Underground Geotechnical Monitoring in Hong Kong
A practical, evidence-led guide to monitoring deep excavations, tunnels, ERSS, utilities and basements—linking ground behaviour, water response and structural movement to timely engineering decisions.
Why underground monitoring matters in Hong Kong
Hong Kong underground works frequently combine restricted sites, deep temporary works, variable ground and groundwater conditions, nearby buildings, utilities and operating transport assets. Monitoring is most useful when it is designed as part of the engineering control process—not treated as a detached stream of readings.
Excavation and lateral support
For deep excavations and excavation-and-lateral-support systems, monitoring may need to track groundwater, wall deflection, support loads and movement or vibration at sensitive receivers. Hong Kong GEO guidance describes these as typical monitoring components, subject to the project design and site conditions.[1]
Tunnelling and sensitive receivers
Hong Kong tunnel guidance highlights hydrogeology, water inflow or pressure, settlement, heave, lateral movement, vibration and the response of nearby sensitive receivers as matters to consider in investigation and design.[2]
Ground-loss risk management
CEDD supplementary guidance for deep excavation calls for attention to groundwater inflow and drawdown, maintained instruments, monitoring of ground, buildings and services, and predefined contingency arrangements where that guidance applies.[3]
Scope note: This page explains engineering good practice and instrument-selection logic. Monitoring frequency, trigger levels, action plans, instrumentation density and authority submissions must be defined by the responsible project professionals for the actual ground model, construction sequence, risk allocation and statutory requirements.
Monitoring applications below and beside the city
The monitoring system should follow the credible failure mechanisms and construction stages for the work. The same instrument can be useful in several settings, but its layout and interpretation change with the engineering question.
Deep excavations and basements
Observe retaining-wall deflection, ground settlement, strut or anchor response, basal movement and groundwater change as excavation and propping advance.
Tunnels and shafts
Track convergence, lining movement, ground response, pore pressure, vibration and the behaviour of assets within the predicted influence zone.
ERSS and temporary works
Connect instrumentation to the staged excavation sequence, temporary support installation, preload or stressing activities, and the project response plan.
Utilities and transport assets
Measure movement or vibration at pipelines, drains, rail structures, roads and other sensitive infrastructure while distinguishing construction response from background variation.
Adjacent buildings and structures
Combine movement, tilt, cracking and vibration data with condition records so that observed changes can be assessed in their structural and construction context.
Monitoring parameters and the decisions they support
Start with the parameter that can confirm or challenge an engineering assumption. Then select the instrument, reference frame, reading frequency and reporting route that can deliver decision-ready evidence.
Lateral ground or wall movement
Why: identify deformation profiles and developing shear zones. Where: retaining walls, soil or rock beside excavations and tunnels. Decision: review excavation sequence, support installation or ground treatment.
Vertical settlement or heave
Why: quantify surface and asset movement. Where: roads, slabs, tracks, utilities and surrounding ground. Decision: compare actual response with predictions and protection criteria.
Convergence and clearance
Why: detect relative movement across a tunnel, shaft or excavation. Where: linings, portals and confined underground spaces. Decision: assess stability, support response and operational clearance.
Groundwater level
Why: observe drawdown, recharge and hydraulic trends. Where: standpipes or instrumented boreholes inside and outside the works. Decision: investigate leakage, dewatering effects or unexpected inflow.
Pore-water pressure
Why: assess effective-stress and hydraulic response at a defined zone. Where: soil layers, interfaces and beneath excavations. Decision: review uplift, piping, stability and dewatering assumptions.
Earth or contact pressure
Why: observe load transfer at soil–structure interfaces. Where: walls, slabs, linings or embedded elements. Decision: evaluate whether pressure development is consistent with the design model.
Strut, tie or anchor load
Why: track temporary-support demand and load redistribution. Where: props, walers, anchors and connection zones. Decision: assess support performance before progressing to the next stage.
Structural strain
Why: infer local stress or load trends when the structural relationship is defined. Where: steel, concrete reinforcement, piles and tunnel linings. Decision: identify load concentration or unexpected redistribution.
Tilt and rotation
Why: capture angular change that may not be obvious from isolated settlement points. Where: façades, columns, walls, equipment and structural frames. Decision: assess differential movement and serviceability implications.
