Tailings dams are among the largest engineered structures on the planet, yet they are often monitored using discrete, point-based sensors. This approach can leave uncertainties on dam performance where no sensors are installed. As ore grades decline and demand for critical minerals grows, mines are producing ever-increasing volumes of tailings. In turn, this increased production in tailings is driving an increasing number of tailings dams failures (Hudson-Edwards et al. 2024). With each failure, the mining industry faces increasing pressure to maintain its social license to operate globally.
Recent advances in distributed fibre optic sensing have made it possible to transform standard telecommunications-grade fibre optic cables into thousands of strain and seismic sensors extending over kilometres. Fibre optic sensing works by sending pulses of light along an optical fibre. As these pulses travel, natural imperfections in the glass fibre scatter a portion of the light back toward the recording device, known as an interrogator. Distributed acoustic sensing (DAS) is one form of fibre optic sensing that uses the phenomenon known as Rayleigh backscattering. By analysing the backscattered light pulses, DAS enables continuous, high-resolution monitoring of physical changes in the surrounding environment.
Lumidas, a technology startup founded on Susanne Ouellet’s doctoral research at the University of Calgary, is harnessing the power of DAS to monitor the performance of tailings dams. Her work, supported by collaborations with multidisciplinary leaders in seismology, fibre optic sensing, and geotechnical engineering, is informed by over a decade of experience in both research and consulting. Ouellet was driven to explore DAS technology in greater depth following the catastrophic 2019 Brumadinho tailings dam failure in Brazil, which claimed over 270 lives. Although the dam was equipped with an array of monitoring systems, none detected signs of instability prior to the collapse.
At the time, Ouellet was working as a geotechnical engineer at BGC Engineering, evaluating technologies to support an emergency response system at a tailings storage facility in northern Canada. The Brumadinho disaster served as a catalyst for her decision to investigate the potential of DAS to fill critical monitoring gaps in the industry.
Seismic Coda waves
Ouellet’s four-year PhD research programme, in partnership with a mine operator, BGC Engineering Inc., LUNA OptaSense, and Mitacs, tested DAS technology in real-world conditions at an upstream tailings facility in Saskatchewan.
Passive seismic techniques like coda wave interferometry (CWI), initially performed at the site using geophones (Ouellet et al. 2022), were shown to work equally well with fibre optics (Ouellet et al. 2025). CWI helps geotechnical engineers to monitor tiny changes in subsurface properties over time. It works by analysing the “coda” portion of seismic signals. These are the later-arriving waveforms produced by multiple scattering events as waves travel through a heterogeneous medium. These scattered waves are highly sensitive to small changes in the material, such as shifts in mean effective stress or variations in moisture content (Snieder, 2002; Grêt et al., 2006).
A key advantage of CWI is that it can leverage the ambient seismic wavefield, generated by natural or anthropogenic sources such as machinery, traffic, or wind, as its energy source. By cross-correlating seismic recordings between pairs of sensors, an approximation of the response of the medium can be extracted. When subsurface conditions change, these alterations appear as small shifts in the travel times of coda waves, which can be quantified as relative changes in seismic velocity (dv/v).
Tailings storage facilities are rich in anthropogenic noise: haul trucks, conveyors, crushers, drilling, and blasting all produce persistent vibrations. While such variable noise environments might pose a challenge for traditional monitoring techniques, CWI only requires a portion of the noise field to remain stable to reliably track dv/v changes (Hadziioannou et al., 2009).
Laboratory experiments have demonstrated that CWI can detect mean effective stress changes smaller than 0.1% in uncemented soils (Dai et al., 2013). This level of sensitivity holds potential for early detection of evolving instability in tailings dams, where small stress variations may signal the onset of structural weakening. Ouellet et al. (2025) applied CWI to a 120 m segment of installed fibre optic cable at the site using DAS to track dv/v changes up to 2% during periods of spring thaw and rainfall. These variations were interpreted in the context of environmental factors such as freeze-thaw cycles, tailings pond level fluctuations and precipitation. Despite the presence of active construction noise, stable ambient noise cross-correlations were obtained using advanced denoising techniques, allowing reliable daily measurements using as little as one hour of data per day. Depth sensitivity analysis suggested that the observed dv/v changes were most pronounced at depths between 7 and 14m, corresponding to the transition zone between the dam fill and underlying foundation.
This research is among the first to integrate DAS and CWI for tailings dam monitoring, highlighting the potential of DAS to dramatically increase spatial coverage, offering thousands of sensors along a fibre optic cable. Such monitoring could inform early warning systems and support risk mitigation efforts by detecting precursory changes that conventional point sensors may miss.
