Solid Liquid Separation in Hydrometallurgy: Equipment Guide

Last updated: 2026-10-18 · Reviewed by: Senior Metallurgical Engineer, {BRAND}

A 50,000 t/d copper concentrator circulates roughly 120,000 m³/d of process water, and correct staging of solid liquid separation is what allows 85–90% of it to be recycled rather than discharged (typical range for a modern plant with thickening, filtration and dry-stack tailings). The equipment choice at each stage — thickener, filter press, vacuum filter or centrifuge — fixes the water balance, the washing efficiency and the tailings footprint at the same time.

Key Takeaways

• Solid liquid separation equipment is selected in stages: thicken first, then filter only what needs a cake.
• Unit area flux (t/m²/d), not tank volume, sizes a thickener; cake moisture, not area alone, sizes a press.
• A well-designed CCD circuit recovers more than 99% of dissolved values at a wash ratio near 2.5.
• Filter presses win on dryness; thickeners win on cost per tonne; vacuum filters suit coarse, free-draining solids.
• Flocculant dosage above about 40 g/t usually signals a feedwell or dilution problem, not a polymer problem.

What Is Solid Liquid Separation in a Hydrometallurgical Plant?

Solid liquid separation is a family of unit operations that divides a slurry into a clarified liquid overflow and a solids-rich underflow or cake, and it sits at the junction of three flowsheet decisions that are usually owned by three different people. The metallurgist cares because washing efficiency determines how much dissolved metal leaves with the tailings. The water services engineer cares because every percentage point of cake moisture changes the plant’s make-up water demand. The tailings engineer cares because underflow density and cake dryness decide whether the facility is a wet pond, a thickened stack or a filtered dry-stack. A flowsheet that treats these as three separate selections almost always over-spends on equipment, because the cheapest overall circuit is found by optimising the train, not each machine.

How to Select and Size a High Rate Thickener

Solid-Liquid Separation

Filter Press Area Calculator from Leaf Test Data

Three definitions are worth stating precisely, because vendor proposals use the same words loosely.

A high-rate thickener is a gravity sedimentation vessel that uses a deep, agitated feedwell and polymer flocculation to settle fine solids at 4–10 times the flux of a conventional clarifier, producing an overflow below roughly 200 mg/L suspended solids and an underflow of 50–65% w/w solids. The high-rate thickener page covers the feedwell and rake design in detail.

A paste thickener is a deep-bed sedimentation vessel that sacrifices overflow polish for underflow consistency, using a tall side-wall and high-torque raking to deliver 65–72% w/w solids with a yield stress in the 20–100 Pa range. The product is not a cake; it is a flowable, non-segregating slurry that can be pumped to a central discharge.

Counter current decantation (CCD) is a washing circuit that arranges thickeners in series so that fresh or barren wash water advances against the direction of solids flow, displacing dissolved values into a progressively enriched pregnant solution. The arithmetic of wash ratio, stage efficiency and soluble loss is set out in CCD circuit design.

The staged logic that follows from these definitions is simple to state and hard to over-apply: thicken everything, filter only what must be dry. A thickener recovers water at the lowest cost per tonne of any device; a press or filter produces a handleable cake but at a capital and operating cost an order of magnitude higher per tonne of solids. Circuits that skip the thickening stage and filter raw leach slurry directly pay for that decision in cloth consumption, polymer use and press area.

Which Design Parameters Actually Size the Equipment?

The parameters below are the ones that appear on every sizing sheet, and they are also the ones most often quoted without a basis. Each carries a wet-basis or dry-basis convention that must be stated, because a 50% feed solid on a wet basis is not the same number on a dry basis, and slurry densities computed from the wrong convention are wrong by more than the design margin.

Table 1 — Key design parameters with typical ranges (indicative, to be confirmed by test work)

Unit area flux deserves the emphasis it rarely gets. A high-rate thickener rated at 6 t/m²/d handles 24,000 t/d of dry solids on 4,000 m² of area — a diameter near 71 m. The same duty on a paste thickener at 1.0 t/m²/d needs 24,000 m², a 175 m diameter, which is why paste thickening is chosen for underflow properties and tailings geometry, not for water recovery economics. The sizing method, including the depth-of-bed and settling-rate corrections that separate a real design from a rule of thumb, is worked through in thickener design calculation.

