MBR at Fifty: What Membrane Bioreactors Actually Changed in Wastewater Practice

MBR, short for Membrane Bio-Reactor, joins the activated-sludge process to membrane separation. On paper that is a small change: replace the secondary clarifier with a membrane. In practice it breaks a coupling that had constrained biological treatment for decades, and almost every advantage and every cost of the technology follows from that one substitution.

The constraint it removes

In conventional activated sludge, solid-liquid separation depends on gravity in the secondary clarifier, and the quality of that separation depends on how well the sludge settles. Settling, in turn, depends on how carefully the aeration tank is run. Because the clarifier has to cope, the aeration tank cannot hold a high sludge concentration — typically about 1.5–3.5 g/L — which caps the biochemical reaction rate.

There are other entanglements. Hydraulic retention time (HRT) and sludge retention time (SRT) are bound together, so raising volumetric load and lowering sludge load pull against each other. Excess sludge is substantial, and its disposal can take 25%–40% of a plant's operating cost. Sludge bulking is a standing risk, and suspended solids escaping in the effluent degrade the product water. MBR decouples SRT from HRT, holds a far higher sludge concentration, allows slow-growing specialists to establish themselves, and by pushing the F/M ratio down cuts excess sludge production, in theory to zero.

Configurations and their trade-offs

The split-type MBR keeps the membrane module outside the bioreactor. Mixed liquor is pumped under pressure to the filtration side; liquid passes through as product water while solids and macromolecules return with the concentrate. It runs reliably, is easy to clean, replace and expand, and generally delivers high flux. The cost is energy: holding cross-flow velocity high enough to limit fouling means large recirculation flows, and the shear from high-speed pumping can inactivate some microbial cells.

The immersed or integrated type puts the module inside the bioreactor and draws permeate by suction. Dropping the recirculation system cuts energy use and shrinks the footprint, which is why it has drawn most of the attention in recent years. The penalties are lower membrane flux, a greater tendency to foul, and membranes that are awkward to clean or replace once fouled.

Two specialised variants are worth knowing. The aerated MBR uses silicone-rubber or hydrophobic microporous membranes in plate or hollow-fibre form to achieve bubble-free aeration with gas partial pressure held below the bubble point, improving contact time and oxygen-transfer efficiency while making aeration control independent of bubble size and residence time. The extractive membrane bioreactor, EMBR, was developed for wastewaters too acidic, too alkaline or too toxic for direct microbial contact; wastewater flows inside the membrane and activated sludge outside, so pollutants diffuse across and are degraded without the organisms ever meeting the raw stream. Because the two sides are hydraulically independent, HRT and SRT can each be held in their own optimum range.

What comes out of the other end

Membrane separation produces water far clearer than any settling tank: suspended solids and turbidity close to zero, bacteria and viruses largely removed. Effluent exceeds the Domestic Reclaimed-Water Quality Standard CJ25.1-89 issued by the Ministry of Construction and can be reused directly as non-potable municipal reclaimed water.

Because microorganisms are fully retained, the reactor holds a high microbial concentration, which raises volumetric load, saves floor area, and lets the plant ride out swings in influent quality and quantity. Nitrifying bacteria survive and accumulate, so nitrification improves and refractory organics get a longer effective residence time. The plant can be built above ground, semi-underground or underground, which makes site constraints much less of an obstacle.

The energy bill nobody escapes

None of this is free. MBR consumes more energy than conventional biological treatment for three compounding reasons: membrane driving pressure has to be maintained, the MLSS concentration in the tank is high so aeration intensity must rise to keep oxygen transfer adequate, and flow velocity across the membrane has to increase to scour the surface and hold flux. Any evaluation should start with that number rather than with footprint.

Where it has been proven

The timeline is longer than the marketing suggests. The first MBR wastewater plant was built by Dorr-Oliver in the United States in 1967, handling 14 m³/d. Japan put a water-reuse system into a high-rise building in 1977 and built 10 m³/d and 50 m³/d plants in 1980; by the early 1990s, 39 such plants were running there with a maximum capacity of 500 m³/d, and more than 100 high-rise buildings used MBR for reclaimed water. Canada's Zenon launched an ultrafiltration tubular MBR and later an immersed hollow-fibre module, with installations across the United States, Germany, France and Egypt at scales from 380 m³/d to 7,600 m³/d. Kubota's plate membranes and Mitsubishi Rayon's hollow-fibre products established the immersed format in commercial use.

Industrial applications broadened through the 1990s. An MBR built in Ohio in the early 1990s for an automobile manufacturer's wastewater ran at 151 m³/d with an organic load of 6.3 kgCOD/m³·d and 94% COD removal, degrading most of the oil and grease. A Dutch fat-extraction plant that had suffered persistent bulking replaced its settling tank with a membrane module and resolved the separation problem. Food, aquaculture, livestock, cosmetics, dye and petrochemical wastewaters all showed good results.

High-strength streams show what retention can do. The Japanese NS system for urine and faeces, built at production scale in Koshigaya in 1985, pairs flat-sheet membrane panels with an aerobic high-concentration activated-sludge reactor, holding sludge at 15,000–18,000 mg/L; by 1994 more than 1,200 such systems were serving over 40 million people. Landfill leachate has been treated by MBR since before 1994, and a unit commissioned in New Jersey at the end of 2000 handled concentrations 50–100 times those of conventional equipment by retaining bacteria at 50,000 g/L; in field pilots with influent COD from several hundred up to 40,000 mg/L, pollutant removal ran above 90%. For drinking water, a plant at Douchy in France produced 400 m³/d from 1995 with nitrogen below 0.1 mgNO₂/L and pesticide below 0.02 μg/L.

Chinese work followed a similar curve. An integrated MBR pilot developed at Tsinghua University passed national appraisal in May 1998, and a practical system for hospital wastewater was commissioned at Haidian Township Hospital in Beijing in June 2000. A demonstration project built at Puchen Building in Tianjin New Technology Industrial Park in September 2000 treats 25 tonnes of sewage per day, occupying 10 m² and consuming 0.7 kWh per tonne, with all effluent reused for toilet flushing and green-space irrigation.

Open questions

MBR is mature but not finished. Two areas still drive development: establishing the maximum economically viable flow rate under current membrane supply and design standards, and building new reactor configurations that cut energy, handle special water qualities, remove nitrogen and phosphorus simultaneously, and stay easy to operate over long runs. Used for advanced treatment, MBR can also sit ahead of RO desalination, replacing the sand-filtration, security-filtration and ultrafiltration train and extending membrane life by dropping the organic load.