Putting the Misconceptions to Rest:

Putting the Misconceptions to Rest: 2010 HWEA Conference Honolulu, HI Brandy Nussbaum I. Kruger, Inc

Clarification/Separation OPTIONS Following MBBR Treatment

Discussion Topics What is MBBR Conventional Clarification High Rate Ballasted Clarification Dissolved Air Flotation Discfiltration Granular Media Filtration Membranes Summary

The Principle of the Moving Bed Biofilm Reactor (MBBR) Technology Aerobic reactor Anoxic reactor

The Moving Bed Biofilm Reactor (MBBR) in Practice

Components to the MBBR Aerobic System AnoxKaldnes Media Media Retention Sieves 304L Stainless Steel Cylindrical Wedge Wire Perforated Plate foam control sieves AnoxKaldnes Aeration Grid 304L Stainless Steel Medium Bubble Maintenance Free No replacement parts

Components to the Anoxic MBBR System AnoxKaldnes Media Media Retention Sieves 304L Stainless Steel Flat Panel Wedge Wire Slow-Speed Mixers Submersible or Top entry Can be speed controlled

MBBR Solutions Highly Flexible MBBR stand-alone MBBR as pre-treatment (roughing, BAS TM ) MBBR as tertiary treatment (polishing, LagoonGuard TM ) MBBR in activated sludge (HYBAS TM /IFAS)

MBBR Biomass Separation Alternatives MBBR - settling MBBR Actiflo (ballasted flocculation) MBBR - Flotation MBBR Media Filtration MBBR Disc Filter MBBR Membrane Filtration

Development of PSD s at various MBBR loadings (HRT s) (Åhl et al, 2006) diff volume% diff number% 8 6 4 2 0 inlet HRT = 1h HRT = 2h HRT = 3h HRT = 4h 0,01 0,1 1 10 100 1000 particle diameter µm 8 6 4 2 0 inlet HRT = 1h HRT = 2h HRT = 3h HRT = 4h 0,01 0,1 1 10 100 1000 particle dimeter [µm] 1. There is a shift towards a higher volume of larger particles with increasing HRT 2. There is at the same time, however, a relative increase in the number concentration of submicron particles with increasing HRT Hypothesis : Flocculation is taking place Erosion of single bacteria or remains of dead cells happens Both are f(loading/hrt)

Relative amounts of COD in the different size fractions (Melin at al, 2005) Amount of COD (%) 80 70 60 50 40 30 20 10 0 >1 µm 0.1-1 µm 30 kd-0.1 µm <30 kd Size fraction a) Majority of COD in particles > 1 µm Increase in the suspended COD with increasing HRT Decrease in colloidal particles (0,1 1 µm) when HRT increase HRT = 0.75 h HRT = 1 h HRT = 3 h HRT = 4 h

Conventional Settling after MBBR Without coagulation Coagulant With coagulation Metal salt (Al, Fe) Cationic polymer Low Al/Fe + cat. polymer High Al/Fe + anionic polymer

Cheyenne Crow Creek WWTP

THE CROW CREEK WATER RECLAMATION FACILITY New Treatment Process Primary Clarifier 1 Screening & Degritting Primary Clarifier 2 Flow Split Anoxic Basins (Pre-Denite) New MBBR Reactors Future Nitrate Recycle Discharge Reuse Filters Alum/Polymer Sludge UV Disinfection Secondary Clarifier 2 Secondary Clarifier 1

PERFORMANCE DATA Cheyenne BOPU Crow Creek WRF TSS Removal - 30 Day Avgerage Influent 500.0 450.0 400.0 350.0 300.0 250.0 200.0 150.0 100.0 50.0 0.0 25.0 20.0 15.0 10.0 5.0 0.0 Effluent Influent Effluent

Actiflo Microsand Ballasted Clarification Sludge Hydrocyclon M Microsand M M M Outlet Coagulant Inlet Injection Flocculation Coagulation Lamella- - Polymer sedimentation Coagulant Microsand Polymer Water Primary particles Flocs

