Pressure Filter vs Filter Press: Cake Moisture & Cost

Last updated: 2026-10-27

In the pressure filter vs filter press decision, cake moisture usually casts the first vote: membrane filter presses typically reach 8–12% w/w on fine hydrometallurgical solids, where pressure leaf filters level out at 14–18%. Cost per dry tonne casts the second vote, and above roughly 20 t/h of dry solids the continuous pressure filtration route wins it.

Key Takeaways

• Membrane filter presses deliver the driest cakes on compressible fine solids, typically 8–12% w/w.
• Continuous pressure filters win on specific throughput and dewatering cost per dry tonne above roughly 20 t/h.
• Cake thickness, not pressure, often decides moisture: thin cakes at 6 bar can beat thick cakes at 16 bar.
• Automation shifts the labour comparison far more than it shifts the CAPEX comparison.
• The full equipment landscape is mapped in our solid-liquid separation hub.

What Is a Pressure Filter and What Is a Filter Press?

The two machine families both push liquor through a cloth under positive pressure rather than sucking it through under vacuum, but they build, hold and remove the cake in fundamentally different ways.

A pressure filter is a filtration machine that operates inside a closed, pressurized vessel, forming cake on fixed cloth-covered elements and removing it by some automatic or semi-automatic action — vibration, blowback, scraping, plate indexing or rotation — so that filtration resumes without opening the vessel to atmosphere. The feed pump, or in some designs a compressed-air blanket, supplies the driving force, typically 4–8 bar for leaf designs and 6–12 bar for plate designs.

A pressure leaf filter is a vertical pressure vessel that houses a bundle of horizontal or vertical leaf elements, each covered with filter cloth, on which cake builds while the vessel is flooded; at the end of the cycle the vessel is drained and the cake is dislodged, usually onto a conveyor or into a slurry tank. These machines are workhorses for clarification-adjacent duties: washed precipitates, catalyst recovery, carbon fines, and coarse crystalline solids where the cake releases cleanly.

A vertical automatic pressure filter is an indexing machine that carries a stack of horizontal cloth-covered plates inside a pressure vessel; the plates clamp for filtration and wash, open a fixed gap, and index forward so each cloth discharges its cake in turn, which converts an inherently batch operation into a nominally continuous one with availability well above 90%.

A filter press is a batch dewatering machine that clamps a stack of recessed plates between fixed and moving heads, pumping slurry into each sealed chamber until the cake fills it; the press is then opened and every plate drops its cake in a single discharge event. A membrane filter press is a variant that fits flexible diaphragms behind each cloth so the cake can be mechanically squeezed at 8–16 bar after feed, an option developed for compressible cakes that resist air-blow deliquoring. A tower press, less common but relevant at the margins, is an endless-belt machine that runs a horizontal cloth through a tower of alternating pressure and wash zones.

The consequence of these construction differences is a duty split that recurs on almost every project: the press family buys its exceptional cake dryness with dead time between cycles, and the pressure filter family buys throughput by accepting a wetter cake.

Side-by-Side Parameters: Pressure Filter vs Filter Press

The table below compresses the two families into the five numbers that decide most selections. Every value is a typical range — to be confirmed by testwork on your own slurry — because particle size distribution and compressibility move all of them by wide margins.

about ASME VIII Div 1 vs Div 2

Three rows deserve comment. First, the membrane press achieves its headline moisture on compressible material — metal hydroxide precipitates, leach residues, fine gypsum — where the squeeze consolidates a cake that air alone cannot dry; on coarse, incompressible sand the membrane adds little and its premium is wasted. Second, the vertical automatic pressure filter’s throughput advantage is partly a cloth-area accounting effect and partly genuine: its thin, evenly formed cakes and automatic discharge keep the effective filtration rate high all day. Third, cake thickness ranges differ by design intent, not accident — the press deliberately grows a thick cake to amortize its dead time, while the pressure filter keeps cakes thin because discharge integrity limits how much cake a leaf or plate can carry without breaking up.

For how moisture targets should be set against downstream handling rather than equipment capability, see our filter cake moisture resource; the measurement basis matters as much as the number.

Batch vs Continuous: How Do CAPEX, Labour, Footprint and Washability Compare?

