Example of MBR application for wastewater reuse City of Key Colony Beach Florida MBR used as pretreatment for reverse osmosis for wastewater reuse2

By 1998, the City of Key Colony Beach's (Fig. 10) municipal wastewater treatment plant had reached its rated capacity, limiting development opportunities. The city anticipated that stringent effluent requirements of 5 mg/L biochemical oxygen demand (BOD) and total suspended solids (TSS), 3 mg/L total nitrogen (TN), and 1 mg/L total phosphorous (TP) would likely be imposed in the future because of the coastal discharge. The existing treatment facility would be unable to achieve this effluent quality without significant capital upgrade. In addition, the community wanted to irrigate the local golf course using recycled wastewater, but substantial salinity removal would be required because of seawater intrusion into the municipal sewer network.

The wastewater treatment technology selected was the ZenoGem® MBR, provided by GE Water and Process Technologies. The MBR process was selected for several reasons, among them the ability to achieve a high quality, particulate free effluent on a very compact footprint and the ability to generate an effluent suitable for direct RO treatment without pretreatment [i.e., MBR effluent silt density index (SDI)<3]. The ZenoGem® process was put in operation in June 1999 and the RO process was put in operation in December 1999. The ZenoGem® process had treated all flows to the wastewater treatment facility since information information provided by GE water Water & process Process technologiesTechnologies.

Examples Reverse Osmosis Projections
Figure 10 Aerial view of City of Key Colony.

commissioning, while the RO unit has been operated for only intermittent periods because the water distribution system has not yet been constructed.

ZeeWeed® are proprietary hollow fiber membranes that are immersed within the bioreactor in direct contact with the mixed liquor. The ZeeWeed® hollow fiber membranes are contained in bundles called modules, which are assembled into cassettes of 8-12 modules. The membrane modules are directly immersed in the aeration tank, in direct contact with the mixed liquor. Through the use of a centrifugal pump, a vacuum varying between 13.8 and 62kPa (2-9 psi) is applied to a header connecting the membrane modules. The vacuum draws the treated water through the hollow fiber membranes. The treated water passes through the hollow fibers and is pumped out by the permeate pump (see Fig. 11, "ZenoGem® Conceptual Process''). All particulate matter and the mixed liquor solids are rejected at the surface of the membrane. The ZeeWeed® membranes are automatically back-pulsed on a regular basis using collected permeate. A coarse bubble air diffuser is located at the base of each membrane module. The airflow provided by the diffuser scours the external surface of the membrane, transferring the rejected solids away from the membrane surface. This airflow also provides a portion of the biological oxygen requirements. Supplemental coarse or fine bubble diffuser grids may be used to supply the remainder of the biological oxygen

Figure 11 Key Colony MBR.

requirements. Sludge is wasted directly from the aeration tank at the operating MLSS concentration between 10,000 and 15,000 mg/L. The high biomass concentration allows the ZenoGem® process to be operated at reduced organic loading rates (i.e., low food/microorganism ratio) and elevated solids retention times (>15 days). Year-round nitrification is ensured because the operating SRT greatly exceeds the minimum SRT required for nitrification, which is typically 5-7 days under winter operating conditions. ZenoGem® bioreactors are ideally suited for denitrification as well. Since the ZeeWeed® membranes eliminate the need for secondary clarification, it is not necessary for the operators to concern themselves with the settling properties of the mixed liquor. The anoxic zone can be sized for optimal nitrogen removal and with the high MLSS concentrations, a total nitrogen removal efficiency of over 90% is readily achieved.

The aeration tank at Key Colony is separated into two trains, with each train divided into three distinct zones separated by concrete baffles (Zones 1, 2, and 3) as presented in Fig. 11. The raw wastewater, after passing through the rotating drum screens, is fed into Zone 1, which has a combined (Train 1+Train 2) operating volume of 210,000 L. A small amount of air is used for mixing in this zone, but the dissolved oxygen (DO) concentration is maintained less than 0.2 mg/L. The majority of denitrification occurs in this zone. The mixed liquor flows by gravity (through a submerged cutout) from Zone 1 to Zone 2. Zone 2 has a combined operating volume of 280,000 L and is aerated at a limited rate to achieve a DO concentration in the range of 0.2-0.8 mg/L. The DO in Zone 2 is maintained at a low enough concentration to allow both nitrification and denitrification to occur in Zone 2, minimizing the ammonia and nitrate concentration entering Zone 3. The mixed liquor flows by gravity (through a submerged cutout) from Zone 2 to Zone 3. Zone 3 has a combined operating volume of 260,050 L. This zone also contains the ZeeWeed® membranes, and is fully aerobic, being aerated by a grid of coarse bubble diffusers to achieve a DO concentration greater than 2 mg/L. Any of the remaining ammonia and soluble carbon (measured as BOD) will be oxidized in this zone. Mixed liquor from Zone 3 is recirculated back to Zone 1 at a flow rate of 6250L/min, approximately four to eight times the influent flow rate.

