-------
I
ro ^
rH ^J*
^^ i^^
r* —•
r*1
in
rH
in H
«-• nj 4J
CO
H
a
o
M
o
CQ
•
-P *
~ -H
^ c
M O
03 -P
rH G
O <
O <
-P
O 'd
oo
(U
J3
CD
cr>
JJ
-P
O
>i P rH
M rH \
a)
•^5g
o
•> m
^Iv
U -P G
PQ tn o
OS >i-H
W -P
(^ (T>
•P 51
G G -P
H Q) G
g s
G
O
U
0)
X!
CQ !H
G EH
9 , S
•H ^ O
CD M
0
-P
nj -P
rH ftj
CD 5
« 0> X!
•P 3
S to rH
I (0 O
fe= & CO
CO
(U
•H
Pn
vo
in
en
899
-------
influent wastewater. The calculated ft against the stage number
N was then plotted in Figures 2 and 3.
The results as shown in both Figures 2 and 3 indicate that
the 1F decreases with increasing the number of RBC stages employed.
Based on the data presented in Figures 2 and 3, the 11-N relation-
ship can be described by two first-order equations as follows:
and
11 = 6.47 Exp.
11 = 6.90 Exp.
-0.140N
-0.313N
(13)
(14)
Substituting Eqs,11 and 13 into Eq.7 or Eqs.12 and 14 into Eq.7, the
RBC nitrification models are obtained. The descriptive equations
are:
0.1692
F =
6.47
aw 0.140N _ 0.2395 _ 0.6229
exp .LI i
(15)
F =
6.90x10
-4 q°-6803 L
2.6452
exp.
0.313N 0.2170
(16)
MODEL COMPARISON AND VERIFICATION
Weng et al have earlier proposed two mathematical models
similar to Eqs.10/15 and 16 (8 , 12) . Their models for the
prediction of BOD and nitrogen removals,in RBC systems, respec-
tively/ are:
_ _ 2.4465 LQ
r — ——~————
0.2815 0.4533
Q
0.3988
0.7004 _ 1.2423
-L
(17)
F =
0.0545 Lf
0.6440 0.1440
Q
0.5300
172760
(.18)
The main differences among Eqs., 10, 15, 16, 17, and 18 are:
(a) the number of RBC stages and the wastewater temperature are;
not included in Weng's model but the operating parameter of disk
rotational speed is considered, (b) the flow and disk surface
area are synonomous with the hydraulic loading in Eqs. 10, 15
and 16 as q=Q/A, and (c) the partial regression coefficient
values for the Weng's model and the models obtained from the
900
-------
present study are different.
The stage number and the wastewater temperature are the
significant variables which should be considered in the modeling
of RBC systems. All models proposed by Weng et al did not in-
clude temperature variable because they were developed from data
at a constant temperature and therefore regression analysis re-
ported it as not significant. Additionally, exclusion of stage
number as an independent variable will not allow usage of the
model for comprehensive design purposes. The rotating disk
speed is included as a model parameter in Eqs.15 and 16 and
although it is important for RBC operation it varies only
slightly between 1-4 rpm, as it is presently designed into the
process of the manufacturer.
The experimental data reported by Hao and Hendrick (21),
Sack, et al (22) , and Antonio ( 6 ) were taken to test Egs.10 and
17, and the results are presented in Figure 4. It is apparent
from Figure 4 that the BOD model developed from this study
yields better agreement than that resulted from the Weng's model.
In addition, data from the studies of Wu (23), Wilcox (18),
and Marsh (24) were used to examine the nitrification models as
proposed in Eqs.15, 16, and 18. The above studies were conducted
by using the full-scale RBC plants having a disk size > 10 ft.
The verifications of Egs.15 and 16 are illustrated in Figures 5
and 6 respectively. However, the results obtained from the
Weng's nitrification model were not shown in Figures 5 and 6 be-
cause the predicted values were greatly different from the
observed data.
CONCLUSIONS
The results of this study show that:
If the survival and growth of sludge organisms in the
rotating biological contactor wastewater treatment systems
is independent of pH and alkalinity, the efficiencies of
BOD and ammonia nitrogen removal can be easily predicted
by simply knowing of stage numbers, hydraulic loading,
influent substrate concentration, and wastewater tempera-
ture. Both BOD and nitrification models presently
developed yield good agreement between calculated results
and observed data.
According to the present study, the BOD model is given
as:
F =
14.2 q
0.5579
exp.
°'32N
°'6837
901
-------
o
o
20
30
40
Tesst Number
ffl
0)
en
-u
0)
M-t
a
H
Figure 4, Field Observations and Model Predcitions
From Eqs. 10 and 17
902
-------
Wilcox et al
L5 ,30 .
Present Model from
Eg, 15
O Observed Data
11 13 0.5 17 19 21 23 25 27
Test Number
Figure 5. Field Observations and Model Predictions
From Eq. 15 903
-------
20
rH
i
O
O
:z
<*
tc
•P
e
0)
H
15
10
_ 30
L OL
O
§
O
23
ro
•p
c
40
30
• Observed Data
O Present Model Prom Eg.16
1 3 5 7 9 JJ. 13 J.5 17 19 23. 23 25 27 29
Test Number
Figure 6. Field Observations and Model Predictions
From Eq. 16
904
-------
When the RBC plant is operated to serve as a secondary treat-
ment system, the nitrification model is
F =
6.47 q
0.1692
exp
0.14N 0.6229 0.2395
In addition, the nitrification model for RBC tertiary waste
water treatment system is
F =
6.90 x IP"4 q°-6803 Lo2'6452
eXP.°-313NT°-2170
The significant aspect of this study is that it demon-
strates the feasibility of using the RBC to control the
removal of carbonaceous BOD and ammonia nitrogen from
municipal wastewater and to optimize the utilization of
disk surface area under a wide range of operating conditions.
These models presently obtained are relatively simple and
can be used by the environmental engineer to design RBC
systems without knowledge of kinetic information.
905
-------
References
1. Eckenfelder, W.W.Jr., "Trickling Filter Design and Performance."
Transactions American Society of Civil Engineers, 128, 1963
2. Lagnese, J.F., "Upgrading With Rotating Bio-discs." Journal
Water Pollution Control Federation of Pennsylvania, vol. 9,
6, 1976
3. Antonie, R.L., "Fixed Biological. Surfaces - Wastewater Treat-
ment." CRC Press, Cleveland, Ohio, 1976
4. Smith, E.D., et al., "Tertiary Treatment of Wastewater Using
A Rotating Biological Disk System," Construction Engineering
Research Laboratory, Department of the Army, June, 1979
5. Borchardt, J.A., et al., "Nitrification of Secondary Municipal
Waste Effluent by Rotating Bio-disks,-" Department of Civil
Engineering, School of Engineering, The University of
Michigan, June, 1978'
6. Antonie, R.L., "Evaluation of A Rotating Disk Wastewater
Treatment Plant." Journal Water Pollution Control Federation,
vol. 46, 498, 1974
7. Clark, J.H., et al., "Performance of A Rotating Biological
Contactor Under Varying Wastewater Flow." Journal Water
Pollution Control Federation, vol. 50, 896, 1978
8. Weng, C., "Biological Fixed Film Rotating Disk for Wastewater
Treatment.." Ph.D. Thesis, School of Engineering and Science,
New York University, 1972 ;
9. Scheible, O.K., "Upgrading Primary Tanks With Rotating
Biological Contactors." Project No. EDGB'-Ol-OO, Clinton
Borgert Associates, Edgewood, New Jersey, November, 1978
10. Famularo, J., et al., "Application of Mass Transport To
Rotating Biological Contactors." Journal Water Pollution
Control Federation, vol. 50, 653, 1978
11. Joost, R.H., "Systematicon In Using The Rotating Biological
Surface Wastewater Treatment Process." Proceedings of the
24th Annual Purdue Industrial Waste Conference, Purdue
University, 365, 1969
12. Weng, C., and Molof, A.H., "Nitrification In The Biological
Fixed-Film Rotating Disc System." Journal Water Pollution
Control Federation, vol. 46, 1674, 1974
13. Antonie, R.L., and Koehler, F.J., "Application of Rotating
Disc Process to Municipal Wastewater Treatment." EPA Project
No. 17050 DAM, Autotrol Corporation, Milwaukee, Wisconsin,
November, 1971
906
-------
14. Torpey, W.N., Heukelekian, H., and Eaplovsky, A.J., "Rotating
Biological Disc Wastewater Treatment Process." EPA Project
No. 17010 EBM, Department of the Interior, Washington, D.C.,
1970
15. Malhortra, S.K., and Williams, T.C., "Performance of A Bio-
disk Plant In A Northern Michigan Community," The 48th Annual
Conference of the Water Pollution Control Federation, Miami
Beach, October, 1975
16. Borchardt, J.A., "Biological Waste Treatment Using Rotating
Discs." Biotech. & Bioengineering Symp., John Wiley & Sons,
Inc., New York, No. 2, 131, 1971
17. Wild, H.E., Sawyer, C.N., and McMahon, T.C., "Factors Affecting
Nitrification Kinetics." Journal Water Pollution Control
Federation, vol. 43, 1845, 1971
18. Wilcox, A.T., "Advanced Wastewater Pilot Plant Treatment
Studies." Reid, Quebe, Allison, Wilcox & Associates, January,
1975
19. Kincannon, P.F., et al., "One Step Nitrification and Carbon
Removal." Water and Sewage Works, vol. 122, 66, 1975
20. Saunders, F.M., et al., "Effect of SRT on Nitrification in
RBC Systemsr" presented at 1979 National Conference on
Environmental Engineering, American Society of Civil
Engineers, San Francisco, California, 1979
21. Hao, O., and Hendrick, G.F., "Rotating Biological Reactors
Removal Nutrients." Water & Sewage Works, vol. 122, 70, 1975
22. Sack, W.A., et al., "Evaluation of the Bio-disc Treatment
Process for Summer Camp Application," EPA Project
#67012-73-022, West Virginia University, Morgantown, WV, 1973
23. Wu, Y.C., "Evaluation of RBC Plant Performance at Cranberry
Township Water Pollution Control Plant," Department of Civil
Engineering, The University of Pittsburgh, Pittsburgh, PA,
1979
24. Marsh, D., et al., "The Coupled Trickling Filter-Rotating
Biological Contactor Nitrification Process: Design
Consideration," presented at the 52nd Water Pollution Control
Federation Conference, Houston, Texas, 1979
25. Pretorias, W. A., "Some Operational Characteristics of A Bio-
logical Disc Unit." Water Research, vol. 5, 1141, 1971
907
-------
Notations
The following symbols, are used in this paper:
2
A = total effective disk surface area,ft.
B = disk rotational speed, rpm
D = submerged disk depth,%
DF = degree of freedom
F = fraction of influent soluble BOD or ammonia nitrogen
remaining in the effluent
K = treatability constant of waste material
L - influent concentration of soluble BOD or ammonia
nitrogen, mg/1
N = number of stages
Q = flow rate,gpd
2
q = hydraulic loading, gpd/ft"
R = physical configuration constant, depends upon
disk diameter, spacing thickness, submergence, etc.
R_ = multiple correlation coefficient
c
S = reactor residence time, hours
T = temperature, C
£ » (K/N)x Rd
IT = Log £ or Log (K/N) Rd
a,b,c,d,e,f,g,h,and r
= partial regression coefficients
908
-------
RBC DESIGN CONSIDERATIONS FOR INDUSTRIAL WASTEWATERS
A.A. Friedman
Associate Professor
Department of Civil Engineering
Syracuse University
Syracuse, New York 13210
INTRODUCTION
The decade of the 1970's can be described as the "decade of change" with
respect to industrial wastewater management. Major changes in environmental
goals were reflected by the introduction and current common usage in our work-
ing vocabulary of terms such as ammonia nitrogen, nitrate nitrogen, BODpo*
phosphorus, and a host of priority pollutants as well as what once would have
been considered unbelievably low requirements for the more familiar BODr and
suspended solids parameters for describing effluent design criteria. As we
leave the decade, the concept of COD as an "indicator pollutant" has recently
been proposed as a regulatory parameter (1). If enacted, this new regulatory
concept may have more impact on some industrial wastewater treatment systems
during the new decade than the impact of the new regulatory requirements en-
acted during the previous decade.
These new regulatory concepts have been accompanied by significant changes
in both cost philosophy and treatment process understanding. During this decade,
our approach to industrial waste treatment has changed from "It can't be done
(because of the cost)!" to a philosophy of "Whatever it costs, it will be done!"
Research during the decade has provided new information and concepts that in
turn have yielded better understanding about the abilities and limitations of
treatment processes.
As a result of innovative thinking and research, the decade of the 1970's
saw the development and introduction of several new processes for wastewater
treatment. The development and widespread implementation of rotating biological
909
-------
contactors probably represents the most significant new treatment process in-
troduced during the decade. However, several RBC failures have created con-
cern about process reliability among engineers trying to meet the challenge
of designing cost and energy effective industrial wastewater treatment sys-
tems. The primary purpose of this paper is to describe the major causes of
process failure so that engineers can take full advantage of the cost effec-
tive potential of RBC treatment without the fear of process failure.
In the early 1970's RBC technology was a developing art with only a few
experimental units available to provide design information for domestic waste-
water treatment situations. Industrial utilization of this innovative and
energy saving process was based on trial and terror practices which resulted
in a few notable failures along the way. In the early part of the decade,
sales personnel claimed that RBC's could treat virtually any biodegradeable
wastewater, under any conditions, to any desired degree of treatment. All
that was required was sufficient media! Claims were also made that the
process (1) had a very high resistance to toxic slugs and shock loads; (2)
could be built at lower costs than alternative treatment systems; (3) could
be operated with substantial power savings compared to alternative treatment
systems; (4-) produced minimal solids yield; and (5) required minimal operator
attention. For design engineers accustomed to hydraulic detention times of
six to twenty-four hours the two hour RBC hydraulic detention time seemed
unbelievable, although it proved to be true for many situations. Unfortunately,
RBC's have not proven to be quite the panacea promised, especially for in-
dustrial wastewaters. Even though they have been applied to or pilot-tested
with a large variety of industrial wastewaters with a great deal of success,
failures with some industrial wastes have occurred and will probably con-
tinue to occur in the future. RBC's have distinct fundamental limitations
that can lead to process failures for some wastewaters. However, the.se lim-
itations are rarely described in the literature. The primary purpose of this
paper is to categorize and describe some of these failure modes and to sug-
gest preventive evaluation procedures for the consideration of designers work-
ing with industrial wastewaters.
FAILURE MODES
RBC failure can be defined as any situation where .the process does not
meet effluent goals or does so in an objectionable manner. Situations such
as process inability to meet effluent BOD and/or ammonia standards, or the
production of solids that won't settle or cannot be readily separated from
the carrier stream, or the production of objectionable odors are examples of
process failure modes. Obviously, media separation, shaft, bearing and
mechanical drive train problems are also previous causes of process failure.
Although designers should be able to learn from previous RBC failures, there
is a paucity of available descriptive information for the following reasons:
1. Design engineers and consulting firms avoid describing their failures
for fear of degrading their public images.
2. Equipment manufacturers won't voluntarily describe previous mechani-
cal failures of their equipment or process failures for fear of in-
terference with future sales. Previous failures are indirectly al-
luded to when manufacturers introduce their "new improved whatchamacallit"
or "thingamajig variation" in their sales literature.
3. Facility owners are understandably unhappy when their wastewater
treatment system fails. Owners try to avoid having their public image
associated with or recognized in conjunction with failures. Frequently,
910
-------
based on the advice of their attorneys, owners refuse to discuss or
permit interested outside parties to visit and inspect these failures
for fear of influencing pending litigation. Smaller industries are
deeply concerned about their capital loss and their relationship
with regulatory agencies. Larger organizations, having more capital
resources, salvage what they can, bury their mistakes and go on to a
new treatment system.
Regardless of the reason for silence, it is human nature to openly de-
scribe successes and forget failures. Several owners and engineers contacted
by the author of this paper were reticent to discuss their process failures
because of "corporate policy." Hence, failure data are not generally avail-
able for inspection and examination. However, failure descriptions are passed
around by the design profession by word of mouth or by pirated copies of clan-
destine letters, memoranda and draft reports intended for "internal use."
There are three general classes of in-plant problems: (1) process,
(2) equipment and (3) operational failures. Process failures result from
inadequate or insufficient pilot testing and poor planning by the design en-
gineer. Equipment failures are less likely in the future, as the major RBC
manufacturers have modified their products to eliminate major equipment
design problems. Operational problems occur due to inadequate maintenance,
or more frequently when the loading conditions are changes without regard to
the effects of these changes on the wastewater treatment system. These three
failure modes may be interactive. For example, significantly increased or-
ganic loading can result in abnormally heavy biomass attachment to the RBC
discs which in turn may result in stressing equipment to levels far beyond
those anticipated during design. Effluent quality will also probably decrease
as a result of the increased loading. Thus, an operational change may bring
about what could be incorrectly described as equipment and process failures.
PROCESS DESIGN CONSIDERATIONS
The following examples of RBC design factors are each related to at least
one prototype RBC failure or potential-fa-i-lure. Industrial owners' represen-
tatives and design engineers, while willing to discuss their problems privately
and "off the record," requested that data and/or identifying wastewater char-
acteristics be held in complete confidentiality for the reasons previously cited.
However, bench and pilot plant RBC data from the open literature will be used
to illustrate the types of problems encountered. Most industrial RBC pro-
cess design problems can be attributed to (1) unique characteristics of the
industrial wastewater, (2) loading conditions, or (3) improper pilot plant
operating conditions and poor data interpretation.
Wastewater Characteristics
RBC's are primarily effective for removing soluble wastewater constituents.
Particulates entering the RBC unit will either pass through the unit without
treatment or become enmeshed in the biomass attached to the discs. Because
the hydraulic detention time in the RBC unit and its clarifier is so short,
small influent organic particles escaping capture in the treatment train can
exert high effluent oxygen demands when measured by standard BOD tests. Al-
ternatively, if enough influent particles are trapped on the discs, (1) struc-
tural problems due to the extra weight can cause mechanical failures; (2) par-
ticle bridging between adjacent discs will prevent the distribution of sub-
strate and more importantly the transfer of oxygen to the biomass, thus' resulting
911
-------
1n process failure; or (3) organic particles enmeshed within the biomass can
create localized oxygen demands that exceed the oxygen transfer capability of
the system and lead to anaerobic conditions. Thus, RBC systems to be designed
for wastewaters containing significant quantities of particles should always
be preceded'with effective particle separation facilities.
Wastewater constituents should carefully be analyzed for a wide range of
plant production schedules and conditions. While composite samples are useful
for evaluating average design conditions, a careful plant analysis is required
to account for peak organic concentrations and maximum flow conditions, along
with nutrient balance assessments for these peak conditions. Because RBC
systems are less forgiving than alternative suspended culture systems having
longer hydraulic detention times (2, 3, 4). equalization should be strongly
considered for systems subjected to frequent peak loading conditions in order
to prevent undesirable transients in effluent quality.
Industrial wastewaters should be very carefully scrutinized for constituents
that may form precipitates in the RBC units as a result of oxidation and/or
pH changes. Recent bench scale RBC studies with tannery beamhouse waste-
waters in the Sanitary Engineering laboratories at Syracuse University's
Department of Civil Engineering serve to illustrate this type of potential
problem. After several months of continuous operation, an inorganic precipi-
tate probably consisting of a calcium sulfate-calcium carbonate complex was
found throughout the biomass attached to the discs. Even though this one-
quarter to three-eighths inch thick layer did not appear to interfere with
soluble COD removal for these experiments, it is likely that the development
of similar biomass-precipitate combinations could lead to unanticipated
mechanical stresses, additional power requirements and possible bridging
conditions in prototype units designed solely on a basis of area! substrate
removal criteria. For situations of this type, either additional pre-treat-
ment or an alternative treatment system should be recommended.
The nature of the organics in the wastewater can have a severe impact
on the choice of treatment system to be employed. Readily biodegradeable
soluble organic materials are ideal substrates for the RBC process. However,
some industrial wastewaters present special problems that may not be recognized
until construction is complete and several months of operation have transpired.
Figure 1 presents steady state operating data obtained with an eight-stage,
bench scale RBC unit treating tannery beamhouse wastes following neutrali-
zation, sedimentation and anaerobic treatment. Each stage had about 4.4
square feet of disc area and the total unit had a surface area of 35.2 square
feet and a nominal hydraulic detention time of 17.7 hours. As with many high
strength industrial wastewaters, soluble COD is the parameter of choice due
to better precision and repeatability than that obtained with the standard
BODg test. Conventional interpretation of these data would suggest that
biological treatment was essentially complete by the end of the third stage
and that the overall removal rates in these three stages were about 5.6 and
3.3 pounds of COD and BODg per thousand square feet, respectively. The re-
actions appear to have been completed in about 6.6 hours and the overall re-
movals are about 83 and 99 percent for COD and BODr respectively. Excess
oxygen and nutrients along with near neutral pH conditions in the succeeding
stages tend to confirm the conclusion of a complete reacfj-ion. The effluent
soluble COD, about 260 mg/1, for this situation would be termed refractory and
ignored for design purposes. However, long term BOD tests, conducted with seed
microorganisms from the RBC effluent yield the data shown in Figure 2. As a '
result of this additional information, it becomes obvious that the biological
912
-------
reactions did not go to completion in the time available in the RBC reactor.
Thus, effluents discharged from a system designed on data similar to those
shown in Figure 1 could readily meet most effluent BODg criteria but would
also result in unexplained oxygen demands in receiving waters.
The most probable explanation for the apparent paradox presented by the
data for this industrial wastewater is that readily biodegradeable wastewater
constituents are removed in a few hours in the RBC unit but the complex
soluble organic materials (probably proteins) remaining in the carrier stream
require more time to be enzymatically broken down into biodegradeable units
than is available within the RBC reactor. Other industrial wastewaters exhibit
similar slow biological degradation characteristics and each wastewater should
be evaluated as described above prior to process selection and design. Where
short RBC hydraulic residence times are not adequate for complete treatment,
alternative treatment methods, including hybrid RBC systems, should be evaluated.
Loading Rates
Overall 8;BC performance with industrial wastewaters is dependent on hy-
draulic and organic application rates as well as the nature of the wastewater.
Process performance at the microbial level is usually controlled by oxygen and
substrate transport into and through the biofilm and/or the transport of
waste products away from reaction sites (5, 6). Organic loading rate in terms
of mass applied per unit area per unit time seems to be the most appropriate
design parameter for predicting effluent quality. Loading rates between about
one-half and three pounds of soluble BODg per 1000 square feet per day will
generally produce an effluent containing less than 20 mg/1 of soluble BODj-
along with readily settleable secondary solids. However, higher loading rates
result in decreasing effluent quality and adverse operating conditions. Sur-
prisingly, evidence is becoming available which indicates that very low load-
ing rates (less than one-half pound BODg/1000 ft2-day) do not necessarily
assure an outstanding effluent quality. This is probably due to lack of a
substrate concentration sufficient to drive organic compounds into the biomass
film. Conversely, at organic loading rates of about five pounds soluble
BODg per 1000 square feet per day, RBC reactors begin to behave as a combined
fixed film-suspended culture system. Under these loading conditions effluent
quality deteriorates in terms of both soluble BOD5 and settleable suspended
solids. Also, strong unpleasant odors are likely to be generated as rapidly
growing suspended microbes deplete the available oxygen in the bulk carrier
stream.
These adverse phenomena are illustrated by the following data obtained
with an eight stage, bench scale RBC unit treating a soluble readily biode-
gradeable wastewater under ideal laboratory conditions. Figures 3 and 4
present data obtained under steady state operating conditions with disc ro-
tational speed being the only operating parameter varied for this set of ex-
periments. Complete data sets from these experiments have been reported else-
where (5). Figure 3 indicates the soluble COD remaining in each stage and
Figure 4 shows the stage dissolved oxygen profiles for the same loading con-
ditions. Dispersed growth particles were found in each stage and in the ef-
fluents for all four experiments. The overall loading rate was about 5 Ibs.
of BOD,- (7.6 Ibs. of COD) per 1000 square feet per day. The units had in-
O
sufficient oxygenation capacity to meet the biological oxygen demand at the
lower two rotational speeds. This in turn encouraged biological sulfate re-
913
-------
duction and generation of strong hydrogen sulfide odors from the unit. However,
more than adequate dissolved oxygen was available for the higher two rotation-
al speeds and is reflected in the better soluble COD removal for these con-
ditions. Note that the first stage(s) is where the critical operating condition
occurs. These data suggest that bypass or step feeding, supplemental air, or
increased rotational speed provided by variable speed drive trains may be use-
ful during peak loading conditions. Similarly, the removal of baffles between
the first two (or more) stages for overloaded prototype plants may help to
alleviate anaerobic conditions by increasing the oxygenation capacity of the
first stage.
Pilot Plant Recommendations
Each industrial wastewater stream is unique. The experiences and infor-
mation obtained from a plant at one location producing a particular product
line, while providing useful guidelines, rarely can be transferred to another
plant producing a similar product. Well designed RBC pilot plant programs can
help engineers to design systems that will provide years of low cost, trouble
free operation. Conversely, a poor or incomplete pre-design evaluation program
can lead to years of continuous operational problems and frustrations. Good
pilot testing programs are not inexpensive, but they are much less costly than
system failure and subsequent remedial actions.
Both bench scale and prototype pilot plant testing are complementary and
have preferred applications. Bench scale laboratory studies can be used to
define the upper limits of treatability for a specific industrial wastewater
stream. Critical conditions involving factors such as precipitate formation,
toxic conditions, temperature effects, or the effects of mixing or segregating
various in-plant wastestreams can be effectively evaluated with bench scale
units under controlled conditions. Unfortunately^ experimental lead time is
usually critical in the design process and preliminary studies are frequently
rushed to completion. The acclimation and replacement of biomass may take up
to several months to develop, mature and produce meaningful results following
changes in loading parameters. This is especially important where the effects
of changed influent loading conditions can be anticipated to affect and inter-
fere with nitrification in latter stages. Complete stage by stage data sets
including dissolved oxygen, pH, alkalinity, nitrogen, solids and the concen-
trations of compounds of specific interest, in addition to stage BOD or COD
data are highly desirable for defining the reactions and interactions taking
place in the RBC unit. Daily scraping is necessary to prevent excessive wall
growth from interfering with data analyses and interpretations. Effluent, data
should include an evaluation of solids settling characteristics as well as the
parameters previously described.
Where a wide variety of process variables are to be evaluated over an ex-
tended range of values, multiple bench scale experimental units should be em-
ployed. While these recommendations are costly to implement compared to
standard quick input-output experiments, they should prevent downstream process
"surprises" that might otherwise be attributed to undefined "changes in the
wastewater characteristics between pilot studies and construction."
Unfortunately, bench scale pilot plants cannot duplicate the hydraulic and
oxygen transfer characteristics of full size RBC units (4, 5). Portable full
size RBC units can be rented from manufacturers for on-site evaluation. Gen-
erally area! removal rates are lower in prototype RBC units than in bench
scale units because the average mixing intensity, surface shear and exposed
liquid film thicknesses are different. At the present time, suitable scale-
up equations based on fundamental principles are not available to relate bench
914
-------
scale and prototype RBC performance. While such expressions may be developed
in the future, only a carefully designed testing program that includes con-
sideration of the elements described above can help prevent unpleasant
"surprises" following construction.
CONCLUSIONS
The RBC process has many applications for the treatment of wastewaters.
Rapidly escalating power costs make the process even more promising for the
long term, cost effective treatment of industrial wastewaters. However, be-
cause of unique wastewater characteristics, gross traditional parameters such
as BODg and suspended solids alone are inadequate for assessing RBC treatment
for many industrial wastewaters. Only complete chemical characterization
studies and careful pilot plant evaluation programs can prevent subsequent
process failures and allow owners to obtain the full benefits potentially
available from the RBC process.
REFERENC-ES
1. Leather Tanning and .Finishing Point Source Category Effluent Limitations
Guidelines, Pretreatment Standards and New Source Performance, Federal
Register, July 2, 1979, p. 38746.
2. Filion, M.P., K.L. Murphy and J.P. Stephenson,"Performance of a Rotating
Biological Contactor Under Transient Loading Conditions," J. Water Poll.
Control Fed., 51, 7, 1925 (1979).
3. Wilkey, R.C. and A.A. Friedman, "Response of Rotating Biological Contactors
To Shock Loadings," Proceedings Fifth Annual Envir. Engr. and Sci. Conf.,
Univ. of Lousiville, Louisville, KY, March 1975.
4. R.W. Wilson, K.L. Murphy and J.P. Stepheson, "Effect of Scale-Up in
Establishing Design Loadings for Rotating Biological Contactors," Presented
At 51st Water Poll. Control Fed. Conf., Anaheim, CA, Oct. 1978.
5. Friedman, A.A., L.E. Robbins and R.C. Woods, "Effect of Disk Rotational
Speed on Biological Contactor Efficiency," J. Water Poll. Control Fed.,
51, 11, 2678 (1979).
6. Famularo, J., J.A. Muller and T. Mulligan, "Application of Mass Transfer
To Rotating Biological Contactors," J. Water Poll. Control Fed., 50, 4, 653 (1978),
915
-------
150O
81000
>o
§
UJ
m 500
I I I I
C,N «l550mg/t COD,720mg/-t BOD5
C^ 280 mg/-t COD, 9mg/( BODg
TOTAL AREA = 35.2 sq.ft.
Q = 26.6 f/day
flHYD= 17.7 hours
-O—
.—A— _
IN I 2 3
FIGURE I BENCH SCALE RBC DATA
4
STAGE
UJ
e °
6 >-
X
o
a
>» ^
4>
o
8
916
-------
1250
1000
8 750
CO
UJ
500
250
R8C EFFLUENT
COD»260mg/-t-
RBC INFLUENT
COD* 1550 mg/t
J_
_L
10
20
25
15
DAYS
FIGURE 2 INFLUENT AND EFFLUENT BOD AND COD
COMPARISONS
25
20 f
15
Q
O
CO
LJ
CD
CO
I
5t
UJ
30
917
-------
250
8
I 334567
STAGE NUMBER
FIGURE 3 BENCH SCALE RBC TREATMENT FOR A CARBONACEOUS WASTE-
WATER
918
-------
80-
s
\ \
o 5.7 rpm
O 12.3 rpm
A 17.7 rpm
a 29.5 rpm
i 1 I
CQ at 228 mg/t
Q a 4.3 gpdsf
*HYDa 1-2 hours
i 1 1
IN
8
34567
STAGE NUMBER
FIGURE 4 DISSOLVED OXYGEN PROFILES FOR A CARBONACEOUS WASTE-
WATER
919
-------
-------
ANALYSIS AND DESIGN OF ROTATING BIOLOGICAL CONTACTORS
By
James A. Mueller
Associate Professor of Environmental Engineering and Science
Manhattan College, Bronx, N.Y., USA
Jack Famularo
Associate Professor - Chemical Engineering Program
Manhattan College, Bronx, N.Y., USA
James Fitzpatrick
Research Engineer, Hydroscience, Inc.
411 Old Hook Rd., Emerson, N.J. 07630
INTRODUCTION
The basic mechanisms controlling the performance of rotating biological
contactors in removing soluble substrate are twofold: (1) the transfer of
substrate and electron acceptor (such as oxygen or nitrate) to the biomass
attached to the rotating surfaces, and (2) the rate of biological growth
and/or reaction occurring in the biomass. To properly understand the func-
tioning of the RBC process, to confidently scale up pilot or bench scale data
to plant design, and to get a first cut evaluation of process modifications
on system performance, it is desirable to combine the mass transfer and
kinetic mechanisms with the system hydraulics.
123
Our research efforts ' 'J over the last three years have incorporated the
above first principles modeling approach to analyze and design RBC systems.
This paper will summarize the results of this effort.
The approach taken is to summarize the model characteristics for both
carbon oxidation and nitrogen removal with sufficient validation presented.
The major thrusts of the paper are to quantify the importance of the various
removal mechanisms acting in the system, indicate a design approach using the
921
-------
model for cabon oxidation, and evaluate the effect of wastewater characteris-
tics and process modifications on system performance.
MODEL DEVELOPMENT
The RBC model is applicable to soluble organic carbon (BOD or COD) re-
moval, nitrification, and soluble nitrogen removal. It consists of material
balance equations that are numerically integrated to obtain concentrations of
the carbonaceous and nitrogenous substrates in the liquid film and biofilm
attached to the rotating media as well as concentrations in the mixed tank
and effluent liquid from each stage. For the mass transfer process, the
elements included in this model are oxygen and substrate transfer through
liquid and biofilms, ammonia stripping, and subsurface aeration for an air
drive system. All mass transfer coefficients in the model are related to
system operating conditions. Reaction rates are also simultaneously related
to electron donor (substrate) and electron acceptor (oxygen or nitrate) con-
centrations requiring no prior assumptions as to which component is control-
ling system performance.
For the model framework, the media is assumed to consist of flat plates
divided into four stationary pie-shaped sectors as shown in Figure 1.
A stationary liquid film is utilized and gradients of all components
through the biofilm are obtained by dividing the biofilm into completely
mixed layers as shown in Figure 2. Since the reaction rate of the autotrophs
is slower than the heterotrophs, deeper penetration of components exist in
nitrification stages compared to those with carbon oxidation. Liquid film
and biofilm are transported through stationary sectors at volumetric flow
rates Q. and Qp respectively, both a function of rotational speed and layer
thickness.
Mass Balance Equations
For each RBC stage, 3 types of mass balance equations are required: (1)
tank equations, (2) biofilm equations and (3) liquid film equations. Refer-
ring to Figures 1 and 2, the non-steady state equation for the tank for each
of the components considered in the model is:
Tank:
T TO
QCS1 - S1U)
w
(ST - S1.) + Rc V + K.aV(S* - ST)
Jo L
(1)
All terms in the above equation are defined in the nomenclature section of
the paper.
922
-------
ROTATIQ^
LIQUID FILM
COATS DISC
Q, EFFLUENT
oT —
INFLUENT, Q,
STO
FIGURE I
SKETCH OF SECTORS IN THE RBC MODEL
923
-------
iz
<
UJ
cc
<
CC
UJ
t/3
MEDIA
s6
BIOFtLM
S
rt 11
LIQUID FILM
GAS PHASE
S1/
*
g7O gSO g5O g4O g3O g2O g| O
Qi
LEGEND:
PROFILE
AVERAGE CONCENTRATION
FIGURE 2
BIOF1LM SCHEMATIC DIAGRAM
924
-------
The mass balance equation for the tank not only includes the (1) advec-
tion terms and (2) mass transfer to the biofilm, which normally contribute
the major portion of the overall concentration change in the tank, but also
(3) reaction due to organisms (VSS) suspended in the tank mixed liquid as
well as (4) gas transfer in the tank when an air drive model is used. To
numerically integrate the above equation and solve for the tank concentra-
tion, S , requires knowledge of the concentration JLJJ the first biofilm layer,
S , as well as that existing in the liquid film, S , just before it reenters
the tank. For simplicity, no separate liquid film is assumed to exist in the
submerged sectors. Thus Kg is an overall transfer coefficient from the tank
contents to the first biofilm layer in the submerged sectors.
To obtain the biofilm concentration, S , in the first layer for each sec-
tor, a material balance on an infinitesimal slice of biofilm yields the fol-
lowing differential equation:
Biofilm:
32S
AS
- s> + Rs = I?
(2)
In the above equation, the diffusion term relates the concentration in
the layer being analyzed to its adjacent layers in the same sector. Trans-
port of mass by diffusion to adjacent sectors is assumed negligible compared
to the advective transport term. For the internal biofilm layers, the above
equation is used for each component. At the liquid-biofilm interface and the
media - (or dormant biofilm) biofilm interface, the following boundary condi-
tions apply:
Liquid-Biofilm interface below water line;
- s)
Liquid-Biofilm interface above water line:
-Ds f = Ks(sL - s)
Media-Biofilm interface:
-D0 -22. = 0
S ay
§ y = s
The first two boundary conditions couple the biofilm concentration with the
tank and liquid film concentrations. The last boundary condition results
since the transport into the media or the dormant biofilm layer is zero.
This effectively limits all reaction to the active biofilm thickness used in
the model.
To finish the equation set, a material balance equation for the liquid
film in each sector above the water line is required as follows:
925
-------
Liquid Film;
(SL° - SL) - ^(
Kr
S) + -(
T
- SL)
3S_
3t
(3)
The above equation contains mass transfer to the biofilm which applies
for all parameters and also mass transfer from the gas phase to the liquid
film which applies only for the gaseous components, oxygen and non-ionized
ammonia.
Reaction and Growth Kinetics
Biological reaction is based on the mixed liquor volatile suspended
solids concentrations existing in the tank, liquid film, and biofilm. Two
organism species are utilized: the heterotrophs using organic carbon as the
substrate and the autotrophs using ammonia as substrate.
In the RBC model, five reacting components are simultaneously considered:
BOD, 02, NO--N, NH*-N, and soluble organic nitrogen (SON). Figure 3 indi-
cates schematically the transport and reaction mechanisms incorporated in the
model. Interphase transport of both oxygen and ammonia between the gas phase
and the liquid film occurs while the disc is rotating above the liquid sur-
face. From the liquid film, the five diffusing species penetrate by molecu-
lar diffusion into the biofilm where biological reaction takes place.
Oxygen is used for organic carbon (BOD) oxidation, endogenous respira-
tion, and nitrification.
Soluble organic nitrogen is hydrolyzed by both autotrophic and heterotro-
phic species. Initial model verification using only heterotrophic species
for organic nitrogen hydrolysis proved inadequate. A review by Painter in-
dicates that hydrolysis by both species is possible although little data is
available for substantiation. Oxidation of ammonia nitrogen to nitrate by
the autotrophs provides energy for autotrophic growth. For model simplicity,
intermediate nitrite production is ignored. Although endogenous respiration
of autotrophs is provided for in the model framework, for simplicity and ease
of mathematical computation, an endogenous rate coefficient of zero was uti-
lized with a net cell yield value for autotrophic growth.
In the inner biofilm layers which have low oxygen concentrations, deni-
trification reactions occur both for BOD removal and endogenous respiration.
Except for organic nitrogen hydrolysis which is independent of other reac-
tions, the rates of reaction of all species are related stoichiometrically to
the rates of heterotrophic and autrotrophic growth. Table 1 presents the
substrate and electron acceptor rate equations for the five species of con-
cern. Note that all rate equations are written as production equations for
uniformity.
The growth rate equations used in Table 1 are defined in Table 2. The
net growth rates of both species are delineated with respect to growth and
endogenous respiration rates using oxygen or nitrate for ease of use in the
substrate and electron acceptor equations. Michaelis kinetics are used for
both electron donor and acceptor allowing either component to control. In
926
-------
FIGURE 3
NITRIFICATION REACTION SCHEMATIC FOR AN RBC SYSTEM
927
-------
TABLE 1
SUBSTRATE AND ELECTRON ACCEPTOR RATE EQUATIONS
1. BOD, S
1
dS1 -1
dt~ = R1 = Y CRX13+ RX143
where: Y, =
(frai>
a = Ao /A Soluble Substrate
f = 1.112 mg 02/mg X
2. Total Ammonia Nitrogen, S,
, _ A Soluble COD
1 A Soluble Substrate '
dS
dt
-RX,
f2 CRX1 + RX23 ~ R5
where: Y? = net
Autotrophic cell yield = 0.17 mg X2/mg NH4-N ; f2 = 0.124 mg NH4-N/mg X
3. Oxygen, S
dS3
dt~ = R3
~f3 CEX13 ~ RX2]
where: a2 = 4.57 mg 02/mg NH^-N oxidized
928
-------
TABLE 1 (Continued)
4. Nitrate N,
dt
Y
where: f. = 0.35 mg NO.,-N/mg Op
5. SON, S
dt
/5 S3
35+ 5m 3+ 3m
929
-------
TABLE 2
GROWTH RATE EQUATIONS
1. Net Heterotrophic Growth Rate, RX
1
S S
a) EX., = R. X. -5—=— ^—|— , Growth Rate Using 0
13 1 1 S1+S1m V 3m 2
S S
b) RX14 = P1 X1
1 1m 3 3m 4 4m
3-^3 — , Growth Rate Using NOo-N
c) EX..- = b. X- -g--^— , Endog. Resp. Rate Using
3 3m
d) EX... = b. X.
14 11
, Endog. Resp. Rate Using NOo-N
3
2. Net Autotrophic Growth Rate, RX,
S2 S3
RX2 = U2 X2 S— S~ > Gr°Wth
ing °
930
-------
the denitrification reactions, an additional oxygen Michaelis type term (Som/
S +S ) is included which has the effect of allowing denitrificatiorf to ocdur
when the oxygen concentration is reduced to values near or below the oxygen
Michaelis constant. Since nitrification is more significantly affected by
low oxygen concentrations, different oxygen Michaelis values are used for
substrate oxidation and nitrification.
To solve the above equations, the organism distribution between hetero-
trophs and autotrophs must be determined when nitrification is present. This
is accomplished by setting the ratio of the organism concentrations in each
stage equal to the ratio of -their growth rates in the biofilm for that stage.
Thus:
X ~ n
1 z RX1
1
where n = number of biofilm segments in all sectors of a stage.
In addition to the above reaction terms, the ionized ammonia fraction
must be estimated for use in the ammmonia stripping equation as follows:
S2N=fNS2 ^
PKa-pH
where f = 1/1 +10 = non-ionized fraction and pK = 10.05 - 0.032
t(°C) (Ifef. 5). a
Mass Transfer Coefficients
The mass transfer coefficients for oxygen and substrate as well as liquid
film thickness were obtained using the theory of withdrawal of a flat plate.
Levich analyzed the situation of the film substrate entrained on a flat
plate vertically withdrawn from a quiescent liquid and deduced that the film
thickness hQ could be represented by the following equation:
2/3
1/2
(6)
Assuming that the physical properties of the wastewater will not differ
appreciably from pure water at 20 c, equation (12) reduces to:
h0 = 6.85 v2/3
where v is the withdrawal velocity.
(7)
931
-------
Due to irregularities in the biofilm surface, the retained liquid film ^
will be thicker than that obtained on a flat plate. Williamson and McCarty
measured surface irregularities and found depths ranging between 16 and 30
microns. Based on these measurements it was decided to add 25 microns to h
to obtain the liquid film thickness &, . Since withdrawal velocity in a ro-
tating system is a function of radial position, an average withdrawal velo-
city at the centroid of mass (2/3 of media radius) was employed.
The mass transfer coefficients are related to <$L through the equations:
Ks =
V2
KL=
Do
V2
(8)
i i
where, D and Dg are diffusivities of oxygen and substrate in water, and S /2
is the diffusion path length from the average concentration to the interface
concentration.
In developing the oxygen transfer equation from air to the liquid film,
the saturation concentration for oxygen, S~, is assumed to remain constant
requiring the partial pressure in the gas phase to remain constant. If one
assumes that the air enters the internal media passages as it leaves the tank
liquid and remains there until discharged as the media reenters the tank, the
decrease in 02 partial pressure is negligible for any reasonable 0^ transfer.
rate. Thus tne assumption of a constant S-, is valid. This is due to the re-
latively low solubility of oxygen and high partial pressure (20.9%) of oxygen
in air.
However, when evaluating ammonia stripping the reverse is true, ammonia
being a highly soluble gas with a partial pressure of zero typically existing
in the atmosphere. If one utilizes the previous stagnant film techniques to
calculate K.? with the saturation value S2 = o, then the amount of ammonia
stripping occurring is greatly exaggerated, since the gas phase composition
would require continuous instantaneous replacement in the internal media
voids.
Since gas flow rate appears to be the major factor affecting ammonia
stripping, the following approach using a saturated gas flow was utilized. A
gas phase flow rate is calculated as a function of rotational speed similar
to the liquid film flow rate. The ammonia partial pressure in the gas phase
is assumed to be in equilibrium with the non-ionized ammonia concentration in
the liquid film of the last aerated sector prior to reentering the tank con-
tents . In equation 3, for the liquid film, the amount of ammmonia trans-
ferred from the gas phase becomes:
- fNS2) = -
KL2fN
RT AA\
(9)
Since p2 = fNS2/H from Henry's Law and the gas flow can be given by G
= RH A^u, then the ammonia transfer coefficient above the water line is:
932
-------
K.
L2
(10)
Subsurface Gas Transfer - Air Drive
The additional aeration occurring for the air drive system can be evalu-
ated using similarities to diffused aeration systems. For diffused aeration,
the oxygen transfer coefficient is related to gas flow and transfer effi-
ciency by the following: -
1.04 G
KLa20V = 100
's20
o [=] MG
hr
(11)
where G = gas flow, scfm, E = % absorgtion efficiency. The above equation
is related to standard conditions of 20 C. Temperature scaleup is typically
given by:
, _t(oC)-20
KLat = KLa20
typically
(12)
For ammonia stripping, the assumption of saturated gas flow is again made
which provides the maximum quantity of ammonia removal possible.
Numerical Evaluation - Method of Solution
Although the tank, biofilm, and liquid flow mass balance equations (1-3)
are presented as time variable or non-steady state equations, they were
solved as steady-state equations for the purposes of the model calibrations
presented herein. Under the assumption of steady-state, the mass-balance
equations reduce to a set of simultaneous non-linear algebraic equations,
which may be written as:
311S1
a12s2 + a13s3
a1nsn = b1
a21S1
a22S2
a23S3
a2nsn = b2
(13)
an1s1 + an2s2 + an3s3 + "• + annsn = bn
or in standard matrix notation:
[A](s) = (b)
where:
s. is the (unknown and to be solved for) concentration in compartment i
a., is comprised of the advective and mass transfer terms, and the reac-
tive and kinetic terms involving s.
933
-------
a*.j is comprised of the advective and mass transfer terms, and the reac-
tive and kinetic terms between compartments i and j, involving s..
b. is the effective driving force due to appropriate influent condi-
tions and boundary conditions,
where a compartment is defined to be a particular component (BOD, NHq> C^'
etc.) in a particular medium (tank, biofilm, liquid film), in a particular
stage of the rotating biological contactor.
Due to the non-linear nature of the mass balance equations and the number
of equations involved (61 cells x 5 components = 305 equations and unknowns/
stage), use of conventional matrix inversion to solve for the unknown is
ruled out. Instead, a modified form of Gauss-Seidel iteration is employed.
The Gauss-Seidel iteration method involves rewriting each of the equations in
(13) in terms of the unknown as follows: ,
S1 =
b1 ~ al2S2 ~
'11
a., s
1n n
b2 ~ 321S1 " a23S3 " ... 32nsn
a22
(14)
sn =
bn ~ as ~ as ~
n1s1 ~ an2s2 ~ ... an n-1sn-1
nn
Starting with an initial solution estimate for (s), the above equations are
iteratively solved until the residual differences between two successive
iterations is within a set tolerance. This promise may be shown as follows:
~ 312S2 ~ a13S3 " •" a1nsn
k+1
22
(15)
. a k+1 k+1
bn ~ an151 - an2S2
sk+1
* " A
n n-1 n-1
934
-------
100
o>
E
80r
60
-------
where:
lc
s = concentration of compartment n at interation k
k+1
s = concentration of compartment n at iteration k+1
It was found necessary to modify the Gauss-Seidel iteration method after
experiencing non-convergence of the system of equations. This non-conver-
gence occurred in compartments where the concentration of one of the varia-
bles of interest was at or below the Michaelis constant for that variable.
The non-convergence is to be expected when that occurs since this is the
region in which the equations are most non-linear. This problem was overcome
by taking the derivatives of the growth rate equation of Table 2 with respect
to the dependent variable and modifying the equations in (15) as follows:
k+1
b ~
al2S2
k
a13S3 ~
k
a1nsn
d(ds/dt)/ds.|
k+1 k+1
k+1 Dn ~ an1s1 ~ an2s2
k+1
3 .. S ,
n n-1 n-1
nn
d(ds /dt)/ds
n n
This modified Gauss-Seidel procedure was successfully convergent in all
cases.
MODEL CALIBRATION
.1.2
Numerous calibration runs1'" were conducted to evaluate model adequacy
using pilot and full scale data for both carbonaceous removal and nitrifica-
tion systems. To date, only data from mechanical drive systems have been
used for calibration purposes. The approach taken is to set Michaelis coef-
ficients in the range of typical literature values and utilize the data to
obtain the remaining kinetic and stoichiometric coefficients.
In the early carbonaceous studies, both low and high Michaelis coeffi-
cients were utilized to simulate zero and first order substrate removal kine-
tics. Stage dissolved oxygen data was used to obtain the substrate removal
coefficient, k = U
X../Y...
Figure 4 shows one of the calibration runs for a
papermill wastewater using zero order kinetics. First order kinetics pro-
vided a similar fit but with a wider range of kinetic coefficients when dif-
ferent wastewaters were analyzed.
936
-------
The RBC model developed in this study provides concentrations within the
biofilm which can be used, in conjunction with substrate and oxygen Michaelis
constants, to determine active biofilm depth. This is illustrated in the
biofilrn profiles shown in Figures 5 and 6. Concentrations employed to con^
struct these profiles were taken from the computer output of the 10 gpd/ft
(407 1/d/sq. m) verification run of Figure 4.
The kinetic equations used in the RBC model cause a reduction in BOD re-
moval rate when either of the ratios, S/CS+S ) for BOD or oxygen drops below
unity. For example, if S3«S~ in regions of the biofilm where S-,/(S.|+S.|m)
is approximately equal to one, reaction is limited by a deficiency of oxygen.
Examination of Figure 5 reveals that this is clearly the case in stage 1.
Substrate concentrations are in excess of 4.3 mg BOD/1 throughout the bio-
film. Since S. = 0.1 mg BOD/1, substrate will reduce removal rate at most
by 2.3 percent. On the other hand, oxygen concentrations dropped well below
S-,m = .001 mg/1 at biofilm depths in excess of 120 microns. For practical
purposes, the active biofilm depth of stage 1 is 120 microns, and substrate
removal in the stage is limited by a deficiency of oxygen. The reverse situ-
ation exists in stage 4. As can be seen in Figure 6, except for one or two
points in the biofilm, the oxygen concentration is much greater than its
Michaelis constant and has a negligible effect in reducing BOD removal rate.
In stage 4, a deficiency of substrate within the biofilm is the rate limiting
factor. The average value of S./CS^+S^) at the 120 micron depth was 0.06,
causing a 94 percent reduction in removal rate.
For the nitrification system, five coefficients must be evaluated from
the data:
estimate i
SON data.
n
11 <
NO
U
and
a*, and
, while i
.. Soluble BODj- and oxygen data were used to
was obtained from ammonia data and k,- from
_ data then served as a check on the system.
Figures 7 and 8 show calibration results for two data sets (on the
Gladstone, Michigan municipal wastewater) . In the runs, no attempt was made
to apply a consistent temperature correction factor to the various coeffi-
cients. Instead the best coefficients which fit the data were chosen and
then used to evaluate the temperature effect. Also in these runs, a yield
coefficient of 0.38 was used for the autotrophs based on thermodynamic con-
siderations. In the remaining runs a value of 0.17 was used for the net ce.
"
yield coefficient. In all figures the units for ^ ,
while the Michaelis constants are in units of mg/1.
H
2,
and
are day
A wide range of nitrogen concentrations occur in Figure 9 over the three
days analyzed possibly due to digester supernatant being returned to the pri-
mary clarifier during the major portion of this run. It was not returned for
the remaining runs. In some runs little BODj- removal was at times measured
in the first stage, requiring higher influent BOD,- values for the best fit of
all stage data.
The variation of the four kinetic parameters obtained from the above
calibrations as a function of temperature is shown in Figure 9. For all par-
ameters, a significant temperature effect is seen. Using the following equa-
tion to describe temperature efffects,
937
-------
64.2 mg BOD/I
BOD
BOD MICHAELIS CONSTANT =0.1 mg/l
30 60 90 120 150 180 210
BIOFILM DEPTH fmicrons) '
4.5 mg Oz/l
DO
OXYGEN MICHAELIS CONSTANT = .001 mg/l
30 60 90 120 150 180 210
BIOFILM DEPTH (microns)
FIGURE 5
BIOFILM BOD AND DO PROFILES. IN STAGE I OF PAPERMILL A
PILOT PLANT AT 10 GPD/FT2
938
-------
28.6 mg BOD/I
BOD
BOD MICHAELIS CONSTANT =0.1 mg/l
3O 60 90 120 150 ISO 210
-BIOFILM DEPTH (microns)
DO
2.2 mg 02/l OXYGEN MICHAELIS CONSTANT = .OOI mg/l
Z!
30 60 90 120 150 180 210
BIOFILM DEPTH (microns) '
FIGURE 6
BIOFILM BOD AND DO PROFILES IN STAGE 4 OF PAPER MILL A
PILOT PLANT AT 10 GPD/FT2
939
60
— 50
E 40
Q* 30
o
CO
. 20
_i
o
w 10
O
30
25
<; 20
o
E
— 15
O
« 10
5
0
12
"5, 10
2 8
UJ
o
I
CO
o
f
= 2.0
/X, = 3.0
S,m = 50
_L
234
STAGE
k 5 = 0.04
S5m=2-°
I
234
STAGE
a | =0.8
b| = 0.15
S^m=0.l
234
STAGE
30
25
20
l5
I
•y 10
z
5
0
Q = 1.51 gpd/ft2
TEMP. = 19 °C
FIGURE 7
NITRIFICATION MODEL CALIBRATION - DATA SET I
MUNICIPAL WASTEWATER TEMPERATURE = 19°C
940
-------
90
~ 75
e GO
t? 45
o
tn
. 3O
OT 15
0
30
25
- 20
c»
E
— 15
1 10
5
0
_ 12
^ 10
e
z 8
IU
X "
o
3 4
O V
to 2
CO
a
f, = 1.5
/i, = 1.0
S,m = 50
234
STAGE
ks = 0.02
234
STAGE
a, = 0.8
b, = 0.07
S3m=°-'
JL
I
_L
234
STAGE
O
2
30
25
2O
l5
10
0
30
25
20
15
10
5
0
_L
I
1
23456
STAGE
Y2 = 0.38
Hr _£L
234
STAGE
Q = 1.02 gpd/ft2
TEMP. = 8.7 °C
FIGURE 8
NITRIFICATION MODEL CALIBRATION - DATA SET 3
MUNICIPAL WASTEWATER TEMPERATURE = 8.7°C
941
-------
I
o
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.8
— 6= 1.104
1
I
5 10 15 20
TEMPERATURE (°C)
25
O.07O
0.0 6 O
0.050
O.O4.0
0.0 3O
0.020
O.OIO
0.008
6= 1.103
k5(20r°-045
1
O 5 10 15 20
TEMPERATURE (°C )
25
0.30
0.20
0.15
r* o.io
0.08
O.O6
0.05
0.04
0.03
0.02
0.015
>-
o
6= 1.106
I
1
1
0.30
0.20
O.I 5
~ 0.10
>-
Q
5 10 15 20 25
TEMPERATURE (°C )
O.08
0.06
0.05
O.O4
0.03
0.02
O.OI5
5 10 15 20
TEMPERATURE (°C )
25
FIGURE 9
EFFECT OF TEMPERATURE ON KINETIC COEFFICIENTS
942
-------
,t rC)-20
a 0 value of 1.1 was chosen to represent all parameters. A value of this
magnitude is typical for the nitrogen data but high for the BOD data. This
may be due in part to the digester supernatant influent to the first data
set. To obtain a greater degree of reliability for these temperature coeffi-
cients additional data sets should be analyzed. For the autotrophic growth
rate, vu, a lower growth rate was required when the yield coefficient was
reduced from 0.38 to 0.17 for the first three data sets; this lower value
used for final verification runs .
Figures 10 and 11 illustrate the profiles resulting through the biofilm
for the first and third stages for data set 1 (Figure 7). Sectors 2 and 4
only, the last sectors in the submerged and aerated portions of the disc, are
presented to show the maximum differences in concentrations. For stage 1,
sharp gradients in both BOD^ and D.O. result through the biofilm. This is a
result of the relatively high BOD reaction rate compared to the mass transfer
rate. Much smaller gradients exist for the soluble organic nitrogen due to
the lower reaction rate. The latter tends to be reaction limited not mass
transfer limited as are BODj- and oxygen. The depth of oxygen penetration is
significantly greater above the water line due to the higher D.O. values ex-
isting in the aerated sectors compared to the tank. Only 10% of the total
biofilm depth of this stage (3180 ) is active in the aerated sectors and 3%
in the submerged sectors indicating the large amount of excess biofilm
carried in this stage. This is contrary to the value of 70% assumed by
Clark, et al . No nitrification occurs in this stage due to the heterotrophs
completely outgrowing the autotrophs as seen later in Figure 12. In stage 3,
nitrification is beginning since the BOD5 has been reduced to a level allow-
ing significant autotrophic growth to occur. The nitrate profile is very
interesting in this stage since it peaks at an intermediate depth in the bio-
film of 100-150 microns. Since nitrate is produced within the biofilm from
nitrification, concentrations in the liquid film and tank are lower than the
peak due to the concentration gradient required for diffusion out of the bio-
film. In the internal portions of the biofilm where the oxygen concentration
is near zero, significant denitrification occurs due to additional BOD,- re-
moval and heterotroph endogenous respiration. The ammonia gradient in the
biofilm is slight, similar to the organic nitrogen, indicating the reaction
controlled nature of nitrification except in the lower depths of biofilm
where oxygen transport limits reaction. Oxygen penetration in the submerged
sectors is about double that in the first stage due to the lower rates in the
biofilm.
In Figure 12, the heterotrophs are seen to predominate entirely in the
first two stages followed by a rapid decline with a proportional increase of
the autotrophs in the latter stages. This is due entirely to the growth
rates of the two species ami the concentrations of BOD and ammonia existing
in, the -biofilm"in these~~itages. The major mechanisms for TKN removal are
also shown in Figure 12. In the first two stages, net growth is the only
significant removal mechanism for TKN. In these first two stages growth re-
moves approximately 2.3 mg/1 TKN but the net removal is about one half that
amount due to TKN production by endogenous respiration of the heterotrophs.
The major fraction (-90%) of TKN removal is due to nitrification in the last
four stages.
943
-------
30
25
^_^
^ 20
01
£
— 15
n
0 10
m
5
S|m = 50 mg/l
TANK
~~ \
\
\
• i i P... \
LIOUID\. \
_ FILM \v
^^^i^-..
-
1 1 1 1 1 1 1
30
25
.-^
"S. 20
E
"" 15
^y
I
••••* 10
^
5
/->
S2m= 2 mg/l
—
— ABOVE AND BELOW
WATERLINE
—
-
1 1 1 1 1 1 I
0 IOO 200 300 400 500 " 0 IOO 200 300 400 500
BIOFILM DEPTH (microns) BIOFILM DEPTH (microns)
lo.u
12.5
c 10.0
o>
E
— 7.5
O
w 5.0
2.5
f\
S5m = 2mg/l
~~
_
—
'-S^=j—
—
II 1 1 ! _ 1 L_
\_»
5
^> 4
. E
_-
*z. 3
1
ro 2
O
z
1
r\
S4m= O.I mg/l
—
—
—
1 t 1 \ 1 1 1 1
0 IOO 200 300 400 500 ~ 0 IOO 200 300 400 500
BIOFILM DEPTH (microns) BIOFILM DEPTH (microns)
12
•^
o> 10
E
Z 8
LJ
X 6
O
O y.
UJ 4
O 2
CO
O
S3m= O.I mg/l
—
—
>
_ \
\
\
\
_ \.
^ N.
1 1 ^--l l^-^l 1 I
LEGEND:
ABOVE WATER LINE -SECTOR 4
BELOW WATER LINE - SECTOR 2
O IOO 20O 300 4OO 500
BIOFILM DEPTH (microns)
FIGURE 10
STAGE I CALCULATED PROFILES OF
BIOFILM CONCENTRATIONS -DATA SET I
944
-------
«
10
• —
^ 8
o>
£
"*""* 6
Q
0 4
CD
2
Q
SIm=50mg/.
—
TANK
\
~ \
\
— LIQUID^S.X
FILM >^\
— ^^^^
^v^^
1 ! 1 1 ! 1 - ! .
30
25
_
—
^. 20
6
"" 15
1
+ 4- 10
5
n
So =2 mg/l
2m
—
—
—
—
! 1 ! 1 II 1
0 IOO 200 3OO 400 500 " 0 IOO 2OO 30O 4OO 50O
BIOFILM DEPTH (microns) BIOFILM DEPTH (microns)
IO
^ 8
c»
e
— 6
O
to 4
2
Q
s.. •«-./-
_
—
~~ii— -— .
— •"
-
1 1 1 1 1 1 1
D
5
^ 4
E
. — .
•z. ' 3
1
p
0
1
r\
,4..OJ^
—
_-Tir^\^
— ''*~ ~~~ ~~'^ ____
, '
-
1 1 1 1 1 1 1
O IOO 20O 3OO 400 500 " 0 IOO 200 30O 400 5OO
BIOFILM DEPTH (microns) BIOFILM DEPTH (microns)
-~
"5, 10
c
z 8
u
o
> 6
X
o
^^^
Q 4
O o
to £•
to
Q
S3m = O.I mq/l
—
—
• .— ., ..^
\
— \
\
\
\
— \
\
\
~" *— — -^ ^^.
N^ N^
1 ! """I"- — L^-^-L_ 1 1
-
LEGEND:
/\nn\/F WATFP i IMF ^FfTrvf? A
BELOW WATER LINE-SECTOR 2
° O IOO 200 300 400 500
BIOFILM DEPTH (microns)
FIGURE II
STAGE 3 CALCULATED PROFILES OF
BIOFILM CONCENTRATIONS - DATA SET
945
-------
100
80
OQ 60
cc
t-
£/)
X
I-
=S
O
ir
CD
40
20
0
AUTOTROPHS
HETEROTROPHS
STAGE
30
25
~ 20
en
E
15
10
Ul
°= 5
0
-5
-10
Q = 0.39 mgd
"Q = 1.51 gpd/ft2
T = 19.0 °C
ENDOGENOUS
RESPIRATION-
\
I
3 4
STAGE
FIGURE 12
DISTRIBUTIONS OF GROWTH AND TKN REMOVAL FOR DATA SET I
946
-------
The distribution of total soluble nitrogen removals for this run is shown
in Figure 13. The major quantity of nitrogen lost from the system is due to
denitrification which occurs mainly in the third and fourth stages when sig-
nificant populations of both heterotroph and autotrophs exist. Some denitri-
fication occurs in the first two stages due to 0.4 mg/1 of NO_-N present in
the influent.
The biofilm contains the major portion of active biomass and thus
accounts for 89 to 97% of the reactions occurring in the system. The tank
reactions are seen to account for 3.^.to 10.8% of the total reaction, the
higher values for the higher suspended solids concentrations carried in the
tanks. The liquid film removals are negligible due to the relatively low
volume compared to the tank volume. In viewing the nitrogen removal data,
less than 4% of the total nitrogen removed from the system occurs by strip-
ping at pH values less than 7.7. At higher pH values, greater ammonia strip-
ping w.ill occur; data set 6 showing 7.1% of the nitrogen removal occurred by
stripping due mainly to the stage 1 pH of 8.7. The remainder of the nitrogen
removal occurs by net growth and denitrification.
At lower temperatures, growth predominates due to the higher dissolved
oxygen values in the tank and lower heterotrophic endogenous decay rates;
both reducing denitrification.
Applications of the RBC Model
After obtaining kinetic rate constants from pilot plant data, the RBC
model can be employed to design full-scale plants to minimize total treatment
cost. A significant component of total treatment cost is the capital expend-
iture for media surface. It is therefore advantageous to obtain a high BOD
removal per unit of media surface area for each treatment stage.
As illustrated in Figure 14, the RBC model can be employed to find opti-
mum operating conditions for carbonaceous removal. Removal increases with
BOD applied up to a maximum value where oxygen limitation results. Consider
the treatment of a wastewater containing 100 rag BOD/1 and zero dissolved oxy-
gen. From Figure 14, it is seen that a maximum BOD removal is achieved by
sizing the first treatment stage so that the applied BOD rate is 7.5 lb/d/
1000 sf (36.7 kg/d/1000 sq. m). Under these conditions, the BOD removal rate
will be 3.9 lb/d/1000 sf (19.1 kg/d/1000 sq. m) corresponding to 52 percent
removal. Thus, the influent to the second stage will be a substrate concen-
tration of 48 mg BOD/1, and a similar procedure can be utilized. This leads
to a "T" type design configuration, a large first stage followed by smaller
later stages. If the design incorporates applied organic loadings much
greater than the breakpoints in Figure 14, excess biological growth should
result on the media since an excess quantity of substrate is available to
provide anaerobic growth in the lower biofilm layers.
Q
In a recent paper,. Chesner and Molof observed that BOD removals decrease
with increasing media diameter of a laboratory RBC system operated at con-
stant peripheral velocity. This is a very significant finding, in view of
the fact that full-scale plants are often designed by scaling-up pilot plant
data at fixed peripheral speed and hydraulic loading.
947
-------
,- 4
01
E
o
2
U)
CC
z
UI
o
cc
1-
o
AMMONIA STRIPPING
234
STAGE
FIGURE 13
DISTRIBUTION OF TOTAL SOLUBLE NITROGEN REMOVAL
FOR DATA SET I
948
-------
o
5
UJ
a:
o
o
DQ
Q
O
00
Q
UJ
_J
Q.
QL
CO
UJ
O
2
cc
ZD
CO
— UJ
UJ
cr
at
o: "
UJ «
Q- o
o
CO 2
O -
2 -
LJ
OC.
Q
O
CD
LJ
O
CO
UJ
{ js ooo'i/a/ai) Q3AOW3d aoa
949
-------
The RBC model predicts this effect for the 6 gpd/ft pilot plant run \
given in Figure 4. In order to maintain constant peripheral velocity, rota-
tional velocity must be decreased as disc size is increased. Detention time
of the liquid film above the water line is increased, resulting in a large
reduction in substrate concentration before reentry of the media into the
mixed liquid in the tank. For a peripheral velocity of 1 fps, the detention
times above the water line are 10.3 and 31.4 seconds for the 6.56 foot (2m)
and 20 foot (6.1 m) discs, respectively. In stage 1, substrate concentration
dropped to 15.4 rag BOD/1 on the 6.56 foot pilot unit, and to 2.2 mg BOD/1 on
the 20 foot disc. Low concentrations in the liquid film coating the biofilm
result in a depletion of substrate within the biofilm. As a consequence,
large portions of the biofilm are inactive above the water line for the 20-
foot media.
The effect on total BOD removal for the four-stage system (Figure 15) is
a reduction from 86 to 77.5 percent in going from the 6.56 foot pilot disc to
a 20 foot plant disc. Pitfalls of this type can be avoided by designing
plants with the RBC model after coefficients have been obtained from pilot
scale data. Otherwise appropriate scaling factors have to be obtained from
comparative studies using different diameter systems on the same wastewater
(Wilson et a!1U).
Afer having sufficiently verified the above approach with pilot and full
scale data, the model provides the framework necessary to analyze additional
operating modes and conditions, thus minimizing the relatively expensive and
time consuming pilot or full scale studies. Utilizing the highest loading
condition for the nitrifying six stage Gladstone, Michigan plant (Figure 7),
Figure 16 shows the effect of providing a high purity oxygen atmosphere in
the aerated sectors of the RBC system. The higher driving force for oxygen:
transfer to the aerated sectors reduces the oxygen limitation on the nitrifi-
cation rate, allowing a lesser number of stages to provide the same effluent
quality. No effect on BOD removal results since the low BOD concentrations
cause the heterotrophs to be substrate limited rather than oxygen limited.
Utilizing air instead of mechanical drive to rotate the media provides an
additional source of oxygen transfer to the system by diffused aeration in
the mixed tank. The oxygen absorption efficiency is probably somewhat lower
than coarse bubble aeration due to the lower liquid depths. As shown in
Figure 17i use of an air drive system with absorption efficiencies of either
2% or H% will provide higher tank dissolved oxygen concentrations, and will:
require fewer stages than a mechanical drive system to achieve similar efflu-
ent qualities. The air drive system at 4% oxygen absorption efficiency would
provide efficiencies similar to those using a 30% oxygen atmosphere in the
gas phase (C = 15 mg/1).
S
The effect of temperature on nitrification efficiency is shown in Figure
18 as a function of hydraulic loading rate for a given influent composition
and system design, the lower temperature causing higher effluent TKN values
at similar loading rates. Optimum utilization of RBC systems will result if
a variable effluent standard for TKN was allowed, higher in the winter than
in the summer based on receiving water quality considerations.
950
-------
5
o
5
UJ
o:
o
o
en
UJ
o
a:
UJ
o.
100
90
80
70
60
50
30
•20
10
MEDIA
DIAMETER
Sj = 60 MG BOD/I
6 GPD/SF
244 I /d /«q. m
2
STAGE
FIGURE 15
EFFECT OF MEDIA DIAMETER ON
BOD REMOVALS AT CONSTANT PERIPHERAL VELOCITY
951
-------
6C
C 50
E 40
a 30
O
CO
. 20
0
<° 10
f,- =2.0
~ /i| = 2.7
Slm= 50
- \
\
\^
— N.
\.
--»>_
I I I ! 1 1 !
30
25
^ 20
E
z l5
1
•^ 10
2
5
r\
£12 =0.045
S2m=2.0
^ 'N^^X.
^^ * ^^x^^
~" ^ 'N ^^^^^
^ v'\ ^^\
~ X>C""\
1 I.I 1 1 1 ~~T-
1123456 " I 1 2 3 4 5 6
STAGE STAGE
30
25
^ 20
E
— 15
z
o
w 10
5
_
_ k 5 = 0.04
5 Tl
—
^.
~~ ' <"«-=3-s=__^_^
- ~"^ -__
"** "™"*fei
1 1 1 1 ! 1 I
30
_ 25
"^ 20
E
15
2
1
K> 10
0
2
5
/•>
•^
y
Y2 = 0.17
_ ^,-~
X
X
— / /'
/ .,•
/ .x"x
/ x ^^
/ '' s^'
/ ^' ^^X^
1 }£-^'\^^\ 1 1 1
1123-456 ~ I i 2 3 4 5 6
STAGE STAGE
~
^ 10
E
2 8
UJ
o
o
Q 4
3
^J O
C/5 ^
IO
O
Q
X
X a, =0.8
, ** bt = 0.18
/' S3m=0.l
f
_ / ^/
- . .— •-''
/ x""
-__, / •
/ ^^
1 ! 1 1 ! 1 1
Q = 1.51 gpd/ft2
TEMP. = 19 °C
LEGEND:
Cs = 9.2 mg/l
. cs =15.0 mg/l
Cs =25.0 mg/l
I 123456
' STAGE
FIGURE 16
EFFECT OF OXYGEN SATURATION ON RBC PERFORMANCE
952
-------
60
~ 50
CT
e 40
o
m
30
20
10
0
25
^ 20
CT
E
— 15
o
w 10
UJ
o
X
o
o
UJ
o
to
5
O
12
10
f = 2.0
lm
=50
!
234
STAGE
= 0.04
5m
I
1
234
STAGE
= 0.8
= 0.18
3m
= O.I
1
t
234
STAGE
30
25
20
15
10
5
0
25
20
10
5
O
p.'z =O.O45
S2m =2.0
1
1
23 4
STAGE
Y = 0.17
I 234 5 6
STAGE
Q = 1.51 ,gpd/ f f2
TEMP. = 19 °C
01=1.5 TO O.8 rpm
E0= 2%, 4%
Gs= 73.1 TO 20.6 scfm
LEGEND:
MECHANICAL DRIVE
AIR DRIVE 2% E0
4% E0
FIGURE 17
MECHANICAL DRIVE VERSUS AIR DRIVE ( TAPERED AERATION)
953
-------
2.0
•o
o.
en
O
Z
Q
<
O
_J
O
oc
o
>-
I
1.2
0.8
0.4
INFLUENT TKN = 23.9 ing/!
INFLUENT BOD5 = 35.0mg/l
6 STAGE SYSTEMS
J_
J_
EFFLUENT TKN (mg/l)
I
8 12 16
TEMPERATURE (°C)
20
24
FIGURE 18
HYDRAULIC LOADING VERSUS TEMPERATURE
FOR VARIOUS EFFLUENT TKN REQUIREMENTS
954
-------
Figure 19 evaluates the effect of typical municipal hourly load and flow
fluctuations on RBC system performance. The biological growth distribution
between autotrophs and heterotrophs was held constant at the distribution
obtained for steady state conditions. The load fluctuations over a day for a
given system design will cause significant fluctuations in effluent quality
over the day, the peak flows and loads causing increased effluent concentra-
tions, while nighttime low flow values would have significantly lower concen-
trations. This agrees well with the data of Filion et al
CONCLUSIONS
1. Organic carbon removal, nitrification, and total nitrogen removal in
the RBC process can be successfully described by incorporating the basic
principles of mass transfer and simultaneous reaction in a steady state
mathematical model.
2. Design of a full scale plant from pilot plant data without use of
the model or appropriate scale-up factors can result in underdesign.
3. From the modeling results and verification studies conducted to
date, it is concluded that:
a) The degree of ammonia stripping in the RBC process is slight
(normally less than H% of total nitrogen removed). This is due to
the relatively low gas exchange rates and low non-ionized fractions
at pH values less than 8.0.
b) The major fraction of total nitrogen removal occurs due to
growth and denitrification, the former predominating during low
temperatures due to high D.O. values. The major portion of denitri-
fication occurs in the middle stages where nitrification is initi-
ated in the presence of significant heterotrophs.
c) Reactions occurring in the liquid film in the aerated sectors
are negligible due to low liquid volume and suspended solids concen-
trations.
d) Tank reactions are a function of the tank MLVSS concentrations
in each stage. Less than 5% of total reaction occurred in the tank
at MLVSS concentrations of 50-80 mg/1 while as much as 11% occurred
at concentration above 110 mg/1.
4. Calibration of the model for air drive systems is required to deter-
mine appropriate 02 transfer efficiencies and effects on kinetic coeffi-
cients .
5. After calibration, the model allows the designer to describe process
performance with one set of temperature-corrected kinetic coefficients for a
particular waste. It puts the description of the process on a rational basis
which will be useful in analyzing results of full and pilot scale studies.
It should provide the basis for design since it allows evaluation of various
process configurations and wastewater characteristics other than those of the
pilot study.
955
-------
60
x] 50
o>
^)40
Q
O
CO 30
UJ
m 20
% 10
f , =2.0
/i =3.0
I 12345
STAGE
o>
E
30
25
20
15
1 10
K5 = 0.045
S5m=2.0
'T 1 I
234
STAGE
o>
E
UJ
O
X
O
Q
UJ
O
to
to
12
10
8
6
4
2
O
a1 =0.8
b' =0.2
2345
STAGE
X
•z.
o>
E
30
25
20
15
10
5
0
30
25
20
15
10
5
0
p-2
- 0.05
I 123456
STAGE
Y2 =0.17
2 3 4 56
STAGE
Q = 1.51 gpd /f»z
TEMPERATURE = 19° C
LEGEND:
CONCENTRATION
(mg/l)
SYMBOL Q BOD5 NH3 -N SON! LOAD
Q 35 16.5 7.4; W
0.5Q 17.5 8.25 3.7 0.25W
I.5Q 52.5 24.75 II. I 2.25W
FIGURE 19
EFFECT OF LOAD VARIABILITY ON'RBC PERFORMANCE
956
-------
NOMENCLATURE
Para- Sub- Super-
meters script script
Identification
Dimensions
EX
G
H
K
KLa
k
M
N
Q
S
S
2
L
5
2
W
L
F
Interfacial area
Total wetted Area for one side
of clean disc
Total aerated area
Diffusivity
Component diffusivity in biofilm
Component diffusivity in liquid
film
0~ absorption efficiency for
air drive
Endogenous respiration rate
Gas flow rate
Air flow rate for air drive
media
Henry's constant for ammonia
Transfer coefficient from gas
to liquid
Transfer coefficient from
liquid to biofilm
02 transfer coefficient in tank
for air drive
SON reaction rate
Molecular weight ammonia
Number submerged sectors
Influent wastewater flow rate
Liquid film flow rate
Biofilm flow rate
L2/T
M/L3-T
L3/T
scfm
mg/l-atm
L/T
L/T
g/mole
L3/T
957
-------
NOMENCLATURE (Continued)
Para- Sub- Super-
meters script script
R
R
RX
S
T
t
V
H
1
2
2N
3
4
5
g
1
2
L
T
0
1
L4
Identification
Reaction rate
Gas constant
Hydraulic radius of channels
Growth rate
Component concentration in
Biofilra
Substrate (BOD5 or COD)
Ammonia Nitrogen
Nonionized ammonia nitrogen
Oxygen
Nitrate nitrogen
Soluble organic nitrogen
Liquid film
Tank
Influent
1st biofilm layer
Last aerated sector
Absolute Temperature
Temperature
Tank volume
Gas volume adjacent to wetted
area, AT
Organism (VSS) concentration
Heterotrophs
Autotrophs
Dimensions
M/L3-T
0.082 1-atm/g-mole- K
L
M/L3-T
mg/1
958
-------
Para- Sub- Super-
meters script script
NOMENCLATURE (Continued)
Identification
Dimensions
Active biofilrn thickness
Liquid film thickness
Rotational speed
rpm
959
-------
REFERENCES
1. Famularo, J., Mueller, J.A., and Mulligan, T., "Application of Mass
Transfer to Rotating Biological Contactors," JWPCF, 50, 635-670, 1978.
2. Mueller, J., Famularo, J., and Paquin, P., "Nitrification in Rotating
Biological Contactors," accepted for publication, 1979 by JWPCF, pre-
sented at 51st WPCF Convention, Anaheim, October 1978.
3. Mueller, J., Famularo, J., and Paquin, P., "Practical Application of
Mass Transfer to Rotating Biological Contactor and Trickling Filter
System," accepted for publication in WATER, 1979, presented at 86th
AICHE National Meeting, Houston, April 1979.
4. Painter, H.A., "Review of Literature on Inorganic Nitrogen Metabolism
in Microorganisms," Water Research, 4, 393-450, 1970.
5. Handbook of Chemistry and Physics, CRC Press, 54th Edition, pp. D130,
1973.
6. Levich, V.G., Physiochemical Hydrodynamics, pg. 681, Prentice-Hall, New
Jersey, 1968.
7. Williamson, K.J., and McCarty, P.L., "Verification Studies of the Bio-
film Model for Bacterial Substrate Utilization," JWPCF, _48 pp. 281-296,
1976.
8. Clark, J.H. et al., "Performance of a Rotating Biological Contactor
Under Varying Wastewater Flow," JWPCF, 50, 896-911, 1978.
9. Chesner, W.H., and Molof, A.H., "Relative Performance of Different
Sized Biological Rotating Dies," paper presented at 49th Annual WPCF
Conference, Minneapolis, 1976.
10. Wilson, R.W., Murphy, K.L., and Stephenson, J.P., "Effect of Scale-Up
in Establishing Design Loadings for Rotating Biological Contactors,"
presented at the 51st Annual WPCF Conference, Anaheim, October 1978.
11. Filion, M.P., Murphy, K.L., and Stephenson, J.P., "Performance of a
Rotating Biological Contactor Under Transient Loading Conditions," pre-
sented at the 50th Annual WPCF Conference, Philadelphia, 1977.
950
-------
PART VIII: UPGRADING PRIMARY AND SECONDARY WASTE TREATMENT SYSTEMS WITH RBC
UPGRADING PRIMARY TANKS WITH
ROTATING BIOLOGICAL CONTACTORS
by
0. Karl Scheible
John J. Novak
Hydroseience
Wes-twood, New Jersey
Introduction
The Borough of Edgewater, New Jersey, operates a primary
wastewater treatment facility which discharges into the Hudson
River. The State of New Jersey, in 1971, imposed a requirement
on Edgewater to upgrade its facility to secondary treatment.
Constrained by severe land limitations, several treatment
alternatives were considered which would minimize plant
expansion.
A process which indicated considerable promise involved the
installation of rotating biological contactors (RBC) in the
existing primary clarifiers. An intermediate floor would be
installed at mid-depth in the tanks. Biological treatment of the
raw wastewaters would be accomplished in the upper RBC sector and
secondary clarification would take place in the sector below the
floor. Because the proposed treatment scheme involved new
concepts, a pilot study was required to confirm the feasibility
of the process and to develop design and cost information. A
program was then developed and financed by Edgewater to evaluate
the RBC/Underflow Clarifier system with a prototype unit.
The installation of the RBC/Underflow Clarifier pilot
facility was completed in May 1973. The process evaluation was
conducted over a period of three years by Edgewater personnel and
results from these studies indicated that modification of the
primary Clarifier to the two-tier treatment process could produce
a secondary treatment effluent commensurate with U.S. Environ-
mental Protection Agency effluent requirements. Realizing its
potential, Edgewater officials sought, and received, a
demonstration grant from the U.S. EPA to improve the existing
facility and to continue the process evaluation under an
intensive testing program.
961
-------
OBJECTIVES AND SCOPE OF STUDY
The objectives of the
evaluation were as follows:
RBC/Underflow Clarifier pilot
(1) Establish the feasibility of upgrading existing primary
sedimentation plants to meet the secondary treatment
requirements of PL 92-500 through the installation of
RBC units directly in primary clarifiers.
(2) Evaluate the degree of pretreatment necessary to
successfully operate an RBC system in this mode.
(3) Evaluate the effects of climatic conditions, diurnal
flow, and total daily load and waste characteristic
variations on process efficiency.
(4) Establish process and plant design parameters, and
capital and operating costs for the application of this
upgrading technique to maximize the use of tankage and
facilities at existing primary sedimentation plants.
This paper attempts to highlight several aspects of the
study. A formal report was prepared and submitted to the
U.S.E.P.A. Municipal Environmental Research Laboratory,
Cincinnati, Ohio, which documents the results of the total
program.
EDGEWATER TREATMENT PLANT
The, Edgewater Sewage Treatment plant has a design capacity of
1140 m /d (3 mgd), and provides primary treatment of combined
sanitary and storm waters. Figure 1 shows a plan of the existing
plant site. The major facilities include an administration
building, pump house, comminutor, grit collector, five primary
settling tanks, chlorine building, and outfall sewer. Sludge is
processed in two anaerobic digesters and two vacuum filters. A
flash dryer is available although not presently used. Land is
limited, comprising only 1.5 acres of usable area.
Table 1 presents a characterization of the raw wastewater for
the one year testing period from March 1977 to February 1978.
962
-------
BOROUGH
OF
?A;CL1FFSIDE PARK
BOROUGH
OF
EDGEWATER
INFLUENT SEWER
o
PROPERTY LINE
SLUDGE DIGESTERS BUILDING
*
CHLORINE BUILDING
COMMINUTOR
GRIT COLLECTOR
PRIMARY
SETTLING
TANKS
SLUDGE
BUILDING
SITE OF
' PILOT FACILITY
ADMINISTRATION
BUILDING
INFLUENT
SEWER
100
20O
300 FT.
Figure 1. Wastewater treatment plant site.
_
963
-------
TABLE 1
RAW SEWAGE COMPOSITION
Average
Flow, m /day (mgd)
BOD5 total, mg/1
BODt- soluble, mg/1
COD total, mg/1
COD soluble, mg/1
TSS, mg/1
TVSS, mg/1
TKN total, mg/1
TKN soluble, mg/1
NH3-N, mg/1
9,920 (2.6)
144
80
350
176
169
137
26
22
13
Range of Values
4,540-31,800 (1.2-8.4)
50-573
22-188
128-772
67-280
36-280
44-206
10- 41
9- 31
3- 21
The. average daily flow was approximately 9,900 m /d (2.6 mgd).
Peak storm flows exceeded 27,000 m /d (7.2 mgd), the maximum flow
capacity of the flow meter. The raw wastewaters were generally
representative of dilute combined sewerage, with an average total
BOD,, of 144 mg/1 and 169 mg/1 suspended solids.
RBC/UNDERFLOW CLARIFIER TEST MODULE
layout of the primary clarifiers and
jure 2. Primary Settling Tank No.
fo 1-hp RRP/tlnf^f»r f 1 nw cl^rifi^r rnnr!(=>
A layout or tne primary ciaririers ana nest moauie is snown
on Figure 2. Primary Settling Tank No. 5 was converted to
evaluate the RBC/Underflow clarifier concept. Raw wastewater,
after passing through the comminutor and grit collector, was
pumped from a point located 3.51 m (11.5 ft) from the influent
side of Settling Tank No. 3.
Figure 3 shows a cross-section of Tank 5 after conversion.
The intermediate floor was installed to provide an underflow
clarifier with a water depth of 1.42 m (4 ft 8 in). Four RBC
units with diameters of 3.61 m (12 ft) were installed in the
21.34 m (70 ft) long by 4.37 m (14 ft 4 in) wide tank with a
water depth of 1.22 m (4 ft) above the intermediate floor.
Fiberglass covers were installed over the RBC units for
protection of the media and attached biomass.
964
-------
TOTAL
PLANT
EFFLUENT
^RBC FLOW METER ^
RBC
EFFLUENT-
FORMER
PRIMARY
TANK
NO. 5 K
r
UJ
_1
Q
O
s
1-
LU
H
O
CD
CC
L
1
> ©x'^1
L
r(D
Shaft 1
(J) ©
Shaft 2
j"— •*. V*~X.
CD C§)
Shaft 3
—
(j) (s)
~?.
C
?\
U
Shaft 4
§
K
D__
Cr)
j"
i \ '
UJ
^
> o
_J z
o ^
I- z
0 H-
o
^ cc
0 ^
u. E
1 ^
\
HXh-q
Q_
^
a.
I
^
I
T '
> J
ro
C>
•z.
z.
\-
-^
cc
<
^
tr
L ^
*\t
l\J
o
n
H
>~
<£
^
E
a.
^
(D
PLANT FLOW METER -~^.
(D WASTEWATER QUALITY SAMPLING SITES
(?) TRACER SAMPLING LOCATIONS
. 1
P
-. PRIMARY EFFLUENT
t
DETRITOR^
L_
O
i£
Z.
\-
>_
(T
ej
E
a.
NORTHSIDE
FLOW
SOUTHSIDE
' r FLOW
^ 1
1
1
-\
|
HIGH FLOW
BY-PASS
Figure 2. Plant flow schematic.
965
-------
UJ
h-
o
UJ
0)
Q
0
60
•U
CO
fO
0)
Vt
3
M)
966
-------
The RBC units were made of high-density polyethlyene. Stages
one, two and three each had a surface area of 1,220 m /m of shaft
length (4,000 ft /ft) and stage four had a surface area of 1,830
m /m (6,000 ft /ft). Each of the four shafts was 4.1 m (13 ft 5
in) long. The units were immersed to a depth of 1.07 m (3 ft 6
in) which provided a to^tal effective, wetted surface area for the
four shafts of 18,200 m (196,500 ft ). The small portion of the
central surface which was not wetted and effectively free of
microorganisms represented 17 percent of the total surface area.
Throughout this paper, loading will be expressed on an effective
surface area basis, i.e., wetted surface area.
The test program anticipated the removal of grit and trash
without removing substantial portions of BOD and SS. Since the
plant detritor could be by-passed at times of high influent flows
to the plant, a modification was introduced to provide the intake
to the:RBC pump at a point 3.51 m (11 ft 6 in) from the head end
of Settling Tank No. 3. This allowed high rate primary treatment
prior to the RBC system.
3 A pump capable of providing a maximum test flow of 5,680
m /day (1.5 mgd) was installed, with a programmer to provide a
variable signal to an electrically operated valve. The valve was
programmed to provide a diurnal hydraulic variation, with the
peak to average ratio equal to 1.5 and the minimum to average
ratio equal to 0.5.
A weir was installed in the influent channel to the RBC
sector to distribute the flow uniformly. Additionally, screens
were attached to the influent channel to catch large fibrous
material. The intermediate floor and the influent channel were
adequately caulked to prevent leakage and/or exchange of effluent
and incoming wastewaters.
EXPERIMENTAL PROGRAM
Phase 1 of the experimental program, studied the RBC system
performance over a wide range of loading conditions. The initial
loading was set relatively low to ensure an effluent quality
greater than EPA requirements. (EPA secondary treatment
standards presently call for 85 percent BOD5 and TSS removal or
monthly average BOD,- and SS concentrations less than or equal to
30 mg/1. Maximum weekly average BOD^ and SS concentrations must
be less than or equal to 45 mg/i.) The loading was then
_
967
-------
increased to yield an effluent quality approximately equivalent
to EPA standards. The third and final loading condition was
chosen to stress the RBC system, i.e. violate the 30/30
standards.
BOD5/SS
The low loading condition was evaluated for approximately two
weeks. The moderate and high loading conditions were each
evaluated over an approximate period of five weeks. Several days
were provided before each analysis period for the system to
acclimate to the change in loading. Typically, this acclimation
period extended over one to two weeks.
Phase 2 involved selection of an optimum system loading rate
based on an analysis of the data collected from Phase 1. This
selection was aided by use of a computer simulation model of
fixed film kinetics with particular application to the RBC
system. The second phase of the program studied long-term.steady
state operation of the RBC at the pre-selected optimum loading.
rate applied during warm temperature conditions and low dissolved
oxygen levels. A two-week acclimation period was provided before
this study period, which lasted two months.
Phase 3 of the experimental program imposed the optimum
loading on the system during winter, cold temperature 'conditions
for a period of 2-1/2 months. The loading was maintained at or
near that evaluated during the summer months.
Sampling
Seven sampling locations were utilized to monitor water
quality through the RBC system. These are shown on Figure 2.
Daily, 24-hour, flow-proportioned composite samples were drawn
from the raw influent, RBC influent, and the final effluent from
the RBC/Underflow Clarifier. Discrete samplers, ISCO Model
Number 1680 Samplers with Multiplexers, Model Number 1295, were
positioned at the RBC influent and underflow clarifier effluent;
a single composite sampler, ISCO Model Number 1580W, was
maintained in the raw influent waste stream.
Periodically,
-i r~
24-hour flow-proportioned composite samples
were drawn from each of the four stages in the RBC system. These
samples were drawn from mid-depth with submersible pumps, and
composited in 18.9 1 (5 gal) containers packed with ice and
insulated.
968
-------
Analytical Program
Table 2 summarizes the analytical schedule followed during
the major phases of the experimental program. The numbers
indicate the number of samples analyzed for that particular
parameter per week.
Studies were also conducted periodically to characterize the
physical and hydraulic operation of the system. These included
tracer analyses, zone and flocculant settling tests, and chemical
addition tests.
During each flow condition, or major modification to the
physical system, a tracer analysis was conducted to characterize
the -hydraulics through the RBC system and to monitor the system
for any physical abnormality such as leakage, etc. Lithium
chloride was evenly distributed across the RBC influent channel
and samples taken with time at selected sampling locations (see
Figure 2). The samples were then analyzed for lithium by
standard atomic absorption spectrophotometer procedures.
Flocculant settling tests were conducted using 2.13 m (7 ft)
high, 15.24 cm (6 in) diameter columns with sampling ports at
0.305 m (1 ft) intervals. Sample (typically from the fourth
stage) was pumped into the column and aliquots drawn at each port
at regular time intervals. Standard jar test procedures were
employed to evaluate the effects and feasibility of chemical
addition to improve solids capture in the RBC/Underflow
Clarifier.
EXPERIMENTAL RESULTS
The results of the three phase program are highlighted in the
following discussions.
Variable Loading
The results of the system evaluation
conditions are summarized in Table 3. The
RBC/Underflow Clarifier system ranged from
(0.28 to 0.67 mgd) . The~ total BOD,, loads to
11.7 and 19.7 g BOD/d/m (1.1, 2.1 and 4.04
at several loading
flows to the
1000 to 2500 m /d
the system were 5.3,
Ib BODc/d/1000 ft )
for the low, moderate and high loading conditions, respectively.
969
-------
TABLE 2
ANALYTICAL SCHEDULE
(NUMBER OF SAMPLES ANALYZED PER WEEK)
Analysis(4)
Raw RBC Stages RBC RBC
Influent Influent 1,2,3&4 Effluent Sludge
Flow
Temperature
PH
DO
BOD,-(T)
BOD^(S)
COD (T)
COD (S)
TS
TVS
TSS
TVSS .,,.
TKN (T) J;J
TKN
NH--*
Recorded
7
7
7
5
5
5
7
7
3
3
3
3
3
Total Sulfide
Grease/Oils
(2)
7
7
7
7
5
5
4
7
7
3
3
3
3
3
1
1
7
7
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(T)
(3 )
P04-P total
PO^-P T-orthol
._. Biweekly Biweekly
^ ' Periodically
Periodically
Recorded
7
7
7
7
5
5
5
7
7
3
3
3
3
3
1
1
Biweekly
When Drawn
When
When
Drawn
Drawn
(1) Only during interstage studies.
(2) Conducted at EPA Laboratories, Cincinnati, Ohio.
(3) Conducted at the Hydroscience Laboratory, Westwood,
New Jersey.
(4) (T) = Total; (S) = Soluble, as defined by filtrate.
970
-------
TAB LE 3
EVALUATION OP LOADING VARIATIONS
FLow, m /d
(mgd)
3 2
Hydraulic Loading, m /d/m
(gpd/ft2)
Influent BOD,-, mg/1 (T)
(S)
Total BOD,, Loading g/d/m^
(lbs/d/1000 ft^)
f
Soluble BOD,- Loading, g/d/m"
(lbs/d/1000 ft )
Influent TSS, mg/1
Temperature, °C
Influent D.O., mg/1
Effluent BOD,., mg/1 (T)
(S)
Effluent TSS, mg/1
Low
Load ing
3/22-4/5/77
Moderate
Loading
4/11-5/13/77
High
Loading
5/23-6/30/77
1060
(0.28)
0.06
(1.4)
92
48
5.3
1.1
2.8
0.6
124
13.0
6.9
14.5
9.7
1440
(0.39)
0.08
(1.9)
148
98
11.7
2.4
7.7
1.6
122
17.2
5.0
23. 2
22.1
2500
(0.67-)
0.14
(3.4)
143
75
19.7
4.0
10.4
2.1
128
23. 2
2.2
55.0
30.7
24
23
57
971
-------
The effluent during the low
the EPA standards for BOD5
the stressed condition caused
and moderate loadings complied with
and suspended solids (30/30), while
a violation for both BOD,- and SS.
Interstage Analysis of the RBC System
Throughout the experimental program, 24 hour flow propor-
tioned composite samples from each stage were analyzed on a
regular basis for soluble BOD5 an<3 COD. Dissolved oxygen
measurements were taken between 9 and 11 AM and were repre-
sentative of the peak diurnal loading condition. These
interstage data were used to calibrate an RBC kinetic model
developed by Hydroscience. This model was in turn utilized to
develop design nomographs, as discussed in a subsequent section.
The model is essentially a series of material balance
equations which are solved to determine substrate and oxygen
levels in the effluent from each stage and in the attached
biofilm. Mass transfer resistances, determined as a function of
operating conditions, are considered in both the liquid phase and
biofilm, and the reaction rate is related to substrate and oxygen
concentrations through the kinetic equations.
Calibration of the model was accomplished by establishing
values of the variables associated with the physical and
biological processes and performing a search for appropriate BODj.
removal and oxygen utilization rate constants. Figures 4 and 5
show the final verification results for each of six study
periods. The model was able to effectively predict soluble BOD5
and DO profiles through the system using a single set of kinetic
parameters for all cases. The D.O. data represent measurements
taken during peak loading conditions. Model D.O. profiles are
presented for both the peak and average conditions.
The analysis indicates that the model is capable of
predicting system performance over a range of hydraulic and
organic loading conditions using a single set of kinetic
parameters. The match of the observed data demonstrates that
hydraulic and mass transfer components respond correctly to
system variations.
972
-------
120
100
80
60
40
20
0
120
1OO
co
E 80
en
E
in
Q
O
CD.
UJ
CD
o
CO
O
CO
UJ
_J
CD
3
O
CO
in
Q
O
CD
UJ
_J
CD
O
CO
60
40
20
0
120
IOO
80
60
40
20
0
LOW LOADING PERIOD
(3/14 and 3/17 )
0 = I.48O m3/d
HL= O.O8I m3/
£
UJ
o
>-
X
o
o
UJ
o
CO
co
o>
E
LU
o
>-
X
o
o
UJ
>
_J
o
CO
CO
O>
E
UJ
CD
>-
X
o
Q
UJ
o
CO
CO
3 4
STAGE
12
10
0
12
10
Q = O.65 mg 02/mg BOD.
b Xv = 3.94 mg 02/ I -mln
k'=O.23min-l
PEAK
0
12
IO
a = 0.65mg 02 /mg BODg
b'Xv = 5.52mg O2/I-min
k = 0.27min-I
a'= 0.65 mg 02/mg BODg
b'Xy= 7.9 mg O2/ l-min
k'=0.34 min-l
AVERAGE-
PEAK
—observed
-predicted
2 3
STAGE
Figure 4. RBC kinetic model verification
based on interstage SBOD and D.O. data
973
-------
120
^ 100
E
to
Q
O
m
UJ
_j
m
o
co
O
CO
UJ
m
:D
o
co
80
60
40
20
0
120
100
80
60
40
20
HIGH LOADING PERIOD
(6/22-7/1/77)
Q= 2,750 m3/d
HL=0.15 m3/
E
0
m
UJ
m
_i
0
CO
1 c\>
100
80
60
40
20
n
—
i
—
r COLD TEMPERATURE PERIOD
( 1
0 =
\-1
_
<
—
>. <
X
i
/5- 2/24/78
1.515 m3/d
)
HL=O.083 m3/d /m2
SL
= 7.8 g SBODc /d/
•n''
TEMP.r ll.4°c"
)
V
\
\,
r^
-1- J_ T
1
1 1
V-^
1°
— *•
1 1
en
E
LU
C3
>-
X
O
Q
UJ
O
CO
CO
UJ
CO
X
o
Q
UJ
O
CO
CO
Q
o>
E
UJ
o
X
o
a
UJ
o
co
co
12
10
0
12
10
0
12
10
2 3
STAGE
Figure 5.
Legend:
Q
0 = 0.65mg 02/mg BODg
b'Xy=7.9 mg 02/l -min
k'=O.36min-l
AVERAGE-
PEAK-
Q = O.65
BOD
b Xy = 7.9 mg 02/ I -min
k'=0.38min-l
AVERAGE-
PEAK-
a = O.65 mg O2/mg 80Dg
b'Xv= 3.35 mg 02/l-mln
'
I
I
I
4 E
-observed
predicted
RBC kinetic model verification
2
STAGE
based on interstage SBOD and D.O.
data
974
-------
f
Oxygen Utilization
A primary role of the rotating media is to provide an
effective means for oxygenation of the fixed biofilm and prevent
anoxic or oxygen-limiting conditions in the removal of substrate.
The system can be approximated by a COD balance (assuming minimal
autotrophic activity) which estimates the total oxygen
utilization for both substrate oxidation and cell synthesis:
02 Utilization =
RBC Influent COD - Effluent COD - COD Wasted
The term (influent COD - effluent COD) is effectively the
TCOD removal rate and. is presented on Figure 6 as a function of
TCOD loading to the system. The COD equivalent of the solids was
then subtracted from the TCOD removal rate and plotted as the net
oxygen utilization rate, as shown on Figure 6. The oxygen
utilization rate, as predicted by the kinetic model is also shown
on Figure 6, and corresponds closely with the curve based on the
COD balance. The flattening of the rates at the higher influent
loadings suggest the system is reaching a limiting condition in
its ability to transfer oxygen.
Active Biofilm
The RBC model is capable of constructing oxygen and substrate
profiles throuh the RBC stages and into the biofilm. The kinetic
equations will cause a reduction in substrate removal when D.O.
becomes limiting. Thus, if the DO is less than an assumed
Michaelis constant in regions of the biofilm, the reaction is
limited by a deficiency of oxygen.
Figure 7 presents an estimation of active biofilm depth for
each stage under high and moderate loading conditions.
As shown on Figure 7, the active film layer is between 300
and 600 pm, typically dictated by oxygen limiting conditions, and
dependent upon loading conditions. This suggests that excessive
growth of biofilm does not result in additional substrate
removal. This was observed in the late summer months when acid
dumps caused considerable sloughing of the attached growth.
Effective treatment was still maintained with a relatively thin
biofilm. A judgment as to whether the biofilm in excess of the
active depth is useful is difficult. While it adds considerably
to the mass to be supported by the shaft, the large solids
975
-------
5.0
4.0
CM
*-
t-
O
o
UJ O
o
T3
3.0
2.0
1.0
25
20
CM
E
I 5
LEGEND:
I 3/22-4/6/77
2 4/II-5/I3/77
3 S/23-6/30/77
4 7/18-9/25/77
5 12/1/77-2/24/78
TCOD REMOVAL
RATE
ESTIMATED 02 UTILIZATION
RATE (MODEL PREDICTION)
ESTIMATED 02 UTILIZATION
RATE(COD BALANCE)
I
0
I
10
20 30
g /day /m2
I I
40
50
2.0 4.0 6.0
Ibs/day/ 1000 ft2
TCOD LOADING RATE
8.0
10.0
Figure 6. Estimate of RBC oxygen utilization rates
976
-------
HIGH LOADING
MODERATE LOADING
en
E
m
O
O
03
LU
_J
CD
_
O
CO
LJ
X
O
Q
LJ
O
CO
CO
70
60
50
40
30
20
10
0
8
I
I
o>
E
in
Q
O
03
LU
_t
CD
ZD
_J
O
CO
0 100 200 300 400 50O
BIOFILM DEPTH
STAGE I
I
D>
E
LU
CO
>-
X
O
Q
LU
O
CO
CO
0 IOO 200 30O 400 500
BIOFILM DEPTH (//.)
HYDRAULIC LOAD= 0.14 m3/d/m2
SOLUBLE SUBSTRATE LOAD= 10.4 g/d/m2
100 200 300 400 500 600 700
BIOFILM DEPTH (p.)
0 100 20O 300 400 500 600 700
BIOFILM DEPTH (fj.)
HYDRAULIC LOAD=0.08 m3/d/m2
SOLUBLE SUBSTRATE LOAD= 7.7 g/d/m2
Figure 7. Biofilm concentrations of substrate
and dissolved oxygen, in successive stages
977
-------
inventory may serve to control net solids production by anaerobic
endogenous respiration.
Summer and Winter Evaluations
A summary of the summer and winter evaluations is presented
in Table 4. There was an average temperature difference of
14.8°C between the warm and cold conditions. The influent
dissolved oxygen averaged 1.4 mg/1 during the summer period and
6.1 mg/1 during the winter.
The results indicated that under equivalent loading
conditions, similar removal efficiencies (as expressed by percent
removal) were experienced during both the summer and winter
evaluation periods.
Temperature will affect several of the mechanisms involved in
the kinetics of the fixed film process, including substrate
removal rates, oxygen saturation values (hence, mass transfer:
driving forces), and the diffusivities of oxygen and substrate.
The minimal impact of temperature on system performance was due
to compensating effects of the various parameters affected by
temperature. Thus, the higher removal rates and diffusivities
experienced in the summer were offset by the low dissolved oxygen
levels and the lower dissolved oxygen saturation value. The
lower kinetic removal rates in the winter were compensated by
high influent dissolved oxygen concentrations and higher
dissolved oxygen saturation values. Since dissolved oxygen
penetration was found to be a limiting factor, imposition of high
dissolved oxygen concentrations and/or higher dissolved oxygen
saturation values will effectively increase the oxygen driving
force, increase-! the active film thickness, and result in
increased substrate removal. Thus, although one would expect
lower substrate removals during the winter due to suppression of
the kientic removal rate, the increased oxygen driving force
provides effective compensation. At Edgewater, this resulted in
substrate removals similar to those experienced in the summer.
Underflow Clarifier Performance
Beyond the fourth shaft (refer to Figure 3), the RBC/Under-
flow Clarifier system consists of two distinct sectors, the
turnaround sector and the underflow clarifier sector. Tracer
analyses indicated that the entire turnaround sector behaved as a
978
-------
TABLE 4
COMPARISON OF SUMMER AND WINTER PERFORMANCE
Hydraulic Loading Rate
m /d/m „
(gpd/fO
TBODg Leading Rate
(lb/d/1,000 ft2)
RBC Influent BOD5
mg/1
Effluent BOD5
mg/1
TBOD5 removal (%)
SBOD5 removal (%)
Temperature °C
Influent D.O.
T
S
T
S
Summer
7/18/77-9/25/77
0. 085
(2.08)
11
(2.3)
134
97
28
23
79
76
26.1
1.4
Winter
12/1/77-2/24/78
0.081
(2.0)
13
(2.6)
158
91
33
24
79
74
11. 3
6.1
979
-------
completely mixed system and that the mixing characteristics in
this sector effectively reduced the volume nominally associated
with the underflow clarifier from 100 to 75 m (26,000 to 20,000
gallons), or by approximately 25 percent. The nominal surfacg
area, i^s., that which is below the intermediate floor is 72.8 m
(784 ft ) . Based on the estimated 25 percent reduction, 2t'"ie
available, or effective, surface area becomes 54.6 m (588 ft ).
Figure 8 presents the correlation of effluent TSS as a
function of overflow rate, based on average observed data from
each of the major sampling periods. The correlation shown on the
Figure, while not particularly uniform, implies an, allowable
effective clarifier overflow rate between 22 and 26 m /d/m (550
and 650 gpd/ft ) to obtain an effluent TSS less than 30 mg/1. At
Edgewater this is equivalent-, to a hydraulic loading rate to the
RBC of 0.065 to 0.08 m /d/m (1.6 to 1.9 gpd/-f t ) , assuming an
effective surface area of 54.6 m (588 ft ). The percent
removals shown on Figure 7 are based on a fourth stage
concentration of 160 mg/1, which was the overall average TSS in
the fourth stage during the experimental program.
A solids inventory was maintained throughout the experimental
program, and, as expected, the data indicated increasing
inventories with increasing BOD,- removal, rates. Figure 9 is a
correlation of the total BOD,- removal rate with total sludge
wastage rate (where total sludge wastage is equal to the sum of
the pumped waste solids and the effluent solids). The slope of
the line is 1.25 Ibs sludge/lb TBOD5 removed. Excluding the
average influent SS of 125 mg/1, this implies a production rate
of 0.3 to 0.4 Ib SS per Ib of TBOD5 removed.
EVALUATION OF PROCESS DESIGN ALTERNATIVES
Pretreatment
Observations and data gathered in the study indicated a need
for pretreatment to remove grit, trash and floatables prior to
the RBC system. In the case of Edgewater, high r^ate primary
treatment, with overflow rates between 285 and 370 m /d/m (7000
to 9000 gpd/ft ) was found to be adequate, achieving between 20
and 25 percent removal of TSS. Rough screening was necessary to
remove large fibrous material which passed through the high rate
primary treatment sector.
980
-------
en
CO
i-
t-
~»
UJ
u.
UJ
ou
50
40
20
10
C
t-Eseivo: ©
1 3/22-4/6/77 .
2 4/11- 5/13/77 /
3 5/23-6/30/77 /
4 7/18-9/25/77 /
5 12/1/77-2/24/78 /
~"" 7 OVERALL AVG TSS
/ RAW INF = 170 mg/l
/ RBC INF = 125 mg/l
©y STAGE FOUR = 160 mg/l
~ v. /
/^\x
_ / ^^~
\ \ \ \ 1
) 10 20 30 40 50 6
m3/day/m2
i i i i i ii
100
90
80
70
60
0
cr.
i
-------
7.0 -
6.0
_ 30.0
5.0
25.0
UJ
CD
Sg 4.0
UJ O
C9 —
D ^
CO \
1-
O
2.0
1.0
35.0
_ N_ 20.0
h- 15.0
— 10.0
- 5.0
TOTAL SLUDGE WASTAGE
EFFLUENT SOLIDS +WASTE SOLIDS
0
O
L
0
DATA SHOWN ARE
WEEKLY AVERAGES
10 15
g/ddy/m2
I I
20
25
1.0 2.0 3.0 4.0
Ibs/day/IOOOft2
TBOD5 REMOVAL RATE
5.0
Figure 9. Correlation of total suspended solids
wastage and TBOD removal
982
-------
Rotating Biological Reactors
RBC fixed film systems function primarily in the removal of
soluble organic material, measurable as soluble BOD,- (and COD).
Thus the design of the system is based on soluble organic loading
and soluble effluent organic requirements. The fraction of the
TBODj- influent loading associated with solids will be removed
from the system by clarification and these removals are related
more to the hydraulic loading of the system. The soluble
removals, however, are directly related to biofilm kinetics and
the ability of the system to transfer sufficient oxygen. The
design sequence assumes, based on the above, that the secondary
clarification sector will provide adequate solids removal
efficiency and reduce TSS levels to within a desired range. The
BOD,, associated with these solids can be computed from measured
BOD5 to TSS correlations and from this the required effluent
soluble BOD^ can be determined.
Single stage design nomographs were developed on the basis of
the kinetic model verifications discussed earlier. These design
curves were developed from the system evaluation at Edgewater
and, as such, should not be directly applied to the design of
systems for treatment of different wastewaters. Application to a
particular wsatewater would require that the appropriate kinetic
parameters be determined and new design nomographs developed.
The curves in this instance are based on an evaluation of a
municipal wastewater system and may be useful in preliminary
design applications and general process sizing for the treatment
of .similar wastewaters.
The design of an RBC system should maximize BOD'removals in
each stage by controlling the BOD loading on the media surface.
Maximizing removals in each stage minimizes the total media
surface area requirements, thereby minimizing the initial capital
expenditure requirements.
* Figure 10 presents a series of single stage solutions based.
on a temperature of 20°C, and an influent DO of 0.0 mg/1. The
reaction kinetics described and verified in the previous section
were used in the development of the curves where the effluent
soluble BODc is determined at the appropriate influent soluble
BOD5 and hyaraulic loading rate. The predicted effluent SBOD5
983
-------
200
ISO
^ 160
CT
E
o
CO
.. 140
I2O
LU
-J
m
J-
UJ
100
80
60
fc 40
bJ
20
INF D.0.= 0.0 mg/l
T= 20°C
HYDRAULIC LOADING
(m3/d/m2)
0.5
0.4
0.3
O.I
I
I
I
I
I
I
20 40 60 80 100 120 I4O 160 180 200 220 240 260
INFLUENT SOLUBLE BOD«, mg/l
Figure 10. Single stage process design solutions relating
effluent SBODC to influent SBOD and hydraulic loading
5 o
984
-------
concentration from the first stage becomes the influent SBOD5 to
the second stage. The iterative use of the design curves allows
the prediction of the effluent from a multi-stage RBC system.
Influent Dissolved Oxygen Effects
A second design curve, Figure 11, presents the effect of
influent DO on the treatment efficiency of the RBC system. The
presence of DO in the influent provides an additional source of
oxygen for the biofilm, and may additionally allow a higher
concentration gradient, enhancing mass transfer into the biofilm.
As indicated on Figure 11, the greater impact occurs at the
higher substrate levels. At an influent SBOD,- of 150 mg/1, an
influent DO of 6.0 mg/1 may allow approximately 12 percent
improvement in BODj- removed in the initial stage.
Secondary Clarification
The pilot study evaluation at Edgewater indicated that a
limiting process condition in the operation of the system was the
solids removal efficiency accomplished by the secondary
clarification zone. The experimental data determined that the
maximum effective overflow rate to obtain an effluent SS of 30
mg/1 was 26.5 m /d/m (650 gpd/ft ). Thus, the hydraulic loading
to the RBC system may be limited by the effective surface area
available in the secondary clarification zone. Tracer analyses
determined this to be 54.6 m (588 ft ) in the Edgewater system
(a 25 percent reduction from the nominal area of 72.8 m ') . The
maximum average flow to the system would therefore be 1,450 m /d
(0.38 mgd) .
40.38 mgd)
nr/d/nr (i
is equivalent to a
93 gpd/ft ) for the
soluble BOD,, of 90
The flow of 1,450 m3/d
hydraulic loading of 0.07
Edgewater system. At an influent soluble BOD5 of 90 mg/1, the
design curves on Figure 10 project an effluent soluble BOD,- of 19
mg/1. Adding the BOD5 associated with the 30 mg/1 TSS, the total
BODc is projected at 26.5 mg/1. Although this will meet the
effluent criteria, the secondary clarifier is effectively
limiting the design of the RBC system to 18,270 m (43,000 ft )
effective media surface area. Denser media, which would allow a
higher organic loading, could not be considered since the
clarifier would become hydraulically overloaded.
985
-------
UJ
cc
O
m
CO
UJ
CO
<
UJ
cc
o
UJ
o
o:
UJ
O.
16
14
12
10
8
6
4
2
0
0= 1500 m3/doy
T= 20°C
INFLUENT
DISSOLVED OXYGEN
6 mg/l
3 mg/l
I mg/l
0
50 100 150
INFLUENT SOLUBLE BOD5, mg/
200
250
Figure II. Single stage process design curves relating
the effect of dissolved oxygen on SBOD removals
986
-------
[
To maximize the organic loading to the RBC sector and
minimize the RBC surface area requirements, consideration must be
given to the design of the underflow clarifier system to
accomplish efficient solids capture. This may involve provision
of additional secondary clarifiers, the use of chemical addition
to improve the efficiency of the existing underflow clarifiers,
or the use of rapid sand filters as a final treatment step.
PROCESS DESIGN EXAMPLE FOR THE EDGEWATER SYSTEM
The following example is presented to demonstrate the use of
the design curves and to further discuss process considerations
relating to the RBC/Underflow Clarifier system. Since the curves
are based on the experimental program at Edgewater, the example
describes the process requirements to upgrade the existing
Edgewater facility to secondary treatment capabilities, based on
the present day waste characterization.
The peak monthly average conditions are summarized on Table
5. These were determined to be the controlling conditions in the
Edgewater process design.
TAB LE 5
PEAK MONTHLY AVERAGES
Flow, m /d
(mgd)
Influent
BOD mg/1 (T)
(S)
TSS, mg/1
SBOD Loading, Kg/d
(Ibs/d)
D.O., mg/1
Temperature, °C
Effluent
BOD mg/1 (T)
* (S)
TSS, mg/1
13600
(3.6)
215
130
260
1750
(3900)
1-5
11-26
30
20
30
987
-------
Design of the RBC system is based on the loadings to the
system, while the clarifier design is considered on the basis of
peak flows. The required effluent criteria which apply to
Edgewater under the stated influent waste characterization are
also presented on Table 5.
Pretreatment
The results of the Edgewater study indicated that the
influent to the RBC sector should not exceed 120 to 140 mg/1
suspended solids. A single primary clarifier at Edgewater would
be used to provide high rate primary treatment for the entire
plant flow. Twenty to twenty-five percent TSS removals can be
expected.
treatment
No
step.
removal of BOD,- is assumed through this pre-
Secondary Clarification
Based on th^e results of the experimental program, an overflow
rate of 24.7 m /d/m (600 gpd/ft ) would be necessary to achieve
the efflyent criteria of 30 mg/1 SS. At peak monthly flow of
13,600 m /d (3.6 mgd), an effective surface area of 6000 ft is
indicated for the Edgewater system.
RBC Organic Removal
Maximum use of the tankage would allow four 4.1 m (13.5 ft)
shafts (3.65 m diameter) per tank. High-density media would be
installed in all but the first shaft in each tank. Conventional
density media would be installed in the first stage. Each shaft
would be 0.46 m (30 in) above the water surface.
Figure 12 presents the solutions for a varying number of
tanks based on the design curves shown on Figures 10 and 11. For
the particular application described on Table 5, the soluble
effluent BOD,- criteria under peak monthly conditions would be met
with a total of nine tanks, each with four RBC shafts^and a total
effective media surface area of 22,600 m (243,000 ft ) per tank^
The total nominal media surface area per tank would be 27,300 m
(294,300 ft ).
Assuming extension of the intermediate floor to provide an
effective surface area of 64.6 m (695 ft ) per tank, nine tanks
would provide a total surface area (effective) of 580 m (6,250
988
-------
O\J
7O
^> 60
E
^
in
Q
O
m 50
LU
|
—- 1
CD
O
CO 4O
I-
2
LU
_1
U- 30
LU
Q
LU
O
£1 on
fl fcU
X
LU
10
Q
FLOW= I0.20O mVday (2.6MGD) INFLUENT
4 STAGES /TANK SBOD5
INFLUENT DISSOLVED OXYGEN = 3.0 mg/l PEAK
ALL STAGE DISSOLVED OXYGEN = 2.0 mg/ 1 AVERAGE
A
/ .
/ /
/
/
/
/
/
/
/
/ /
*
ESTIMATED /
SOLUBLE BOD5 /
_ REQUIREMENTS
PEAK 7-da/= 27 mg/l /
t
PEAK MONTHLY: 2O mg/l
/
AVERAGE DAILY; 14 mg/l
s
\ \
f J
/
/
/
u>
z
z
X
0)
1
\ /
/
/
[_
0)
H
Z
|
r /
/
/
/
/
/
f
s
//
/
en
H
2
7:
0)
I
(ISOmg/l)
PEAK
MONTHLY
AVERAGE
(ISO mg/l)
DAILY
AVERAGE
(90mg/l)
45 mg/l
H
2
-------
ft2). This would be in line with the required secondary
clarifier surface area under peak monthly flow conditions.
Summary of Process Design Evaluation
The process design of both the underflow clarifier and the
RBC sectors was controlled by the peak monthly average condition
at Edgewater. The design is summarized on Table 6. A total of
ten tanks would be necessary at Edgewater, one for high-rate
pretreatment, and the remaining nine modified or newly
constructed as RBC/Underflow Clarifier systems.
Process Design Modifications
Specific improvements can be made in the design of the
Edgewater plant which may result in a reduction of total required
tankage. Two methods suggested are aeration to DO levels of 5.0
mg/1 thoughout the system, and the use of chemical treatment to
improve solids capture efficiency.
Aeration can be provided by a supplemental air supply.
Although not evaluated directly during the Edgewater study, the
potential impact of interstage aeration was simulated by the use
of the design Figures 10 and 11. The results were superimposed
over results of the initial Edgewater design example (Figure 12)
and are displayed on Figure 13. The simulation indicated that
provision of interstage aeration alone did not significantly
improve the design.
Tests were conducted during the study which indicated that
the addition of FeClg to the fourth stage mixed liquor would
significantly enhance the settleability of the solids. Bench
scale tests indicated that it was necessary to provide a
sufficient period of agitated contact between the waste and
coagulant prior to the clarification zone. The results of these
studies showed that within the clarifier operating range of 20 to
m
25
would
/d/m (overflow
allow an
rate), chemical
effluent TSS of 15 to 20
addition
mg/1.
(20 mg/1 FeCl3)
A rapid mix zone would need to be provided at Edgewater to
assure efficient chemical treatment. Alternatives may involve
injecting the FeCl-, solution directly above the air header if
990
-------
TABLE 6
PROCESS DESIGN SUMMARY AT EDGEWATER UNDER PRESENT CONDITIONS
Process Conditions:
Peak monthly average
TBODg loading
SBODj. loading
SBOD^
Temperature
Flow (based on loading)
Influent DO
Flow to clarifier
Process Design Parameters:
TBOD5 loading rate
SBOD5 loading rate
Equivalent hydraulic loading rate 0.05 m /d/m
2,120 kg/day 94,670 Ibs/day)
1,320 kg/day (2,900 Ibs/day)
130 mg/1
20°C
10,200 m /day (2.7 mgd)
3.0 mg/1
13,600 m /day (3.6 mgd)
10.4 g/d/in 9
(2.1 lb^d/1,000 ft )
6.5 g/d/m (effective)
(1.3^1b/d/l,000 ft )
•y r— »J / -i / £
Clarifier overflow rate at peak
Monthly Hydraulic Flow
(1.2 gpd/ft")
23.5 m3/d/m2
Process Design (using existing Edgewater tank design):
2
Total nominal RBC media
Surface Area
Total effective RBC media
Surface Area
Shafts/Tank
Total RBC tanks
Total Intermediate Floor
Surface Area
246,000 m'
(2.65 x910u ft*-
203,400 m K ?
(2.2 x 10 ft )
4
580 m2 (6,200 ft2)
991
-------
u>
Q
o
03
tu
_1
CD
O
OT
Ul
U.
u.
IU
o
UJ
f-
o
UJ
a.
X
at
7O
60
50
40
30
IO
O
FLOW= 10,200 mVday (2.6 MGD} INFLUENT
4 STAGES /TANK SBOD5
PEAK
LEGEND- 7- day
_ INF D.O.= 3.O mg/l AVERAGE
ALL STAGE D.O. = 2.0 mg/l ..
