0071776257_ar004-RBC McGrawHill

173

Chapter 4

Rotating Biological Contactors

1.0 IntRoduCtIon 174

2.0 PRoCess desIgn ConsIdeRatIons 178

2.1 Media surface area 179

2.2 pH and nutrient Balance 180

2.3 oxygen transfer 180

2.4 Flow and Loading Variability 182

2.5 operating temperature 183

2.6 solids Production 183

2.7 toxic and Inhibitory substances 184

3.0 RotatIng BIoLogICaL ContaCtoR desIgn MetHods 184

3.1 Monod Kinetic Model 184

3.2 second-order Model 186

3.3 empirical Model 187

3.4 Manufacturers’ design Curves 189

3.5 Comparison of Model Predictions 190

3.6 Predicted Performance versus Full-scale data 191

3.7 temperature Correction 192

4.0 RotatIng BIoLogICaL ContaCtoR nItRIFICatIon ModeLs 194

5.0 denItRIFICatIon aPPLICatIon 197

6.0 PHysICaL desIgn FeatuRes 1986.1 Physical Layout 1986.2 tank Volume 1986.3 Hydraulics and Flow

Control 1986.4 Media 1996.5 drive systems 2006.6 Covers 2016.7 Biomass Control 201

7.0 RotatIng BIoLogICaL ContaCtoR desIgn exaMPLes 2027.1 secondary treatment

design example 202

(continued)

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)Copyright © 2011 The McGraw-Hill Companies. All rights reserved.

Any use is subject to the Terms of Use as given at the website.

Source: BIOFILM REACTORS

174 Biofilm Reactors

7.2 advanced secondary treatment design example 203

8.0 PRoBLeMs and CoRReCtIVe aCtIons 204

8.1 Inadequate treatment Capacity 205

8.2 excessive First-stage Loadings 205

8.3 excessive Biomass growth 206

8.4 Loping of air-drive systems 206

8.5 High Clarifier effluent suspended solids 207

8.6 Corrosion of Media supports 207

9.0 PILot-PLant studIes 207

10.0 ReFeRenCes 208

1.0 IntRoductIonthe rotating biological contactor (RBC) process is a biological process that has been used widely for secondary and advanced secondary wastewater treatment. the pro-cess has been used for five-day biochemical oxygen demand (Bod5) removal, nitrifica-tion, or combined Bod5 removal and nitrification. as a secondary treatment process, it has been applied widely, where the average effluent limitations of 30 mg/L Bod5 and total suspended solids (tss) are required, by permit. as an advanced second-ary treatment process, it has been used often in conjunction with effluent filtration, where there are limits of ≤10 mg/L Bod5 and tss and for effluent ammonia-nitrogen (nH3-n) levels as low as 1 mg/L. the RBC process has been used for aerobic pre-treatment of industrial process wastewater, and it also has been used in an anoxic mode for denitrification.

the RBC process is a form of fixed-film or attached-growth biological treatment. a circular disk of polystyrene or polyvinyl chloride is mounted on a horizontal shaft and partially submerged (typically 40%) in a tank holding the wastewater undergo-ing treatment. Plastic packing material, in a cylindrical basket or cage mounted on a horizontal shaft, also has been used. the media is slowly (1 to 1.6 rpm) rotated in the direction of the influent flow, to expose the biological film to food and nutrients and to provide necessary oxygen. excess biological cell mass is “sloughed” from the media and removed from the wastewater forward-flow by clarification. Figure 4.1 provides a graphic representation of the process.




Rotating Biological contactors 175

a modification of the RBC process is the submerged biological contactor (sBC), in which the RBC is submerged up to 70 to 90%. advantages may include lower shaft loadings, ability to retrofit the process to existing activated sludge aeration tanks, use of a dual air-header scouring system (which reduces biomass thickness), and ability to use larger-diameter media bundles. the design approach recommended is simi-lar to that recommended for air-driven RBC systems. the sBCs also have been con-sidered for application in denitrification processes where the shaft is mechanically driven and the introduction of air into the bulk liquid is minimized. In general, how-ever, the sBC has seen limited application.

Rotating biological contactor units commonly are manufactured in standard units, with a media bundle of approximately 3.5 m (12 ft) in diameter and 7.5 m (25 ft) in length. standard-density units have a media surface area of approximately 9300 m2 (100 000 sq ft) per shaft, whereas high-density units typically have a media surface area of approximately 13 900 m2 (150 000 sq ft) per shaft. standard-density units are used for Bod5 removal, where biological growth is relatively thicker, and more open media are necessary to maintain open passageways for wastewater flow throughout the media. High-density units most commonly are used in nitrification applications, where biofilm growths are relatively thin. some designs will use lower surface area in the initial stages, followed by higher surface-area media. a more recent development

Cover

Rotating biologicalcontactor

Interstagebaffle

EffluentSludgeFood

Nutrients

Influent

Oxygen

Degradation products

Figure 4.1 general representation of RBC process.





is the use of plastic packing material instead of circular discs. the packing material, contained in a cylindrical cage mounted on the shaft, provides a very high specific surface area for biofilm growth.

Rotating biological contactor units are covered to prevent algae growth, excessive heat loss in cold weather, and uV deterioration of polyethylene media. Prefabricated fiberglass, Quonset-hut types of covers commonly have been used, although conven-tional building structures also are used for covering the units.

the RBC process typically is configured with several stages arranged in series, with one or more shafts composing each stage and with one or more parallel trains of shaft stages to provide the needed media surface area. each train typically consists of several shafts installed in a common tank, with baffles installed between shafts to separate the shafts into distinct stages. the number of stages required depends on the degree of treatment desired, with one or two stages provided for roughing applica-tions and six or more stages provided where advanced treatment with nitrification is necessary. shafts typically are installed perpendicular to the direction of wastewater flow, with baffles for staging. For plants designed to treat relatively small flows, the shafts may be oriented in the direction of wastewater flow, with baffles provided to allow more than one stage per shaft. Figure 4.2 shows the general arrangements typi-cally used for the RBC process.

shaft rotation most typically has been provided by mechanical-drive systems. air-drive systems using diffused air and an array of cups, fixed to the periphery of the media, which capture some of the diffused air resulting in rotation from the buoy-ancy effect, also have been used. the mechanical drive system may be considered a constant-speed/variable-torque system, where shaft revolutions per minute are relatively constant, and torque requirements of the drive system vary with biomass growth and other conditions. the air-drive system may be considered a constant-torque/variable-speed system, where, at a given air rate, the torque applied to the shaft is relatively constant, and the shaft revolutions per minute vary with biomass growth and other conditions.

the RBC process has a number of potential advantages, including simplicity and the need for minimal operator attention, low energy costs, low overall costs, and rapid recovery from shock loadings. However, there have been numerous examples of RBC process failure that have resulted from the following:

structural failure of the shaft, media, or media support systems,•

Less-than-anticipated treatment performance,•





excessive development of nuisance organisms,•

development of excessive or uneven biomass growth,•

Inadequate performance of air-drive systems for shaft rotation, and•

Misapplication of pilot plant data.•

Primarily, these failures have been attributed to the following:

Misapplication of the process or use of inadequate design criteria;•

original designs for shafts and media support systems that were inadequate;•

Inadequate upstream treatment;•

Failure to consider the effect of sidestream loadings in the process design;•

Lack of a thorough understanding of the long-term performance efficiency •and characteristics of the process in full-scale use;

Typical paralleltrain configuration

End flow for small plantsTypicaladvancedtreatment

Typicalsecondarytreatment

Figure 4.2 Rotating biological contactor process general flow arrangements.





development of design standards and criteria on the basis of limited data, •much of which was developed using small-scale systems; and

general lack of a conservative approach in developing design relationships for •the process.

Many of the problems that were associated with earlier installations of the RBC process are now better understood, and the process has been used successfully at numerous locations. However, some regulatory agencies have established poli-cies that restrict or prevent the use of the RBC process in their areas of authority. therefore, the designer should investigate the applicability of the process in the area of the project before considering the RBC process.

In addition to the anticipated benefits or advantages of the RBC process, it is essential that the designer consider the limitations of the process, long-term perfor-mance of the process that can be expected in full-scale applications, and other prac-tical knowledge that is based on experience with the RBC process during the last 2 decades. as with any process, proper sizing and process design, a conservative and flexible design approach, and careful consideration to the specification of materials and components are necessary, if the process is to be applied successfully.

