\ I /
United States
Environmental Protection
Agency
Municipal Environmental
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-84-106 July 1984
Project Summary
Design Information on Rotating
Biological Contactors
Richard C. Brenner, James A. Heidman, Edward J. Opatken, and Albert C.
Petrasek, Jr.
A document summarizing design
information for rotating biological
contactors (RBC) was developed to
provide in-depth data on the critical
features of this process and the key
factors affecting its operation and
performance. The information provided
is intended to supplement and qualify
that available from RBC manufacturers
and the published literature. Topics
addressed include process and design
considerations for carbonaceous re-
moval, nitrification, and damnification;
equipment reliability and service life;
power requirements for air driven and
mechanically driven RBC units; and
general system design considerations
involving structural, hydraulic, and
operational flexibility. Practical, usable
design information was emphasized
along with important theoretical con-
cepts.
This Project Summary was developed
by EPA's Municipal Environmental
Research Laboratory. Cincinnati. OH.
to announce key findings of the design
information document that is available
in its entirety in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
RBC's for municipal wastewater treat-
ment have been introduced relatively
rapidly to the United States. The result
has been widespread application of a
technology that is not familiar to many
design engineers. Out of necessity, many
RBC designs were initially based solely
on proprietary-generated empirical design
procedures. The lack of a comprehensive
appraisal of modified media configurations
left design engineers and equipment
purchasers with an inadequate basis for
comparing performance of alternative
equipment designs and effectively relating
anticipated performance to loading. As
interest in the process has increased,
more complex, deterministic design
approaches have begun to appear in the
technical literature.
The purpose of this project was to
prepare a design information document
that would supplement commonly ac-
cepted RBC design methodology (such as
manufacturers' design manuals and
deterministic models) by providing appro-
priate qualifiers and summaries of
operating and performance data not
readily available to the design community.
Important design parameters and rela-
tionships (or lack of them) are discussed
to promote a more rational RBC design
approach. Many aspects of the informa-
tion contained in the document are
equally applicable to industrial wastewa-
ter treatment. The document is not
intended to serve as a cookbook-style
design reference or to replace any of the
design guides mentioned above.
The full report considers the equipment
and design practices of the five current
U.S. RBC manufacturers — Autotrol (now
Envirex), Clow, Crane-Cochrane, Lyco,
and Walker Process.* Significant varia-
tions in shaft, media configuration, and
media attachment designs are evident
among the five and must be taken into
account in designing an RBC system.
• Information was originally presented
on the equipment and design practices of
a sixth U.S. RBC manufacturer. This
manufacturer has recently stopped
•Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use.
-------
marketing RBC units and is referred to
hereinafter as Manufacturer X. Descrip-
tions of Manufacturer X's equipment
have subsequently been removed from
the full report and this Project Summary,
but published references to and discus-
sions of its previous design methods have
been retained where appropriate.
Equipment Description and
Performance
The most serious equipment problem
that can affect an RBC system is the
structural breaking (failure) of a shaft. A
shaft failure usually involves temporary
loss of the affected RBC unit and perhaps
its entire train, damage to a portion of the
media, and replacement of the broken
shaft with a new shaft along with
salvaged and/or new media. Numerous
failures of first-generation shafts in this
country have been attributed to poor
fabrication and welding practices. Im-
proved second-generation designs have
corrected most of the early manufacturing
deficiencies. Shaft failures have also
occurred because of fatigue resulting
from overstressing of the member's load-
carrying capacity.
A matrix of estimated shaft fatigue
limits and shaft fatigue lives was devel-
oped for the shaft cross-sections shown
in Figure 1 as a function of shaft
attachment detail, shaft wall thickness.
and combined media and biofilm load.
Appropriate fatigue provisions of the
American Welding Society (AWS) Struc-
tural Welding Code for Design of Tubular
Structures (Section 10) and Design of
New Bridges (Section 9) were used to
determine the fatigue limit of each shaft
design and the estimated shaft life
when imposed loads exceed fatigue
limits. A shaft member subjected to a
stress range below its fatigue or endurance
limit will not fail (with a high degree of
confidence) as a result of structural
fatigue.
Fatigue loading limits for RBC shafts
can be expressed as a function of the
thickness of the biofilm attached to the
rotating media. Estimates of allowable
biofilm thickness that keep the imposed
load below the respective fatigue limits of
the six shafts varied depending on shaft
structural dimensions and characteris-
tics. The ranges calculated were 75 to
>150 mils (0.075 to >0.15 in.) for
standard density media and 50 to 125
mils (0.05 to 0.125 in.) for high density
media.