INF D.0.= 5.0 mg/l /
ALL STAGE D.O. = 5.0 mg/ 1 / /
— / . /
/ /
/ / /
ESTIMATED /
SOLU8LEBOD5 //
_ REQUIREMENTS /'
(REO'MTS SHOWN IN PARANTHESES //
ASSUME CHEMICAL TREATMENTTO >
AN EFFLUENT TSS= 15 mg/l ) /
(PEAK 7-doy = 42 mg/l) //
~ /'
/' /
/
It
"~" y
PEAK 7-day= 27 mg/l /
(PEAK MONTHLY: 27 mg/I) h
I
PEAK MONTHLY = 2O mg/l
(AVERAGE DAILY= 19 mg/l) /'
AVERAGE DAILY: 14 mg/l
X
1 1
v /
ff
/-
ID
H
Z
CA
/ /
/
/ y
/
s£
CO
z
X
(A
/ /
/ /
1 '
,
x^"""
'
z
(a
1
/ /
/ ,X
/x7^
/
./]
X^xx
-H
Z
1
' / /
/ / .
/ /
/ /
/
/
/ X
**
Ul
H
Z
X
CA
1
(150 mg/l)
PEAK
MONTHLY
AVERAGE
( !3Omg/l)
DAILY
AVERAGE
(90 mg/l)
45mg/l
H
Z
30 mg/l
22 mg/l
NPDES
REQ'D
TBOD5
500
1,000
1,500
m'/day
2,OOO
I
2,500
3,000
O.I
O.2
O.3
0.5
O.4
MGD
EQUIVALENT FLOW/TANK
0.6
O.7
0.8
Figure 13. Process design at Edgewater with
aeration and chemical treatment
992
-------
supplemental air is being provided or by installing a separate
baffled stage on the extended intermediate floor, with adequate
mechanical mixing.
If the effluent solids are maintained at 15 mg/1, soluble
BODC effluent requirements change significantly. These are shown
The B0°c: associated with the 15 mg/1 TSS is
3 mg/1. Thus the daily, peak monthly, and peak
SBOD5 requirements become 19, 27 and 42 mg/1,
As shown on Figure 13, the peak monthly condition
but the tankage requirement for the RBC system is
seven tanks (vs. nine in the initial design),
on TTigure 13.
assumed to be
7-day average
respectively.
again governs,
now reduced to
assuming provision of supplemental air.
COSTS
Cost analyses were conducted of alternative design sequences
at Edgewater. These costs are based on conditions at Edgewater,
including 1977 loading estimates and removal rate coefficients
determined during the field program. It is important to realize
that costs will be sensitive to these parameters. Thus higher
removal rate coefficients would induce lower capital and
operating costs.
In the first alternative, one of the existing five primary
sedimentation tanks would be converted to a high rate pre-
treatment tank, while the remaining four would be converted to
the RBC/Underf low Clarifier process. new tankage (approximately
equivalent to the existing tankage) was then added to provide the
requirement for additional surface area in both media and
underflow clarification to meet secondary effluent objectives.
This alternative is shown in the upper display on Figure 14. The
unit cost for this upgrading procedure is estimated to be 0.077
$/m ($0.29/1,000 gal), considering both operating and
maintenance, and amortized captal costs.
A second alternative considered was high rate pretreatment,
standard RBC tankage (no underflow clarifier), and utilization of
the existing primary tanks for secondary clarification as shown
in Figure 14. The unit cost of this scheme is estimated to be
$0.061/m^ ($0.23/1,000 gal), which is somewhat less than the
above RBC/Underflow Clarifier design alternative. Land
requirements (included in these costs), however, would be 50
percent higher.
993
-------
INFLUENT
i '
EFFLUENT
q
r>
RBC
13.5 Ft.
Shafts
R
E
T
R
E
A
T
M
E
N
T
a
0
RBC
13.5 Ft.
Shafts
Q
RBC
25 Ft.
Shafts
Ml * It It It
" ? ? ? ?
n
j
?
a
RBC
18 Ft.
Shafts
t j k
CASE I - RBC / UNDERFLOW CLARIFIERS
ESTIMATED ANNUAL COST ft 0.29 PER 1,000 GALLONS
EFFLUENT
INFLUENT
PRETREATMENT
****** * * *
SEC
ON DA
RY
CLARIFIERS
RBC
25 Ft.
Shafts
(Typ)
t
1
CASE 2 - SEPARATED RBC UNITS
ESTIMATED ANNUAL COST $ 0.23 PER 1,000 GALLONS
Figure 14. Mechanical drives schematic layout.
994
-------
SUMMARY
2-
5.
The RBC/Underflow Clarifier concept was demonstrated to be an
effective secondary treatment process, capable of meetina
NPDES secondary treatment ceffluent requirements of 30 ma/1
TBODj- and TSS, or 85 percent TBODc and TSS removal, whichever
provides the qreater degree of treatment.
The peak monthly loadinq at Edqewater controlled the process
desiqn. The influent peak monthly total BODr was 215 ma/1.
with a cor respond inq soluble BOD,- equal to 130 mq/1 .
Limiting organic loadings for the RBCT sector were determined
to be 16L.4 q TBOD^/d/irr (2.1 lb/d/1000 ft2) and 6.5 a
SBODc/d/nr (1.3 lb/d/1000 f tz) .
,Pretreatment
study period
of the raw wastes was required throuahout- the
to remove qrit, trash, and floatables. Rough
screeninq was necessary to remove larae fibrous material
which passed through the high rate primary treatment sector-
Tracer analyses indicated that each RBC staae. as.defined bv
baffle placement, behaved as a completely mixed tank. The
mixing characteristics of the turnaround sector reduced the
effective volume of the clarifier bv approximatelv 25
percent.
The overall organic removal efficiency of the RBC svstem was
limited by oxygen availability, as determined bv use of an
RBC mathematical kinetic model. Oxygen utilization curves
developed from COD balances and the model indicated that- the
system reached a limiting condition in its abilitv to
transfer oxygen at the higher influent organic loadina rates.
The overall seasonal effects were minimal based on the
evaluation of the system under summer•and winter conditions.
The temperature differential experienced was 15°C. Althouah
temperature affects several mechanisms involved in the
kinetics of the fixed film process, the minimal overall
impact experienced over this larae temperature differential
was due to compensating influences of the various parameters
affected by temperature.
995
-------
7.
Pre-aeration was investiqated usinq the kinetic model. since
the system at Edqewater is characterized bv decreasi'na
orqanic load with proqressive staging, the provision of
pre-aeration to the influent of the RBC/Underflow process
would not have a significant impact on removal efficiencv.
Interstage aeration would achieve greater substrate removals.
Underflow baffles effectively stage the RBC system into a
series of completely mixed tanks. Baffling also created
higher velocities along the intermediate floor and minimized
solids accumulation. At an initial baffle clearance of 15.?.
cm (6 in) , velocities were not sufficient to prevent
considerable accumulations on the floor. Reduction of the
baffle clearance to 5 cm (2 in) effectively prevented further
solids accumulation.
A linear corre-lation of total suspended solids wastage as a
function of TBOD,- loading to the system was determined. On
average, biological solids growth was estimated to be 0.38 g
SS /g BOD5 .removed.
A correlation of effluent TSS and underflow clarifier
overflow rate was constructed on data collected over the
entire program. The correlation implies an allowable
overflow rate between 22 and 26 nr/d/rtr (550 and 650
to obtain an effluent TSS less than 30 ing/1.
10.
11
Chemical addition studies showed that ferric chloride
addition to the fourth stage effluent would effectivelv
improve solids settleability. ;
Cost analyses were conducted of alternative design
at Edqewater. The unit cost for the RBC/Underf low
upgrading procedure was estimated to be $0.08/m"
gal) .
seauences
r.lar if ier
($0.3/1 000
ACKNOWLEDGEMENTS
0. Karl Scheible is a Senior Engineer at Hvdroscience, Inc.
and was responsible for management of the Edgewater project.
John J. Novak is a Project Engineer at Hydroscience, Inc.
The assistance of Alonso Gutierrez of Clinton Bogert and
Associates, Fort Lee, N.J. in the cost evaluations is acknowledct-
ed with appreciation.
996
-------
UPGRADING TRICKLING FILTER EFFLUENTS
WITH A RBC SYSTEM
By
Calvin P. C. Poon
Professor of Environmental Engineering
Howard K. Chin
Graduate Student
University of Rhode Island
Kingston, Rhode Island 02881
Edgar D. Smith
Environmental Engineer
Walter J. Mikucki
Chemical Engineer
Army Construction Engineering Research Laboratory
Champaign, Illinois
INTRODUCTION
Many secondary treatment plants in the United States were
built two decades ago. The treated effluents of these plants
cannot meet the current effluent quality standards. Several
alternatives can be taken to improve the effluent quality.
Among these alternatives are: (1) construction of a new treat-
ment plant for the existing and future loads; (2) expansion of
the existing plant; and (3) addition of an upgrading retrofit
system to the existing plant. The last alternative is deemed
desirable because of lower cost requirements and added opera-
tional flexibilities.
This paper describes a pilot plant study using a rotating
biological contactor (RBC) system to remove organic as well as
997
-------
to induce nitrification from a trickling filter effluent. The
application of a RBC treatment system to upgrade secondary
effluent is relatively new in the United States. Only a few
plants have been in operation for more than one year. Lagnese
(1978) described the use of a RBC system operated in parallel
as well as in series with trickling filters. According to his
findings, it was suggested that soluble BOD (SBOD) instead of
total BOD should be used in the RBC process design. Reh (1977)
also suggested the same approach since they considered that
particulate BOD was removed by various physical phenomena and :
its removal was not really a function of the biological oxida-
tion activity of the biomass. Lagnese (1978) also reported in
his study a large portion of the effluent BOD from the RBC ;
process following a trickling filter was nitrogenous BOD (NBOD).
Sullivan (1978) reported the operation of a few treatment plants
using RBC systems to upgrade secondary effluent qualities.
Since only a scanty description of the operation and a meager
amount of data were presented in his report, the RBC performance
for such application cannot be properly evaluated.
PILOT PLANT INSTALLATION AND OPERATION
The pilot plant module was a standard 0.5-m diameter, 4-
stage RBC unit manufactured by Autotrol Corporation,^Wisconsin.
In operation, approximately 40 percent of the 23.3 m total
contact surface area was submerged in wastewater. The single
speed motor furnished for the module unit was replaced with a
variable speed motor so that the effects of rotating velocity
on unit performance could be studied. Chemical supplements
could be added, if needed, from reservoirs holding the prepared
chemical solutions to the pilot unit using chemical pumps. A
clarifier (polyethelene tank) of 0.51-m diameter was equipped
with a 0.5-m long V-notched overflow weir. The effluent from
the clarifier was discharged into the chlorination chamber of
the existing trickling filtering plant. The entire pilot unit,
with the exception of the clarifier, was housed in a sheet-metal
shed.
It took 9 days to develop a thin layer of biomass covering
the entire outside of the media. Both the wastewater temper-
ature and the ambient temperature during this startup period
were at 11°C. Commencing on January 10, 1978, the pilot plant
was run continuously in a 7-month span. Random grab samples
were taken from the pilot plant influent, effluent, and some- :
times from each of the four stages for chemical and biological
analyses. In addition the amount of biological slime on the
rotating media was determined from time to time by scraping
the culture from a 20-cm surface area. Sludge volume index
was determined from 30-min. settling in an Imhoff cone. Graduate
998
-------
cylinder batch settling test was used to determine the sludge
settling velocity as well as the solid flux from which the pro-
per size of a clarifier-thickener could be determined.
RESULTS AND DISCUSSION
Soluble BOD Removal The effluent SBOD concentration
obtained from the pilot plant operation is illustrated in
Figure 1. The soluble nitrogenous BOD,- (SNBOD ) line is ob-
tained by substracting the soluble carbonaceous BODg (SCBODg)
from the SBOD5. When the effluent contains 15 mg/1 SBOD- and
an average of 15 mg/1 suspended solids (the later contributes
from 12 to 18 mg/1 of BOD to the effluent, averaging 15 mg/1),
the equivalent total BOD5 of the effluent is 30 mg/1 which
meets the current standards of secondary effluent. For this
effluent of 15 mg/1 SBOD,., Figure 1 shows that an influent SBOD,-
-loading up to 7.8 g/m -day can be applied. Assuming a value
of 10 mg/1 of SBOD- allowed in tertiary treatment effluents,
the influent SBOD,-. loading would be reduced to 3.5 g/m -day.
To compare the RBC performance in this study with that of
a similar study by Lagnese (1978), Table 9 in his report shows
that on the average an effluent SBODg of 16 mg/1 (range-11 to
21 mg/1) was obtained for a hydraulic loading of 0.04 m /m -
day and an influent SBOD
ing was therefore 7.57 g
v of 60 mg/1.
/m -day. Fig
The influent SBOD5 load-
Figure 1 of this study shows
an average of 15 mg/1 effluent SBOD5 at the same loading and
therefore the results of these two studies are comparable.
The removal of SBODj. versus influent SBODj- loading is
shown in Figure 2. It can be seen that the removal -in this
study (Curve A) was between 55 to 60 percent at an influent
SBOD5 loading of 4.88 g/m2-day or above. Below this loading
the removal percentage was lower. Data were taken from
Lagnese's study and plotted as Curve B for comparison. The
removal in Lagnese's study was higher by a few percentage points
possibly because the influent SBODj. concentration at 60 mg/1
was higher than the average of 25 mg/1 in this study. Also
shown are Curve C (data taken from Reh (1977)) and Autotrol Corp
Design Manual (1972) for RBC treatment of primary effluent
with higher SBODj- concentrations) and Curve D (Autotrol predic-
tion model curve, with low SBOD,- concentration at 25 mg/1) . The
evidence thus far indicates that the SBOD,- removal percentage
obtained, by RBC treatment decreases with decreasing influent
SBODj. concentrations.
Two reasons can be offered to explain the lower percentage
removal of SBODj... Trickling filters remove the more readily
degradable fraction of the wastewater, leaving the more resis-
tent fraction to the RBC. Also the low concentration of the
influent SBODj- could result in a depletion of BOD within the
biofilm, a phenomenon known as hindered mass transfer.
Famularo (1978) have found that low BOD concentrations in the
999
-------
liquid film coating the biofilm result in a depletion of BOD
within the biofilm, leading to inactive biofilm and inefficient
BOD removal.
Soluble Nitrogenous BOD5 Figure 1 shows that the SCBOD,-
of the RBC effluent was quite steady, varying from 2 to 8 mg /I
when the influent SBOD,- loading varied significantly from
0.49 to 7.8 g/m -day. Within the same range of SBOD,- loadings,
however, the effluent also contained from 1 to 13 mg /I SNBODg.
As a result the effluent SBODj. concentration was increased by
50 to 160 percent. Calculations using data presented in
Table 5 by Lagnese (1978) show that the effluent SBODc was
increased by nitrogenous BOD5 by 33.3 to 115 percent in the RBC
unit. If nitrogenous BOD5 can be excluded from consideration
in the secondary effluent standards, the allowable influent
SBOD,. loading can be increased to 61 g/m^-day by extrapolation
from Figure 1.
Effects of Low Temperature and Hydraulic Shock The RBC
pilot plant was tested in a low temperature series with the
wastewater temperature ranging from 5 to 10 C. Figure 3 shows
data points relating the effluent SBOD,- concentration to the
influent SBOD- loadings. Also presented in Figure 3 are data
points collected from the hydraulic shock series, usincr 120 to
220 percent of a moderate hydraulic load of 0.045 m /m^-day and
at a low influent SBOD,- concentration.
The two regression lines overlap and indicate that the
effluent SBOD,- increases very rapidly with increasing influent
SBOD,- loadings under the stress of low temperature or hydraulic
shoos. By keeping the same 15 mg /I of allowable effluent
SBOD,- in Figure 1 and 3 and compare the allowable influent
SBOD,- loadings, a correction factor of 0.41 is obtained (3.2
versus 7.8 g/m-day). It is worth noting that this correction
factor 0.41 was for the low temperature range 5 to 10 C and
for the RBC unit receiving low SBOD- concentrations and conse-
quently maintaining a low biological population.
Temperature coefficients 9 = 1.03, 1.04, and 1.06 are
often used respectively for activated sludge, trickling filter,
and lagoons. It is noted that the revised design manual by
Autotrol (1978) suggests a temperature correction factor of 0.76
for a low temperature of 7.2 C which was the average of the low
temperature series in this study. The temperature coefficient
9 = 1.05 applies to the condition from which Autotrol developed
their correction factor. A RBC unit, being a very efficient
heat transfer system according to Murphy (1975), and treating a
low BOD wastewater with a low biological population in the
present study, would be affected more by temperature similar to
lagoons. Because of the unique application of RBC in this stud^>
it is suggested that a temperature coefficient 6 = 1.07 be used,
or a correction factor of 0.42 which fits very closely the data
presented in this study.
1000
-------
Q- 01
2 S
00
I
*
E
-o
ro
O
to
4-1
C
o
OJ
3
rH
4-t
(3
•H
Q
O
CQ
O
CO
4J
C
0)
-P (T3
0) O
•H Q
X! O
to ffl
O
-------
(t
Q
V
O m
in O
O
00
in
a
•H
03
rH
rd
>
O
g
0)
Q
O
CQ
0)
r-t
•§
H
O
W
CJ
-------
J*
u
O
-C
U
°E 3
°. -*=
—J (/I
• •
in
I
CM
E
00
c
to
Q
O
CQ
0)
4-1
S-l
O
X!
(fl
O
(U
S-l
Q)
Q)
O
H
4J •
(0 en
fi
Q
O
PQ O
XI O
3 O
•H X!
O W
O
-H
0)
4-1 T3
IW >t
w x:
M
1003
-------
With hydraulic shock loadings in effect, the shortened
detention time (44 to 25 minutes) and the low F/M ratio
magnified the hindered mass transfer effect. It was observed
that the thickness of the biofilm on the RBC media was reduced
slightly. In other words, the RBC unit began to drift to
match the population to the low soluble BOD5 concentration.
Effects of Orgariic Shock To. simulate organic shocks,
loadings from 4.3 to 15.4 g SBOD5/m -day (124 to 444 percent
of the average load) were applied. The highest SBOD,. concen-
tration of the RBC influent was 72 mg /I which was considerably
higher than the average of 25 mg /I. Figure 4 shows that if
there was a discernable effect, the shock organic loading
actually improved the treatment performance slightly. The
result is not at all surprising since RBC is known to be able
to handle high BOD loadings successfully with good removal.
From 17 to 28.3 g total BOD^/m^-day can be removed without any
adverse effect according to Lagnese (1978) and Poon (1979).
Data reported by Reh (1977) show no adverse effect at 17 g
SBOD5/m^-day. Apparently the biofilm can keep up with the
higher loadings by removing the SBOD5 at the rate it enters
the RBC unit. It is worth noting that nitrification activity
is stronger at lower organic loadings and consequently the
masking effect of nitrogenous BOD is more pronounced. This
explains why a better performance might be obtained at higher
SBODg loadings.
Soluble MH0-N Remova'l and Effects of Low Temperature and
High Organic Loading Data shown in Figure 5 indicates that
the effluent soluble NEU-N concentration increases with in-
creasing influent soluble NH3~N concentration. A similar rela-
tionship was found by Antonie (1971). Other studies by Reh
(1977) and Reid (1974) also imply similar results. The rate
of nitrification decreases with temperature decrease throughout
the range from 30 to 5°C, according to EPA (1973) . Murphy
(1975) reported a good fit of nitrification rate data in their
RBC study, using the Arrhenius relationship over a temperature
range of 7 to 25 C. The present study was started in winter
and nitrification was not very well established throughout the
winter and early spring. The low temperature of the wastewater
ranged from 5 to 10 C (average 7.7°C) as opposed to the average
of 18.8°C for the normal condition in Figure 5. By comparing
curves A and B, it can be seen that the effluent soluble NH.,-N
concentration was approximately 38 percent higher than that
under normal condition for any influent loading within the
range indicated in Figure 5.
High organic shock was simulated in this study by applying
concurrently a high hydraulic loading and a moderate BODj. con-
centration. High hydraulic loadings from 0.18 to 0.245
m /m -day resulted in short detention times from 43 to 32
minutes. According to EPA (1973) 60 minutes would be required
1004
-------
\
\
\
£ £
\
X.
ro
"O
E
t>o
c
•o
u
O
fi I
t>*
-p
c
fi
-H
rtJ
in
Q
O
-------
e*
0-2
• ox
I
r*
-o
e«
o
I
Z
c
o
3
O
•H
JJ
2
-p
c
0)
u
fl
o
o
0)
O -H
01 13
C
0)
-H I
M-I
iw
0)
n
(U ^l
0) X5
S 3
-P H
(1) O
42 tn
04 -U
•H C
A QJ
tn d
fi rH
O H-4
•H C
-P •*
(0
H T3
0) fl
|/2ui-N-£HN
3
H
1006
-------
for nitrification, removing 25 to 30 mg/1 NH3-N to 2.5 to
5.0 mg/1 in a suspended growth system. For complete NEU-N
removal, 1.5 hours are required at pH 8.5 and 7 hours at
pH 6.5. Using a RBC as a secondary treatment unit, Antonie
(1971) reported that approximately 60 minutes would be required
to accomplish 85 percent NEU-N removal. At higher loadings a
RBC unit would need a minimum of 1.5-day detention time for
nearly complete NEU-N removal, according to Lue-Hing (1974).
Another unfavorable factor in this study was the higher BOD
concentration in the RBC influent. Bacteria that remove
carbonacous BOD grow at a much faster rate than nitrifiers.
Consequently they outgrow and reduce the nitrifiers in the
RBC unit since no biological sludge is recycled. The BOD
limit in RBC systems for successful nitrification is 30 mg /I
according to Antonie (1972), and 14 to 20 mg/1 according to
Antonie (1971). In this study series, the influent total BOD5
ranged from 35 to 131 mg /I, averaging 89 mg /I. The short
detention time coupled with the higher BOD concentration resul-
ted in the reduction of NEU-N removal as shown in Curve B,
Figure 5.
Nitrification Kinetics
The soluble NEU-N removal increased
with the influent soluble NEU-N loading as illustrated in
Figure 6. The removal was approximately 50 percent under nor-
mal condition (0 to 0.73 g/m2-day loading), and 83 percent at
higher loadings. At low temperatures or high organic shocks
(concurrent high hydraulic loading and moderate BOD concen-
tration) , the removal was reduced to 20 percent.
„ The removal term in Figure 6 expressed in g .soluble NEU-N/
m -day can be called the unit rate of nitrification, similar
to that defined by Murphy (1975) as mg TKN/m -hr. removal. The
temperature effect on the unit nitrification rate was (1-20%/
50%) = 60% reduction when the temperature was dropped from
18.8°C (average of k^ormal condition) to 7.7°C (average of low
temperature series). Estimated from Figure 4 of Murphy's work
(1975)f The temperature effect on the unit nitrification rate
was 55 percent reduction for the same temperature drop. For a
suspended growth system, a reduction of 63 percent of the unit
nitrification rate can be estimated from Figure 1-13 of the
EPA publication (1973). It thus appears that the temperature
effect on the unit nitrification rate is relatively the same
for both RBC and the suspended growth systems.
On several occasions NEUC1 was added to the RBC influent
to simulate a NEU-N overloading situation. Alkalinity in the
form of sodium carbonate was also added in addition to NH-Cl
in one occasion. The performance data in this series were
included in preparing Figure 7 covering a wide range of soluble
NEU-N loadings. Figure 7 suggests that S-shaped or logistic-S
curves best fit the data, expressed as:
1007
-------
-P
0)
s
(U
•o
m •
tn
c a
-------
CO
Cn
a
•H
w
3
O
•H
-P
<0
•H
i
n
33
2
0)
3
r-4
O
1009
-------
(1)
R =
max
1 + m-e
b-L
in which R = unit rate of nitrification* soluble NEUN removal
in Ib./lOOO ft2-day or g/m -day
max
m =
maximum unit rate of nitrification, same unit,
m and b are constants, where
"taax " Ro
R
b = - In ^
R (R
o max
n
R, (R - R )
1 max o
in which n =
Ro'Rl'R2
intervals between L ,
• o
loadings
soluble NH...-N removal at loadings L , L,
all in lb/1000 ft2-day or g/ni -day ° J
and L
2'
Based on the data presented in Figure 7 and for n = 0.3 lb/
1000 ft2-day (loadings 0.1, 0.4 and 0.7 lb/1000 ft2-day and
corresponding removals at 0.05, 0.285, and 0.516 lb/1000 ft2-da^
the developed constants are m = 10.28, and b = 7.84 (Curve A).
Consequently the logistic-S curve applicable to this study is:
R =
0.564
1 + 10.28e
-7.84(AL)
(2)
where AL = difference between any applied loading and 0.1 lb/
1000 ft -day.
A logistic-S equation was found to fit the nitrification data
in a natural stream, according to Wezernak (1968). The equa-
tion was developed with both nitrifying bacteria and substrate
exerting a concentration effect on the rate of nitrification.
Such equation would be most applicable to an environment when
both nitrifying organisms and NH -N concentrations are low.
The low soluble NEU-N loadings and the thin biofilm developed
on the RBC media in this study justify the use of this logistic-
S equation.
Curve B in Figure 7 was developed for conditions of low
temperature, high organic loading, or NH3-N overloading with
insufficient alkalinity in the influent. Curve B can be
expressed by the following equation:
1010
-------
R =
0.243
(3)
1 + 13.29e
-6.44(AL)
Comparing Curves A and B, there is a reduction of 57 to
75 percent of the unit nitrification rate within the range of
influent loadings of 0 to 1.1 Ib soluble NH3~N/1000 ft2-day.
While both kinetic equations 2 and 3 are considered adequate
in predicting the unit rate of nitrification, the constant
values m and b as well as the R values were developed
specifically for this study. O^Ser RBC plants of similar
application should develop specific values of m, b, and
for their own use.
max
Biofilm'Characteristics and Sludge Synthesis The biofilm
thickness was highly uneven at all 4 stages. Thickness of less
than 0.5 mm to 2.0 mm was observed at the first stage. The
difference of thickness was less significant at the other stages
but nevertheless clearly visible. The average growth of the
4 stages were 1.08, 0.79, 0.52 and 0.44 g/1000 cm disc surface
area.
The amount of sludge synthesis was determined indirectly
by substracting the 4th stage MLVSS concentration from the
influent MLVSS concentration and dividing the difference by
the SBOD5 removed. The calculated value is 0.28 g MLVSS/g of
SBODj. removed. This value is low compared to the average of
0.5 g/g BOD5 removed, indicating that RBC as an upgrading retro-
fit unit has a much lower biomass synthesis. This low synthesis
value reflects a possible long sludge retention time in the
RBC unit.
Sludge Retention Time Unlike a complete-mix suspended
growth system, the sludge retention time of the biofilm can
only be determined indirectly by setting up a mass balance
equation as follows:
change of
film biomass
4th
stage
MLSS
influent
MLSS
soluble
BOD
removed
x
growth
yield
coefficient
If the value of the change of film biomass calculated is
positive, solid sloughing takes place. A negative change means
increase of biofilm thickness and therefore no sloughing. The
growth yield coefficient was calculated previously to be
0.28 g MLVSS/g SBOD,., or 0.31 g MLSS/g SBOD5 using a ratio of
MLVSS/MLSS =0.9 for the average condition in this study. The
calculation using the data of a typical day is as follows:
= 31.5 mg/l-29mg/l-(2.4 mg/1 x 0.31)
= 1.75 mg/1 (sloughing)
biomass
(7/18/1978)
1011
-------
3 2
Since the hydraulic loading on the same day was 0.147 m /m -day
and there were 163.7 g of biomass in the biofilm,
3 2
Rate of sloughing = 1.75 mg/1 x 0.147 -2 - x 23.3 m
=5.95 g/day.
= 163'7
m -day
= 27'5
The average sludge retention time for the biofilm (163.7 g
biofilm in 125 liters of sewage in the RBC tankage, equivalent
to 1310 mg/1 MLSS) was 20.2-day in this study. The existing
MLSS concentration in the RBC tankage was on the average 24 mg/1
with a sludge retention time of 0.037 day (identical to the
average hydraulic retention time in the RBC unit) . The com-
bined retention was therefore:
Average RBC sludge _ (1310x20.2) + (24x0.037) = , g „ ,
retention time 1310 +24 ' Y
By comparison, the 19.8-day sludge retention time was
longer than that of activated sludge processes and shorter
than that of aerated lagoon processes. Since biological solid
synthesis is inversely proportional to sludge retention time,
it is then expected that sludge systhesis in the RBC unit should
be between 0.5 (for activated sludge processes) and 0.15 (for
aerated lagoons) . This is in agreement with the synthesis value
of 0.31 found in this study.
DESIGN EXAMPLE
Data developed from this study are used to design a retro-
fit RBC system to upgrade trickling filter effluents. Design
criteria are listed below.
(a) Acceptable secondary effluent standards:
15 mg/1 suspended solids
15 mg/1 soluble BOD5
Acceptable tertiary effluent standards:
1.0 mg/1 suspended solids Cdual .
9.0 mg/1 soluble BODj.
2.0 mg/1 soluble NH~N
filter)
(b) Effluent quality standards to be maintained throughout
the year without any special allowance for cold
weather conditions .
1012
-------
(c) High density media up to 132,000 ft surface area
for a 12-ft dia. module, 25-ft long shaft, and 4
stages.
(d) Use 20.4 m3/m2-day (500 gpd/ft2) overflow rate for
RBC clarifier design.
(e) Figure 3 alone for secondary treatment design, while
Figures 3 and 5 both for tertiary treatment design.
(f) Trickling filter effluent and loading:
SBOD|- concentration
Design flow 1.0 mgd (3785 m /day)
22 mg/1
83.24 kg/day (183.5
Ib/day)
8 mg/1
30.26 kg/day (66.7
Ib/day)
SBOD5 loading
Soluble NH.,-N cone.
Soluble NH^-N loading
Secondary
treatment
Tertiary
treatment
Disc surface
2
m
40,584
(436,860 ft)
67,300 o
(724,300 ft)
based on SBOD-
loading
97,870 9
(1,053,480 f t )
based on
soluble NH3-N
loading
Scale up
Factor
1.5
1.5
1.2
No. Of RBC
module
12 ft dia.
25 ft shaft
4-stage
4
6
8
Clarifier
Surface Area
2 2
m ft^
185.8 2000
use dual media
filter with or
without final
clarif iers
If RBC is to upgrade trickling filter effluents to meet
tertiary standards, nitrification controls the size selection
rather than BODj- removal since the allowable soluble NH.,-N
loading is much lower.
ACKNOWLEDGEMENT
This study was supported by funds provided by the U.S. Army
Construction Engineering Research Laboratory, Champaign,
Illinois.
1013
-------
REFERENCE
Lagnese, J.F., Evaluation of RBC used to upgrade municipal
plant to secondary standards, paper presented at the Tech. Conf,,
WPCA of Pa., Pittsburgh, April, 1978.
Reh, C.W., et. al., An approach to design of RBCs for treatment
of municipal wastewater, paper presented at the ASCE National
Environ. Engr. Conf., Nashville, July, 1977.
Sullivan, R.A., et. al., Upgrading existing waste treatment
facilities utilizing the BIO-SURF process, company report,
Autotrol Corp., May, 1978.
Autotrol Corp., Design manual for wastewater treatment systems,
1972.
Famularo, J., et. al.. Application of mass transfer to RBC,
J. WPCF., 50, 653, 1978.
Autotrol Corp., Design manual for wastewater treatment systems,
revised, 1978.
Murphy, K.L., et. al., Nitrogen control: design considerations
for supported growth systems, paper presented at the 48th Annual
Conf., WPCF., Miami Beach, Oct., 1975.
Poon, C.P.C., et. al.. Factors controlling RBC performance,
J. WPCF., 51, Part 1, 601, 1979.
Reid, Onebe, Allison, Wilcox & Associates, Inc., Nitrification
systems for NH3~N removal, company report, 1974.
EPA Tech. Tran. Seminar Publ., Nitrification and denitrifi-
cation facilities, Wastewater treatment, Aug., 1973.
Lu-Hing, C., et. al., Nitrification of a high ammonia content
sludge supernatent by use of rotating discs, 29th Proc. Purdue
Ind. Waste Conf., Purdue Univ., 245, 1974.
Antonie, R.L., et. al., Evaluation of a 0.5 mgd BIO-SURF
municipal wastewater treatment plant, paper presented at the
45th Annual Conf., WPCF, Atlanta, Oct., 1972.
Antonie, R.L., et. al., Application of rotating disc process
to municipal wastewater treatment, EPA report 17050 DAM 11/71,
35, 1971.
Wezernak, C.T., et. al., Evaluation of nitrification in streams,
J. San. Engr. Div., Proc. ASCE, 94, 883, Oct., 1968.
1014
-------
UPGRADING EXISTING WASTE TREATMENT FACILITIES
UTILIZING THE BIO-SURF PROCESS
By
Richard A. Sullivan
Manager, Process & Application Engineering
Autotrol Corporation
and
Robert J. Hynek
Manager, Process Verification & Pilot Plant Program
Autotrol Corporation
Introduction
A Targe number of treatment plants around the United States have been forced
to upgrade their effluent quality due to either hydraulic or organic over-
loading, or the establishment of more stringent effluent requirements by the
Federal Government.
Since 1973 Autotrol Corporation has been developing new and innovative methods
of using BIO-SURF, a Rotating Biological Contactor process, to upgrade exist-
ing treatment plants with minimal capital expenditures and little increase in
the operational costs of the existing facility.
This paper presents the R.B.C. process fundamentals and various techniques
Autotrol has utilized to upgrade existing primary clarification, trickling
filtration and activated sludge plants.
1015
-------
The Bio-Surf Process
The BIO-SURF process is a Rotating Biological Fixed-Film Reactqr. It employs
large diameter corrugated and flat polyethylene sheets thermally welded to-
gether to form a media network for biological development. These polyethylene
units are placed on a square steel shaft and the shaft and media are placed
in a tank until approximately 40% of the media surface is submerged below the
wastewater level. Micro-organisms naturally present in the wastewater
adhere and develop on the media until a biological mass, approximately 1/16"
thick, is developed on the media. The media is continually rotated alternate;-
ly exposing the biological mass to the atmosphere and then to the wastewater.
During the exposure of the biological mass to the atmosphere, oxygen present
in the atmosphere is absorbed into the biological cultures present. When the
biology rotates below the wastewater surface, the micro-organisms present on
the media will assimilate and remove the soluble carbonaceous and/or nitro-
genous impurities present in the waste.
The Rotating Biological Contactor, the BIO-SURF system, has been employed
widely in the United States over the past four years, and has proven itself
to be a viable and economic method for secondary treatment and nitrification
and more recently denitrification. The main advantage of the system are its
relatively simple operation and its significantly reduced power costs.
In the past the media has been rotated by a mechanical drive mechanism, i.e.
a motor and gear reducer. More recently, Autotrol Corporation has developed
an Air Drive system. This equipment consists of a series of cups or vanes
attached to the periphery of the BIO-SURF media. A supplemental air source,
installed below the media, releases air at low pressure into the varies.
Through the buoyant force of the air captured in the vanes the media is
rotated. Controlling the air flow will change the shaft rotational speeds,
thus providing a variable speed drive system.
It has been determined that the Air Drive system, AERO-SURF, has a number of
biological advantages over the mechanical drive system. This is due to:
1. Higher dissolved oxygen within the reactor liquid;
2. Thinner biological cultures and its associated thinner
anaerobic layer;
3. Elimination of sulfur bacteria and corresponding lower
removal kinetics; and '.
4. The combination of the above factors results in an
increase in removal efficiency per unit area.
Description of Upgrading Techniques
In employing the rotating biological contactor technology for expanding or
improving existing.treatment plant efficiency or capacity, a myriad of
applications have developed. These R.B.C- systems include series and
parallel operation with existing suspended growth and fixed growth systems
and direct placement of R.B.C. equipment within primary clarifier and
activated sludge basins. Various flow schemes are described in Chart 1 and
Figures 1-11.
1016
-------
EXISTING LOCATION
- Edgewater, NJ
fr M .
cd cu o
S 'H •
^
o
M
H
U
rH
PH
%
pLj
0
Pq1
O
§
/ -rl |3 <4H pq
5H CU -rl
PH la !-4 ^ &
cd i — i
5-i cd i — 1 M-4
cu u cu o
4-) O 5-1
IH 4-1 >, 3 4-J
<3 5-4 00 d
5-4 Cd -rl CU
• 0 13 pq B
CJ -rl d CU
• 5H O CU Cj
pq PH O CU cd
CU Cn rH
p4 t<3 en * — ' PM
^~ •N - S~\
<-l CN
rH
pr"]
to
0
M
la
C_3
W
E-i
O
la
t-H
3
PH
O
iS
cd
'o 0
B
OJ CU
0) Pi
5-1
oo en
CU 3
Q 0
CU
CU CJ
en cd
cd d
(U O
5-1 rQ
0 5-4
d Cd
M O
4-1
^H
£-4
M
i-J
M
U
pq
Pq
O
w
[>H
EH
d
CU
£=J
4-1
cd
CU
5-4
EH
K**»
5-1
cd 4-1
S d
•rl Cd
5-1 rH
PM PM
rH
cu
CU rH
rH rH
14-1 Cd
MH M
cd *~~^ cd
pq CO PL,
CO
d cd -H
Cd 4-1 CN
E-I d
o cu cj
!>> N 5-4 •
5-4 -H 3 pq
cd 5-4 00 •
B O -H Pi
•H EU pq
5H CU
PM .^"1 cu cj
4-1 0) Cd
d vH en rH
M |5 *• — ' PM
,^^
r-l
4-J
•H
CJ
cd
P.
cj
U
•H
rH
3
cfl
m
*T3
^
frj
,
4-1
•H
•H
o
cd
pq
00
Cj
•rt -"
4-J
CO
•H
X
w
(4_I
0
CO
fa
o
4J
CJ
cd
CU
Pd
B
•H
pq
rH
CO -H
0) pq
•H
cu a)
en X
•H
d Pq
•H
00
• d
CJ -H
. 4-1
pq co
Pi ^
cu
O ,d
cd 4-i
rH -H
PM IS
/~N
CN
rH
cd
MH >
O O
B
cu pi
5-j
00 CO
CU 3
Q 0
CU
CU O
co cd
cd d
cu o
5-1 rP
0 H
d cd
M CJ
CO
JH"
CU
4-J
rH
•H
PH
00
d
•rl
rH
^
O
•H
5-1
E-i
».
(U
•H
- North Huntingdon, PA
1m
CO -H
Q) pq
•rl
5-4 13
4-1
o
rH Cd
cd cu
cj Pi
•H
oo ,d
O 4-1
rH fe
o o
•H 5-4
PQ O
5-1
o
4J CO
cj 0)
cd -H
a) r-, 5-1
Pi o cu
•-H cn
a)
oo ts d
-d -H
r^ f""'"^
rH
cn aj o
rl •
13 3 pq
CU 00 •
4-1 -H pi
cd pq
> cu
•rl 0) O
4-1 CU Cd
CJ Cn rH
•H
4-J
CJ
,
CU
•H
- Cadillac, MI
CU
00
3
rH
cn
13
cu d
4-1 'H
cd cd
> rl
4-1
O CO
4-1
•H
O
cd
cd
d
o
•H
4_1
cd
a
•H
•H
£-1
4-1
•H
&
CO CO
3 d
o o
•rl -H
5H 4-J
cd cd
> 0
•H
CO 1
4J
•H
O
CU
oo
3
0 -rl
rH CJ
i d
•rl
4J
o •
cd O
m •
3 •
cn pi
^^
>
4-1
•H
CJ
cd co
ft 3
cd o
u a)
CJ
o cd
•rl d
rH 0
Cd rO
t_j t |
13 cd
>! O
p~{
CH
cu o
CO
cd cu
cu
CU
•H
.d
CJ
0
O -H
§1 i
cd
Pi CJ
1017
-------
r
RAW >
SEWAGE
PRIMARY
TREATMENT
BIO-SURF
PROCESS
SECONDARY
CLARIFIER
.FINAL
* EFFLUENT
Figure 1
RAW
SEWAGE
PRIMARY
CLARIFIER
Figure 2
SLUDGE
PUMP
PLASTIC. INTERSTAGE
COVER S\ A BAFFLES
INFLUENT
FROM
PRE-TREATHENT
SECONDARY CLARIFICATION UNDER FALSE BOTTOM CW]
Figure 3
1018
-------
RAW
SEWAGE
PRIMARY
TREATMENT
BIO-SURF
PROCESS
SECONDARY
CLARIFIER
Figure 4
RAW
SEWAGE
PRIMARY
CLARIFIER
BIO-SURF
PROCESS
SECONDARY
CLARIFIER
Figure 5
RAW
WASTEWATER
PRIMARY
TREATMENT
EXISTING
SECONDARY
CLARIFIER
_» FINAL
EFFLUENT
BIO-SURF
ANOXIC
DENITRIFICATION
Figure 6
1019
-------
§<
0)0
as
Z"-
OE
01
8
U-W
tE(/>
I3UJ
WO
Qg
mo.
CO
z
cc
01
1
f~
0)
60
•H
so
>a
!-.=
U-W
DCW
IDUJ
D)
C
•6
g
C
TO
Q.
0)
f
"w
CO
a)
a.
>-UI
5S
15
I
E
39VM3S MVd
o
TO
o>
TO
"5
O
TO
0)
£
O
CO
30VAA3S MVH
1020
-------
g
§
AOZ
QZUJ
m0
53
W
it
i<
£d
I
3OVM3S MVU
1021
-------
Application of the R.B.C. process to upgrading existing facilities is part-
icularly attractive due to the following system features of the R.B.C. design.
1. Low hydraulic head loss - the head drop through R.B.C. tankage is less
than 6" and therefore facilitates its utilization in existing treatment
flow trains in various configurations.
2. Shallox* excavations and tankage requirements - tankage is only 6' deep
and therefore can be easily employed in locations where high groundwater
and bedrock conditions are encountered. The shallow tankage requirements
also facilitate employment within existing structures such as primary
clarifiers and aeration basins.
3. Minimal effects on Solids Loading to secondary clarifiers - R.B.C.
systems MLSS concentrations are extremely low and do not vary with
seasonal fluctuations as compared to other attached growth systems or
suspended growth systems. The impact of solids loading on downstream unit
processes such as secondary clarifiers or final filters is minimal.
Higher hydraulic rates and/or elimination of solid separation unit
processes have proven feasible.
4. Modular Construction - Since the R.B.C. units are constructed in a
variety of unit lengths and diameters which can be applied in a module-
type approach, site-land availability presents minimal problems in employ-?
ment of the R.B.C. concept.
Selection of a particular flow scheme from the various alternatives depends
upon:
1. Site conditions (i.e.: land availability, hydraulic gradient of existing
facility and soils conditions);
2. Sizing and maximum beneficial utilization of existing plant unit
processes;
3. Economic consideration of tank modifications and/or construction;
4. Potential ease of future upgrading.
Upgrading Existing Primary Treatment Facilities
The concept of upgrading existing primary to secondary treatment without
adding secondary clarifiers was first demonstrated by Autotrol Corporation
at Edgewater, New Jersey. The Edgewater Sewage Treatment Plant has provided
primary treatment as a mechanism for removing an average of 30% BOD and 52%
suspended solids from a mixture of domestic and industrial wastewaters. The
plant was built in 1954 and expanded in 1958 to a design capacity of 2.5 MGD.
The plant site is extremely limited comprising only 1.5 acres of usable land
area. The major original unit operations include pumping, comminution,
grit collection, primary sedimentation and chlorination. Treated wastewater
then flows to the Hudson River via an outfall discharge pipeline. Sludge is
processed with two anaerobic sludge digesters and two vacuum filters. The
treatment plant is currently processing its 1958 design capacity.
In 1972 the Interstate Sanitation Commission and the Environmental Protection
Agency required that the plant •meet secondary treatment standards, i.e. effluent
monthly averages of'30 mg/1 BOD and suspended solids, or a minimum of 85%
BOD and suspended solids removal.
1022
-------
A concept for converting this plant to secondary treatment level was developed
by Autotrol Corporation employing the BIO-SURF process. One of five existing
primary treatment tanks, 70-feet long, 14-feet 4-inches wide, with an average
side-water depth of. 9-feet, was segregated and modified. This involved the
division of the primary tank by an intermediate horizontal baffle; thus, the
primary tank was divided into upper and lower chambers. The BIO-SURF units
were placed in the upper zone to perform biological treatment of the influent
wastewater, and original sludge collection mechanisms remained in the lower
zone to collect suspended biological cultures which had sloughed from the
R.B.C. units. Minimal primary treatment was employed prior to wastewater
entry into the biological zone. The minimal primary treatment consists of
grit removal and a much reduced primary clarifier zone with an overflow rate
of 5,000 gpd/ft^ at average flow-. Startup occurred in 1974.
Performance for the year 1976 is shown in Table 1. This summary shows that
the modified tank is currently treating 500,000 gpd with an average effluent
of 29 mg/1 BODcj. This corresponds to an average effluent BOD and suspended
solids removal of 89%.
Thus, by making minimal modifications to primary treatment facilities, the
BIO-SURF process was able to convert an existing primary tank to a secondary
treatment facility. The major modifications required to perform this function
are:
1. The installation of a horizontal baffle, or false floor,
separating the existing primary tank into separate zones;
2. The incorporation of Rotating Biological Contactors in the
upper zone for removal of soluble carbonaceous or nitrogenous
impurities in the wastewater. These Rotating Biological
Contactors must be made to a specified length to conform to
the width of the existing primary treatment tank;
3. Modifications to the sludge scraping' mechanisms in the
existing primary tank by re-directing the collector chain and
putting in chain guides in the subjacent sedimentation zone.
Upgrading Existing Trickling Filter Plants
The BIO-SURF process has also been used widely in upgrading existing trickling
filter plants. General flow diagrams for utilizing this concept would place
the R.B.C. either in series operation with the trickling filter to upgrade
the plant to meet, more stringent effluent requirements, or in parallel opera-
tion with the trickling filter to upgrade the plant for increased flow or
organic loading beyond the original design capabilities. (See Figures 4-7).
Presently in operation Autotrol Corporation has sewage treatment plants where
the Rotating Biological Contactor system has been employed between the
currently operating trickling filter and secondary clarifiers. Birdsboro,
Pennsylvania, and North/Huntingdon, Pennsylvania are two examples.
1023
-------
TABLE I
RESULTS OF PLANT SCALE TESTS - EDGEWATER, NEW JERSEY
PRIMARY TANK UPGRADING
MONTH YEAR
JAN 1976
FEE "
MAR "
APRIL "
MAY "
JUNE "
JULY "
AUG "
SEPT "
OCT "
NOV "
DEC "
FLOW INF. BOD
0.560
N.R.
N.R.
0.564
0.537
0.480
0.540
0.395
N.R.
0.446
0.433
0.400
128
122
179
240
182
283
325
300
367
324
283
302
INF. S.S.
106
120
152
194
185
272
336
266
313
237
231
310
EFF. BOD
24
24
31
28
27
26
33
39
27
34
21
33
EFF S.S.
23
20
24
24
25
20
22
53
20
23
18
26
, % BOD
81
80
83
88
85
91
90
87
93
90
93
89
% S.S.
78
83
84
88
86
93
94
80 ;
94
90
92
89
AVG.
0.484 253
227
29
25
89
89
1024
-------
North Huntingdon, Pennsylvania, initially installed a rock trickling filter
plant designed for 50% removal of BOD at a. flow of 1.5 MGD. The entire plant
was comprised of primary clarifiers, high, rate rock trickling filters,
secondary clarifiers, chlorine contact tanks, anaerobic digesters and sludge
drying beds. Due to increased hydraulic flow and more stringent effluent
requirements, it was necessary to increase the removal capability to 85%
BOD removal at a flow of 1.75 MGD; thus the upgrade consisted of simultaneous
increases in hydraulic capacity and treatment efficiency. The BIO-SURF
system installed to meet the new conditions consists of four 20-foot long
R.B.C. shafts providing 305,000 ft^ of effective biological surface.
Although the R.B.C. currently operates in parallel with the rock filter system,
the plant was designed to function either in series or in parallel operation
with the original flow train. During new plant evaluations, it was observed
that superior performance with regard to specific removal rate was attained
when the system operated in the parallel mode (Figure 4). It is believed
that ,the reason for better removals during the parallel operation mode was
due to the nature of the substrate being more biodegradable in substance.
Also, the higher substrate concentration increased the diffusion gradient to
the microorganism. As a consequence the plant was able to significantly
upgrade the effluent quality to meet a total effluent BOD of 29 mg/1. Results
of performance tests are shown in detail in Table II, for series operation
and parallel operation.
The City of Birdsboro, Pennsylvania installation is similar to that of North
Huntingdon, with the exception that the existing trickling filter was initially
designed to provide greater percentages of BOD reduction. Results from the
trickling filter showed effluent BOD and suspended solids of approximately
56 mg/1 of BOD and suspended solids. The City of Birdsboro wished to upgrade
their facility to meet effluents of approximately 25 mg/1 Ultimate Oxygen
Demand which required oxidation of both carbonaceous and nitrogenous impurities.
Eight BIO-SURF units were installed between the existing trickling filter and
secondary clarifier to upgrade the original system to meet the desired effluent
quality. These units are currently operating at an overall hydraulic loading
of 1.2 gpd/ft^. Typical results are shown in Table III. In summary, then,
data in Table II and Table III confirm the capability of the BIO-SURF process
to upgrade plants where the trickling filter was initially designed as a
roughing process, as at the North Huntingdon facility, or to provide nitrifica-
tion to overloaded secondary plants, as at Birdsboro, Pennsylvania.
Upgrading Existing Activated Sludge Plants
1. Carbonaceous Reduction - As mentioned in the previous discussion, the
R.B.C. process can be installed prior to existing suspended growth
reactors to reduce the soluble BOD load to the suspended growth reactor.
An example of such an installation is Crawfordsville, Indiana, which has
a flow diagram similar to (Figure 7). Four shafts, which can operate as
either single or'multi-stage systems, follow the primary clarifier and
precede the suspended growth reactor to reduce the influent soluble BOD
concentration to the 25 . to 50 mg/1 range. This greatly reduces the
soluble load to the suspended growth system and since the solids generated
1025 .
-------
TABLE II
SOLUBLE BOD,. RESULTS - NORTH HUNTINGDON, PENNSYLVANIA
HYDR.
LOAD
(GPD/SF)
TEST B
3.1
3.0
3.1
3.2
3.1
3.2
3.1
3.1
3.1
AVERAGE
INFL
TOTAL
(MG/L)
(SERIES,
98
90
70
98
102
85
68
70
88
85
TEST C (PARALLEL,
3.3
3.1
3.1
3.7
3.0
4.0
6.5
3.2
5.2
AVERAGE
140
123
119
121
118
121
60
140
107
117
_>
FOR
RBC PERFORMANCE
. BOD5 to BIO-DISC
SOLUBLE
(MG/L)
SUMMER)
69
63
41
67
81
57
46
54
60
60
i
SUMMER)
86
87
86
87
87
79
30
99
67
79
RATIO S/T
(%)
70
70
59
68
79
67
68
77
68
70
61
71
72
72
74
65
50
71
63
67
EFFL.
TOTAL
(MG/L)
24
25
21
34
35
36
23
25
34
29
26
—
25
23
31
34
22
32
36
29
BOD5 FROM
SOLUBLE
(MG/L)
14
13
11
19
21
21
15
12
15
16
20
—
12
12
15
20
11
14
21
16
BIO-DISC
RATIO S/T
(%)
58
52
52
56
60
58
65
48
44
55
77
—
48
52
48
59
50
44
58
55
1026
-------
Autotrol Corporation
Reference No. 256-9
Table III
Laboratory Analysis Summary
Birdsboro, Pennsylvania
BODQ
o
BODg (Carbon)
BODT
CODg
CODT
D.O.
HC03
%-Tot.
-Sol.
NH3-N
N03-N
Org . N-Sol
-Insol.
pH
SS-Tot.
-Vol.
% VSS
0.622 MGD, 1.13 gpd/ft^
October 11, 1979
P.C.E.
35
36
85
80
175
_
125
-
15 .'2
8.5
7.0
6.7
-
7.1
63
49
(77.8)
, 19°C
mg/1
T.F.E.
21
19
56
70
130
-
110
-
11.4
7.3
9.1
4.1
-
7.1
76
38
(50.0)
S.C.E.
5
6
12
35
60
-
55
-
4.9
1.6
20.0
3.0
-
7.2
40
10
(25.0)
1027
-------
by the R.B.C. system is relatively small, ha.s little effect on solids
inventory within the suspended growth reactor. Thus, by employment of
the R.B.C. system in this, manner the soluble load to the suspended growth
reactor is significantly reduced allowing for increases in flow or degree
of treatment. It is also anticipated that performance will demonstrate
an over-all increase in stability against shock load conditions.
2. Nitrification - The BIO-SURF process has also been employed to upgrade
existing activated sludge plants to provide nitrification. Operating
experience at such plants include Cadillac, Michigan, Hinkley, Ohio and
Sarasota, Florida.
A flow diagram of the Sarasota, Florida installation is shown in Table IV.
The BIO-SURF process is generally employed after the secondary clarifiers
to provide additional BOD removal as well as nitrification. The system,
because of its low solids generation during the nitrification stage, can
discharge its effluent directly into the receiving water body or if
higher degrees of suspended solids removal are required into tertiary
filters. The latest data from this plant are shown in Table V. As can
be seen by these results, extremely low BOD5 and ammonia concentration
have been obtained from the Sarasota, Florida installation.
SURFACT PROCESS
In 1976 a new concept for upgrading existing suspended growth reactors using.
Rotating Biological contactors was developed by the City of Philadelphia :
Water Department. This concept is called "The SURFACT Process" because it is
a biological system resulting from the combination of rotating biological
fixed film surfaces (SURF) and an activated sludge system (ACT). The purpose
of merging the Rotating Biological Contactor with an activated sludge system
is to provide additional removal of BOD or nitrification, and to provide process
stability and flexibility. Studies since 1976 indicate also that the R.B.C.
and the activated sludge system act, for the most part, independently of one1
another and BOD removals of 90% are possible in the same tankage where conven-
tional activated sludge had obtained minimal treatment efficiencies.
The SURFACT concept involves only a slight modification of existing aeration
tanks and installation of R.B.C. units in the upper regions of these tanks.
By so doing, the Rotating Biological Contactors greatly increase the effective
mixed liquor suspended solids concentration, (as much as 33% to 50%), and
increases the removal capability in the aeration system without changing
existing tank volumes, return sludge sizing, blower sizing, or any other aspect
of the existing facilities. Also, since the equivalent increase in mixed
liquor suspended solids is held on a fixed, film reactor, the solids loading to
the secondary clarifier is unchanged from the original design, thereby avoiding
major modifications to the secpndary clarifier. Rotation of an air drive
R.B.C. can be provided by the roll of the existing aeration tank or by a
supplemental air header placed below the media. Mechanical drive is another
alternative.
As a result of preliminary testing conducted by the City of Philadelphia in
1974 and 1976, utilizing the AERO-SURF and the SURFACT approach, it was
1028
-------
Autotrol Corporation
TABLE V
Gulf Gate, AWT, Sarasota, Florida
Year Month
1977 Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
]978 Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
1979 Jan
Feb
Mar
Apr
May
Jun
Jul
Average
Flow MGD
.864
.860
.853
.792
.760
.783
.786
.900
1.133
.900
.890
1.000
1.060
1.120
1.140
.937
.883
.852
.969
1.080
.922
.905
.926
.980
1.230
1.170
1.180
1.080
.990
.973
.974
.964
+ .128
BODftng/l)*
Infl.
177
198
195
190
198
183
217
207
233
272
310
251
219
242
164
242
248
208
169
153
189
191
210
199
185
208
210
231
208
201
190
210
+33
Effl.
1.3
2.7
2.2
1.0
0.7
0.5
0.6
0.7
1.1
1.0
1.0
1.4
1.3
1.9
2.0
1.0
1.1
0.9
1.7
1.2
0.7
0.9
1.1
0..9
0.8
0.9
2.9
1.5
0.8
0.6
0.7
1.2
+.6
SS (mg/1) Total NCme/l)
Infl.
278
279
274
288
310
282
338
245
242
293
346
251
212
220
186
246
244
217
178
196
204
212
243
205
266
273
254
282
299
278
230
254
+42
Effl.
1.1
1.5
1.6
1.7
1.0
1.0
0.5
0.6
0.8
1.3
0.8
0.2
0.4
1.1
1.5
0.3
0.5
0.5
0.7
0.6
0.5
0.4
0.4
0.4
0.3
0.6
1.0
0.5
0.4
0.6
0.8
0.8
+0.4
Effl.
1.4
2.3
2.0
1.0
1.4
1.7
1.8
2.1
0.9
1.2
1.3
0.6
0.9
3.3
2.4
1.4
1.5
1.4
1.7
1.7
1.1
1.2
1.2
1.4
1.7
1.4
3.9
2.0
1.3
1.2
1.1
1.6
+0.7
*Total
1029
-------
Table IV
GAG Utilities, Sarasota, Florida.
''GulfGate & South Gate STP
Raw Wastewater (Lift station)
1
Grit Removal
Return
Activated
Sludge
Aerobic
Digestion
Sludge "*=
Disposal
Equalization
Aeration
Filtration
BIOSURF
Nitrification
Dentrification
Final
Clarification
Final Effluent
Chlorination
1030
-------
decided that a plant-scale SURFACT demonstration facility should, be evaluated
at the Philadelphia Northeast Plant. This treatment plant consists of grit
removal, primary clarification, twenty existing aeration tanks and secondary
clarifiers, and anaerobic digestion to process the.resultant sludge from these
unit operations. The plant-scale demonstration incorporated twenty-two AERO-
SURF shafts into two parallel aeration basins. Each of these shafts contain
media 12-foot in diameter with. 100,000 square feet of effective surface area.
The aeration basin was originally .des-igned as a two-pass modified channel.
Each pass was modified with eleven air-driven R.B.C. shafts. To receive the
effluent from the SURFACT aeration tank a secondary clarifier at the existing
facility was completely isolated by the installation of new return sludge
pumps.
The evaluation at the Philadelphia plant began in October, 1977 and, after
mechanical start-up and process stabalization, collection of data was started
shortly after January 1, 1978. The results of these tests are shown on
Table VI.
It can be seen from Phase III and Phase IV of the operation that the SURFACT
system has been able to consistently achieve BOD removals of 88% and 90%
soluble BOD removal in excess of 95%, and suspended solids removal in excess
of 80% at loadings of 74-pounds and 63-pounds of BOD per 1,000 cubic feet of
aeration capacity. More recent evaluations demonstrate that loadings of
100# BOD5/1000 cubic feet provided an effluent BOD quality of 20-25 mg/1 BOD.
The F/M ratio necessary to achieve these results is approximately 0.5 with
sludge age of approximately 2.7 to 3.6 days.
After 18 months of plant scale evaluation, it was concluded that the SURFACT
Process was the most economical alternative for the upgrading of the
Northeast Philadelphia facility. A total of approximately 180 shafts will be
installed in the twenty passes.
In summary, the BIO-SURF, AERO-SURF and the SURFACT systems have proven
complete flexibility in adapting to any existing plant's flow scheme in order
to upgrade the facility. In addition, these processes have been able to
perform this function in a consistently economical manner.
1031
-------
CQ —
o as
8£
0.0
1032
-------
TABLE VI
SURFACT
% REMOVAL
- SS
- BOD
- SBOD
EFFLUENT, MG/L
- SS
- BOD
- SBOD
TOTAL % REMOVAL
- SS
- BOD
- SOR, GPD/Ft.2
PHASE II
75
82
91
25
32
6
88
83
720
- WOR, 103GPD/Ft.2 82
DT, HOURS
- PRIMARY
- AERATOR
- FINAL
0.8
3.0
2.1
FINAL, SOLIDS LOADING 10
lbs/day/ft.2
LBS BODA/1,000 Ft.3 74
F/M
- W/RBC 0.5
- W/0 RBC 1.0
SRT
-W/RBC 3.5
-W/0 RBC 1.4
Ft.3, AIR/Gal. 1.0
Ft.3, AIR/lb BODA 600
LB 02A/LB BODR 0.5
KWH/LB BODR 0.6
PHASE III
84
88
95
26
22
5
87
90
720
82
0.8
3.2
2.2
17
74
0.3
0.6
4.*2
2.7
1.1
700
0.5
0.6
PHASE IV
83
90
95
23
20
6
90
600
68
0.8
3.6
2.4
18
63
0.25
0.4
5.6
3.6
1.4
840
0.6
0.6
1033
-------
-------
ROTATING BIOLOGICAL CONTACTOR PROCESS FOR
SECONDARY TREATMENT AND NITRIFICATION
FOLLOWING A TRICKLING FILTER
By
Roy D. Miller, Ph.D., P.E.
Charles I. Noss
Arnold Ostrofsky
Robert S. Ryczak
US Army Medical Bioengineering Research and Development Laboratory
Fort Detrick, Frederick, MD 21701
Introduction
The practice of biological nitrification has been spurred by enactment of
environmental Legislation as well as increased activities of regulatory agencies.
Discharge limitations have been placed for ammonia—nitrogen in the National
Pollutant Discharge Elimination System (NPDES) with Public Law 92-500 (currently
the Clean Water Act of 1977, Public Law 95-217). Department of the Army instal-
lations must apply for and receive discharge permits for wastewater effluents.
In many cases, existing NPDES permits impose ammonia-nitrogen limitations on
Army-produced wastewaters, and it is anticipated that future permits will contain
comparable, if not more stringent, discharge limitations.
Existing NPDES permits also impose secondary standards for biochemical oxygen
demand of less than 30 mg/L, with the limiting concentration based on the con-
ditions of the receiving waters. Biological treatment processes, which allow
nitrification, decrease the BOD level resulting from the carbonaceous and nitro-
genous materials. As oxidation of carbonaceous organics and ammonia-nitrogen
occur within the same process, stringent discharge requirements for both param-
eters are frequently obtained.
1035
-------
The purpose of this study was to evaluate performance of the rotating bio-
logical contactor (RBC) for BOD,, reduction and ammonia-nitrogen removal where
the RBC followed a trickling filter. This treatment scheme was selected for
evaluation based on several factors. It has potential utilization in existing
facilities, a relatively low energy and operating cost, simplicity of operation,
and flexibility with respect to expansion or upgrading of treatment facilities.
The pilot RBC in this study received effluent from a high-rate trickling filter
that performed at less than secondary effluent standards (i.e., trickling filter
secondary effluent BOD_ was routinely greater than 30 mg/L). Ammonia-nitrogen
and BOD(. removal efficxencies across the EBC were determined at varied hydraulic
loading rates ranging from 1.5 to 4.3 gpd/sq. ft. of RBC surface area. Other
parameters which were varied included the influent pH levels, organic loading
rates, and suspended solids concentrations. Performance was evaluated during
both summer and winter conditions.
Background
US Army wastewater treatment systems consist primarily of trickling filters .
as secondary treatment processes, a few activated sludge systems, and several
extended aeration package plants. The .Army has unique situations in which
treatment plants are often flow underloaded due to decreases in the size of the
Army population during peacetime. Also, consolidation of activities from sev-
eral installations to only one installation, summer training of Reserve and
National Guard troops, and maneuvers of troops can cause drastic seasonal
changes in loadings on treatment plants. Drastic diurnal changes in loadings ;
can be caused by civilian work forces that contribute wastes during normal work-
ing hours, but not at other times. In addition, size of Army treatment plants
(0.1 to 5.0 MGD), dictate the need for simplicity of operation and maintenance.
Therefore, advanced wastewater treatment (AWT) techniques applied to municipal
wastewater treatment systems may not be applicable to Army treatment systems.
The RBC process was developed during the past two decades. In the last few
years, the RBC process has been chosen over trickling filter and activated sludge
systems in some designs. Reportedly, the RBC process allows for a longer and
more intense contact time than the trickling filter process, thereby increasing
BOD,, removal. Continuous sloughing of biomass is a feature of an RBC process,
whereas excess growth on a trickling filter can cause pondings. A high degree of
of treatment within an RBC can be achieved by proper design of aeration capa-
bilities and effluent retention without unnecessary recycling of effluent to
maintain minimum wetting rates. The RBC process is less likely to be affected
by organic shock loads or a hydraulic surge as would the activated sludge pro-
cess, which relies on sludge recycling to maintain the proper food-to-micro-
organisms ratio. Lastly, because of the high density of the biological solids
in the RBC process, purportedly higher flow rates can be used in the secondary
clarifier, yet maintain a comparable or thicker sludge than the activated sludge
process.
LITERATURE REVIEW
RBC Treatment Process
The RBC process consists of a series of plastic disks of which 40 percent
of the surface area is immersed in wastewater (see Figure 1). As the disks
1036
-------
rotate, the entire media surface develops a culture of microbiological organisms.
The organisms adhere and multiply to form a uniform growth referred to as a
fixed-film. The biomass supported by the plastic media picks up a thin layer of
nutrient laden water as it rotates through the wastewater. The film of water
trickles over the microorganisms which remove dissolved organics and oxygen.
The rotation of the media through the wastewater not only allows for aeration
and mixed liquor, but also provides shear forces which cause sloughing of excess
growth.
RBC units commonly operate in series with the number of units depending on
the organic and/or hydraulic load to be treated. The function of the first
stages is to remove organic material, with subsequent stages removing ammonia
when nitrification is used to meet effluent NHL-N standards. Nitrification does
not begin until the soluble-BCD,, level and corresponding large populations of
heterotrophic organisms have been adequately lowered. The actual reason that
heterotrophs and autotrophic nitrifiers do not co-exist in equal quantities
throughout successive RBC stages is not clearly understood, but it is reported
that the activities of the two populations do not occur simultaneously. ' '
The amount of nitrification achieved has been correlated to the hydraulic load-
ing of the system, usually expressed as the volume of wastewater applied to a
square measure of surface area per day. One to 4 gpd/sq. ft. have often been
used as standard^loading rates for pilot plants and full scale wastewater treat-
ment facilities. ' ' The change in hydraulic load also changes the organic
load as more food is introduced to the active component of .the waste treatment
system.
Sufficient dissolved oxygen (DO) must be available in the wastewater within
the RBC system to insure adequate treatment for BOD_ removal and nitrification.
Wastewater DO levels of 1 to 2 mg/L are generally considered to be the minimum
requirement to avoid DO limiting conditions. Frequently, RBC systems have been
designed to provide oxygen mans transfer via disk rotation through -the waste-
water and air. However, in some cases, this has been considered a shortcoming
of the process since supplemental oxygen must sometimes be provided to prevent
DO limiting conditions.
Oxygen Demand of Wastewater
*
A major criterion used to determine the extent of pollution of receiving
waters is the measurement of oxygen required for the stabilization of organic
matter present in the system. The total amount of oxygen necessary to stabi-
lize a waste is referred to as the oxygen demand. The ultimate oxygen demand
includes not only the amount of oxygen required to stabilize oxidizable carbon-
aceous materials, but also that which is required to microbially transform
ammonia-nitrogen to nitrate-nitrogen. For untreated domestic sewage there is
little oxygen demand by nitrifier populations for the first 8 days of stabili-
zation in the BOD test. Therefore, the BOD test is normally considered as
representing the oxygen demand of carbonaceous material. However, total BOD_
can be poor indication of treatment where a significant population of nitrify-
ing bacteria are present. For sewage that has received secondary treatment and
nitrification, conversion of ammonia to nitrate in the BOD bottle may signifi-
cantly increase the BOD_ measurement and erroneously indicate a lesser degree
of treatment than that actually received.
1037
-------
The purpose of biological treatment relative to carbonaceous material is
the conversion of soluble organics to particulate bacteria. However, the
unfiltered BOD,, test represents a measure of soluble as well as insoluble
organic matter and NH_-N oxidation. Biological treatment need not be applied
to removal of colloidal and suspended organics. Suspended solids that contri-
bute to oxygen demand can be removed by physical-chemical processes-such as
gravity settling and filtration. The practical consequence is that optimal
treatment for removal of oxygen demand may be removal of suspended solids and
not biological treatment. The use of filtered and unfiltered BOD- tests should
indicate relative fractions of oxygen demand as originating from soluble or ;
particulate material. In addition, the filtered BOD,, test should not undergo
nitrification, because initial nitrifying populations in the BOD bottle would
be reduced to insignificant levels and the BOD bottle is subsequently seeded
with raw sewage (i.e., heterotrophic bacteria). Thus, the unfiltered BOD,.
test can be an unreliable parameter from which to judge biological treatment
performance.
Nitrogen Control
The principle of biologically induced nitrogen removal in wastewater treat-
ment facilities is based on the activity of populations of autotrophic nitri-
fying and denitrifying bacteria and their capability to sequentially oxidize
and reduce nitrogen from ammonia to nitrate to nitrogen gas. Nitrification is
the oxidation of NH_-N to nitrate, and denitrification is the reduction of
nitrate to nitrogen gas. Different types of microorganisms are required for
each action. The extent of their use in wastewater treatment depends upon the
end objective. Nitrification is used to control wastewater effluent levels of
ammonia, but both nitrification and denitrification must be used to control
total nitrogen levels in wastewater effluents. Although process technology for
ammonia-nitrogen removal includes breakpoint chlorination, ammonia stripping,
ion exchange, and nitrification/denitrification, this paper deals only with
nitrification.
In addition to nitrification/denitrification, microorganisms other than the
nitrifiers and denitrifiers require nitrogen for growth. The amount of nitrogen
assimilated during oxidation of carbonaceous material has been generally placed
at 5 percent of the oxygen demand (i.e., BOD to N = 20 to 1). The consequence
is two fold: (1) nitrogen must be present for biological oxidation of carbon-
aceous material, and (2) removal of ammonia-nitrogen during biological treatment
of wastewaters may be due to assimilation, not necessarily due to nitrification.
The importance of nitrogen control in wastewater effluents is its impact on
receiving waters. As ammonia becomes oxidized to nitrate, the dissolved oxygen
level of water is decreased. Ammonia-nitrogen at concentrations of 0.25 to
0.30 mg/L are lethal to fish within 14 to 21 days. Nitrate is readily avail-
able for assimilation by plant life, causing algal blooms when present in too
large a quantity. Also, nitrate can cause methemoglobinemia in infants when
contaminated water is used as a drinking water supply.
1038
-------
Nitrification
The two microbial genera usually associated with nitrification are
Nitrosomonas and Nitrobacter. Both genera of organisms are autotrophic nitri-
fying bacteria indicating that energy for growth is derived from the oxidation
of inorganic nitrogen. The oxidation of ammonia to nitrate is a two step pro-
cess requiring both organisms for the conversion. Nitrosomonas transforms
ammonia to nitrite while Nitrobacter further oxidizes nitrite to nitrate. The
overall oxidation of ammonia by these organisms is given by the following
equation:
- Nitrosomonas
Nitrobacter
(1)
As ammonia is oxidized, carbonate is utilized, as nitrate formation occurs,
carbonic acid is produced. This microbiologically induced change in the car-
bonate buffering system results in the destruction of alkalinity at a rate of
7.1 mg' (as CaCO ) per mg of ammonia oxidized. As the nitrification process
reduces the alkalinity and increases the carbonic acid concentration, the pH
of the wastewater may drop as low as pH 6.0, and adversely impact the rate of
nitrification. This decrease in pH can be minimized by aeration to strip COp
from the wastewater, or by insuring the presence of excess alkalinity.
Primary environmental conditions for optimizing rates of nitrification are
pH and temperature. The reported pH optima cover a wide range, but the con-
sensus is that as the pH decreases,, the rate of nitrification declines. Sawyer,
et al., and Engel and Alexander have reported pH optima for nitrification
between 8.0 and 9.0, and 7.0 and 9.0, respectively. Painter has stated that,-
nitrification processes cease at or below pH 6.3 to 6.7. Poduska and Andrews
have shown that abrupt changes in pH from 7.2 to 5.8 markedly reduced the
ammonia oxidation by nitrifiers while the reversal in pH restored the original
nitrification rate.
Temperature optima for nitrification are generally reported by various
authors at about 30°C with a range of 28-35°C. ' ' ' ' ' Temperature
influences heterotrophic and autotrophic microorganisms, thereby affecting
secondary treatment and nitrification efficiencies. The nitrification rate is
more temperature sensitive than the rates for organic removal. Nitrification
rates decrease about 50 percent for each 10 C drop in wastewater temperature
below about 30°C. For example, the nitrification rate at 10 C would be about
half that of 20 C. Secondary treatment efficiency is less likely to be affected
by temperature changes, probably due to microbial population diversity and other
system constraints. Organic removal rates for fixed-film processes should
decrease about 25 percent for each 10 C drop in wastewater temperature below
about 30°C. For example, the rate of biological activity in a trickling filter
process, at 10°C would be about 75 percent of that of 20°C. However, the actual
temperature effect on a biological process is probably characteristic only of
that system.
MATERIALS AND METHODS
The pilot RBC consisted of four compartments in series. The 0.5 meter
plastic disks provided 250 sq. ft. of surface area for microbial attachment.
1039
-------
The disks are rotated through liquor at 13 rpm with 40 percent of the fixed-
film submerged at any point. A schematic of the pilot RBC process is shown in
Figure 1.
Pilot studies used domestic wastewater from the Fort Detrick housing area.
The wastewater was shredded by a grinder pump and pumped into a 250 gallon
equalization tank which was periodically replenished through a float level
switch control. This enabled a relatively constant flow of partially settled
wastewater to be pumped into the primary clarifier. The wastewater flowed by
gravity through the primary clarifier and was then pumped to a high-rate trick-
ling filter. The trickling filter contained a 2-inch irregular stone media
with effective media depth of 4 ft. and 2.25 sq. ft. of filter surface. Efflu-
ent was collected at the bottom of the filter in a wet well and pumped to the
secondary clarifier. This pretreatment scheme was used to simulate an existing.
trickling filter plant. The RBC followed trickling filter treatment and was
used as an upgrading technique for additional BOD,, reduction and nitrification
(ammonia-nitrogen removal).
Performance of the RBC was evaluated over a two-year period without temper-
ature control in order to determine both summer and winter efficiencies for
BOD- reduction and ammonia-nitrogen removal. RBC performance was evaluated at
normal and elevated pH levels of pH 7.1 and pH 8.7 for the RBC influent (trick-
ling filter secondary effluent).
The RBC influent flow was varied to provide hydraulic loading ranging from
1.5 to 4.3 gpd/sq. ft. of RBC surface area. Treatment efficiency of the trick-
ling filter was varied to provide different levels of BOD,, in the RBC influent.
Levels of suspended solids in the RBC influent were varied by having secondary '
settling before the RBC and having no settling between the trickling filter and
the RBC.
Sampling and Analyses
Sample points included 24-hour composites of primary influent, trickling
filter effluent (RBC influent) and RBC effluent. Analyses of test parameters
in the four stages of the RBC unit were determined from grab samples.
Measurements of flow, temperature, dissolved oxygen, pH, suspended solids,
total organic carbon, and ammonia-nitrogen were made daily. Alkalinity and
chemical oxygen demand (COD) measurements were made four times per week. Bio-
chemical oxygen demand (BOD,.) analyses were performed twice weekly.
Total organic carbon measurements were made on a Beckman Model 915 Total
Organic Carbon Analyzer. Ammonia-nitrogen concentrations were measured with an
Orion specific ion electrode. Dissolved oxygen and BOD determinations were
made using a Delta Scientific Model 2110 Dissolved Oxygen Meter and probe.
Chemical oxygen demand, total Kjeldahl nitrogen, and phosphorus analyses were
made using a Technicon Auto Analyzer II system according to Technicon methods.
Filtered samples were filtered through fiberglass filters. All other analyses
were performed according to Standard Methods.
1040
-------
RESULTS AND DISCUSSION
Hydraulic Loading
The RBC was initially evaluated for secondary treatment and nitrification
following a high—rate trickling filter during the fall, winter, and spring of
1976-1977 at hydraulic loadings of 1.5, 2.0, 3.0, and 4.0 gpd/sq. ft. of RBC
surface area. Secondary treatment efficiency was evaluated using filtered and
unfiltered BOD,., and filtered TOG. Nitrification efficiency was evaluated
using NH -N, TKN, NO~/NO.,-N, and alkalinity. Suspended solids, pH, temperature,
and dissolved oxygen levels within the RBC stages were monitored. Sample points
included raw wastewater, trickling filter effluent and RBC effluent. Limited
sampling was conducted within stages of the RBC to evaluate progression of
treatment.
Secondary treatment performance of the RBC is shown by data in Table 1 in
which the RBC removed 20 mg/L of TOC. Table 1 also shows environmental con-
ditions of the study at 1.5 gpd/sq. ft. The wastewater temperature averaged
13.4 C during this period in January and February of 1977, and the median pH
for RBC influent was 8.6. Natural (biological) recarbonation and nitrification
within the RBC process depressed the pH to about 7.5 in the RBC secondary
effluent.
Careful consideration should be used in evaluating treatment efficiency of
the biological processes based on unfiltered BOD . The oxygen demand of sus-
pended solids makes unfiltered BOD levels higher than filtered BOD- levels
for raw wastewater and trickling filter secondary effluent. For highly treated
RBC secondary effluent, nitrification could add to the oxygen demand of sus-
pended solids to make unfiltered BOD,- greater than filtered BOD-. Because of
the effects of suspended solids and nitrification on oxygen demand, unfiltered
BOD- should not be used alone to evaluate biological treatment efficiency. The
difference between values of unfiltered and filtered BOD- must be recognized
because NPDES permits only require analysis for unfiltered BOD- without distin-
guishing the soluble loading from the particulate loading of a biological
treatment process. The difference between unfiltered and filtered BOD_ values
indicate that pilot studies did not optimize removal of suspended solids. This
fact? is also shown by the suspended solids data in Table 1. Data for ammonia-
nitrogen in Table 1 indicate that nitrification was essentially complete for
the RBC secondary effluent and could contribute little to the unfiltered BOD-.
Therefore, reductions in unfiltered BOD values should be attempted by removal
of suspended solids and not through biological treatment.
Figure 2 shows ammonia removal through biological activity in the trickling
filter and RBC processes, where the RBC hydraulic loading was 1.5 gpd/sq. ft.
The RBC secondary effluent ammonia level of 0.5 mg/L indicates this to be a
highly nitrified wastewater. As discussed in the Literature Review section,
removal of nitrogen can be due to both bacteriological assimilation and nitri-
fication. Generally, nitrogen would be assimilated in the amount of about 5
percent of the quantity of carbonaceous oxygen demand utilized, or about 5 per-
cent of the change in filtered BOD- across the treatment process. This nitro-
gen loss/removal should show up in the total Kjeldahl nitrogen (TKN) test
(e.g., organic nitrogen plus ammonia) but may not decrease the ammonia level
because microorganisms could assimilate nitrogen directly from organic com-
1041
-------
pounds or convert organic nitrogen to ammonia. Removal of nitrogen beyond as-
similation should be due to nitrification. The amount of nitrification can be
determined from NO--N/NO--N analyses, since ammonia is oxidized to nitrate.
Nitrite and nitrite-nitrate levels are shown in Table 1. The alkalinity test :
can also be used as a check on the amount of nitrification, because 7.1 mg/L
of alkalinity are destroyed per 1 mg/L of NH -N oxidized. Alkalinity levels
are shown in Table 1.
Figure 2 shows the RBC influent and RBC secondary effluent levels of ammonia
to be 18 and 0.5 mg/L, respectively. This shows that 17.5 mg/L of ammonia-
nitrogen were removed across the RBC process, from both assimilation and nitri-
fication. Unfortunately, TKN analyses were not available for this particular
study because of equipment problems, and the total nitrogen story cannot be told
here. Table 1 shows that 41 mg/L of filtered BOD,, were utilized within the RBC
process, therefore, 5 percent of 41, or about 2.0 mg/L, of nitrogen would be
assimilated by heterotrophic organisms; additional nitrogen would be assimilated
by autotrophic organisms. Remaining ammonia would be available for nitrifica-
tion. Table 1 shows that 14 mg/L of NO,,-N/NO«-N was formed and represents
ammonia oxidized by nitrification. Table 1 also shows 100 mg/L of alkalinity
destroyed across the RBC process, and this corresponds to 15 mg/L of ammonia
nitrified (e.g., 100 divided by 7.1). In summary, of the nitrogen present in
RBC influent, an excess of 2.0 mg/L was assimilated and about 15 mg/L was oxi-
dized. This agrees well with the actual levels of ammonia observed. Ammonia
removal was essentially complete in RBC secondary effluent at 1.5 gpd/sq. ft. .
even though temperature (13.4°C) was poor for nitrification; however, pH had re-
mained near optimum for nitrification. The median pH levels were 7.5 in RBC
secondary effluent.
Figure 3 contains results of further studies of secondary treatment by the
RBC process following a trickling filter. Supporting data are contained in
Table 1 for hydraulic loadings of 2.0, 3.0, and 4.0 gpd/sq.-ft., respectively.
Secondary standards were achieved at all hydraulic loadings as shown by filtered
BOD_ levels for RBC secondary effluent at all hydraulic loadings. Filtered BOD^
utilized by biological activity within the RBC process (e.g., RBC influent
values minus RBC secondary effluent values) were 36, 39 and 38 mg/L at 2.0, 3.0
and 4.0 gpd/sq. ft., respectively. This corresponds to an organic removal rate
of about 1.3 Ibs. filtered BOD5/1000 sq. ft.-day at the 4.0 gpd/sq. ft. hydrau-
lic loadings. The organic removal rates should increase if the organic loading
were higher. Likewise, filtered TOG utilized within the RBC process were 21,
17, and 17 mg/L at 2.0, 3.0 and 4.0 gpd/sq. ft., respectively. The ratio of
filtered BOD5 utilized to filtered TOG utilized averaged 2.0:1. Wastewater tem-
perature increased slightly at each increased hydraulic loading as shown in
Table 1. RBC influent pH levels remained at about pH 8.5. It can be concluded
from Figure 3 that an existing trickling filter can be upgraded to secondary
standards by use of an RBC process at hydraulic loadings up to 4.0 gpd/sq. ft.
and at organic removal rates in excess of 1.3 Ibs. filtered BOD^/IOOO sq. ft.-
day.
Figure 2 shows RBC treatment performance for nitrification at hydraulic
loadings of 1.5, 2.0, 3.0 and 4.0 gpd/sq. ft. RBC secondary effluent was highly
nitrified at 1.5 gpd/sq. ft. and at 2.0 gpd/sq. ft. for temperatures above 13 C
(Table 1). Nitrification was essentially complete at 3.0 and 4.0 gpd/sq. ft.
Wastewater temperature was relatively poor for nitrification at 1.5 and 2.0
1042
-------
gpd/sq. ft. (e.g., 13.4 and 15.1 C, respectively) as shown by data in Table 1.
However, pH levels were near optimum for nitrification at those loadings. RBC
secondary effluent pH was about pH 7.5. Wastewater temperature and pH levels
were relatively good for nitrification at hydraulic loadings of 3.0 and 4.0
gpd/sq. ft. as shown by data in Table 1. Continued good performance by the RBC
process at the higher hydraulic loadings is probably attributable to environ-
mental conditions of wastewater temperature and pH.
Combined use of Figures 2 and 3 and Table 1 with supporting data are neces-
sary to fully evaluate nitrification across the RBC process. Figure 3 showed
that about 40 mg/L of filtered BOD was utilized across the RBC process at each
hydraulic loading; this represents about 5 percent of 40 or 2.0 mg/L of nitrogen
assimilated by heterotrophic organisms. Additional nitrogen would be assimi-
lated by autotrophic nitrifying organisms. Data in Table 1 show nitrate levels
present in RBC secondary effluent, and show the alkalinity destroyed by the RBC
process to compare favorably with ammonia levels. For example, at the hydraulic
loading of 4.0 gpd/sq. ft., 63 mg/L of alkalinity were destroyed, corresponding
to 8.9 mg/L of ammonia-nitrogen removal by nitrification. To balance nitrogen
losses at 4.0 gpd/sq. ft., 8.9 mg/L would be added to 2.0 or more mg/L, showing
total nitrogen assimilated and nitrified to be about 11.0 mg/L plus 2.0 mg/L in
RBC secondary effluent, or 10 mg/L ammonia-nitrogen removed at 4.0 gpd/sq. ft.
Organic nitrogen was unaccounted for and would represent a fraction of the
nitrogen assimilated or present as ammonia in the RBC secondary effluent.
The effect of wastewater temperature on secondary treatment and nitrifica-
tion across the RBC process can be seen by use of Figure 4. This Figure shows
the temperature profile and RBC secondary effluent ammonia levels during the
period December 15, 1976, to January 11, 1977, at a hydraulic loading of 2.0
gpd/sq. ft. The RBC secondary effluent was highly nitrified during the period
of December 15—28 as evidenced by an average ammonia—nitrogen level of 1.0 mg/L.
Ammonia nitrogen levels in the RBC secondary effluent increased to about
6.0 mg/L during the period of January 3-11, after wastewater temperatures had
dropped to about 10 C. Decreased ammonia removal lagged behind this temperature
drop, but the effect was substantial in lessening the degree of nitrification.
Both organic removal and nitrification rates should have been affected by the
drop in wastewater temperature. Decreased nitrification rates should be about
half for each 10 C drop, while organic removal rates (secondary treatment)
should be about 75 percent for each 10 C drop. Combined temperature effects
from decreased organic removal and nitrification rates resulted in the increased
ammonia levels shown in Figure 4. Treatment progression within the RBC process
should show those effects, since organic removal occurs in the initial stages
and nitrification in the later stages.
Table 2 shows the progression of nitrification within the RBC process at the
1.5 gpd/sq. ft. hydraulic loading. Grab samples were used for data presented in
Table 2 and should represent a valid picture of RBC stage-by-stage performance,
because flow equalization had occurred within treatment processes before RBC
treatment. Nitrification was occurring in Stage 1 and subsequent stages as evi-
denced by formation of nitrate and removal of ammonia. It is also evident that
nitrification and organic removal were occurring simultaneously within the ini-
tial stages because the filtered BOD,, and TOG levels dropped across the RBC pro-
cess (Table 1), even though organic removal data was not taken within RBC stages.
1043
-------
Ammonia was removed in greater quantity than nitrate formed; the difference,
about 3.0 mg/L, is attributed to assimilation by both heterotrophs and auto-
trophs. Dissolved oxygen levels remained ideal for nitrification in all RBC
stages. The pH drop from stage to stage was caused by biological recarbonation
plus destruction of alkalinity by nitrification.
Figure 5 shows progression of nitrification within the RBC process at var,i-
ous hydraulic loadings of 1.5, 2.0, 3.0 and 4.0 gpd/sq. ft. Rates of nitrifi-
cation appeared to reach a maximum level at each loading, and then decreased in
the last RBC stage. This points out several issues of ammonia-nitrogen removal.
First, nitrification rate is influenced by organics removal, and organic
removal seems to take precedence. Ammonia removal in Stage 1 shows this
influence by organic removal. Ammonia removal in Stages 2 and 3 indicate that
nitrification rates are relatively unaffected by ammonia concentrations above
about 5 mg/L, while State 4 results indicate that nitrification rate has de-
creased due to limited concentrations of ammonia. That is, nitrification rates
approximate a zero-order reaction above about 2.5 mg/L and first-order reaction
below 2.5 mg/L. This explains why stringent NPDES permit limitations of
<2.0 mg/L makes compliance difficult.
No Intermediate Settling
Further studies of the RBC process for secondary treatment and nitrifica-
tion following a trickling filter were made at hydraulic loading of about 3.0
gpd/sq. ft. without intermediate settling between the trickling filter and RBC
process. These studies were conducted to evaluate the possibility of upgrad-
ing existing trickling filter plants by use of RBC processes while maintaining
existing secondary clarifiers to settle out RBC effluent. This scheme should
result in savings on total wastewater treatment plant upgrade by making maxir
mum use of existing facilities. These studies were conducted with RBC influ-
ent pH elevated to about pH 8.0 to create good environmental conditions for
nitrification, and separately with neutral pH for the RBC influent to evaluate
pH effect on nitrification.
Table 3 contains data for secondary treatment performance of the trickling
filter and RBC treatment processes at 3.0 gpd/sq. ft. without intermediate
settling between the trickling filter and RBC process. Filtered BOD,, values
show that the RBC secondary effluent met secondary standards. It can be con-
cluded from data in Table 3 that the RBC process can be used to upgrade an
existing trickling filter to secondary standards without intermediate settling
between the trickling filter and RBC process.
Progression of treatment within the RBC stages at 3.0 gpd/sq. ft. following
a trickling filter without intermediate settling is shown in Figure 6. This
shows that rates of nitrification were relatively low in Stage 1 and higher in
Stages 2 and 3, and again lower in Stage 4. This is similar to results of
Figure 5 where intermediate settling had been used before the RBC process.
Again, these rates of nitrification show that organic removal affects nitrifi-
cation rates in the initial RBC stages where secondary treatment takes place
and nitrification rates are highest at ammonia levels above about 5 mg/L
where the rate of reaction is reported to be zero-order. Nitrification rates
decrease when the ammonia level drops below about 5 mg/L so that the latter
1044
-------
stage of treatment is less efficient, and stringent NPDES permits become more
difficult and costly to meet. It can be concluded from use of Figures 5 and 6
that nitrification across an RBC process is not greatly influenced by inter-
mediate settling between the trickling filter and RBC process.
The effect of pH on nitrification across the RBC process is shown in
Figure 7, where the RBC process followed the trickling filter without intermed-
iate settling. RBC secondary effluent ammonia levels increased from about
1.5 mg/L with elevated pH adjustment before the RBC process to about 10 mg/L
at neutral pH levels. Good nitrification within the RBC process continued for
about 4 days after elevated pH changed to neutral levels. Then, poor nitrifi-
cation continued for the duration of the study at neutral pH. Figure 7 shows
the benefit of chemical feed for pH and alkalinity control for nitrification.
Reportedly, elevated pH primarily increases the ratio of nitrifiers to hetero-
trophs in the system. This favors simultaneous nitrification and organic
removal in initial RBC stages and higher nitrification rates in latter stages
for elevated pH levels, but not for neutral pH.
Further study of the pH effect on nitrification was conducted by repeating
the neutral pH conditions for RBC performance without intermediate settling.
Data are presented in Table 3. Figure 8 shows ammonia levels for RBC influent
and RBC secondary effluent. RBC performance for nitrification was relatively
good with elevated pH as indicated by an effluent ammonia level of 1.6 mg/L,
but poor with neutral pH as indicated by effluent ammonia levels of 12.7 mg/L
and 8.8 mg/L for the two studies. Supporting data in Table 3 for alkalinity
and TKN confirm the poor nitrification at neutral pH levels; 66 mg/L of alka-
linity were destroyed indicating 9.3 mg/L of ammonia nitrified at neutral pH
values.
Data in Table 3 for filtered BOD,, and TOC indicate that secondary efflaent
standards were achieved by the RBC process at neutral pH. RBC-secondary efflu-
ent filtered BOD,, values were 8 and 10 mg/L for the two studies, and 21 and
22 mg/L of TOC were removed by the RBC process. This secondary treatment per-
formance by the RBC process at both neutral pH levels compares favorably with
RBC performance at elevated pH levels of about pH 8.0 shown by data in Table 3.
It can be concluded that elevated pH levels of about pH 8.0 do not affect
secondary treatment performance (BOD,, removal) of the RBC process.
Organic Loading
Further studies of the RBC process for secondary treatment and nitrifica-
tion were made at hydraulic loadings of about 3.0 gpd/sq. ft. to evaluate the
effect of organic loading on secondary treatment and nitrification. These
studies evaluated the effect of partial and complete secondary treatment by an
existing trickling filter, versus no secondary treatment, prior to an RBC
process for secondary treatment and/or nitrification. These studies should
help answer the question arising in design upgrades, "Should existing facilities
be abandoned, or used?" The result should be that increasing organic loading
increases effluent ammonia levels unless additional RBC surface is available
at higher organic loadings, because substantial organic removal must occur
before nitrification begins. ' Therefore, the use of existing facilities
for organic removal should substantially reduce surface area requirements of
1045
-------
the RBC process for further secondary treatment plus nitrification. For ex-
ample, Antonie showed that a 6 MGD, mechanical drive RBC plant for secondary
treatment to 20 mg/L BOD5 would cost $3 million (in 1976) while such a plant
that provided hoth secondary treatment and nitrification to 1.0 mg/L ammonia .
would cost $4.5 million. This points out the value of existing facilities, :
such as trickling filters, that provide organic removal, during design upgrade.
In the example cited, an existing trickling filter providing secondary treat-
ment could be worth up to $3 million.
Earlier studies at elevated pH levels using intermediate settling between
the trickling filter and RBC process were based on hydraulic loadings. However,
the organic and ammonia concentrations remained essentially constant, so that
increases in hydraulic loading also increased organic loading. Results of these
studies at elevated pH levels showed that increased hydraulic (and organic)
loadings had little effect on effluent ammonia levels (Figure 4). However,
temperature increased with increases in loadings during those studies and
partially accounted for additional nitrification at higher organic loadings.
In addition, it was observed that simultaneous organic removal and nitrification
occurred in the same stages of the RBC process at elevated pH levels, and helped
account for the good nitrification at high loadings.
Studies were conducted at neutral and elevated pH levels and two temperature
values to further evaluate the organic loading effect on secondary treatment and
effluent ammonia levels. The RBC process was evaluated at 3.0 gpd/sq. ft.
following a trickling filter at wastewater temperatures of about 25 C and 17 C
at pH levels of 7.1 and 8.5 respectively. Also, the RBC process was evaluated
without prior biological treatment at wastewater temperatures of about .25°C
and pH 8.8. Results of RBC treatment performance for secondary treatment are
shown in Figure 9, and nitrification results are shown in Figure 10. Supporting
data are contained in Table 4.
Figure 9 shows that RBC secondary effluent met secondary standards in all
cases. However, this does not imply that organic removal within the RBC process
was the same in all cases.
Comparable filtered BOD removal was observed when the trickling filter
preceded the RBC at 17 and 25°C. However, considerably more NH_-N was removed
at 17 C, not as a result of temperature, but because the influent pH had been
raised to pH 8.5. When the trickling filter was not used, applying an in-
creased organic load to the RBC, a greater reduction in filtered BOD occurred.
However, the ammonia removed was decreased when the increased organic load
was applied to the RBC.
Figure 10 shows that nitrification was near completion at 25 C with lower
organic loading on the RBC process; whereas, nitrification was relatively poor
at the lower temperatures with higher organic loading to the RBC process.
These results indicate higher organic loadings, which in turn increase effluent
ammonia levels, because substantial organic removal must occur before nitrifi-
cation begins. Lower temperatures seemed to have a similar effect on effluent
ammonia levels, like that observed with increased organic loading.
1046
-------
SUMMARY AND CONCLUSIONS
The RBC process performed effectively as a system to upgrade an existing
trickling filter plant when used for secondary treatment over a range of hydrau-
lic loads from 1.5 to 4.3 gpd/sq. ft. Secondary treatment was less noticeably
affected by wastewater temperature than was nitrification. Results also indi—
cated that elevating the RBC influent pH to about pH 8.0 increased the nitri-
fication rate. The increase in nitrification rate decreases the RBC surface
area requirements, resulting in less costly wastewater treatment plant upgrades.
An existing trickling filter plant can be upgraded to secondary standards
for BOD by using a rotating biological contactor (RBC) process for organic
removal without intermediate settling between the trickling filter and RBC
process.
An existing trickling filter plant can be upgraded for ammonia removal by
using an RBC process for nitrification without intermediate settling between
the trickling filter and RBC process.
The use of existing trickling filter plants for partial or complete
secondary treatment decreases the surface area requirements of RBC's for
secondary treatment and/or nitrification, resulting in less costly upgradings.
Chemical feed, or elevation of pH levels to about pH 8.0 in RBC influent,
increases the rate of nitrification and decreases RBC surface area require-
ments, resulting in less costly improvement and more consistent attainment
of ammonia removal.
Wastewater temperature affects both secondary treatment and nitrification
within the RBC process, with nitrification rates being more sensitive to low
temperatures than organic removal.
ACKNOWLEDGMENTS
The authors wish to extend special thanks to Mr. Kenneth A. Bartgis,
Engineering Technician; SSG Felix B. Legaspi, Jr., Engineering Assistant;
and Charles F. Harrison, Physical Science Technician, for their technical
assistance for the duration of the project.
DISCLAIMER
"The opinions or assertions contained herein are the private views of the
authors and are not to be construed as official or as reflecting the views of
the Department of the Army or the Department of Defense."
1047
-------
LITERATURE CITED
1. Wild, H.E., Jr., C.N. Sawyer, and T.C. McMahon, "Factors Affecting
Nitrification Kinetics," _J. Water Pollut. Control Fed., 43:1845 (1971).
2. Steiner, C.G., "The New Rotating Disk Process," Advance Publ. Copy (1978).
3. Hockenbury, M.R., G.T. Daigger, and C.P. Grady, Jr., "Factors Affecting
Nitrification," J. c»f _the Env. Engr. Div.: 9 (1977).
4. Antonie, R.L., D.L. Kluge, and J.H. Mielke, "Evaluation of a Rotating Disk
Wastewater Treatment Plant," J_. Water Pollut. Control Fed. 46:498 (1974).
5. Antonie, R.L., "Rotating Biological Contactor for Secondary Wastewater
Treatment," Presented at Gulp/Wesner/Gulp WWT Seminar, South Lake Tahoe, NV
(1976).
6. Sawyer, G.N. and P.L. McCarthy, "Chemistry for Sanitary Engineers,"
McGraw-Hill Book Company, New York, NY (1967).
7. Busch, A.W., Aerobic Biological Treatment of Waste Waters, Oligodynamics
Press, Houston, TX (1971).
8. Clark, J.W. and Viessman, Jr., "Water Supply and Pollution Control,"
International Textbook Company, Scranton, PA (1965).
9. Smart, G., "The Effect of Ammonia on Gill Structures of Rainbow Trout,"
J. Fish Biol., 8:471-475 (1976).
10. Gruener, N. and H.I. Shuval, "Toxicology of Nitrites," Environmental
Quality and Safety, 2:219-229 (1973).
U- Process Design Manual for: Nitrogen Control, U.S. EPA Tech. Transfer (1975).
12. Sawyer, C.N., H.E. Wild, and T.C. McMahon, "Nitrification and Denitrifi-
cation Facilities Wastewater Treatment," U.S. EPA, Tech. Transfer (1973).
13. Engel, M.S. and M. Alexander, "Growth and Metabolism of IT. europeae,"
J.. Bacteriol.. 76:217 (1959).
14. Painter, H.A., "A Review of the Literature on Inorganic Nitrogen Metabo-
lism in Microorganism, " Water Res_. , 4:393 (1970).
15. Poduska, R.A. and J.F. Andrews, "Dynamics of Nitrification in the Activated
Sludge Process," 29th Industrial Wastes Conf., Purdue Univ., IN (1974).
16. Buswell, A.M., T. Shiota, N. Lawrence, and I.V. Meter, "Laboratory Studies
on the Kinetics of the Growth of Nitrosomonas with Relation to the Nitrification
Phase of the BOD Test," Appl. Microbiol., 2:21 (1954).
1048
-------
17. Deppe, K. and H. Engel, "Untersuchungen uber die Temperaturabhangigkeit
der Nitratbildung durch Nitrobacter winogradskii Buch. bei ungehemmtem und
gehemmtem Wachstum," Zentbl. Bakt. Parasitk de II. 113, 561-568 (I960).
18. Laudelout, H. and L. vanTichelen, "Kinetics of the Nitrite Oxidation by
Nitrobacter winogradskii," J. Bacteriol., 79:392-42 (1960).
19. Balakrishnan, S. and W.W. Eckenfelder, "Nitrogen Relationship in Biological
Waste Treatment Processes - II, Nitrification in Trickling Filters," Water Res.,
3:167 (1969).
20. Haug, R.T. and P.L. McCarty, "Nitrification with the Submerged Filter,"
U.S. EPA Grant #17010EPM (1971).
21. Haung, C.S. and N.E. Hopson, "Temperature and pH Effect on the Biological
Nitrification Process," Presented at the New York WPCA, New York, NY (1974).
/
22. Standard Methods for the Examination of Water and Wastewater, 14th Edition,
American Public Health Association. American Water Works Association, Water
Pollution Control Federation (1976).
1049
-------
o
*
-*
^^
e
4J
•
«)
•CJ O
CO CO
en
1
ca
H
"3
O
O
M
3 ^
C
is
1
>
OH
H
OO vo
•vfocMincMmincoHH
voo .COHH « • « «
iHCM T-HCMOOVDOO
H H
CT\ CM in t — I in co in I in crv
CO • H CM <1* • • •
H H r* o -*
H
mmcMCMHCMCOHH-*
-vTCTi «COOH • • " '
m H CM oo m o o
H H
CM in
ocor^mcor^m I CMVO
CM«HCOCM« « •
o H r-. o m
i~n
OO •
> d
cd
S -H
Cd TlJ
CJ d)
a a
• •
Cd rQ
1050
-------
TABLE 2. NITRIFICATION WITHIN THE RBC TREATMENT PROCESS
AT VARIED HYDRAULIC LOADING RATES
RBC Stage Number
pH
Parameter
DO
I.
II.
III.
IV.
1.5
1
2
3
4
2.0
1
2.
3
4
3.0
1
2
3
4
4.0
1
2
3
4
gpd/sq. ft.
9.4
3.6
1.8
0.9
gpd/sq. ft.
13.4
8.4
3.7
1.8
gpd/sq. ft.
11.1
9.7
6.2
3.6
gpd/sq. ft.
12.7
9.6
5.3
2.3
8.0
7.5
7.4
7.5
7.9
7.7
7.6
7.5
8.2
7.7
7.5
7.4
7.9
7.4
7.2
7.2
4.6
5.1
5.9
7.2
3.5
4.1
4.6
5.0
5.1
4.9
5.1
5.3
4.5
4.2
4.2
4.8
13.2
12.7
12.4
12.2
12.7
12.2
11.9
11.7
17.5
17.3
17.2
17.3
20.2
19.8
19.7
19.7
4.68
9.60
14.56
17.24
_
—
—
-
4.70
5.73
9.15
13.60
_
—
-
"™
1051
-------
ca
s
g
w
1
H
g
0
e
g
»
•u
«•<
ft
D*
W
t)
&
OJJ
O
CO
Si
8
§
S
o
cc
w
p*
§
£?
B
"
o
S
•
co
I
x-v H-4
r^« IIH
Is* W
c\
rH O
COKl'Sf • «O • •
i-H o\ r^ CM r~. •*
H CM CM
X-N
1
1-H
•--^ o
60 O
s
13 x~> v^ ^
a) ,-a a>
X"S 13 M X~N ^**. X— \ p*> 5-1
H3(U(DiJbOi-34J 3
-~-M-P~-~S-^'H 4J
MQJi-HbO^-'MS cd
B 4-> -H 0 0 -H !-i
v.-'r-liHS-xJZj^-x,-! 0)
inn c i cd a.
cB^^ s *rg ^^ §
pq H !S H <1 P< H
S
Id
a
cd
•H
T3
a\
S
•
cd
1052
-------
CO
Q
«3j
o
H-3
CJ
M
fe
3
0
o
w
M
1
Q
4-J
H_j
•
cr
CO
""O
p.
&o
o
•
CO
PQ
P3
•
"*
PL)
§
O cu
in 3
CM rH
MH
*> MH
CD W
CO
CU ^
a
O 4J
M 0
P-i CU
X-N Q) t-4 }-i CU
1-4 T3 M -^ ^-s CU !*, $-••
^^ CU CU 60 h4 4-J 4J 3
60H4-1 S-^rH-H 4-1
0CUrHv_-'60'H0 ctt
V-* 4J -H B PL, -H M
rH «4H JZ! v_x p.) a)
IOH 0 |. « Cti fl<
QpntD COSJ Q ^cct &
SMSorHtrJ S
H pq 3 H O
-------
Top View
Effluent
Well/Outlet
n
Stage #4 Stage #3
Stage #2
Stage #1
I Inlet
Wet Well
Drive
System
Side View
Wastewater Level
Figure 1. Diagram of Rotating Biological Contactor.
1054
-------
4->
E
Ol
t
LU
|
|
-M
E
i — i
» •• CU
- 3
t _
q-
UJ
+->
q~^
E
i — i
-t->
3
q"^
£ -
+-> -
QJ • •»••»• »»»»«»»»*^
q^
E
i — i
4j
*••• •••••» * E
OJ
q-
UJ -
4J 1
E • • • »»••••
OJ " *
T- OJ
en r— 3
-a i- q-
03 "a aj
0 >,
s-
o in ea
•i- 3 -a
r— 0 E
3 -i— O
03 S- CJ
S- ro OJ
"O > to
) J ^ f ^
03 CQ
0 E
« OJ O
CM cn+J
o
S- 13
•i— i-
1 O
•r- 0 >>
1.5
'erformance for Ammon
nmonia-m'trogen are c
ng rate, respective!
C_5 (0
CQ q- o
o: o «—
•
ro
in
CM
o
CM
in
LO
OJ
[/6ui
1055
-------
cu
0)
CD
LlJ
d)
UJ
CU
OJ
O)
ro
"O
O-
cn
cn
fO
-P T3
C (O
CD O
O
CJ rei
OQ S-
OL -O
to _E
d) O
4-> CO
(O O)
DC
•O E
3 CU
rd o
"O O
•J= >?
to E
3 CL)
O 3
O
•i- S- '
fOM-
(O S-
fO
fO O
> O
ecu
to
cu
o; o
cn
co
OJ
o o
en -M
s- -o
o cu
<*- s-
— ro
Q.
cu E
o o
E O
fO
E CU
S- Q
CU O
n_ DO
o s-
D3 O
CO
cu
L/6ui eQOa
1056
-------
NH3-N, rng/1 -
i i i i i i
to
Q
fO
4->
03
to
E
C
ro
s-
cu
o
C
CJ)
C_J T3
D2 CL
o; en
CM
OJ
S-
to cr.
s- c:
CD T—
O--O
E res
OJ O
M- O
C -r-
O tO
OJ S_
cu
s_
- 30 '
1057
-------
12
10
8
CO
6 _
4 _
2 _
O 1.5 gpd/sq. ft
2.0 gpd/sq. ft
A 3.0 gpd/sq. ft
A 4.0 gpd/sq. ft
Stage #1
Stage #2 Stage #3
RBC Stages
Stage #4
Figure 5. Progression of Nitrification within the RBC Process
at Various Hydraulic Loadings.
1058
-------
res
o
TFE
Stage #1
Stage #2
Stage #3 Stage #4
0 NH3-N, mg/1
• pH
^ dissolved oxygen, mg/1
Figure 6. Progression of Nitrification within the RBC
Process at 3.0 gpd/sq. ft. following a Trickling
-Filter without Intermediate Settling.
1059
-------
r
NH3-N, mg/1
o
CS1
o
oo
in
CM
o
C\J
oo
cr
to
Q.
O5
o
*
co
4-»
(O
to
to
OJ
o
o
ca
cu
to
to
o
o
c:
o
o
£
M-
O
O
CO
CU
CD
1060
-------
-------
CU
UJ
QJ
=5
CD
4-
LU
CU
3
r—
<*-
CU
3
CU
u +->
(O
to
(U
O
£
Q_
O
ca
OL
OJ
O
S-
C7> CU
C Q.
•i—
r— +J
O CU
H- (T3
c_> cu
>P s-
CM
-a
a> •>
-a s-
O) i—
s- -r-
o_ u.
o
ca
O
o
O
CO
O
LO
o
CO
o
CM
o
o
CU
c/o
O)
S-
L/Bui «saog
1062
-------
CD
4_
LU
4->
E
CU ""
3
i—
E
CU r—
3
r— L_
f-
= .
t—4
_! ,
S - 5 S g
* • *
4->
C
M-
co
D;
cr
to
T3
QL
CT;
_ CD
ro
» -s
•r- CO
1 — C/5
^ CU
0 0 o
-Q P*» CQ
i — C£
•o
cu •> cu
T3S- _E
OJ CU +J
O -*->
CU r— lf-
S- -r- 0
n. u_
cu
Of k
u
CO E
a: ta
o
H-
~ S-
cu
D_
4->
E g
*— U^
£ -1
a a>
••- s-
t *-
~~
0 E
>O 0
£i in T-
CM +J
•o ro
cu « S
T3 S- -r-
CU CU C|-
0 4J -r-
CU i— s_
S- -i— +J
2_ u_ -T-
0 =•
n
3i
C3
o o cu
•~ J_
3
rn
l/6ui eue6ouq.LM - eiuoiuuiy
1063
-------
-------
PART IX: DESIGN AND OPERATION
A DESIGN APPROACH FOR ROTATING BIOLOGICAL
CONTACTORS TREATING INDUSTRIAL WASTEWATERS
By
W. W. Eckenfelder, Jr.
Distinguished Professor
Environmental and Water Resources Engineering
Vanderbilt University
Nashville, Tennessee, USA
Louis Vandevenne
Senior Research Scientist Cebedeau - Liege
Lecturer in Department of Environmental
Engineering - Liege University
This paper presents a design model for rotating biological contactors
treating industrial wastewaters based upon the multiple zero order organic
removal concept. This same model has previously been applied to trickling
filters and the activated sludge process. The basis for the model is that
the organic removal rate in each stage is proportional to the concentration
of organic remaining in that stage. Consider an industrial wastewater
containing a mixture of organics of varying removal rates. The more readily
degradable constituents will be removed first resulting in a higher overall
removal rate in the early stages followed by a progressively decreasing removal
rate in subsequent stages. This is defined by the relationship:
in which
Q
A
S
s
K
Q/A (SQ - S) = K S
hydraulic loading gal/day (m /day)
2 2
filmed area ft (m )
influent soluble BOD, mg/1
effluent soluble BOD, mg/1
proportionality constant
1065
-------
A maximum rate will result from oxygen limiting conditions. Other models
are compared for their application to various wastewaters.
A design example for an industrial wastewater is presented.
NOMENCLATURE
S
Si
n
Q
A
EA =
Q/A =
k
K
Kf =
V
P
r =
c =
concentration in substrate of the influent (mgBODc/1)
concentration in substrate of the effluent of stage i
(mgBOD5/l)
number total of stage
o
flow applied to the system of n stages (m /d)
2
wetted area of the discs in every stage (m )
wetted area of the discs for the total system (n stages)
3 2
hydraulic loading for every stage (m /m .d)
kinetics constant (d )
3 2
proportionality constant (m /m d)
saturation constant in the Monod relationship (mgBOD,-/!)
o
volume of the trough (m )
removal efficiency per stage (percent)
coefficient of correlation
2
maximum substrate removal per one stage (g BOD5/m .d)
DEVELOPMENT OF THE DESIGN MODEL
The design model is based upon the multiple zero order organic removal
concept that has been applied to trickling filters and the activated sludge
process treating domestic and industrial wastewaters (TISCHLER et al, 1969).
The basis for the model is that the organic removal rate in the trough is
proportional to the concentration of organic remaining in the trough. This
is defined by the relationship:
Assuming complete mixing in the trough and the mass of fixed micro-organisms
is proportional to the wetted area of the discs. A mass balance on the substrate
for the trough, as shown in figure 1-a is given by:
- V- Q ($„ - S,) - KS, A
dt
1
(2)
1066
-------
which at steady state becomes
Q/A (SQ - S.,) = K S
1
(3)
Considerations of relationship (1) and (2) leads to the fact that in this
model mass transport of oxygen and substrate, are not explicitly included and
that organic removal by suspended micro-organism in the trough is negligible.
This last assumption is reasonable if the detention time in the trough is
short and the concentration of the biomass is low.
The relationship (2) and (3) is identical to the KORNEGAY model (1972),
where the Monod relationship is used for the organic removal rate, if one
assumes that S. is negligible with respect to the saturation constant Kf.
In this model the constant K is the proportionality constant between the
removal rate and the concentration remaining. Implicit in the definition is
the fact that more readily degradable organics are removed first thereby
resulting in a decreasing removal rate as the residual concentration decreases.
Biofilm properties are incorporated in the coefficient. It is necessary to
consider reactor hydraulics. It's known (HERBERT, 1964, KORNEGAY and ANDREW
1969) that a series of completely mixed reactors provide a valid approximation
of a plug-flow system. This consideration is illustrated (See Fig. 2) by the
response of a system of RBC's with two identical stages to a pulse input
tracer (POPEL 1964). These considerations lead to the conclusion that
relationship (3) is only applicable stage per stage, the general form of
relation (3) for a system of n identical stages is:
= KSi
(4)
At steady state the concentration of substrate S- leaving stage i in terms
of the influent concentration SQ is given by:
S =
i = (1 + K A/Q) 1
(5)
It is important to note that in relationship (5) Q/A refers to each stage and
not to the total area of the staged system.
The plug-flow mode provides a definite advantage for staged operation
as indicated by consideration of relationship (5).
MODEL PREDICTIONS
For a series of identical RBC stages when flow is constant the relation-
ship (5) shows that the removal efficiency per stage, p, is a constant.
Adams et al (1975), Grau et al (1975) and Grady and Williams (1975) showed
the effect of influent concentration for the activated sludge process treating
highly variable strength wastewaters. The following development assumes a
moderately constant strength wastewater.
This is defined by the relationship:
log (Si/s ) = 1 log (1 - p)
(6)
A plot of the logarithm of S./S versus the stage number, i, would give a
linear plot as shown in Figure °3) with a slope equal to the removal efficiency,
p. This graphical approach is very useful, as described later, to study
the substrate concentration profile through a stage RBC system.
1067
-------
The multi-zero order assumption used in the model implies that the removal
efficiency is independant of the feed concentration and that the primary
design factor is the hydraulic loading rate per unit disc area. The effect
of hydraulic loading rate, expressed in terms of specific contact area,
is shown in Figure (4).
These curves are similar to the empirical relationship of POPEL (1964)
for a fixed influent concentration and to the design charts of the STENGELIN
Corporation widely used in Europe to design RBC plants.
The usual interpretation of this model is to plot the removal rate
per unit of disc area (j. ^ _ s_} versus s _ _ Jhis win give a linear plot
as shown in Figure (5) with a slope equal to K.
A maximum removal rate can be expected when the biofilm growth is at a
maximum rate under the experimental conditions. This maximum rate will
result from oxygen or other limiting conditions.
When oxygen is limiting, the capacity of the unit can be increased
by increasing the rotational speed WELCH (1968). This limitation is generally
observed at high loadings in the first, stages in a staged RBC system. For
cost optimization, however, one must consider that the power consumption
increases exponentially with rotational speed.
MODEL VERIFICATION
The model has been tested with data available in the literature by
ECKENFELDER (1977) and recently by EDELINE and VANDEVENNE (1979).
The multiple zero order assumption, relationship (6) has been verified
in several cases, TORPEY (1971), BIRKS (1971), LABELLA (1972) and KINCANNON
(1974) in Figures 6 to 9.
We observed in some cases a discontinuity in the plat for the first
stage, where the removal rate is higher than that predicted from relationship
(4). In this case as indicated in Figures 8 and 9, the first stage must be
considered separately. A second discontinuity is also observed in some cases
for the last stages of a RBC system where the removal rate decreases. This
expresses a quantitative and/or qualitative modification of the fixed biomass,
probably due to a variation in the nature of the substrate. This leads to a
modification of the proportionality constant, K.
These considerations result from the staged system. The more readily
degradable constituents are removed first resulting in a higher removal
rate in the early stages followed by a progressively decreasing removal rate
in subsequent stages. Typical results are presented in Figures 10 and 11.
It has been observed (EDELINE and VANDEVENNE, 1979) that when one only
uses cumulative data, S- - S , these data can be interpreted by the model
in some cases. In cases this is an empirical relationship that does not
justify an extrapolation for a staged system with a variable number of
stages. The value of K in this case is higher than the true value because
it includes the plug-flow kinetics effect and eventually a higher removal
in the first stage. The true value of K can be obtained by linearisation of
relationship (5) as indicated in Figure 12. This fact has been also observed
by the authors with the Kornegay model.
1068
-------
The results are summarized in Table 1 for domestic wastewater and in
Table 2 for industrial wastewaters A better coefficient of correlation could
be obtained in some cases if one had not forced the line to pass through
zero. The residual line give a BOD,- removal limited to an effluent BOD(-
from 3 to 10, which is reasonable wnen one considers the dynamics of the
biological process.
COMPARISON WITH OTHERS MODELS
EDELINE and VANDEVENNE (1979) have compared this model with the Kornegay
mode. In may cases the Kornegay model gives a better coefficient of correlation
(plotting Q-rg- versus 1/S). However, the proposed model offers the simplicity
of having onTy one constant.
When the flow is constant, the design model presented in this paper
gives identical predictions as the model of HANSFORD et at (1978), that consider
the biological reaction and mass transfer to the biofilm, since the authors
assume fist order kinetics that has the same formulation as relationship (1).
When the hydraulic loading varies a comparison of the results published by
the authors from the ANTONIE et al (1971) data as presented in Figure 13,
the predictions appear similar (the coefficient of correlation is not given
by the authors). This may imply that mass transfer processes are not influenced
by a modification of flow applied to the RBC. This fact can be related to the
results of KORNEGAY (1972) that showed that flow rate does not have as
significant effect on the constant of saturation.
DESIGN EXAMPLE
The following data was obtained on a
effluent. The test unit had 4 stages.
RBC for a bleached pulp mill
Flow to the unit
2
g pd/ft*
2
4
6
8
10
3 2
in /m d
O.O82
0,163
0,245
O,326
Q,4O8
Influent
76
72
81
78
83
B
Stage 1
26
41
42
52
60
OD5 (mgO2
Stage 2
10
24
25
38
47
/I)
Stage 3
7
16
15
25
36
Stage 4
6
12
10
18
30
Temperature : 18 "C
1069
-------
TABLE 1
DOMESTIC WASTEWATERS
Reference
Antcwi* and Koehler
Ban HuAt Tan
Clark at al
Warfci
Popel
Stengelin
Torpey
Temperature
CO
12,5-20
(1)
(1)
(I)
(1)
(1)
16 -26
K
3.2,
m /m a
O,2O4
0.633
0,301
O,41O
0,428
0,679
0,701
C
g LBODr/m d
-
35
«
-
32
105
-
r
0,9059
0,8747
0, 7420
O,664O
O,8O12
-
O, 94O4
(1) In these studies temperature follow seasonal variation
2 32
Note :-with flow in gpd/ft , multiply the value of K (m /m d) by
24,51.
-»F -= 32' + 5/9 °C
1070
-------
TABLE 2
INDUSTRIAL WASTEWATERS
References
Antonie et Welch
Bifks et al
/
Bin tan j a
Cebedeau
Chen et al
Chittenden et al
Cochrane et al
Eckenfelder
Kincannon et al
Labella et al
Thomas et al
ffillard et al
Wastevater
Dairy
Cheese
Synthetic subs-
trate (caseine)
Malte-house
Paper-mill
Meat (pretreated
by anaerobic
lagoon)
Potato
Bleached pulp
mill
Meat
Winery
Distillery
Hardboard
Tempera-
ture
(°C)
12-20
7-20
2O
20
3O-35
(20)
15-18
18
-
*»
14-19
K
3.2,
m /m d
0,082*
0,039
0,055*
0,037
0,177
0,343
0,082
0,556
O,O26
0,126
0, 15O(6stages)
0,435
OfO56
C
g BODs/m d
-
24
36*
-
-
•"
40
-
60
170
-
r
0,8909
0,9543
0, 7998
0,4634
0, 5282
0,8182
O,82O9
0,9525
O,9050
0, 9485
0,9150
0,9647
0,9779
* Data in DCO
** Relative to the total RBC unit.
Note :-with flow in gpd/ft2, multiply the value of K (m /m d) by 24,51
- "F - 32° + 5/9 •€
1071
-------
The proportionality constant is calculated by plotting Q/A (S- •, - S.)
versus Si.
The hydraulic load per stage, Q/A, is given by multiplying the hydraulic
load relative to the unit by 4.
From Figure 14, K = 0.556 m3/m2d (13.6 gpd/sq ft)
Design an RBC plant for 80 percent BOD removal assuming an influent
BODg of 85 mg/1 and a required effluent BOD of 17 mg/1.
The corresponding Q/A to meet S = 17 mg/1 can be calculated by the
relationship:
Sn _ 1
S0 (1 + K A/Q)'
with Sn = 17 mg/1
n = number of stages of the RBC unit
The results are plottted in Figure 5.
The total area of discs is calculated based on the hydraulic load for the
total system, Q/y/\:
Q_ = 9, f J
EA A
where n is the number of stages in the RBC unit. These results are plotted
in Figure 16.
To meet S =17 mg/1 in a three stage RBC system, the hydraulic load
for the systenfis 0.26 mS/n^.d (6.37 gpd/sq ft).
If the influent flow is 2,000 m3/d (0.528 mgd) the total surface area is:
o:aS"J/&d -7-692 m2 <=82-800 s"ft)
The final number of stages can be decided by a cost optimization.
When the experimental data shows a maximum removal capacity, the organic
loading for each stage should be checked. If this loading is higher than the
maximum, the hydraulic loading should be decreased.
Also when one observes a high removal rate in the first stage, this
stage should be considered separately and the other stages calculated as
indicated in the example.