Crack movement
Why: separate opening, closing or shear change from a static visual record. Where: existing buildings, linings and concrete elements. Decision: correlate crack behaviour with construction and environmental conditions.
Vibration
Why: characterise transient ground or structural motion. Where: sensitive buildings, utilities, tunnels and equipment. Decision: manage work methods and assess events against project criteria.
Acceleration and dynamic response
Why: resolve higher-frequency structural behaviour where velocity alone is insufficient. Where: structures, machinery zones and transport assets. Decision: investigate dynamic sources, resonance or operational effects.
Deep axial movement
Why: locate movement with depth rather than only at the surface. Where: soil, rock masses, foundations and tunnel influence zones. Decision: distinguish shallow movement from deeper ground response.
Construction activity and environment
Why: provide the context needed to interpret instrument changes. Where: excavation stages, pumping records, rainfall, temperature and work logs. Decision: test whether a change is construction-related, environmental or instrumental.
Instrument health and data continuity
Why: unreliable telemetry can resemble stable ground or sudden movement. Where: sensors, cables, loggers, power and communications. Decision: verify anomalies, schedule maintenance and preserve an auditable record.
Match the instrument to the engineering question
Selection depends on required accuracy, reliability, response time and site conditions—an approach reflected in Hong Kong GEO guidance.[1] The matrix is qualitative: project-specific range, resolution, installation detail and acceptance criteria must be checked against the design and manufacturer documentation.
| Instrument / method | Primary parameter | Typical underground use | Reading mode | Key strength | Important limitation |
|---|---|---|---|---|---|
| Manual inclinometer | Lateral displacement profile | Retaining walls, slopes, ground beside excavations | Manual survey | Depth-resolved deformation profile | Intermittent; access and operator practice affect repeatability |
| In-place inclinometer (IPI) | Lateral displacement at instrumented depths | Walls or ground needing frequent updates | Automated or hybrid | High-frequency trend and alarm capability | Measures only configured zones; requires stable reference and maintenance |
| Vibrating-wire piezometer | Pore-water pressure | Defined strata, beneath slabs, around dewatering works | Manual logger or automated | Zone-specific pressure response and robust long-term signal | Installation and saturation quality are critical; not a complete vertical profile alone |
| Standpipe piezometer | Groundwater head | Baseline and long-term groundwater observation | Manual or water-level logger | Direct, understandable water-level reference | Response may be slow in low-permeability ground and the screened zone must be known |
| Water-level sensor | Continuous water level | Standpipes, sumps or wells | Automated | Captures rapid drawdown, recharge and pumping cycles | Needs barometric, drift and fouling checks appropriate to the sensor |
| Rod extensometer | Axial displacement at selected anchors | Ground above tunnels, foundations and excavation bases | Manual or automated | Locates movement with depth | Interpretation depends on anchor integrity and reference stability |
| Multi-point borehole extensometer (MPBX) | Multi-depth axial movement | Rock or soil around underground openings | Manual or automated | Separates movement between several depth intervals | Complex installation; fewer measurement axes than a full spatial survey |
| Settlement point / marker | Vertical movement | Ground, road, slab and utility corridors | Survey | Simple spatial coverage and clear datum relationship | Intermittent and vulnerable to disturbance; datum stability must be controlled |
| Precise levelling | Elevation change | Buildings, rails, roads and settlement arrays | Manual survey | Traceable network measurement over many points | Line of sight, access, closure and survey procedure govern quality |
| Survey prism | Three-dimensional point movement | Façades, retaining walls, tunnel portals and structures | Manual or robotic total station | Coordinates movement across a broad network | Requires line of sight; atmospheric and reference-prism stability matter |
| Automated total station (ATS/RTS) | Automated 3D displacement | Dense networks around excavations, tunnels and assets | Automated | Scalable, frequent remote observations | Occlusion, refraction, vibration and poor reference geometry can degrade results |
| Tiltmeter | Rotation | Buildings, retaining walls, columns and machinery | Manual or automated | Responsive to angular change | Temperature, mounting and local deformation must be separated from global movement |
| Electrolevel | High-sensitivity tilt | Beams, tunnel segments and structural arrays | Automated | Suitable for linked differential-rotation systems | Needs careful installation, temperature management and stable cabling |
| Crackmeter / jointmeter | Relative crack or joint movement | Existing structures, linings and concrete joints | Manual or automated | Directly tracks a selected discontinuity | Local reading does not alone establish structural cause or overall movement |
