Decanter Centrifuge Mineral Slurry: Duty Selection Guide

Last updated: 2026-10-27

A decanter centrifuge mineral slurry duty earns its place when solids are too fine for screens and a cake drier than thickener underflow has value: expect 2,500–3,500 g, cake at 15–25% w/w moisture on crystalline solids, and operating cost of US$1.5–4 per dry tonne. This guide maps where the machine fits, how to size it, and what it costs to run, within our solid-liquid separation equipment family.

Key Takeaways

• A decanter centrifuge wins on fine-particle, medium-tonnage duties where thickener underflow alone stays pumpable and filter presses cycle too slowly.
• Differential speed, not bowl speed, is the real control lever for cake dryness; set it from scroll torque.
• On flocculated duties, polymer spend routinely exceeds power cost, so flocculant conditioning quality is an economic parameter, not a detail.
• Scroll flight tip wear is the true maintenance driver; tungsten carbide tiles can double or triple major-interval life.
• Feed dilution ahead of the machine is often the cheapest capacity gain available — cheaper than buying a larger bowl.

Decanter Centrifuge Mineral Slurry Fundamentals: Bowl, Scroll, Differential and Pond

A decanter centrifuge is a continuously fed solid-bowl sedimentation machine that separates solids from liquid by spinning the slurry hard enough for particles to settle against a rotating wall, then conveying those solids out with an internal screw. The machine — also called a solid bowl centrifuge or scroll centrifuge — replaces gravity with 2,000–4,000 times gravity, which is why it settles in seconds what a thickener settles in hours, and why it can produce a compacting, conveying cake rather than merely a dense slurry.

Slurry Yield Stress

Solid-Liquid Separation overview

The anatomy is worth fixing precisely, because every selection argument later in this article refers to one of five components:

1. Bowl. A cylindrical-conical tube, typically 350–1,400 mm diameter, spinning at 1,600–4,000 rpm so that the wall sees 2,500–4,000 g on industrial mineral duties. The cylindrical section is the settling zone; the conical section is the “beach” up which solids are dragged clear of the liquid before discharge.
2. Scroll. The scroll is a helical screw conveyor that turns inside the bowl at a controlled speed offset, hugging the settled solids from the cylinder, up the beach, and out through discharge ports. Flight tips run against the cake, which is where wear concentrates.
3. Differential speed. Differential speed is the speed difference between scroll and bowl — usually 3–30 rpm on mineral duty, set through a gearbox or back-drive — that determines how fast settled solids are conveyed out and how long the cake is drained on the beach.
4. Pond and weir. The pond is a ring of liquid that stands inside the bowl between the beach shoulder and adjustable weir plates; its depth (commonly set by 20–80 mm of radial weir adjustment) fixes the residence time available for settling and the dry length of beach left for drainage.
5. Flocculant injection. Feed enters through a stationary tube into the pond, and flocculant injection — either upstream in a conditioning pipe or down the feed tube itself — is a dosing step that agglomerates sub-10 µm particles into fast-settling agglomerates the bowl can actually capture.

The separation physics is unglamorous: particles travel radially outward at a velocity set by Stokes’ law multiplied by g-force, and anything that reaches the bowl wall before the liquid reaches the weirs is captured. This creates the machine’s defining trade. Raise bowl speed for better capture and drier cake, and you raise power, stress and wear. Deepen the pond for clearer centrate, and you shorten the dry beach and wet the cake. Speed up the scroll to handle more solids, and the cake leaves wetter and the conveyor torque climbs. Every operating decision in a decanter centrifuge mineral slurry application is a position on one of these trade curves, which is what the next sections quantify.

Where Does a Decanter Centrifuge Mineral Slurry Duty Belong in a Flowsheet?

The short answer: after a thickener, on duties between roughly 5 and 150 m³/h per machine where the solids are too fine for screens and screws, and where a cake drier than thickener underflow has value — whether that value is wash efficiency, transportability, reagent recovery or product quality. A high-rate thickener remains the correct first stage for bulk water removal; its underflow is what a decanter usually feeds on, and its sizing logic is covered separately in our high-rate thickener selection guide. The duty map below gives indicative envelopes — every entry is a typical range, to be confirmed by bench or pilot testwork.