SUMMARY AND CONCLUSIONS
A model based on a multi-zero order kinetics, is developed and the method
for evaluating the constant described. Validations and limitations are des-
cribed on the basis of published data. Constants are given for domestic and
some industrial wastewaters.
The simple relationship reasonably fits experimental data and can be
used to design rotating biological contactors. In some cases, special emphasis
on the first stage removal is necessary. A maximum removal that may be due
to oxygen limitation must be considered also.
1072
-------
At this time, several models may describe the removal relationship in
a staged RBC unit. For this reason, the authors recognize that in some
cases empirical relationships or design charts may be better for the design.
It is essential to record during a pilot plant study the substrate concentration
profile in a staged RBC unit.
1073
-------
REFERENCES
L. F. Tischler, W. W. Eckenfelder (1969), Linear Substrate Removal in the
Activated Sludge Process, Advances in Water Pollution Research 361,
Pergamon Press, Oxford, England.
B. H. Kornegay (1972), Modeling and Simulation of Fixed Film Biological
Reactors, Mathematical Modeling In Environmental Engineering, 8th
Annual Workshop, Nassau, Bahamas.
D Herbert (1964), Multistage Continuous Culture, Continuous Cultivation of
Microorganisms, Edited by I. Malek, K. Beran, and J. Haspodka, Academic
Press, New York, p. 23, 1964.
B. H. Kornegay, J. F. Andrew (1969), Application of the Continuous Culture
Theory to the Trickling Filter Process, Proceeding of the 24th Ind. Waste
Conference, Purdue University.
F. Pope!(1964), Aufbau, Abbauleistung und Bemessung von Tauchtropfkorper
SchWeitz,'Z. f. H.ydrologie XXVI (2), 394-407.
C. E. Adams, et al (1975), A Kinetic Model for Design of Completely Mixed
Activated Sludge Treating Variable-strength Industrial Mater, Water
Research, 9, 37.
P. Grau, et al (1975), Kinetics of Multi-component Substrate Removal by Activated
Sludge, Water Research, 9, 637.
C. P. L. Grady, Jr., D. R. Williams (1975), Effects of Influent Substrate
Concentration on the Kinetics of Natural Microbial Population on
Continuous Culture, Water Research, 9» 171.
F. Popel (1964), Leistung. Berechnung und Gestaltung von Tauchtropfkorper-
anlagen, Stuttgarter Berichte z. Siedlungswasserwirtschaft, 11,
(Oldenbourg, Munchen).
F. M. Welch (1968), Preliminary Results of a New Approach in the Aerobic
Treatment of Highly Concentrated Wastes. 23rd Purdue Ind. Waste Conf.
May 1968.
W. W. Eckenfelder, Jr., (1980), Principles of Water Quality Management. CBI
Publishing Company, Boston, Mass.
F. Edeline, L. Vandevenne (1979), Cinetique de L'Epuration dans les Biodisgues,
CEBEDEAU, Janvier-Fevrier 1979, 422-423.
W. H. Torpey, H. Heukelekian, A. J. Kaplovsky, R. Epstein (1971), Rotating
Discs with Biological Growths Prepare Wastewater for Disposal or Reuse.
J. Wat. Pollut. Control, Fed. 2181-2188.
C. W. Birks, R. J. Hynck (1971), Treatment of Cheese Processing Wastes by
the Bio-disc Process, Proceeding of the 26th Ind. Waste Conference,
Purdue University.
1074
-------
S. A'. Labella, I. H. Thaker, J. E. Tehan (1972), Treatment of Winery Hastes
by Aerated Lagoon, Activated Sludge and Rotating Biological Contactor,
Proceedings of the 27th Ind. Waste Conference, Purdue University.
D. F. Kincannon, J. A. Chittenden, E. S. Stover (1974), Use of Rotating
Biological Contactor on Meat Industry Wastewaters, Proc. S. Nat. Symp.
on Food Process. Wastes, EPA 660/2/74058.
D. F. Kincannon, E. L. Stover (1976), Evaluating Rotating Biological Contactors
Performance, Water & Sewage Works, March 1976.
G. S. Hansford, J. F. Andrews, C. G. Grieves, A. D. Carr (1978), A Steady-State
Model for the Rotating Biological Disc Reactor. Water Research, 12, 855-868.
R. L. Antonie, F. J. Koehler (1971), Application of Rotation Disc Process
to Municipal Wastewater Treatment. Clean Water (EPA) 17050 DAM 11/71.
Ban Huat Tan, M. Giffard, G. Martin (1976), Epuration Biologique des eaux
Usees Urbaines a_u Moyen de Disques Biologiques Tournants, TSM - L'eau
281-284.
J. H. Clark, E. M. Moseng, T. Asano (1978), Performance of a Rotating
Biological Contactor Under Varying Wastewater Flow, JWPCF, 50, 806-821.
E. Marki (1964), .Ergebnisse von Versuchen der EAWAG mit dem Tauchtropfkorper
Schweitz, Z. f. Hydrologie XXXVI (2), 408-427.
R. L. Antonie, F. M. Welch (1969), Preliminary Results of a Novel Biological
Process for Treating Dairy Waste, Proceeding of the 24th Ind. Waste
Conference, Purdue University.
H. H. Bintanja (1976), The Use of 0? in a Rotating Disc Process. W. Res.
10, 561-565.
H. T. Chen, E. E. Frederickson, J. F. Cormack, S. R. Young (1974), Four
Biological Systems for Treating Integrated Paper Mill Effluent, TAPPI,
57, 111-115.
J. A. Chittenden, W. J. Wells (1971), Rotating Biological Contactors Following
* Anaerobic Lagoons. JWPCF, 43, 746-754.
M. W. Cochrane, K. A. Dostal (1972), RBC Treatment of Simulated Potato Processing
Wastes. NERC (EPA), Corvallis.
J. L. Thomas, L. G. Koekrsen (1974), Activated Sludge - Bio-disc Treatment of
Distillery Wastewater. U. S. Environmental Protection Agency, 66/2-74-014.
H. K. Willard, W. F. Eckerle, R. H. Scott (1972), Feasibility of Rotating
Disc Treatment Process for Hardboard and Insulation Board Wastewater,
Preliminary Report, NERC (EPA), Corvallis.
1075
-------
-------
EMPIRICAL OXYGEN TRANSFER RELATION IN
THE RBC PROCESS
by
B. F. Severin
Graduate Student, Department of Civil
Engineering, University of Illinois
Urbana, Illionis
H. Brociner
Process Engineer
Environmental Division
J. E. Dumanowski
Laboratory Supervisor
Environmnetal Equipment Division
Itasca, Illinois
J. T. Su
Research Engineer
Center Engineering Laboratory
Santa Clara, CA
M. M. .Gurvitch
Chemist
Chemical Research and Development Center
Princeton, N.J.
F. M. C. Corporation
INTRODUCTION
Maintenance of bulk solution dissolved oxygen concentra-
tions in the Rotating Biological Contactor process (RBC) has
been found to be important in retaining process efficiency
in treating simulated dairy wastes (1), synthetic sewage (2),
petroleum wastes (3), and in nitrification (4). Two hypothe-
ses can be made to explain the observed phenomenon. The first
would suggest a loss of efficiency of metabolism to low
oxygen concentrations. The second would suggest a population
change occurs on the disc such that organisms which can with-
stand low oxygen concentrations begin to predominate resulting
in a net loss of process efficiency. Dissolved oxygen
concentration limits have been observed to be between 1.0 to
2.1 mg/1. The lack of understanding of oxygen transfer in
pilot and full scale RBC plants has apparently resulted in
1077
-------
cases where the full scale plant was unable to meet the pre-
dicted removal efficiency (5).
Two procedures have been suggested to scale-up oxygen
transfer effects in pilot and full scale RBC systems. Oxygen
transfer to the bulk solution by rotating discs has been
presented as a function of disc surface area turnover (2, 5,
6). This concept has been used to argue for pilot plant
scale-up based on disc rotational speed (rpm; revolutions
per minute). However, other literature suggests pilot plant
scale-up based on disc tip speed as the design parameter (7) .
Oxygen transfer to the bulk solution is governed by
water adhering to the discs, as well as mixing in the tank.
Neither of the two models for scale-up, i.e., tip speed and
rpm, fully consider the role of tank mixing in transfering
oxygen. In order to understand the role of tank mixing in ;
the transfer of oxygen, it is necessary to model the effects
of tank volume and geometry on oxygen transfer. Zeevalkink
et al. (8)- used an empirical relation based on disc sub-
mergence depth to correlate a series of oxygen transfer data
taken with flat, polystyrene discs.
Data presented in this report show that neither tip-
speed nor rpm scale-up adequately defines oxygen transfer
relations between pilot RBC systems tested with clean tap
water and clean discs. An empirical relation between oxygen
mass transfer and Reynolds numbers calculated from tank
geometries is developed. A scale-up factor relating rota-
tional effects in pilot and full-scale systems is proposed.
RBC PILOT PLANTS
Oxygen transfer was evaluated in three RBC pilot plants;
Bensenville, Illinois; Princeton, New Jersey; and Santa
Clara, California. All RBC discs were made with the FMC
Corporation, Environmental Equipment Division, corregated
media design shown in Figure 1. A total of eight test cases
at the three pilot facilities were studied for the develop-
ment of the oxygen transfer model. A test case is defined
as a variation in the number of discs, the water depth in a
tank, or an alteration in the distance between the bottom
of the tank and the bottom edge of the discs.
The Bensenville RBC pilot plant, located at the South
Wastewater Treatment Plant, Bensenville, Illinois, is a full
size, twenty-five foot shaft, with media divided into four
stages (66 sheets of media per stage). The full shaft carries
84,000 square feet of media. The discs are 11.69 ft. in
1078
-------
MEDIA DETAIL
HIGH DENSITY POLYETHYLENE
FIGURE !
1079
-------
diameter. Oxygen transfer tests were made in the first and
third stage of this pilot plant. Data presented are the
average of the two results and, therefore, represent a single
stage of the four stage system. Two test cases were studied
at this pilot plant. In test case B the water depth was
maintained at the normal operating depth. In test case BL
the water level was lowered by six inches.
Four test cases were studied at the FMC Central Engineer-
ing Laboratory facility in Santa Clara, California. In the
first three test cases, CIO, C22 , and C30 , ten, twenty-four
and thirty discs were used, respectively. In the fourth
test case, C30H, the shaft containing 30 discs was raised
by five inches to increase the distance between the bottom
of the .tank and the edge of the media. The water level was
also raised by five inches in order to maintain the same
percent submergence of the discs . The media used in this
pilot facility was 11.83 feet in diameter.
Two test cases were studied at the FMC Chemical Research
and Development pilot facility at Princeton, New Jersey. This
pilot plant utilized 3.95 ft. diameter media. Eight sheets
of media were used on the shaft in test case P8, and ten
sheets were used in test case P10 .
A summary of the physical data for all eight test cases
is presented in Tables 1 and 2. In Table 1 the disc diameter1
is represented by (2 x r) where r is the disc radius. The
number of discs is n, the rank volume is V. , and the tank
water depth at the center of the tank is H. The tank width
is given as w. In table 2 the disc section length is L.
The active disc area, a, is calculated in Equation 1:
2 2
1) a = mmr (r - rw )
where r is the distance from the center of the RBC shaft
w
to the water and m is a factor relating the surface area
of one sheet of plastic media to the surface area of a
circle with the same radius. The active disc volume, V,,
is expressed in Equation 2 :
2)
Vd = fLir (r -
where f is the void volume of the media. The active volume
and the active area are the volume and area of media which
become submerged during one revolution of the shaft.
1080
-------
co
EH
EH
O
J
H
CM
IS
H
CO
><
ffl
S3
Kx* *" *
££4 ft ,*•—%
—,3 Q -P
j> ^j r™i m
— En & —
4J
4-1
tt! K
^"^. ?V* TV! r i
ffi J3 EH ft
EH IS O
i ) (^ h-* i j
4-J ^-i t— ' -H
^ ^ ^^
w
o — .
« 2! EH
w w 2;
EH O W
ri! g W
H 03 ft
Q D <—
to
CO
o
OH CO
W H
— D fa
2 O
4J
j^ OS
O W
X CO EH
H Ed
» L^ fV} ^Q y^ |_f^
CM oo ro ro ro m
rO CM r-i rH iH iH
rorooorororororo
VD VD 0 CM O OOO 0
vo >sD r-H CM ro ro
-------
H
EH
I
O
o
EnS
W O
O W EH
IS O EH
^flj P O
H EH W CQ
— cn
H O «
Q W 12
HriJ
Q EH
vo
en
CO
*
0
vo o
CP> in
00 CN
• •
o o
o
in
CN
•
O
O
in
CN
•
o
o P- r->
r*» ^o VD
VO H rH
• • •
O 0 0
EH
1
Hi
H
O
n
H
^-% t>
•OH
> EH
>— O
U
cn
H
Q
»«»
|VJ
gj
D
Hi
O
00
*
f-
rH
in
CO
*
O
CT*
^
00
•
cn
r~
VD
*
in
t-»
r— 1
m
*
cr>
m
CN
in
•
CTi
ro
CN
ro
ro
•
r—
VO
Pi
o
13
rt!
O
H
W
• •
ro \^ c^
CN rH
O O ro ro o
in in oo oo m
• • • • •
in m o .-i CN
0 CN O
J rH CN ro
m ffl o u o
CN
oo
CN
Cfi
o
in
•
CN
M
o
ro
U
«=r in
cn in
CN ro
r» ro
vo oo
• •
o o
o
00 H
PH ft
1082
-------
The tanks at Bensenville and Santa Clara were semicir-
cular in design. The tank at Princeton was an angular
design, such that the distance between the media and the tank
wall was fairly constant. The Princeton tank, therefore,
approached a semicircular design. The distances between
the tank and media for all test cases is approximated by the
distance, E, in Table 2.
METHODS AND PROCEDURES
Oxygen transfer was evaluated using clean RBC discs in
clean tap water with the non-steady state aeration test. The
dissolved oxygen was first removed by the addition of approx-
imately 120 mg/1 sodium sulfite. Cobalt chloride (0.5 mg
Co/i) was used as a catalyst. As the discs rotated, the
oxygen content of the test water slowly increased. Oxygen
concentrations were determined with YSI dissolved oxygen
probes and meters (Yellow Springs Instrument Co., Yellow
Springs, OH) in tests conducted at the Princeton and Santa
Clara pilot plants. The modified (sodium azide) Winkler dis-
solved oxygen test (9) was used to determine oxygen concentra-
tions at the Bensenville pilot plant. The oxygen saturation
value was determined, in most cases, by allowing the test to
proceed until a constant dissolved oxygen concentration was
achieved. In several cases the saturation concentration was
determined from tabulated values in Standard Methods (9). The
oxygen transfer coefficient KTa, for each test was calculated
as the negative slope of a plot of the natural logarithm of
the oxygen deficit versus time. The oxygen deficit is the
difference between the observed oxygen saturation concentra-
tion and the oxygen concentration at the sample time. Data
representing between 10 and 90 percent of saturation were used
to calculate KTa. All tests were performed with water temper-
atures near 20°C. Corrections of K a to K aon, the oxygen
-Lj \r* ^f W
transfer coefficient at 20°C, were made using the mathemati-
cal procedure discussed in Standard Methods (9).
RESULTS AND ANALYSIS
Oxygen transfer coefficients, KTaon (min ), at several
Li Z U
rpm values, were determined for each of the eight test cases.
Oxygen transfer rates, OTR (gO~/min) were calculated for each
test using Equation 3 and the proper unit conversions:
• 3) OTR = KTa,,nV,C
L 20 t s
where C_ is the saturation value of oxygen at 20°C (9.2 mg
S
The results for KTa,,0 and OTR are reported in Table 3.
Jj /U
1083
-------
TABLE 3:
OXYGEN TRANSFER RESULTS
PILOT
PLANT
B
BL
CIO
C22
C30
C30H
rpm
0.8
1.2
1.6
2.4
1.2
1.6
2.4
1.52
1.71
2.26
1.52
1.71
1.88
1.52
1.71
1.88
1.52
1.71
1.88
KLa20
(min™1)
0.0170
0.0265
0.0450
0.1150
0.0315
0.0521
0.1314
0.0135
0.0155
0.0245
0.0323
0.0399
0.0507
0.0361
0.0471
0.0622
0.0339
0.0444
0.0594
OTR
OXYGEN
TRANSFER RATE
(g02/min)
1.44
2.26
3.83
9.77
2.32
3.84
9.67
0.484
0.555
0.878
1.138
, 1.406
1.787
1.
I,
26
64
2.17
1:
1,
38
81
2.42
P8
P10
1.2
3.0
5.0
7.5
1.2
3.0
5.0
7.5
0.011
0.048
0.101
0.272
0.019
0.059
0.133
0.314
0.015
0.066
0.138
0.374
0.026
0.080
0.180
0.425
1084
-------
Several attempts were made to linearize the oxygen trans-
fer results from the eight test cases. None of the attempts
was successful until a method of estimating fluid mixing was
developed. This method is based on a Reynolds number calcu-
lated from tank geometry. The Reynolds number is developed
by considering a reference cross-section of the RBC tank and
media. The reference cross-section is a plane observed
parallel through the shaft and perpendicular to the ground
(Figure 2). The Reynolds number, Re, is calculated using
Equation 4:
4) Re = vR/U
where the ne_t water velocity through the reference cross-
section is v (ft/min), the hydraulic radius of the cross-
section is R (ft), and u is the kinematic viscosity.
Several assumptions are required in order to calculate
the Reynolds number. The first assumption is that the media
acts as a pump to move water from one side of the tank to
the other across the reference cross-section. An estimate
3
of the pumping capacity of the media, Q(ft /min), is given
by Equation 5.
5) Q = rpm x V,
A second assumption is that all of the water pumped across
the reference cross-section must flow back to the other side
of the tank. The media is viewed as an impassable barrier
to water flowing in the direction opposite to the disc rota-
tion. All the flow must pass around the media through the
: 2
area, A (ft ), between the tank wall and the media. This area
4s calculated using Equation 6.
6) A = (E x L) + .HxCW - L)
The estimated net water velocity, v (ft/min), is given by
Equation 7.
7) v = Q/A
The hydraulic radius, R, of the area between the tank wall
and the media is found using Equation 8:
8) R = A/p
where p (ft) is the wetted perimeter of the reference cross-
section calculated using Equation 9.
9)
= 4H-2E+w
1085
-------
r SHAFT
CENTER TO
WATER LINE
PW
H
-DISC
RADIUS
WATER -
DEPTH
CROSS
SECTIONAL
AREA ,A
(SHADED)
I
DISC
I
SHAFT
MEDIA SECT.
LENGTH
Uf
I
WATER
LEVEL
^--^ HEAVY LINE
>> *{P) WETTED
^ PERIMETER
E
I
DISC EDGE
TO TANK
BOTTOM
TANK WIDTH
CROSS - SECTION REFERENCE
VIEW
FIGURE 2
1086
-------
Table 4 presents the data required to calculate Reynolds num-
bers for each test case.
Calculation of a dimensionless parameter to describe
the oxygen transfer data was fashioned after Wu, et al., (10).
These authors used a dimensionless parameter of the form given
in Equation 10 to describe oxygen transfer by mechanical
mixers.
10)
= OTR/yd
The dimensionless oxygen transfer number, Non,
is calculated
for'data standardized to 20°C. The dynamic viscosity is y,
and d is a characteristic distance in the reactor design.
A characteristic distance, d, for correlating RBC data
was found to be the volume of the tank divided by the active
disc area (Equation 11) .
11)
d = V /a
Equation 10 can be modified by replacing expressions for OTR
and d. This is shown in Equation 12.
12) N
20
KLa20 Cs a
y
A summary of N-Q and Re values calculated for all the
test data is given in Table 5 and plotted in Figure 3. The
plotted line has the equation:
13) N2Q = 5.7 x 10"8 Re1'815
The correlation coefficient is 0.994. Further development
of dimensionless parameters was not pursued since N_n and
JRe were adequate to produce a linear relation for the avail-
able data.
Replacement of N2Q in Equation 13 by the original para-
meters outlined in Equations 1 and 12 and replacement of Re
in Equation 13 by the original parameters outlined in Equa-
tions 2, 4, 6, 1, and 9, results in an equation expressed in
readily measured terms (Equation 14).
14)
P
= 5.7 x 10
1.815
1087
-------
DIMENSIONLESS OXYGEN TRANSFER [N20]
o ° ~ _ E
52 •_ o o o
/I
7
/
c
/
X
/*
/
f
/
o
a
A
A.
X
P8
PIO
CIO
C22
C30
C30H
8
BU
00 10000 100000
REYNOLDS NUMBER - Re
DIMENSIONLESS OXYGEN TRANSFER
VERSUS REYNOLDS NUMBERS
FIGURE 3
1088
-------
-p
MH
Q E
W EH
EH [ .
i EH S
W H
!2 PS
H
rH
ro
cr>
CM
ro
CN
CM
ro
ro
CN
O
CM
m
CN
CO
CO
HtN
EH -P
CJ »W
W —
CQ
<;
co w
co «
o <
PS
o
ro
o
rH
CO
CO
rH
.in
CO
CFl
o
r-
•
o
O
H
PS
CO
D
H
Q
CN
ro
•
O
m
vo
ro
rH
oo
in
CN
oo
in
en
ro
CN
rH
*
O
oo
O
IS
g
£H .*— v
EH fl
PS H -H
W CJ E
EH O\
-------
TABLE 5: DIMENSIONLESS OXYGEN TRANSFER AND REYNOLDS NUMBERS
PILOT
PLANT
B
BL
CIO
C22
C30
C30H
P8
P10
rpm
0.8
1.2
1.6
2.4
1.2
1.6
2.4
1.52
1.71
2.26
1.52
1.71
1.88
1.52
1.71
1.88
1.52
1.71
1.88
1.2
,0
,0
7.5
1.2
3.0
5.0
7.5
^20
4.63
7.22
12.26
31.33
8.13
13.44
33.90
0.57
0.65
1.03
2,
3,
98
69
4.68
4.55
5.93
7.84
4.27
5.60
7.49
0.04
0.18
0.39
1.05
0.09
0.29
0.64
1.51
Re
20200
30300
40400
60600
30530
40700
61060
8320
9360
12370
17490
19680
21640
23190
26090
28680
22440
25240
27750
1590
3980
6630
9940
1930
4820
8050
12080
1090
-------
Rearrangement of terms and the substitution of L by the ratio
n/q, where q is the number of discs per unit length, results
in Equation 15 :
0.815
15)
1.815
where
—8
16) K, = 5.7 x 10 (dimensionless)
17) K2 = 35.32 (mg ft3/g£)
The dimensions of all the parameters in Equation 15 are those
given in the nomenclature section.
Equations 13 and 15 can be expected to be valid only for
the media design and tank geometry used in these tests. It
is speculated that the same general form of the correlation
between N2Q and Re will describe results for other forms of
media, but different coefficients will be found. A media
design closer to a flat plate should give oxygen transfer
results with a lesser dependence upon tank mixing, as des-
cribed by the Reynolds number. This is presumably due to
differences in the water pumping capabilities of media with
different designs.
DISCUSSION
Conventional Scale-up Methods
Figures 4, 5, and 6 show the oxygen transfer results
expressed as milligrams oxygen per liter per minute (mg 02/
i min) plotted against the conventional scale-up parameters
of tip speed, surface area turnover rate, and rpm, respec-
tively. It is instructive to compare pilot configurations
with approximately the same ratio of surface area to tank
volume. The pair of pilot configurations which have the
closest ratio of surface area to volume are P10 and C30,
2 3
with ratios of 68.3 and 69.4 ft /ft , respectively. When
comparing the operation of the plants at constant tip speeds,
P10 transfers oxygen (mg 02/& min) at a rate approximately
three times that of C30. When operating at constant surface
area turnover rates, P10 transfers oxygen at a rate 100
times that of C30. If constant values of rpm are compared,
1091
-------
0.01
10
100
TIP SPEED (ft/min)
OXYGEN TRANSFER (mg 02/l-min) versus
TIP SPEED (ft/min)
FIGURE 4
1092
-------
IO
O
o>
OC
LU
U.
CO
CE
LU
O
>- 0.10
§
0.01
100
1,000 10,000
SURFACE AREA TURNOVER
-------
to
.:. 1.0
N
o
O>
cc
til
o:
tu
CD
> 0.10
/
0.01
o P8
• PIO
a CIO
® C22
• C30
^ C30H
* 8
x. BL
O.I
1.0
10
RPM
OXYGEN TRANSFER (mg 02/l-min) versus
ROTATIONAL SPEED (rpm)
FIGURE 6
1094
-------
between the approximate rpm values of 1.5 to 1.9, C30 trans-
fers oxygen at a rate 1.5 to 3 times the rate of P10. From
these analyses, it is seen that none of the parameters of
tip speed, surface area, or rpm give proper oxygen transfer
correlations for scale-up of rotational effects. Scale-up
based on tip speed and surface area turnover overestimate
the transfer capacity of the large diameter plant. Scale-up
by rpm underestimates the transfer capacity of large diameter
plants. None of these methods gives as strong a correlation
between oxygen transfer and rotational effects as does the
relation between N2Q and Re shown in Figure 3 and Equation 13.
Scale-up By N2Q and Re
At the present state of knowledge, it is necessary to
assume that if two RBC systems transfer oxygen at the same
rate (mass of oxygen per unit area per volume) in clean tap
water with the media free from biomass, then the oxygen
transfer will be the same when the two systems are operated
under similar conditions of hydraulic loading, organic load-
ing, and substrate removal. Little data concerning the effects
of biomass on the transfer of oxygen to the bulk solution are
available to test this assumption.
If the above assumption is made, then Equation 15 may
be used to establish the test rpm of a pilot system. If
the full scale and pilot systems contain the same ratio of
surface area to volume, then it is only necessary to match
the oxygen transfer coefficient, KTaon, of the pilot system
Li <£U
to that of the full scale RBC. If the surface area to volume
ratios in the two systems differ, then the values of (BLa-gV /a)
must be matched. The experimental loading capacity of the
pilot RBC is then found. The maximum loading is achieved
when the minimum dissolved oxygen limit in the bulk solution
can no longer be maintained. Since the first stage of an
RBC is subjected to the highest oxygen demand, the critical
area for testing the rotational effects on oxygen transfer
is in the first stage.
The above statements concerning K a2Q and (K a20V./a)
were derived by considering mass balances of oxygen and
organic substrate in a single stage of an idealized RBC.
It is assumed that the single stage is completely mixed.
When pilot and full scale RBC systems are compared, it is
assumed that constant dissolved oxygen levels must be main-
tained, and that the rate of oxygen depletion and substrate
removal by the biomass is dependent upon the surface area per
volume of media available.
1095
-------
The drawback to this scale-up method is that in order to
determine the oxygen transfer coefficient of the full scale
RBC, it is necessary to assume the tank dimensions and amount
of media in one of the stages. This requires that tank
design parameters are available from the RBC manufacturer.
Some rationale for the design of the full scale plant must
be considered, i.e., some estimation of the surface area
and staging requirements, prior to the initiation of the
pilot testing.
FUTURE RESEARCH NEEDS
In order to use the N2Q - Re correlation to scale-up
oxygen transfer effects, it must be assumed that two RBC
systems with equal oxygen transfer capacities in clean water
will have the same oxygen transfer capacities when operated
under similar conditions of hydraulic and organic loading.
This assumption can be tested with side-by-side operation of
a pilot and a full scale RBC which have been preset for equal
oxygen transfer capacities.
Tests need to be designed to determine where the oxygen
demand of the organic loading is exerted, i.e., what fraction
of biological oxygen demand is removed ,at the expense of
oxygen available immediately from the atmosphere and what
fraction is removed at the expense of oxygen dissolved in
the bulk liquid. It may be possible with this information
to predetermine the maximum hydraulic and organic loading
to an RBC through mass balance considerations.
The
- Re relation was developed from data taken in
tanks which were semicircular in design (Santa Clara and
Bensenville) and of angular design (Princeton) , such that,
the disc edge at all points below the water level was approx-
imately the same distance to the tank wall. The effects on
oxygen transfer of square bottom tanks: and of multiple
shafts in the same tank need to be studied.
The N_n - Re relation should be confirmed with 2 foot
and 6 foot diameter pilot plants . The ability to rely on
2 foot diameter pilot plants rather than having to use 6 foot
systems would increase the available design data by allowing
for easier piloting of a waste stream.
The Npg - Re relation was not derived from fundamental
physical considerations, but was developed empirically. A
development of the oxygen transfer capacity of rotating
discs from fundamental physical considerations should be
undertaken .
1096
-------
The N20 - Re relation should be used in'conjunction with
power tests and organic removal rates to help optimize RBC
performance and tank design.
CONCLUSIONS
1) Oxygen transfer data were collected from three RBC
pilot plants. Eight test cases using variations in water
level, number of discs per tank, and distances between the
media and tank bottoms were studied. The results from these
tests have been used to develop an empirical correlation
between a dimensionless oxygen transfer grouping, N2Q, and
a Reynolds number, Re, incorporating net flow relations in
the tanks. The dependence of N~0 on Re is suspected to be a
function of the media design. The closer the design is to a
flat plate, the smaller the pumping capacity of the discs and
the smaller the dependence of the N,.,,. coefficient on the
Reynolds term.
2) Conventional methods of analysis of the scale-up
of rotational effects between small pilot plants and full
scale plants, i.e., tip speed, rpm, and surface area turnover
rates, failed to provide a correlation between oxygen trans-
fer data from the eight pilot test cases.
3) A method of scaling the rotational effects between
pilot and full-scale RBC systems is suggested where the
desired operational rpm of the pilot plant is calculated from
tank geometries of the pilot and full scale RBC systems .
If the two systems have similar surface area to volume ratios,
then the scale-up factor is to equalize the oxygen transfer
coefficient, K_aon(min ). If the surface area to volume
Lt £\J
ratios are not equal, then the scaling factor is the parameter
(K_aoriV./a) , where V, is the volu
ij e.\i t C.
the active area of the RBC discs.
(K_aoriV./a) , where V, is the volume of the tank and (a) is
e.\i t C.
NOMENCLATURE
2
a active surface area of the RBC (ft )
A area between the RBC media and the tank at the reference
2
cross-section (ft )
oxygen concentration (mg 02/1)
1097
-------
E
f
H
K
KLa20
L
m
n
N20
OTR
q
Q
r
rw
R
Re
rpm
p(mu)
oxygen saturation concentration (mg 09/1) ; 9.2 mg/1 at
20°C
ratio, tank volume/active disc area, V./a, (ft)
distance from disc edge to tank bottom (ft)
void volume; volume of water per volume of disc, 0.935
(dimensionless)
water depth of center of tank
-8
experimental constant, 5.7 x 10 , (dimensionless)
conversion factor, 35.32 (mg x ft / g x 1)
M "I
oxygen transfer coefficient (min )
oxygen transfer coefficient corrected to 20°C (min )
length of RBC disc section (ft)
area of RBC disc per area of a circle with the same
2 2
radius, 3.02 ft /ft , (dimensionless)
number of RBC discs (dimensionless)
dimensionless oxygen transfer number
oxygen transfer rate corrected to 20°C (g 02/min)
wetted perimeter; solid boundaries at reference cross-
section (ft)
number of discs per length (n/ft) = 12
theoretical pumping rate of the RBC (ft /min)
RBC disc radius (ft)
distance from RBC shaft center to waterline (ft)
hydraulic radius, A/p, (ft)
Reynolds number (dimensionless)
rotational speed (min )
viscosity (19.19 g/ft min) at 20°C
1098
-------
u (nu)
v
Vd
Vt
w
-5 2
kinematic viscosity (65.44 x 10 ft /min) at 20°C
average water velocity (ft/rain)
active volume of RBC discs (ft )
volume of test tank (ft )
tank width (ft)
REFERENCES
1) Welch, F. M., "Preliminary Results of a New Approach in
the Aerobic Biological Treatment of Highly Concentrated
Wastes." Proceedings 23rd Purdue University Industrial
Waste Conference, p. 428 (1968).
2) Chesner, W. H., and Molof, A. H., "Biological Rotating
Disc Scale-up Design: Dissolved Oxygen Effects."
Progress in Water Technology, Vol. 9, p. 811 (1977).
3) Dunn, R. 0., McCarthy, W. C., and Godlove, J. W., "An
Evaluation of Oxygen Transfer in Laboratory, Pilot, and
Commercial Rotating Biological Contractors." 85th Annual
Meeting of the AIChE, Philadelphia, PA (June, 1978).
4) Borchardt, J. A., Kang, S. J., and Chung, T.. H. , "Nitri-
fication of Secondary Municipal Waste Effluents by
Rotating Bio-Discs." EPA-600/2-78-061, (June, 1978).
5) Friedman, A. A., Robbins, L. E., and Woods, R. C.r
"Effect of Disc Rotational Speed on RBC Efficiency."
Proceedings 33rd Purdue University Industrial Waste
Conference, p. 73 (1978).
6) Bintanja, H. H. J., Van Der Erve, J. J. V. M., and
Boelhouwer, C., "Oxygen Transfer in a Rotating Disc
Treatment Plant." Water Research, Vol. 9, p. 1147
(1975).
7) Antonie, R. L., Fixed Biological Surfaces - Wastewater
Treatment, CRC Press, Inc., Cleveland, OH (1976).
8) Zeevalkink, J. A., Kelderman, P., Visser, D. C., and
Boelhouwer, C., "Physical Mass Transfer in a Rotating
Disc Gas-Liquid Contactor." Water Research, Vol. 13,
no. 9, p. 913 (1979).
1099
-------
9) Standard Methods for the Examination of Water and Waste-
water, 14th Edition, APHA, AWWA, and WPCF, eds., Wash-
ington, D.C. (1976) .
10) Wu, Y. C., Yang, C. H., and Kao, J. F., "Dimensionless
Parameters Help Describe Oxygen Transfer Rate." Water
and Sewage Works, p. 56 (October, 1978).
1100
-------
COMPARISON OF FULL SCALE RBC PERFORMANCE
WITH DESIGN CRITERIA
by
Charles Chi-Su Chou
Robert J. Hynek
Richard A. Sullivan
Autrotol Corporation
Milwaukee, Wisconsin
I.
INTRODUCTION
It is estimated that there are over three hundred wastewater treatment
plants using rotating biological contactors since the corrugated surface
was first developed and applied in the United States. The majority of the
installations in North America have been designed for the removal of BOD
and/or ammonia in domestic sewage. In the past decade, design methodology
and loading criteria of the RBC process have been defined and consolidated
through a great number of pilot studies and field monitoring programs.
iVbre recently, Autotrol Corporation has followed the monthly process perfor-
mance of many RBC treatment facilities in the United States, so that the
field operating conditions and treatment efficiency can be updated and
analyzed.
1101
-------
By keeping close contact with the operating personnel, not only the
appropriate operational measures have been recommended to the plant super-
intendents, but the process optimization could be realized according to
changing environmental conditions. In adition, the field data are regularly
examined against design criteria in order to verify the current process
design method and to reveal problem areas which affect system performance.
In this report, chronological field data from seven RBC sewage treat-
ment plants will be discussed. These plants have been in operation for one
to six years. They are serving the communities of Gladstone, Michigan,
Toppenish and Woodland, Washington, Rhinelander, Wisconsin, Georgetown,
Kentuclcy, Caddilac, Michigan, and Fort Khox, Kentucky. The first five were
designed for secondary treatment with RBC following either primary clarifiers
or fine screens. In the latter two plants, RBC's are used for nitrifying
activated sludge effluent (Caddilac), and for removing BOD and ammonia
following primary treatment (Fort Khox).
The monthly performance reports from the seven plants for the past one
to three years are summarized. The actual performance has 'been compared
with the designed effluent requirements or the NPDES permit conditions.
All the RBC plants being discussed employ mechanical drive systems for
rotation and the nominal speed has been 1.6 rpm. The wastewater flows are
either normal or parallel to the shafts. Both flat-bottom and contoured
tanks are used. Although standard density medium is used in all the plants,
latter stages at Rhinelander and Fort Knox are also equipped with High
Density media.
II. SECONDARY TREATMENT AFTER GRAVITY PRIMARY CLARIFICATION
The unit operations of the plants at Gladstone, Michigan, Woodland and
Toppenish, Washington are shown in Table 1. The original Bio-Surf design
flow, surface area, and staging are also tabulated. The wastewater flows
in all three plants are parallel to the RBC shafts, and trapezoidal tanks
are used. Conventional secondary clarifiers are employed for the removal
of RBC effluent suspended solids. Secondary clarifier effluents are chlor-
inated and discharged. The sludge is anaerobically digested at Gladstone ,
and Toppenish, but aerobically treated at Woodland. The digester supernatant
is recirculated to the head end of the plants. Each shaft is divided by
one cross- sectioned baffle, resulting in two stages per shaft.
The monthly average BOD and suspended solids data of the three plants
are shown in Table 2, 3 and.4 respectively. The summer and winter perfor-
mance of each plant is compared with the design requirement.
1102
-------
Gladstone Treatment Plant
This is the first rotating contactor sewage treatment plant built in
the U.S. with an EPA construction grant, and has been in operation since
March of 1974, for upgrading the original primary treatment. The design and
NPDES requirement on BOD and SS is 30 mg/1. Phosphorus removal is achieved
by alum and polymer coagulation in two rectangular primary clarifiers.
As shown in Table 2, this plant has not only met the discharge limits
on BOD and SS, but has also produced a highly nitrified effluent. The
average ammonia nitrogen in the summer months is less than 1.1 mg/1. From
April to June 1979, when the average flow was 1.0 MGD, the same, as the
design flow, the effluent BOD and SS were 15 mg/1 and 17 mg/1 respectively,
well below the permit level. Some of the earlier field data and background
information of this plant was documented by Antonie (1). In general, the
Gladstone plant has been doing very well.
Toppenish Treatment Plant
This 1.3 MGD plant was designed for secondary treatment with discharge
requirement of 30 mg/1 and/or eighty-five percent removal of both BOD and
suspended solids. The plant was started up in the summer of 1978. In 1979,
the hydraulic loading during summer months already surpassed the design
level due to infiltration of irrigation water (Table 3). For all the flow
and temperature conditions experienced, the process performance has been
quite stable. The average effluent BOD and SS are 6 mg/1 and 8-9 mg/1,
which account for over ninety-three percent removal. In addition, nitrifiers
are believed to dominate on latter stages as indicated by biological slime
determinations, although the extent of nitrification awaits to be assessed
by ammonia analysis.
Woodland Treatment Plant
This small treatment plant serves a community of twenty-two hundred
people. The two-shaft RBC process has been on stream since January of 1975
and requires minimum attention. Each shaft is outfitted with three inter-
stage baffles resulting in 4-stage treatment. The monthly average BOD and
suspended solids data from August 1978 to July 1979 are summarized in Table
4. The effluent BOD and SS averaged 20 mg/1 and 14 mg/1 for the twelve
months. Although the plant is receiving less than the designed hydraulic
loading, the waste strength has been substantially higher than domestic
waste concentrations and the performance was quite satisfactory despite
water temperature declines to 45°F temperature range.
1103
-------
III. SECONDARY TREATMENT AFTER FINE SCREENING
The two plants discussed herein, Rhinelander, Wisconsin and Georgetown,
Kentucky, both employ fine screens (Hydrasieves) ahead of RBC process for
removing debris from raw wastewater. The flow is perpendicular to the
shafts. Raw BOD and suspended solids loads are essentially unchanged by the
screening and thus, solids loadings are somewhat higher than at plants hav-
ing primary clarifiers. There are two RBC tanks in both plants. The
Rhinelander plant has flat-bottom tanks with vertical interstage baffles,
whereas the Georgetown plant has a contoured tank baffling configuration.
The unit operations and design information of the two installations
are shown in Table 5. The lower design hydraulic loading at Rhinelander is
due to the more stringent design effluent requirement and lower water temp-
erature in the winter. Process performance at these plants has been very
satisfactory and relatively unaffected by the additonal solid loading in
the contactors, as the plants continue to produce effluent quality below
permit requirements.
Rhinelander Treatment Plant
The average flow from July 1978 to August of 1979 is seven percent
higher than the 1.9 MGD design flow as shown in Table 6. The influent and
effluent BOD are almost identical with the projected levels. The influent
suspended solids is 11 mg/1 higher than the designed 160 mg/1. It is
reasoned that the nature and concentration of the influent SS and the higher
overflow rate in the final clarifiers result in the average 20 mg/1 effluent
SS, which is 4 mg/1 higher than the expected design figure. However, the
effluent has been consistently in compliance with the 30 mg/1 discharge
permit conditions, and a few of the operational problems such as malfunction-
ing grit chamber, faulty digester operation, and insufficinet ventilation in
the RBC building were corrected in the past year.
Georgetown Treatment Plant
The RBC process was started in June, 1975. The wastewater is of medium
strength in BOD and SS. A summary of recent monthly data is shown in Table
7. The performance has been quite stable as indicated by the consistent BOD
and SS removal of over 90% for the twelve months surveyed. The average final
effluent BOD and SS are both 16 mg/1, well below the 30 mg/1 discharge limit.
Again, the lack of primary solid removal does not affect the wastewater
treatment efficiency of the rotating contactors, although no consideration
was taken for the extra solids load in the original design.
IV NITRIFICATION OF ACTIVATED SLUDGE, EFFLUENT
The advanced treatment plant at Cadillac, Michigan, employs an activat-
ed sludge system with FeCl3 addition for BOD and phosphorus removal. The
Bio-Surf process is used for nitrifying secondary clarifier effluent with
final polishing provided by sand filters. (A flow diagram of this installa-
tion is shown in Figure 1). The plant was designed for 2 MGD and the RBC
section was started up in March of 1976. Eight shafts with a total surface
1104
-------
area of 7.6 x 10^ sqaure feet are installed in two parallel flat-bottom
tanks. The flow is normal to the shafts through slots at the bottom of
interstage baffles. The nitrification process has been operated in a
four-stage mode.
The NPDES requirement on effluent ammonia nitrogen is 1.5 mg/1 from
June to October. During the summer of 1976, the effluent ammonia nitrogen
slightly exceeded the limit. It was later revealed by a thorough tracer
study funded by the Autotrol Corporation, that the RBC nitrifying capacity
was penalized by short circuiting along the bottom of the tank. Subsequent-
ly, a simple adjustment was made to the baffle dimensions to improve resi-
dence time distribution in June of 1977. The effluent ammonia nitrogen
since that time has been below the NPDES permit shown in Table 8.
Since nitrification is only required during the summer, the biological
contactors are stopped during the winter. By doing so, savings in energy
and 0 § M labor can be achieved without affecting other sections of the
treatment plant. Usually, the rotating contactors at Cadillac are restarted
in April so that active nitrifying biomass can be developed to meet stringent
effluent ammonia requirements from June 1st. Dissolved oxygen in RBC tanks
varies from 3.5 to 4.0 mg/1, and alkalinity declines 8.0 mg/1 per part per
million of ammonia nitrogen oxidized. The pH in RBC effluent is about 7.6
during the summer months, when the influent pH is 7.8. Thus, the nitrifica-
tion proceeds at a rather favorable pH range.
V BOD AND AMMONIA REMOVAL FOLLOWING PRIMARY SETTLING
The 6 MGD sewage treatment plant at Fort Khox, Kentucky was designed
for combined removal of BOD and ammonia. After the conventional primary
treatment, wastewater is distributed to six flat-bottom RBC tanks. There
are three standard media and three High Density media shafts in each bay,
with verticle interstage baffles between adjacent shafts. Thus, the RBC
process is operated in six-stage mode. The High Density media on stages 4
through 6 is supposed to support the relatively thin nitrifying biomass.
The thirty-six Bio-Surf shafts were started in November of 1977. The
performance data from November of 1978 to October of 1979 are shown with the
effluent requirements in Table 9. Both the effluent BOD and the suspended
solids have met the design quality. The winter ammonia nitrogen also met
the discharge limit of 5 mg/1, but the average summer effluent ammonia is
2.7 mg/1, exceeding the required 2.0 mg/1 standard.
A field examination of the biomass on the last three stages reveals
that the apparent autotrophic!biofilm is extremely thin and the color is
darkei" than the average nitrifiers on the contactor surface. It is suspected
that the unique biofilm appearance and the inadequate nitrifying activity
might have been caused by the sub-neutral pH in RBC tanks, especially toward
the latter stages where pH as low as 6.5 has been detected. The range of
the pH of RBC effluent is also shown in Table 9.
i
Earlier studies have shown that the nitrifiers either suspended growth
or attached thin film would be severely inhibited when the pH drops below
1105
-------
6.8 (2) (3). A subsequent full scale kinetic study conducted at the
Autotrol Corporation also demonstrates that the nitrification rate declines
from 0.31 Ib. of ammonia nitrogen removed per thousand square feet per day
at pH 7.0 to 0.17 lb/10002ft. day at pH 6.5. At Fort Knox, the critical pH
in the nitrification stages is the result of insufficient buffer capacity of
the wastewater. Although the average influent total alkalinity is about
160 mg/1, the pH tends to decrease below 7.0 when the first forty percent of
the alkalinity is consumed by the nitric acid produced. The amount of alka-
linity consumed is equivalent to the removal of about 9 mg/1 ammonia nitrogen.
At this point, the nitrification rate would slow down significantly at
ammonia concentration less than 6 mg/1 due to both the unfavorable pH and
the decreasing ammonia concentration. As one would expect, nitrification
kinetics is first order with respect to ammonia when its level falls below
5 mg/1 and removal proceeds at slower rates.
VI COMPARISON OF BIO-SURF PERFORMANCE WITH DESIGN CURVES
The process design of the RBC treatment plant discussed above were
based upon Autotrol's criteria, which have resulted from a great deal of
effort in piloting, monitoring full scale operation and kinetic modeling (4)
(5). The major design curves for both BOD and ammonia removal are attached
as Appendix I and II. Since the fixed-film biological contactors absorb and
oxidize the impurities in the soluble forms, the general.design practice
has been the derivation of the required surface area from the influent
soluble BOD (and/or ammonia), and the desired effluent soluble BOD ammonia
oxidation (Appendix III and IV). If soluble BOD data is not available,
it is estimated by subtracting the suspended portion according to the
empirical formula shown in Appendix V.
For ease of comparison, the hydraulic loading, the calculated influent
soluble BOD, the field effluent BOD and ammonia are shown in contrast with
the predicted effluent quality in Table 10. Among the seven plants, Glad-
stone, Toppenish, Rhinelander, Georgetown and Cadillac are producing better
effluent than predicted. Woodland's effluent BOD has also been meeting the
NPDES permit requirement, however, it shows some deviation from the predicted
level which might be due to unusual sludge handling practices. This plant
awaits to be further assessed when the hydraulic loading approaches 3 to 4
gpd/ft2 in the future.
At Fort Knox, the effluent BOD is essentially the same as predicted.
The actual ammonia nitrogen, however, is higher than the calculated figure.
As discussed earlier, this discrepency is reasoned to be caused by the un-
favorable pH in the nitrification section. It is concluded that the nitri-
fying capacity of the equipment is penalized by the acidic environment.
VII CONCLUSIONS
1. All the seven plants surveyed have met the NPDES permit require-
ments on effluent BOD and suspended solids.
2. The RBC plant treating activated sludge effluent has met both
the NPDES limit and design predictions on effluent ammonia
1106
-------
nitrogen.
3. Five out of the six plants designed for BOD removal have per-
formed up to Autotrol's process design criteria. The only
exception is the underloaded Woodland treatment plant, which
awaits further assessment when hydraulic loading approaches
design level.
4. The data suggests that the elimination of primary settling of
suspended solids does not affect the BOD removal capacity of
the RBC process.
5. The Fort Knox treatment plant, designed for combined BOD and
ammonia removal, is meeting the effluent BOD, SS, and the
winter ammonia requirement. The average summer time effluent
ammonia is 0.7 mg/1 higher than the required 2.0 mg/1. Investi-
gations point to retardation effects on nitrification at the
sub-neutral pH values observed at this installation.
1107
-------
VIII REFERENCES
1. Fixed Biological Surfaces - Wastewater Treatment
by Ronald L. Antonie - 1976 CRC Press, Inc.
2. Process Design Manual for Nitrogen Control
USEPA Technology Transfer - October, 1975
3. Huang, C. S., and N. E. Hopson, "Temperature and pH Effect on
the Biological Nitrification Process". Presented at the Annual
Winter Meeting, New York Water Pollution Control Assoc., N.Y.C.,
January, 1974.
4. Famularo, J., J. A. Mueller and T. Mulligan, "Application of Mass
Transfer to Rotating Biological Contactors" JWPCF 50: 653-671,
1978.
5. Mueller, J. A., and P. Paquin, "Nitrification in Rotating Biolog-
ical Contactors" Presented at 51st Annual Conference of Water
Pollution Control Federation, Anaheim, California, October 1-6,
1978.
1108
-------
FIGURE 1
PROCESS FLOW DIAGRAM OF CADILLAC WASTEWATER TREATMENT PLANT
PLANT INFLUENT
Filtrate
Vacuum Waste Sludg
Filter
1
Sludge Disposal
GRIT CHAMBER
1
1
e PRIMARY
Chemical Coagulants
FLOCCULATOR - CLARIFIERS
I
AERATION TANKS
i
FINAL CLARIFIERS
I
RECYCLED SLUDGE
ROTATING BIOLOGICAL CONTACTORS
I
SAND FILTERS
1
CHLORINE CONTACTS TANKS
FINAL EFFLUENT
1109
-------
TABLE 1
UNIT OPERATIONS AND DESIGN DATA
OF GLADSTONE, TOPPENISH AND WOODLAND
SEWAGE TREATMENT PLANT
PLANT
1.
2.
3.
4.
5.
6.
7.
8.
9.
UNIT OPERATION
Bar Screen
Comminutor
Grit Chamber
Primary Clarifier
Bio- Surf
Design Flow, MGD
Design SA ft2
No. of Stages
No. of Shafts
Secondary Clarifier
Disinfection
Sludge Digestion
Sludge Dewatering
vs Disposal
GLADSTONE
Yes
Yes
Yes
(non- aerated)
Yes
1.0
5.2 x 105
6
6
Yes
Chlorination
Anaerobic
Land
TOPPENISH
Yes
Yes
Yes
(non- aerated)
Yes
1.30
5.9 x 105
4
6
Yes
Chlorination
Anaerobic
Land
WOODLAND
Yes
Yes
Yes j
(aerated) j
Yes
0.56
1.4 x 105
4
2
Yes
Chlorination
Aerobic
Land
1110
-------
ft
ft
w
bflft
hJ
•3-tOtOCNlrHOrHOrHrHrH^
V
T— I
OO <3- r-H O OO rH
LO \r> o oo LO o cxi
t—1 t—! rH i—I rH CNJ rH
LO
I • I
r-H rH iH
CD
0-j
K)
CT>
rH
CN1
W
ft
w
b$ft
1
CO
CO
H
CJ
hH
S
§ 8
CO
CTJ
CO
CJ
ft
ft
bfllft
§
LOLOVOOOOOOOOOO
T-HCXIrHrHrHrHrHrHrHrHrHrH rH t-H rH rH t—I
VOLOOOrHtOCXIOOtO \O CXI
o^t>-t>-i>-c^ooocr>Oi— i
\OC3OChr- l
CTl
O
00
^
OCDOOOCDCDOOOOCD CDOCDrHrHOOO O
oo
cr>
IS)
OO
OO l>»
OO
oo
r-~ oo oo O)
— oo t~^ c^~ o^ r-^
CD CD rH
Ofl P-< bfl O;
crj CD cc> *^^
P
00 OO 1^ fH -
^~ t^ oo oo •- , CD^
O-r-i CD>sd § j. ^S
f-iCDCD fx,f-i—-CT>C^F~-
CD rQ rQ H CtJ
1111
-------
co
co
o
to
o
to
1
-
O §
^ y,
CO v— '
jp*_j I
O
to
o
to
i
ft
o
•
I-l
g
•H
CO
-------
TABLE 3
SUMMARY OF MONTHLY DATA AT
TOPPENISH, WASHINGTON
PERIOD
1979
January
February
March
April
May
June
July
August
September
October
November
'Summer Average
May to
October
FLOW
MGD °
0.73
0.73
0.72
0.90
1.33
1.50
1.57
1.61
1.34
1.05
0.89
1.40
Winter Average:
November to 0.79
April
Deisgn **
Condition
1.3
F
54
55
57
57
62
64
64
65
66
64
60
64
57
BODr
RAW
177
153
146
118
183
126
86
109
100
114
175
120
154
mg/1
FINAL EFFL.
7
6
5
*
8
4
8
6
4
4
5
6
6
30
NPDES Permit** 1.3
30
S.S. mg/1
RAW
104
159
178
136
183
189
162
125
124
169
149
159
145
FINAL EFFL.
9
6
16
*
6
5
11
. 7
8
13
5
8
9
30
30
Estimated RBC influent soluble BOD = (o.75) (Raw BOD) -(0.6) (0.4) (Raw SS)
* Sampler malfunctioned, available data not included in average
** And/or 85% reduction of raw BOD and SS
1113
-------
TABLE 4
SUMMARY OF MONTHLY DATA
WOODLAND, WASHINGTON
PERIOD
August '78
September '78
October '78
November '78
December '78
January '79
February '79
March '79
April '79
May '79
June '79
July '79
Summer Average
Winter Average
Design
Condition
NPDES Permit
tLU
MOD
0.207
0.225
0.223
0.214
0.249
0.247
0.310
0.302
0.307
0.289
0.282
0.272
0.250
0.272
0.56
IW
OF
66
64
61
56
50
48
45
47
48
50
52
56
58
49
BODT
RAW
243
247
249
205
200
271
213
176
178
162
160
103
194
207
, mg/1
EFFL.
20
19
23
18
22
20
20
22
18
19
20
22
20
20
30
30
SS,
RAW
296
287
430
358
184
235
211
225
174
170
177
203
260
231
mg/1
EFFL.
16
10
15
19
15
11
18
18
8
11
14
14
13
15
30
30
1114
-------
r
TABLE 5
UNIT OPERATIONS AND DESIGN DATA OF
PLANTS USING SCREENS FOR INITIAL TREATMENT
PLANT
UNIT OPERATION
1. Bar Screen
2. Grit Chamber
3. Fine Screen
4. Bio- Surf (RBC)
Design flow, MGD
Design Surface, ft.^
Number of Stages
Number of Shafts
5. Secondary Clarifier
6. Disinfection
7. Sludge Digestion
8,. Sludge Dewatering
RHINELANDER, GEORGETOWN,
WISCONSIN KENTUCKY
Yes
Yes
Yes
1.9
11 x 10s
5
10
Yes
Chlorination
Anaerobic
Belt Press
Yes
No
Yes
3.0
8 x 10s
4
8
Yes
Chlorination
Anaerobic
Sludge Bed
1115
-------
TABLE 6
SUMMARY OF MONTHLY DATA AT
RHINELANDER, WISCONSIN
PERIOD
July '78
August '78
September '78
October '78
November '78
December '78
January '79
February '79
March '79
April '79
May '79
June '79
July '79
August '79
Average :
Design
Condition
NPDES Permit
FLOW
MGD °F
2.08
2.27
2.98
3.07
1.85
1.45
1.34
1.44
1.75
2.12 46
2.09
2.36
1.87 59
1.75
2.03
1.9
BOD .
RAW
122
152
187
154
144
119
153
144
130
116
120
191
155
121
143
145
, mg/1
EFFL.
13
14
16
18
18
14
18
15
15
12
10
16
12
11
14
15
30
SS,
RAW
163
191
171
131
166
167
207
164
154
168
132
243
181
162
171
160
mg/1
EFFL.
23
25
23
21
22
19
24
22
17
23
15
22
11
14
20
16
30
Estimated RBC influent soluble BOD = Raw BOD - (0.6) (Raw SS)
1116
-------
TABLE 7
SUMMARY OF MONTHLY DATA AT
GEORGETOWN, KENTUCKY
EODr (ing/1)
SS (mg/1)
PERIOD
July ' 78
August '78
September '78
October '78
November ' 78
December '78
January '78
February '79
March '79
April '79
May '79
June '79
Annual Average
December to
April '79
Design
Condition
NPDES Permit
FLOW, MGD
1.08
1.24
1.11
1.19
1.36
1.87
1.69
1.62
1.31
1.65
1.06
1.00
1.35
1.63
3.0
3.0
RAW
202
212
206
201
204
225
214
224
206
213
239
205
213
216
FINAL EFFL.
15
14
15
14
14
16
14
17
16
18
20
16
16
16
30
30
RAW
202
207
216
202
205
195
204
204
205
207
216
201
205
203
FINAL EFFL.
17
14
16
12
15
14
12
16
17
18
18
17
16
15
30
30
Estimated RBC Influent Soluble BOD = Raw BOD - (0.6) (Raw SS)
1117
-------
TABLE 8
SUMMARY OF MONTHLY DATA AT
CADILLAC, MICHIGAN
Ntfr-N, mg/1
PERIOD
November '77
December '77
January '78
February '78
March '78
May, '78
June '78
July '78
August '78
September '78
October '78
June '79
July '79
August '79
September '79
October '79
FLOW, MGD.
1.36
1.31
1.42
1.47
1.46
1.36
1.44
1.36
1.41
1.42
1.52
1.62
1.61
1.61
1.50
1.51
WASTEWATER
TEMPERATURE OF
60
55
53
52
53
58
65
68
69
67
63
63
66
67
66
63
RBC INFL.
10.2
16.7
16.3
15.5
15.0
4.3
9.6
6.6
8.6
8.8
6.8
13.1
8.9
8.6
5.9
11.0
RBC EFFL.
1.2
2.6
2.2
2.4
'2.7
1.0
2.0
0.6
1.5
0.9
0.4
1.4
0.7
0.8
0.6
1.5
FINAL
EFFL.
2.0
2.3
2.5
0.7
1.8
0.6
1.5
0.8
0.2
1.0
0.4
0.6
1.2
Summer Average
June to October
Summer NPDES
Requirement
1.50
66
8.5
1.0
0.86
1.5
1118
-------
cn
E-i
<;
<
| g
i3 H
H fTl
•^
C5 **
*^? K^1
r^— i pSj
ft ;z
o, S
>< r->
PH P*!
§ O
*g| t^H
5g|
^j
co
§
^^
bflPlH
£ PH
fTl
[J_|
•\
S H-3
^,g
5j- ^H
*
ft
W
sj
T~H £S
bjQPn
Egg
hH
QJ
PQ
s
_5
ft
ft
w
t *i
2
rH rH
\ft
S
/~^ ! "|
O ft
« g
I-H
o
§
.
ft X
ft p.
w
ft
or
.-3 fr""<
ft ly
Q
O
s
e
LO i>. CD >o cn
•s}- rH CXI CXI K)
CD tO tO t — •*
CXI rH rH rH CXI
OO f~~- O C~~ O
!>^ t>- LO rH tO
rH rH rH
CD to vO CXI to
rH rH rH CXI
CXI ^f OO
1^ I>- rH tO O
OO f- i— 1 rH rH
o oo oo oo o^
C-^ \O vO vO VO
cn cxi vo cn LO
LO tO rH rH 1^
to ^* ^f* *vt* to
oo oo cn
^5r~s r~
cn
f-i fn t>^t>-
CD CD ^ fn -
£3 £3 c$ £3 .^
^ S § J3 H
^Q i-j ft g
cxj
•*
to
CX)
t^
cn
dD
rH
LO
OO
CD
£~^.
00
•*
cn
c~^
*.
rH
•H
5-1
•§"
cn
CXI
cn
rH
t-^
CD
rH
O^
0*1
CTi
C3
C*^*
^vf"
LO
t-o
cn
^.
•.
^
^
CX|
CXJ
LO
CXI
O
cn
rH
cn
CXI
rH
rH
i-H
t^^
CD
O
^
cn
[^,
».
CD
r|
cn
rH
to
CX|
!>.
VO
rH
OO
CXI
rH
rH
rH
t^
cn
i— 1
*
cn
!>»»
•.
rH^
r^
l>-
CXI
CXI
rH
LO
t —
rH
CX|
rH
CD
vO
CD
t>!
"^
vO
rH
CD
cn
to
cn
o-
•_
-P
JA
Go
^
LO
>*
cn
LO
oo
cn
rH
00
vD
i
LO
*
vO
CD
f — .
^j-
cn
to
cn
—
fH
CD
£3
CD
p
CO
to
CXI
00
CXI
o
CD
rH
f^.
CD
«^j-
rH
CD
i>!
i
00
*
vO
CD
t^^
rH
LO
to
cn
*.
!H
CD
f>
0
s
t^ CXI
CXI K)
cn oo
rH rH
t>- OO
tO CD
rH rH
cn ^t-
rH
cn
cn CD
o
rH
o cn
I — vo
LO OO
OO CD
to •*
!H ^
CD CD
C3 J^)
g .s
CO ^g
CD 0
to/) fcxQ
oj cd
CD CD
s 3
CXI LO
P!
rH
O CD <-H
to to R
1
LO
0
1
§
PQ
rH
CD O -P
rH CXJ o
E — '
+-*
PH
-------
^
rH
V
o
rH
OJ
O
rH
O O
|
w
M
M oo t-oo oooo
a
g
o
rV
S
i
O
I—I
B
en
S
§
31-1
'
W CM I
LOLO mm
i^o
<«
•vf 10
to en
•* ^t-
o o
oo
*
rH
rHO
O
o-i
VO rH
OO CD
O O
OO
§
I
^
(fl
f-1 rH
S
8
-P
&
00
S:
CD
00
fH «
O
CD
CD
OO
rH t-
8
O
Si
a
f-» 5-1
M CD
O
P^H
1120
-------
APPENDIX 1
<
o
or
C/>
>o
UJU
GUJ
LL.
— LU
CC
a
-j
< UJ
2%
O
COX
UJ
O)
UJ
O
O
in
in
A
LU
Q.
LU
I-
LU
o
in
'aoa aianios
I I I !
o
CO
o
CJ
LL
Q
O.
Z
Q
O
_i
O
cc
o
>-
1/OIAI 'QOa 1V1O1
1121
-------
APPENDIX 2
CO
cc
UJ
UJ
I
111
o
o
UL
o
O
LL.
DC
u.
o
A
UJ
tc
I
cc
LU
Q.
i-
oc
LU
I
LU
CO
I
Q
o.
O
O
z
Q
<
O
O
cc
Q
CM
n/ow 'N-
1122
-------
APPENDIX 3
Ill
cc
Is
CQ
£§
£*•
LU
O
o
LJLJ
or
DC
O
O
UJ
QC
I
QC
UJ
Q.
2
UJ
I-
UOlOVd NOI103UUOO
1123
-------
APPENDIX 4
TEMPERATURE
CORRECTION FOR NITRIFICATION
cc
2
o
g
o
HI
cc
cc
o
o
TEMPERATURE°F
1124
-------
APPENDIX 5
SOLUBLE BOD
Soluble BOD is determined by conducting a standard BOD test
on wastewater that has passed through the fiberglass mat filter
used for a standard suspended solids test. This gives results
comparable to a millipore filter but the test is faster and easier to
perform. Although a workable test with consistent results, this
procedure is not a measure of materials in "true solution"
because some colloidal materials will pass through the fiberglass
filter. For conducting soluble BOD tests refer to the procedure in
the Appendix of this manual.
For design purposes, soluble BOD test data on the wastewater to
be treated is, by far, the most useful information. If the wastewater
flow exists, whether it is a raw waste or primary effluent, it is' a
simple matter to run BOD tests on the filtrate from the standard
suspended solids test. For purposes of preliminary studies, initial
cost estimates, or if the wastewater does not exist, it is possible to
estimate the soluble BOD concentration from historical data on
total BOD and suspended solids, or by assuming appropriate
values for these parameters. Soluble BOD can be estimated as
follows:
Soluble BOD = Total BOD — Suspended BOD
Suspended BOD = K x Suspended Solids
Soluble BOD = Total BOD — K x Suspended Solids
where
K = 0.5 to 0.7 for most domestic wastes
= 0.5 on raw sewage where suspended solids values are
significantly higher than the total BOD values e.g., BOD
= 200 mg/l and SS = 250 mg/l (typical of strictly
residential area)
= 0.6 on raw sewage where suspended solids values are
about the same as the total BOD values (typical of
municipalities with commercial and industrial
discharges)
= 0.6 on primary treatment effluents'
= 0.5 on secondary treatment effluents
1125
-------
-------
FIRST USA AIR DRIVE RBC UNITS
OPERATIONAL EXPERIENCE AND PERFORMANCE
INDIAN CREEK WASTEWATER TREATMENT PLANT
CINCINNATI (CLEVES), OHIO
By
M. M. Schirtzinger, President
M. M. Schirtzinger & Associates, Limited
Consulting Environmental Engineers
Chillicothe, Ohio
PLANT DESCRIPTION
The Indian Creek Wastewater Treatment Plant provides secondary treat-
ment for an average daily flow of 0.5 million gallons per day (mgd) and
receives peak flows up to 1.7 mgd of domestic sewage. Six air—driven rotary
biological contactors (RBCs) (Aero-Surf units supplied by the Autotrol
Corporation) are preceeded by grit removal and two 40-foot diameter pri—
Inary clarifiers and are followed by two 40—foot diameter secondary clari-
fiers. Secondary sludge is returned to the primary clarifiers and combined
with primary sludge. Sludge is removed from the primary clarifier to a
sludge holding tank arid is trucked away to another plant location for ulti-
mate disposal.
The Indian Creek Wastewater.Treatment Plant (WWTP) is located on the
banks of the Ohio River at the foot of a monument to William Henry Harrison
and is bordered by park land on two sides. The plant was designed by
M. M. Schirtzinger & Associates, Limited (MMSAL) and is operated by the
Cincinnati Metropolitan Sewer District (MSD).
The plant receives domestic sewage from the communities of Cleves and
North Bend and the southwestern part of Hamilton County. Flow through the
plant is by gravity. Raw waste strength is low in biological oxygen demand
(BOD) averaging from 100 to 130 milligrams per liter (mg/1), and no indus-
trial waste is received for treatment.
1127
-------
FIRST USA AIR DRIVE RBC
The plant was originally designed using mechanically driven 20-foot
shafts in three stages. Just prior to the beginning of construction, the
plant was modified to use the air drive system because of the additional
reliability which the air drive units would provide, the power savings
which could be expected, and because the flood elevation made protection
of the mechanical drive units difficult.
Two sets of three single-stage 20-foot shaft length RBC units are
arranged for parallel flow. Each RBC has 80,000 square feet of surface
area (a total of 480,000 square feet). Each set of three RBCs is in a
common tank with dividing baffles to provide three separate stages.
Ammonia reduction is not required at this plant. Air is supplied by a
single Lamson blower with a 40 horsepower (H.P.) driver.
The original mechanically driven RBC units would have been equipped
with 7.5 H.P. drive units requiring a total connected horsepower of 45.
The RBC air drive blower is provided with a 40 H.P. motor; however, it is
throttled to use 25 H.P. or 4.16 H.P. per drive. This results in a .savings
in the daily power use. However, the precise savings in power is not known
since the power draw for the mechanical drives is unknown and because a
portion of the air supply from this blower is used to aerate the feed well
of the primary clarifiers.
ADVANTAGES OF THE AIR DRIVE UNITS
The advantages of the air drive units are not limited just to power
savings as other positive benefits are available which include: (1) indi-
vidual speed regulation of each stage, (2) fewer moving parts and electri-
cal equipment which are subject to corrosion, (3) augmentation of the bio-
reaction due to the air supply, (4) the use of air for other unit processes,
(5) the elimination of effluent reaeration, (6) no decrease in efficiency
during wintertime operating temperatures, and C7) safety. These advan--
tages are more fully discussed as follows:
1. Because the air flow to each stage can be regulated by a throttling
butterfly valve, the revolutions per minute (rpm) of each stage can
be individually adjusted. Thus, each stage rpm can be proportional
to the BODc applied. The first stage -would have the highest rpm.
The optimum speed of the units ideally would be 1.3 rpm, 1.1 rpm,
and 1.0 rpm. For units with low BOD^ load, air flow and power can
be reduced and speed can be as low as 0.8 rpm. The relationship
between speed of rotation and quantity .of'air is shown oh
Figure 1.
2. Since there are no mechanical or electrical parts inside the RBC
enclosure, corrosion and oxidation of chain drives, gear reducers,
1128
-------
FIGURE I
250
AIR REQUIREMENTS
BASED ON 25-FOOT MEDIA ASSEMBLIES
AT AMBIENT PRESSURE AND TEMPERATURE
200
CO
5
UJ
cc
HI
a
150
UJ
03
™ 100
u.
o
u.
o
50
HI-DENSITY MEDIA
STAN
DARD MEDIA
0.8
0.9 1.0 1.1 1.2
ROTATIONAL SPEED, RPM
1.3
1.4
1129
-------
motors, and electrical connections is eliminated. The air header
is corrosion resistant PVC pipe. Valves are located outside the
enclosure above ground and are easy to adjust and maintain.
3. The use of air augments the unit processes by providing the at-
tached growth system with a surplus of oxygen. For high strength
wastes and heavy biomass, increased air is required as shown on
Figure 2. At the Indian Creek plant, dissolved oxygen (DO) in-
creases through each stage for a total DO increase of 1 to 3 mg/1.
4. Because air is available, it can be used for other unit processes
in the plant such as to aerate grit chambers and feed wells of
the primary clarifier; and all the air can come from a single air
blower unit. This results in capital cost savings because the
number of motors, starters, wire, etc. can be reduced.
5. Because of the air added for driving the RBC units, effluent re-
aeration is not required. Sufficient DO is added to the waste
flow to maintain DO in the final effluent above 5 mg/1 as required
in the Indian Creek plant NPDES permit.
6. In cold climates, the covered RBC units are not subjected to freez-
ing or a decrease in biological activity as a result of cold ambient
air temperatures. Tests with ambient air temperature of Q°C show
that there is an increase of 1°C in the wastewater from the pri-
mary clarifier effluent to the discharge of the final KBC stage.
This added heat comes from the heat of compression generated in
the air blower.
7. Since all lubrication and adjustments can take place outside the
RBC enclosure, personnel do not have to enter into the RBC en-
closures and thus safety hazards are eliminated.
DISADVANTAGES OF THE AIR DRIVE UNITS
The only disadvantage to air drives would occur on a long term shutdown
of the RBC units whereby an unbalanced biological growth on the wetted por-
tion of the RBC units becomes so heavy that the unit would not again rotate
with only the buoyancy provided by the air drive system. If such a situa-
tion occurred, it would be necessary to totally dewater the RBC units so
that the biomass sloughed off and the shaft was again in balance to resume
rotation.
This disadvantage would be avoided by the use of a backup air blower
and with standby power. These backup units are normally provided for in
most plants and would be required for continuity of operation on Environ-
mental Protection Agency financed projects.
1130
-------
FIGURE 2
CO
h-
LU
UJ
ID
O
LLt
IT
O
CO
tr
o
o
to
in
CM
o
o
o
in
in
CM
U-
o
o
o
Q
m
_i
o"
z
Q
O
_J
Q
O
m
tu
CQ
co
in
o
o
in
o
CD
o
CM
O
O
o
o
o
CO
o
CO
,01 U3d UIV 1N3ISWV JO
1131
-------
PLANT PERFORMANCE
Average performance of the Indian Creek plant for the reduction of
BODc, suspended solids, ammonia nitrogen, and nitrate nitrogen removal as
well as other parameters is shown on Table I.
the various loading rates shown.
Data listed is averaged for
TABLE I
Parameter
BOD
BOD5 (Soluble)
COD
Nk (Soluble)
NH3H
Organic N
Suspended Solids
Suspended Solids
(Total)
(Volatile)
Percent Reduced at Loading Rates.
0.65 gpd/sf 1.04 gpd/sf 2.08 gpd/sf
97.0%
95.4%
82.0%
91.2%
100.0%
42.9%
94.8%
97.1%
89.5%
91.6%
54.0%
90.0%
83.3%
90.2%
50.2%
85.5%
COMPARISON WITH DESIGN DATA
Actual performance of the plant compared to data used for design pur-
poses shows that the Autotrol design data is reliable for air drive units.
Comparing the BODc loading and effluent obtained from the operation at the
Indian Creek plant shows excellent correlation with the design data pre-
sented in Figure 3 for soluble BODc loading in the range of 40 to 70 mg/1.
CORRELATION OF OTHER DATA
Correlation with other operational, parameters at a loading rate of
0.65 gpd/sf is shown in Table II for the shafts designated as North (N-l)
and South (S-l), etc.
TABLE II
North Shafts
N-l
N-2
N-3
South Shafts
S-l
S-2
S-3
Biofilm Thickness,
inches
0.067 0.013 0.011
0.040* 0.011* 0.005*
Effluent Soluble
BOD5, mg/1
(Influent: 33 mg/1) 12
1132
-------
FIGURE 3
o
CO
o
LU
tr
Q
O
CO
a:
LU
LU
O
H
CO
LU
O
a
-j
o
o
m
LU
_j
m
D
h-
LU
U_
Z
\
X.
'S
"^
-
..
-..
\
s
s^
x,
•-
\
X
x
X
•^
•--
\
s
s
*s,
X
•-^
— -
-^_
^_
•~-
s
s
s
*•>
"--
..
~.
\
s
XJ
^
-.
^
---
-«>
\
's
•x
•-
^
-.
S
s
X
-.-.
• ^
..
"*"*
-
>.
s
X
v
^
N,
"^
-_
-
_.
•..
S
\
pt
N
s
^>
^,
•»
\
V
\
'
\
V
s
•^
*•
v
^
-
s
\
s
w
s
\
s
s
^.
V
\
\
\
\
s
s^
x
'•-.
\
\
\
^
\
s
X
X
•s.
"^
\
•t
\
\
\
y
s
[s
v%
,
^
*•.
*
\
v
.
\
K
s
V
s.
N
-•
\
\
\
\
\
\
X
x_
"•-
\
\
V
\
\
K,
\
•
-.
•-~.
\
\
\
s,
\
-^
•-..
-•
•^
•..
s
s
>
V
\
LL.
o
IO
in
fii
oc
D
rr
LU
Q.
S
l-
nr
in
K-
*i
*>
^f
LU
h-
in
-------
TABLE II
(Continued)
North Shafts
South Shafts
Applied Carbon, Ibs/
1000 ft2 i
Specific Removal
Rate, Carbon i
NHo-N, mg/1 (Influent:
15.4 mg/1) :
Applied HN3-N, Ibs/
1000 ft2 0.250 0.180 0.026
N-l
1.54
1.34
.1.1
N-2
0.20
0.13
1.6
N-3
0.07
0.04
0.0
Specific Removal
Rate, NH3-N
Shaft RPM
*Estimated
0.070 0.154 0.026
1.18 1.20
1.36
S-l
0.54
0.44
2.8
0.250
S-2
0.10
0.05
0.0
0.045
S-3
0.05
0.03
0.0
0.000
0.205 0.045 0.000
1.54 1.28 1.36
OTHER ENERGY CONSERVATTON FEATURES OF THE INDIAN CREEK PLANT
In order to maximize the efficiency of all power use at the Indian
Creek plant, advantage was taken of the heat generated by the RBC blower
units. This is used for building heating by having the furnace intake
air obtained from the blower room. In the summer, room ventilation is
achieved by using the blower intake air from the laboratory and pump room
areas.
REMOTE OPERATION
The plant is designed for automatic unattended operation. No adjust-
ments are made to the rotational speed of the RBC units, and all sludge
withdrawal is performed by pumps on timer controls. Intermittent recycle of
secondary sludge and the withdrawal of primary sludge to holding tanks is
achieved automatically. Plant effluent parameters such as DO, temperature,
suspended solids, and pH are continuously monitored and transmitted to the
Cincinnati Metropolitan Sewer District Mill Creek Plant. Chlorine residual
is monitored and controlled with a closed loop chlorine residual analyzer.
1134
-------
RELIABILITY
Since its initial operation in January 1978, the plant has operated
without mechanical failures nor interruption of flow for any reason.
Blocking of air lines or difficulties with air-driven units has not been
experienced nor has the sludge handling equipment malfunctioned during
that period of time. The only difficulties experienced to date have been
with remote monitoring sensors for DO, pH, and suspended solids.
SUMMARY AND CONCLUSIONS
Based upon the experienced to date at the Indian Creek plant, air
driven RBC units are highly reliable providing the designed level of
treatment efficiency at low power input rates. Air drives have no dis-
advantages and have numerous advantages over mechanically driven units.
These advantages include (1) individual speed regulation of each stage,
(2) fewer moving parts and electrical equipment which are subject to cor-
rosion, (3) augmentation of the bioreaction due to the air supply, (4) the
use of air for other unit processes, (5) the elimination of effluent re-
aeration, (6) no decrease in efficiency during wintertime operating
temperatures, and (7) safety.
Acknowledgement s
The author hereby acknowledges the assistance and cooperation of the
Cincinnati Metropolitan Sewer District, Richard A. Vanderhoof, Director,
in providing data and cooperating in making plant changes.
The graphs and other data presented in this paper were by courtesy
of the Autotrol Corporation.
1135
-------
-------
DESIGN AND OPERATION OF TWO ROTATING BIOLOGICAL
CONTACTOR PLANTS AT FUNDY NATIONAL PARK,
NEW BRUNSWICK, CANADA
By
T. Viraraghavan
Environmental Engineer
R. C. Landine
Senior Environmental Engineer
E. L. Winchester
Environmental Engineer
ADI Limited
Fredericton, N. B., Canada
And
G. H. Jenkins
Senior Design Engineer
Parks Canada
Hali fax, N. S., Canada
INTRODUCTION
Rotating biological contactor (RBC) systems were selected to
upgrade existing wastewater treatment plants (septic tanks)
treating wastewater from two campgrounds at Fundy National Park,
New Brunswick, Canada, during the summer months. As part of the
design and selection process, an investigation was carried out
on the performance of RBC equipment in general and RBC equipment
manufactured by a Canadian manufacturer in particular. Results
of this investigation are presented in the paper. The plants
1137
-------
became operational in the summer of 1979. The design features,
the operational controls and the performance of these plants are
described in the paper.
TREATMENT ALTERNATIVES CONSIDERED AND THE CHOICE
The two treatment plants, one at Headquarters campground and the
other at Point Wolfe Campground, were to be designed to treat
flows of 18 600 gpd and 26 000 gpd respectively and analysis re-
sults indicated that the septic tank effluent was relatively
weak in strength (BOD less than 100 mg/1) on occasions sampled.
Package plants incorporating suspended growth biological systems
like extended aeration, contact stabi 1 i zati on were not considered
suitable to treat a weak wastewater with fluctuating flows as it
would be difficult to develop optimum MLSS for operation and
these plants would require more skilled operation. Though a
well-designed physical-chemical treatment system would be able
to accomplish the desired results, such was not considered be-
cause it would require considerable supervision as well as
skilled operation. It was felt that fixed growth systems would
normally be acceptable treatment systems in this case. However,
trickling filter systems, because of possible odour and fly
problems and relatively lower BOD removal efficiencies, were not
recommended.
Rotating biological contactor systems were aesthetically more
acceptable; these were found to be more efficient, comparatively,
in respect of BOD removal and were found to be ideal for summer
camp applications treating weak wastewaters and fluctuating flows
RBC systems were found to require minimum start-up time, viz.,
only one week, and the operation of these plants is quite simple.
Results of a study done by Sack and Phillips (1) on the evalua-
tion of a RBC system for summer camp application indicated that
these systems are especially suitable in such applications.
Based on the above analysis, it was proposed to use RBC systems
to upgrade the septic tank effluent in the two locations.
SUMMARY RESULTS OF GENERAL INVESTIGATION OF RBC EQUIPMENT
A general investigation of RBC equipment, and their performance
was conducted. Results of the investigation are outlined below.
Autotrol Corporation, Milwaukee, Wisconsin pioneered the manu-
facture of RBC equipment in North America. RBC equipment are
presently manufactured by others listed below, in addition to
Autotrol.
1. Environmental Systems Division,
Geo. A. Hormel and Company
Coon Rapids, Minnesota
1138
-------
2. Tai t/Bio-Shafts Inc.,
Oconomowoc, Wisconsin
3. Envirodise Corporation (Clow)
Beacon, New York
4. CMS Equipment Limited
Mississauga, Ontario
There are a few more manufacturers of RBC units but these were
not evaluated.
It was learned that RBC units installed as of August, 1978 in the
states of Wisconsin and Maine and in the Province of Alberta were
manufactured by Autotrol. There were six RBC plants in the
Province of Alberta, all reported to be performing satisfactorily
with only a few minor problems of adjustment and with no problem
involving either shaft malfunction or failure. All four RBC
plants either built or under construction in Maine were Autrotrol
units and one was operating for ayear with no significant mech-
anical problems (subsequently, in the fall of 1979 2 out of 5
shafts failed at the Fort Fairfield plant). All the RBC units
in operation in Wisconsin were Autotrol units. Other RBC units
are presently included in the specifications as these are reported
to be comparable in design. It was understood that the Wisconsin
Environmental Department was reasonably happy with Autotrol plants
(mechanical aspects and secondary treatment requirements). A
perusal of design bulletins and drawings of the RBC manufacturers
showed that these were very similar with only minor variations.
This investigation revealed that generally some problems with the
shaft surfaced in some locations and necessary rectifications were
carried out by the manufacturer as part of development work. It
was considered prudent to check the design of these shafts in
detail during shop drawing stage and to take sufficient care during
installation to minimize any problems.
SUITABILITY OF ROTORDISK UNITS
The suitability of Rotordisk units and related equipment manu-
factured by CMS Equipment for use at Fundy National Park was
specifically assessed as this equipment was found to be very com-
petitive. Information on the performance of seven installations
obtained through telephone enquiry, as part of the investigation,
revealed that three units exhibited mechanical faults -- two
related to the shaft and one related to the gear reduction unit.
Two of these units were the first and the third installations
of CMS Equipment. Enquiries revealed that the fabrication of the
Rotordisk units had improved with time and experience gained in
the early installations. The process performance was generally
satisfactory in all cases. It was recommended, {based on this
enquiry as well as review of the general specifications and design,
1139
-------
the CMS RotorOisk be considered as an approved equal. Based on
this recommendation, and the competitive nature of their quotes
(approximately 30% economical) Parks Canada selected RotorDisk
units for use at Fundy National Park. During a review of shop
drawings, the drive shaft sizes were increased to the next
higher sizes to be safe against fatigue failure.
DESCRIPTION OF PLANTS, FLOW PATTERN AND DESIGN CRITERIA
Each treatment system consists of an existing septic tank, a
rotating biological contactor unit with clarifier and a chlorin
ation system. An arrangement is provided in both the systems
to recycle clarifier effluent through the system during periods
of low flow. Sludge collected from the clarifier and scum
removed from either the recycle tank or the chlorine contact
tank are returned to the septic tank. Flow
plants are presented in Figures 1 and 2
the two plants are shown in Table 1.
diagrams for
Desi gn cri teri a
the
for
TABLE 1 - DESIGN CRITERIA
Parti culars
Headquarters
Campground Treatment
Plant
Desi gn populati on
Total BOD assuming BOD
contribution of 0.05 Ib/cap/day,
Ib/day
Maximum daily flow, gpd
Peak fl ow , gpd
BOD of raw wastewater, mg/1
SS of raw wastewater, mg/1
Expected BOD removal efficiency
of septi c tank %
BOD of septic tank effluent, mg/1
Expected BOD removal efficiency
in RBC (at 12-8°C) %
Expected BOD of plant effluent, mg/1
Expected SS of plant effluent, mg/1
Septic Tank
Capacity of existing septic tank,
gallons
Septic zone capacity, gallons
Equalization zone capacity, gallons
Expected sludge accumulation in an
operating season, gallons
Surface area of tank,
18
45
532
27
600
600
172
165
30
120
84
19
10-20
14
9
4
100
300
800
200
270
Point Wolfe
Campground
Treatment Plant
740
37
26 000
74 400
172
173
30
120
87
16
10-20
30 000
13 000
17 000
5 600
450
1140
-------
TABLE 1 - DEISGN CRITERIA (CONT'D)
Parti culars
RBC-RotorDisk Unit
Headquarters
Campground Treatment
Plant
gpm
Capacity of feed mechanism,
RBC media diameter, ft
RBC media area, ft2
Design hydraulic loading (average
gpd/ft2
Design BOD loading (average)
lb/d/1000 ft2
Secondary Clari fi'er
15
flow)
Surface
Overflow
gpd/ft2
Overflow
gpd/ft2
Chlorine
area, ft*
rate (average
rate (peak fl
Contact Tank
flow)
ow)
Capaci ty,
Detention
minutes
(peak flow) minutes
Chlorine dosage, mg/1
gal Ions
time, (average flow)
(maximum)
20
7
000
1.2
1.2
75
250
610
1 300
100
41
10
Point Wol fe
Campground
Treatment Plant
27
10
24 000
1.1
1.1
110
240
680
800
45
15
10
OPERATION AND CONTROL OF TREATMENT PLANTS
Sack and Phillips (1) recommended that recycle of sewage from
the final clarifier during periods of low flow should be made
an automatic function in order to keep the biomass in an
optimum condition. The two RBC plants at Fundy National Park
incorporate this special design feature of effluent recycle
to septic tanks at times of low flow as well as sludge and scum
return to the septic tanks. These aspects are elaborated in
thi s section.
HEADQUARTERS PLANT OPERATION AND CONTROL
The pumps deliver the septic tank effluent to the RBC module.
The average pumping rate the RBC module is controlled in the
range of 12 to 20 gpm, preferably closer to 12 gpm, through
(1) throttling the valve on the by-pass line which relieves
back to the sump or (2) through throttling both the by-pass
valve and the valve on the pump discharge line. These have to
be made by manual adjustments, and once adjusted, they are
marked/recorded and left undisturbed generally. The recycle
pump is controlled by the flow rate over the V-notch weir within
a specific pre-set range.
1141
-------
The pumping rate for the recycle pump is controlled such that
a maximum flow of 10 gpm is obtained; this is achieved by
throttling the two valves (one by-pass and the other on dis-
charge) similar to the case of primary pumps. The recycle
pump will operate when the flow rate at the v-notch weir is
less than 5 gpm as sensed by the flow recorder. The pump will
cease operation when the flow increases to 8 gpm. Within the
flow rate of 0 to 8 gpm leaving the treatment plant, the
recycle pump will be controlled by float switches.
Sludge from the clarifier is automatically transferred to the
sludge and scum tank through a pipe from the revolving clarifier
mechanism. Another pipe is used to transfer the scum from the
recycle tank to the sludge and scum tank, when the scum collector
is operated manually. The sludge and scum pumps deliver the
sludge and scum to the inlet end of the septic tank.
Measuring, recording and totalizing of the effluent flow rate
is provided utilizing a bubbler tube immersed in a 22^Q V-notch
weir located in the recycle tank. The back pressure produced
by the bubbler system (directly proportional to the liquid
level flowing over the weir) is sensed by the recorder (housed
in a separate enclosure) which extracts the weir formula for
recording and totalizing. An electronic signal-, proportional
to the flow, is transmitted to pace the chlorinator. Also
located within the recorder instrument case are signal contacts,
set to open and close at adjustable flow rates for operation
of the recycle pump.
POINT WOLFE PLANT OPERATION AND CONTROL
The septic tank effluent flows by gravity to the RBC module in
this case. The recycle pump operation and control is similar
to the one described for Headquarters Plant. The v-notch for
flow measurement is placed in the chlorine contact tank in
the present case. The scum is also collected from the chlorine
contact tank. Sludge and scum pumping system and the control
system are similar to the ones at Headquarters Plant.
BUILDING FOR RBC PLANTS
The rotating biological contactor should be covered or enclosed.
The enclosure serves two main functions. Operation in low
temperature climates requires a housing to limit wastewater
temperature reduction or heat loss. Also, exposure of the
rotating surfaces to direct sunlight and to rainfall should
be prevented to stop the growth of algae on the outer surfaces,
and to prevent rain wash-off, respectively. The enclosure also
helps to prevent deterioration of the plastic medium by ultra-
violet rays from sunlight.
1142
-------
Wood frame buildings are provided to house the RBC units,
chlorination system, recycle system, flow measurement units and
electrical controls. The structures not only provide weather
protection for the units and their associated controls, but
also help to maintain aesthetic appearance of the Park, with
these buildings conforming to others in the area. No heating is
provided for the buildings. Air within the enclosure is at a
temperature approximately equal to that of the wastewater. At
lower ambient temperatures, the high humidity within the buildings
will usually result in condensation on the walls and ceiling.
Condensation can be minimized by the provision of forced
ventilation of the enclosure, drawing fresh air through louvres
and removing the moisture-laden air using an exhaust fan.
Corrosion-resistant materials are specified for use within
the buildings. Interior wood surfaces have been Pentox-treated
to resist rot. Reappli cati on of Pentox will be required over
the years.
The building floors are sloped toward peripheral channels at the
exterior walls for drainage during washing down. Hose bibs
connected to the Park potable water system, are provided for
this purpose. The floors of the buildings are insulated from
frost eliminating the need to disconnect piping during the
winter shut down of the plants.
PERFORMANCE OF THE RBC PLANTS
Headquarters Plant
The plant became operational in the summer of 1979. The plant
was inspected in late August to observe its performance. The
biomass growth on the RBC was peculiar in that it was quite
excessive on the first module, with very much less on the next
module and almost nothing on the last module. In the first
module, about 25% of the gaps between the woven media sheets
were filled with biofilm (sludge). Although there were
malodours in the room, foul odours were evident when the sludge
deposits on the RBC were disturbed. The biomass was grey-brown
and had a furry or hair-like appearance. It was believed that
a leak in the septic tank to the wet well contributing septic
liquor and sludge was causing the peculiar development of
biomass on the RBC unit. The RBC influent has a strong unpleasant
odour suggesting a strong wastewater and this was borne out
by the results of analytical tests given below.
BOD SS
Sample mg/1 mg/1
RBC influent
RBC effluent
Final clarifier effluent
Efficiency of removal
165
48
30
82%
47
35
15
685
1143
-------
point Wolfe plant
This plant also became operational in the summer of 1979. The
performance was observed in late August. The biological growth
on the RBC units appeared healthy and normal, with biomass
decreasing progressively from the first module to the fourth
module. The thickness of the biomass on the face plates of the
first module was approximately 0.12 in., decreasing to 0.04 in,
on the fourth unit. The interstices of the woven mesh media
were 70-80% occupied in the first module, with a light film on
the solid portion of the media in the last module. The results
of analytical tests performed at this plant are shown below.
Sample
RBC influent
RBC effluent
Final clarifier effluent
Efficiency of removal
BOD
mg/1
50
32
10
80%
The percentage removals in respect of BOD and SS are in the
same range for both the RBC plants. Preliminary results thus
indicate that the performance of these plants is generally
good. However the final effluent quality in the case of
Headquarters Plant is not as good due to the much stronger
wastewater applied. It is believed that the final effluent
BOD will fall below 20 mg/1 once the leak in the septic tank
and recycle problems at Headquarters Plant are set right.
A comprehensive study of
would be desirable.
REFERENCE
the performance of both these plants
1. Sack, W.A., and Phillips, S.A., "Evaluation of the Bio-Disc
Treatment Process for Summer Camp Application" EPA-670/2-
73-022, U S Environmental Protection Agency, Washington,
D.C., August 1973.
1144
-------
H
co
cr
LU
i-
tr
o
<
LU
X
a:
o
CD
o
LU
cr
CD
, 1145
-------
Q_
O
tr
o
<
Q
OJ
UJ
a:
ID
-------
THE GLADSTONE, MICHIGAN EXPERIENCE
PERFORMANCE OF A 1,0 MGD RBC PLANT
IN A NORTHERN MICHIGAN COMMUNITY
by
Theodore C. Williams, P, E,
Chairman of the Board
Ralph J. Berner, Jr., P, E.
Senior Studies Manager
Williams & Works
Grand Raipds, Michigan
INTRODUCTION
In March of 1974, the first federally funded municipal wastewater treatment
plant utilizing Rotating Biological Contactors (RBC) was placed in
operation in Gladstone, Michigan. The plant was designed by Williams &
Works of Grand Rapids, Michigan to replace an existing primary treatment
plant at the same site. Because, in many respects, this plant was the first
of its kind in the United States, the performance of the system in a
moderately severe northern climate has received a relatively high degree of
scrutiny by engineers and manufacturers engaged in the design of other RBC
installations. The design considerations for this plant, operating data for
four years of operation, and some general observations about RBC
installations based on the Gladstone experience will be presented in this
paper.
1147
-------
DESIGN CONSIDERATIONS
Gladstone, Michigan is a predominantly residential community of about 5,000
people located in the southern part of Michigan's Upper Peninsula, on the
shore of Lake Michigan, approximately 10 miles north of Escanaba. The
climate in the area is characterized by cool, dry summers and cold, snowy
winters. The mean annual temperature is 42°F (5.5°C). Gladstone derives
its municipal water supply from Lake Michigan and wastewater temperatures
are frequently in the mid 40's (°F) during winter months.
The 1.0 mgd (3,785 cu m/day) RBC plant was designed to provide secondary
treatment and phosphorus removal on the site of an existing primary
treatment plant. A system incorporating housed rotating biological contac-
tors was selected on the basis of aesthetics, operational advantages, and
data from successful pilot plant testing.
The general design parameters were as follows:
TABLE I
GENERAL DESIGN PARAMETERS
Gladstone, Michigan
Population
BOD5
TSS
Average Daily Flow
Peak Flow
Effluent Quality Limitations:
BOD,-
b
Total Suspended Solids (TSS)
Total Phosphorus (TP)
10,000
1,670 Ib/day (758 kg/day)
2,000 Ib/day (908 kg/day)
1.0 mgd (3,785 cu m/day)
2.88 mgd (10,000 cu m/day)
30 mg/1 (30-day average)
30 mg/1 (30-day average)
20% of influent TP
General pilot plant tests of the RBC process were conducted at the
University of Michigan in Ann Arbor in 1968-1969 under the supervision of
Professor J.A. Borchardt. The pilot plant consisted of three stages in
series, each with fifty, 4-foot (1.22 m) diameter, 0.5-inch (1.27 cm) thick
1148
-------
was operated for approximately one year using raw wastewater and primary
effluent from the Ann Arbor municipal treatment plant. Some results of the
pilot plant tests are summarized in Table II. In general, this pilot plant
experience offered the following design considerations:
• Primary treatment is necessary to prevent accumulations of debris on
RBC shafts and to increase overall BOD,- removals.
• The RBC shafts must be rotated at a speed sufficient to entrain
sloughings and mixed liquor solids. At low speeds, solids accumu-
lations within the RBC tankage caused the depletion of dissolved oxygen
and created odor nuisances.
• An average of 80 to 93 percent overall BODS removals were consistently
achieved at hydraulic loadings of 1 to 4 gpd/sq.ft. and at temperatures
ranging from 48° to 63° F.
• At similar hydraulic loadings, BODS reductions through the RBC process
dropped about 1% for each 1° C drop in wastewater temperature.
A schematic flow diagram of the treatment facility is shown in Figure 1.
The design parameters of the unit processes and appurtenant equipment are
summarized in Table III. Raw wastewater is screened and pumped to two grit
chambers, then comminuted and settled in a single, rectangular, primary
clarifier salvaged from the old plant. Primary clarifier effluent is split
to two parallel RBC paths, each with three shafts and six stages. Mixed
liquor within the RBC tanks flow through 2-foot (0.61 m) diameter submerged
openings to subsequent stages. RBC effluent is dosed with liquid alum for
phosphorus precipitation prior to pumping to twin rectangular clarifiers.
*
Anionic liquid polymer is added between the pumps and the secondary
clarifiers. Secondary effluent is chlorinated prior to discharge to Lake
Michigan.
Grit is manually cleaned from the grit chambers and disposed of at a
sanitary landfill. Settled primary and secondary sludges are pumped to a
primary anaerobic digester that is heated and equipped with gas recircula-
tion and mixing apparatus. Settled, digested sludge from the secondary
digester is withdrawn by tank truck to sludge drying beds off-site. Dried
sludge is disposed of in a sanitary landfill. Supernatant from the
secondary digester is drained by gravity to the raw wastewater wet well.
1149
-------
in
in
f_^
I— <
LU
— )
CO
?
i— re
r- >
GJ cy
^ CfC
o
E^
cc
H-
-
or
g"T^
^5»7
S-^
=3
CO
j
»-•"
f"^
^g**
LU
01
i
in
f^
o
CO
OJ
•O en
5- re
CO •»->
CO
OJ
•o en
c re
CM 4->
CO
CO*— >
CO O
i •
in o
• (JQ
CO **—^
r^.
in cr>
^^ •
1 T—
CO CO
r— * — »
ve-
in en
1 CO
CO*—
•~
0)
+J en
c/> re
'" CO
tf^ !»— ^
f^ en
i in
CO*"*
•• •
s ^™*
*r-" H-*
S» ^H-
0. LU
en ••"%•
CO O
c— 0
O^ **^
*3*
^
fO
a;
^J-^
CM en
i in
VO c—
cn^-'
o
o
CD
1—
z:
3
0-
^
i
»— i
o_
1
CO S- ••— >
4J CO CX •
l/J 4-> £ LU
3: S K-» —
&} <^*^»
in vo
i in.
in
-o c: —
O> O IO
40 t— 01
re 4-> 4->
E C 3
•i- CO C
tn
1 r—
CM — '
r—
^^ ^™"*
in CM
i ^j*
in —
CM
o
J^ ir>
1 CO
tp
"*^.
CM VO
i in
COr—
C^ ^"^
r—
O^ *"'*^
in vo
1 LO
in
vo
CO
in xn>-
cn vo
i •
•Sf CM
• cn
vo —
CO
CO —
i— vo
i •
VO CO
v~"
vo m
r-~ *
1 T—
CO i—
— •
0 ^
in CM
CM •
1 O\
f" • r~*
f— *-_J
CO ^"^
OJ
1 VO
^"
^•^
i— VO
1 r—
in i—
CO. — >
VO CM
1 VO
vo
CM
•^ X~"N
en CO
1 •
CM VO
• CO
CM
^
CM «3-
*5l" »
1 f~~
t— CM
I—- — •
in —
VO CM
i ro
in — *
•~
O"—
CD m
f** •
1 CM
r-« in
CO- —
00 ^"^^
cnr-
r"™» r"™
I r—
^^ ^— ^
m
CM
CM CM
1 VO
f~ --i ^
I—
CM *" *^
in o
i in
CO
V)
Qj
to
OJ
to
CL
O
i— C- »
3t- CT-
CO ~O V)
t- re ~~^
•o o "a
a.
en
o
o
in
•a
a.
in CT>
vo
• o
CM O
CD
-a
o
i.
o_
4^
10
J—
VO
o
CM
' 1
p»M
CM
^
cn
vo
co"
*^.
^3*
1
CO
«3-
•o
a.
CM CT>
co
• O
f— O
O
«
in
-o
Q.
CM O
in
CO
o^
vo
^^
^~
CO
^x»
in
i
CM
r~
•~^
in
cn
vo
•*>.
CM
p^
•^.
CM
r—
1
l>»
^^
•^»
r—
^~
tf)
CO
3
re
>
CO
CR
re
s_
CO
>
1150
-------
INFLUENT
LANDFILL
SLUDGE
DRYING
BEDS
SECONDARY
DIGESTER
PRIMARY
DIGESTER
BAR SCREEN
RAW SEWAGE PUMPS
GRIT CHAMBERS
PARSHALL FLUME
COMMINUTOR
PRIMARY CLARIFIER
ROTATING
BIOLOGICAL
CONTACTORS
(LIQUID ALUM FEED)
SECONDARY PUMPS
(POLYMER FEED)
FINAL CLARIFIERS
TO LAKE MICHIGAN
CHLORINE
CONTACT
CHAMBER
GLADSTONE, MICHIGAN
WASTEWATER TREATMENT SYSTEM
Figure 1
1151
-------
TABLE III
UNIT PROCESS AND EQUIPMENT DATA
Equipment
Raw Sewage Pumps
Primary Clarifier
Design Data
3 @ 1,400 gpm (88.3 I/sec), 2 variable
speed, and 1 constant speed lag pump
One rectangular 70,000 gallon (265 cu m)
capacity with 1.5 hour detention and 895
gpd/sq ft (36.5 cu m/day/sq m) overflow
rate
RBC Units
Bio-Surf by Autotrol. Two parallel
paths with three shafts and six stages
in each, 515,500 sq ft (47,890 sq m) of
media, 90 minutes retention and 1.94
gpd/sq ft (0.097 cu m/day/sq m) hydrau-
lic loading rate
RBC Effluent Pumps
Secondary Clarifiers
Same as raw sewage pumps
Two rectangular, 2.75 hour detention and
620 gpd/sq ft (25.3 cu m/day/sq m)over-
flow rate
Chlorine Contact
Two baffled tanks with retention time of
35 minutes
Primary Digester
One fixed cover, 100,000 gallon (378.5
cu m) capacity with 15.5 days detention
at design loading, heated, mixed by gas
recirculation
Secondary Digester
One floating cover, 100,000 gallon
(378.5 cu m) capacity with supernatant
discharge to raw wastewater wet well
Tank Truck
Drying Beds
One 2,500 gallon (9.4 cu m) capacity
Adjoining sanitary landfill site, 600' x
200' (183 m x 61 m) without underdrains
Liquid Alum Storage
Two tanks, each 2,600 gallon (9.8 cu m)
capacity
1152
-------
INITIAL OPERATING EXPERIENCE
The operation of the plant was begun on March 1, 1974. At the same time, an
18-month testing and shakedown program was begun to monitor the performance
of the plant and to make operational improvements where necessary. During
this test period, wastewater flows averaged 0.755 mgd (2,875 cu m/day) and
were influenced occasionally by sewer infiltration and inflows. At start-
up, wastewater temperatures averaged 45°F (7.2°C), then gradually increased
to the low 60's (°F) by mid-summer.
Influent BODg concentrations ranged from 129 to 219 mg/1 and influent total
suspended solids concentrations ranged from 101 mg/1 to 168 mg/1. The
average influent BOD^ and total suspended solids concentrations for the test
period were 164 mg/1 and 132 mg/1, respectively.
The results of this initial 18-month test period are summarized as follows:
• Because the influent wastewater temperature was quite cold at startup,
it took 18 days before heavy sloughings of biomass were observed and
several months to achieve a steady state operation.
* During the last 12 months of the testing period (September 1974 through
September 1975) with the addition of 70 mg/1 of alum and 0.8 mg/1 of
polymer, effluent BOD^'s were generally less than 10 mg/1; effluent
suspended solids averaged 15 mg/1; and effluent total phosphorus
concentrations were less than 1.6 mg/1.
* In January 1975, design flow conditions were simulated for two weeks by
shutting down one of the parallel RBC paths and one of the final
clarifiers. The effluent BOD,- for this two-week test period averaged
19 mg/1 with an average wastewater temperature of 47 F (8.3 C).
• Because of low hydraulic loadings (1.0 to 2.1 gpd/sq.ft.), nitrifi-
cation occurred; 50% to 60% reductions in ammonia nitrogen occurred
during winter months and 85% to 95% reductions occurred during summer
months.
• After one of the RBC paths was restarted in February 1975, after being
shut down for full scale tests, nitrifiers did not reestablish dominant
cultures on latter stages until June, when wastewater temperatures
approached 60° F (15.6°C).
1153
-------
• Excellent BOD5, suspended solids and ammonia nitrogen removals
permitted reductions in effluent chlorine doses, from 6-7 mg/1 to 2 -
3 mg/1.
• Recycling of secondary sludge through the primary clarifier was
required to thicken the combined sludges from 2% - 3% solids to 4% - 5%
solids content.
• Slug releases of digester supernatant to the head end of the plant were
detrimental to overall nitrification results. Controlled release of
digester supernatant is considered essential to achieve consistent
ammonia nitrogen reductions.
• No significant differences were noticed in the digestibility of
sludges with or without alum addition. Phosphorus release, volatile
solids reduction, or gas production were not significantly affected by
alum addition. The average total sludge production was 0.83-lb of
total solids per pound of BOD,- removed, including chemical sludge
production.
• Measured power usage by RBC shafts ranged from 5.6 hp to 3.5 hp. The
average running horsepower per shaft was 4.2 hp.
PLANT PERFORMANCE (1976-1979)
The overall performance of the plant, subsequent to the 18-month test
program, is summarized in Table IV. The data covers the period from January
1976 through December 1979. Table V shows a comparison of plant performance
under summer and winter conditions. Data for these tables was obtained from
monthly operating reports prepared by the plant operator.
This data shows that the plant is consistently removing over 90% of influent
BODg even under winter conditions and at hydraulic loading rates of between
1.0 and 1.4 gpd/sq.ft. Wastewater strength is affected by increased
infiltration and inflow during warm weather months. During winter months,
effluent BODg concentrations increase as wastewater temperatures decrease
to lows of 45°F. Effluent dissolved oxygen concentrations are consistently
above 7 mg/1.
1154
-------
TABLE IV
PERFORMANCE SUMMARY
Parameter
Flow (mgd)
Hyd. Loading (gpd/sq.ft.)
BOD5 (mg/1)
- Influent
- Effluent
- % Removed
TSS (mg/1)
- Influent
- Effluent
VSS (mg/1)
-- Influent
- Effluent
Total Phosphorus (mg/1)
- Influent
- Effluent
- % Removed
Ammom'a-N (mg/1 )
- Influent
- Effluent
- % Removed
1976 1977
.635 .725
1.23 1.41
182 143
6 7
97% 95%
135 122
15 16
111 95
8 7
7.8 6.3
1.4 1.1
82% 83%
15.5
3.5
77%
Effluent Dissolved Oxygen (mg/1) 7.3 7.6
Raw Sludge Pumped (lb VSS/day)
576 544
Digester Gas Prod, (cu.ft./day) 4,307 3,981
Chemicals Used (Ib/day)
- Chlorine
- Alum
- Polymer
13 19
435 485
1.5 1.7
1978
.612
1.19
155
7
95%
139
17
112
8
.4.7
0.9
81%
17.4
2.1
88%
7.6
512
3,954
! 12
; 365
1.5
1979
.712
1.38
129
12
91%
118
16
92
•• 9
3,5
.0.9
74%
15.7
2.7
83%
8.2
544
3,718
12
262
1.4
1155
-------
Effluent total suspended solids concentrations remained fairly consistent
throughout the period at an average of about 16 mg/1. The hydraulic profile
of the plant is such that secondary solids are pumped twice prior to removal
with settled primary solids. The addition of polymer assists secondary
sedimentation, and lower suspended solids concentrations could probably be
obtained in the final effluent if flocculents were not disturbed by
centrifugal pumps.
More than 80% of influent total phosphorus concentrations are removed by the
addition of alum (hydrated aluminum sulfate). Plant operators attempt to
maintain a liquid alum dosage rate sufficient to provide 1.5 moles of
aluminum to 1.0 mole of influent total phosphorus.
The plant was not designed to accomplish nitrification. However, because it
is currently operated at hydraulic loadings of less than 2 gpd/sq.ft.,
nitrification does occur. Ammonia analyses on weekly grab samples show that
average influent ammonia concentrations of 16 mg/1 to 18 mg/1 are reduced to
2 mg/1 to 4 mg/1 through the plant. As might be expected, less nitrifi-
cation occurs during winter months.
In 1979, an average of 720 pounds per day of sludge was pumped to the primary
digester with an average solids content of 3.7% by weight. The volatile
content of the sludge averaged 74% by weight. Total sludge production,
including chemical sludges, averaged 1.04 pounds per pound of BOD,- removed,
or approximately 1,000 pounds per million gallons treated. Approximately
262 pounds per day of alum was added to remove an average of 2.6 mg/1 of
total phosphorus. Stoichiometrically, an average of about 90 pounds per day
of chemical sludge was produced in 1979.
OPERATING COSTS
Energy and Chemical Costs
The energy and chemical uses and costs associated with the plant operation
for 1978 are summarized in Table IV. The costs shown per million gallons
reflect actual prices Gladstone paid for the respective commodities in 1978.
1156
-------
TABLE V
COMPARATIVE PERFORMANCE SUMMARY
Winter vs Summer Operation
Flow (MGD)
Hydraulic Loading (gpd/sq.ft.)
Waste Temp (°F)
BOD5 (mg/1)
- Influent
- Effluent
- % Removed
TSS (mg/1)
- Influent
- Effluent
Total Phosphorus (mg/1)
- Influent
- Effluent
- % Removed
Ammonia-N (mg/1)
- Influent
- Effluent
- % Removed
Effluent Dissolved Oxygen (mg/1)
Average Winter
Values ]_/
0.530
1.03
46
181
10
94
140
16
5.8
0.8
86
20.0
4.9
76
8.1
Average Summer
Values 2/
0.726
1.41
64
128
9
93
130
16
4.7
1.0
79
15.1
1.0
93
7.3
!_/ From January, February, March, 1977 through 1979 reported data.
2/ From July, August, September, 1977 through 1979 reported data.
1157
-------
CD
CM «*• cn
p- f>. «*
• • •
p— VD CO
i— «*
•V*
o r-«
o «a-
• •
r-~ «*-
CM
vo
CM
CO
CM
in
0
o
o
o
10
LO
o
in
to
.
e t/i
03 -O
o
CM
10
CT>
in
oo
vo
»— <
s»
UJ
CD
eC
P
LU
CO
Z3
I
5c
o
I1- 4
1
0
0
s
"'*•
>-
s
LU
LU
CO
F*^
t^"
cn
c
(O
cn
*fv»
JC
O
•p-
s
•t
o>
c
o
4->
in
•o
(0
5
CD
s:
r-§
(U CO
Lu O
o
o
*~~
i i i
&
•1—
O CD
Z^
j % "f-»
•?•* — L»
u 2:
OJS^
Lu
co in
vo o
CM r— I
•-
co in
r>. o
i «— t—
«
CO
CO
CM
O
CM
in
00
o
CM
5
•
O
V)
s_
id
,
J_
CO
•o
c
o
o
OJ
CO
cn
c
• »r»
CL
Q.
03
cn
to
£
03
CO
s
rO
ca
E
S
+->
in
>,
co
C_J
cn
ea
c
o
•p-
+J
CO
c
•r*
i.
O
!c
CJ
r™ •
CO
>
B£
0)
.0:
in
3
i-
O
-E
O.
in
0
.c
o.
.+J
c
OJ
E
+j
(O
V
S- i-
OJ J—
E
S >> 03
3 i— CJ
p- O -O
eC CL. 3
p^
CO
in
t.
0)
4->
ey>
CO
0]
3=
cn
c
•p—
•o
•p-
3
CQ
S
m
i—
CO
>-
CO
_1
<:
F=
o
h-
o
«f-
s
o
i— OJ
LU r—
JQ
>> T-
r— cn
•p- «p-
CO f—
o cn
o>
0) z:
cn
CO
J- u
0)
>
< z
1158
-------
Electricity consumption is of primary interest and the table shows
electrical power consumption in terms of raw sewage pumping, secondary (RBC)
treatment, and anaerobic digestion. The power consumption for the RBC units
actually includes power consumption by clarifier drives, intermediate
pumps, lights, ventilating equipment, etc. and computes to an average of
approximately 38 running horsepower at any given time. Of this,
approximately 80 percent (30 hp), including power draw by the RBC units,
could be considered independent of actual flows. Based on measured power
consumptions by the six RBC drive units in 1975, the RBC system draws an
average of 25-26 horsepower at, any given time.
To gain more perspective on energy consumption at Gladstone, as related to
consumptions by other types of small wastewater treatment plants, Table VII
displays similar breakdowns for six treatment plants in Michigan. Because
of wide variations in commodity costs from community to community, the costs
shown in Table VII have been adjusted to a common base as follows:
Electricity
Fuel
Natural Gas
Gasoline
Chemicals
Chlorine - Effluent
- Alma Purifax
Lime
Alum
Ferric Chloride
Polymer
$0.042 per KWH
$0.210 per 100 cu.ft.
$0.700 per gallon
$0.120
$0.097
$0.036
$0.050
$0.073
$1.880
per pound
per pound
per pound
per pound
per pound
per pound
It is noteworthy that the net energy usage by the Alma extended aeration
system and the Gladstone RBC system are about the same. However, the
Gladstone plant was currently only receiving about 60 percent of its design
flow, whereas the Alma system was at about 94 percent of capacity. Net
energy usage by the two systems at design flow conditions, with appropriate
assumptions about variable power usage, is forecast as follows:
1159
-------
o
•
CO
00
ID
.
vo
en
CO
•51-
*
f^
CO
CO
CM
•P
CJ Q)
o> E -P
•O 4-» W
3 re o
r— CJ
CO J-
cvj
o
•
CM
(O
CO
•
CO
CO
CM
•
o
CM,
t/1
8
CD
C£
UJ
H-l UJ
Z co
^ u.
p O
in
I
LU
UJ
C£
UJ
> tn
(O
•O
C -
o « o
00,
CM
VO
O
10
10
CO'
*
cr>
CM
r*.
CM
in
N.
CO
.
CM
in
On- to;
to O- r
10
0
CO
a*.
vo
CO
O
CD
S «/»
c cr
r*. o o
to o f-
O O">+->
. re re
O r— Ol
^—' «f—
C S-
O) O S-
r— *r- >—i
r— •»->
•»- re -o
>-o o
+J T— O
C Xi—
o o u.
CJ
o
CD W
^ o o
CO Of-
CM re re
^^ o s-
|2 43""*
^ -o re
o!"x a.
i— o to
•o
•a i i
CM
VD
O
CD
CM
VO
CO
C T-
1- O 4J
COM
O Ol C7>
tt) re «r-
I $1
C5 CJ ,
-o
re c.
s- o
•—* CJ Q.
0 i— s_
>— X O 3
UJ O- Q.
ea
Sill
O
CD co
S CJ
01
CM T-
SE°
• CJ O
O -»->f-
""^ >> o
CJ IO J-
C CJ
o o re
*J CO C
co c£ «c
•o
re i i
<3
eg
re
co
J«
CO ••->
• >> CO
E wo
a> u
•u +J
w c +J
>> CJ-r-
co E c
4-> 3
i/> re
•r- CJ CJ
•M 4-> re
CO
O O i—
i— CJ
re J=
CJ +J
S o i-
*H-.0
o o
*~* -r- +J
va
CO
3
i-
O
O. .
CO CO
O •!->
J= C
Ci O
f ^J
o re
«t-
< U CO
•M CJ'^
O J= -O
zi— re
•O r-
re to
CJ V)
-------
I
Alma
Gladstone
1978
Flow
% of
Design
94
61
1978
Electrical
Use,
KWH/MG
1,209
1,105
Adjusted
Electrical Use
at Design Flow,
KWH/MG
1,150
700-800
Thus, electrical power use for secondary treatment at Gladstone should be 30
percent to 35 percent less than that at Alma as the design flows for each
plant are approached. Comparison of the two plants must be tempered by an
appreciation for higher degrees of treatment achieved at Alma.
dverall Operation and Maintenance Costs
Table VIII is a conceptual operation and maintenance budget for a system
such as Gladstone's for fiscal year 1980. The Gladstone plant is staffed by
one superintendent/operator and four shift operators. The plant is manned
16 hours per day, five days per week; eight hours per day on weekends and
holidays. Approximately 30 percent of the labor budget at Gladstone is
allocated to the water sypply system; however, the conceptual budget assumes
total assignment of labor to the wastewater treatment system. Labor costs
for the five-man staff are estimates for 1980, but are considered reasonable
for current average operating labor in this region of the country. Energy
and chemical costs reflect Gladstone's 1978 costs, increased by 20 percent.
The equipment replacement line item is in deference to current EPA user
charge requirements for funds to replace key process equipment items over
the service life of a facility. This budget suggests a treatment cost of
approximately $600 per million gallons of wastewater treated.
SUMMARY
In summary, the Gladstone experience has been a good one. Excellent
treatment results are being obtained; plant effluent quality is well within
design limits. Consistent nitrification at loadings below 2 gpd/sq.ft. is
being achieved as an added bonus toward high quality treatment.
The RBC system has been very consistent in achieving wastewater treatment
t
and in permitting stable solids handling procedures at Gladstone. Few, if
1161
-------
TABLE VIII
CONCEPTUAL OPERATION AND MAINTENANCE BUDGET
FY 1980
Labor/Benefits
Power/Fuel s/Chemi cal s
Routine Plant Maintenance
Parts Inventory/Supplies/Mi scellaneous
Equipment Rental
Equipment Replacement Fund
TOTAL
$ 96,000
»
33,000
3,000
4,000
3,000
15,000
$154,000
Cost per million gallons treated
I/ Based on treating 260 MG per year.
1162
-------
any, problems have occurred with the day-to-day operation of this RBC
secondary treatment system. The process has been economical and has
required a minimum of operator attention over the past four years.
ACKNOWLEDGEMENTS
The authors wish to thank the City management of Gladstone, Michigan and Mr.
Willard Morley, Plant Superintendent, for making plant operation data
available for this paper.
REFERENCES
"Energy Considerations i_n Wastewater Treatment for Small Communities"; S.K.
Malhotra, A.R. Posthuma; Williams & Works, Inc.; Paper presented at the
Fourth Mid-American Conference on Environmental Engineering Design; August,
1979.
1163
-------
-------
PART X: NITRIFICATION AND DENITRIFICATION
The Effect of Organic Loading on Nitrification
in KBC Waste-water, Treatment Processes
K. Ito and T. Matsuo
Department of Urban and Sanitary Engineering,
The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113, Japan
INTRODUCTION
Though organic substances, themselves, are not harmful to nitrifying
bacteria, the nitrification in the biofilm attached to the rotating disc
may be influenced by them. Under high organic loading, the number of
nitrifying bacteria in the biofilm will be decreased, not because o'f sub-
strates, dissolved oxygen, or essential nutrients competition, but because of
the difference of growth rate between the nitrifying bacteria and the aerobic
heterotrophs. The aerobic heterotrophs will grow very rapidly under high
organic loading, and the biofilm on the disc will thicken and eventually be
sloughed off by gravity and shearing stresses acting on the rotating biofilm.
As a result of frequent sloughing, the.nitrifying bacteria may be washed out
from the attached biofilm.
One prominent difference between the ordinary conventional trickling
filter and the rotating biological contactor (RBC) is the longitudinal mixing
characteristics of their flow patterns. In the trickling filter, the flow
pattern is thought to be that of plug flow. But in the RBC, the flow pattern
may be closer to that of complete mixing. Because of this difference,
nitrification in the RBC may be more sensitive to effects of variations of
organic loading.
1165
-------
In this paper, some results of experiments are presented to discuss to
what extent organic loading may affect nitrification, how to maintain the
stable nitrification in the RBC, and how to proceed with denitrification in it.
EXPERIMENTAL EQUIPMENT AND METHOD i
Three different types of continuous flow laboratory scale RBC units were
used. One was a four-stage KBC unit and the others were single-stage EEC units;
Schematic flow diagrams are shown in Figs.l and 2, respectively. Dimensions
and other specifications are given in Table 1.
Six series of experiments
1.
were carried out. Classi-
fications of experiments
and the composition of
influents used are "briefly
summarized in Table 2.
Analytical methods
for determining water
quality parameters used in
these experiments and tech-
niques of bacterial enume-
ration are summarized in
Table 3. The amount of
biomass attached to the
discs was weighed directly
after half a minute of de-
watering. The thickness
of "biofilm near the rim of
a disc was measured at sev-
eral points "by slide cali-
per. Biomass sloughed off
from discs was retrieved
with a fine mesh nylon net,
dried, and weighed to eval-
uate the rate of biomass
sloughing in the EEC.
Table
Dimensions of the RBC units
HBC unit classification
Number of stage*
Number of disc*
Dice Diameter
Material
Interval between dlio»
Contactor volvo*
Fraction of
area submerged
Rotational apecd
Typ»(A)
4
10 x 4
25 em
FTC.