| LVDT / displacement transducer | Short-range relative displacement | Joints, bearings, supports and laboratory-style local measurements | Automated | Continuous and responsive local measurement | Mounting frame and working range must suit the expected movement |
| Strain gauge / vibrating-wire strain gauge | Strain | Props, piles, steelwork, reinforcement and linings | Manual or automated | Direct local structural-response trend | Stress or load inference needs material, section and temperature information |
| Load cell | Axial force | Anchors, struts and selected supports | Manual or automated | Measures force at the installed load path | Alignment, seating and load redistribution can affect representativeness |
| Pressure cell | Earth or contact pressure | Soil–structure interfaces, slabs and tunnel linings | Manual or automated | Observes local contact-pressure change | Installation disturbance and arching make a point reading hard to generalise |
| Geophone | Particle velocity / vibration event | Buildings, utilities, tunnels and sensitive receivers | Continuous or event-triggered | Established construction-vibration monitoring method | Sensor coupling, frequency content and event attribution require review |
| Accelerometer | Acceleration / dynamic response | Structures, plant interfaces and transport assets | Continuous or event-triggered | Useful for dynamic and higher-frequency behaviour | More data and processing; acceleration is not interchangeable with velocity |
| Convergence monitoring | Relative closure / geometry change | Tunnels, shafts and headings | Tape, total station or automated sensors | Direct measure of opening response | Reference geometry, access and construction obstructions must be controlled |
| Distributed optical fibre sensing | Distributed strain or temperature | Long alignments, linings, pipelines and embedded assets | Automated | Dense spatial coverage along the fibre | System choice, bonding, spatial resolution and interpretation are application-specific |
| Data logger and telemetry gateway | Acquisition, health and transmission | Automated multi-sensor systems | Automated | Centralises timestamps, alarms and remote access | Does not improve a poor sensor installation; power, network and cybersecurity need planning |
Compare instruments before choosing a system
Instruments that appear to measure the same phenomenon often answer different questions. Pairing complementary methods can improve spatial coverage, update frequency or independent verification.
Manual inclinometer vs in-place inclinometer
Manual: a fuller depth profile at survey intervals, useful for locating deformation zones. In-place: frequent readings at preselected depths, useful for trends and alerts. A hybrid arrangement can retain profile checks while automating the critical zones.
Standpipe vs vibrating-wire piezometer
Standpipe: an intuitive groundwater-head observation for a defined screened interval. Vibrating wire: zone-specific pore-pressure response with convenient automation. Cross-checking nearby methods can help distinguish hydraulic change from instrument behaviour.
Precise levelling vs automated total station
Levelling: strong vertical-control networks with procedural closure checks. ATS/RTS: frequent 3D observations across many prisms. Line of sight, reference stability and atmospheric effects must be managed for robotic systems.
Prisms vs tiltmeters
Prisms: point displacement in a survey coordinate system. Tiltmeters: local rotation at high update frequency. Used together, they can show whether rotation is consistent with wider structural translation or differential movement.
Crack gauge vs LVDT
Manual gauge: simple periodic crack-width reference. LVDT or automated jointmeter: continuous relative movement at a selected location. Neither method alone identifies the structural cause, so movement and activity context remain essential.
Geophone vs accelerometer
Geophone: commonly used for particle-velocity-based construction vibration. Accelerometer: resolves acceleration and dynamic structural response. Their outputs and criteria are not interchangeable; choose the quantity needed by the project assessment.
Manual, automated or hybrid monitoring?
The most defensible answer is driven by risk, change rate, access, required response time, sensor suitability and the consequence of data loss. Automation supports timely awareness; it does not replace engineering review or field verification.
Manual monitoring
Useful when change is slow, access is safe and planned, a broad survey network is needed, or the method benefits from direct field observation. Quality relies on stable datums, competent procedure, repeatable setup and prompt reporting.
Automated monitoring
Useful for frequent readings, restricted access, rapidly changing construction stages, overnight work or near-real-time alerts. It needs resilient power, communications, clock synchronisation, health checks, maintenance and an escalation workflow.
Hybrid monitoring
Combines automatic trend detection with independent manual checks and broader periodic surveys. Hybrid systems are often valuable where a critical response needs frequent observation but spatial completeness and verification still matter.