Two lines of this table decide more projects than the rest. Washability favours the press because a closed chamber forces wash liquor through a uniform cake of known thickness, so soluble recovery at a wash ratio of 1.0–1.5 is routine; in a leaf bundle, the outer leaves wash differently from the inner ones, and the wash ratio creeps to 2.0 or higher for the same recovery. Availability favours the pressure filter because cake discharge never requires opening the machine — no plate shifting, no cloth-dropping, no operator-dependent discharge quality. Cloth life is broadly similar, but the labour content of a press cloth change (a full set in one shift, two people) versus a leaf vessel (element by element, working inside a confined space) is a real operating difference that vendor datasheets never show.

Where a press is being compared against thickening-centrifuging options rather than against other filters, our thickener vs filter press comparison and the decanter centrifuge for mineral slurry guide bracket that wider choice.

Which Machine for Which Duty? Decision Rules

The duty, not the vendor catalogue, decides this question. For high-value, low-tonnage fine precipitates — cobalt hydroxide, copper powder, precious-metal sludges — a membrane filter press is preferred because the cake is compressible, wash requirement is high and product value makes the extra moisture recovery worth full automation at modest installed area. For gypsum from a neutralisation or citric-acid circuit at 5–15 t/h of dry solids, a vertical automatic pressure filter is preferred because the feed rate is steady, the cake is dischargeable, and the machine washes, dewater and discharges without ever opening, which an operator-hungry batch press cannot match at this tonnage. For lithium carbonate slurries where the cake must carry minimal mother-liquor contamination, a membrane filter press is preferred because chambered displacement washing at low wash ratio protects the lithium yield in ways a leaf bundle’s uneven wash cannot. For coarse, abrasive flotation tailings feeding a dry-stack tailings facility at hundreds of tonnes per hour, a continuous horizontal pressure filter (with vacuum belt filters evaluated in parallel) is preferred because specific cake throughput two to four times a press keeps the filter building and the machine count sane, and the moisture difference between 15% and 18% w/w rarely changes stackability. For corrosive, high-temperature leach slurries where cloth handling is the dominant maintenance pain, a horizontal pressure plate filter with overhead cloth wash is preferred because the closed vessel protects both the operator and the cloth.

The through-line: as tonnage rises and product value falls, the decision migrates from the press to the pressure filter; as wash demand and moisture specification tighten, it migrates back.

A Seven-Step Selection Procedure for the Pressure Filter vs Filter Press Choice

1. Fix the dry-solids mass balance. Establish feed t/h dry solids, feed solids concentration after upstream thickening (our paste thickener design guide covers the feed-conditioning half of this), and the required cake tonnage. Every downstream comparison is meaningless until these numbers are fixed.
2. Set the moisture target from downstream handling. Work backwards from haulage, stackability or smelter specification — not from the driest number any vendor has quoted. A target of 12% w/w versus 16% w/w can double the machine cost.
3. Characterise compressibility and cake resistance. Run leaf or piston press tests across the candidate pressure range. Compressible cakes (resistance rising steeply with pressure) favour membranes and thin cakes; incompressible cakes discharge the advantage of high pressure.
4. Compute specific cake throughput for both candidates. Convert area, cake thickness, cake density and cycle time into kg dry solids per m² per hour for each machine on your data, then divide installed duty by that number to get machine count and installed area. Never accept a vendor’s throughput figure without its cycle-stage breakdown.
5. Price the wash requirement. Convert the target soluble recovery into wash ratio and wash liquor volume for both geometries; the press’s wash efficiency advantage shows up as smaller wash-water and evaporative-load systems upstream and downstream.
6. Cost labour and automation honestly. Model the operator-minutes per tonne for a manual press, an automatic press and the pressure filter, including cloth-change and discharge events. This is where the two families diverge most.
7. Run a side-by-side pilot campaign on real slurry. Confirm moisture, throughput, wash efficiency, filtrate clarity and cloth life for both candidates before committing; our pilot testing team runs exactly this paired campaign. A decision taken without step 7 is a decision taken on the vendor’s best slurry, not yours.

Why Does a 16 Bar Press Not Always Beat a 6 Bar Pressure Filter?

Engineers arrive at this comparison expecting pressure to be the whole story, and the expectation survives exactly one pilot campaign. Filtration and deliquoring time scale roughly with the square of cake thickness divided by the pressure drop, so what matters is the ratio of thickness to driving force, not the force alone. A membrane press at 16 bar holding a 30 mm cake faces a drainage path three times longer than a pressure filter at 6 bar holding a 10 mm cake; the press needs over twice the specific dewatering effort to reach the same moisture, and on incompressible feeds it may never reach it faster. Thin cakes at modest pressure repeatedly out-perform thick cakes at heroic pressure on a kg-per-m²-per-hour basis — this is the single most counter-intuitive result in the pressure filter vs filter press comparison and the reason cake thickness is the first lever we reach for in sizing reviews.

