Decanter Centrifuge vs Filter Press: Which for Your Slurry

Last updated: 2026-11-08

In the decanter centrifuge vs filter press decision, the decisive criterion is cake moisture: a membrane filter press typically delivers 15–25% w/w cake moisture on hydrometallurgical solids, while a decanter centrifuge reaches 25–40% on crystalline, granular solids — and worse when fines dominate. Particle size and duty cycle decide next; everything after that is economics. This page works the comparison through systematically, within the wider discipline of solid-liquid separation.

Key Takeaways

• Cake moisture target is the master criterion: filter press 15–25% w/w; decanter 25–40%.
• Decanters need solids above 20–30 µm with a real density difference; fines kill recovery.
• Expect 1–2 kg dry polymer per tonne of solids on a decanter; a press often needs none.
• Decanters run continuously and unattended; presses are batch machines with cloth maintenance.
• Dry-stack tailings and wash-heavy duties favor presses; crystalline salts and mobile duty favor decanters.

Decanter Centrifuge vs Filter Press: Principles and Definitions

Four definitions frame the comparison, because most selection mistakes begin with a fuzzy idea of what each machine does.

A decanter centrifuge (decanter, solid-bowl scroll centrifuge) is a horizontally oriented, continuously fed machine that separates solids from liquid by sedimentation inside a rotating bowl at 2,000–4,000 g, then removes the settled solids with an internal screw conveyor turning slightly faster than the bowl. Liquids overflow over adjustable weirs; solids discharge as a damp, granular cake at the conical beach — feed in, centrate out, cake out, all at once.

A filter press is a batch pressure filtration machine that forces slurry into a stack of recessed plates clad with filter cloth, building cakes in each chamber at feed pressures typically between 6 and 16 bar (0.6–1.6 MPa). When the cake is formed, the feed stops; with membrane plates the cake is then squeezed for extra liquor recovery before the plates open and the cakes drop by gravity. Full background is on our filter press for mining page.

A membrane (diaphragm) plate is a recessed plate whose cloth-supporting surface is backed by an inflatable diaphragm that compresses the formed cake, cutting cake moisture by 2–6 percentage points versus a plain chamber plate (typical range, confirm with test work). Squeeze pressures of 8–16 bar are common in mining practice.

Cake moisture (residual moisture) is the mass fraction of liquid that remains in the discharged cake, expressed on a wet basis; 20% w/w moisture means 800 kg of dry solids per tonne of wet cake. Measurement conventions and moisture mechanisms are treated on our filter cake moisture page.

The two machines therefore dewater by entirely different physics — sedimentation and compaction for the decanter, percolation through a compressible cake under pressure for the press — which is why the honest comparison is made parameter by parameter, duty by duty.

How Do the Two Machines Compare on Key Parameters?

Table 1 puts the two machines side by side on the parameters that drive selection. Every value is a typical industry envelope, to be confirmed with test work on your own slurry.

Three rows deserve emphasis. The g-force row explains the moisture gap: sedimentation compresses a cake only as hard as the centrifugal field allows, and the liquid film carried out on granular solids is irreducible below roughly 25% moisture for most hydromet solids. The recovery row is the quiet killer in decanter selection: what cannot sediment leaves in the centrate, and that fraction must be caught downstream or bled out. And the wash row decides reagent-recovery duties outright — a press washes by displacement, a decanter’s co-current wash needs roughly ten times more liquor (typical ratio, confirm with test work).

How Much Cake Moisture Does Each Machine Actually Deliver?

Cake moisture is where the decanter centrifuge vs filter press gap is widest, and it is governed by particle size distribution, cake compressibility, and the dewatering force applied.

Filter press 15–25% w/w. At 6–16 bar, hydraulic pressure forces liquor through the cake until capillary and surface-held water dominate; the membrane squeeze then collapses voids mechanically, which is why membrane presses land in the 15–25% band on hydrometallurgical precipitates, gypsum-type cakes, and leach residues (typical range, confirm with test work). Plain chamber plates sit 3–6 points wetter. The physics — capillary retention versus bound water, and how the specific cake resistance controls how hard you must push — is on our cake resistance page.

