Membrane bioreactors add duties conventional STPs never have: suction-limited permeate extraction, high-rate return sludge, membrane tank mixing and chemic
Membrane bioreactors earn their place where reuse quality, footprint or conventional-clarifier limits demand them: mixed liquor is filtered directly through ultrafiltration membranes that hold biomass on one side and pass permeate clear enough for reuse, deleting the secondary clarifier and its footprint while raising biomass concentration far beyond conventional levels. The biology is familiar — aeration, organic removal, nitrification — but the hydraulics around the membrane tank create duties no conventional plant contains, and each demands its own pump logic. Permeate extraction is the defining one. Water must be drawn through membrane pores under vacuum or low-head suction, and the driving force available is limited by transmembrane pressure — a fraction of a bar — so the permeate pump operates in a regime where every metre of lost head is a fraction of the total budget, not an efficiency footnote. Its flow is further governed by flux: litres through each square metre of membrane per hour, a budget set by membrane area and fouling state, which the pump cannot exceed without dragging fouling forward.
Return sludge is the second added duty, and it inverts conventional sizing. A conventional plant returns maybe half to all of its average flow to maintain biomass; an MBR circulates mixed liquor through the membrane tank at multiples of feed flow to keep membranes scoured and solids suspended across the tank, so RAS pumps handle high solids concentrations at high rates, running nearly continuously with duty-standby coverage. Then the supporting cast: membrane aeration is usually blowers, but tank mixing and decant duties may still need pumps; cleaning-in-place systems alternate sodium hypochlorite and citric or sulfuric acid through the membranes, and their transfer and recirculation pumps see chemistry that destroys unverified elastomers; and permeate systems themselves — backwash pumps reusing filtered water to clean the membranes — close the loop. Selecting pumps for an MBR means reading each duty against flux, transmembrane pressure and cleaning chemistry, and the plants that do get membrane systems whose performance decays slowly and predictably; plants that specify MBR pumps from a conventional-STP catalogue meet fouling, cavitation and elastomer failures their conventional experience never taught them to expect.
The membrane tank's pumping duties, duty by duty
Each duty in an MBR carries its own selection contract:
- Permeate extraction — the duty physics defines: water pulled through hollow fibre or flat sheet membranes under suction, with available head capped near transmembrane limits of a few tens of kilopascals; permeate pumps are selected for low head, accurate flow pacing and gentle treatment of a liquid that is already clean — typically self-priming or flooded-suction centrifugal machines on VFD, sized so their comfortable duty range brackets the flux budget; every suction-side loss is designed out, because a metre of line loss on a conventional pump is trivia and on a permeate system is a tenth of the driving force; the NPSH discipline from wastewater pump selection runs here at its most sensitive, on clean water with almost no head to spare.
- Return activated sludge — high rate, high solids: RAS pumps circulate mixed liquor at 8,000–12,000 mg/L through the membrane tank at ratios the membrane supplier's process design fixes, commonly several times the feed flow; the machines are conventional in family — non-clog submersibles or dry-pit solids-handling pumps — but duty-standby coverage is stricter than conventional practice because membranes scoured by circulation foul fast when circulation stops; the RAS sizing logic from sewage pump redundancy applies with an MBR multiplier on both rate and consequence.
- Membrane tank mixing and decant support: where tanks need supplemental mixing beyond membrane aeration, jet mixing loops fed by pumps keep solids off tank floors between aeration cycles; duty here resembles the equalization-basin mixing interplay — pumps drawing from the same tank whose process they serve, selected so their drawing zone and the membranes' zone cooperate rather than compete.
- Cleaning chemical transfer and recirculation: maintenance and recovery cleaning push sodium hypochlorite and organic acids through membrane modules; transfer pumps move chemicals from storage to day tanks, and recirculation or injection pumps see the chemicals at working concentration; elastomer and metallurgy verification against both chemistry families — oxidiser and acid alternately — is non-negotiable, and the chemical-duty pump selection discipline from ETP plants applies directly, because a stator or seal specified for one chemistry meets the other within a fortnight.
- Permeate backwash and reuse duty: some designs backwash membranes with their own permeate, pumped back through the fibres in reverse; backwash pumps handle clean water at moderate pressure with flow accuracy that matters — over-backwashing wastes flux budget and treated water, under-backwashing lets foulants consolidate — so the machines are selected like precise chemical metering of clean water, with VFD pacing slaved to the membrane control system's cleaning sequence.
