Suspended Solids in Copperbelt PLS: What 100 to 250 mg/L Means for Your Filtration Specification

Short answer: Pregnant leach solution from an agitated leach on the African Copperbelt carries 100 to 250 mg/L of suspended solids, and after clarification most of it still runs upwards of 50 mg/L and frequently over 100 mg/L. A South American heap leach circuit delivers PLS below 30 mg/L without clarifying anything at all. That gap is not an operating failure. It is what the region's ore grade and rainfall make inevitable, and it is the number a filtration specification here has to be written against, rather than the flow rate.

The two numbers, and the gap between them

Two surveys of Copperbelt solvent extraction, both published by the Southern African Institute of Mining and Metallurgy, give the figures at two different points in the circuit.

The 2023 survey by Tinkler and Sole gives the raw stream. Agitated leaching is carried out in stirred tanks at 30 to 40°C, so the PLS contains high levels of suspended solids, which they put at 100 to 250 mg/L.

The 2016 survey of Copperbelt operating practice gives the stream after treatment, and it is blunt about what treatment achieves: heap leach operations typically deliver PLS containing less than 30 mg/L TSS, while even after clarification most agitated leach PLS will contain upwards of 50 mg/L and frequently over 100 mg/L TSS.

Read those side by side and the picture is uncomfortable. Clarification on the Copperbelt is doing real work, taking a stream in the low hundreds of mg/L down to somewhere upwards of 50. It is also handing solvent extraction a feed that begins above where an unclarified heap leach circuit finishes.

Neither paper states a removal efficiency for that step and this article will not invent one. The useful comparison is not a percentage. It is that two circuits producing the same cathode feed solvent extraction at upwards of 50 mg/L on one side and under 30 on the other, and only the first of them is paying for a clarification stage to get there.

Why Copperbelt feed is not heap leach feed

None of this is a choice made badly. It follows from the orebody and the weather.

Copperbelt ores are rich, at 1.5 to 5 percent acid-soluble copper, with cobalt as a byproduct credit. The wet season delivers more than 1000 mm of rain concentrated into about five months. High grades and that rainfall together favour agitated leaching over heap leaching, and once you have chosen agitated leaching in stirred tanks, the solid liquid separation has to happen somewhere.

It happens in countercurrent decantation, and the 2016 survey identifies that as the origin of the problem: a disadvantage of solid liquid separation by CCD is the high level of total suspended solids that reports to the PLS, which greatly increases crud formation and the associated losses of extractant and diluent. In percolation leaching systems, the survey notes, the ore bed itself acts as a filtration medium, so TSS values are much lower.

That sentence is worth sitting with, because it reframes the whole comparison. A heap leach circuit is not better at clarification. It has a filter built into its leach. Thousands of tonnes of ore do the duty for free, before the solution ever reaches a pump. An agitated leach has no equivalent, so every kilogram of fines that CCD fails to settle arrives at solvent extraction.

Two further characteristics of the feed make the solids harder to deal with than the raw number suggests. The ores are high in silica, which the 2023 survey links directly to significant crud formation and to a range of other adverse physical effects through the hydrometallurgical circuit. Dissolved and hydrated silica forms polymeric structures as temperature and residence time increase, which alters solution viscosity and impedes the transfer of copper across the aqueous and organic interface. Silica is also usually a major component of crud itself. Separately, the PLS is often calcium saturated, because the valuable minerals sit in high acid consuming dolomite and limestone.

So the particle burden is fine, partly colloidal, chemically active, and sitting in a solution close to saturation on another species. A clarity target written for a different orebody does not transfer.

What the industry says this costs

The 2016 survey does not hedge on the consequence. The rate of build-up of crud, both at the settler interface and on the bottom of the settlers, it states, is by far the most significant operational issue facing the African agitated leach SX operations. If not actively managed, the crud will literally fill up the settlers over a six to twelve month period, and most agitated leach SX operations regularly take individual settlers off line to remove bottom crud.

There is a reagent number attached. Extractant consumption in the region runs at about 3 to 6 kg per tonne of cathode, against 1 to 4 kg per tonne for South American heap leach circuits whose PLS, in the 2023 survey's phrasing, has low levels of suspended solids owing to being filtered through the heap. Some of that difference is entrained organic. The rest is the organic lost with crud.

The survey's own conclusion on what to do about it is one sentence: minimizing the entry of solids into SX is obviously the best approach, although this is difficult to accomplish in practice.

Once crud has formed, it stops being a filtration question. It is a three phase separation belonging to centrifuges, and the recovered organic then goes to clay treatment. We set out that boundary in detail in the copper SX-EW duty map, and it has not moved: a pressure filter's contribution to crud is preventive and it happens one stage upstream.

Why clarification alone is not closing the gap

The most useful line in the 2016 survey is the one about the equipment already installed for this duty. Several operations have installed pinned-bed clarifiers on the PLS streams; these, it records, have seldom been effective, although there are examples where TSS are consistently reduced to less than 20 mg/L.

Both halves of that sentence matter, and they should be read together rather than picked between.

The target is achievable. Somebody is holding under 20 mg/L on Copperbelt PLS, consistently, which means the number is not a laboratory figure and there is no physical reason a circuit here has to live at 100.