Cake moisture behaves differently: it is not an independent design variable but an outcome of feed size distribution, cycle time and cloth condition. Quartz-flocculated leach residues at 45–75 µm typically filter to 16–20% moisture in a membrane-squeeze press; the same solids on a vacuum belt filter rarely reach better than 24–28%. The measurement conventions and their effect on dry-stack acceptance criteria are discussed in filter cake moisture.

How Do You Choose Solid Liquid Separation Equipment: Thickener, Filter Press, Vacuum Filter or Centrifuge?

The four equipment families overlap less than their brochures suggest. Each has a settling or filtering characteristic that makes it the economic winner for a defined feed, and the selection question is answered correctly by first asking what the product stream must be, then asking what the feed allows.

Table 2 — Equipment family comparison (typical values, indicative)

The conditional conclusions follow directly from the table, and each is stated as a decision rule rather than a preference.

• Choose a high-rate thickener when the duty is bulk water recovery and the downstream step tolerates a 55–65% w/w underflow. It wins whenever feed tonnage is above roughly 5,000 t/d of dry solids and the product does not need to be conveyed or stacked as a cake.
• Choose a paste thickener when the tailings plan requires a non-segregating, high-density discharge and a cake is not justified by water scarcity or climate. It wins over a high-rate unit when underflow must exceed 65% w/w, and loses when area is expensive, because its flux is a fraction of a high-rate unit’s.
• Choose a filter press when the cake must be dry enough to truck, stack without containment or re-pulped without free liquid — generally below 20% moisture — or when soluble losses in cake moisture carry payable metal. The filter press wins on dryness and on filtrate clarity; it loses on throughput per unit area and on cloth labour. The trade-off against thickeners is set out explicitly in thickener vs filter press.
• Choose a vacuum belt or drum filter when the solids are coarse, crystalline and free-draining — copper EW residue washed on a belt filter, or crystallised salts behind a drum — and when continuous operation with wash stages matters more than ultimate dryness. It wins on availability and simplicity; it loses whenever particle size drops below roughly 45 µm, where cake formation becomes slow and cracking draws air instead of liquid.
• Choose a centrifuge only when the feed blinds every cloth tried — very fine clays, high-viscosity laterite liquors — and the duty is small enough that its high cost per tonne is acceptable. In base-metal circuits it is the rarest of the four families.

Where the choice sits between thickening and pressing — the most common fork in a new flowsheet — the deciding numbers are cake moisture acceptance, wash-water value and tonnage, and the comparison is quantified in thickener vs filter press. For the whole-train view, including how filtration integrates with turnkey plant delivery, the equipment decision should be made against the CCD and tailings alternatives at the same time, never machine by machine.

Why Do CCD Circuits and Dry-Stack Tailings Decide the Water Balance?

The equipment table above prices machines; the water balance prices the flowsheet. Two circuit arrangements dominate modern practice, and both are chosen by running the same arithmetic: how much value is lost in residual solution, and how much water must be imported.

In a CCD circuit, the wash efficiency calculation is unforgiving of shortcuts. At a wash ratio of 2.5 with 90% stage efficiency, a five-stage circuit recovers above 99.5% of dissolved values; at a wash ratio of 1.8 the same five stages recover roughly 98%, which on a 50,000 t/d copper leach at 1.5 g/L copper in solution is a soluble loss measured in tonnes of copper per day. Because each CCD stage is a thickener, the circuit’s capital cost scales with the number of stages and the flux, which is why flocculant performance — and therefore the feedwell design — has a leverage on the whole circuit far beyond its cost per gram.

Dry-stack tailings extend the same logic to the end of the flowsheet. A filtered, dry-stack facility eliminates the tailings pond, recovers 85–90% of the water that a pond would hold or evaporate, and is increasingly a permitting requirement rather than an option in water-scarce regions. The engineering constraint is cake moisture: the stack must support itself at the placed moisture, typically requiring 12–18% w/w for fine residues, and that number in turn fixes the press sizing and the wash ratio upstream. The stacking criteria, slope stability and regional practice are covered in dry-stack tailings applications; the supporting sizing methods for each device are collected in our resources library.