1 CB
28.8 litres
(7.2 x 4')
0.45
15 rja
Tjrp.(B)
1
10
25 OB
we
1 CB
7.2 litrea
0.45
15 rp»
Tjrp«(c)
1
5
30 «•
FTC
SOB
lO.f litree
0.49
2-10 rpm
Reservoir
Clarifier
Fig. 2. Schematic flow diagram of the Type(B)
or Type(c) contactor
Reservoir
Recirculating Pipe
Clarifieri
Fig. 1. Schematic flow diagram of the Type(A) contactor
1166
-------
Table 2. Classifications of experiments
Classifications of experiments
(l) Experiment
to estimate the growth rate
of nitrifying bacteria
( 2 ) Experiment
to calculate the biomass
sloughing rate
( 3 ) Experiment
to estimate critical organic
loading for nitrification
(k) Experiment
to count numbers of
nitrifying bacteria
(5) Experiment on denitrification
using a recirculating system
( 6 ) Experiment
to proceed with both nitri-
fication and denitrification
KBC unit type
Type(B)
Type(C)
Type (A)
Type(A)
Type (A)
with reeircu-
1 at ion
Type(B)
Composition of
influent sources
dnfy)2soii, Na2co3
nutrient salts
soluted in tap water
synthetic dry milk
soluted in tap water
(NHl^SO^, Na2C03
synthetic dry milk
soluted in tap water
(WHi|.)2S01j., Na2C03
synthetic dry milk
soluted in tap water
(KHll)2SOl;, Wa2C03
synthetic dry milk
soluted in tap water
HHijCl, NaHC03
nutrient salts
glucose or methanol
soluted in tap water
* synthetic dry milk composition: C 50 %, N 2 % , P Q.k %
several vitamins and minerals are included
Table 3
Analytical methods
Measurement
Analytical method
Organic nitrogen
Ammonium nitrogen
Nitrate nitrogen
Nitrite nitrogen
COD(Cr)
TOG
Dissolved oxygen
Viable count of
nitrifying bacteria
Viable count of aerobic
heterotrophic bacteria
Kjeldahl method
Ammonium electrode method
Cadmium column reduction method
GR method
Standard Methods
Standard Methods
Azide modification of
Winkler method
Glass electrode method
MPN technique
Spread plate technique
with Sakurai?s agar medium
1167
-------
EXPERIMENTAL RESULTS
Observed growth rate of nitrifying bacteria on discs
Concerning environmental conditions for nitrification, dissolved oxygen
concentration and pH were maintained higher than 5 mg/1 and between- 6.8 and
7-8, respectively. The water temperature was held between 15°C and 25°C, and
the influent ammonium was mostly oxidized to nitrate. The thickness of bio-
film, assumed to consist mainly of nitrifying bacteria, was very thin, less
than 1 mm, and rather uniform over the surface of a disc. The sloughing rate
of the biomass was less than 0.1 gSS/day-m2 as shown in Fig.3.
From results of viable counts of
ammonium oxidizing and nitrite oxidizing
bacteria in the biofilm, we could esti-
mate the number of these bacteria at 10
to 10^/m^ and 10T to 10^/m2, respectively.
The relationship between the ammonium
nitrogen removal rate and the ammonium
nitrogen loading is shown in Fig.it. The
obtained maximum removal rate was
k gMlj-N/daym2. A comparison between the
observed specific growth rate of nitrify-
ing bacteria and the observed specific
removal rate of ammonium nutrogen is shown
in Fig.5> The yield coefficient could be
determined to be 0.15 from these results.
It vas significant that both the maximum
specific growth rate of nitrifying bacte-
ria and the maximum specific removal rate
1 -i"
OS
0.2
3 1
3 d
II
II
Arcnonium nitrogen loading', -
1Q
Fig.U. Relationships between
ammonium nitrogen removal '
rate and ammonium nitrogen
loading
S'aripheric velocity: 19.6 cra/s
Airjnonium nitrogen concentration!
in bulk liquid
10
20
30
| 0.03
u
1 0.02
H'
B 0.01
(in vet basis)
temoniua nitrogen concentration
in bulk liquid- stir/i '.
100 200 300 400
tsi5hcA liionass weight per unit area
500
Fig.3. Sloughing rate and
mean biofilm thickness
10 20 30
Fig.5. Comparison between the observed
specific growth rate of nitrifying bac-
teria and the observed specific removal
rate of ammonium nitrogen
1168
-------
of ammonium nitrogen in the Mo film on the disc were as low as 0.03 and
0.2/day, respectively, and these values were about one tenth of those(1^2 A 3J
in dispersed cultures, respectively.
Biomass sloughing rate
In this series of experiments, water temperature and pH were maintained
between l8°C and 20°C and between 7*0 and 7-5, respectively. In the case of
the rotational speed being 9-7 rpm, the dissolved oxygen concentration was
greater than 3 mg/1.
As the organic loading was increased, the thickness of Mofilm increased
rapidly and the dominant species changed into filamentous organisms.
Relationships between the attached biomass weight and the organic loading ex-
pressed in COD(Cr), and the sloughing rate of the biomass from biofilms are
shown in Fig.6 and 7, respectively. When the COD. loading was increased to
higher than 20 g/day-m2, the sloughing rate increased to 2 gSS/daym2.
Generally speaking, biofilms grew very rapidly but at the same time they were
sloughed off very frequently.
2r
S 2
Peripherie velocity: 15.2 cm/s
•s I
?, 1
Peripherie velocity: 15-2 cm/s
gCOD/daym2
7
6
-------
* 3000
» 2000
s §
5 Si
1000
30
1 |20
Q
8 10
5
j?!
13
2
»-
0
COD loading: 35 gCOD/daynp
Hydraulic loading: 0.11 np/day-m2
Rotatinal speed
rpm
10
5 X) cra/s 15
Peripherie velocity
Fig.8. Effects of rotational speed
Peripherie velocity: 19.6 cmfs
Kieldal-H loading: 3.0-5.8 g/daym2|
Dissolved Oxygen: U-6 rog/1 ,};
pH : 6.9-8.5 :
Water temperature: 20-23°C ;
COD loading for one stags gCOD/day-m2
Fig.9. Critical organic loading
for nitrification
Critical organic loading for nitrification
A relationship between the production rate of oxidized nitrogen (WOX-N
= N02-N + N03-N) and COD loading is shown in Fig.9.
From the "balance of removed total nitrogen, produced NOX-N, and nitrogen
"taken into increased biomass, we could not find any evidence of the occur-
rence of denitrification. Therefore, the production rate of WOX-N was nearly
equal to the nitrification rate and it was concluded that nitrification could
not proceed when COD loading was increased to higher than 20 gCOD/daym^.
/ , , ._ ;
Viable counts of nitrifying bacteria
The time dependent variation of viable counts of nitrifying bacteria and
aerobic heterotrophic bacteria in biofilms were'examined. This was carried
out using the second stage unit of a four-stage RBC, because it was necessary
that there exist simultaneously a certain amount of nitrifying bacteria and
aerobic heterotrophic bacteria as an initial condition. After the organic
loading was increased suddenly, the biofilm grew very rapidly and frequent
sloughing began. As the amount of biomass on discs increased, viable counts
of aerobic heterotrophic bacteria increased, but those of nitrifying bacteria
decreased as shown in Fig.10.
Denitrification by the use of a recirculating system
As seen previously, denitrification did not occur in the usual one- ;
through operation. There have been some reports which give attempts to pro-r
ceed with denitrification by recirculating the effluent to the first stage
of the RBC. In the present series of experiments, correlations among the
1170
-------
removal efficiency of NOX-N, the recirculation ratio, and the organic loading
•were investigated concerning the removal of total nitrogen from the water as
a whole system.
•bacteria ,
A' relationship "between the NOX-1T
removal efficiency and COD loading,
which was loaded on the first stage
under the constant hydraulic, loading
of 0..08 m-Ydaym2, is shown in
Fig. 11. Although the observed NOX-N
removal efficiency is not identical
to the denitrification efficiency in
a strict sense, we can see that de-
nitrification proceeded very well
under higher organic loading than
20 gCOD/day.m2.
can expect about
90% NOX-N removal under the COD load-
ing of ^0 gCOD/day^m2 in the first
stage.
The total nitrogen (T-N) removal
efficiency through a whole stage was
calculated from the "balance between
concentration of T-W of influent and
that of effluent. A comparison of
results with and without recirculation
is shown in Fig. 12.
50-r
OJ
3"
s 40
0)
° in
"s 30
If ^
1$ 10
Recirculation ratio: O.M-0.53
T-N loading for one staget 3.5->i.9
through a whole stage
5 gCOD/day.m2 10
10
COD loading .
20 30 40
for one stage
Fig.12. Comparison of T-U removal
efficiency through a whole stage
with and without recirculation
Fig.10. Variation of viable counts
of nitrifying bacteria with time
under sudden high organic loading
100
50
Recirculation ratio: O.Uo-0.53
-e-
COD loading
in the first stage
0 K) 20 30 40
Fig.11. relationship between MX.-N
removal efficiency and COD loading
in the first stage
In the case without recirculation, the T-N removed from water might be con-
verted to biomass. Accordingly, we can expect about a 10 per cent increase
in T-N removal efficiency by denitrification of the recirculated water in the
first stage. Relationships between the recirculation ratio and the NOX-K
1171
-------
removal efficiency in the first stage, T-N removal efficiency through a whole
stage, and concentration of ammonium nitrogen (KH^-N) in effluent water are •
shown in Fig.13.
Nitrification and denitrification on the same disc
To investigate the possibility of simultaneously proceeding with "both
reaction of nitrification and denitrification in the tdofilm on the same disc,
a preliminary experiment was carried out. According, to results of this exper-
iment as shown in Fig.lH, we could not confirm this possibility when glucose
was used as an organic source. However, confirmation was possible, as antic-
ipated, when methanol was used as the organic source.
100r
|
50
Methanol used, as,, an, organic source
2*0
? 1.0
-j_a through a «hole
COS
first stage:
3S-U8 g/dayin2
t ] Effluent HHj,-N
i^M Effluent 163-11
ESSSJ 1<2 (via N03-H)
g
.i §^|
5 3 I
•H R) •
so ra
^*"*^_N
, 1 •
i:-:'-:-3 Biosynthesis;
S
S
'
i
" B
0 10 20
Glucose used as an organic source
i :
Heoirculation ratio
Fig.13. Relationships between
recirculaiion ratio and NOX-W
removal efficiency,'T-H removal
efficiency, and effluent
concentration
DISCUSSION
2.0
1.0
COD loading
for one stage
10
20
gCOD/day-m2
Fig.l^. Conversion of influent ammonium
nitrogen with methanol and glucose used
as an organic source
The obtained yield coefficient of nitrifying bacteria in this study is
nearly equal to the value of O.l6 which can be calculated from the following
empirical equations . ' 5 )
Nitrosomonas
55
76 02 + 5 C02 =
109
Nitrobacter
UOO NOg + 195 02
5 C02
2 H2° =
52 H20
H
However, the observed maximum specific growth rate of nitrifying bacteria
was apparently small. The effective thickness of a biofilm for nitrification
was estimated using the mathematical kinetic model as summarized in Tables k
and 5.
1172
-------
Table U. Mathematical kinetic model for zero order reaction
Ammonium nitrogen
Dissolved oxygen
Mass balance : equation
Boundary conditions
Substrate concentration
for zero order reaction
Effective biofilm
depth ( Le )
Remarks
= 0 at z = Le,
Ci0 at z
- R
- R
0 at Z = L
CD
developed and listed
in different expressions
by Williamson and McCarty
^S'
LP =
assuming that the removal rate of
oxygen is proportional to that of
ammonium nitrogen and the dissolved
oxygen concentration decreases to
0 mg/1 under the oxygen transfer
limitation
Nomenclature
ci»C*
R, R*
k
I>b,Db
z
Le
cio
//m
' y
X
M, N*
D$
" concentration of ammonium nitrogen and oxygen
in the biofilm, respectively [M /L^]
= reaction rate of ammonium nitrogen and oxygen .respectively [M/LJT]
proportional constant
= diffusion coefficient within the biofilm
for ammonium nitrogen and oxygen, respectively [L2/T]
= distance into biofilm from the biofilm surface [L]
= effective bjQfilm depth [L]
= c"oncentration °t ammonium nitrogen at the biofilm surface [M/IP]
= the maximum specific growth rate of nitrifying bacteria [1/Tj
= yield coefficient
= bacterial concentration in the biofilm [M/L^]
= surface flux of ammonium nitrogen and oxygen, respectively [M/T'L3]
= diffusion coefficient through water for oxygen [L2/T]
= stagnant liquid layer depth [L]
Co = dissolved oxygen concentration in bulk liquid [M/L ]
Table 5. Effective thickness of the biofilm calculated from Eqs.(l)and(2)
Eq. (1)
Eq. (2)
Effective biofilm thickness '(L,J
30 - 110 JU
m
20 - 120
Terminal concentration of
ammonium nitrogen ( C± at
when the concentration of
ammonium nitrogen in the
bulk liquid is 10 mg/1
= Lj
5-9 mg/1
6-8 mg/1
Parameter values used
Y = 0.16
k = H.25
Area factor = 1.0
°b'Dw =
X =
C* = 5 - 8 mg/1
2.5 cm2/aay(5)
0.2 - O.U /day
30 - 50 mg/cm3(6)(7)
= 50 - 100 //
Diffusion coefficient through water for ammonium nitrogen =
1173
-------
These tables show that the biofilm is under the condition of oxygen transfer ,
limitation, and it can he inferred that the effective thickness was approx-
imately 50 Urn to 100 X/m, about one tenth of a whole thickness of the biofilm,;
which explains the small value of the observed specific growth rate. :
Therefore, it can be assumed that nitrification can proceed only in the very
thin surface layer of a biofilm, and that the continuous rapid growth of
aerobic heterotrophs and the frequent sloughing of biofilms should eventually
cause the extermination of nitrification in the unit. As a numerical value
of the critical organic loading which exterminates nitrification, a value of
20 gCOD/day-m^ was obtained for one stage of the KBC used in this study. At
this organic loading, the sloughing rate of biofilms reached 2 gSS/daym^.
This rate was far beyond the maximum growbh rate of nitrifying bacteria of
O.T gSS/day-m^ which was calculated from obtained values of the maximum re-
moval rate of ammonium nitrogen and the yield coefficient.
It seems very promising to proceed with denitrification by using organic
substances in influent water. However, it appears necessary to maintain a
high organic loading, more than 20 gCOD/day-m^ for one stage of the RBC used:
in this study, in order to proceed with denitrification. For the case of
high organic loading, it may be very effective to recirculate the effluent '•
to the first stage of a multistage EEC to remove nitrogen without any other :
addition of organic substances and special facilities. When methanol only
was used as the organic material, the biofilm appeared to be stable enough to
maintain the nitrifying bacteria in the surface layer and the denitrifying
bacteria in the internal layer because of the rather small growth rate of the
aerobic methylotrophs . In this special case, both nitrification and denitri-
fication may proceed simultaneously on the same disc.
SUMMARY
The results obtained above can be summarized as follows :
(l) Nitrification proceeds mainly in the surface layer of the biofilm
and nitrifying bacteria in this layer may be washed out easily by frequent
sloughing of biofilms under higher organic loading.
As nitrification is rather sensitive to organic loading, it is recom-
mended that multistage RBC be used where each stage is connected by cascade j
to prevent the ill-effects of intermixing and to maintain stable conditions .
for nitrification.
(3) Under higher organic loading, the recirculation of the effluent to
the first stage unit may proceed with denitrification and will improve the
nitrogen removal efficiency. \
(k) Both reactions of nitrification and denitrification can be done siT
multaneously by the use of methanol as the only organic source, but this may
not be expected when other ordinary organic substances are employed.
1174
-------
ACKNOWLEDGMENT
Thanks are due to Mr.Y.Misawa for his assistance. This work was.finan-
cially supported by the Special Project Research (1979, the primary investi-
gator: J.Matsumoto) of.the Ministry of Education,Japan.
REFERENCES
1. Williamson,K.J., and McCarty,P.L.,"A Model of Substrate Utilization by
Bacterial Films" J.WPCF, kQ, 9-2k (1976)
2. Toya,Y."Studies on the Biological Denitrification Process (l)" Jr.Japan
Sewage Works Association, 7, 21-1*2 (1970)
3. Stanier,R., Aderberg,E. , and Ingraham,J.,"The Microbial World" Hh Ed.,
Prentice-Hall Inc., Hew Jersey (1976)
h. Kojima,A., et al.,"Nitrogen and Phosphorous Removal and Eutrophication
Control" (in Japanese), I.P.O., Tokyo (1977)
5. Williamson,K.J., and McCarty,P.L./'Verification Studies of the Biofilm
Model for Bacterial Substrate Utilization" J.WPCF, ^8, 281-296 (1976)
6. Hoehn,R.C., and Ray,A.D./'Effects of Thickness on Bacterial Film"
J.WPCF, 1^5, 2303-2320 (1973)
7. Grieves,C.G. /'Dynamic and Steady State Models for.the Rotating Biological
Disc Reactor" Ph.D. thesis, Dept. of Environ. Syst. Eng., Clemson Univ.
(1972)
1175
-------
-------
I
NITRIFICATION ENHANCEMENT THROUGH PH CONTROL
WITH ROTATING BIOLOGICAL CONTACTORS
By
James M. Stratta
Graduate Student
and
David A. Long
Associate Professor of Civil Engineering
The Pennsylvania State University
University Park, PA
INTRODUCTION
Recent emphasis on nitrogen control in receiving bodies of water has
required the removal of ammonia nitrogen from the effluent of many waste-
water treatment facilities. The desire to utilize wastewater treatment
technology compatible with existing facilities, minimize energy requirements,
insure ease of operation, and achieve the desired ammonia removal efficien-
cies has resulted in the current examination of rotating biological contactor
(RBC) systems for possible widespread utilization in such tertiary treatment
system applications (1-7).
One of the potential problems associated with such fixed film nitri-
fying systems is the reduced rate of nitrification which occurs when treat-
ing low alkalinity-low pH wastewaters (1,2). It is the intent of this
paper to discuss a current research effort which will ultimately assist in
the development of design criteria for RBC systems for use in nitrifying
low alkalinity-low pH wastewaters.
1177
-------
BACKGROUND
The organisms which convert ammonia to nitrate are mainly a mixed
population of chemoautotrophic bacteria which initially convert ammonia
to nitrite and then to nitrate. Energy is derived from this process which
is used to support other metabolic processes. The Nitrosomonas and •
Nitrobacter bacteria are the two primary genera responsible for this nitri-
fication process. The respective oxidation reactions are presented below
(1). As can be seen from these reactions, the nitrifying process results
in the creation of acid which effectively neutralizes the alkalinity in the
wastewater. Theoretically, 7.1 mg/£ of alkalinity is destroyed for each
mg/Jt of ammonia oxidized. The destruction of alkalinity results in an
increasing pH depression as the amount of ammonia oxidized becomes greater.
The actual magnitude of the pH depression is somewhat mitigated by the
removal of carbonic acid through the stripping of C02 from the wastewater
surface (1).
Ammonia Oxidation:
NH* +1.5
, „ ,,„«- Nitrosomonas
+ 2 HC03 *•
(1)
Nitrite Oxidation:
' + 0.5 02 Nitrobacter, NQ-
(2)
Overall Reaction:
+ 2 02 + 2 HC03-
(3)
The observed level of alkalinity varys significantly among wastewaters.
The major influencing factor in the amount of alkalinity present is the
drinking water supply which ultimately discharges into the wastewater treat-
ment system. Relatively high natural alkalinities are associated with certain
ground water supplies and much lower alkalinities normally are associated with
surface supplies. Domestic usage adds approximately 100 mg CaCOg/ft to the
natural alkalinity of the carriage water (8). Therefore, the amount of
alkalinity in a domestic wastewater may range from 100 mg CaCQ^/H or less to
several hundred mg CaCO^/^l. The net effect of such variations in alkalinity
is to provide a different buffering capacity for"each wastewater treatment
system. Domestic wastewaters may have from about 12 to 25 mg/£ of ammonia
nitrogen. The range of alkalinity destroyed during the nitrification of such
ammonia concentrations is 85 mg CaCO-j/£ to 178 mg CaC03/£. Obviously the pH
depression can be very slight for low ammonia-high alkalinity wastewaters or
very significant for high ammonia-low alkalinity wastewaters. The effect of
this relationship on nitrification becomes more significant as the amount of
alkalinity drops or the level of ammonia increases.
A number of publications have, to varying degrees, addressed the effect
of pH on the nitrification process (1,2,3,7,9,10,11,12). The reported data
show a wide variety of optimum pH. In general, the nitrifying organisms
1178
-------
function in an optimum mode at approximately pH 8.0 to 8.5 The nitrifying
performance then decreases with decreasing pH to about pH 5.5 where nitri-
fication essentially ceases. The reported rates of nitrification are rela-
tively ambiguous between pH 5.5 and 8.0. Unfortunately, data gathered to
date on RBC wastewater treatment systems largely have failed to establish
the difference between long and short term effects. The research efforts on
the effect of pH on nitrification with wastewater treatment systems in most
cases have been short term studies (10). One of the problems involved in
investigating the nitrification process is the slow growth rates of the
organisms. This factor has been a major reason for the inability to date to
properly establish the long term effect of pH change relative to short term
pH changes; The ability of an organism to successfully acclimate to lower
pH values has a significant bearing on the degree of biological nitrification
which is to be expected for any particular wastewater treatment system. If
engineers are to be able to design RBC biological nitrification systems
which optimize the rate of nitrification, more definitive research is neces-
sary to clearly establish the relative rates of nitrification for both short
term and long term pH changes. Only after the long term nitrification rates
have been established can the benefit of pH adjustment be made clear.
The literature also fails to address the efficacy of alternative chemi-
cals and chemical feed schemes for the adjustment of pH in RBC systems in
order to achieve optimum nitrification. Such pH adjustments have been
attempted for both RBC and activated sludge systems (2,13). Unfortunately,
these studies have been inconclusive in establishing the benefit of chemical
addition. Studies on RBC systems which have treated high strength ammonia
wastewaters have reported success in practicing alkaline addition (14,15).
However for those RBC facilities which are designed to nitrify low alkalinity-
low pH domestic wastewaters, additional research in this area would assist de-
sign engineers in their attempts to provide the greatest rate of•nitrification
at the least initial capital expenditure and operation and maintenance cost.
SCOPE OF RESEARCH
The objectives of this ongoing research effort are:
1. Establish the relative rates of nitrification for domestic waste-
water treatment within an acclimated RBC fixed film system as a function of
pH (between pH 6.0 and 9.0).
. 2. Observe and characterize the relative changes in the attached bio-
logical film as a function of pH.
3. Evaluate the efficacy of chemical addition to improve nitrification
within an RBC fixed film system through the maintenance of an optimum pH.
4. Evaluate prime candidate chemicals for pH controlled nitrification
for the RBC.
5. Develop economic design criteria for the RBC for pH controlled
nitrification.
1179
-------
RESEARCH PLAN
RBC Pilot Facility
Secondary effluent from a trickling filter at the Penn State University
(PSU) Wastewater Treatment Plant was selected for use in this evaluation. A
schematic of the PSU wastewater treatment facility is shown in Figure 1.
Approximately 1.7 m^/day of wastewater which is recirculated around the PSU
trickling filters is pumped to a pilot 0.5 m RBC pilot facility where tempera-
ture and flow are adjusted as desired. The character of the wastewater
entering the RBC is described in Table 1. The pilot RBC facility is shown
in Figure 2. Operating characteristics of the pilot RBC are presented in
Table 2.
The operation of the RBC pilot unit is maintained at a hydraulic loading
of 81 £/m2/day (2 gpd/ft^), a rotational speed of 13 rpm, and a temperature
in the first stage of 20°C. Preliminary pilot plant data which are presented
in Table 1 demonstrate that the character of the wastewater entering the RBC
is relatively low in BOD5 and suspended solids. These data also show that
some nitrification is occurring in the PSU sewage treatment plant. The
pilot RBC system is functioning essentially as a separate stage nitrification
process rather than a combined carbon oxidation-nitrification system.
TABLE 1. RBC INFLUENT WASTEWATER CHARACTERISTICS
BOD5 (soluble)
Suspended Solids —
Temperature
Flow
NH3-N
NO~-N + NO~-N
6 mg/£
18 mg/£
21°C
1.7 m3/day
14 mg/£
3 mg/£
Phase I - Nitrification as a Function of pH.
The initial phase of the research has been devoted to the establishment
of an RGB system which is capable of nitrifying the natural wastewater con-
taining approximately 290 mg CaCC>3/£ of alkalinity. Figure 3 shows a typical
diurnal pattern of the ammonia-nitrogen entering the PSU wastewater treatment
plant. The dampening of this diurnal pattern by.primary treatment is illus-
trated in Figure 4 which also shows the diurnal pattern of ammonia-nitrogen
within stage 1 of the pilot RBC. The ammonia-nitrogen removal performance
of the pilot RBC from start-up is shown in Figure 5. In approximately 2
weeks after start-up, the RBC was producing an effluent with less than 0.05
mg NHo-N/Jl. Figure 6 demonstrates the degree of treatment provided during
steady state operation. Obviously, at the current hydraulic loading rate of
81 Jl/m^/day, the first stage of the RBC is the most important and removes the
greatest amount of ammonia nitrogen. Stages 2 and 3 provide additional
ammonia removal; however, stage 4 provides no real additional treatment. The
pH levels in stages 1 and 2 were 7.5 and 7.6 respectively. However, as the
amount of ammonia oxidized in stages 3 and 4 declined, the pH in the later
1180
-------
PREAERATION
PRIMARY
CLARIFICATIO N
SECONDARY
C LAR I FICATION
CHLORINAT10N
Figure 1. Penn State University Wastewater Treatment Plant.
1181
-------
FLOW
CONTROL
PILOT R B C
U STAGES
Figure 2. Pilot RBC Wastewater Treatment Plant.
stages rose due to the stripping of C02. The pH in stages 3 and 4 were 7.8
and 7.9, respectively.
TABLE 2. PILOT RBC OPERATING,. CHARACTERISTICS
Disc Diameter •
Disc Area, Total
Number of Stages
Discs per Stage
Rotational Speed
Peripheral Speed
Hydraulic Loading
Hydraulic Residence Time
Wastewater Temperature
0.5 m
21.3 m2
4
9
13 rpm
0.34 m/sec
81 £/m2/day
116 min.
20° + 1°C
Based upon these preliminary data, subsequent phase 1 research will not
utilize the entire RBC as a four stage treatment system. Instead, the
influent flov? will be increased four fold and split equally into each of the
1182
-------
Figure 3. Diurnal Ammonia Variation Entering the Pennsylvania State
University Wastewater Treatment Plant.
1183
-------
20
16
12
RBC INFLUENT
STAGE 1
AM
PM
1
AR-
AM
Figure 4. Diurnal Ammonia Variation Entering the RBC and in Stage 1 of
the RBC.
1184
-------
h-
z
LU
08
CM CO
1
o
CM
1
CO
1
CM
|
CO
1
-------
LU
O
<
H
CO
o
CQ
DC
a
o
•H
M
o
CJ
•H
(3
o
- NOIlVaiN3ONOO
vO
Q)
3
•H
1186
-------
four stages. Therefore, each of the original stages will function as a
single stage system. Because the first stage is the most critical and capa-
ble of adequately demonstrating the effect of pH, four different pH's and
alkalinity conditions will be evaluated simultaneously. Each set of discs
will receive the same wastewater, at the same temperature, the same influent
dissolved oxygen level, and the same influent concentration of soluble ZOT>$.
Only the pH and alkalinity will be adjusted .for each stage.
The pH adjustment phase will be used to examine RBC performance at pH
7.5, 7.0, 6.5 and 6.0. This part of the study will be terminated when steady
state conditions are achieved. The experiment then will be repeated at pH
7.5, 8.0, 8.5 and 9.0, In each part, the pH 7.5 stage will serve as a
control. The performance of the RBC stages will be evaluated on the basis
of ammonia oxidation, solids generation, film establishment and population
of nitrifiers on the fixed film surfaces. Each of the stages will start with
a completely clean set of discs so that each stage will develop a fixed film
nitrifying population which favors the particular pH of that stage. It is
anticipated that differences in the nitrifying performance will be reflected
in differences in the respective populations.
Biological Film Monitoring and Microbial Enumeration
The effect of altering the pH will result in changing the performance
characteristics of the RBC system. This change will be reflected in the rate
and magnitude of film buildup and the microbial composition of the disc film.
The RBC pilot has been modified to facilitate the examination of the attached
film. Each of the four stages is composed of eight molded polyethylene discs
attached to a common central shaft. Each disc has approximately 0.63 m^ of
surface area available for film attachment. An additional flat plexiglass
disc has been added to each state thereby providing an additional area of
0.27 m^. Sections of Mylar are attached to each of the plexiglass discs to
allow for easy removal. The biological films are removed from the Mylar and
broken up by blending and evaluated for ammonia oxidizing and nitrite oxidiz-
ing bacteria as well as for total and volatile solids. The Nitrosomonas are
grown in an ammonium-calcium carbonate media and the Nitrobacter are grown
in a nitrite-calcium carbonate media (16). The method of enumeration utilizes
a 96 well tissue culture plate (17,18). Because of the slow growth rate of
the organisms the microbial enumeration procedure requires approximately 28
days of incubation before the final MPN values can be determined. Preliminary
data at pH 7.5 which demonstrates the biological slime buildup and microbial
enumerations are provided in Figures 7 and 8, respectively. The biological
film development continued throughout the seven weeks of the preliminary
operation of the pilot RBC. The heaviest growth was noted on the first stage
with decreasing growth in the later stages. Although the nitrification per-
formance of the unit had stabilized by mid-November, the film continued to
develop throughout the balance of the initial test period. This observation
indicated that only a. fraction of the biological film was actively engaged in
ammonia and nitrite oxidation. The enumeration of ammonia oxidizing and
nitrite oxidizing bacteria as shown in Figure 8 was performed after seven
weeks of operation. In stages 1 and 2, the number of Nitrosomonas bacteria
were greater than Nitrobacter bacteria; however, this trend was reversed in
stages 3 and 4. This change apparently was due to the disappearance of
ammonia-nitrogen in stages 3 and 4 and lingering concentration of nitrit^e-
1187
-------
CO
,uio/6m -
CM
0)
a
H
•H
S
0)
6
0)
1)
p
3
•H
O
•H
O
r>
cu
60:
•H
1188
-------
CM
£
o
10C
10'
106
1
NITROSOMONAS
2
NITROBACTER
3
RBC
STAGE
Figure 8. RBC Microbial Enumerations.
1189
-------
nitrogen. Both populations decreased by approximately an order of magnitude
across the four stages of the RBC pilot plant.
Phase II - pH Adjustment
The last phase of this research effort will be to examine alternative
schemes for pH control so that nitrification can be optimized. After phase I
is completed, the optimum pH level will be identified. The wastewater will
be altered through chemical addition to simulate a low pH-low alkalinity
wastewater. Alternative chemical additives will be evaluated in enhancing
the nitrification process. The alternative chemicals to be evaluated are
lime, soda ash, sodium bicarbonate, and sodium hydroxide.
SUMMARY
The results of the two year research effort will be used to provide
more definitive guidance on the long and short term effect of pH for fixed
film RBC nitrification. The results also will more clearly establish the
optimum pH level required for nitrification with full scale RBC facilities
as well as the means by which proper pH adjustment can be made to insure
optimum nitrification.
ACKNOWLEDGMENT
This research is supported by the U.S. Army Medical Research and
Development Command under Contract No. DAM 17-70-C-9110.
1190
-------
LITERATURE CITED
1. Process Design Manual for Nitrogen Control, U.S. EPA, Oct. 1975.
2. Hitdlebaugh, John A., "Phase I - Water Quality Engineering Special
Study No. 32-24-0116-79, Sewage Treatment Plant Evaluation - Summer
Conditions, Fort Knox, Kentucky," U.S. Army Environmental Hygiene
Agency, 16 March 1979.
3. Borchardt, Jack A., et al., "Nitrification of Secondary Municipal
Waste Effluents by Rotating Bio-Discs," EPA Report 600/2-78=061,
June 1978.
4. Hewitt, T., "Nitrification of a Secondary Municipal Effluent Using
a Rotating Biological Contactor," Wastewater Treatment Section,
Pollution Control Branch, Ontario Ministry of the Environment,
April 1978.
5. O'Shaughnessy, James C., and Frederic C. Blanc, "Biological Nitrifica-
tion and Denitrification Using Rotating Biological Contractors,"
Water Resources Research Center, University of Massachusetts, July
1978.
6. Antonie, Ronald L., Fixed Biological Surfaces - Wastewater Treatment,
CRC Press, Cleveland, 1975.
7. Miller, Roy D., et al., "Rotating Biological Contactor Process for
Secondary Treatment and Nitrification Following a Trickling Filter,"
U.S. Army Medical Bioengineering Research and Development Laboratory,
Ft. Detrick, MD, June 1979.
8. Metcalf and Eddy Inc., Wastewater Engineering: Collection, Treatment,
and Disposal, McGraw-Hill, New York, 1972.
9. Haug, Roger T., and Perry L. McCarty, "Nitrification with Submerged
Filters," WPCF, Vol. 44, No. 11, p. 2086, 1972.
10. Mitchell, Ralph, Water Pollution Microbiology, Wiley-Interscience,
New York, p. 391, 1978.
11. Painter, H. A., "A Review of Literature on Inorganic Nitrogen Metabolism
in Microorganisms," Water Research, Pergamon Press, Vol. 4, pp. 393-
450, 1970.
12. Wild, Harry E., et al., "Factors Affecting Nitrification Kinetics,"
WPCF, Vol. 43, pp. 1845, 1971.
13. Heidman, James H., et al., "Carbon, Nitrogen, and Phosphorus Removal in
Staged Nitrification - Denitrification Treatment," EPA 670/1-75-052,
June 1975.
1191
-------
14. Lue-Ling, Cecil, et al., "Biological Nitrification of Sludge
Supernatant by Rotating Discs," WPCF, Vol. 48, p. 25, January 1976.
15. Prakasara, T. B., et al., "Nitrogen Removal From Digested Sludge
Supernatant Liquor Using Attached and Suspended Growth Systems,"
Proceedings - 32nd Purdue Industrial Waste Conference, 10-12 May
1979.
16. Alexander, M. and F. E. Clark, "Nitrifying Bacteria." In C. A. Black
(ed.), Methods of Soil Analysis, Part 2, Am Soc. Agron., Madison,
WI, 1965.
17. Olem, Harvey, Rotating-Disc Biological Oxidation of Ferrous Iron in
Acid Mine Drainage Treatment, Ph.D. Thesis, Civil Engineering
Department, The Pennsylvania State University, 1978.
18. LaBeda, David P. and Martin Alexander, "Effects of SC^ and NC>2 in
Nitrification in Soil," J. Environ. Qual., Vol. 7, No. 4, p. 523?
1978.
1192
-------
NITRIFICATION OF MINICIPAL WASTEWATER
USING ROTATING BIOLOGICAL CONTACTORS
by
James C, O'Shaughnessy
Associate Professor of Civil Engineering
Northeastern University, Boston, Mass., U.S.A.
Frederic C. Blanc
Associate Professor of Civil Engineering
Northeastern University, Boston, Mass., U.S.A.
Peter Brooks
Engineer, Water Pollution Control Board
Commonwealth of Virginia, Richmond, VA, U.S.A.
Alan Silbovitz
Engineer, Weston & Sampson Engineers
Boston, Mass., U. S. A.
Richard Stanton
Engineer, U.S. Internal Revenue Service
Boston, Mass., U. S. A.
1193
-------
INTRODUCTION
The need for nitrogen control has been realized In recent years to up-
grade the quality of the waters receiving wastewater effluents. One method of
upgrading wastewater effluents Is using biological nitrification within the
treatment process.
The biological process in which ammonia nitrogen is converted to nitrate
is a two-step process in which nitrosomonas bacteria convert ammonia to ni-
trate and nitrobacter and then convert the nitrite nitrogen to nitrate nitro-
gen. The stoichiometric equations for this synthesis of oxidation process can
be summarized as follows- (1) :
55 NH^ +76 02+ 109 HCO Nitrosomonas \
400
2
4-
C H?N02 + 57 H20 H- 104 lUCO..
cells
+4
+ 195 0_ Nitrosobacter
C,H^NO + 3 H 0 + 400 NO "
•J I £~ & J
Since carbonic acid is formed in this process, and bicarbonate levels are
reduced, as Bicarbonate alkalinity is consumed the pH will drop. It can be
theoretically calculated that approximately 7 mg/1 of alkalinity will be de-
stroyed for every milligram of ammonia nitrogen that is oxidized.
Presently, Rotating Biological Contactors (RBC's) are being used for the
secondary treatment of domestic wastewater, industrial wastewater, for septage
treatment, and for mitrification and denitrification. There are many firms
manufacturing and marketing such devices across the nation. Differences exist
between the manufactured units currently available. Such differences as surface
configuration, materials, sizes and the surface to volume ratio of the reaction
chambers exist in various designs and applications.
The design considerations used in the selection of RBC's are as follows:
1. Influent wastewater characteristics
2. Hydraulic loading
3. Operation -temperature
4. pH control
5. Effluent wastewater requirements
6. Disc velocity
7, Size and surface area of media
8. Substrate Loading rates
It is important that these parameters be investigated carefully before
selection or sizing the units.
1194
-------
Most of the available literature to date deals with reduction of
within an RBC unit. Limited data has been published on the performance of
RBC units when used to achieve biological nitrification of a domestic waste-
water.
The primary objectives of the research study presented in this paper
were:
Perform a comparative analysis of the nitrification efficiencies of
the rotating biological contactor system with an operating two-stage
activated sludge system.
Establish treatability parameters involved with an RBC system.
These parameters include organic and nitrogen loading rates, sludge
production, surface area requirements for proper F/M ratios.
EXPERIMENTAL FACILITIES AND PROCEDURES
Both laboratory bench scale units and field pilot units were used for
data collection.
The laboratory study was conducted in the Environmental engineering lab-
oratories located at Northeastern University. The bench scale units used in
this study consisted of two pairs of contactors with a surface area equal to
9^08 ft^/stage. There were ten (10) discs/stage. Volumes were equal to .7 gal/
stage. These units were constructed at Northeastern University using plastic
media supplied by EPCO-Hbrmel.
The wastewater flow scheme is shown in Figure 1. The units were fed sec-
ondary treated effluent from the Marlboro Easterly Wastewater Treatment Plant.
Two units 05 and D) received normal secondary effluent while two units (A and
G) received the same effluent but the feed was "spiked" with ammonia chloride
so that the ammonia concentration was 40 mg/1 NH3~N. Fresh effluent was trans-
ported to the laboratory every other day.
2
'Units A and B were .four stage units with a total surface area of 36.3 ft
and liquid volume of 2.8 gal. Units C and D were single stage units with an
expanded liquid volume chanber. Throughout the study, the effective liquid vol-
ume of Units C and D was 2.8 gal.
Grab samples were taken at intervals, based on flow and detention time
per stage, so that the flow could be followed through the units.
Figure 2 shows the set-up for both a single stage unit and a four stage
unit.
Four different pilot RBC systems were used during the field portion of this
study. The units were located at the Marlboro Easterly Wastewater Treatment
Plant in Harlboro, Massachusetts, Each of the units received effluent pumped
directly from the secondary clarifier of the plant's activated sludge treat-
ment system. The pilot units were located in the treatment plant's underground
service tunnels.
1195
-------
0)
H
CO
a
o
^_
A
o
cd
O
u
en
O
1-1
1196
-------
!=
o
ID
3
U
Li
— *
O
o
»
-
.
.
in
r-
-* *3
fB 3 -
u «•* i
J*-*-~ ~
/^
/
/
; >
V
v
\ —
a]
- — — — -~.
I
1
'
, 1
1
™1
*
_(
J
J
*
C
3
3 C
) £«
-~,
I~ -«. °
N \ /
1 | I i
/ C
/ / >
— ^
s
XJ
O ""
3
a
0)
en
^
JS
a
u
o
5
^
•u
C
CJ -*
_
^
l
s
0
-
O
1
- — —
1
1
•
_
I I I
J _
1
J
1
. _ _
__ j
1
-
1
~1
1
1
j I
1
-
W
*> 0)
...J r4
5 bOO
^ CO CtJ
o O
a r* ...j
j_j i^
4J ij
M jj ^-1
-* LLJ
^_ tL|
" JD m
1
Ji a) a»
> CO CO
•a ?~»
-i J-J CO
w O 4->
ctf C!
eg pq
>»3 K
*»
CN
j-q
g
O
fe
1197
-------
The first system (Unit 1) utilized was a four ft. diameter four stage
system supplied by the Environmental Pollution Control (EPCO), a subsidiary
of the George A. Hormell Company of Austin, Minnesota. There was 1,570 square
feet of disc surface area, equally spaced in each, stage. The unit was run
with a liquid volume of 150 gallons. The volume to surface ratio of the unit
was 0.095 gal/ft2.
The second unit (Unit 2) utilized a two ft. diameter four-stage contac-
tor which, was supplied by the Autotrol Corporation of Milwaukee, Wisconsin.
This unit had a total surface area of 250 square feet, the liguid volume was
27 gallons, there were 36 bio-discs equally spaced within the four stages,
and a surface to volume ratio was approximately 0.12 gal/ft.
The third unit (Unit 3), constructed by the Department of Civil Engin-
eering at Northeastern University, consisted of a two ft. module containing
40 discs, 10 discs per stage with a total surface area of 327 square feet.
The discs were constructed of EPCO media. The unit's liquid volume was 129
gal., resulting in a volume to surface ratio of 0.4 gal/ft2.
The fourth unit (Unit 4) was a single stage unit, constructed by North-
eastern University with a surface area of 169 ft and liquid volume equal to
12.4 gal. giving a ratio of 0.07 gal/ft2.
Operation of the units varied, with Unit 2 being run as both a four stage
unit, and also as two separate two stage units.
Grab samples were taken at intervals, based on flow and detention time.
During the study, all ammonia values are average values b-ased on three samples
taken at three consecutive detention times. Analyses were completed within
24 hours of sampling.
The nitrification pilot units were operated a total of seven months, be-,
ginning in September and ending in December of 1976; and again from April
through June of 1977. The first time period saw Unit 1 run as two separate
two stage units. One of the.two stage units acted as a "control" unit and did
not receive additional alkalinity addition. All other units received addition-
al alkalinity. Unit 2 was run as a four stage unit, and Unit 4 as a single
stage unit. During the second time period, of the nitrification pilot study,
"Units 1, 2, and 3 were run as four stage units, all receiving additional al-
kalinity.
All samples were grab type samples. Analyses were either run immediately
after collection, or the samples were refrigerated to 4°C. All analyses were
completed within 24 hours of collection on the refrigerated samples. All water
quality testing followed Standard Methods (2).
BENCH SCALE STUDY RESULTS
The nitrification bench study was run for approximately two months follow-
ing a one month start-up period. The influent to Unit B (4-stage) and Unit C
(single stage) averaged between 10 to 15 ing HEhpN/l for the remainder of the
study. The ammonia influent and effluent concentrations for all of the units
1198
-------
are shown in Figures 3 and 4.
Both Unit B and Unit C had effluent ammonia nitrogen concentrations less
than 1.0 mg/1 (with the exception of one data point) throughout the study
period. Unit C had a slight increase in ammonia nitrogen, in the effluent
following increases in influent ammonia concentrations. However, during
periods of constant loading rates, the ammonia nitrogen concentration of the
units was less than 0.1 mg NI^-N/l.
Unit A was able to maintain low ammonia effluent concentrations (less than
0.3 mg HH3-N/1) during May. The loading rates during this period were low
and relatively stable. Unit C (single stage) followed a similar trend during
the same time period.
The increased unit loading in early June resulted in high ammonia concen-
trations in the effluents, as a result of this shock load. Both units recover-
ed from the shock loading increase, however, the four stage unit did recover
more rapidly. The four stage unit (Unit A) also achieved a higher degree of
treatment of the wastewater during this portion of the study. This early June
loading increase in Units A and D resulted from increased ammonia concentration
only while the hydraulic loading was fairly constant.
The peak in ammonia effluent concentrations which occurred in the middle
of May was associated with increased hydraulic loading rates on the RBC units.
The hydraulic residence time per stage of unit was reduced from approximately
3.0 hours/stage to 1.5 hours/stage. This hydraulic loading increase also
caused the daily ammonia loading rate to double. The two four-rstage units had
little change in effluent quality while both single stage units (Units C and
D) had higher ammonia effluent concentrations. Since all of the units had the
same liquid volume, the results indicate that staging will help reduce the •
effects of a simultaneous increase in flow and ammonia concentration.
Ammonia Loading Rates and Nitrate Appearance
The ammonia nitrogen loading rates to each unit were calculated and the
results show in Figure 5. The loading rates are expressed as Ibs. NHo-N
applied/day-1000 ft2. The Ibs. NH3-N converted/day-1000 ft2 was also calcula-
ted. The data shown in Figure 5 excludes data from periods just after sharp
loading changes, and as a result, represents data recorded during periods of
stable operation. The mathematical relationship for Ibs. of ammonia nitrogen
applied vs. Ib. ammonia nitrogen converted is shown on Figure 5. This equation
was calculated by the "least squares method", for loadings up to 0.8 Ibs. NH^-N
applied/1000 ft^-day. The correlation coefficient of the data was 0.99. Most
of the ammonia nitrogen applied -to the RBC units at such loading rates was con-
verted nitrate nitrogen.
The conversion of ammonia nitrogen represents nitrogen which was converted
to nitrate nitrogen by the nitrifying bacteria plus ammonia nitrogen converted
into cellular nitrogen compounds associated with new biomass production.
1199
-------
en
CP
o
-------
m
4J
c
Q)
3
H
4-<
C
H
P
rH
C
D
r
4->
C
Q)
0
.H
-------
— < a o Q
jj jj i> jj
-H —<—<—(
C C C C
O
O
\o
o
in
o
o
o
o •-
<
2
c
E
O T3
O 0)
-H
Cr-l
QJ D-
tsO D.
0<
Vi
4O CO
•H -H
2 C
O
« g
•H g
c<
o
E to
E-d
< C
3
W O
T3 P^
d
3 •
O M
o
o
o
o
m
o '
CN
O
I- HN -SHI
1202
-------
Table 1 represents a summary of the ammonia nitrogen conversion and the
nitrate nitrogen appearance data for this study period. An average of 93 per
cent of the ammonia nitrogen which was removed from the influent feed appeared
as nitrate nitrogen in the treated effluents. This data indicates that most
of the converted ammonia nitrogen is nitrate nitrogen, a result of nitrifica-
tion, and will remain in the effluent. Only seven percent of ammonia mass
balance appeared to be associated with new biomass.
In caluclating the mass balance, when nitrate appearance was higher than
ammonia conversion, ammonia was assumed to be totally converted to nitrate.
TABLE Ij PERCENT OF CONVERTED AMMONIA NITROGEN
WHICH APPEARED AS NITRATE NITROGEN.
UNIT A
UNIT B
UNIT C
UNIT D
95%
98%
88%
90%
MEAN VALUE FOR ALL UNITS
9.3%
The percent ammonia nitrogen converted is plotted against ammonia nitrogen
loading rate in Figure 6. In all cases, the four stage units converted 95 per-
cent of the ammonia nitrogen. The maximum loading rate for four stage units
was 0.2 Ibs, NH3~N/1000ft^-day. The single stage units had 95 percent ammonia
nitrogen conversion for loadings up to 0.8 Ibs, NH.
exceptions.
t-N/1000ft2-day, with two
These efficiencies .represent steady state periods of operation, without
shock loading effects. Higher loading rates could not be reasonably evaluated
in the laboratory study due to the logistical problems involved with transpor-
tation of the influent.
Alkalinity and pH
Additional alkalinity was added to the influent to insure that sufficient
inorganic carbon would be available for nitrification. Alkalinity additions
resulted in influent alkalinity concentrations of 250 to 350 mg/1 as CaC03 for
Units E and C. Units- A and D which has higher ammonia nitrogen concentration
in the influent had average alkalinity concentrations of approximately 500 mg/1
as CaCO^. Effluent alkalinity was greater than 150 mg/1 as CaCO_ in all cases,
The pH values ranged between 8.0 and 9.0. Both Na2C03 and NaHCO, were
used as a buffer and also to supply inorganic carbon. Influent pH values
remained constant over two to three week periods, and there were not any sudden
pH changes over the study period. The average pH value was 8.5 for influents
during May.
1203
-------
Q
W
6
o
53
63
H
a
o
CM
100
95
90
0
FIGURE 6:
• Unit A
V Unit B
@ Unit C
A unit D
4 Stage Data
4 State Data
Single Stage Data
Single Stage Data
1
~o72 ot4 oTe oTs i.o 172
LBS. NH3-N APPLIED/1,000 FT2 - DAY
PERCENT AMMONIA NITROGEN CONVERTED TO NITRATE
NITROGEN VS. AMMONIA APPLIED (BENCH SCALE UNITS)
1204
-------
Organic Nitrogen
Throughout the s-tudy-, Influent organic nitrogen concentrations averaged
4.4 mg 1F/3L» .wfcila effluent*concentrations- averaged 1,2 tog N/l.
Units A and B, the four stage units, had less organic nitrogen in their
effluent samples than did the singli stage units. All units averaged less than
2.0 mg N/l in the effluent samples, while the four stage units averaged organic
nitrogen concentration below 1.0 mg N/l. Complete data for this portion of the
study is given in O'Shaughnessy and Blanc (3).
Staging
Several times during the study, samples were taken in each of the four
stage units.
The results indicated that most of the nitrification occurred in the first
stage. In all cases, at least 90 percent of the ammonia nitrogen conversion
occurred in the first stage. The average first stage ammonia nitrogen conver-
sion vds 95 percent of the ammonia applied. The data indicated that for steady
state condition, most of the nitrification occurred in the first stage.
PILOT STUDY- RESULTS
The pilot studies conducted at Marlboro Easterly W.T.P. were run over two
separate time periods. However, discussion of all of the results will be com-
bined in this section. The two separate time periods were Fall, 1976, and
Spring, 1977. Wastewater temperatures in the units were in the same range
(10°C to 16°C) for both study periods.
Fall Study Results
During the 1976 study period, four separate units were run. They were
respectively:
1, N.U. single stage unit - 0.07 gal/ft2;
2
2. EPCO-HORMEL two-stage unit - 0.09 gal/ft (used as a control unit
without any alkalinity additions);
3. EPCO-HORMEL two stage unit - 0.09 gal/ft2;
4. AUTOTROL four stage unit - 0.12 gal/ft2.
Data from the study indicated that nitrification of secondary treated
municipal wastewater was incomplete unless additional alkalinity was added.
Alkalinity additions were always directly to the first stage of a unit, except
in the case of the control unit.
1205
-------
Figure 7 is a plot of data from the two separate two stage units. The
unit receiving alkalinity additions averaged 9.1 percent ammonia nitrogen
conversion over a two-month period, while the identical control unit only
averaged 22 percent ammonia conversion. The influent alkalinity over the two
month period averaged 80 mg as CaCO.,/1, and the pH averaged 7.0. An average
of approximately 185 mg CaCO,/l of alkalinity was required to carry out the
biological oxidation of the ammonia nitrogen. The poor ammonia conversion
ratio of the control unit when compared to the other two stage unit indicated
that alkalinity addition was required for a high level of nitrification to be
completed.
Ammonia Loading Rates
Figure 8 shows the loading rate and effluent ammonia concentrations for
the four stage EBC unit. As in the bench scale study, the. ammonia effluent
concentrations followed the loading rate. Sudden.increases in ammonia loading
rate resulted in increased ammonia nitrogen concentrations in the unit's; efflu-
ent. For all units, the lowest effluent ammonia concentrations occurred when
a constant loading rate below 0.30 Ibs NH^-N/lOOOft - day was maintained. The
sharp increase in effluent ammonia during early December can be attributed to a
sudden drop in temperature below 10°C.
The conversion of ammonia in both the two stage unit and the first two
stages of the four stage unit was fairly constant CFigure 9). An average of
90 percent of the applied ammonia was converted for a loading rate of 0.40 Ibs
NH3~N/100Qft2rKiay. Loading rates of 0.20 Ibs NH3~N/1000ft2-day and less pro-
duced between 94—99 percent ammonia nitrogen conversion (Figure 9). There
was a noticeable reduction of the percent ammonia converted as the ammonia
loading rate was increased. Only 80 percent ammonia conversion was recorded at
a loading of 0.5 Ibs Nlfcj-N/lOOOft2-day. The data plotted in Figure 9 repre-
sent loading rates and conversion of ammonia within two stage units.
The first stage loading rates and conversion efficiency was calculated for
each unit. Neglecting shock load and start-up data, the following average am-
monia conversion percentages were calculated for the study period:
Single stage unit - 56 percent at 0.51 Ibs NH3~N/1000ft2-day,
9
2-stage unit - 65 percent at 0.61 Ibs NHo-N/lOOOft -day,
4-stage unit - 73 percent at 0.78 Ibs NH3~N/1000ft2-day.
While there was a large range in both ammonia nitrogen loading rates, the
majority of ammonia nitrogen was converted in the first stage.
The data plotted in Figure 10 represents ammonia nitrogen conversion and
loading rates for both two and four stage units. An average of 90 percent of
the applied ammonia nitrogen was converted for loading rates up to 0.6 Ibs
NH3-N/1000ft2~day. This data includes both two and four stage units. For
periods of relatively steady state loaiding, the four stage unit averaged 96
percent ammonia conversion, while the two stage unit averaged 91 percent am-
monia conversion. Based on this data, the four stage unit appeared to achieve
better ammonia nitrogen conversion than the two stage unit. Since each utili-
1206
-------
I
•
tn «
O
o
Q
z
I
C
0)
CO
•P
•H
r 13
C
O
e
in
O
O -U
•H -H
to C
H-H
Q)r-l
> Cd
C^J
Or-l
r-»-
w
o
OOO'T/aaiHHANOD S—EN •SSI
1207
-------
AVO -
,ij 000'1/N- HK SET - OKJOVO'J i
(0
C
O
CO
)-i
J-J
c
0)
O
c
O
c
Q)
3
r-l
U-l
C C
co rs
c o
•i-i i— i
13 -r-t
CO P-,
O
C co
CU 4-)
(1/S-CKN
3D!;cr VTNOKKY iKSiilaiS
CO O)
-HJ2
C -U
O
1 O
00
o
M
PK
1208
-------
>•
a
E-
t.
O
O
2
i
•H
c
O
<
CO
T3
c
E
P-l CO
CD
m bO
•O CO
0)
•H O
C J3
0) 4->
WJ-H
•U
0)
CO 0)
^4 >
C C
O O
go
•I C
OJ
W bfl
T3 O
O
M
000*I/aSlH2AiIC3 K-^HN -
1209
-------
o
o
o
g
M
-SSI
-------
zed approximately the same surface area, multiple stages appeared to cause
increased treatment efficiencies,
However, the two units had different volume to surface ratios. This
resulted in the four stage unit having a 33 percent greater hydraulic resi-
dence time than the two stage unit. The effect of hydraulic resident time
will be discussed in a later section,
SPRING STUDY - LOADING RATES
During April, May and June, 1977, three units were run on the secondary.
effluent. All three units were four stage units. They were, respectively:
o
1. EPCO-HORMEL four stage unit - 0.09 gal/ft ;
O
2, AUTOTROL four stage unit - 0.12- gal/ft ;
ty
3. N.U. four stage unit - 0,4 gal/ft /
Throughout this study there were several power outages and other mechani-
cal problems both with the units and pumps. As a results, the ammonia loading
rates to each unit were more varied with time than occurred in the Fall, 1976
study period. As in both the bench and Fall pilot study, as the unit loading
increased rapidly, the ammonia nitrogen concentration in the effluent also in-
creased. However, shortly following the rapid loading increase, the ammonia
concentration in the effluent returned.to below 1.0 mg NH3-N/1. During perxods
of relatively constant ammonia nitrogen loading, the units had effluent concen-
trations below 1.0 mg NH3~N/1.
In addition to flow and loading variation, the BOD5 concentrations in the
unit influent also varied throughout the study. The average BOD5 influent
concentration for the Fall study was 10 mg/1, while the Spring study concentra-
tions averaged 22 mg/1. Prior to May 20, 1977, influent BOD5 concentrations
averaged 16 mg/1; and after May 20, the average BOD5 influent concentratxon
was 29 m°7l The reason for these different values was a change in the mode
of operation at the Marlboro Easterly. W.T.P. In the Fall study, the influent
averaged 25.5 mg NHo-N/1, while in the Spring the average was 17.8 mg NH3-N/1.
The increased BOD5 concentrations with a reduction in NH3~N concentration caus-
ed a reduction in the amount of ammonia nitrogen converted in the first stage of
the four stage units. While the Fall pilot and bench scale units converted the
majority of ammonia nitrogen in their first stage, the following results were
noted during the Spring study.
AVERAGE FIRST-STAGE DATA
Unit
EPCO-HORMEL
AUTOTROL
N.U.
Ratio: 7
gal/ft/
0.09
0.12
0.4,
% NH -N
Converted
13 .
25
35
1211
Ibs NH -N/lOOOft
Applied
0.76
1.10
1.29
-day
-------
With an increased BODc loading on all of the units, the conversion of
ammonia to nitrate in the first stage was sharply reduced. The unit with the
largest hydraulic residence time achieved the greatest percentage of ammonia
nitrogen conversion in the first stage, while the unit with the least hydraulic
residence time achieved the lowest percent ammonia conversion in the first
stage.
The data in Table 2 shows the relationship between influent BODc concen-
trations and the ammonia nitrogen conversion for various four stage units.
The overall ammonia conversion was good for all units, but the increased
BOD5 loading caused the first stage conversion reaction to either be carried on
simultaneously with BOD5 reduction, or shifted to latter stages within the unit.
The net result was reduction of the percentage of nitrification carried out
within the first stage. The longer the wastewater remained within the first
stage, the greater the percent nitrification achieved.
While the location of active ammonia conversion within the unit shifted to
latter stages, overall unit performances remained good. Figure 11 shows the
application and conversion rates for ammonia nitrogen in the three four stage
units. The N.U. four stage unit achieved the greatest ammonia nitrogen conver-
sion for loadings up to 0.6 Ibs NH3-N/1000ft -day. Since the volume to surface
ratio of the N.U. unit was approximately four times that of the other units, the
wastewater remained in each stage of the N.U. unit approximately four times
the other units. The unit with the next highest volume to surface ratio
achieved the next highest ammonia nitrogen conversion.
The data indicated that the larger the liquid volume to media surface
ratio, the higher the conversion of ammonia nitrogen within the unit. The in-
creased hydraulic residence time within the unit also helped reduce residual
BOD5 within the first stage, allowing more surface area for nitrifying bac-
teria.
However, all of the suspended solids within the unit must be kept in
suspension once they "sluff-off" from the RBC media. If the volume to surface
ratio is too large, dead space and associated solids deposition problems
may arise. In the N.U. pilot unit, submersible pumps were utilized in
each stage to keep the solids in suspension. The scale up factor associated
with full scale operations, settling characteristics of the nitrifying
sludge, basin geometry, and disc rotational speed will all affect the
determination of what is the maximum practical volume to surface ratio
that should be used in full scale plant design.
A larger volume to surface ratio will also help reduce the negative effects
of sudden peak hydraulic and ammonia nitrogen loadings, The use of staging
and optimization of volume to surface ratios are required for optimal unit
design.
Nitrate Appearance
The data for nitrate appearance is listed in Table 3. The nitrate analysis
was run only during the Fall pilot study, The .data indicated that 86 percent of
the converted ammonia nitrogen appeared as nitrate nitrogen in the two stage
1212
-------
u
E-1
..^
03
cn
o
t-i
^
:__(
2
w
o
o
CJ
2
O
O
P£
H
M
2
<
M
O
S
<£
Q
CO
2
O
I ,J
H
^5
P^
2
w
o
o
C-l
1
J
O
o
CO
csl
fd
£x
S
C-4
*
•%*•
1
CO
IT
2:
cn
£
C
£
UJ
^.
OJ
cn
cx)
co
OI
cn
ro
00
CM
Ol
cn
(O
5>
'
0)
cn
-S
cx>
> t
•Zi
»— *
LO • —
Q \
O cn
CO £
H—
1— 1
!^
1:15
UJ
H-
O
LO CM
o CD"
LO CM
• •
o o
LO CM
• •
o o
LO **O
• ••
— o
CM CO
• •
LO CM
LO CO
CM r—
CM <£>
0 0
cc ca
~T~ ^— f—
CJ O CD
•z. t— i—
1 1 1 •"*""> ~~^
py^ ^^
f*"^ P1**
"v*«* ^^»
c^j m
+-~ c\i
"**•*. ***•*»
O O
•— r—
W"
CO
•
«3"
o
•
f^^
0
CO
. ' —
J±
CO
CD
o:
i—
CD
t—
ZD
r-v
r^«»
•*^
VO
CM
'JO
•* CO CM
CM CO LO
•a- co CD
f » *
CM CO ID
O CM CM
• • •
LO LO CO
'
O O CM
^J~ f^» VO
r— r— i
CD r-~ co
CM i — CM
LO CO < —
—I
O —I —1
o; uj uj
i — is: 12:
CD CD c±: o a:
1— C_3 O (_> O
rs a_ rn a. z
vo
(JO LO
CD' i-^
r— r^*
•- »
i — CM
LO
••
i— LO
O
*
C\J \£>
l™ ~ ^tf"
r— i—
r*-. C3
, — OJ
£O VO
• *
^7^ ~~*
» *
z 2:
r**1* r*^»
r^* r^
^***. "x"*-
kO i~~
ou ro
LO 'O
1213
-------
§
I
r-j
c-.
fc,
O
o
irvi
w
z
1
CO
T3 4-1
d o
3rH
O -H
CO 0)
> t>0
n)
T) 4->
0) C/D
o
*
c
m
o
B
AVC -
1214
-------
UJ
CJ3
oo
Qi
O
o
K
CJ
CD
cx
CL
ro
CO
o
O)
>
c
o
o
CTv
o
o
LO
CO
CO
CO
CO
CO
CO
co
CO
<*o
o*.
CO
CO
*»*
• c:
,— cj
01 i—
O
i
CO i
o <
CO
tO
CTi
co
CJ
CTi
CVI
CNJ
CT> -
CD
CNJ
CD
CVJ
CO
CNJ
Oi —
Oi —
O-i
p-1
w
H
H
t—i
•z;
PC
-------
unit effluent. The four stage unit had an average conversion rate of 88 per
cent. For the pilot study, 87 percent of the converted ammonia nitrogen ap-
peared as nitrate nitrogen. The bench scale study averaged 93 percent, These
results- indicate that approximately 90 percent of the converted ammonia nitro-
gen is associated with nitrate nitrogen in the unit effluent, while about 10
percent is associated with biomass growth requirements.
Solids Production
A summary of the solids data collected during the Spring pilot study is
given in Table 4. The values are average values based on the data
•collected* The N,TJ. unit had lowest suspended solids concentration in the
effluent samples. These samples were taken just as the effluent left the RBC
units and represent a mixed sample. When the samples were allowed to settle
for one hour, the average suspended solids concentration was approximately the
same for all three units (9 mg S.S./l). The N.U. unit had the lowest suspended
solids production and lowest suspended solids production to NH--N conversion
ratio. By comparing .the solids production of each unit to the relative hydrau-
lic residence time in each unit, the data indicates that as a unit's volume
to surface area increases the net suspended solids production will decrease for
a given ammonia nitrogen application rate. This phenomena is analogous to the
waste sludge production vs. aeration time in the activated sludge treatment
process. The longer detention times allow for solids production to be minimi-
zed.
The average total suspended solids concentrations ranged from 16-31 mg S.S.
/I for these units. These low concentrations raise questions concerning both
the need and effectiveness of a clarifier following RBC nitrification units.
NITRIFICATION: RBC VS ACTIVATED SLUDGE
The bench and pilot studies indicated a high level of nitrifica^
tion can be achieved using rotating biological contactors. Since the RBC units
were treating Marlboro Easterly W.T.P. secondary effluent, a comparison between
the RBC effluent and the activated sludge effluent was made. The Marlboro East-
erly facility accomplishes nitrification of its secondary effluent with a sepa-
rate activated sludge system. The ammonia nitrogen concentrations in the final
effluents from the Marlboro Easterly W.T.P. were recorded and compared with the
RBC effluents.
During the period of the Fall, 1976 pilot study, the activated sludge
system had an average NE^-N effluent concentration of 0.63 mg/1, and the four
stage RBC unit had an average NH^-N concentration of 0.62 mg/1. The Spring,
1977 study period average NH -N effluent concentration for the activated
sludge system was 0.07 mg/1^ and the four stage RBC units ranged from 0.61 mg/1
to 0.96 mg/1. In this study the RBC units were subjected to variable loading
rates. The average JOT^r'N effluent concentrations- for the RBC units exclude
effluent values associated with sharp peaking loads. While the RBC effluent
was higher in ISHo-N concentration than was the activated sludge effluent, a
good quality effluent was achieved with the RBC units. A more constant load-
ing of the RBC units would have achieved an even higher degree of nitrification
1216
-------
o
CM
»
o
CO
CD
CM
CM
CVJ
o
LO
CO
O
CM
o
o
cr>
CO
cr>
CM
ce: 03
CD en
1
O
d. C3
O cn
cc
I— CM
o •—
ZD C5
O>
1217
-------
within the units.
SUMMARY AND-CONCLUSIONS
Loading rates of up to 0.2 Ibs NH3-N/1000ft2-day achieved 94 percent ammo-
nia conversion within the unit. Throughout the study periods a constant re-
lationship was observed vetween ammonia nitrogen converted and ammonia nitrogen
applied. The data indicated that ammonia nitrogen application rate should be a
prime design parameter. Along with ammonia nitrogen loading rate, the volume
to surface ratio of the RBC unit also had an effect on the unit's efficiency.
A larger volume to surface ratio increased ammonia conversion while causing a
decrease in solids production. The increase in unit liquid volume causes an
increased hydraulic retention time for the wastewater flowing through the unit.
The net effect was similar to those of extending the aeration period in the
activated sludge system.
2
The maximum volume to surface ratio evaluated was 0.4 gal/ft . It was
necessary to use submerged pumps within each stage to keep the floe in sus-
pension. The problems of scale-up design, basin geometry, and rotational speed
were not evaluated.
Future studies on plant scale models should be conducted so that RBC nitri-
fication systems can use optimal design information. Over-design of an activa-
ted sludge nitrification system will cause the MLSS to settle poorly. The net
results will be a higher suspended solids concentration in the clarifier efflu-
ent, and problems maintaining a proper food .-microorganism ratio in the aera-
tion tank, Since an RBC nitrification system will not be adversely affected
from either of these two problems, it will be advantageous to optimize the
volume to surface ratio.
The solids data from this study indicated that a clarifier following the
RBC units would be less effective than a filter system. Again, maximizing the
RBC volume to surface ratio should increase the length of filter-run, and de-
crease waste solids handling.
Ammonia nitrogen loading rate and the unit's volume to surface ratio
should be used as the major design parameters. A hydraulic loading rate based
on gpd/ft2 is only an indirect approach for constant volume to surface ratios.
Use of this hydraulic loading technique will not allow the designer to maximize
the unit's efficiencies in terms of ammonia conversion and solids production.
Throughout these studies, 90 percent of the converted ammonia nitrogen
appeared as nitrate nitrogen in the unit's effluent. The nitrifying bacteria
biomass and other associated losses within the RBC unit accounted for 10 per
cent of the converted ammonia nitrogen. Spot checks revealed very low nitrate
nitrogen values in the effluent (less than 0.1 mg N02/l) and the first stage of
each RBC unit was always aerobic.
Since aerobic conditions were maintained within the .units, and staging
helped with ammonia conversion efficiencies, a minimum of two stages is recom-
mended in RBC design. As a majority of the ammonia nitrogen was converted in
the first stage, factors such as loading variation, volume to surface ratio,
1218 :
-------
and BOD- loading on the first stage should be evaluated to determine staging
requirements.
REFERENCES
1.
2.
3.
"Process Design Manual for Nitrogen Control,"
Protection Agency - Technology Transfer.
U.S. Environmental
Standard Methods for Examination of Water and Wasteuater,
13th Edition, American Public Health Association, New York, 1971.
O'Shaughnessy, J.C. and Blanc, F.C. "Biological Nitrification and
Denitrification Using Rotating Biological Contactors," Publication
No. 97, Water Resources Research Center, University of Massachusetts,
Amherst, MA, 1978.
1219
-------
-------
PILOT SCALE STUDIES ON THE NITRIFICATION
OF PRIMARY AND SECONDARY EFFLUENTS
USING ROTATING BIOLOGICAL DISCS
AT THE
METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
By
David R. Zenz
Coordinator of Research
Eugene Bogusch
Sanitary Chemist
Max Krup
Sanitary Chemist
T.B.S. Prakasam
Project Manager
Cecil Lue-Hing
Director of Research and Development
Metropolitan Sanitary District Of Greater Chicago
Chicago,'Illinois, U.S.A.
1221
-------
INTRODUCTION
The Metropolitan Sanitary District of Greater Chicago (District) has
been investigating the process of microbial nitrification with different
unit operations as a means of removing NH^-W from its effluents. This
report presents and discusses the results of pilot scale studies involving
the use of rotating biological disc units for achieving nitrification of
primary effluent (single-stage nitrification) and an activated sludge
effluent (two-stage nitrification).
Nature of Process
Microbial nitrification is accomplished by a group of chemoautotrophic
bacteria. The oxidation of NH^-N and W02-W by nitrifying bacteria may be
expressed as:
o
ammonia oxidizers
2H+ +
(1)
and
NO, + | 0,
nitrite oxidizers
10,
(2)
Nitrification of sewage can be maintained in aerobic biological treat-
ment processes provided that the following conditions are satisfied:
1. A low enough net growth rate to maintain an adequate population
of nitrifying bacteria.
2. Sufficient dissolved oxygen to sustain an adequate rate of
nitrification. '.
3. Satisfactory pH to maintain an adequate rate of nitrification.
U. The absence of substances that inhibit nitrification.
As a system treating a primary effluent, the rotating disc system (EDS)
is expected to remove carbonaceous material as well as achieving oxidation
of ammonia. Thus, the EDS is expected to provide an environment suitable :
for the growth and maintenance of organic matter oxidizers and autotrophic
nitrifying bacteria.
In a system treating secondary effluent, the RDS system is expected to
remove mainly NH^-N, since the preceeding secondary process will remove
considerable amounts of BOD and SS. The RDS system, in this case, would be
expected to contain mainly nitrifying bacteria.
The net growth rate of the nitrifying population of an RDS would depend
on the NH^-N loading to the system and the degree of attachment achieved- by
the biological growth. Generally, fixed growth systems have lower net growth
rates than suspended growth systems. Thus, considerably higher solids
retention times are possible in fixed growth systems than in suspended
growth systems at comparable hydraulic retention times.
1222
-------
Aeration is provided by rotation of the partially--submerged discs.
Oxygen is absorbed by the thin film of liquid in accordance with Henry's law.
LITERATURE REVIEW
Although the RDS process has been used quite extensively in Europe for
more than two decades mainly in small communities for secondary treatment,
its developmental work in the U.S.A. began only in 19&5. A critical review
of this processs and its application to the treatment of municipal and
industrial wastes is presented by Antonie (l).
In addition to the utilization of this process for secondary treatment,
design criteria also have been developed and reported for nitrification of
wastewater based on the data obtained in pilot-scale experiments conducted
at various locations (1,2). These criteria have been delineated for both
carbon and ammonia oxidation in combined as well as separate wastewater
treatment stages. Two important design parameters are the required surface
area of the rotating discs and hydraulic retention time to achieve a certain
percent of NH,-IT removal from wastewaters. Design curves have been developed
relating these parameters to a desired percent removal of NH.-N contained in
sewages of different BOD values. Since the temperature of wastewater
influences the growth rate of nitrifying bacteria, temperature correction
factors have also been developed to scale down the hydraulic loading rates
to ensure nitrification during winter operation (1,2).
The RDS process has been reported to successfully achieve nitrification
of a sludge lagoon supernatant liquor, which contained a high concentration
of NH.-IT (3). In this study, a pilot-scale rotating biological disc unit
(132 liters) was able to oxidize 99-6$ of the 918 mg NH.-N/L contained in
the wastewater at a loading rate of 0.2 kg NHi-N/m of disc surface area/d
at 10 C. The ammonia removal averaged 99-8% at 20 C at a loading rate of
0.7 kg NH.-N/m /d at 20 C indicating that at higher temperatures, higher
nitrification rates are possible. As a follow-up to this study, successful
denitrification of the nitrified sludge lagoon supernatant liquor has also
been reported utilizing a submerged rotating biological disc unit (U).
A communication received from Autotrol in February, 1979, one of the
companies manufacturing rotating disc systems, indicates that it has provided
equipment for U5 installations which are designed for single-stage nitrifi-
cation and for another 11 which are designed as separate-stage nitrification.
These are among a total number of 278 RDS installations which are either
under operation, construction, or design (5). The actual number of RDS
systems that are operational in the U.S.A. will undoubtedly be more than
these numbers, since there are other manufacturers providing these systems.
MATERIALS AID METHODS
Experimental Apparatus
The rotating disc pilot unit was manufactured by the Autotrol Corporation.
Figure 1 is a schematic of the pilot unit. The 35 gallon (132 l) semi-
circular tank was approximately 2 feet (0.6l m) in diameter and 5 feet (1.53 m)
in length. The tank was divided into four compartments of equal volume.
Lightweight 18-5/8 inch (0.^7 m) diameter polyethylene discs were mounted on
1223
-------
UJ
o
<
o
X
o
ce
Ul
UJ
cc
CD
ce
H-
co
Q
Q_
CO
o
cc
UJ
X —
cc
h-
UJ
UJ
o
UJ
X
o
CO
1224
-------
a horizontal shaft whichptransversed the length of the tank. The discs
provided 250 ft (23-2 m ) of surface area for microbial growth.
Approximately hO% of the disc area was submerged in the tank (3).? The disc
surfaceparea to liquid volume ratio in the pilot unit was 53-^ ft /ft
(53.U m /m ). A 1/U HP electric motor was used to rotate the discs at 7 rpm.
A Moyno pump was utilized to control the influent flow to the RDS and, hence,
the hydraulic retention time.
For the studies of primary effluent nitrification, a clarifier was used
during the latter stages of the testing program (Tests 13 and lU) to
determine sludge production. The clarifier was pyramidal in shape. It's ?
size was 6 feet (1.83 m) in height with a surface area of 2.25 ft (0.21 m ).
The volume of the clarifier was 30 gallons (ilk l).
•Chemical Analysis
In the studies of single-stage nitrification using the RDS process,
influent and effluent samples were analyzed for pH, suspended solids, volatile
suspended solids, Kjeldahl nitrogen, ammonia, nitrite, nitrate, chemical
oxygen demand and biochemical oxygen demand.
In the studies of two-stage nitrification using the RDS process, samples
of influent and effluent were analyzed for total Kjeldahl nitrogen, ammonia
nitrogen, nitrite, and nitrate.
The chemical analyses used were according to Standard Methods (6)
except that modifications were made for the analysis of COD and the nitrogen
series (TKN, NHi-N, NO -IT, and HO -IT). These modifications were necessary
to adopt these analyses to the automated Technicon procedures used at
District laboratories (7)-
SOURCES OF SEWAGE
Nitrification of Primary Effluent
In the studies of nitrification of primary effluent, the RDS pilot plant
was located at the West-Southwest Plant of the District. As a source of
sewage, the RDS pilot plant received primary effluent from primary clarifiers
of the Southwest Plant. Characteristics of this primary effluent for the
period of July, 197^- to June, 1975 are contained in Table 1. As can be seen,
the effluent has sufficient alkalinity to satisfy the needs of the auto-
trophic nitrifying bacteria in oxidizing the WH, -IT contained in the primary
effluent and to prevent any pH changes which will occur because of hydrogen
ions produced in the oxidation of ammonia to nitrate. Of course, if a
significant drop in pH occurs, it would be unfavorable to the nitrification
process. The other constituents present are well below levels considered
toxic to carbon and nitrogen oxidizing organisms.
Nitrification of Secondary Effluent
The activated sludge effluent from Battery C of the Calumet Sewage
Treatment Plant of the District was used during this phase of the investi-
gation. This conventional activated sludge plant has an influent and
effluent with the characteristics shown in Table 2 (average values for the
1225
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
TABLE 1
CHARACTERISTIC COMPOSITION OF SOUTHWEST TREATMENT PLANT
RAW SEWAGE AND PRIMARY EFFLUENT*
BOD
COD
NH3-N
PH
SS
TS
Cd
Cr
Cu
Pb
Zn
Org-N
Hg, yg/1
Alkalinity
Raw Sewage
131
367
12.68
7.3
207
832
0.03
0.39
0.09
0.12
0.67
11.11
0.49
225
. Primary Effluent
88
277
12.79
7,5
186
739
0.02
0.19
0.07
0.08
0.37
7.92
0.24
225
All values are in mg/1 except pH and Hg.
* Values were averages for the period July 1974 - June 1975,
1226
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
TABLE 2
CHARACTERISTIC COMPOSITION OF CALUMET TREATMENT
PLANT RAW SEWAGE AND ACTIVATED SLUDGE EFFLUENT.**
Parameter
Suspended Solids
BOD
NH^N
Org-N
Total Solids
COD
pH
Alk.ali.nity
Phenols
Hexane Solubles
Cyanide
Cadmium
chromium
Copper
Lead
Mercury
Zinc
Raw Sewage
272 mg/1
204 mg/1
16.7 mg/1
122 mg/1
1034 mg/1
357 mg/1
7.3
NA*
206 ppb
53 ppb
NA
0.03 mg/1
0.2 mg/1
0.5 mg/1
0.2 mg/1
0,3 ppb
0 . 6 mg/1
Activated Sludge
Effluent
27 mg/1
25 mg/1
11.2 mg/1
3.9 mg/1
783 mg/1
104 mg/1
7.4 '
240 mg/1
6 ppb
14 ppb
0.10 mg/1
<0.01 mg/1
<0.01 mg/1
0.13 mg/1
0.01 mg/1
0.13 ppb
0.02 mg/1
* NA = Not Available
.** Values, were the averages for the year 1975,
1227
-------
year 1975). The alkalinity is sufficient to meet the requirements of the
nitrifying "bacteria and to prevent pH drop due to the nitrification process.;
No constituent is present at levels which could interfere with nitrification.
RESULTS
Primary Effluent Nitrification
During the course of this study, the effect of hydraulic retention time;
on the degree of nitrification achievable was primarily studied. The
hydraulic retention times were able to "be manipulated by adjusting the flow
rate of the influent to the unit. However, no efforts were made to vary the
temperature of the influent. The influent temperatures given in Table 3
were the average temperatures recorded during the indicated period.
A summary of the data obtained at various hydraulic retention times is
given in Table 3 with respect to the percent ammonia removals achieved and
the concentration of effluent NH, -IT observed.
It is obvious from Table 3 that there was considerable variation in
effluent ammonia levels for Tests 6 to 12. The RDS system exhibited wide :
fluctuations in performance for sewage temperatures.of 15 C or less and
hydraulic retention times of 2.00 hours or less. Even at a hydraulic
retention time of U.O hours and a sewage temperature of-27 C (Test 3),
effluent ammonia ranged up to 5-^ mg/1, although the average effluent
ammonia was 2.3 mg/1.
Effect of HRT
Hydraulic retention time, initially maintained at 5-0 hours at the
beginning of the study, was progressively decreased to 1 hours by Test 8. .
At hydraulic retention times of 2.5 hours and greater, Figure 2, ammonia
removals were greater than 80% and effluent ammonia concentrations ranged
from 0.6 to 2.5 mg/1. However, ammonia removals fell when the hydraulic
retention time was lowered to 1.25 hours. The ammonia removal at this
hydraulic retention time averaged in the range of 52 to 78%, and the effluent
MH.-N concentration was in the range of 2.8 to 6.9 mg/1. When the HRT was :
lowered further to 1 hour (Test 8), ammonia removals were drastically
reduced to 16$ resulting in an effluent NH,-N concentration of 13.1 mg/1.
It should be noted that the temperature was 9 C during Test 8, which was
the lowest during the study. The highest ammonia loading rate of O.UU also
occurred during this test. At a later stage in the study when temperatures
were within the range of 15 to 23 C, ammonia removal was considerably
higher (Tests 10 and 12) than at 9°C (Test 8) and was about 50%. However, |
at a temperature of 19 C and a HRT of 1 hour (Test 11), ammonia removal was
only 18$. In all the tests conducted at a HRT of 1 hour, only partial
nitrification was achieved, and the effluent NH,-N concentration was in the
range of 5 to 13.1 mg/1.
In Test 9» when the hydraulic retention was increased to 2 hours from
the 1 hour employed in Test 8,. NH.-N removals increased from 1.6% to 72%.
Test 9 lasted for two months at an average temperature of 13°C. The
effluent NH.-N concentration averaged 3.0 mg/1.
1228
-------
o
o
EC
U
K
E-<
B
Cfi
b
0
O
M
PS
CO
H W
O
X <
PS H
<
M
z
to
<
H
O
0
^
U
S
1
2
BD
H t-3
fc.
*C W
S OS
PS S
eCK '
H ft
B
O O *
az £
z a
°gO
w^3
PS 2 O
D a fo
HC/J •*
Bj r^
S CJ o*
HCO iH
^ M
Q -
Z J S
f£ (^ O
U >0
BH
CSC3
KO
fa O
OH
EH
00
HZ
CuH
fa fri
Hg
O
'w ^
O l?>
R
H ."
5 f
^ n^*
Z
Z .H
a" o
K e
01
K
Z
1 --J
O O1
Z
1 fH
WV.
Z g'
Z
1 .H
g1"!
rH
W tn
E^ S
(1)
H
S
W
cL
e o
g 0
\r-a
Z 4J
S3 0 Q
Z o
o
EH
C£ M
O
rH COO
rH rH r-1 rH CM
11 II II I 1 11
*
(M \
o r-
C) rH
1 1
o in
CO rH
yj
rH
^r co
'T ^0
rH CM
(M^-
O O
VC rH
in ro
r- r-
CM rH
CM CM
C C
o o
3 3
in m
M rj
cn
0
o
rH
1
in •_•>
CO —I
i-> ,°M
x,\
CO
W O
/H rH
1 1
-TO
CM O
CM
in iH
us r*-
CM CA
^r «n
0 0
0 0
o n
n in
rH CM
CM rH
C G
0 O
3 3
IM
«-H
en CM
p- -^
rH n
0 0
3 3
IH m
M W
cn
i-H
CN
O
O
rH
1
in in
(N CM
CM CM
\X.
rH
rH
T CM
CO rH
rH rH
1 1
in CM
CO rH
^
in
in cn
T ro
o m
co ro
o o
CO O
o m
rH
CO rH
VO CM
rH fO
C G
3 3
»w m
H W
CO
CN
cn
o
o
i-H
m in
ro cc
\\
CM
rH
S
(0
CO
t
i
*4
f3
&
.G
-p
UH
O
U
0)
3
«-H
(3 •
>+J
to
0 Q)
JC +J
4J
CD
rH J5
rH 4J
^ 0
QJ G
•H rt
T3 JH
C 3
•H *O
W .C
•^
^G "*H
4)
0 W
O (Tj
rH i>
§S
«
1229
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
HYDRAULIC RETENTION TIME VERSUS AMMONIA REMOVAL
PRIMARY EFFLUENT NITRIFICATION
Figure 2
UJ
o
oc
o_
z
UJ
(C
i
*
X
100 r-
90
80
70
60
50
40
30
20
10
0(18° C)
0(22° C)
0(I5°C)
0(13° C)
(23°C)0
0(15° C)
0(I8°C)
0(23°C)
0(20° C)
0(25°C)
(19° C)
AVERAGE TEST TEMPERATURES GIVEN IN PARENTHESES
, i i i I
HYDRAULIC RETENTION TIME, HOURS
1230
-------
The effect of BTH,-IT loading rate on KH,-W removal is presented in
Figure 3. This shows that NH.-IiF removals werep72$ or greater at an NH, -U
loading rate of less than 0.2 lb EFH, -EF/1000 ft /day. At these loading
rates, the effluent NH,-N concentration was in the range of 0.6 to 3.0 mg/1.
Loading rates higher than 0.2 caused only partial nitrification, and the
general trend as expected was an increase in the effluent EH, -ET concentration
as the loading rate was increased as a result of lowering the hydraulic
retention time.
Solids Production and BOD and Suspended Solids Removal in the
Nitrifying RDS Unit
In Tests 1 to 12, the feasibility of nitrifying domestic wastewater
was studied with no effort being made to clarify the wastewater after it
was nitrified. In the following two tests, viz 13 and ik, wastewater was
nitrified at temperatures of 22 C and 18 C, respectively, and at a
hydraulic retention time of 1.67 hours. In Test lH, small plastic paddles
were installed on the perimeter of the rotating discs in each compartment in
order to keep the solids in suspension and minimize their accumulation at
the bottom of each compartment; where as in Test 13, no such paddles were
provided. The results of these two tests are presented in Tables h and 5_.
The results described in Table k reflect the overall performance of the RDS
in terms of BOD, COD, total solids, suspended solidsj NH,-N removed through
the system with and without paddles on the rotating disc unit. With the
exception of COD, the removals of suspended solids, BOD, TKH, and NH.-U
were good; the unit with the paddles exhibiting slightly higher removals than
the one without them. In Test 1^, i.e. the test conducted with the unit
containing the paddles, the percent NIL-IT, BOD, and suspended solids removals
were 87, 83, and 8U, respectively, in comparison to the percent removals of
80, 83, and 77 for the same parameters in Test 13.
The solids production data presented in Table 5 is expressed as
suspended solids produced per a) unit of influent suspended solids, b) unit
BOD removed, and c) unit COD removed. In the computations, the solids
produced is expressed as the sum of, the sludge withdrawn from the clarifier
and the solids carried over in the effluent.
All three determinations of sludge production show that significantly
higher sludge production resulted with paddle operation than without it.
The apparently lower sludge production noticed when the RDS unit was
operated without paddles may partly be due to the loss of solids accumulated
in the unit via biological degradation. Anaerobic decomposition of the
accumulated solids was apparent as evidenced by black sludge deposits in the
compartments of the RDS, When paddles were installed, such black sludge
deposits did not form, and a much grayer sludge resulted than the one
encountered in the unit without paddles.
The sludge production data (Table 5) in terms of Ibs of suspended solids
produced per Ib of influent suspended solids, is directly comparable to the
data reported by Obayashi et_ al_ (8) for the District's Southwest Plant.
They found this to be the best sludge production parameter (correlation
1231
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
AMMONIA LOADING VERSUS AMMONIA REMOVAL
PRIMARY EFFLUENT NITRIFICATION
Figure 3
UJ
o
cc
100 i-
90
80
70
60
50
UJ
o:
z 40
i
x
z
30
20
10
J20°C)
0 0(I8°C)
0(25°C)
0(23°C)
0(I8°C)
0(I5°C)
0(22° C)
0(I2°C)
©(I3°C)
(23«C)0
0(I5°C)
0(19° C)
©(9° C)
AVERAGE TEST TEMPERATURES GIVEN IN PARENTHESES
, , . i I
O.I
0.2
0.3
0.4
0.5
NH4-N LOADING (LBS NH4-N/IOOO FTVDAY)
1232
-------
THE METROPOLITAN SANITARY DISTRICT OP GREATER CHICAGO
TABLE 4
PRIMARY EFFLUENT NITRIFICATION
R.DS PERFORMANCE, TESTS 13 AND 14
JUNE 30 - OCTOBER 28, 1976*
Test 13(no paddles)
June 30 - August 10
Test 14(w/paddles)
Sept. 3 - Oct. 28
HRT, hours
Temperature, °C
NH^-N Loadings
Influent TKN, mg/1
Effluent TKN, mg/1
TKN reduction , %
Influent NH4~N, mg/1
Effluent NH^-N, mg/1**
NH4~N reduction, %
Influent NO--N, mg/1
Effluent NOif-N, ,mg/l
Influent NO-.-N, mg/1
Effluent NO^-N, mg/1
Influent COD, mg/1
Effluent COD, mg/1
COD reduction, %
Influent TSS, mg/1
Effluent TSS, mg/1
TSS reduction, %
Influent BOD, mg/1
Effluent BOD, mg/1
BOD reduction, %
Influent Alkalinity, mg/1
Effluent Alkalinity, mg/1
Influent pH
Effluent pH
1.67
22
0,19
16.9
4.2
75%
11.2
2.2
80%
0.25
0.67
0.2
9.0
277
146
47%
94
22
77%
83
14
83%
188
132
7.4
7.3
1.67
18
0.16
15.8
3.6
77%
9.5
1.2
87%
0.37
0.61
0.4
5.6
263
121
54%
96
15
84%
90
15
83%
194
145
7.4
7.4
* Values are averages for the test.
** Range for Test 13: 0.5-10 mg/1, Test 14: 0.0-5.6 mg/1.
1233
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
TABLE 5
RDS SLUDGE PRODUCTION
JUNE 30 - OCTOBER 28, 1976
Sludge' Production
Test 13(no paddles)
June 20 - August 10
Test 14(w/paddles
Sept. 3 - Oct. 28
Suspended Solids Produced
Influent Suspended Solids
Suspended Solids Produced
BOD Removed
Suspended Sol'ids Produced
COD Removed
0.97
1.38
0.70
1.31
1.58
0.84
1234
-------
coefficient; r , of 0.97) in a nitrification study conducted over a four-
month period at the Southwest Treatment Plant utilizing an aeration battery
receiving a flow of 220 mgd. They reported the sludge production to be in
the range of 0.68 to 0.76 Ibs per pound of influent suspended solids with an
average of 0.72. Clearly, the sludge production from the RDS system is
considerably higher than for the single-stage activated sludge system .tested
by Obayashi _et_ al_. It appears that, in this case, the activated sludge
system produced less sludge than-a corresponding RDS process. One possible
reason for the lower sludge production in the activated sludge process is
that the battery used in the above study is a newer battery and perhaps is
more efficient in terms of oxidizing further the sludge produced than the
RDS unit. It is also possible that the activated sludge had a lower F:M
ratio than the RDS unit during the respective periods used for the comparison
of sludge production.
Secondary Effluent Nitrification
As in the previously discussed study, the only controlled operational
parameter was the hydraulic retention time. Other significant variables
such as influent ammonia nitrogen concentrations and sewage temperatures
were, of course, uncontrolled. The RDS was, therefore, operated at specific
HRTs with the objective of determining under what conditions satisfactory
.ammonia removal could be achieved.
The results of this study were obtained under the following conditions:
1. The activated sludge effluent had ammonia nitrogen concentrations
in the range of 8 to 20 mg/1 during the period of this study,
and averaged l^J- mg/1.
2. Sewage temperatures ranged from 9 to 25 C.
3. The effect of hydraulic retention time was tested within the
range of 0.6 to 5 hours.
The results of this study are summarized in Tables 6 .and 7.- Nine
different modes of operation were monitored. The operational modes were
established at various combinations of hydraulic retention time and
temperature.
Table 6 presents the following operational data; hydraulic retention
time, ammonia loading, sewage temperature, length of test periods, and
effluent ammonia nitrogen concentration. Table 7 shows the following
analytical data; influent and effluent total Kjeldahl nitrogen, ammonia
nitrogen, nitrite nitrogen., and nitrate nitrogen. Information pertaining
to these parameters is given as the average and range for each tested
operational mode. Average ammonia removal in percent is also given in
Table 7.
The operation of the RDS was begun at a five-hour HRT. At this time,
sewage temperatures averaged 25 C. Within 5 days, the oxidation of ammonia
to nitrite was established. However, an additional 10 days were required
to achieve complete oxidation of nitrite to nitrate.
1235
-------
5 -*fU
MS1*-1
1236
-------
CHICAGO
K
W
s
8
g
s
FH
WITARY DISTKIC
TABLE 7
to
METROPOLI
I
Dw
o
2
O
H
e* en
ITRIFICA
MT BY RD
PORMATIONS IN H
3 SLUDGE EFFLUE
en w
II
NITROGEN
ACT
Average
NII4-N
Reraova 1 , %
*
E
2
1 01
r tn
C a
0)
5
01
C
r-i (S
\«
I
T
s
rH 1C
ST
2
1 O
•5 o>
<
O
en
c
a
-< a
\
&
e
, 2 0
w a>
E* a
o
<
c
r-l
C.
fi
o
w
4J
a
£
<*
y
ON
CM 00
CO CO
0 0
r-t
I I
O -V
«3- ro
o o
\o •*
O -ef
O O
in •
>H H
1 1
0 O
CN in
o o
t- p-
o vo
o -a-
Cn r-}
rH
1 I
CN M
C 0
i-H
CN 03
in -^r
i
n in
CM
1 |
G\ n
\D (N
CO
i~-t
i t
ro co
o r-
o tn
rH fN
rH CN
cn to
co cn
0 0
t I
-3-0
00
00
SfrH
O O
in wJ
CT\ rH
rH rH
I I
CO rH
m o '
n CN
rH
O^ O%
CN r-t
t t
CN r*-
vo ^f
co ro
7 Influent
Efflueni
rH
ro
CO
o -r-
CO *
rH CN
1 1
O
O CN
0 •*
CN r-
0 r?
CN O
T ro
O rH
1 1
i-H rH
O O
0 0
r-t «=3-
O O
O 0
o o
CN rH
t 1
O CN
r-. o
vo ro
rH
rH CO
CN rH
I t
r- in
rH CN
cn in
C c
o -i
3 *3
r~t >H
4J ^4
f ; '4^
H U
CO
ro
o
r-
VC VO
r-i
I I
O rH
O -
O H
•-H'rH
O ro
CO O
CO O
O CN
1 1
O rH
o o
o o
rH f-
O Q
o cn
ro r-
CN
1 [
\o r>-
rH O
rH -
CN rH
rH
CN 0
CN r-l
1 1
I-- rH
rH
co r-
rH r**
9 Influent
Efflueni
1237
-------
Tables 6 and 7_ indicate that HE,-N removals remained relatively
constant (88.5$ to 9k.h%) at HRT values of 5, 3, 2, and 1.3 hours (Tests
1, 2, 3j and 7). The average temperatures at these HRT values were 25, 21,
21, and 20 C, respectively. The average effluent NE,-H concentration during
these tests was less than 1.3 mg/1 and ranged between 0.1 and 11.6 mg/1.
When the HRT was decreased to 1 hour in Test U, the percent ammonia
removal decreased to 76.8$. The average effluent ammonia nitrogen concen-
tration over 5^ days of operation in Test k was 3.5 mg/1 and ranged
between 0.8 and 9-3 ag/1. This decreased ammonia removal occurred not only
because of the reduction of the hydraulic retention time to 1 hour "but also
because of the low average sewage temperature of 12°C prevailing during this
experimental period.
When the HRT was subsequently increased to 2 hours in Test 5, effluent
NHv-N concentrations averaged 1.0 mg/1 and ranged between 0.6 and 5-6 mg/1.
This effluent NH.-H concentration resulted at an average sewage temperature
of 13 C. Although the average 1.9 mg/1 effluent HE,-H concentration
represents a substantial decrease over the 3-5 mg/1 HH.-H concentration of
Test U, the percent ammonia removal increased only slightly from 76.8% to
79-3?». This resulted from sharply depressed influent ammonia levels.
Test 6 represents continued operation at the two-hour HRTbut at
slightly higher sewage temperatures. The average temperature of Test 6 was
17°C as compared to 13°C in Test 5. Effluent NH,-H averaged 2.2 mg/1 and
ranged from 0.5 to 6.8 mg/1, not significantly different from that of Test 5-
Percentage ammonia removal increased slightly because of higher influent
ammonia concentrations than in Test 5.
At a 0.8-hour HRT and an average temperature of 16 C, effluent ammonia
nitrogen averaged 2.3 mg/1 and ranged from 0.2 to 10.0 mg/1. Percentage
ammonia removal averaged 83.!$. This performance of the unit at 0.8 hours
was comparable to that obtained in Test 6 which was conducted at a, HRT of
2 hours. As a result of lowering the HRT to 0.8 hours in Test 8, the UH,;-H
loading was approximately three times that of the one employed in Test 6.
At a 0.6-hour HRT and an average temperature of 21 C, effluent HH.-H
averaged 5-1 mg/1 and ranged from 0.7 to 7.^ mg/1. Ammonia removal averaged
70.3$. This mode of operation yielded the highest effluent HE,-H concentra-
tion and the lowest percentage ammonia removal. Since sewage temperatures
were high, averaging 21 C, the decline in effluent quality must be
attributed to the short hydraulic retention time and the high ammonia
loading.
In spite of the condition of a constant hydraulic flow employed in
various test runs, the RDS often exhibited inconsistent performance. This
is shown by the effluent ammonia nitrogen data (average and range) given!in
Table 6. With the exception of Tests 1, 2, and 3, the RDS performance was
frequently diminished as indicated by the high effluent HH.-H concentrations,
which are several times larger than the average effluent HE.-H concentration.
At hydraulic retention times of 2.0 hours or less and sewage temperatures of
12 to 20 C (Tests U to 8); effluent ammonia levels ranged up to 11.6 mg/1,
although average ammonia levels were less than 3.5 mg/1. In Test 6, for:a
hydraulic retention time of 2 hours and a sewage temperature of 17 Cj the
1238
-------
r
average effluent ammonia levels vere 2.2 mg/1, but daily values ranged
as high as 6.8 mg/1.
From a design criterion standpoint,pthe relationship between the
substrate loading rate (ib NIL -11/1000 ft /day) and ammonia removal is
important. Such a relationship is presented in Figure U from the data
obtained in the pilot scale study. Figure ^ shows that as the ammonia
loading is increased, the ammonia removal decreases. This is reasonable,
since the finite surface area, of the EDS will support a limited amount of
biological growth.
Also apparent in Figure U is a trend for reduced ammonia removal at
low sewage temperatures. This is a result of decreased growth rates of the
nitrifying bacteria at low temperatures.
Figure 5 shows the relationship between hydraulic retention time and
ammonia removal. As the HET is reduced, ammonia removal is also reduced.
Reduction of the KRT stresses the nitrification process in two ways. Since
the HRT is equivalent to the nominal treatment time in the reactor, less
time will be available for BTH. -IT removal as the HRT is reduced. In addition,
reduction of the HET increases the sewage loading to the EDS. Since the
influent ammonia concentration varies over a relatively narrow range, the
ammonia loading will generally increase in direct proportion to the sewage
flow. Thus, a reduction of the HRT will decrease treatment time and
increase the ammonia loading.
DISCUSSION
In general, the results obtained with respect to the hydraulic
retention times and hydraulic loading rates required for achieving nitrifi-
cation in this study of nitrification of both primary effluent and secondary
effluent agree with those summarized and reported by Antonie (l). From the
results of various pilot plant studies, he found that the nitrification of
primary effluent was able to be achieved (90% NH,-N removal) with a
hydraulic loading rate of less than 1.5 to about 3 gpd/ft for primary
effluents containing WH,-N concentration within the range of Uo to 15 mg/1,
respectively. These results were obtained above a temperature of 13 C, and
the BOD of the incoming wastewaters used in the different studies was within
the range of 250 to 80 mg/1. The results reported for the nitrification of
secondary effluents at temperatures greater than 13 C and effluent BOD of
less than 20 mg/1, indicate that about 90% of the NH,-W was removed within a
hydraulic loading rate of'less than 2,0 to 3.5 gpd/ft^, when these effluents
averaged an initial 1H,-N concentration of ho to 10 mg/1. The higher
hydraulic loading rates were possible at lower inlet EfH.-N concentration.
In this study, the HRT for achieving an average NH. -N removal of 85% or
greater from primary effluent was found to be in the .range of 1.7 to 5 hours
(-0,5 to 2 gpd/ft ) for sewage temperatures of 15 to 25 C. In general,
decreased HET and temperature decreased ammonia removal. For- example, at
an HRT of 1 hour and sewage, temperatures of 9 to 23 C, the average percent
NHi-N removal was within the range of .16 to 51%, the higher removal occurring
in most instances at higher temperatures... The EDS effluent NH, averaged 5 to
13.1 mg/1 for individual, test runs, while a wide fluctuation was observed on
a daily basis within the range of 1.3 to 25-6 mg/1. Of all the runs, only
1239
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
EFFECT OF AMMONIA LOADING ON AMMONIA REMOVAL
FOR SECOND STAGE ROTATING DISC SYSTEM
SECONDARY EFFLUENT NITRIFICATION
Figure 4
100
LU
O
cc
Ul
Ou
1
UJ
cc.
90
80
70
60
50
40
0(25°C)
0(2I°C)
°(20°C)
© (20°C)
O (18° C)
0(I3°C)
0(I6°C)
©(I2°C)
O (2I°C)
AVERAGE TEST TEMPERATURES GIVEN IN PARENTHESES
I I I I
0.2
0.4
0.6
0.8
1.0
NH4-N LOADING (LBS NH4-N/IOOO FT2/DAY)
1240
-------
THE METROPOLITAN SANITARY DISTRICT OF GREATER CHICAGO
EFFECT OF HYDRAULIC RETENTION TIME ON AMMONIA REMOVAL
FOR SECOND STAGE ROTATING DISC SYSTEM
SECONDARY EFFLUENT NITRIFICATION
Figure 5
100 r-
90
ui
o
o:
80
70
UJ
cc
60
50
40
0(20°C)
O (20° C)
0(I6°C) 0(I8°C)
O
0(I2°C)
0(2I°C)
0(25°C)
AVERAGE TEST TEMPERATURES GIVEN IN PARENTHESES
i i I I
HYDRAULIC RETENTION TIME, HOURS
1241
-------
runs 2 and k yielded effluents with the least fluctuations in effluent
Fluctuations in effluent NH. -N levels occurred even under the above-
noted operating conditions which yielded 85$ or greater NH, -N removals
from the primary effluent. Effluent NH.-N concentration averaged 2.5 mg/1
or less , "but daily concentration ranged from 0.1 to 10 mg/1 .
The degree of nitrification achieved for secondary effluent lay the
pilot scale EDS unit was comparable to that of primary effluent , in that
higher HRTs and temperatures yielded higher NE. -N removal than lower HRTs
and temperatures. The hydraulic loading rates were within the range of 0.7
to 2 gpd/ft for achieving nitrification within an average temperature range
of 13 to 25 C for the various runs. However, the wide fluctuations of RDS ;
effluent NIL -N concentration noted on a daily basis in the case of primary
effluent nitrification were also observed in the secondary effluent
nitrification. For example, at a hydraulic retention time of 2 hours and
a sevage temperature of IT C, effluent ammonia averaged 2.2 mg/1. However,
the levels of effluent ammonia was as high as 6.8 mg/1.
It should be noted that these fluctuations in effluent HE, -N levels
occurred in these RDS pilot-scale studies , despite the fact that HRT i
remained constant within each of the test runs. Therefore, the fluctuations
must be caused by daily changes in influent NH. -N concentration and/or by
the lack of an adequate mass of nitrifiers on the disc surfaces to oxidize
the incoming NH.-N loading within the HRT of the unit. It- is, therefore,
reasonable to conclude that its performance will be even more erratic when
it is subjected to the diurnal fluctuations occurring in sewage flow at a
treatment plant.
A recently reported study by Filion, Murphy, and Stephenson (9) -alluded
to the greater sensitivity of the RDS unit to variation in influent total
Kjeldahl nitrogen loading due to fluctuation in flow and HRT. They
indicated such performance was not as good as activated sludge nitrification
systems. In previously reported nitrification studies performed by the
District on activated sludge systems utilizing combined and separate-stage
nitrification, the effluent NH.-N concentration did not fluctuate as widely
and was consistently less than 2.5 mg/1 once nitrification was initiated
and sustained ( 10,11).
The sludge production in the RDS unit used for the nitrification of
primary effluent was considerably higher than the sludge production observed
in a. plant-scale nitrification battery of the West-Southwest Treatment Plant
(1.3 lb vs 0.72 Ib of suspended solids/lb of influent suspended solids).
This higher production of sludge solids by the RDS unit is obviously
undesirable .
In view of the foregoing, although the RDS pilot-scale studies at the
W-SW and Calumet Sewage Treatment Plants have demonstrated the feasibility
of achieving nitrification of the primary and secondary effluents of these '
plants, respectively, the installation of the RDS units for full-scale
nitrification was not recommended. This is due to the wide fluctuations of
effluent NH. -N concentrations noted on a daily basis and the observed higher
sludge production rates with the RDS unit contrary to the stable and lower
1242
-------
level of effluent NH.-N levels and sludge production in the combined and
separate-stage activated sludge nitrification systems tested by the District.
CONCLUSIONS
Primary Effluent Nitrification
1. Ammonia removal efficiency of the RDS system correlated directly
with hydraulic retention time and temperature, and inversely with
ammonia loading rate. .
2. At hydraulic retention times of 1.25 to 2 hours and sewage
temperatures of 12 to 15 C, the RDS system appeared stressed,
producing ammonial removals of 52 to JQ% and effluent ammonia
levels ranging from 0.1 to l6.U mg/1. Ammonia loading for these
conditions ranged from 0.15 to 0.32 Ibs NH.-N per 1000 ft per day.
3. The study did not yield data on low sewage temperature (10 to 13 C)
operation of the RDS consistent with good nitrification (>90%
removal). Best low temperature performance was achieved at a
temperature of 12 C, a HRT of 1.25 hour§, an ammonia loading of
approximately 0.3 Ibs NH,-N per 1000 ft - day, and an average
ammonia removal of lQ>%. However, performance was erratic with
effluent ammonia concentrations ranging from 0.1 to 10.2 mg/1.
U. Since greater than 90$ ammonia removal was only achieved at
hydraulic detention times of U.O hours or greater, and -ammonia
loadings of 0.11 Ibs WH.-N/1000 ft - day or less, and only then at
temperatures above 18 C,. it must be assumed that lower sewage
temperature would require at least these levels of HRT and NH,-N
loading.combinations in order.t© produce comparable ammonia removal.
5. Sludge production averaged about 1.50 Ibs of suspended solids
produced per Tb of influent solids.
Secondary Effluent Nitrification
1. Ammonia removal was found to be directly related to sewage
temperature, hydraulic retention time, and inversely to ammonia
2. At a hydraulic retention time of 1 hour and a sewage temperature
of 12 C, the RDS system appeared stressed, producing ammonia
removals averaging 11% and effluent ammonia levels 0.8 to
9.3 mg/1. Ammonia loadings for this condition averaged O.U2 Ibs
NH^-N per 1000 ft - day.
Hydraulic retention period of 2 hours, ammonia loading of
13 lbs/1000 ft - day, and a sewage temperature of 13 C produced
an average effluent of'1.9 mg/1 of STR-N. However, ammonia
removal averaged only 70,3$' and effluent ammonia levels
fluctuated widely (0,6 to 5.6 mg/l).
Ammonia removals exceeding 90$, were achieved at hydraulic Retention
periods of 2 to 5 hours and loadings up to .22 lbs/1000 ft - day
1243
-------
"but only with sewage temperature at 21 C or at>ove. In addition,
only at these temperatures, hydraulic retention times, and ammonia
loadings were effluent ammonia concentrations consistently below
2.5 mg/1.
5. For consistent ammonia removals above 90% at sewage temperatures
of 10 to 13°C, hydraulic retention'times exceeding 2 hours and an
ammonia loading of .22 lb/1000 ft - day or less appear necessary.
Because of the sludge production associated with the use of the RDS
system for primary effluent and the fluctuations in effluent ammonia
concentrations associated with nitrification of both primary and secondary
effluent, the District rejected this system for use in its treatment
facilities.
1244
-------
ACKNOWLEDGMENT
The assistance given by Mr, G. Cargill and his associates in maintaining
and operating the EDS pilot plant at the Calumet Treatment Plant is greatly
appreciated. The authors are also grate'ful for the analytical support
given by the' staff of the Quality Control Division of the Research and
Development Department.
DISCLAIMER
The results presented herein were obtained by the Research and
Development Department of the Metropolitan Sanitary District of Greater
Chicago.' Publication of this report does not'signify that the contents
necessarily reflect, the views and policies of the District, nor. does
mention of trade names -o"r commercial products constitute endorsement or -
recommendation for use.
1245
-------
REFERENCES
1. Antonie, R. L., Fixed Biological Surfaces - Waste-water Treatment,
The Rotating Biological Contactor, CRC Press Inc., Cleveland, Ohio,
RP 200, 1976.
2. USEPA Technology Transfer Process Design Manual for Nitrogen
Control, October 1975.
3. Lue-Hing, C.; Obayashi, A. W. ; Zenz, D. R.; Washington, B. ;
and Sawyer, B. M. ; Biological Nitrification o£ Sludge Supernatant
"by Rotating Discs, JWPCF U8:25-H6, 1976.
U. Prakasam, T. B. S. ; Robinson, W. E.; and Lue-Hing, C.; Nitrogen
Removal from Digested Sludge Supernatant Liquor Using Attached and
Suspended Growth Systems, Proc. 32nd Industrial Waste Conf.
7^5-758, 1977.
5- Personal Communication from Jeffrey A. Cowee, Autotrol Inc.,
Milwaukee, Wisconsin, February 1979-
6. Standard for the Examination of Water and Wastewater, APHA, AWWA,
WPCF, lUth Edition, 1975.
7. The Metropolitan Sanitary District of Greater Chicago, Research
and Development Department. Technicon Automated Analysis Manual,
1972.
8. Obayashi, A. W. ; Washington, B. ; Lue-Hing, C.; Net Sludge Yields
Obtained During Single-Stage Nitrification Studies at Chicago's
West-Southwest Treatment Plant, Proc. 32nd Purdue Industrial
Waste Conf. 759-769, 1977.
9. Filion, M. P.; Murphy, K. L.; and Stephenson, J. P.; Performance of
a Rotating Biological Contactor Under Transient Loading Conditions.
JWPCF 51:1925-1933.
10. Zenz, D. R.; Richardson, G. ; Lue-Hing, C.; and Sawyer, B.;
Determination of the Design Parameters for a 30 MGD Two-Stage
Nitrification Plant. Paper Presented at the National Conference on
Nitrogen Control, (USEPA), Chicago, Illinois 1976 (Metropolitan
Sanitary District of Greater Chicago, Report 76-18, 1976).
11. Prakasam, T. B. S.; Lue-Hing, C.j Bogusch, E.; Zenz, D, R.;
Pilot Scale Studies on Single Stage Nitrification, JWPCF,
51:190U-19l85 1979-
1246
-------
USE OF ROTATING BIOLOGICAL CONTACTORS FOR NITRIFICATION
AT THE CITY OF GUELPH WATER POLLUTION CONTROL PLANT
GUELPH, ONTARIO, CANADA
By
Paul M. Crawford, P. Eng.
Gore & Storrie Limited
Toronto, Ontario, Canada
INTRODUCTION
The City of Guelph is situated on the Speed River about 100 km
west of Toronto. A tributary of the Grand River, which empties into
Lake Erie, the Speed River had a peak flow of 3450 x 106 m3/d, a low
flow of 36.3 x 103 m3/d, and a mean flow of 606 x 106 m3/d in 1978.
Prior to 1979, the Guelph Water Pollution Control Plant consisted
of three separate activated sludge type secondary treatment plants,
including phosphorus removal, with a total capacity of 46,000 rrf'/d.
The plant effluent was consistently of good secondary effluent quality at
approximately 15 mg/1 of BOD5 and suspended solids and less than 1 mg/1
of phosphorus as P. Because of the design loading of the activated
sludge system, however, nitrification, that is the oxidation of ammonia
to nitrate, did not occur consistently in the plant.
In the summer of 1975, intensive water qualities surveys were
undertaken by the Ontario Ministry of the Environment on the Grand River
Basin. The data generated by these surveys was used to establish the
relationship between dissolved oxygen concentration in the streams and
waste loadings from the various water pollution control plants in the
Basin. It was concluded that minimum dissolved oxygen concentrations of
1247
-------
4 mg/1 would not be realized in the Speed River downstream of the Guelph
Water Pollution Control Plant until the nitrogenous oxygen demand of the
plant effluent was significantly reduced.
The Ontario Ministry of the Environment proposed the effluent
requirements, summarized in Table 1, to be the minimum degree of treatment
required in order to meet the dissolved oxygen criteria in the Speed River.
For the Guelph effluent, the BOD20 to BOD5 ratio is about 1.5 to 1. This
means that the ammonia effluent requirement is 2.0 mg/1.
PROCESS SELECTION
When it became apparent to the City of Guelph that nitrification
would become an effluent requirement, studies to determine the organic
loading reduction required in the existing plants to achieve nitrifica-
tion were initiated. While a limited degree of nitrification was
achieved, the single stage process lacked stability. It was therefore
decided to investigate a two-stage system whereby a separate process
would be "added-on" to the existing secondary treatment system.
While there are numerous processes for nitrogen removal, it was
determined that for the Guelph application, biological nitrification
utilizing an attached growth system would be the least cost solution.
Rotating Biological Contactors, or RBC's, were selected as the most
economical system for the following reasons:
1. low energy requirement
2. no need for final settling tanks if followed by effluent ;
filtration
3. small land area requirement
PILOT TESTING
In order to obtain design information, a pilot testing program
utilizing an RBC was initiated. The City of Guelph procured, installed
and operated equipment suitable to carry out the pilot testing as well
as performing sample collection and analysis. A pilot scale RBC was
leased from the Bio-Systems Division of Autotrol Corporation. It was
a 36 disc unit with a total surface area of 23 m2, media diameter of
0.5 m, a shaft length of 1.32 m and a hydraulic capacity of 0.13 m3.
The tank for the unit was baffled so that there were four separate
stages each with an equal number of discs. Influent wastewater for the
unit was the effluent from the existing Plant No. 2.
The result of the pilot testing are displayed graphically
in Figure No. 1. It shows the specific removal rate of ammonia
nitrogen as a function of effluent ammonia concentration-. Most of
the data was taken at an average wastewater temperature of 18°C.
Corresponding removal rates at 16°C and 13 C were then7 determined
using temperature correction factors for nitrification determined by
Autotrol Corporation. The specific removal rate follows the classic
1248
-------
Monod relationship with a zero order rate at high ammonia concentrations
and a first order rate for low ammonia concentrations with the-knee of
the curve at approximately 6 mg/1 ammonia nitrogen.
Three different rotational speeds were used during the
testing period. These corresponded to full-scale peripheral speed,
half full-scale speed and full-scale rotational speed.
For design purposes, only the half full-scale peripheral speed
data was used as the full-scale peripheral speed data was considered
to be too optimistic in comparison with other experience and the full-
scale rotational speed data was not fully developed due to the length
of the testing period.
Diurnal flow conditions existing at the plant were simulated
during the pilot testing. For the ranges studied, it appeared that
the nitrification process was not upset.
From the data produced during the pilot study, it was determined
that at the average wastewater temperature of 16°C, 32 RBC's would be
required to achieve the effluent requirement of 2.0 mg/1 ammonia
nitrogen with an influent concentration of 20 mg/1 ammonia nitrogen
and a wastewater flow of 55,000 m3/d.
EQUIPMENT PROCUREMENT
Having established the design criteria for construction of the
plant to meet the effluent requirement for nitrification, the City of
Guelph advertised public tenders for the "Supply, Delivery, Supervision
of Installation, Placing into Operation and Testing of Rotating Biolo-
gical Contactors". The contract was for the RBC's only and excluded
drives and covers.
Specific attention was given in the specifications for the shaft
bearings. Greased lubricated spherical roller bearings with spring
loaded lip seals and a B-10 life of 200,000. hours were required. Also
specified was the inclusion of a hydraulic load cell in the bearing
holising biomass thickness on the RBC's.
Recognizing the importance of energy consumption, the tender
documents required that the equipment supplier guarantee their equipment's
energy consumption for the purposes of both tender evaluation and per-
formance guarantee. Autotrol Corporation guaranteed an energy input
of 2.05 kw per RBC.
In addition, a process guarantee was established as shown in
Table 2. The guaranteed minimum specific ammonia removal rate was
purposely selected at about 10% less than the zero order rate which
the pilot study results indicated for 13°C wastewater temperature in
order to account for any scale-up factors. The minimum ammonia con-
centration of 6 mg/1 was specified to ensure that the specific removal
rate was in the zero order range and thus not ammonia concentration
1249
-------
limited. The average influent soluble B005 was selected by the equipment
supplier, Autotrol Corporation. The minimum wastewater temperature of
13 C was chosen to correspond to the specific ammonia removal rate. The
dissolved oxygen concentration limit of 2.0 mg/1 was selected as it is
the generally accepted minimum for dissolved oxygen concentration not to
be rate limiting for nitrification. The maximum biomass thickness of
3.0 mm was specified in order to limit the stress on the steel central
shaft to an allovable level.
During testing, it was required that both the energy consumption
and process guarantees be met simultaneously.
Three tenders .were received and after lengthy evaluation, a
contract was awarded to Autotrol Corporation for 32 air-driven RBC's.
The air-drive option was selected for the following reasons:
- simple, infinitely variable speed control system
- considerable capital cost savings over mechanical drive system
- greatly reduced operation and maintenance costs
FACILITY DESIGN
Plant Layout
A plan of the layout of the Nitrification and Filtration Facilities
is shown in Figure No. 2. The secondary effluent from Plants No. 1 and 2
vras diverted to the Plant No. 3 chlorine contact tank which was converted
to a secondary effluent pumping station. All of the secondary effluent
is pumped from there to the RBC tanks where the nitrification process
occurs. The flow then enters the Automatic Backwash Filters for
filtration. Finally, after chlorination and detention in the chlorine
contact tank, the flow is measured by a Parshall flume and discharged to
the Speed River.
Tankage
To allow the maximum practical degree of staging, the thirty-two
RBC's were arranged in four groups of eight in series. The tanks were
designed in accordance with the recommendations of the manufacturer,
having flat bottoms with the inlet and outlet ends filleted to prevent
stagnant areas.
The flow enters the tank through a V-notch weir to provide even
distribution across the width of the tank. Precast concrete flow
deflectors are installed between adjacent RBC's to prevent short
circuiting across the floor of the tank. The direction of rotation of
the RBC's is counter-current to the flow direction. Figure No. 3 shows
a sectional view of the RBC tanks.
1250
-------
Aluminum baffle plates,with a 100 mm high slot at the bottom are
provided between each of the last five adjacent RBC's. The purpose of
these baffles is to provide staging downstream of the RBC at which
ammonia concentration is less than 6 mg/1. The staging, which in
effect creates a completely mixed tank reactor for each RBC, allows
maximization of the effluent ammonia from each RBC. Since ammonia
nitrogen concentration is rate limiting at values less than 6 mg/1
maximizing concentration also maximizes the ammonia removal rate.
Air Drive System
The air drive system consists of three blowers, a discharge air
header with a branch, to each RBC and individual butterfly valves and
diffuser headers for each RBC.
; Three centrifugal blowers, each with a capacity of 1.43 m3/s at
a discharge pressure of 20 kPa are provided. To obtain the maximum air
drive speed of 1.2 rpm for each RBC requires that two blowers be in
operation with a third used as a standby. With only one blower in
operation the maximum rotational speed for all RBC's is about 0.9 rpm.
Based on the pilot study data, the required process efficiency can
likely be obtained in the full-scale installation at a rojtational
speed of 0.8 rpm. i
Each blower is equipped with a motorized butterfly valve for
throttling. Placing this valve on the suction side of the blower allows.
reduced energy consumption when throttling. These valves are operated
by a controller which maintains any selected pressure in the discharge
header. This feature allows the speed of any individual RBC to be
adjusted without affecting the speeds of the other RBC's since the
pressure upstream of the control butterfly valve at each RBC is constant.
The diffuser headers are fabricated in thin-wall stainless steel
pipe and utilize "Snap-Cap" type coarse bubble diffusers. The headers
are removable from the tanks for cleaning or diffuser replacement
without removing the fiberglass covers.
Since no part of any RBC is visible during operation, proximity
switches are used as speed sensors to detect rotational failure of an
RBC.
Covers
The RBC's are protected from weather by one-piece fiberglass covers.
Eye-level rectangular ports are provided on one side of the covers. They
allow for detection of plugged diffusers by observation of the air cups
and manual rotation of the RBC's if required. Removable, press-fit
covers allow for bearing maintenance or replacement and access for
rotational speed, determinations.
1251
-------
The covers are supported on concrete beams which span the width of
the tank. Open grating is provided between adjacent beams for a walkway
between each RBC.
OPERATIONAL START-UP AND TESTING
Flow Conditions
The RBC's were started up on secondary effluent on 1979-09-19.
The flow to the RBC's for the next four weeks was normal flow from the
existing secondary treatment plants which included diurnal variations.
In preparation for testing for both process and energy performance,
two of the four RBC tanks were shut off and the pumping station was set
on manual control to provide a constant flow of secondary effluent to
the two remaining RBC tanks, one of which was to be tested. The flow
was chosen such that it was slightly less the future design flow of
13,750 m3/cl per tank.
The aluminum baffle plates were placed after the fourth RBC in the
tank under test to effectively isolate the first four RBC's. With a
surface area of 13.,750 m2 per RBC and a flow of 13,250 m3/d through the
tank, a removal of 6.2 mg/1 of ammonia nitrogen was required from the
influent to the effluent of the fourth RBC in order to'meet the specific
removal rate guarantee.
It was assumed that each of the four RBC's under test would
operate at the same specific removal rate if their rotational speeds
were the same. It woul.d therefore be necessary only to determine the
energy input to one RBC for guarantee purposes.
Analytical Equipment
Twenty-four hour composite samplers were installed with sample
tubing at the RBC influent and at the slotted opening at the bottom
of the baffles at the fourth RBC effluent. The samplers were set to
operate in the time proportional mode with one sample being taken each
hour. Samples for ammonia determination were preserved with concentrated
sulfuric acid added to the sample jar prior to sampling.
