| AI & digital twins
Left: Rocky Reach Dam in Washington state, US. Years of operational data records were collected and analysed to develop a digital twin © Antone Abbott Jr /
Shutterstock.com
The digital twin will help minimise the risk to perform on the real operation, such as load rejection, over-speed test and vibration at the unit start or stop stage. The team anticipates future projects expanding on the application of digital twins, addressing broader concerns such as biological buildup like sludge in coolers, along with the challenge in making them more environmentally friendly which are now common worries for many utilities. Looking to the future, the vision for hydropower
systems is a data-driven one where data analytics and predictive maintenance algorithms drive asset management. It is one where there are no unplanned outages and lost revenue but rather outages determined by data-driven maintenance schedules and equipment replacements. The project is funded by DOE’s Water Power
Technologies Office. For more details see https://
hydro.digitaltwin.labworks.org/.
Mining matters With limited application of digital twin technology
within the mining industry over the past few years, the concept has not yet been fully utilised in tailings dam safety management, maintenance, and cost- effectiveness. However, researchers from South Africa and Zimbabwe have been developing a conceptual framework for a smart tailings dam stability monitoring tool. The framework relies upon digital twin simulation and machine-learning (ML) techniques, and comprises four main components: real-time data collection, digital twin modelling, ML-based early detection and prediction, and intelligence-driven decision-support. Compared to traditional monitoring methods such
as routine inspections and time-based maintenance, this approach offers the advantage of not requiring direct human input for the acquisition, transmission and processing of data. As a result, much larger datasets can be manipulated and processed efficiently without any human error. Furthermore, the
proposed method does not suffer the limitation of requiring physical presence at a mining site in order for data and output information to be accessed. Since the model is virtual, many of its operations can be carried out or accessed remotely, including data transmission, processing and trend visualisation. Finally, the proposed approach allows for the cost-effective simulation of various parameter combinations to determine their collective impact on the stability of tailings storage facilities and potential for failure. There are several potential improvements and extensions that can be made to the proposed framework to enhance its performance and efficiency. As the authors Mwanza et al suggest, one
improvement could involve incorporating forecasts of weather elements, such as rainfall and wind speed, as well as seismic events, into the prediction model in order to improve its accuracy. Other improvements could be to incorporate site metadata, such as age of facility, dam dimensions, and material of construction into the model, as well as adoption of deep learning algorithms. This would likely improve the accuracy and reliability of the model even further. The authors believe their proposed framework is an important contribution in current efforts to mitigate against tailings dam failures in the mining sector. It has the potential to transform the way in which tailings dam stability is monitored by allowing for more accurate and real-time monitoring of structures and ground conditions. Adoption of the model could lead to safer and more efficient mining operations, as well as better protection of environment and mining communities. Compared to traditional monitoring methods, it offers the means for developing a more accurate and comprehensive model of the tailings dam, and can provide early warning detection for the prevention of catastrophic failures. Future research should focus on addressing the framework’s limitations and on developing strategies to make the framework more cost-effective and scalable for widespread adoption in the mining industry.
References
Park, D.; You, H. A Digital Twin Dam and Watershed Management Platform. Water 2023, 15, 2106.
https://doi.org/10.3390/ w15112106
Hydropower Digital Twins Solution Helps with Operator Challenges | News Release | PNNL
Mwanza, J; Mashumba P; and Telukdarie A. A Framework for Monitoring Stability of Tailings Dams in Real-time Using Digital Twin Simulation and Machine Learning. Procedia Computer Science 232 (2024) 2279–2288
www.waterpowermagazine.com | January 2025 | 17
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45
