What Makes Hollow Fiber MBR Criteria Different
Hollow fiber and flat plate are both submerged MBR configurations, but they impose different upstream and operating envelopes, and a flat-sheet datasheet copied onto a hollow fiber project will mis-size screening, aeration and maintenance. The EPA membrane bioreactor fact sheet is explicit on the split: "hollow fiber membranes typically require 1- to 2-mm screening, while plate membranes require 2- to 3-mm screening" (Wallis-Lage et al. 2006, cited in the EPA fact sheet). Hollow fibers are grouped in bundles manifolded into cassettes; the EPA fact sheet describes GE/Zenon ZeeWeed units as "hollow-fiber, tubular membranes configured in bundles" with a number of bundles connected by manifolds into units that "can be readily changed for maintenance or replacement."
Fouling control is the second divergence. EPA states that "all the principal membrane systems used in MBRs use an air scour technique to reduce buildup of material on the membranes," but adds that GE/Zenon systems add a back-pulse step: "permeate is occasionally pumped back into the membranes to keep the pores cleared out. Back-pulsing is typically done on a timer, with the time of pulsing accounting for 1 to 5 percent of the total operating time." Hollow fiber therefore carries one extra fouling-control lever that flat sheet does not, at the cost of 1–5% of operating time and a permeate-side backwash pump. The other lever is geometry: EPA puts the maximum membrane pore size at "about 1 micron (0.001 millimeters)" so microorganisms remain in the system; the 0.1 µm PVDF specification common in commercial hollow fiber cassettes sits an order of magnitude tighter, with corresponding consequences for fouling rate and air-scour demand. The cassette is also a maintenance criterion: manifolded bundles are designed to be lifted out for service, so deck height, crane access and cassette count are datasheet items rather than site-layout details.
Influent and Biological Criteria for a Hollow Fiber MBR
The biological envelope for a hollow fiber MBR is set by PD CEN/TR 15897 (BSI British Standards), which fixes two baselines: carbon removal is mandatory and complete nitrification is recommended (S1 per scraped S2). PD CEN/TR 15897 also lists the filterability risks the biological design must mitigate — short circuits, elevated extracellular polymeric substances (EPS) and incomplete biodegradation (S1 per scraped S2). The reason those risks matter more in an MBR than in conventional activated sludge (CAS) is that the membrane completely retains biomass, so the secondary clarifier is removed and the S2 pharmaceutical study (Springer, Anal. Bioanal. Chem.) reports that "MBR biomass can be raised to about 20 mg/L, an order of magnitude beyond the typical mixed-liquor envelope of CAS." Higher MLSS cuts the aerobic volume for a given food-to-microorganism ratio, increases the solids inventory the membrane must hold, and forces the designer to set an SRT/HRT balance that nitrifies without starving the biology. The S3 source (Seven Seas Water) states that MBR tank SRT and HRT typically range from days to weeks, well above CAS, and PCI Membranes (S4 per scraped S2) states the resulting footprint can be up to 50% smaller than a CAS plant.
What the datasheet cannot take from a generic table is the influent envelope: temperature, pH, salinity, fats/oils/greases (FOG) and the C/N ratio. The scraped sources do not publish default numeric ranges for these parameters, so the specifier should pull them from the client's influent characterisation or from a pilot rather than from a vendor brochure. Two equipment items that the influent criteria typically drive are a GX Series rotary mechanical bar screen at headworks for fine screening, and a DAF unit for FOG and suspended solids reduction ahead of the membrane tank, so the biological tank receives a feed that does not poison the nitrifiers or smother the cassettes. PD CEN/TR 15897 also treats the SRT/HRT window as a design lever because long SRTs support nitrification and the biodegradation of slowly degraded compounds — a benefit confirmed in the S2 study, where most of the monitored pharmaceuticals were removed to greater than 80% during MBR treatment.
Hollow Fiber Membrane Criteria: Pore Size, Geometry, Flux, Back-Pulse
The membrane-side numbers a specifier needs are pore size, fiber geometry, cassette layout, back-pulse duty and net flux, anchored in EPA and PCI Membranes evidence. EPA states that the maximum membrane pore size is set during manufacturing at "about 1 micron (0.001 millimeters)" so that microorganisms remain in the system; commercial 0.1 µm PVDF hollow fiber is the common MBR specification, with 0.4 µm used in some flat-sheet references (Kubota panel, 0.4 µm, per the S2 lab study). EPA lists three MBR geometries — "hollow fibers grouped in bundles," "flat plates," and tubular — and the S4 source (PCI Membranes per scraped S2) explicitly rules out spiral-wound for MBR duty because of its sensitivity to suspended solids. EPA also states that GE/Zenon systems use "hollow-fiber, tubular membranes configured in bundles" that "can be readily changed for maintenance or replacement," and that Zenon offers a 10-year membrane guarantee, with other suppliers offering 3–5 years; some guarantees are tied to screen size.