2.0 PRocess desIgn consIdeRatIonssoluble and particulate carbonaceous Bod5 (cBod5) components of wastewater are reduced in the RBC process, by a combination of oxidation by the biofilm and syn-thesis into new cell mass. soluble waste components, nutrients, and oxygen are trans-ferred to the biofilm from the bulk liquid, and degradation products are transferred from the biofilm to the bulk liquid. Particulate waste components are enmeshed by the biofilm and hydrolyzed before degradation can occur. For the overall process to work successfully, the proper environmental and process conditions must be pre-sent. Principal RBC process design considerations include the following:

Media surface area,•

pH and nutrient balance,•

oxygen transfer,•

Flow and loading variability,•

operating temperature,•

excess biomass production, and•

toxic and inhibitory substances.•





2.1 Media surface areaadequate media surface area is essential to the process, to provide sufficient biofilm to affect treatment and operate at specific substrate use rates (e.g., the mass of Bod5/unit and time/unit media surface area). the substrate removal rate must be matched by the oxygen-transfer capability of the system. sufficient media and proper opera-tional controls are necessary to reduce the potential for nuisance organism develop-ment and limit excessive biofilm buildup, which could lead to structural damage to the shaft or media.

selecting the media surface area to be provided to meet design objectives is a balance between a conservative assessment of the capability of the process and eco-nomics. Because higher specific substrate use rates occur at higher bulk wastewater substrate concentrations, staging the process can provide the same degree of treat-ment that could be provided by a single-stage system having a larger surface area. a single-stage system would require that the bulk Bod5 concentration be limited to a relatively low value. this would reduce Bod5 removal rates, because, in the lower range of Bod5 values, specific substrate-use rates are related to concentration (first-order kinetics). therefore, a more economical design can be provided by stag-ing, where the earlier stages are loaded more heavily organically and have relatively high bulk fluid Bod5 concentrations and higher substrate-removal rates. However, care should be taken so that loadings to the first stages are not high enough to cause inadequate oxygen transfer, nuisance organism development, overweight shafts, and associated process problems.

early graphical methods (prepared by RBC suppliers) for determining media surface area requirements for the RBC process typically have yielded surface areas less than those required to meet design requirements for Bod5 removal based on actual long-term plant operating data. However, graphical methods evolved from a minimal database and pilot units, which later proved to be optimistic in the sizing of RBCs. Recently developed models, which are more reliable for defining surface-area requirements for Bod5 removal and nitrification, are presented and discussed later in this chapter. the reader also is referred to Chapter 3 for a discussion of modeling techniques for trickling filters, which also are applicable to the design of the RBC pro-cess. In addition, Chapter 11 provides a general understanding of modeling biofilm systems.

a lack of adequate consideration of the effect of loadings from recycle streams, such as solids-processing unit operations or filter backwash, have been reported to be factors in unsatisfactory process performance at several plants. However, this is not





a concern that is particular to RBC plants, and it is not caused by the RBC process. therefore, a designer should develop a complete mass balance of the process, so that the effect of recycle sidestreams on influent flow and organic loadings to the RBC process are considered, and their effect on the determination of required media sur-face area is incorporated to the design.

2.2 pH and nutrient Balanceas with any biological treatment process, a proper pH and nutrient balance must be provided for effective treatment by the RBC process. the optimum pH for bacterial growth has been reported to be 6.5 to 7.5 (Metcalf and eddy, 1979). Where nitrifica-tion is required, optimum performance has been observed at a slightly alkaline pH, with the optimum pH being 7.5 to 8.5 for suspended-growth systems (sawyer et al., 1973). For RBC systems, nitrification rates have been reported to fall off rapidly, with the pH decreasing from 7.0 to 6.0 (Brenner et al., 1984). Well-acclimated RBC systems, where the pH is maintained in a narrow range, may be able to tolerate pH values somewhat below 7.0. However, with the minimal buffering at this pH, a stable pH is difficult to maintain.

For systems required to nitrify, the alkalinity of wastewater is important because of the consumption of alkalinity. the autotrophic organisms responsible for nitri-fication consume inorganic carbon (alkalinity). the conversion of the base ammo-nia (nH3) to nitrate (no3

-) releases hydrogen ions (H+), thereby producing an acid (ammonium oxide [nHo3]). approximately 7 mg of calcium carbonate (CaCo3) alka-linity are required for each milligram of ammonia-nitrogen oxidized. It has been rec-ommended that a residual alkalinity of 50 to 100 mg/L be maintained in the effluent of a nitrifying plant (envirex, Inc., 1989). this alkalinity level typically produces a pH of 7.0 or higher. Chemical addition may be necessary to wastewater where natural alkalinity is insufficient.

as microorganisms providing treatment require nutrients for cell growth, an ade-quate supply of available nutrients is necessary for proper performance of the pro-cess. a minimum mass ratio of 100:5:1 (Bod: nitrogen:phosphorus) typically is used.

2.3 oxygen transferthe rate of oxygen transfer must be sufficient to maintain fully aerobic conditions in the RBC process. organic loadings above the oxygen-transfer capability of the sys-tem result in reduced performance, odors, and development of nuisance organisms.





Maximum oxygen-transfer rates of 6.8 to 7.3 g oxygen/m2·d (1.4 to 1.5 lb/d/1000 sq ft) have been reported (scheible and novak, 1980), based on tests conducted with full-scale, mechanical-drive systems. studies of supplemental aeration applied to mechanical-drive RBCs indicated an oxygen-transfer efficiency of 2 to 2.5% for applied air (Chou, 1978). If it is assumed that the effective oxygen-transfer rate for an air-drive RBC is 2.5%, a 9300-m2 (100 000-sq ft) air-drive unit supplied with 7 m3/min (250 scfm) would have an oxygen-transfer rate of approximately 8.3 g oxygen/m2·d (1.7 lb/d/1000 sq ft). Maximum oxygen-transfer rates from 6.8 to 8.3 g oxygen/m2·d (1.4 to 1.7 lb/d/1000 sq ft) are consistent with the maximum reported ammonia-nitrogen removal rate of approximately l.5 g/m2·d (0.3 lb/d/1000 sq ft), based on 4.6 kg oxygen/kg nH3-n oxidized.

Historically, pilot testing of the RBC process was performed using small-scale units operating at the same media peripheral velocity (approximately 18 m/min [60 ft/min]) as full-scale units. data on process performance collected on small- diameter units can be misleading because of the higher oxygen-transfer capability of the smaller units compared with full-scale units. Reh et al. (1977) estimated that a 0.5-m-diameter (1.5-ft-diameter) test unit provided an oxygen-transfer rate 1.6 times that of a 3.2-m-diam (10.5-ft-diam) unit that also was tested. similar scale-up problems also were noted by other researchers (Brenner et al., 1984). For these reasons, data col-lected on small-scale systems should be used only for treatability considerations and not to establish sizing criteria.

Manufacturers of RBC equipment historically have recommended maximum organic loadings to the first stage of RBC systems of approximately 20 g soluble Bod5 [sBod5]/m2·d (4 lb/d/1000 sq ft) of media. this equates to an organic loading of 39 g Bod5/m2·d (8 lb/d/1000 sq ft) on a total Bod5-basis for a typical domestic wastewater with 50% soluble Bod5/Bod5 in the primary effluent. However, oxy-gen-transfer capabilities were reported to have been exceeded at first-stage loadings of more than 15 g sBod5/m2·d (3 lb/d/1000 sq ft) (McCann and sullivan, 1980). a survey of 23 plants indicated nuisance organism development because of oxygen-limiting conditions for first-stage loadings in excess of approximately 31.2 g Bod5/m2·d (6.4 lb/d/1000 sq ft) of media (Weston, Inc., 1985). one RBC manufacturer currently recommends maximum first-stage loadings of approximately 12 to 15 g sBod5/m2·d (2.5 to 3.0 lb/d/1000 sq ft) for unaerated RBC5 (envirex, Inc., 1989). this rate equates to 24 to 29 g total Bod5/m2·d (5 to 6 lb/d/1000 sq ft), assuming 50% sBod5/Bod5 (15 g sBod5/m2·d [3 lb/d/1000 sq ft]). the manufacturer (envirex, Inc., 1989) indicates that loadings below these levels provide a safety margin against





unexpected loads and prevent sulfide formation. another manufacturer (Lyco, Inc., 1992) recommends maximum first-stage loadings of 12 to 20 g sBod5/m2·d (2.5 to 4.0 lb/d/1000 sq ft).

the designer should consider limiting organic loading to approximately 29 g Bod5/m2·d (6 lb/d/1000 sq ft) (total Bod5-basis) to any RBC stage. this criterion will limit loadings to levels within the oxygen-transfer capability of the system and reduce the potential for growth of nuisance organisms.