South Adams WWTP MBBR-ACTIFLO Pilot Influent characteristics Coagulants and dose Effluent characteristics Process Rise rate 3 mg SS/l Turb FeCl 3 or Al s (SO 4 ) 3 + anionic polymer 1 mg SS/l Turb. mg P/l MBBR + 55 85 m/h 12.4 mg Fe/l + 0,8 mg/l 3 < 2 ACTIFLO most tests at 110-150 130-160 65 m/h 4 7.6 mg Al/l + 0,6 mg/l 6.2 mg Fe/l + 0,6 mg/l < 5 5 < 2 0.12 2 2.1 mg Fe/l + 0,45 mg/l 10 MBBR + Settling + 120 m/h 13.2 mg Fe/l + 0,5 mg/l 1.4 <0,1 ACTIFLO 100 m/h 3,5 4,5 10-25 6.0 mg Al/l + 0,6 mg/l < 2 >0.3 MBBR + Settling + 90 m/h 13.2 mg Fe/l + 0,5 mg/l < 1 <0.1 ACTIFLO Turbo 115 m/h 3,5 4,5 10-25 9.6 mg Fe/l + 0,5 mg/l 1.2 <0.1 115 m/h 9.1 mg Al/l + 0,6 mg/l 1.4 0.1 1 Anionic polymer M155 2 Lowest value achieved - at high Al dose (14.4 mg Al/l) 3 Rise rate, hydraulic surface load (m 3 /m 2. h) calculated on the settler foot-print area

Summary of Pilot Results, Quebec, Can (John Meunier, VWS, Can) Actiflo following MBBR MBBR load : Actiflo rise rate: FeCl 3 -dose : Polymer dose : 4 25 g BOD 5 /m 2 d 40 120 m/h 30 115 µl/l 0.5-1.2 mg/l Parameter units Influent Effluent % removal Total P (mg/l) 1-5 0.2-0.7 Turbidity NTU 58-154 2.5-6.8 TSS (mg/l) 94-196 13-24 86-88 COD (mg/l) 176-361 31-58 82-84 BOD (mg/l) 53-155 5-11 91-93 ph --- 7.1-7.5

Example Drammen WWTP Inlet 3 1 2 Line 1 Line 2 Outlet Line 3 Upgrading a chemically enhanced primary plant to become a secondary plant by the use of MBBR and ACTIFLO 4 1: Flocculation 2: Sedimentation 3: Thickeners 4: Sludge silo Line 4 Line 5 Line 6 Inlet Kaldnes MBBR 1-1 Kaldnes MBBR 1-2 DAF 1 Outlet Inlet Kaldnes MBBR 1-1 Kaldnes MBBR 1-2 Outlet Kaldnes MBBR 2-1 Kaldnes MBBR 2-2 DAF 2 Kaldnes MBBR 2-1 Kaldnes MBBR 2-2 3 1 Kaldnes MBBR 3-1 Kaldnes MBBR 3-2 DAF 3 2 3 1 Kaldnes MBBR 3-1 Kaldnes MBBR 3-2 Kaldnes MBBR 4-1 Kaldnes MBBR 4-2 DAF 4 Kaldnes MBBR 4-1 Kaldnes MBBR 4-2 4 Kaldnes MBBR 5-1 Kaldnes MBBR 6-1 Kaldnes MBBR 5-2 Kaldnes MBBR 6-2 DAF 5 DAF 6 4 5 2 Actiflo 1 Actiflo 2 1: Kaldnes MBBR 2: DAF 3: Sludge silo 3% TS 4: Sludge silo 6% TS 1: Kaldnes MBBR 2: Actiflo 3: Thickeners 4: Sludge silos 6% 5: Free space (chemical handling)

Two MBBR ACTIFLO in Bergen Under Construction Bergen WWTP design: r BOD = 11,5 g BOD 5 /m 2 d v f = 60 m/h (rise rate in settling zone)

Handeland, Skreia MBBR/Actiflo Plants BOD5 results Handeland WWTP 2007-2008 350 300 Influent concentration Effluent concentration Concentration BOD5, mg/l 250 200 150 100 50 a. Handeland WWTP 0 BOD 5 results Skreia WWTP 2005-2008 800 40 700 BOD5:Inlet BOD5:Outlet 35 Inlet concentration, g O/m 3 600 500 400 300 200 30 25 20 15 10 Outlet concentration,g O/m3 100 5 0 0 b. Skreia WWTP Date

MBBR-Flotation (DAF) Historically a very popular option for compact MBBR separation in Scandinavia Data below from the Johnstown, CO LagoonGaurd MBBR- DAF Installation Parameter DAF influent Range Average DAF Effluent Range Average Turbidity, NTU 18-80 40 2-28 15 BOD mg/l 24-35 27 3-13 10 TSS mg/l 19-62 35 8-25 11