Back-pulse and intermittent permeation are documented design levers rather than operating details. The S2 lab study ran a Kubota flat-sheet panel at 8 minutes permeation / 2 minutes relaxation; the same intermittent-permeation principle applies to hollow fiber but is paired with the 1–5% back-pulse duty that EPA attributes to GE/Zenon. Generational efficiency is a supplier-specific input: the PCI Membranes second-generation cassette launched in 2022 achieved "+10% membrane packing density and −5% scouring energy versus the prior generation" (S4 per scraped S2), and a hollow fiber specifier should request equivalent generational efficiency data from the supplier being bid. Net flux, recovery and backwash interval are project-specific inputs that the scraped sources do not give as a default range, so the writer treats net flux as a value set against the supplier's published curve and the influent fouling index rather than copying a generic number. The reference operating point is the HydropureWater DF series flat-sheet module (0.1 µm PVDF, 80–225 m² per module, 32–135 m³/d per module, integrated aeration box) — a flat-sheet benchmark against which a hollow fiber supplier should be asked to demonstrate flux per cassette, packing density per cassette, and aeration energy per m³ of permeate.
| Membrane-side criterion | Value or scope | Source / date |
|---|---|---|
| Pore size (EPA maximum) | About 1 µm (0.001 mm), set during manufacture | EPA MBR fact sheet |
| Common hollow fiber specification | 0.1 µm PVDF | Commercial MBR datasheets (per scraped S2) |
| Reference flat-sheet pore (Kubota, lab) | 0.4 µm | S2 lab study (Anal. Bioanal. Chem.) |
| Membrane geometries | Hollow fiber, flat plate, tubular; spiral-wound not suitable | EPA fact sheet; PCI Membranes per scraped S2 |
| Back-pulse duty | 1–5% of total operating time (timer-controlled) | EPA MBR fact sheet |
| Intermittent permeation (Kubota, lab) | 8 min on / 2 min relaxation | S2 lab study |
| PCI Membranes 2nd-gen cassette (2022) | +10% packing density, −5% scouring energy vs prior generation | PCI Membranes per scraped S2 (2022) |
| Membrane guarantee | Zenon 10 years; other suppliers 3–5 years (tied to screen size) | EPA MBR fact sheet |
| Net flux, recovery, backwash interval | Project-specific; not given in scraped sources | Request from supplier |
The cassette selection is anchored to a flat-sheet reference such as the HydropureWater DF series flat sheet module when the design calls for submerged flat-sheet operation, with the equivalent RO/UF membrane filter elements covering the downstream polishing train where reuse is required.
Aeration, Hydraulic and Pretreatment Criteria
Aeration serves two roles in a hollow fiber MBR — oxygen supply to the biomass and continuous scouring of the membrane surface — so the air system must be sized for the higher of the two demands. The S3 and S4 sources both make the rule explicit: scour air, not process air, usually governs the blower at high MLSS, because the air demand for membrane scouring scales with membrane area and is independent of biological loading. EPA adds the hollow-fiber-specific layer: GE/Zenon systems add back-pulsing on a timer, with the pulsing time accounting for 1 to 5 percent of total operating time, so the backwash pump and the backwash flow split are part of the hydraulic envelope from day one. The PCI Membranes second-generation cassette with −5% scouring energy is the kind of evidence a specifier should ask a hollow fiber supplier to match, because scouring energy at constant MLSS is the single largest operating cost in most MBRs.
