Rotary Vacuum Filter

A rotary vacuum filter is a continuous filtration machine that dewaters a slurry on the outside of a rotating, cloth-covered drum held under internal vacuum. The drum turns through a vat of feed slurry, cake forms on the cloth as filtrate is pulled through the drum internals, the cake is lifted out of the vat, washed and dewatered with air drawn through it, and then discharged by a scraper, a belt, or a roll before the surface re-enters the slurry. One drum, running day and night at a fraction of a revolution per minute, replaces the fill, clamp, squeeze, and dump cycle of batch pressure filtration with a steady stream of cake and filtrate.

This page treats the rotary vacuum filter as a purchasing object: how drum submergence, vacuum level, cloth or precoat selection, discharge method, and wash zone design translate into filtration flux, cake moisture, and cloth life. It sits beside the batch filter press and the continuous decanter centrifuge in the dewatering decision, and next to the liquid ring vacuum pump that is almost always its largest utility consumer.

Cylindrical industrial filtration vessel hardware installed in a process line, representative of the continuous separation equipment class that includes rotary drum vacuum filters

Photo: Agne27, CC BY-SA 3.0, via Wikimedia Commons

The decision concentrates in five places: the solids loading and particle size of the feed, the required cake dryness and wash quality, the discharge method the cake physics allow, the vacuum level and its utility cost, and the cloth or precoat regime that keeps flux from collapsing between change-outs. Get those right and the drum size follows.

Chapter 1 / 06

What a Rotary Vacuum Filter Is and What Is Being Decided

The industrial machine this page describes is most often called a rotary drum vacuum filter, and the abbreviation RDVF appears throughout vendor literature. It consists of a horizontal drum, typically 0.6 to 4 meters in diameter and up to 6 meters long, rotating at 0.1 to 3 revolutions per minute while partially submerged in a vat of slurry. The drum face is divided into narrow compartments, each covered with filter cloth and connected through a rotating control valve to an internal filtrate pipe. As each compartment passes through the vat, vacuum applied from inside draws filtrate through the cloth and deposits a layer of cake on the outside. The compartment then rises through the atmosphere, where the same vacuum pulls air through the cake to dewater it and, where required, wash water sprayed on the surface percolates through to displace mother liquor.

What separates the rotary vacuum filter from the rest of the dewatering field is continuity at modest driving force. A filter press drives filtration at 6 to 16 bar in batches and reaches drier cake; a decanter centrifuge drives it with several thousand g of centrifugal force and handles soft, fine solids that blind cloth; the rotary vacuum filter drives it at 0.6 to 0.85 bar of vacuum, endlessly, with washing performed in the open air where operators can see it. The trade is a wetter cake, typically 15 to 40 percent moisture by weight depending on the solids, bought with simple mechanics, gentle solids handling, and a duty cycle that never stops for a plate pack to be opened.

The machine is a system, not a drum. Behind every successful installation stand a feed tank with agitation, vat level held by a level transmitter and a control valve on the feed line, a chemical dosing skid when flocculant or coagulant is needed, a vacuum pump sized for the air in-leakage as much as the filtrate, a filtrate receiver with a mist eliminator that keeps liquid out of the vacuum pump, a cake conveyor, and a cloth wash or precoat system. When engineers say a rotary vacuum filter trial failed, the root cause is as often in this balance of plant as in the drum itself. The five decisions this page works through, drum size, discharge method, filter medium, vacuum and wash design, and materials of construction, are the ones that determine whether the installed system meets its flux and moisture guarantees.

Buyers from mineral processing, pigment and chemical manufacture, sugar refining, wastewater plants, and pharmaceutical drying operations all buy the same machine for different reasons. A mineral concentrator wants throughput per square meter and cloth life measured in weeks. A pigment plant wants a bright, well-washed cake and will pay for precoat discipline. A wastewater utility wants unattended operation and cake dry enough to reduce hauling cost. The specification language differs, but the physics on the drum face is the same, and the paragraphs that follow keep all three buyers in view.