2.4 Flow and Loading Variabilitytypically, higher effluent concentrations during the higher loading periods of the day are counterbalanced by lower effluent values during low-loading periods. However, a high degree of variability in influent flow or organic loading may disrupt the RBC process. Fixed-film processes, in general, are more susceptible to breakthrough of substrate, as a result of high peak loadings, than are suspended-growth processes. Loadings can increase to the point where breakthrough from the RBC process occurs. at this point, the substrate loading per unit surface area exceeds the substrate removal rate per unit surface area. this potentially could occur as a result of diurnal variations, in which case, it may be more cost-effective to provide flow equalization tanks for load equalization rather than provide additional RBC units. another poten-tial solution is to store peak day loads for feeding into the process at night, when the influent loads are minimized.

It has been recommended (envirex, Inc., 1989) that flow equalization is incorpo-rated to the design if peak-to-average flow ratios exceed 2.5. the designer also should be aware of variability in organic loadings, including the effect of recycle loads from solids-processing unit operations. Particular attention should be paid to the sBod concentration in the recycle from dewatering operations. Providing flow equalization may compensate for both flow and loading variability and eliminate the requirement for additional RBC tankage and media. this is especially true if stringent effluent limitations must be met. In the absence of equalization, a high variation in hydraulic or organic loading to the process would need to be addressed by providing addi-tional RBC media and tankage.

In addition to daily variations, the designer should consider peak daily loadings, which may result from industrial operations or other variable conditions. Because effluent limitations must be achieved during peak-loading periods and not just under average conditions, appropriate safety or peaking factors should be incorporated to the design. as with any process, the degree of safety factor incorporated to the design





also is influenced by the details of the discharge permit (instantaneous limit, monthly average, annual waste load allocation, etc.).

2.5 operating temperatureRates for the removal of Bod5 and nitrification are reduced by cold wastewater tem-peratures in RBC systems. It is important that the designer carefully review long-term data on wastewater temperatures and select a design temperature representative of critical cold-weather operating periods. For situations in which effluent limitations vary seasonally, each seasonal condition should be checked to determine which period is critical for design purposes.

Biochemical oxygen demand removal and nitrification rates typically are reported by RBC equipment manufacturers as being relatively constant above 13°C (55°F) and declining significantly below this temperature (envirex, Inc., 1989; Lyco, Inc., 1992; Walker Process, Inc., 1992). Relationships developed to adjust media-sizing criteria for reduced temperature operation are presented later in this chapter.

2.6 solids Productionexcess solids production must be estimated to size the solids-handling facilities of the plant. solids production in the RBC process is a function of the synthesis of new cell mass, cell mass decay caused by endogenous respiration, inert solids present in the wastewater, and solids loss from the system caused by tss in the clarified effluent.

In general, solids production from the RBC process can be expected to be similar to other fixed-film processes. one manufacturer (envirex, Inc., 1989) reports that waste solids production, based on clarifier underflow, ranges from approximately 0.4 to 0.6 kg/kg total Bod5 removed by the process. Lower values represent lightly loaded systems with higher endogenous oxidation of cell mass, and higher values represent more highly loaded systems. net solids yields typically will range from 0.5 to 0.8 kg/kg Bod5 removed from the process. the designer should use a conservative approach to solids production estimates, to account for periods of high biomass release from media and cold-temperature operation, where endogenous respiration will be lower.

typical secondary sludge concentrations for the RBC process are reported to be 2.5 to 3% (envirex, Inc., 1989), with thickening in a continuous, low-volume solids withdrawal clarification system. envirex, Inc. (1989) also reports concentrations of 4 to 5% achievable by gravity thickening of the RBC secondary sludge or co-thickening with primary sludge in primary clarification. as with other aerobic biological waste-water treatment processes, solids volatility can be expected to range from 80 to 95%.





It generally is good practice to use clarifiers with conventional mechanical solids-removal mechanisms with the RBC process. spiral blade sludge collector mechanisms also should be considered. siphon-type clarifiers typically are not required and will unnecessarily dilute settled solids. thickening of RBC solids in the clarifier can lead to performance problems. Because the quantity of solids to be removed is relatively low and thicker, uptake pipes or openings are prone to plugging.

2.7 toxic and Inhibitory substancestoxic or inhibitory substances, if present in sufficient concentration in plant influent, will have a negative effect on wastewater treatment. In addition, excessive quantities of such substances may have a deleterious effect on solids management and efflu-ent quality because of pass-through. the designer should review current technical literature for information on such substances if they are suspected of being present in concentrations that are deleterious to biological systems.

3.0 RotatIng BIoLogIcaL contactoR desIgn MetHods

numerous predictive models and equations have been proposed for estimating the removal of Bod5 in RBC systems. several of these proposed relationships are described in this section. actual plant performance data also are compared with the predictions of one model and with design curves published by two RBC manufactur-ers. some models and manufacturers’ design approaches are based on sBod5, and others are based on total Bod5. For a consistent presentation, all models and design approaches are presented graphically on a total Bod5-basis, assuming that sBod5 of primary effluent makes up 50% of the total Bod5 for both RBC influent and settled effluent.

3.1 Monod Kinetic Modelas part of a study funded by the u.s. environmental Protection agency (Washington, d.C.) (u.s. ePa) (Weston, Inc., 1985), a design approach was presented based on Monod growth kinetics, as described by Clark et al. (1978). using a mass-balance approach, assuming steady-state conditions, and making other simplifying assump-tions, the following relationship was obtained:

R = (Fi/Ai)(So – Si) = PiSil(Ki + Si) (4.1)





WhereR = substrate removed per unit media surface area per unit time,FI = wastewater flowrate to stage i (gpd),Ai = area of stage i (sq ft),So = influent sBod5 to stage i (mg/L),Si = effluent sBod5 from stage i (mg/L),Pi = area capacity constant for stage i (gpd/sq ft-mg/L), andKi = half-velocity constant for stage i (mg/L).

Based on an analysis of interstage sBod5 data from 11 RBC facilities, values for the maximum removal rate, P, and the half-velocity constant, K, were determined by plot-ting the data in the linearized form of eq 4.1, as follows: [1/R = (K/PS + 1/P)]. the major-ity of the facilities were air-driven, and none of the facilities were considered organically overloaded (first-stage total Bod5 loadings were below 31 g Bod5/m2·d [6.4 lb/d/1000 sq ft]). the P and K values were determined for stages 1 through 4, as follows:

(1) stage 1—P1 = 1000 gpd/sq ft-mg/L and K1 = 161 mg/L (2) stage 2—P2 = 667 gpd/sq ft-mg/L and K2 = 139 mg/L (3) stage 3—Ps = 400 gpd/sq ft-mg/L and K3 = 82 mg/L (4) stage 4—P4 = 100 gpd/sq ft-mg/L and K4 = 25 mg/L.

using the above coefficients, eq 4.1 can be used to solve for the Si value exiting a given stage, as follows:

(stage 1) (stage 2) (stage 3)

S

HL S K P HL S K P HL K Si

o o o=− − + − − + ×{[ ( )] } {[ ( ) ] [ ( ) ( )]} .

1 1 1 2 2 22

32

304 55

12( )HL (4.2)

where HL1, HL2, HL3 = hydraulic loading rate for stage 1, etc. (m3/m2·d [gpd/sq ft]).

In using the model for design purposes, information must be available, or assumptions must be made, for the ratio of soluble to total Bod5. the work presented in the u.s. ePa study (Weston, Inc., 1985) assumed a sBod5 fraction of 0.5 for both the RBC influent and settled effluent, which is a typical value for domestic waste-water undergoing RBC treatment. the u.s. ePa study (Weston, Inc., 1985) reported that the model was typically conservative (predicted lower organic removal than observed), with the closest correlation for RBC influent values of approximately 100 mg/L Bod5 or less. the study was reported as being a “first step in collecting input data, which should be used to test the model over a variety of flows and loadings.”





Figure 4.3 presents a graph of solutions of the Clark model for various RBC influent concentrations, as a function of organic loading rate for the assumptions noted. soluble Bod5 is assumed to be 50% of total Bod5 in the development of this figure.

3.2 second-order Modela second-order kinetic model was developed (opatken, 1980) based on an analysis of interstage data from two full-scale facilities. Predictions of the model were com-pared with data from nine other full-scale plants, and there was general agreement (Brenner et al., 1984). an equation was developed based on the Levenspiel equation (Levenspiel, 1972), having the following form:

Ckt Cktn

n=− + + −1 1 4

21

0 5[ ( )] .

