Tailings Management – Dewatering of Slurry Tailings at Disposal Site

Tailings dams are an essential part of mining, mineral processing and power generation. However, risks including a catastrophic dam collapse mean that tailings dams present serious safety, environmental and social consequences. Most risks, however, can be alleviated through reducing the volume of water in the tailings dam. Furthermore, any reduction in water stored in the tailings dam results in more sustainable use of water resources.

This project investigated the feasibility of dewatering tailings at the CHPP tailings dam through a passive open channel, or flume. A pilot-scale flume of 0.1m width and 12.6m length was built and tested under three configurations and at different flow rates of material. The designs all consisted of four 3m long flume sections separated by four 0.15m weirs. Additional configurations considered weir roughness and collection hoppers prior to the weirs. The experimental results indicated that weir roughness and collection hoppers do not influence the performance under steady-state operation.

A steady-state model of flume performance was then developed utilising the particle size distribution as an input. The model assumed Stokes' settling and constant solids density and fluid viscosity, after which the measured results could be used to solve for the unknown constant(s). Good agreement was obtained between measurements and prediction and the model was successfully validated for a second sample of tailings. The good agreement supported the assumptions in the development of the model. Mathematical analysis of the flume model showed that a constant level of settling performance is achieved for a constant ratio of throughput to flume footprint area. Furthermore, for most practical cases, the flume footprint area was shown to grow exponentially with increasing settling performance.

It was also shown that the model could be applied in the development of flume design charts, the result of which achieved the objective of the project. However, the advantage of a mathematical model also allowed for a direct analysis on the parameters influencing settling performance of the flume. Mathematically, the particle diameter was the most sensitive parameter; however, the experimental work also demonstrated that the slurry viscosity may vary substantially between samples, which was successfully captured by the model.

To address these potential sources of optimisation, the Boycott effect and flocculation were considered through testing conducted in a scale-up flume. The scale-up flume was 0.3m wide and contained four 2m flume sections. These were separated by four 0.30.3m settling chambers which contained 55 mm inclined tubes. It was found that channel width should not exceed 0.1m due to the obtained velocity profile promoting an undesirable biased settling near the flume wall. The inclined tubes tested were found to have negligible impact on performance due to their narrow width potentially resulting in shear instabilities and mixing. However, when flocculation was also considered in combination with the inclined tubes, there was a substantial increase in flume performance over the isolated use of either inclined tubes or flocculation.