The second practitioner’s lesson concerns compressible cakes and the membrane squeeze. On a compressible cake, extending the feed-pressing stage delivers diminishing returns because the cake’s own structure throttles flow long before equilibrium is reached; the voids collapse, resistance rises, and the last moisture points cost disproportionate hours. A membrane squeeze sidesteps this by applying mechanical stress directly to the cake structure, consolidating it the way a filter press pressing stage never can — which is why a 90-second squeeze at 15 bar routinely beats an extra 20 minutes of feed pumping at the same pressure on zinc residues and fine hydroxides. Squeeze beats duration; duration never fully beats squeeze on genuinely compressible material.

The third lesson is diagnostic: cloth blinding announces itself as rising cycle time, not rising moisture. As the weave loads with fines and scale, formation slows first; the plant compensates by extending stages, and moisture stays nominally on-spec for weeks while throughput quietly erodes 10–20%. Plants that trend filtration time per cycle catch the blinding at the cloth wash; plants that trend only moisture discover it at the annual shutdown. The same logic drives why automation changes the labour comparison more than the CAPEX comparison: a manual membrane press needs an operator at the plates for every one of its 24–35 daily cycles, roughly one full-time position per press around the clock, while an automatic press with PLC-sequenced plate shifting, cloth wash and drip trays needs supervision, not attendance — and the pressure filter needed no attendance at all from the start. That labour delta, compounded over five years, is routinely larger than the CAPEX difference between the machine families, yet it appears in no quotation. Where the press’s cycle discipline is the deciding variable, our filter press cycle optimization resource quantifies the stage-by-stage trade-offs.

What Does a Five-Year Cost Structure Look Like?

The figures below are illustrative, built on stated assumptions: duty of 10 t/h dry solids at 85% availability, power at USD 0.10/kWh, labour at USD 25/h fully burdened, capital charged at 12% per year over five years, cloth life 9 months, air at USD 0.02 per Nm³ — typical ranges, to be confirmed by testwork and vendor quotation.

Read as a five-year total, the press family typically spends 4.9–9.1 USD per dry tonne and the pressure filter family 5.2–9.8, with the press cheaper wherever its superior moisture genuinely removes a downstream drying step, and the pressure filter cheaper wherever high throughput dilutes its capital over more tonnes. The number that moves either machine out of contention is rarely on this table: it is the drying step the moisture target does or does not eliminate, and it should be priced first.

What Does Side-by-Side Trial Data Show?

The following results are indicative and anonymised, assembled as a composite of paired pilot campaigns on representative slurries; they demonstrate the shape of the decision, not a guarantee for any specific feed.

On a zinc leach residue at 55 wt% feed solids, a pilot membrane press with a 25 mm chamber and 15 bar squeeze produced cake at 9.6% w/w moisture at a specific throughput of 46 kg dry solids per m² per hour; the same slurry on a pilot vertical automatic pressure filter at 8 bar reached 11.8% w/w at 128 kg/m²·h. On a neutralisation gypsum, the split was 10.4% w/w at 61 kg/m²·h for the press against 12.5% w/w at 185 kg/m²·h for the pressure filter. On a washed lithium carbonate slurry with a 1.5 wash ratio, the press recovered 92% of soluble lithium in the wash filtrate while the leaf bundle needed a 2.2 ratio to reach 90%. Cloth life ran 7–9 months on both machines on the gypsum, but leaf cloth changes took three times the labour hours of the press set change.

The pattern held across every paired campaign we have run: the press won moisture and wash efficiency by 2–3 percentage points, the pressure filter won specific throughput by a factor of two to three, and the tie was broken by tonnage and by whether the extra two moisture points eliminated a downstream cost.

Frequently Asked Questions

What cake moisture can each machine realistically achieve?

On compressible fine solids such as leach residues and metal hydroxides, membrane filter presses typically reach 8–12% w/w and pressure filters 12–18% w/w. On coarse, incompressible tailings, both families converge toward 14–18% w/w and the choice shifts to throughput. Every slurry has its own floor, so treat published ranges as typical only and confirm with leaf tests on your material.