The hydraulic envelope is anchored in three EPA rules. First, "peak design flows should be no more than 1.5 to 2 times the average design flow"; if peak flows exceed that limit, additional membranes or equalisation (external basin, or internal equalisation via elevated tank levels) must be added. Second, for full-treatment MBR installations, the "N+1" concept is recommended: "one additional membrane tank/unit beyond what the design would nominally call for," to give operators flexibility and ensure sufficient operating capacity during maintenance (Wallis-Lage et al. 2006, cited in the EPA fact sheet). Third, all MBR systems require 1- to 3-mm fine screens immediately before the membranes, depending on the MBR manufacturer; for hollow fiber that is the tighter 1–2 mm window, which raises the headworks load but protects the smaller fiber lumen. The pretreatment train typically combines a rotary mechanical bar screen at headworks, grit removal, a DAF unit for FOG, and a polishing fine screen at the membrane tank, with a chemical dosing system available for phosphorus precipitation or for pH trim ahead of the membranes — see the automatic chemical dosing system as a reference configuration. The hydraulic design must also cover the permeate/backwash flow split and a backwash recovery loop so that cleaning water is not lost to drain (S2). Pressure-driven sidestream loops are appropriate for small or hard-to-treat industrial streams (PCI Membranes per scraped S2); for hollow fiber, submerged is the default geometry at medium-to-large flows.
Hollow Fiber vs Flat Sheet vs Sidestream: When to Choose Hollow Fiber
The geometry decision should be made against a single matrix, not against three separate vendor datasheets. EPA states that both hollow fiber and flat plate MBRs are commercially proven, that hollow fiber uses "hollow-fiber, tubular membranes configured in bundles" and flat plate uses panel-and-manifold cassettes, and that screening requirement is 1–2 mm for hollow fiber and 2–3 mm for flat plate (Wallis-Lage et al. 2006). Hollow fiber adds a back-pulse cycle for fouling control; flat-plate systems rely on air-scour and relaxation only — the extra lever consumes 1–5% of operating time and a permeate-side pump. PCI Membranes (per scraped S2) gives the matching geometry rule: pressure-driven sidestream membranes suit smaller installations and tough-to-treat industrial wastewaters, while submerged configurations suit medium-to-large installations. PD CEN/TR 15897 (S1 per scraped S2) is explicit that above 10,000 m³/d, large MBRs are described as tending to use separated membrane tanks rather than an integrated reactor, and that "membrane interchangeability between manufacturers is not possible at the module level without considering the full system."
| Parameter | Hollow fiber (submerged) | Flat sheet (submerged) | Sidestream (pressure-driven) |
|---|---|---|---|
| Screening before membranes | 1–2 mm | 2–3 mm | Coarser; depends on loop |
| Fouling-control mechanism | Air scour + back-pulse (1–5% of operating time) | Air scour + relaxation only | Cross-flow velocity; no air scour |
| Pore size (typical) | 0.1 µm PVDF | 0.1–0.4 µm PVDF | UF range |
| Default flow envelope | Medium-to-large (submerged) | Medium-to-large (submerged) | Small or hard-to-treat industrial |
| Maintenance access | Cassette lift-out via manifolded bundles | Panel swap-out | Loop isolation, no cassette |
| Module interchangeability | Not at module level without full-system review (PD CEN/TR 15897) | Same constraint | Same constraint |
The decision rule: choose hollow fiber for medium-to-large flows where back-pulse is available and the operator can keep a tighter 1–2 mm screen; choose flat plate for flows with higher FOG/rag load where 2–3 mm screening and easier plate swap-out reduce maintenance risk; choose sidestream for small or hard-to-treat industrial streams where pressure-driven cross-flow is acceptable. Above 10,000 m³/d, separate membrane tanks become the default and the design should be evaluated at the full system level, not at the cassette level. For a packaged plant in the small-to-medium municipal range, an integrated MBR membrane bioreactor system matches the lower-scale envelope set by PD CEN/TR 15897.
Validation, Effluent Quality and Compliance Floor for 2026
The 2022 J. Env. Management envelope (per scraped S2) sets a defensible compliance floor for a 2026 hollow fiber MBR datasheet: greater than 99% TSS removal, greater than 91% COD removal, and 68.3–99.7% removal across a range of microcontaminants. The S2 pharmaceutical study (Springer, Anal. Bioanal. Chem.) reported 98.7% TSS and 90.4% total COD removal on a laboratory-scale MBR, with most monitored pharmaceuticals removed at greater than 80% but carbamazepine as the persistent outlier at less than 20% — a result that is explicit that even a well-run MBR will not fully eliminate persistent micropollutants, so a polishing step (RO or an advanced oxidation process) is the defensible answer to a regulator. The S4 source (PCI Membranes per scraped S2) notes that the MBR effluent is essentially free of suspended solids with reduced bacterial and viral content, so minimum disinfection is required and the permeate is suitable as feed to a downstream reverse osmosis train for water reuse, or for discharge to sensitive receiving bodies. Field evidence from the S5 Water Environment Research 2026 study (per scraped S2) documents that MBR-treated wastewater has been used successfully for safflower irrigation under Mediterranean conditions, with MBR-irrigated yield statistically equivalent to freshwater irrigation, subject to salinity and sodicity monitoring.
