Why tailings monitoring is non-negotiable now
The catastrophic tailings dam failure at Brumadinho, Brazil in January 2019 reset the industry's expectations overnight. Out of it came the Global Industry Standard on Tailings Management (GISTM), which raises the bar on governance, transparency and — critically — performance monitoring across a facility's whole lifecycle. For South African operators, a well-designed monitoring programme is now central to demonstrating conformance: it improves documentation and traceability, and ensures timely escalation when trigger levels are reached.
The engineering reason is simpler still: a tailings dam telegraphs distress. Rising pore pressure, creeping displacement, a dropping freeboard or a shifting phreatic surface all show up in instrument data before a failure. The problem with manual, periodic readings is the gaps between them — exactly when conditions can change. Real-time monitoring closes those gaps.
Important: monitoring supports — it never replaces — the engineer of record and your tailings management governance. The instrumentation plan, trigger levels and response actions must come from the responsible engineering professionals. This guide is about the IoT layer that makes their plan continuous and auditable.
What you monitor on a TSF
- Pore water pressure — piezometers at depth; rising pore pressure is a primary stability indicator.
- Movement & deformation — inclinometers, survey prisms / GNSS and displacement sensors track embankment movement and slope change.
- Phreatic surface — the position of the water table within the dam, inferred from piezometer arrays.
- Pond level & freeboard — level sensing to confirm you keep the required freeboard and manage the water balance.
- Flow & seepage — decant, return-water and seepage flows, plus rainfall, for the water balance.
The exact instrument set comes from the dam's design and failure modes — not from a generic checklist. The IoT job is to read those instruments reliably, continuously, and in one place.
Instrument by instrument
| Parameter | Instrument | What it reveals | GISTM / TARP role |
|---|---|---|---|
| Pore water pressure | Vibrating-wire piezometers at depth | Rising pore pressure = falling effective stress = the primary stability signal | Trigger levels per piezometer set by the engineer of record |
| Displacement | Inclinometers, survey prisms, GNSS, InSAR | Creep, toe bulging, crest settlement | Rate-of-movement triggers; escalation to evacuation |
| Phreatic surface | Piezometer arrays, standpipes | Where the water table sits in the wall | Compared against design assumptions |
| Pond level and freeboard | Level sensors, cameras | Overtopping risk, decant management | Operational trigger; storm-event response |
| Seepage flow and quality | Weirs/flumes with level, pH/conductivity | Internal erosion, changed flow paths | Environmental and stability signal at once |
| Rainfall and weather | On-site station | Storm loading context for every other reading | Event-based review |
How to set up real-time monitoring
- Work from the engineer of record. The monitoring plan, instrument locations and trigger levels follow the TSF design and the responsible engineer's requirements.
- Instrument the key parameters. Piezometers, inclinometers, survey/displacement, pond level and freeboard, and flow — connected to edge devices.
- Connect and buffer at the edge. TSFs are large and often remote, so low-power wireless with solar and edge buffering keeps data flowing across power cuts and signal gaps — no lost readings during the weather events that matter most.
- Set trigger levels (TARP). Configure green/amber/red thresholds tied to a Trigger Action Response Plan, each level with named owners and defined actions.
- Alert, record and report. Route exceedances to the responsible people in real time, and keep an auditable, exportable record for GISTM conformance and the dam-safety regime.
Why real-time beats periodic readings
Manual monitoring gives you a snapshot when someone walks the dam and reads the instruments. Real-time monitoring gives you the trend — and the alert — the moment pore pressure climbs or movement accelerates, including overnight, over a long weekend, or during the heavy-rainfall events that drive many failures. It also builds the continuous, time-stamped record that conformance and incident investigation depend on. The same wireless approach is cost-effective and removes the safety exposure of sending people onto a questionable embankment to take readings.
At addanode this runs on the in-house addaNet platform as part of our mining work — reading geotechnical and water instruments, buffering through outages on solar where needed, and alerting against your trigger levels, alongside occupational hygiene and mine water monitoring in one picture. Because we build the hardware and software in-house and support it locally, the monitoring layer fits your dam's instrumentation plan rather than forcing a generic kit.
This guide is general orientation on the IoT/monitoring layer. Tailings governance, instrumentation design, trigger levels and response actions must be set by the engineer of record and aligned to GISTM, the South African dam-safety regime and your appointed professionals.