I’ve specified membrane bioreactors for three industrial projects and talked clients out of them on five others. The MBR vendors don’t like me much. But the technology isn’t a magic solution — it’s a tool with specific strengths and clear limitations.
Here’s what I’ve learned about where MBRs earn their keep, and where they become an expensive mistake.
What an MBR Actually Does
A membrane bioreactor combines biological treatment with membrane filtration. Instead of using a secondary clarifier to separate treated water from biomass, the MBR uses submerged membrane modules — either hollow fiber or flat sheet — with pore sizes of 0.01–0.1 microns. A permeate pump pulls clean water through the membranes while the biomass stays in the reactor.
The result: treated water that’s essentially free of suspended solids and bacteria, with a mixed liquor suspended solids (MLSS) concentration of 8,000–15,000 mg/L — two to three times higher than a conventional activated sludge system. Higher biomass concentration means a smaller tank for the same treatment capacity.
Sounds great. But here’s what the glossy brochures don’t emphasize.
The Three Real Costs of MBR
Membrane fouling. Membranes foul. Always. The question is how fast. Extracellular polymeric substances (EPS) produced by the biomass stick to the membrane surface. Inorganic precipitates — calcium carbonate, struvite, iron oxides — form scale. The result is declining permeability over time, requiring more frequent chemical cleaning.
In a well-designed MBR treating municipal wastewater, you might clean the membranes every 6–12 months. In an industrial MBR treating wastewater with high oil and grease, or variable pH, or certain solvents, you might clean every 4–8 weeks. Each cleaning cycle means downtime and chemical cost. The operators need to understand trans-membrane pressure trends and clean before permeability crashes — reactive cleaning is always more expensive than preventive.
Energy consumption. The aeration system in an MBR serves two purposes: supplying oxygen to the biomass (process air) and scouring the membrane surface to control fouling (membrane air). The membrane air is typically 50–70% of total aeration energy. A conventional activated sludge plant might consume 0.3–0.6 kWh/m³ of treated water. An MBR typically consumes 0.8–1.5 kWh/m³. For a plant treating 1,000 m³/day, the difference is $15,000–40,000 per year in electricity alone.
Membrane replacement. Membranes don’t last forever. In industrial service, expect 5–8 years before replacement. The replacement cost is typically 40–60% of the original membrane system capital cost — so you’re not buying the membranes once, you’re buying them every 7 years or so.
When MBR Is the Right Choice
MBR makes strong economic sense in these situations:
Tight discharge limits. If your permit requires BOD below 10 mg/L, TSS below 10 mg/L, and turbidity below 1 NTU, an MBR delivers this consistently. A conventional system with tertiary filtration can also meet these limits, but with more unit processes and more operator attention.
Space constraints. The higher MLSS means the biological reactor is 30–50% smaller. If land is expensive or unavailable, MBR saves real money on civil works. I’ve seen MBRs built inside existing buildings where a conventional clarifier simply wouldn’t fit.
Water reuse. If the treated water needs to go to reverse osmosis for reuse, the MBR provides excellent RO feed water quality with SDI (silt density index) consistently below 3. That means longer RO membrane life and fewer RO cleaning cycles.
Difficult-to-settle biomass. Some industrial wastewaters — pharmaceutical, chemical, some food processing — produce biomass that settles poorly in a conventional clarifier. The MBR eliminates settling as a concern entirely.
When MBR Is the Wrong Choice
High oil and grease. If your wastewater contains more than 50 mg/L of oil and grease consistently, the membranes will foul rapidly. You’ll need extensive pretreatment — dissolved air flotation, oil-water separation — before the MBR, and even then, membrane life will be shorter than projected.
Highly variable flow or load. MBRs handle some variation, but a plant that goes from zero flow to full flow in 30 minutes, or sees 5x swings in organic loading, will struggle. The biological system can’t adapt that fast, and the membranes suffer when sludge characteristics change rapidly.
Limited operator expertise. An MBR requires a higher level of process understanding than a conventional plant. The operators need to understand TMP trends, permeability calculations, chemical cleaning protocols, and the relationship between MLSS and membrane performance. If your operators’ primary experience is with conventional activated sludge, budget for significant training — and expect some painful lessons along the way.
The Pilot Test Decision Framework
If you’re considering MBR for an industrial application, run a pilot. Not a bench-scale test — a pilot with actual membrane modules treating a side stream of your actual wastewater for at least 3 months. During the pilot, track:
– Permeability decline rate (LMH/bar/day or similar)
– Chemical cleaning frequency and chemical consumption
– Energy consumption (kWh/m³) under normal operation
– Effluent quality at different MLSS concentrations
– Membrane autopsy findings if permeability crashes
The pilot will tell you more in 3 months than a year of reading technical papers. And if the vendor won’t support a pilot, that’s information too.
MBR is a powerful technology that has transformed what’s possible in industrial wastewater treatment. But it’s not a universal solution. The best MBR installations I’ve seen are the ones where the decision was made based on the specific wastewater characteristics and treatment goals — not on a vendor’s promise that “MBR can handle anything.” Because it can’t. And knowing what it can’t handle is just as important as knowing what it can.