TUNNELS. MOVEMENT. ENGINEERING JUDGMENT.

Tunnel & Underground Geotechnical Monitoring Intelligence

GeoSmar reviews tunnel and underground monitoring data, ground movement, groundwater and adjacent-asset response to help project teams distinguish credible change, understand trends and focus engineering action.

Tunnels & Underground Works

Tunnel monitoring is a control system for ground, structure and construction — not a collection of instruments.

Underground works change stress, groundwater conditions and the support provided by the ground. The monitoring system therefore has to do more than record movement. It should test design assumptions, show whether construction is behaving as expected, identify effects on nearby assets and support timely engineering response when behaviour changes.

Ground & construction Ground model, tunnel method, excavation stage, face position, dewatering, support installation, grouting, adjacent works and known asset constraints.
Monitoring evidence Survey, settlement, convergence, inclinometers, extensometers, piezometers, structural sensors, vibration, automated systems and wider-area ground-motion information.
Engineering intelligence QA/QC, trends, rate of change, trigger review, cross-checking between instruments, interpretation and clear recommendations for the next technical decision.
GeoSmar’s role begins above the field-measurement layer. A project can retain its existing instrumentation contractor, survey team and data-acquisition systems while GeoSmar provides monitoring strategy, independent review, diagnostics and recurring engineering interpretation.

The monitoring questions

The instrument should be selected only after the engineering question is clear.

The U.S. Federal Highway Administration states that tunnel instrumentation should verify design assumptions, monitor construction performance and help avoid or mitigate problems. Singapore LTA’s current civil design criteria likewise require instrumentation to verify assumptions, confirm predicted behaviour, assess effects on buildings and utilities, provide a performance record and enable contingency measures in time.

Ground

How is the ground responding?

Surface settlement, subsurface deformation, volume-loss response, lateral movement, heave and groundwater change may need to be understood together.

Tunnel

How is the excavation and lining behaving?

Convergence, lining displacement, load, strain or support-system response may be relevant depending on the tunnelling method and design assumptions.

Assets

What is happening to nearby structures and utilities?

Buildings, operating tunnels, railways, bridges, utilities and other third-party assets may need their own movement, distortion, vibration or structural monitoring.

Ground conditions & geology

The same tunnel geometry can produce very different monitoring needs in different ground.

This is a global industry page rather than a page for one identified tunnel project. GeoSmar therefore does not assign a site geology that has not been provided. Project-specific advice should be based on official geological information, the ground investigation, the interpreted ground model and the actual tunnelling method.

Ground information to review before monitoring design

  • Stratigraphy, weathering profile and strength or stiffness variation
  • Groundwater levels, pressure regime and permeability
  • Compressible layers, fill, weak seams or soft deposits where documented
  • Rock mass discontinuities, faults, cavities or karst where documented
  • Existing tunnels, basements, piles, utilities and buried structures
  • Expected excavation and support sequence
  • Ground treatment, dewatering, grouting or freezing where applicable

Why geology changes the monitoring logic

Soft-ground tunnelling may place more emphasis on settlement, pore pressure and volume-loss response. Rock tunnelling may require attention to convergence, discontinuity-controlled movement and support performance. Mixed ground, permeable deposits, weak interfaces or existing underground structures can change both the expected mechanism and the instrument locations required to observe it.

Official example: London Clay at Crossrail Hyde Park
Crossrail Learning Legacy documents a research monitoring site where new Crossrail tunnels passed beneath existing London Underground tunnels in London Clay. The monitoring scheme included surface and borehole instrumentation, rod extensometers, in-place inclinometers and multi-level vibrating-wire piezometers. The case demonstrates how subsurface movement and groundwater response were studied around closely spaced urban tunnels.
Official example: Chalk tunnelling toward Woolwich
Crossrail’s published chalk case reports manual levelling studs for surface and shallow sewer monitoring, hydrostatic levelling cells for buildings and bridges, and extensometers for subsurface monitoring. Monitoring frequency changed as the TBM entered and left the monitoring zone of influence.
Official example: high-permeability ground at Singapore TEL Marina Bay
Singapore LTA states that ground freezing was used before tunnelling works at Marina Bay station because high soil permeability increased the risk of water seepage. The lesson for monitoring design is that groundwater and ground-treatment behaviour can be as important as displacement when permeability is a controlling risk.
Site geology should never be invented for SEO. For a named future tunnel project, GeoSmar should replace this general discussion with verified geological units, GI findings and project-specific monitoring implications drawn from official or client-provided records.