These estimates of shaft fatigue limits
were made assuming that no welding
deficiencies or shaft deterioration existed
as a result of corrosion. The weldment
provisions of the AWS Structural Welding
Code should be used as an absolute
minimum to ensure satisfactory weld-
ment fabrication. Epoxy coatings are
Clow
-28-in. O.D.
Walker Process
16-in. square
(outer
dimension)
ion)
Weld
Locations-~-
— 1 in.
Autotrol
Lyco
Crane- Cochrane
outer face
-24 in. to —
outer face
Weld
\Locations__|_
Lyco/Hormel
Figure 1. Cross-sections of RBC shafts.
routinely used by RBC manufacturers to
protect shafts from the corrosive effects
of wastewater. Available information
from coating manufacturers indicates
that epoxy coating materials have an
expected life of 5 to 10 years or more, but
long-term data are limited.
Shaft fatigue protection may also be
provided by increasing shaft section
modulus (additional shaft wall thickness)
to decrease the applicable cyclic stress
range and compensate for possible metal
loss resulting from corrosion. In selecting
and specifying RBC shafts, the designer
should require evidence from the shaft
manufacturer that satisfactory fatigue
stress protection (including protection
from possible corrosive effects) has been
provided to cover the design life of the
RBC system under expected loading
conditions, either by providing reliable
protective coating material or by modifying
the shaft design to increase the section
modulus.
Design techniques can be used to
minimize shaft overstressing from exces-
sive biofilm growth. These techniques
include (1) incoming load manipulation to
bypass temporarily the organically over-
loaded stages (usually the first two)
and/or to distribute that load to greater
media surface through step feeding or
removal of baffles and (2) variable speed
mechanical drives to reduce biofilm
thickness by means of increased shear.
For some mechanically driven systems,
strategic placement of supplemental air
headers to increase biofilm shearing
rates may also represent a cost-effective
method. A secondary benefit of variable
speed drives and/or supplemental air
would be the ability to increase tempora-
rily the concentration of bulk liquid dis-
solved oxygen (DO) in specific stages.
The plant operator must be provided
with sufficient flexibility to respond to
changing load and process conditions. In
addition to mechanisms for manipulating
load and increasing DO concentration
and biofilm shearing potential, recom-
mended operational tools include (1)
shaft load cells for estimating biofilm
weight (particularly in the critical lead
stages), (2) DO monitoring equipment to
aid in process control, (3) flow control
devices to equalize hydraulic and organic
loads to parallel trains, and (4) air flow
measuring equipment to adjust air flow
rates properly in air drive RBC installa-
tions.'
High density polyethylene (HOPE) is
used by all manufacturers in the fabrica-
tion of RBC media. The method of forming
a media pack and attaching it to the shaft
determines whether an RBC assembly
-------
must be removed from its tank for field
replacement of damaged or deteriorating
media. In designing an RBC system, the
engineer should consider the operation and
maintenance ramifications of field media
replacement. If a particular design
requires removal of the RBC assembly
from its tank to perform this task,
permanently installed shaft removal
equipment may be cost effective. The
alternative is to bring in an over- or off-
the-road crane as needed. Present media
pack attachment designs of Clow, Crane-
Cochrane, and Lyco do not require
removal from the tank for media replace-
ment, but current Autotrol, Tait/Bio-
Shafts, and Walker Process designs do.
A standard RBC assembly consists of
media approximately 12 ft in diameter
supported on or from a 27-ft-long shaft
(25-ft media length). Standard density
media are defined as those with a
nominal surface area of 100,000 sq
ft/25 ft of media length. High density
media are formed by reducing the space
between repeating plastic corrugations
by 33 percent, thus achieving a media
configuration with 50 percent more
available surface area (i.e., a nominal
surface area of 150,000 sq ft/25 ft of
media length).
Standard density media are normally
used in the front or lead stages of an RBC
train where organic strength is highest
and biofilm growth is thickest. High
density media have been used primarily
in the middle and final stages where
lower organic concentrations will not
promote and sustain thick biofilm growth.
Excepting for second-step nitrification
applications where the organic load is
known to be low, use of high density
media in lead RBC stages should be
avoided because of potentially excessive
biofilm development per unit volume of
media (resulting in shaft overstressing)
and the increased possibility of inducing
biofilm bridging between adjacent media
sheets. Recommended rules of thumb are
to limit media biofilm thickness, if pos-
sible, to 75 mils (0.075 in.) for standard
density media and 50 mils (0.05 in.)
for high density media to provide an
acceptable margin of safety against shaft
overload and possible fatigue failure
These recommendations do not apply to
Walker Process and Lyco because of the
much higher fatigue limits of their shafts.