The hydraulic load cells and pump were used to determine biomass
thickness starting on date of start-up and throughout the testing period.
A temporary orifice plate flowmeter as shown in Figure No. 4,was
installed upstream of the butterfly control valve at the third RBC in
the tank under test. Pressure and temperature readings taken at the
flowmeter allow determination of both ambient air flow and power input-
to the RBC in accordance with the equations given in Appendix I.
Dissolved oxygen measurements were made at the RBC influent and
at each RBC. Wastewater temperature was also determined in the RBC tanks.
1252
-------
Perfo rma nee T es t i ng Res u1ts
Prior to commencement of the performance testing, twenty-four hour
composite samples of influent and effluent samples were analysed for
ammonia nitrogen to determine whether the system had stabilized at a
specific ammonia removal rate in excess of that guaranteed.
On the basis of such testing, the performance tests were commenced
on 1979-11-19. It was apparent, however, by 1979-11-21, that the results
were not meeting the guarantee requirements and consequently, the testing
was stopped.
It was suspected that the reason for the loss of nitrification was
excessive carbonaceous oxygen demand loading on the RBC's which inhibited
the nitrification population. A simultaneous increase in biomass thick-
ness tended to confirm this suspicion.
The existing secondary plants had been somewhat upset prior to
the testing and 'attempts were made to improve the secondary effluent
quality by increasing the aeration tank air supply. At the same time,
ammonia removal was monitored using grab samples in order to determine
when the testing might resume. More promising results were obtained
and testing recommenced on 1979-12-03. The test results are summarized
in Table 3.
There was less than 4% difference between average rotational speeds
for the four RBC's under test and the maximum difference was less than
10%. From this, it may be assumed that each RBC had the same specific
removal rate.
The average ammonia removal across the four RBC's was 6.1 mg/1
which is less than 2% below the process requirement of 6.2 mg/1. Given
the accuracy of the sampling, analysis, flow measurements and surface
area calculations, the RBC's were considered to have met the performance
guarantee.
The RBC influent pH averaged 8.0 slightly below the accepted
nitrification optimum of 8.5. Typical high influent alkalinity
concentrations of 300 mg/1 as CaC03 did not allow a significant pH
depression due to the alkalinity consumption of the nitrification
reaction.
The average suspended solids and BOD5 results of 15 and 13 mg/1
respectively are typical of the secondary treatment plant performance.
The average soluble BODs concentration of 9 mg/1 was slightly less than
the 10 mg/1 maximum limit in the process guarantee.
The average wastewater temperature of 14°C was slightly above the
minimum required of 13°C. It is also of interest that there was .no
change in wastewater temperature from the influent to the effluent of
RBC No. 4 in spite of ambient air temperatures about 0°C and the uninsulated
fiberglass covers.
1253
-------
The RBC influent dissolved oxygen concentration averaged 4.9 mg/1
while the RBC No. 4 effluent averaged 5.5 mg/1, an increase of about
12$. This is due in part to the air dirve system and in part to the
oxygen transfer capability of the RBC itself. These concentrations far
exceed the minimum required of 2.0 mg/1. The influent concentrations
are significantly higher than normal secondary effluent because of the
aerating effect of the axial flow pumps which lift the secondary effluent
to the RBC tanks.
The average energy consumption rate for the tested RBC No. 3 was
1.11 kw. This is almost 46% less than that guaranteed. It should also
be noted that this power requirement is at 1.13 rpm. If the process
performance can be .met at 0.8 rpm as in the pilot study work,the power
requirement would be reduced to about 0.5 kw which is about 76% less
than that guaranteed.
The biomass thickness on RBC No. 1 in the test tank is plotted as
a function of time in Figure No. 5. The data exhibit the classic micro-
organism growth pattern1. Initially, there is a lag phase as the x
bacteria begin to establish themselves. This is followed by a constant
growth phase in which the biomass increases at an expoential rate as all
nutrients are in excess. As the nitrifying biomass concentration increases,
the ammonia concentration becomes growth limiting and the declining growth
phase begins. The system then reaches an equilibrium stage in which the
biomass thickness remains relatively constant with minor fluctuations due
to variations in influent BOD5 and ammonia concentrations,
Figure No. 6 illustrates biomass thickness during the test period
as a function of RBC tank length. The biomass thickness is linearly
related to RBC tank length. This is attributable to the decrease in
ammonia concentration through the tank.
Testing was also carried out to determine power input and ambient
air flow as a function of rotational speed of the RBC. With the butterfly
control valve fully open, the discharge header pressure was varied. The
resulting power inputs, ambient air flows and rotational speeds were then
recorded. A linear regression was performed on both sets of data expressin
power and air flow as power functions of rotational speed,, The regression
analysis is summarized in Table 4.
For the power input the best correlation is with the third power of
rotational speed. This is in agreement with the laws of dynamic
similitude for rotating systems2in which the following equation is given
for the dimension!ess group ir3 :
TT3
(1)
where P = power
p 5 density
N = rotational speed
D H diameter
1254
-------
In this case p and D are constant and
P a N3
The test data points and the linear regression curve are shown in
Figure No. 7.
(2)
For the ambient air flow the best correlation is with the second
power of rotational speed. This can be rationalized by recognizing that
the net torque on the RBC is proportional to the ambient air flow in the
range of the testing or
QaT (3)
where Q = ambient air flow
T 5 net torque on RBC
It follows that:
V
V P a T x N (by definition) (4)
and P a N3 (2)
.*. T a N2 (5)
and Q a N2 . (6)
The test data points and linear regression curve are shown in
Figure No. 8. Also shown is the comparable data as developed by
Autotrol Corporation3. The test data is about 10% lower than the
Autotrol data. It should be noted that the Autotrol data is for
design purposes and is therefore conservative.
Future Testing
The wastewater temperatures are presently (1980-02-06) too low to
optimize the system performance. When the temperatures rise-again in the
summer of 1980, additional testing will be carried to determine the optimum
baffle arrangement for staging and the minimum possible rotational, speed to
minimize energy costs.
COSTS
Energy Costs
The electrical input power at the blowers at two different RBC
rotational speeds were determined. With this data the overall efficiency
of the air supply system and the energy cost was calculated. These
values are summarized in Table 5.
1255
-------
The efficiency of the air supply system decreases significantly
at RBC rotational speeds above 0.9 rpm because two blowers must be in
operation. The energy cost is reduced by 61% by reducing rotational
speed from 1.1 rpm to 0.8 rpm.
Total Annual Cost
The total annual costs are summarized in Table 6. The capital
cost is based on the contract prices for equipment and tankage. The
electrical energy is based upon operating the RBC's at 1.1 rpm but
this may be reduced to 0.8 rpm which will significantly reduce the
energy cost. The maintenance costs are estimated at about 1% of the
equipment capital cost. The labour cost is based on one man full-time
for operation and maintenance.
The cost per unit wastewater flow of 2.7<£/m3 is about 16% of the
total annual cost of a complete secondary treatment plant. It should be
noted that these costs do not include for any solids separation'after
the RBC's. At Guelph, solids separation is accomplished through dual
media filtration but only about 15% of the solids loading on the filter
is from the nitrification process while about 85% is from the existing
secondary treatment system. The pumping cost, which is site specific,
is also not included in the above costs.
ACKNOWLEDGEMENT
We gratefully acknowledge the assistance of the City of Guelph's
Engineering Department Staff; Mr. R.D. Funnell, P.Eng., City Engineer;
Mr. K.J. Bull, P.Eng., Operations Engineer and Mr. J.A. Sanvido,
Superintendent of Water and Sewage Works in the preparation of this
paper.
1256
-------
APPENDIX I
The following equations were used to determine ambient air flow
and power,input to the RBC under test.
P = 1.011 m
where P E
m E
m =
1
Pa
0.283
(7)
: power (kw)
= air mass flow rate (kg/s)
T2 = air temperature six pipe diameters downstream
of orifice plate (OK)
PS = absolute barometric pressure (kPa)
p2 = absolute air pressure six pipe diameters downstream
of orifice plate (kPa)
V2. pgc Ap x 1000 (8)
= compressibility factor (defined below)
C-j E discharge coefficient (determined from tables * as
a function of Reynolds number at the .orifice)
i\2 = orifice area (m2)
3 = orifice to pipe diameter ratio
p = air density at one pipe diameter upstream of
orifice plate (kg/m3)
Ap = differential pressure measured,at 1.0 pipe
diameters upstream and 0.8 pipe diameters
downstream (kPa)
gc = proportionality factor in Newton's second law
(kg-m/N-s2)
Y = 1 - (0.41 + .0.353")
(9)
where pi =
k =
P = Pi
RTi
where R =
absolute air pressure measured at one pipe
diameter upstream of orifice plate (kPa)
specific heat capacity ratio for air
gas constant for air (J/kg-°K)
air temperature measured at one pipe diameter
upstream of orifice plate (°K)
1257
(10)
-------
(11)
where Q = ambient air flow rate (m3/s)
fs= ambient air temperature (°K)
1258
-------
LIST OF REFERENCES
1. "Unit Processes of Sanitary Engineering", L.G. Rich, J. Wiley
and Sons, 1963, pp 9-14.
2. "Essentials of Engineering Fluid Mechanics", R.M. Olson, Inter-
national Textbook Company, Second Edition, 1966, p 179.
3. "Autotrol Wastewater Treatment Systems Design Manual", Autotrol
Corporation, 1979, p C-15.
4. ASME Research Report on "Fluid Meters, Fluid Meter Applications",
6th Edition, p 206.
1259
-------
TABLE 1
EFFLUENT REQUIREMENTS
For river temperatures less than 10 C
BOD5 - 440 kg/d
Ammonia - Best Practicable Treatment
Phosphorus - 1 mg/1
For river temperatures above 10°C
Total Oxygen Demand (TOD) *
Phosphorus
* TOD = BOD5
1190 kg/d
0.5 mg/1
x - +4.57 x NH3 - N
tSUUs
Effluent loadings from expanded Water Pollution Control
Plant to ensure dissolved oxygen in the Speed River
downstream o.f the plant not less than 5 mg/1 more than
20% of any day nor less than 4 mg/1 at any time.
1260
-------
TABLE 2
PROCESS GUARANTEE CONDITIONS
Specific Ammonia Removal Rate
Ammonia Concentration
Average Influent Soluble BODs
Wastewater Temperature
Dissolved Oxygen Concentration
Biomass Thickness
1.5 g/m2*d
6.0 mg/1
10 mg/1
13°C
2.0 mg/1
3.0 mm
1261
-------
O E
CQ E
CO *->
s =
an E
O^ O CTv O
1262
-------
TABLE 4
SUMMARY OF LINEAR REGRESSION
OF POWER INPUT AND AMBIENT AIR FLOW
vs
ROTATIONAL SPEED
Assumed Equation
Linear Regression Result
Correletion Coefficient
P = mN + b
P = mN2 + b
P = mN3 + b
P = mN" + b
Q = mN + b
Q = mN2 + b
Q = mN3 + b
P
P
P
P
Q
Q
Q
= 1.88N - 1.02
= 0.99N2 - 0.15
= 0.70N3 + 0.13
= 0.54N1* + 0.27
= 99N - 49
= 52N2 - 3.2
= 37N3 + 11.6
0.9941
0.9981
0.9988
0.9975
0.9975
0.9997
0.9989
P = power input (kw)
Q = ambient air flow (X10"3 m3/s)
N = rotatiorjal speed (rpm)
-------
TABLE 5
SUMMARY OF ENERGY COSTS
RBC
Speed
(rpm)
0.8
1.1
Power Required
at each RBC
(kw)
0.49
1.06
Total Power
Required at RBC
(kw)
15.7
33.9
Blower Power
Required
(kw)
28
72
Eff.
56
47
Energy Cost
($/a )
(@ 2.5<£/kwh)
6130
15770
1264
-------
TABLE 6
TOTAL ANNUAL COST
FOR RBC NITRIFICATION SYSTEM ($Can. 1979)
Capita] Cost
Rotating Biological Contactors 1,540,000
Air Drive System 145,000
Covers 121,000
Tankage and Miscellaneous 1,550,000
Engineering @ 9% 302,000
Total Capital Cost 3,662,000
Annual Amortized Capital Cost @ 13%
Operating Cost
Electrical Energy 16S000
Maintenance 175000
Labour 30,000
Annual Operating Cost
Total Annual Cost
Cpst per unit wastewater flow
476,000
63,000
$ 539,000
2.7<£/m3
1265
-------
0
cx>
o
o
O
O
o
0 0
1
o
o
ro
>
o
i
i
OF GUELPH
L_
5
CE
o
j— •
o
^j
h-
—^
*£.
O
o
_!
O
<£>
3
O
DQ
O
t
g
o:
z E
(5 ^
P ^^~
< 8 •
c °>
_j < t
< Z 0
> o o
o s ~i
^ ^ LU
Lu < >
a:
1= a:
0 ^ <
y LU i
ill
CO LU <
01
k Z
1 U
_'^
^ O
S5 X C
' • •••
z
. CO
12 1^
NTRATION /s
LU
Q g
o
0
-CO LU
O
a:
h-
-CD Z
V \
o o
Q
to
Q
OJ
CM
z
o
Oldl03dS
1266
-------
&
co
LJ
o
a
\-
QL
b
u.
(0
o
u.
o:
LJL
O
_J
Q_
LU
a:
ID
CD
U.
-------
CO
ro
01
LU
a:
ID
C9
Lu
a:
o
o
o
o
o
o
3
O
CD
CD
z
I
o
a:
h-
o
Id
CO
1268
-------
01
Z
UJ
cc
S2
Lu
o:
UJ
LJ
3
u.
LU
UJ
o
u_
o:
o
1269
-------
o
co
<
O c
SS3N>D1H1
cvl
o"
o
00
O
cp
O
10
o
o
to
O
in
CM
z
UJ
(T
ID
O
U-
UJ
1270
-------
u,
CD
o
O
g
O
o
-J
<
o
I CD
3
O
CD
CO
CO
LU
(T
7)
LU
O
CO
01
cc
o
CO
CO O
-------
1.2
1.1
1.0
0.9-
OB-
~ O7-
CITY OF GUELPH
ROTATING BIOLOGICAL CONTACTOR
PERFORMANCE TEST RESULTS
POWER INPUT VS
ROTATIONAL SPEED
CL
0.6-
0.5-
0.4
P=0.70 N^O.13
r= 0.9988
0.7 08 0.9
1.0
1.1
1.2
N (rpm)
FIGURE N27
1272
-------
M
10^
IO
'O
O
90
80
70
60
50
40-
30
20
10
CITY OF GUELPH
ROTATING BIOLOGICAL CONTACTOR
PERFORMANCE TEST RESULTS
AMBIENT AIR FLOW VS
ROTATIONAL SPEED
AUTOTROL
DATA
0
0.7 0.8 0.9
fSj(rpm)
1.0 I.I 1.2
FIGURE N28
1273
-------
-------
DENITRIFICATION OF NITRIFIED MUNICIPAL
WASTEWATERrUSING ROTATING BIOLOGICAL CONTACTORS
by
Frederic C. Blanc
Associate Professor of Civil Engineering
Northeastern University, Boston, Massachusetts, U.S.A.
James C. 0*Shaughnessy
Associate Professor of Civil Engineering
Northeastern University, Boston, Massachusetts, U.S.A.
Daniel J. Connick
Engineer^ E.D.P., Inc-.,. Cambri'dge, Massachusetts, U.S.A.
Donald Wood
Engineer, L.E.A. Inc., Boston, Massachusetts, U.S.A.
INTRODUCTION
Biological denitrification has been developing rapidly as a means of re-
moving nitrate nitrogen from wastewaters to meet total nitrogen standards when
applied for controlling eutrophication in surface waters.
This research utilized rotating biological contactor (RBC) units to per-
form bench scale and pilot plant denitrification studies of a nitrified domes-
tic wastewater. Tests were performed to determine the effects of hydraulic
loading, nitrate concentrations and temperature, and to find the optimum meth-
anol addition requirement.
Specifically, the objectives of this study were: (1) to observe removal
efficiencies of nitrate nitrogen under various loading rates; (2) to determine
the effect of nitrate concentration within the reactor on removal rates; (3)
to determine the effect of temperature on the rate of reaction; and (4) to
ascertain the quantity of organic carbon required, as methanol, for optimum
removals.
1275
-------
BACKGROUND
Denitrif ication is ..the Biological process in which nitrate is partially
reduced to gaseous end products and partiailly converted to cell material.
The conversion of nitrate to cell material is by the assimilatory mode,
which follows the stoiciometric equation; (1)
522 2 + 3HCO,
where methanol is used for the organic carbon source,
Dissamilatory denitrif ication follows a series of, enzymatic reactions
which can be simplified for design purposes as a two-step process. The first
step involves the conversion of nitrate to nitrite, and the second step con-
verts the nitrite to nitrogen gas as follows;
N03 + 0.33CH3OH = N02 + 0.33C02 + 0,67H20 Step 1 (1,2)
N02 + 0.5CH3OH => 0,5N2 + 0,5C02 4- 0.5H20 + OH"* Step 2
where methanol is used as the organic carbon source. The overall transforma-
tion is then represented by;
N03 + 0.833CH3OH = 0.5N2 + 0.833C03 = 1.167H2-0 + OH~
Since hydroxide ions are produced in this reaction and carbonic acid is pre-
sent, an increase in bicarbonate alkalinity is expected during denitrif ication.
Various factors such as; temperature, alkalinity, pH, organic carbon
source, and process operating parameters, determine the performance of any
denitrif ication unit,
Dawson, et al (3) performed batch, bench-scale, mixed reactor studies
using pure cultures of Psuedomonas, denitrif icans to determine . temperature
dependencies. The rate of reaction for this test was plotted as a function !
of temperature and could be represented by an Arrhenius curve . The results
of this study are shown in Figure 1,
Alkalinity production for denitrif ication may be defined stoichiometric-
ally as 3.57 mg of alkalinity as CaCO- per mg nitrate reduced to nitrogen
gas. (1) Good denitrif ication .will take place if the pH is' in the 6.0 —
8»0 range. The optimum pH is in the range of 7.0 to 7.5.
Some idea of proper process operational parameters may be obtained by
examination of research involving suspended cell reactors. Moore and
Shroeder (A) , investigated the effects of cell residence time on cell yield
and the rate of dentrif ication, within an activated sludge system. They
concluded that a six day cell residence time was near optimum for low pro-
duction and thus low methanol requirements. Although greater than nine day
cell residence time could reduce these parameters, it would require a sub-
stantial unit volume increase for similar flow capacities. The importance
of a complete~mix system was noted due to the detrimental effect of short
1276
-------
0.2
s_
3
o
o>
CJ
0.15
-o
cu
>
.o
cu
s_
<£> 0.1
o
TO
d)
O.O5
95 # Confidence Limits
10
15
20
25
30
Temoerature C
Figure 1 Denitrification Temperature Dependency
R.N. Dawson Data (3)
1277
-------
circuiting on cell residence time.
Various investigations (5,6,7,8,9) using fixed film reactors Have also
been performed. For example, Buddies et al C5) used a. mixed media filter op-
erating as a packed bed denitrif ieation chamber in addition to a solids sepera-
tion device. Methanol was fed at a, 3.5 mg CH OH/mg nitrate nitrogen ratio.
Greater than 95 percent dfenitrification was attained with this system at a flow
rate of 2.5 gpm/sf with, an influent nitrate level at about 8.7 mg/1. The
reaction rate could be determine to be 78,2 mg/sf/min. Backwashing was nec-
essary every 24 hours and required 10 percent of the total flow volume.
The use of RBC*s for idenitrification has only recently come under investir-'
gation. Davies (10) bas reported on the use of a partially submerged rotating
disc unit (RDU) operated anaerohically for denitrification. The unit provided
an area of 11,25 sq m for bacterial growth with, a surface area to liquid volume
ratio of 789.5 sq m/cu m. All parameters were controlled to be constant, when
not being specifically analyzed. Nitrate influent and effluent levels were
60 mg/1 and 10 mg/1 respectively, the influent being trickling filter effluent
spiked witbuKNO,. The pH was fixed at 7.0, the temperature maintained at 20°C,
and the C:N ratio kept constant at 3;1 by the addition of methanol. Results
of the study indicated that the maximum rate of denitrif ication was 260 mg
NO--N reduced per square meter per hour.
Murphy, et al 0-1), reported on a comparison of the denitrification capa-
bilities of a suspended growth system, a submerged RBC unit, and four upflow
submerged packed columns.
The suspended growth system operated at a,2.5 hour detention time with
an SRT of 3 to 9 days, The RBC unit had a total surface area of 23.2 sq m,
was divided into four equal stages and operated at 4 liters per minute giving
a 100 minute detention time.,
Results of the RBG unit indicate that the denitrification rate was inde- •
pendent of nitrate and nitrite concentrations. The unit did provide efficient
and predictable removals at all temperatures normally encountered. Data was
reported as removals per unit volume, however, a more useful form for compar-
ison is on a surface area basis. From the reaction rate temperature dependency
curve, removals at 5°C and 20°C could be determine to be 93 and 435 mg/sq m/
hour, respectively. The effluent from this unit was reported to quite low in
suspended solids, indicating the possibility of eliminating further clarifica~ ,
tion.
EXPERIMENTAL EQUIPMENT
This paper presents the results of laboratory bench, and pilot plant re-
search investigations which evaluated some of the design parameters involved
in denitrifying a nitrified municipal wastewater effluent using rotating bio-
logical contactors.
Initial laboratory bench scale experiments were conducted at Northeastern
University's Environmental Engineering Laboratories over a period of 12 weeks
in the spring of 1976,
1278
-------
Feed.
Tank
Deaeration
Chamber
Anaerobic
Biodisc A
Feed Pump
Effluent
Methanol Feed
I
Anaerobic
Biodisc B
FIGURE 2 : Flow Scheme for Laboratory Denitrification Units
1279
-------
C
Q)
H-l
5-1
Q)
4J
C
O
O
O
-JJ
•H
C
O
•
ra
g
3
Oi
C4
o
4-> W
C 0)
(1) -H
O
-P
-H r-l
C (O
D C
-H
• fa
UJ
• g
CO
-P
O
C
O
•H
-P
KS
O
•H
-P
•H
C
Q)
Q
0)
S
Q)
x:
o
CO
o
r-1
on
0)
bO
•H
1280
-------
_
Two airtight, single stage units were constructed of clear plexiglass.
Each housed a stainless steel shaft supporting ten discs giving a total contact
area of 9.08 ft/unit for Biomass growth. Disc media was polyethylene, supplied
by EPCO-HQBMELL and designed to Be 40% submerged. The wastewater volume of
each unit was approximately .8 gallons, giving a fixed volume to surface ratio
of 0.09 gal/ft^'. The units and a flow- scheme are shown in Figure 2.
Influent was supplied via constant head peristaltic pumps through inert
tygon tubing. Methanol, originally fed with an electrolysis pump, was later
also supplied - fey peristaltic pumps.
Other unit components consisted of an influent deoxygenation chamber, in
which. N~ gas was diffused through the wastewater, and an effluent clarifier
designed to prevent air from entering the disc unit via the effluent line.
The single stage pilot plant portion of this project was conducted at
the Marlborough. Easterly Treatment Plant. A ten week study was conducted in
the Fall of 1976 using clarified activated sludge from secondary tanks, which
was further treated by pilot plant RBC nitrification units, as dentrification
RBC unit influent. The unit was then washed down and reset for a 24 week
study running directly from the final effluent (prior to chlorination) of the
Marlborough Plant,
One single stage unit was utilized for the pilot study. A stainless steel
shaft housed 12 polyethylene discs 2T in diameter, also of material supplied
by EPCO-HOBMELL, which combined with the wetted surface area of the unit pro-
vided 126.5 ft2 for the attachment of microbial growth. The shaft was also
attached to an external variable speed drive and rotated at 3 1/8 RPM under
normal operating conditions. This unit was operated with the discs totally
submerged in the wastewater which itself had a volume of 58,6 gal. The vol-
ume to surface ratio for this unit was 0.46 gal/ft2. Figure 3 schematically
illustrates the pilot unit operation,
A two stage denitrification pilot plant unit was operated at the Marlborough
Easterly Treatment Plant from July 1977 to November 1977. A central stainless
steel shaft held 20 polyethylene discs, 10 in each stage. The discs were 2
feet in diameter and were made of material supplied by EPCO-HOKMELL. The spa-
cing of the discs centerline to centerline was approximately one inch and each
disc section was held in place by two plexiglass end pieces. The polyethylene
discs provided 88.8 sq. ft. of surface area/stage for microbial growth. The
surface area provided by the wetted walls of the unit and the plexiglass end
pieces provided an additional 37.7 sq, ft. per stage for a total surface area
of 126.5 sq, ft, per stage. The discs were submerged in a total volume of 69
gallons for each stage for a volume to surface area ratio of 0.546 gallons/ft^.
The shaft was attached to an external variable speed drive and rotated at
eight EPM under normal operating conditions. The unit was operated with the
discs totally submerged in wastewater and the unit was covered.
ANALYTICAL METHODS
All samples were grab type samples. Analyses were either run immediately
after collection, or the samples were refrigerated at 4°C. All analyses were
completed within 24 hours of collection oh the refrigerated samples.
1281
-------
TABLE 1: SUMMARY OF DENITRIFICATION IN BENCH SCALE UNITS:
Nitrate Concentrations
mg N03-N/1
DATE,
5/10
12
18
20
6/02
6
15
17
21
29
7/01
3
7
8
9
13
14
Average
Value
Inf
23.0
25.0
20.0
20.0
20.0
11.0
13.2
23.0
20.0
16.0
13.6
12.4
14.4
15,2
15.6
21.6
18.4
17.8
Unit A
Eff
2.0
1.8
1.0
1.4
2.3
1.1
0.2
0.4
0.0
0.5
0.4
0.2
0.3
0.6
0.4
0.3
0.8
Unit B
Eff
---.
—
3,6
8.0
5.0
0.4
2,8
0.6
0.5
0.2
0.8
0.3
0.6
0.3
0.4
1,8
Percent
N03 Removal
Unit A Unit B
91
93
95
93
88
86
98
98
100
"•».
97
97
99
98
96
98
98
95
-•**
-"
—
82
60
45
97
—
86
96
97
97
94
98
96
99.
98
88
1282
-------
All laboratory analyses were conducted in accordance with the methods out-
lined in Standard Methods (12). Nitrate nitrogen was measured by the chroma-
tropic acid method after Standard Methods (12).
EXPERIMENTAL RESULTS AND DISCUSSION
I - Laboratory Bench Scale Study
The denitrification bench scale study was run for two and one-half months,
from early May through mid*-July, 1976, in order to determine the feasibility
of RBC denitrification and aid in pilot plant design. The hydraulic loading rate
was held fairly constant throughout the study; however, the nitrate nitrogen
concentration varied. The reason for the variation was that the feed influent
used was the effluent from a nitrification RBC bench study.
The major problem encountered during the bench scale denitrification study
was achieving the proper methanol addition to the influent, so-,a sufficient
carbon source would be .^available to the denitrifying bacteria. In each case
where either of the units did not achieve a high level of denitrification, there
was insufficient methanol addition to the influent.
In evaluating the data, if insufficient carbon feeding was apparent, then
that data were not used,
During the bench scale study, most of the loading rates were below 50 mg/m^-
hr (0.24 Ibs N03»-N/1000ft2-day). Unit A had an average nitrate nitrogen efflu-
ent concentration of 0.81 mg/1 with an average value of 95 percent of the .nitrate
nitrogen removed.
The influent averaged 1.9 mg TKN/1 with the effluents averaging 1.3 mg
TKN/1 and 1.6 mg TKN/1 for Units A and B, respectively. A summary -of total
denitrification performances is listed in Table 1.
The data indicated that a high degree of denitrification was achieved using
rotating biological contactors. Over 90 percent of the nitrate nitrogen was
converted to nitrogen gas, without any buildup in.total nitrogen concentrations
in the effluent. Near anaerobic conditions were maintained within the units
and total nitrogen removal averaged greater than 85 percent.
II PILOT PLANT STUDIES
A - Single Stage Unit
A single stage denitrification pilot unit was run at Marlborough, Massa-
chusetts Easterly WTP from October 1976 through June 1977, The unit was a
single stage unit, which had the discs totally submerged in the wastewater
stream, as described in a previous section of this paper. Methanol was used
as the carbon source during this study.
During 1976 the denitrification unit was fed nitrified effluent from the
RBC nitrification units. During 1977, the Unit was fed nitrified effluent
for the Marlborough Easterly WTP activated sludge process. This effluent was
collected prior to chlorination.
1283
-------
•P
fi
a>
W
t)
M'H
CU
V4 O
4J-r»
•H -P
0)
4J
a)
•H O
530
s
s
M
5vu)
.1284
-------
8
£5
t
ro
g
tn
g
6 _
5 -
4 _
3 _
2 -
1 -
80 120 160 200 240
Time (Minutes)
FIGURE 5: Nitrate Concentrations vs Time
For No-Flow Conditions
1285
-------
z
I
3/12/77 Data
3/20/77 Data
4/30/77 Data
Y = 0.0433X +• 0.269
r = .978 N = 30
X = 23.1Y - 6.2
I
40 80 120 160
Mg N03-N REMOVED/SQM./HR
200
240
FIGURE 6: Nitrate Nitrogen Concentrations vs
Nitrate Removal Rate for No Flow
Conditions
1286
-------
Representative nitrate nitrogen concentration data collected throughout
the study period are plotted in Figure 4. Because of three major factors:
1. the variability of the unit influent nitrate concentration;
2. the clogging problems in the pumping system, and thus high flow
variations j
3. and the low nitrate nitrogen loading rates studied in the bench
scale study,
loading rates on the pilot unit were varied throughout the study period. Gen-
erally, influent flow rates were held fairly constant for a. minimum of one-
week period. Steady state flow- was achieved more often during the 1977 por-
tion of the study.
Since the pilot unit was single stage, and the effects of staging could
not be evaluated, higher loading rates and resulting removals were determined,
rather than the lower loading rates studied with bench scale units.
During the operation of the pilot unit, it became apparent that the ac-
tual rate of nitrate removal was related to the nitrate nitrogen concentration
within the unit. In order to evaluate the rate of reaction as affected by
nitrate concentration, several batch tests were run. For these tests, the in-
fluent was shut off to the unit, and the nitrate level within the unit recor-
ded with time. The results were parabolic curves of decreasing nitrate levels
with time. A typical run is. plotted in Figure 5, The slope of each curve at
any nitrate level indicated the actual removal rate at that level. The data
obtained from. no~flow conditions was used to calculate the nitrate removal
rates, and the rates were plotted vs. nitrate concentrations and shown in
Figure 6. Fow low nitrate levels the reaction rates for the three dates tested
are close to equivalent, however,, at the higher nitrate concentrations, there
was larger variation,
Based on this data, the removal rate, at any oeprating level of nitrate,
within the unit, was calculated to be:
R = 23.1 N - 6,2
where: R is the mg/sq m/hr removed and
N is the nitrate level mg NO,,-N/1 in the unit.
Being a complete mix system, the effluent level would also be the same as that
within the unit.
The correlation coefficient for this curve is 0.98 and therefore, over
95% of the variation in the rate of removal is accounted for by the change
in the nitrate level within the reactor unit.
As the RBC is designed to be a continuous flow unit, a similar comparison
with .flow-through conditions was necessary.
1287
-------
c
o
•H
O
C
o
o
.p
c
8
4J
•2. 3
4J
ra
ri 2
Y = .0653X - 1.23
r = .90
X = 15.4Y + 18.9
C
L
J_
40 80 120 160 200
NO,-N Removed mg/m2/hr
240
FIGURE 7: Effluent Nitrate Concentrations vs Nitrate
Nitrogen Loading Rates
1288
-------
240
200
160
t>
o
o
o
z
120
80
40
Bench Study Data
Pilot Study Data
Y = 0.53X +17.0
r = 0.90
I
40
80 120
-N Applied
160
200
FIGURE 8:
Nitrate Nitrogen Removed vs Nitrate
Application Rates
1289
-------
II II
> n
0)
•U
CO
10
0)
W)
< 53 to
2 «
I 0) 4J
"i 4-». cd
•H O
85 H
4-1 td
Q)
a
O
M
P-4
1290
-------
I
=4
en
o
O
fa
l4
I
Q
§
O
PM
O
O
O
03
7VAOK3H 3iVHIIN
1291
-------
With the unit operating at various flow rates, temperatures, and nitrate
influent and effluent levels, actual removal rates were determined. Although
some oxygen entered the'unit, and at times was. not completely removed, its
effects were not initially considered. Figure 7 shows the nitrate nitrogen
removal rate vs, the nitrate nitrogen concentrations in the effluent for data
collected in a flow-through situation. As expected, due to the varying con-
ditions listed above, there was somewhat more scatter within the data than
when under no^flow conditions. The correlation coefficient was 0.90, such
that 81% of the removal rate variation could actually be attributed to the,
change in nitrate level.
The data plotted in Figure 7 neglect correction for both temperature and
any dissolved oxygen which may be present in the unit..
Figure 8 is a plot of nitrate nitrogen removal for various nitrate load-
ing rates. As the influent nitrate loading increased, the nitrate removal de-
creased.
The percent nitrate removal vs, nitrate application rate is plotted in
Figure 9 and the data indicate that 90% nitrata-N removal can be expected at
loading rates less than 55 mg/sq m/hr. The removal may be approximated from
the linear relationship shown on Figure 8. This relationship was for loading
rates up to 250 mg N03~N/m2-hr. or 1,22 Ibs. NO -N/1000ft2-day.
Nitrate removal data from the single stage unit was. plotted vs. hydraulic
loading rates (l/m2"hr) in Figure 10. As the hydraulic loading, and thus the
nitrate loading, increased, the percent nitrate removal decreased. Hydraulic
loading rates of 8.5 1/m -hr. had at least 60% nitrate removal while hydraulic
loadings below 2.0 l/m^hr. had 90 to 95% nitrate removal. The relationship
was calculated as being:
% Nitrate Removal = 102 ~ 4.9 Q,
where Q = l/m2/hr.
The plotted data was not corrected for temperature, and was in good aggree-
ment with Davies' data (10). Again this data was obtained using a single stage
unit with variable nitrate and temperature values.
In this study, methanol (CH-OH) was used as the carbon source. Varying
amounts of methanol were added to the wasteflow in order to determine optimum
carbon requirements. This data can be found in detail in reference (13). The
results indicated that in a single stage unit, addition of more than 80 mg
COD/1, carbon addition did not increase nitrate removal efficiencies, irre-
spective of the nitrate loading rate,
Figure 11 shows the ratio of methanol addition to nitrate removal for
different loading rates and removal efficiencies. The nitrate application
rates ranged from 40-180 mg NO,.<-N/m -^ir, The methanol requirement is based
on the COD requirement for NO~~N removal and D.O, removal within the unit.
For this study, the methanol requirement ratio in terms of COD was calculated
to be 1.37. The ratios shown in Figure 11 are for CH^OH requirements (not
COD) added to achieve NO-HST removal. These ratios are for influent nitrate
1292
-------
o o o o
s?
o
o
01
a
O
ts
o
03
(D
a
•3
a
>
o
o
i
•c
•o
•a
u
a, .
a •->
I E
MX.
O tr>
2 E
is
e:
a
c
CO
c
c
4-1
•H
g
O
CO
O
•H
4J
0)
•P
rt
J3
O
4J
O
§
(1)
CO
o >
o o
«w 6
m cu
O
O
O
o
O
M
aivaiiN
1293
-------
o
t/>
CO
C3
I— 1
u_
oo
o>
Dl
CO
(O CD
o>
1294
-------
CO
O
CD
OJ
to
CO
CD
CO
-P
to
s-
CM
LU
o;
3
o
O
UD
CD
CD
LO
I
CO
CD
V
o «*
3 S
8
CO
en CM
•r— ^^.
-o en
to £
O >—*
.
O)
CD
O
CM
O
C3
CD
LO
OJ
CD
CD
CM
O
to
CD
CD
CD
LO
N-SON SB
3-sai.j
(aq/ui/6ui)
1295
-------
LO
CO
CO
LO
CM
CO
•»->
c«
CM
fd
o>
en
05
to
CT>
evi
LO
-a
ns
QJ
UJ
o:
C3
»—I
U_
O
LO
CD
LO
CM
CVJ
LO
LO
M-£ON SB
(uq/ ui/Bui)
1296
-------
LL!
ta
>
o
65
CD
ra
CO
UJ
cs:
oo
CD
un
CM
CD
CD
CM
o
in
I
oo
CD
o
CO
+->
n3 *•"••»*
oc s-
J=
CD —,
E CM
^ ^
tO C7>
5-a
4->
E
CD
CD
CD
CD
LO
O
CD
O
OO
o
vo
CD
CM
O
1297
-------
nitrogen concentrations of 22 mg NOo
The data shown in Figure 11 indicates, that the nitrate removal efficien-
cies in the RBC unit was a function of Both, nitrate loading rate, and the
carbon to nitrate ratio used. In this study, for the loading rates investiga-
ted, the optimal methanol to nitrate ratio varied between 2.4 - 3.0, depending
on nitrate loading rate.
Although an exact solids production monitoring balance was not continually
conducted, the solids production was low. This is to be generally expected in
anaerobic processes. The average value of effluent suspended solids was 3.4
mg/1.
Temperature dependency in the range of 6 to 16°C was also levaluated in
this study. The results indicate that it could be described by the following
expression;
B - Two Stage Pilot Unit
The two stage pilot plant was operated at the Marlborough W.T.E. during
the summer and fall of 1977, one year after the single stage operation. Each
stage was essentially the same in surface area and volume as the one stage unit;
his unit was operated at a higher rotational speed and had a greater spacing
between discs.
Figures 12, 13, 14, and 15 show the curves of nitrate loading rates vs.
removal rates and nitrate loading rates vs. percentage removal rates for the
two stage unit. Figures 12 and 13 present first stage results while Figures
14 and 15 present the two stage unit results. To compare the effects of stag-
ing for the two stage investigation in 1977, Figures 13 and 15 may be examined
simultaneously. These figures indicate that at an application rate of 1 lb^
of N03~N per 1000 ft2 of surface area ,(205 mg N03~N/m2-hr) , the single stage
would remove 63 percent of the nitrate while a two stage unit would remove
83 percent. In both instances, the contact times were the same.
CONCLUSIONS
The following conclusions were reached based on the studies described in
this paper:
1. Denitrification of a nitrified wastewater can be readily accom-
plished using rotating biological contactors .
2, There is a linear relationship between the amount of nitrate ;
applied and the amount removed. This can be expressed by the
equation; NR = .59 NA + 18.8, with the variables expressed as
mg NOo^N/sq m-hr.
3. Nitrate concentration in the unit has a significant effect on
removal rates. The higher the level of nitrate within the
unit, the higher the removal rate will be.
1298
-------
5.
90% nitrate reduction can be expected at loading rates of up
to 55 mg NOg^-N/sq m-hr under normal operating conditions in a
.single stage reactor. For a two stage reactor, 90% nitrate re-
duction will occur at loading rates up to 100 mg NOo-N/sq m-hr.
Removals will increase with increasing detention times for any
unit at a given influent nitrate concentration.
6, Methanol requirements for 'optimum removal can be estimated as:
CH OH mg/1 = 2.6 N03^N mg/1 + 0.91 D.O. mg/1. Lesser amounts
limit removals and excess amounts produce high effluent BOD values.
7. The temperature dependency can be estimated as:
K, . K2Q 1.075
8. Solids production was low at all loading rates during these in-
vest iga tions.
9. Staging will improve the performance of RBC denitrification.units
1299
-------
REFERENCES
ln) Process- Design 'Manual;For Nitrogen Control. EPA Technology Transfer
Oct., 1975.
2,) McCarty, P.L., and P. St Amant, "Biological Denitrification of Wastewater
By Addition of Organic Materials", Proc. 24th Ind, Waste Conf., Purdue
Univ., W. Laffayette, Ind., Ext, Ser. 135-1272 (1969)
t
3.) Dawson, R.N., and F,A, Pretorius, "The Temperature Dependency of Bio-
logical Denitrification", Water Research, Vol. 6, pp, 71-83, 1972.
4.) Moore, S.F., and E.D. Shroeder, "An Investigation of the Effects of Cell
Residence Time on Anaerobic Bacterial Denitrification", Water Research,
Vol. 4, pp 685-694, 1970
5.) Duddles, G.A., Richardson, S.E., and E.F, Earth, "Plastic Medium Trick-
ling Filters for Biological Nitrogen Control", JWPCF, Vol. 46, No. 5,
pp 937-946, May, 1976
6,) Requa, D.A., and E.D. Shroeder, "Kinetics of Packed Bed Denitrification",
JWPCF, Vol. 45, No 8, pp. 1696-1707, Aug. 1973
7.) Jewell, W.J,, and R.J. Cummings, "Denitrification of Concentrated Waste-L
waters", JWPCF, Voo. 47, No. 9, pp2281-2291, Aug., 1975
8.) Jeris, J.S., and R.W, Owens, "Pilot Scale High Rate Biological Denitrifi-
cation", JWPCF, Vol. 47, No. 8, pp 2403-2507, Aug., 1975
9.) Savage, J.St,.',and J.J. Chen, "Operating Experiences With Columnar Denitri-
fication", Water Research, Vol. 9, pp 751-757
10.) Davies, T.R., and W.A. Pretorius, "Denitrification With a Bacterial Disc
Unit", Water REsearch, Vol. 9, pp 459-463, 1975
11.) Murphy, K.L-, Sutton, P.M., Wilson, R.W., and B.E. Jank, "Nitrogen. Control:
Design Considerations for Supported Growth Systems", JWPCF, Vol. 49, No. 4,
pp 549-557, April, 1977
12.) Standard Methods for the Examination of Water and Wastewater. 14th
13.)
Edition APHA, AWWA, WPCF, 1975
Blanc, F.C. and O'Shaughnessy, J.C, "Biological Nitrification and Deni-
trification Using Rotating Biological Contactors", Publication No. 97,
Water. REsources Research Center, University of Massachusetts, Amherst,
MA , July, 1978
1300
-------
NITROGEN REMOVAL IN ROTATING BIOLOGICAL CONTACTORS
WITHOUT THE USE OF EXTERNAL CARBON SOURCE
By
Hallvard 0degaard
Associate professor
B«j0rn Rust en
Research assistant
Division of Hydraulic and Sanitary Engineering
The University of Trondheim, 703U Trondheim-NTH, Norway
Int ro duet ion
The main pollution problem resulting from wastewater discharges in
Norway is eutrophication of the inland lakes and the fjords. In order to
fight eutrophication, phosphate removal is now carried out in most treat-
ment plants. The role of nitrogen in eutrophication has been more or less
neglected, and treatment plants for nitrogen removal are non-existant in
Norway. The interest in nitrogen removal is however growing caused by the
fact that nitrogen has been found to be growth limiting nutrient in some
eutrophic lakes.
The interest for nitrogen removal combined with the rising popularity
of the biodisc in Norway, has initiated research at The University of
Trondheim on the use of biodisc-plants for the removal of both organic
matter, phosphorus and nitrogen.
1301
-------
NITBOGEN REMOVAL IN BIODISC-SYSTEMS
The literature on biological nitrogen removal in general is vast, at
least compared to the number of actual plants that have been built through-
out the world. The majority of the papers onnitrification/denitrification,
however, deals with suspended cultures. Most of the denitrification
studies that have been carried out with fixed film cultures, have dealt
with anaerobic filters or lately, fluidized beds. In the last few years
some authors have reported from investigations of denitrification in bio-
disc systems /I, 2, 3, V.
The flow diagram normally used when biodiscs are to be used for
both nitrification and denitrification, is the one shown in Pig. 1.
-te-
ChfeOH
i
' ,. ! l
f
V * )
tt \y
Aerobic
niirifier
Anaerobic Sedimen
denitrifier tation
Fig.
1. Normal process for nitrogen removal
in biodisc plants.
The advantage of this system compared to a suspended.culture system
or to a system with anaerobic filter as denitrifier, is that only one
sedimentation unit is needed.
The main disadvantage of the system, is the need for an additional
carbon source to obtain denitrification. As in other denitrification
systems, methanol would normally be used. There are several disadvantages
with methanol addition:
1. It increases operation costs. The enormous increase in methanol
price the last few years, has made this point very important.
2. It increases investment cost since an advanced control system
is necessary in order t° add the correct amount of methanol.
3. Surplus methanol not used-for denitrification would increase
the effluent BOD. The amounts normally added corresponds to
80-100 mg BOD/I in municipal sewage, which could be detrimental
to the receiving water if denitrification for some reason did
not take place or in case of failure in the control system men-
tioned above.
The goal with our research was to unnecessitate the need for an
external carbon source and use the raw water itself as such. This has
successfully been done in suspended culture systems in the so-called
Bardenpho-process proposed by Barnard /5/> where nitrate-rich mixed liquor
from an activated sludge tank, is recycled to an anaerobic suspended cul-
ture tank ahead of the aerobic one.
In the process presented in this paper, the same principle has been
utilized in a biodisc-system, see Fig. 2.
1302
-------
r
Q
•Qr=rQ CNr
Fig. 2. Proposed process for nitrogen removal in
biodise pfonts.
The ammonium-rich influent passes through a tank with submerged bio-
discs where anoxic conditions will exist. From this tank it enters an
ordinary aerobic biodisc unit where nitrification takes place. Nitrified
water is then recycled to the inlet of the ..anoxic tank where denitrifica-
tion takes place and where raw water can act as carbon source.
The maximum possible nitrogen removal that can be achieved with the
process, may be calculated from a mass balance assuming complete nitri-
fication and denitrification in the two steps:
Removal:
CNi ~ CI
'100% =
and
r+1
100$ =
100%
are nitrogen concentrations in the influent,
where C^. ,
in the effluent and in the recycled flow respectively.
A removal of 90% requires for instance a recycled flow of 9 times the
influent flow, while Q0% is achievable with a recycle ratio of r = k.
EXPERIMENTAL ARRANGEMENT
The experiments were carried out in two plexi-glass biodisc units
arranged in parallel, each with a capacity of 1-10 1/h (Fig. 3).
Experiments with three different types of raw water have been
carried out :
1. Artificial wastewater
2. Municipal wastewater (sewage)
3. Leachate from sanitary landfill
Artificial wastewater
The reason for starting out with artificial wastewater was to ensure
influent characteristics to be as stable as possible in order to investi-
gate the possibilities of the process. The water was composed to be as
similar to municipal wastewater as possible.
1303
-------
Pump
Disc
Disc
Com
Disc
Raw
water
tank
1m3
— C
Pu
in
)
-v^»-
[_
I
mP .
300mm
Denitrifier
diam. 300mm
number 10
>. number 4
area 1 .41 m2
volume 32.81
- 1
v 1
f~ll El. motor \ / .
y V V r
600mm " "200mm
Nitrifier
300mm
20
4
2.82m2
23. 6 I
ional speed 15-16 rev/mi n 15-16rev/min
50mm
Fig. 3. Experimental unit.
Table 1 Composition of artificial waste water
Glucose
Yeast extract
Dried milk
Urea
MHuCl
Na2PO^«12H20
200 ing/1
30 "
120 "
30 "
60 "
100 "
KHC03
NaHC03
MgSO^'THzO
FeSOit-THzO
MnSOit*H20
CaCl2«6H20
Bentonite
50 mg/1
130 "
50 "
2 "
2 "
3 "
Ho "
This water gives the following results when analyzed:
Total COD (CODt)
Soluble COD (CODS)
Soluble BOD (BODss)
Tot P
Tot N
Suspended solids
Alkalinity
mg 0/1
300 "
188
9 mg P/l
36 mg W/1
90 mg/1
130 mg CaC03/l
Sewage
The sewage was taken from an apartment house during morning hours in
order to have high ammonium concentrations. The water showed, however,
after presettling, to have a relatively low soluble organic content
(COD££ltereJ ~ 100-l60 mg O/l) compared to the ammonium content (NH^-U ~
35-65 mg H/l). As will be commented later, 'this gave us some problems
since COD became the limiting factor for denitrification. In part of the
experiments this relationship between COD and N was corrected for by in-
troducing dairy effluent (dried milk) to the sewage.
1304
-------
Leachate from Landfill
The high amount of ammonium in leachate may cause both oxygen defi-
cit and eutrophication in the receiving vaters, and a couple of such
situations in Norway have intensified research on treatment of leachate
from landfills.
A review through the literature /6/ will show that treatment of
leachate is very troublesome, and that anaerobic treatment has been used
with some success. ¥e found it interesting, therefore, to test the pro-
posed process on leachate.
The leachate used in this experiment was that of the landfill of
Trondheim. The experiments were carried out during a period of heavy rain-
fall that diluted the leachate. Table 2 shows composition of the leachate
during 1975 and 1976 compared with the composition during our experiment.
Table 2
Composition of leachate
Parameter
COD mg/1
Tot N
NHzj-N
Tot P
Fe
PH
1975 - 76
1250 - 1650
80 - 100
65 - 80
0.7 - 0.8
105 - lUO
6.8
Actual experiment
UOO - 800
80 - 130
70 - 125
0.6
100 - 1U5
7.0
The difference is especially significant for COD while the differ-
ence for the other parameters are minor.
In order to prevent the problems with ironhydroxide-precipitation
on the biofilm, most of the iron was precipitated in advance by adjusting
the pH to 8, aeration and filtration.
1305
-------
RESULTS
Hitrogen removal
In Fig. U are shown experimental results from situations where nitri-
fication and denitrification limited "by any factor. The results
are compared with the theoretically possible removal as mentioned earlier.
The figure shows that the results fit the theoretical removal quite
nicely. This demonstrates that given the situation that no factor (such
as pH, Oz-concentration, BOD-loading, toxic substances, and so on) limits
nitrification or denitrification, the proposed process will work as anti-
cipated. Concequently, the process is an interesting alternative to the
different nitrogen removal processes used today.
Artificial wastewater
Sewage
TTT100%
Fig. 4. Nitrogen removal against resirculation ratio.
Nitrification
Several factors are influencing the degree of nitrification, such
as content of biodegradable organic matter, oxygen, PH and temperature.
Most workers agree that the oxygen content in the water may be a
limiting factor with regard to the nitrification rate. Weng and Molof
HI found that nitrification took place only if more than 2 mg 02/1 was
present. In our experiments the oxygen content was normally in the
range of 3-6 mg 02/1 and never below 2.5 mg 02/1 in the nitrifier.
Optimum pH for nitrification has been found to be around 8 /8/ but
nitrification is achieved in the range of pH = 6.5-9.5. In all our ex-
periments the pH was in the range of 7.0-8.3.
Optimum temperature for nitrification has been found to be in th<=
range of 28 - 36 °C /9/ where municipal treatment processes seldom operate
Antonie /2/ has postulated that the nitrification rate is constant over
/10/ fJS! d!!Crease~ ^ daereasiag temperature under that. Pretorius
/10/ found Q10 (= U2o/Uio)'=1.25 in his biodisc experiments.
1306
-------
No temperature control was used in the experiments presented here.
During the artificial waste water period, the temperature was in the
range of 13-16 °C, while it varied more in the sewage and the leachate
period (10-16 °C).
In order to achieve nitrification, the content of biodegradable
matter has to be brought down sufficiently.
Prior to the experiments with the proposed treatment process (Fig.2),
some experiments with the process shown in Fig. 1 were performed in order
to investigate the acceptable organic loading to achieve nitrification.
These experiments were carried out with the artificial wastewater.
Fig. 5 shows the relationship between the degree of nitrification
(per cent removal of HH^-N)- and the content of soluble organic matter
•'filtered) ^n "fc^ie effluent.
NH4-N
% Removed
100
80
60
40
20-
. 2-
10 —
Artificial
wastewater
0 20 40 60 80
mgCODs/t
Fig. 5. Effect of soluble COD on
the degree of nitrification.
It is demonstrated that a soluble COD-content of less than 30 mg
COD/1 was needed to ensure complete nitrification. Antonie /2/ reported
that nitrification started when the BOD-concentration was brought down to
30 mg BODs/1, while it was fully developed below 8-10 mg BOD5/I. Weng
and Molof /7/ found that nitrification occurred at COD-concentrations
below 50 mg COD/1. Both of these findings correspond to a soluble COD-
concentration of about 25 - 35 mg COD/1 and are in agreement with our
results.
In Fig. 6 is plotted the degree of nitrification against organic
loading of soluble COD per unit area of biofilm. It is demonstrated that
the load had to be kept below 25 g CODs/m2-d to ensure full nitrification.
1307
-------
%
100
80
60
40-
20-
NH4- N
removed
Artificial
wastewater
0 25 50 75 100
gCODs/m2d
Fig. 6. Effect of organic loading
on the degree of nitrification.
To evaluate the results on nitrification in the experiments with
the proposed anoxic-aerobic process is a little complicated because the
recirculated water contributes both to the COD-loading and the ammonium-
loading. The COD that enters the nitrifier is probably composed of one
easy "biodegradable fraction from the raw water and one heavy bio-
degradable fraction from the recirculated water.
In the sewage experiment the BODs in the recirculated water for one
run was found to be as low as 3.5 mg BOD5/I, while the COD was 39 mg
COD/1. It may thus be stated that the organic load from the recirculated
water does not contribute much to the total biodegradable organic load
that will determine the degree of nitrification.
In the experiments presented in Figs. 5 and 6, recirculation was
not used. In the experiments with the proposed process higher CODS-
values could be tolerated to ensure nitrification. This is partly due
to the fact that non-biodegradable COD is recirculated. In the leachate
experiments, nitrification took place even at COD-values of 150-200 mg
COD/1.
In Fig. 7 is plotted the nitrification rate against ammonium-loading
for situations where nitrification took place (was not limited by organic
loading). The spread in the results may be explained by the lack of
temperature control.
It is interesting to note that for all the types of raw water tested,
there seems to be a critical loading of aboxrt 100-150 mg UHit-N/m2-h
over which the nitrification rate is at its maximum and independent
upon the loading. This maximum nitrification rate is about 100-120 mg
NHi»-Nremoved/m2 *k "both for the artificial wastewater and the sewage,while
1308
-------
it is considerably lower for the leachate, around ^0 mg NHit-Nremoved
NH4 - N removed
NHL - N removed
2
i
120
80-
40-
n
mg/m • h ^
*
,»_ . 120
/* •
£ • •
/ 80-
/ 40-
/ Artificial wastewater
' ..... , . — , — te. n
mg/rri • h
•
• •
«
^ A
/" • ' '•
/ •
/•*
/
/ Sewage
1 1 , , 1 , , ^B-
0
TOO
200 300
TOO 200 300
NH4 - N applied, mg/m2- h NH4 - N applied, mg/m2-h
NH4 - N removed
Jlmg/m • h
40-
20-
>
f
Leachate
fc~_
0 200 4.00 600
NIH4 - N applied, mg/m - h
Fig. 7. Nitrification rate against ammonium-loading.
In Fig. 8 is shown the nitrification rate in each step of the unit,
which may be assumed to be complete mixed.
Even though the spread in the results are great, it may be concluded
that the nitrification process follows Monod-kinetics, with zero-order
kinetics at higher ammonium concentrations and first order, at lower con-
centrations. It seems that zero order kinetics governs the process at
concentrations higher than k—6 mg HHi»-lT/l in normal sewages and higher
than 20 mg NH^-N/1 in landfill leachate.
1309
-------
120.
80
40
NhL- N removed
mg/m • h
Artificial wastewater
048
- N removed
12 16 20
mg NH4 - N/l
24
mg/
m
120-
80-
40-
t
NH4 - N removed
mg/m • h
40-
20-
/ Sewage
First comp. only
- «
Leachate
First comp. only
8
12
mg NH. - N/l
1 1 1 1 1 r
20 40 60
mg NH. - N/l
Fig. 8. Nitrification rate against ammonium-concentration.
In Table 3 the nitrification rates found in this study is compared
to what has been found in other studies.
Table 3
Nitrification rates
Raw water
Nitrification rate
Reference
Sewage with
addition
Municipal sewage
Artificial wastewater
Municipal sewage
Landfill leachate
77-5 mg NH\-N/m2'h
bb mg TKN/m2-h
max. 120 mg
max. 110 mg
max. UP mg N£U-N/m2-h
Pretorius /10/
Murphy et al. /ll/
This study
1310
-------
Denitrification
Several factors are influencing denitrification, such as oxygen con-
centration, pH, temperature, carbon source, nitrate-loading and ratio
between loading of organic matter and nitrate.
The denitrification has to be carried out at anoxic conditions. True
anaerobic conditions in the liquid does not seem to be necessary /12,U/.
In the experiments presented here, conditions were normally anaerobic
but denitrification took place even at oxygen-concentrations in the water
around 1 mg Oz/l. It was not necessary to cover the free water surface
to the atmosphere.
Optimum pH for denitrification seems to be pH = 7~7-5 /8/. In our
experiments the pH varies between 7 and 8 in the wastewater experiments
and between 8, and 8.5 in the leachate experiments.
Methanol is normally used as carbon source. Several authors have
investigated nitrate/nitrite-removal as a function of methanol consump-
tion for fixed film systems. Christensen and Harremoes /12/ reviewed
the different investigations, and concluded that empirically, a methanol
to nitrogen ratio of 3 g CH3OH/g N03-I!T was necessary to reach more than
90% nitrate removal.
In suspended cultures the theoretical methanol consumption would
be 2.U7 g CH3OH/g TO3-N /13/, which is close to what has been measured
by many authors /12/.
The most interesting alternative to methanol is undoubtedly the raw
sewage itself. Warkis, Rebhun and Sheindorf /lU/ have recently published
results from suspended culture experiments where different carbon sources
(methanol, sodium acetate and chemically treated raw sewage) were used.
They concluded that by expressing the organic matter as soluble BOD, a
critical value of 2.3 mg BOD/mg NOX-U existed when 100$ denitrifica-
tion was to be reached regardless of what carbon source was used.
In the experiments presented here, soluble COD was, for different
reasons, used as a measure for the organic matter. In the artificial
wastewater experiments the COD-concentration did not seem to be limiting
for denitrification. In the municipal wastewater experiments, sewage from
morning hours was used. This had an abnormal high nitrogen concentration
compared to the soluble COD-concentration. Complete denitrification was
consequently not obtained. In Fig. 9 the nitrate removal is plotted
against the COD/NOX-W ratio (COD in raw water and total HOx-W applied).
It is demonstrated that almost complete denitrification was obtained
when the ratio of soluble COD in the raw water to the ETOX-N concentration
into the anaerobic unit was over 7, and that the relationship -between the
removal and this ratio was linear for ratios lower than J.
This made us add dried milk corresponding to a soluble COD addition
of 100 mg CODS/1 in the proceeding experiments. (Bio-disc systems are,
by the way, normally used in Norway when dairy effluents are treated to-
gether with municipal sewage.)
The fact that the influent COD is reduced through the anoxic step
is another valuable aspect about the proposed process. Reduced organic
loading on the aerobic step makes it possible to reduce the necessary
area in this step in order to achieve the required COD to obtain nitrifi-
cation. Altogether, this means lower investment cost compared to a situa-
tion were the denitrifier is proceeding the nitrifier.
1311
-------
nN°x~ Nremoved
100
60
20
Sewage
—i 1 1 1 1 1 1
2 6 10 14
mg COD inf j. /mg NOX - N
Fig. 9. Degree of denitrification
against COD/NOX-N
ratio.
In order to find how much COD that was consumed by denitrification,
the^ratio between CODCOnsumed and (NOx-N)consumed was calculated. This
ratio should be expected to be constant for the different runs. In order
to visualize the results the ratio is plotted against recirculation ratio.
Even if there is^a considerable spread in the results, Fig. 10 shows that
2.5-3.5 as COD is consumed for each mg WOX-H removed by denitrification.
maCODremoved
o
•H
4J
4-
2-
mgNOx - Nremoved
Sewage
•8
Fig. 10. COD consumed during
denitrification.
In Fig. 11 are plotted the NOX-N removal (%) versus HOX-W loading
for the different waters.
In the artificial wastewater experiment, a maximum load of k g
NOx-N/m2.d could be tolerated in order to achieve more than 90$ denitri-
fication. The corresponding load in the sewage experiment seemed to be
a little bit higher, 6 g NOx-W/m2'd.
1312
-------
100-
60-
NOX - N ,
1 removed 100-
...T*«V so
<
•»«
i
** * vs
24*
20-
0
Artificial wastewater
8 12 16 20
gNOx- N/m2-d
%n
100-
60-
20-
NOV- N
J\
removed
NOX- N
i i removed
Sewage
100
80
60-}
* Leachate
0
8 12 16
g NOX - N/m2- d
0 24 6
g NOX - N/m2-d
Fig. 11. Degree of denifrification against NOX - N - loading.
In the leachate experiment more than 9®% denitrification was
difficult to achieve. This might tie caused by some unknown limitation.
To achieve Q0% denitrification a maximum load of 2 g !TOx-lT/m2*d could
be used.
In Fig. 12 is plotted the denitrification rate (mg NOx-lSremc,-v-ed./m2' •&)
against the NOX-N load (mg NOx-Wapplied/m2-h.) .
It is demonstrated that denitrification rate increases linearly with
the load. In the sewage and the leachate experiments it is difficult to
state what is maximum rate. In the sewage experiment it seems to be in
the range of 350 - U^O mg NOx-N/m2-h.
In the artificial wastewater experiment the maximum value is more
pronounced at a value of 150 mg HOx-W/m2.h.
The reason for this discrepancy is difficult to state, but we believe
that the denitrification rate in the artificial wastewater experiments was
limited by nitrate concentration at high denitrification rates.
Watanabe and Ishiguro /3/ found no maximum denitrification rate
even if they measured rates as high as 69^- mg N03-N/m2-h, while Cheung
reported maximum denitrification rate of 676 mg NOx-N/m2-h.
The denitrification rates foxm£ in this experiments are compared to
what is found in other experiments ia Table U.
1313
-------
NC^ -N removed
NOX - N removed
t
200-
100-
0
tmg/m
«
;r
.h \
• 200.
' * * 100-
Artificial wastewater
— , k. 0
^ mg/m -h
S
/
f Leachate
w -- "^
0 200 400 600 800
mg/m -h
100 300
mg/m2-h
I
400-
300
200^
TOO
NCL -N removed
mg/m2 h
Sewage
Table k
0 200 400 600 800 1000
NOX - N applied rng/m • h
Fig. 12. Denitrification rate against NOX - N loading.
Denitrification rates
Raw water
Trickling filter effluent
with KNO 3 -addition
Effluent from biological
treatment with NaNOa-
addition
Nitrified sewage with
KNO a-addit ion
Artificial wastewater
Sewage
Landfill leachate
Carbon
source
Methanol
Methanol
Methanol
Raw water
Raw water
Raw water
Denitrification rate
max. 260 mg N03.-N/m2»h
max. 676 mg NOx-N/m2*h
i' 69^ mg N03-N/m2«h
max. 150-200
mg HOx-W/m2-h
max.350-U50
mg WOx-N/m2-h
1 ) 150-200
mg WOx-N/m2-h
Reference
Da-vies and
Pretorius /!/
Cheung /h/
Watanabe and
Ishiguro /3/
This study
it H
ti it
1) Not maximum level.
1314
-------
_
DISCUSSION
The results of these experiments show that the proposed process
should be an interesting alternative when nitrogen is to be removed from
wastewater.
The advantages of the process are:
1. Low operation cost since external carbon source is omitted
2_ Low capital cost since only one sedimentation unit is needed,
and since the aerobic biodisc can be reduced as a conse-
quence of organic matter being consumed during denitrification
in the anaerobic step.
3. No danger of destroying the effluent BOD as a consequence of
excessive methanol addition.
The possible disadvantages are:
1. The ratio between soluble COD loading (g CODs/h) on the denitri-
fier from the influent and the NOX-H loading (g UOx-lT/h) on the
denitrifier should be £ 7.
2. High total nitrogen removals (>90$) requires a high recircula-
tion ratio (QjR/Q > 9), 80$ removal is however reached at a
QR/Q of k. Recirculation may however be arranged very cheaply
by a bucket-type pump mounted on the biodisc shaft.
3. If nitrification for some reason is inhibited, the ammonium
loading on the nitrifier increases rapidly as a consequence of
recirculation. This accelerates the reduction in nitrification.
If there are signs that nitrification has been reduced, the
recirculation ratio should be lowered until the nitrification
is brought back to the normal state before recirculation is
brought back to normal.
When designing a plant with the proposed process, three factors are
to be determined, the recirculation ratio, the biodisc area for the nitri-
fier and the biodisc area for the denitrifier.
The recirculation ratio is determined by the percentage of nitrogen
removal that is needed after the equation:
•100$
One should be conservative and use a bit higher recirculation ratio than
that is found by this equation to account for the possibility that nitri-
fication is not 100$.
The area of the aerobic nitrifier should be designed after normal
design criteria in order to obtain nitrification on biodiscs. The organic
load, however, may be reduced by 2.5 g COD for each g NOX-N that is to be
removed in the denitrifier. In normal sewage this would mean a reduction
in organic load corresponding to a reduction in the concentration of the
influent of 70-80 g COD/m3 or 50-60 g BODs/m3 (soluble organic matter).
It is, however, important that the nitrifier is conservatively designed
in order to ensure full nitrification. There is no need to add anything
to the organic load from the recirculated water because this water has a
very low concentration of biodegradable organic matter since the water
already has passed the nitrification step.
1315
-------
The area of the denitrifier should "be designed for a NOX-N load
of about k g NOx-N/m2-d to ensure full denitrification.
When these guidelines are followed, one would normally find that
the necessary area for the denitrifier is less than half of the area of
the nitrifier.
CONCLUSIONS
Nitrification/denitrification experiments with an anaerobic-aerobic
biodisc system with recirculation have been carried out using artificial
wastewater, sewage and landfill leachate as raw water. The following con-
clusions can be made:
1. It has been demonstrated that the proposed process (Fig. 2) are
able to give nitrogen removals in sewage corresponding to
RN = (r/(r+l))-100$ (r = recirculation ratio) without the use
of an external carbon source.
2. Nitrification rates of 120, 110 and ho mg NHtf-N/m2.h have been
measured for the artificial wastewater, the sewage and the land-
fill leachate respectively.
3. Denitrification rates of 150-200, 350-^50 and 150-200 mg NOx-M/m2-h
have been measured for the artificial wastewater, the sewage and
the landfill leachate respectively.
k. It has been found that the consumption of organic matter during
denitrification is near 3 g COD/g WOx-N removed.
5. In order to achieve full denitrification by the process, it has
been found that the ratio between soluble COD-loading (g CODs/h)
on the denitrifier from the influent and the NOX-N loading
(g NOx-N/h) on the denitrifier should be ^ 7- This corresponds
to a CODs/N-ratio of 6-7 in the raw water.
6. In addition to the possibility of the process to treat sewage,
the experiments also demonstrated that landfill leachate can be
treated by the process. It is, however, recommended that iron
in this type of water is removed prior to the biodisc process
in order to avoid ironhydroxide precipitation on the biofilm.
7. The main advantage of the process is that addition of external
carbon source is omitted and that the process would be economic-
ally favourable both from an investment and an operation view-
point .
ACKNOWLEDGEMENT
The authors wish to thank The Royal Norwegian Council for Scientific
and Industrial Research for financial support. Also we want to thank
Jostein Skjefstad, G0ril Thorvaldsen, Bente Storebraten, Solfrid Rotstigen,
Bj0rn Strand and Ove Rust en for taking part in the laboratory work.
1316
-------
REFERENCES
/!/ Davies, T.R. and Pretorius, W.A.:
Denitrification with a bacterial disc unit.
Water Research, Vol. 9, pp. ^59-^3, 1975-
/2/ Antonie, R.L. :
Fixed biological surfaces - wastewater treatment.
CRC Press, Inc., 1976.
/3/ Watanabe, Y. and Ishiguro, M.:
Denitrification kinetics in a submerged rotating biological
disc unit.
Progress of Water Technology, Vol. 10, No. 5/6, pp.187-195,1978.
A/ Cheung, P.S.:
Biological denitrification in the rotating disc system.
Water Pollution Control, Vol. 79, No. 3, pp. 395-^08, 1979-
Barnard, J.L.:
Biological denitrification.
Water Pollution Control, Vol. 72,' No. 6, pp. 705-720, 1973.
'Robinson, H,0. and Maris, P.J.:
Leachate from domestic waste: Generation, composition and
treatment. A review.
Technical report TR 108, Water Research Centre, England, 1979.
Weng, C. and Molof, A.H.:
Nitrification in the biological fixed-film rotating disc system.
Journal of Water Pollution Control Federation, Vol. h6, No. 7,
pp. 167^-1685, 197^.
Environmental Protection Agency:
Process design Manual for nitrogen control.
Office of Technology Transfer, EPA, Cincinnati, USA, 1975.
Sharma, B. and Ahlert, R.C.:
Nitrification and nitrogen removal.
Water Research, Vol. 11, No. 10, pp. 897~925, 1977-
Pretorius, W.A.:
Nitrification on the rotating disc unit.
Progress of Water Technology, Vol. 7, No. 1, pp. 2U3-2U9, 1975.
Murphy, K.L., Sutton, P.M., Wilson, R.W. and Jank, B.E.:
Nitrogen control: design considerations for supported growth
systems. Journal of Water Pollution Control Federation,
Vol. "1*9, PP- 5^9-557, 1977.
Christensen, M.H. and Harremoes, P.:
Biological denitrification of sewage: A literature review.
Progress of Water Technology, Vol. 8, No.U/5, pp.509-555, 1977.
Me Carty, P.L., Beck, L. and Amant, P.St.:
Biological denitrification of wastewaters by addition of
organic materials. Proc. of the 2Vth Ind. Waste Conf.,
Purdue University, 1969.
Narkis-, N. , Rebhun, M. and Sheindorf, Ch. :
Denitrification at various carbon to nitrogen ratios.
Water research, Vol. 13, No. 1, pp. 93-98, 1979.
Ill
/8/
/9/
/10/
/ll/
/12/
/13/
1317
-------
-------
NITRIFICATIOW/DENTTRIFICATION STUDIES
WITH ROTATING BIOLOGICAL CONTACTORS '
By
Allan G. Smith
and
R. K. Khettry
Ontario Ministry of The Environment
1.0 Introduction
Considerable research has been conducted over the past ten years
on the use of "biological methods for the removal of nitrogen from waste-
water. Variations of the activated sludge (A/S) process have "been
utilized to first achieve nitrification (oxidation of ammonia to nitrate),
then denitrification (the reduction of nitrate to nitrogen gas).
Nitrification/denitrification systems using the activated sludge
principal consumes considerable energy and demands close operator attention.
With the current demands for nitrification by the Ministry of
Environment (MOE) in some areas of southern Ontario (l) an alternative
to the activated sludge system has been sought. The Rotating Biological
Contactor (RBC) offers less energy consumption and operator attention
than A/S systems and studies have shown comparable consistuent removals
(2). As a result, following pilot-studies, a full-scale RBC section has
been installed for nitrification of secondary effluent at the Guelph
Ontario WPCP.
1319
-------
The requirements for nitrification in Ontario are all site-specific.
Generally, the concerns are the nitrogenous oxygen demand upon the
receiving stream (mainly river systems) and the toxic effect of ammonia
(in the unionized form) upon aquatic life.
There are some denitrification facilities treating high nitrate
industrial wastewater, but to the author's knowledge, only one municipal
plant (i.e. Penticton, B.C.) is currently being planned in Canada.
Hitrogen removal from domestic wastewater is at the moment of low priority, :
trailing recent increased emphasis on better phosphorus removal.
Currently all of the municipal treatment plants expanded for nitrogen
control' in Ontario, will only include nitrification. If a step is required
to reduce effluent nitrate (again site specific), for additional nutrient
removal, a tertiary denitrification system run on nitrified secondary
effluent would be a logical choice.
1.1 Past Studies
Investigations into the application of the RBC for tertiary de-*
nitrification have not been widely reported, but there are a few worth
noting.
In 197^ Davies and Pretorius (3) published work done with an RBC
unit converted for denitrification. By sealing the top of the unit,
oxygen transfer to the discs was reduced creating an anoxic atmosphere
conclusive to denitrification. The discs were rotated at 6 rpm and sub-
merged to k5% of their area.
Parameters such as pH, temperature and methanol addition were
examined as to possible effects on RBC denitrification efficiency. An
optimum pH range of 7 to 8.5 for denitrification was found with Qo% of
maximum nitrate removal obtained at pH 6, and 70% at pH 9. An optimum
temperature range of 10-30°C was reported and a methanol to nitrate
(C/N) ratio of 2.6/1 was found adequate for complete denitrification. '.
Davies and Pretorius (3) commented, that a larger volume of liquid
could "be treated with the discs fully submerged, but corrosion of bearings :
can "be a problem. Partially submerged discs would have more unit airspace
and allow better release of nitrogen gas from the discs bio-growth.
More recent studies by Soyupak (h) (Master of Engineering Thesis),
covered areas such as the effects of hydraulics and temperature on de- \
nitrification reactions within a totally submergenced disc unit. Pilot-
plant size units with disc areas of 28.km. (306 ft ) were tested at flow-
rates of 2.5 L/min. (0.6 Igpm) and k L/min. (0.9 Igpm) with the units
operated single-stage or h stages in series. Hydraulic responses with
the addition of dye at U different temperature levels were monitored after
establishing biological growth on the discs.
Methanol and influent nitrate concentration were held constant at
a C/N ratio of 1.5/1.
1320
-------
Out of this work, the results showed that each compartment (stage)
of the RBC reacted (hydraulically) similiar to two Continuously Stirred
Tank Reactors (CSTR) in series with interchanging flow.
Denitrification reaction rates were found to "be zero order for
the level of influent oxidized nitrogen in these studies.
The temperature dependency for denitrification using the RBC
appeared not to be influenced by the hydraulic loadings studied. However
denitrification rates were more sensitive to temperature variations than
a CSTR unit or packed column reactors under similiar operating conditions.
Comparisons made on a volumetric bases, showed that the RBC
produced higher denitrification rates than CSTR with a mixed liquor
suspended solids (MLSS) range between 1500 and 2500 mg/L, or a packed
column.
Pilot-plant studies by Antonie published in a text book on RBC's
(5), indicated the possible nitrate removal efficiencies attainable at
various hydraulic loadings. For example, using fully submerged discs at
a wastewater temperature of >_ 13 C (55°F) and nitrate concentrations up
to 25 mg/L, 96% nitrate removal were predicted at a hydraulic rate of
0.15 m3/m2/d (3 gpd/ft2). Reduced nitrate removals down to 50% can be
expected when increasing the hydraulic loading to O.Up m3/m2/d (10 gpd/ft2)
(See Figure l). Methanol addition was implemented in these experiments,
but the amount was not stated. However, the author did suggest that a
floating cover could be placed on the water surface over the discs, to
restrict oxygen transfer, thereby reducing methanol requirements.
After reviewing the literature it was felt that more specific
information was required; particularly on a combined nitrification-
denitrification RBC system.
2.0 Objectives of Study
A pilot-study was formulated and implemented to gain first-hand
information as to the viability of an RBC system for nitrogen removal.
The study was conducted at the Ministry of Environment Ontario Experimental
Facility (O.E.F.) in Brampton.
of:
The particular objectives of this project were to study the effects
(a) a nitrification RBC module operation on ultimate performance
of a denitrification RBC module.
(b) submergence of the denitrification module discs on nitrate
removal efficiency.
1321
-------
I ID
H
o
a
o
•H
IS
o
•H
fi
4J
ti
<0
•H
fr C!
c q
* I
3 M
0>
O 4J
•rl IH
H nJ
§ -O
S-4 Q)
•d o
& €
I
o
o
VD
1
(1)
I1
(U
EH
I CM
H
00
s
tn
C
O
•H
o
o
o
CO
O
O
CM
N-
1322
-------
3.0 Experimental Setup And Procedures
3.1 Pilot-Plant Description
The pilot-plant consisted of two complete RBC modules arranged in
series (Figure 2). The first module for nitrification had a total disc
area of 156 m2 (l680 ft2) and comprised of four equally sized stages
(compartments). The plastic discs were driven by an electrical variable
speed motor, capable of delivering up to 9 rpm.
A portion of un-nitrified secondary effluent from a fullscale
activated sludge process at O.E.F., was used as feed for the nitrification
module. Previous studies by Hewitt of this Ministry (6) showed that
0.12 m3/m^/d (2.k Igpd/ft2) would produce full nitrification during winter
operation. An hydraulic loading of 10 L/min. (2.2 Igpm) approximating
this value, for the nitrification RBC module, was selected for this study.
The module was fed to give a plug-flow (k stages in series) mode of
operation.
The second disc module for denitrification had a total disc area
of 23 m2 (250 ft ) and received a portion of nitrified effluent, from the
first RBC module at feedrates between 2.5 and 5.2 L/min. (0.56 to 1.2 Igpm)
These flowrates produced a hydraulic to surface loading range of between
O.l6 to 0.3^ m3/m2/d (3.2 to 7 Igpd/ft2) to a total of h stages in series.
Prior to the study, alterations were made to the module effluent
piping to facilitate control of liquid level; from k^% to full disc
submergence. For the same purpose, the walls of the module were also
extended, the disc drive-shaft lengthened, and new bearings with snuffing
boxes were installed to prevent leaks. The disc rotation speed was fixed
at 6 rpm.
Sampling taps were installed along the base of the module, one at
each RBC stage. A clear plastic cover was also placed over the top of
the unit to reduce oxygen transfer to the discs, but to allow visual
inspection.
Methanol was pumped to the first stage of the denitrification
module at a predetermined concentration ratio to influent nitrate -Rf
(C/ET ratio) of 3 to 1. This ratio was calculated using a formula
developed by McCarty (7) which is stated as follows:
Cm = 2.U7 W + 1.53 W. + 0.87 D
where Cm = mg/L methanol
N. = NO -W concentration mg/L
N. = UOp-BT concentration mg/L
D = influent dissolved oxygen concentration mg/L
An influent nitrate -N of 15 mg/L, nitrite -N of 1 mg/L and DO of
8 mg/L were assumed for this study. Using these values in the above
1323
-------
o
§
Q)
rH
10
-------
TABLE 1
OEP RBC STUDY SCHEDULE
DENITRIFICATION
RUN # DATE
W.W.T. C FEEDRATE L/MIN. MODE C/N RATIO
1
2
3
4
5
6
7
8
9
Oct. 2 26/78
Nov. 1-Dec. 15/78
Jan. 2-Feb. 9/79
Feb. 12-19/79
Feb. 19 -Mar. 19/79
Apr. 19~May 2/79
May 7-June 8/79
June 11- June 22/79
June 2 5- July 11/79
Specialized Run involving
19
17
11
7
7
11
14
15
16
variation
3
5
3
5
5
5
5
5
2.5
of nitrification
Sub.
Sub.
Semi-
Sub
Semi-
Sub
Semi-
Sub
Seini-
Sub
Sub
Semi-
Sub
Semi-
Sub
module rpm
2.3/1
2.0/1
1.6/1
1.9/1
2. .6/1
3.2/1
3.7/1
3.4/1
3.4/1
through
*
a total of 5 individual tests.
1325
-------
equation a methanol dosage of
was required.
mg/L (C/W of 3/1) to the influent flow,
The nitrification and denitrification RBC modules were housed in
an insulated trailer heated by a small electrical heater, to prevent
freezing of effluent lines and samples .
3.2
Study Schedule
A series of experimental runs were conducted from October 1978
to July 1979. A total of 9 runs were completed in which the nitrate
removal efficiency of the denitrification RBC module was compared under
semi-and total disc submergence, through naturally varying wastewater
temperatures.
Disc rotational speed in the nitrification module was fixed at
U rpm for runs 1 to U and 2 rpm in runs 6 to 9•
More detailed investigation was given to the nitrification RBC
module during run #5• The module was operated at 1 to U rpm to determine
the effects of disc rotational speed on module DO and resultant nitri-
fication. The effects of the last stage DO (nitrification module) level
on nitrate removal in the denitrification module was also studied.
Half-way through the study, the methanol stock was found to
be 20% lower concentration then expected. Consequently, for the first
four runs methanol addition was 1/3 lower-than the requirement.
A schedule showing modes of operation for the RBC study and
corrected C/ET ratio is contained in Table 1. The study was primarily
formulated on the operation of the denitrification RBC module.
3.3 Analyses And Tests
Twenty-four hour composite samples were taken (min. 3 days/wk.)
of nitrification module influent and effluent and denitrification
effluent. Allyl-thiourea (ATU) was added to each sample container to
a concentration of 1 mg/L, as a preservative for nitrification.
The following analyses were performed:
- Five-day Biological Oxygen Demand (BOD )
- Suspended Solids (SS)
- Total Kjeldahl, ammonia, nitrite and nitrate nitrogens
(TKN, NH^-W, NO -N and WO -W)
- Chemical Oxygen Demand (COD)
- Alkalinity as CaCO~
1326
-------
Onsite tests routinely consisted of stage by stage measurement of
DO and temperature twice per day. These measurements were taken more
frequently during intensified work in run 55 and continued into runs
6 to 9.
Dissolved oxygen measurements on the denitrification module were
conducted, by first evacuating a 1 Litre flask(containing a DO probe)
with helium, and then introducing wastewater from the sampling taps on
the side of the unit.
U.O Discussion of Results
h.l Nitrification EEC Module
(a) Runs 1 to 5
For the most part, complete oxidation of ammonia was accomplished
in the nitrification module, producing a mean effluent ammonia of < 1 mg/L,
at a mean influent ammonia concentration of ih mg/L. (See Table 2).
The resulting range of hydraulic loading to the nitrification module for
the study was 0.076 to 0.102 m3/m2/d (1.5 to 2.0 Igpd/ft2).
Exceptions to this were shown in runs 3 and 5. In run 3, the
effluent ammonia concentration rose slightly to approximately 1.3 mg/L
and coincided with a 6°C drop in temperature. The highest influent
ammonia to disc surface loading (1.68 gNHK-N/m2/d) also occurred during
this run (Table 2).
As stated earlier, more intensive work was done on the nitrification
module in run 5- Routine DO measurements (during runs 1 to k) in the
nitrification module showed levels often exceeding 8 mg/L and at times as
high as 11 mg/L in the last two stages. These DO levels were higher than
expected, and in turn introduced more demand for organic carbon, then
orginally accounted for.
Consequently run 5 was formulated to see what effects various
disc rpm had on nitrification efficiency, module DO level and ultimate
nitrate removal in the second module.
Run 5 produced the highest mean effluent ammonia (2.1 mg/L) with
an approximate 10% drop in removal efficiency resulting from reduced
disc rpm. Surveys showed a marked difference in DO concentration in
various RBC stages upon altering the disc rpm. A mean RBC DO of 3.6 mg/L
resulted at 1 rpm compared to 8.7 mg/L at h rpm (See Table 3).
Ammonia oxidation, at liquid temperatures ranging from 6.5 to
to 8.7°C, improved as the disc rpm was raised. Effluent ammonia -W ranged
between 2-5 and 2.7 mg/L at 1 and 2 rpm, but fell below 1 mg/L at 3 and
k rpm.
1327
-------
I
1
H
fe
a
H
H
•d
CO
CTl
CO
ID
CN
co
CO
oo
CO
ft
in
CTl
1328
-------
TABLE 3
SUMMARY OF NITRIFICATION RBC
DO PATTERNS AND EFFICIENCY
(Mean Values) Run 5
Disc RPM And Test #
Flow 1/min
Igpm
Igpd/ft2
o
Temp C
RBC DO
Stage #1
(mg/L)
#2
#3
#4
NH -N
Influent g/d
Removal g/d
Removal %
Effl. con.
mg/L
••
7.9
1,85
1.54
8,1
4.4
3.2
3.3
3.5
205
174
85
2.7
(a) (b)
8,3 9.2
1.84 2.04
1,53 1.70
7.6 7.1
6.9 8.9
5,5 7.6
5.2 7.1
6,3 6.9
191 172
160 139
84 81
2.6 2.5
8,2
1.82
1.52
6.5
8.6
7,8
8.5
9,5
154
146
95
0.6
7,3
1.62
1,35
8.7
8.8
8.0
8.1
10,0
168
160
95
0.8
1329
-------
The tests at 2 rpm were carried out for a longer duration and
divided into two sub-programs (a) and (b), to determine the effects of
disc mass loading. At similiar rpm and temperature, little difference
in nitrification was noted between a disc ammonia loading of l6d g/d
and 139 g/d (Table U).
(b) Runs 6 to 9
Following run 55 the disc rpm was maintained at 2 for the remainder
of the study. It was felt that the expected rise in temperature during
the coming spring and summer would compensate for any loss of nitrifica-
tion at a lower disc rpm. The analytical results from runs 6 to 9 show
consistent effluent ammonia of less-than 1 mg/L, at mean wastewater
temperatures ranging from 11 to l6°C (Table 2). These results were
achieved at a mean RBC DO ranging from U.3 to 7-3 mg/L. The effect of
reduced oxygen demand was shown in run 7 (Table 2) when the influent
ammonia concentration dropped to 6.7 mg/L, coincidental with a low
BOD5 of 9 mg/L. As a result, the mean RBC DO rose approximately 2.0
mg/L above that of runs 6, 8 and 9-
l».2
Denitrification EEC Module
(a) Run 1 to
As stated earlier, resultant C/W ratio in the first H runs was
lower-than predicted because of a dilute methanol supply. For this
reason the initial runs were only compared, on the basis of denitrifica
tion efficiency, with each other.
A drop in nitrate removal from 73% to 31% occurred when the
feedrate of the denitrification module, at full disc submergence, was
increased from 3 to 5 L/min. (Runs 1 and 2). On a weight to disc area
basis, the nitrate loading rates for run 1 and 2 were 2.68 and U.87
g N03-N/m2/d respectively (Table _5_) .
A further drop in nitrate removal resulted in runs 3 and U when
the module was changed to the semi-submerged mode . Values of 26% and
2&% removal were produced in runs 3 and h respectively, at nitrate
loadings slightly higher than the previous two runs (Table 5.) .
Taking into account a slightly higher nitrate loading in run h
(compared to run 2), it appeared that both semi and submerged modes
produced identical results at the higher loadings (U.87 and 5-^9 g NQ-^-'
m^/cl) at similiar C/N ratio. However better results were shown in the
submerged mode at the lower (3 L/min) feedrate. ¥astewater temperature
dropped from a mean of 19°C to 7°C from runs 1 to U, and probably was
partially responsible for a loss in denitrification during the semi-
submerged runs.
More definite comparisons resulted between submerged and semi-
submerged modes of operation, in the latter part of the study at a
narrower temperature range.
1330
-------
EH
CQ S
>H 0
W H
> EH
PH *^
CQ 1-
o h
0 K
O H
FQ S
«
|
oo
CQ
W
C5
. «!
EH
CQ
W
D OJ
1
1
=te
§
L1A
-4-
t—
-3-
OJ 00
-3- u\
OJ
^t
OO VC
• *
oo _=r
H
^J-
CO VD
o-i _^"
0
UA
OJ VD
-* LTN
Ol H
i-3l W
H
£5 C5
VD
OO
t—
H
00
O LTN
• •
LT\ O
H
00
t—
O~\ VD
oo . o
H
O
t—
_* VD
OO O
H
O
t—
LTN ITN
OO CD
H
SI n
1 M
H
g ci!
^ PH
t—
OJ
ir\
VD
VD
VD OJ
LTN t—
OD
-3"
0 0
-^- VD
VD
OO
ON co
OJ -*
0
-d-
VD OO
oo _=r
ol H
1-31 M
W
§ ^
oo
OO
-3-
OO
VD
rH t-
VD VD
OO
-*
ON oo
OO _=!•
OJ
oo
t— LPv
• •
OJ OO
VD
oo
00 J-
oo j-
0| H
i-3| W
W
@ E?
ON
1331
-------
0)
u
8
<:
u
H
pq
3
g
H
§
CN
O -P
•tn
U
O
cn
en
in
(X5
10
CO
Oi
CM
CN
CO
ro
CO
1332
-------
(b) Run 5
As a result of raised nitrification RBC DO (discussed earlier),
DO concentration also rose in all stages of the denitrification module.
As shown in Figure 3, values above 1 mg/L DO resulted in the denitrifica-
tion module as influent DO (nitrification effluent) rose from 3-5 to
6.8 mg/L. A further increase to slightly more than 2 mg/L DO was observed
in the denitrification module, when the nitrification module DO rose above
8 mg/L.
Little or no change was noted in nitrate removal between test 1
and 2(a) (test number also corresponds to nitrification disc rpm) at the
same C/N ratio with a slight rise in module DO (Table 6). However a
marked improvement did occur in test 2(a), when the C/N ratio was raised
to 3.3, and illustrated the benefits of raising methanol dosage.
A drop in nitrate removal (68% to l8$) resulted in Test 3 at a
similiar C/N ratio to test 2(b) when a rise in module DO occurred.
Similiar results were shown in test k (2h% nitrate removal) to previous
runs 3 and U in the submerged mode (compared Tables 5 and 6).
It appeared from these tests, that a mean denitrification module
DO of less than 2 mg/L, combined with a C/N ratio (3.3/1) close to the
theoretical requirement (i.e. in test 2(b)), produced good nitrate
removal. Consequently the nitrification RBC module rpm was lowered to
2, for the remainder of the study and the C/N ratio was maintained at
approximately 3/1.
Runs 6-9
Under the revised operating conditions in runs 6 to 9 a wide
range of nitrate removal efficiency took place, at C/N rations ranging
from 2.8 to 3.2 (See Table 7)- A wastewater temperature range of 12 to
l6°C occurred during these runs. Similiar removals were observed in run
6 at identical module operation (semi-submerged) to run 55 (test 2(b))
at a liquid temperature of 11°C. However, when the denitrification
module was changed to full disc submergence in run 7> an increase in
nitrate removal resulted to 91% to a denitrification rate of 3.Hi gm
nitrate/m2/d at a temperature of lU°C (See Table 7).
An immediate loss in nitrate removal efficiency to hO% occurred
when the denitrification module was reverted to the semi-submerged
mode (Run 8) at l6°C (Table 7).
In order
conditions, the
(run 8) to 2.27
to 2.3% occurred
to 2.8 may have
nitrate removal
to achieve better denitrification under semi-submerged
disc loading was reduced in the last run from U.39
g nitrate/m^/d. An unexpected drop in nitrate removal
as a result of this change. A slight drop in C/N ratio
contributed to this loss in efficiency, but higher
was expected with reduced loading.
1333
-------
10
Effects of Nitrification RBC
RPM on Nitrification and
Denitrification Module DO Levels
(RUN 5)
FIGURE 3
a
<3)
1
-O
Nitri fication
Denitri fication
Disc RPM
Nitrification Module
1334
-------
TABLE 6
SUMMARY OF DENITRIFICATION RBC
DO PATTERNS AND EFFICIENCY
Nitrification Module Run 5
Disc RPM and Test #
Flow 1/min
NO -N
Inf. g/d
Temp C
>
C/N Ratio
RBC
DO Bay #1
(mg/L)
#2
#3
#4
NO -N
Removal g/m /d
%
Effluent
mg/1
5.6
137
9.0
2.0/1
1.3
0.9
1.1
0.8
1.59
27
12
(a)
5.2
120
8.6
2,0/1
1.2
1.3
1.6
1.1
1.55
30
11
(b)
5.5
107
7.3
3.3/1
1.8
1.6
1.1
1.2
3.14
68
4.4
5.5
119
7.0
3,0/1
2,8
2.4
2.2
2.1
0.91
18
12
5.6
137
9.6
2.6/1
2.2
2.3
1.6
1.8
1.42
24
13
1335
-------
I
I
O
g
o
H
EH
H
fa
a
W
o ^
& CJ>
ffi >D
CO
H
u
°B
•d
i-H X.
O X
g
U $
•d
CM
Q)
4J
td
M
•M
•H
fl
§
Hi
o c
W -H
til.
^
0
3
=tt=
rt
CM ^J1 VD
H rH rH
VD rH 0
r*- ^ c^
» «• *
CM CO H
^ rH 0
^D O*i ^*
X. X. X
O CM CM
CO CO CO
VO <£> CTl
co r- co
• • k
^J* CO ^
o o in
• * •
ID in »*
^O O
to to
1 1
(D "p 0)
CO W W
10 r~ oo
H
•*
CM
i-H
CO
CM
rH
^N,
00
CM
c^
CM
•
CN
0
CM
•8
CO
i
to
0.
1336
-------
In the semi-submerged mode runs, a loss in denitrification
efficiency resulted when the disc loading was reduced (compare runs
6 to 9)- These results were coincidental with what occurred, less
dramatically in runs 3 and U.
DO surveys of the denitrification module through the last four
runs showed mean values below 1.5, with the lowest at 0.8 mg/L in run
6 (Table 8).
DO levels between the module liquid surface and cover (module
atmosphere) showed mean values between 5-1 and. 8.7- The highest DO
concentrations appeared in run 7 with full disc submergence and
reduced air space over the module discs.
lo correlation was found between RBC denitrification efficiency
and DO concentration ranges occurring in the module during runs 6 to 9-
k.3 Removal of BOD , S.S. and TW (Total Nitrogen)
2
BODg removal within the nitrification module, at h rpm (Runs 1
to U), averaged 28% with influent levels of 25 mg/L and effluent values
of 18 mg/L. Similiar values and reductions were observed for suspended
solids (See Table 9).
Little change was noted in either BOD^ or S.S. from influent
to effluent in the denitrification module. A somewhat different
pattern was shown in runs 5 to 9 at lower (2) disc rpm and influent
constituent concentrations. A mean BOD5 reduction from 12 to 7 mg/L
resulted but S.S. rose from 8 to 11 mg/L. Therefore, soluble BODc
reduction through the nitrification module resulted during the latter
runs.
A gain in S.S. through the denitrification module was shown;
probably due to solids sloughing from the discs (Table 9)- A long
stringy loosely-attached growth was evident on the discs and bottom
sediment was more pronounced than in the nitrification module.
For the most part, a slight gain in BOD^ was noted in the
denitrification module effluent, and was mainly attributed to increased
suspended solids concentration.
h.k Alkalinity
Previous studies by the author (8) on an activated sludge
nitrification/denitrification system showed a wt./wt. ratio of calcium
alkalinity removed per ammonia oxidized of 7-1. In terms of denitrifica-
tion, a wt/wt. ratio of h.2 of alkalinity returned to the system per
nitrate removal was also observed.
In comparison, the RBC pilot-study showed calcium alkalinity loss
ratio of 1.6 for nitrification and a gain ratio 3-7 for denitrification.
1337
-------
CO
ta
w
2
O
H
8
H
fe
a
M CO
II
•si1
CM
=a=
in
en CN
• •
^ in
co
•3< en
• »
o o
m co
»* •
o o
O H
CO
sis
CO
r> co
• *
CM O
vo in
• *
o •-!
co r--
• »
O i-J
CO "*
• •
o oi
CM
ro o
• *
rH CO
sis
ID
•
ID
CM O
• •
in o
co in
CT> •
O CM
o in
• •
H CM
CM in
•• •
H CM
SI
tD
CO
in
en
VD
.00 cs
• •
O rH
CO CO
CM
O ID
in
co r~
• •
i-l H
SI!
•s
0)
u
1338
-------
TABLE 9
KBC BOD , SS AND TKN RESULTS
(MEAN VALVES)
Nitrification
Runs 1-4
BOD
SS
TKN
Runs 5-9
BOD
SS
TKN
Study
Mean
BOD
SS
TKN
Inf.
25
21
20
12
8
15
Eff.
18
15
3.5
7
11
3.4
% Rem
28
29
82
42
-62
77
19
-17
80
Deni tri f icat ion
Eff.
17
16
3.9
10
18
3.7
% Rem
6
- 7
-11
-43
-36
- 9
-19
-22
-10
1339
-------
5.0 Conclusions
The following conclusions were derived from the RBC pilot-plant
study at O.E.F.
(a) Close to full ammonia oxidation was attained in the
nitrification RBC module with some benefit shown on
raising disc rpm above 2.
(b) DO concentration in the nitrification model increased with
raised disc rpm and in turn lowered nitrate removal in the
denitrification module.
(c) A nitrate removal efficiency of 91% with effluent nitrate
level close to 1 mg/L, was attainable when the denitrification
module was operated at full disc submergence. Lower nitrate
removals were experienced with the semi-submerged mode at
comparable liquid temperatures and C/W ratio.
(d) A reduction in BOD, both soluble and insoluble, resulted
through the nitrification module with a slight gain in
suspended matter shown in the denitrification model.
(e) Calcium alkalinity loss per weight ammonia oxidized (ratio)
slightly higher than previous studies was shown; whereas
alkalinity released per weight nitrate removal was less than
previously observed in activated sludge process studies.
1340
-------
References
1. Ministries of the Environment and Natural Resources (Ont.)
Thames River Basin ¥ater Management Study 1975-
2. Beak, T.W.
"An Evaluation of European Experience with the
Rotating Biological Contactor" WPCP EPS-U-WP-73-^
Oct. 1973."
3. Davies, T.R. and Pretorius W.A.
"Denitrification with a Bacterial Disc Unit"
Jn. Wat. Res., Vol. 9, pp U53.-^63 Pergamon Press 1975.
k. Soyupak, S.
"Continuous Rotating Biological Contactor for
Denitrification of Wastewater"
Master of Eng. Thesis (1976) McMaster University, Hamilton,
Ontario, Canada.
5. Antonie, R.L.
"Fixed Biological Surfaces - Wastewater Treatment"
Text CRC Press Inc. Publisher 1976/Cleveland, Ohio.
6. Hewitt, T.
"Nitrification of a Secondary Municipal Effluent
Using A Rotating Biological Contactor"
Ministry Environment Ontario Res. Pub. #71 April 1978.
7. McCarty, P.L., Beck, L. and St. Amant P.
"Biological Denitrification of Waste-waters by
Addition of Organic Materials,"
Proc. 2Vth Purdue Ind. Waste Conf., Pg 1271-1285, 1969-
8. Smith, A.G.
"Nitrification - Denitrification of Wastewater Using
A Single-Sludge System" Vol. 2
Canada/Ontario Res. Pub. #96, 1979-
1341
-------
-------
PART XI: SELECTIONS AND ECONOMICS
DESIGN CONSIDERATIONS FOR A 16 MGD
ROTATING BIOLOGICAL CONTACTOR TREATMENT FACILITY
By
William F. Barry, P.E.
Project Engineer
James W. Heine, P.E.
Division Director
OWEN AYRES & ASSOCIATES INC
Consulting Engineers
1300 West Clairemont Avenue
Eau Claire, Wisconsin
General
Eau Claire, Wisconsin is located approximately 100 miles east of the
Twin Cities of Minneapolis - St. Paul, Minnesota. Eau Claire is one of the
fastest growing metropolitan areas in the State of Wisconsin. The estimated
1980 population for the City is 50,000. The project population in the
design year of 2000 is 101,716.
The climate in Eau Claire is one of extreme. Temperatures vary from
above 95°F in the summer to below -30°F in the winter. Daily temperature
variances of 30°F are common. The winter climate often is below 0°F for
periods of two months or more. A typical winter will have snow covered
ground for a three to four month period.
In 1969 the City took the initial steps to upgrade an overloaded
primary treatment facility to provide secondary treatment of its wastewater
prior to discharge to the Chippewa River, a tributary of the Mississippi.
OWEN AYRES & ASSSOCIATES INC was hired as consulting engineer for the
project. A preliminary engineering report completed in May, 1971 analyzed
1343
-------
the treatment alternatives available to the City and recommended a treatment
facility utilizing activated sludge treatment with solids incineration. By
mid-1972 a design utilizing the report concept was approximately 60% complete.
With the passage of Public Law 92-500 in October, 1972 funding of the
project became uncertain and design work was ceased. By late 1974 new
funding regulations were being finalized and the City applied for funding
to conduct a "201" facilities plan. The application was approved and work :
was restarted in early 1975 at the initial conceptual stages. The following
numbers and costs are based upon the 1976 facilities plan and pilot studies
conducted in 1976 and 1977.
Although frustrating to all parties involved, the delay in funding was
beneficial. An opportunity was afforded for the analysis of newer technology
and an analysis of more energy efficient systems was now feasible.
RECEIVING WATERS
The ultimate receiving water for the Eau Claire wastewater treatment
plant is the Chippewa River. The River has its origin in a number of lakes
and swamps in the northwestern part of the state. The drainage area is
oriented in a northeast to southwest direction extending from near the
Upper Michigan boundry to the confluence with the Mississippi River. The
Chippewa drainage basin encompasses a total drainage area of 9468 square
miles, being second in size in the State only to the Wisconsin River Basin.
The total length of the basin is about 170 miles. The Eau Claire urban
area is located about 55 miles from the Mississippi River, or in the lower
third of the drainage basin. The Eau Claire River is a major tributary, its
confluence with the Chippewa being in downtown Eau Claire.
Upstream from the urban area are approximately twenty-two municipalities,
several milk processing plants and three paper mills that discharge from
point sources various concentrations of wastewater to the Chippewa River or
its tributaries.- Many additional non-point wastewater sources exist upstream
from Eau Claire.
STANDARDS AND GUIDELINES
The intrastate standards applicable to the Chippewa River in the Eau
Claire area are those dealing with recreational use, fish, and other aquatic
life. In summary, these standards are:
Dissolved Oxygen: The dissolved oxygen content in surface waters
not be lowered to less than 5 mg/1 at any time.
Temperature: There shall be no temperature changes that may adversely
affect aquatic life. Natural daily and seasonal temperature
fluctuations shall be maintained. The maximum temperature rise at
the edge of the mixing zone above the existing natural temperature
shall not exceed 5°F for streams. The temperature shall not exceed
89°F for warm water fish.
1344
-------
pH: shall be within the range of 6.0 to 9.0, with no change greater than
0.5 units outside the natural seasonal maximum and minimum.
Toxics: Unauthorized concentrations of substances are not permitted
that alone or in combination with other materials present are
toxic to fish or other aquatic life.
Bacteriological: The membrane filter fecal coliform count shall not exceed
200 per 100 ml as geometric mean based on not less than 5 samples
per month, nor exceed 400 per 100 ml in more than 10% of all samples
during the month.
BOD,.: (monthly average)
(weekly average)
Suspended Solids:
(monthly average)
(weekly average)
30 mg/1
45 mg/1
30 mg/1
45 mg/1
The above standards are considered best practicable waste treatment
technology to meet the state water quality standards. Any wastewater
treatment scheme proposed must be capable of meeting these effluent standards
as a minimum.
WASTEWATER PROJECTIONS
The three major contributors of wastewater in the Eau Claire urban
area are those eminating from domestic, industrial and infiltration/inflow
sources. A projection of the volume from the three sources was made for
the service area through the year 2000.
Domestic wastewaters were defined as all flows from residential,
commercial and public sources. The inaccuracy of the existing venturi flow
meter at the Eau Claire wastewater treatment facility made use of this data
questionable. To obtain a reasonable estimate of past and future wastewater
flume, billing records available from the Eau Claire water utility were
used. Per capita water consumption records were analyzed for the years
1955 through 1975. A linear regression analysis was used to predict water
consumption in the year 2000 as shown by Figure 1. The year 2000 domestic
water usage was projected to be 142 gallons/capita/day. Assuming 80% of
the water usage eventually reaches the wastewater treatment plant, the per
capita wastewater contribution will be 114 gpcd. The anticipated domestic
wastewater flow will be 11.6 MGD.
Industrial wastewaters are the most difficult of all flows to accurately
project. Two major industries are currently in operation within Eau Claire
and discharge to the municipal facility wastewater system.
National Presto Industries is an ordinance plant which manufactures
military projectiles. The market for the products from the plant vary
widely. If the plant were to operate at full production the company estimates
a wastewater volume of 608,000 gpd. This was used as a year 2000 design
flow.
1345
-------
UJ
cc
:D
CD
U_
in
UJ
o
o
CO
(0
UJ
a:
UJ
o
oo
1346
-------
Uniroyal, Inc. manufactures various types of rubber tires in Eau
Claire. During a peak production day an estimated.flow of 850,000 gpd is
discharged into the sanitary sewer system. The wastewater is from wet
collectors which trap carbon blank solids and from the cafeteria and restrooms.
Uniroyal does not anticipate any substantial increase in production and a
year 2000 flow of 850,000 gpd was used for design purposes.
Numerous other smaller industries are located within the City of Eau
Claire. The average annual daily water usage for the industries is 0.9 MGD.
The City is currently pursuing an aggressive program of obtaining small
industrial facilities within the City. Based on industrial land available
within the community an anticipated industrial wastewater flow from the
smaller industries is 2.4 MGD in the year 2000.
An infiltration/inflow analysis prepared for the City of Eau Claire
revealed excessive clearwater sources. After corrective.measures are
completed it is estimated that clearwater will be reduced to 0.8 MGD.
The anticipated year 2000 wastewater flows are summarized in Table 1.
YEAR 2000 WASTEWATER FLOWS
TABLE 1
Source
Volume (MGD)
Domestic Wastewater
Uniroyal, Inc.
National Presto Industries
Other Industry
Infiltration/Inflow
TOTAL
16.3
Other parameters of importance that required consideration were the 5-
day biochemical oxygen demand (BOD ) and the suspended solids concentration
of the incoming wastewater. Past records available from area wastewater
treatment facilities showed that a reasonable value of these parameters was
250 mg/1 or 33,735 Ib/day each.
LIQUID HANDLING
Only secondary treatment processes will be discussed in the section
although a similar type analysis was conducted within other areas of the
plant. Based on existing performance of the City treatment plant it was
estimated that 35% of the incoming BODj. would be removed in the primary
treatment portion of the facility. This leaves an organic loading (BOD,-)
to the secondary process of 21,900 Ibs/day.
1347
-------
Three secondary processes were analyzed in detail. These were an
activated sludge process,rotating biological contactor (RBC) and an activated
biological filter (ABF). Only the RBC process design will be detailed
here.
A schematic diagram of the RBC system is shown on Figure 2. The
process consists of large diameter corrugated plastic media which is mounted
on a horizontal shaft and placed in a concrete tank. The media is slowly
rotated while approximately 40% of the surface area is submerged in the
wastewater. Microorganisms naturally present in the wastewater adhere to
the rotating surfaces. In rotation, the media carries a film of wastewater
into the air where it trickles down the surface and absorbs oxygen from the
air. Organisms in the biomass then remove both dissolved oxygen and organic
materials from the film of wastewater. Further removal of dissolved oxygen
and organic material occurs as the media continues rotation through the
wastewater in the tank. Unused dissolved oxygen is mixed with the contents
of the wastewater.
Using manufacturers standard performance curves it was estimated that
a hydraulic loading of 3.33 gallons/day/ft.2 of media surface would produce
the required effluent quality with a wastewater at 50°F. This design
required the use of 4.8 million square feet of surface area. This would
require the use of 48 shafts each containing 100,000 square feet of surface.
The shaft dimensions are approximately 11'-10" diameter by 25 feet long.
A similar design was done on the other two secondary systems. The
estimated construction costs, operation and maintenance costs and present
worth values for the three secondary systems is summarized in Table 2. The
costs include clarifiers and return sludge equipment where required.
SECONDARY TREATMENT COST ESTIMATES - 1976 DOLLARS
TABLE 2
Process
Activated Sludge
(Complete Mix Mode)
Rotating Biological
Contactor
(Mechanical Drive)
Activated Biofilter
Construction
Costs
(Secondary Units Only)
$3,024,300
$3,137,200
$2,445,200
Annual Present Worth
Energy Plant
Cost (Liquid Handling)
$202,800
$ 97,300
$156,600
$12,981,250
$12,072,450
$12,083,150
1348
-------
o
<
UJ
UJ
o:
CD
<£.
LU
LU t
— d
< o
2%
ID
LU
tE
O
I-
o
z
o
o
o
CD
o
_l
o
00
^s
^8
PJ
UJ
a:
li-
-------
Based on the accuracy of the estimates all three treatment schemes
proved to be cost competitive. To verify the assumptions made to arrive at
the above estimates, pilot plants were run on both the RBC and ABF systems.
PILOT PLANT TESTING
Two pilot plants were installed for testing. One plant consisted of
RBC unit rented from the Autotrol Corporation, Milwaukee, Wisconsin. The
second unit was fabricated at the City of Eau Claire wastewater treatment
facility with equipment purchased from Neptune Microfloc, Inc. to establish
a ABF system. Both units were constructed in the spring and summer of
1976.
The Autotrol pilot plant model rented is known as a two meter pilot
plant. The overall dimensions of the unit were 14' x 8' x 81 and the
active media surface area is 7900 sq. ft. Feed wastewater for this unit
was obtained by pumping primary effluent from the City of Eau Claire municipal
wastewater treatment plant to the unit wet well immediately preceding the
media. From the wet well the wastewater was lifted into the unit by means
of a bucket pump supplied with the pilot unit. The active surface area
within the pilot plant was divided into four stages of equal surface area.
The unit was driven by a motor connected with a gear reducer and chain
system to rotate the discs. A schematic of the unit is shown by Figure 3.
The activated biological filtration system was constructed in an
existing grit chamber that is not required for current plant operation.
The tower dimensions were 4" square by 14' in height. Primary effluent was
pumped into a wet well of dimensions 3' x 4' in depth. The water was then
pumped into the top of the tower. The underflow from the tower returned
into the wet well to be remixed with the incoming fluid.
Excess flow in the wet vreii overflowed into an aeration basin with
dimensions 41 square by 7' deep. After aeration the wastewater flowed into
a final clarifier of dimensions 8' long by 4T wide by 7' deep equipped with
a sloped bottom and sludge hopper. The clarified liquid overflowed a V-
notch weir and was returned to the municipal treatment plant flow. Sludge
collected within this compartment was transferred back to the wet well by
means of a submersible pump. A flow schematic is shown by Figure 4.
.Both pilot plants were placed into operation and both produced an
effluent quality much less than was predicted. To investigate whether the
problem was one of process deficiency or wastewater composition, a bench
scale, batch fed activated sludge unit was placed into operation. This
unit, in two months of daily batch feeding on the twice per day basis, was
unable to produce a substantial buildup of mixed liquor suspended solids.
From the result it was concluded that the operational problems of both
pilot plants were caused by a retardent present in the wastewater entering
the Eau Claire treatment facility. A totally unexpected occurence based on
a knowledge' of the contributing area to the treatment plant.
1350
-------
to
UJ
or
o
h-
o
<
h-
2
O
O
o
o
o
o
03
-------
UJ
o
OT
05
-------
In order to determine what wastewater constituent might be causing the
lack of treatability, a sample of biomass from both the RBC and ABF pilot
systems was collected for detailed laboratory analysis. The results of
this testing indicated the potential for heavy metal toxicity. At this
point it was decided to trace the metal contaminants to their source by
sampling within the collection system. The City was divided into six
smaller regions and sampling conducted. No unusual metal concentrations
were found at any of the sampling locations.
However one location revealed the existance of extreme pH variations.
These pH fluctuations were showing up in a lesser magnitude approximately
four hours later in the influent to the wastewater treatment plant. The
cause of the pH fluctuations was traced to a local industry which produces
printed circuit boards. The industry was contacted and cooperated extremely
well in controlling these severe pH fluctuations.
Both pilot plants were again activated in spring of 1977. The ABF
unit still did not perform satisfactory. The RBC unit operated well at
constant and varying flow rates. Analysis of the data collected showed the
requirement for a design loading of 3.06 gpd/ft. This requires 5.4 million
square feet of surface area or 54 large shafts. For symmetry reasons a
design based on the use of 56 RBC shafts was recommended.
RBC DRIVE SYSTEM
The system design utilized a drive.system consisting of two operating
air compressors, piping and diffusers. The selection of the air drive
system was based primarily on the maintenance requirements of 56 motors,
gear boxes and variable speed drive boxes vs. the two air compressors and
piping.
As an informational item, contractors bidding on the project were
required to supply an alternate bid for a mechanically driven RBC system.
Based on first cost a savings of $38,000 was possible by using the mechanical
drive system. A present worth analysis was then conducted to determine the
life cycle cost difference for the two systems. For reference purposes the
mechanical drive system was used as a reference point (i.e. capital cost of
mechanical drive = 0, capital cost of air drive = $38,000). The analysis
is based on a total of 56 shafts with the following assumptions.
General Assumptions
Electrical power costs. $0.0228/KWH (from 1976 Eau Claire Urban area
Facilities Plan.) Cost in 1980 when the system is to be placed into operation
is $0.00277/KWH.
Labor Costs at $8.00/Hour.
Interest rate 6-5/8%.
Present Worth of annual sum (P/a) .06625 = 10.9099
20
- Twenty year design life.
1353
-------
Operating time is continuous.
One horsepower - .746 KW.
Shaft maintenance time (without drive) is 2.5 minutes/day/shaft.
A 125 HP motor operating at 104 B.H.P. is
A 7-1/2 HP motor operating at 5 B.H.P.
90% efficient.
is 88% efficient.
Blower maintenance time is 1/2 hour/day.
Mechanical Drive maintenance time is 3 minutes/drive/day.
Air Drive System
Two blowers with a 125 HP motor
rating are required.
Power draw is 104 B.H.P.
A 90% efficiency can be expected
at this loading.
Energy consumed is 86.2 KW per
motor.
Yearly power cost is $41,840.
Present worth of power is $456,000.
Daily maintenance time is 3.3 hours.
Yearly maintenance cost is $9,640.
Present worth of maintenance is
$105,000.
vs. Mechanical Drive System
Fifty-six drive motors of 7.5
H.P. are required.
Power draw is 5 B.H.P.
An 88% efficiency can be expected
at this loading.
Energy consumed is 4.2 KW per
motor.
Yearly power cost is $57,400.
Present worth of power is $627,000.
Daily maintenance time is 5.1 hours.
Yearly maintenance cost is $14,900.
Present worth of maintenance is
$162,000.
Item
Installation
Power
Maintenance
Total
Difference
PRESENT WORTH COST SUMMARY
TABLE 3
Mechanical Drive
0
627,000
162 ,,000
$789,000
$190,000
Air Drive
38,000
456,000
105,000
$599,000
Neither system requires any new technology based on operating components.
Blowers, gear boxes, etc. have been used successfully in wastewater plants
for years. Based on the cost analysis the air drive system was recommended
for construction.
STATUS
The Eau Claire Wastewater Treatment Plant is currently under construction.
An overall view of the plant is shown on Figure 5. The facility is expected
to be placed into operation in the Fall of 1980. The secondary portion of
the plant consists of 56 air driven rotating biological contactors followed
by gravity sedimentation. Total estimated construction cost of the plant
is $13,600,000.
1354
-------
EAU CLAIRE
WASTEWATER TREATMENT PLANT
SITE PLAN
FIGURE 5
•OWEN AYRES 8 ASSOCIATES INC
1355
-------
-------
I
AN EVALUATION OF THE COST-EFFECTIVENESS
OF THE ROTATING BIOLOGICAL CONTACTOR
PROCESS IN COMBINED CARBON OXIDATION AND NITRIFICATION APPLICATIONS
By
Jeffrey L. Pierce, P.E.
Vice President
Lee A. Lundberg, P.E.
Principal Engineer
Schneider Consulting Engineers
98 Vanadium Road
Bridgeville, Pennsylvania 15017
INTRODUCTION
Environmental engineers are often confronted with effluent requirements
calling for reductions in both biochemical oxygen demand (BOD 5) and ammonia
nitrogen (NHs-N) prior to stream discharge. A number of biological and
physical-chemical processes are available for NHs-N removal including ammonia
stripping, breakpoint chlorination, ion exchange, suspended growth biological
systems and attached growth biological systems. In municipal applications
advanced BOD 5 removals (55%+) are virtually always achieved through the use
of suspended growth or fixed growth biological reactors. For this reason,
the most cost-effective process chains in applications calling for the
simultaneous removal of BODs and NHa-N are typically completely biological.
This paper will compare the cost-effectiveness of the rotating
biological contactor process to other biological processes for combined
carbon oxidation and nitrification applications. Processes to be given
consideration include: synthetic media trickling filters, conventional
media trickling filters, and the single-stage activated sludge process.
The evaluation will be based on a typical northern climate municipal
wastewater.
1357
-------
BACKGROUND
All approaches to biological nitrification can be considered either
combined carbon oxidation-nitrification processes or separate stage
nitrification processes. They can be further subdivided into suspended
growth or attached growth processes.
Suspended growth processes are those in which biological solids are
suspended in a liquor by some mixing mechanism. The conventional activated
sludge process is the most common example of a suspended growth process.
In attached growth processes, the bulk of the biomass is fixed to some
permanent media held in the reactor. The trickling filter and the Rotating
Biological Contactor are common examples of attached growth processes . The
suspended growth processes require post-clarification and the return of the
biological solids to the head of the reactor to maintain an adequate biomass
inventory. Attached growth systems will also generally require post-clari-
fication for compliance with effluent requirements; however, since an ade-
quate inventory of biomass remains fixed to the reactor's media recycle
for the purpose of inventory maintenance is not required. Partial recycle
of clarified effluent is practiced in some attached growth processes,
most notably in trickling filtration, but to meet other requirements or
achieve other technical advantages.
In a separate stage nitrification process, the BODs and NHs-N removal
functions are largely confined to separate reactors. The carbon removal
stage precedes the nitrification stage to lower the organic load and to
facilitate the development of a population of nitrif iers in the second
stage. Intermediate clarification is essential when a suspended growth
first stage is employed. It is optional when an attached growth first
stage is employed.
DESIGN CONSIDERATIONS AND CRITERIA
A cost-effectiveness analysis of alternative nitrification processes
cannot be undertaken without first discussing design criteria. The follow-
ing paragraphs review design considerations for both suspended growth and
attached growth processes.
Conventional Media Trickling Filtration
The ability of conventional media trickling filters to provide nitri-
fication in a single-stage is not widely recognized in this country. The
authors have had conversations with individuals in the regulatory, research
and engineering communities who have argued that
A) Deep beds are needed (20 feet +) to provide complete nitrification,
B) Nitrification cannot be effected in conventional trickling filters
because of low wastewater residence time, and
C) Nitrification can only be effected if the influent to the
trickling filter is less than 30 to 40 mg/liter
1358
-------
An extensive review of the literature belies the above assertions and
demonstrates the efficacy of conventional trickling filtration as a combined
carbon oxidation-nitrification process. Further, it is possible to develop
empirical design criteria based .upon available pilot- and full-scale operat-
ing data.
Between 1944 and 1961 a number of articles were written summarizing
investigations into nitrification in conventional media trickling filters.
To be of utility to the current Paper, a researcher must have reported at
least the following: observed influent and effluent NE^-N concentration,
influent BODs concentration and hydraulic load, and the type/volume of media
in the filter. Further, only investigations that loaded filters with
municipal primary effluent were considered appropriate for use. Four
significant research efforts were found that provide required information:
1) the National Research Council's investigation1 of twelve operating full-
scale, single-stage filters across the United States in 1944; 2) Stones'
extensive 1961 investigation2 of large-scale pilot plants at Salford,
England; 3) Grantham's pilot-scale studies3 at Gainsville, Florida in 1953;
and 4) Burgess' pilot-scale work1* at Corvallis, Oregon in 1961. The body of
data collected during these research efforts' covers a wide range of organic
loadings, climatic conditions, operating modes and types of filter media.
The empirical design criteria proposed herein will relate percent
removal to the pounds of BODs applied per day per 1000 ft2 of filter media.
Such a criteria is not only convenient but it also has a sound conceptual
basis. The amount of biomass held in a reactor should generally be portional
to the amount of surface area provided by the media in the reactor. A precise
estimate of the biomass available within an attached growth reactor would
require knowledge of the thickness of the slime coating the media, the
effective media surface area and the fraction of volatile versus inert solids
in the slime. As detailed information of this nature is unavailable, a
direct proportionality between biomass and media surface area must be presumed.
At high influent BOD 5 loads, heterotrophic organisms can be expected to
dominate the slime found on the media. As the ratio of BODs load to the
available surface area decreases, nitrifying organisms can begin to
dominate the slime beginning with the lowest layers of the media. Several
hypotheses can be advanced to explain the reduced inventory of nitrifying
organisms at the higher BOD5 loading rates and in the upper layers of filters
under moderate loading rates. One attractive hypothesis is based on the
observed natural variation in growth rates of the heterotrophic and nitri-
fying organisms responsible for BOD 5 and NHs-N removal. The heterotrophic
organisms grow more rapidly and have a greater yield than the nitrif iers .
The higher sluffage rate occurring during BODs removal causes a continual
displacement of the nitrif iers from the film preventing the establishment
of an adequate nitrifying population. This scenario is analygous to an
activated sludge process operated at comparatively low mean cell residence
times where the nitrifying organisms are being wasted at a faster rate than
they can regenerate. A second reasonable hypothesis is based on the
relationship of the availability of dissolved oxygen to the rate of ammonia
oxidation. The rate of ammonia oxidation decreases with decreases in
dissolved oxygen concentration. The nitrifiers are more sensitive to
depressed dissolved oxygen concentrations than the heterotrophic organisms
1359
-------
responsible for carbon pxidatipn. In filters with higher BODg loadings,
lower dissolved oxygen levels might logically fee anticipated and. thus the
nitrifiers a,re again put a.t a competitive disadvantage.. Both of the . abpye
hypotheses are speculative and in any event a firm definition' of the.
operable mechanism or mechanisms, is not critical to this Paper as an
empirical design criteria is being prpposed.