Monitoring architecture

A credible tunnel monitoring system watches the excavation, the ground and the assets around it.

The exact system depends on the design and risk register, but a useful monitoring architecture usually separates what is being observed into layers so that one dataset can be checked against another.

Surface & ground

Settlement, heave, lateral deformation, ground-loss response and wider movement patterns above and around the tunnel.

Subsurface

Movement with depth, extensometer response, groundwater or pore pressure and deformation near the excavation influence zone.

Tunnel & support

Convergence, lining displacement, support load, strain, crack or joint movement where the design requires those parameters.

Third-party assets

Existing tunnels, railways, buildings, foundations, bridges, utilities and other structures requiring independent movement or distortion criteria.

Manual survey Automated total station GNSS Inclinometer Extensometer Piezometer Convergence Tilt Crack / joint movement Vibration InSAR context

Instrument selection

Different instruments answer different tunnel-monitoring questions.

The table below is an early-stage engineering discussion framework, not a specification. Final instrument type, accuracy, range, frequency, redundancy and installation detail must be tied to the project’s ground model, predicted behaviour, asset tolerance and response plan.

Engineering question Possible instrument / method Typical use in underground works Important interpretation issue
Is the ground surface settling or heaving? Precise levelling, settlement points, automated total station, GNSS, hydrostatic levelling where appropriate Settlement troughs, station boxes, shafts, adjacent roads and structures Stable reference, baseline, survey geometry, rate of change and relationship to face / excavation position
How is lateral movement changing with depth? Manual inclinometer, in-place inclinometer, ShapeArray-type systems Cut-and-cover works, shafts, retaining systems, slopes and ground beside tunnels Point of fixity, casing behaviour, baseline, depth correlation and apparent versus real movement
How is vertical or horizontal subsurface strain distributed? Rod extensometer, magnetic / multipoint extensometer Ground above tunnels, tunnel interaction zones, deep settlement or heave profiles Anchor depths, reference stability and interpretation against the expected deformation mechanism
Is groundwater or pore pressure changing? Vibrating-wire piezometer, standpipe / Casagrande-type piezometer where response is adequate Tunnelling, shafts, dewatering, ground treatment and permeable strata Response time, installation zone, hydraulic connectivity, temperature and construction sequence
Is the tunnel lining converging or distorting? Optical targets, automated survey, convergence arrays, displacement sensors Existing tunnels, sprayed-concrete lining, mined caverns and new underground structures Reference system, cross-section geometry, installation timing and the displacement that occurred before first reading
Are structural loads or strains changing? Strain gauges, load cells, pressure cells, structural sensors Supports, linings, struts, anchors, critical existing structures Calibration, temperature, load path and distinction between local and global response
Are cracks, joints or interfaces moving? Crackmeter, joint meter, optical displacement monitoring Existing tunnels, buildings, utilities and brittle structures Movement direction, temperature, installation stability and relation to overall distortion
Is vibration affecting an asset? Vibration monitor / geophone / accelerometer Mining, blasting, breaking, TBM-related activity and sensitive adjacent assets Event timing, source, frequency content and the applicable project criterion
Is movement extending beyond the instrumented corridor? InSAR-derived ground-motion information Historical baseline, corridor screening and wider surface movement context Line-of-sight geometry, coherence, observation cadence, geolocation and correlation with ground instruments
LTA’s current design criteria give useful examples of specification detail. The criteria require reading-frequency tables, project-specific accuracy and range, manual confirmation of automated-total-station performance, and inclinometers extending beyond the tunnel or excavation influence zone to a point of fixity. These are examples of how monitoring design moves from “instrument name” to an auditable engineering requirement.