(4.3)

100

90

80B

OD

5 re

mov

al, %

70

60

50

400.5 1.0 2.0

5075

100125

150

Roatating biologicalcontactor influent

total BOD5

3.01.5 2.5 3.5

Organic loading, lb BOD5/d/1000 sq ft

Figure 4.3 Clark RBC predictive model, Monod kinetics, Bod5 removal: (1) values are total Bod5; (2) 50% sBod5 assumed; (3) >13°C (>55°F) temperature; (4) three stages, with 50% of media area in the first stage; and (5) standard-density media (lb/d/1000 sq ft × 4.882 = g/m2·d).





WhereCn = sBod5 effluent from stage n (mg/L); k = reaction constant (L/mg·h); and t = average hydraulic retention time (HRt) in stage n (hours).

a k rate of 0.083 L/mg·h was determined to be appropriate for municipal waste-water, and a tank volume/media area relationship of 4.9 L/m2 (0.12 gal/sq ft) of media was used to determine HRts.

Figure 4.4 is a graph of solutions for opatken’s model (1980) for various influent Bod5 concentrations and the assumptions indicated as a function of organic loading. again, 50% sBod5 was assumed for both RBC influent and effluent.

3.3 empirical ModelBecause the RBC process is an attached-growth process similar to the aerobic biotower and trickling filter processes, mathematical relationships developed for those pro-cesses can be used to predict the performance of RBC systems. Benjes (1977) pre-sented an empirical relationship to predict the performance of RBCs. the relationship

125

150

100

90

80

70

60

BO

D5

rem

oval

, %

50

40

300.5 1.0 2.0 3.0 3.5

Organic loading, lb BOD5/d/1000 sq ft1.5 2.5

100

75

50

Rotating biologicalcontactor influent

BOD5

Figure 4.4 opatken RBC predictive model, second-order kinetics, Bod5 removal: (1) values are total Bod5; (2) >13°C (>55°F) temperature; (3) 50% sBod5 assumed; (4) three stages, with 50% of media area in the first stage; and (5) standard-density media (lb/d/1000 sq ft × 4.882 = g/m2·d).





was based on the Velz equation (1948) (see Chapter 3) and schulze (1960), as applied to trickling filters and biotowers. the relationship predicts the fraction of influent Bod5 remaining in the effluent from the process, as a function of media volume and hydraulic loading, as follows:

SS

ee

i

k V Q= − ( / ) .0 5

(4.4)

Where Se = effluent total Bod5 (mg/L),

Si = influent total Bod5 (mg/L),e = the natural number 2.7183,

V = media volume (m3 [cu ft]),Q = average flowrate (m3/m2·d [gpm]), andk = reaction constant, 0.30.

to establish the reaction constant, Benjes (1977) reviewed operating data from a number of plants. Figure 4.5 is a graph of solutions to the Benjes model for various influ-ent concentrations, for the assumptions shown, as a function of organic loading rate.

125150


total BOD5

100

75

50

95

85

75

65

55

450.5 1.0 2.0 3.0 3.5

Organic loading, lb BOD5/d/1000 sq ft1.5 2.5

90

80

70

60

BO

D5

rem

oval

, %

50

Figure 4.5 Benjes RBC predictive model, Bod5 removal: (1) values are total Bod5; (2) >13°C (>55°F) temperature; and (3) standard-density media (lb/d/1000 sq ft × 4.882 = g/m2·d).





3.4 Manufacturers’ design curvesManufacturers of RBC equipment use design relationships that were developed from experimental and full-scale studies. envirex, Inc. (1989) uses a design approach based on a family of curves to predict effluent Bod5 as a function of hydraulic load-ing rate and influent sBod5. slightly higher performance is predicted for air-drive units compared with mechanical-drive units. design curves were generated from an experimentally developed relationship between effluent sBod5 and organic loading (kgsBod5/m2·d) of the media. Figure 4.6 shows these design relationships, assuming 50% sBod5 and the other assumptions shown. Figure 4.6 also shows the relationship recommended by envirex, Inc. (Waukesha, Wisconsin), which applies to the efflu-ent from upstream biological treatment processes and relatively low RBC influent Bod5 (<100 mg/L). Figure 4.6 also shows the design relationships used by Lyco, Inc. (Marlboro, new Jersey), another RBC manufacturer. Lyco, Inc., uses total Bod5 load-ing as the basis for its design methodology. a third manufacturer, Walker Process, Inc., does not publish design curves and relies on a computerized design technique (Walker Process, Inc., 1992).

In addition to the general application of their design procedures to establish media surface-area requirements, manufacturers use other factors, including maximum

50

40

30

20

102.0 3.0 4.0 5.0

Air drive

Mechanicaldrive

5.5Organic loading, lb BOD5/d/1000 sq ft

2.5 3.5 4.5

45

35

25

15

Effl

uent

tota

l BO

D5,

mg/

L

Figure 4.6 Manufacturers’ RBC design basis, Bod5 removal: (1) values are total Bod5; (2) >13°C (>55°F) temperature; and (3) 50% sBod5 assumed (envirex, Inc., 1989) (lb/d/1000 sq ft × 4.882 = g/m2·d).





organic loadings to individual stages of the system, temperature corrections for cold-weather operation, wastewater characteristics, and other site-specific informa-tion. In some cases, site-specific factors dictate the application of safety factors in determining media area requirements (envirex, Inc., 1992).

3.5 comparison of Model PredictionsWith respect to predictive relationships presented in the preceding sections, a com-parison can be made of the organic loading that would result in a given level of treatment for each of the relationships. table 4.1 shows this comparison for typical domestic wastewater of 125 mg/L total Bod5 in the primary effluent, 50% of which is soluble, and for two different levels of treatment—76% total Bod5 removal to achieve secondary treatment and 90% total Bod5 removal.

as indicated in table 4.1, there is a wide range in predicted organic loading rates applicable to the two treatment levels, as predicted by the models or design approaches. the Benjes model predicts the largest surface-area requirement in each case and would require approximately 1.9 times the media area for the secondary treatment application and more than 3.0 times the media area for the advanced treatment application than that predicted by the model or design approach that yields the lowest media-area requirement. therefore, the design engineer should use caution in the selection and application of design models in this situation.

Table 4.1 Comparison of organic loadings for models.

Predictive model or design approach

Loading (kg total Bod5/m2·d [lb/d/sq ft])

76% removal 90% Removal

Clark model 12.7 (2.6) 6.8 (1.4)

opatken model 15.1 (3.1) 3.9 (0.8)

Benjes model 11.2 (2.3) 3.4 (0.7)

envirex, Inc., mechanical drive 19.5 (4.0) 10.7 (2.2)

envirex, Inc., air drive 21.0 (4.3) 12.2 (2.5)

envirex, Inc., <100 mg/L influent 13.2 (2.7) 7.3 (1.5)

Lyco, Inc. 17.6 (3.6) 8.8 (1.8)

Walker Process, Inc. 15.1 (3.1) 8.3 (1.7)





3.6 Predicted Performance versus Full-scale dataFigure 4.7 compares performance predicted by the manufacturer’s design relation-ships (envirex, Inc., 1989; Lyco, Inc., 1992) with data from 16 full-scale plants (Weston, Inc., 1985) and data from 11 additional full-scale plants (doran, 1994). as indicated, the performance of the 27 full-scale plants was significantly lower than predicted by the manufacturer’s design method (envirex, Inc., mechanical and air drive and Lyco, Inc.) at a given organic loading rate. the envirex, Inc., design curve for biological effluent and low influent Bod5 concentrations (<100 mg/L total Bod5) compares more favorably with the plant data, but still would predict greater Bod5 removal than achieved at a given organic loading.

Figure 4.8 compares performance predicted by the Benjes relationship with data from the same 27 plants as those in Figure 4.7. Figure 4.8 shows the Benjes solutions for 75, 125, and 175 mg/L RBC influent total Bod5 and the data from the 27 plants shown in subsets of 50 to 100, 100 to 150, and greater than 150 mg/L total Bod5 in full-scale plant influent. as indicated in Figure 4.8, predictions from the Benjes rela-tionship compare well with data from the 27 plants over a wide range of conditions.

as shown in Figure 4.8, some of the plant data indicate better performance, and some of the data indicate worse performance, compared with the performance

120

100

80

60

40

20

00 1 2 3

Organic loading, lb BOD5/d/1000 sq ft

Effl

uent

tota

l BO

D5

mg/

L

4

Mechanicaldrive

Plant data

Airdrive

5 6 7 8

Figure 4.7 equipment manufacturers’ predictions versus full-scale plant data, Bod5 removal (lb/d/1000 sq ft × 4.882 = g/m2·d).





predicted by the Benjes relationship. the Benjes model, with k = 0.3, would appear to indicate expected average performance, but would not be suitable for a conservative design approach, without applying a safety factor.