Decanter 25–40% w/w on friendly solids. At 2,500–3,500 g, a decanter compacts granular, crystalline solids — gypsum, jarosite-type residues, coarse crystalline salts — to 25–35% moisture, and coarse free-draining fractions can dip below that. Two effects pull moisture upward: solids retain a surface film of liquor that no g-force removes, and the screw conveyor and beach re-shear the cake on its way out, releasing fines that migrate to the centrate and leave the discharged cake coarser but damper in net.

Fines make both numbers worse — the decanter much more so. When the feed carries a heavy sub-20 µm fraction, the press responds with a slower cycle and a slightly wetter cake; the decanter responds with collapsing solids recovery, because those fines never sediment in the residence time available. A slurry carrying 30–40% sub-20 µm material can turn a 30% moisture decanter cake into a 40%+ cake capturing under 85% of the feed solids (typical behavior, confirm with test work). Hence feed PSD analysis is the first laboratory measurement in any selection exercise.

Which Machine Handles Fine Solids Better?

The press, decisively. A decanter is a sedimentation device: solids recovery depends on each particle’s settling velocity under the bowl’s g-field, which scales with the square of particle diameter and with the solids-to-liquid density difference. Three practical thresholds follow (typical values, confirm with test work):

• Particles below roughly 20 µm, at density differences below a few hundred kg/m³, settle too slowly to capture reliably — recovery can fall below 90% on their fraction.
• Meaningful capture of 20–30 µm fractions requires high g-force, deep-pool operation, and often polymer; the decanter centrifuge on mineral slurries page covers these settings.
• Bowl capacity scales through the sigma theory; our centrifuge capacity and sigma theory page translates a bench settling result into bowl size.

The press has no such threshold. Finer solids blind the cloth and stretch the cycle, but with flocculated feed and adequate pressure they are captured at 97% and above. The trade is time: a fine, high-resistance cake may need 2–4 h cycles where a crystalline cake completes in under an hour — a sizing question, not a feasibility question.

One upstream lever applies to both: pre-thickening. Feeding either device from a thickener underflow at 35–45% w/w solids (typical, confirm with test work) shrinks the machine size materially; the trade-offs are on our thickener versus filter press page.

How Do Polymer, Power, and Maintenance Economics Compare?

The operating-cost comparison is three line items, and the decanter centrifuge vs filter press ranking flips between them.

Polymer dosing. A decanter usually cannot capture fine fractions without flocculation: typical dry polymer demand is 1–2 kg per tonne of dry solids, dosed as a 0.1–0.3% active solution into the feed line (typical range, confirm with test work). At $2.5–4 per kilogram, that is roughly $2.5–8 per tonne of solids — often the largest operating item on a decanter duty. A filter press frequently runs polymer-free because the cloth does the capturing; where conditioning helps a slow cake, 0–0.5 kg/t is typical. Dosing strategy and arithmetic are in our flocculant dosage guide.

Power. The decanter pays for its g-force continuously: 2,500–4,000 g of rotating mass absorbs roughly 2.5–4 kW per m³/h of feed across the main and back drives (typical range, confirm against vendor curves). A press consumes power in pulses — the feed pump against 15 bar for most of the cycle — at roughly 0.8–1.5 kW per m³/h installed. On energy per tonne of dry solids, the press typically wins by a factor of 2–3; the worked example below puts numbers on that.

Wear and maintenance. The decanter’s wear concentrates where cake is scraped and conveyed: screw conveyor flights (hardfaced, rebuilt on a rolling schedule), the beach section, and the main bearings, which set availability. The press’s wear concentrates in consumables: cloths replaced every 1,500–3,000 cycles in abrasive mining service (typical, site-dependent), plate gaskets, and the hydraulic closing pack. Maintenance labor differs in character as much as magnitude — the press needs attention at every discharge; the decanter needs almost none, which is why thin crews choose it for remote duty.

What Are the Typical Failure Modes of Each Machine?

Both machines fail in patterned, diagnosable ways; the lists below make a site walk-down faster.