Thickener vs Filter Press

Filter Press Area Calculator from Leaf Test Data in practice

Three placement rules follow from the table. First, put the machine where cake dryness or washing has a quantified value; if the answer to “what is 5% extra cake dryness worth?” is nothing, a thicker underflow is cheaper. Second, never let a decanter replace bulk thickening — feeding it dilute, unflocculated slurry at high tonnage is the most common flowsheet error we see, and the fix is upstream thickener duty, not a bigger centrifuge. Third, be honest about tailings strategy: a decanter can dewater fine tailings, but it does not make stackable dry-stack cake at mine tonnages — that argument, and the press-based answer, belongs to tailings design and to our pressure filter versus filter press comparison.

Which Parameters Decide Decanter Centrifuge Selection for Mineral Slurry?

Six parameters decide the selection, and they interact, so the table below should be read as a coupled system rather than a menu.

Differential speed deserves its own paragraph, because in operation it is the real control lever. Bowl speed is usually fixed at commissioning; pond depth is a maintenance-window adjustment; flocculant responds slowly. What operators touch every shift is Δn, typically via back-drive torque control. Run the differential too fast and the cake rides through the drainage beach too quickly — a duty that delivered 18% moisture at 8 rpm will deliver 26% at 25 rpm, and the surplus water goes back to the circuit. Run it too slow and solids accumulate in the bowl until the scroll conveyor torque spikes and the machine trips on its torque alarm, at best a nuisance and at worst a dig-out with the bowl half full of compacted solids. The correct operating point is a torque setpoint, not a speed setpoint: hold scroll torque at 40–60% of rated and let Δn float.

Flocculant economics are the second thing practitioners learn the hard way. On a flocculated duty, polymer at 0.5 kg/t and US$2.5–4/kg costs US$1.25–2.0/t of dry solids, while the machine’s power share is often only US$0.3–0.8/t. Flocculant spend can exceed power cost by a factor of two or three, so conditioning quality — dilution to 0.05–0.1%, adequate ageing, injection where shear is gentle but mixing complete — is worth more engineering attention than the motor schedule. Under-dosing shows up as centrate turbidity; over-dosing shows up as a slippery, poorly conveying cake that defeats the scroll.

Wear is the third hard lesson. Scroll flight tips run continuously against the settled cake, and on crystalline, abrasive solids the tips are the true maintenance driver — not bearings, not the gearbox. When the flights lose their outer edge, conveying capacity falls, torque rises, and cake dryness quietly degrades weeks before anything fails. Hardfaced flights are the standard answer; tungsten carbide tiles brazed along the flight tips typically extend scroll life two- to four-fold on gypsum and tailings duty, and on a 24/7 machine the tiles routinely pay back within one avoided major overhaul. Materials of construction follow the liquor: 2205 or 2507 duplex for chloride-bearing brines and acid-plant sludge, 316L or carbon steel with exposed welding for neutral duties — a decision worth an order of magnitude in capital cost, so make it from liquor analysis, not catalogue habit.

Selection Logic by Duty: Five Conclusions

Treating each duty from the map above on its own terms:

