When it comes to safety around tailings storage facilities (TSF), it is imperative to have an effective monitoring program in place.

According to Klohn Crippen Berger (KCB) associate and senior geotechnical engineer James Penman, the starting point for dam safety monitoring and management is understanding the critical failure modes for the structure.

“Dam safety failure modes are ways in which the dam fails, to the point that containment is lost and tailings and/or water flow out,” he said.

Catastrophic dam failures typically happen through the three major failure modes of overtopping, instability and internal erosion (or piping).

Image from a drone survey of a dam and its beach.

“Slope instability has been the most prominent failure mode in recent times, namely the Samarco and Brumadinho dam failures in Brazil.

“Overtopping is less prominent in tailings dams which, in contrast to water dams, should be purposefully run to keep the water level below the maximum safe limit; however, there are well documented failures where overtopping was a key failure mode, such as Merriespruit in South Africa and Mount Polley in Canada.


“The third failure mode grouping of internal erosion (or piping) – where a hydraulic gradient displaces material from the fills or foundations opening a ‘pipe’ to the impoundment – is less prevalent as a catastrophic failure mode for tailings dams, due to lower hydraulic gradients but still can occur and must be mitigated.”

The recent Global Industry Standard on Tailings Management (GISTM) includes dam safety monitoring as a specific principle (Principle No. 7) under the umbrella topic of “design, construction, operations and monitoring of tailings facilities”.

The GISTM, released by the Global Tailings Review last year, was a joint initiative by the International Council on Mining and Minerals (ICMM), the United Nations Environment Program (UNEP) and the Principles for Responsible Investment (PRI).

The Review was convened following the failure of the TSF at the Corrego Do Feijao iron ore mine near Brumadinho, Brazil, in 2019, in which 259 lives were tragically lost, with 11 remaining missing, and in which billions of dollars in clean-up costs and fines have been levied against the dam owners.

The GISTM aimed to establish a minimum standard for the construction, operation and management of tailings facilities worldwide. It recommends designing, implementing and operating monitoring systems to manage risk at all phases of the facility lifecycle, including closure.

KCB supported the forensic failure assessments of recent failures in Brazil, Australia and Canada, and the knowledge gained by KCB was reflected in feedback provided on the GISTM and in KCB’s approach to tailings management.

“Our involvement with the forensic analyses of recent dam failures – which drove the need for the GISTM – gave us unique insights into development of failure modes and potential warning signs to be monitored and managed” Mr Penman said.

“Our goal is to develop practical, effective monitoring programs to provide our clients – and potentially affected downstream communities – the peace-of-mind that their tailings facilities are operated in a safe, sustainable manner.”

There are two key reasons for setting up these monitoring systems. Firstly, it provides design confirmation by monitoring performance against the expected behaviour of the design – an approach known as the Observational Method and one of the key elements of the GISTM.

Secondly, it informs emergency contingency measures and responses by monitoring potential failure mode precursors.

Dams should have robust Trigger-Action- Response-Plans (TARP) which link to the monitoring and observation data through a series of pre-set trigger levels which then prompt the dam operators and owners to take necessary mitigative action.

These trigger levels are developed based on the performance objectives and risk management plans for these tailings facilities, describing actions to be taken if trigger levels are exceeded (performance outside the normal range), to prevent a loss of control in the system.

The range of actions should be pre-defined as far as practical, based on the magnitude of the exceedance of the trigger levels.

“In coming up with a monitoring program or system, you must have the confidence it can track things as they happen and be able to inform timely action so that you can implement preventative contingency plans and notify key stakeholders in an appropriate timeframe,” Mr Penman said.

A rainfall gauge at a decant pond.

A combination of site observations, manual inspections by operators as well as surface and sub-surface instrumentations is recommended.

“While there can be a tendency to overmonitor a dam with the latest gadgets and technology, it’s really about understanding fundamentally how the dam might fail and what you need to monitor to prevent this. As the saying goes, ‘what gets measured gets managed’,” Mr Penman said.

“For example, if the key failure mode is overtopping it is crucial to understand how much water you can store in the dam at all times, this would drive how often the pond level needs to be measured or whether you choose to utilise satellites or drones to provide daily data on the pond size.”

The focus should be quality of monitoring versus the quantity of monitoring.

“I think it’s much better to have two to three metrics you monitor really well –and that have clear linkage back to failure modes – rather than measuring 20 different things with the risk of overwhelming the site responsible engineer or designer with too much information to respond appropriately,” Mr Penman said.

“Risks and uncertainties can derail an operation but you can manage these well if you have a good monitoring program developed as part of your initial design and construction when the costs of setting up these systems becomes marginal.”

The current range of monitoring devices used by TSFs globally include piezometers, video cameras, satellite imagery, total stations and prisms, drones, inclinometers, wireline extensometers, laser scanners, tilt meters, fibre optics, radar and many more.

Operators on the ground are equally instrumental in the monitoring program of TSFs, rather than a single reliance on technology.

“I think operators with an understanding of dam failure modes are the most important part – we can have lots of data and information however it’s the people on the ground everyday who are best placed to see the changes and potential issues while they perform inspections,” Mr Penman said.

“This includes things like dam seepage and leaks that are hard to monitor via piezometers no matter how many you have.

“There might be several hundred metres between each instrumented dam section, and within that area there might be seepage flows through historical defects in the dam, such as road sheeting leftover from previous lifts, that would not be picked up by instrumentation.”

About Klohn Crippen Berger

KCB helped revolutionise the design of modern tailings facilities 60 years ago and today provides solutions for some of the largest, most technically challenging TSFs in the world.

Originally founded in Canada, KCB has done extensive work in Australia, Papua New Guinea, Indonesia, Peru, Brazil, North America and Canada.

Tailings dam management is at the core of what KCB does in Australia with about100 employees in its Perth, Brisbane and Newcastle offices.

The company also offers groundwater management, geochemistry, surface water management and mine closure services to the mining industry.

KCB has supported the successful design, operation, and closure of tailings facilities for more than half a century by focusing on providing clients a quality over quantity.