Tailings storage failures at Samarco, Mount Polley, and Brumadinho have focused intense attention on tailings dam safety. Technology is playing an increasingly important role in monitoring these facilities and identifying risks before they become disasters.

The Monitoring Imperative

Tailings storage facilities (TSFs) are among mining’s largest engineered structures. They store billions of tonnes of process waste behind embankments that can exceed 100 metres in height. When they fail, the consequences are catastrophic.

Traditional monitoring relied on periodic surveys, visual inspections, and point sensors. While these remain important, they have limitations:

• Infrequent data: Monthly or quarterly surveys miss rapid changes
• Point measurements: Individual sensors can’t characterise large structures comprehensively
• Human factors: Inspections depend on individual expertise and attention
• Limited coverage: Remote areas of facilities may receive less attention

Modern monitoring systems address these limitations with continuous, comprehensive data collection.

Satellite Monitoring

Satellite-based monitoring has become standard practice for major tailings facilities:

InSAR (Interferometric Synthetic Aperture Radar): Radar satellites measure surface displacement with millimetre precision. Regular passes create time series showing how facilities move and deform.

Optical monitoring: High-resolution satellite imagery supports change detection. Comparison between images reveals erosion, seepage, and other visible changes.

Coverage and frequency: Satellites can monitor facilities regardless of location or access. Revisit frequencies of days to weeks provide regular updates.

The CSIRO has been active in developing and validating satellite monitoring techniques for mining applications, including tailings facilities.

Ground-Based Sensor Networks

Satellite monitoring is complemented by extensive ground-based instrumentation:

Piezometers: Measure pore water pressure within embankments – a critical stability parameter. Modern systems report continuously via telemetry.

Inclinometers: Measure subsurface displacement. Movement within embankments can indicate developing instability before surface signs appear.

Seismometers: Detect ground vibration from various sources. Microseismic monitoring can identify internal erosion or other processes.

Weather stations: Precipitation and temperature affect facility behaviour. Correlating conditions with other measurements helps interpretation.

Weirs and monitoring wells: Track water flows and levels around facilities. Changes can indicate seepage or drainage issues.

Modern installations network these sensors, providing integrated real-time data to operators and engineers.

Drone and Autonomous Surveys

Uncrewed aerial vehicles (UAVs) have transformed site-level monitoring:

Photogrammetry: Regular drone surveys create detailed 3D models of facilities. Comparison over time reveals changes in geometry.

Thermal imaging: Thermal cameras detect temperature differences that can indicate seepage or internal processes.

LiDAR scanning: Laser scanning from drones provides precise elevation data for deformation analysis.

Visual inspection: Drones access areas that are difficult or dangerous for personnel. High-resolution imagery supports detailed visual inspection.

Some operations have implemented automated drone systems that fly programmed routes on schedule, providing consistent monitoring without operator involvement.

Data Integration and AI

The challenge with modern monitoring isn’t data collection – it’s making sense of the data flood:

Dashboard systems: Integrated platforms combine data from multiple sources into unified views. Operators can see facility status at a glance.

Trend analysis: Software tracks measurements over time, identifying changes that might escape notice in raw data.

Threshold alerting: Automated systems generate warnings when measurements exceed defined limits. Escalation protocols ensure appropriate response.

Anomaly detection: AI systems can identify unusual patterns that don’t trigger specific thresholds. Machine learning recognises deviations from normal behaviour.

Correlation analysis: Advanced systems identify relationships between different measurements, helping engineers understand facility behaviour.

team400.ai working with mining companies are developing increasingly sophisticated systems that learn from historical data to identify emerging risks.

Regulatory Evolution

Tailings monitoring requirements have tightened significantly following recent failures:

Global Industry Standard on Tailings Management: Published in 2020, this standard establishes requirements for monitoring, governance, and disclosure that are becoming industry benchmarks.

Independent review requirements: Many jurisdictions now require independent expert review of tailings facilities and monitoring systems.

Public disclosure: Companies are increasingly required to disclose tailings facility details and monitoring results to communities and investors.

Consequence classification: Facilities are classified by potential consequences, with higher-consequence facilities subject to more stringent requirements.

These regulatory changes are driving investment in monitoring technology and capability.

Implementation Challenges

Advanced tailings monitoring faces practical challenges:

Legacy facilities: Older facilities may lack instrumentation provision, making retrofits difficult and expensive.

Remote locations: Many tailings facilities are in areas with limited power and communications infrastructure.

Data management: Continuous monitoring generates enormous data volumes. Storage, processing, and archival require careful planning.

Skills requirements: Interpreting modern monitoring data requires expertise in remote sensing, data science, and geotechnical engineering.

Alert fatigue: Systems must be calibrated to avoid excessive false alarms that lead operators to ignore warnings.

Beyond Monitoring: Predictive Capability

The frontier in tailings safety is moving from monitoring to prediction:

Stability modelling: Numerical models incorporating real-time monitoring data can assess stability continuously, not just during periodic reviews.

Failure mode analysis: AI systems are being developed to identify precursor patterns associated with specific failure modes.

Weather integration: Forecasting systems can assess how predicted conditions might affect facility stability.

Risk quantification: Probabilistic approaches provide ongoing risk assessments that support prioritisation and decision-making.

The goal is early warning – identifying developing problems with enough lead time for intervention.

Industry Progress

The mining industry has invested heavily in tailings monitoring since recent high-profile failures:

Major mining companies have hired specialist staff, implemented new monitoring systems, and increased governance oversight.

Technology vendors have expanded product offerings, with several new satellite monitoring services and integrated platform providers entering the market.

Independent tailings review boards have become common, providing ongoing expert oversight of monitoring and operations.

These investments are improving safety, but the work is ongoing. Tailings facilities require perpetual attention, and monitoring systems must evolve with technology and understanding.

The industry’s goal is zero tailings failures. Technology is an essential tool in achieving this, but it must be combined with sound engineering, robust governance, and genuine commitment to safety.