• CCD underflow polishing. The thickener has already recovered 90%+ of the water; what remains is wash economics. A decanter with wash-water injection on the beach can cut soluble losses by 30–60% versus underflow direct to residue, which on a copper or PGM circuit usually pays for the machine in under two years. Conclusion: strong fit, size for wash efficiency rather than maximum g-force, and model the gain against your CCD wash ratio baseline before committing.
• Gypsum and lithium carbonate. Crystalline or near-crystalline solids that settle fast and drain well: the decanter’s best duty. Gypsum at 12–18% moisture and lithium carbonate at 18–25% are routine, and because these are products or filterable intermediates rather than residues, the machine often sits at the front of a dewatering train feeding a dryer or a washer. Conclusion: default choice in this size band; check liquor corrosivity and crystal habit (needle-like gypsum needs gentler feed pumping to avoid degradation) with testwork.
• Clay and slimes tailings. The machine that is perfect on gypsum struggles on clay, and the reason is fundamental: clays settle slowly even at high g, demand 2–5 kg/t of flocculant, and produce a cake whose yield stress is high but whose permeability is low, so beach drainage barely works. Conclusion: a decanter is viable for fine, low-tonnage slimes streams (typically under 15–20 t/h dry solids per machine) where a transportable 65% solids cake has value; it is the wrong machine for high-tonnage coarse tailings, where screens, cyclones and presses win on every metric that matters.
• Acid-plant sludge and effluent solids. Volumes are small and corrosivity is high; the duty is volume reduction for disposal. Conclusion: fit, but only with the alloy, seal and vapour-tightness decisions made from the liquor analysis; a carbon-steel machine saved 20% at purchase and rebuilt at 18 months is the most expensive sludge pump you will ever buy.
• Hydroxide precipitates (Ni/Co/Cu, manganese). Gelatinous hydroxides compact poorly and blind media, which is exactly why centrifuges are considered — no media to blind. The honest limits: capture is good with flocculant, but cake moisture of 40–60% means the product stays a paste, and scroll conveying of a gelatinous paste is torque-intensive. Conclusion: use the decanter as a washing and classification step ahead of a final dewatering device, or accept paste discharge where the downstream step tolerates it; do not specify one to make a dry hydroxide cake.

The generalisation behind all five conclusions: centrifugal dewatering wins where particles are fine (roughly 1–100 µm), tonnage is moderate (up to ~50 t/h dry solids per machine), and the liquid must be recovered cleanly; it loses where tonnage is high, particles are coarse and permeable (screening and dewatering by drainage are cheaper), or the duty demands a stackable, low-moisture cake at mine scale. Feed dilution deserves a final word here, because it is the quiet capacity lever in almost every decanter centrifuge mineral slurry duty: settling velocity falls sharply at high solids fraction, a duty running at 40% feed solids often gains 20–40% capacity by diluting to 25–30% with recycled centrate or process water — the cheapest capacity gain most plants own, and far cheaper than a second machine.

How Do You Scale Up From Bench Centrifuge Tests to Full Size?

Scale-up rests on the sigma concept, and it is worth understanding simply and honestly. Sigma theory says that a sedimentation centrifuge can be characterised by a single number, Σ, with units of area — an “equivalent settling area” — such that feed rate divided by Σ (Q/Σ) is the same for a bench machine and a full-scale one when the same capture of the same feed is achieved. In practice Σ is the physically settled area multiplied by a machine efficiency factor, and this is where honesty matters: flocculated mineral slurries break agglomerates under shear, cake re-entrains, and real machines rarely match ideal sigma behaviour, so competent design applies an efficiency of roughly 45–65% and confirms it on the actual slurry. Sigma gets you a defensible first sizing; only testwork on your feed gets you a bankable one. The procedure we follow:

1. Characterise the feed. Particle size distribution down to 1 µm, solids density, liquor density and chemistry, temperature, and upstream variability — a duty that swings from 20% to 40% feed solids will not behave like its average.
2. Run bench spin tests. A laboratory bottle centrifuge at 2,000–4,000 g on representative, flocculated samples gives capture versus time and cake solids versus g-force; run at three g-levels and two flocculant doses minimum.
3. Compute the bench sigma. From the bench bowl geometry and speed, calculate Σ for each condition and the Q/Σ that achieved target capture and cake moisture — this is your scale-up datum.
4. Pilot on the real machine. A skid-mounted pilot decanter (350–450 mm bowl) on a several-day campaign of real feed is the only way to capture floc shear, wear behaviour and centrate quality honestly; this is the core of our pilot testing service.
5. Optimise the operating triangle. On the pilot, map cake moisture and centrate solids against differential speed, pond depth and flocculant dose; pick the operating point with margin, not the heroic corner of the map.
6. Fix the flocculant system. Confirm make-up water quality, dilution ratio, ageing time and injection point on the pilot rig; a polymer plant specified from the pilot data avoids the most common capacity shortfall.
7. Apply design margin honestly. Size the production machine at no more than 70–80% of the demonstrated pilot flux, define the scroll torque setpoint from pilot data, and carry a stated basis — never silently — into the performance guarantee.