A short tour of named duties makes the machine concrete. In phosphoric acid plants, belt discharge drums dewater gypsum at hundreds of square meters of installed area, with rubber-covered compartments resisting fluoride attack. In sugar refineries, rotary vacuum filters wash and dry carbonation and sulfitation muds, where cake purity and liquor recovery decide sugar yield. In corn and wheat starch plants, precoat drums, their aid blended by a weigh feeder and their feed pushed by a slurry pump sized for thick underflow, recover gluten and steepwater solids from dilute liquors. In metal refineries, precipitated copper, nickel, and uranium concentrates leave scraper discharge drums at moisture contents set by smelter contracts. Each duty names the same four knobs, submergence, vacuum, cloth, and discharge, and the numbers differ by an order of magnitude while the machine barely changes.

Chapter 2 / 06

How Drum Rotation, Submergence, and Vacuum Do the Work

Filtration on a rotary vacuum filter is a cycle divided by geography. Imagine the drum face as a clock: the submerged arc is the cake formation zone, the arc above the slurry level is the drying zone, and within the drying zone a wash spray band and a final dewatering band sit before the discharge point. Each compartment on the drum face traverses this geography once per revolution, so the cycle time for any element of cake is the rotation period, typically 20 seconds to 10 minutes. Slowing the drum lengthens every zone at once; the control valve, with its bridge blocks, is what lets the designer allocate the arc between forming, washing, and drying independently of speed.

Cake growth in the formation zone follows the same law as pressure filtration: filtrate flow declines as the cake thickens, roughly in proportion to the square root of time for incompressible solids. Because every compartment spends the same submerged arc time in the vat, the cake leaving the vat has a uniform thickness set by feed solids concentration, submergence, and vacuum. Drum submergence, the fraction of the drum face submerged, runs from about 10 percent for fast-filtering crystalline solids to 40 percent or more for slow, fine slurries that need most of the drum just to build a workable cake. High submergence trades drying arc for forming arc, so it is a moisture decision as much as a throughput decision.

The vacuum system defines the driving force and a large share of the operating cost. Most drums operate at 200 to 500 millibar absolute, which is 0.6 to 0.8 bar below atmosphere, produced almost universally by a liquid ring vacuum pump because it tolerates corrosive vapors and carryover; only small, benign-vapor duties are served acceptably by a rotary vane pump, which oil contamination of condensate eventually disqualifies on process side. The pump must handle two streams: the filtrate vapor and air pulled deliberately through the cake during the drying zone, plus unavoidable air in-leakage at the control valve face and shaft seals. The two-stage separator train, a filtrate receiver above and a separator or demister ahead of the pump, exists to keep liquid slugs out of the pump; a wet vacuum pump loses capacity precisely when the filter loads up.

Washing happens in the open-air arc. Wash water, either fresh or a cleaner process liquor, sprays onto the exposed cake and the same vacuum pulls it through, displacing mother liquor in the cake pores. Displacement washing is far more efficient per ton of water than re-slurring and refiltering: a wash ratio of 1 to 1.5 kilograms of water per kilogram of cake solids commonly removes 80 to 95 percent of soluble impurities when the cake is permeable. The limit is channeling, where wash water finds preferred paths and leaves pockets of mother liquor behind, which is why wash distribution weirs and wash troughs that deliver a uniform curtain matter more than wash volume.

Drying in the final arc pulls air through the wet cake, and the moisture left at discharge is set by the balance of air rate, time, and the capillary entry pressure of the cake. Fine, compressible cakes hold moisture stubbornly; steam hood assistance, hot air injection, or a mechanical press roll can each knock several moisture points off when the duty justifies them. Filtration flux, expressed as dry kilograms per square meter of drum area per hour, is the number that sizes the drum: well-behaved mineral slurries run 300 to 1,000, pigments and fine chemicals 50 to 200, and flocculated biological solids 20 to 80. Vendors size from a leaf test on the actual slurry, and buyers should expect that test, not a catalog curve, behind every quoted flux.

Chapter 3 / 06

Drum Designs, Discharge Methods, and Duty Boundaries

The discharge method is the first fork in the road because the cake, not the buyer, chooses it. Scraper discharge, the classic arrangement, uses a doctor blade set a millimeter or two from the cloth to peel the cake off as the drum surface passes. It suits cakes at least 3 to 5 millimeters thick that hold together, from mineral concentrates to crystalline salts. Belt discharge replaces the fixed scraper with an endless filter cloth that leaves the drum at the discharge point, runs over a small roller that cracks the cake off by flexion, then passes through wash rollers before returning to the drum. The self-cleaning journey of the cloth makes belt discharge the choice for sticky, blinding slurries such as metallurgical precipitates and sewage sludge.