Early media designs were subject to
degradation from the ultraviolet compo-
nent of sunlight. The use of anti-oxidants
in HOPE formulations and protective
covers have essentially eliminated this
problem. Prolonged exposure to high
temperature environments and concen-
trated organic solvents can also lead to
media degradation and/or brittleness.
Cases of media rupture resulting from the
poorly understood phenomenon of stress
cracking have also been reported. Media
can be broken while in service as a result
of lateral shifting in assemblies using
radial arm support systems if inadequate
bracing and tightening capability are not
provided. On the average, however,
media failures have not been encountered
as often as shaft failures.
The drive assembly has proven to be the
most reliable equipment component in a
mechanically driven RBC system. Occa-
sional broken drive chains are the only
problem of note. Operating experience
with air driven systems is too limited to
date to make definitive projections of
equipment performance. The critical
aspects of air driven system performance
currently appear to relate more to
achieving desired air distribution among
multiple shafts and maintaining uniform
rotational speeds of individual shafts
than they do to reliability of the air
delivery equipment itself.
In a few severe cases where biofilm
growth on RBC media has become
grossly uneven or unbalanced, lack of
positive rotational control with air driven
systems has brought shaft rotation to a
complete halt. Providing the operator
with load manipulation capability and/or
equipment to achieve rapid biofilm
stripping (e.g., high pressure water
injection) are the most effective techniques
for responding to these situations.
The full report summarizes the struc-
tural and mechanical performance of 17
RBC plants surveyed using questionnaires
and onsite visits. Data are presented for
the three major RBC system component
units (shaft, media, and drive).
Organic Removal
Under most circumstances, mass
transfer is the dominant factor affecting
organic removal in an RBC system.
Mass transfer resistances associated
with both the liquid phase and the biofilm
result in significant concentration gradi-
ents from the bulk liquid to biological
reaction sites on the media, and frequently
these resistances control system perform-
ance. Oxygen transfer becomes rate
limiting and controls the overall reaction
rate m heavily loaded systems. After all
of the factors affecting substrate removal
and oxygen transfer are considered, no
reason exists to expect substrate removal
from the -RBC bulk liquid to follow any
simple mathematical model.
Oxygen loading to any stage of an RBC
system should not exceed values compa-
tible with the oxygen transfer'capability of
the system. COD balance studies conduc-
ted on a full-diameter RBC unit have
indicated a maximum oxygen transfer
rate of approximately 1.5 Ib Oz/day/1000
sq ft. Most oxygen transfer in an RBC
takes place in the atmospheric portion of
the rotational cycle.
The presence of Beggiatoa organisms
on RBC media is associated with biofilm
DO depletion and is considered an
indication of organic overloading. In the
absence of DO, sulfide essential for
Beggiatoa growth is generated by means
of sulfate reduction and/or anaerobic
decomposition deep within the biofilm.
When designing a mechanically driven
RBC (System, research and field observa-
tion^ indicate that safe, conservative
firstfstage loading figures to avoid
orgahic overloading are 2.5 Ib soluble
BODs/day/sq ft or 6 Ib total BODs/day/
100d sq ft. These figures may be increased,
but the designer must recognize that this
increase may lead to heavier-than-
normal biofilm growth, bulk liquid DO
depletion, development of sulfide oxidizing
nuisance organisms, deterioriation of
overall process and/or mechanical
performance, etc.
A first-stage loading exceeding the
above figures may be justified depending
on the degree of operational and main-
tenance attention the plant will receive,
the structural capacity of the selected
shaft, the ability to strip excess biomass
from the media, the levels of sulfur
compounds in the RBC system influent,
the media surface area required in the
remaining stages, and the ability to vary
the operational mode of the plant. The
loading, however, should generally not
exceed 4 Ib soluble BOD5/day/1000 sq ft
or 8 Ib total BOD5/day/1000 sq ft.
Mass organic removal rates in an
overloaded RBC stage may exceed those
of a stage operating in a proper loading
range if sulfide levels are low enough to
avoid extensive Beggiatoa growth. In
many instances, however, mass removal
rates in overloaded stages may actually
decrease as Beggiatoa organisms begin to
compete seriously with desirable hetero-
trophs for oxygen and space on RBC
media. In extreme cases, the takeover of
the first stage of an overloaded RBC
system by Beggiatoa can shift the load to
the next stage,-leading to a progressive
Beggiatoa takeover of the entire system
and significant deterioration of effluent
quality.