Alarm design: a trigger should have a named recipient, defined verification method, required response time, engineering owner and documented action. Raw threshold notifications without context can create both false confidence and alert fatigue.
From risk register to verified response
A monitoring plan becomes useful when every reading can be traced to a risk, a construction activity and a decision route.
Define objectives
State the failure mechanisms, sensitive receivers, design assumptions and decisions the data must support.
Establish baselines
Collect sufficient pre-construction observations and record existing condition, environmental variation and datum stability.
Select instruments
Match parameter, range, resolution, response time, survivability and reading mode to the actual engineering question.
Design the layout
Place instruments relative to geology, groundwater, construction stages, predicted influence zones and critical assets.
Install and commission
Document location, orientation, depth, sensor identity, calibration information, zero readings and acceptance checks.
Integrate activity records
Align monitoring timestamps with excavation, support, pumping, grouting, tunnelling, traffic and weather records.
Validate data
Use automated quality flags, reference checks, duplicate methods, field observations and engineering plausibility review.
Apply response controls
Route verified changes through the agreed alert–action–alarm or project-specific trigger and contingency process.
Close out and archive
Confirm residual trends, retain an auditable dataset and document what the measurements showed about the ground model.
Groundwater monitoring is a system, not a single water level
Deep excavation and tunnelling can interact with multiple aquifers, perched water, defects, recharge sources and dewatering systems. Interpretation should connect groundwater observations with geology, pumping, inflow, rainfall, tide where relevant, and movement response.
Inside and outside comparison
Instrumentation on both sides of a cutoff or retaining system can help identify hydraulic gradients, drawdown propagation and whether inflow behaviour matches the expected seepage model.
Depth-specific response
Separate screened intervals or piezometer zones can show whether change is confined to one layer. A single open water level should not automatically be treated as the pore pressure everywhere.
Movement correlation
Groundwater and pore-pressure trends become more informative when reviewed beside settlement, wall movement, base response and excavation or pumping milestones.
Anomaly verification
Unexpected drawdown or rise should prompt checks of sensor health, datum, barometric or tidal effects where applicable, pumping records, local leakage and nearby instruments before conclusions are drawn.
Construction controls
Monitoring can support decisions on pumping rate, recharge, grouting, sequencing, leakage investigation and contingency actions—but those controls remain project-specific engineering decisions.
Long-term transition
Plan how temporary dewatering observations will transition through structural completion, recharge and final groundwater recovery, including who owns the data and response after handover.
Separate construction response from environmental and operational effects
A technically sound monitoring system records enough context to explain variation. Temperature cycles, rainfall, tide, traffic, train operations, plant vibration, survey visibility and maintenance activity can influence readings without representing the same ground behaviour.
Baseline duration
Baseline monitoring should capture representative operating and environmental conditions where practicable, not merely a convenient single reading before work begins.
Time alignment
Use consistent clocks and time zones across sensors, site diaries, equipment logs and construction records so events can be correlated without ambiguity.
Reference controls
Protect survey benchmarks, reference prisms and supposedly stable anchor points. If the reference moves, the entire dataset can appear to move with it.
Condition observations
Photographs, visual inspections and condition surveys provide evidence that a remote trend alone cannot supply, especially for cracks, finishes and exposed utilities.
Data qualification
Flag maintenance, lost communication, obstruction, power interruption, reset, re-zeroing and suspected damage so analysts do not treat compromised data as valid ground response.
Independent checks
Where consequences are significant, use a different measurement principle or field check to confirm important changes before escalating or dismissing them.
Turn sensor data into an auditable engineering record
A digital platform should preserve provenance, expose instrument health, connect changes to construction events and help authorised people act. Dashboards are useful only when the underlying measurements and workflow remain trustworthy.
Data provenance
Retain sensor ID, location, orientation, units, calibration or verification records, raw and processed values, correction history and the person or system responsible for changes.
Quality-control rules
Identify missing values, flatlines, outliers, implausible rates, reference drift, low battery and communication loss without silently deleting the original data.
Role-based alerts
Send verified information to the correct engineering and construction roles, with acknowledgement, escalation and a record of actions—not only a generic email blast.
Construction correlation
Overlay excavation levels, support installation, pumping, blasting, grouting, tunnel advance and inspections so trends can be interpreted against the work sequence.
Resilience and security
Plan local buffering, backup power, network fallback, access control, data retention and recovery. Temporary loss of a dashboard should not erase the monitoring record.