Is a membrane filter press worth the extra capital over a chamber press?

On compressible cakes, yes: the squeeze delivers 3–5 percentage points of moisture and often a shorter cycle than extended air blowing, which pays back through eliminated drying or haulage savings. On coarse, incompressible feeds, no: membranes add capital without dewatering benefit because the cake structure does not respond to mechanical consolidation. Confirm compressibility with a piston press test before specifying membranes.

Why does cycle time creep up weeks before cake moisture rises?

Cloth blinding reduces permeability progressively, so cake formation slows first. Operators extend stages to compensate, moisture stays on-spec, and throughput erodes silently until the cloth is washed or replaced. Track filtration time per cycle against a clean-cloth baseline and schedule washes when it drifts 15–20% high; that discipline recovers capacity months before a moisture alarm would.

Can pressure filters run continuously, or only semi-continuously?

Indexing designs — vertical automatic pressure filters, horizontal plate filters and tower presses — approach true continuous duty because cake discharge happens inside the pressure vessel while filtration continues elsewhere in the machine. Pressure leaf filters are genuinely semi-batch: the vessel must be opened and drained each cycle, so their effective availability sits between a batch press and an indexing machine, typically 85–92%.

How does filter cloth life and replacement effort compare?

Cloth life is broadly similar across both families at 3–12 months depending on blinding aggressiveness, but replacement labour differs sharply. A press cloth set is changed in one shift by two people; leaf elements are re-clothed one at a time, often inside the vessel. On duties requiring monthly washing and quarterly changes, this maintenance labour difference alone can reach USD 0.2–0.5 per dry tonne.

Which machine is easier to automate?

The pressure filter, structurally: its discharge is built into the machine and needs no external interaction. A filter press can be automated to the same standard with a PLC-sequenced plate shifter, cloth wash and cake hopper, but it is an add-on that costs capital and needs maintenance. Fully automated, both families run with supervision only; the manual press is the configuration that destroys the labour comparison.

How much wash water does in-press washing actually need?

A membrane or chamber press washing a uniform cake typically achieves 90–95% soluble recovery at a wash ratio of 1.0–1.5 times the residual cake void volume. Leaf bundles and indexed plate filters need 1.8–2.5 for the same recovery because wash distribution is less uniform. At high tonnage this difference sizes the wash-water circuit, the downstream evaporation or CCD load, and sometimes the water balance of the whole plant.

Conclusion

The pressure filter vs filter press question resolves on two numbers measured on your own slurry: cake moisture in percentage w/w, and specific cake throughput in kg per m² per hour. The membrane filter press wins the moisture vote on compressible fines and turns that margin into money whenever dewatering eliminates a drying or freight cost; the continuous pressure filter wins the throughput vote and turns that margin into money whenever tonnes are large and labour is real. Neither machine wins by default, and quotations alone will not settle it, because both families look best on the vendor’s slurry rather than yours. Put representative feed through a paired pilot campaign, measure both numbers under your own wash and pressure programme, and let the arithmetic pick the machine — talk to Lexin-tech about pilot testing before the flowsheet is frozen. Broader guides on thickening, washing and dewatering are collected in our resource library.

References

1. Rushton, A., Ward, A. S. & Holdich, R. G. (2000). Solid-Liquid Filtration and Separation Technology, 2nd edition. Wiley-VCH, Weinheim.
2. Svarovsky, L. (2000). Solid-Liquid Separation, 4th edition. Butterworth-Heinemann, Oxford.
3. Wakeman, R. J. & Tarleton, E. S. (2005). Solid/Liquid Separation: Principles of Industrial Filtration. Elsevier, Amsterdam.
4. Tarleton, E. S. & Wakeman, R. J. (2007). Solid/Liquid Separation: Equipment Selection and Process Design. Elsevier, Amsterdam.
5. Perry, R. H. & Green, D. W. (eds.) (2019). Perry’s Chemical Engineers’ Handbook, 9th edition, Section 18: Liquid–Solid Operations and Equipment. McGraw-Hill, New York.
6. SME Mineral Processing and Extractive Metallurgy Handbook (2019). Society for Mining, Metallurgy & Exploration, Englewood, CO.
7. Wills, B. A. & Finch, J. A. (2016). Wills’ Mineral Processing Technology, 8th edition, dewatering chapters. Elsevier, Amsterdam.

Changelog

2026-10-27 — first published.