Revealing hidden failure
In collaboration with the British Geological Survey and Luna OptaSense, Ouellet’s doctoral research (Ouellet et al. 2024) also showcased the power of DAS to reveal previously undetectable landslide processes at the Hollin Hill Landslide Observatory in the UK. By repurposing a buried 925m fibre optic cable into a dense network of DAS strain sensors, the team achieved nano-strain sensitivity and sub-minute temporal resolution, capturing near-surface deformations with less than 1mm of displacement during a three-day rainfall event.
Five key landslide processes were identified: the initiation of strain at the scarp, triggering of a rupture zone, retrogression toward the slope crest, a flow-lobe surge at the toe, and subsequent stabilisation. These observations were derived from strain-rate spatiotemporal images that resolved dynamic slope behavior at the meter scale, extending beyond the capability of conventional geotechnical or remote sensing methods. The study further validated DAS-derived displacement estimates by comparing them to collocated inclinometer data, with results in strong agreement. These results illustrate how distributed fibre optic sensing could help reveal new insights into tailings dam behaviour, tracking both slow deformations and sudden changes with high fidelity.
Making DAS practical for tailings dams
DAS adoption is not yet widespread for geotechnical monitoring in the mining industry for tailings dams, which may be due to the perceived complexity and high initial investment. That’s where Lumidas comes in. Lumidas is developing purpose-built data processing pipelines to translate DAS data into intuitive dam performance indicators. These include strain anomalies, seismic velocity changes, and other early warning signals. Outputs are visualised through dynamic heat maps and spatiotemporal overlays on a web-based platform.
The system is currently under development in a two-year pilot project in collaboration with multiple industry and academic stakeholders, including Teck Resources Limited, BGC Engineering Inc., Norwegian Geotechnical Institute, University of Calgary, Nerve-Sensors and FEBUS Optics. This initiative is focused on creating a scalable, cloud-based monitoring framework adaptable to dams of all sizes and geographies.
The Lumidas monitoring system is designed for tailings dam operators, mining companies, engineers, regulators, and researchers. It provides a shared platform for managing risk, improving operational decisions, and increasing transparency with stakeholders. Whether deployed in remote northern Canada or in densely populated mining regions, the system is designed to scale. Future developments will include expanding the range of fibre optic deployments to different dam geometries and automating anomaly detection with AI, alongside future pilots designed to further validate and benchmark the platform under a variety of field conditions.
References
Dai, S., Wuttke, F., & Santamarina, J. C. (2013). Coda wave analysis to monitor processes in soils. Journal of Geotechnical and Geoenvironmental Engineering, 139(9), 1504–1511. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000872
Grêt, A., Snieder, R., & Özbay, U. (2006). Monitoring in situ stress changes in a mining environment with coda wave interferometry. Geophysical Journal International, 167(2), 504–508. https://doi.org/10.1111/j.1365-246X.2006.03097.x
Hadziioannou, C., Larose, E., Coutant, O., Roux, P., & Campillo, M. (2009). Stability of monitoring weak changes in multiply scattering media with ambient noise correlation: Laboratory experiments. The Journal of the Acoustical Society of America, 125(6), 3688–3695. https://doi.org/10.1121/1.3125345
Hudson-Edwards, K., Kemp, D., Torres-Cruz, L.A., Macklin, M.G., Brewer, P.A., Owen, J.R., Franks, D. M., Marquis, E., & Thomas, C.J. (2024). Tailings storage facilities, failures and disaster risk. Nat Rev Earth Environ 5, 612–630 (2024). https://doi.org/10.1038/s43017-024-00576-4
Ouellet, S., Dettmer, J., Mikesell, T. D., Lato, M., & Karrenbach, M. (2025). Tailings dam performance monitoring by combining coda wave interferometry with distributed acoustic sensing. ASCE Journal of Geotechnical and Geoenvironmental Engineering. https://doi.org/10.1061/JGGEFK.GTENG-13066
Ouellet, S., Dettmer, J., Lato, M. J., Cole, S., Hutchinson, D. J., Karrenbach, M., Dashwood, B., Chambers, J. E., & Crickmore, R. (2024). Previously hidden landslide processes revealed using distributed acoustic sensing with nanostrain-rate sensitivity. Nature Communications, 15, 6239. https://doi.org/10.1038/s41467-024-50604-6
Ouellet, S.M., Dettmer, J., Olivier, G., DeWit, T. & Lato, M. Advanced monitoring of tailings dam performance using seismic noise and stress models. Commun Earth Environ 3, 301 (2022). https://doi.org/10.1038/s43247-022-00629-w
Snieder, R. (2002). Coda wave interferometry and the equilibration of energy in elastic media. Physical Review E, 66(4), 046615. https://doi.org/10.1103/PhysRevE.66.046615
