How ore sorting can be beneficial to tailings

CSIRO has developed sensors for large-scale ore sorting that rapidly determines ore grade to cost-effectively sort minerals of value from waste, and this clearly offers the advantage of a lower overall mass throughput to processing circuits. An ongoing aspect of ore sorting research is to recognise distinctions between ores that will lead to good or problematic tailings, the latter likely to contain higher proportions of clays (in particular clay particles with finer sizes or high aspect ratio shapes, such as smectites, that are inclined towards gelation in suspension). The problematic phases greatly reduce settling fluxes and produce low solids underflows; this again affects throughputs, especially in washing circuits. At some operations mine mapping leads to such ores either being stockpiled as too difficult, not mined at all or being blended with better behaved ores. If ore sorting could reliably identify problematic phases that lead to poor tailings behaviour, it’s possible ore sorting could provide a better basis for ore blending that then enables more productive treatment within existing thickening capacity.

What do plant operators need to do to adapt to changes in ore quality entering the plant?

Most unit operations work best when receiving consistent feeds. From a flotation perspective, higher quality shouldn’t require much adaptation, but thickening can require some adjustment. If tailings come from flotation at a lower solids concentration, that’s potentially a significant advantage, given most operations are solids dilution deficient prior to feedwell flocculation – this could contribute to reduced flocculant consumption and improved quality of the water being returned to the circuit. However, if an existing thickener is designed to deliver that prior to flocculation, there is potential for over-dilution, which can be detrimental to performance. Reduced volumetric flows to thickeners can also then fall outside the optimal feedwell operating range (flows that are too low can be as bad for flocculation as too high). This can require changes to dosing strategies and possibly even some minor design changes, but operationally it is easier to deal with low flows than with high flows.

When ore sorting is predominantly aimed at delivering higher grades, it could potentially mean tailings that contain more problematic gangue phases, which then may require adjustment to how tailings are treated. It will be on a case-by-case basis, and some characterisation of how the tailings responses change would be required, but overall positive benefit would still be expected. Demonstrating the potential to sort ores with an aim to actually minimise downstream tailings problems would be more challenging, as it would require considerable testwork to prove such benefits can be achieved while maintaining product recoveries. This would also only be of value when a site has a strong desire (or need) to find a way to effectively deal with substantial reserves of an otherwise difficult-to-process ore.

While ore sorting provides the potential for water savings, high throughputs and tailings storage, such benefits can only be realised in applications where these issues are constraints. Using Chilean copper operations as an example, their tailings circuits are typically designed to achieve a thickener underflow solids concentration target that would then deliver a specified water return, but often the thickeners don’t achieve that target. Lower volumetric/mass flows of tailings could allow longer residence times within thickeners, and therefore more chance of achieving underflow targets.

CSIRO has extensive experience in optimising the performance of existing thickeners, often where they are designed for a specific range of operating conditions or a feed type, but now have to deal with something much different. We have unmatched capability for quantifying flocculation performance and identifying the best reagents for an application, and we combine that with a range of tools for measurements on full-scale units and advanced computational modelling capability to then predict the best possible design modifications. This can be used to both show the potential for benefit from ore sorting and then optimise the tailings circuit to ensure this benefit is fully realised.