Since startup, the effluent BOD, TSS, and total nitrogen concentrations have remained below 5, 5, and 3 mg/L, respectively, and with alum addition, it is possible to achieve total phosphorus concentrations less than 1 mg/L. ZenoGem® effluent turbidity has consistently measured < 0.2 NTU. The RO unit has been operated for only intermittent periods because the water distribution system has not yet been constructed. There has been no evidence of RO membrane fouling, indicating that the ZenoGem® effluent was entirely suitable for direct feed to the RO. For the period of time when the RO unit has been operated, the reduction in conductivity has averaged greater than 98%. The total dissolved solids (TDS) of the RO permeate has been measured at less than 100 mg/L, resulting in a high-quality water entirely suitable for irrigation purposes.

Since the commissioning of the Key Colony plant, there have been a number of improvements to the membrane cassette configuration. Some of these improvements have already been implemented at Key Colony. For example, with the addition of automatic valves, the membrane air can now be cycled between the two trains, resulting in a close to 50% reduction in aeration requirements. As the membrane aeration makes up a substantial portion of the total operating costs of the ZeeWeed®/ZenoGem® system, this reduction in net aeration requirements will equate to significant operating cost savings over the life of the plant. Other improvements to the design include more efficient spacing of the individual membrane elements in the cassettes, resulting in the ability to increase membrane surface area per cassette by over 20%, while at the same time improving the efficiency of the membrane aeration system, resulting in a reduced membrane cleaning frequency. This new cassette configuration has been in operation at Key Colony since August 2000, and the performance of the new membrane cassettes has equaled or exceeded the performance of the original cassette configuration.

Table 7 Summary of results from MBR studies on wastewater reuse

Source and objective

MBR system details

MBR effluent quality


Ahn et al. [30]: Compared the performance of direct membrane filtration and MBR systems (municipal wastewater)


MLSS: 4000-7000 mg/L Membrane

Submerged hollow fibers


Pore size: 0.1 mm

COD: <9 mg/L BOD: <0.8 mg/L TOC: <3.8 mg/L Total suspended solids: < 0.3 mg/L Turbidity: <0.1 NTU Total nitrogen: < 10 mg/L Ammonia-N: < 2.2 mg/L Total phosphorous: <1.6 mg/L HPC assay: 0.02-4.2 x 104CFU/

100 mL Coliphage: 0PFU/100mL

Fouling resistance increased much faster in direct membrane filtration than in the MBR. Also, filtration resistance in the MBR was one order of magnitude lower than in the direct filtration system

Al-Malack [12]: Investigated effect of MLSS and organic loading rate on system performance (synthetic wastewater)

Bioreactor Volume: 20 L MLSS: 3000-15,000 mg/L OLR: 0.1-1.2 kg COD per kg MLSS per day Membrane

Submerged air sparged Tubular: 1.27 cm ID Material: polyester Pore size: 20-40 mm

Total suspended solids: 0 mg/L

Total coliform: 1.7 X 107 to 9 X 107

MPN/100mL Sludge production: 0.26 mg VSS/mg COD

Poor COD removal occurred immediately following sudden increases in organic loading rate. Increase in MLSS concentration increased the COD removal efficiency

Chae et al. [34]: Characterized effectiveness of MBR at organic and nitrogen removal at different HRTs (municipal wastewater)


Vertical anoxic(an)-aerobic(ar) system MLSS: 8700 mg/L(an),

Polytetrafluoroethylene Pore size: 0.45 mm

Total suspended solids: 100% removal COD: 94-97% removal Total nitrogen: 62-76% removal Total phosphorous: 42-77% removal Total coliform: 11 counts/100mL E.coli: 2 counts/100 mL