And the usual answer is underperforming. This is the region's own literature reporting on the region's own installed base, not a vendor's comparison. When a technology is installed widely and works at a few sites, the variable is rarely the technology. It is how well the duty was characterised before the equipment was sized.

There is corroborating evidence in what operators spend money on upstream. One Copperbelt operation has dosed silica coagulants ahead of both its copper and its zinc solvent extraction circuits, specifically to remove silica from the PLS and avoid severe clogging of its filters. Coagulants can bring down colloidal silica and particles below about half a micron, which is the fraction that settles slowly or not at all. That is a plant paying for upstream chemistry to keep a downstream filter clear, which is a fair measure of how hard this feed is to filter.

Specifying against a real feed number

The gap between a circuit at 20 mg/L and one at 100 is mostly a specification gap. Five things carry it.

Start from the measured solids loading, not the flow rate

Ask for TSS in mg/L at the point the filter will see it, with the sampling point stated, and the clarity the next stage needs in the same units. An enquiry that names a flow rate, a temperature and nothing about solids cannot be quoted properly by anybody, and what comes back will be sized on assumption.

Add the seasonal spread. A wet season that delivers most of its rain in five months does not produce a constant feed, and the number that matters for sizing is the one at the worst month, not the annual average.

Size the area to the solids load and the cycle you will accept

Filtration area follows the solids the filter has to hold and how often you are prepared to discharge it, not the throughput on its own. Two circuits at the same m³/hr with a fourfold difference in solids loading are not the same duty and should not carry the same area.

Large areas are available where the duty needs them. Units of 200 m² and 500 m² can be done. Our own delivered capacities in copper run from 5 to 350 m³/hr, and our copper SX-EW experience includes an installation in the DRC at the top of that range.

Tune the precoat and the body feed to the load

Precoat and body feed are the adjustable part of the specification, and a high solids feed is exactly the case where the adjustment is worth making. The medium is a flux calcined diatomite filter aid. As a general rule of thumb, precoat runs at about 1 kg per m² of filtration area.

Body feed is dosed against the solids arriving, which is the variable this feed makes large and makes seasonal. Right sizing the precoat and tuning the body feed rate to the actual load is how filter aid consumption comes down, and it cannot be done from a flow rate alone. The trade-offs between the two modes are set out in precoat versus body feed filtration.

Treat discharge frequency as an availability question

More solids in means more cake, and more cake means more discharge cycles. At a low solids loading, discharge is a housekeeping item. At Copperbelt loadings it is a production number, and it belongs in the evaluation rather than in a maintenance schedule nobody costed.

A cake discharge cycle typically costs zero to one hour of downtime, and hydraulic open and close with the closure interlocked is what keeps it at the low end. We looked at how that arithmetic works in what a cake discharge cycle actually costs.

Specify materials for the solution, not just the solids

The conditions travel with the feed. Leaching at 30 to 40°C, calcium saturation, and the chloride the downstream electrolyte specification tolerates all bear on material selection. For reference, copper electrowinning requires an advance electrolyte at roughly 45 to 50 g/L copper, about 150 g/L sulphuric acid, under 2 g/L iron, under 0.1 g/L manganese and about 30 mg/L chloride to make LME Grade A cathode consistently.

We have built pressure filters in SS316 and SS316L, and also in Hastelloy and titanium.

What to put in the enquiry

Eight lines. If a specification carries these, any competent supplier can size against it, and the quotes become comparable.

1. Measured TSS in mg/L at the point the filter will see it, and where that sample was taken.

2. Required filtrate clarity, in the same units.

3. Flow rate, with the wet season and dry season spread, not an annual average.

4. Solution temperature at the filter.

5. Chloride, calcium and free acid.

6. A solids characterisation including silica content and an indication of particle size, since the sub half micron fraction behaves differently from the rest.

7. What the next stage actually needs, as distinct from what the current specification says.

8. Any coagulant, flocculant or other reagent already being dosed upstream, and where.

The last one is regularly left out and it changes the answer, because surfactant chemistry added for one purpose does not stop acting at the next unit operation.

The point

Copperbelt PLS is a harder feed than a specification written for another orebody assumes, for reasons that sit in the geology and the weather rather than in the plant. The industry's own survey says solids control is the region's dominant operating problem, says preventing solids entering solvent extraction is the best approach available, and says the equipment usually installed to do it has seldom been effective while a few sites hold under 20 mg/L.

That is a specification problem before it is an equipment problem. Our duty is the solid liquid separation itself: suspended solids out of PLS and out of electrolyte, on leaf or precoat, from 5 to 350 m³/hr, with about 1,000 units installed worldwide across 460+ installations and 25 units in copper refining. Crud, once formed, and entrained organic are somebody else's equipment, and we say so in writing because an engineer specifying a circuit needs the boundary more than they need the pitch. Where that duty sits in a refinery circuit as a whole is covered in our complete guide to electrolyte filtration for copper refineries.

If you have a measured TSS number for your PLS or your electrolyte and want to know what it implies for area, precoat and cycle time, send it with the eight lines above and we will work through it.