The authors have re-expressed' the work of the National Research Council ,
and Stones on a pounds. BOD5 applied per 10QO ft^ of media basis and have
plotted the data on Figures X and II. The work of Stones implies that
recirculation has a beneficial effect on nitrification. A 1:1 recirculation
ratio was employed. The National Research Council's data confirms Stones'
observations in the lower loading ranges and provides information on the
effect of higher organic loads on nitrification. In general, each of the
National Research Council data points represents the average of four months
of data at an individual plant. Plants from both cold and warm weather
climates and plants with and without recirculation were covered in their
survey.
Figures I and II, along with information from Gainsville and Corvallis,
was combined with more recent unpublished information that the authors ' have
in their files from five additional full-scale rock trickling filters to
prepare Figure III titled "Basis of Design: Conventional Media Trickling
Filters" . It is proposed that the upper boundary of the envelope of per-
formance be employed in applications involving warm wastewaters and involving
recirculation. The lower boundary is to be employed in applications in-
volving cold wastewaters and contemplating no recirculation. In all cases,
single-stage filters with minimum bed depths of six feet may be presumed.
Figure III shows that loadings 'lower than 0.5 //BODg/lOOO ft2-day
are necessary to achieve 90% NHs-^N removal.
Synthetic Media Trickling Filters
Synthetic plastic media was originally developed for trickling filters
in the mid 1950 *s. Because of its relatively high ratio of surface area
to volume and because of its increased void space, it gained acceptance in
the early 1960 *s as a recognized substitute for conventional media. There :
are two prominent types of synthetic media in common use: 1 - corrugated
PVC sheet modules which are cut on-site and then laid into the filter bed,
and 2 - plastic rings which are dumped into the filter bed.
In attempting to establish a design criteria for combined carbon
oxidation-nitrification in synthetic media trickling filters the authors
found virtually no published research, or full-scale operating data. Most
work to date has, assumed that the influent feed to the nitrification
reactor would be secondary effluent or its equivalent. An attempt is made
in the following paragraphs to develop a design criteria for combined carbon
oxidation-nitrification synthetic media reactors through the use of' available
information.
1360
-------
CONVENTIONAL MEDIA TRICKLING FILTRATION
AT SALFORD, ENGLAND BY STONES
100
90-
80-
70
LU
aS 0"
Z
'» 40
I
Z
3o--
20--
1O--
WITH RECIRCULATION
WITHOUT RECIRCULATION
0.1 0.2 0.3 0.4 O.5 0.6
ORGANIC LOAD (* BOD5/1,OOOFT2.DAY)
0.7
FIGURE I
1361
-------
100
90-
80-
70 -
_l
> 60
O
2
lil 50
OC
co
I
40
30
20
10
0
CONVENTIONAL MEDIA TRICKLING FILTRATION
BASED ON NRC INVESTIGATIONS
0.4
APPROXIMATE ENVELOPE
OF PERFORMANCE
0.8 1.2 16 2.0 2.4
ORGANIC LOAD <* BOD5/1,OOOFT2-DAY)
2.8
FIGURE 31
1362
-------
^ Q CO
g LU OC
S2 2 w
CO , H
iu -J d
Q < Q.
CO
CO
<
00
> o
z £
O H
o
N
ui
DC
g
u.
_o
CM*
IO
CM
I-
o
o
o
IO
o
o
m
IO
~d
o o o o o o oo oo
oa>cON-cou)
-------
In 1969, Bala Krishnan5 reported on investigations conducted at the
University of Texas. Secondary effluent from a contact stabilization plant
was applied to a 6 foot deep, 1/2 foot diameter pilot tower. Sewage was
applied to the tower at loading rates of 10, 12.5, 20 and 30 MGAD. Samples
were drawn at 1, 3, 4, and 6 foot depths and tests were run for the various
species of nitrogen including ammonia. Bala Krishnan concluded that
percent nitrification was governed by hydraulic load on the filters (removal
decreased at higher loadings) and that the amount of ammonia removal in-
creased with filter depth. An alternative method of viewing Bala Krishnan's
data is to express it on the previously introduced a #BOB5/1000 ft2-day
basis. Figure IV accomplishes this task assuming that the BODs concentration
of the secondary effluent averaged 25 mg/liter. Bala Krishnan's conclusion
that percent ammonia removal is governed by hydraulic loading is only valid
if the influent wastewater's BOD5 concentration is a fixed value. The design
criteria proposed in his paper (based on hydraulic load) is only applicable
to feed wastewaters that are essentially equivalent to secondary effluent.
In June 1973, Buddies6 reported on the results of extensive experimen-
tation at the Midland, Michigan Wastewater Treatment Plant. A 21.5 feet
deep, 3 foot diameter, pilot tower was constructed on the grounds of the
Midland Treatment Plant. Data was collected over a period of several months
during x^hich various hydraulic loads were imposed on the filter, ranging
from 0.5 to 2.0 GPM/ft2 (based on filter surface). The wastewater feed was
unchlorinated secondary effluent having a BOBs concentration in the range
of 15-20 mg/liter. Both warm and cold weather months were included within
the study. Like Bala Krishnan, Buddies concluded that nitrification was
governed by the hydraulic loading rate and that the amount of ammonia removal
increased with filter depth. It was implied that a tower depth of over
20 feet is required to effect complete nitrification (NHs-N < 1-2 mg/liter).
Buddies' data has be re-expressed on a #BOBs/1000 ft2*day basis and is
plotted on Figure IV. A cold weather (45°F) and warm weather (65°F) curve
is presented. Buddies' implication that a 20-foot tower is required for
nitrification is probably not justified. It is true that increasing
removals of NHs—N are observed moving down the tower, but the phenomenon
is more likely the result of a decreasing overall system load at each point
rather than being related to wastewater contact time in the filter. In
other words, it is likely that a 15-foot tower would perform as well as a
20-foot tower providing each held an equivalent amount of media.
The only research directly relevant to combined carbon oxidation-
nitrification found in the literature was an article written by Stenquist^.
Studies were conducted at the Stockton, California Treatment Plant over
the months of July - Becember, 1972. A pilot plant very similar to that
used by Buddies was employed. Unlike Buddies' arrangement, the pilot plant
was fed primary effluent. The early months of the study period corresponded
to the canning season and little nitrification was observed. Buring the
later months of the study, organic loading decreased and high degrees of
nitrification were observed.
Performance during the later months can be divided into two loading
ranges and can be re-expressed as 94% removal at 0.52 #BOBs/1000 ft2-day
and 89% at 0.82 #BOBs/1000 ft2-day. Stenquist's data has also been plotted
on Figure IV.
1364
-------
COMBINED CARBON OXIDATION AND NITRIFICATION
IN SYNTHETIC MEDIA FILTERS
_l
EMOVA
QC
z
CO
X
Z
*
i uu -
90-
80-
70-
60-
50-
4O -
30 -
20 -
10 -
0 -
- \ X"""X-
\\ x STENQUEST 1972
\ ^
\ \
K\ \
\*\ 45 F > 65 F
>_ V\ DUDDLES 1973
\\
• ^v
\ *^V^
^
V^
V "^s^ BALAKRISHNAN 1970
X. • **-•». 80 F
••*—- — Urns
i i i i
i i i i
0 0.5 1.0 1.6 2.0
BODg/I.OOOFT^-DAY
FIGURE 33T.
1365
-------
In reviewing Figure IV significant differences in reported performance
are apparent. The differences can largely be explained by variations in
the researcher's experimental arrangements. Duddles' and Stenquist's
investigations were conducted using fairly large in-field pilot plants
while Bala Krishnan employed a small laboratory-scale unit. In addition,
Duddles and Stenquist used commercially available corregated PVC trickling
filter media, where Bala Krishnan employed a high density (58-76 ft2/ft3)
mixture of rings and saddles. There is, however, a significant difference
between the curves derived from the data of Stenquist and Duddles. Stenquist
shows that only one-third the media recommended by Duddles is required to
achieve 90% ammonia removal. The discrepancy can probably be traced to
the fact that Duddles employed secondary effluent in his studies while
Stenquist employed primary effluent. To be technically consistent with the
design criteria developed for conventional media trickling filters, only
Stenquist's data should be considered. Duddles' investigation was not
actually a study of combined carbon oxidation-nitrification, but rather
a study of nitrification in a separate stage reactor. For low BODs yaste-
waters, the #BOD5/1000 ft2-day criteria proposed herein is probably kn
inappropriate measure of performance. It is apparent that a great djeal of
additional research is required to define a firm design criteria for combined
carbon oxidation-nitrification in synthetic media trickling filters.
A basis of design for synthetic media trickling filters has been
proposed on Figure V. For purposes of this paper, it was assumed that
Stenquist's data provides an upper limit of the envelope of performance
while Duddles' data provides the lower limit. Although this approach
is somewhat conservative, the insufficency of currently available data
does not permit the development of a less stringent basis of design. As
in the case of conventional media filters, the lower boundary of the envelope
shown on Figure V will be employed in cases of cold wastewaters and the upper
boundary will be employed in cases of warm wastewaters. In all applications
it will be assumed that a minimum tower depth of 15 feet and some degree of
recirculation will be employed. These provisions are necessary to minimize
the danger of short circuiting and to provide adequate wetting of all media
particularly at the lower organic loading rates.
Figure V shows that loadings lower, than 0.6 #BOD5/1000 ft2 -day are
necessary to achieve 90% NHs-N removal.
Rotating Biological Contactors
The design criteria for the rotating biological contactor (RBC) process
is well developed and supported by a large body of pilot-scale and full-scale
operating data. The bulk of this information has been collected and "inter-
preted by the major manufacturer of RBC media, the Autotrol Corporation.
Probably the most comprehensive and up-to-date design criteria for the RBC
process is set forth in Autotrol's 1979 Design Manual.8 A two step
procedure is recommended for designing combined carbon oxidation-nitrifica-
tion facilities. First, sufficient media is provided to lower the soluble
BOD5 to 15 mg/1 (approximately 30 mg/1 total BOD5) . Empirically developed
graphs are provided that relate influent soluble BODs concentration to efflu-
ent soluble BODs concentration as a function of media hydraulic loading
1366
-------
Ul
CC
2 < CO
^ — rr
® QUJ
CO U I-
lil 2 =J
Q Ou.
fc H C5
0^2
CO X _j
g
u.
.q
"oi
CM
a.
o
o
o
O
m
iq
"o
2ggP2922Po
1367
-------
(GPD/ft2) at 55 °F. Temperature correction factors are provided for waste-
waters varying in temperature from 55°F to 409F. Additional media is then
provided to achieve the desired degree of nitrification. Again, empirically
developed graphs are provided that relate influent NHg-N concentration to
effluent NHs-N concentration as a function of media hydraulic loading
(GPD/ft2) at 55°F. Temperature correction factors are provided for waste-.
waters varying in temperatures from 55°F to 42°F. The total media require-
ment is the sum of the BOD 5 and NHs-N media requirements.
It is interesting to compare the above methodology to that originally
recommended by Autotrol in their 1972 Design Manual. ^ A two step procedure
was recommended. A graph was presented that directly related effluent
BODs concentration to anticipated percent NHs-N removal. As an illustration,
a 20 mg/1 BODs concentration was associated with a 70% NHs-N removal. With
a target BODs concentration identified the second task was to provide suffi-
cient media to attain that concentration. A graph was provided relating
influent BODs concentration to various percent removals of BODs as a
function of media hydraulic loading (GPD/ft2) . Temperature correctiqn factors
were again provided for wastewater under 55°F.
To further complicate matters, R.L. Antonie of the Autotrol Corporation
presented in June 1974 what could be considered Autotrol 's interium design
criteria. 10 Antonie presented a one step design methodology graphically
relating influent BODs concentration to percent ammonia removal as a function
of media hydraulic loading (GPD/ft2) at 55 °F. Temperature correction factors
were provided for wastewaters in the range of 55°F to 40°F.
The 1974 and 1979 revisions to the Autotrol approach to nitrification
design provided not only changes in the philosophy of their approach to the
sizing of combined carbon oxidation-nitrification reactors, but more
importantly proposed substantially more conservative loadings to achieve
equivalent NHs-N removals. Actual field experience with full-scale operating
facilities such as Gladstone, MI11 had apparently convinced Autotrol that
their original 1972 anticipations for the process were overly optimistic.
To facilitate a direct comparison between the 1972, 1974 and 1978 RBC
design criteria and the criteria advanced for the other attached growth
reactors, Autotrol' s RBC design criteria has been re-expressed by the Author's
on a #BODs/1000 ft2*day basis. Figure VI provides a re-expression of
Autotrol' s criteria. It was assumed that the influent total BODs concentra-
tion was 120 mg/1, the influent soluble BODs concentration was 72 mg/1 and
the influent NHs-N concentration was 18 mg/1. Figure VI shows that the load-
ing rate to achieve 85% nitrification was halved between 1972 and 1979.
Viewed alternately, the amount of media required doubled. Currently, loadings
lower than 1.3 #BODs/1000 ft2 -day are required to achieve 90% NHa-N removal.
This loading is still much higher than the 0.5 to 0.6 #BOD5/1000 ft2*day
derived previously for other atttached growth reactors. This is apparently
attributable to more effective use of. the theoretically available media
surface area and/ or greater availability of oxygen.
1368
-------
CO
DC
O
UJ
Z
CD
O
o
CO
O
h-
<
O
U- O
II
Q
z o
< o
N
28
CO
h-
<
I-
o
o
O
00
o
00
o
h-
i
HI
cc
ID
o
o
_ »
CO
O
O
m
1VAOW3U N- HN ±N3Od3d
1369
-------
For purposes of this Paper, criteria set forth in Autotrol's 1979
Design Manual8 will be employed to size the RBC facilities. The reader
is referred to that text for a thorough discussion of their BBC design
criteria and design considerations.
Suspended Growth Processes
The two-stage activated sludge process for carbon oxidation-nitri-
fication was promoted by the United States Environmental Protection Agency
in 1973 as the only process capable of providing complete nitrification on
a year round basis in northern climates.12 Experience has shown this
assertion to be incorrect. Complete nitrification can be provided by
properly designed single-stage activated sludge processes in northern climates.
This does not, however, imply that the two-stage process is without merit.
Sufficient technical and economic justification may exist in certain instances
to make it the process of choice. Single-stage pure oxygen activated sludge
is another suspended growth process that deserves serious consideration in
combined carbon oxidation-nitrification applications. In this discussion
the two-stage process is distinguished from the single-stage process by
the introduction of intermediate clarification.
The extent of nitrification in a single-stage activated sludge process is
dependent on population dynamics. The mean cell residence time of the system
(on a carbonaceous basis) must exceed the reciprocal of the growth rate of
the nitrifying bacteria. Nitrification can only be maintained when the
growth rate of the nitrifying bacteria is rapid enough to replace the organ-
isms lost through sludge wasting. The nitrifier growth rate is extremely
sensitive to temperature, dissolved oxygen and pH. A simple expression
been advaced to calculate maximum nitrifier growth rate under varying
environmental conditions.13
M - 0.47
0.098(T-15°C)
pH >_ 7.2
(Equation 1)
For a dissolved oxygen concentration of 2.0 mg/1, a wastewater temperature
of 47°F and the application of a 225% safety factor, a design solids reten-
tion time of 15 days would be indicated. A mean cell residence time of 15
days is readily achievable in properly designed activated sludge processes
that treat normal strength domestic wastes., The higher mean cell residence
time can only be attained by maintaining a higher solids inventory in the
system. Options available to increase the inventory include maintenance of
a higher MLSS concentration in the aeration basins and/or increased tank
volume (longer hydraulic detention times). The higher MLSS concentrations
may require the adoption of more conservative final clarifiers become solids
flux limited. Further, the aeration system must be upgraded to provide
greater mixing and to satisfy the increased oxygen demands due to ammonia
oxidation and endogenous respiration of the carbonaceous solids associated
with the higher mean cell residence times; In short, carbon oxidation-
nitrification can be achieved by the single-stage activated sludge process
at the expense of a number of design concessions. Two-stage air activated
1370
-------
sludge and single-stage oxygen activated sludge may offer technical and econo-
mic advantages depending on the specific circumstances of an individual appli-
cation.
An economical concept for the two-stage process with which the Authors
have had experience employs the first stage as a high-rate roughing unit
not preceded by primary clarifiers which produces an effluent BOD5 of
50 to 60 mg/1. Deletion of primary clarification and the reduction in
detention time in the first stage to less than 2.5 hours reduces the construc-
tion cost of implementing a two-stage system. A two-stage system would become
increasing advantageous in cases of extremely low wastewater temperature or
in those cases requiring a high safety factor. In such situations, mean cell
residence times in excess of 20 days would be required in single sludge
systems. The higher sludge ages would become increasingly more difficult
and costly to attain.
Single-stage pure oxygen activated sludge can be employed as an aid in
. maintaining the higher sludge ages necessary for single-step nitification.
With pure oxygen it is easier to maintain.high levels of MLSS in the
aeration basins lessening the need to add tank Volume to aid in increasing the
inventory of solids in the system. Proponents of pure oxygen claim improved
settling characteristics of the mixed liquor reducing the final clarifier
surface area required at high solids loads. Further, a case can be made
that the maintenance of higher dissolved oxygen levels in the aeration
basins stimulate the rate of growth of the nitrifiers and reduce the required
mean cell residence time required to maintain an adequate inventory of nitri-
fiers. Offsetting the above advantages, however, is the cost of construct-
ing covered aeration basins and providing on-site oxygen generation equipment.
For purposes of this Paper the single-stage air activated sludge process
will be compared to the previously introduced attached growth reactors. It
was felt that the single-stage process provides a representative cost for
the suspended growth processes. Further, it is less complex than the two-stage
and pure oxygen processes and in that sense is closer in overall operability
to the attached growth systems.
Design of the single-stage air activated sludge process will be pre-
dicted upon maintenance of the sludge age required by Equation 1. A
rational approach will be employed to establish the solids inventory
and final clarifier surface overflow rates necessary to maintain the
required sludge age.
BASIS OF COMPARISON
The influent to the combined carbon oxidation-nitrification treatment
facilities was assumed to have the following characteristics:
BOD5 (Total)
BOD5 (Soluble)
Suspended Solids
120 mg/1
72 mg/1
90 mg/1
1371
-------
NH3-N
Temperature (Winter)
Temperature (.Summer)
18 mg/1
47 °F
659F
The above characteristics were selected as being representative of a typical
primary effluent. The following set of effluent requirements were adopted
for of the current evaluation:
BOD5 (Total)
Suspended Solids
NH3-N (Winter)
NH3-N (Summer)
20 mg/1
20 mg/1
4.5 mg/1
1.5 mg/1.
These requirements represent a typical combined carbon oxidation-nitrifica-
tion application where the regulatory agency has allowed a seasonal variation
in the ammonia standard. Many agencies allow such a variance in considera-
tion of the increased difficultly in attaining stringent ammonia removals
during cold weather and in consideration of the larger stream flow generally
available during the winter season.
The following assumptions will be in effect during all cost comparisons:
A) Sufficient land is available for construction of all the alterna-
tives and no difficult rock or subsurface conditions exist to
interfere with construction.
B) The topography at the site is of such a contour that flow through
the suspended growth and RBC processes can be accomplished by
gravity but that intermediate pumping is required for both forms
of trickling filtration.
C) The amounts and characteristics of the sludge produced by each
of the alternatives, and conversely the sludge handling/disposal
costs, are roughly equivalent. An exception to this assumption
was made in the case of the activated sludge process for which
flotation thickening facilities were provided.
D) The cost of providing post-treatment disinfection is the same
under each of the alternatives.
E) Construction costs are based on an ENR index of 3250 and include
a 25% contingency to cover engineering, financial and legal project
costs. Operating costs were based on $0.045 per KWH, $10.00 per
labor man-hour and $1.75 per dry pound of polymer.
F) Present worth calculations were based on a discount rate of
7-1/8% and a recovery period of 20 years.
1372
-------
SUMMARY OF COST ANALYSES
The sections that follow present an analysis of cost based on the above
assumptions. '
Conventional Media Trickling Filtration
The conditions under study require a 75% removal of NHs-N at 47PF and
a 92% removal of NHs-N at 65 °F. Figure III indicates that a 92% removal
of NHs-N can be accomplished 'during warm weather at a loading rate of less
than 0.35 #BOD5/1000 ft2-day. A 75% NHs-N removal during cold weather requires
a loading rate of less than 0.175 #BOD5/100Q ft2'day. The cold weather
case governs and the lower loading rate must be adopted for design.
Figure VII provides an estimate of construction cost and total present
worth cost for conventional media trickling filtration as a function of media
surface area and design flow. The estimate presumes the use of circular
single-stage trickling filters having an average media density of 15 ft2/ft3
and a bed depth of eight feet. Intermediate pumping facilities were
provided assuming 25 ft of total head with no allowances for recirculation.
Final clarifiers were sized on the basis of 800 GPD/ft2 with typical allowances
provided for sludge pumping and piping facilities to return the biological
sludge to the head of the primary tanks for co-thickening.
Synthetic Media Trickling Filtration
Figure V indicates that a 92% removal of NHs-N can be accomplished
during warm weather at a. loading rate of less than 0.5 #BOD5/10002'day,
while a 75% removal of NHs-N removal during cold weather requires a loading
rate of less than 0.3 #BOD5/1000 ft2-day. Again, the cold weather case
governs. .
Figure VIII provides an estimate of construction cost and total present
worth cost for synthetic media trickling filtration as a function of media
surface area and design flow. The estimate presumes the use of circular
single-stage trickling filters packed with corrugated PVC sheet modules cut
on-site and laid into the filter bed. A twenty-four foot deep bed was provided,
the top two-thirds being packed with a standard media having a density of
30 ft2/ft3 and the lower third being packed with a high density 44 ft2/ft3
media. Intermediate pumpage facilities were provided assuming 41 ft of total
head with a sufficient recycle rate to guarantee a minimum hydraulic loading
rate of 0.75 GPM/ft2 to the surface of the filter. A minimum hydraulic load-
ing rate of 0.75 GPM/ft2 is necessary to guarantee complete wetting of the high
density media to make full use of its surface area. Provisions for final
clarification and sludge handling were identical to those of the conventional
media trickling filters.
Rotating Biological Contactors
A 92% removal of NHs-N can be achieved during the warm weather condition
when the loading rate is less than 1.25 #BOD5/1000 ft2«day, while a 75%
removal of NHs-N during the cold weather condition requires a loading rate
of less than 1.1 #BOD5/1000 ft2-day. Like the previous two attached growth
processes, the cold weather condition governs.
1373
-------
o
-o
co
PI
UJ
DC
0
u.
o
d-
u>
q
d-
o
Q
0
O
_j
u.
2:
0
co
u
o
o
*•>
10
o
-w
CM
CO
Z
O
O O
o o LLI
o
-o
CM
01
o
-o
H
U_
UJ
DC
<
tu
o
u.
DC
CO
<
O
UJ
-s
o
10
o
CM
(SNomw) isoo nvioi
1374
-------
U)
tO
-CM
CO
Z
O
o
-o
CM
U.
HI
DC
<
UJ
o
<
U.
a:
13
CO
Q
UJ
'CM
10
o
CJ
o
cvt
1SOO 1V±O±
1375
-------
Figure IX provides an estimate of; construction cost and total present
worth cost for mechanical drive rotating biological contactors as a, function
of media surface area and design flow. The estimate includes the construction
cost of the KBC tankage, final clarifiers at 8QQ GPD/ft2, K£C equipment and
cover installation and the necessary pumpage equipment/piping to return the
waste biological sludge to the head of the primary tanks for co-thickening.
Suspended Growth Nitrification
Equation 1 coupled with a 225% safety factor indicates that a mean
cell residence time of 15 days is required at 47 °F to assure nitrification
in a single-stage suspended growth system. A 15'day mean cell residence
could typically be maintained in an activated sludge system with the
following design parameters:
Mean cell residence time:
Hydraulic detention time:
MLSS:
Recycle ratio:
15 days
8.1 hours
3100 mg/1
60 %
Figure X provides an estimate of construction cost and total present worth
cost for an activated sludge process designed on the above basis. The
use of fine bubble, porous ceramic plate difussers having an oxygen transfer
efficiency of 26.5% was assumed. Sufficient oxygen was provided to satisfy
both the carbonaceous and nitrogeous oxygen demands. Final clarifiers were
sized based on a surface overflow rate of 600 GPD/ft2. Dissolved air flota-
tion facilities are necessary for thickening the waste bilogical solids from
the process and these units have been factored into Figure X. Included in
the construction cost are estimates for aeration tankage and equipment;
final clarifiers; return and waste sludge pumping facilities; air, sludge
and wastewater piping; blower and flotation equipment; and a control building.
COMPARATIVE ANALYSES
Figure XI provides a graphic comparison of the present worth cost of
the four nitrification alternatives over a range of flows from 3 MGD to
50 MGD. The following conclusions can be drawn from Figure XI:
A) The RBC process is the least cost attached growth nitrification
alternative over the entire range of flow. Conventional media
trickling filtration was the most costly alternative and synthetic
media trickling filtration was the second most costly.
B) The single-stage activated sludge process is the least cost nitri-
fication alternative at flows in excess of 8-9 MGD.
C) At flows under 10 MGD, the Attached' growth processes become more
attractive. The REG process overtakes the activated sludge
process at around 8-9 MGD. The 'synthetic media trickling filters
and the conventional media trickling filters overtake the activated
sludge process at 3-4 MGD.
1376
-------
o
O"'
10
CD-
CO
Q
O
O
g
CO
ui
a
CO
H
co
o
o
I-
cc
o
LU
CO
111
cc
Q.
o
d-
o
Z)
DC
H
CO
O
o
LU
DC
-)
O
"ro
CO
z
O
111
o
o
LU
^co «
CM
H
LL.
LU
oc
LU
O
<
U_
CC
Z>
CO
Q
LU
oc
Q. £=
- -10
Z O DC
CD Z
DC 111
o
10
o
CO
o
CM
isoo nvioi
1377
-------
h-
co
O
o
X
H
OC
O
in
O
Q
^ CO
I- Q
Z LU
CO <
LU >
DC fl
CL Q
Q <
x
CO LLI
o
~c>
M
m
cc
'o
CO
a
u.
z
2
CO
LU
O
o
"o
O CO HI
o
IO
o
co
Nl) ±SOO 1V1O1
1378
-------
I-
co
o
o
DC
O
z
LU
CO
LU
DC
0.
LU
K- <
•o
10
Ul
OC
3
CD
o
•c>
O
O
•
"
oc
UJ
111
s
o
'to
o
u>
*>
o
CM
o
«»
Nl) J.SOO 1V1OJ.
1379
-------
It must be remembered, that the above cpsts are npt site specific and that
they axe developed based pn'an assumed'set of'typical design conditions.
Further, preliminary- cost analyses such as. thp.s.e'. developed herein are not
a substitute for rigorous cost analyses conducted on a project by project
basis. With.these1considerations in mind, only a few definite conclusions
can be drawn:
A) The KBC process'must be seriously considered and subjected to
detailed cost analyses for all applications requiring combined
carbon oxidation-nitrification involving flows less than
15 MiGD. The RBC process appears to be competitive with the
single stage activated sludge process to at least that capacity.
B) At flows less than 6 MGD, the synthetic media trickling filtration
process should also be given serious consideration. The attached
growth processes become increasingly attractive at the lower
design flows.
A thorough cost-effectiveness analysis must consider more than simply
present worth cost. A separate evaluation of comparative construction
costs and operating costs is relevant because the lowest present worth
alternative may not provide the lowest net local user charge. Alternatives
involving higher construction costs but having lower operating costs may
result in lower net local user charges depending on the structure of a
project's financing. Figures VII, VIII, IX and X present construction cost
in addition to total present worth cost. The present worth of annual operation/
maintenance can be developed from these figures by subtracting construction
cost from total present worth cost. Present worth annual operation/maintenance
can be converted into annual operating/maintenance cost by division by the
discount rate (10.49 - 20 years @ 7-1/8%) employed in this evaluation. In
general, the attached growth reactors are more costly to construct but less
costly to operate.
Non-monetary considerations such as reliability and operability should
also be addressed in a thorough cost-effectiveness analysis. The attached
growth reactors obviously have an advantage over the suspended growth process
in the area of operability. They require less operational attention and are
self-regulating. The suspended growth process requires close attention
to sludge wastage, air demands and the recycle ratio. Such vigilence is
necessary to regulate mean cell residence time, to satisfy oxygen demands and
to preserve the efficiency of the final clarifiers. All processes can be
called upon to provide reasonably reliable treatment. Because the attached
growth reactors place less reliance on operator judgement, they could be
considered to be more reliable. The design criteria for both the EBC process
and the suspended growth process is reasonably well established. The
criteria for both forms trickling filtration is, however, not as well
established and the conservative analyst may desire to consider the trickling
filtration processes somewhat less reliable than the other processes on this
basis.
There are some variations to the above processes, not yet discussed, that
could enhance their cost-effectiveness. As an alternative to mechanical
drive RBC's, the Autotrol Corporation is offering an air drive process which
1380
-------
offers reductions, inboth., capita.l and operating cost relative to th.e
traditional mechanical driye. In a companion Paper,•llf the. Authors. have
shown a present worth, cost s.aytngs, ya,rying fronj'6.5% a,t 3 MGD. to 16% at
50 MGD. relative to the traditional niechanica.l dri.Ve» Th.e sayings, are
achieved through, low,e.r poorer consumption and a reduction in the. number
of shafts of media required to reach, equivalent BQDs and NHs-N removals.
The number of shafts are reduced due to the increased use of high density
media and a slightly- higher loading rate per square foot of media. It
was also shown in analyses, contained in that Paper that pure oxygen
activated sludge was very competitive with air activated sludge at higher
flows. It is expected tha.t in the nitrification mode that pure oxygen
would be even more attractive and that pure oxygen may be the suspended
growth process of choice at flows over 10 MGD.
In conclusion, this study indicates that the RBC process is a viable
and economic alternative for combined carbon oxidation-nitrification
for wastewater flows possibly as high as 15 MGD. At higher flow rates
the single stage suspended growth processes (air or oxygen activated
sludge) appear to have a cost advantage. Because the RBC process has
been shown to be cost-effective over a wide range of flows it should be
seriously considered in future facilities planning work that involves
combined carbon oxidation-nitrification.
It has also been shown that RBC design criteria has undergone a
gradual evolution between 1972 and 1979. These changes have outdated
information contained in fairly recently published and widely used
engineering references including EPA's Process Design Manual For Nitrogen
Control (1975),13 Antonie's Fixed Biological Surfaces-Wastewater Treatment^
(1976),15 and Metcalf ^ Eddy's Wastewater Engineering; Treatment,
Disposal and Reuse; 2nd Edition (1979).1S The criteria contained in
these texts is no longer valid and its application will result in the
provision of insufficient media for nitrification and in unsatisfactory
RBC performance.
References
1. * National Research Council, Division of Medical Science: Report of
the Subcommittee on Sewage Treatment, Committee on Sanitary Engineering;
May, 1946.
2. T. Stones; Investigation on Biological Filtration at Salford;. Journal
of the Institute of Sewage Purification, No. 5; p. 406; (1961).
3. G.R. Grantham; Trickling Filter Performance at Intermediate Loading
Rates; Sewage and Industrial Wastes, 23, No. 10; .p. 1227.; (1951).
4. F.J. Burgess, et a.1; Evaluation Criteria for Deep Trickling Filters;
Journal of the Water Pollution Control Federation, 33, No. 8; p. 787;
(1961).
5. S. Eala Krishnanj et'.al; Nitrogen Removal by Modified Activated Sludge
Process; Journal of the ASCE Sanitary Engineering Division; p. 501;
(April, 1970).
1381
-------
6. G.A. Buddies, et al; Application of plastic Media Trickling Filters
for Biologica.l Nitrification Sys.tems; USEPA Environmental Protection
Technology Series,. EPA-R2-73-19.9; . (June, 19731,
7. R.J. Stenquigt, et al; Ca.rbon* Oxidation - Nitrification in Synthetic
Media Trickling Filters; Journal of the Water Pollution Control
Federation, 46, No. 10; p. 2327; (1974).
8. Autotrol Corporation; Wastewater Treatment Systems Design Manual;
1979.
9. Autotrol Corporation; Bio-Surf Design Manual; October, 1972.
10. R.L. Antonie; Nitrification and Denitrification with the Bio-Surf
Process; Presented at the Annual Meeting of the New England Water
Pollution Control Association, Kennebunkpbrt, MA. (June, 1974),
IJ.. S.K. Malhotra, et al; Performance of a Bio-Disk Plant in a Northern
Community; Presented at the 1975 WPCF Annual Conference, Miami, FLi;
(October, 1975).
12. United States Environmental Protection Agency; Nitrification and
Denitrification Facilities: Wastewater Treatment; August, 1973;
p. 4.
13. United States Environmental Protection Agency; Process Design Manual
for Nitrogen Control; October, 1975; p. 4 - 44.
14. L.A. Lundberg and J.L. Pierce; Comparative Cost-Effectiveness Analysis
of Rotating Biological Contactor and Activated Sludge Processes For
Carbon Oxidation; Presented at the First National Symposium on RBC
Technology, Seven Springs, PA (February, 1980).
15. R.L. Antonie; Fixed Biological Surfaces - Wastewater Treatment; 1976;
p. 56.
16. Metcalf & Eddy, Inc; Wastewater Engineering: Treatment, Disposal,
Reuse, 2nd Edition; 1979; p. 720.
1382
-------
COMPUTERIZED COST EFFECTIVE ANALYSIS
OF
FIXED FILM NITRIFICATON SYSTEMS
By
Paul T. Sun
Steve R. Struss
Clark Dietz Engineers, Inc.
Urbana, Illinois
Murdock J. Cullinane, Jr.
U.S. Army Engineer Waterways Experiment Station
Vicksburg, Mississippi
INTRODUCTION
Trickling filter tower systems and rotating biological contactors have been used
extensively in the nitrification of municipal secondary effluents. In order to
make realistic cost comparisons between these two systems, preliminary process
designs have to be carried out first and capital as well as 0 & M. costs are
estimated based on these preliminary sizing calculations. Several parametric
type cost curves are available in the literature which relate cost to some
sizing factor; such as volume of media or wastewater flow treated (1), (2), (3)
and (4). These types of curves have several disadvantages. First, they are
out-dated almost as soon as they are published. Based on the authors'
experience, using general cost indices to update these curves is not very
successful. Second, the general cost curve approach is not sensitive enough for
site-specific estimation, thus limiting the usefulness of the curves in
upgrading existing treatment works.
A computerized approach may be used to accurately estimate the capital and 0 & M
costs of both trickling filter and rotating biological contactor systems. Cost.
models for these processes have been developed as a part of a comprehensive
computer system; CAPDET (Computer Assisted Procedures for the Design and
Evaluation of Wastewater Treatment Facilities). (1)
1383
-------
General Description of CAPDET
The development of CAPDET is jointly funded by U.S. Army Corps of Engineers and
U.S. Environmental Protection Agency. It is intended to be a screening tool
capable of providing a methodology whereby a large number of alternative
wastewater treatment systems, each capable of meeting specified effluent
criteria, can be simultaneously ranked on the basis of cost effectiveness.
A new cost estimating approach, a "modified cost element method" is used in
CAPDET. In the true cost element approach, construction details are well enough
defined to adequately estimate the quantities of materials, manhours of labor,
etc. necessary to build and operate the facility. Whereas, the modified cost
element approach limits the detail by selecting only those cost elements which
are a major impact on the total cost of the facility. Thus, total cost
associated with a unit process is expressed as a sum of major component parts
plus a percentage. The cost estimating procedures in CAPDET, for most unit
processes, were developed on the premise that cost elements would identify at
least 85% of the capital costs and 75% of the operation, maintenance and repair
costs. The remainder of the costs would be expressed as a percentage of the sum
of element costs.
The major cost items for construction of most unit process can generally be
categorized as follows:
Earthwork
Concrete or Steel Basins
Installed Equipment
Piping
Building
Electrical and Control System
Miscellaneous Support Facilities
Detailed design formulae are developed to calculate these material quantities
based on the design flow, influent pollutant concentrations, effluent limita-
tions and user specified design parameters.
The cost of materials generally used in construction can be priced with readily
available unit costs (6). However, equipment for the wastewater treatment
system constitutes one of the largest items of fixed capital costs. It is
desirable therefore to maintain up-to-date equipment cost data for CAPDET. With
a limited number of unit cost input entries, it is very difficult to maintain
reliable cost data.
The following description outlines a procedure which produces an accurate
estimate of equipment cost without necessitating a vast number of unit price
inputs. The installed equipment cost may be considered in three components:
the purchase cost of the equipment, installation labor cost, and other minor
costs, such as electrical works, minor piping, foundations, painting, etc. The
purchase cost of process equipment is a function of size or capacity. , To
minimize the number of inputs required, a standard size (or capacity) unit is
selected and the purchase cost of all other size (or capacity) units of that
type is expressed as a fraction or multiple of the standard unit purchase cost.
This cost ratio versus size relationship has been developed for each major item
of equipment required in the design. These relationships assume the form shown:
1384
-------
(COST)
<
(COST)
A
_c
A
(1)
where
A = some characteristic size measurement such as volume, area, horse power,
weight.
0 and S = subscripts designating other and standard sizes, respectively.
F = a function of.
The exact form of the cost-versus-size relationship and the selection of the
standard sizes for each major equipment item are determined based on manufac-
turer's information and available literature. In most cases these size-cost
relationships are relatively unaffected by inflation and other cost changes.
The CAPDET user has two options by which the purchase cost of equipment can be
escalated to account for inflation. The first option is for the user to obtain
from equipment manufacturers the current purchase cost of the standard size
equipment. The purchase cost of any other size equipment is then automatically
escalated by the cost versus size relationships which have been developed in the
model. The second option is to escalate the purchase cost by the use of cost
indices. Only one input is required for this process: the Marshall and Swift
Equipment Cost Index from Chemical Engineering magazine. The 1977 first quarter
purchase prices of the standard, size equipment are stored in the model and are
updated automatically if the M & S cost index is input into the program. The
latter of the above methods is the least accurate, however, it provides default
values when current prices are not available.
Equipment installation cost is estimated by multiplying the manhour
requirements obtained in the design by user input labor rates. The other minor
costs for each type of equipment consist of piping, concrete, steel,
instruments, electrical, insulation, painting, insurance, taxes, etc. The cost
of these items are estimated as a percentage of purchase costs and will vary
with the type and size of equipment. These percentage values are established
based on design experience, engineering judgement, manufacturer's input, and
previously published literature.
The total construction cost of a unit process is thus the sum of the cost of
installed equipment and the general construction material cost such as
earthwork, reinforced concrete in place, piping, etc.
The operation and maintenance for a wastewater treatment facility can be divided
into several major categories: power, operation labor, maintenance labor,
chemical costs, and repair material costs.
The electrical power consumption has been determined for each unit process. The
power consumption for the treatment facility is simply the totalized power
consumption for the unit processes. The power consumption is converted to a
cost by multiplying the power consumption in kilowatt-hours per year by the user
selected unit price input for electric power. The user input unit cost should
be obtained from the utility supplying power to the proposed facility.
1385
-------
The operation and maintenance labor costs can be divided into four groups:
administration and general labor, operation labor, maintenance labor, and
laboratory labor. Recommended staffing for different levels of manpower
required for each of the four labor groups are established by the model. By
utilizing staffing charts provided in the literature, weighted average salaries
for each labor group may be established. To reduce the number of inputs, the
weighted average salaries are expressed as a percent of operator II salary.
The maintenance and operation labor requirements are calculated for each
individual unit process. The total requirement is the sum of the requirements
for each unit process used in the treatment facility. The total annual labor
cost is the sum of the labor costs for each of the four labor groups.
The cost of operation and maintenance materials and supplies is calculated for
each unit process as a function of the construction cost.
The total annual operation and maintenance cost is the sum of the electric power
costs, operation and maintenance labor costs, operation and maintenance
material and supply costs, and chemical costs.
In order for the model to present a meaningful comparison of alternatives, a
cost evaluation utilizing "time-value of money concepts" is necessary. Simply
stated, the method chosen for the economic evaluation is to compare annual costs
computed over a fixed evaluation period. A minimum evaluation period of 20
years is utilized in the CAPDET model. The cost evaluation has been presented
such that individual unit processes or total treatment facilities may be
evaluated.
Current Study
The following sections describe in detail the design and cost estimating
procedures for the fixed film nitrification processes. Finally, the CAPDET
model is used to compare the cost effectiveness of the two fixed film nitrifi-
cation processes under various conditions.
TRICKLING FILTER NITRIFICATION SYSTEMS
The trickling filter process has been successfully utilized in a number of cases
where nitrification of secondary effuent was required. Synthetic media, both
low and high specific surface area types, are used in supporting, the surface
growth of nitrifiers. These nitrifiers have a very low yield factor, i.e.,!low
sludge production rate; the BOD and suspended solids concentrations of the
effluent are essentially the same as the influent. Therefore, final clarifiers
are not generally provided in fixed film growth nitrification systems. Also,
due to the light weight and high structural strength of the synthetic media,
trickling filter towers can be built as tall as 28 feet without external
support, thus making them economical in achieving a high degree of
nitrification.
However, because of the height of modern trickling filter towers, gravity feed
which is common among the old rock filters becomes a rarity. Some type of
intermediate pumping is necessary to deliver the secondary effluent to the top
of the filters. Therefore, a pumping station should be considered as part of
the trickling filter unit process and accordingly, the associated capital and
operation and maintenance costs should be included.
1386
-------
In this section, procedures for trickling filter tower design are outlined. The
cost estimating procedures are also described. Procedures for the design and
cost estimate for intermediate pumping station are discussed in the next
section.
Process Design Procedures
Calculate Media Surface Area Requirement
Very few trickling filter nitrification studies have been published. Reference
(7) provides the best summary of existing information. It is adopted in the
CAPDET model for process calculations. The following equations can be used to
calculate the required surface area based upon effluent concentration and
wastewater temperature. If different concentration limitations are imposed for
summer and winter conditions, the larger of the two surface area requirements is
adopted as the appropriate design.
First, a critical ammonia effluent concentration is defined.
NC = 4.5 - 0.115 (T) (2)
where
N = Critical ammonia effluent concentration, mg/1 as N. This gives the break
points in the lines given in Reference (7).
T = Wastewater temperature, °C
The required surface area is then calculated:
If N «£ N
If N > N
e c
S = 6900 - 190 (T) (3a)
S = (21250 - 527 T) - 3410 Ng (3b)
where
S = Required surface area per pound of NIL, - N oxidized per day, sq. ft.
N = Effluent ammonia concentration, mg/1 as N
Calculate Volume of Media Required
V, = - x (N - N ) x 8.34 x Q (4)
d n o e avg.
where
V, = Volume of media required, cu. ft.
1387
-------
N = Influent ammonia concentration, mg/1 as N
Q = Averaged design wastewater flow, mgd
n = Specific surface area of the media selected. Values range from 28 to 41 sq.
ft./cu. ft. depending on manufacturer's specifications
Calculate the Cross-Sectional Area and Depth of the Tower System
A surface hydraulic loading rate of 0.75 gpm/sq. ft. at design flow is recom-
mended by the manufacturers (8) (9). The cross-sectional area then would be
SA = 925.3 x Q
avg.
(5)
where
SA = Cross-sectional area of filter, sq. ft.
And the depth of the tower is then calculated:
D = Vd/SA
(6)
where
D = Depth or height of the media, ft.
The maximum recommended height of a tower, without external structural support,
is approximately 28 feet. If the calculated depth, D, is less than 28 feet, the
process design has been accomplished. A recirculation ratio of 1:1 at Q is
adequate. The pumping system should therefore be designed to handle twice"the
amount of the design flow or the peak flow, Q , , whichever is larger. However,
under certain circumstances, it is possible tnat the calculated height would be
larger than the maximum allowable value. An iteration procedure is provided in
CAPDET to increase the tower cross-sectional area thus decreasing the height to
acceptable levels. The required surface loading rate for complete wetting is
provided by recirculation. These calculations would yield the design firm
pumping capacity of the intermediate pumping station, Q-r.p, expressed in mgd.
FP'
System Design
Field experience indicates that two towers are usually provided until the cross-
sectional area of each tower reaches 17,700 sq. ft. or (150 ft. in diameter).
When individual tower cross-sectional area, SA/2, is larger than this value,
three or more towers are used. The surface area of the filter tower is limited
by the available sizes of the rotary distribution arms. These arms are
generally in the range of 20 to 200 feet. Figure 1 provides an experience curve
which indicates the number of towers generally designed under various
conditions. After the number of towers,
individual tower can then be calculated;
N , is selected, the diameter of each
1388
-------
DIA = 1.128 (
1/2
D x N
(7)
where
DIA = Diameter of individual tower, ft.
N = Number of towers
Estimate of Construction Costs
Construction of a typical modern trickling filter includes the media, rotating
distribution arm system with its central column support, underdrain system, and
the outside wall. The plastic media usually is supplied and installed by the
manufacturers and the cost of the installed system is estimated as dollars/ cu.
ft. Polyester fiberglass, light weight steel, and precast double-tee concrete
construction have all been used as the media containment structure. In CAPDET,
study, it is assumed that a 6-inch cast-in-place reinforced concrete wall is
utilized. The media support system consists of precast beams and concrete
support posts as suggested by one of the manufacturers (9). The underdrain
system includes the drainage floor and channel, sidewall with air openings and
louvers. The distributor system is supported by a concrete column extending from
the floor to the top of the tower. The following procedures are designed to
estimate the material takeoff of these major items and subsequently, the
construction cost by using unit price inputs.
Material Takeoff
The major cost items in a trickling filter system include cost of earthwork,
cost of reinforced concrete in place, cost of medium and cost of distributor
arm. Table 1 summarizes the algorithms for calculating the quantities of
material required. These quantities are related to the specific dimensions of a
selected tower.
Cost of Equipment
Two major items are involved here, the cost of the media and the cost of the
distributor arms. The cost of the plastic sheet media ranges from $2.50 to
$4.00 per cu. ft. depending on the specific surface area. The estimate of the
cost of the media, $ ,. , is straight forward
' media' *
$ .. = 0.785 x (DIA)'
vmedia
x D x N x UPIMC
(8)
where
UPIMC = Media unit cost, a user input in dollars per cu. ft.
Calculating the cost of distributor arm system is more complicated. The
purchase cost of this type of equipment can be related to the cost of a standard
size arm, in this case, a distributor arm with a diameter of 50 ft.
1389
-------
s
L
3 O O O <0
T tO CVJ —
103 5 I04 5 I05 5 I06
CALCULATED NECESSARY CROSS SECTJONAL AREA (3A),SQ. FT.
CO
o:
UJ
o:
UJ
UJ
cc
CD
U_
o
01
U-
o
o
1
UJ
CO
UJ
o:
£
I
Q
UJ
O
O
cr
QL
1390
-------
Table 1
Quantities of Material for the Construction
of a Trickling Filter System
I. Earthwork Quantities, V , (cu. yd.)
V = 1.48 x 10"3 (DIA)3 + 0.208 (DIA)2 + 3.28 (DIA) + 14.9
II. Reinforced Concrete Wall, V , (cu. yd.)
_ ' cw' •*
i. External wall
V = 0.0581 (D + 3) (DIA) + 0.467 (DIA)
C.W C
ii. Media support system
If DIA < 40 ft.
V = 5.33 x 10"5 (DIA)3'08 + 7.41 x 10"4 (DIA)2'963
c ws
If DIA>-40 ft.
Vcws = 3'64 X
iii. V = V + V
cw ewe cws
0.023^(DIA)
2>°35
III. Reinforced Concrete Slab, V , (cu. yd.)
If DIA <70 ft.
V = 0.019 (DIA)2 + 0.37 (DIA) + 0.151 D + 0.3
cs
if DIA:>?O ft.
V =0.019 (DIA)2 +0.63 (DIA) + 0.6 D + 1.2
cs
V = Volume of earthwork, cu. yd.
ew ' J
V = Volume of reinforce'd concrete wall in place, cu. yd.
cw y ' 3
V = Volume of reinforced concrete slab in place, cu. yd.
cs * ' '
1391
-------
The relationship is;
$
Arm
1.367 + 0.01265
(DIA)] x $
Arm - 50
(9)
where
$. = The purchase price of a distributor arm with a diameter of DIA feet,
Arm dollars
$. _n = The standard size arm cost to be specified by the user, dollars
Construction Cost
The construction cost of the tower system: minus the usual miscellaneous costs
such as contractor's profit, engineering fee, general site work and others, can
be estimated by using standard construction unit costs. The relationship is:
$„-, = 1.10 f$ .. + N. (1.32 $. + $ + $ + $
YTF L media t vArm vew vcw y
cs
(10)
where
$_ = Construction cost of trickling filter systems, dollars
J.JB
1.10 = Cost increase accounting for the minor items such as piping, stairway
and etc.
1.32 = Cost escalation accounting for the installation of the arm system
and
(lla) !
(lib)
$ = V x UPIEW
vew ew
$ = V x UPICW
cw cw
$ = V x UPICS
Ycs cs
UPIEW = Unit price input of earthwork, dollars per cu. yd.
(lie)
UPICW = Unit price input of reinforced concrete wall in place, dollars per cu.
yd.
UPICS = Unit price input of reinforced concrete slab in place, dollars per cu.
yd.
Estimate Operation and Maintenance Cost
The operation and maintenance (0 & M) cost of running a trickling filter system
includes the cost of manpower and that of repair and maintenance material. The
tower itself does not require any significant amount of energy. Although in
certain areas of the U.S., towers are equipped with mechanical ventilation
devices, the energy required to operate these fans is minimal. Thus, the only
energy required to operate the trickling filter nitrification process would be
pumping which will be described in the next section.
1392
-------
0 & M Manpower
Reference (1) supplies 0 & M manpower requirements with respect to the surface
area of the towers:
If SA < 11,000 sq. ft.
OMH = 10 (SA)°'3° (12)
and
MMH = 11.5 (SA)0-26 (13)
If SA>11,000 sq. ft.
OMH = 0.2 x (SA)0'7 (14)
MMH = 0.046 x (SA)°-86 (15)
The total operation and maintenance manpower requirements would be:
OMMH = OMH + MMH (16)
where
OMH = Operation manpower requirement, manhrs/yr.
MMH = Maintenance manpower requirement, manhrs/yr.
OMMH = 0 & M manpower requirement, manhrs/yr.
SA = Cross-sectional area of towers, sq. ft.
Operation and Maintenance Material Cost
This item includes repair and replacement material cost and cost of major
maintenance work performed by outside contractors. It is approximated by a
percentage of the construction cost of the tower system (1).
$MR = 0.006 x $TF (17)
where
$,„ = Annual maintenance and repair material cost .
Total 0 & M Cost
The total 0 & M cost can be easily estimated by using a unit price input for the
labor rate.
$om = (OMMH)(MC) + $m
where
1393
-------
$ = Annual 0 & M cost, dollars/yr.
MC = Manpower cost, doliars/manhr.
INTERMEDIATE PUMPING STATIONS
An intermediate pumping station is an integral part of a modern trickling filter
system. The secondary effluent from the final clarifier will flow into a wet
well and then be lifted by pumps to the top of the tower. The pumps not only
have to deliver the wastewater volume but also to create enough head to cause
the distributor arms to rotate. Ordinarily, constant speed centrifugal pumps
are employed. The logic for selection of a constant speed pump is that a more or
less constant hydraulic loading to the towers is always maintained in field
practice.
The intermediate pumping station includes the pumps and their associated piping
and controls, the wet well, dry well and motor housing. The preliminary design
and cost estimating procedures are described briefly here.
Process Design Procedures
In pumping system design, two parameters will have to be defined first; design
flow or firm pumping capacity and the total dynamic head. The design firm
pumping capacity,
trickling filter.
0,,,,, has been specified by the logic in the section on
cr
The total dynamic head of this pumping system would include (1) the static head,
(2) the frictional head created by flow velocity through the pipes and (3) the
dynamic head required to rotate the distributor arms. Experience has shown that
the following equation gives a reasonable prediction for various flow and site
conditions:
TDK = D + 12
(19)
where
TDK = Total dynamic head, ft.
D = Height of the filter media, ft.
Pump Selection
In selecting the number of pumps within the pump station, the rule-of-thumb
summarized in the following will be used.
Selection of Number of Pumps Within
a Pump Station
, (mgd)
Less than 5 mgd
5-50 ragd
50 - 150 mgd
more than 150 mgd
Number of Pumps, N
2
3
4
5
1394
-------
The number selected, N , includes one standby pump. For simplicity, the same
size pumps are used in mis model, then, each individual pump would handle:
QTP
q = 694.4 N*\ 1 (20)
P
where
q = Capacity of each pump, gpm
It is customary to add a safety factor to the calculated pump capacity. Values
from 10 to 50 percent have been suggested (10). The following equation gives a
relationship between the suggested safety factor and the sizes of pumps.
SF = 1.9 - 0.2 Log q (21)
and SF is always larger or equal to 1.1
where
SF = Safety factor for pump selection
Thus, the design flow for each pump would be
qd = (SFp) x (q) (22)
where
q, = Design capacity of the selected pump, gpm
Experience has shown that the pump discharge nozzle velocity should be in the
range of 11 to 14 feet per second for wastewater-applications. And the sizing
procedure would be to calculate the pump's discharge nozzle diameter.
d = 0.1927 q, (23)
p M
d = Pump sizes in inches of discharge nozzle diameter. The available pump
P sizes are 4", 6", 8", 10", 12", 14", 16", 18", 20", 24", 30", 36", 42",
48", 54", 60" and 72". A step function is employed in CAPDET to select
the appropriate pump from the available sizes.
Drive Selection
Electric motors are the most commonly employed pump drivers in a wastewater
treatment plant. The cost of a motor is related to its rated horsepower and
rotating speed. The rotating speed is usually governed by the characteristics of
the pumps selected. An important index of centrifugal pumps is the specific
speed parameter. It is considered as a shape factor.
1395
-------
The specific speed is defined as (10):
r x
NS = r X - (24)
(TDH)u./i
where
NS = Specific speed
r = Impeller rotating speed, rpm
A typical value of 4000 has shown to be most appropriate for sewage applications
(10). Thus the rotating speed for the motor would be:
r =
4°°°
C25)
On the other hand, the brake horsepower to be delivered by the motor is directly
related to the flow and the head;
(qd) (TDK)
BHp = 3959.7 (E)
(26)
where
BH = Brake horsepower of the motor, H
P P
E = Pump efficiency
Pump efficiency is a variable changing with the head it is pumping against.
However, it is a rule that the larger the pump the more efficient it is. And a
general relationship has been developed as follows (12):
"5
E = 0.63 + 2.42 x 10" x (qd)
and E is always less than or equal to 0.90
(27)
Again, a step function is used in CAPDET to select the motor from commercially
available sizes.
Estimate of Construction Costs
The major construction items of an intermediate pumping station, ^besides the
cost of pumps and equipment, consists of the wet well, dry well and motor
control housing. The wet well and dry well are usually underground structures
which can be priced by estimating the quantities of earthwork and reinforced
concrete required. The cost of the building housing the motors and their
control panels can be approximated by multiplying a unit cost to the square
footage of the building.
1396
-------
Material Takeoff
The general algorithms for calculating material quantities for the pumping
station construction is summarized in Table 3. The relationships are developed
from field data. They are related to the firm pumping capacity, (X^.
£ c
Cost of Equipment
Two major cost items are to be developed for each pump station; the cost of pumps
and the cost of the drive units. The other costs such as piping, control,
ventilation, dehumidification can be substantial. However, historically, the
"minor" costs have been estimated as a percent of the cost of pump and driver.
The purchase price of a single pump is a nonlinear function of the pump size
expressed as the diameter of the discharge nozzle. The purchase cost of pumps
can be related to the cost of a standard size pump and 16 inch pump has been
found to be the most appropriate. The relationship is shown as follows: (12)
If d < 16 inches
P
If d > 16 inches
P
where
$p= 0.041
($p.l6)
$ = 0.01 (d )1'663 x ($ )
vp P P ~ 16
(28a)
(28b)
d = Pump size, inches
$ = Purchase price of pump at size d , dollars
$ , = Purchase price of a 16 inch-0 pump, dollars
The purchase price of an AC motor is a function of the power it can deliver, BH ,
and its synchronous speed. Again, the price of any motor can be related to that
of the standard. In this case, it is an induction motor with 100 Hp capacity and
1200 rpm speed. The relationship is shown as: (13)
(BH + 5)
0.85
$ = 3.9
m
0.75
- 100
>
(29)
where
$ = Purchase price of motor, dollars
•m - 100
= Purchase price of the standard motor, dollars
1397
-------
Table 3
Quantities of Material for the Construction
of an Intermediate Pumping Station
I. Earthwork Quantities, V , (cu. yd.)
SW
If QFp < 20 mgd
= 1,519 (QFp)
0.22
If QFp > 20 mgd
Vew = 591
V0.538
II. Reinforced Concrete Wall, V , (cu. yd.)
cw
If QFp < 10 mgd
Vcw = 77'8
If QFp > 10 mgd
V0.147
=50.3 (QFp)
0.337
III. Reinforced Concrete Slab, V , (cu. yd.)
=80.4 (QFp)
0.408
IV. Pump Station Building Area, A, , (sq. ft.)
= 19.7 (QFp) + 640
1398
-------
Guthrie (14) gives the following factors to be used in estimating field instal-
lation cost for pumps and their drive units;
Items
Equipment Cost (pump and motor)
Field Materials
Piping
Concrete
Instruments
Electrical
Insulation
Paint
Field Labor
Material Erection
Equipment Setting
Percent of Equipment Cost
100.0 %
29.6 %
3.9 %
2.9 %
30.3 %
2.5 %
0.8 %
59.0 %
8.9 %
Total Direct Cost
Construction Cost
237.9 %
The construction cost of an intermediate pumping station can thus be estimated
by:
$ps = 1.15 x
Np x (2.38)($p
$ ) + $
vm ew
$
cw
$ + $T
cs \
(30)
where
$p,, = Construction cost of intermediate pumping station, dollars
1.15 = Cost increases accounting for the minor articles such as overhead crane,
hvac, etc.
and
$b = Aj^ x UPIBC
(31)
Where UPIBC is the unit price for building expressed in dollars per square feet.
Other symbols have been defined earlier.
Estimate of Operation and Maintenance Costs
The major items of the intermediate pump station 0 & M cost are: 0 & M manpower
requirement, energy and material and supply cost. They are estimated individ-
ually in the following subsections.
0 & M Manpower
The available information relates the 0 & M manpower requirement with the firr
pumping capacity of the station (3) .
1399
-------
If 0 < QFp < 7 ragd
OMH = 440 Q
MMH = 360 Q
0.129
!FP
0.148
:FP
If 7 < QVD < 30 mgd
TFP
OMH = 294 QT
0.335
MMH = 255 QFp
0.325
If 30 <• QFp < 80 mgd
OMH = 40 Q
0.866
!FP
MMH = 86 Q.
0.646
!FP
1.012
0.881
(32)
(33)
(34)
(35)
(36)
(37)
'' (38)
(39)
If QFp>80 mgd
OMH = 21
MMH = 31 QFpv
where
OMH = Operation manpower requirement, manhrs/yr.
MMH — Maintenance manpower requirement, manhrs/yr.
Q__. = Firm pumping capacity, mgd
jjJt
Total manpower requirement OMMH would be
OMMH = OMH + MMH (40)
Annual Power Consumption
As mentioned previously, essentially the only power requirement for the
trickling filter system is the pumping process. The usual operating procedure
is to keep the hydraulic loading rate around 0.75 gpra/sq. ft. Thus, the
operating flow would be:
Q =0.589 (Nt)(DIA):
(41)
where
q = Operation flow, gpm
The pumping energy would be:
KWH = 1.66 x
(qJ(TDH)
(42)
1400
-------
where
KWH = Annual power requirement, kwhr/yr.
Operation and Maintenance Material Cost
This item includes repair and replacement material cost and the cost of major
maintenance works to be performed by outside contractors. It can be expressed
as a percentage of the construction cost of the intermediate pump station (1) .
(43)
where
$,_ = Cost of maintenance and repair material, dollars/yr.
Total 0 & M Cost
$ = (OMMH)(MC) + (KWH) (EC) + $m
oni riK
(44)
ROTATING BIOLOGICAL CONTACTOR NITRIFICATION SYSTEM
The RBC is a relatively new process.
nitrification on secondary effluents.
It has also been utilized in promoting
The basic principles governing the performance of the RBC systems are the same
as those for trickling filters. The only major process difference between these
two systems is that for the RBC process the aeration and mass transfer
mechanisms are provided by rotating the media in and out of the wastewaters,
whereas for trickling filter systems, aeration and mass transfer mechanisms are
provided by passing the wastewater over the media.
The following subsections provide the design and cost estimating procedures for
the RBC system.
Process Design Procedures
Figure 2 shows the experimental relationships between surface nitrification
rate and the exposed ammonia concentration for effluents (15). As oiie can
observe, the rate is related to the temperature and the effluent ammonia
concentration desired. The information presented in Figure 4 can be modeled by
using the same technique proposed by Keinath (16). The resulting design
equation is as follows:
- 15,125
TCF
8.6 In ()
e
(NQ - N )
(45)
where
A,, = Surface area required per million gallons of wastewater flow, sq. ft./mgd
1401
-------
I-6
cvT
.5
LU
cc
I
1
to
X
z
.3
.2
0
GUELPH,ONTARIO NITRIFICATION
AT VARIOUS TEMPERATURES
MADISON ,WISC. DATA
€5°F —
55°F
55°F
1.5 3.0 4.5 6.0 7.5 9.0 10.5 12.0 13,5 15.0
NH3-N CONG, mg/1
FIGURE 2
AMMONIA REMOVAL RATE
FOR RBC SYSTEMS
1402
-------
TCF = Temperature correction factor
and TCF, Temperature Correction Factor can be defined as:
When T is larger than 18° C
TCF = 1.4
When T is smaller than 18° C
TCF = 0.0787 x T
(46a)
(46b)
If different effluent concentration limitations are given for summer and winter
conditions, the surface area requirement for each condition must be determined
and the final design based on the condition that requires a larger media
surface.
The required media surface area would be:
A = Q x A.,
avg. N
(47)
where
A = Required media surface area, sq. ft.
Estimate of Construction Costs
Working from the required media surface area, a physical layout of the plant is
assumed from which, construction costs, such as excavation, equipment and
concrete, are derived.
The first step in determining capital costs is to calculate the number of RBC
shafts required to supply the design media surface area. Since generally a
large amount of surface area is needed to siipport the less efficient nitrifying
biomass, shafts having areas of 150,000 ft. each are commonly used. This shaft
has physical dimension of 11'-10" in diameter and 25' in length. The specific
surface area is approximately 54.5 sq.ft./cu.ft. (17) The number of shafts
required may be calculated as follows:
A
Ssh
(48)
where
N , = Number of shafts necessary
A = Total media surface area required, sq. ft.
A , = Media surface area per shaft, sq.ft./shaft, 150,000 sq.ft./shaft
For practical reasons the value for N , is rounded up to the next whole number to
avoid working with fractional shafts in subsequent calculations.
1403
-------
Material Takeoff
For simplicity of design, it is assumed that shafts will be arranged in groups
of eight. Each group is called a bank and consists of two end shafts and six
intermediate shafts. Any shafts in excess of a multiple number of eight are
placed in a partial bank of from one to seven shafts as needed. Many other
configurations are possible; however, varying the configurations does not
affect earthwork, concrete, and other related material requirements signifi-
cantly. The number of full and partial banks, N, , is, therefore, the number of
shafts divided by eight and rounded up to the next whole number.
Nb = Nsh/8
(49)
where
N, = Number of banks
D
Each bank will consist of two end shafts and from zero to six intermediate
and intermediate shafts, N. , are
shafts. The number of end shafts,
calculated as follows:
N
es
is •
N = 2 N,
es b
N. = N , - N
is sh es
(50)
(51)
Using these figures and the associated volumes they represent, the earthwork and
concrete requirements are calculated.
where
V = 130 N + 142 N. (52)
ew es is
V = 23 N + 20.7 N. (53)
cs es is
V = 11.5 N + 8.6 N. (54)
cw es is
V = Volume of earthwork required, cubic yards
GW
V = Volume of concrete slab required, cubic yards
CS
V = Volume of concrete wall required, cubic yards
Construction Cost
The bare construction cost of the RBC system consists of the costs of earthwork,
concrete works, installed media and other minor items. The RBC shafts are
purchased as a complete unit from the manufacturers. Usually the package
includes the media, protective fiberglass cover, driving unit and other miscel-
laneous items. A factor of 15% is added to the RBC media purchase cost for
installation.
1404
-------
The construction cost of a RBC system would therefore be:
fl.15 (N , ) x $ ,+ $ + $ + $ 1
[_ v sh' ysh vew vcw vcs
(55)
where
$ = Construction cost of RBC system, dollars
KisC
1.10 = Cost increase accounting for minor items such as site cleaning, piping,
electrical work and etc.
Estimate of Operation and Maintenance Costs
The major components of the RBC operation and maintenance cost are the 0 & M
manpower requirement, energy and material and supply cost.
0 & M Manpower
Reference (17) provides an estimate of manpower for the operation and
maintenance of the RBC system;
For a plant with less than 30 shafts
OMMH = 52 N (1.25 - 0.025 N ) (56a)
sp sp
For a plant with more than 30 shafts
OMMH = 26 x N , (56b)
sn
where
OMMH = Operation and maintenance manpower, manhrs/year
Annual Power Consumption
A power consumption of 5 Hp/shaft is used in estimating the power consumption of
the RBC system. (18)
KWH = 32,630 x Ngh (57)
where
KWH = Annual power consumption, kwh/yr.
Operation and Maintenance Material Cost
This item includes repair and replacement material cost and the cost of major
maintenance works performed by outside contractors. It is expressed as a
percent of installed cost of the RBC equipment.
= 0.01
(58)
1405
-------
Total 0 & M Cost
where
= (OMMH)CMC) + (KWH)(EC).+
(59)
$ = Annual 0 & M cost, dollars/yr.
MC = Manpower cost, dollars/mannr.
EC = Electricity cost, dollars/kwhr
COST EVALUATION OF FIXED FILM NITRIFICATION SYSTEMS
The CAPDET model was used to perform the comparative cost evaluation of the two
fixed film nitrification systems described above. The inputs used for the cost
evaluation are summarized in Table 4. The analysis was performed for wastewater
flows ranging from 0.5 to 50 mgd. The CAPDET model allows change of any or all
of the parameters presented by Table 4 in order to reflect local design
conditions. However, space limits the quantity of data which can be presented
in this paper. In addition to the design factors used, the assumption is also
made that the peak flow is twice the average flow in design of the intermediate
pumping station associated with the trickling filter tower.
The results of the cost analysis conducted during this study are presented
graphically by Figures 3 and 4. As stated, this study was designed to evaluate
the affects of increasing flows on the relative cost effectiveness of the two
fixed film nitrification systems.
The values presented in these figures represent the bare construction, operation
and maintenance costs of the individual unit processes. Contractor profit and
overhead, mobilization and sitework as well as engineering costs are not
included.
Under the conditions presented in Table 4, the RBC system is more cost effective
than, the trickling filter system when the design flow is less than 3 mgd.
However, for larger plants the trickling filter system becomes more competitive.
It is hypothesized that because of the modular nature of the RBC system, its
"economy of scale" is not as favorable as that of the trickling filter towers.
However, the air drived RBC system might change the cost picture significantly.
Another point that has to be emphasized here is that the process design equa-
tions for trickling filter nitrification systems were based on pilot studies
under controlled environment. No full scale plant performance data is available
to substantiate their accuracy. Whereas, the design equations for RBC was
derived from full scale plant studies. It is highly possible that the surface
areas requirement for tickling filter may have to be increased based upon future
findings. The CAPDET model can be easily upgraded to accommodate the constant
expanding knowledge of the process kinetics of fixed film nitrification
processes.
1406
-------
Table 4
Conditions Sepcified in the Cost Evaluation
of the Fixed Film Nitrification Systems
I. Influent, Effluent Conditions
i. Influent oxidizable nitrogen concentration: 25 mg/l-N
ii. Effluent ammonia limitation: summer 1 mg/l-N @ 24°C
winter 4 mg/l-N @ 9°C
II. Trickling Filter Nitrification System
i. Media cost: UPIMC = $3.50/cu.ft. @ 41 sq.ft./cu.ft.
ii. Standard size distributor arm cost: $. _n = $39,000
III. Intermediate Pumping Station
i. Standard size pump cost: $ _-,/- = $16,500
ii. Standard size motor cost: $ = $2,850
iii. Pump house building cost: UPIBC = $30/sq. ft.
IV. RBC Nitrification System
i. RBC medium unit cost: $39,000 @ 150,000 sq.ft./shaft
V. General Civil Works
i. Unit price of earthwork: UPIEW = $5.00/cu.yd.
ii. Unit price of reinforced concrete wall: UPICW = $300/cu.yd.
iii. Unit price of reinforced concrete slab: UPICS = $270/cu.yd.
VI. Operation and Maintenance Unit Costs
i. Labor rate: MC = $7.00/manhr.
ii. Cost of electricity: EC = $0.04/kwhr.
VII. Parameters Used In Economic Analysis
i. Interest rate: 6 7/8% annual
ii. Return period: 20 years
1407
-------
2
107
RBC CAPITAL
X
-TF CAPITAL
10*
>
•*&
cc
o
te"
8
RBCOaM
10*
iU_DJLM
iff1
10'
5 10 5 10
WWTP PLANT DESIGN FLOW, mgd
FIGURE 3
CAPITAL AND OSM COST FOR THE
FIXED FILM NITRIFICATION SYSTEMS
1408
-------
TREATMENT COST, 8/1000 GALLONS
0 P
D i_ ro
\
\
\
V
\
\
V
\
\
\
\
RBC SYS
/—TRIG
/
\
V
\
\
V
X^_
— -— .
TEMS^
)
X,
N
Xy
» ,
"""•-•-..,_
1 . •-
— «^»™^
mmt*m
0.5
5 10 20
WWTP PLANT DESIGN FLOW mgd
50
FIGURE 4
COMPARISON OF TREATMENT COSTS OF THE
TWO FIXED FILM NITRIFICATION SYSTEMS
1409
-------
SUMMARY
A detailed description, of the CAPDET model for the fixed film nitrification.
systems is presented here. The model is so designed with many flexibilities are
incorporated. Engineers can obtain an accurate cost analysis with few user
inputs. They can manipulate the unit price inputs to suit the local conditions,
if desired. For instance, for sites with difficult soil characteristics, a high
unit price value for earthwork can be.used, thus, yielding more realistic cost
comparisons. CAPDET is also a valuable research and education tool. It can
provide students the opportunity to study the cost savings of certain process
improvement and the optimization of overall treatment plant design.
1410
-------
REFERENCE
(1) Patterson and Banker, Estimating Costs and Manpower Requirements for
Conventional Wastewater Treatment Facilities, EPA 17090 DAN. 10/1971.
(2) USEPA MCD53, Innovative and Alternative Technology Assessment Manual, EPA
430/ 9-78-009.
(3) Metcalf ,and Eddy, Inc. "Water Pollution Abatement Technology:
Capabilities and Cost, Public Owned Treatment Works" PB-250690-01,
NTIS Springfield, VA. 1976.
(4) Benjes, H. H. , "Small Community Wastewater Treatment Facilities
Biological Treatment Systems". Prepared for the EPA Technology
Transfer Seminar on Small Wastewater Treatment Systems, March 1977.
(5) Clark Dietz Engineers, Inc., CAPDET Task I_I Report, contract No. DACW
39-77-C-G027, 1977.
(6) Dodge Guide, McGraw-Hill Information Systems Company.
(7) Brown and Caldwell. Process Design Manual for Nitrogen Control, USEPA
Technology Transfer, 1975.
T>
(8) Envirotech Co., "Surfpac Plastic Media Biological Oxidation Process".
1975.
(9) B. F. Goodrich, "Vinyl Core Media".
(10) Kraassik, et al, Pump Handbook, McGraw-Hill, 1976.
(11) Metcalf and Eddy, Inc., Wastewater Engineering Collection, Treatment,
Disposal, McGraw-Hill, 1972.
(12) Worthington Pumps, Public Works Engineers Manual, 1977.
(13) General Electric Co., Wastewater System Manual, 1976.
(14) Popper, H. Editor, Modern Cost Engineering Techniques, McGraw-Hill Co.,
1970.
(15) Wanielista, M. P. and W. W. Eckenfelder, Jr., Advances in Water and
Wastewater Treatment: Biological Nutrient Removal, Ann Arbor
Science, 1978.
(16) Keinath, T. M. and M. P. Wanielista, Mathematical Modeling for Water
Pollution Control Processes, Ann Arbor Science, 1975.
(17) Antonie, R. L. Fixed Biological Surfaces ^ Wastewater Treatment, CRC
Press, Inc., 1976.
(18) Wesner, G. M. et. al., Technical Report: Energy Conservation in Municipal
Wastewater Treatment, EPA 430/9-77-011 1978.
1411
-------
-------
. COMPARATIVE COST-EFFECTIVENESS ANALYSIS
OF ROTATING BIOLOGICAL CONTACTOR. AND
ACTIVATED SLUDGE PROCESSES FOR CARB.QN OXIDATION
By
Lee A, Lundberg, P.E.
Principal Engineer
Jeffrey L, Pierce, P.E.
Vice President
Schneider Consulting Engineers
Bridgeville, Pennsylvania
INTRODUCTION
The use of cost-effectiveness analysis as a decision-making tool in
facilities planning for wastewater treatment works is required under the
U.S. Environmental Protection Agency's regulations governing its construc-
tion grants program as currently promulgated under the Clean Water Act
of 1977 (P.L. 95-317). This tool has been applied routinely for municipal
treatment works since the original regulations were issued pursuant to the
Federal Water Pollution Control Act of 1972 (P.L. 92-500). When properly
•applied, it can provide a powerful insight into the total costs associated
with competing alternative courses of action and, when expanded upon, can
yield a system formula which, may be analyzed for sensitivity to differential
inflation factors, between the resources required for each alternative.
Unfortunately, the various methods historically used for cost-effectiveness
analysis have.been inconsistent throughout the engineering community due to
different interpretations, of the. regulatory guidelines; therefore,' the
resultant alterna,tlye .selection process/'has. tended to be, in many cases, more
subjective tha.n objective in .nature. -'In certain instances, these inconsis-
tencies -may have led to.either' the premature rejection of the Rotating
Biological Contactor (R.RC) process fpom, consideration at the screening level
or the weighting .of'the" cost-effectiveness analysis so as to place the RBC
1413
-------
process at a competitive disadvantage Against other more "conventional"
wastewa,ter. treatment processes..
The purpos.e of. this.. study is. t
-------
In order for a cost-effectiveness analysis to be truly useful, it must
include all potential process chains which are capable of meeting the
required effluent standards and likely to be cost-competitive. Once the
competing process chains have been selected, the analyst must collect and/or
develop sufficient information to enable him to estimate, as accurately
as possible, the complete economic impacts which would result from the
selection of any one of the competing alternatives. All process-dependent
facilities should be included and all non process-dependent facilities such
as laboratory buildings, maintenance garages, administrative buildings, etc.
should, for the purposes of analysis, be excluded, unless the costs for
these facilities could change as a result of the process which is selected.
The degree to which the various process details are analyzed generally will
not seriously affect the ultimate outcome of the analysis, provided that at
least a minimum level of detail is used and, further, that all competing
alternatives are developed to the same level of detail. One should avoid
the practice of developing greatly detailed estimates for certain processes,
including careful enumeration of equipment, facilities, operation and main-
tenance requirements and construction costs (the "quasi-design" approach)
while, at the same time, developing cost estimates for the remaining
competing processes based on published cost curves (the "pseudo-design"
approach). The effect of intermingling approaches at this juncture in the
facilities planning process could invalidate the results generated by the
analysis and, as a consequence, render possibly questionable ultimate
conclusions. Following completion of the cost-effectiveness analysis, sub-
jective considerations may be addressed again, if necessary, as in the case
where the analysis has yielded more than one alternative with the lowest
present worth or very close values, to within the accuracy of the estimates.
BASIS FOR EXAMPLES
To demonstrate the potential range of design flow capacities over which
RBC processes may be cost-competitive, sample cost-effectiveness' analyses have
been prepared for the upgrading of hypothetically existing primary treatment
plants to meet secondary effluent limitations of 30 mg/1 BODs and 30 mg/1
suspended solids for design flow capacities of 3, 5, 10, 30 and 50 MGD. The
influent to the secondary treatment facilities from the existing primary
portion of the plant has been assumed to have the following characteristics:
BOD5 (Total):
BOD5 (Soluble):
Suspended Solids:
140 mg/1
85 mg/1
90 mg/1
The foregoing reflects primary treatment removals of approximately 30%
for BOD5 (Total) and 55% for suspended solids. It has also been assumed
that there is sufficient land available for construction of any of the poten-
tial processes; that the land is of suitable contour to permit flow through
the secondary facilities by gravity; that existing chlorine contact tanks
may be. incorporated into the process without reconstruction; and that all
excavation may be performed by machines, with no hard or difficult rock or
subsurface conditions to interfere with construction. From the standpoint
of unit redundancy, in general, two final clarifiers have been provided for
1415
-------
the 3 MGD case, four final clarifiers have been provided for the 5, 10 and
30 MGD cases; and six final clarifiers have been provided for the 50 MGD
case. Operating costs have been estimated on the basis of $0.045 per
kilowatt-hour (KWH) of electrical power consumed; labor costs have been
estimated on the basis of $10.00 per man-hour; and polymer costs have been
estimated on the basis of $1.75 per pound of dry polymer consumed. All
present worth calculations have been made on the basis of a discount rate of
7-1/8% for a period of 20 years. The following sections will briefly sum-
marize the basic design criteria used for the development of cost estimates
for the various generic process analyzed.
RBC PROCESSES
RBC processes fall into the general category of attached film processes
and utilize corrugated plastic media stacks mounted on shaft assemblies.
RBC equipment units are available in shafts of various standard lengths and
in either standard density or high density designs. Shaft rotation may be
produced either by a mechanical drive assembly or an air drive assembly.
With mechanical drive RBC's, each shaft has its own motor, gear box, drive
chain, etc. and with air drive KBC's, each shaft is rotated by the buoyancy
of air which is captured in air cups attached to the perimeter of the media.
Air is supplied from a centralized blower system through a pipe and diffuser
network. For ease of comparison the estimates presented herein for air
drive and mechanical drive RBC processes have been based on equipment of
the same manufacturer and the design parameters have been developed on the
basis of the manufacturer's published design procedures.1 Four variations
are possible for the RBC processes and are included herein. They are as
follows:
1. Air Drive RBC using all high density media shafts.
2. Air Drive RBC using an appropriate combination of standard and
high density media shafts.
3. Mechanical Drive RBC using an appropriate combination of standard
and high density media shafts.
4. Mechanical Drive RBC using all standard density media shafts.
For all RBC processes, shaft lengths of 25 feet have been assumed and
an allowance for fiberglass covers has been included. The basic design
loadings have been selected at levels appropriate to meet the effluent
limitation of 30 mg/1 BOD5 at a wastewater temperature of 47°F. Final
clarifiers have been designed on the basis of an average surface overflow
rate of 800 gallons/day/square foot. The basic design parameters used in
this study are as follows:
Organic
Loading:
Hydraulic
Loading:
Air Drive
Mechanical Drive
Air Drive
Mechanical Drive
1.7 #BOD5 (soluble)/I,000 ft2 - Day
1.6 #BOD5 (soluble)/I,000 ft2 - Day
2.4 gallons/ft2 - Day
2.25' gallons/ft2 - Day
1416
-------
No provision has been made for separate thickening facilities for the
RBC processes considered herein. Allowances have been made for providing
the necessary pumping equipment, piping, etc. for returning the waste sludge
to the head end of the primary clarifiers for cosettling with the primary
sludge. It is anticipated that this approach would yield a combined sludge
stream having a concentration in the range of 4 to 6% solids by weight. This
would normally be suitable for most solids handling schemes and, further,
would be comparable to the blended sludge concentrations which would be
expected to result from thickening of the waste activated sludges separately
and then blending them with the primary sludge.
The estimated total construction costs, miscellaneous costs, annual
operation and maintenance costs and total present worth for each of the RBC
processes at each of the five design flows are presented in Table I.
Included in the construction costs are estimates for the RBC tankage and
equipment; final clarifiers; waste sludge pumping facilities; exterior and
underground piping; blower and control building (as required); and miscel-
laneous modifications and/or re-piping necessary to enable cosettling of
RBC waste sludge with primary sludge in the primary clarifiers. As can be
seen from Table I, the air drive RBC processes exhibit a slightly lower
present worth than mechanical drive RBC processes with the differential
present worth between the various designs becoming larger at higher design
flows. The present worth of air drive and mechanical drive RBC processes
using combinations of standard and high density media shafts are generally
within 10% of each other throughout the range of design flows studied.
AIR ACTIVATED SLUDGE PROCESSES
Air activated sludge processes are among the oldest of the processes
included in this analysis and represent the most commonly used process for
domestic wastewater treatment in the past. There are many variations
among air activated sludge processes which differ in their removal
efficiencies; required detention times; methods of feed introduction;
food to microorganism ratios; recycle ratios; organic loading; solids
residence times and many other parameters. As discussed earlier, the
purpose of this study is not to present the absolute optimal design for
any particular process but, rather, to present a typical facilities design
for each process in order to establish a basis for comparison. Other designs
could be developed which might yield slightly different configurations and/or
construction/operating costs; however, such a detailed presentation is
beyond the scope of this study. The actual design conditions and effluent
limitations will usually dictate the appropriate choice of design criteria
and should be carefully evaluated on a case by case basis.
Air activated sludge processes utilize a suspended biomass, rather
than an attached film as used in RBC processes, to effect wastewater treat-
ment. The principal source of the oxygen necessary for biological activity
is from atmospheric air. Since air contains only about 21% oxygen, it is
necessary to compress large volumes of gas consisting mostly of nitrogen
and other inert materials in order to deliver the required quantity of oxygen
to the process. Further, due to the reduced saturation value of oxygen dis-
solved in wastewater from air, these processes are inherently less efficient
1417
-------
TABLE I
COST-EFFECTIVENESS ANALYSIS
ROTATING BIOLOGICAL CONTACTOR PROCESSES
(ESTIMATED YEAR 1980 COST IN $1,000'S)
AIR DRIVE RBC
MECHANICAL DRIVE RBC
DESIGN FLOW = 3 MGD
Construction Cost
Contingencies, Misc. (25%)
TOTAL COST
Annual Operation & Maint.
TOTAL PRESENT WORTH
DESIGN FLOW - 5 MGD
Construction Cost
Contingencies, Misc. (25%)
TOTAL COST
Annual Operation & Maint.
TOTAL PRESENT WORTH
DESIGN FLOW » 10 MGD
Construction Cost
Contingencies, Misc. (25%)
TOTAL COST
Annual Operation & Maint.
TOTAL PRESENT WORTH
DESIGN FLOW - 30 MGD
Construction Cost
Contingencies, Misc. (25%)
TOTAL COST
Annual Operation & Maint.
TOTAL PRESENT WORTH
DESIGN FLOW « 50 MGD
Construction Cost
Contingencies, Misc. (25%)
TOTAL COST
Annual Operation & Maint.
TOTAL PRESENT WORTH
All High
Density
Media
$ 1,460
365
$ 1,825
$ 72.6
$ 2,585
$ 2,210
550
$ 2,760
$ 108.8
$ 3,900
$ 3,490
870
$ 4,360
$ 162.2
$ 6,060
$ 8,330
2,080
$10,410
$ 333.1
$13,905
$13,370
3,340
$16,710
$ 513.0
$22,090
Combined
Media
$ 1,540
385
$ 1,925
$ 72.6
$ 2,685
$ 2,295
575
$ 2,870
$ 110.4
$ 4,030
$ 3,800
950
$ 4,750
$ 166.7
$ 6,500
$ 8,920
2,230
$11,150
$ 334.4
$14,660
$14,525
3,630
$18,155
$ 522.0
$2.3,630
Combined
Media
$ 1,430
360
$ 1,790
$ 93.6
$ 2,770
$ 2,410
600
$ 3,010
$ 138.2
$ 4,460
$ 3,800
950
$ 4,750
$ 190.9
$ 6,755
$ 9,710
2,430
$12,140
$ 390.3
$16,235
$15,780
3,945
$19,725
$ 611.6
$26,140
All Standard
Density Media
$ 1,485
370
$ 1,855
$ 95.4
$ 2,855
$ 2,410
600
$ 3,010
$ 138.2
$ 4,460.
$ 3,9.65
990
$ 4,955
$ 195.4
$ 7,005
$10,165
2,540
$12,705
$ 403.8
$16,940
$16,445
4,110
. $2Q,555
$ 634.1
$27,210
1418
-------
in terms of oxygen transfer than are processes utilizing relatively pure
oxygen, as discussed later. A variety of dissolution equipment is available
for use with air activated sludge systems. Three variations are included
in this study which represent a reasonable cross-section of the available
equipment classifications. They are as follows:
1. Fine Bubble Diffused Aeration - This alternative uses porous
ceramic plates located on an air piping distribution grid at the
bottom of the tank which introduce a uniform system of fine
bubbles throughout the biological contact tanks. The assumed
oxygen transfer efficiency for this equipment used in this study
is 26.5%.
2. Coarse Bubble Diffused Aeration - This alternative uses open-type
diffusers located along a pipe header at the bottom of the tank
which introduce coarse bubbles throughout the biological contact
tanks. The assumed oxygen transfer efficiency for this equipment
used in this study is 9.5%.
3. Draft-Tube Mechanical Aeration - This alternative uses a combined
system of blowers, air spargers and mechanical mixers to effect
oxygen transfer. The mechanical mixer consists of a draft tube
with an internal impeller which forces wastewater downward from
the surface of the tank and accelerates it past an air sparge
ring where relatively coarse bubbles are broken up and entrained
in the downward flow. The assumed oxygen transfer efficiency for
this equipment used in this study is 50%.
For air activated sludge processes, the final clarifiers have been
designed on the basis of an average surface overflow rate of 600 gallons/
day/square foot for the 3, 5 and 10 MGD cases and 800 gallons/day/square
foot for the 30 and 50 MGD cases. This is due to the fact that the antici-
pated peak:average flow ratios are, for the lower design flow cases, such
that the performance of the final clarifier in terms of suspended solids
removals may be adversely affected to the extent that the higher surface
overflow rate may be inadequate for clarification during these peak flow
perit>ds. The remaining basic design parameters used in this analysis for
air activated sludge processes are as follows:^
Food: Microorganism Ratio
(#BOD5/#MLVSS-Day):
Organic Loading
(#BOD5/1,000 ft3-Day):
Hydraulic Detention Time
(Hours):
MLSS (mg/1):
MLVSS (mg/1):
Average Recycle Ratio:
Diffused
Aeration
0.27
42
5.0
3100
2480
60%
Mechanical
Aeration
0.35
42
5.0
2400
1920
60%
1419
-------
Thickening of waste activated sludge is included in the cost estimates
for air activated sludge processes. Dissolved air flotation units have
been provided and sized on the basis of 2.25 ///hour/square foot of flotation
area. One complete standby unit has been incorporated into the design to
provide sufficient reliability; for the accommodation of peak loadings; and
to enable the plant personnel to reduce the operating time required for
thickening by using all units as may be desired. Polymer conditioning
equipment has been included and the estimated consumption of polymer has
been assumed to average 4 pounds of dry polymer per ton of dry solids
processed.
The estimated total construction costs, miscellaneous costs, annual
operation and maintenance costs and total present worth for each of the air
activated sludge processes at each of the five design flows are presented
in Table II. Included in the construction costs are estimates for the
aeration tankage and equipment; final clarifiers; return and waste sludge
pumping facilities; exterior and underground piping; blower and control
building; and dissolved air flotation facilities. As can be seen from
Table II, the fine bubble diffused aeration and draft-tube mechanical
aeration designs exhibit similar values for present worth, while the coarse
bubble diffused aeration facilities have a slightly higher present worth
value, ranging from 3 to 10% higher with the differential increasing at
higher design flows. This is due to the significantly lower oxygen transfer
efficiency inherent in the coarse bubble diffused air systems which requires
more power for the process blowers.
PURE OXYGEN ACTIVATED SLUDGE PROCESSES
In pure oxygen activated sludge processes, as with air activated sludge
processes, the treatment of wastewater is effected by bringing it into inti-
mate contact with a suspended mass of biological microorganisms within a
tank designed to provide sufficient detention time to assure that the desired
degree of treatment will be attained. These treatment processes utilize
pure oxygen rather than air as the source of oxygen for use in biological
metabolism and may have certain cost advantages over air activated sludge
processes, particularly in large applications. Due to the fact that
essentially pure oxygen (90-95%) is used, the natural driving force availa-
ble for dissolution of oxygen into the biological suspension is much greater
than would be the case using air. Using pure oxygen, the saturation value
of oxygen dissolved in wastewater is approximately 4.4. times greater than
with air. This further enables the active biomass to be maintained at
higher levels than is feasible with conventional air systems; permits higher
dissolved oxygen levels to be maintained in the mixed liquor; and permits
the reduction in the volume of the biological contact tanks, thereby reduc-
ing the construction costs of these process units. From an operating cost
standpoint, this results in a reduction of the amount of electrical
energy required for the dissolution system in comparison with conventional
air systems.
The principal drawback to pure oxygen activated sludge processes is due
to the fact that the air supply used in conventional air systems is free and
the high purity oxygen which is required in oxygen systems, unfortunately,
1420
-------
TABLE II
COST-EFFECTIVENESS ANALYSIS
AIR ACTIVATED SLUDGE PROCESSES
(ESTIMATED YEAR 1980 COST IN $1,000'S)
DESIGN FLOW =3 MGD
Construction Cost
Contingencies, Engr., Misc. (25%)
TOTAL COST
Annual Operation & Maintenance
TOTAL PRESENT WORTH
DESIGN FLOW =5 MGD
Construction Cost
Contingencies, Engr., Misc. (25%)
TOTAL COST
Annual Operation & Maintenance
TOTAL PRESENT WORTH
DESIGN FLOW = 10 MGD
Construction Cost
Contingencies, Engr., Misc. (25%)
TOTAL COST
Annual Operation & Maintenance
TOTAL PRESENT WORTH
DESIGN FLOW = 30 MGD
Construction Cost
Contingencies, Engr., Misc. (25%)
TOTAL COST
Annual Operation & Maintenance
TOTAL PRESENT WORTH
DESIGN FLOW = 50 MGD
Construction Cost
Contingencies, Engr., Misc. (25%)
TOTAL COST
Annual Operation & Maintenance
TOTAL PRESENT WORTH
Fine
Bubble
Diffused
Aeration
Coarse
Bubble
Diffused
Aeration
Draft-Tube
Mechanical
Aeration
$ 2,085
520
$ 2,605
$ 141.4
$ 4,090
$ 8,530
2,130
$10,660
$ 601.6
$16,970
$13,040
3,260
$16,300
$ 912.0
$25,870
$ 2,050
510
$ 2,56Q
$ 150.0
$ 4,135
$ 8,405
2,100
$10,505
$ 747.0
$18,340
$12,850
3,210
$16,0.60.
$ 1,160.0
$28,230
$ 1,9.95
5QQ
$ 2,495
$ 138.0
$ 3,9.45
$ 2,975
745
$ 3,720
$ 190.6
$ 5,720
$ 2,930
730
$ 3,660
$ 212.6
$ 5,890
$ 2,870
720
$ 3,590
$ 190.1
$ 5,585
$ 4,770
1,190
$ 5,960
$ 299.8
$ 9,105
$ 4,705
1,175
$ 5,880
$ 345.1
$ 9,500
$ 4,640
1,160
$ 5,800
$ 300.1
$ 8,950
$ 8,380.
2,095
$10,475
$ 608.6
$16,860
$12,860
3,215
$16., 075
$ 9.24.6
$25,775
1421
-------
is not. Generally, the only available means for obtaining oxygen is by
providing on-site generation facilities. These facilities require a sub-
stantial initial capital investment and use large amounts of electrical
energy for the production of oxygen. Therefore, a portion of the savings
in electrical energy for dissolution (in some cases all of the savings)
is eliminated through the addition of the oxygen generation facilities.
In large plants, however, an operating savings in electrical energy consump-
tion will generally result using pure oxygen systems. This savings often
offsets the additional capital costs over the planning period enough to make
pure oxygen activated sludge processes attractive.
Two principal variations are available with pure oxygen activated
sludge processes. One uses covered tanks with an enriched atmosphere in
the space above the liquid and one uses open tank construction. The
covered tank alternative is staged into a series of three compartments and
transfers 90% of the oxygen supplied to the tank through the three stages.
The open tank alternative provides 90% oxygen transfer within the liquid
depth by introducing the oxygen in extremely fine bubbles near the bottom
of the tank. Further, two types of oxygen generation units are available:
one a cryogenic generator, the other a pressure swing adsorption (PSA)
generator. In general, a cryogenic generation system costs more initially
but requires less electrical energy than does a PSA generation system.
Selection between these two types of systems within the range of design
flows included is not always obvious and, therefore, both will be presented
for illustration purposes. The four variations included in this study are
as follows:
1. Open Tank construction with a cryogenic oxygen generation system.
2. Covered Tank construction with a cryogenic oxygen generation
system.
•3. Covered Tank construction with a PSA oxygen generation system.
4. Open Tank construction with a PSA oxygen generation system.
For pure oxygen activated sludge processes, the final clarifiers
have been designed on the basis of an average overflow rate of 600 gallons/
day/square foot for all cases included in this study. Although with air
activated sludge processes it was deemed appropriate to increase the design
overflow rate at the higher design flows, it is felt that solids flux con-
siderations could be controlling in the pure oxygen activated sludge pro-
cesses and that the more conservative design is indicated within the range
included in this study. The remaining basic design parameters used in this
analysis for pure oxygen activated sludge processes are as follows:'1''5
Food:Microorganism Ratio
(#BOD5/#MLVSS-Day):
Organic Loading
(#BOD5/1,000 ft3-Day):
Hydraulic Detention Time
(Hours):
Open
Tanks
0.70
155
1.35
Covered
Tanks
0.75
175
1.2
1422
-------
r
MLSS (mg/l):
MLVSS (mg/l):
Average Recycle Ratio:
Open
'Tanks
4,375
3,500
40%
Covered
' "Tanks
4,575
3,660
40%
Thickening of waste activated sludge is included in the cost estimates
for pure oxygen activated sludge processes. Dissolved air flotation units
have been provided and sized on the basis of 3.0 ///hour/square foot of flota-
tion area. As with air activated sludge processes, one complete standby
unit has been incorporated into the design. Polymer conditioning has been
included and the estimated consumption of polymer has been assumed to average
4 pounds of dry polymer per ton of dry solids processed. The average under-
flow concentration from the clarifiers has been assumed to be 1.75% solids
by weight. In the larger cases studied, centrifugation also was considered
for thickening. In the 30 MGD case, a supplementary cost-effectiveness
analysis indicated that the dissolved air flotation and centrifuge alter-
natives were comparable in terms of present worth and in the 50 MGD case,
the centrifuge exhibited a slight present worth advantage; however, since
centrifuge data was based upon certain assumed polymer dosages and loading
rates which would only be verifiable through pilot testing, it was
believed more realistic to presume the use of dissolved air flotation for
thickening in all cases studied.
The estimated total construction costs, miscellaneous costs, operation
and maintenance costs and total present worth for each of the pure oxygen
activated sludge processes at each of the five design flows are presented
in Table III. Included in the construction costs are estimates for the
oxygenation tankage and equipment; final clarifiers; return and waste sludge
pumping facilities; exterior and underground piping; oxygen control building;
oxygen generation and storage facilities; and dissolved air flotation
facilities. As can be seen from Table III, generally all of the pure oxygen
activated sludge facilities exhibit almost the same present worth for any
given design flow, within very close limits. There is a slight consistent
advantage favoring the PSA oxygen generator alternatives, although it is not
significant. In general, the selection of oxygen generator type would be
more reasonably made on the basis of system needs, such as the turndown
requirements. Cryogenic generators are usually limited in turndown to about
60 to 70% of rated capacity, while PSA generators may be more or less turned
down almost to zero output.
COMPARATIVE ANALYSES
In conducting cost-effectiveness analyses, the annual operation and
maintenance costs play an important role in determing the selected alterna-
tive. These costs are converted into equivalent present worth by multiply-
ing by an appropriate factor (10.49187 in this study). The single most
significant annual operating cost, particularly for the larger design flows,
is the cost of power. Table IV presents the total annual power requirements
for each of the processes and design flows considered in this study. In
general, the RBC processes have a significant advantage in terms of annual
1423
-------
TABLE III
COST-EFFECTIVENESS ANALYSIS
OXYGEN ACTIVATED SLUDGE PROCESSES
(ESTIMATED YEAR 1980 COST IN $1,000'S)
CRYOGENIC
Open
Tanks
DESIGN FLOW ^ 3 MGD
Construction Cost $ 2,810
Contingencies, Misc. (25%) . 700
TOTAL COST $ 3,510
Annual Operation & Maintenance $ 151 . 9
TOTAL PRESENT WORTH $ 5,105
DESIGN FLOW = 5 MGD
Construction Cost $ 3,695
Contingencies, Misc. (25%) 925
TOTAL COST $ 4,620
Annual Operation & Maintenance $ 204.7
TOTAL PRESENT WORTH $ 6,770
DESIGN FLOW - 10 MGD
Construction Cost $ 5,385
Contingencies, Misc. (25%) 1,345
TOTAL COST $ 6,730
Annual Operation & Maintenance $ 317.6
TOTAL PRESENT WORTH $10,060
DESIGN FLOW - 30 MGD
Construction Cost $ 9,475
Contingencies, Misc. (25%) 2,370
TOTAL COST $11,845
Annual Operation & Maintenance $ 614.8
TOTAL PRESENT WORTH $18,295
DESIGN FLOW » 50 MGD
Construction Cost $14,160
Contingencies, Misc. (25%) 3,540
TOTAL COST $17,700
Annual Operation & Maintenance $ 916.7
TOTAL PRESENT WORTH $27,320
GENERATOR
Covered
Tanks
$ 2,705
675
$ 3,380
$ " 149.9
$ 4,955
$ 3,495
875
$ 4,370
$ 201.3
$ 6,480
$ 5,330
1,330
$ 6,660
$ 311.6
$ 9,930
$ 9,600
2,400
$12,000
$ 598.6
$18,280
$13,970
3,490
$17,460
$ 890.8
$26,805
PSA' GENERATOR
Covered
Tanks
$ 2,575
645
$ 3,220
$ 152.4
$ 4,820
$ 3,355
840
$ 4,195
$ 205.6
$ 6,350
$ 5,160
1.290
$ 6,450
$ 319.2
Open
Tanks
$ 2,595
650
$ 3,245
$ 152.8
$ 4,850
$ 3,440
860
$ 4,30.0
$ 206.2
$ 6,465
$ 5,090
1.270
$ 6,360
$ 32Q.8
$ 9,800. $ 9_,725
$ 9.,340
2,335
$11,675
$ 619.1
$18,170
$13,610
3.400
$17,010
$ 923.4
$26,700
$ a,Q35
2,260.
$11,295
$ 623.1
$17,.830
$13,570.
. 3,39.0
$16,960
$ 929.8
$26,715
1424
-------
Q
O C5
sr
st-
in m m
sr sr sr
vo o m
en en
o
vo
oo
U-1
CM
ov
m
CPl
en- oo
o' m o o
en in oo CM
vo o r«. eri
o o
m g
m
in
o
m
oo
o
o
CM
CM CM CM CM
in
in
o
CM
en
in
SI-
CM
oo in
in
in
m
o
CM
o
in
oo
o
in
oo
m
ov
o
en
oo
in
vo
oo
10 1
o in
CM in
si- sr
o o
o o
m m
in m in
in in in
oo si- o\
in o m o
si- r-. \o oo
o\ oo ON o\
en
CO
CO
w
cj
8
PH
n m
E2 •*
a CM
cd
•H
rcJ
0)
J£{
JJ
•H
CO
CO fj
PS CU
0 O
O ,£5
rH
rH
HH "^1
•53
O 1
M
O 0)
0 >
i-l -H
O S-i
M Q
m
M
O -H
£55 <^
I — !
H
H rH
Q
p£!
in
«^*
CM
rd
•H
Tj
CU
g
13
CU
a
•H
«")
g
0
O
1
0)
•H
M
Q
M
•H
C
M 41
H Pd
q
<<
Pi rH
M
o
CJ
en
o
oo
m
^-^
a
S
60
X
O
"^
CO
PM
CO
ed
H
a
cu
ft
O
sr
1425
-------
power consumption. Throughout the'range .of design flows, analyzed herein,
the RBC processes generally require, from 4Q. to 6.5% .pf the,povrer required.
for pure oxygen activated sludge processes and from'23'to 60% of the power
required for air activated sludge processes. Further, air drive RBC pro-
cesses generally consume somewhat less power than do mechanical drive
RBC processes.
Table V presents the total present worth for each of the processes and
design flows considered in this study. For the 3, 5 and 10 MGD cases, all
RBC processes exhibit a significant cost advantage over the air and pure
oxygen activated sludge processes. At 30 and 50 MGD, the air drive RBC
processes have a significant cost advantage over mechanical drive RBC
processes and the air and pure oxygen activated sludge processes, all of
which are closely grouped, with the exception of the coarse bubble diffused
aeration version of the air activated sludge process.
Following the preparation of the cost-effectiveness analysis, it may be
necessary or desirable to examine the subjective considerations for those
processes which are grouped within 5 or 10% of the present worth of the
alternative having the lowest present worth value prior to selecting the
final plan. Based on the information presented herein, for the 3, 5
and 10 MGD cases, this would mean the comparison of air drive and mechanical
drive RBC processes and for the 30 and 50 MGD cases, this would mean the
comparison of essentially all of the alternatives with the exception, perhaps,
of the coarse bubble diffused aeration version of the air activated sludge
process. In terms of operational simplicity; power requirements; waste
sludge thickening; and labor costs; all of the RBC processes have, in the
author's opinion, a'definite subjective advantage. The various activated
sludge alternatives may have somewhat of a subjective advantage in terms
of flexibility; reliability; and shock load assimilation capacity; however,
not a significant enough advantage to warrant the skewing of the analyses.
Land area requirements are comparable for RBC and air activated sludge
processes; however, pure oxygen activated sludge processes have the advantage
there. In terms of comparing the air drive and mechanical drive RBC processes,
the air drive RBC processes have, in the author's opinion, the advantage in
terms of maintenance, power requirements and turndown capability. The
foregoing illustrates some of the various subjective considerations which
may be considered following the completion of the cost-effectiveness analysis
step. None of these considerations are sufficiently important to reverse
the results of the analysis and, therefore, it is recommended that they be
addressed briefly, for completeness, and then the final selection should be
made based on the results of the cost-effectiveness analysis.
DISCUSSION AND CONCLUSIONS
Recognizing that the estimates contained herein, although rigorously :
developed, are budgetary in nature and not site specific for any particular
set of circumstances, the authors acknowledge that variations are to be
expected from the costs presented in this study. There are many assumptions
built into these cost estimates and they may not necessarily suit the pre-
ferences of other analysts or be appropriate for a given situation. The
purpose of this study has been to investigate, over a significant range of
1426
-------
w
1-4
Q
o o
in g
Q
O O
en g
o
0 0
r-l g
Q
m o
g
Q
co ca
CO
CO
w
o
o
pq
PH
0
OS
O
CM
CM
•co-
in
o
o\
ft
en
rH
•co-
co
vO
o
n
vO
•co-
o
o
CTv
CO
<0-
m
00
m
CM
•co-
ed
•H
Tj
cu
g
^
4-1
•H
CO
co d
P3 0)
0 Q
EH
CJ jj
<< oO
H -iH
J3 53
O
CJ rH
r-J
<£j
U 1
M
O CU
O >
rJ -H
O H
M Q
PQ
M
" 3
125 <;
HH
H
H rH
5
o
en
vo
n
en
CM
•co-
co
vo
VO
•I
-d-
rH
•co-
o
o
in
r,
vo
•CO-
CO
en
O
*^-
•co-
in
OO
vo
CM
•co-
ed
•H
13
cu
g
13
CU
d
•H
"i
o
o
1
cu
>
•H
M
«
^J
3
CM
O
-d-
rH
*
vO
CM
•CO-
CO
CM
«*
vO
rH
•co-
in
in
P^.
^
vO
•CO-
CO
vo
n
^J-
•co-
0
p^
rt
CM
•co-
ed
•H
TJ
CU
g
13
CB
d
•H
"s
o
1
0)
M
Q
T— 1
rd
O
•H
ed
0
o>
g
en
O
rH
CM
*
p^.
CM
•CO-
CO
OY
f.
vo
rH
•co-
in
o
o
M
r~^
•co-
co
vo •
•H W
to C5
O - O
o
Cd CO
CJ
•H Q
d w
cd EH
rC!
CU M
CJ
*d" P^
3
o
f^-»
oo
wt
m
CM
•co-
co
o\
vo
rH
•co-
in
o
•«
O"\
•co-
co
CM
*
in
<0-
o
CT>
0
*»
CO
rH
O
0
m
A
o>
•co-
0
CT>
OO
in
•co-
in
en
^d*
•CO-
pj
O
•H
4J
ed
CU
^1
13
CU
9
m
•H
Q
cu
rH
fl
3
PP
cu
CO
ed
O
0
CM
in
r^
f^
•%
in
CM
o
vo
oo
^
vO
rH
•CO-
CO
m
CTi
*>
CO
•co-
in
oo
m
m
•co-
in
cy\
en
•co-
rt
O
•H
ed
M
CU
rH
ed
O
iH
cd
cu
S
0)
43
g_(
1
4-1
ed
M
Q
en
w
tD
S
CO
n
w
EH
M
EH
U
^3
w
0
B
o
2
PH
O
CM
cn
•V
p^
CM
•co-
in
OV
CM
00
rH
•CO-
o
VO
o
•X
o
rH
•co-
o
1^.
f^
n
vO
•co-
in
0
rH
m
•CO-
^—^
Cj
CU
60
^"»
X
O
o
•H
0)
60
O
!>*
CJ
'*~s
m
A!
rj
ed
H
d
CU
co"
rH
in
o
oo
r>
VO
CM
•CO-
CO
OO
CM
OO
rH
•CO-
CO
cn
CTl
A
OS
•CO-
CO
oo
ft
vo
•co-
in
in
CPi
<3*
•co-
/*^
d
cu
60
^
^
o
o
•H
d
cu
60
o
^v
M
CJ
v-^
CO
r«
d
ed
£H
X)
cu
M
cu
t>
0
CJ
CM
o m
O rH
«t ^
VO vO
CM CM
•CO- -CO-
CO O
r^ en
rH CO
oo r^
rH rH
•CO- -CO-
o m
O CM
oo r-
cr> CTV
•co- -co-
O m
in oo
en -
^^
d
cu
60 ^
X cu
O 60
s
O Qj
o o
* •
c^O -3-
1427
-------
design flows, the relative cost-effectiveness '.of RBG processes j,n. Delation
to activated sludge processes.. for a typi.ca.1 carbon oxidation application. The
results of this study indicate that RB-C processes are,' indeed, cost-competi-
tive with activated sludge processes' throughout the range of design flows
investigated. Therefore, the RBG processes should he given careful consi-
deration in any similar application.
Further investigation is warranted in several areas with regard to
RBC processes and cost-effectiveness in general which either surfaced during
the preparation of this study or were observed by the authors during its
preparation. First of all, the range of design flows could be extended
upwards, perhaps to 100 MGD, since no distinct breakpoint was observed
beyond which RBC processes were no longer preferred on the basis of their
present worth values. Secondly, the use of inclined plate—settler types
of clarifiers may be warranted and could greatly reduce the costs for
RBC processes. Use of the plate settlers would result in the additional
benefit of reducing land area requirements for the RBC processes to levels
comparable with pure oxygen activated sludge processes. The need for waste
sludge thickening may require additional study to determine whether or not
it is required as an additional cost item for RBC processes or as a deduc-
tion from activated sludge processes. Further, as more data is generated
from centrifuge manufacturers and/or operating installations, the use of
dissolved air flotation exclusively for thickening as presented in this
study may require reexamination.
In conclusion, this study indicates that RBC processes are viable
alternatives for carbon oxidation; are cost-competitive and cost-effective
throughout the range from 3 to 50 MGD and should be carried through the
detailed analysis performed as a part of future facilities planning work
for proposed BODs removal projects.
References
1. AUTOTROL Wastewater Treatment Systems - Design Manual, AUTOTROL
Corporation - Bio Systems Division, Milwaukee, Wisconsin (1979).
2. Wastewater Treatment Plant Design - A Manual of Practice, Water
Pollution Control Federation, Washington, D.C. and the American
Society of Civil Engineers, New York, NY, Lancaster Press, Inc.
(1977).
3. Komline, T.R., "Sludge Thickening By Dissolved Air Flotation in
the USA", Presented at 'Flotation For Water and Waste Treatment',
a Water Research Centre Conference, 8-10 June 1976, Felixstowe,
Suffolk, England.
4. Stetzer, Richard S., "FMC Pure Oxygen - An Effective Solution to
Wastewater Treatment and Process Applications", Presented at the
70th Annual AIChE Meeting, November 13-17, 1977,-New York, NY.
5. "Oxygen Activated Sludge Wastewater Treatment Systems - Design
Criteria and Operating Experience", Environmental Protection
Agency - Technology Transfer, EPA 625/4-73-003a, August, 1973.
1428
-------
WORKSHOP ON RBC RESEARCH NEEDS *
Chairman: Dr. J. A. Borchardt
Assistants: Dr. Y.C. Wu, Dr. Ed. D. Smith, Dr. R. D. Miller,
Mr. E. J. Opatken, Dr. W. A. Sack, Dr. C.P.L. Grady, Jr.,
Dr. S. K. Banerji, Dr. D. F. Kincannon, Dr. A. A. Friedman,
Dr. W. G. Characklis, Dr. F. M. Saunders, Dr. J. C. Haung,
Dr. P.C. Poon, Mr. R. L. Antonie, Mr. G. R. Fisette,
Mr. G. E. Flann, Mr. J. T.- Madden, Mr. B. Joost, Mr. M0 Creston,
Dr. C. G. Steiner, Prof0 W; W. Eckenfelder, Mr. J. F. Lagnese,
Dr. E. J. LaMotta
DR. BORCHARDT: We are very pleased and grateful to see so many of you
come and I know there's a lot of competing interest for your time. We
hope that this will be a productive meeting. I thought perhaps that we
would organize this in such a fashion that we would address five separate
topics and for that purpose we have picked five subchairmen to -each
discuss one of these topics for a few minutes. And then we are going to,
for that particular subject, call for discussions, questions, statements
from the floor for another 10 to 15 minutes<> That allows about 20 to
25 minutes for each subject, hopefully that five of them you see we
should finish in an hour and a half or something of that sort. So that is
the plan for the evening. Now the subjects and subchairmen that we are
going to use, we have three of them here on the podium and I'm expecting
the other two any minute. The first subject is scale factors from pilot
to prototype and Professor Grady from Purdue University is going to be
the subchairman in charge of that particular subject. Professor Grady
is here on my left on the end at the present. The next subject we are
going to talk about; kinetics, internal and external diffusion problems.
Professor LaMotta is going to take that subject. Then the third group
will be RPM turbulence surface effects—that sort of thing, and Professor
Friedman has agreed to chair that; he's not here yet as I see it, and
then operating problems. Mr. Legnese of Duncan-Legnese Associates is
going to take the operating problems. And then we have research support
and I know you are all interested in research support and Ed Opatken has
agreed that he will discuss this and answer all questions. So you see,
you have a full evening ahead of you and I hope that you are not think-
ing about other problems, that you're concentrating on this research.
If you were here yesterday you heard about Murphy's Law and O'Toole's
corollary and a few other things. I suppose you all realize that Mrs.
Murphy has a law as well. Mrs. Murphy's law says in effect that if any-
thing can go wrong it will go wrong and when Dad is out of town, so you
can expect the plumbing is broken or the fuses are blown or something at
home, but just relax until the meeting is over and then worry about
those problems. Now we still don't have our other two subchairmen but
I'm not going to worry about that for I know we have enough experts in
the group here to handle them. So our first subchairman tonight will
be Dr. Grady from Purdue University. He is going to talk about scale-up
factors from model to prototype.
*The discussions from this workshop were typed from a tape of the workshop.
Consequently, problems with content and clarity may be apparent in certain
sections.
. 1429
-------
DR. GRADY: I am going to talk very briefly hopefully because I feel
on scale factors there is a very strong interaction. I am talking
about scale factors. I am talking about each of the second and third
topics today, because really scale factors or scale-up is basically a
problem of identifying the fundamentals within the system so that we
can then know how those fundamentals are affected by system signs.
You really can't talk about scale up without talking about turbulence ,
diffusion and all of these other problems. So I guess what I would
like to do for a moment is to perhaps reiterate some of the things
that I said this afternoon, some of you I know were not here. The
course of that discussion was that by using relatively fundamental
mathematical models of the type that Jim Mueller uses and other people
are working on within this field, you can take into account scale
factors. Obviously all of you have had fluid mechanics background
somewhere along your way and you know that scale factors are handled
in a number of ways in a fluid system. There is really no reason why
we cannot handle scale factors in the same way within our system even
though they are going to be more complex because they involve more
factors. But since my point of view and I guess I am trying to give
you something to start, to get the discussion going, because certainly
I do not have any answers on the thing. But from my point of view the
best way to approach scale—up is first to develop fundamental mathe-
matical models and then to use those fundamental mathematical models
to study the effect of scale for a large number of different situations.
Only through doing that can one begin to see generality which can then
be simplified so that those rather complex mathematical models can be
reduced to simpler forms which contain the most important factors.
Then once that has been done, we can then go about rationally designing
prototype studies so that those prototype studies can then be scaled up.
So one did ask the question what exactly do we mean by scale-up? Are
we strictly talking about size of the system or are we also talking
about process engineering of the systems. And I would hope as you
consider the question of scale-up, that you will also consider the
question of process engineering because I heard a lot of people asking
questions this afternoon about such factors as what are optimal feed
policies? What are optimal size reactors for different stages within
the system? These are questions that can be answered for any given
system again once you have a relatively fundamental mathematical model,
because then you can play with that system and can come up with the
either feed pattern, with the loading pattern, with the reactor config-
urations that would be most cost effective for any given set of kinetic
parameters for that system. In addition, one of the areas of research
in the field of chemical engineering right now which is very important
as far as we are concerned is how to take into account uncertainties
during design and again once you have a fundamental mathematical model
for the process then you can begin to apply some of the new research
techniques to consider the fact that all of our parameters which go
into these models will have uncertainty associated with them. In
addition, we all face uncertainty when it comes to loading on our plant,
what will happen in the future on those plants. So there is a growing
body of knowledge now as to how to handle those things in the design
process. One will hope that this also will become a research need or
a research application within our field as well. So as I look at scale-
up then, I see scale-up in a broader sense and research needs for scale-
up in a broader sense than just taking a prototype unit and figuring
1430
-------
how to increase its size. I would hope that we would be doing research
that will tell us more about mass transfer coefficients, how they are
influenced by rotational speed, by disk diameter. I would hope that we
would be doing research that helps us know more about oxygen transfer
within the system because once we know that, then we can look at the
benefits that would come from air drive or pure oxygen systems as Jim
Mueller talked about this afternoon. I would hope that we would continue
to do research on the effect of rotational speed and disk configuration
on the fitness of the liquid layer that is carried up into the aerated
sector. I would hope that we could continue to do research on the effect
of disk configuration on the distribution of that layer since it is going
to be moving by gravitational forces. I would hope that we would continue
to do research on the biomass thickness within the system as a function
of rotational speed and loading upon that system. I would imagine that
the other speakers will discuss that problem as well. So I feel at a
point in time where we have an opportunity to develop an understanding
of these systems that would help us make extrapolations which have not
been possible in our field in general in the past. And with that I guess
I will ask you for your comments because certainly you have a lot more
experience with scale up than I do. So with that I will turn it over
back to Jack and let him fill your comments and your questions.
DR. BORCHARDT: We have five microphones available, one in the back,
one here, one up front here and one over here. Please avail yourself
with these microphones, give your name and state your question if you
have a question, or statement. If you have had a problem or whatever
it is, we would like to know it. This is your chance to express yourself.
Is there anyone who would like to make a comment relative to scaling of
pilot to prototype?
DR. SMITH: There is one obvious scale-up factor that, at least I think
it's very obvious, but .1 hear very rarely anyone ever mentioning it and
it is the fact that the matrix design for instance you can design, say
you need 24 units. Are you going to put it in a two by twelve matrix or
three by eight matrix, four by six matrix, six by four matrix, etc.,
one by '24, and it is very important based on my experience and some other
people here when we talked about it. This is very important because if
you choose the wrong matrix, often times you overload your first units
and they will go anaerobic and therefore you will lose your nitrification
capacity, etc. I think this is a very important scale-up problem and
should be addressed. Does anybody else have any comment on that?
DR. GRADY: Well, that's the thing I was talking about Ed, when I was
talking about reactor engineering within our system, is because once
you have a decent mathematical model then that's what it allows you to
do and you can look at all of those combinations with a very small capital
outlay as far as operating costs on the computer is concerned. And then
it just depends upon how much faith you have in the model.
MR. HENNESSY: One of the things that I have noticed today is that people
have not even determined what the real parameters are that determine real
operation of the bio-surf and even at that they have not even agreed what
the answers are once they come up with the parameters. What I would like
1431
-------
to know is do you really feel that at this time you can walk into a
treatment, a sewage plant that has to be upgraded and merely say this
is a cost effective system and we can put it into operation and it
will do this for you. You know, can you really guarantee that you
are spending the people's money properly?
MR. WONG: I am addressing to the manufacturers. We have many, many
manufacturing plants and we are going to use the biological discs data
and from what I have heard the last two days I am not too sure where I
stand from the standpoint that we have facility to be constructed. We
know we will have certain waste streams to be treated and we run a pilot
study with 2-ft. discs and I wondered...and we of course, we run it at
a very specific configuration that we take and specific conditions that
we take, and I am just wondering whether our tentative data package is
sufficiently comprehensive to be capable of going out on open bids to
the manufacturing communities. It seems to me from the discussion and
from the presentation that it almost necessitates that we have to tailor
our package to the pilot work that we have done and know what should be
the configuration of the discs that we have used and to the type of flow
rate and drawing rate, etc.. And we just come out with the generalities.
I wonder whether such a package would be sufficient for competitive bids.
I don't know how the manufacturers that are represented here view these
conditions.
MR. HYNEK: It's a. very good question and I would refer back to Al
Friedman's presentation this afternoon that he has raised very logical
questions, what is the work of a pilot plant, and we've discovered the
same thing. A small two-foot diameter pilot plant is an excellent tool
for determining what type of biology we will get, what type of removal
you will get, but aside from educating the customer as to the process
and I could say an advantage, simplicity of startup and quick degree of
treatment. It is a learning tool particularly in the industrial sector.
The customer does not realize the impact of his manufacturing process on
waste treatment, and in the process of a pilot study he learns very
quickly the economic impact of what he has been doing and it leads to
modification and improvement of his product process. We relied very
heavily on the small pilot plant because it was an effective tool to get
out in the marketplace, get someone to understand the process. As soon
as we could we switched to full scale data and basically our approach
today is, and Jim Mueller mentioned that Autotrol funded mass modeling
work, and I've been involved in that. I think this is a great step
forward. Basically we use the small models today to get a feeling for
it from our point of view because we have a large experience in data
bank. We know what the small ones can do. We have learned from full
scale installations what the big ones can do. When push comes to shove,
you're coming to us for an appraisal of what the full scale plant has
to be. We rely on the mass model information we have learned because
it is realistic. We rely on our full scale data bank, it is realistic,
and you are going to have to stop looking at ten-pound removals. You
are going to have to look at the realistic world. I do not know if I
have answered the question but I tried damn hard.
1432
-------
The use of mathematical models is a separate goodie. That work separately
from actual pilot plant data. So I'm an advocate of pilot data but why
emphasize the limitations of scale-up for turbulence and oxygen and
that is my thought.
MR. FRIEDMAN: There was some good work done in Canada a couple of years
ago on municipal waste waters where they scaled from. ..so I remember from
half meter to about six-foot. The scale-up factor required an additional
16% area. The question then was what happens when you go prototype and I
kicked this around with the primary author several times because 80% of
the times, we were both scared stiff point blankly, because we do not know
how to scale-up. I think both of us would argue that somewhere between
25 and 30% of linear scale factor out of ignorance.