Figure 4.9 shows a family of curves representing k rates from 0.2 to 0.4 superim-posed on the same plant data in Figures 4.7 and 4.8. a k of 0.27 would encompass approximately 75% of the data in Figure 4.9 and would be analogous to a safety fac-tor of approximately 25% in media volume compared with the use of a k of 0.30. a k of 0.25 would encompass approximately 85% of the data of Figure 4.9. Larger safety factors may be dictated by more stringent maximum daily or weekly effluent limita-tions, high peak-to-average organic-loading ratios, and other site-specific factors.

3.7 temperature correctionCold wastewater temperatures can be expected to reduce the rates of microbial growth and substrate use. the van’t Hoff-arrhenius relationship commonly has been applied to biological systems to adjust kinetic constants with temperature.

KK

T

20

= ( − 20)θ Τ (4.5)

Organic loading, lb BOD5/d1000 sq ft1

Data50–100 mg/L

> 150 mg/L100–150 mg/L

2 3 4 5 6

75

125

175


BOD5

7 8040

50

60

70

BO

D5

rem

oval

, % 80

90

100

Figure 4.8 Benjes RBC model predictions versus full-scale plant data, Bod5 removal: (1) values are total Bod5; (2) k = 0.30; (3) >13°C (>55°F) temperature; and (3) standard-density media (lb/d/1000 sq ft × 4.882 = g/m2·d).





WhereKt = reaction rate at temperature, T (°C);K20 = reaction rate at 20°C; andθ = temperature coefficient.

theta values from 1.01 to 1.05 for Bod5 removal have been applied to attached-growth biological treatment systems. For the RBC process, it has been reported (Weston, Inc., 1985) that actual full-scale field data are sparse and have not provided adequate verification of the temperature coefficient.

Benjes (1977) evaluated data from several full- and pilot-scale plants to deter-mine the effect of temperature on the k rate for eq 4.4. From a typical k of 0.3 at tem-peratures of more than 13°C, the k was reduced to typical values of approximately 0.2 at 7°C and 0.15 at 5°C.

all major manufacturers of RBC equipment (envirex, Inc., 1989; Lyco, Inc., 1992; Walker Process, Inc., 1992) provide adjustment factors in their design approaches for adjusting media surface area requirements for design temperature, as shown in table 4.2.

Hydraulic loading rate, gpd/sq ft

21.510.50

0.1

0.2

0.3

Se/

Si 0.4

0.5

0.6

Plant data0.20

0.25

0.30

0.35

0.40

Benjes, Inc., k

2.5 3 3.5 4

Figure 4.9 Benjes RBC predictive model, effect of k rate on predicted performance (standard-density media assumed) (envirex, Inc., 1989) (gpd/sq ft × 40.74 = L/m2·d).





With the correction factors in table 4.2, the media surface area determined for 13°C (55°F) is divided by the applicable correction factor to determine the required media surface area at the design temperature. nitrifying systems are predicted to be more sensitive to cold temperatures than systems designed for Bod5 removal. this is a similar situation to experiences with suspended-growth systems. Cold-temperature solids retention times are more than twice as high as those required at warmer tem-peratures for effective nitrification, depending on system conditions (u.s. ePa, 1977).

4.0 RotatIng BIoLogIcaL contactoR nItRIFIcatIon ModeLs

Rotating biological contactor systems have been designed to provide nitrification in combined Bod5 removal and nitrification systems, where sufficient media surface area is provided to accomplish both Bod5 removal and ammonia oxidation. Less commonly, the RBC process has been used to nitrify effluent from other types of bio-logical treatment processes. the total Kjeldahl nitrogen (tKn) concentration should be used when determining the ammonia concentration, because most of the organic nitrogen will be converted aerobically to ammonia.

For typical domestic wastewater, RBC manufacturer design information (envirex, Inc., 1989; Lyco, Inc., 1992; Walker Process, Inc., 1992) indicates that nitri-fication begins in the RBC process when the total Bod5 concentration has been reduced to 30 mg/L or less (15 mg/L sBod5 or less). Brenner et al. (1984), based

Table 4.2 Media surface adjustment factors.

(°c [°F])

temperature correction factor for Bod5 removal ratios

envirex, Inc. Lyco, Inc. Walker Process, Inc.

18 (64) 1.00 1.00 1.00

13 (55) 1.00 1.00 1.00

10 (50) 0.87 0.83 0.87

7 (45) 0.76 0.71 0.73

6 (42) 0.67 0.66 0.65

4 (40) 0.65 0.62 —





on an analysis of plant data, reported that maximum nitrification rates may not be achieved by the RBC process in some cases until sBod5 values are reduced to approximately 5 mg/L.

a relationship based on Monod kinetics was proposed by Pano et al. (1981), as follows:

Z kC

K CNi

i

=+

n (4.6)

Where Z = nH3-n removal rate in RBC stage (g/m2·d [lb/d/1000 sq ft]), kn = maximum nH3-n removal rate (g/m2·d [lb/d/1000 sq ft]), Ci = nH3-n concentration in RBC stage (mg/L), andKn = nH3-n removal half-saturation constant (mg/L).

the values for kn and Kn of 0.478 and 0.4, respectively, were developed from pilot-scale data at 15°C (59°F) (Brenner et al., 1984).

the design basis used by one RBC manufacturer (envirex, Inc., 1989) is based on a zero-order ammonia oxidation rate at 13°C (55°F), of 1.5 g/m2·d (0.3 lb/d/1000 sq ft) of media for substrate concentrations of more than 5 mg/L nH3-n, with a first-or-der ammonia removal rate for substrate concentrations of less than 5 mg/L. envirex, Inc. uses a family of design curves, as shown in Figure 4.10, to determine the required media surface area for nitrification.

Based on an evaluation of full-scale data at elevated temperatures, Brenner et al. (1984) report that the maximum zero-order rate of 1.5 g/m2·d (0.3 lb/d/1000 sq ft) does not increase with increasing temperatures above 13°C (55°F). this agrees with the recommendations of two manufacturers (envirex, Inc., 1989; Lyco, Inc., 1992) that surface-area adjustments should not be made for temperatures of more than 13°C. It is possible that ammonia removal rates of more than 1.5 g/m2·d (0.3 lb/d/1000 sq ft) are limited by the oxygen-transfer capabilities of the pro-cess. also, high predation may be a factor in reducing nitrification rates at elevated temperatures.

Based on an evaluation of limited data, Brenner et al. (1984) concluded that the average performance of full-scale data evaluated was close to the design basis of one manufacturer (envirex, Inc.), although the plant data were scattered widely, with some data exhibiting ammonia removal rates substantially less than 1.5 g/m2·d. this suggests that a safety factor is needed for design.





Lyco, Inc. (1992), recommends that the following specific nitrification rates be used for the indicated ranges of ammonia removal: removal range of nH3-n = 45 to 5, 5 to 3, and 3 to 2 mg/L for removal rates of 1.5, 1.2, and 0.75 g/m2·d, respectively.

With the above approach, Lyco, Inc. further recommends that the mass of ammo-nia (kilograms per day) in each of the ranges be computed; that the required media area (square meters) be computed for each range, by dividing the appropriate mass by the removal rate; and that the required media area for each range be summed to determine the total media surface area requirement for nitrification. this approach is similar to one from envirex, Inc. (1989), in that a maximum ammonia removal rate of 1.5 g/m2·d is assumed, and the removal rate is assumed to decrease for ammonia concentrations of less than 5 mg/L.

In sizing RBC systems for nitrification, it is important to recognize that organic nitrogen present in the process also may be available for nitrification because of its hydrolysis to ammonia during the process. Randtke et al. (1978) determined that sol-uble organic nitrogen from full-scale activated sludge plants averaged 1.5 mg/L and

6

5

4

3

2

1

00 1.0 1.5

30 25 22 20 19 18 17 16 15 1413

12

11

10

2.0 2.5Hydraulic loading, gpd/sq ft

Effl

uent

am

mon

ia–n

itrog

en, m

g/L

Wastewater temperature > 13 °C (>55 °F)Influent ammonia–nitrogen, mg/L

3.0 3.5 4.0 4.5 5.0

Figure 4.10 Manufacturers’ design basis, nitrification of domestic wastewater (envirex, Inc., 1989) (gpd/sq ft × 40.74 = L/m2·d).





that approximately 1 mg/L of this was refractory organic material present in the raw wastewater. the amount of refractory organic nitrogen can vary significantly, if there are industrial contributions to the wastewater. on the basis of their studies, Barth and Bunch (1979) proposed that the following quantity of influent tKn is assumed to be available for nitrification and used as the basis of the RBC sizing:

tKna = tKnt – 1 mg/L – (0.055 Bod5r) (4.7)

WheretKna = tKn available for nitrification (mg/L),tKnf = total tKn in RBC influent (mg/L), andBod5r = Bod5 removed in RBC process (mg/L).

this relationship assumes that approximately 1 mg/L of the influent tKn is refrac-tory and not available for nitrification and that 0.055 times the Bod5 removed by the process is used for the synthesis of new cell mass. the remaining tKn (the tKna value) would be available for nitrification and be used as the basis of the RBC sizing. Brenner et al. (1984) also stressed the importance of considering the tKn content of solids processing recycle streams (digester supernatant, dewatering filtrate, heat pro-cessing liquors, etc.) and inclusion of these loadings in the process design.