What Does a Decanter Centrifuge Mineral Slurry Duty Cost per Dry Tonne?

With the assumptions stated, the arithmetic is transparent. Basis: 100 m³/h of feed at 30% w/w solids, slurry density 1.28 t/m³, i.e. 38.4 t/h of dry solids at nameplate duty; 8,000 operating hours per year; power at US$0.10/kWh; flocculant at US$3.00/kg delivered; indicative 2026 pricing, to be confirmed by quotation and testwork.

• Power. Main drive plus auxiliary load of 1.0 kWh/m³ feed = 100 kW average = US$10.0/h = US$0.26/t dry solids. Note the cube-law penalty: the same duty at 3,500 g instead of 2,500 g costs roughly 2.7× the main-drive energy.
• Flocculant (where used). 0.5 kg/t × US$3.00/kg = US$1.50/t; on crystalline duties needing little or no polymer this line collapses toward zero.
• Wear parts. Scroll protection, feed-tube liners, bearings: US$60,000–150,000/year on continuous abrasive duty = US$0.20–0.50/t, halved in practice where tungsten carbide tiles are fitted.
• Maintenance labour and spares. US$0.15–0.35/t including scheduled overhauls amortised.
• Wash water and centrate handling. Negligible when centrate returns to the thickener; add US$0.05–0.15/t if wash water is fresh.

The totals land at US$0.6–1.0/t for clean crystalline duty (gypsum, lithium carbonate) and US$1.5–4/t for flocculated fine duties (clays, hydroxides, sludges) — dominated, tellingly, by reagent and wear rather than by power. Those figures assume the machine runs near nameplate; at 50% utilisation every per-tonne line roughly doubles, which is why duty definition matters more than price negotiation.

An Indicative Operating Record: Copper CCD Polishing Duty

The following is an anonymised, indicative record drawn from a blended picture of installed duty, not a specific client. A copper leach plant ran a 550 mm bowl decanter on CCD final-stage underflow polishing: 90 m³/h feed at 32% w/w solids, flocculated at 0.35 kg/t, 2,600 g, differential speed under torque control at 10–14 rpm. Sustained results over a six-month window were: centrate suspended solids of 0.2–0.4% w/w (95–97% solids capture), cake at 70–73% solids w/w (27–30% moisture), and soluble copper loss in residue cut by an estimated 45% versus the previous direct-discharge baseline. On wear, the hardfaced scroll reached 1,900 hours before conveying capacity visibly faded; retrofitting tungsten carbide tiles extended the observed tip life to beyond 5,500 hours at similar duty, supporting the lifecycle argument made earlier. Treat all numbers as indicative — the transferable lesson is the pattern, not the decimals: torque-controlled differential, modest flocculant, and wear protection decided the economics.

Frequently Asked Questions

What cake moisture can a decanter centrifuge achieve on mineral slurry?
On crystalline solids such as gypsum, expect 12–20% w/w moisture; on lithium carbonate, 18–25%; on fine clays and hydroxide precipitates, 35–60%, sometimes paste-like. Cake moisture is governed mainly by differential speed, pond depth and beach length, with g-force second. Exact figures are feed-specific and must be confirmed by bench or pilot testwork on representative slurry.

How is a decanter centrifuge different from a screen-bowl or peeler centrifuge?
A decanter is a continuously fed sedimentation machine with no medium, suited to fine, moderate-tonnage slurry. A screen-bowl centrifuge adds a screening section for coarser, permeable solids and drier cake; a peeler is a batch machine for crystalline products at high purity. Rough guide: below about 100 µm think decanter, above it think screen bowl, and for product-grade crystals consider peelers.

