healthy tunnel shows movement that decreases and stabilises once the face has advanced and the ring is closed.
  • The rate of convergence and the total are compared with the predicted values and the trigger levels. A rate that stops falling or increases signals a problem: over-stress of the rock or the support, a nearby fault, or water.

Trigger levels

Three levels are agreed in advance:

Level Meaning Response
Alert Behaviour differs from the prediction Increase monitoring, inspect, review
Alarm Approaching a limit Stop the advance, install additional support, notify the engineer
Action The limit is exceeded Evacuate if necessary; emergency measures; the design is reviewed
Worked ExampleExample — reading convergence

A horizontal span of a tunnel is measured as 8.000 m at installation. After 30 days the distance is 7.985 m.

Convergence = 8.000 − 7.985 = 15 mm, or 15/8000 = 0.19 % of the span.

If the daily rate has fallen from 3 mm/day in the first week to 0.2 mm/day, the movement is stabilising and the support is adequate. If the rate rises back to 2 mm/day after a period of quiet, the alert level is reached and the cause (a new fault, water, or a change in the excavation) is investigated.

Back-analysis: the measured displacements are used to calibrate the ground properties in a numerical model, which refines the design for the following stretches.

This chapter is in the syllabus of

Open an exam to see where this chapter sits in its syllabus, and to practise it.