Chapter 3 presents additional information regarding various kinetic models used to design fixed-film systems.

5.0 denItRIFIcatIon aPPLIcatIonBecause an sBC is almost entirely submerged, oxygen transfer can be minimized. thus, an anoxic environment suitable for denitrification can be established. applications of sBCs for denitrification are limited, but one interesting example is the Wallingford Wastewater treatment Plant (WWtP) in Connecticut (Bradstreet et al., 2009). the existing 30 000 m3/d (8.0-mgd) WWtP incorporated a pre- denitrification system to their existing RBC plant by recycling settled solids from the final clarifiers to a basin with 2.1 hours HRt at a capacity of 25 000 m3/d (6.5 mgd). Pilot studies confirmed the ability of the process to establish a mixed liquor in the basins. after a period of operation during which the performance was opti-mized, an average annual effluent total nitrogen concentration of less than 8.0 mg/L was achieved. efforts to improve performance focused on minimizing the dissolved oxygen recycle to the pre-denitrification zones, providing adequate wastewater carbon for denitrification, and establishing the optimum mixed-liquor suspended





solids (MLss) levels (1800 to 2000 mg/L in winter and 3800 to 4000 mg/L in summer). Higher MLss levels tend to stress the final clarifiers, because, as is typical with RBC plants, they were not designed for such loadings. a post-denitrification system using sBCs with supplemental carbon addition was planned to decrease the effluent total nitrogen concentration to less than 5.0 mg/L. the design recom-mendation from the manufacturer was based on a loading of 14.4 g/m3 (0.9 lb no3

-

/1000 cu ft) safety factor of submerged biological contactor (sBC) media surface area.

6.0 PHysIcaL desIgn FeatuRes6.1 Physical LayoutIn planning physical layout, a designer should consider equipment access for main-tenance purposes and future plant-expansion requirements. access to bearings, load cells, drives, air valves, and other equipment requiring routine inspection and main-tenance should be provided for safe and convenient operation and maintenance by WWtP staff. the designer should consider the potential need for shaft removal for repair and replacement and access isles between trains for crane access. tank drains should be provided to facilitate dewatering of tankage for inspection and mainte-nance. Portable blowers or other provisions for adequate ventilation should be pro-vided for safe entry to any confined space.

the future construction of parallel trains or other processes to provide future capacity or enhanced treatment should be considered during design. adequate space should be reserved on-site for these facilities, as appropriate. “Knockouts” can be pro-vided in channels, division boxes, and so on, to facilitate future plant construction.

6.2 tank Volumethe tank volume to contain the RBCs typically is sized at 4.9 L/m2 (0.12 gal/sq ft) of media for low-density units. tanks for high-density units have been designed to be of the same size and configuration as for a low-density shaft having the same physi-cal dimensions to simplify layout and for economic construction.

6.3 Hydraulics and Flow controlFlow through the RBC process is by gravity from the process influent to the clari-fier effluent. In hydraulic design, the designer should consider providing effective





flow-splitting between parallel units, the ability to isolate process trains or to bypass tanks arranged in series to facilitate maintenance, and the ability to step-feed the first stages of the process to control first-stage organic loading. the ability to provide step-feeding may be of particular importance to multi-shaft first stages, to allow for more uniform loading to the shafts making up the stage.

Flow commonly is split to parallel trains by providing a low headloss influent channel perpendicular to the train flow path and by providing free discharge weirs at the influent end of each train. aeration of the influent channel often is provided in such designs to reduce the potential for septicity and maintain solids in suspension in low-velocity influent channels. Feed to the influent channel may be best accom-plished in a symmetrical fashion, to reduce potential problems with uneven distribu-tion of loading to process trains.

Isolation of process trains can be accomplished readily by using gates or stop logs at the influent and effluent ends of the trains. a free discharge on the effluent end and wide low-head influent weirs improve the capability of effective hydraulic isolation for maintenance. this is because a free discharge will prevent the effluent from backing up into an isolated channel (should isolation gates not be installed), and a wide low-head weir will minimize the increase in the water-surface elevation in the RBC trains remaining in service.

6.4 MediaRotating biological contactor units are provided in low-, medium-, and high-den-sity configurations, with low-density units having a media surface area of approx-imately 118 m2/m3 (36 sq ft/cu ft), and high-density units having a media surface area of 180 m2/m3 (55 sq ft/cu ft). this compares with 100 m2/m3 (30 sq ft/cu ft) (standard density) and 140 m2/m3 (42 sq ft/cu ft) (high density) for plastic biotower media.

Low-density media are used in the first stages of RBC systems designed for Bod5 removal, to reduce potential media clogging problems and overweight prob-lems caused by higher biomass accumulations. High-density media typically are used in the latter stages of systems designed to achieve relatively low Bod5 val-ues and in portions of systems designed for nitrification. Medium-density media are recommended by one manufacturer (Lyco, Inc., 1992) as a transitional media to be installed following the initial stages of the process, where low-density media are used, and preceding high-density media used in low Bod5 loading stages and for nitrification.





6.5 drive systemsBoth mechanical- and air-drive systems are available for RBC systems. Mechanical-drive systems consist of an electric motor, speed reducer, and belt- or chain-drive components separately provided for each shaft. Multistage speed reducers also are used instead of belt or chain drives. typically, 3.7-kW (5-hp) mechanical drives have been provided for full-size units (media approximately 3.5 m in diameter and 7.5 m in length). air-drive units involve remote blower facilities delivering air to header systems located under each shaft. air headers are equipped with coarse-bubble dif-fusers, and approximately 4.2 to 11.3 m3/min (150 to 400 scfm) of air are provided per standard size shaft, depending on media density, revolutions per minute, wastewater temperature, slime thickness and characteristics, and other conditions. the quantity of air required for air-drive systems is difficult to predict and can vary considerably from shaft to shaft and over time, depending on environmental and loading condi-tions. Mechanical-drive units have been designed for operation from 1.2 to 1.6 rpm, and air-drive units have been designed for operation from 1.0 to 1.4 rpm.

Rotational speeds should be consistent, and an even biomass growth should be maintained, to avoid uneven shaft weight. unbalanced growth may cause cyclical loadings in mechanical-drive systems and loping (uneven rotation) in air-drive sys-tems. a loping condition often accelerates and, if not corrected, may lead to inade-quate treatment and the inability to maintain rotation in the unit.

For mechanical-drive systems, materials, and motor designs should be selected that will resist the corrosive effect of a humid operating environment. Rotational-speed flexibility should be provided by alternative pulley or sprocket ratios or by providing variable-speed drives.

air-drive systems should provide ample reserve air supply to maintain rota-tional speeds, restart stalled shafts, and provide short-term, relatively high rotational speeds to control excessive or unbalanced biomass growth. available data indicate that an air rate of 11.3 m3/min (400 scfm) or more per shaft may be needed to main-tain 1.2 rpm in a heavily loaded situation (Brenner et al., 1984). Large-capacity air cups (150 mm) typically are provided in the first stages of the process. these air cups pro-vide the capability to exert a greater torque on the shaft and reduce loping problems. smaller-capacity air cups (100 mm) typically are provided on those stages receiving low Bod5 loadings and those designed for nitrification (envirex, Inc., 1992).

With air-drive systems, manual adjustments to the air rate to a given shaft may affect the air delivered to other units and their rotational speed. this problem has been overcome at some installations, by providing rotational-speed sensors and automatic





adjustment of valve settings to individual shafts to maintain selected revolutions per minute. such systems also have been designed to provide sensing and alarms for uneven rotation or loping conditions.