Decanter centrifuge failure modes:

1. Vibration and bearing temperature rise. Symptom: frame vibration above baseline, hot bearing housings. Cause: cake imbalance from feed surges or uneven beach buildup; degraded grease. Remedy: stabilize feed, check pool depth, service bearings on condition.
2. Screw conveyor torque trips. Symptom: back-drive torque spiking, tripping the machine. Cause: flight wear, cake packing at the beach, or a solids spike beyond bowl capacity. Remedy: inspect and rebuild hardfacing; verify feed density interlocks.
3. Collapsing solids recovery. Symptom: centrate turbidity climbing over weeks. Cause: rising sub-20 µm fraction in the feed or polymer under-dosing. Remedy: re-run feed PSD, retune polymer dose and dilution before touching bowl geometry.
4. Slimy, wetter cake. Symptom: moisture creeping up 3–5 points. Cause: polymer overdose producing weak, water-holding flocs, or pool depth too deep, drowning the beach. Remedy: dose-response test at the feed well; adjust weir rings.
5. Cake discharge choking. Symptom: cake port buildup, vibration, stoppage. Cause: sticky, clay-bearing solids smearing the beach, or an undersized chute. Remedy: reduce feed rate; review beach angle and scroll pitch with the vendor.

Filter press failure modes:

1. Cloth blinding. Symptom: cycle time lengthening cycle by cycle, moisture rising. Cause: fines and crystalline scale wedged into the weave. Remedy: scheduled cloth wash (chemical or high-pressure), then re-specify the weave if wash intervals fall below ~200 cycles.
2. Sticky cakes failing to discharge. Symptom: cakes hanging in chambers, manual scraping each cycle. Cause: cake too wet or plastic — insufficient squeeze, high clay content. Remedy: raise squeeze pressure or time; add an air-blow drying step.
3. Plate pack leaks. Symptom: turbid filtrate or dripping between plates. Cause: worn gaskets, cloth folded at the sealing face, pack misalignment. Remedy: re-align the pack, replace gaskets, retrain the discharge procedure.
4. Uneven cake formation. Symptom: one side of the pack filling faster, thin cakes at the ends. Cause: feed-eye blockage, blinding on individual plates, pump starvation. Remedy: flush feed eyes each cycle, replace blinded cloths, verify pump suction.
5. Hydraulic closing-pack faults. Symptom: pack failing to hold pressure through the cycle. Cause: pump or accumulator seal wear. Remedy: pressure-hold test, seal kit replacement on schedule.

The pattern is worth naming: the decanter fails mechanically (wear, vibration, torque), the press fails functionally (blinding, sticking, leaking). A maintenance crew strong on rotating equipment favors the decanter; a crew strong on scheduled consumable management favors the press.

What Does Each Machine Cost to Buy and Run?

Capital cost indications, deliberately broad because materials of construction and regional pricing move the number: a membrane filter press sized for a 10 t/h dry solids mining duty falls in the order of $300,000–600,000 installed-ready, ex-cake handling; a comparable single large decanter falls in the order of $400,000–800,000, but usually requires a standby unit on critical duty, widening the practical gap (indicative figures only, confirm with quotation). Operating cost per tonne of dry solids is where the comparison becomes concrete; the worked example below computes it line by line. Pressure equipment’s premium is expanded in our vacuum versus pressure filtration comparison.

Decanter Centrifuge vs Filter Press: The Scenario Decision Matrix

Table 2 lists the recurring selection situations in hydrometallurgical slurry dewatering, with the recommendation and reason in one line each.

To state each as a firm recommendation: For dry-stack tailings destined for permanent, trafficable storage, the filter press is preferred because a 15–25% moisture cake compacts under its own weight and hauls without free liquor. For crystalline salt recovery from hot brines, the decanter centrifuge is preferred because coarse crystals at high density difference dewater cleanly under centrifugal force, continuously against the crystallizer’s output. For wash-heavy duties where dissolved reagent values must be recovered, the filter press is preferred because a displacement wash uses a small multiple of cake void volume, where decanter washing needs a dilution volume an order of magnitude larger. For mobile and contract dewatering campaigns, the decanter is preferred because a single continuous skid needs no cloth inventory and near-zero attendance. Where the feed is dominated by sub-20 µm precipitates, the filter press is preferred because solids recovery, not moisture, is the binding constraint.