Live infrastructure & third-party assets

Urban tunnelling becomes harder when the assets above and beside the tunnel cannot stop operating.

Operating railways, historic structures, bridges, utilities and occupied buildings often have tighter tolerances and more complex response procedures than the tunnel itself. Monitoring therefore needs to address both absolute movement and distortion relevant to the asset.

Operating tunnels & railways

Automated monitoring, redundancy, stable references and rapid communication become important when underground works pass close to live transport infrastructure.

Buildings & foundations

Settlement alone may not describe damage potential. Differential movement, gradient, tilt and distortion may be more relevant to the structure being protected.

Utilities & buried assets

Access can be limited and the asset may continue operating throughout tunnelling. Instrument installation, protection, remote acquisition and response ownership need early agreement.

Singapore LTA example: Circle Line 6 beneath protected and operating assets
LTA states that more than 600 instruments were installed and monitored around the clock around the former Tanjong Pagar Railway Station during CCL6 tunnelling. It also states that close to 100 instruments monitored Keppel Viaduct during underpinning and tunnelling. These figures illustrate the monitoring density that may be required when tunnelling interacts with significant third-party assets.
Singapore LTA example: Downtown Line close to live MRT tunnels
LTA reports that some Downtown Line tunnels were constructed only metres from operating rail tunnels and that hundreds of instruments monitored live tunnels 24/7. The project context shows why data availability and response timing can become as important as instrument accuracy.
Crossrail example: utility convergence monitoring
Crossrail Learning Legacy documents the use of Shape Accel Arrays to monitor displacement and convergence in the Ranelagh Sewer while twin-bore tunnel drives passed below. The published case emphasises reliable monitoring of third-party assets, installation planning and compatibility with the asset owner’s requirements.

Trigger levels & engineering response

A trigger level is useful only when the action, authority and response time are defined with it.

FHWA warns that leaving allowable movement and mitigation decisions undefined can create technical arguments and commercial disputes. Crossrail published examples show how Green, Amber and Red trigger states were linked to specific reviews, inspections and construction responses.

Before construction

Define the basis

Predicted behaviour, tolerable asset movement, baseline, trigger parameter, monitoring frequency and decision ownership should be established before the trigger is needed.

During construction

Review trend, not only threshold

Rate of change and consistency with nearby instruments can justify early review even before a formal threshold is crossed.

After exceedance

Link data to an action plan

Notification, data verification, inspection, engineering review, increased frequency, mitigation or work stoppage should be tied to the agreed trigger framework.

Crossrail Stepney Green: an example of action-linked triggers
Crossrail Learning Legacy describes Green, Amber and Red trigger levels for sprayed-concrete-lined caverns. The Green response included review and possible monitoring-frequency increase; Amber required inspection and engineering review; Red required tunnelling to stop and an emergency engineering review. The exact percentages used there were project-specific and should not be copied to another tunnel without its own design basis.
Why distortion can matter more than absolute settlement
Crossrail’s Bond Street case notes that damage to buildings, structures and utilities is often associated with distortion rather than displacement alone. Monitoring and trigger design should therefore match the damage mechanism of the asset being protected.
Do not copy trigger values from another project. Trigger criteria must be derived from the current design, impact assessment, asset tolerance, regulatory requirements and agreed response plan. Published case-study values are useful for understanding the framework, not for reusing the numbers.

Monitoring QA/QC

Fast data is not useful if the reference, baseline or instrument condition is uncertain.

Tunnel monitoring often combines manual survey, automated survey and geotechnical sensors. A disciplined QA/QC workflow reduces false alarms and prevents an apparent movement from becoming an engineering conclusion before the data has been checked.

Before first reading

Confirm instrument identity, coordinates, orientation, calibration, installation record, reference point, baseline period and expected sign convention.