Engineering review
Automated calculations can prioritise attention, but competent review is required to assess causation, spatial consistency, construction relevance and the appropriate response.
External reference cases for underground monitoring
These independent published cases illustrate monitoring strategies and lessons relevant to Hong Kong. They are not GEOOE or GEOORIGIN projects, and no participation, supply or client relationship is implied.
Central–Wan Chai Bypass tunnel
A Hong Kong Engineer technical article describes an observational approach around the Causeway Bay Typhoon Shelter portal and mined tunnel. Reported instrumentation included automated deformation monitoring, strain gauges, piezometers, extensometers, utility points, geophones and tunnel convergence observations, supported by baseline monitoring and alert–action–alarm notifications.
Kwu Tung Station above an operating railway tunnel
A 2024 HKIE activity report notes construction of the station above the operating East Rail Line tunnel box. It identifies groundwater control and continuous tunnel monitoring as necessary during and after excavation, illustrating the importance of protecting an asset that remains operational.
Crossrail ground-response instrumentation
Crossrail Learning Legacy documents a research installation in London Clay near existing Central Line tunnels. Rod extensometers, in-place inclinometers and multi-level vibrating-wire piezometers were selected to examine ground response, with paired instruments and a nearby standpipe supporting comparison and long-term verification.
Crossrail Paddington station-box excavation
A Crossrail Learning Legacy paper describes excavation above segmental tunnels using detailed modelling, real-time tunnel-movement monitoring and daily reporting correlated with construction activity. Monitoring informed adjustments to excavation constraints, with inspection and independent cross-checking used alongside the automated evidence.
Thomson–East Coast Line at Orchard
Singapore’s Land Transport Authority states that settlement and movement at Orchard MRT Station were monitored in real time around the clock during technically complex works. The public project account also describes ground improvement and underground construction methods used in the area.
Keiyo Line Tokyo Underground Station
A published Japanese paper describes station construction that crossed an existing Yokosuka Line shield tunnel. The tunnel was excavated around, exposed and temporarily underpinned, and a measurement and control system was introduced to monitor its behaviour and respond to unexpected change.
Metro station deep excavation
A peer-reviewed Procedia Engineering paper presents a long-term real-time monitoring and safety-evaluation strategy for a metro station during deep excavation. The documented monitoring categories included retaining-wall deformation, ground-surface settlement and strut force, integrated through a software platform.
Source: PolyU Institutional Research Archive / Procedia Engineering, 2011[10]
What these cases suggest for Hong Kong projects
The value lies in the monitoring logic, not in copying another project’s thresholds or layout. Ten practical lessons recur across the verified references.
Design from the mechanism
Select measurements that can test the predicted ground, groundwater and structural response.
Start the baseline early
Capture normal environmental and operational variation before critical construction begins.
Instrument the interfaces
Pay attention where new works interact with existing tunnels, utilities, buildings or temporary supports.
Use complementary methods
Combine spatial surveys, depth profiles, local sensors and field inspection where consequences justify it.
Correlate with activity
A reading becomes more useful when aligned with excavation, pumping, support and tunnelling records.
Protect the reference
Benchmark and reference movement can invalidate an otherwise precise monitoring network.
Plan for data loss
Power, communications, access and sensor damage are foreseeable risks that need fallback arrangements.
Verify before interpreting
Check instrument health, adjacent sensors, construction events and physical observations before assigning cause.
Connect alarms to action
Pre-agreed ownership and contingency steps make monitoring operational, not merely informational.
Retain an auditable record
Installation, corrections, approvals and responses should remain traceable through construction and handover.
A monitoring system built around the decision
GEOOE approaches underground monitoring by clarifying the engineering question first, then defining the sensing, acquisition, verification and reporting chain needed to answer it. The project designer and responsible engineers retain control of design assumptions, trigger values and construction decisions.
Scope and risk mapping
Translate construction stages, ground and groundwater risks, sensitive receivers and response times into a measurable scope.
Instrument and network design
Compare manual, automated and hybrid options, installation constraints, reference arrangements, redundancy and survivability.
Installation quality records
Build traceability around sensor identity, position, orientation, borehole or mounting detail, baseline and commissioning checks.
Data acquisition and dashboards
Configure logging, telemetry, health flags, visualisation and role-based notifications around the required update frequency.