A Worked Comparison: 10 t/h of Dry Solids (Synthetic)

The example below is synthetic and illustrative — representative figures for a base-metal residue duty, not drawn from any specific plant — and every arithmetic step closes so you can substitute your own numbers.

Common duty. A leach residue circuit produces 10 t/h of dry solids. The slurry is thickened to 38% w/w solids, giving 10 ÷ 0.38 = 26.3 t/h of slurry at 1.35 t/m³, i.e. 26.3 ÷ 1.35 ≈ 19.5 m³/h of feed. The destination is dry-stack; cake moisture target is ≤20% w/w.

Option A — membrane filter press. Test work on the residue at 15 bar gives a filter yield of 150 kg dry solids per m² per hour (synthetic test datum). Required area: 10,000 ÷ 150 ≈ 67 m²; install 70 m². At a 40-minute cycle, each square meter handles 150 × (40/60) = 100 kg dry solids per cycle, so the machine runs 10,000 ÷ (70 × 100) ≈ 1.43 cycles/h — a 42-minute effective cycle, consistent with the margin. Cake at 20% moisture: wet cake = 10 ÷ 0.80 = 12.5 t/h. Power: pump, squeeze and blow-down average ≈ 22 kW, i.e. 2.2 kWh/t dry solids. Polymer: light conditioning at 0.3 kg/t × $3/kg = $0.90/t. Wash water: 0.8 m³ per tonne of wet cake.

Option B — decanter centrifuge. The residue carries a 25% sub-20 µm fraction; bench sedimentation with 1.5 kg/t polymer projects 93% solids recovery (synthetic test datum). Captured solids: 10 × 0.93 = 9.3 t/h. Cake at 32% moisture: wet cake = 9.3 ÷ 0.68 = 13.7 t/h. Sigma-method sizing points to a large-format machine (≈550 mm bowl at ≈2,850 rpm, about 2,500 g) near its limit; specify two units, one duty and one standby. Power: main plus back drive ≈ 65 kW absorbed, i.e. 6.5 kWh/t dry solids. Polymer: 1.5 kg/t × $3/kg = $4.50/t. Lost solids: 0.7 t/h reports to centrate and must be re-fed to thickening.

Closing the comparison. At $0.08/kWh metered at the switchboard (assumption stated): Option A electricity = 2.2 × 0.08 = $0.18/t; Option B = 6.5 × 0.08 = $0.52/t. Polymer: $0.90/t versus $4.50/t. Consumables add roughly $1.5–2.5/t for cloth and gaskets on the press versus $1–2/t for hardfacing and bearings on the decanter (indicative, site-dependent). Against a ≤20% moisture transport spec, Option B’s 32% cake fails the dry-stack criterion outright, and its 7% solids loss would demand a polishing step. For this synthetic duty the press wins on every binding criterion — moisture, recovery, and energy — while the decanter’s advantages buy nothing against a moisture specification.

Frequently Asked Questions

Can a decanter centrifuge ever match filter press cake dryness?

Rarely on hydrometallurgical solids. A decanter’s moisture floor is set by the liquid film that centrifugal force cannot strip from particle surfaces, typically landing at 25–40% w/w on crystalline solids. Only coarse, free-draining, high-density solids approach the dryness of a squeezed press cake. Below about 25% moisture specification, plan on pressure filtration or a drying step after the decanter.

How much polymer does a decanter centrifuge use per tonne of solids?

Typical dry polymer demand is 1–2 kg per tonne of dry solids, dosed as a well-aged 0.1–0.3% active solution into the feed line (typical range, confirm with test work). Under-dosing shows as turbid centrate; overdosing as slimy, wetter cake. At common reagent prices, polymer is often the largest operating cost item on a decanter duty, so a dose-response curve from your own slurry beats any vendor default.

