Both the short and long-term behaviour of tailings are mostly determined by the dewatering of the different phases present and in particular by those of the smallest particle size. Declining ore grades lead to the processing of higher tonnages, and combined with finer grinding for values liberation, mean that tailings management is increasingly challenging (Edraki et al., 2014). In the short-term this is reflected in the greater difficulty in returning water to the circuit, the greater risk arising from storing larger volumes of dilute tailings, and increases in associated costs, but there are also longer-term issues in mine closure and rehabilitation.

Recent focus given to tailings has come after catastrophic TSF failures, raising concerns with such facilities all around the world and the tailings stored within them. In reviewing such responses, Santamarina et al. (2019) states:

“This liquefaction of the impounded materials may suggest to regulators and the public that the problem lies with the impounded materials themselves. However, in the absence of internal collapse or induced shear (e.g. because of a seismic event), liquefaction and outflow of ponded ash and tailings occur after the dam has failed. Thus, liquefaction does not cause the failure, but rather the disaster that follows.”

Tailings storage facility (TSF) failures Catastrophic dam wall failures are usually either a civil engineering issue in their design and construction, or a consequence of poor management practices. Chronic issues with walls may be caught early with better monitoring. Of course, seismic activity does occur and machinery can also cause tailings liquefaction. However, addressing TSF structural issues should not distract from separate efforts to influence tailings properties such that their potential for liquefaction is reduced, and the size of TSFs themselves may be scaled-back. Tailings research with CSIRO Mineral Resources has been towards this objective.

Defining tailings

Tailings consistency varies considerably depending on the particle concentration in water:

Low solids concentration (low density) slurries that segregate when flowing. Thickened to higher slurry concentrations that are non-segregating.

Pastes that have stress and flow like toothpaste when discharged from pipes. Cakes formed from filtration, no longer exhibiting any fluid-like properties.

Particle size distributions are broad, but sizes <20 µm (particularly clays) are most problematic, being harder to dewater. Water within tailings slurries is often more useful than the solids, which are typically of low value. While past inefficiencies in precious metals processing has led to success from re-processing old tailings (e.g. for gold or platinum), the presence of trace metals or residual value phases in tailings rarely justifies re-processing without other circumstances being highly favourable). Water quality can also be highly variable, and reagents or other dissolved species potentially building up on re-use and impacting on any discharge into the environment.

Where CSIRO tailings research fits in

The life of a typical mining and processing site, from exploration through to post-closure activities, is represented in the image below. Tailings are clearly a major aspect from Operation onwards, but will also need to be considered during Planning and design (and therefore Construction). While Operation is shown as only one of six stages, it actually encompasses so many other sub-stages within the process flowsheet, summarised here as going from Mining through to the Tailings circuit, which critically returns water to the Hydrometallurgy/Processing circuit. The latter circuit and the steps prior all can impact on the volume and properties of the final tailings that need to be treated. Ore sorting seeks to enhance the grade of the material then processed, but any elimination of a significant fraction of gangue solids also reduces the volumes going to tailings. The liberation of values during Comminution is important towards maximising recoveries, but fine grinding can then adversely affect tailings properties; controlled grinding should achieve liberation without greatly elevating tailings slurry volumes or the flocculant dosages required to treat them. Hydrometallurgical routes are often fixed in place early during Planning and design, and limited consideration of tailings impacts leads to adoption of conventional processes; Proactive leaching refers to novel hydrometallurgical routes (e.g. using alternative lixiviants) that become more viable or attractive options when their positive impacts on tailings properties are able to be recognised and quantified in the planning phase.

The primary method for solid-liquid separation in tailings circuits is by flocculation and gravity thickening, processes in which CSIRO has unmatched expertise and capability. CSIRO has been world-leading in using computational fluid dynamics (CFD) to enhancing feedwell flocculation in gravity thickeners, successfully applied to full-scale operations around the world, mainly for tailings/residue circuits. Systematic CFD studies of feedwells have led to design rules that have formed the basis of appraisals, as well as the ability to produce advanced performance response surfaces that may form the basis for feed-forward model-based control.

Part of the success from feedwell CFD was drawn from a deep understanding of flocculation fundamentals, with >25 years of CSIRO study in this area highlighting how the key sensitivities affecting tailings flocculation are distinct from traditional principles developed in academic colloid chemistry studies of wastewater clarification. This knowledge and the tools developed for performance characterisation are now being used to examine the impact of alternative polymers and high salinity process waters, which is also linked to advanced rheological characterisation of the flocculant solutions. The highly viscoelastic flocculant solutions significantly impact upon flocculant mixing through solutions and slurries, on which CSIRO has conducted the first detailed measurements and modelling, as well as proposing changes to sparge geometries to promote mixing. Feed-forward control is also predicated on having suitable feed sensors, which CSIRO continues to examine. Many operations attempt to control to a target underflow solids concentration, but thickener performance is more productively related to underflow rheology, for which CSIRO has developed a robust instrument we believe gives much more detailed information than commercially-available alternatives. CSIRO also has extensive capability in the area of high solids tailings transport, most notably the current development of a sensor for pipeline bed load detection.

Filtration and centrifugation are now gaining considerable attention, either after thickening or even as alternative options for tailings treatment. The success of both for tailings treatment is also strongly influenced by flocculant mixing, which becomes even more problematic at higher solids concentrations. CSIRO has studied topological mixing as a way to achieve more efficient distribution of flocculant and aggregation at high solids, both as a large-scale continuous process and as a small-scale test for understanding polymer impacts on deposition.

References

• J.S. Adiansyah, M. Rosano, S. Vink, G. Keir (2015). A framework for a sustainable approach to mine tailings management: disposal strategies, Journal of Cleaner Production 108, 1050-1062.
• M. Edraki, T. Huynh, T. Baumgartl, L. Huang, M. Andrusiewicz, K. Tungpalan, M. Tayebi-Khorami, E. Wightman, S. Palaniandy, E. Manlapig (2014). Designer tailings–an integrated model for tailings management. Proceedings, Life of Mine 2014 Conference, Brisbane, pp. 16-18.
• J.C. Santamarina, L.A. Torres-Cruz, R.C. Bachus (2019). Why coal ash and tailings dam disasters occur. Science 364(6440), 526-528.