Roll discharge presses a secondary roller against the cake at the point where the drum surface moves away; adhesion to the faster or oppositely running roller lifts thin, tacky cakes that a scraper would smear, typical of latex, dye, and fine pigment duties at cake thicknesses of 1 to 3 millimeters. Precoat discharge is a different regime altogether: before feeding slurry, the drum builds a 50 to 100 millimeter bed of diatomaceous earth, perlite, or cellulose on a fine backing cloth, and the slurry forms a thin layer on top of that precoat. A slowly advancing blade slices off the slurry cake together with a fraction of a millimeter of precoat per revolution, presenting fresh precoat surface every cycle. Precoat rotary vacuum filtration clarifies liquids that would blind any cloth outright, at the cost of consumable precoat and a cake diluted with filter aid.

Within these discharge families, vendors offer metallurgy and sealing options that define duty boundaries. Drums in phosphoric acid service use 316L or higher alloy faces and rubber-covered compartments; chloride duties push toward titanium or fiberglass-reinforced plastic; sugar and food duties demand polished stainless and food-grade cloth. Gas-tight designs with inert blanketing exist for solvent filtrates, and enclosed filter bodies appear where volatile organic carbon limits apply to the drying zone. The enclosed variant matters because the drying zone is an open air path: whatever vapor the vacuum system pulls passes to the pump and its seal liquid, so emissions permitting belongs inside the filter specification, not after it.

Duty boundaries against the neighbors are worth stating plainly. The machine is not a filter press: when the requirement is 25 percent moisture or drier on a fine chemical, or when cake must be washed to near-zero soluble content in a closed system, batch pressure filtration wins and the drum concedes. It is not a decanter centrifuge: when solids are fine and soft, feed varies hour to hour, and floor space is short, the centrifuge wins on robustness per square meter. And when the duty is truly a clarification job, dissolved or sub-micron solids at parts per million, the right machine is more often a precoat drum in its clarifying mode, a candle filter, or downstream polishing with a precision filter cartridge, not a cloth drum run thin. Within the wider industrial filter family the drum is also not a self cleaning filter of the backwash class: those continuously regenerate a granular or mesh medium in line and never produce a cake. On wastewater duties the drum usually follows a clarifier or a dissolved air flotation cell that has already thickened the feed, and a sludge pump transfers that thickened biosolids stream to the vat; trying to filter dilute 0.5 percent solids wastes the machine on water.

Discharge methodTypical cake thicknessCake character it fitsWatch-outs
Scraper (doctor blade)3 to 25 mmCoherent cakes: minerals, salts, sugarCloth wear at the blade gap; thin cakes smear
Belt (cloth leaves drum)2 to 15 mmSticky or blinding cakes: precipitates, biosolidsCloth tracking and wash rollers add maintenance
Roll1 to 3 mmTacky thin cakes: latex, pigments, dyesRoll covering and speed ratio need tuning
Precoat (advancing blade)1 to 3 mm cake on 50 to 100 mm aidClarification and blinding slurries; bright liquorsFilter aid cost; cake contaminated with aid
Chapter 4 / 06

Key Specification Parameters

Four groups of numbers define a rotary vacuum filter purchase: the sizing numbers that set drum area, the process numbers that set performance guarantees, the utility numbers that set operating cost, and the mechanical numbers that set materials and code boundaries. The sizing set starts from dry solids throughput and leaf-test flux, which together give required filter area, and continues through drum submergence and speed range. The process set is cake discharge moisture, wash solubles target, and filtrate clarity. The utility set is vacuum pump capacity and motor list, wash water rate, and any steam or hot gas consumption. The mechanical set is materials of construction, cloth specification, control valve trim, and the vessel code the casing must meet.