The consequence for equipment selection is that the water balance is set at three points — thickener underflow, CCD wash ratio and cake moisture — and the three cannot be optimised independently. A project that sets cake moisture from the tailings design, then sizes presses to meet it, then discovers the CCD stages are oversized for the reduced wash demand has spent capital twice on the same litre of water.

What Goes Wrong in Operating Solid Liquid Separation Circuits?

The failure modes below account for most production losses in separation circuits, written in the form in which they reach the control room: symptom, cause, action. They are listed in rough order of frequency.

1. Cloth blinding. Symptom: cycle time drifts upward over weeks and cake comes off wet and thin. Cause: fine solids or precipitated salts (gypsum, jarosite, silica) lodged in the cloth weave, accelerated by over-flocculated feed that forms soft, poorly draining flocs. Action: run a cloth permeability test against a new swatch each month, back-flush or acid-wash on a schedule matched to the fouling chemistry, and verify flocculant preparation dilution before blaming the cloth.
2. Muddy thickener overflow. Symptom: overflow suspended solids climb from under 100 mg/L to 500 mg/L or more, fouling downstream solvent extraction or EW. Cause: excessive flux relative to settling rate, a flooded feedwell, or polymer under- or overdosage that destabilises the floc bed. Action: check the settling rate of a fresh feed sample against design, verify feedwell dilution water is flowing, and step dosage down before stepping it up — muddy overflow from overdosage is more common than from underdosage.
3. High cake moisture. Symptom: cake moisture rises 2–4 points above the test-work basis, threatening stack stability or haulage. Cause: cloth blinding, insufficient membrane squeeze pressure, feed particle size drifting finer as the ore changes, or cycle time shortened to hold throughput. Action: re-run a bench pressure filtration on current feed, restore squeeze pressure first (cheapest), then extend cycle time, then consider cloth change — in that order of cost.
4. Flocculant overdosage. Symptom: dosage creeps above 40–50 g/t, underflow viscosity rises, raking torque increases, and flocs carry entrained liquor that worsens washing. Cause: poor make-up practice (undissolved polymer), incorrect dilution water quality (hardness, residual oxidant), or operators compensating for a feedwell problem with polymer. Action: audit make-up and aging, measure solution viscosity in the feedwell, and re-tune dosage against a series of settling tests on fresh feed rather than against yesterday’s dosage.
5. Underflow sanding. Symptom: underflow density fluctuates, pumps cavitate and raking torque spikes. Cause: coarse size fraction settling faster than the rake removes it, typically after a crusher screen failure or a change in ore hardness. Action: verify the size distribution upstream, raise rake speed temporarily, and never let the bed level fall below the rake as a “fix” — the compacted bed that results is far harder to recover than the sanding event itself.
6. CCD stage imbalance. Symptom: one stage runs hot (low density overflow) while the next runs thick, and soluble loss rises at the last stage. Cause: interstage pumps set asymmetrically, or wash water distribution drifting from design. Action: verify each stage’s underflow density against the design staircase and re-balance wash water from the last stage forward — the last stage controls soluble loss and must be kept thinnest.
7. Gypsum and scale fouling. Symptom: feedwell, launders and cloth lose capacity in parallel across the circuit. Cause: supersaturation of calcium sulphate or carbonate as temperature and pH change through the train. Action: sample scale, match an antiscalant or acid dosing point to the supersaturation location, and schedule mechanical cleaning before capacity loss reaches the thickener flux limit.

Five of the seven are process-side rather than mechanical, which is why a separation circuit’s reliability is bought in the metallurgical laboratory and the control philosophy, not only in the equipment specification.

What Does Solid Liquid Separation Cost, and Where Is the Money?

Cost ranges below are indicative 2025 figures, FOB equipment ex-works, excluding installation, electrical and civil works, and should be treated as screening numbers only (basis: carbon steel wetted parts for neutral duty; higher alloys scale the figures by the material factor of the duty).