Why is differential speed more important than bowl speed in daily operation?
Bowl speed is normally fixed, while differential speed (Δn) is adjusted every shift against scroll torque. Δn sets how long cake drains on the beach and how fast solids are conveyed out: too fast gives a wet cake and lost capture, too slow builds a bowl load until the torque alarm trips. Running on a torque setpoint of roughly 40–60% of rated is the practical answer.

When does a decanter beat a filter press — and when not?
A decanter wins on continuous duty, fine particles that blind cloth, moderate tonnage (to roughly 50 t/h dry solids per machine), and where centrate clarity matters. A press wins where tonnage is high, cake must be very dry and stackable, and batching is acceptable. The full trade-off is set out in our pressure filter versus filter press comparison; confirm with testwork.

How much flocculant does a mineral slurry decanter duty consume?
Clean crystalline duties may need little or none; typical flocculated mineral duty runs 0.1–1.0 kg/t of dry solids, and clays or gelatinous hydroxides can need 2–5 kg/t. Because polymer at US$2.5–4/kg usually costs more per tonne than the machine’s power, dosing precision and conditioning quality are economic parameters, and overdose can worsen cake handling.

What wears out first on a decanter centrifuge?
Scroll flight tips, which run continuously against the abrasive cake, are the true maintenance driver — ahead of bearings and gearboxes. As tips wear, conveying capacity and cake dryness degrade before anything fails. Tungsten carbide tiles on the flight tips typically extend scroll life two- to four-fold on gypsum and tailings duties and are usually justified on continuous abrasive service.

Can feed dilution really increase decanter capacity?
Yes, and it is often the cheapest gain available. Settling velocity falls steeply as slurry viscosity and hindering rise with solids fraction, so diluting a 40% w/w feed to 25–30% with centrate or process water can lift capacity by 20–40% at the same g-force. The trade is extra liquor through the circuit, so dilute with process water you were already handling, not fresh make-up.

What testwork should precede ordering a decanter centrifuge?
Minimum: particle size analysis, bench bottle-centrifuge tests at three g-levels with flocculant screening, and a pilot campaign on real feed mapping cake moisture and centrate quality against differential speed, pond and dose. Size production machines at 70–80% of demonstrated pilot flux. Skimping on the pilot step is the most reliable way to buy a machine that misses guarantee.

Conclusion

A decanter centrifuge mineral slurry duty is a strong, defensible choice where solids are fine, tonnage is moderate, and a cake drier than thickener underflow has real value — gypsum and lithium carbonate dewatering and CCD underflow polishing are its natural home, while high-tonnage coarse tailings belong to other machines. Size it from honest sigma-based scale-up confirmed on real feed, run it on torque-controlled differential speed, spend engineering attention on flocculant conditioning and scroll wear protection, and the machine will deliver cake at 15–30% moisture for US$1.5–4 per dry tonne with predictable maintenance. If you are sizing a duty now, talk to us about a pilot testing campaign, or browse the rest of our resources for the surrounding circuit.

References

1. Leung, W. W.-F. Industrial Centrifugation Technology. New York: McGraw-Hill, 1998.
2. Ambler, C. M. “The Evaluation of Centrifuge Performance.” Chemical Engineering Progress 48 (1952): 150–158.
3. Rushton, A., A. S. Ward, and R. G. Holdich. Solid-Liquid Filtration and Separation Technology. 2nd ed. Weinheim: Wiley-VCH, 2000.
4. Svarovsky, L. Solid-Liquid Separation. 4th ed. Oxford: Butterworth-Heinemann, 2000.
5. Green, D. W., and R. H. Perry, eds. Perry’s Chemical Engineers’ Handbook. 8th ed. New York: McGraw-Hill, 2008.
6. U.S. Geological Survey. Mineral Commodity Summaries 2025 — Gypsum. Reston, VA: USGS, 2025. https://www.usgs.gov/

Changelog

• 2026-10-27 — First published.