ParameterTypical rangeWhy it matters
Filter area (drum face)0.5 to 100 m2Sizing variable; flux from leaf test times area fixes throughput
Drum speed0.1 to 3 rpm (to 6 on some designs)Sets cycle time; variable speed is standard on modern units
Drum submergence10 to 40 percent of faceDivision between cake forming arc and wash or dry arc
Operating vacuum200 to 500 mbar absoluteDriving force; also sizes the liquid ring vacuum pump
Cake thickness at discharge1 to 25 mm by methodMust suit discharge hardware; thick dries better, thin washes cleaner
Cake moisture (wet basis)15 to 40 percent typicalGuarantee point; drives haulage or downstream drying cost
Filtration flux20 to 1000 kg/m2h dry solidsThe number that must come from a leaf test on real slurry
Cloth life2 weeks to 6 monthsLargest consumable driver after labor and power

Wash design deserves its own line in the specification. State the wash liquor, its quality, the permissible ratio of wash to dry solids, and the residual soluble target in the discharged cake, because these four numbers fix the wash zone arc, the spray design, and often the difference between meeting and missing a purity guarantee. Where the cake is compressible, state whether steam hood, press roll, or hot air options may be offered as alternatives in the bid, since vendors differ in which drying aids they can attach to their drum frame. Ask each bidder to state the leaf test or pilot protocol behind the flux and moisture numbers in the proposal.

Vacuum utility deserves equal precision. The specification should require the vacuum pump capacity in cubic meters per hour of air at the design vacuum, stated separately from the filtrate rate, together with seal liquid flow and temperature limits for a liquid ring machine. Compressors this size are modest, but air in-leakage through an aging control valve can double the required pump size, and a pump sized without in-leakage margin will lose vacuum exactly on the hottest afternoon when cooling water warms and vapor load rises. The filtrate receiver volume, separator demisting, and barometric leg or pump-back arrangement belong in the same package so that responsibility for carryover is unambiguous.

Chapter 5 / 06

Selection Decision Factors and RFQ Language

Selection starts with the slurry, honestly characterized. Particle size distribution including the fraction below 10 um, solids concentration by weight, liquid viscosity and temperature, and the nature of the solids, crystalline, amorphous, fibrous, oily, or biological, together predict cloth blinding, cake cracking, and discharge behavior. A laboratory leaf test simulating the drum cycle, form, wash, dry, gives flux and moisture; a bench precoat run tells whether filter aid is needed at all. Plants that skip this step and size from a similar duty elsewhere inherit both drums that run out of submergence in summer and cloth that blinds in a week.

The discharge decision then falls out of the cake. Coherent mineral and salt cakes point to scraper discharge with a polypropylene or polyester cloth in the 20 to 100 um retention class. Sticky or blinding duties point to belt discharge and its continuously washed cloth. Thin tacky films point to roll discharge. Liquors that must run bright, or slimy feeds like fermentation broth and activated carbon, point to precoat operation with diatomite or cellulose, sized for the precoat consumption per ton of product. Where the solids are the product, precoat contamination may itself rule the method out and push the duty toward a decanter centrifuge or a pressure filter.

Materials and emissions close the selection. State the chloride, fluoride, acid, or solvent exposure honestly, because welding a 316L drum for a chloride service guarantees a stress corrosion story later. Enclosed and gas-tight bodies, inert gas blanketing of the vat, and closed loop seal liquid systems are all standard options with real cost, and they belong in the RFQ when volatile organics or toxic liquors are present. For plants with steam in the battery limits, a steam hood on the drying arc is cheap moisture reduction; for plants without, it is an expensive accessory.

A complete RFQ for a rotary vacuum filter states, in one page of process data: slurry flow and solids concentration; particle size distribution and solids density; liquid viscosity, temperature, and vapor pressure or boiling behavior under vacuum; required cake moisture and wash residual solubles; wash liquor and permissible wash ratio; materials of construction with chloride or fluoride limits; vacuum utility available and its cooling water temperature; preference or prohibition of precoat and filter aid; continuous duty factor and turndown; and the applicable emission or vessel code requirements. Vendors quoting against that page are quoting the same machine, and the flux, moisture, cloth life, and power numbers in their proposals become comparable line items rather than marketing prose.