Table 3 — Indicative equipment cost ranges (2025, FOB, screening basis)

Three cost observations change decisions. First, flocculant is the largest recurring item on a thickened duty — at 25 g/t and USD 2.5/kg, a 24,000 t/d circuit spends about USD 550,000 per year on polymer, so a dosage audit pays for itself repeatedly. Second, vacuum power on belt filters is frequently underestimated: at 8–15 kWh per tonne of dry solids, a mid-size belt filter consumes more energy than the thickener that should precede it, which is the economic argument for thickening everything and vacuum-filtering only the coarse fraction. Third, cake moisture is worth money twice — as water not purchased and as haulage tonnes not moved — and a press sized one moisture point conservative can cost less over ten years than the water it makes unnecessary.

FAQ

Q1. When is a paste thickener better than a high-rate thickener?
Choose a paste thickener when the underflow must exceed about 65% w/w and behave as a non-segregating slurry — typically for centralised or high-density tailings discharge where a filtered cake is not justified. Accept that its unit area flux is 3–10 times lower, so the diameter is much larger. When water recovery alone drives the decision, the high-rate thickener is almost always cheaper per tonne of water recovered.

Q2. What feed solids concentration should I target ahead of a filter press?
Thicken the feed to the highest concentration that still pumps reliably and flocculates well, typically 30–50% w/w for leach residues. Higher feed solids shorten the fill and drain phases of the cycle and raise cake solids directly, but above roughly 55% the slurry becomes hard to distribute evenly across the plates and cycle time savings evaporate. Confirm the optimum with bench pressure filtration on the actual feed.

Q3. Can a CCD circuit replace washing on a filter press?
It can, and the choice is economic rather than technical. A CCD circuit recovers dissolved values continuously with low labour, but leaves the tailings as slurry and requires a pond or thickened discharge. Filter-cake washing gives a dewatered product and very low soluble loss, but at higher capital and operating cost. Where tailings must be dry-stacked anyway, press-based washing usually wins; where a pond exists, CCD wins.

Q4. Why is my cake moisture higher than the vendor’s guarantee?
The three most common causes are a feed size distribution finer than the test-work sample, cloth blinding that has reduced effective drainage area, and membrane squeeze pressure that never reaches design because of air or hydraulic faults. Check the squeeze pressure first because it is free to inspect, then run a particle size analysis against the sizing basis, then examine the cloth. Vendor guarantees assume the sizing basis holds.

Q5. How many CCD stages do I need?
Work backwards from the acceptable soluble loss. At 90% stage efficiency and a wash ratio of 2.5, four stages give about 99% recovery and five give above 99.5%; at a wash ratio of 1.8 you need one more stage for the same result. Each stage is a thickener with full capital and operating cost, so the practical answer is usually four to five stages, with the wash ratio tuned by test work rather than by stage count.

Q6. Is solid liquid separation equipment always specified as separate machines, or can one device do both?
Thickeners and filters are almost always separate devices in base-metal practice, because they are optimised for different products — clear overflow versus dry cake. Combination attempts exist (deep-cone paste thickeners feeding presses directly) and work where the press feed benefits from dense, well-flocculated feed. But a single device asked to deliver both polished overflow and a stackable cake generally delivers neither, and the staged train remains the default for good reason.

Q7. What test work do I need before sizing the equipment?
At minimum: settling rate tests at several flocculant dosages to establish flux, bench pressure filtration or vacuum leaf tests on actual slurry to establish cake moisture and cycle time, and wash-curve tests to fix the CCD stage count or cake wash requirement. Two to four weeks of testing on representative samples costs a fraction of one oversized machine and is the cheapest capital a separation project ever spends.

Conclusion

Solid liquid separation in a hydrometallurgical plant is a staged train, not a machine purchase: thicken everything, wash it counter-currently where values are dissolved, and press or vacuum-filter only the tonnage that must arrive dry. The equipment that wins is the one that matches the feed’s settling behaviour and the flowsheet’s moisture requirement at the lowest cost per tonne of water recovered — and that winner changes between projects, which is why the test work comes first.