DR. HOVEY: Personally addressing the fellow from the army's problems,
you've got what...10, 15 different configurations ,of pilot plants and
media. You got ten I don't know how many manufacturers sending things
out to bid and your starting point, your pilot operation is manufacturer
A, manufacturer C, a manufacturer D pilot plant. Perhaps what we need
is a standard configuration from which we start and we have done it for
say seven tests and all kinds of ASTM standards you know, and this is the
apparatus and maybe what we need is a standard pilot plant from which we
would develop our pilot data and then it becomes the manufacturer's respon-
sibility. You know you must match this pilot plant data with his full
scale materials that you are going to bid on the job with.
DR. KETCHUM: We are working on the project with RBC and -we have done the
pilot scale work and they are going to do the design criteria, and the
big question is okay, we have the pilot scale data that contains up to a
thousand mg/1 of formaldehyde, has a very selected culture degrading it,
and we have all this nice data. What do we do with it? For fun, you are
the experts. How do we scale-up? What kind of a biodisc do we use? What
is it a dissolved oxygen limited system. We know that. What kind of
disc configuration is the best for dissolved oxygen limited system? What
kind of loading rates are best? It's a very easy degradable type of ma-
terial. It goes very rapidly. We say 70, 80 percent degradation in our
flash chamber. What do you do from here? Can you go to a 50% disc space
in the first chamber and be able to make it run or is it going to do with
dissolved oxygen limited. What do you do with it?
DR. FRIEDMAN: I really do not have an answer for you and also I really
do not believe we have an honest set of scale-up factors at the present
time that we can apply.
DR. KETCHUM: We have gone out to four or five manufacturers and we have
come back with different configurations from each one. I mean vastly
different—nothing even resembles each other. So putting it all together,
we kind of came up with a configuration of our own. And another question
I would like to ask is okay, if you do a system that has 50 to 60% of its
area in the first disc, we know that we can get 80 to 90% BOD reduction in
that area saying that we are not dissolved oxygen limited. The pilot scale
study shows that the third and fourth discs which are running in at 2-1-1
configuration had very little growth on it. The third disc removed some
BOD and the fourth just went out for the ride. How can we in a full scale
1433
-------
unit design, so that we can get better loading of those discs and higher
BOD reductions so we can get lower effluent BOD valve.
DR. O'SHAUGHNESSY: You started this mess, I'll let you finish it up.
In the beginning when we were talking about scale-up factors and we were
going round and round, and there has been one pertinent comment tonight
on scale of factors, all of us are running around doing our own little
thing with this unit and that unit and they are not the same, and I think
if we would go into a laboratory and take any wastewater and put it in a
settling column test with a six inch diameter, and go to a two and half
foot diameter it would come out with different settling curves. And what
do we use for scale of factor on the domestic wastewater? Well, we go
and find out. Why in the heck don't we start with some standard pilot
units on domestic wastewater so we can get some information to look at a
certain geometry on a fullscale plant and find out what the heck scale-up
factors are, instead of beating each other over the head. Forget indus-
trial waste, we will never get that answered.
MR. WONG: As a customer I wish that various manufacturers will form a
league or an association or society to come up with a typical pilot model
from the standpoint that if the customer wants to have open bid on the
job and if the pilot model is followed, these manufacturers will be able
to bid on the job without any reluctance. If they can come up with that
type of agreement among themselves, it would help the customers a lot.
DR. BORCHARDT: Now, we are going to pass on to the next subject which
perhaps will be a little bit easier for us to accept, and this has to
do with kinetic factors internal and external diffusion, and the discus-
sion will be led by Dr. LaMotta, University of Massachusetts.
DR. LaMOTTA: What I would like to do is summarize the most important
concepts relating to external diffusion and internal diffusion and both
needs for the mathematical modelling. From my viewpoint, first of all
the mass transfer can be studied in two general cases, the disc has :
surfaces exposed to the air which means we are getting three phases; the '
gas stage, liquid stage, and the biofilm, and we have another case when
the disc is emerged in the trough and we only have two phases; the liquid
film and the biofilm. Then if we consider first the most difficult case,
with the disc in the air, we have three phases we have to consider; the
transfer of oxygen from the gas phase to the liquid film, then the trans-
fer of oxygen from the liquid film to the biological film. So we have
two types of mass transfer processing, and none of them have been addressed
so far in the literature. Most of the models and most of the studies have
concentrated on the transfer of oxygen from the gas to the bulk solution
in the trough. I do not remember seeing any paper in which we have dis-
cussed what happens with the dissolved oxygen, for instance, on the liquid
film in the air. What is the effect of biological utilization of dissolved
oxygen on the liquid film in the air? Well, some people may say if the
rotational speed is slow then you get dissolved oxygen saturation in the
liquid film. Well, if that is the case, one may think the rotational
speed definitely has an effect. However, that problem has not been ad-
dressed yet. With respect to the transfer of the dissolved organic from
the bulk of the liquid, when I say bulk of the liquid I mean in the liquid
film, to the biological film, again we ignore exactly what is going on.
1434
-------
r
MR. WONG: I do not think the question is really answered from the
standpoint of I really still do not know what to do? That this really means
that, from what I heard, is a pilot is only to build the confidence
of their customer from that standpoint, I really do not need the
pilot. If I do not need a pilot and of course we have to prepare
a tentative package to go on bid, what do manufacturers need in the
tentative package to be able for them to provide the service?
DR. BORCHARDT: Well I think you must realize, that Autotrol has been
getting data for perhaps the past fifteen years. So now they do not
need pilot plant work as much but when they first started they desper-
ately needed pilot work.
MR. WONG: When you talk about pilot work, I think you will agree that
each stream is merely different...And Autotrol I wonder if they have
ever had a stream that contains explosive and repellant particles. You
know it is different., I'm just wondering whether our pilot really serves
the purpose. What does it intend to do?
DR. BORCHARDT: Now I think that we might discuss this all evening and
not really reach a plateau of mutual agreement. There is one thing about
industrial wastes such as you have mentioned, in general, they are
probably sterile and you just do not turn the pilot plant on and say
whatever results I get that is what the prototype will do for me. It is
a question of developing a seed that seed will be happy in that environ-
ment. Now many times the environment changes suddenly maybe, but the
environment does change,, You have heard an awful lot about Beggiotoa.
For example, I've seen plants where overnight Beggiotoa will appear and
then in a similar sharp fashion go away. That means the biomass increases
from 1/8 to 3/16 up to perhaps a half an inch. The mixed liquor solids
jump up from 250 to 2500 purely because the substrate has changed suddenly
and you cannot model that easily because it is one of those things that
depends on proper seeding and so forth and especially in an industrial
waste this is a difficult thing. I have run an industrial pilot for
months and couldn't get the thing to start working well, even though I
kept throwing seed at it. So these kinds of things do take a little bit
as we call it empathy. You have to understand the system and work with it.
Now I would say that is a pretty tough question to say are you spending
people's money properly,, So I think shouldn't really debate that. Let us
say in order to meet an established standard I think that is something
else again. I think we can accomplish that. I am not sure that your
question meant that; however, it would be my opinion that in most cases
a good consulting firm that has the tools that have been mentioned here
at the meeting would certainly be capable of telling you how to meet
this state of objective with this RBC process.
MR. MADDEN: It would seem to me that if we were going to start to look
at scale-up, we might first address the possibility of securing data that
would address scale-up as it has to be addressed, that is to say, take a
plant that has similar characteristics of waste water of the various dia-
meters or various pilot sizes and secure absolute data, hard data, from
which you could then produce your mathematical models. I think to start
from the mathematical model end and work toward the plant reality or the
treatment facility as it will ultimately be designed, may in fact be the
1435
-------
wrong direction. I do not think there is sufficient data available
from any of the manufacturers or from any of the waste treatment
plants that are now on line that can produce the quality of data
required to employ mathematical models, something you can scale-up
from a two foot diameter, or four, or six, or ten, and ultimately
design a 24 or 60 or 80 shaft job on it. Also it should be pointed
out that most of the facilities, I know speaking for our facility,
most of the operational facilities that we have on line are nowhere
near either design flow; they may not be near organic stream flow;
they may have idiosyncracies that were not considered at the time
the project was designed. So to secure that data and then force the
data into a position so that you can develop a mathematical model
and accept probability or a mathematical probability as the criteria
for then saying it is going to work, it is going to fit, it just does
not fall into...it just seems that that is not the way to go.
DR. BORCHARDT: Well I think in general we are not actually here to
debate with you that particular issue if that is your conviction. In
general the use of mathematical models I think is here to stay. In
general what you do is use the model to indicate to you the variables
that you have to define with more finesse which you then proceed to
do, and then you work back and forth until you get something that seems
to you to work, and in general that I think is what is happening in
this field. At first we started out empirically but now we are moving
more and more in the direction of getting models. I myself have used
two-foot models, 18 inch, one-foot models, four-foot models and on up
to the 11 and 12-foot prototypes. So that I do feel we are using models
in that fashion and I feel that they are doing a great job for us in
helping us. Frankly I feel that is the way we must go.
DR. MOLOF: I am an advocate of models. We had a scale-up system. We
showed that RPM was the scale-up choice but you cannot do it because
this couldn't take it. Now we have a way of doing it. The problem...
let me just remind everyone about scale-up. The bacteria that grow on
the disc of a 3-foot are the same bacteria that grow on the 12-foot disc.
If we can duplicate, we don't have to duplicate the bacteria. They grow
the same size. We find out on the 3-foot disc or the 2-foot disc or
1-foot disc what the properties of the system are and domestic waste
changes. It has infiltration inflow, dilution stream. I've done a lot
of work with different types of sewage and different types of industrial
wastes. I can learn a lot by pilot, and let me explain I was in industry
and I was in a big company and they were wise guys like some of you and
they said let us skip the pilot plant. It went from laboratory to full
scale plant. They had to write it off for several million dollars a year
later. I think that construction is too expensive, that if we have any
doubts, go to a small pilot plant. They are not that expensive. You
learn a heck of a lot. You learn we are going to go anaerobic. You
learn do we have enough oxygen. Let me just comment what we have to
learn to scale. You have to learn to scale mixing. Now mixing is some-
thing that the field has been working on for 50 years. We should have
that expertise in mixing so we have to study that as Dr. Grady said we
have to study turbulence in these reactors; we have to study fundamentals
such as oxygen transfers. Oxygen transfer is the other thing. If we
can scale oxygen, we can scale turbulence. I can go from a six inch to
a 12-foot and be right every time, and that is what our goal should be.
1436
-------
Therefore my belief is in that part of the rotating disc exposed to
the air, we know nothing, absolutely nothing. And if you want to
make a good progress in understanding the behavior of the biological
film, and in the rotating biological contacter with the biofilni in it,
we had better begin to do research in that area. Most of our research
has concentrated on the second case which is the innermost portion of
the disc. Still then in that case, the assumption made in most of the
mathematical models has been that the mass transfer coefficient K-^ and
all those correlations presented in the papers. Can we directly cal-
culate dissolved oxygen in the bulk of the liquid given that we know
the geometry of the disc being considered. One of the basic assumptions
in the mathematical models has been that the liquid film that we have
in the air. If the oxygen transfer occurs instantaneously upon the
entry in the trough that is false. As Dr. Grady said this afternoon,
that is a lie. Then, in the mathematical model we make that assumption
which is a lie. You see, nobody knows exactly what the effect of not
making that incorrect assumption will be in the performance of the model.
Therefore.my suggestion would be let us make the more difficult realistic
assumption that there is not any completely exchange of the liquid film
in the trough. In speaking of a. disc boundary layer when the disc re-
enters in the trough, and that disc boundary layer thickness is larger
than the thickness of the liquid film of the disc in the air. So we
have to study the mass transfer coefficient of the oxygen and this or-
ganic from this boundary layer, from this boundary layer to the bulk of
the liquid and to the biological film. We ignore, absolutely ignore the
mass transfer coefficient in this case. This is what I think the most
important research needs in the future. With respect to internal diffu-
sion, all of us knows the mathematical models have to use the coefficient
of the diffusibility of the substance inside the biological film. The
first big problem we have in dealing with the subject of the sewage.
We asked the question, the diffusivity of what? Some of the mathematical
models have used the diffusivity of glucose. But we know that is not
realistic. And then the question is, should we use, if it is a compound
or complex organic, like in the case of sewage where for the most part
the mathematical model, modelling the diffusion of substance inside biolo-
gical films. In my opinion we should try to understand first the pheno-
menon, using the statistics compounds that can be directly measured
instead of using indirect preparation of parameters of COD, BOD, or TOG.
We do not know anything about the effects of film thickness on the effec-
tive diffusivity of the substance. In most mathematical models the
effective diffusivity continues to be a constant. We have known in the
case of suspended growth, the effective diffusivity is the function of
particle size diameter. Similarly you might expect the effective diffu-
sivity is not a constant, it may be affected by film thickness. This
knowledge has not been explored yet. If you pull everything together
in a mathematical model of the process, then you have to begin consider-
ing all these complicated facts. They may lead us to a logic equation.
We saw in the first equation of that happening due to the biological
portion of the film exposed to the air of the liquid film exposed to the
air, and we have to combine those effects with what is happening inside
the trough, and we wound up with a logic differential equation. For
the verification of the model, we have done by using BOD, TOG, or COD,
have a parameter to measure concentration. In my viewpoint that is
inappropriate, simply because BOD, TOG and COD are global parameters,
operational parameters that do not let you know what is happening with
1437
-------
each one of these specific compounds that we have indeed. So even
though I advocate strongly for mathematical modelling simply to better
understand what are the most important factors affecting the performance
of the unit, I wouldn't recommend using directly that mathematical
model as a mathematical formula and use BOD, TOC or COD as a parameter
to measure concentration. With this I would like to open the discussion
on several topics. I know that there are people who have not done work
in this area simply because it usually involves high mathematical differen-
tial equations and most practicing engineers do not like to deal with
that. However, there are some people who have done some work in this
area and I would like to discuss this problem with them.
DR. GRADY: I wanted to throw out one other thing, along these same
lines for those of you that are working physically in these areas to
think about: that is, is it possible to come up with a chemical system,
perhaps an heterogeneous catalysis system, that is much simpler from
the reaction standpoint, that could serve as a prototype model that
could be studied in order to get at these basic physical mass transfer
parameters in a simpler manner than is required with our complex biolo-
gical system.
DR. LaMOTTA: Well, in that case even the physical would make sense...
there might be the criticism that we are neglecting the biological ac-
tivity. Some people even believe that there is an active transport of
substances endued by, say the cells which are inside of the biological
form. And so we would be neglecting these effects, that is the only
danger. But I believe that you are right. We may understand better the
system if we work with an inert system, and we would then study only the
physical parameters involved in the process.
DR. WU: In order to avoid the unnecessary argument which may occur at
the second symposium, the methods or techniques employed for the measure-
ment of the RBC system operating parameters should be defined. Otherwise,
the kinetic constants obtained from different studies may be different
although they are derived from the same model/equation. For this reason,
the research funding agency such as EPA or NSF should look into this
problem and make suggestions to those persons who are planning to conduct
the RBC research in the future.
DR. CHARACKLIS: I would like to comment first just on modelling in general,
With too many people the word modelling kind of provides them with a shiver
up and down the spine when they hear the word, and I think the problem is
that all of us deal with models, everyday, not just with our mathematical
modelling, and that is, for example;the example I use with my students
frequently is everybody has a model of how a car operates; my wife's is
very plain. She models the automobile; the car operator by turning the
key and toward the ignition. And it is very plain, every once in awhile
she has to look at the gas gauge and she knows that it needs some gasoline
and it goes on. The mechanic, despite some of our experiences, probably
has a different idea of how the automobile works, and a little more in
depth. I think the modelling that Les is talking about is critical and
I think it is critical from one standpoint: it is not necessarily going
to solve design problems. Good engineering is going to solve design
1438
-------
problems. But modelling, I think, is going to provide a framework for
understanding the system and for interpreting the data we get. It does
not provide the answer. And modelling is not something that ends at
some point, it is an overview process which is continually upgraded by
new results. From the standpoint of the specific questions that we are
talking about now, I think BOD is really providing a barrier for further
progress in understanding what is happening with this RBC. The only
thing unknown of that believe in Science are the conservation equations;
it is what everything derives from that we do. You cannot do a material
balance on BOD. So how the hell are you going to do any modelling using
BOD then, if you cannot even do a material balance with it? You cannot
get a diffusion coefficient with it I will promise you that. Or if you
do I would love you to explain to me what it means. Then, finally, let
me point out, the diffusion coefficient, if you measure it even with a
specific compound, if you are lucky you are going to get plus or minus
fifteen percent...if you are lucky. Those quantities in Perry's Handbook
for chemical engineering, plus or minus fifteen percent.is real good
T-* that is with very defined systems. Still diffusion coefficient is
important in this fundamental knowledge. When you get out on the field
you are going to take that model and you are going to cut it down. You
are going to go from the mechanic's model of the automobile to my wife's
model of the automobile, but in true you are going to have some understand-
ing, some fundamental understanding of what is happening in doing so.
You are not going to be doing its modelling with exponential formula too
where units are not even considered in order to design a system. Kinetic
is always important. I bet you can count the number of reactions on your
hand, the chemical reactions that we know of today, whose kinetics were
determined at priory. Everything has got to be done empirically and it
is done within a relatively small range so we can use relatively simple
models. Make yourself something like what Dr. Grady was talking about
which has three or four parameters to something that has maybe two, it
could go up to a small range. Dr. Grady is looking for more comprehensive
models. And that is admirable and it is necessary, but the lady that
has a problem back here, she needs good engineering. That is as far as
I see it, engineering is never going to be cut and dried. If it is I
am going to get the hell out, and there is not going to be any fun then.
Because of that is what these people are paid for, so I think there is
a problem here thinking everything is black and white, and you know...
That's about all I have to say.
MR. ATHAVALEY: Referring to which mechanisms actually controls there
could be a lot of debate on that. One could study whether it is transfer
from gas phase to liquid phase or the mechanism is in the interphase. If
you open any kinetic book you will find there a chapter on retention time
distribution, and all the constant data were determined on pilot scale or
lab scale, are then taken into account when actual distribtuion of reten-
tion time is taken into consideration. I took a look on configuration of
the plate or disc which are on display and the main problem there is dis-
tribution as referred to solid distribution or with wastewater retention
depend on how this disc really performs, and for this reason there should
be a combination of some integral relationship of kinetics or the practical
application of kinetics are empirical, and combination of kinetics with
fine distribution and the aid of particular species is important. I discuss
that point the scale-up and design engineering with regards to kinetic
reactor, with regards to fluidized bed reactor, are done on the basis of
1439
-------
retention time distribution. The technique used for that is chromatography
technique or photographic technique, so on so forth. I am not very clear
at this stage how this could be applied to the disc reactor which is
complicated and the reactors I have just mentioned. I am strictly talking
from kinetics chemical engineering viewpoint.
DR. MUELLER: I would like to make a few comments on it myself. With
respect to the model, I think academia for a long time has had a bad name
with respect to model because the reactor you do not understand, and we
tend to identify them with the utmost degree, that is where we make our
money. We make our money from the research project. Why lick the problem?
That is one aspect of the model, that is one aspect of the understanding
the system. I personally think we have got to do that more and more. The
more we learn on the academic money and the more the field will know what
is going on with the system and they will ultimately design it better.
That is one aspect. The second aspect is using your smart from academia
and saying here is what we know of today of our modelling and here is what
we all know about all the parameters we got to have in this thing. Now,
with respect to glucose diffusion, I will be damned if I know how BOD dif-
fuses, but I do use glucose because I have nothing better to go on. I
think my kinetic parameters are incomplete and I got a lot of ignorance in
the model, I am not sure how all these other kinetic coefficients readily
do close, and I will be damned if they do not fit the model which uses
glucose as a BOD diffusion parameter. Now, I get numbers for those kinetic
coefficients. I cannot swear that those are the numbers of my need and
I am not sure of the ultimate truth, but I can swear that those fit the
models under different operating conditions, engineer to the other aspect,
not the academic, not the research aspect of which I am a part. The en-
gineering aspect, I want to evaluate that system this is the best I have.
I have got a model, I have got imperfect information on it, hell, I am
going to assume some numbers. Now, if I can fit that numbers for that
set of data and I can get another set of data on the same waste on a dif-
ferent operating conditions, that is fit and I have another set of data
on the same waste, with all those numbers I get on that data, the A, the
K, the B, they will all be off, but I have got a fair degree of confidence
that I can extrapolate this condition and that condition to that condition,
fairly wide swing and still predict the point. Now, as an engineer with
the best capability they have today, in my hands, from the academia and
from anybody else, I can use that to extrapolate the other conditions:
little question on how accurate it will be, but it is the best I got.
Now, I think that is where we are today with the models. We can use it
for design, we can bite the bullet, and bite somebody's coefficients and
say if there is ever, we can apply this model for this case and for that
case and so far we haven't said we can. There is a lot of room here. But
we did it academically. The story today are selling equipment, Clow in
the market and Autotrol in the market FMC in the market, and some others
in the market. Independence, of how long it takes to solve his academic
problem it is going to be a lot of work and a lot of money on his initial
work that is going to get some solutions, just we have known today. So,
it has got to be a marriage in it too. It has got to be a marriage of
fighting upon accepting the best knowledge we have today. Based upon
this information engineers can select the system on a market the best
we have today. And we also have to say to ourselves that we do not know
it all and we have to go out to get more data.
1440
-------
MR. OPATKEN: I would just like to bring up one point. We know the COD
reduction at each stage. We know the DO level coming in, we know the
DO level coming out. We know how much oxygen has been transferred. We
also, with that same thing, have the mass transfer coefficient from it.
I do not know where the problem is.
MR. HARRIS: I do not know anything about RBC's and so that it gives me
advantage. I can ask a dumb question and perhaps get away with it. I
have heard some comments and encouragements that RBC's will give a higher
degree of performance than a trickling filter, so my question readily is
why should that be since both units are essentially the fixed film reactor?
I do not understand at all.
DR. BORCHARDT: I think I would go to the origin of that and ask that in-
dividual the question. I like to think of an RBC as a horizontal trick-
ling filter myself.
MR. HARRIS: Rather than a fixed-film reactor.
DR. BORCHARDT: Sure, well that is what it is.
MR. HARRIS: Why should they in a sense be different?
DR. BORCHARDT: I do not believe they are different except that we are
accustomed to recirculation in trickling filters and we are not in our
RBC, that is the difference.
MR. HARRIS: The recirculation...
DR. BORCHARDT: The recirculation factor, yes.
MR. HARRIS: Is the one that makes any difference that is observed?
DR. BORCHARDT: That is the difference. We are not accustomed yet to re-
circulating. I think there are times when we should be in the RBC we
are not, so far.
DR. CHOU: One recommendation as far as the research priority is concerned,
we have talked about a problem of uniform methodology as far as measuring
the organic component or components, and I do not know the percentage of
nitrification required in this country. I think there is an important
market and the ammonia analysis is uniform and there is easier thesis to
work with so I recommend some research effort can be channeled into this
particular area and that is my comment.
MR. DENNIS: In response to your question about the trickling filter I
also do not know a lot about RBCs, but it is kind of appealing to think
of an RBC as , or a trickling filter, as a secondary biological contactor
and I think the reason why the performance in the RBC is better is because
the reactor is longer on its plug flow system, whereas in the trickling
filter you have about maybe six feet of filter stone.
1441
-------
DR. BORCHARDT: I want you to all realize that we are not up here because
we know the answers to the question, it is we were selected because they
thought we could listen politely to your questions, and tell you there
are no answers.
DR. CHARACKLIS: I have a question too. I have been working with biological
film system from a completely different standpoint, and at least from a
different problem. Two of the problems we are concerned with:first, the
problem is essentially energy lost, frictional resistance decreased, the
film growth in the tube and secondly heat transfer resistance increased
by the fact that there is a biolfilm inserted. Now, you know when we
get into this problem you look into the literature for heat transfer
coefficient and mass transfer coefficient for example, and you find that
correlation in turbulence flow which is all you can use : correlations :
of these mass transfer and heat transfer coefficients always talks to
friction factors in a flow system. Somewhat is easy, you have measurements
to make too. My question is we use in one of our systems a rotating drum
and grow a. film on it, and the measurement of frictional resistance is
simply applied a torque meter between the motor and the drum. It is a
very nice indirect indicator of how much film is there but it is a direct
indicator of the amount of frictional resistance at the surface, which
should be closely correlated to mass transfer and in our case for example
heat transfer. I wonder if anybody has ever tried keeping a torque or
in fact whether the system was sensitive enough to detect these changes
in torque on the shaft due to the biofilm on the surface of them.
MR. NICKLE: From municipal wastewater with four shafts and four stages,
does it matter whether the shafts are loaded perpendicular to the shaft
or actually parallel to the shaft from a hydraulic standpoint and process
standpoint.
DR. BORCHARDT: Well, I think no. I loaded them both ways and actually
the mixing within the stage is so complete, one time 1 measured this and
I found that I got seven turnovers for each flow through, in other words,
there was complete mix in each stage and it does not matter whether you
put it in longitudinally or across the set of discs, if I understand
your question correctly.
DR. MUELLER: I would like to discuss the dimple effect for the moment.
From the actual, our initial workdown in that area with the Autotrol
media, if they grow too thick a film, the densest part of the media and
cut down active surface, and we can cut it down significantly with large
film thickness. We do not cut it down with normal thicknesses of a six-
teenth or so, you start the disc thickness of half an inch really hurts,
you have go no surface area left for transfer of oxygen or substrate, but
the principle effect of the function really of the thickness of your bio-
film. The more thickness you get in the greater the surface area, but if
you got a lot of biofilm thickness on it you .Just wipe it out like that.
So there is got to be a joining of both effects, surface area by dimples
made, and biofilm thickness. And that will optimize the system.
DR. BORCHARDT: Now, up to a certain point that film was completely porous
on a microscopic basis, but it is that point that we do not know about
that you are speaking of. Apparently, the anaerobic region on the dimple
surface is wasted, and I think I am inclined to agree with you.
1442
-------
DR. BORCHARDT: Well, let us pass on to the next subject, because I
feel that all of these things are so interrelated that it is very hard
to separate them and we will just discuss this next subject and perhaps
you will see what I am talking about. In any way, for Group III, we
have selected a brief discussion here as RPM in Turbulence. You all
recognize the fact that many of these have different surfaces, obviously
turbulence is an important aspect of this whole thing. We are going to
turn this over to Dr. Friedman from Syracuse University.
DR. FRIEDMAN: I can sum it up very quickly. I do not know, I do not
think anybody else does either. But let me give you some ideas. All
the models I heard today assumed the uniform liquid film across the
surface of the disc and yet when I go into a laboratory or go out in the
field to look, I notice that there is no such thing as a uniform depth
of the liquid film over the surface at any point on the surface, or any
two points. Now this has a lot of implication in terms of diffusion of
oxygen from the gas phase into the liquid phase, or at least to the
biomass. So that is one thing we need to improve in our modelling. Now,
I attempted to do this a number of years ago when I went and talked to
a resident hydraulic engineer, he was a darn good one; he scratched his
head a little bit and went away and thought about it. He came back and
said why don't you go upstairs and try chemical engineering. So I went
upstairs and talked to the resident chemical engineer who was presumed
to be good in hydraulic, and he sort of scratched his hands awhile, and
said why don't you go and ask Dr. Lee. The person who produced a simple
plate immersed half way it would not be too bad, but when we do not im-
merse it all the way or some other percentage we really have a tough sit-
uation. Now, I am saying from a modelling point of view to start with,
we do not know how to describe the film thickness. Now, we got into
another area yesterday, I heard was a. kind of heat, how far apart should
these discs be. Then the next question is what is the effect of the
dimples on the disc, each manufacturer has a different set of dimples.
At least that is what I call them. I am assured by one manufacturer that
they have an absolutely beautiful computer designed that selected this
particular configuration, and another one told me today we happend to
stumble into it, but it is the best one. I am being facetious but I am
really pleading ignorance. I do not know, and I do not know anyone else
that really knows. Certainly this is going to be critical in terms of
scale-up. And the answer is we really do not know, going back to playing
games with industrial waste on how to scale the stuff. I think, and we
have not tried it yet, nobody has seriously taken a series of three
different sized scale-up models starting with two-foot, perhaps six-foot
and then full scale and runs through a series of different parameters in
terms of testing and looking at least obtaining some empirical.factors
that might be useful while we are waiting to develop our models to a
level that we can really believe in them for predictive purposes, and
not have to be waste speicifc and calibrate each model one on a parti-
cular waste. I would like to make one answer to Dr. Characklis' comment
in terms of why BOD-5; because it is legally required, Bill, that is the
only rational reason. With that, I .would like to quit and open it up to
the floor.
_
1443
-------
DR. MUELLER: I guess I do not agree with you.
DR. BORCHARDT: You will accept the first 120 microns.
DR. MUELLER: That has to be the case. It is the next four thousand
microns that I am really concerned about. And under the anaerobic
conditions. What you do, you shut off the whole channel for any kind
of transfer. Once you clog it you shut it. Maximum 90% and probably
half-way, you should have a good portion of your surface area, and we
can account for that sort of thing, but the optimal design for all
manufacturers is to keep it to a certain level so you optimize the
area available for your dimples, the active dimple area if you can
provide. That is not that big.
DR. FRIEDMAN: Jim, I am not arguing with you, but I would ask you,
no I won't either, I won't embarrass you by asking which one has the
best dimples. I do not think anybody can honestly answer that right now.
MR. MADDEN: Well, I would like to respond to the gentleman who asked the
question concerning the energy usage and the biofilm thickness, and there
has been a certain amount of work done by all the manufacturers in deter-
mining what the amount of energy required, what the torque of measurements
would be with the increase in biofilm thickness, but there is the consid-
eration that more study has to be done because the biofilm thicknesses
change in character and if you get a long and stringy filamentous growth,
you have one set of problems, and if you have a just a lesser stringy
growth you have a different value of torque absorption, but none of the
data that we have accumulated would be with the acceptable in what you
are looking at because it is not as fine, it is not as tuned as you
would like to see it. For example, we would do it with one meter meas-
uring energy to the motor end and find out exactly how much current is
used in that equipment. What you are probably looking at is something
like the use of mechanical gauges and that like, and we did some of that
work. It is not in depth to the extent I think you should look at, but
we are looking at more of that, and I think other manufacturers are doing
the same thing. Somebody else had a statement and I think it really
belongs to the manufacturing pocket, concerning axial or longitudinal
application. And speaking for myself, we would prefer longitudinal
application as opposed to having it actually, and we do think you get
better mixing in that way in spite of what you said.
DR. BORCHARDT: I will accept that.
DR. O'SHAUGHNESSY: Up until now I have heard all the things we do not
know about RBCs and what is going on. If you look at the other things
we did not talk about at this conference: activated slude. Over the past
ten years, you realized that as this is the thing developed for building
a lot of these plants and we are still merry go around, and they get
messed up, but in the activated sludge system, you are able to change
F/M ratio as long as you can to get enough oxygen in there. Now, we
have heard models and we have heard diffusion, we have heard which dimple
has the best effect and so forth. I have a simple question since we really
do not know how to optimize anything from scale-up to whatever we want to
do, and you want to optimize this design at when you get an oxygen deficient
1444
-------
system situation in the first stage, where do you put variable speed
motor or variable speed rotation in every unit so when we do screw up
at least we will not be sued quite so bad. Any participant can respond
to that.
MR. ANTONIE: We have looked at variable speed driver for a decade, and
they are ferociously inefficient. We have earlier found that most me-
chanics are not good enough, and if you want to double your power consump-
tion you can go into a variable speed drive. But it is very imprac-
tical. In regard to biofilm characteristics and energy consumption, that
is whether it is a relatively filamentous or relatively thin layer, biofilm
thickness control is a very important parameter, and our recent experience
using air driven and air supplementing contactors, has shown a great deal
of effect in keeping that biofilm as thin as practically possible and
necessary, and you also want to make sure that you have got the full ef-
fectiveness dimple surface area.
MR. MADDEN: Regarding the variable speed drive feature. It may be worthwile
for us to look at that from a viewpoint of study and something to consider
in mathematical modelling, but one of the things we find that is a poor
feature from an operational point is that to leave a trickling plant
operator in the position of the doctor or the judge or the jury on varying
the speed of a trickling plant, with two, four, five MGD behind him,
gets down to be a very bad situation. Somebody decides that they want
to run slow or faster and you can have a whole set of operational problems
that we could probably get into in the next discussion area, but that is
something we think that is a bad feature of going into variable speed
drive or multi-variable speed drive within a waste water treatment system
of RBC.
DR. BORCHARDT: Yes, that is going to be the next area, so any of you
that are talking about operational problems please give the next speaker
here a little chance, a little opportunity.
DR. HAUNG: Well, we have talked a lot about biofilm thickness. I think
when you talk about biofilm thickness, refer to the thickness and say it
is kind" of misleading, for we are talking about how much of the active
microorganisms which are involved in the anaerobic oxidation of the organic
materials is in the wastewater, and all the time that people find no matter
if the biofilm is thick or thin they all have the good BOD, COD or BOD re-
duction in the treated effluent. That is because no matter, you have the
thin or thick one, even the thick one only the very top surface is useful.
The rest of the deep inside are really junk because they are anaerobic and
are partially effective in oxidation work they will do partial of the job
and of which is not accomplished too much of the COD reduction ...think of
these in terms of the oxygen transfer from the liquid or from the air phase
to the sludge phase. Which one is more important in the external phase,
or in the gaseous phase or the' liquid; phase? It is my personal opinion
that the gaseous phase is much more efficient. I am not a chemical engi-
neer or a physical chemist but in the laboratory when I use a YSI DO probe,
first I have to calibrate the probe, I stick the probe into the water, and
just shake it for a few seconds, then dissolved oxygen reach the saturation,
so I personally believe the oxygen transfer from the gas into the liquid
1445
-------
film is much faster than from the bulk solution into the stationary
stagnant layer.
MR. JOOST: I do not get out too often, but I have got a couple of comments.
As you know, activated sludge has been around a long time, lot of accounts
have been made for mathematical modelling, still have alpha and beta factors,
and all of a sudden now we want to pinpoint the RBC because this is going
to be the panacea of all the problems, Well the RBC system happens to be
just an ordinary biological process. The bottom lines is that the operator..
he is the guy that has got to produce the effluent, and no matter what we
tell him by a mathematical model which is a steady state formula based on a
given film thickness which really does not occur, they put on another alpha
or beta factor. This is not going to help him operate the plant, that is
one comment. The second comment was that dimple surfaces and air drive and
supplemental air. I did not get a chance to speak this afternoon, we had a
lively discussion. I would like to mention Jim Madden, we are the other Com-
pany that tests bioshaft, that failed to get mentioned by the way. But any-
way as far as the...way back when the Europeans developed, excuse me, com-
mercialized the rotating disc unit they used a flat surface media; poly-
styrene foam. And when Autotrol and their development program came out with
the configurated disc they you know, kind of haw-haw and laugh-laugh. But
it is really not going to fly. But they do build a hell of a good pilot
plant because we tested the pilot plant against the flat surfaces we had,
sometimes the Autotrol pilot unit produced better than we did, other times
the flat surface produced better. I guess it was about '71 or '72, and
Ron you can correct me on this, when it first came...when you first came
out with the manual about comparing the flat surfaces with the configurated
discs, to say that the result of Pewaukee study demonstrated that the unit
surface area was giving you the same performance as the flat surfaces,
nothing was mentioned about RPM, and I think at that time you had maybe
about a 1.9 RPM. The two-meter discs test at Pewaukee, you had about a
1.9 RPM and comparative to the flat disc surfaces, which had the same per-
formance per unit area, but the flat surface was operating at 0.8 RPM,
which, yeah, you can get the same performance but it took you a little
more energy, a little more compact time to obtain the same performance.
When you first came out with the air drive unit there were a lot of comments,
in fact I even wrote a little dissertation on the physical disc design and
I got a letter from Autotrol that is a no-no, that regard to the industries
I had to back off on that, but that did come out with that patent, and in
the patent the figures show that the air drive unit which can produce the
supplemental air which can produce the same performance per unit area at
a. 0.8 RPM versus a mechanical drive at 1.6 RPM is just giving us the sur-
faces back which we robbed in the flat surface area. Now, flat surfaces
can get the unit's performance at 0.8. When you got to the configurated
surfaces you had to go to 0.6 to get that same process performance per unit
area. The air drives, according to the patent now, you are telling me I
can go back to the 0.8 RPM which is where we thought about it, and when
someone mentioned about the dimpled discs and you know whether they fill
up or not, I think Jack, when we were doing some studies at Ann Arbor, we
both kind of agreed that the plane surfaces as probably the ultimate in
surfaces.
DR. BORCHARDT: That is all we had.
MR. JOOST: You are right, that is all we had. And now that we can do the
configurated surfaces whether it be channeled, dimpled or pimples, as I
1446
-------
understand the latest one is the jelly roll, we still are trying to approach
the same process performance per unit area as on the flat surfaces.
Those are my two comments.
MR. GERHARD: Now we have had thanks to Bob Joost through the work you
did at Ann Arbor and other places Jack, we have had I do not know if
it is as much as or more than other manufacturers but we have certainly
had one heck of a lot of experiences on variable speed driver and the
effect on the standard, so-called standard domestic waste, and more so
even on industrial waste which has some wild characteristics. We know
and have the physical evidence to verify that the DO in the mixed liquor
will change depending upon whether you increase or decrease the RPM or
rotational speeds. The BOD removal or reduction process efficiency
changes, the suspension of the solids coming in and/or generated within
the reactor, changes with speed. The temperatures were found on the hot
industrial wastes will be effected, the cooling effect of the disc is
substantially changed by vibration when you have higher versus lower
speed, the effect on large scale units as well. So from the standpoint
of what was set up by Mr. O'Shaughnessy, yes, we have done this, I
would expect other manufacturers to have done this, and I have seriously
suggest to the basis protocol of this meeting this evening as to the EPA
or other sponsored research that has to be one of the absolute musts.
And then, Jack, the other thing that you have done at,or .not you neces-
sarily but the Ann Arbor facility has done, as you explained there, the
NSF situation, National Sanitation Foundation or whatever its proper
title is, can set up a program to compare various package plant manu-
facturers designs in so-called prototype size units to find out if
number one, I presume they met their own claims and number two, how
they met the identical criteria of other manufacturers say their stuffs.
With this, would this now be a practical thing for the Army, if you will,
for EPA who is by far the biggest funder of all the projects to the tune
of 75 or more percent. It is perhaps some grants from the states or
regions which will also be involved in that they stand to benefit by
standardizing as much as they can between different dimples or pimples
or whatever the case may be.
DR. BORCHARDT: Bob, the NSF tests against the criteria, the criteria is
established by the industry and regulatory agencies and I suggest that
you get the industry organized along with the various regulatory agencies
to establish criteria performance and I am sure NSF will be very happy
to test against that criteria.
MR. GERHARD: Well, I would like to ask Mr. Opatken also to express the
opinion of EPA on the situation like that is there or was there or will
there possibly be some considerations of that type of thing.
MR. OPATKEN: There is always consideration.
DR. BORCHARDT: We will have a general discussion after everything is over.
Mr. Lagnese of Dunca, Lagnese and Associates in Pittsburgh is now going to
discuss operating problems and many of the things that we have talked about
the kind of plant operation. Once we have completed that and the discussion
that goes with it, the floor will be open for general comments just as long
as you want to stay here, so bear with us.
1447
-------
MR. LAGNESE: Before I mention about operations, sneak in here my own
observations as we attempt to find the ways to fine tune our modelling
and our predictions so that all our answers come out to be right. Is
that I look back at the other process activated sludge and everything,
I wonder if the RBC and a lot of the things today that were encouraged
and innovated process does not._ Maybe would benefit by the ten states
Standard mentality that I think all of us fought against, most of us
I think of being the generation that wanted rationality and so on, but
when you think back, I think of the activated sludge process has devel- ,
oped in the same genesis as this does, that they would have the same
problems. But they had the benefits in a way as I look back now and I
just sort of here scratching my head, is that some wise people just
said there would be six hours aeration. We would put oxygen in and/or
air in the terms of so many cubic feet per pound or per cubic feet of
space and so on, and I guess what I am saying basically we got to where
the activated sludge still has many of its problems but not problems I
think. I think the things we are trying to do by just over-design and
the RBC and to its credits, and I guess to the credits of the profession
today and our very cost effective mentality that we are working under.
We will fix up the design and says positively this is. the thing that
will work, and we are frustrated. I hear argument back and forth about
things that I agree with but that is not going to do it. When we sit :
back I am sure the EPA would not like to, as they struggle to get every-
thing down to the so there is no waste, but I think as we look back we
have to keep that in mindo I think we have gotten to where we are with
lots of things, clarification, activated sludge, whatever by over-design.
I never see a plant, RBC or otherwise, that when it is half-loaded it
works x
-------
would argue that there is no need to speed up the operation of a bio-
logical system by the continuation of object because he has no bulking
problem, no solids flux problem or no DO problem. Today again, the ad-
vocate of more operating control of RBC would counter quickly with
examples to the contrary. It seems to me that the dividing line in
the argument relates to the adequacy of the manufacturers' designs, the
manufacturer, the design and the application. The past record is cer- ;.
tainly not too commendable in this regard. The problems of under-design
and facility malfunction are there in sufficient number to certainly
provide a sound argument for the needs of more operating flexibility to
offset an obviously limited design state-o"f-the-art for RBC. On the
other hand maybe we are over a mountain and have learned our past mistakes
and have now reached a point of better predictive design capability and
can finally offer the RBC as a handoff operating process. In soliciting
your views on this dilemma let me first provide my shopping list of
possible operating controls which could be divided on RBC systems. Of
course, one would be a course of variable speed rotating controls.
Secondly the device to measure the rate of biomass on the RBC to both
process controls and to protect structural stability, integrity. There
is the possibility of step feed arrangement, and we maybe could fit
into that recirculation by step feed if it is an easier design approach.
Stage control which the problem of changing stages to fit a given condition
and from my own personal experience have had to get down and remove my
filaments by pulling out a piece of wood or a piece of baffle, and I
suspect we could probably design something that could facilitate the change
of staging by something simpler than what we have had to do. Dissolved
oxygen and temperature monitoring, temperature in both directions which
are concerns in industrial, sometimes being too high and nitrogen problem
being low. Heat addition impossibilities, we have seen in once case an
industrial application of the need to actually control temperature from
getting too low. And, standby supplemental aeration options, I hesitate
to mention that after being in the earlier session, but I think we would
have to say that that is certainly an option possibly that could help
control operations. I am sure that many of you could from personal ex-
periences add other operating controls, would take away some of the ones
I have mentioned, which might prove useful to a particular problem you
have encountered. With that type of perspective, I invite your comments
as to the direction which should be taken to the providing of operating
control options in the design and manufacturing of RBC.
MR. VAUGHN: I have heard in the last two days a litter discussion concern-
ing the micro—organisms which grow on the media, and I have gained the im-
pression that it is quite important to observe the nature of the organisms
which are growing, yet I find that most manufacturers are advocating the
use of a quonset hut cover which at best seems to provide only access to
the bearings on these units. I would like to suggest that research cer-
tainly should be done concerning the type of cover and the type of housing
installations provided, especially when we look at scaling-up problems,
because most of the research work is done on units which are more fully
housed than what we find in the field, and certainly would invite some
comment from the equipment manufacturers on that. Well, certainly on the
effect that that cover has on the environment in which that unit is oper-
ating, termperature of winter and summer, and from an operator's standpoint
if it is important for the operator, which I assume it should be, since
it is important for the researcher, to observe the film that is occurring
1449
-------
on these un-'ts. Then I ask the question how can you do it with the present
housing? How do you expect that operator to even be able to see what is
happening on his RBC when his access to the unit is only at the bearing
ends, and this to me for an operator control would appear to be an impor-
tant feature.
MR. HYNEK: I will answer that question directly. I will start it off
from a practical standpoint which has been realized by Region Seven of '.
the EPA where where a lot of the RBCs started off with buildings, at
a point in time when covers were not available. And at Region Seven
you have the State of Iowa for example, cold temperature in winter time,
and the operators they have to grease the bearings, they like the idea
of being able to walk around between the shafts and so forth. While
Region Seven no longer advocates buildings, high humidity, corrosion,
coupled with OSHA requirements, lighting requirements, heating require-
ments and the bottom line was heating. One plant spent a thousand dol-
lars a month to keep that plant at the near normal temperature. With
respect to the cover in terms of access for an operator, let me go back
to what I feel are the basic tools an operator has to know if his plant
is operating correctly; eye, ears, nose and smell, he can look inside the
door and that biology on the first face of the reactor is a reflection ;
of biology that is occurring in the reactor not exactly the same by any
means; there is a different hydraulic profile, different hydraulic
exposure, but it does reflect changes just as well as what is going on
inside. So, as eyes can tell day to day changes, and this has been
common experience. His ears, if he does not hear any noise, he knows
something has stopped because the drive motor on the mechanical system is
there0 On the air drive system there is a kissing sound from the water
moving with the bubbles and so forth. He knows, just like that, he
knows there is something different if he is there day by day. So this
is the basic virtue of the RBC process, from an operator's standpoint.
His own senses can tell him very quickly if something is wrong. With
respect to the atmosphere within the covers, there has been detailed
technical data developed early on, is there enough oxygen diffusion in
and out. There has never been a proven case of a lack of oxygen. The
closest in a quantitative term is possibly five percent differential,
and I challenge anyone in the audience to do it any better than it has
been done0 For building however, we have had documented cases where the
oxygen content in the building because they hook the light in the building
with the fan0 When this light switch went off I think, and the operator
went home, the fan went off. Ten percent of oxygen in the building; you
couldn't light a. cigaret, you could not maintain normal feelings, that
you were in a good oxygen environment. Performance however, not measur-
ably different. I hope I have answered some of your questions. And you
can take samples from the radius of the media, look at it under a micro-
scope, and see the microflora and monitor that on a daily basis, if you ,
really want to get into the fine art of seeing the subtle difference.
DR. FRIEDMAN: Could I ask a question. That is fine, Bob, all these
signs of what we do, but the question really is suppose you find something
wrong then what do you do?
1450
-------
MR. HYNEK: It is a good question, and it has been a common question.
Well, if there is something wrong in the sense that he senses that
there is something wrong, we have' found that many basic unit processes
either up front or behind have been a factor. We have several manufac-
turers present that I think we can liken to pioneers because we are
early in this game, and so you try to answer these questions. We
have found from extensive field evaluations of all equipment, full scale,
that there are problems up front with respect to solids removal. There
are problems with solids removal after in the secondary clarifier because
the operator does not want to pump properly, he does not have really
simple automatic equipment to put her on a timer basis. The digesters
cannot get the anaerobic digesters up to speed: they have waited for
extended months of time, they might have local problems. One recent
investigation ocean outfall, when the tide comes in it backs and floats
the head works, and they have to elevate the waste water to make sure
it will go through the plant, so there really are a lot of serious
problems outside this new black box, and I think we are getting hit
from all corners, from consulting people who are sensitive to their
customer, from our sales representatives who are sensitive to the con-
sultant, and the customer, they tend to panic. I think they have got
to slow down and cool down and realize that you can expect this panacea
concept which perhaps the early-on RBC created, and well we are now
dealing with biological phenomena, bug is bug is bug. It can do so
much, but mechanically these other systems can impact on the biology,
and they have to recognize that.
DR. SMITH: The Army has been recently mandated to cut back their enery
consumption by twenty percent by 1985 or so, and in lieu of the fact
that there does exist a common problem inherent to RBCs cold weather
winter operations and due to the fact that we are trying to push solar
heating, solar collection, etc., I was just wondering if there has been
any activity in trying concretive efforts to make the covers actual
solar collectors. I am not saying make them clear because you would
have algae growth of course, but maybe some sort of mechanism. Paint-
ing them black would be simple except during the summer they may over-
heat. I was just wondering whether that would be an obvious and
simple way to solve the energy problem, feeding, and also cold weather
problems. Somebody have an answer to that?
DR. FRIEDMAN: I think the main reason for utilization of RBCs in the
country is the energy utilization and the reduction of energy consumption.
The EPA was wise enough to fund the energy plant in Maine which incorpor-
ates some of the panels over the discs, the first totally energy free
plant in the United States, using anaerobic digested gas to heat the
building and deriving its energy source from the solar panels. At this
particular time we have discussed many things with regard to RBCs in
operational problems but I think the topic of discussion of why RBCs
are being used are basically its simplicity of oepration, its guarantee
of a flexibility should be incorporated in the system, and the capability
to remove BOD at an extremely low horsepower. We are discussing a lot
of problems but I guess maybe one of the topics of discussion should be
optimization of total energy uses in biological waste treatment, and I
Ehink that the statement that was made regarding to power, consideration
1451
-------
of the Wilton.Maine Plant potential future studies with regard to energy
utilization using RBCs should be considered.
DR. YU: I have a question to be directed to the manufacturer. I do
not know whether the design from the different manufacturers have a
standardization in terms of all of their equipments or not. Based
on the past experience, some of the big companies due to business
reasons, want to shut down certain divisions. So when you shut down
the divisions field service is no longer available to what is exactly
the duration of a biological disc, so in case you need it and it
happened to be that division of the company that was shut down, where
do you get it? And, you know, this has more interest than those
kinetic constants. When you need it, K does not mean anything.
DR. BORCHARDT: That sounded like a statement and not a question.
DR. YU: Well, I say the question is directed to the manufacturers,
whether they have a standardization or not; when you get it from the
A Company, if the A Company shuts down, you get it from the B, and
you do not expect property owners have to take care of their own
problems.
MR. ANTONIE: With the standard RBC components we originally designed
and were marketing in '72 and since, have sort of become the standard
at least from the standpoint of shaft length and tank and so forth.
And I think you would find that any plant that has been built since
1972, you will be able to find two or three RBCs that could fit
that structure.
DR. BORCHARDT: That is, if you go out of business, as I understand
his question.
MR. ANTONIE: Incidentally, people that have gone out of the wastewater
treatment business did not do so because of competition, they did so
because there is no money being spent.
MR. HARRIS: I asked the question earlier which was not answered> so let
me rephrase it. Is there a fundamental operating advantage ,of the RBC
over the trickling filter?
DR. BORCHARDT: Well, we heard your question the first time, and there
were a couple of attempts. We have not guaranteed an answer to every
question, you understand that.
MR. HARRIS: Yes, I understand. If the answer is no, then I would point
out that there was in the past couple of days one paper at least given
comparing the energy requirements between trickling filters, RBC and
activated sludge. The RBC as I recall was somewhere between the trick-
ling filter and the activated sludge. Now, if there is no particular
cost advantage for the RBC, and there is a definite energy advantage
for the trickling filter, why don't we have a first national conference
on trickling filters?
DR. FRIEDMAN: I think your reasonings for standard rock trickling
filters are no longer in vogue; it is terribly expensive to handling
1452
-------
rock and be sure it does not crush. There is also a difference in
space requirements for the two; precisely they are certainly space
limited. The RBC installation would be smaller.
MR. HARRIS: Yes,' the standard filter is on the rock but there is
another type that does not use rock.
DR. FRIEDMAN: That is a packed bed filter. You did not ask that.
That is much more expensive. Do you agree with that?
MR. HARRIS: I did not' have the basis for agreeing or disagreeing.
DR. BORCHARDT: Yes, I would have to say that if you use plastic media,
the trickling filter would be much more expensive.
DR. WU: There are two papers that will be presented at tomorrow's
session to talk about economic evaluation of all kinds of systems—
activated sludge, trickling filter and RBC. I think the authors may
answer your question better from their paper presentations.
DR. CHESTER: I want to get back into the operation of controls a little
bit. We had some.discussions this morning on supplemental air and ro-
tational speed, and I am not.quite certain what the exact effect of
each one is, or what each one does individually. I happen to think
they both probably do the same thing if you add air to increase rotation
of speed, increase the turbulence, increasing the net transfer. In
terms of any operational control in the most efficient manner, and I
am trying to question out the manufacturers at this stage, it would
seem that having supplemental air standby supply and mechanical drive,
is probably the most efficient way to go about operating the system
because you do not have to run that air to drive the system continuously
and you can only put it on and put it on when you feel it is absolutely
necessary from a profit control standpoint.
DR. BORCHARDT: Well now that I think is a statement that you want a
yes or no to.
DR. CHESNER: What do you feel about that?
DR. BORCHARDT: Well then what would you say, yes or no?
DR. CHESNER: There is a difference of opinion in terms of whether we
should increase or whether the operational control should be an increase
in rotational speed or should we have a natural system driving this biodisc
by itself and supplying it simultaneously?
DR. BORCHARDT: I have a feeling that up to this time the air drive system
has been used because it was thought that the air drive system has advan-
tages which would recommend it. I think there is some areas which the
positive mechanical drive might suit the situation better and I think this
is an engineering question which has got. to be answered as the problem
comes up, I am not sure that answers your question either, but I do feel
that is an engineering decision. This is not always so, one way or another,
you see, so yes or no is not the answer.
1453 ...
-------
DR. CHESNER: I think the real issue here is energy though.
DR. BORCHARDT: Well you can make energy one issue but you have to have the
thing run. It does not run when you are saving energy. That is not the
answer you said.
DR. CHESNER: The answers with the mechanical drive, you are probably
going to keep that system running. The decision is to add an operational
control but you want the operational control, whether that would supple-
ment the air. That seems to be the only control unit that you have.
Supplement air in increasing rotational speed and it is not an intention
that operating the system with mechanical drive and having supplemental air
drive as stand-by is the most energy efficient way of operating the system,
energy efficient way having some control over the system.
MR. LAGNESE: Jack, I do not disagree with that. I think in the area we
are talking about, of what should we do with operational control, I think
it is a good question and I, the way I look at it is that we are going to
continue to try to pay for performance that we are not going to give more
than what we need. I think we have got to talk about operating control
like that. But I think the alternative to make it bigger and you do not
need stand-by supplemental air or fast procedures, but I think for some
reason we have gone into this thing that we are going to try to come as close
to it as we can. From what I have seen we had better then put in operating
control and I think I would buy your logic that we are talking about one
of the other and I think I agree with you that maybe the supplemental air
is the easy way to go.
DR. BORCHARDT: You mean you want to add both the mechanical drive and
the supplemental air in case you need it?
MR. LAGNESE: That is right. This is an operational stand-by for the
possible that we cannot pay for design or waste conditions to the exact
conditions. I think the alternative is to make the system bigger. But
that is not the way it is going.
DR. MOLOF: I would like to talk' about this idea of the dimples and all
that stuff. But relate that back to the manufacturers' getting together
to create a standard. And in the drinking water field the manufacturers
of home water treatment units formed an organization with a name called
"Wackysack". I hope you guys did better than that. And what they did
was they did not let EPA run it. A fellow from EPA was on the board,
they wrote to the National Sanitation Foundation and they remained a
private industrial group setting their own standards for quality. And
I think this might work very good with the group here. The fellow
in EPA was Frank Bell, the National Science Foundation with Dr. Nina
McCollum. You can find out and get together and do this thing but you
have a model that has been set up already, and they really set tough \
standards. I would like to talk about the dimple effect on area because ^
I have had the haunting nightmare that we are playing a numbers game with
area. And I think one of our problems is that engineer is going to
get involved with is relating pounds of BOD and I know you do not like
that word, some people do not like it, but what is the pounds of BOD per
thousand square feet per day? What is the square feet on Autotrol, on
1454
-------
anyone's unit when there is a half inch of biological growth on it. How
do you explain it? This is the problem that the manufacturer is going
to have to answer because I do not think any of them can define what
the effective area under actual operation is. I do not expect an
answer but just accept it as a challenge for me to them, if they cannot
define the actual operating area. Now maybe that is why our design
engineers are in trouble. They are told of an effective area and they
design it and they overload it and they need supplemental air. I would
like to bring you up the fact that speakjLng on operation, really gave a
very good viewpoint to the fact that the operator has very little
control which means the design engineer has almost full control. You
poured it out correctly in an activated sludge plant but they have
enough valves that turn and buttons to push. The design engineer can
just give them to you and you can operate the plant. You are absolutely
right that this plant you lose that control and I always search for ways
that people can adjust to conditions that were unforseen. I think
that is a very good point and you ought to follow it up. But that
does not make that one comment to help Mr. Friedman. About 1975 I put
in a proposal with somebody to study what they think between 3-foot
and 12-foot and they thought I was nuts.
MR. MADDEN: I would like to comment on some of the things that Bob
Hynek poimfced out concerning the operations of difficulty. We have
found in the majority of our operating plants on line and discussing
the various operatives that many of the problems in the RBC system do
not originate in the RBC system. I would say upstream of the RBC in
terms of pre-treatment or advanced treatment in terms of clarification,
infiltration or whatever we consider as the system. And we have found
that upon looking at the problem that it is not something where the
design engineer took it to account perhaps of the addition of the
supernatant coming out of a centrifuge or something of that nature, or
a mis-design in terms of dumping this centrifuge supernatant directly
into an RBC system without passing '.through the primary clarification.
But when we looked at the problem and we found operating facilities and
why they have not been many, and they have not been terribly complicated,
we can in almost every case trace them to problems such as these older
problems and the like, all have their bases in either upstream or down-
stream treatment and really not in the RBC system. Concerning trickling
filter, it has been our experience, we have one of our people operate a
trickling filter and RBC plant, and the general comment has been that
the trickling filter is the rock filter that we have experienced has
not been able to achieve the degrees of treatment that contemplates
with the RBC biological contactor. And while there are others who may
differ with this viewpoint, the fact remains without high rate trickling
filter or high rate recycles, you do not get the degrees of treatment
that you do with the continued ones through RBC systems. I think that
might help out the gentleman from Shell Oil regarding his question about
RBC and trickling filters.
DR. BUNCH: I would like to make a plea for all the little villages and
hamlets of 10,000 or 15,000 people who have a $50,000 per year budget
to run the fire department, the ambulance, the water department, the
sewage department„ We stand here tonight and talk about your operator's
experience, and say we need more operators. I would hope that we can get
the RBC plants to the point that you do not have a round-the-clock operators.
1455
-------
Any time you have to have a round-the-clock operators, you are going to
triple your cost and labor, and labor is going to be one of the big items
in the future. And I would hope to think that the RBC system could cut
down on the number of operators. And the second thing as far as energy
being all that important, so important that the local scale, because
on a national scale the amount of electricity required or the energy
required to operate all the sewage plants in the United States, there
is a very few percentage of that total energy budget. In fact, it is
about...to operate a sewage plant would equalize leaving a 60-watt bulb
on 24 hours a day per capita.
MR. DIAPER: I have not got that much experience on RBC but I do have
experience at another location with a device employed for the other
purpose in wastewater treatment which we are pioneer, for the microstrainer.
And even through similar problems that you are going through with now
with the RBC's, and I agree with Joe about the need for conservation in
design and you have to have a safety factor because you cannot cope with
these unprecedented contingencies that occur in the sewage treatment
plant upstream. But these as with the microstrainers are fitting into
the variable speed drive. And first of all we had it manually operated
and then we alternated it and we controlled it by the head differential
across the strainer. Obviously you are not going to get head differential
across the RBC and maybe you will, but surely there is some parameter on
the RBC that you could use such as film techniques or load or viscosity
to control the speed of the motor automatically. I certainly think you
should give the operator that opportunity to control the speed manually
even if you can do it automatically. And the second thing also arising
from my experience with the microstrainer is after being pioneering for
several years, the standards will be developed for design and they will
be developed from a simple experience based on ten states standards.
And I wonder no one discusses it as this discussion has pointed out that
there would be a better control method, as it has been pointed out, it
would be a way controlling the surface area for giving flow if the 10
state standards would be included in the design section of the RBC.
DR. BORCHARDT: The British have a. large DO electrode that they use to
control submergence of surface aerators. Such a DO electrode might in
the first stage of the RBC be used, to be used to control speed which
would possibly help us with some of our low DO situations. But as you
heard from Autotrol regarding the use of variable speed drive the pri-
mary results are extremely expensive and so that is one of our problems.
MR. DIAPER: Well I think you are going to get towards the condition that
Joe is pointing out that you are going to have to consider the safety
factors in the design to get away from these operating problems.
DR. GRADY: I want to make one very simple comment. Of course some of you
may agree that I have made a lot of simple comments, but anyway, you now
talk about factors of safety. One way to handle uncertainty is what is
called a "minimax" design and that is you take the worse situation which
represents maximum size installation but then you optimize the cost of
that system in order to minimize the design for that particular circum-
stance. And I think right now given the uncertainties that we face on
all of that parameter value, but this is really the approach that we
should all be talking throughly about these systems; however, no one knows
1456
-------
how to put in good operational controls. We do not know what is going
to happen from now on.
MR. GERHART: In an example of an installation at a petrochemical complex
that faces that given problem of control variability and flexibility on
a plant that had not been built yet, you could not even have a pilot
plant study, let alone model or anything else, and they wanted a process
guarantee which, of course, it is absolutely not but what they said you
could offer us in the way of flexibility because we will not spend more
money than we have to. And the net result was that two-thirds of flexi-
ble system, one involving in an equipment with six shafts in a single
flow stream. Now they fit a parallel channel so that they could bypass
any third or step feed as they desired. This is a very simple method
of control, not a very costly method of control. Second thing they did
was to fit in the outflow in stages one and two. The DO flowed in stages
one and two separately would indicate to the operator that if it dropped
below, and I do not know where they set it, 1-1/2 two parts of water, but
certainly no less than one, that if he dropped below that level in any of
the two stages, then he should increase his RPM. But those things as we
know from the personal observation of the pilot large scale systems. But
the big thing was that it could load cells under the three end bearings on
the first floor shaft in the train. But if the weight of the total operat-
ing shaft exceeded some value and I do not know what value they set. If
we set five BOD average with 500 or 1,000 parts per million we thought ,at
least the work done by Belgium's study. Any way they put the load cells
into the free end bank of the first four shafts. If any one of them exceeded
a certain weight, it would then be indicative of a thick biomass film if
you will, that they should then speed up or change up the RPM, or one way
or the other I suppose. You know that the biofilm can be thickened down
by increasing RPM or shearing forces.
DR. FIREDMAN: Do you have estimated the additional cost of energy to the
overall facility?
MR. GERHART: The price that I quoted was about ten percent. That is what
they told me. The engineer who designed the systems provided this infor-
mation but it is not definite.
DR. FRIEDMAN: That seems like a very marginal price to pay for that. But
I would like to point out we are really not talking about factors of
safety we are talking about what a Professor of mine used to call factors
of ignorance, and that is what we are really addressing.
MR. GERHART: Well when you talk about a factor of plus or minus fifteen
percent, man that is the difference between a successfully operating plant
and a lousy plant. And in terms of the operator that discussed his problem
a little bit ago, if he has got the local enforcement agency on his back
because he is fifteen percent over his permit, he does not give a stick or
damn about the K-factor or anything else. He says what can I do about it,
and one little fellow in Iowa, who did not have our installation, he got
so damn desperate he went out and got some garden hose I understand, poked
some holes in the damn thing and tied up to the various flow systems. But
by gosh he got his thing up to make a difference.
MR. HYNEK: I concur with Bob Gerhart's remarks regarding the low cells.
We have done a lot of work with it. The technology is all here, it is
1457
-------
out there you can buy it. It can be done with simply a mechanical hydraulic
jack operation, or you can put it on the electronic -system and read it
off the panel. So that question has already been addressed and
taken care of. It is a little tough to sell it, because they do
not'realize the importance of it. One of the other things I would
like to mention that has not come up here, is that we have had
experience, I will not call it bitter, but recognize that the hy-
draulic distribution, when you have a central distribution point
and you have several tanks through which the wastewater has to flow;
there has been evidence of poor engineering in terms of distribution
of the flow. And you will overload one bank of shaft and unload
another, and if particularly you are into nitrogen and you have a
poor operating plant, potentially„ The other thing is, we have main-
tained the program with staying in contact with the operators, and
we are hearing all stories from the standpoint that the little
localities will not pay the buck. So the guy that will run the
plant correctly, puts some dedication into; time after time we have
cases where we, in fact there is one on the record where the State
Department of Environment Control actually took the man to court for
falsifying the records. He was not qualified to run an anaerobic
digester and he was fudging all the data to make the plant look good.
Another case, one of our plants ran four months with no attention:
as testimony, there was never any snow tracks or shovel marks through
the shafts for four months, no lubrication. He was fired. So we
have had problems at all levels, it is .not just the RBC.
DR. BORCHARDT: The last subject, Mr. Opatken is going to talk about
research support funds. I know you are all dying to hear it.
MR. OPATKEN & MR. BASILICO: We started "out with a sub-chairman section
and a couple of committee members and we met yesterday last evening.
And the first item on the agenda was my committee members; how I
wanted them to get to the committee, that suggests me immediately changed
the subject. So I am left here by myself. When
etc., etc....we regret to inform you that...this is not standard. It
is not a form letter but it does occur. But there is a reason why it
occurs. People submit a good research proposal, and the proposal will
be on a. subject which may be reaching the top of a program plateau. It
is an on-going program; a program that has been developed, objectives
have been defined and there is movement towards reaching those objectives.
You come in with a proposal. It just has a very poor chance at that par-
ticular time of being funded. You have filed at a time with high visi-
bility. But what you have to begin with is the fact that what you want
is a program that is on the rise, a program that will become visible,
then your chances are much better. I offer that just as advice because
that is about the way the game is played. The pie is cut up and it is
cut up on the basis of priorities. I will say that if you have a project
that you feel would offer a significant improvement, it certainly would
again be considered, but I will not define significant,, Now what if you
do have ideas set forth. The best method is to contact the Wastewater
Research Division. Get in touch with the person who is involved with
the particular area that you are interested in. You make that contact.
It is nice then to discuss wljat you have to offer, the data that you
have and we like to then set up'1 a presentation by you to a group of peers
1458
-------
for evaluation at that particular time. Then a project is then evaluated'
in an informal manner and it is then encouraged or discouraged without
going through the formality of a full-plan proposal. That is one of
the fine methods. Another method is to submit your pre-proposal. But
again:.! say, you know what you are working on is the program priority
level. Well, the Office of Research Development in EPA has recently
come out with an exploratory research program for university professors
who are interested in research. Basically that kind of work is going
to be the peer review of proposals, and the reviews would be based on
certain criteria. And the Ad Hoc reviewers will be selected from a
group of people that have been nominated. Some have been self-nominated,
some are still being nominated. They are the experts in control
technology, industrial, municipal areas, and so forth. The mechanism
for funding has been organized and the information to be announced by
some EPA research pamphlets and flyers. I have got a draft right now
it lets you know how the system works and the type of reviews and the
type of various applications that we process and so forth. The funding
actually have taken money out of one pocket and put it in the other to
pay for this new exploratory research program. The thing that got
this research program initiated is that during past four or five years
our program has been criticized for lack of exploratory type of research.
Somebody said the EPA was only interested in applied research so that
we really did not apply our program to funding to University type research.
The EPA administrative officers now have decided to take the bull by the
horns and apply the money from research project to set up a research
grant program that does give the universities a chance to come in with
some wider range type of proposals that are related to various control
technology we are interested in today, and I think that gets us to the
point that you.may ask me how these priorities get set and so forth. I
do not know where I think what industry does, you do, or you try, and
as I would like to say to you I would like to get as a program manager
for research. I am worried about the fiscal eight-two and so forth.
since we have to run the program that addresses the immediate
priorities that were set two years ago and I guess what I would like to
see is some kind of perceptible proposal relation to research needs at
least come out of this symposium. Maybe somehow the workshop chairman
can summarize the results of discussion in terms of research needs, to
get some good research proposals from this group and eventually as a
result of panel review and so forth. We have to make a decision for
what subject should be funded; fund a project that is going to emphasize
a bigger or better model. My impression is, we have got it from the
industries and from the university, and so forth. From the proposals
we should know what are the real needs, what are the priority needs
and they could help us in the formulation of a very useful program for
wastewater treatment. In fact, we need some more research and as I
s'ay it is. supposed to be made on RBCs. However, we first have to define
the objective: why do you need to do that. It takes all kinds of stuff
to set up a program just because of .the kinds of procedures required. I
think it has got to come from outside and needs your input. That is the
crisis of today. If you have questions, I will be glad to answer,
DRo BORCHARDT: I guess we are all from without, is that what yOu are
saying?
1459
-------
HR. OPATKEN: Well, I have to say so I guess. Our program direction
now aims at the operating programs, the construction grant programs,
the regional offices guaranteeing what the states, what the consulting
engineers, the priority program nails down the toxic pollutants
control technology and our research program is geared toward the
solution for toxis waste disposal. In fact, our construction program
now has two or three hundred RBC plants and you know they are having
problems and maybe they ought to look into you know, better control
parameters and criteria for design. So our research program is more
geared to obtain input from outside. Dr. Borchardt suggests to in-
stitute a research committee. For instance the research committee
on wastewater treatment research and it has been a problem in our
program from the regional people and hopefully we plan to have a
completed strategy from this committee. And we do plan to send the
proposal out for peer review. AS we would like to send it to ASCE
RBC Committee first because the Committee has already taken the
position to study the RBC research needs. I think that could go a
long way in helping the guy in the EPA research center.
DR. BORCHARDT: Well, you chaps all realize then that that suggestion
you made for research is now going into the computer list as a priority
project and we know the EPA will fund grants in the future. The group
did not realize they were involved in that sort of thing tonight or a
lot more of them would have gotten up and had a say what is happening.
MR. BASILLICO: No, I think the nature of this conference really has
brought up a lot of good discussion on research needs in these areas,
and most conferences do not end up like this I guess.
DR. BORCHARDT: Well, are there any comments now on this idea of research?
Anyone else have research funds that need to be spent?
MR. VESSIO: As far as my involvement with the biological contactor is
concerned, most of my experience and expertise has been with the appli-
cation of mechanical surface aerators in municipal industrial processes
and sitting in the back of the room I was listening to the various discus-
sions tonight and listening to the other people who were conducting this
particular meeting of today. I see a. very close relationship between the
development of the RBC and the history of the mechanical surface aerator
device in this country. As a matter of fact the RBCs are at that posi-
tion that the mechanical surface aerators were marketed in the early 1950's,
Yoomans Brothers which is now part of Clow Corporation were the first
company to market on a new scale mechanical surface aerators in this;
country and did a very admirable job with it. Mechanical surface aeration
device just like the RBC.'s in Europe and in fact they utilized in 1929,
but were never really introduced into this country until the early 1950's.
And the first aerators in this country were actually marketed on a license
from a person in England. And the expertise was infant just as it is
here with the RBC. And it was not until 1965 and by this time there
were two or three other manufacturers of surface aerators who got into
the scene just like you have here with the RBCs; you have some initial
works done by Allis Chalmers, a lot of works done by Autotrol; there
were companies like in Envirodisc that came on the scene, TAIT bioshaft,
1460
-------
Walker process and so forth, each one making different claims, or claiming
some feature of superiority of their particular product. In 1965 at
the TAPII Convention, Yeomans Brothers developed a standard for testing
mechanical surface aerators and many equipment manufacturers utilized that
method. That particular procedure is now a part of standard method.
It is also written up in the manual published by the water pollution
control federation. The whole point for what I am saying here is that
why can't there be a standard for determining efficiency for the various
RBC designs extensively used today? An interesting paper today gave
the right terminology which reference was made to closed pack media
versus open pack media, which caused some lively discussion and I think
that this is extremely important. Because the little bit of exposure
that I have had with RBCs indicates to me that it is going to be as
it is right now I think, or it is certainly training to be, the equip-
ment of choice because it has several advantages of operational and
power-wise and so forth, from any of the .processes, that have been used
in the past and are even being tested today. We have found some of the
wastewater treatment processes which are not really economical, I
would suggest to the gentleman from the EPA that the true consideration
be given to testing for efficacy the various designs extent and being
offered today by all the manufacturers, whether dimple flat sheets,
jelly rolls, whatever. That these units be tested side by side on
similar type of waste and conduct an independent determination as to
what designs seem to have the optimal efficiency.
DR. BORCHARDT: Very fine. I gather from what he said that there is
hope for the future, that there are problems to be solved, and hopefully
there will be resources to solve the problems. Mr. Opatken says yes,
there will be. Well, we are now a half hour overtime followers and
unless there is some violent objections I am going to call the meeting
to an end.
1461
-------
-------
APPENDIX A
List of Participants
1463
-------
LIST OF PARTICIPANTS AND CONFERENCE ASSISTANTS
ABRAMS, Joel I.
Civil Engr. Dept. Chairman
University of Pittsburgh
Pittsburgh, PA 15261
AKE, James R.
Dr. of Facilities Engr.
DFAE Sewage Plant
Fort Bragg, NC 28307
ALEXANDER, Charles J.
Clow Corporation
PO Box 68
Florence,, KY 41042
ALI, Farrukh
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA15261
ALLEN, James R.
Canton Borough Authority
PO Box 237
Canton, PA 17724
ALLISON, Russell J.
Gulf Research & Development
PO Box 2038
Pittsburgh, PA 15230
ANDERSON, Bruce
Dr. of Facilities Engr.
DFAE Sewage Plant
Fort Bragg, NC 28307
ANDERSON, Douglas
FMC Corporation
1800 FMC Drive West
Itasca, IL 60143
ANTONIE, Ronald L.
Autotrol Corporation
5855 North Glen Park Rd.
Milwaukee, WI 53209
ASNER, Jeffrey
Md. State Envir. Health Admin.
201 W. Preston St.
Baltimore, MD 21201
ATHAVALEY, Arun S.
Subsurface Disposal Corp.
5555 West Loop South
Bellaire, TX 77401
BACHTEAL, Robert M.
McNamee, Porter & Seeley
2223 Packard Rd.
Ann Arbor, MI 48104
BALANCE, J.
Perma Engineering Sales Ltd.
Box 12, Group 200, R.R.#2
Winnipeg, Manitoba R3C2E6
BANERJI, S.K.
Dept. of Civil Engr.
University of Missouri
Columbia, MO 65201
BAO, Charles
Transviron, Inc.
1624 York Rd.
Lutherville, MD 21093
BARRY, W.F.
Owens, Ayers & Associates,Inc.
1300 W. Clairemont Ave.
Eau Claire, WI 54701
BASILICO, James V.
US-EPA
Research & Development
Washington, DC 20460
BATE, Robert A.
Lakeland Engineers,Inc.
6701 Seybold Rd.
Madison, WI 53719
BAYNARD, Albert
Dept. of Public Works
100 New Churchmans Rd.
New Castle, DE 19720
BECKMAN, William R.
The Munters Corporation
PO Box 6428
Fort Myers, FL 33901
BEIMERS, Charles J.
Williams & Works, Inc.
611 Cascade West Parkway
Grand Rapids, MI 49506
BEISEL, Kinney E.
Bluefield Sanitary Bd.
PO Box 998
Bluefield, WV 24701
1464
-------
BELSCHNER, Dale Lee
Washington Suburban Sanitary Comm.
434 jyiillshire Driye
fillersyille, JXlb 21108
BERGLUND, David E.
S.E.A. Consultants
54 Canal St.
Boston, MA 02114
BERGS, Mary A.
Davy Engineering Co.
115 South 6th St.
La Crosse, WI 54601 '
BERNER, Ralph
Williams & Works, Inc.
611 Cascade West Parkway
Grand Rapids, MI 49506
BERRINGER, Robert
Plum Boro Sewage Dept.
Nassau Drive
Pittsburgh, PA 15239
BISSELL, Paul K.
The Mack Company
Olde Courthouse Bldg.
7 Court St.
Canfield, OH 44406
BLANC, Frederick C.
Civil Engr. Department
Northeastern University
Boston, MA 02115
BONER, Marc C.
Stanley Consultants, Inc.
2600 Century Parkway N.E.
Atlanta, GA 30033
BONHOTE, Dominique J.
Autotrol Ltd.
Aeschenvorstadt 57B
CH-4051 Basel, Switzerland
BORCHART, J.A.
Dept. of Civil Engr.
University of Michigan
Ann Arbor, MI 48107
BORCHART, John
Gilbert Associates
PO Box 1498
Reading, PA 19603
BRACEWELL, Lloyd W.
Swanson-Oswald Associates
594 Howard St.
San Francisco, CA 94105
BRODERICK, Steve
Waldor Pump & Equip. Co.
9700 Humboldt Ave. S.
Minneapolis, MN 55431
BROOKS, Peter M.
Public Health Engr.
109 Governor St.
Richmond, VA 23219
BROWN, William E.
Wright-Pierce Architects/Engrs.
99 Main St.
Topsham, ME 04086
BRYANT, V.
Eastman Kodak Co.
901 Elm Grove Rd.
Rochester, NY 14650
BUCKLEY, David B.
Research Engr.
Tufts University
Medford, MA 02155
BUDJINSKI, W.
City of South Bay
PO Box 130
South Bay, FL 33493
BUNCH, Robert L.
US-EPA
26 West St. Clair
Cincinnati, OH 45268
BURLBAUGH, Alfred G.
Northwest Engineering,Inc.
3597 East State
Hermitage, PA 16146
BURRETT, Bill
Ray Lindsey Co.
PO Box 8124
Prairie Village, KS.66208
BURWINKEL, John W.
Busch Co.
4907 P.enn Ave.
Pittsburgh, PA 15224
CAMERON, W.L.
Swanson-Oswald Assoc.
594 Howard St.
San Francisco, CA 94105
CANADAY, James T.
Alexandria Sanitation Authority
Box 1205
Alexandria, VA 22313
1465
-------
CHARACKLIS, W,G,
Dept. of Civil Engr.
Engineering Mechanics
Montana State University
Bozeman, MT 59715
CHESNER, Warren H.
Roy P. Weston
1044 Northern Blvd.
Roslyn, NY 11576
CHIESA, Giovanni
Castagnetti S.P.A.
Via Fabbrichetta, 65
Grugliasco, Torino (Italy) 10095
CHOU, Chi-Su
Atitotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
CHRISTIAN, H.B. Jr.
Environmental Systems Div.
Geo. A. Hormel & Co.
11501 Yellowbrick Rd.
Coon Rapids, MN 55433
CHRISTY, Robert W.
Depollution Div.
Ralph B. Carter Co.
PO Box 214
Ocean City, NJ 08226
COAST, Morgan K.
Cerrone & Vaughn, Inc.
401 Main St.
Wheeling, WV 26003
COLL, James
Sverdrup & Parcel & Assoc.
801 North Eleventh
St. Louis, MO 63101
COLLIER, James R.
Utah Div. of Environmental Health
150 West North Temple
Salt Lake City, UT 84110
COLLINS, Anthony G.
Lehigh University
Bethlehem, PA 18015
CONNER, James M.
Conner Water & Waste Equip.,Inc.
3295 Babcock Blvd.
Pittsburgh, PA 15237
CONNER, James M. Jr.
Conner Water & Waste Equip.,Inc.
3295 Babcock Blvd.
Pittsburgh, PA 15237
COSTELLO, Albert J.
Lower Lacka. Valley Sanit. Auth.
PO Box 67, Coxton Rd.
Duryea, PA 18642
COTTER, Richard J.
Pantech Engineers, Inc.
340 Liberty St.
Franklin, PA 16323
COULTER, Robert
C.M.S. Equipment Ltd.
5266 General Rd. Unit 12
Mississauga, Ontario L4W1Z7
COWEE, Jeffrey D.
Autotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
CRAWFORD, Paul M.
Gore & Storrie Ltd.
1670 Bayview Ave.
Toronto, Ontario M4G3C2
CREAGHEAD, Joseph H.
CLOW Corporation
56 Industrial Rd.
Florence, KY 41042
CROUCH, Gary S.
Anderson & Associates, Inc.
100 Ardmore St.
Blackburg, VA 24060
CUMMISKEY, R. Thomas
BESCO
PO Box 328
Doylestown, PA 18901
CUOMO, Frank A.
Ralph B. Carter Co.
192 Atlantic St.
Hackensack, NJ 07602
DABROWSKI, John A.
Illinois E.P.A.
2200 Churchill Rd.
Springfield, IL 62706
1466
-------
BANNER, Jim
Ray Lindsey Co.
PO Box 8124
Prairie Village,
KS 66208
DAVIDSON, Roger
FMC Corp. Environment Equip. Div.
1800 FMC Drive West
Itasca, IL 60143
DAVIE, Richard L.
Autotrol Corporation
5855 N: Glen Park Rd.
Milwaukee, WI 53209
DAVIS, C.R.
Walker Process
840 N. Russell Ave.
Aurora, IL 60506
DAVIS, Gary W.
Walker Process
840 N. Russell Ave.
Aurora, IL 60506
DECARLO, Dale A.
Burgess & Niple, Ltd.
5085 Reed Road
Columbus, OH 43220
DEE, William P.
Malcolm Pirnie, Inc.
6161 Busch Blvd.
Columbus, OH 43229
DEISS, Richard A.
Richard A. Deiss & Assoc.
RD #1, Alden St. Ext.
Meadville, PA 16335
DENNIS, Robert W.
Exxon Research & Engineering Co.
PO Box 101
Florham Park, NJ 07932
DEPOULI, William H.
Bowe, Walsh & Associates
1 Huntington Quadrangle
Melville, NY 11747
DESCHAMPS, Jean-Claude
Ministry of Public Health
148 Boulevard de la Resistance
1400-Nivelles, Belgium
DESHPANDE, Sharad
IIT Chicago, Illinois
157 W. Stevenson
Glendale Hts., IL 60137 .
DIAPER, Tony
Crane-Cochrane
Box 191
King of Prussia, PA 19406
DICKINSON, John P.
1728 Central Ave.
Fort Dodge, IA 50501
DIFRANCESCO, James V.
LTV, Fibercast Div.'
255 Parkway Dr.
Pittsburgh, PA 15228
DOMINIE, Kenneth
Dept. of Consumer Affairs & Env.
Elizabeth Avenue
St. Johns, Newfoundland, Canada
DOORENDOS, Butch
Dept. of Environmental Quality
Bldg. 900 East Grand
Des Moines, IA 53019
DUFF, E.Roy
Henry P. Thompson Co.
4866 Cooper Rd.
Cincinnati, OH 45242
DUFFERT, Charles M.
Geo. A. Hormel & Co.
11501 Yellowbrick Rd.
Coon Rapids, MN 55433
DUHAMEL, Young
Bendlin-Duhamel Assoc., Inc.
94 Valley Rd.
Montclair, NJ 07042
DUPONT, Robert R.
Kansas University
1741 W. 19th Apt. 7B
Lawrence, KA 66044
DURANCEAU, Vern
Walker Process
840 N. Russell Ave.
Aurora, IL 60506
1467
-------
DUST, John
178 Pryor
City of Atlanta
Atlanta, GA 30303
ELLIS, David
Malcolm Pirnie, Inc.
6161 Busch Blvd.
Columbus, OH 43229
FARRINGTON, Paul
Buchart-Horn Consulting Engrs.
400 Market St.
Lewisburg, PA 17837
FEBBO, Louis
Lower Lacka. Valley Sanitary Auth.
PO Box 67, Coxton Rd.
Duryea, PA 18642
FEDERICO, John G.
Greeley & Hansen
1818 Market St.
Philadelphia, PA 19103
FEDOTOFF, Roy C.
Stearns & Wheler
10 Albany St.
Cazendovia, NY 13035
FETCH, John J.
Capital Controls Div.
Dart Industries
201 Advance Lane
Colmar, PA 18915
FISETTE, George R.
Ralph B. Carter Co.
192 Atlantic St.
Hackensack, NJ 07602
FITCH, Larry
Sanitation Engr.
Environmental Engr. Div.
State Office Bldg.
MOntpelier, VT 05602
FITZPATRICK, James W.
Industrial Representative
43 Black Rd.
Sault Ste Marie, Ontario P6A6J8
FLANN, Gary E.
ESD-Geo. A. Hormel & Co.
11501 Yellowbrick Rd.
Coon Rapids, MN 55433
FOLLWEILER, Jeffrey L.
Md. State Envir. Health Admin.
201 W. Preston St.
Baltimore, MD 21201 1468
FORBES, Eugene J.
Dept. of Environmental Protection
122 Washington St.
Hartford, CT 06115
FORGIE, David J.L.
University of Saskatchewan
Saskatoon, Sask. Canada
FORRELLI, James R.
Engr./Mgr. Morgantown Sanitary Bd.
389 Spruce St. •
Morgantown, WV 26505
FOSTER, Richard E. :
Williams-Russell & Assoc.
250 Piedmont Ave.
Atlanta GA 30308
FREEMIRE, Roy S.
Freemire & Assoc.,Inc.
Suite B-8 9150 Rumsey Rd.
Columbia, MD 21045
FRIEDMAN, A.A.
Dr. Engr., PE
Dept. of Civil Engr.
Syracuse University
147 Hinds Hall
Syracuse, NY 13210
FRYMIER, Manning H.
Cerrone & Vaughn, Inc.
401 Main St.
Wheeling, WV 26003
FYOCK, Timothy B.
Neilan Engrs.,Inc.
150 W. Union St.
Somerset, PA 15501
GAASCH, Jack F.
Wastewater Treatment Plant
Fredonia, NY 14063
GARG, Brij M.
Dept. of Envir. Resources
Commonwealth of PA
Harrisburg, PA 17120
GASS, Don
Autotrol Corporation
5205 Ironwood
Milwaukee, WI 53217
GERHARD, Robert E.
ESD-Geo. A. Hormel & Co.
11501 Yellowbrick Rd.
Coon Rapids, MN 55433
-------
GILLESPIE, James P.
E-Systems Inc. ETAG
7700 Arlington Blvd.
Falls Church, VA 22046
GOOD, Larry D.
SIELCO, Inc.
309 Washington St.
Columbus, IN 47201
GOSSETT, Richard G.
Autotrol Corporation
5855 N. Glen Park Rd.
Milwaukee, WI 53209
GOUTY, Otis D.
Kelly, Gidley, Blair & Wolfe,Inc.
1260 Greenbrier St.
Charleston, WV 25311
GRADY, C.P. Leslie Jr.
Purdue University
Civil Engr. School
West Lafayette, IN 47907
GRAHAME, Arthur W. Jr.
Burde Associates
PO Box 247
Paramus, NJ 07652
GRATZ, John
5703 Forbes Ave.
Pittsburgh, PA 15217
GREENE, Wayne C.
Hercules Inc.
Radford, VA 24141
GRIFFITH, Lynn H.
Glace & Glace Inc.
2771 Paxton St.
Harrisburg, PA 17111
GROVER, William A.
Dept. of Environ. Protection
Statehouse Station 17
Augusta, ME 04333
HAHN, Roger A.
Chief, Sanitation Branch
FED Bldg. 603
Fort Ritchie, MD 21719
HALLHAGEN, Anders
Mgr. Environmental Protection
Berol Kemi Ab
Stenungsund, Sweden S-444 01
HAMILTON, Harold J.
Dir. of Public Works
Havre de Grace, MD 21078
HANKES, Robert W.
Crane Company
800 Third Ave.
King of Prussia, PA 19406
HANLON, Pat
Rawdon Myers, Inc.
10814 Millington Ct.
Cincinnati, OH 45242
HANNA, David J.
Stearns & Wheler
10 Albany St.
Cazanovia, NY 13035
HANSEN, Nancy
Dept. of Civil Engr.
University of Pittsburgh
5703 Forbes Ave.
Pittsburgh, PA 15217
HARGENS, Dean A.
Shive-Hattery & Assoc.
PO Box 1050, Hwy. 1 & 1-80
Iowa City, IA 52244
HARRELLE, Lomax
City of South Bay
PO Box 130
South Bay, FL 33493
HARRIS, D.S.
Shell Oil Co.
PO Box 3105
Houston, TX 77001
HENNESSY, Thomas J.
Wash.-E.Wash. Joint Authority
62 E. Wheeling St.
Washington, PA 15301
HERKER, Amil Jr.
Clear Lake Sanitary Dist.
Box 282
Clear Lake, IA 50428
HERTSCH, Frank F.
City Attorney
Havre de Grace, MD 21078
HILL, A. Judson
ARCO Environmental Inc.
100 RIDC Plaza
1469 Pittsburgh, PA 15238
-------
HIMES, David L.
Deputy Dir. of Public Works
Havre de Grace, MD 21078
HINCHBERGER, James L.
Butler County Water & Sewage Dept.
130 High St.
Hamilton, OH 45012
HITDLEBAUGH, John A.
US Army Envir. Hygiene Agency
Aberdeen Proving Ground, MD 21010
HOAG, George
Dept. of Civil Engr.
University of Connecticut
Storrs, CT 06268
HOEPLE, Ronald A.
Walter E. Deuchler Assoc.
230 S. Woodlawn
Aurora, IL 60506
HONAKER, Robert T.
Bluefield Sanitary Bd.
PO Box 998
Bluefield, WV 24701
HOUSTON, James D.
New Castle County Dept. of
Public Works
100 New Churchmans Rd.
New Castle, DE 19720
HOVEY, Wendell H.
Dept. of Civil Engr.
University of Connecticut
Storrs, CT 06268
HSI, Eugene Y.
Transviron, Inc.
1624 York Rd.
Lutherville, MD 21093
HSIEH, Hsin
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
HUANG, Ching-San
USAF Occupational & Envir.
Health Lab
USAF OEHL/ECW,Brooks APB
San Antonio, TX 78235
HUANG, Jiunn-Min
11 Cypress St.
Tenafly, NJ 07670
HUFF, Dewey D.
Parrott, Ely, & Hurt Consult. Engrs,
620 Euclid Ave.
Lexington, KY 40502
HUGHES, Glenn G.
Alden E. Stilson & Assoc.
170 N. High St.
Columbus, OH 43215
HULTBERG, Kermit J.
Oper.-Sewage Treatment Plant
200 E. 3rd St.
Jamestown, NY 14701
Hynek, Robert J.
Autotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
IANNONE, John J.
Roy F. Weston
1044 Northern Blvd.
Roslyn, NY 11576
IEMURA, Hiroshi
Nippon Autotrol K.K.
Shuwa Onarimon Bldg.
1-11 Shinbashi 6-Chome
Minato-Ku, Tokyo, Japan
ITO, Kazuo
Dept. of Urban & Sanitary Engr.
The University of Tokyo
Hongo, Bunkyo-ku
Tokyo, Japan 113
JAFFER, Sham
Greeley and Hansen
222 South Riverside Plaza
Chicago, IL 60606
JAIN, R.K.
Environmental Div.
US Army Construction Engr.
Research Lab
Champaign, IL 61820
JANK, Bruce E.
Wastewater Tech. Centre
Box 5050
Burlington, Ontario L7R4A6
1470
-------
JENKINS, David
Univ. of Calif., Berkeley
659 Davis Hall
Berkeley, CA 94720
JOHNSON, Carl
Alexander Potter Associates
One World Trade Center,Suite 2637
New York, NY. 10048
JOHNSON, William R.
Kelley, Gidley, Blair & Wolfe,Inc.
1260 Greenbrier St.
Charleston, WV 25311
JONES, Warren H.
Jones-MacCrea, Inc.
1625 Burnet Ave.
Syracuse, NY 13217
JOOST., Robert H.
Tait, Ine.:-
500 Webster St. PO Box 1045
Dayton, OH 45401
KEEL, James S.
Henry P. Thompson Co.
4866 Cooper Rd.
Cincinnati, OH 45242
KELLY, William R.
Duncan, Lagnese & Assoc., Inc.
3185 Babcock Blvd.
Pittsburgh, PA 15237
KELTON, Robert T.
TEE PAK
915 N. Michigan
Danville, IL 61832
KINNER, Nancy E.
Dept. of Civil Engr.
University of New Hampshire
Kingsbury Hall
Durham, NH 03824
KINZIE, Daniel M.
1102 Buckingham Ave.
Norfolk, VA 23508
KISWARDY, Paul S.
UkS.Steel Corporation
600 Grant St.
Pittsburgh, PA 15230
KITCHENS, Judith F.
Chief, Pollution Technology
Atlantic Research Corp.
Alexander, VA 22314
KNOWLES, Patrick R.
PCF Sales Corporation
2210 Koppers Bldg.
Pittsburgh, PA 15219
KNUDSEN, John R.
I. Kruger A/S
Njzfarregade 10
7500 Holstebro, Denmark
KOELSCH, Lester M.
Autotrol Corporation
5205 Ironwood
Milwaukee, WI 53217
KORMANIK, Richard A.
Envirex Inc.
1901 S. Prairie Ave.
Waukesha, WI 53186
.KESCHL, Dennis L. KOSHY, Akanod M.
ME Dept. of Environmental Protection Dept. of Environmental Quality
State House, State House Sta. 17 900 East Grand
Augusta, ME 04333 Des Moines, IA 53019
KHETTRY, Rajib K.
Ministry of the Environment
135 St. Clair Ave. W.
Toronto, Ontario M4V1P5
KROEKER, Edwin J.
Stanley Assoc. Engineering Ltd.
11748 Kingsway Ave.
Edmonton, Alberta T5GOX5
KIM, Byung Jon KRUTH, Lawrence F.
NYS Dept. Environmental Conservation Franklin Associates, Inc.
Two World Trade Center RD 5 Box 360
New York, NY 10047 Somerset, PA 15501
KINCANNON, D.F.
Dept. of Civil Engr.
Oklahoma State University
Stillwater, OA 74074
KRYPINSKI, Kenneth C.
U.S. Steel Research
125 Jamison Lane
1471 MOnroeville, PA 15146
-------
KUTCHER, Thomas J.
Oper. Consultant Services,Inc.
PO Box 41081
Cincinnati, OH 45241
LA GREGA, Michael
Dept. of Civil Engr.
Bucknell University
Lewisburg, PA 17837
LAMOTTA, E.J.
Dept. of Civil Engr.
Univ. of Massachusetts
Amherst, MA 01002
LAMPART, Joseph
Envirotech Corporation
4735 Campbells Run Rd.
Pittsburgh, PA 15205
LANG, Mark E.
Colorado State University
Engineering Research Ctr.
Fort Collins, CO 80523
LARSEN, Milton D.
Wilson & Company
PO Box 1648
Salina, KA 67401
LA RUE, John E.
Howard K. Bell, Inc.
354 Waller Ave. PO Box 546
Lexington, KY 40585
LA VIOLETTE, Sherman
Donohue & Assoc., Inc.
4738 N. 40th St.
Sheboygan, WI 53081
LEDFORD, E.
Butler County Water & Sewage Dept.
130 High St.
Hamilton, OH 45012
LEWIS, John C.
Stream Engineers, Inc.
5505 S.E. Milwaukie Ave.
Portland, OR 97202
LIEN, Daisy
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
LIGNELL, Hakan
Nova Corporation
Vallingbyvagen 208, Box 81
S-162 12 Vallingby, Sweden
LIM, Lam K.
US EPA
401 M St., SW
Washington, DC 20460
LINWOOD, Bill
Dr. of Facilities Engineering
DFAE Sewage Plant
Fort Bragg, NC 28307
LONG, David A.
Penn State University
212 Sackett
University Park, PA 16802
LOZANOFF, Martin
Philadelphia Water Dept.
3900 Richmond St.
Philadelphia, PA 19137
LUCE, William A.
Nielsen, Maxwell & Wangsgard
624 North - 300 West
Salt Lake City, UT 84103
LUND, Don E.
McNamee, Porter & Seeley
2223 Packard Rd.
Ann Arbor, MI 48104
LUNDBERG, Lee A.
Schneider Consulting Engrs.
98 Vanadium Rd.
Bridgeville, PA 15017
LUYTKIS;~ Geir I.
Autotrol Ltd.
Aescheuvorstadt 57B
CH-4051 Basel, Switzerland
MACDONALD, Francis A.
Mixing Equipment Co., Inc.
135 Mt. Read Blvd.
Rochester, NY 14611
MADDEN, Bill
Clow Corporation
20 Main St.
Beacon, NY 12508
1472
-------
MADDEN, James
Clow Corporation
20 Main St.
Beacon, NY 12508
MAEHLING, Kevin L.
ERC/LANCY
3725 N. Dunlap St.
St. Paul, MN 55112
MAHER, Peter
E.G. Jordan Co.
PO Box 7050 Downtown Sta.
Portland, ME 04101
MAHMUD, Zahid
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
MAHONY, F.R.
57 Water St.
Hingham, MA 02043
MAHONY, James
57 Water St.
Hingham, MA 02043
MANNING, Lester H.
Town of Hanover
239 Central Ave.
Silver Creek, NY 14136
MARCIL, Gaston B.
John Mennier Inc.
6290 Perinault
Montreal, Quebec H4K 1K5
MATSUO, Tomonori
Dept. of Urban & Sanitary Engrs.
The University of Tokyo
944 Kains Ave.
Albany, CA 94706
MCCANN, Kevin J.
Autotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
MCMILLAN, Bruce
Perma Engineered Sales Ltd.
Box 12, Group 200, R.R.#2
Winnipeg, Manitoba R3c2E6
MEANS, Dennis
N. Dennis Means P.E.
1575 State Rd.
Webster, NY 14580
1473
MIKAELS, Karl E.
WPCF thru Vattehhygien, Sweden
c/o AB Zander of Ingestrom
10223 Stockholm, 'Sweden
MILLER, Roy D.
USA Environmental Hygiene Agency
Building 4411
Fort Meade, MD 20755
MILNER, B.W.
Project Planning Consultants, Ltd.
3767 Howe Ave.
Halifax, N.S. Canada B3L4H9
MOAN, Armand Y.
Degremont S.A.
183 Av. du 18 Juin 1940
Rueil-Malmaison, France 92508
MODESITT, Don
Alexander Potter Associates
One World Trade Center, Suite 2637
New York, NY 10048
MOLOF, Alan H.
Dept. of Civil Engr.
Polytechnic Inst. of New York
333 Jay St.
Brooklyn, NY 11201
MORETTO, Tom M.
Ted C. Miller Associates,Inc.
2140 S. Ivanhoe
Denver, CO 80222
MORGAN, James M. Jr.
Dean of the Faculty
Virginia Military Institute
Lexington, VA 24450
MORLEY, Barry
US Army Engineering Div. (RGAR)
Rd. 601 South
Near Berryville, VA 22611
MUELLER, James A.
Hydroscience, Inc.
363 Old Hook Rd.
Westwood, NJ 07675
NARDONE, Joseph A.
Lower Lacka, Valley Sanitary Auth.
PO Box 67, Coxton Rd.
Duryea, PA 18642
NEUFELD, Ronald
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
-------
NEUWORTH, Mark
Dept. of Environmental Conserv.
50 Wolf Rd.
Albany, NY 12233
NICOLL, Harry
The J.P. Bergren Co.
Cleveland, OH 44118
NIKOLICH, Mark P.
Kappe Associates
201 Penn Center Blvd.
Pittsburgh, PA 15235
NOSS, Charles I.
US AMBRDL
Fort Detrick
Frederick, MD 21701
NUGENT, George J.
Autotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
NUNGESSER, Philip W.
City of Atlanta
Rm. 303, City Hall
Atlanta, GA 30303
OAKES, Larry A.
A.G. Dunbar Company, Ltd.
2745 Dutch Village Rd.
Halifax, N.S. Canada
O'BRIEN, James H.
Shell Development Co.
Westhollow Research Center
333 Highway Six South
Houston, TX 77082
OCHOA, Alfred A. Jr.
Dow Chemical
Bldg. A-1127
Freeport, TX 77541
O'CONNELL, John E.
Haley & Ward, Inc. Engrs.
25 Fox Road
Waltham, MA 02154
0DEGAARD, Hallvard
Assoc. Professor
University of Trondheim
7034 Trondheim/NTH, Norway
OLEM, Harvey
Tennessee Valley Authority
Chattanooga, TN 37401
1474
OPATKEN, E.J.
Wastewater Research Div.
US EPA
26 W. St. Clair St.
Cincinnati, OH 45268
ORWIN, Leonard W.
Metcalf & Eddy, Inc.
1029 Corporative Way
Palo Alto, CA 94303
O'SHAUGHNESSY, James C.
Dept. of Civil Engr.
Northeaster University
Boston, MA 02115
OULTON, Ralph
Edward C. Jordan Co.
562 Congress St.
Box 7050 Downtown Sta.
Portland, ME 04021
PACCHIONI, Joseph P.
Pantech Engineers, Inc.
340 Libert St.
Franklin, PA 16323
PACK, Michael R.
Williams-Russell & Assoc.,Inc.
250 Piedmont Ave.
Atlanta, GA 30308
PAGORIA, Philip
Dept. of Civil Engr.
Old Dominion University
Norfolk, VA 23508
PANO, Abraham
Environmental Engr. Div.
Utah State University
Logan, UT 84322
PARKER, Robert A.
Tait, Inc.
500 Webster St. PO Box 1045
Dayton, OH 45401
PATRICK, H. Wayne
TEE PAK, Inc.
PO,Box 11925
Columbia, SC 29211
PERRY, Wayne C.
C.E. Maguire Inc.
31 Canal St.
Providence, RI 02903
PHEBUS, Charles F.
City of Glenwood Springs
806 Cooper
Glenwood Sprinas, CO 81601
-------
PIERCE, Jeffrey
Schneider Conservation Engrs.
98 Vanadium Rd.
Bridgeville, PA 15017
PLACE, John P.
John P. Place, Inc.
PO Box 10926
Pittsburgh, PA 15236
PLACE, Mark
John P. Place, Inc.
PO Box 10926
Pittsburgh, PA 15236
POHLKOTTE, Robert H.
Pohlkotte Engineering Co.
331 Sudbury Lane
Ballwin, MO 63011
POON, Calvin P.
Dept. of Civil Engr.
University of Rhode Island
Kingston, RI 02881
PRICE, Roger L.
B.C.M., Inc.
1200 Centre City Tower
650 Smithfield St.
Pittsburgh, PA 15222
PROSKO, Melvin D.
Stanley Assoc. Engineering Ltd.
11748 Kingsway Ave.
Edmonton, Alberta, Canada T5GOX5
PROVENZANO, Anthony J.
Dept. of Civil Engr.
Old Dqminion University
Norfolk, VA 23508
PULVER, Bernard G.
Harnish & Lookup, Assoc.
615 Mason St.
Newark, NJ 14513
REYNOLDS,. James H.
Div. of Environmental Engineers
Utah State University
Logan, UT 84322
RICHIE, James R.
Environmental Systems Div.
Geo. A. Hormel & Co.
11501 Yellowbrick Rd.
Minneapolis, MN 55607
ROEBER, John
Clow Corporation
PO Box 68
Florence, KY 41042
ROSENBERRY, Richard M.
O'Brien & Gere Engrs,,
1304 Buckley Rd.
Syracuse, NY 13221
Inc.
ROTH, James
Butler County Water & Sewage Dept.
130 High St.
Hamilton, OH 45012
RUGGLES, Gordon C.
Hazen & Sawyer
PO Box 30428
Raleigh, NC 27612
RUSHBROOK, Edward L. Jr.
Anderson-Nichols, Co., Inc.
6 Loudon Rd.
Concord, NH 03301
RUTA, Herman J.
Allan Engineering Co., Inc.
4031 W. Kiehnau
Milwaukee, WI 53209
RYAN, James G.
Duncan, Lagnese & Assoc.
3185 Babcock Blvd.
Pittsburgh, PA 15237
RYAN, Richard M.
Autotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
SACK, Bill A.
Dept. of Civil Engr.
West Virginia University
Morgantown, WV 26505
SAMPSON, Robert C.
Wash.-E.Wash. Joint Authority
62 East Wheeling St.
Washington, PA 15301
Sanvido, John A.
City of Guelph, Waterworks &
Water Pollution Control
City Hall, 59 Garden St.
Guelph, Ontario N1H3A1
1475
-------
SAUNDERS, P.M.
Dept. of Civil Engr.
Georgia Institute of Technology
Atlanta, GA 30332
SCHAMBER, Arlen
Environmental Systems Div.
Geo. A. Hormel & Co.
11501 Yellowbrick Rd.
Coon Rapids, MN 55433
SCHEIBLE, Robert E.
Scheible & Associates
1926 Waukegan Rd.
Glenview, IL 60025
SCHGLZE, Rich
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
SCHWING, Thomas T.
US EPA
26 W. St. Clair St.
Cincinnati, OH 45268
SERPA, Charles E.
Anderson-Nichols & Co.,
150 Causeway St.
Boston, MA 02114
Inc.
SETTLES, Jim
Jefferson Cty. Public Schools
809 Quail St.
Lakewood, CO 80215
SEVERIN, Elaine F.
Dept. of Civil Engr.
University of Illinois
4140 Civil Engineering Bldg.
Urbana, IL 61801
SHAMITHS, Gregory
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
SHEMONSKI, Anthony T.
Whitman, Reguardt & Assoc.
Ill N. Charles St.
Baltimore, MD 21202
SHIRTZINGER, M.M.
M.M. Shirtzinger Assoc.
1550 Western Ave.
Chillicotte, OH 45601
SHRODE, Larry D.
Freemire & Assoc., Inc.
9150 Rumsey Rd.
Columbia, MD 21045
SIKORSKI, Ted
Clow Corporation
PO Box 68
Florence, KY 41042
SINCLAIR, R.D.
Venus Industries Ltd.
108 Vodni Ave.
Winnipeg, Manitoba
SMITH. Ed. D.
Environmental Div.
US Army Construction Engr.
Research Lab
Champaign, IL 68120
SMITH, Jesse L.
McGill & Smith
119 W. Main St.
Amelia, OH 45102
SMITH, Thomas G.
C. MS. Equipment Ltd.
5266 General Rd. Unit 12
Mississauga, Ontario L4W1Z7
SONGER, Thomas
40 Uni-Tec Inc.
1234 East College Ave.
State College, PA 16801
SPEARGAS, David T.
Freemire & Assoc., Inc.
9150 Rumsey Rd.
Columbia, MD 21045
SRINIVASARAGHAVAN, R.
Greeley and Hansen
222 South Riverside Plaza
Chicago, IL 60606
STAHLMAN, R.L.
Chief Oper.-Sewage Treatment Plant
200 E. 3rd St.
Jamestown, NY 14701
STEED, Leon
Duquesne Light Co.
435 Sixth Avenue
Pittsburgh, PA 15222
1476
-------
STEINER, C.G.
Aquamotion, Inc.
PO Box 23006
Minneapolis, MN 55423
STEPHENSON, Joe P.
Wastewater Technology Center
PO Box 5050
Burlington, Ontario L7R4A6
STEPNOWSKI, James J.
E-Systems, Inc. ETAG
7700 Arlington Blvd.
Falls Church, VA 22046
STEWART, Richard A.
NIRA Consulting Engrs.,Inc.
950 Fifth Ave.
Coraopolis, PA 15108
STOCKBRIDGE, Joseph
NYS Dept. of Environmental Cons.
50 Wolf Rd.
Albany, NY 12233
STOLCH, Klaus
Stanley Assoc. Engineering Ltd.
11748 Kingway Ave.
Edmonton, Alberta T5GOX5
STOVER, Enos L.
Metcalf & Eddy, Inc.
50 Staniford St.
Boston, MA 02114
STRANGE, D.
Rodale Press/Aquaculture Proj .
RD #1 Box 323
Kutztown, PA 19530
STRATAKIS, Nick
NIRA Consulting Engrs., Inc.
950 Fifth Ave.
Coraopolis, PA 15108
STRATTA, James M.
US Army-Penn State Univ.
1084 Saxton Dr.
State College, PA 16801
/
STUDE, C.T.
Consoer, Townsend & Assoc.
3710 Hampton Ave*
St. Louis, MO 63109
STURDEVANT, G.E.
J.R. McCrone Jr., Inc.
20 Ridgely Ave.
Annapolis, MD 20678
STURDEVANT, Harry F.
J.R. McCrone Jr., Inc.
20 Ridgely Ave.
Annapolis, MD 20678
SU, John T.
FMC Corporation
1185 Coleman Ave.
Santa Clara, CA 95052
SUGAR, J. William
Union Carbide Corp.
PO Box 44
Tonawanda, NY 14150
SULLIVAN, Richard A.
Autotrol Corporation
5855 N.Glen Park Rd.
Milwaukee, WI 53209
SUN , Paul T.
803 Centennial Dr.
Champaign, IL 61820
SUTTON, Paul M.
Dorr-Oliver Inc.
77 Havemeyer Lane
Stamford, CT 06904
SWAIN, Robert V.
Glace & Glace, Inc.
2771 Paxton St.
Harrisburg, PA 17111
TAMPLIN, Judy C.
Clermont County Water & Sewer Dist,
2275 Bauer Rd.
Batavia, OH 45103
TATE, William B.
E.I. du Pont de Nemours & Co.
4501 N. Access Rd.
Chattanooga, TN 37415
THISSEN, C.
Walker Process
840 N. Russell Ave.
Aurora, IL 60506
1477
-------
THOMPSON, Edwin E.
McGill S Smith
119 W. Main St.
Amelia, OH 45102
TOSCANO, L.
Clow Corporation
PO Box 68
Florence, KY 41042
TRIPPENSEE, P.W.
Trippensee & Co.
4906 U.S. 27 S.
Sebring, FL 33870
TUCKER, Ben J.
Hoosier Fiberglass
2011 S. 3rd St.
Terre Haute, IN 47802
UNZ, Richard F.
Pennsylvania State Univ.
212 Sackett Bldg.
University Park, PA 16802
USINOWICZ, Paul J.
Fritz Lab #13
Lehigh University
Bethlehem, PA 18015
VANDEVENNE, Louis
CEBEDEAU
2, rue Armand Stevart, B.4000
Liege, Belgium
VAN GORDER, Steven D.
Rodale Press/Aquaculture Proj.
50 Wolf Rd.
Albany, NY 12233
VAN SANTVOORD, Philip
NYS Dept. of Environmental Cons,
50 Wolf Rd.
Albany, NY 12233
VAUGHN, Donald R.
Cerrone & Vaughn, Inc.
401 Main St.
Wheeling, WV 26003
VESIO, Mike
Tait, Inc.
500 Webster St. PO Box 1045
Dayton, OH 45401
VIET, Hiep Trinh
Ministere de 1'Environnement
2360, Chemin Ste-Foy
Ste-Foy, Quebec
VITEK, Frank W.
Marketing Mgr. Cochrane Div.
Crane Company
800 Third Ave.
King of Prussia, PA 19406
VRABEL, Robert A.
Kappe Associates
559 S. Braddock Ave.
Pittsburgh, PA 15221
WACHTER, D.H.
J.P. Bergren Co.
1991 Lee Rd.
Cleveland Hts., OH 44118
WALASEK, James B.
US EPA
26 W. St. Clair St.
Cincinnati, OH 45268
WALL, Danny S.
Howard R. Green Company
PO Box 9009
Cedar Rapids, IA 52409
WARD, Roger C.
HNTB
6927 Mohawk Lane
Indianapolis, IN 46260
WARNEK, Ed
EPA
2200 Churchill Rd.
Springfield, IL 62706
WASHERMAN, Arthur L. Jr.
Environmental Products Assoc,
171 Murphy Rd.
Hartford, CT 06101
WATT, J.C.
Catalytic, Inc.
1500 Market St. CSW
Philadelphia, PA 19102
WEBER, Susan C.
Dept. of Natural Resources
Box 7921, 101 S.Webster St.
Madison, WI 53707
1478
-------
WETZEL, Edward D,
Fritz Engr. Lab-Bldg. #13
Lehigh University
Bethlehem, PA 18015
WILKENS, William H.
Camp Dresser & McKee Inc.
One World Trade Center
New York, NY 10048
WILLIAMS, Douglas E.
W.H. Klinger & Assoc.
617 Broadway
Quincy, IL 62301
WILLIAMS, M.L.
School of Engineering
University of Pittsburgh
Pittsburgh, PA 15261
WOLF, John S.
John P. Place, Inc.
PO Box 10926
Pittsburgh, PA 15236
WOLLMANN, Albert M.
CH2M Hill
1930 Newton Square
Reston, VA 22090
WONG, Kit Y.
US Army (ARRADCOM)
Dover, NJ 07801
WOOD, Jerry R.
Jack G. Raub Co.
125 Baker St.
Costa Mesa, CA 92626
WU, Yeun C.
Dept. of Civil Engr.
University of Pittsburgh
Pittsburgh, PA 15261
YAZICI, Muammer
CMS Equipment Ltd.
5266 General Rd., Unit 12
Mississauga, Ontario L4W1Z7
YEIGH, Larry E.
Napier-Reid Ltd.
10 Alden Rd. Unit 2
Markham, Ontario L3R1E2
YENDELL, Kevin E.
Union Carbide Corp.
Linde Div. PO Box 44
Tonawanda, NY 14150
YU, Ta-shon
MD. State Environmental Health Adm.
201 W. Preston St.
Baltimore, MD 21201
ZALOUM, Ronald
Environnement Canada
1550 Quest de Maisonneuve,4th PI.
Montreal, Quebec H3G1N2
ZENZ, David R.
Research Chemist
Metropolitan Sanitary Dist.
100 East Erie St.
Chicago, IL 60611
ZIMMERMAN, Robert R.
Perma Engineered Sales Ltd.
Box 12, Group 200, R.R. #2
Winnipeg, Manitoba R3C2E6
ZWIERZ, Ken
FMC Corporation
1800 FMD Drive West
Itasca, IL 60143
1479
-------
-------
APPENDIX B:
FLOOR DISCUSSIONS AFTER EACH SESSION
Session 1. GENERAL DISCUSSION
Presiding: J. I. Abrams
Department of Civil Engineering
University of Pittsburgh
MR. COSTELLO: The question was did we inspect the preconstruction sampling
survey for the fourteen facilities that indicated variations in wastewater
loading.
DR. CHESNER: Answer is we did not.
MR. DABROWSKI: The question was, was there existing on paper the location
of each of the individual facilities.
DR. CHESNER: I did not but if you request to me personally I will let you
know.
MR. CONNER: It is very dangerous to take energy data especially in terms
of energy per unit flow and just use it with blank checks. The reason is
you have to consider the efficiency and the performance of that individual
facility with the energy utilization. Some of the facilities we deal with
energy survey that there was such little data, available concerning energy
requirements especially considering the importance of energy these days.
DR. CHESNER: I would say that the large variations between RBC and trickling
filter even though we did not have a lot of data, tends to indicate that
these numbers were pretty accurate in terms of differences in energy that
RBC were less energy efficient than trickling filter but certainly more
energy efficient than activated sludge units. In our field testing we will
be getting more information on RBC units in terms to establish relationships
between flow and energy utilization and between efficiency and energy utili-
zation. However, we still have that gap in terms of trickling filter and
conventional activated sludge data.
MR. DABROWSKI: Your company has been asked by EPA to evaluate RBC wastewater
treatment process. What is your idea to proceed on this important assignment?
DR. CHESNER: I would say that depending on existing applications and 'sort
of, get a try to score that question. We would certainly listen to all man-
ufacturers and right now in the process of designing an RBC facility to treat
seepage which might be the first facility I think in the country that is
eventually doing that. We would certainly entertain each manufacturer and
listen to what they had to say and them come up with the best judgment.. I
do not know if. I answered your question satisfactorily.
1481
-------
MR. HERKER: The question was where design curve indicated a percent removal
of total BOD and what fraction of that total BOD was soluble and what fraction
was insoluble.
DR. CHESNER: There are different design methods; some utilize total BOD and
some utilize soluble BOD. That top method just utilizes total BOD. Unless
you can predict the soluble BOD and your system it really comes out...There
is not a great significance in using a total BOD or soluble BOD. You have
to be able to predict the soluble BOD and if you can you have probably been
dealing in a more accurate method.
DR. YU: From the data you have presented it is quite difficult to generalize
the plant performance as a function of the process operating conditions.
Can you comment on that2
DR. BANERJI: I really cannot from the data we have. I cannot draw a
conclusion but the data...we are not addressing to that question right
now but maybe eventually it will come that...But that does not give you
a clear picture because that is not the only consideration. But this is
just a sample of data we have but not more data available and you have
got to look at individual cases now, than make a general statement. But on
this BOD you cannot use it, so I do not think you can generalize it depending
on the waste you have got.
DR. YU: I have some difficulties in understanding of your data presentations
in one of the table showing the plant performance. Can you again discuss
that?
DR. CHESNER: There are a number of columns on that table; one that I did
not get into was effluent criteria problem and those were problems that the
plant, at least 50% of the time is not meeting its effluent criteria. The
other problems or some of the reported difficulties were associated with
instability in the system.
1482
-------
Session 2. PROCESS VARIABLES AND BIOFILM PROPERTIES
Presiding: R. Neufeld
Department of Civil Engineering
University of Pittsburgh
MR. ATHAVALEY: I am asking the question from the viewpoint of chemical
kinetics as opposed to mass transfer you have presented here. As we all
know the depletion of initial biological oxidation demand or chemical
oxygen demand is the final goal that quality should meet. Now considering
the equations that you have here, are there any means to correlate this
oxygen transfer to percentage reduction, say for instance BOD.
MR. KIM: No. This is solely physical factors in clean water test. There
must be some limits of this system which it can transfer oxygen into the
liquid film in the bio-reactor. So this does .not include in our study.
DR. FRIEDMAN: This disc is heavily loaded, there indeed is going to be
some sort of a correlation between the factors that he has presented.
For actual RBC treatment system I do not anticipate that you would see
any difference.
DR. MOLOF: In an activated sludge system everybody sells an aerator based
on pounds per horsepower'hour, as well as detention time, F/M, etc...
In the disc field all we ask is a square foot area. We do not ask the
question what is its aeration capacity. We are trying to introduce to
this meeting the concept of saying, not only how much square feet area
you have, what is the aeration capacity of your units on certain pounds
per horsepower 'hour. I believe we need that incorporated into understanding
these systems a little better.
MR. MOORE: In summary do I understand your relationship to say this, that
the oxygen transfer capabilities are reduced as the space between the discs
reduces, that is the smaller the space between disc media the less efficient
the area's capability. In addition, what you are saying on heavily loaded
systems, the thicker the film gets obviously the less efficient in terms of
later on in the aeration capability. You get the distance between disc
surfaces decreasing.
MR. KIM: The K£a is per unit volume and it is related to the volume of
the reactor. This is very simple system in which you have liquid and
air, and you pump water up in the air and bring it back. Again, this is
a very simple system which you pump up water in the air and saturate it
to some degree and bring it back. So the "F" is a function of reactor
volume.
MR. MOORE: You are saying that it does not relate the distance between
the discs but it is a function of the reactor volume.
MR. KIM: It is related to the distance of the disc. When I derived this
equation, the only factor left in this equation is distance between the
discs. But the above mentioned factor is somewhat related to the volume
of the reactor.
1483
-------
MR. BACHTEAL: Would you keep the mass flowing rates on COD or BOD.
I missed it when you went through it.
DR. ORWIN: In terms of BOD we are talking, three point six pounds per day
per thousand square feet and seven pounds per day per thousand square
feet. We tried to bracket some of the rules of thumb as Al Friedman says
are meant to be broken. We tried to at least as the first set of experi-
ments to look at common accepted values.
DR. SAUNDERS: As I understand, you ran sixteen experiments over a twelve
months time period. Does that mean each one was made over a three-week
time period?
DR. ORWIN: No, it does not. We initially, when we were in the experimental
program, we would start it up and let it run a week and then take maybe
the end of the second week we would start taking some measurements to evalu-
ate some of our parameters to see where we were. And we found generally
after two weeks we were at something we could call steady state, and we
felt very comfortable about getting steady state. As we progressed we
learned more about the system. We felt you know, we got to where we could
say okay, we are going to change parameters to this and this is what is
going to happen. We were still trying to show that objectively. We felt
we found we could shorten the runs to almost two weeks, a week coming to
what we called steady state and then essentially a week of experimentation.
The systems we were using responded quite quickly and quite stabily, to
changes.
DR. SUANDERS: You indicated in your initial slide, a rather dramatic
change in microbial population, at least the color of the mass in one of
your initial slides, between stages, would indicate a rather, at least,
the potential of the most dynamic changes taking place over an extended
time period with that unit. If we look at solids, like with activated
sludge systems and that sort of key values that we usually use in evaluat-
ing those systems, I do not think that we would consider a seven to fourteen
day time period long enough to turn a population over to get a stable
result from the system. I have done similar units in similar situations
in terms of nitrification with the units that did not nitrify and going to
the units that did nitrify, it took something like a month to get that
population in, you see the changes taking place. You see the changes in
the nitrite, nitrate ammonia balances happening but it is an extended time
period over which those change.
DR. ORWIN: We were seeing in that first stage the white gelatinous mass,
we were seeing almost 95 to 98% carbon removal right there. Our second
stage which was the light brown stage, seemed to be, I would say, a mixed,
that was where the transition was taking place. The mixed bag - we would
see sometimes predominant carbon removal and sometimes predominant ammonia,
sometimes nitrification taking place depending on our conditions. But
our system responded quite quickly. We were quite surprised. We were
monitoring pH in the different stages. We were monitoring alkalinity in
the different stages; we were monitoring ammonia and nitrogen, nitrates
and others. We were doing COD analyses and they came right in and settled
down. We were pleasantly surprised. It did not take a long time for tran-
sition. It may be a function of the type of feed we were feeding it, which
was not a very complicated feed, a glucose, and we were not feeding it
organic nitrogen; on might want to challenge us on that. Given what we
were feeding it, it responded very quickly.