During monitoring

Check continuity, missing readings, jumps, drift, environmental effects, survey-reference stability, damage, maintenance records and consistency with nearby instruments.

Before reporting a trigger

Verify the measurement, compare the trend and rate, check construction activity, review independent datasets and record what is observation versus interpretation.

Singapore LTA’s published standards require instrument track records, calibration certificates, maintenance and inspection schedules, protection against damage, and prompt replacement of malfunctioning instruments. LTA’s current design criteria also require manual monitoring to confirm continued accuracy where automated total stations are used.

Contract & data interfaces

Tunnel monitoring contracts fail at interfaces when responsibility is assumed instead of written down.

The monitoring specification should make technical and commercial responsibilities visible. FHWA explicitly links poorly defined action levels to disputes over mitigation and payment. For remote-first review, the data interface also has to be clear enough that GeoSmar is not being asked to interpret an incomplete or undocumented dataset.

  • Who designs and signs off the monitoring plan?
  • Who installs, protects, maintains and replaces instruments?
  • Who owns the raw data and provides access?
  • Who establishes and approves the baseline?
  • Who defines Green / Amber / Red or other response levels?
  • Who verifies an apparent trigger before escalation?
  • Who has authority to slow, change or stop construction?
  • What monitoring frequency applies to each construction stage?
  • What redundancy or manual verification is required?
  • How are damaged, moved or re-baselined instruments handled?
  • How quickly must data and reports reach each stakeholder?
  • What happens when access is restricted by live operations?
  • How are Engineer-of-Record and statutory responsibilities preserved?
  • How are instructed mitigation and resulting commercial consequences documented?
Independent review is not statutory approval. GeoSmar can provide technical interpretation and independent monitoring review. Where local law or contract conditions require approval by a licensed local professional, the Engineer, Designer or Engineer-of-Record, that formal role must remain with the duly appointed party unless GeoSmar has been formally appointed and is qualified for that jurisdiction.

Official public case studies

Published tunnel projects show how monitoring changes with geology, construction method and the assets at risk.

The cases below are drawn from official project-owner or government sources. They are technical references only; they are not presented as GeoSmar projects or as endorsements of GeoSmar.

Government project owner · Singapore LTA

Circle Line 6 — tunnelling beneath significant existing assets

LTA reports more than 600 instruments monitored the former Tanjong Pagar Railway Station around the clock during tunnelling. Close to 100 instruments monitored Keppel Viaduct during underpinning and tunnel construction. The case shows the scale of monitoring required when new tunnels interact with protected and operating infrastructure.

Official LTA source ↗

Government project owner · Singapore LTA

Thomson–East Coast Line — Orchard and Marina Bay

LTA states that settlement and movement around Orchard MRT were monitored 24/7 with real-time instruments during micro-tunnelling and mining. At Marina Bay, ground freezing was used before tunnelling because high soil permeability increased seepage risk. The case links movement monitoring directly with ground treatment and groundwater risk.

Official LTA TEL source ↗

Project-owner technical legacy · Crossrail Ltd

Hyde Park — new tunnels below existing tunnels in London Clay

Crossrail documents surface and borehole monitoring around twin EPB tunnel drives beneath existing London Underground tunnels. The instrumentation included rod extensometers, in-place inclinometers and multi-level vibrating-wire piezometers, providing a detailed view of ground response in London Clay.

Crossrail Learning Legacy ↗

Project-owner technical legacy · Crossrail Ltd

Woolwich area — TBM-induced settlement in Chalk

Crossrail’s published chalk case describes manual levelling studs for surface and shallow sewer monitoring, hydrostatic levelling cells for buildings and bridges, and extensometers for subsurface response. Monitoring frequency increased when areas entered the TBM zone of influence.

Crossrail Chalk case ↗

What these cases have in common: no single sensor carried the entire monitoring argument. The useful engineering picture came from matching instruments to mechanisms, increasing frequency when risk increased, protecting critical assets and linking measurements to a formal review process.