Balance of plant belongs in the same purchase decision even when it sits on a different purchase order. The agitated feed tank prevents solids settling that starves the vat; the flocculant dosing skid, when used, can double flux on biological and fine chemical duties; the cake conveyor or screw must accept the discharge moisture actually guaranteed; and the filtrate pump-back or barometric leg must be matched to the receiver elevation. A drum quoted alone, without these interfaces written down, is a drum that starts up late.

Spare parts strategy is part of selection because the consumables are predictable. A first purchase should carry at least one full cloth set plus a spare roll of belting where belt discharge is used, a control valve wear plate and bridge block set, discharge blade stock, wear shoes for the agitator, and a seal kit for the vacuum pump. Lead times on large alloy drums are measured in months, but almost everything that actually stops a rotary vacuum filter is a consumable that fits on a shelf. Buyers who negotiate the cloth specification and its alternates at purchase time, rather than accepting a single source weave, hold the lever on the largest recurring cost the machine will ever have.

Acceptance at commissioning should be written into the purchase as a performance test, not left to goodwill. The test runs at design feed rate and solids concentration for an agreed shift length, logging drum speed, submergence, vacuum, wash ratio, throughput, cake moisture, and filtrate clarity against the guarantees. Two numbers do most of the talking: dry kilograms per square meter per hour and cake moisture at discharge, both corrected to the design feed. A performance test clause converts the vendor leaf test from a sales device into an engineering baseline the plant can trend against for the next twenty years.

Chapter 6 / 06

Installation, Standards, Failure Modes, and Maintenance

Installation is straightforward heavy rotating equipment: a level foundation, crane access for the drum and its agitator, clearance to pull the control valve, and enough headroom above the vat to rig the cloth or precoat system. The filter gives its best performance when the vat agitator keeps feed concentration steady, so the feed tank, its agitation, and the valve that controls vat level deserve commissioning attention equal to the drum. Where the drying zone is enclosed for emission control, verify the enclosure drains and the seal liquid circuit before first vacuum, because a flooded separator is the classic first-week outage.

Standards invoked in this equipment class are mostly mechanical and emissions codes rather than a filter-specific product standard. Vessels and casings that see pressure or vacuum are commonly designed to ASME Boiler and Pressure Vessel Code Section VIII rules or the local pressure equipment equivalent; electrical parts in solvent or dust service follow ATEX or the applicable hazardous area code; and food and sugar duties add sanitary design practice. Vendor literature from ANDRITZ on the Krauss-Maffei TSF drum family and from Komline-Sanderson on rotary drum vacuum filters is the practical reference for zone arrangements and discharge options, and the EPA monitoring knowledge base treats the rotary drum filter among liquid stream filtration devices for regulatory monitoring context. None of these replaces a leaf test, and a buyer should treat any flux guarantee offered without one as provisional.

The recurring failure modes are few and well documented. Flux decay across a shift almost always means cloth blinding or feed chemistry drift, answered by cloth washing intensity, occasional chemical cleaning in place, or a change of cloth weave. Wet cake at discharge with normal vacuum means the drying arc is starved, from too fast a drum speed, too deep a submergence, or a wash spray overrunning the drying zone. Vacuum that will not hold points to the control valve face, wear plates, or shaft seals. Cake that will not discharge cleanly points to the discharge gap, roll speed ratio, or a cloth that has glazed. Each symptom has a cause and a fix, and the maintenance plan is built around them.

SymptomLikely causeFirst action
Flux down shift over shiftCloth blinding, fines in feed, chemistry driftIncrease cloth wash; check flocculant dose; CIP with the liquor the cloth vendor specifies
Cake suddenly wetDrying arc shortened, wash overrunning, vacuum lossSlow the drum; trim wash spray band; check control valve and pump capacity
Vacuum will not reach setpointControl valve face wear, shaft seal leaks, pump capacityInspect wear plate and bridge blocks; smoke-test seals; verify pump air capacity
Cake smears at scraperThin or tacky cake, wrong gapDeepen submergence or slow drum; reset blade gap; consider roll or belt discharge
Filtrate cloudyCloth pinholes, precoat breakthrough, seal bypassDye-test the cloth; verify precoat blade advance; check compartment gaskets

