Flocculants and tailings

Can flocculant choice have a major impact on tailings treatment?

by Dr Chris Vernon

The introduction of synthetic, high molecular weight, water-soluble polymer flocculants was a turning point in tailings dewatering within the minerals industry.

Gravity thickening had previously relied on coagulants and natural products (e.g. starch) to aggregate particles and accelerate settling.

Synthetic flocculants enable much larger aggregates to form, leading to much higher mass throughputs, as well as the potential to greatly reduce thickener diameters.

In the decades since their introduction, flocculant molecular weights increased, the range of charge densities available expanded, make-up was facilitated by offering products in different physical forms and alternative functionalities were explored.

Cationic products are common in wastewater treatment, but higher molecular weights are more accessible for anionics, and the copolymers of acrylamide and acrylates still dominate mineral tailings applications to this day.

Improved flocculant chemistry doesn’t equate to increased advantage

Each week there are literally dozens of papers or patents on preparing and demonstrating new or improved flocculant products with different functional chemistries, highlighting there are a number of ways to promote particle aggregation.

However, just because a product can work as a flocculant doesn’t mean a need is then addressed or an advantage derived from its full-scale application.

This is a common flaw in many otherwise admirable synthetic studies, and manifests in two ways:

• Poor techniques applied during flocculation and in quantifying performance, leading to results that are hard to interpret and don’t adequately capture relevant behaviour.
• Poor appreciation of the impact of flocculant polymers across the tailings treatment process, from feedwell flocculation and bed compaction in thickeners through to tailings discharge. It’s much more than which product gives the highest settling rate at the lowest dosage.

When does flocculant selection make a difference?

Let’s consider a few scenarios where flocculant selection can make a difference.

1. Feeds needing high solids dilution prior to flocculation

Fine particle flocculation leads to fractal-like aggregate structures, meaning that when they are large, they are highly porous and fragile.

Large, fast-settling sizes often requires solids dilution, particularly in tailings with clays present.

Achieving the required lower solids concentration prior to flocculation with water from the thickener’s clarification zone can involve eductors or forced (pumped) dilution, but can lead to excessive shear during flocculation.

But what if a different flocculant could increase aggregate density?

A more compact aggregate structure means more mass in the same effective volume, so requirements for solids dilution are then reduced. Given many applications struggle to get such dilution, this could be significant.

Common acrylamide-acrylate copolymer flocculants typically don’t offer much variation in aggregate density, but incorporating a different functionality can achieve this, as we’ve shown in a detailed laboratory study of flocculation kinetics for a model feed.

Those kinetic results were then used in computational fluid dynamics (CFD) predictions of aggregation with the two flocculants in an open-with-shelf feedwell design.

The CFD image below is a snapshot in time for just one condition. It clearly shows a wide variety of particle/aggregate paths.

This modelling study considered 50000 paths for each set of conditions, which included aggregate growth and breakage achieved with each flocculant at the same dosage at four different solids concentrations.

CFD then predicted aggregate settling rates as they passed through and out the feedwell (see below).

The flocculant known to produce larger aggregate sizes at a higher solids volume fraction in lab tests (BASF Rheomax^® DR 1050) was indeed predicted to greatly increase mass throughput, but only at high feed concentrations and only dosages that could produce fast settling.

Therein lies the key point – the application requires denser aggregate structure.

When dosages are low and aggregates are not large, the effective volume of flocculated solids may be insufficient to have a major negative impact on flocculation.

This is the case for most older thickeners, where large tank diameters (and hence low liquor rise rates) create less demand on settling flux.

Solids dilution is a much greater priority in modern paste thickeners, for which it is hard to imagine low settling flux conditions being applicable.

2. Maximising thickener underflow solids concentrations

Seeking to control thickeners to an underflow solids concentration is common practice, when in reality it’s the underflow rheology that reflects operating conditions within the thickener.

In the schematic above, the grey curve illustrates the standard yield stress response to solids concentration for unflocculated solids, with an exponential rise from a quite fluid state to cake-like behaviour.

The yield stress obtained from a thickener (and hence where you are on such curves) depends on the thickener type (dimensions, rakes and rake drive), operating conditions (bed height, throughput) and the aggregation state.

For the latter, the curve is always pushed left to lower solids, as represented by ‘Flocculant 1’ for typical operation.

While aggregates won’t be as large as first formed in the feedwell, significant structures will persist in beds and through to the underflow.

However, structures will be further degraded on pumping, with the vertical dashed line showing potential for a sharp drop in yield stress.

An underflow that may have initially had the ideal consistency for paste disposal then effectively becomes water-like.

A flocculant that produces a denser or weaker structure (the latter possibly from requiring a lower dosage) could lead to the response illustrated for ‘Flocculant 2’, with a higher solids concentration achieved.

Such solids will still be thinned on transport, but because the concentration corresponds to a steeper portion of the unflocculated solids curve, the suspension retains more of its yield stress, i.e. it won’t become water-like.

Is this really significant?

The difference in solids concentration may only be in the order of 1 wt% (larger if combined with optimised feedwell flocculation).

However, it can be crucial towards being able to deposit tailings as paste. It is unlikely to impact noticeably on the volumes of water returned to the process from thickening, but it is an advantage to deliver solids to a TSF at a higher concentration that is less fluid like.