Present procedures for scaling up pilot
test data collected on less-than-full-
diameter RBC units to full-scale design
are inadequate. Mass removal data
-------
generated on less-than-full-diameter
units operated at equivalent f ieldscale tip
speeds (60 fpm ±) tend to be optimistic
because of the higher rates of oxygen
transfer and atmospheric surface renewal
inherently achieved. The distortion is
magnified as the scale decreases.
Deterministic mathematical models
that attempt to incorporate all of the
factors affecting the RBC process have
been used in some design situations in
conjunction with the results of smaller-
scale RBC pilot studies carried out on the
wastewater in question. Though these
models provide considerable insight into
the variables and ranges of variables
affecting RBC performance, their suc-
cessful application to design depends
greatly on accurate calibration of model
coefficients.
Very small-scale RBC pilot units (1.5-ft
diameter or less) are more useful in
determining the basic treatability of a
wastewater than in establishing full-
scale design parameters. If small-diam-
eter units must be used to collect design
data, however, it is important that each
stage be loaded below the oxygen
transfer capability of a full-scale unit to
minimize scale-up considerations. Full-
diameter (12 ft ±) or nearly full-diameter
RBC pilot studies are recommended where
feasible, since they offer the highest
probability of obtaining reliable RBC
design data.
Additional factors affecting organic
removal in RBC's include, among others,
the ratio of tank volume to media surface
area (detention time), influent wastewater
strength, influent flow and load variability,
wastewater temperature, and staging
configuration. Most RBC manufacturers
have standarized ratios of tank volume-
to-media surface area at 0.12 gal/sq ft,
since higher ratios did not improve BOD
removal at equal hydraulic loading rates.
For municipal wastewaters, percentage
BOD removal normally increases with
increasing wastewater strength; the con-
verse is generally true for high-strength
industrial wastewaters.
Loading variations are not assumed by
the manufacturers to affect process per-
formance adversely at peak-to-average
flow ratios of 2.5 or less. The Autotrol and
Clow design manuals recommend using
either peak flow and load conditions for
design or flow equalization for ratios
above 2.5.
The manufacturers universally contend
that about 55°F wastewater temperature
does not affect organic removal design.
Below 55°F, varying degrees of decreased
biological activity are predicted by the
manufacturers.
The optimal number and size of stages
and the overall staging configuration for a
given design are frequently difficult to
predict. Flexible hydraulic designs that
enable the plant operator to change the
size and number of stages (e.g., step
feeding and stage bypass provisions,
removable baffles, etc.) will enhance the
potential for maximum performance from
any RBC facility.
No single best design procedure or set
of relationships has been found that can
universally predict RBC organic removal.
Empirical design approaches used by four
manufacturers (Autotrol, Clow, Lyco, and
Manufacturer X) exhibit considerable
variation in organic loading versus
predicted effluent quality (Figure 2).
Except for Lyco, these empirical design
techniques are based on predictions of
effluent soluble BOD5. Effluent total
BOD5 projections are then made assum-
ing a set ratio (e.g., 0.5) of soluble-to-total
BODs in the effluent. This assumption
ignores (1) the impact of solids settling
characteristics on final clarifier perform-
ance and (2) the wide variations in
effluent suspended solids and paniculate
BOD5 that are possible with any,given
concentration of effluent soluble BODs. In
the absence of full-diameter pilot plant
data for the design in question, any RBC
design guidelines should be used with
discretion.
To devise an improved method for
estimating organic removal, soluble
organics interstage data were evaluated
using second order kinetics. Interstage
and final effluent soluble BODs values
predicted using this approach are com-
pared in the full report with measured
values obtained at nine full-scale, air
driven RBC plants. The predicted and
measured values are in good agreement
for seven of the nine plants. The lack of
close agreement at one of the other two
plants could be explained by inadequate
oxygen transfer in the first stage to handle
the high influent organic loading. The
lack of close agreement at the ninth plant
could not be explained.
The second-order predicted values
more closely approximated measured
soluble BODs concentrations than did
values predicted by Autotrol's empirical
organic removal design method for air
driven RBC's. These results indicate that
second-order kinetics may offer an
improved basis for predicting interstage
soluble organic removals in RBC systems
that are not oxygen transfer limited; they
also suggest that further evaluations be
conducted for both mechanically and air
driven options as additional interstage
data become available. As with manufac-
turers' empirical techniques, the second-
order kinetic approach only addresses the
impact of soluble organics on effluent
quality.