Verification and reporting
Combine automated quality checks with field evidence and engineering review so important changes can be confirmed promptly.
Lifecycle support
Plan maintenance, recalibration or checks, damaged-instrument response, data retention and the transition to long-term monitoring.
Explore GEOOE’s geotechnical instrumentation, ground monitoring, ERSS monitoring and technical resources, or review the wider applications library.
Underground monitoring FAQ
What should be monitored for a deep excavation in Hong Kong?
The scope should follow the project’s credible failure mechanisms and sensitive receivers. Common parameters may include retaining-wall and ground movement, settlement, groundwater level or pore pressure, support loads, structural response and vibration. The responsible project professionals must define the actual layout, frequency and criteria.
How early should baseline monitoring begin?
Early enough to characterise normal instrument, environmental and operational variation before critical works. The appropriate duration depends on the asset, expected seasonal or tidal effects, construction programme and authority or contract requirements.
Is automated monitoring always better than manual monitoring?
No. Automation is valuable when update frequency, access or response time demands it. Manual methods can provide broader spatial coverage, direct observations and independent checks. Many higher-risk programmes benefit from a deliberately designed hybrid system.
What is the difference between groundwater level and pore-water pressure?
A standpipe typically indicates hydraulic head over its screened response zone, while a piezometer measures pressure at a more defined location or zone. The results are related but should not be treated as interchangeable without understanding installation details and ground conditions.
How should alert, action and alarm levels be set?
They should come from the project design, predicted behaviour, asset criteria, observational-method plan and construction contingencies. Generic thresholds copied from another project are not a substitute for project-specific engineering assessment.
How can false alarms be reduced?
Use stable references, commissioning checks, instrument-health rules, environmental and activity context, rate and spatial-consistency checks, and an agreed verification workflow. Preserve raw data and flag suspect readings rather than silently removing them.
Can one instrument confirm that underground works are safe?
No single reading proves overall safety. Monitoring supports risk management when it is combined with design review, construction control, inspection, competent interpretation and predefined response actions.
Discuss an underground monitoring scope for Hong Kong
Share the construction stages, ground and groundwater risks, nearby assets, required response time and reporting expectations. GEOOE can help structure an instrumentation and data-acquisition approach around those project needs.
Technical references
Sources were reviewed before publication. They support the specific guidance and external cases cited above; they do not imply GEOOE involvement in the referenced projects.
- [1] Civil Engineering and Development Department, Geotechnical Engineering Office. GEO Publication No. 1/2023. See the excavation monitoring discussion and instrument-selection considerations.
- [2] Civil Engineering and Development Department, Geotechnical Engineering Office. Amendment No. GG2/01/2017 to Geoguide 2. Tunnel investigation and design considerations.
- [3] Civil Engineering and Development Department, Geotechnical Engineering Office. Technical Guidance Note No. 49, Revision B, 5 February 2024. Deep-excavation ground-loss and sinkhole risk mitigation.
- [4] Conrad Ng, Peter Poon and Tim Leung. Use of NATM on Central–Wan Chai Bypass tunnel. Hong Kong Engineer, September 2016.
- [5] Hong Kong Institution of Engineers. Technical visit / activity report on Kwu Tung Station works. Hong Kong Engineer, November 2024.
- [6] Michael S. P. Wan and Jamie R. Standing. Lessons learnt from the installation of field instrumentation to monitor ground response to tunnelling. Crossrail Learning Legacy, 3 November 2014.
- [7] Crossrail Learning Legacy. Digging for monitoring gold: tunnel monitoring during Crossrail Paddington station box excavation.
- [8] Singapore Land Transport Authority. Thomson–East Coast Line. Official project information, including the Orchard monitoring account.
- [9] J-STAGE / Concrete Journal. Construction of the Keiyo Line Tokyo Underground Station crossing the existing Yokosuka Line shield tunnel, Volume 28, Issue 2, 1990. Original article in Japanese.
- [10] L. Ran, X. W. Ye and H. H. Zhu. Long-term monitoring and safety evaluation of a metro station during deep excavation. Procedia Engineering 14, 785–792, 2011. DOI: 10.1016/j.proeng.2011.07.099.
Editorial note: Technical application guide prepared for GEOOE. Instrument suitability, installation detail, performance requirements, trigger values and response actions must be verified for each project by the responsible qualified professionals.