As discussed in one of my earlier blogs, the current focus on TSF safety and management shouldn’t just be on TSF walls, but also on optimising the properties of the materials being stored behind them.

3. Improving the quality of overflow water returned to the process

Some have published on the potential for flocculants in returned thickener overflow to have adverse impacts on upstream processes (e.g. flotation).

However, most solid phases readily take up any polymer present. For flocculant to reach an overflow at an appreciable level is a sign of incredibly poor operation (i.e. extreme overdosing, inadequate flocculant dilution and very poor mixing hydrodynamics).

The exception may be in clarifiers where feed solids concentrations are low. Even then it’s a relatively straight-forward process to ensure proper dosing and hydrodynamics.

A bigger concern may be residual solids that report to the overflow.

High clarity overflows (say <100 ppm solids) are essential if overflow then goes to solvent extraction, electrowinning or precipitation, or if the overflow is to be discharged into a natural water body.

Requirements can be much less stringent if the water just goes back to comminution or flotation, but what if there’s one particular phase that is not well flocculated and thereby its proportion rises in the overflow solids?

Fine quartz and well-dispersed clays can be in the ultrafine range (<1 µm), and high molecular weight flocculants are intrinsically inefficient at capturing particles of such sizes.

Clays will adsorb flocculants on their surfaces, but require extreme dosages unless partially aggregated (hence the advantage of pre-coagulation).

Fine quartz may not respond to conventional flocculants unless their surface layers are modified. There is the potential for a recirculating load to develop and adverse impacts may then grow.

The best known example of flocculant choice impacting on overflow quality is in the flocculation of bauxite residue from the Bayer process.

For some bauxites the use of flocculants with the hydroxamate functionality (which favours interactions with iron oxide surfaces even in highly caustic liquors) has proved essential towards achieving the high clarity needed prior to alumina precipitation.

Other examples are known, although few are as universally adopted.

Clarity is one area where the “one flocculant fits all” approach has the biggest impact. We now find operations that had clarity as a low priority are now giving it greater consideration.

Improving fines capture is a fertile area for proposing alternative flocculants.

It is important to understand that novel functional chemistries don’t just offer the potential for better interactions with surfaces.

In solution they may also give a different viscoelastic response which can influence their mixing and distribution through a particle suspension. Their backbone chains are often shorter, so they can provide a greater number of chains at the same dosage on a mass basis, offering better fines capture.

Testing alternative products will often focus on selected conditions that show a benefit. But, proper testing will map a wider range of conditions that may then highlight practical limitations.

Few studies take all these factors into account,. That’s without even thinking about the impact of high salinity in the process waters.

4. Polymer selection giving a step-change in dewatering

Flocculants are needed to promote settling from very dilute suspensions.

Once you’ve got to high solids, residual structures from flocculation can inhibit consolidation.

While the polymers degrade with time, a number of research groups have explored temperature-sensitive polymers as a way to accelerate this.

By far the biggest problem with such work is that at the temperatures supposedly favouring flocculation most of the proposed polymers are poor flocculants.

A more promising alternative is in-pipe flocculation. Here flocculants are added at very high dosages (1-2 kg/tonne) to tailings already at underflow concentrations.

When done right, further dewatering is then achieved on deposition.

Even though the process remains poorly understood, its applied at a massive scale in the Canadian oil sands industry.

Given the very high dosages, it’s no surprise that there’s an impressive effort underway to seek alternative polymers for these applications.

Such developments are hampered by the fairly basic procedures often applied at a lab-scale for testing at high solids making it very hard to isolate the effects of products from those of variably poor mixing.

We recently published our approach to better control the applied mixing conditions, which has been used on model tailings to show both how dosages can be reduced and the impact of flocculant molecular weight.

Work with this approach is continuing with the University of Queensland to establish if the structures formed in tailings then have implications for closure.

Improving flocculant choice and understanding with help from CSIRO

There is no doubt flocculant choice can impact on tailings treatment.

The important question to ask though is “do you have an application where it needs to have an impact?“

If that is the case, do you have the procedures in place to quantify such effects in testing, and are you then able to modify your operations to realise such benefits?

Flocculant choice plus optimised flocculation has the potential to improve the properties of suspensions discharged to a TSF, which reduces risk and could ease TSF management concerns, in turn benefiting social-license-to-operate.

CSIRO doesn’t make, sell or promote flocculant products.

However, with our extensive range of tools for aggregate characterisation and performance testing, we are uniquely positioned to identify products that may offer advantages in a tailings application and to properly quantify the benefits of novel products.

We have the capacity to:

• identify what is the best of a given range of reagents/flocculants
• understand under which conditions it will work best (or when it won’t) and
• advise on how to implement it at full-scale to get maximum benefit.

Critically, decades of experience from applying CFD to full-scale processes provides us with experience in making sure lab testing can be extended to practical applications.

If you’re interested in reducing TSF management risks through improved flocculant choice, we are uniquely positioned and interested in collaborating with you.

Email me at Chris.Vernon@csiro.au to explore how we can help you.

I also invite you to join me when I present at AusIMM Lithium and Battery Metals 2020 digital conference later this month.

Although the focus is on lithium, many other elements are essential in making lithium-ion batteries.

I will examine the role of interconnected elements in the battery value chain, the production of those, and progress that Australia could make in the value-add from simple commodities, to battery grade chemicals, to engineered precursors, to batteries themselves.

Subscribe and receive more updates on our mineral processing research.