Nitrogen Control
Nitrification
The factors affecting nitrification in the
RBC process are representative of those
that affect any attached growth biological
process. They include influent organic
concentration, influent nitrogen concen-
40
a 20
,o
10
AUTOTRQL
AND
CLOW
Lyco ffSOJ
Note: Soluble BODs/totalBODs =
Lyco (JOO}
Lyco 1225)
MFR X 1301
<(20)
Note: Numbers in f ) are
influent soluble BODi
concentrations for
Manufacturer X and total
BODs concentrations
for Lyco.
Figure 2.
1 2 3 4 5 6 7
Organic Loading Rate lib total BODs/day/ WOO sq ft)
Effluent BODs as a function of organic hading for selected RBC manufacturers
design techniques.
-------
tration and composition, wastewater
temperature, DO concentration, pH and
alkalinity, and influent flow and load
variability, among each others.
Most empirical design procedures are
based on the assumption that significant
nitrification does not begin in an RBC
system until bulk liquid soluble BOD5 has
been reduced to 15 mg/L, or total BOD5 to
30 mg/L. In combined carbon oxidation/
nitrification units, BOD5 values of this
magnitude may first be encountered in
the second, third, or fourth stages,
depending on influent strength, organic
loading rate, and wastewater tempera-
ture. In separate-stage nitrification RBC
applications, the soluble BOD5 concen-
tration of the wastewater entering the
RBC reactor is usually well below 15
mg/L and substantial nitrification is
typically evident in the first stage.
Analysis of field data indicates that while
incipient nitrification is generally observed
in RBC stages with soluble BOD* concen-
trations of 15 mg/L ± maximum nitrifica-
tion rates are not attained until bulk liquid
soluble BODs drops to 10 mg/L or less.
Numerous investigators have found
that RBC nitrifier growth kinetics and
ammonia nitrogen oxidation follow the
classic Monod expression. When substrate
(ammonia nitrogen) concentration is low
compared with the half-velocity constant,
ammonia nitrogen oxidation approaches
first-order removal. Conversely, when
substrate concentration is high in relation
to the half-velocity constant, oxidation
approaches a constant zero-order removal
rate.
The zero-order ammonia nitrogen
removal rate observed for RBC's and the
ammonia nitrogen concentration transi-
tion point at which it begins depend on
scale and wastewater temperature. Zero-
order removal rates as high as 0.8 Ib NH3-
N /day/1000 sq ft have been measured
on small pilot units. For full-diameter
RBC's, the maximum attainable zero-
order removal rate is generally acknow-
ledged to be about0.3 lbNH3-N/day/1000
sq ft. As with organic removal, higher
inherent rates of atmospheric surface
renewal and oxygen transfer are the
probable reasons for the higher mass nitri-
fication rates attainable with pilot RBC
units.
Because of the varying degrees of
hydrolysis that can occur in sewer lines,
RBC nitrification designs should not be
based on influent ammonia nitrogen
concentration alone. To do so risks serious
undersizing if substantial amounts of
organic nitrogen are present.
Nitrification rates drop sharply at
wastewater temperatures below 55°F.
The manufacturers have found, however,
that wastewater temperatures above 55°F
do not affect RBC nitrification. Nitrification
temperature correction factors recom-
mended by four manufacturers are
given in the full report.
Minimum DO levels of 3 to 4 mg/L are
recommended in RBC nitrifying stages to
ensure that nitrification is not limited by
oxygen transfer info the biofilm. Lower
DO levels are more likely to occur in the
transition stage from organic removal to
incipient nitrification and in the next
succeeding stage than in the final stages.
The beneficial effect on DO of compressed
air injection in air driven RBC's is
approximately offset by their lower
atmospheric oxygeriation (compared with
mechanically driven units), which results
from reduced rotational velocities (typi-
cally 65 to 75 percent of the standard
mechanically driven velocity of 1.6 rpm).
The literature recommends optimal pH
values for nitrification ranging from 7.0to
9.0. If pH drops much below 7.0 in
unacclimated systems, the nitrification
rate will be seriously retarded, decreasing
to zero somewhere between pH 6.0 and
5.0. Since nitrification reactions consume
alkalinity, alkaline addition may be
required for RBC nitrification systems
treating wastewaters with low or highly
variable alkalinity.
Nitrification rates have been enhanced
through upward pH adjustment in pilot-
scale RBC units, with the optimum rate
occurring at pH 8.5. Similar response has
not been demonstrated on full-diameter
RBC's.