Validation should mirror PD CEN/TR 15897's whole-system view (S1 per scraped S2): pilot trials, full-system acceptance, and O&M handover, not cassette-level tests in isolation. Real-time sensors and a control loop should detect the filterability watch list — short circuits, EPS spikes and incomplete biodegradation — not just verify effluent chemistry at the bench. Chemical cleaning and backwash chemistry, frequency and the backwash recovery loop are part of the criteria, and the datasheet should specify whether backwash recovery returns to a holding tank with recirculation or to waste. Membrane life is a cost-effectiveness lever tied directly to screen size, so the 1–2 mm hollow fiber screen directly supports the longer Zenon-class warranty the EPA fact sheet describes. For reuse applications, the polishing train is the industrial RO polishing train referenced for the design, with a UV sterilizer as the disinfection barrier.
Frequently Asked Questions
What is the 2026 cost driver breakdown for a hollow fiber MBR package?
The scraped sources do not publish a 2026 unit CAPEX or OPEX figure for an MBR package, so a price copied from a generic table would be a guess. The defensible approach is to break the cost into its drivers and request quotes against each: membrane area (set by net flux and MLSS), aeration energy (set by scour demand and biological loading — ask the supplier for the specific scouring energy per m³ of permeate, the same metric the PCI Membranes 2022 second-generation cassette improved by −5%), cleaning chemicals and backwash recovery, sludge handling, and the pretreatment train. Regional benchmarks for budget envelopes are available in the 2026 MBR cost, compliance and supplier guide and in the Norway MBR cost and compliance reference, which together give the figures a buyer should ask a supplier to match.
How do I select a hollow fiber MBR supplier against PD CEN/TR 15897?
Evaluate on full-system performance rather than datasheet parity. The specifier should require references at the target scale (above 500 PT and, if relevant, above 10,000 m³/d), documented effluent quality against the 2022 J. Env. Management envelope (greater than 99% TSS, greater than 91% COD), after-sales membrane supply and replacement lead time, and the supplier's own aeration energy data. The PCI Membranes 2022 second-generation cassette (+10% packing density, −5% scouring energy) is the kind of generational efficiency data a specifier should ask for in writing. PD CEN/TR 15897 (S1) is also clear that membrane interchangeability between manufacturers is not possible at the module level without considering the full system, so the cassette geometry and aeration box should be locked to a single supplier rather than mixed across vendors.
How do I size a hollow fiber MBR from net flux and MLSS?
Size from the consolidated criteria in the matrix and tables above: influent load, MLSS target, SRT/HRT, net flux and membrane geometry. Then check the result against PD CEN/TR 15897's two thresholds (S1): above 500 PT the design is treated as a custom MBR, and above 10,000 m³/d the design should use separate membrane tanks. The S2 lab reference gives a 0.4 µm Kubota panel at 0.106 m² per panel with ~21 L active bioreactor volume as a bench-scale operating point; the HydropureWater DF series flat-sheet module gives a commercial reference of 0.1 µm PVDF, 80–225 m² per module, 32–135 m³/d per module with an integrated aeration box. Net flux, recovery and backwash interval are not given as defaults in the scraped sources, so those values must come from the supplier's published curve against the project-specific fouling index, not from a generic datasheet.
What is the compliance risk for trace organics in a 2026 hollow fiber MBR?
Set the compliance floor at greater than 99% TSS and greater than 91% COD from the 2022 envelope, and add downstream polishing (RO or AOP) for persistent compounds such as carbamazepine, which the S2 study shows MBR alone does not remove (less than 20% removal). The S5 2026 safflower irrigation study (per scraped S2) shows reuse is feasible subject to salinity and sodicity monitoring, so the compliance criteria should include a salinity envelope and a monitoring plan, not only a pollutant list. Disinfection with a UV sterilizer ahead of the polishing train or reuse point is the standard barrier; the MBR permeate itself is already low in bacteria and viruses per the EPA fact sheet.