6.6 coversthe RBC process requires covering for protection from atmospheric conditions, media deterioration from uV light, and algal growth. Rotating biological contactor systems have been installed in buildings or under prefabricated fiberglass-reinforced plastic (FRP) covers. the FRP covers should be equipped with doors, to allow inspec-tion of the growth on the media. Buildings have been constructed of masonry, treated wood, and pre-engineered metal. the FRP covers typically are designed in sections to facilitate shipment to the job site and allow the subsequent removal of cover sections for RBC repairs or shaft removal. For designs using a building structure to house the RBC units, provisions also should be made for removing roof sections, shaft removal and replacement, or other major repairs.

the designer also should consider adequate ventilation, humidity and conden-sation control, heat loss, and corrosion caused by the humid atmosphere within the building or cover. For this reason, some FRP cover designs provide bearing and drive locations outside the environment of the RBC.

6.7 Biomass controlexcessive biofilm thickness can result in process impairment because of excessive or uneven shaft weight, loping in air-drive systems, media clogging, excess energy con-sumption, and nuisance growths and odors caused by anaerobic conditions within the slime layer.

Facilities should be provided in the design of the system to monitor shaft weight, as an indication of biomass buildup, and as a means to control shaft weight, if bio-mass develops beyond the range recommended by the manufacturer. Load cell devices commonly are provided to allow manual weighing of the shaft with a hand hydraulic pump and a pressure-sensing device. electronic strain gauge load cells also are available.

excess biomass may be controlled by removing interstage baffles or step-feeding to reduce the stage organic loading; increasing rotational speed; temporarily remov-ing a train from service and starving for a period of time; supplemental aeration; alternately reversing the rotational direction; or chemically stripping the media. the design should allow one or more means for biomass control.





7.0 RotatIng BIoLogIcaL contactoR desIgn exaMPLes

7.1 secondary treatment design examplePreliminary work has established the following design loadings (influent to the RBC system), including internal recycle loadings:

average flow = 19 000 m• 3/d (5.0 mgd), influent Bod5 =125 mg/L;

Peak flow = 37 900 m• 3/d (10.0 mgd), minimum temperature = 10°C (50°F); and

effluent requirements = 30 mg/L total Bod• 5.

(1) From eq 4.4, for k – 0.30 (warm weather), In(Se/Si) = −k(V/Q)0.5

In(30/125) = −1.43 = −0.30(V/Q)0.5

4.77 = (V/Q)0.5

22.7 =V/Q, cu ft/gpm V = 22.7(5.0)(106/1440) = 78 900 cu ft

(2) number of shafts (2750 cu ft/shaft), (78 900 cu ft)/(2750 cu ft/shaft) = 28.7 shafts. add approximately 25% for factor of safety, 28.7(1.25) = 35.9 shafts adjust for 10°C (50°F) temperature, using 0.87 correction factor, 35.9(1/0.87) = 41.3 shafts (4 150 000 sq ft)

(3) size first stage for a maximum of 6 lb Bod/d/1000 sq ft, 5 mgd (8.34 lb/gal) (125 mg/L) = 5212 lb Bod5/d (5212 1b Bod5)/(6 lb Bod5/d/1000 sq ft/d) = 868 700 sq ft or nine low-

density (100 000 sq ft each) media shafts.

(4) Choose 10 trains of four (would reduce safety factor to 21%).

5212 lb Bod /d

[(10 shafts) (100 000 sq ft/shaft)] lb B5 = 5 2. ood /d/1000 sq ft5

this is an acceptable loading. the baffle between the first and second shafts could be removed to reduce loading further, and three stages would remain.





7.2 advanced secondary treatment design examplea facility having the above design loads and flows, with RBC influent ammonia and tKn of 25 and 35 mg/L, respectively, should meet the following treatment stan-dards: effluent Bod5 = 20 mg/L and effluent nH4-n = 5 mg/L.

(1) shafts required for Bod5 removal down to 30 mg/L (onset of nitrification) equal 40 (from above example).

(2) determine ammonia to use for design purposes. 35 mg/L influent tKn – 1 mg/L refractory effluent tKn – 6 mg/L tKn

removed by synthesis [0.055] [125 – 20] = 28 mg/L

(3) Because effluent ammonia is 5 mg/L, base is on 0.3 lb nH4-n/d/1000 sq ft.

5(8.34) (28 to 5)lb/d(0.3 lb/1000 sq ft/d)

3 197 000 sq = fft

add approximately 25% safety factor.(3 197 000 sq ft) (1.25) = 4 000 000 sq ft

10°C (50°F) temperature correction factor = 0.78(4 000 000 sq ft) / (0.78) = 5 150 000 sq ft

number of 150 000-sq ft shafts(5 150 000 sq ft)/(150 000 sq ft/shaft) = 34 shafts

(4) select arrangement. select 11 trains of four low-density and three high-density shafts per train.

this will provide a total of 869 000 m2 (9350 000 sq ft) of media, which compares with the computed total of 864 000 m2 (9 300 000 sq ft) (386 000 + 478 000 m2 [4 150 000 + 5 150 000 sq ft]).

(5) Check media required to achieve Bod5 reduction to 20 mg/L. %Bod5 removed = [(125 to 20)/125]100% = 84%

From Figure 4.5, choose an organic loading rate of 1.25 lb Bod5/d/l000 sq ft.

adjust for temperature. (1.15 lb Bod5/d/1000 sq ft) (0.87) = 1.00 lb Bod5/d/1000 sq ft Include 25% safety factor.





(1.00 lb Bod5/d/1000 sq ft) / (1.25) = 0.80 lb Bod5/d/1000 sq ft Compute loading based on above design. (5212 lb Bod5/d)·[(9350)1000 sq ft] = 0.57 lb Bod5/d/1000 sq ft

therefore, the selected design is acceptable from an overall Bod5-removal standpoint.

8.0 PRoBLeMs and coRRectIVe actIonssignificant problems that have been experienced in applying the RBC process and cor-rective actions that have been used to reduce those problems are discussed in this sec-tion. Problems addressed in the following section include inadequate treatment capacity that must be resolved by loading reduction or increased plant capacity, high-effluent suspended solids and associated Bod5, and corrosion of media-supporting systems.

Where treatment performance is inadequate, the cause or causes of the inade-quate treatment should be understood fully before designing process improvements. For example,

High effluent soluble substrate concentrations may indicate inadequate RBC •capacity, high loadings, or process upset;

Low-effluent soluble substrate concentrations, but high total effluent substrate •concentrations, indicate inadequate solids removal;

nutrient imbalances, excessive pH swings, and highly variable organic load-•ings indicate problems caused by industrial discharges;

depressed pH and low alkalinity in systems designed for nitrification indicate •the need for supplemental alkalinity;

High instantaneous flows may disrupt the process, by limiting retention times •in the RBC system and interfering with clarification processes.

Plant surveys should be performed to allow the RBC process performance to be characterized fully, and data should be collected from significant sources of plant loading, as appropriate, to provide the necessary information for implementing an effective remedy for poor process performance. Before proceeding with significant plant expansion, the designer should investigate the feasibility of influent loading controls, enhanced preliminary-primary treatment, controls or separate treatment of recycle streams, or improved suspended-solids control, to improve the overall per-formance of the RBC system.





8.1 Inadequate treatment capacitya number of RBC facilities have been constructed that have not achieved the degree of treatment anticipated during design. typically, this has been traced to inadequate consideration of recycle loadings to the plant from solids processing, performance lower than that predicted by the RBC vendor’s design curves, or unanticipated indus-trial or other loadings.

the resolution of treatment capacity problems has been accomplished by con-structing additional treatment units or taking other appropriate steps to reduce sys-tem organic loading. the following steps potentially could apply:

Control or separate treatment of recycle sidestream loadings;•

Implementation of industrial waste pretreatment programs;•

Improvement of the degree of preliminary and primary treatment provided;•

Installation of additional RBCs;•

Construction of biotowers in series or parallel with the RBCs; and•

Construction of activated sludge facilities, in series with or parallel to the RBCs,•

additional RBC units would reduce the organic and hydraulic loading on the RBC facilities and increase the degree of treatment provided. Reduced loadings also would tend to mitigate problems with oxygen deficiency, nuisance microorganism growth, and excessive shaft weight, which are symptomatic of overloaded systems. the construction of other biological treatment systems, either in parallel or upstream of the RBCs, will accomplish the same results.

With any of the above augmentations of the RBC process, the maximum capa-bility of the RBCs would need to be assessed, and the most economical and practical process additions defined. Blending the effluents from a down-rated RBC system and another process, such as activated sludge, may apply, in some situations.

8.2 excessive First-stage Loadingsthe following actions should be considered to resolve excessive first-stage loadings:

Remove interstage baffles to increase the media area provided in the first RBC •stage;

Increase RBC rotational speed to increase oxygen transfer and encourage •media sloughing;





add supplemental aeration to mechanical-drive systems;•

step-feed process influent to bypass the first stage, with some of the flow and •loading, thereby spreading the organic loading between the first stages of the process; and

add chemicals to the raw wastewater to enhance Bod• 5 removal during pri-mary treatment.