The impact of flow and mass loading
variations is usually more severe on the
nitrification efficiency of RBC's than on
organic removal performance. Nitrifiers
have long generation times compared
with heterotrophs. Furthermore, external
accumulation and/or internal storage of
substrate for delayed metabolism does
not take place with nitrifiers as with
heterotrophs. Consequently, influent
surges in flow or unoxidized nitrogen
concentration (either of which increases
nitrogen mass loading) will be accompa-
nied by similar delayed (roughly equal to
reactor detention time) spikes of unoxi-
dized nitrogen in the effluent unless
adequate RBC surface is provided to
compensate for expected variations. In
design situations that specify consistently
low effluent ammonia nitrogen residuals
(1 to 2 mg/L), the amount of media
surface required should be estimated
with and without prior flow equalization
to determine which option is more cost
effective.
Since the mid-1970's, Autotrol has
conducted extensive testing and data
evaluation to model ammonia nitrogen
oxidation in RBC's. Their current empirical
procedure for full-diameter RBC units
operating at wastewater temperatures of
55 °F or greater is based on a Monod-type
curve (Figure 3). This curve projects first-
order removal (oxidation) of ammonia
nitrogen at concentrations in the stage
liquid below about 5 mg/L. Above 5 mg/L
NH3-N, removal is claimed to proceed at a
zero-order rate of approximately 0.3 Ib
NH3-N/day/1000 sq ft. This rate is
consistent with the maximum potential
nitrification rate (0.33 Ib NH3-N/day/1000
sq ft) that can be calculated based on an
assumed peak oxygen transfer capability
for 12-ft diameter media of 1.5 Ib
O2/day/1000 sq ft (see the earlier
section on Organic Removal).
Applied to a staged RBC reactor, the
above design basis predicts that when
soluble BODs drops below 15 mg/L,
ammonia nitrogen will be oxidized at a
constant (zero-order) rate as it passes
through succeeding stages down to a
bulk liquid concentration of about 5
mg/L, and thereafter at an approximate
first-order rate. Nitrification performance
data from three full-scale RBC facilities
were evaluated for wastewater tempera-
tures of 55 ± 2°F. The resulting best fit
curve of ammonia nitrogen removal rate
versus ammonia nitrogen concentration
closely approximates the Autotrol design
curve (Figure 3).
Denitrification
Denitrification can be accomplished
with RBC's using two different process
configurations. In the first, representing
the traditional biological denitrification
approach, organic removal and nitrifica-
tion occur upstream of the denitrification
reactor in either a lead-stage(s) RBC unit
or some other type of biological system. A
supplemental carbon source (commonly
methanol) must be added to the waste-
water to provide the required energy for
microbial denitrification in the denitrifi-
cation (anoxic) stage. A short-term, final,
aerated RBC stage or other polishing
biological unit may also have to be added
to oxidize any residual methanol not used
for denitrification.
A possible alternative RBC denitrifica-
tion configuration would be to place the
anoxic stage before the organic removal
and nitrification reactor(s). This technique
is well known for activated sludge
systems, but thus far it has not been used
in any full-scale RBC installations. The
organic carbon naturally present in the
incoming wastewater provides the neces-
sary energy for reducing nitrate nitrogen
-------
0.4
0.3
•S 0.2
0.1
Wastewater Temperature = 55°F
1.5 3.0 4.5 6.0 7.5 9.0 10.5
Stage NH3-N Concentration (mg/L)
12.0 13.5 15.0
Figure 3.
Second-generation Autotrol ammonia nitrogen removal rate curve for full-scale
RBC's (from: Antonie, R.L.. "Nitrogeii Control with Rotating Contactors," Autotrol
Corporation, Undated; Reprinted with*permission).
recirculated to the anoxic stage from the
downstream reactor(s). Though the
maximum potential denitrification effi-
ciency at recirculation-to-plant influent
flow ratios of 1 to 3 is reportedly only
about 75 percent, this method has the
advantage of not requiring a supplemental
carbon source.
To keep the denitrification stage
anoxic, the RBC media must be complete-
ly submerged inthewastewater. Rotation
is achieved with mechanical drives at an
angular velocity of approximately 1.6
rpm. At this time, Autotrol (now Envirex)
is the only manufacturer marketing
RBC's for denitrification.
Autotrol's design procedure is based on
pilot studies indicating that in the
presence of an adequate carbon source
the denitrification rate in RBC's is
independent of bulk liquid nitrate nitrogen
concentration down to 1 mg/L, with a
zero-order removal rate at 55°F of
approximately 0.9 Ib NOa-N/day/IOOO
sq ft. Other pilot plant results, however,
have exhibited a first-order relationship
between nitrate removal rate and bulk
liquid nitrate nitrogen concentration in
the 0- to 6-mg/L range.