What particle size is too fine for a decanter centrifuge?

As a working rule, fractions below roughly 20 µm — at modest solids-to-liquid density differences — sediment too slowly to capture reliably at 2,000–4,000 g, and recovery on that fraction can fall below 90%. Recovery degrades as the sub-20 µm share grows; above roughly 30% of the feed, a decanter is usually the wrong machine unless polymer dosage rises sharply. Run a feed PSD and bench settling test before sizing any bowl.

Is a filter press being batch a real disadvantage?

It depends on the surrounding plant, and usually it does not. Batch operation suits variable feed rates and lets operators check cake quality at every discharge; surge capacity equalizes flow to conveying. What batch operation does cost is labor, buffer tankage, and cake-discharge automation if uptime demands it. Continuous decanters avoid those needs but forfeit the press’s drier cake and wash efficiency.

Can filter press cakes be washed?

Yes, and well. Once the chambers are full, wash liquor is introduced through the feed eyes and displaces mother liquor through the cake before the squeeze. A wash volume of 1.5–3 times the cake void volume typically recovers the large majority of soluble values (typical range, confirm with test work). Decanter washing, being co-current dilution, needs roughly an order of magnitude more liquor for similar recovery — decisive in reagent-recovery duties.

Which machine costs more to maintain?

They spend money differently. The press consumes cloths every 1,500–3,000 cycles in abrasive mining service (typical, site-dependent), plus gaskets and hydraulic-seal kits — predictable consumables with scheduled labor. The decanter consumes hardfacing rebuilds, beach refurbishment, and bearing overhauls — shop work with crane time and spare-bowl logistics. Compare maintenance cost per tonne of dry solids on your own duty cycle, not list prices.

What test work should precede the purchase decision?

Three tests on your actual slurry, in order: a particle size distribution of the feed solids; a bench settling and polymer dose-response test (which projects decanter recovery and polymer demand); and a leaf or mini-press test at candidate squeeze pressures (which yields filter yield and achievable cake moisture). Those results feed the sizing arithmetic in the worked example above and turn quotations into comparable, plant-specific numbers before capital is committed.

Conclusion

The decanter centrifuge vs filter press choice resolves into three questions asked in order: how dry must the cake be, how fine are the solids, and how continuous must the duty run. A filter press answers a moisture specification — 15–25% w/w is its home territory — and defends recovery on fine, flocculated solids; a decanter answers a continuity and labor specification, dewatering crystalline, granular solids acceptably at 25–40% moisture. Where wash recovery matters, the press wins on reagent economics; where the machine must run unattended or follow a mobile campaign, the decanter wins on operability. Size both options from your own test data — PSD, bench settling, and leaf tests — before inviting quotations, and let the binding criterion make the call. If you are evaluating either machine for a specific slurry, request a quote with your feed analysis, solids rate, and cake specification, and our process team will review the duty with you. Further comparison notes sit in our resource library.

References

1. Svarovsky, L., Solid-Liquid Separation, 4th edition, Butterworth-Heinemann — general reference on centrifuges, pressure filtration, and selection methodology.
2. Records, F. A. and Sutherland, K., Decanter Centrifuge Handbook, Elsevier — scroll centrifuge geometry, capacity theory, and wear practice.
3. Rushton, A., Ward, A. S. and Holdich, R. G., Solid-Liquid Filtration and Separation Technology, 2nd edition, Wiley-VCH — filtration theory and press practice.
4. Wakeman, R. J. and Tarleton, E. S., Solid/Liquid Separation: Principles of Industrial Filtration, Elsevier — cake formation, washing, and deliquoring fundamentals.
5. Society for Mining, Metallurgy & Exploration (SME), Mineral Processing and Extractive Metallurgy Handbook, sections on dewatering and tailings management.
6. Fuerstenau, M. C. and Han, K. N., Principles of Mineral Processing, SME, chapters on solid-liquid separation.
7. American Filtration and Separations Society (AFSS), short-course and technical literature on centrifuges and pressure filtration equipment selection.

Changelog

• 2026-11-08: First published.