Routine maintenance centers on the cloth, the control valve, and the discharge hardware. Cloth life from two weeks on abrasive mineral duty to six months on gentle chemical duty is the dominant consumable cost, and recording cloth condition at each change against feed conditions is how plants learn which weave earns its price. The control valve wear plate and bridge blocks are inspected at cloth changes; discharge blades, belts, and rollers are dressed or replaced on condition; and the vat agitator bearings follow the plant lubrication schedule. A monthly hour-meter review of drum speed, vacuum, and cake moisture against the commissioning baseline catches slow drift before it becomes a capacity problem, and a spare cloth set on site is cheap insurance against a two-week lead time.

Instrumentation for that review is modest and worth specifying up front: a vacuum transmitter on the receiver, a flow meter on the wash water line, a temperature transmitter on the seal water return, a load cell or belt weigher on the cake conveyor to close the throughput balance, and drum speed as a hard number in the control system rather than a dial position. Seal water flow and temperature to the vacuum pump deserve interlocks, because a liquid ring machine that runs dry or hot loses vacuum silently until the cake turns wet. Plants that trend these six signals weekly find their drums boring, which is exactly what a continuous filter should be. The drive train deserves the same respect: a drum this size turns on a gearbox driven by an ac motor through a vfd, and the slow, torque-rich duty means the gearbox oil analysis tells you more about drum health than any operator observation, while cake leaving the enclosed discharge chute often passes through a rotary airlock on its way to the conveyor.

FAQ

How does a rotary vacuum filter differ from a filter press for dewatering?

The drum filters continuously at 0.6 to 0.8 bar of vacuum and discharges cake every revolution, while the press batches at 6 to 16 bar and reaches a drier cake with longer cycles and more operator attention. Choose the drum for continuous throughput, visible cake washing, and simple mechanics; choose the press when maximum dryness or closed wash is the economic driver.

What is precoat filtration and when do I need it?

Precoat operation builds a 50 to 100 mm bed of diatomaceous earth or cellulose on the drum and slices a fraction of a millimeter off with each revolution together with the thin slurry cake. It is the answer for slimy or blinding feeds and liquors that must run bright, at the cost of filter aid consumption and a cake diluted with that aid.

How is drum size actually determined?

From dry solids throughput divided by filtration flux, with flux measured in a leaf test on the real slurry at the real temperature and vacuum. Well-behaved slurries run 300 to 1000 kg per square meter per hour, difficult fine slurries 50 to 200, and flocculated biosolids 20 to 80. A drum sized without a leaf test is a guess wearing a drawing number.

Which discharge method should I specify?

Let the cake decide: 3 mm and coherent takes scraper discharge, sticky or blinding takes belt discharge with its self-washing cloth, thin tacky films take roll discharge, and clarification or slimy feed takes precoat. Specifying the wrong method shows up immediately as smeared cake and a cloth that never comes clean.

How much vacuum and air does the system need?

Most drums operate between 200 and 500 mbar absolute, and the liquid ring vacuum pump must be sized for the air deliberately drawn through the cake plus in-leakage at the control valve and seals, not just the filtrate vapor. Undersized air capacity shows up as lost vacuum and wet cake on hot days.

What typically goes wrong after commissioning?

Progressive cloth blinding, a control valve wear plate that leaks air, and wash sprays that overrun the drying arc account for most early problems. All three are addressed with cloth wash and cleaning in place, valve plate renewal at cloth changes, and a trim of the wash band, rather than major surgery.

How do I keep cake washing efficient?

Wash at a ratio of 1 to 1.5 kilograms of water per kilogram of dry solids with a uniform curtain from a trough or weir, not a coarse spray, and confirm the wash zone arc is not being squeezed by formation or drying demands. Displacement washing in the open arc removes 80 to 95 percent of solubles on permeable cakes, far more efficient than re-slurrying.

Sources

  1. ANDRITZ - Krauss-Maffei TSF vacuum drum filter
  2. ANDRITZ - CORES corrosion-proof vacuum drum filter
  3. Komline-Sanderson - Rotary drum vacuum filter
  4. Komline-Sanderson - Rotary drum vacuum filter bulletin (PDF)
Ask SpecForge AI