How GeoSmar can support underground works

GeoSmar is designed to review what the tunnel monitoring means, even when another contractor performs the field work.

GeoSmar is the market-facing brand of Rauz Caucasus LLC and is structured for remote-first international delivery. For tunnel and underground projects, the strongest fit is the technical layer between field measurements and the project decision.

Monitoring Intelligence

Recurring engineering review

Review trends, rates, trigger status, anomalies, groundwater and movement across multiple instruments and construction stages.

Independent Review

Separate technical assurance

Review monitoring plans, contractor reports, data quality, trigger logic and whether the engineering explanation is supported by the evidence.

Diagnostics

Investigate unexplained readings

Examine sudden settlement, inclinometer shifts, inconsistent survey, piezometer change or conflicting instruments before the project treats them as real movement.

Monitoring Design

Plan the monitoring logic

Support monitoring philosophy, instrument selection, layout concepts, baseline requirements, reading frequency, trigger framework and reporting structure.

InSAR

Add wider ground-motion context

Where suitable, compare tunnel monitoring with satellite-derived surface movement to understand historical or spatial deformation outside the dense ground network.

Reporting

Reduce repetitive review work

Structure automated charting, threshold checks and data screening so engineers can spend more time on interpretation and the decisions that matter.

Remote-first Vendor-neutral Independent review Existing contractors supported Project-specific scope

Frequently asked questions

Tunnel monitoring questions that should be answered before the data starts arriving.

Which instruments are normally used for tunnel monitoring?
There is no universal tunnel instrument set. Depending on the mechanism and asset, a project may use settlement survey, automated total stations, GNSS, inclinometers, extensometers, piezometers, convergence monitoring, tiltmeters, crackmeters, strain or load sensors, vibration monitoring and other project-specific systems. Instrument selection should follow the engineering question and ground model.
How often should tunnel monitoring be read?
Frequency should change with risk, construction activity and the required response time. Singapore LTA’s current design criteria require the monitoring design to state reading frequencies and criteria for increasing or decreasing them. For real-time monitoring during active tunnelling under buildings or critical structures, LTA defines less than three hours between successive readings of the same instrument; individual projects may require much shorter intervals.
Can automated total station data be accepted without manual checks?
Not automatically. Singapore LTA’s current criteria require provision for regular manual monitoring to confirm the continued accuracy of instruments monitored in real time using automated total stations. The correct verification regime for another jurisdiction should follow that project’s specification and regulatory requirements.
Should trigger levels be the same on every tunnel project?
No. Trigger levels should reflect predicted behaviour, asset tolerances, design assumptions, contractual requirements and the agreed action plan. Published Crossrail examples are useful for understanding the Green / Amber / Red process, but their numerical values are project-specific.
Can GeoSmar review monitoring performed by another contractor?
Yes, subject to data quality and scope. GeoSmar’s intended model allows field installation, routine readings and maintenance to remain with an existing local contractor while GeoSmar provides independent data review, diagnostics and engineering interpretation.
Can InSAR replace tunnel settlement monitoring?
No. InSAR can provide historical and wider-area surface-movement context where suitable, but it does not replace subsurface deformation, groundwater, tunnel convergence or high-frequency local monitoring. It is most useful as a complementary evidence layer.
What information does GeoSmar need for a first tunnel review?
Useful starting material includes the project brief, tunnel alignment and sections, construction method, ground investigation, groundwater information, affected assets, monitoring plan, trigger criteria, instrument schedule, baseline records, recent data and construction progress. The scope can then be matched to the specific question the project needs answered.

Start a technical discussion

Have tunnel monitoring data that is difficult to explain?

Send GeoSmar the tunnel or underground-works brief, monitoring plan, sample data and the engineering question you need to answer. A first review can determine whether the most useful next step is independent monitoring review, data diagnostics, monitoring-design support, InSAR context or recurring monitoring intelligence.

Scroll to Top