Typical design values for DAF following MBBR (Ødegaard et al, 2009) Tank depth Surface overflow rate (m 2 /m 3. h) At design flow, At maximum design flow, 2 3 m Q dim 5 Q maxdim 10 Dispersion pressure : 400 600 kpa (4-6 bar) Air saturation: 80-90 % Dispersion water flow (% of Q maxdim ) : 10-25 % (depending on SS in and air saturation)

Average Results from three Scandinavian Plants using DAF after MBBR Parameter Nordre Follo WWTP Gardermoen WWTP Sjölunda WWTP Design values Design flow (m 3 /h) 750 920 7920 Max. flow (m 3 /h) 1125 1300 15840 Temp. ( o C) 6-14 4-14 8-20 Plant size Tot. MBBR vol. (m 3 ) 3710 5790 6230 Flocculation vol. (m 3 ) 230 180 3960 Flotation area (m 2 ) 150 215 2000 Year documented 2008 2009 2009 Average in-out conc. and treatment efficiency In Out % In Out % In Out % SS (mg/l) - - - - - - 282 12 96 BOD (mg/l) 123 3.8 97 274 2.2 99 231 10 96 COD (mg/l) 453 49 89 725 33 95 559 62 89 Tot N (mg/l) 32 6.5 80 62 8.7 85 41 10 76 Tot P (mg/l) 4.2 0.14 97 8.8 0.11 99 5.3 0.30 94

MBBR - Microscreening The Hydrotech Disc Filter Directly after MBBR or for polishing Sieve openings: 10 100 µm Operational head-loss: 10 Backwash during operation Disc Filter plant after post DN MBBR at Rya WWTP, Gothenburg, Sweden

Results from the Rya WWTP Discfilter (Mattson et al, 2009) Mesh pore size, µm Feed, mg SS/l Effluent, mg SS/l Capacity, m 3 /m 2 filter. h Average St.dev. Average St.dev. Average St.dev. Number of Samples 18 27.5 14.5 5.0 1.8 13.7 7.2 37 10 30.5 10.8 3.5 1.3 4.8 2.2 22 30 25 10 micron 18 micron Filtration velocity (m/h) 20 15 10 5 0 0 10 20 30 40 50 Influent SS (mg/l) Filtration rate (based on total filter mesh area) versus influent SS (Persson et al. 2006)

Discfilter With Chemical Dosing Effluent SS (mg/l) 200 180 160 140 120 100 80 60 Gardermoen WWTP 40 micron Nordre Follo WWTP 40 micron Gardermoen WWTP 20 micron Effluent SS (mg/l) 50 45 40 35 30 25 20 15 Gardermoen WWTP Nordre Follo WWTP 40 10 20 5 0 0 50 100 150 200 250 300 350 Influent SS (mg/l) 0 0 10 20 30 40 50 60 70 80 90 100 Polymer dose (mg PE/g SS) Effluent SS versus influent SS without any pre-coagulation/flocculation Effluent SS as a function of polymer dosing.

Sand filtration after MBBR Various uses: As polishing step Directly after nitrifying or denitrifying MBBR Ex.: Klagshamn WWTP 8,2 m/h at Q max Effluent : < 5 mg SS/l, 0,2 mg P/l Directly after high-rate (secondary treatment) MBBR- Cheyenne Crow Creek

Membrane filtration after MBBR Different strategies tested a. MBBR submerged hollow fiber UF membrane (i.e. Zenon ZeeWeed) b. MBBR Discfilter Contained hollow fiber UF membrane c. MBBR submerged membrane in reactor with settling zone d. MBBR DAF Contained hollow fiber UF membrane

Conclusions 1. Particle size distribution (PSD) of biomass leaving MBBR: a. Large mass fraction of relatively large particles (30 300 µm), but a high number of small particles (0.1 1 µm). The easiest way to deal with the latter is by coagulation ahead of the separation reactor b. The PSD is shifted towards larger particles when the organic area load is decreasing (HRT is increasing). Hence the higher the area load, the better is the effect of pre-coagulation 2. All of the commonly used separation methods may be used. a. MBBR allows the use of a variety of separation methods providing greater flexibility that the AS processes miss b. DAF is well proven c. Microsand ballasted lamella settling (Actiflo) is of increasing interest and results in an extremely compact plant d. A very compact solution is also achieved with the use of Discfilter di tl ft th MBBR li hi t ti t

Final Thoughts