- Feed and biological-train duties unchanged: upstream screening, anoxic mixing, aeration blowers and waste sludge pumping follow conventional logic — the screening discipline and sludge-transfer selection carry over directly; the MBR's unique requirement is that these conventional duties feed a membrane system intolerant of upstream lapses, so screening upstream of membranes is held to finer standards than conventional practice, because hair and fibres that survive conventional screening wrap hollow fibres and never let go.
Selection decisions specific to membrane plants
The specification calls that separate reliable MBR pumping from fouled expectations:
- Permeate pumps sized to flux, not to plant flow: the membrane supplier's flux budget — litres per square metre per hour across total installed membrane area — caps permeate flow; pumps sized beyond the budget invite operators to run flux higher than fouling control supports, and the penalty arrives as accelerated fouling that cleaning cannot fully reverse; the pump selection conversation starts from the membrane datasheet, not the hydraulic flow diagram.
- Suction geometry engineered like a vacuum system: flooded suction preferred wherever tank geometry allows; self-priming arrangements verified against the actual suction lift with margins the transmembrane budget can afford; suction lines short, large-dimensioned and free of high points that trap air — every design choice made against a head budget an order of magnitude tighter than conventional pumping; the air-handling lessons from pump station priming apply at their most unforgiving.
- RAS capacity at design ratio with standby and proving: circulation pumps sized to the membrane process design's return ratio, installed as duty-standby pairs with automatic alternation, and their standby honesty proven on schedule; membranes that lose circulation for hours consolidate foulant and recover slowly, so RAS redundancy is treated as membrane protection, not pump convenience.
- Pacing integrated with the membrane control system: permeate extraction follows tank level and flux schedules, RAS follows the process ratio, backwash follows cleaning sequences — all slaved to the membrane system's PLC rather than running on independent float logic; MBR pumping is a process-control performance in which each pump is an instrument, and control integration is a specification line, not a commissioning afterthought.
- Chemical-duty verification across the cleaning cycle: every pump that touches cleaning chemicals — transfer, injection, recirculation — carries elastomer and metallurgy selections verified against hypochlorite at cleaning concentration and organic acids at recovery-cleaning concentration, with rinse-state operation between; the failure mode to design out is the pump specified for one chemistry meeting the other on a Tuesday.
- Solids tolerance on everything that sees mixed liquor: any pump drawing from biological tanks — RAS, mixing, decant — carries the non-clog solids discipline: open passages, hardened shafts, sealing that survives 12,000 mg/LMLSS slurry, because mixed liquor is sewage in concentrated form and every conventional sewage-pumping wear lesson applies at higher intensity.
- Monitoring that watches flux and head, not amps alone: permeate flow trending against transmembrane pressure is the membrane system's health chart, and pump-side instrumentation — flow meters, suction pressure transmitters — feeds it; a plant whose permeate pumps lack accurate flow measurement is flying its most expensive asset blind, and the instrumentation budget belongs in the pump scope.
Common mistakes in MBR pumping
These recur as membrane plants meet conventional-STP habits:
- Permeate pumps sized on plant flow with conventional margins. Oversized permeate machines let operators chase flow beyond the flux budget, and membranes repay the ambition with fouling curves that steepen beyond recovery; the flux datasheet, not the flow diagram, sets the pump size, and the discipline is process design, not conservatism.
- Conventional suction assumptions on a suction-limited duty. Designs that add normal suction-line losses, lift and fittings to a permeate circuit spend head the membrane budget does not contain; symptoms are flow starvation at unchanged pump speed, and the diagnosis surprises teams until someone tabulates the head budget — at which point the redesign is line geometry, not bigger pumps.
- RAS treated like conventional return sludge. Single RAS pumps sized to conventional return ratios leave MBR tanks under-scoured and membranes fouling on schedules the supplier's warranty never assumed; MBR return ratios and duty-standby coverage are membrane requirements, and retrofitting them after fouling damage costs more than every RAS pump ever specified.
- Cleaning pumps specified for one chemistry. Hypochlorite-rated elastomers meeting citric acid cycles, or vice versa, degrade on contact, and the degradation pieces travel to membranes as debris; every chemical-path pump carries a two-chemistry verification, and rinse-state duty counts because rinse water between cycles carries both residuals.
- Independent float logic on an integrated process. Permeate, RAS and backwash pumps each running their own float settings fight the membrane PLC's cleaning and relaxation sequences, and the fights show up as flux instability; control integration under the membrane system's PLC is the design, and local autonomy is the fallback, never the default.
- Ignoring hair and fibre at screening. Membranes reject what screens miss, and hollow fibres collect stringy material into wraps that hands must remove module by module; screening ahead of MBRs runs finer than conventional practice with redundancy on screening units, because the membrane tank cannot be defended by pumps — only by what the screens let through.