If you are building or re-rating a leach, CCD or tailings circuit and want the equipment train sized against your own liquor and solids, send us the feed analysis and tonnage through our turnkey hydrometallurgical plants page, and we will return an equipment selection with design parameters and an indicative water balance.

References

1. Wills, B.A. & Finch, J.A., Wills’ Mineral Processing Technology, 8th ed., Elsevier — sedimentation, thickening and filtration chapters.
2. SME Mineral Processing Handbook, Society for Mining, Metallurgy & Exploration — dewatering and solid-liquid separation sections. smenet.org
3. USGS, Mineral Commodity Summaries 2025 — copper and nickel concentrate and mine production context. usgs.gov
4. Jewell, R. & Fourie, A. (eds.), Paste and Thickened Tailings — A Guide, Australian Centre for Geomechanics, University of Western Australia.
5. ALTA Metallurgical Services proceedings — CCD circuit practice, uranium and copper SX feed clarification papers (Melbourne, various years).
6. CSIRO Mineral Resources — flocculation and dewatering research for mineral tailings. csiro.au

Changelog

• 2026-10-18 — Initial pillar refresh: equipment family comparison, design parameter table, conditional selection rules, CCD and dry-stack water-balance integration, troubleshooting list, cost screening table.

{
“@context”: “https://schema.org”,
“@graph”: [
{
“@type”: “TechArticle”,
“headline”: “Solid Liquid Separation in Hydrometallurgy: Equipment Guide”,
“about”: [“solid liquid separation”, “high-rate thickener”, “filter press”, “counter current decantation”, “dry stack tailings”],
“datePublished”: “2026-10-18”,
“dateModified”: “2026-10-18”,
“author”: {“@type”: “Person”, “name”: “Process Metallurgist, {BRAND}”},
“publisher”: {“@type”: “Organization”, “name”: “{BRAND}”}
},
{
“@type”: “FAQPage”,
“mainEntity”: [
{“@type”: “Question”, “name”: “When is a paste thickener better than a high-rate thickener?”, “acceptedAnswer”: {“@type”: “Answer”, “text”: “When the underflow must exceed about 65% w/w as a non-segregating slurry, typically for high-density tailings discharge; the high-rate thickener is cheaper for water recovery alone.”}},
{“@type”: “Question”, “name”: “What feed solids concentration should I target ahead of a filter press?”, “acceptedAnswer”: {“@type”: “Answer”, “text”: “Thicken to 30–50% w/w for leach residues; above roughly 55% the slurry distributes poorly across the plates and cycle savings disappear. Confirm with bench pressure filtration.”}},
{“@type”: “Question”, “name”: “Can a CCD circuit replace washing on a filter press?”, “acceptedAnswer”: {“@type”: “Answer”, “text”: “Yes, and the choice is economic: CCD is cheaper continuously but leaves slurry tailings; press washing suits dry-stack duty with very low soluble loss.”}},
{“@type”: “Question”, “name”: “Why is my cake moisture higher than the vendor’s guarantee?”, “acceptedAnswer”: {“@type”: “Answer”, “text”: “Usually finer feed size than the sizing basis, cloth blinding, or squeeze pressure below design. Check squeeze pressure, then particle size, then cloth.”}},
{“@type”: “Question”, “name”: “How many CCD stages do I need?”, “acceptedAnswer”: {“@type”: “Answer”, “text”: “Work back from acceptable soluble loss: at 90% stage efficiency and wash ratio 2.5, four stages give about 99% recovery and five above 99.5%; lower wash ratios need more stages.”}}
]
},
{
“@type”: “BreadcrumbList”,
“itemListElement”: [
{“@type”: “ListItem”, “position”: 1, “name”: “Home”, “item”: “/”},
{“@type”: “ListItem”, “position”: 2, “name”: “Solid Liquid Separation”, “item”: “/solid-liquid-separation/”},
{“@type”: “ListItem”, “position”: 3, “name”: “Solid Liquid Separation in Hydrometallurgy: Equipment Guide”, “item”: “/solid-liquid-separation/”}
]
}
]
}