Power Consumption
In mechanically driven RBC systems,
power is used to overcome internal
resistances and losses in the motors,
friction losses in gear reduction and drive
belts or chains, friction losses in shaft
support bearings, and the drag forces
resulting from rotation of the media and
attached biof ilm through the wastewater.
The drag forces, which represent the
largest power drain, are affected by the
amount and shape of the media .surface,
rotational speed, wastewater viscosity,
and the type and amount of biological
growth.
In air driven systems, compressed air
discharged beneath the RBC media rises
and is captured by air cups mounted to
the periphery of the media. The resulting
buoyant forces provide the torque neces-
sary for media rotation. Power is required
to overcome (1) losses in the compressor,
compressor motor, air headers, control
valves, and diffusers and (2) the static
head of wastewater in the RBC tank.
Power factor is an important parameter
in determining the energy cost associated
with operating an induction motor. Power
factor is defined as actual power divided
by apprarent power. Polyphase watt
meters measure actual power (kW)
drawn, but demand charges are based on
apparent power (kVA). Most electric
utilities have demand charge schedules
that penalize customers with power
factors of less than 0.9. The use of
capacitors to increase the power factors
of induction motors can significantly
reduce power costs in RBC plants,
particularly plants that have large
numbers of mechanically driven units.
The Upper Mill Creek treatment plant in
Butler County, Ohio, uses mechanically
driven units equipped with power factor
correction capacitors. Measurements
made on seven units indicated that the
capacitors were increasing power factor
from an average of about 0.5 to 0.99. The
resulting 2.5-kVA savings in apparent
power is worth $17.30/month/shaft in
lower electricity costs at an assumed
demand charge of $6.92/month kVA.
Based on an approximate installed
capacitor cost of $200, the payout period
in this case would be about 12 months.
Field measurements of the power
required for mechanically driven RBC
units are reported for 16 facilities (92
shafts) in the full design information
dopument. The media surface area for the
shafts monitored varied from 100,000 to ™
180,000 sq ft. The measured mean power
requirement was 2.98 hp/shaft, with a
standard deviation of 0.71 hp/shaft and
recorded high and low readings of 5.10
and 1.62 hp/shaft, respectively.
Of the 92 mechanical drive units
monitored, a total of 55 were equipped
with media having a surface area of
100,000 to 128,250 sq ft. Media having a
surface area in this range are generally
regarded in the industry as representing
standard density media. The media
surface area of the other 37 units varied
from 138,000 to 180,000 sq ft, a range
characteristic of high density media. The
average measured power consumption
for the standard density units was 2.80
hp/shaft with a standard deviation of
0.62 hp/shaft. For the high density units,
the average power requirement recorded
was 3.22 hp/shaft with a standard
deviation of 0.79 hp/shaft.
An earlier EPA survey indicated that
manufacturers' estimated power require-
ments for mechanically driven RBC's
range from 2.7 to 3.4 hp/shaft for
standard density media and from 3.5 to
4.2 hp/shaft for high density media. The
above field-measured values agree well
with the manufacturers' esitmates for ^
standard density media and are slightly m
lower than the estimates for high density ~
media.
When field-measured, mechanical
drive power levels exceed the means
indicated above for standard and high
density media by one to two standard
deviations ot more, the operator should
investigate whether the higher power
consumption is being caused by equip-
ment problems, heavier-than-normal
biofilm growth, or both. Potential equip-
ment problem areas include improper
alignment, inadequate lubrication, ex-
cessive rotational speed, excessive belt
tension or belt slippage, and general wear
and deterioration of the drive components.
Power consumption in air driven RBC
systems is affected by motor and blower
characteristics and line losses as well as
rotational speed and biofilm growth. The
type of blower and its proximity to the
RBC trains can significantly affect overall
power requirements. Information on
rotational speed versus air flow relation-
ships is thus of greater value to the design
engineer than are indiscriminate, overall
power measurements.
Because the type of information dis-
cussed above is essentially unavailable to
the design community at large, studies
were conducted at two air driven installa- _
tions (Lower East Fork and Indian Creek) fl
in the Cincinnati, Ohio, area. In general, ^
-------
Autotrol's design curves underestimated
the actual air requirements measured at
the two plants. Load cell readings at
Lower East Fork did not indicate that the
biofilm being carried on those units was
excessive. Though data from two plants
are not sufficient to describe the rotational
responses that may prevail at other RBC
air driven installations, they do indicate
that the recommended design relation-
ships are not always applicable.