8.3 excessive Biomass growthexcessive biomass growth typically results from excessive organic loadings. If the system capacity is inadequate to allow biomass control by the methods described, the following actions should be considered:

Chemical stripping of the media with caustic soda or other appropriate chem-•ical treatment to remove slime buildup,

alternate reversal of the RBC rotational direction, and•

discontinuation or reduction of flow to the overweight train to starve biomass •and reduce mass by endogenous metabolism.

using the above procedures may have a negative short-term effect on the treat-ment efficiency. also, the above procedures temporarily will increase loadings to the units remaining in service and may accelerate biomass growth on those units. In facili-ties subject to chronic overloading conditions, it is important to carefully monitor shaft weights, by making routine load cell measurements. this is particularly true during the winter (lower endogenous rate) and during periods of high plant loadings.

8.4 Loping of air-drive systemsLoping results from uneven biomass growth, which causes a non-uniform rotational speed. When established, a loping condition may be difficult to control in a given shaft. decreasing the loading to the RBC stage or increasing the air rate and rota-tional speed may prove successful. Installing water feed to the air cups (water-assist) at a point past their uppermost rotation is a manner of increasing rotational torque. this practice will increase the effective loading on the shaft. For a shaft with repeated loping problems and with 100-mm (4-in.) air cups, the air cups may be replaced with larger 150-mm (6-in.) cups. another solution, such as chemical cleaning or starving, may be necessary to resolve a serious loping problem.





8.5 High clarifier effluent suspended solidsFine colloidal solids typically will be flocculated and enmeshed during biological treatment, resulting in their removal in secondary clarification. overflow of fine sus-pended solids may be a chronic problem in maintaining compliance with effluent limitations. success in enhancing suspended solids removal in clarification has been reported in Lancaster, Wisconsin, where recycle from the final clarifiers is recircu-lated to the influent of air-drive RBCs to form a mixed liquor. the flocculent nature of the mixed liquor is reported to have enhanced the control of fine suspended solids, resulting in a substantial improvement in effluent quality (doran, 1994).

Certain RBC process modifications for enhanced effluent suspended-solids con-trol, such as installation of RBCs in activated sludge aeration tanks and recirculation of secondary solids to a RBC contact zone, may be subject to patent restrictions.

the use of polymer and other coagulant aids during clarification and filtration of the clarifier effluent also may be used for enhanced suspended-solids control. an excessive overflow rate or the design of the clarifier may be the cause of a high efflu-ent suspended solids concentration.

8.6 corrosion of Media supportsCorrosive failure of media support systems has been reported at some WWtPs. this situation may be the result of site-specific water chemistry, biological degradation products, original materials selection, incompatible welding materials, or other fac-tors. the fact that some WWtPs report that this problem only occurs in particular portions or stages of the process suggests that biological activity may be an impor-tant factor. Welding repairs or selective material replacements may be required to maintain the integrity of the system. Cathodic protection also has been used, with apparent success. Manufacturers have continued to improve the structural design of the RBC, by modifying the materials of construction and the details of construction.

9.0 PILot-PLant studIesPilot-plant studies are recommended before application of the process to indus-trial or municipal wastewater containing a significant industrial waste component. Pilot-plant studies should be conducted using full-scale media, to minimize scale-up problems. Pilot-plant studies should be conducted for extended periods, to include seasonal effects, long-term effects of deposition of organic and inorganic materials on the RBC media, and other factors important to process application.





Pilot-plant investigations should be designed to include all of the factors antici-pated with full-scale installation, including daily changes in waste strength and flow. In addition, the effects of solids-processing sidestreams and other in-plant waste streams should be included.

to provide the data needed to fully evaluate the process performance, the fol-lowing data collection should be included, as appropriate:

Influent and effluent total Bod• 5 and influent and effluent sBod5 (from each stage);

dissolved oxygen, tKn, ammonia, and nitrate/nitrite stage-to-stage profiles;•

temperature trends and profiles;•

Flow and flow variability;•

shaft weight trends; and•

air requirements for air-drive systems.•

10.0 ReFeRencesBarth, e. F.; Bunch, R. L. (1979) Biodegradation and Treatability of Specific Pollutants,

ePa-600/9-79-034; u.s. environmental Protection agency: Washington, d.C.

Benjes, H. H., Jr. (1977) small Community Wastewater treatment Facilities—Biological treatment systems. Prepared for u.s. environmental Protection agency (Washington, d.C.) Technology Transfer National Seminar on Small Wastewater Treatment Systems, Culp/Wesner/Culp, el dorado Hills, California; u.s. environmental Protection agency: Washington, d.C.

Bradstreet, K., et al. (2009) Proceedings of the 82nd Annual Water Environment Federation Technical Exposition and Conference, orlando, Florida, oct 17–21; Water environment Federation: alexandria, Virginia, 1255–1276.

Brenner et al. (1984) Design Information on Rotating Biological Contactors, ePa-600/2-84-106; u.s. environmental Protection agency: Cincinnati, ohio.

Chou, C. C. (1978) oxygen transfer Capacity of Clean Media Pilot Reactors at south shore. autrotrol Corporation: Milwaukee, Wisconsin.

Clark, J. H.; Moseng, e. M.; asano, t. (1978) Performance of a Rotating Biological Contactor under Varying Wastewater Flow. J. Water Pollut. Control Fed., 50, 896–911.





doran, M. d., strand associates, Inc., Madison, Wisconsin (1994) Personal communication.

envirex, Inc. (1989) specific RBC Process design Criteria. envirex, Inc.: Waukesha, Wisconsin.

envirex, Inc.: Waukesha, Wisconsin (1992) Personal communication.

Levenspiel, o. (1972) Chemical Reaction Engineering; Wiley & sons: new york.

Lyco, Inc. (1992) Rotating Biological Surface (RBS) Wastewater Equipment: RBS Design Manual; Lyco, Inc.: Marlboro, new Jersey.

McCann, K. J.; sullivan, R. a. (1980) aerated Rotating Biological Contactors: What are the Benefits? Proceedings of the 1st National Symposium on Rotating Biological Contactor Technology, Vol. I, ePa-600/9-80-046a; Champion, Pennsylvania.

Metcalf and eddy, Inc. (1979) Wastewater Engineering: Treatment, Disposal, and Reuse; Mcgraw-Hill: new york.

opatken, e. J. (1980) Rotating Biological Contactors—second order Kinetics. Proceedings of the 1st National Symposium on Rotating Biological Contactor Technology, Vol. I, ePa-600/9-80-046a; Champion, Pennsylvania.

Pano, a., et al. (1981) The Kinetics of Rotating Biological Contactors Treating Domestic Wastewater, Water Quality series uWRL/Q-8104; utah state university, Logan, utah.

Randtke, s. J.; Parkin, g. F.; Keller, J. V.; Leckie, J. o.; McCarty, P. L. (1978) Soluble Organic Nitrogen Characteristics and Removal, ePa-600/2-78-030; u.s. environmental Protection agency: Cincinnati, ohio.

Reh, C. W.; et al. (1977) an approach to design of RBCs for treatment of Municipal Wastewater. Paper presented at American Society of Civil Engineers National Environmental Engineering Conference, nashville, tennessee.

sawyer, C. n.; Wild, H. e., Jr.; McMahon, t. C. (1973) Nitrification and Denitrification Facilities, Wastewater Treatment, u.s. ePa technology transfer; u.s. environmental Protection agency: Cincinnati, ohio.

scheible, o. K.; novak, J. J. (1980) upgrading Primary tanks with Rotating Biological Contactors. Proceedings of the 1st National Symposium on Rotating Biological Contactor Technology, Vol. II, ePa-600/9-80-046b, Champion, Pennsylvania.

schulze, K. L. (1960) Load and efficiency of trickling Filters. J. Water Pollut. Control Fed., 32, 245–253.





u.s. environmental Protection agency (1977) Process Control Manual for Aerobic Biological Wastewater Treatment Facilities, ePa-III/a-524-77; u.s. environmental Protection agency: Washington, d.C.

Velz, C. J. (1948) a Basic Law for the Performance of Biological Filters. Sew. Works J., 20, 607–617.

Walker Process, Inc., aurora, Illinois (1992) Personal communication.

Weston, Inc. (1985) Review of Current RBC (Rotating Biological Contactor) Performance and Design Procedures, ePa-600/2-85-033; u.s. environmental Protection agency: Cincinnati, ohio.




0071776257_ar004-RBC McGrawHill

Documents

1st national

rotating biological

rotating biological

1000 sq ft

environmental

plastic packing

submerged

processing