1484
-------
DR. SAUNDERS: Did I understand you to say that you pH was eight point five?
DR. ORWIN: We adjusted second stage pH to eight point five. We wanted to
enhance nitrification.
DR. SAUNDERS: On what basis?
DR. ORWIN: Well, on the basis of the literature which was saying that
was pH what eight and above was desirable.
MR. BUCKLEY: How did you measure film thickness?
DR. MOLOF: The film thickness was measured with a contrast microscope
where we calibrated it and then just ground the depth, the focus, at a
different depth.
MR. BUCKLEY: Any way of evaluating at the film thickness?
DR. MOLOF: What film?
MR. BUCKLEY: Active film thickness. Is there any way of evaluating active
film thickness that you can think of?
DR. MOLOF: At this time the only thing that we could do, this was in the
very early seventies, was to do the volatile solids, which as you know,
does not really relate to the active. But what we feel is that film
thickness and age show where the film does not become active anymore. After
that Initial 300 microns, it does not appear that the other material is
doing the work.
MR. WATTS: I think I derived from what you said that in the operation of
the unit, that you did not really change this speed or adjust for to get
rid of the oxygen limitation. Is that right?
DR. MOLOF: We adjusted the disc speed from 10 to 40, but it is not in
this paper.
MR. WATTS: Well, I mean are some of the things we are seeing as the load
increases, systems below a certain level not being oxygen limited and systems
above a certain level oxygen limiting takes place?
DR. MOLOF: That is exactly right.
MR. SETTLES: I may have missed it; did you identify what your mass loading
rates were and how many COD per thousand square foot?
DRo MOLOF: Yes, I did, but I will be happy to repeat it. I said that the
three grams per day which was what we were feeding for each stage, was about
15 pounds a COD per day per thousand square feet.
DR. REYNOLDS: Identified the DO limiting sections, substrate sections,
did you measure dissolved oxygen and mixed liquor in the tanking sections
that were oxygen limited, and if so, what were those values, what value did
the dissolved oxygen have to reach before it was no longer oxygen limited?
1485
-------
DR. MOLOF: If the fellow who did the work was here he could answer your
question. I do not...this was done in 171-*72 and he did not want to
deliver it. So, I had to. We measured the DO but I do not remember the
exact value. Incidentally, Dr. Chesner who spoke this morning has published
data that shows a below about one point five to two. You will get an inhi-
bition in reduction with discs, but that was separate work.
MR. STRATTA: I had a couple of questions. First of all, with respect to
nitrifiers, did you try to enumerate any of the nitrifiers on your film?
DR. HOAG: No, we did not. We did not do any plating counts at all, and
we really were not able to use them with photo micrograph or microscopic
examination.
MR. STRATTA: And the second question I had, you made reference to having
some problems in breaking up your film with blending techniques. What did
you eventually use to break up your film so you could enumerate the organisms
you saw?
DR. HOAG: Well, by scraping part of the bacterial film off into a slide, we
could just separate it into various sections and on the slide itself, using
a very large cover glass we were enabled to expose various edges of the
sludge particles and there were some liquids associated with the biofilm
that was on the disc as it was turning. And so we did get, we were able to
examine the organisms that way.
MISS KINNER: What was the magnification of the plot on that shot you showed?
DR. HOAG: I think that was about 500 power.
MISS KINNER: Five hundred. And the other question I had for you is, you
talked a little bit about the role of filaments and their increase in the
second stage. What do you postulate is their, the reason for their dominance?
DR. HOAG: Well, I think the most obvious thing would be that they were prey-
ing on the bacteria that were oxidized in the carbon in the first stage.
DR. UNE: I would like to make a suggestion that you eliminate much of the
bacterial part of your work when you do your thesis. First of all, you do
not have many species and what you see in there are nitrifying bacteria N-
debria in flat form, but the debria is not a nitrifier. It is a specific
species of organism or genus organism and this is part of the problem that
engineers have in effect bolt down on the field over the many years, and,
the microbiology of waste treatment. And I would like not to see it con-
tinued in the case of the RBC. I could not really get a good look, at your
photo micrography but I am very suspicious that you do not have Sphaerotilus
or Nocardia in those pictures.
DR. HOAG: Well, as I warned, when I started, it is very difficult to
identify the bacteria without plating them. You really cannot be sure by
a microscopic examination.
DR. MOLOF: I wanted to ask you, did you notice any characteristics of
RBC sludge if you look into a microscope and say the system is operating
right?
1486
-------
DR. HOAG: I think I did, J;n especially in the first stage and second
The second stages when we were really loading the system at a high carbon
loading rate and a low hydraulic slow rate, there were really no other or-
ganisms other than the bacteria in the first stage, but when we decreased
the loading rate or increased the flow, we would find a few species affiliates.
MR. PERRY: When you took the samples, did you try taking samples at different
locations on the discs?
DR. HOAG: Yes, we did. The results that I was presenting here were all
from similar locations on all the discs. However, a few times we did
examine various sections of the biofilm. We found similar cultures but
there were slight decreases in the microfauna population, as you went
from the periphery of the discs to the center of the discs.
MR. PERRY: Did you try to quantitate the amounts .that you took? You said
you took it with a slide. Did you just reach in and scrape the biomass?
DR. HOAG: We really had a difficulty in, we were able to quantify the
amounts that we took off the disc itself but since we really were not
enumerating them, the numbers per hundred ml we really did not need to
know what the amounts of the biofilm were rotating on.
1487
-------
-------
Session 3. MUNICIPAL WASTEWATER TREATMENT
Presiding: Edward D. Smith
U.S. Army Construction Engineering Research Laboratory
DR. BANERJI: I understand that there was some problems in disc media.
MR. DUPONT: When I was there, there was one shaft that was out. The end
of one had sheared off at I believe, it was at the drive unit. But it
was laying sideways. All the other shafts in that path were turning so
that it would not get lopsided but that had happened once before and also
happened after I was there. So there is a definite problem. I think they
went back and rewound it and replaced the shaft. I do not believe they planned
on it because they had to tear the building roof off to go near it and replace
it. So they had major problems. I do not know if they are having any more,
but the data is dated. I wish I had more information but I am still going
to school. But that is it, they did have problems and it was structural
problems.
DR. SMITH: I am very sensitive to you but municipal plants; it is not only
the RBC plants activated sludge, and trickling filter whatever, often times
the operators only collect the data that they need to fill out their MPS permit
and they do not collect the data that they need to evaluate the performance
of their plant. And I think that is a very important consideration that the
plant operators; what we should all try to encourage them to start doing that.
Well the next paper stated that the water problem, water shortage by year
2000 is going to be, make the enery shortage look like a Sunday picnic. Now
I do not know if you agree to that. Certainly in certain places in the United
States that is true. Many islands have a lot of water problems and all of
this type of research is in its infancy. I think it is very clear that one
aspect of water conservation will be utilizing saline water for whatever,
and I think this is a very important research. Also I think it is very
interesting that we are using very sophisticated technology. For instance,
she mentioned the electron microscope to study RBCs. And I think we need
to continue in that area. And I know Dr. Poon who did the initial work on
saline wastewater and RBCs has a lot of questions, I hope. Just one time
I note that there is a lot of need for research on nitrifying bacteria and
saline research in case you are looking for a topic NSF or EPA,
DR. SACK: I remember, I listened in at the early part of the session. What
kind of salts did you add, could you tell me?
MISS KINNER: Yes, we added Utility Marine Mix artificial sea salt. We
originally had thought of adding straight sea water but found we were having
a problem with diluting out our BOD, diluting general organic and to avoid
that we mixed up a concentrated salt solution with these utility salts.
DR. SACK: We have used high salt concentrations with coal gasification
waste water and I wondered if some of the blobs on your screen, the dark
ones, could have been precipitation when some of your added salts hit some
phosphorous that you might also have added, or may not have added. Secondly,
we have seen those cysts in there that you labeled as undescribed cysts and
in the past I speculated they may actually be very stress protozoa or some-
thing of this nature that were getting ready to expire from the scene or
had expire, and wondered if you found them present only under highly saline
conditions and not under, not just the sewage.
1489
-------
MISS KINNER: We found them present in both fresh and salt water conditions
and Art Moore, I do not know if you are familiar with him, he is a proto-
zoologist at the University, and he examined those cysts and he does not
think they are stress for design. He just thinks that there is some kind of
an insisting stage probably not even of a protozoan. He is thoroughly
convinced that they are not protozoans at all.
DR. SACK: Does he think they are alive?
MISS KINNER: Yes, he thinks they are alive and we are not sure what forms
they are but we are 'pretty sure that they are alive. And they seem to be
present just randomly and that may be a function of the fact that we just
missed them. Perhaps we missed it, but this is not a quantitative analysis
and we have missed them. And the salinity that stresses as I say, they were
apparent in both saline and non-saline conditions. As to the blobs on some
of those pictures, I think also some react, it could be precipitate. Some
of them comes from the fact that we did have a lot of material adhering to
the disc that for the particular under certain conditions, and we think that
that could be non-living material. Yes definitely, possibly precipitate,
but we are pretty sure it is not living and in culturing we have been able
to kind of eliminate some of those factors.
DR. SACK: It has been reported many times that high salts such as from
industrial wastes cause effluence to become somewhat turbid. Did you notice
that? If you did I did not hear you say that.
MISS KINNER: We did not have a problem with turbidity. That may be because
the salt concentration that we were using were generally fairly low and we
did put the salinity up to about six percent. We did find an increase in
the turbidity of the waste and in fact the BOD was at that time, not up to
the 30 milligrams per liter standard but quickly adjusted to that.
DR. SMITH: Did you think it was relatively high turbidity index for saline?
MISS KINNER: Yes. We were actually...if I showed you a fresh water and a
saline sample, I will bet you could not even tell the difference.
DR. SMITH: I bet I could not either.
DR. POON: I have a lot of questions but I have only two general ones in salt
expectations. First one, when, in most of them of course I use extra sea
water and mix it with the sewage. And naturally it would have a lot of
marine algae and sewage organisms on the biofilm. The disc biofilm has a
lot of marine algae, some are vine colored, green color, it is very colorful.
I regret because of the short time involved, I do not have time to study the
nitrogen removal and phosphate removal. And I would recommend anyone who
would do this, we should look into that. Maybe there is additional nitrogen
and phosphate removals better than using fresh water. I do not know if you
have data on the quantitative basis at what concentrations they may be, I
should say the fluctuation of the concentration would have any effect on
the growth of various organisms because this is the only thing that I found
within the range from zero, I should not say zero but a hundred parts per
million up to about 12,000. If I ran the system long enough to the steady
state, there would be no effect at all. But if I let the concentration
fluctuate, the hydraulic loading will highly affect the BOD removal.
1490
-------
MISS KINNER: We do have some data and we did not find any inhibitory
effect with changing concentration, but of course these things were
brough up on changing concentrations since we changed the quantity. It
was constantly varying between 2 percent and 4 percent,, So we did not
acclimate them at all to a straight salinity as.part of the work. They
are constantly in a varying salinity. But we did not observe any in-
ability of the microorganism to treat the waste under those variations
as I explained in the last question. We did see some limited ability
to meet on an initial period of acclimation and then we did after that
6 percent or 5 percent.
DR. POON:
after?
Where were these experiments of concentration we have looked
MISS KINNER: We did have percent solids at all on that, so we just
measured the influent and effluent salinity. That is what we were mon-
itoring and it was pretty much between 10 and 12 points per thousand.
DR. POON: And your soluble BOD's were around 15 in effluent?
MISS KINNER: Yes 10 to 15.
DR. POON: And you may or may not .meet 5 mg/.l effluent requirement in
BOD totals?
MISS KINNER: Well we did do totals as a matter of fact but I just did
not mention in here. They were all under the 30 milligrams per liter.
I just drew a purpose for this study. I would also say that besides the
bacteria we had, the protozo, we were pretty sure that those were not
in green form present. We did not get the .algae .at ..all. So this could
have been there in limited numbers. But anyway, we are using a salt
water waste. The higher organic concentration should knock off the
bacteria and this has also been seen by marine microbiologist Dr. Johnson
at University of New Hampshire.
MR. VESIO: At the point where you were testing, you switched over to a
saline solution where you have a salt water intrusion. Was there a drop
off in the overall treatment effect until that point, or if there was, how
long did it take for the bacteria to acclimate themselves and bring the
treatment level back to what it was initially before salt water intrusion?
MISS KINNER: Well first of all let me say that we initially started the
experiment, we used a COD as a monitor, a daily monitor, instead of BOD
so that we can have lag time. And this COD did not work at all in salinity,
so that the first experiment that we had, we actually assumed that after
one week we were at a steady state and found out that were the BOD that
we were in fact. That steady state we were getting ridiculously high
answers to the COD, if you had not promised to run one of those, we should
have had similar results. I spoke to a gentleman who said they actually
ran TOC which we are now switching to. So it is hard for us to say how
long it really took to get the steady state. The second one that we
switched to the salt. We did find that we got the steady state faster than
a week, in fact, we were getting good BOD removal in four days or so, but
we just do not have, the data because of these testing problems that we ran
into on solids as to how long things have to acclimate. From a microbio-
logical viewpoint, just salt sampling that I was doing, there did not appear
to be any specific change in the populations but we were observing as a
1491
-------
die off of rotifer and protozoa.
MR.VESIO: Does 6-7% salt concentration significantly affect bio-activity in RBC?
MISS KINNER: Yes, we did that at seven percent. There were variations
because of the initial problems in running the system. We had some vari-
ations that we kind of were fitting into there up to six percent salinity.
And the system did take a few days corrections, kind of stepped to a quota
effluent and get most of the BOD did handle that way. But the removals
were somewhat higher, excuse me, the removals were somewhat lower. The
BOD were somewhat higher during that period, and they probably would
not meet and actually load for this municipality.
MR. FREEMIRE: Is there any subsequent follow up on performance as to
adding up the alkalinity to adjust pH or aerating?
MR. HITDLEBAUGH: As & result of the work that we did, the Army Corps of
Engineers both from our district had put in a project to add supplemental
aeration to the system and that was, that is supposed to be under way now
but I have not checked on it recently, and hopefully that will be in next
summer. I am not sure what the status is as far as any pH adjustment with
the system. They may have been waiting to see what the effect of the
supplemental air had to see how much of an effect that would have as far
as improving all systems.
MR. BRACEWELL: I did not quite understand the impact of temperature. You
said that at 13 degree centigrade that you were having an effect on your
removals. The data during the winter time was about as good or better from
the summer. And so how do you know, what makes you say the temperature
was important and how important was it?
MR. HITDLEBAUGH: Maybe that was more confusing. What I was trying to say
was that during the summer time the temperature is 26 degrees centigrade,
much more optimum conditions for biological activity. During the winter
time, the temperature was 13 degrees centigrade and much more an inhibitive
factor on biological activity. Even now, I am saying a negative effect,
from summer to winter even though the temperature has an apparent negative
effect because of the fact that we did not seem to have any problems with
the DO limiting conditions, for some seemed to be better. So what I am
trying to say is that the fact that there was more dissolved oxygen limit-
ing conditions in the winter time, over compensates for any negative effect
due to a lower rate of temperature.
MR. BRACEWELL: Did you see any interim effects, for example in the fall
when maybe the temperature was .2 degrees centigrade?
MR. HITDLEBAUGH: Yes, definitely.
MR. BRACEWELL: Was not a DO limited?
MR. HITDLEBAUGH: Yes, definitely.
MR. BRACEWELL: Is that a BOD removal?
MR. HITDLEBAUGH: The plant did meet and I mentioned the word periodic, we
were involved in the process from August through January and in fact Auto-
trol even went down and did some, a little bit of testing and in fact
during their test, the plant was needed experiment. So there was a few
1492
-------
periodic problems and we picked, we chose the study to look at the peak
summer and the peak winter conditions to see the extremes there.
MR. SHAMITKO: We have received some reports that some of the shafts were
breaking with some of the Autotrol biodiscs. That is my first question.
Has that been corrected and if it has been corrected, what was the original
problem? And my second question deals with your samplings for the BOD
testing. Was the sample taking directly after the waste had flowed through
the disc or after it went through the final clarifiier?
MR. HYNEK: You were speaking of my impression of the talk.
MR. SHAMITKO: Either gentlemen.
MR. HYNEK: The constant situation we were reporting to you raw wastewater
and final effluent because that is what was reported by, according to the
State of Michigan. In the Cadillac situation we had information at both
points, the secondary clarifier effluent, the biosurf effluent, and the
final filter effluent for ammonia. Did I answer your question?
MR. SHAMITKO: Yes sir.
MR. HYNEK: In the case of Cadillac it was flow comppsited sampling auto-
matically. Very pretty plant. We all had shaft difficulties and we have
faced the steel failing, the construction of a shaft requires a welding
procedure with a backing strip when you bend two U-shape metals together.
And the failure was placed to an interruption of this backing strip and
not according to specifications by the way, and we have since terminated
our arrangement with that particular supplier. We are currently purchasing
our shafts from steel-making people in Japan and with a very superior shaft,
and there is a definite program with Autotrol Corporation and its customers.
But those that have failed are being replaced. It is not necessarily all
at Autotrol expense but it is a very, as you can imagine, a very serious
situation when one of these large shafts break and causes the customer
concern. But they have been very tolerant and we have been open as we can
with them to explain what happened, why it happened, and what we are
prepared to do to back up our product.
•*
MR. SHAMITKO: One other question. The follow-up of the first answer you
gave. The BOD removal rate, was there a big difference from the BOD
measure after went directly to the RBC as opposed to after it went to the
clarifier? Was there a big difference between the BOD amounts?
MR. HYNEK: Well, we do not normally get separate data like that from the
customer but the thing is how you take your sample at the biosurf effluent.
If you do not settle it you will get quite a large number indicative of
the endogenous respiratious. What we do usually at that point is to settle
it for 30 minutes and decent the supernatant .& clarification and then compare
that number to an actual clarifier performance. For example, they are
adding alum-.for phosphorus removal. The clarifier is going to do pretty
good. And our set of samples probably would not settle or give a BOD number
as low because of the little bit more of suspended solids contribution.
MR. GARY: I think the area of the small package treatment plants including
RBCo In the state of Pennsylvania, the state may accept your treatment
system if the system is approved by NSF. Is your company planning to get
1493
-------
the NSF to approve your equipment?
MR. VESIO: Well we certainly thought about it. We have mdt done or
pursued NSF approval with this product as yet. We have mixed emotions
about it. The NSF approval even though we realize that many states do
require a pre-engineered type of system, what we like to think is the fact
that number one, in the RBC approved of themselves in terms of performance
capability. Number two, we certainly would be the first to admit that
there is nothing magical or mystical about the concept that we have come
up with. We are fortunate in that we are able to take the puristream
technology with regard to the steel tankage and diffuse their portion
of their system, as well as Bob Joest and the TAIT Bioshafts expertise
in terms of the size in the RBC, and take the two technologies together
and that is a good system. If we find that because of that is certainly a
pre-engineered type of package. If we find that NSF testing of some
type would definitely be required, quite frankly we would have no recourse
but to go along with it because we have a tremendous amount of confidence
in this particular product in the concept.
MR. CROUCH: What kind of mechanical collector do you have in that final
clarifier?
MR. VESIO: On most of our models we use a hopper bottom type clarifier
such as the typical arrangement show. When we get up to above 150,000
gallons per day, you have the option of going with a rectangular suction
type of collector mechanism or a rectangular chain flight collector. But
those are often standard and flows from 150,000 gallons per day and up.
DR. POON: I wish you could elaborate how do you size RBC system in the
final clarifier?
MR. VESIO: The RBC themselves, we feel, that we have a pretty conservative
design in that in the first stage reactors, we do use a hydraulic loading
rate for domestic strength sewage, and in the first stage units, the
hydraulic loading rate ranges from 1.76 to 2.0 gallons per square foot
per day. Now as far as the clarifiers are concerned, as indicated the
surface settling rate does vary depending upon the flow. What we were
striving to do was to supply the clarifiers with two hours retention
time. Now some states hawe their own criteria for -clarifiers and my feeling
in this particular thing since I had input to this particular concept was
that the two hours is plenty really even in an activated sludge system,
two hours is plenty of retention time with any clarifier. With a RBC system,
in most instances especially if it is a lightly loaded system which just
about our model upon start-up and quite frankly unless something unusual
happens in the development of the community, these systems do not usually
come out to full loads for some length of time. And because of this the
ambiant temperature is going to be given nitrifiers and in the tail end of
the third stage and in the first stage and a first stage reactor. You
are going to run into a problem if you get into an extended retention
time in the final clarifier where denitrification is going to begin to
occur. The Autotrol takes that into consideration; the post-aeration
aspect problems is great and if you keep in mind that we are talking
applications' involving recreation parks, subdivisions, and so forth. But
it is only in which the problems are apparent in the inheritance of the
extended package plants. So basically in summary, is that in our large
first stage reactors we do not exceed two gallons per square foot per day
in hydraulic loading and because the clarifier is a standard size to give
1494
-------
you two hours retention in both the primary and the final clarifier.
If the state regulatory agencies or such that no retention time is
required, you have the flexibility to do this because of the marginal
configuration. This is true for the flow equalization chamber, it is
true for the aerobic digestion chamber. As I indicated the two percent
on flow equalization to keep the separate capital in aerobic digestion.
There are states like Ohio, states like Indiana where surge capacities
in Indiana, in a school project they require 67 percent surge capacity at the
time of school. With this particular margin of configuration, you can
give them 67 percent. So Ohio has a very elaborate program that you go
through, based on an average daily flow and a 3.3 peaking factor and you
go through the calculation that they prescribe, and you could come up with
a flow equalization chamber of about 37 to 40 percent in volume. So you can
change the volumes as needed to meet the state requirements or application
requirements.
DR. SMITH: We would just like to take 30 seconds to sort of summarize
what we found from this conference. I think that RBC technology is
demonstrated as that to be an advanced technology and indicated by the
number of manufacturers and competitiveness among these and this sort
of gives us a choice. Also you notice there seems to be a real wide
interest in RBC and there seems to be so many applications for RBC that
most other technologies do not have that many applications, lechate treat-
ment, nitrogen and phosphorus removal. I just returned from talking to
the puricycle people in Colorado and' they had a very interesting scheme
where like homes in the mountains where you do not have a well and you
cannot have a septic tank, but they have come up with a treatment channel
with a little RBC unit in it and some very rather sophisticated technologies.
And it is monitored with by method of process or technology and it costs
about $8,000. That is actually cheaper they claim than buying a $30,000
small package treatment plant. It takes their sewage and treats it to
portable water quality. And I am certainly not an oponent of portability
at this time. There is a lot more research that must be done, but I think
the little black box that we have heard so much about is finally here and
in that NSF is evaluating it at this time. Also you learned from some of
the papers that there is a lot of clearance in modifying some of this RBC
technology and it is nice that we have RBC technologies of all different
sizes for households or trailer park or service stations on up to hundred
million gallons per day plant.
1495
-------
-------
Session 4. BIOKINETIC STUDIES
Presiding: C. P. L. Grady, Jr.
Department of Civil Engineering
Purdue University
DR. JENKINS: Mike, I really do not know how to phrase this question
but the impression I get is that you are forcing the biodisc system to
be like an activated sludge plant by intentionally scraping off materials
from the surface of the biodisc and turning it away just as you would waste
activated sludge, but I guess my general question is how does this, how
does the system that you describe, this method of sludge wasting really
replicate what goes on in a biodisc system in the field where this type
of wastage does not take place. And I guess the specific part of the
question would be, how would you relate your sludge wasting rates conducted
as you did, to the nitrification values you obtained with them, to what
you would get in a field scale system where hydraulic loading or organic
loading seems to be the parameter, used to correlate with whether you
get ammonia or not, or oxidation or not.
DR. SAUNDERS: I do not in any way mean to indicate that I would like to
have the operator go out and scrape his disc system. That is not the
purpose, that was not the intention. The purpose was to measure the
rate of growth in the system and to see of the growth rate as we would
relate it to SRT values, or any other way we like, relate it to the per-
formance of the system with respect to nitrification, so we are not talking
about this as a means of operating the system or in any way applying it
to practice. I do not want that to be indicated. The other side of it is,
it says simply the data that we have for nitrlfiers can be applied to these
systems. And that was really our purpose, to simply see if the relation-
ships that we developed forcing the systems would in fact indicate that
the kinetic data that we reliably used in activated sludge systems could
be applied here. There is a tremendous amount of information that has
got to be combined with this. We have got to look at the heteretrophic
organisms and the rate at which they grow .in the systems, .the orate to 'which
they are .being sloughed frtan fche system, there is a tremendous amount of
work that needs to be done. Probably one of the most important things
that needs to be done in this regard is to look at the dissolved oxygen,
because now we have been looked at probably the primary reason for lack of
nitrification, or having nitrification in the plug flow system, the first
one is overloaded to produce anaerobic conditions. If we do not have dis-
solved oxygen we do not get nitrification until the subsequent unit. The
results that we have got, again, if we listen all of them this afternoon,
they have used this approach and my contention is it is a valid approach,
to look at kinetic growth relationship within the culture system. You have
got to combine it with other relationships, with the growth of heterotroph,
experimental determine and also relationships for oxygen uptake and oxygen
profiles within the film.
DR. POON: I am not too sure about your last statement. If we ignore the
consideration of suspended solids effects mi safety because such a high
solids exist in the first stage of the RBC, oxygen demand would be very
high. That could probably, easily reaching- a critical factor in the
design. That is just'one comment.
1497
-------
DR. KINCANNON: I think you are very true on that end. Again, at least
my feeling is we should be looking at the suspended solids and again I
made the comment that, well, maybe a factor of safety force, but like
you say, it could be a big problem too,
DR. BOON: I want to ask you, did you look into the sludge retention time
in the system, as a whole, and then try to differentiate between -two
fixed and suspended biofilm?
DR. KINCANNON: Well, I tried to omit the problem that I had in trying to ;
make the calculations. Now, again, I take the different approach than
what Mike did in the last paper. He takes the total solids. I am inter-
ested in only the solids that I considered being produced. I look at
sludge retention as being the reciprocal of a growth rate and not the time
the sludge is held in a unit. You know, it is a growth rate, when you
have this solid accumulation such as this, it is difficult to determine
the actual amount of growth rate. I tried to make some calculations on
this and ran into a lot of problems. Wow, I agree very much with what
Bill Characklis mentioned earlier, that you finally reach an equilibrium
level so that your growth is actually constantly being lost, and if you,
and I do this with a biological solids all the time, you do not have to
worry about this accumulation, but the suspended solids in the effluent
basically measures what is being produced that day.
DR. POON: That would be a constant sloughing off on the biofilm sometimes
in fact some of the particulate form would be trapped by the biofilm,
There is a constant change between sloughing and attachment to the biofilm.
Well, my calculation for sludge retention time is close to 20 days under
perhaps a similar loading as yours. I just want to see if you have any
calculations concerning it.
DR. KINCANNON: Well, no, well, I made them on this but I was, like I
say, I was concerned about the numbers because of the solids being retained
here.
DR. REYNOLDS: You mentioned an active biomass thickness, I think it was
200 microns. How did you determine that?
DR. KINCANNON: From the literature. I did not determine this on our system
itself.
DR. DENNIS: When you calculated your growth rates you subtracted the
delta COD from the effluence? I believe that was your technique, you got
the refractory compounds by a delta COD technique.
DR. KINCANNON: This was for the substrate removal or SB. SB was the
total COD minus this residual COD. To me this more nearly represents
the amount of material that is available for biological degradation. In
other words, we have this -residual which, if you let ifc aerate for a week
you are not able to lower it below that.
MR. DENNIS: So, you use the delta COD to calculate the effluence, -and
that was what you used to get the unit of substrate utilization rate?
DR. KINCANNON: Yes.
1498
-------
MR. DENNIS: I do not think you subtracted it from your influence
concentration, or did you?
DR. KINCANNON: I did not.
MR. DENNIS: Would not that tend to bias your results to a higher substrate
utilization by subtracting it?
DR. KINCANNON: Well, there is a question of whether you should subtract
it from the influent. There we were using glucose as the substrate and
assuming this is completely bio-degradable, that is the reason I did not.
MR, KIM: What I am asking is if you include R in your equation do you think
that your model becomes more useful?
DR. HAUNG: Well, I do not really know.
MR. KIM: Actually, this film thickness is proportional to the tip velocity
rather than rotational velocity. Why do you use the rotational speed instead
of the tip velocity?
DR. HAUNG: Well, I have N and R time together. Of course, it is also a
function of kinetic speed too. I did not neglect that.
MR. KIM: If you use tip velocity maybe you can generalize your model.
DR. HAUNG: Well, if I use different disc sizes I may do that. In this
study I only have used one single size.; As Dr. Friedman stated last month
that the rotating speed should be considered at the same time for scale-up.
That is very important. You do not want to talk only tip speed it should
include both the radius of the disc and rotating speed.
MR. KIM: In Dr. Friedman's paper last Noyemeber he mentioned that the
rotational speed was a very important function in his research, right?
However, your model is really related to the film thickness. The thickness
is a function of tip velocity, rather disc speed. But that
is not my point.
DR. OfSHAUGHNESSY: What was the minimum ammonia effluent concentration you
achieved?
DR. HAUNG: You mean the whole test?
DR. 0'SHAUGHNESSY: In your effluent?
DR. HAUNG: The minimum?
DR. OfSHAUGHNESSY: Yes, what was the lowest ammonia concentration?
DR. HAUNG: In the effluent the ammonia concentration is about one mg/1.
DR. 0TSHAUGHNESSY: Did you go below one? Most of the time you were above two?
DR. HAUNG: Yes, most of the time.
1499
-------
DR. OfSHAUGHNESSY: Let us go back to the previous question. In other
words, we have done fixed-film reactors also. To get zero on the kinetics
above 2 milligrams per liter, but once you go down below two or three it
goes to first-order kinetics.
DR. HAUNG: Yes, that is a possibility.
DR. O1SHAUGHNESSY: And looking at the way you had it, I think the model
would be valid for your higher concentrations but to try to extrapolate
that to get that final removal might be dangerous.
DR. HAUNG: I am not saying your comments are wrong, but some further
studies should be conducted to reach the possible conclusion.
DRo 0'SHAUGHNESSY: What was your minimum influent concentration?
DR. HAUNG: Minimum concentration is about 3 or 4 mg/1.
DR. 0*SHAUGHNESSY: Your minimum influent concentration was 5 in the
model held as compared to only a hundred milligrams per liter.
DR. HAUNG: No, I did not use both concentrations to fit that small unit.
I did not try that. I used a very high concentration in the full scale
study up to 600 mg/1.
1500
-------
Session 5. AIR DRIVE AND SUPPLEMENTAL AIR
Presiding: W. A. Sack
Department of Civil Engineering
West Virginia University
MR. GERHARD: I noticed in the beginning of your talk that you indicated
that there were two methods of control. One was speed and the other was
aeration. You have given a history of some eighteen to twenty four months
with work on supplemental aeration, but you did not mention at all the
speed control. Did you, is there a reason for that?
MR. SULLIVAN: Yes sir. We feel from a standpoint of power evaluation,
believe that RBC an increase in speed increases power exponentially. We
do not feel it is feasible to increase the dissolved oxygen within
the reactor by increasing speed because the power requirement due to an
exponential rate on the BOD removal film
MR. GERHARD: The contrary situation might exist, though, because you cannot
get air for nothing either and in terms of energy you have to provide
blowers certainly with the sufficient size. As I understand Alexandria
situation was a mechanical drive and you added air to it, which improved
the plant efficiency certainly at the expense of the air blowers. Can
you positively state without having pursued it in depth as you have your
aerosurf, that the balance of energy is in favor of air as opposed to
perhaps a minor change in the rotating RPM with a minor change in horsepower?
MR. SULLIVAN: Based on our conclusions of work with air and at one point
six RPM, determining the power evaluations of increased speed, it is my
conclusion that at this time that we can definitely state that we definitely
do not want to increase speed on RBCs.
MR. GERHARD: Is there any kind of perhaps kind of a mechanical reason
that you do not want to exceed a sixty foot per minute tip speed or whatever?
MR. SULLIVAN:
economics.
The basis for my statement is based on pure energy operating
MR. BURNER: In your 18-month testing at South Shore did you look at the
differential and settleability characteristics of the sludges?
MR. SULLIVAN: There was during, I do not want to say every test stage, but
during the majority of test days discrete particle settling test done on
the effluent from fourth stage. The results indicated that the settleability
of overflow rate reach an effluent concentration or approximately the same.
In some cases the mechanical drive did performance and in other cases the
air drive did out the performance but the ranges of overflow rate required
to get a given effluent concentration at the same rate they did it by ap-
proximately fifty gallons per day per square foot, which I feel is the ac-
curacy of the test.
MR. WONG: You have not mentioned in your discussion regarding the amount of
air required for supplemental aeration
1501
-------
MR. SULLIVAN: The results here do not present all the data evaluated.
There is another paper called the aerosurf process that does go into all
the data but I could tell you that the range of applications are there
on the data that was presented, was approximately a hundred to two hundred
CFM per 'shaft at the low loading of approximately a hundred CFM to two
hundred CFM was only present on upstream stages generally downstream. We
were in the range of eight CFM, but an average of first stage of approxi-
mately 140 CFM, 150 and downstream 80 to 120. It is very difficult to
give that an overview. What you have to do is to examine nine months
data to find out what the organic loading are to find how to match your
organic profile with the amount of influent flow but the data is available.
MR. .PATRICK: How do you arrive at biomass thickness?
MR. SULLIVAN: All the biomass thickness that was conducted at South Shore
was conducted by rotor shaft bearing. In one of the system, was jacked
up you know the dry weight of the media with biofilm on it immersed with
a 40% immersion. We then weigh the bearing and we get a reaction weight
and that tells us the total slime on the media.
MR. PATRICK: And how do you equate that to a thickness?
MR. SULLIVAN: Well you have x amount of square foot on it, the media, as
present and each cubic foot of biology was approximately, weighs approxi-
mately 62.4 pounds of weight. You have a hundred thousand square foot.
You know the amount of biomass present has a weight of "x" multiply a
hundred thousand square foot times the thickness to come up with that
weight, it comes up with the biofilm thickness.
MR. PATRICK: And is there some advantage to minimizing biomass thickness
other than operational considerations?
MR. SULLIVAN: We have definitely found in the minimization of the anaerobic
layer tends to promote a biology which was higher in life and has a better
BOD removal in the aerobic layer.
MR. PATRICK: How about maintaining the structural.integrity of the system?
MR. SULLIVAN: Well, there is no question you do reduce slug you reduce
structural integrity, so there is no question about that.
MR. PATRICK: Do you have a cutoff weight on that?
MR. SULLIVAN: Yes, thirty thousandths.
MR. PATRICK: What was that?
MR. SULLIVAN: The test data that the Autotr9l has performed were being an
operating shaft life of fifty years of biomass thickness of ninety thousands
of an inch. We have conducted all our structural requirements at ninety
thousandths. For a. process standpoint I think the number I quoted was sixty
thousandths, because ninety thousandths biofilm thickness is too thick.
MR. HALLHAGEN: How does the temperature affect the supplemental air system?
1502
-------
MR. SULLIVAN: The work here was presented independent of temperature.
It is obvious that warmer wastewater temperatures and warmer the air
temperatures the-.amount of air will increase. The data here was over an
eighteen month scan so as a result it did take in winter and summer oper-
ations, but we did not go through a number of seasonal changes, so I
really have not determined the effects of wastewater temperature on the
amount of air put into the reactor. It would take probably another year
to do that. There is no question in my mind however, that a lower waste-
water temperature the less air would be required and at higher wastewater
temperatures more air would be required.
MR. RUSHBROOK: Does Autotrol see itself moving away from mechanical drive
units in favor of the air-drive?
MR. SULLIVAN: There is no question from the process standpoint that
automated feels that aero-reaetory is the right way to go from the bio-
logical standpoint. I think all our recommendations are that way and have
been that way for apporximately six months to a year. We are recommending
in certain installations retro-fitting of existing installations with
aeroreactors. We believe it has definite process benefits. Yes, we are
going on it»
MR. MOORE: I have two questions. One question is with the air-drive.
•What is your recommendation of maximum size with BOD loading compared to
mechanical2 The other question is, it seems to me that as you get into
second, third and fourth stages of the process that you are not dealing
with the biofilm thickness problem and thereby it seems that a need for
air-drive is minimized.
MR. SULLIVAN: The current design with the recommendations are approximately
five pounds soluble BOD, with the air and four pounds soluble BOD maximum
stage level with the mechanical. You can see that the data 'presented indi-
cates about six pounds before you get a DO sag. But we design it right
now on a five pound, basically because of the data we do not have a temper-
ature, extremely high race with the temperatures. Well in essence if the
biofilm is thin the effects of air are minimized and you can see the reduc-
tion. There is not a significant increase of BOD removal. Autotrol feels
at this time that energy is a big function of the RBC system. The capability
to incorporate air and to reduce rotational speed, and to minimize energy on
the downstream stages is critical in the future. Air drive gives us that
opportunity. From a process standpoint, downstream at low loads, we do not
see a definite need for process improvement. However, coupled with the
reduction of speed and the optimization of energy, we feel that air is
critical.
DR. CHESNER: Is there any real concentration of dissolved oxygen that
should be maintained in the reactor?
MR. SULLIVAN: Very difficult question to answer. Off the top I would say
about a part of a half a part of dissolved oxygen within the fluid. The
critical point is really dependent upon the biology present on the media.
When we see a biological shift we try to raise dissolved oxygen if it is
a half a point or a part. In other words, when a biology shifts from a
truly aerobic culture to an organism, it is like Beggiotoa or something.
It is oxidizing sulphur deriving its energy sources from that other than
1503
-------
carbon. We tend to change the environment and whether we change the en-
vironment 'the only control that we have is increasing the dissolved oxygen
within the reactor. So the dissolved oxygen is not the critical criteria.
As an off the cuff statement I would say about a part to a part and a half,
but the real key function is the biological organisms present the change
in that organism.
JR. JANK: Just a continuation on that question in Figure 2 you showed the
DO level with the...applied organic loading. In order to get one milligram
per liter of DO you showed a loading of one pound.
MR. SULLIVAN: No, it is a delta in dissolved oxygen for instance if I was
entering the reactor it begins to sag dissolved oxygen. If I was entering ;
the reactor two it determines when the biology utilize oxygen. If I was
entering the reactor two when I, said I would have a negative. If I was
going up I would have a positive. What I am finding out is when a mechanical
driven system I am sagging in about three pounds soluble and on the air driven
systems I am still increasing dissolved oxygen at approximately six pounds
soluble at an increase in the RPM in the reactor up to a six point soluble.
It is not concentration, it is delta concentrations.
MR. BEISEL: In your studies there is a cost for the supplemental air both
the capital blowers and piping plus your operating costs. Was this cost
effective per pound of BOD removed, in other words the additional cost,
if I wanted to convert my present system?
MR. SULLIVAN: Well, I guess the question is on necessity. I think on the
newest systems there is no question on mine with the potential of energy
reduction and downstream speed control and flexibility. We feel it is a
necessary capability in an RBC system with existing plants. It depends on
how the effluent is doing. We have existing plants that we definitely are
recommending in order to achieve our full quality to incorporate supplemental
air. If you are currently having no problems with the mechanical .equipment
there is an awful lot of plants mechanically around the country now, right
now, doing extremely well. I would not put in there, but if something happens
that you are unknown of I think the incorporation of air could provide a
significant benefit to the reduction kinetics. I would not recommend it to
take every mechanical drive right now to combine with air. I do not think
that that is necessary.
MR, BEISEL: I have a 3% MGD plant with 24 Autotrol units and I have had
some operational problems caused by other things in the plant, such as
high strength supernatant from the thickener that is giving me some problems.
I have finally gotten around it but I have all mechanical drives, have no
way to supplement air without additional blowers and piping.
MR. SULLIVAN: The first step that Autotrol recommend is at this particular
time, to optimize the system as much as possible with the mechanical. In
the event that cannot be done the incorporation of air is probably a desirable
end result. But I do not want to tell you at this time that you should go
out and spend "x" number of dollars on this.
MR. BEISEL: Well, I only have about four months operating experience on
the plant. It started up in October.
1304
-------
MR. SULLIVAN: Maybe we could discuss it in detail afterwards.
MR. BROWN: Did you evaluate how much air it took to ratate the given speed
or power requirements?
DR. SRINIVASARAGHAVEN: We certainly did not do anything about power require-
ments because the air was provided by a portable compressor so it would
not give very much indication. In terms of air requirements, it escapes
my mind now but I think the one point six RPM took 180 CFM. At one RPM it
was close to 120 or 130 CFM0
MR. BROWN: Will air be the sole source of power to drive these or will there
be any form of mechanical-assist?
DR. SRINIVASARGHAVEN: We have not provided any other source. We are en-
visioning a mechanical device that would be serving each train of RBC that
you could hook up in case you have problems to start the RBC. But each RBC
shaft will not be provided with a separate motor of some sort.
MR. WARD: Was that an overall loading of one point four pounds of soluble
BOD per thousand square feet?
DR. SRINIVASARGHAVEN: That is correct. The first stage loading would be
four times as much.
MR. WARD: I thought you showed six stages though, instead of four.
DR0 SRINIVASARGHAVEN: But in the pilot unit I am talking about the loading
was four times as much in the actual installation. You could operate it
anyway you want. You could lift the baffle, you do not intend to operate
six stages. You could operate any number of stages from zero, one to six,
so the loading would depend on how you wish to operate, but you would not
intend to operate any higher than five pounds. First stage loading.
MR. WARD: I missed it if you presented it, what is your general conclusions
about reducing the RPM of the latter stages?
DR. SRINIVASARGHAVE: The pilot unit was not set up to reduce the RPM for
each stage. You could reduce the speed only if all stages...the one RPM
test indicated test data, indicated performance is just as good as one point
six RPM. So unless there is difficulty in meeting the effluent standard
we will recommend that they operate at one RPM with the provision it is
fairly simple to increase the RPM, so they could increase it.
MR. SERPA: On your soluble BOD was this a calculated value or an estimated
value based on the total BOD?
DR. SRINIVASARGHAVEN: It is a test value. We measured the soluble BOD and
projected for the future the design values projected at the test program
values it actually measured.
MISS ROSENBERRY: I have two questions. The first, if you put the RBC over
the activated sludge tank, is there any problem in getting at the air
header if you have to2
MR. COWEE: I would say that depends a great deal upon the type of air
header that you have in the system. I am not really familiar with activated
1505
-------
sludge systems so I cannot specifically answer your question. I would
say there is a good possibility under some circumstances that could be
the case.
MISS ROSENBERRY: And the next question is, is there any change in head by
adding the RBC?
MR. COWEE: Not significantly.
MR. GERHARD: I have got a couple of questions on the Philadelphia thing,
having listened to the talks several years ago at the Philadelphia WPCF
Convention. From the claims made at that time, seem now to have been
either changed, altered, corrected or something as I understand it, the
fine bubble diffusers were put into the Philadelphia twenty-two shaft
RBC installations to bring the process up to the required design improvement
level. I heard from another consulting engineer who was interested in the
same or a similar process that the chances were good had it been checked,
that just the addition of the fine bubble of diffusers system alone would
have given the improved capacity or treatment levels, and that the RBC air
drive units were superfluous. That is the first question. To what extent
then do you know or have you compared the improved process performance without
the RBC is running. That is number one. Number two, I understand also
that the same installation is preparing to take bids on 180 more shaft with
mechanical drive and would you care to comment on that?
MR. COWEE: I think I will take the last question first. I am not really
capable of answering that question because since it is now a system that
is coming up for bids it is being handled through the marketing department.
I had not been related with the Philadelphia system I would say now for
a good five, six months. I do not know what is going on as far as hanky
panky related to bidding the system at this point. The only indication that
I have or that I have heard of it at this point, was that due to the require-
ments of maintaining a very specific RPM profile on the air drive, the oper-
ations personnel were not interested in continuing to do that type of fine
tuning on the system, but all the testing that we have done and that we have
seen done by Philadelphia has indicated that there is no need to continue
doing that fine tuning. So it is more or less a political question at this
point, between the operation people as to whether or not that is the case.
That is my personal feeling. I really cannot say specifically what is going
on there, so I cannot answer your question directly. I wish I could remember
all three of your other first question.
MR. GERHARD:
fine bubble?
Was there a comparison of the RBC not in operation with the
Was there an actual comparison made?
MR. COWEE: You mean having the RBC in the tank without turning them and
using the fine bubble, the RBCs. will not sit in the tank without turning.
Just the oxygen pickup from the air pickup from the fine bubble system or
the course bubble system and the hydraulic role does impart motion to the
RBC. So they have to be tied down in some fashion so as not to rotate. I
do not think that has been done, I am not aware of it at this point.
MRo GERHARD: How about without the supplemental air condition?
turned off to see what the natural effectiveness?
Was that
1506
-------
f
MR, COWEE: I have been told that the people at Philadelphia had run the
RBC system in the aeration tanks without using the supplemental air
header and were able to get significant rotation of speeds on the shafts
and were able to achieve treatment on the shafts that way. Once again,
I have not direct knowledge of that. I was not involved in that testing.
MR. GERHARD: You indicated that you were using about 3h out of the total
available horsepower of 7% which would leave 12 horsepower extra motor
capacity, that you then used to drive blowers.
MR. REH: That is correct. Yes sir.
MR. GERHARD: Was any attempt made at least on one or two units to increase
the RPM by simply changing sheaves or sprockets on a chain whichever way
you wanted to do it, utilizing a full 7% horsepower than in rotational speed
to see what effect it would have. And as a side comment on that, we have
done so on pilot and full scale units, we have found that the biofilm thick-
ness is very definitely controlled by RPM because of the additional shearing
forces and stresses of the higher speeds. We found additional aeration, we
found additional sludge settling characteristics change, in other words
better suspended solids and so forth. It seems incredible that you did not
at least try this. And I tried to ask the gentleman this morning the same
question and it seemed to be an evasive answer. You had an opportunity to
do so, was it done, and what were the results?
MR. REH: I will not evade your answer. The direct answer to that is no,
we did not try. It has not been excluded, the discs were there and will
be there until project engineers come home. This was an experimental program,
only so many things can be done over a given period of time. We did discuss
that option with Autotrol and it was felt that with what we could get in terms
of increased revolution with what we had to turn it with, we would be better
in the short run to go with what we did. There is no reason that we cannot
go back and try it and we probably will.
MR. PRICE: It seems that you really did not have a choice. You give a cost
analysis to show that supplemental aeration is the cheapest way to go, but
with biofilm thicknesses as thick as they were, if you did not control the
biofilm thickness it seems like you .were looking for Impending .shaft .failures»
MR. REH: THat is a good point, and the Autotrol folks again were helpful
in making some load cell weight measurements and the equipment they were
using was some prototype stuff and they are not quite satisfied with all
the numbers but they did indicate that if we kept up with the thickness of
the growth we would probably be increasing our potentials for shaft failures,
so yes, we may be trying to justify the obvious, that is a good point.
MR. PRICE: The other question is did you do any DO profiles, that is was the
improvement due to higher DO or was the improvement due to the shearing
process of the aeration.
MR. REH: Two fold. This system in my opinion was oxygen limited from the
word "go". We found the oxygen measurements before aeration, Roger, generally
were in the range of one to two tenths on th'e '-influent side and in no case
did they exceed a half a part per million or so on the effluent side. When
we begin to aerate we found we consistently found that we consistently main-
tained a dissolved oxygen in excess of two and sometimes as high as four and
five parts per million, so we receive a benefit from the actual aeration of
1507
-------
the system as well as from the turbulent supply. It is two-fold benfit.
MR. JOHNSON: I understand from what you said there, that the baffles made
no difference. Are you saying that you tried baffles between each shaft
and without, and you are saying there was sort of a plug flow effect in
either case?
MR. REH: Correct. The baffles, these were between each of the four shafts
in the series we have a. set of five removable baffles plates and we tried
every conceivable configuration from all baffles in place, to no baffles
in place and the results there indicated less than two parts per million
soluble BOD. And I think plus or minus two parts is questionable in that
sort of a test anyway.
MR. GERHARD: Again, I have some comments, Jim. First of all, we conducted
pilot plan studies of a similar nature and experienced similar results, so
we corroborate your findings on your Cold Spring thing. Secondly, we also
found that by not really heard it described as a closed media as such, that
before but none the less there is a definite difference in water passages
and air passages in the different manufacturers of plastic designs. We have
likewise not experienced problems with the oxygen limitation on our units
and where we have had this situation we have had no problems with meeting
performance requirements and so forth. Do you know or have you checked out
or do you know of any place where a study is planned where a comparison
between manufacturers' media was planned or contemplated, and if so, would
you comment on it?
MRo MADDEN: I understand that the Environmental Protection Agency in their
Cincinnati office is doing or planning to do a study. The definite study
I am not certain of, but one of our people was contacted by the EPA and they
did spend some time with us. And it seemed that gist of their whole thought
was to carry on a side by side study, wherein they would have manufacturers
participate in something like an NSF situation and determine whether on a
hundred thousand square feet of manufacturer!""A";.'RBC'is equal or better than,
or not as good as, a hudnred thousand square feet of manufacturer "B" material.
We have had a fair amount of experience in examining both your material and
material from Autotrol and there is no question in our mind that there is a
serious deficiency in the design and the availability of air to get inside the
closed media, and it is something that has to be addressed. It is something
the gist of the presentation here was to say look follower, we do not agree
with supplemental air. If you get up in loading rates of 10, 12, 13 pounds
soluble applied per thousand square feet of shaft, there is a trend at that
point where you could look at supplemental air. In the case of RBC plant
already discussed, they had a tremendous amount of odor problems in the ini-
tial start-up and one of the panaeceas suggested was to put in supplemental
air. And they did put in supplemental air and it did not really help except
it changed the air from inside the plant and pushed it ouside the plant. So,
there was a lot of flack and I believe there are some people here who are
involved in that facility, and they could even comment more than I could
because you know we are a manufacturer and we look at the problem, we and
when we can as Autotrol does, and as your firm does, we will go back and we
will say look, this is what we think you should do if it is a problem that
is outside the scope of what we could categorize as our responsiblity. If,
and we all have had them, if you have media problem or you have a shaft
problem I think it is certainly a credit to the manufacturers in this industry
that we go back and sweep up our mess. And the important thing for you all
1508
-------
to understand is that just because we have had some shaft problems and
just because we may have different opinions about the way things are
done, the RBC process is in fact a neat process. It is so good that I
sincerely believe it is going to take over activated sludge in a very
short time, as long as we can keep the equipment cost down and maybe
keep the brass knuckles off between the manufacturers.
MISS KINNER: I was wondering if you had done any culturing or any micro-
scopic observation on the films that you had there.
MR. MADDEN: No, we did not do that, Nancy. It would take people of your
caliber to do that kind of study, and unfortunately when we started the
study in all sincerity, it was sincerely a marketing study. What we were
trying to find out was, was there a difference in the magic aurora of
supplemental air, and quite honestly there is not any. We do not see it
and we do quote project and design project based on non-supplemental air
on a side by side basis with supplemental air, with air drive and we do
believe the non-supplemental air system can be just as good as the latter
system.
DR. SRINIVASARAGHAVEN: You kept saying that there is no difference in
soluble BOD removal at Alexandria with and without supplemental air.
MR. MADDEN: Excuse me, no, I said there is a difference.
was reported to us through you and our Mr. Friedman.
That information
DR. SRINIVASARAGHAVEN: Right. The information that you got was not done
on a continuous basis, it was probably taken from spot samples through
your program established. The best way to compare the two would be to
have a separate clarifier taking the supplemental air units through settling
processes whereas the mechanical unit effluent went through the clarifier
and then the total BOD removal itself. I think if it is done on a separate
train basis you take the supplemental air effluent through the clarifier
you take the mechanical unit effluent to the clairfier the effluent BOD.
That process, I am certain, will be a difference in the total BOD coming
out of the clarifier. Another point that you are talking about in terms
of loading, Alexandria is designed as a very high loaded system, which
you may be aware of. The design loading was ten to eleven gallons per day
per square foot. During the pilot test the loading was the order of six-
teen pounds of BOD5 in the first stage and therefore the supplemental air
effective in Alexandria because of the high loading conditions. For low
loading conditions the situation may be different.
MR. MADDEN: I agree with you. The only point I would like to make or
restate is that when we took an absolute comparison of supplemental air,
no supplemental air, for our equipment we can look at it and say there is
two absolutely similar or absolutely exact comparisons of wastewater and
performance, and you can draw a conclusion from that. It is very, I agree
with you, that it is bad to pick up Alexandria, pick up a piece here pick
up a piece ther, mix it all in a bag and say well this is the result, and
that is not what I am trying to accomplish here. What I did want to show
is that when we took information that Autotrol published through the Purdue
1509
-------
Industrial Waste Conference, we wanted to show that there was no benefit
to air drives, and again, I go back to what I .said to Nancy before, that
this was first and foremost a marketing look. Was there a benefit of
going to air drive, was there a benefit for us to research and develop
our own air drive and with the problems that we see with air drive such
as rotational speed control problems followed with increased flow through
a closed media system, we are not convinced of the worth of it. And that
is what I am trying to tell this very profound group here this morning,
that you have all seen really one side of the picture and what I am telling
you is the other side of the picture, and it certainly bears looking at. ',
MR. DIAPER: I have two questions. Did you in any way measure the thickness
of the film and do you have any restriction on that thickness. And secondly,
did you try the effects of speed change on the growth on the media?
MR. MADDEN: First question, no, we did not measure it like Autotrol by weigh-
ing the shafts. We observed it, and on any given day, Tony, I could go
and look at it, or one of the guys who was doing the study could go and
look at it and could make an observation,, At times we tried to call some-
thing yes, there is a little bit of a lighter growth associated with the
air drive, and we can draw that conclusion at this study. But you would
have to look very closely to pick up any discernible difference in the
growth thickness or the growth characteristic and as we mentioned earlier,
perhaps a microscope study where we did in fact look at the higher life
forms to see what was actually happening there, that could be initiated,
but with the aerobic growth which was dissolved oxygen not sinking through
the second stage, if you notice there is an increase in dissolved oxygen
from the first stage to the second stage, without supplemental aeration.
And that kind of says what is happening in the RBC process associated with
our open media. There was no quote sag in dissolved oxygen wastewater. We
have never with the exception of one plant, experienced slides that you
have seen here this morning concerning the Beggiotoa growth that I have
seen. Now I have seen these types of situations and when I look at them,
I say holy smokes, you know if that happened to our stuff I would be very
concerned.
MR. DIAPER: Do you have any restriction on the film thickness?
MR. MADDEN: We have designed for .a quarter inch over the entire media
surface area. Now, we have not yet ever had that type of growth with the
exception of maybe one location where we did have what was estimated to
be three sixteenths using the same method that was mentioned by Mr. Sullivan,
i.e., you take the surface area, take the weight of the machine and calculate
out by process of assuming that the density of the growth is 62 pounds per
cubic foot. You had another question, too.
MR. DIAPER: Did you try the effects of speed variation on plant performance?
MR. MADDEN: On one project we have two speed motors and we did not try at
Cold Springs facility. Cold Springs facility does have a variable frequency
drive system that we can change the speed and quite honestly you know, there
1510
-------
are nine thousand things to study. What do you study next? And one of
them would be speed and that is going to be an ongoing study for us; the
speed versus supplemental air. The project will run through two experi-
ments and we will have an accumulated data of the effects of temperatures.
And there should be some interesting results which we will present at a
later conference.
DR. SRINIVASARAGHAVEN: What are the loading rates employed and what is
the cover effect on RBC?
MR. MADDEN: Well, soluble BOD, three to five pounds per thousand sqaure
feet of surface area at first stage loading, lip to sometimes, now one fa-
cility is taking something seven. The other facility took 8.8 pounds per
thousand square feet of surface area. Well we load Cold Springs up to
fourteen pounds, both sides. There was no Beggiotoa none whatsoever.
Another thing that has to be looked at, and perhaps some of the people like
Mr. Sullivan who is standing up in the back patiently waiting for a question,
just a minute, one of the things that perhaps should be looked at is what
effect do covers have on Rotating Biological Contactor Process. And I will
give you a case in point. Somebody visited our plant installation, and
that is a twenty-four shaft job, it is in a building that has a relatively
small cubic footage in terms of what you would normally expect foj: an RBC
building but it is fairly tight. And in that facility, these two guys
got busy. And one says I have to go out so he went ouside. And it suddenly
dawned on these two guys that there was no oxygen in that building. And
now there is a huge sign on the front door, "Do not enter unless the fans
are running". Now one of the things we are going to look at, is what is
happening on your cover; are the covers going to be oxygen limited, is there
going to be sufficient oxygen under the cover to keep this process going?
You are absolutely right. Humidity can be horrendous. I have seen some
of our own plants, and some of Autotrol's facilities and some others, and
humidity control is a problem. And that is one of the reasons why design
consultants go to covers because they eliminate this problem with air trans-
fer of heating, cooling etc. And condensation occurred inside the building.
MR. SULLIVAN: There is a number of parameters that you have on the board
there. I think it is a very, very important problem. Your making state-
ments then with regard to biofilm thickness, is critical. To define that,
biofilm thickness at room temperatures at which the organisms' growth is
critical. The soluble organic loading that you quoted for Autotrol's design
conditions a three pound soluble. The actual data that has been gathered
shows a zero water break point four pound soluble. Even that design is zero
water dissolved oxygen limiting break point.
MR. MADDEN: Excuse me, let me stop you. I forget the question or the state-
ment. We did not find that. We see a trend in that direction. But we did
not find at four was zero. We found that we were still achieving not quite
first order reaction beyound four pounds. But okay, that is a significant
point. We do not agree with you there. And that is just a question of
that, this is ours, this is yours.
1511
-------
MR. SULLIVAN:
gathered?
What are the other things; how long is the data that you
MR. MADDEN: Since June.
MR. SULLIVAN: And how many plants has it been based on?
MR. MADDEN: Well the pilot study done at Cold Springs is one plant. I
mean that is just, one-two-three-bang, that is all.
MR. SULLIVAN: I think it is advisable that a research of the existing in-
stallations on RBC, both from a carbon removal standpoint, temperature stand-
poing, carbon and nitrogen standpoint, be evaluated. Some of the papers being
presented here with regard to Ft. Knox, with regard to Alexandria, with regard
to influent problems, not necessarily generated on the media, with regard to
biofilra thickness, all play a role in the design of RBC systems. I think all
of this information has to be correlated in order to come up with a proper
design. The organisms have to be studied, biological changes, effects of DO',
effects of slime thickness. And from a standpoint of pro-motion of the RBC
indsutry I think this is critical. Yet when you gather data, it seems to me
that we do have to take biofilm thicknesses. We have to know the types of
biology present, we have to know the effects of temperature, we have to know
the desired oxygen concentration in the reactor, we have to know the biofilm
thickness.
MR. MADDEN: I agree with you. I think all that is very important. I think
though that most of that is important for you and for us and for other com-
panies that are promoting this equipment to basically consulting engineers
who have the same problems you and I have. We have to meet a payroll every
week, we have to get the drawings out, we have to contend with the EPA and
for the group that is here from the academia, they want to know what kind
of a bug is living there. All they want to know is that when Autotrol, or
Clow or Hormell, or other brands say that this is what it will do, that in
fact they will feel comfortable with that. And I think you are right. It
is incumbent upon us as manufacturers to do the work and I did not intend
that this should be the end-all be-all for RBC Supplemental Air. But one
of the things that led us to the study was in looking around at our operating
plants. We do not see this kind of problem. So then we say well maybe some-
thing is not hitting us on the head and that was the reason we started with
this study. And as I said earlier, since your firm was enamored with the air
drive and possibly because a lot of consultants do like adding a little bit
of air in the wastewater treatment system, hey that is the best end of the
world. But if that is the end of the world perhaps since Clow sells aerators
too, make the tanks bigger, throw out the RBC and put aerators in. You know
it is the other side says we cannot see 200 CFM of air, the manufactured
at the inefficiencies' in a blower system clumped in the bottom of a tank and
quite honestly, you made us look at it and we are still looking at it.
DR. CHOU: I am in response to you mentioning the soluble BOD removal at
South Shore. South Shore has about one third to a half of the BOD contribu-
tion from the industry, a single industry which is animal ..group. We found
these soluble BOD at South Shore less than the average that we have found for
1512
-------
domestic waste. And also we found that the soluble BOD removal could reach
even higher than ten pounds depending on what type of wastewater you are
talking about. So there is one factor I think you have missed, which is
the waste characteristic in response to waste treatability.
MR. MADDEN: You are right. We studied one facility, you also studied one
facility. In that facility, you do not state any place in the publication
at Purdue, what actually is happening.
DR. CHOU: No, we stayed at South Shore, exactly, and we also mentioned that
load from the industry is one third to a half. Because we have to get the
paper accepted by putting in the industry.
MR. MADDEN: Well, I understand. But, one of the things that you are missing
is, and we recognize it, I am sure you fellows recognize, is that different
wastes have different biological oxidation rates, and that industrial wastes
can move the growth farther down the flow-path, or move it up closer to the
front, hence a better reduction or a better reaction occurring on the media.
We recognize that and again I will state, this was one study, one look at one
location. We are continuing the study but it is worthwhile in our opinion
to say that we do not, we have not had this experience problem that made
such a strong point with me for this morning. And I do expect that we do have
to respond for it.
DR. CHOU: Yes, we think it is very important to address to the audience
because if we only mention the one side of the fact we could be misleading.
And in regard to the oxygen transfer, you mentioned that the structure, the
internal structure of Autotrol's product could be inferior to yours in re-
garding to oxygen transfer. I would like to ask what is the percentage of
oxygen transfer to the liquid phase, versus the oxygen uptake in the air
phase, because this is a three phase, involving the solid biomass and the
liquid phase and the air phase. So have you done any work as regard to the
oxygen transfer in each phase?
MR. MADDEN: No we have not, and I do not know, but let us go to the other
side of that. We have examined the bundles and what is growing there. And
we have pulled yours and we have seen what is growing or not growing on yours.
So all I can say is that there appears to be a difference and that is why we
are looking at this difference and we are going to continue to look at it.
DR. CHOU: I would like to address some data rather than the appearance because
your biomass looked to me like Beggiotoa - all white. It is not black but
it is all white.
MR. MADDEN: The basic media itself was white, that is to say that the poly-
ethylene itself prior to the time we started using carbon black for U.V.
addition, it was absolutely white, bone white. And the last slide that you
saw that was nitrogenous bacteria and it was a white media, so I cannot make
it what it is not you know, there it is.
DR. CHOU: Now it is my turn to present a data in terms of part of oxygen
per horsepower per hour in comparison with the surface area- I am trying to
1513
-------
make the point that the dictating transfer is in the air phase rather
than in the liquid phase. We have done some work on the liquid phase.
In response to your question, your mentioning of the ethereal structure of
Autotrol's product, I do not think that is a legitimate attack because we
feel that most of the oxygen is transferred through the air. And there is
no limiting as far as oxygen diffusion in the air. So it does not make any
difference what type, you can use a flat disc or dimples. One more contrl-
butionl That is one thing we have done. We also have learned from Japan,
the enclosure oxygen-limiting condition. They have done work on O.R.P.o This
is window area, to define what is the requirement on window area to avoid the
oxygen-limiting condition. So we have data for that.
1514
-------
Session 6. INDUSTRIAL WASTEWATER TREATMENT
Presiding: J. B. Walasek
Wastewater Research Division
U.S. Environmental Protection Agency
DR. PAGORIA: A couple of questions please. When you say 10% seawater I
just want to be sure I understand. Do you mean a hundred parts per thousand
salinity or do you mean approximately three times the strength of seawater?
When you say 10%.
MR. LANG: I mean 10% of the salinity concentration normally found in seawater.
DR. PAGORIA: So you are talking about 3.5 parts per thousand. That is good.
the temperature decline was caused by what; evaporation?
MR. LANG: Yes, temperature change from 28 to 30 degrees at sea level.
DR. PAGORIA: And then one last thing. The source of nitrogen, I am sorry
I missed.
MR. LANG: One four.
MR. CHRISTIAN: I would like to know what pH measurements here 10% salinity
corresponded to.
MR. LANG: The 10% we have pH of 6.5 to 7.8.
MISS KINNER: What was your hydraulic loading on that?
MR. LANG: The hydraulic loading we did it in meters of 100 square meters
per day. That was 6,100.
DR. POON: I could not see the data very well. I wonder if you have been
able to show if you have increased your ammonia nitrogen loading. Would that
compensate some of the inhibitory fact of this chloride? I seem to find in
the removal of organic relationship that seem to exist. For example I increased
the BOD loading to RBC, I could reduce the inhibitory fact of chloride from
the removal. Now does that apply to ammonia?
MR. LANG: The effect of high chloride concentration on organic and nitrogen
removals in RBC should be similar to the case in the activiated sludge process.
However, we did examine this effect in our study.
DR. JENKINS: What was the type of ammonia in these systems to growth nitrifiers?
DR. REYNOLDS: There was a pretty good nitrification reduction and that may
indicate some problems with the Ks and
DR. JENKINS: I was going to suggest if nitrification were taking place you
would expect it perhaps largely in the end of the disc system which might
account for your inability to develop satisfactory growth Constance based on
COD or BOD removal.
.1515
-------
DR. REYNOLDS: I think you are right, Dave. And I appreciate that comment.
One thing I ought to point out though Dave, for your information, cheese
wastes often are nitrogen limited and we had a concern about that in this
particular case. We did do some C/N/P ratios and it was indeed, carbon
limited. Your comment is well taken and I appreciate it.
MR. ATHAVALEY: When the time engineer look at your equation, you are refer-
ring to constants which pertain to the particular substrate for example COD
as discussed here. Now, if I want to know what would be BOD and nitrogen
effect on these constants and if that is the case, are these constants inter-
related or could any attempt be done in that direction to achieve those number
of kinetic constants.
DR. REYNOLDS: As long as you were talking about the carbonaceous removal it
is a fairly easy transition to go from BOD to COD or to TOC. That is the
coefficients and the constants are easy to come up with. They are, if you will,
just a fudge factor away. When you try to go from carbonaceous to nitrogenous
removal, in other words nitrification, then a whole set of different parameters
would have to be developed.
MR. WONG: Have you done any toxicity test for the RBC effluent to pond?
DR. BRACEWELL: No, we did not observe any toxicity in the pond systems so we
did not actually have the time and money for us to do a toxicity test by
ourselves.
MR. BECKMAN: Where did you collect the samples?
DR. BRACEWELL: We measure total TOC on the influence and effluent waste stream.
MR. BECKMAN: Do you have any idea what the feedbacks are going to be on this
gas evaporation system? Are they going to be better in four or five years?
DR. BRACEWELL: They currently incinerate now. But compared to what the
current costs are. It will pay back within four to five years.
MR. BECKMAN: What are their thoughs for that now. Do you think it will be less?
DR. BRACEWELL: It is possible, nothing sure. Although the cost is going up
more it can be paid back within four to five years. Originally we were so
upset, they were going to have the natural gas supply shut off.
MR. ANDERSON: Can you comment on your method of sampling?
MR. WATTS: Well we did take what were the samples available. We were able to
get one what was a sample because it rained on. The total solids most of which
are total dissolved solids from the process,. The soluble contribution from
that is not that. There may be a certain amount on the process path of separ-
able organics that will go down the sewer and there is a lot of separator as
an interval part of equalization basin. But from the present knowledge that
we have, those things would be on the process pattern up to decide and if
1516
-------
they were, they would go down the sewer and their part of the material balance.
In other words say there were more wash down and things like that in material
balance in the original dry water data.
1517
-------
-------
Session 7. CONCEPTS AND MODELS
Presiding: R. D. Miller
U. S. Army Environmental Hygiene Agency
MR. HYNEK: I noticed your logic in reducing the amount of surface area in
successive stages and I believed you said correctly a fifty percent reduction.
And then you have said in the trail which provided successive less oxygen.
However in your case you have cut time, in their case they have not. You have
indicated no recognition of the changing degradability factor in a complex
waste like municipal waste. The easy stuff goes fast, the last five parts
go slow. Could you comment on that please?
MR. STEINER: The reference to the Minneapolis application was for the ex-
clusive purpose of showing the disc as a mass transfer device. That is to
say that as more treatment is achieved there is less aeration required in
the successive stages and therefore, the conclusion was made that with
respect only to aeration that the succeeding stages could be made smaller.
The intention was not to compare the residence time with any other factors.
With respect to your comment on the decreasing Q value or the lessening of
treatability in successive stages, I certainly agree that the top BOD comes
out a lot easier than that which .was remaining. If you were to plot however
the data generated at a place like Pewaukee, you are not getting down to two,
three, four and five parts BOD you are up a little bit higher than that. You
will find that the existence of refractories simply does not make itself
known when you go through plots of data. I do hope it is heredity of the
data does not seem to get to the program correctly.
MR. HYNEK: You have used the applied ammonia per thousand square feet with
indication of over tweny-five pounds. Is this correct?
MR. STEINER: That is correct.
MR. HYNEK: In your paper, the units are somewhat off and I would request or
make this question; do you have data to back up your position or as to your
logic?
MR0 STEINER: Yes and I will refer to Dr. Borchardt's EPA paper for the
data that will substantiate the curve. Also the data for these from Chicago
study on nitrification as well as others.
MISS BERGS: You said the RBC design is based wholly on organic loading. Is
there any minimum of retention time that you would have to include hydraulic
loading?
MR. STEINER: Our work with detention time was well over one minute or less
than a minute I cannot say. But the detention time was down to one minute.
I have found no significance of detention time with respect to removal down
to one minute detention time.
DR. O'SHAUGHNESSY: I agree with your statement that the RBC treatment plant
design should be based upon organic loading. However, for nitrification, our
1519
-------
early work indicated that decay of ammonia nitrogen as defined is ineffective.
But when you change your surface to volume ratio on your reactor it will have
an impact on nitrification kinetics.
DR. WU: I would like to learn some basic knowledge from your excellent work.
What you have mentioned in your presentation is the attachment of the cells
on the surface of biodisc is due to the formation of polysaccharides. If
this is true, how do you make a suggestion to the process design engineer in
regard to the optimization of cell attachment in the fixed growth RBC system?
DR. LAMOTTA: We do not know yet. We are going to perform more research work
in this area so that the optimization can be achieved. However, the objective
of this paper is to study the substances which can be used to precoat the disc
surfaces. As a result of the application of surface coating agent, we can get
an immediate cell attachment and the acclimation time required for biomass to
develop on the disc surface will be significantly reduced.
DRe WU: As far as the effect of magnesium limitation on cell properties is
concerned, the other studies indicated that the production of cell polyssac-
charides increased as the degree of magnesium restriction also increased. Can
you comment on your experimental results that show the cell attachment on the
disc surface reduces as a result of the reduction of magnesium content in the
wastewater?
DR. LAMOTTA: You said that the addition of magnesium into the wastewater is
not going to improve the production of cell polysaccharides. It may be true
but however, the addition of magnesium in the present study is for the enhance-
ment of cell bridging and/or attachment to the disc surface.
DR. UNZ: The addition of magnesium and calcium to the wastewater is a good
idea in terms of the improvement of cell attachment. I think calcium is much
stronger than magnesium in this case.
DR. MUELLER: You have got a real .nice .definition .and I ';agree' .with it a hundred
percent. I guess it is a philosophical question of when you take existing
knowledge perceptive and apply effectively to get what you can out of it and
when you continue to do research. I think applying the first principle we
have today making some assumptions about the reaction coefficients, and I am
sure I will show you again this afternoon the additional knowledge as to what
some of the important factors are and where those research efforts should be
directed. :
DR. GRADY: I cannot say that any better, Jim.
DR. BOON: I have two quick questions. One is concerned about the addition of
lime to RBC sludge for the improvement of solids settlability. Does the chemi-
cal addition impair the sludge digestability? Secondly, do you have any problem
in the winter, any calcium deposits because of so much lime in it?
MR. NOSS: A friend of mine is now doing a study on the digestion study of
the sludge produced. He is also going to be incorporating into the paper the
amount of sludge produced, the characteristics and dewaterability of RBC sludge.
1520
-------
That paper will probably be coming out about the end of this year. But I
have not attempted to put any of that data into this particular presentation.
However, it does not seem to be any problem with handling the sludge and when-
ever our sludge production is not all that great compared to what we would
normally see, in other words, with the increase in the amount of secondary
sludge produced when it is mixed back with the primary sludge we did not
really have any problem in handling it. To answer your second question, we
have no problem with calcium deposits. We did not see anything falling out
into the RBC. To be truthful about it, the use of lime to form the calcium
carbonate which put down the phosphorus and the appetite in the primary clari-
fier. We have no evidence of any residual built up in the RBC system. We
will further investigate this to see any build up before the end of the project.
MR. SMITH: What temperature range was used for your experiments?
MR. NOSS: We ran this particular experiment inside one of the buildings at
Fort Detrick. Therefore, at lowest temperature was fifteen degrees because
the building was heated in the winter. The temperature range was between
fifteen and twenty-five degrees.
MR. MEANS: In the first part of your presentation when you talked about
feeding the flow into the four trains; do you know the application of alkal-
inity was prior to the first train and after the last train in the pH change
from 9.5 down to about 7?
MR. NOSS: We did the standard calculation on alkalinity destruction with
respect to ammonia removal and we measured alkalinity ranging generally
from 200 to 250 mg/1. We were seeing essentially the magic number of seven
miligrams per liter of alkalinity destroyed per miligram of ammonia removed.
MR. BACHTEAL: From your experience, those which you cannot write about, but
at least you can talk about in the abstract, would you prefer to see staging
or the loading rate such that one module would not handle the load? Would
you prefer to see two or three modules or four modules in one stage or alter-
natively to split the flow in two, three, or four streams such as you fed
25% of the flow to module A which would be a first stage module and 25% to
module B which would also be a first stage module?
DR. FRIEDMAN: I have mentioned that question before. My initial reaction
to that question would be I hate to put all my eggs in one basket so I would
be in favor of splitting the streams. If something goes wrong with one of
them, we do have and will have in the future drive problem, I would rather
not have all my eggs sitting and waiting on in stage 1 and they should be
distributed to stages 2, 3 and 4 behind it.
MR. SUTTON: Can you explain a little further your current data that you had
on influent versus effluent in long term BOD and you related that to an
existance in the effluent BOD concentration much greater than the influent.
And I believe you were relating it to hydrolysis occurring in the reactor
and therefore getting a BOD would not show up in the sample of the influent.
1521
-------
DR. FRIEDMAN: This is a particular wastewater that is incredibly high in
protein but as such as the total protein which I call the readily biode-
gradable, I was very careful to call that when I started out. That disa-
pears very quickly from the RBC. No further treatment or reduction of
either COD or BODr with the rest of the six stages beyond second stage was
found. Subsequently, because of this we had very strong evidence but not
complete. We ran a complete protein test and the result of our lab tests
showed that at least 80% of soluble COD was protein. It is obvious from
this study that we are not breaking down all protein despite every opportu-
nity to do so in the RBC system. And yet it did break down under a long-
term BOD test. That is the only part I was trying to make. Eventually
it is going to break down and it is going to be in the receiving stream.
MR. SUTTON: Why did not it show up on your effluent sample as well? The
influent BOD sample did not behave the same as your effluent. I do not
understand.
DR. FRIEDMAN: This might be what you would call a refractory component of
that. Refractory is not a good word in terms of BOD but a more difficult
portion.
MR. SUTTON: You feel that it is because of the shorter retention time of
the RBC plant that dictates against it in this case?
DR. FRIEDMAN: Well, it was not talking about the same kind 6& the waste-
water and running into a suspended culture system and giving a very long
mean cell residence time. My personal opinion is I have no proof of it,
that after the protein bonded on the cell surfaces, it takes awhile to
break them down. When we have a large amount of sludge with a very long
detention time it has an opportunity to do that. When you take the same
concentration of the wastewater and one of these systems you do find the
breakdown of this protein.
MR. IANNONE: You briefly went over the carbonaceous part of your talk. I
was wondering if you feel that the carbonaceous portion of your model is
verified at this time, and are you designing systems or would you design
systems, in accordance with your model prediction? Before you answer that
question we have taken some of your data and compared it to existing design
curves and you seem to continually fall below the predicted curves.
DR. MUELLER: Yes, I would use this to design and this is the only way we
do design RBC systems. However, our analysis is still a specific plant and
still a specific waste. On the initial carbon work was done on industrial
wastewater two different types of paper mills. We also did it for domestic
x^astewater and we got a different K rate. The K rate is a function of the
wastewater we are dealing with. The nitrogen work already anticipates that
the K rate or growth coefficient, the nitrifier, should be constant. The
growth coefficient with the carbon will not be constant. That is a function
of the wastewater. As Dr. Friedman mentioned we do not have a system here
which we can just take a design without putting some linfits on that design.
1522
-------
Our results in the model simulation vertified the results for that giving
wastewater. I certainly wish anyone in this room to feel they can take my
result and use it for designs in their system. If I do not have data on
your system and a -K rate from your system, I cannot use my model unless some-
body uses somebody else's K rate. How does that simulate your rate? It
beats me.
MR. IANNONE: Do you think the K rate for domestic wastewater is going to
vary that greatly?
DR. MUELLER: To a large extent, yes.
DR. LAMOTTA: Does the kinetic behavior of the biological solids differ
between the fixed growth and the suspended growth system?
DR. MUELLER: Yes, we have noticed some differences in kinetic behavior
between the fixed-film RBC system and the suspended-growth activated sludge
system. According to our experience, the kinetic coefficients resulted
from carbon work differed only slightly but a remarkable difference was
found in nitrogen work.
1523
-------
-------
Session 8. UPGRADING PRIMARY AND SECONDARY WASTE TREATMENT
SYSTEMS WITH RBC
Presiding: B. J. Garg
Department of Environmental Resources
Commonwealth of Pennsylvania
MR. SULLIVAN: Just some basic questions with regard to clarifier efficiency.
In the course of the evaluation was a long-tube settling test conducted? The
clarification rates for secondary solid separation that you suggest in the
paper on the order of 500 to 600 gallons per day per square foot; recent pub-
lications have shown around 1500 to 2000 gallons per day per square foot to
reach effluent 30 solids. Could that be a function of the turn-around section,
or were there two settling tube tests conducted to determine actual solids
drop velocity?
DR. MUELLER: In that section which as you know was taken into account in cal-
culating overflow, the actual overflow rate was about 500 even 450 at this
point, which is varying slightly. They did conduct a tube settling test, to
generate that curve but the data that we applied to this was the data from
the five main field tests, that would be out of solid data from the field
testing itself.
MR. SULLIVAN: I guess the question is, is there significant difference from
the actual field test on solid settling velocity to long-tube settling test,
and could that be a function of clarifier turn-around sectional modifications
of the physical 'configuration of the system?
DR. MUELLER: I do not know if there was any difference between the two and
what they found. I presume that but I cannot agree with. Again I think you
put the very long tub properly and we knocked out 25 percent of the remainder.
Why the solids did not settle as well as what you found in the field I do not
know. There is a piece of data which represents the result. Theoretically,
except for that the clarifier function is almost based on solid settling. I
really cannot answer that what is the apparent difference.
MR. SULLIVAN: One other question in design basis. Your final design selection
is based on a 130 soluble BOD. What was the input from the industrial applica-
tions on that?
DR. MUELLER: That is a good point, I have not got the foggiest. You have an
idea on that?
MR. SULLIVAN: Well, basically what we are doing is designed into a peak
monthly and because of the industrial application out there you have a high
organic surge that is about seventy percent during the peak month over the
average month and hydraulically about fifty percent during a peak month over
the average month. So your design basis is about 2.3 times the average monthly
condition and the economics reflected peak month design that is 2.3 times the
surface as a result forces you out of the existing primary tanks.
DR. MUELLER: I agree with you yes, no doubt about it.
MR0 SULLIVAN: And it has a significant impact on economics, factor of two.
1525
-------
DR. MUELLER:
on that.
No doubt about it at all, I agree with you a hundred percent
Was this the installation
MR. DIAPER: Jim I enjoyed your paper on Edgewater.
where we did the micro-strainer tests?
DR. MUELLER: Not in conjunction with our study.
MR. DIAPER: I remember the consultant Dick Tolbert rented a small unit
from us and took it up to Edgewater and ran some tests downstream of the RBC.
The point I wanted to make was that if you were looking for a solution to the
removal of solids the micro-strainer would be a very effective on because the
RBG produces a stringy type solids that is easily filterable. In fact we have
got flow ratings of twenty GPM per square foot, and drop solids from a hundred
to the low ten on a single pass. The installation of a micro-strainer there
for 3.6 MGD would need about two ten foot diameter by ten foot long machines,
and the interval costs would be about a hundred thousand dollars. ;
i .
DR. MUELLER: That is a definite possibility, instead of going with s'ay a
sand filter on the end to get the effluent quality you do a micro-strainer.
You might be in better shape.
DR. SRINIVASARAGHAVAN: The ammonia and BOD loadings throughout the study
have been pretty low compared to many of the other studies and I understand
why. But that kind of a loading where there is absolutely no limitation
oxygen transfer especially in a small scale unit, should there be any consid-
eration for scale-up at all? Why does the scale-up factor? And the second
question is you said the soluble BOD may not be the best parameter to monitor
the RBC performance. So I would like for you to comment on this BOD consid-
eration for RBC design also.
DR. POON: Well, first on scale-up, I really have no... I really am not sure
what should be the scale factore. Therefore I said if the manufacturers
suggest a certain range I would feel confident that you can go down to the
lower range off their suggested scale factor. I do not mean to suggest this
would be the scale factor recommended to everybody. And the second question
now is concerning my question on the soluble BOD parameter. Of course the
reason for people using the soluble BOD is that they consider the RBC only
removed soluble BOD and the particular one would probably go through intact
because the removal would be more effected by the clarif ier in the RBC system.
Also a look at my data, I thought I might have a transparency at hand to !
show if you take grab sample when random sample daily, a lot of times, a lot
of particular BOD in the trickling filter effluents coming out and go through
the RBC system, but they do not go through the system intact. And sometimes
the particulates BOD incorporate into the biofilm, but a lot of times there
is a lot of sloughing out of the biofilm contribute to the particular BOD in
effluence before it reaches the clarif ier. So I have data to show many times
negative value, in other words no sloughing at all, and actually build up
biofilm look at that particular sample. And so I have negative value, zero
1526
-------
value and positive value for sloughing. But what I am trying to say is par-
ticular BOD does not go through the RBC system any time. That is the internal
changes within the system. So I think that soluble BOD may not be the best
parameter to monitor the performance of the RBC.
DR. SRINIVASARAGHAVAN: The plant performance has a correlation with the
BOD; would it be any better relationship between the performance and the
loading?
DR. POON: I do have the curves using total BOD. I will not say it is better
but I could use that data for the design also.
MR. STRATTA: I notice you presented some data on mass of biofilm per square
centimeter. Will you describe how you got that data, scrapings or what have
you, and where it came from the discs?
DR. POON: The method is simple. Just scraping off the amount of biofilm
from the known surface area and dry it and we weigh it.
MR. STRATTA: Was it representative of the disc?
DR. POON: I would not say it is representative of, that is why I say we do
the best. I notice that the biofilm is not even within the same stage and
as I have listened to many speakers that in the modelling of the RBC they
assumed uniform thickness throughout the RBC and that never happened. If
you go through the RBC facility you do not see any uniform thickness at all.
MR. STRATTA: I listened to your paper with great interest. You did not
mention much about suspended solids. Were you trying to reach a standard
on suspended solids?
DR. POON: Our suspended solids concentration is quite low, most of the time
below 100; I would say between 50 and 100.
MR. STRATTA: Were you trying to reach a standard on suspended solids?
DR. POON: I was saying that the average suspended solids coming out of RBC
clairfier was 15 mg/1.
MR. STRATTA: Fifteen?
DR. POON: Yes, so I allow under the worst part of a siuation another 15 mg/1
BOD contributed to the final effluent.
MR. DIAPER: The other question I had was I saw your picture of Philadelphia
and I heard that there was a problem with Philadelphia. Was there a failure
of several shafts?
MR. McCANN: There was some problems with the shafts at the installation, I
believe one or two of them had problems.
1527
-------
MR. DIAPER: Could you expand a little on the nature of the problem?
MR. McCANN: This question has been raised earlier in the conference and
I think we addressed it at that time. There was a problem with some of
the shafts. We think we resolved the problem and our testing shows that
the shafts' current design is adequate and there is no problem with it.
The City of Philadelphia's solution, they have resolved the problem in
their own mind and they have no problem about the use of the equipment.
MISS KINNER: Do you have any problem that has an RBC system before and
after they have activated sludged?
MR. McCANN: I believe we have one or two installations. And when you say
roughing, what usually happens is that the nitrification upgrades and the
roughing becomes a symmetric because what usually is required is secondary
treatment and the activated sludge is converted over into a nitrification
treatment system.
MISS KINNER: If you got your need in nitrification, would there be any
problems in our loading the RBC in activated sludge basin?
MR. McCANN: No, the system would be designed, the engineers' requirements
would usually be that the effluent from the RBC system be such that it would
provide sufficient waste to maintain the activated sludge system. In the
RBC system design we would attempt to maintain the loading so you did not
have an overload condition. So it really depends upon wastewater condition.
MISS KINNER: Are those two things compatible?
MR. McCANN: Yes.
MR. WARD: What I am wondering is whether you have had any experience follow-
ing your RBC process with a tube settling module. This is what we are doing
currently and we have come up with a situation where the RBC has the selective
organism which adheres tenaciously to plastic. This follows through to the
tube settling modules and we get a very heavy growth which eventually begins
to ferment and produce gases and cause the modules to rise to the surface. I
was just wondering if you have had experience with this type of a problem in
other installations?
MR. McCANN: Again, the tube settlers for whatever reasons are seldom utilized
in our installations or I would presume in most biological treatment systems.
We have done some experimental work with it. Usually the tube settler is
utilized for a fine tuning of the effluent. 'Usually it comes from a secondary
clarifier and you normally will not have the heavy growth build up. Once you
start to develop the growth some means of removing it, usually just dropping
the liquid level, causes the growth to slough off. Our experience has been
that the growth is not that tenacious as far as its clinging to the settlers.
MR. WARD: We tried a combination of chlorine application immediately before
the tube settlers in order to reduce the growth in addition to periodical
flushing and it is such a strong growth at a speed that we can drop our tanks
1528
-------
while there is not enough to provide a shearing force to really do any good
to begin with it. And I was just wondering if you had any other...
MR. McCANN: To my knowledge, again, we have only utilized the tube settlers
following a clarification procedure so that the loadings you are talking
about, the concentrations onto the tube settler are on what range? What
range are you using?
MR. WARD: Entering, or a secondary clarifiers, we have a fifty parts per
million but most of that is settleable, so that leaving the tube settler
modules we have about ten parts per million of BOD or less and comparable
suspended solids in a ten to fifteen mg/1 range. But the growth that we
get would amaze you at how thick and strong it becomes. Every piece of equip-
ment following the RBC process. It is not a problem on the walls and the
flight guard and so forth, and our clarifiers we do have a problem with the
tube settlers and we were just wondering whether there really has been anything.
MR. McCANN: Not to my knowledge.
MR. ODEGAARD: I have two statements regarding the separation of the sludge.
In Norway there are several plants that use microstrainer as the separation
of the biodisc sludge. Is not the same type that you use here in the United
States? And about the sedimentation or tube settler used for military instal-
lations sludge in Norway, and there have been considerable problems on disc
sludge with Nomella. This problem has been solved by using a vibrator that
vibrates very gently the Nomella.
MR, WALL: I noticed you have a plant in Ontario and one in Pennsylvania
and one in Peoria, Illinois that had the fiber glass covers in sort of a
winter climate. Do you have problems with the doors condensing and freezing
shut in these cooler climates on the fiberglass cover?
MR. McCANN: We have not experienced any problems with doors freezing in cold
climates at all. The interior of the ambient temperature underneath the
cover usually assumes the temperature within one or two degrees of the water
temperature so that there is usually no problems with a freeze—up in the cover.
MR. DIAPER: I am interested in the solids removal from the RBC. As I pointed
out earlier the microstrainer has been used for that purpose. What was the
system you used at Cadillac, Michigan?
MR. McCANN: Cadillac, Michigan uses a mixed media filtration system.
MR. DIAPER: Does that go onto the RBC without an intermediate clarifier?
MR. McCANN: The filtering consists of activated sludge followed by secondary
clarifiers and then during the summer operation for nitrification RBC and
then directly onto the sand filters.
MR. DIAPER: What sort of removal efficiency are they getting?
1529
-------
MRo McCANN: As far as the solids?
MR. DIAPER: The suspended solids.
MR. McCANN: The suspended solids come into the system usually vary only
within one or two parts of the incoming suspended solids so they very seldom
exceed ten milligrams per liter of suspended solids coming into the RBC and
onto the sand filters. So the loading extremely low but the solids coming
off the sand filters are under five milligrams per liter.
MR. DIAPER: Many years ago when Autotrol was first getting into this business,
I did some microstrainer tests with Pewaukee and we had 100 mg/1 coming into
the microstrainer and about ten going out without any intermediate clarifier.
I wondered if you have given any consideration to the combination of an RBC
and a microstrainer?
MRo McCANN: We are not in the microstrainer business, so we are not usually
really concerned with what is following our system. The engineer's choice
is to what he will use. Any opposition using microstrainers is not ours.
MR. ANTONIE: The question that was asked before with regard to discs before
activated sludge plant. We did not report on it because there was not enough
data available, but there is a plant in Crawfordsville, Indiana with discs
prior to activated sludge and there is a nitrification plant and very specific
with regard to the soluble BOD desired to the nitrification system. In the
capability of bypassing a number of shafts and actually putting on surface
area against the load to reach an effluent concentration designed hydraulically.
For instance, if you start up the plant half design you have the capability to
use half the number of shafts or a fourth the design with three-quarters. So
generally if you are looking for a design soluble concentration with the acti-
vated sludge system on suspended growth, you should hydraulically design flex-
ibility into it so you can control the concentration.
DR. POON: When you use RBC in conjunction with your trickling filter, should
it be put in front of the unit? In front of the trickling filter or after
or in parallel? And if you do have an opinion one way or the other, what is
the reason?
MR. McCANN: The utilization of the RBC is in a trickling filter application,
usually the most confining of the two activated sludge trickling filters. To
my knowledge, we are not, they are seldom if ever utilized in front of a
trickling filter. You would usually not utilize a trickling filter or a
final tuning of the effluent criteria,, We would more likely to be utilized
following the trickling filter for extremely fine effluent, usually either
for nitrification or again for extremely low effluent criteria BOD. As far
as the parallel series, we have not...again it would depend upon the hydraulic
level. I have seen requests for both and seen them operated as both. And it
really just depends upon the client, we have really no preference either way.
1530
-------
DR. SRINIVASARAGHAVAN: Could you translate some of the data plotting that
you show for organic and ammonia loading rates in pound?
DR. MILLER: We did some of it at one time and the data is available.
could do that but, no I cannot right now.
You
DR. POON: I did some calculations. I am really asking two questions.
Use of RBC we do not anticipate any filter fly but in my case I put the RBC
right next to the trickling filter and the trickling filter has the filter
fly. As a result, I have a lot of filter fly in the RBC system. Did you
experience that?
DR. MILLER: No experience to problems with filter fly in the trickling
filter. We have problems with sludge worms. We went over to get what was
identified again, visible observations where there were sludge worms and ac-
tually this problem even came into the RBC process itself. At one point it
looked like red meat balls in the RBC stages. I have not experienced on a
full scale or not, but again it was a very short time in the heat of August
with us, for about one month and it did not seem to have any actual effect
on the process natural conditions.
DR. POON: What is your opinion of lime chemical addition to control pH and
of increasing nitrification rate in the recarbonation system. In very cold
regions would there be any problem with scaling at that position with calcium?
DR. MILLER: I really do not think so. Mr. Stratta of Penn State University,
he is continuing to work on chemical addition. Currently he is working on a
doctorate at Penn State and the specific thing he is working on is chemical
addition in the RBC process.
DR. POON: I saw the slide in Dr. Friedman's presentation yesterday. I think
in the recirculation area they do show a lot of calcium deposit on the first
and second stage of the RBC. So I just thought that in a very cold region it
could be a problem with lime addition.
1531
-------
-------
Session 9- DESIGN AND OPERATION
Presiding: M. F. Saunders
Department of Civil Engineering
Georgia Institute of Technology
MR. OPATKEN: The value of KJ-OQ is related to the viscosity and therefore
that is saying that if we get an increase in viscosity we are getting an
increase in mass transfer. I find hard to correlate those two.
MR. SEVERIN: As it turns out there are several ways you can plot this
data, and I have been told that I should use a Schmidt number and you
can manipulate a Schmidt number which is a Ki. Rather than getting into the
nittygritty chemical engineering of this thing, I would rather point out the
major lesson I think is that not only do we have to look at the different
media designs but I think geometry is playing a very important role in the
oxygen trnasfer.
DR. REYNOLD: Would you please get into a little more detail on how you came
up with emperical dimensionless number that you were plotting that is the
N__. Why did you include the variable that you did as opposed to some
other set of variables that could have been picked. This sort of relates
back to Ed's question about fluid viscosity. How do you get your things
together and you made it fit and I wish you would go into just a little
more detail.
MR. SEVERIN: We found the N,,,, value as given by Dr. Wu in a recent paper
of his in Water and Sewage Works about one year ago, where you see that
dimensionless parameters to ascrible mechanical mixing and how he finds
the oxygen transfer constant. So after finding it we were looking for a
number that would be the dimensionless, and involved some of the parameters
we were looking for and it happened that we saw Dr. Wu's paper and used
that dimensional scoop.
MR. STUDEE: One concern I have had about a potential problem with the air
drive is that it might be foaming problem in the RBC contactor unit.
Evidently you did not have that problem here from what you have said, but
I just wonder if you considered that and if anybody else in the audience
who has had experience with that.
MR. HYNEK: Yes, we did. My concern with the RBCs back before we actually
designed this plant, had to do with foaming was one of them, grease
formation was another. That was part of the reasoning for aearting the
primary clarifier. We thought by putting air in the primary clarifier you
could remove partially the grease that might be in the wastewater and get
it to come out in the skimmers. I was concenred a little bit about foaming
particularly because the plant is underloaded. And wastewater has a ten-
dency to foam more with addition of the air.
MR. JAFFER: I was curious if you have had any observations on rotational
speed as far as eveness and how it might be corrected if it is uneven.
1533
-------
MR. HYNEK: We have been able to adjust but it is not easy to adjust the
balance because once you change one of'these butterfly values affects all
the other butterfly values' settlings. So it is a little bit of a frust-
rating operation to try to achieve any one particular RPM because it is
a matter of continuously going along and adjusting the butterfly valves.
At the present time through we have been able to, you have the second set
of RBC, the rotational speed at the present time, I am just reading this
to you, on the first stage is at 0.95, on the second stage it is 0.8
and on the last stage it is 0.62. On the first stage now which is rece-
iving all the load, the rotational speed is 1.7 on the first stage, 1.2
on the second stage and 1.0 on the third stage. So we have been able to
achieve that with a little bit tampering around actually with these butter-
fly valves. Once they are set though the thing remains exactly that way.
We have not experienced an unveness, so that the things go, you know, flop-
flop-flop; have not seen that at all, they run perfectly smoothly.
MR. HANKES: I am a little bit confused as to the data and the comparsion
of horsepower drawn on the air drive versus perhaps mechanical drive.
You mentioned units being taken out of service and operating at a slower
speed. I wonder if these considerations are taken into accounted when
making a comparison between mechanical and air drive?
MR.BERNER: The way we got the air drive on the particular plant, the plant
was originally designed for mechanical drive. An the set-up was 7% horse-
power drive motor with pretty much of a standard Autotrol design. Autotrol
came along and said we will supply the air drive for this plant at no
increase in cost to you and furthermore we will guarantee the thing, so
that if for reason our air drive does not work we will take the air drive
off and put the mechanical drive units on at no additional expense. From
our point of view that looked like a pretty good deal because we had an
irritating problem trying to get the molds for the mechanical drive above
the flood elevation. We had long chains, and it was just kind of a micky-
mouse arrangement. So the air drive was a realy advantage to us, it did not
cost us anything. On the 40 horsepower motor that drives the six RBCs,
we have actually throttled that down to a place now where it is pulling
33 amps or 25 horsepower. The full load ampage would be 52 amps. That is
4.2 horsepower per shaft. And that does not count we have got two
air headers going from that single blower to aerating the primary clarifier,
so we do not exactly how much air that is. However, we think we can get
this down to 2-3 horsepower per shaft.
MR. HANKES: Part of my confusion comes from the fact that yesterday we were
discussing, I believe that the required horsepower on mechanical drive on
the order of 3-4 horsepower operating at 1.6 EPM. We also mentioned that
the horsepower versus feed was perhaps an exponential realtionship, and
with these units operating at power would be that you are comparing against.
MR. BERNER: Well I do not know what the actual mechnaical horsepower would
be. I think my main single concern with the air drive unit would be the
restarting of a unit which had somehow sat for a long time maybe because of
power failure, and I think that a lot of the horsepower that is in the
mechanical drive units is put in there just, you know,- to be able to restart
the thing if it happens to be down for a period of time. But the disadvantage
of that course is you to pay for that horsepower.
1534
-------
MR. HANKES: One final question. Do you have provided any supplemental
means for starting the units other than air, if they are down for an
extended period of time?
MR. BERNER: I think you were down for a long period of time we would
have to reset the system and restart the system again. We do not have
any other supplemental means.
MR. HOEFLE: Did I hear you characterize a shaft length as twenty feet?
MR. HYNEK: Twenty Feet.
MR. HOEFLE: They are twenty feet length, not twenty-six?
MR.HYNEK: No, that is right, twenty feet.
MR. GROVER: You indicated you have got a six stages unit and indicated
that when nitrification starts up in the summer, so it takes awhile to
develop that culture. Have they followed that culture, have they followed
the build up of nitrites and then conversion of nitrites and get some
quantitative estimate of the time period it takes to get nitrification?
MR. BERNER: No, not specifically at Gladstone.
MR. COVER: And where does nitrification occur within that series of reactor
system? Where do they see it picking up?
MR. BERNER: They see it picking up in the fifth and sixth stages.
MR.COVER: And not prior to that?
MR. BERNER: No, not to my knowledge.
MR. STUDEE: Would you company be willing to provide an order with say a
five year guarantee on structural failure of your unit?
MR. HYNEK: We have on a selective basis yes. As a matter of policy no
we do not, unless everybody is required to provide a bond. We find our-
selves decidedly disadvantaged at times, since we do not have five years
experience with a multitude of failures we find ourselves penalized
because we do not have that experience. And so a consequence we are many
times required to provide a bond in lieu of the experience. On a selec-
tive experience we have. We would look at each individual job and decide
that. As a matter of corporate policy we perfer not to. We do however
we provide process warrenty.
MR. LUND: You have said that you are aerating the primary clarifier so
as to enhance grease removal. What type of equipment do you have then
ahead of the plant?
1535
-------
MR. BERNER: Grit removal is just a long rectangular channel with mechani-
cal sludge collector. Actually our design of these primary clarifiers
takes into consideration several things because of the infiltration in the
sewer system, the clarifiers are oversized, what that means is though that
in summer we had a low flow period of time like during July or August, we
had a very long detention time in those clarifiers. In order to keep
the clarifiers from going septic and that sort of thing, the aeration
in the center feed-drive clarifier does a nice job of that. So we get
a lot of benefits from aerating the primary clarifier, it is our favorite
design.
MR. HYNEK: One minor correction. All of the consumption measurements
you referred to in 1975 were done by an independent electrical engineer and
subsequently these were found to be error. We had sponsored those initial
prior measurements to get the field data and subsequently did this at other
locations and found that the prior measurements were dramatically lower than
those at Glastone. We then went back to Gladstone about two years later
and made measurement again found they were about half of what the original
measurements indicated. So rather than a consumption of approximately four
horsepower per twenty-foot long contactor, it was really about two or two
and a half horsepower per unit. So the total energy consumed by the six
contactors is about twelve to fifteen horsepower.
MR. BERNER: I should appreciate it Mr. Hynek if you could send that data
to me in the near future.
MR. MORGAN: You indicated the nitrogen removal problem was caused by pH.
I was wondering did you test it in terms of alkalinity being the controlling
factor?
MR. HYNEK: Well, the inflow alkalinity is approximately 150 mg/1 and of
course the alkalinity is a good indication of the nitrification capacity
but it is not all. We have noticed that approximately 7 mg/1 decline for
each mg/1 of ammonia removed. And so after the removal of ammonia in the
first stage of the RBC, the alkalinity was reduced down to less than 100
which is the beginning of the subneutral pH. So I wonder if whether I have
answered your question?
1536
-------
Session 10.
Presiding:
NITRIFICATION AND DENITRIFICATION
E. J. Opatken
Wastewater Research Division
U.So Environmental Protection Agency
MR. ODGAARD: I am not going to question but I would like to show you a graph
which is very similar to what was shown right now. These are the figures that
I presented earlier in the conference. What I wanted to show is that the
nitrification rate that was found in this study is exactly the same nitrifica-
tion rate as I found in my study. Here is given as milligrams per square meter
per hour. You gave yours in grams per square meter per day. When I calculated
these I found four and that was the same as you counted. Another thing I wanted
to show is that you showed that you got denitrification when you recycled.
If you want to have of course good denitrification, you should use like this;
you should use an anaerobic reactor before because then you can have the total
denitrification as shown in this graph where the total nitrogen removal is shown
versus the recirculation ratio. From that balance the possible nitrogen removal
will be R divided by R+l. And these are the results of the study which fits
this plot quite well.
MR. HYNEK: Are you doing any microbial study now with the RBC system?
DR. LONG: We have not. I guess the best way to answer the question is, part
of our protocol identifies the kind of microbial work we anticipate doing.
At the present time we have not specifically incorporated in that protocol
into account, but it is anticipated that we will be looking at those in associ-
ation with a nitrosonomas and nitrobacter. If the answer is yes we will be
attempting to do that also.
MR. VESIO: Do you intend to try to use some high concentrations of ammonia
and test those reactions on the fixed film?
DRo LONG: Our present thrust is towards domestic wastewaters, if we have
enough funel, we may be making it higher concentrations.
DR. SAUNDERS: What kind of pH control system are you going to work with and
what range or ranges of value are you going to investigate?
DR. LONG: Our efforst on this study are a wide spectrum of pH initially in
terms of acclamating the film. We are going to have from six all the way up
to about nine, that is our initial efforts. Based upon what we see, in that
first phase, we will then attempt to zero in on an optimum range or an optimum
point. But we are going to wait until we see what the results are of our first
phase and at that point make a judgment as to what we are going to be shooting
for. In response to your first question, we have been looking at alternative
chemical additions and possibly even aeration, various points of addition,
perhaps even combinations. Again that will be determined a little bit later
too.
DR. SAUNDERS: Are you going to control pH in each fill or just pH in the input?
1537
-------
DR. LONG: One of the configurations being envisioned now is to take a look
at pH control coming into the influent, an alternative maybe also in stage
three. We have now identified ts control pH control in one, two, three and
four, that might not be too feasible. Right now we are anticipating, in an-
ticipation of two schemes in terms of points of addition. We may find as we
progress with our work that that does not quite do it and we might have to go
to more points or perhaps just the first point might be satisfactory.
DR. MOLOF: The question of how much did these aeration or pH adjustment affect
nitrogen balances in determining the activity of the biological oxidation and
nitrification.
DR. LONG: We will attempt to identify the nitrogen balance that occurs and
that loss which might be due to elevate pH in the stripping.
DR. HAUNG: Now what kind of medium do you use to control pH?
DR. LONG: There is no state of the method for identifying nitrosomonas and
nitrobacteria. However the most commonly referenced media that is used is
that of Dr. Alexander out at Cornell which he published in A.P. Black micro-
biology series about a decade ago. That is in essence both the media and the
indicator that we are using. In addition to that we have made remodification
to his media. We are adding a metal solution which he does not identify, but
we feel is perhaps appropriate.
DR. YU: You are talking of adding addition of metals.
do you use?
What kind of metals
DR. LONG: I can provide you with that perhaps after. I do not have that
list but there is about five metals in the solution that we add.
MR. ANDERSON: Would you differentiate between nitrite and nitrates?
you did, what test would you use?
If
DR. O'SHAUGHNESSY: We spot checked for nitrite and nitrate.
more than one tenth of a milligram per liter.
There was never
MR. ORWIN: I was wondering if you checked the cost effectiveness of increas-
ing your volume of surface ratio. You said you got better reaction with
everything but the greater volume. But that is going to cost you money. Do
you know whether it is worthwhile doing?
DR. O'SHAUGHNESSY: Let me cite this just a minute before I answer your ques-
tion. At the point four gallons per square foot of a rectangular reactor with
the RBC, we put a small pump to keep everything in circulation. Now we are
not trying to optimize the geometry. I do not even know if you can get it
that high for a full scale unit. I think before you would want to go that
far, you would try it out with some type of prototype, but .you have to make
sure your suspended solids will stay in there and not settle out. The cost
effectiveness, just off the top of my head, you are talking about a little i
bit more concrete.
1538
-------
MR. ORWIN: You are actually talking about three times or three or four times
the volume, so it could be fairly expensive,,
DR. O'SHAUGHNESSY: Yes, that would be an engineering decision, right? I
do not know what the largest one you can use is, no, we did not, I am sorry.
DR. SRINIVASARAGHAVAN: I have two questions. One is in connection with the
slides that you showed. You increased the surface to volume ratio. The
treatment efficiency increased but at the same time your loading values are
also increased.
DR. O'SHAUGHNESSY: No, the loading values are based on pounds of ammonia
nitrogen per thousand square feet per day, that is constant. Each unit
has the same surface area.
DR. SRINIVASARGHAVAN: According to the loading numbers that you showed the
two units were increased.
DR. O'SHAUGHNESSY: In the first part of the study, we had the Autotrol and
the Hormell unit. They were pretty close in terms of surface to volume ratio;
0.09 and 0.12. Only in the spring portion when we drastically increased the
surface volume ratio was that different.
DR. SRINIVASARGHAVAN: So, I wonder what the conclusion was so if you increased
the loading the treatment efficiency increasing 15 to 20 to 35 or something
like that. I was wondering if at the influence of the increasing volume to
surface area ratio as you increased the ammonia influent concentration and
therefore you got higher reduction,
DR. O'SHAUGHNESSY: No, we kept loading it .at the same rate.
DR. SRINVASARGHVAN: Second question; what are the conclusions from the study
that the staging does not influence the ammonia treatment efficiency?
DR. O'SHAUGHNESSY: No, we found that there was a definite impact of staging
when you get fluctuations, you can see that. A lot of time indicated that
the staging did make a big difference except in the two units. The fourth
stage every time there was an upset because of temperature. Second, once we
got onto the field we had much better performance as soon as we had two stages.
So you would recommend a minimum of two stages than probably four or something
like that. Yes, staging is something you would want.
MR. BELSCHNER: Did you measure suspended solids into an outer reactor and
how did you determine any net sludge yield if any from the nitrification
system?
DR. O'SHAUGHNESSY:
at each stage.
Yes, we took effluent and influent and suspended solids
MR. BELSCHNER: Was there a net sludge yield for nitrification?
DR. O'SHAUGHNESSY: We had just what was produced in terms of pounds of sus-
pended solids versus pounds of ammonium removal.
1539
-------
MR. BELSCHNER: Was that based on solids coming out of the reactor or the
difference between xrtiat went into it and came out?
DR. O'SHAUGHNESSY: The difference between the two.
MR. BELSCHNER: That looks unusually high. Typically we find a net solids
destruction of secondary effluent solids going through the process when
nitrifying.
DR. O'SHAUGHNESSY: We check the influent minus the effluent, plus, you know,
what was there. And that is what it came out to be.
MR. BELSCHNER: That is an unusual result. With respect to that question,
what would a. solid concentration in the reactor systems, Jim? You show
effluent concentrations higher than the reactor concentrations.
DR. O'SHAUGHNESSY: I may not have that information right here. They were
not large, well under 50 milligrams per liter as I remember. We have the
data. I do not know just what it is, I would say fifty. There was very
little food in the system.
MR. BELSCHNER: Did you make the measurement of ammonia?
DR. O'SHAUGHNESSY: We measure ammonia using a specifie'.electrode .right "at
the plant.
MR. BELSCHNER: Is that an acceptable technique or...
DR. O'SHAUGHNESSY: Yes, we standardized the program in monitoring the
system. Incidentally, I should point out how we will take samples. We
took many samples based on hydraulic flow time. We would, if it was eight
o* clock in the morning and we knew it was one hour detention time we would
take the next one at nine, ten, eleven, and so forth and follow that plug
through. And if they came out not close enough we would not take the third,
and results are based on average value.
MR. BELSCHNER: When you are measuring efficiency of the system, you are
talking about efficiency of conversion or removal of ammonia under dynamic
loading conditions as well as organic loading in the system. It would seem
like you are attributing to nitrification a certain amount of conversion, a
variable amount of conversion of nitrogen to cell mass in the system. Won't
you maybe look at nitrates as a by-product to determine the process efficiency?
One figure I thought was bad, maybe I misinterpreted.
DR. O'SHAGHNESSY: No, I am sure, the effluent nitrate concentration was an
average from the study; both studies, 90% of the influent ammonia concentration,
so 90% came out in our effluent as nitrate, 10% went into solid mass or de-
nitrification or whatever else you want to attribute it.
DRe ODEGAARD: I would like to know how do you plot your results, because it
was not obvious that any maximum nitrification rate, and if you are going to
design a nitrification system we have to know a little bit about the kinetics
1540
-------
of the nitrification. So my question is, did you find any when you increased
your loads; did you find any plateau where you had a biological reaction?
Did you go into each of the steps and try to find a Monod relation?
DR. O'SHAUGHNESSY: No, as I tried to point out in the beginning, this was
intiated in 1976. One of the concerns is can we achieve and what do we
need to achieve one milligram per liter of a secondary effluent, one milli-
gram of ammonia in these wastewaters. We found this was what we needed to
go up to get equilibrium to find out what would be the limiting factor. You
would be well above the loading rates obviously as the other information has
been presented here. And you would be close to one milligram per liter.
And this was not that big a study, so we did not look at what you are inter-
ested in.
DRo ODEGAARD: I mentioned this because I think it is interesting. If you
go into the literature and we heard from the Japanese presentation today and
my presentation, our figure and all the figures that I found in the literature
are very similar. So the maximum nitrification rate is similar. Then of
course, we have an obvious difference in conclusion.
DR. O'SHAUGHNESSY: All well, On the paper presented this morning, I went
to calculate just roughly. I am estimating that you know, loading about
ten pounds of ammonium nitrogen per thousand square feet per day, and the
highest level we got was point eight pounds.
DR= ODEGAARD: I have a problem converting.
DRo O'SHAUGHNESSY: I think if you look at that I think you will find its
value almost identical to us or two off in terms of where I am. We are
interested in what is going to happen to achieve less than one milligram
per liter as required by a discharge permit.
MRo MOLOF: You have four stages RBC system. What are the observations in
terms of nitrification versus temperature in each stage? Did you make this
study?
DR. ZENZ: We did and I did not present that information because of time
consideration. During the test period where you saw, when we showed infor-
mation on 80 percent, we were getting nitrification in the last stages.
Now some of those higher removals with the higher temperatures, we were not
getting a lot of nitrification in the very last stage. So it would shift;
it would generally depend on temperature, but in the numbers that I am giving
you winter time temperatures we were getting 80% removal. The entire unit
is nitrified and it is you know, I am not giving you information whether it
is unneeded stages at that the matter of the question you gave me.
MR. COULTER: In your last slide you had 80% removals at cold temperatures
at both point two and point four loadings. This is a substantial difference
in terms of hardware required. Did you have any supplementary data in
terms of alkalinity or anything else that might explain that difference?
1541
-------
DR. ZENZ: No, I cannot really explain it. There were changes in influent
ammonia concentration, unfortunately I cannot remember if those went with
those particular places. There was some scatterings in the data. I cannot
really tell you right now if I could postulate a reason for that off the
top of my head.
MR. WATT: Do you have any leakage of ammonia through your nitrification
system?
DR. ODEGAARD: No it was quite low, It was totally nitrified.
MR. ATHAVALEY: Since we are talking here about anaerobic denitrification
what is the temperature factor in removing the ammonia nitrogen? Did you
study the temperature effect?
DR. ODEGAARD: No we did not. As I said we did not even have a temperature
controlled, but the temperature in these experiments were between 12 and 16
degrees centigrade. But you would normally not use heating of anaerobic
step anywayo So I do not think it is...of course the temperature influence
is interesting to all of us.
MR. HAUNG: Did you use any organic compound in denitrification step to
achieve Img/lof ammonia nitrogen in the process effluent?
DR. ODEGAARD: You mean to take out the rest of ammonia? No, I think there
is a plant in Japan which is similar to this one, but in addition, have a
second step where they use methane to take out the last part.
DR. HAUNG: It appears that in a tremendous hydraulic loading to the denitri-
fier, and I wonder would there be a minimum retention time upon this applica-
tion. Would that affect your results?
DR. ODEGAARD: That is an interesting question because we worried about that.
But the only problem we saw is that when you operate the recirculation to
very, very high levels we tried to go as high as twenty, then we got problems.
As long as we stayed under ten and the plant is designed of the normal criteria.
so that the retention time as such is short, then we did not have any problems.
DR. HAUNG: What is the retention time?
DR. ODEGAARD: That is a good question. I think, I do not have it in here
right now but it is, well, that depends upon what you are asking about be-
cause the theoretical detention time through the plant is I guess approxi-
mately three hours.
MR. BALANCE: What was the retention time meant in your case?
DR« ODEGAARD: Well, if you look at one particle, how long the particle is
retained in the system. But then the water was circulated all the time so
it comes back, the same particle comes back many times of course. So that
how do you define detention time?
1542
-------
I
Session 11.
Presiding:
SELECTIONS AND ECONOMICS
J. Miller
Department of Civil Engineering
University of Pittsburgh
MR. JOOST: Keep this in your mind, the power consumption for the mechanical
drive is quite high to me.
MR. BARRY: I have heard numbers anywhere from three to three and a half to
four. That has been down since we did this analysis. So again, it is going
to be an analysis like this has to be done on a case by case basis on all
rates and the best information on it to find.
MR. BERNER: Can you comment on your solid handling processes at this plant?
MR. BARRY: The solid handling for RBC sludges are going off the landfill
immediately. The primary sludge would be pumped directly to anaerobic di-
gester. The secondary sludge there are provisions for it to go to either
a gravity thickener with or without chemical polymer or directly to the
digesters. The two stages digesting system would be recovering the gas
and running into generators from that system after the solids have digested
they will be taken off the plant site by tanks, trucks, and disposed of on
the farmfieldSo We found the farmers are very receptive to this humus waste
material because of the nature of the waste product. And anything the farmers
can get that will hold moisture in the soil they are very anxious to get.
DR0 SACK:
that?
How did you handle labor costs, number of operators, things like
MR0 PIERCE: Well, actually the number of operators is somewhat debatable
between analysts^ We felt that in general the RBC processes should have
fewer operators and if there was any differentiation between the two the
mechanical drive RBC installations, particularly the large ones would have
perhaps an additional one operator. In general we only added for the acti-
vated sludge processes sufficient operators to handle the thickening aspects.
The estimation was the largest design plus three more people operating an
activated sludge plant than operating RBC plant.
MR. BAO: Do you have any comments about Dr. Sun's conclusion about what
he said, his recommendations he said the activated sludge plant would be
cheaper than RBC.
MR. LUNDBERG: I do no recall his paper. He is not here to defend himself
on any comments. I think he was talking about synthetic trickling filters
versus the RBC in that application. I do not think he discussed activated
sludge in his paper and somebody can correct me if I am wrong on that. I
am expecting this question was going to arise. I really do not necessarily
see a conflict there because we are talking about three separate applications
where Pierce was talking about carbon oxidation alone. I was talking about
combined carbon oxidation and nitrification in a single stage and he was talk-
ing about simply nitrification alone in these reactors as a simple process
1543
-------
onto itself. So to begin with we were not talking about the same thing.
The other thing that crossed my mind was the design criteria that was uti-
lized in his paper versus mine. Now it was two different applications. 1
think that if we tried to sit down and compare apples to apples, that he was
maybe using less conservative loading rates for the synthetic media filters
than I did. Now I am not sure so as I said, first of all I really do not
see it as a bumping heads type of thing, because we thought it fanned out
rather nicely that there were three papers here that talked about three
separate types of applications where Pierce and I had two of them and we
opened up the list of people to the symposium that goes with two-thirds of
the paper complement that we were talking about. Those are the only two
things I can see we have to look at specifically at the cost factors he used
and the design criteria to really figure out what the story was.
MR. GROVER: The last speaker, did you make any assumption in the different
sludge dewatering characteristics that purported to exist within the pure
oxygen design? Would you assume a better dewatering characteristic maybe
due to chemicals addition?
MR. PIERCE: Well, actually we did not get into the handling aspects. In
this study it was basically liquid treatment and thickening. We did assume
a higher underflow concentration in the pure oxygen case of perhaps one
point seven five percent underflow versus point seven five percent underflow
for the air, and we did include a difference in the dissolved air flotation
loading rate. Our opinion was that regardless of the sludge handling train
utilized, the liquid treatment of thickening could pretty much be separated
and any...there is a great deal of debate about how much sludge one gets
from an RBC process and whether it is or is not as dewaterable. I think in
the worst case against RBC's they would all be pretty much the same. So the
sludge handling could be addressed separately in a facilities plant.
MR. SUGAR: I guess I am going to have trouble with all of your curves shown
previously. I do not know if you have realized that in particular, the curves
you show with the trial versus the PSA for the pure oxygen plant. I guess :
I am surprised that there is such a similar delta that you are talking about,
because our internal analysis we were trying to find whether there should be
a PSA or a trial. You definitely tend to go with the PSA plant from the
cost effective point of view, attempting power in the smaller range plant.
And we have obtained a trial plant very much below twenty tons a day. On
the other side of the coin, you really would not consider using a PSA. That
would be an excessive to that and therefore I am somewhat surprised at that
delta with that range process range.
MR. PIERCE: Well actually there are probably two explanations for this.
It is a secondary treatment application and it has a relatively dilute low
concentration of BOD. We did basically center this study around a similar
study done at a thirty MGD range and our oxygen generation requirements were
on the order of about eighteen tons at thirty MGD. That particular analysis
had indicated the PSA had a slightly lower present worth value and being con-
cerned about the small delta on here, it is really not that small if you are
1544
-------
only considering the oxygen generation facility or the aeration tanks. But
bear in mind that there is a large block added to all of those and the
thickening facility, the final clarifiers and all the other unit process
equipment for the whole plant if you were to take those, the differential
and apply that entirely to the difference in oxygen generation facilities,
which it pretty much is you would probably find that the delta is much more
what you think it might be0 I cannot really explain why in the range it
fits the.ooit was still a PSA plant and I do not think that again it was in
plus or minus ten percent. It could have gone either way.
MR. ANTONIE: I would like to have a point clarified regarding a question that
was asked earlier, Mr. Pierce, regarding, in fact I do not believe the con-
sulting engineers are still here0..He asked what the impact in your conclusions
would be, shorter evaluation period the sort of payback period, and he speci-
fically asked that he was concerned with industrial wastewater applications,,
I would like to point out, you can correct me if I am wrong Mr. Pierce, that
your evaluation was based solely on the kinetics of municipal wastewater and
that you cannot extrapolate your report on industrial waste.
MR0 PIERCE: I do not know if that was his specific intent but what you are
saying is true. We assumed a typical municipal primary effluent, so depending
on the fraction of soluble versus insoluble BOD in waste there is such a
variety of other factors that the only way you could interpolate you know,
off of this graph and say that for one MGD industrial installation many
things apply here..0I think he may just be giving that as an illustration
of the fact of what would happen if somebody placed more budget upon capital
costs than on the operating costs. But as you did say, I think there is a
lot of danger in trying to take things somebody else did, particularly cost
curves, and apply it to your particular situation. And I think Lee and I
both feel that nobody should take the cost curves that we developed and take
a look at them and say okay, if it is a 30 MGD plant, this is the construc-
tion costo It has to be done on a case by case basis., There are just so
many variables as far as we are concerned, and you can correct me if I am wrong.
One of the cost curves we have developed is totally useless except to give
you a general indication. I think what Lee wants: to try to show was that
the RBC process should not be reasonably thrown .d.ut the window as many people
have .done .because .we .do thaink that the cost for a number of applications
brought across a wide range of flow.
•ft-U.S. Government Printing Office: 1980 M-654-177,
1545
-------
-------
-------
-------
-------
-------