The evolution of monitoring tailings dams across the globe

By Paul Davies, Commercial Manager, Australasia, GroundProbe Australia

How the mining industry monitors tailings dams has evolved over the years with advancements in technology, AI and research and development. In this article, GroundProbe’s Paul Davies discusses the approach to preventing tailings dam disasters with expert, Leo Probst. Here, they reflect on the past catastrophic tailings dams events and how new advancements in technology can ensure optimal safety and reduced risk for both miners and communities.

I recently caught up with Leo Probst, Senior Geotechnical Engineer - Research and Development at GroundProbe about what the industry has learnt about the monitoring of tailings dams since the fatal Samarco dam collapse in Brazil, in 2015.

Currently supporting the research and development (R&D) of new products for tailings dam monitoring, Leo’s industry experience spans various years across mines in South America. Prior to joining the GroundProbe team in 2019, Leo was an adopter of our technology and is working support progress when it comes to monitoring tailings dams.

During our chat we explored the ways the industry has changed with the adoption of new technologies and systems to help miners decrease their risk. The key takeaways are:

• Tailings dams can have complex structures and require more monitoring than previously considered;

• We need systems that can analyse both long-term and short-term trends, including integration of reporting with the mine’s Triggered Action Response Plan (TARP);

• Robust infrastructure and communications are essential to enable mines to make timely decisions. Communities can be impacted negatively when monitoring systems are not robust.

Example One: The Samarco dam collapse, 2015

The Samarco dam collapse, also known as the Mariana dam disaster, occurred on November 5, 2015, in Brazil. It was a catastrophic environmental disaster involving the failure of the Fundão Dam, which was an iron ore tailings dam located near the town of Bento Rodrigues in the state of Minas Gerais.

What happened:

• An estimated 32 million cubic meters of toxic slurry was released to the nearby Bento Rodrigues and several other downstream communities.

• It resulted in the deaths of 19 people, and it was one of the worst environmental disasters in Brazilian history.

• The mining companies invovled have faced significant legal and financial implications.

Example Two: The Brumadinho tailings dam collapse, 2019

The Brumadinho tailings dam collapse was a catastrophic event that occurred on January 25, 2019, in Brazil. To date, this was the deadliest mining disaster in the country’s history.

What happened:

Photo Source: By Ibama from Brasil - Brumadinho, Minas Gerais, CC BY-SA 2.0.

• The dam collapse released an estimated 12 million cubic meters of mud and mining waste into the nearby area.

• The disaster resulted in the deaths of at least 270 people, including mine workers and nearby residents.

• The disaster led to significant outrage and legal consequences for the companies involved. There are now stricter regulations and better oversight of mining activities in Brazil. The recovery effort after the collapse was complex and dangerous due to the unstable terrain and the risk of further collapses.

• The dam disaster is a reminder of the risks associated with tailings dams in mining operations and the importance of stringent safety measures, regular inspections, and environmental safeguards.

Q&A: Looking at the past to prepare for the future

Paul: What was your perspective with monitoring tailings dam failures versus open pit mining failures?

Leo: At the time the industry had almost 20 years’ experience with geotechnical radar monitoring of open pits, but much less radar time in monitoring for dams. The mechanisms for failure between the two types are vastly different.

Paul: Did you consider monitoring of tailings dams to be higher risk than open pits?

Leo: Yes, we needed to consider whether monitoring of dams with radar was too high a risk. It was decided to work with customers and develop solutions for their requirements.

There was room for improvement in the way dam structures were being monitored. You could not drag out the proven techniques that worked in open pits and apply them to tailings dams. They had different failure mechanisms and the technology available at the time was not always adequate.

Paul: Can you give me an example of these differences?

Leo: Sure. Some believed at the time that tailings dam failures could occur without precursor signs. It is now clear that precursor indications are likely to be demonstrated in a structure prior to a collapse.

Traditionally, monitoring data was collected once a week, a month, or every six months. These technologies could not give us the information we needed. We started to develop better procedures to monitor tailings dams to understand the risks and monitoring design for long term, slow deformation, broad areas.  

In the past, real-time monitoring of tailings dams was not common. InSAR (Interferometric Synthetic Aperture Radar) satellite readings were taken every 11 days. Measurements of in-ground sensors would also be periodic. It would take months of collection for a report to be produced. These reports gave back-analysis but were insufficient to capture fast-moving changes.

A dam can change from being long-term stable, to a fast-moving quick failure. Upstream tailings dams are especially risky. The solution needs to cover both real time and long term.

Paul: How did you develop a solution?

Leo: We addressed this in two ways. Firstly, we brought in sensors that could measure more frequently. For example, ground-based SAR (Synthetic Aperture Radar) and other sensors.

We provided more analysis tools to review long-term and short-term trends. This allows us to act quickly if something went wrong.

Unlike open pit mines, where the solution would be obvious, each tailings dam requires a customised approach to cover both scenarios of long-term and critical applications. Integration of the sensor reporting with the TARP would help us manage the combination of risks and confirm these trends.

Paul: How did these changes affect mine operations?

Leo: This is very important to highlight. We learnt the importance of understanding the infrastructure of communication supporting a tailings dam. This was a challenge because tailings dams are isolated and lack both power and communications.

Paul: How can you transmit data from a high-definition sensor, prepare the data in a useful way, and then communicate that information so that operators can make a considered decision to manage their risk?

Leo: The system for collection and transmitting this data needs to be robust and cannot stop for any reason.

Paul: I would imagine if the customer lost communication with their data the impact would be big.

Leo: Absolutely. We started monitoring after the collapse two years ago in South America. There was a risk that a dry stack of material would affect community infrastructure outside the mine causing significant disruption.

There was an additional complexity with this collapse with the interruption to electricity supply. The radars were not affected, because they rely on autonomous power. However, the mine communications network was going to turn off. Our team could not provide 24/7 monitoring. So, we provided a person on site with power to ensure the monitoring continued.

Paul: Any final thoughts?

Leo: There are now a lot of players working together. The diversity of suppliers is good, however makes things more complex. Everyone including miners, regulators, governments, communities and suppliers are connected. We can all learn something.

Conclusion

Providing community support is important. When communities live nearby, they need to be supported in knowing how to evacuate.  

For example, the TARP (Trigger Action Response Plan) for a tailings dam may require an entire community, or even city be evacuated if the alarm is triggered. These evacuations are usually more complex than at an open pit mine where only the mine operators need to evacuate. There can be implications when false alarms (or unwanted alarms) occur, causing widespread disruption to communities.

Regulators can now check for themselves using InSAR without sensors on the ground. Eventually the information will be online for anyone to have access – miners, regulators, and communities. Everyone can look at this data and benefit from the information.

One final reflection. We believe the Internet of Things (IoT) is the future for everything related to tailings dams. We need to increase our understanding of the behaviour of these structures, to ensure stability and safety across the board, for miners and communities.