General Plant Design
Considerations
In general, housing RBC units within a
building (as opposed to using individual
covers for the RBC units) is undesirable
because of the high humidities and the
corrosive atmospheric conditions asso-
ciated with H2S release. Condensation
problems have been encountered on
interior building walls in cold climates,
and the associated high humidities and
ventilation requirements increase heating
costs. If buildings are chosen to house an
RBC unit, the building design must
provide for removal of a shaft/media
assembly should repair or replacement
prove necessary. In contrast to individual
fiber glass covers, full building cover
normally provides more convenient
access to RBC's for routine maintenance
and visual observation.
In all RBC designs, access to individual
shafts for repair or possible removal mus t
be considered. Bearings should also be
accessible for easy removal and replace-
ment and should also be accessible for
easy removal and replacement and
should be equipped with oversized grease
cups to minimize manpower requirements
for lubrication.
Whenever multiple-process trains are
employed, provision for positive and
measurable flow control to individual
trains is essential. Use of single, long
influent channels with slide gate control
for individual trains makes it difficult for
the operator to locate flow maldistributions
and implement appropriate corrective
procedures. Splitter boxes and/or weirs
are low-cost solutions to this problem.
Adequate flow control equipment is
especially important if individual trains
are fed from a single channel, with some
trains rotating with and other trains
against the direction of plant flow.
Feed and discharge flexibility should be
considered in RBC design. Step feed
capability can relieve overloaded first
stage(s) and potentially decrease or
eliminate excessively thick biofilm growth.
Removable baffles may also be effective
in this regard. In underloaded plants, the
final stages frequently must be operated
to keep suspended solids in suspension
rather than to provide additional treat-
ment to meet effluent standards. Thus,
the capability for temporarily bypassing
these terminal units would yield savings
in energy and operation and maintenance.
Load cells can provide useful operating
and shaft load data, especially in the first
stages. Where parallel trains are in
operation, they can pinpoint overloaded
or underloaded trains. Stop motion
detectors, rpm indicators, and clamp-on
ammeters are also potentially useful
monitoring instruments.
The use of deep channels leading to
and exiting from RBC tanks has resulted
in solids deposition and subsequent ac-
cumulation at a number of installations.
Providing for channel aeration (3.5
scfm/linear ft) or using channel configura-
tions that promote adequate scouring
velocity should overcome this problem.
Equipment warranties can be negoti-
ated with the manufacturers, and in some
cases, extended equipment warranties
have been obtained. This possibility
should be thoroughly considered in
equipment specifications. Although RBC
manufacturers continue to make improve-
ments in their equipment, major equip-
ment problems and failures have occurred
at some installations.
How sidestreams from other unit
proceses affect RBC performance must
be considered. Anaerobic digesters
increase ammonia nitrogen loadings, and
sludge conditioning processes such as
heat treatment contribute increased
organic and ammonia nitrogen loadings.
Whenever septic tank discharges are part
of the influent wastewater or when any
unit processes are employed that may
produce sulfide ahead of the RBC units,
the additional oxygen demand associated
with sulfide must be considered in
system design. Prechlorination or pre-
aeration may represent potentially cost
effective means of eliminating the oxygen
demand posed by incoming sulfide. High
influent grease loads require the use of
primary clarifiers instead of screens.
Nitrification is slow to develop in cold
temperatures, and 8 to 10 weeks may be
required before an RBC nitrification
system approaches equilibrium condi-
tions. Where seasonal standards for
nitrification are required, the final design
must consider the transition time and
temperatures needed to develop an ade-
quate nitrifying population.
Most existing air driven installations do
not have provisions for measuring and
controlling air flow to individual RBC
units. Furthermore, some plants cannot
easily verify whether some of the air
driven diffusers have plugged. Operating
an air driven facility under such blind
conditions makes it more difficult to
respond appropriately to operational
problems that may arise. The question of
how much plant blower capacity should
be provided is open to debate. A critical
need exists for more information on the
normal range of air flows that should be
expected with air driven units.
The EPA authors Richard C. Brenner. James A. Heidman, Edward J. Opatken.
and Albert C. Petrasek. Jr.. are with the Municipal Environmental Research
Laboratory, Cincinnati. OH 45268.
The complete report, entitled "Design Information on Rotating Biological
Contactors," (Order No. PB 84-199 561; Cost: $ 17.50, subject to change) will
be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA authors can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
*USGPO: 1984-759-102-10622
-------
United States Center for Environmental Research BULK RATE
Environmental Protection Information POSTAGE & FEES PAID
Agency Cincinnati OH 45268
PERMIT No. G-35
Official Business
Penalty for Private Use $300
l_OU w IlLLtY
REGiUN V EPA
_t
230 a OtAHdCKN SI
CH1CAGU 1L 60604
------- |