Nitrification efficiency decreased when hydraulic retention time (HRT) reduced to 4 h. Higher range of removal efficiencies corresponded to an HRT of 10 h and lower range corresponded to an HRT of 4 h

Cote P. et al. [50]: Demonstrated effects of high biomass concentrations, sludge age, and bioreactor configuration (municipal wastewater)


Two systems: one aerobic, one anoxic-aerobic MLSS: 5000-15,000 mg/L SRT: 10-50 days Membrane

Submerged air sparged hollow fibers MWCO: 200,000 Da

Total suspended solids: 100% removal COD: > 96% removal Ammonia-N: 80-99% removal Total nitrogen: 36-80% removal Total phosphorous: 15% removal Total coliform: >6 log removal Viruses (bacteriophage): 4 log removal

Sludge production was 0.25 kg TSS/kg COD removed, and was approximately 50% lower than for typical conventional system. Lower range of nitrogen removal for aerobic system and higher range for anoxic-aerobic system

Fatone et al. [38]: Demonstrated feasibility of MBRs in wastewater reuse applications (municipal wastewater)


Anoxic-aerobic system MLSS: 4800-9000 mg/L HRT: 6-8 h


Submerged hollow fibers Pore size: 0.04 mm

Total suspended solids: 0 mg/L COD: 4-11 mg/L Ammonia-N: 0.1 mg/L NO^-N: 0.9-5.2 mg/L Total nitrogen: 1.1-5.4 mg/L Total phosphorous: 31-57% removal

Approximately 80%

removal of trace aromatic hydrocarbons in wastewater (i.e., acenaphthylene, acenaphthene, fluorene, anthracene, chrysene)

Source and objective

MBR system details

MBR effluent quality


Guglielmi et al. [27]: Compared performances of MBRs and conventional tertiary systems in wastewater reclamation for irrigation (municipal wastewater)


Anoxic-aerobic system Volume: 7.9 m3 MLSS: 12,000 mg/L SRT: 12-15 days Membrane

Submerged hollow fiber Pore size: 0.04 mm

COD: 12.9 + 3 BOD: 2.0 + 0.5 TKN: 1.9 + 1.5 Ammonia-N: 0.6 + 0.5 Total nitrogen: 8.1+4.5 Total phosphorous: 1.2 + 0.8 TSS: < 1 NTU E. coli: 1 CFU/100mL

MBR proved to be more reliable than conventional activated sludge system with tertiary filtration, especially for microbial contaminants

Investigated ability of

MBRs to removal pharmaceuticals

(pharmaceutical manufacturing wastewater)

Bioreactor Volume: 2.7 m3 HRT: 14.8 h SRT: 20 days Membrane

Submerged hollow fiber Pore size: 0.04 mm

COD: 94% removal

BOD: 99% removal

Total suspended solids: non-detectable

Total nitrogen: 60% removal

Total phosphorous: 38% removal

Some hormones and oral contraceptives were removed to near or below detection limits. However, several key pharmaceuticals were resistant to MBR treatment

Innocenti et al. [13]: Determined effect of SRT and MLSS on performance of an MBR (municipal wastewater)


Batch anoxic/aerobic MLSS: 4000-17,000 mg/L SRT: 10 to > 200 days Membrane Pore size: 0.02 mm Submerged hollow fiber

COD: 19-40 mg/L Total suspended solids: 0 mg/L Ammonia-N: 0.2-0.5 mg/L Total nitrogen: 6.2-13.3 mg/L Total phosphorous: 0.9-1.1 mg/L

Sludge production at 10, 190, and >200 days was 0.56, 0.08, and 0.02 g MLVSS/g COD, respectively. Total nitrogen removal optimal at intermediate MLSS concentration

Jefferson et al. [52]:


COD: 2.5-31 mg/L

Slightly better removal for

Compared performance

Volume: 0.035 m3 (S),

Turbidity: 0.2 NTU

external MBR.

of external (E) and

0.38 m3 (E)

Total coliform: 3-5 log removal

However, the MBRs

submerged (S) MBRs

MLSS: 200 mg/L (S),

were operated under

(municipal wastewater)

430 mg/L (E)

considerably different




e 3

Flat sheet (S), tubular (E)

r a

Pore size: 0.4 mm (S),


4000 kDa (E)

i or e

Jefferson et al. [53]:


COD: <10 mg/L

Performance independent

t or

Compared performance

Volume: 0.035 m3

Total coliform: 7 log removal

of MLSS concentration


of MBRs, biological

HRT: 12 h

in the range of 400-

zr fD o

aerated filters (BAF) and


8000 mg/L. MBR met


membrane aeration

Submerged plate and frame

reuse requirements 100%


bioreactor (MABR)

Pore size: 0.4 mm

of the time, while BAF

(synthetic gray water)

and MABR systems did not

l c t o

Kumar et al. [54]:


BOD: <2 mg/L

Only results from MBR

s o

Investigated the use of an


TOC: <10 mg/L


Wa s

MBR prior to RO


Turbidity: <0.13 NTU


treatment (municipal


Total coliform: <2 MPN/100mL

t er


Flat sheet submerged

Coliphage: <10 PFU/100mL (80% of


Pore size: 0.4 mm

the time)

s e

Source and objective

MBR system details

MBR effluent quality


Li et al. [55]: Compared the performance of singlestage bioreactor configuration and powdered activated carbon (PAC) addition on nitrogen removal in MBRs (synthetic wastewater)

Bioreactor Volume: 17.2 L MLSS: 4100-13,500 mg/L Membrane

Submerged hollow fibers Pore size: 0.2 mm Polyvinylidene

NH+-N: 92-98% removal Total nitrogen: 30-65% removal

Higher MLSS

concentrations (in excess of 12,000 mg/L) led to improved nitrogen removal by enhancing anoxic microenvironments. PAC addition did not affect nitrogen removal

Liu et al. [56]: Investigated the effect of PAC addition on the performance of MBRs (municipal wastewater)


Attached growth MBR Volume: 21 L


Submerged hollow fibers


Pore size: 0.05 mm

COD: <25 mg/L BOD: <1.5 mg/L Ammonia-N: < 0.8 mg/L Total nitrogen: < 15 mg/L

PAC addition decreased the effluent DOC to <15 mg/L and reduced the extent of fouling

Lozier and Fernandez [57]: Characterized performance of MBR as pretreatment for RO (municipal wastewater)


Submerged hollow fibers Pore size: 0.04 mm

Total suspended solids: 0.3-0.4 mg/L Turbidity: 0.16-0.27 NTU COD: 13.8-15 mg/L BOD: 0.57-1.05 mg/L Total phosphorous: 0.18-3.38 mg/L Ammonia-N: 0.11-5.77 mg/L Total nitrogen: 16-27 mg/L

MBR effluent exceeded RO feed water quality criteria

Monti et al. [15]: Compared performance of MBR and conventional systems configured for enhanced nutrient removal (municipal wastewater)


Volume: 2.2 m3

SRT: 10 days

MLSS: 6000-8000 mg/L



Submerged hollow fiber Pore size: 0.04 mm

COD: 90% removal (average) Sludge yield: 0.23-0.28 g VSS/g COD Ammonia-N: nondetectable Total phosphorous: 1.85+ 0.065 mg/L

Sludge yield of MBR approximately 15% lower than for conventional system. Greater denitrification was observed for a conventional system, possibly due to anoxic conditions in the secondary clarifier. Substantially lower effluent phosphorus concentrations were achieved in MBR

Oota et al. [28]: Determined the ability of MBRs to remove viruses and select endocrine disrupting compounds (municipal wastewater)


Anoxic—aerobic system HRT: 6 h


Submerged flat plate Pore size: 0.4 mm

BOD: <5 mg/L TOC: <3.8 mg/L E.coli: 100% removed Viruses (coliphage): 5 log removal Nonylphenol: 0.1 ng/L Bisphenol A: 0.02-0.03 ng/L DEPH: <0.2 ng/L Benzophenone: 0.01 ng/L 17b-Estradiol: nondetectable

Hypothesized that viruses are removed by adsorption onto biomass and retention by foulant layer that forms on membrane surface

Compared performance of MBR and conventional systems at low SRT and HRT (municipal wastewater)

Bioreactor Volume: 225 L HRT: 5-15 h SRT: 2-7 days

COD: >90% removal Total nitrogen: >65% removal (when SRT >3 days)

Consistently better performance of MBR. Overall nitrogen removal declined substantially when the SRT was less than 3 days

Source and objective

MBR system details

MBR effluent quality


Spring et al. [36]: Compared ability of MBRs and conventional activated sludge systems at removing EDCs and other trace contaminants (municipal wastewater)