- Measurement starvation. Plants that skipped permeate flow meters and suction pressure transmitters to save instrument budget cannot trend flux or transmembrane pressure, cannot verify cleaning effectiveness, and discover fouling through capacity loss; the instruments are the membrane system's vital signs, and their cost is rounding error against membrane replacement.
Frequently asked questions
Why does MBR permeate pumping run at such low pressure, and what does that mean for pump selection?
Because the driving force is capped by membrane physics, not by pump capability. Ultrafiltration membranes pass water under small transmembrane pressures — typically well under a bar across the membrane — so the entire head budget available to a permeate pump is a few metres, unlike conventional pumping where dozens of metres give designs slack. Every suction-side loss, lift and fitting consumes a meaningful share of that budget, which is why permeate systems are engineered like vacuum applications: flooded suction where possible, short and generous suction lines, no air traps, and pump duty points verified against the actual transmembrane limit. The pump itself is usually unremarkable — a low-head centrifugal machine on VFD — and the engineering lives in the circuit around it. The operational consequence is discipline: permeate flow is paced to the flux budget the membrane design sets, and the pump that could deliver more is deliberately not allowed to, because exceeding flux purchases fouling that cleaning can only partly reverse.
How is MBR return sludge pumping different from a conventional STP's RAS?
Three differences matter, and all of them push toward heavier duty. First, rate: MBR designs circulate mixed liquor through the membrane tank at ratios of several times feed flow — far above conventional return practice — because circulation scours membranes and keeps solids suspended across the tank, so RAS pumps run at high continuous loads rather than intermittent conventional duty. Second, concentration: mixed liquor at 8,000–12,000 mg/L is thicker than conventional return sludge, demanding the solids-handling construction — open non-clog passages, hardened shafts, robust sealing — that sewage pumping practice already knows, applied at higher intensity. Third, consequence: a conventional plant whose RAS stops degrades slowly; an MBR whose circulation stops begins consolidating foulant on membranes within hours and recovers slowly even after flow resumes. That is why MBR practice treats RAS pumps as membrane protection equipment — duty-standby pairs, automatic alternation, standby proving on schedule — rather than as ordinary sludge pumps with a spare on the shelf.
What should be verified when selecting pumps for the membrane cleaning system?
Verification against both chemistry families in the cleaning cycle, plus the transitions between them. Maintenance cleaning uses sodium hypochlorite at working concentration; recovery cleaning uses citric or sulfuric acid; between them, rinse water carries residuals of both. Every pump in the chemical path — storage transfer, day-tank duty, injection and recirculation — needs elastomers and metallurgy confirmed against oxidiser, against acid, and against the mixed residuals of rinse state, because the commonest failure is a pump specified for one chemistry meeting the other two within the same fortnight. Beyond materials, the cleaning pumps' flow accuracy matters: cleaning protocols specify chemical dose and contact flow, and pumps that approximate both deliver under-cleaned membranes that fouling reclaims quickly. Finally, the controls: cleaning sequences belong to the membrane PLC, and the pumps execute its steps — isolation, fill, soak, recirculate, rinse — as instruments rather than as independent machines with their own float logic. Verify materials across the whole cycle, accuracy to protocol, and integration to the PLC, and the cleaning system protects the membranes as designed.
Can a conventional STP be upgraded to MBR by keeping most existing pumps?
Partially, and the honest accounting separates what carries over from what the membranes bring. Biological-train equipment — anoxic mixers, aeration blowers, feed pumping, waste sludge duty — often carries over with verification that its duty points still suit the higher mixed-liquor concentrations an MBR runs. What almost never carries over is the membrane tank's own machinery: permeate extraction is a new duty with suction-limited physics no conventional plant contains; RAS must be re-rated to MBR circulation ratios and redundancy; screening must be upgraded to membrane-grade fineness because hair that conventional screens tolerated wraps fibres; and cleaning systems are entirely new chemical infrastructure. The economic logic of MBR retrofits usually assumes reuse of civil works and biological equipment while adding the membrane package — pumps included — as a designed system from the membrane supplier's process design. Teams that try to stretch conventional RAS pumps and screening into membrane service meet fouling and wrapping problems that cost more than the correct package would have; the retrofit saves civil works, not membrane-tank pump engineering.
Need help selecting the right pump?
Share your flow, head, liquid type, solids, site layout and duty cycle with Flow Chem Pumps. Our team can help you shortlist the right pump.