SRT: > 40 days Membrane

Submerged hollow fibers Pore size: 0.04 mm

Cholesterol: 96% removal (average) Coprostanol: 100% removal

(nondetectable) Stigmastanol: 100% removal

(nondetectable) Bisphenol A: <12.6ng/L Estrone: <1.2 ng/L 17ß-Estradiol: <1.1 ng/L 17a-Estradiol: <1.6 ng/L

MBRs could remove approximately 10% more cholesterol, coprostanol and stigmastanol from municipal wastewaters than conventional treatment

Tam et al. [40]: Compared performance of MBRs and conventional activated sludge+MF as pre-treatment for RO (municipal wastewater)

Bioreactor Anoxic-aerobic configuration Membrane Pore size: 0.4 mm Submerged hollow fiber

BOD: <2 mg/L COD: 17.5 mg/L (average) Total suspended solids: < 2 mg/L Ammonia-N: 0.4 mg/L (average) NO^-N: 1.9 mg/L (average) E. coli: 7 log removal (average) Viruses: 4.7 log removal (average) THMFP: 9-13.5 mg/L Total estrogens: 80% removal

Removal of viruses decreased significantly following membrane cleaning, and increased over time thereafter

Tao et al. [31]: Investigated Bioreactor Turbidity: <0.2 NTU

effect of anoxic-aerobic Various anoxic-aerobic TOC: <5 mg/L

zone configurations on configurations Ammonia-N: < 1 mg/L

performance of MBR MLSS: 4000-13,000 mg/L Total nitrogen: <12 mg/L

(municipal wastewater) HRT: 4.5-12h

SRT: 14-28 days

Membrane pore size did not affect permeate quality

oo at


Submerged hollow fiber and flat sheet Pore size: 0.035-0.4 mm

Wintgens et al. [45]: Investigated ability of full-scale MBRs to remove EDCs (landfill leachate)

Bioreactor Volume: 180 m3 Anoxic-aerobic configuration MLSS: 25,000 mg/L Membrane External tubular Cross-flow: 5 m/s

Bisphenol A: over 99% removal

Removal attributed to adsorption onto biomass and subsequent biodegradation

Wong et al. [29]: Determined contribution of membrane, biomass and foulant layer to virus removal (synthetic wastewater)


Volume: 19 L

MLSS: 6000 mg/L


Pore size: 0.4 mm

Submerged hollow fiber

Removal increased over time as membrane became fouled. Approximately 0.5 and 0.6 log removal was attributed to retention by the membrane and removal by the biomass, respectively. Approximately 1.3-1.8 log removal attributed to foulant layer

Yoon et al. [33]: Compared performance of MBRs configured for enhanced nutrient removal and


Anoxic-anaerobic-aerobic Volume: 15.9 m3 SRT: 20-60 days

BOD: 0.3-2.1 mg/L Ammonia-N: 91.9% removal (average) Total nitrogen: 3.8-17.1 mg/L Total phosphorous: 0.06-1.4 mg/L

Relatively similar BOD, TN and TP removal efficiencies with MBR and conventional

Source and objective

MBR system details

MBR effluent quality


conventional systems


activated sludge system

with chemical


followed by coagulation/

phosphorous removal

Submerged air sparged

flocculation and gravity

(municipal wastewater)

hollow fiber


Pore size: 0.4 mm

Zhang and Farahbakhsh [3]: Compared microbial removal efficiency in MBRs can conventional activated sludge systems (municipal wastewaters)


MLSS: 10,000 mg/L HRT: 6 h SRT: 18 days Membrane

Submerged hollow fibers Pore size: 0.04 mm

Coliforms: 5.7 log removal Coliphage: 3.1-5.5 log removal

MBRs can achieve better microbial removal in fewer steps than conventional activated sludge. Effluent from MBRs and conventional activated sludge with tertiary filtration were relatively similar in quality

Was this article helpful?

0 0
100 Golf Tips

100 Golf Tips

If you want to become a golf player, it is a good idea to watch professional golf players playing the sport. When you watch them, you would become more inspired in getting better with your game. Aside from that, you could also take note how they carry themselves on the field, as well as how they make their swings.

Get My Free Ebook


  • tanja
    Can we use concrete baffle in the anoxic zone for the bioreactor?
    4 months ago

Post a comment